ADVANCES IN PROTEIN CHEMISTRY VOLUME V
This Page Intentionally Left Blank
ADVANCES IN PROTEIN CHEMISTRY EDITED BY M. L. ANSON
JOHN T. EDSAU
Continental Foods, Inc., Hoboken, New Jersey
Harvard Medical School; Boston, Mascrchurettr
Associate Editor for the British Isles KENNETH BAILEY Trinity Cdlege, Cambridge, Enghnd
VOLUME V
1949 ACADEMIC PRESS INC., PUBLISHERS NEW YORK, N. Y.
COPYRIGHT Cc)
l(.)#) BY ACADEMIC PRESS
INC
ALL ~ I ( : € I T s RESERVED NO PART OF T H I S BOOK MAY BE REPKODUCED IN ANY FORM BY PHOTOSTAT, MICROFILM, OR ANY OTHER MEANS, WITHOUT WRITTEN PERMISSION FROM T H E PIIIILISlIERS.
ACADEMIC PRESS INC 1 I I FIFTHAVENUE
NEWYORK3,N. Y .
llnited Kingdom Edition Puhlished by ACADEMIC PRESS INC. ( L O N D O N ) LTD. BERKELEY SQUARE HOUSE.LONDONw. I
First Printing, 1949 Second Printing, 1963
I’RINTED IN T H E lJNITED STATES OF AMERICA
CONTRIBUTORS TO VOLUME V
JAMESB. ALLISON,Rutgers University, New Brunswick, New Jersey ALBERTCLAUDE,The Rockefeller Institute for Medical Research, New York, New York* JOSEPHS. FRUTON,Department of Physiological Chemistry, Yale University, New Haven, Connecticut K . H. GUSTAVSON,Swedish Tanning Research Institute, Stockholm, Sweden J . W. H . Luaa, Department of Biochemistry, University of Melbourne, Melbourne, Victoria, Australia HAROLDP. LUNDQREN,Western Regional Research Laboratory, U . S. Department of Agriculture, Albany, California THOMAS L. MCMEEKIN,Eastern Regional Research Laboratory, U. S. Department of Agriculture, Philadelphia, Pennsylvania B. DAVIDPOLIS,Eastern Regional Research Laboratory, U. S. Departm m t of Agriculture, Philadelphia, Pennsylvania G. R. TRISTRAM, S i r William D u n n Institute of Biochemistry, University of Cambridge, Cambridge, England * Present Addreee: Institute Jules Bordet, Bruxelles, Belgium.
V
This Page Intentionally Left Blank
CONTENTS CONTRIBUTORS TO VOLUMEV .
......................
v
The Synthesis of Peptider BY JOSEPES FRUTON. Department of Physiolopkul Chemistry. Y a k University. New R a m . Connedicvt 1. Introduction and Nomenclature . . . . . . . . . . . . . . . . . . . 1 I1. General Methods of Peptide Synthesis . . . . . . . . . . . . . . . . 6 I11. Special Aspecte of Peptide Synthesis . . . . . . . . . . . . . . . . . 34 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
.
Amino Acid Composition of Purided Proteins BY G . R . TRISTRAM. Sir William Dunn Znatituk, qf Biochemistry. University of Cambridge. Cambridge. England I Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 I1. Purposes of Amino Acid Analysis . . . . . . . . . . . . . . . . . . 86 I11. Methods of Analysis . . . . . . . . . . . . . . . . . . . . . . . . 86 IV Establishment of Accuracy and Specificity of Methods of Analyaie. . . . 99 V Comparison of Analytical Methods . . . . . . . . . . . . . . . . . 106 VI . Structure of Proteha aa Revealed by Amino Acid Analysis . . . . . . . 125 VII . Amino Acid Composition of Certain Proteins . . . . . . . . . . . . . 129 VIII . General Discussion . . . . . . . . . . . . . . . . . . . . . . . . 141 I X . Conclusion and Summary . . . . . . . . . . . . . . . . . . . . . . 148 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
.
. .
Biological Evaluation of Proteins BY JAMES B . ALLISON. Rutgers Universily. New Brunewick. New Jersey
I . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 156 I1. Evaluation through Nitrogen Balance . . . . . . . . . . . . . . . . 167 I11. Evaluation through Growth . . . . . . . . . . . . . . . . . . . . . 173 IV. Evaluation through Tissue Regeneration . . . . . . . . . . . . . . . 180 V. Evaluation through Amino Acid Analysis . . . . . . . . . . . . . . 192 VI . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
Milk Proteins
.
.
BY THOMAS L MCMEEKINAND B DAVIDPOLIS. Eastern Regional Rcsearch Laboratory. Philadelphia. Pennsylvania
I . Introduction . . . . . . . . . . . . . . . I1. Protein Distribution in Milk . . . . . . . I11. Separation and Properties of Milk Proteins . IV. Proteins of Whey . . . . . . . . . . . . . vii
. . . .
. . . .
. . . .
. . . .
. . . .
. . . . . . . 202 . . . . . . . . 202 . . . . . . . . 203 . . . . . . . 210
viii
CONTENTS
.
V Amino Acid Composition of Milk Proteins . . . . . . . . . . . . . . 219 VI . Encymee in Milk . . . . . . . . . . . . . . . . . . . . . . . . . 219 VII . Relationship of Milk Proteins to Serum Proteins . . . . . . . . . . . 223 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225 Plant Proteins
BY J . W . H . LUOO,Department
of Biochemistry, University of Melbourne, Melbourne, Victoria, Australia
I. I1. I11. IV
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 230 Bulk Proteins of Plants, Plant Organs, etc . . . . . . . . . . . . . . . 232 “Individual” Proteins of Plants . . . . . . . . . . . . . . . . . . . 259 Modes of Occurrence of Protein in Plants . . . . . . . . . . . . . . . 265 V . Protein Metabolism in Plants . . . . . . . . . . . . . . . . . . . . 269 VI . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 294 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
.
Synthetic Fibers Made from Proteins
BY HAROLD P. LUNDOREN, Weetern Regional Research Laboratory, U . S. Department
of
Agriculture, Albany, California
I . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 305 I1. General Considerations of Proteins Chain Behavior . . . . . . . . . . 307 I11. Preparation of Fibers from Proteins . . . . . . . . . . . . . . . . . 311
IV . Molecular Basis for Mechanical Properties of Fibers Made from Proteins 327 V. Summary and Conclusion . . . . . . . . . . . . . . . . . . . . . . 345 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346 Some Protein-Chemical Aspects of Tanning Processes BY K . H . GUSTAVSON. Swedish Tanning Research Institute. Stockholm. Sweden
I . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 354 I1. Chemistry of Collagen . . . . . . . . . . . . . . . . . . . . . . . 356 I11. Keratolysis and Action of Alkali on Hide . . . . . . . . . . . . . . . 375
IV . General Aspects of Tanning . . . . . . . . . . . . . . . . . . . . . 378 V Reaction of Chromium Compounds with Collagen (Chrome Tanning) . . 379 VI . Vegetable-Tanning Process . . . . . . . . . . . . . . . . . . . . . 394 VII . Reaction of Condensed Sulfo Acids (Syntans) with Collagen . . . . . . 402 VIII . Tanning Power of Aldehydes . . . . . . . . . . . . . . . . . . . . 405 I X . Quinone Tannage . . . . . . . . . . . . . . . . . . . . . . . . . 411 X General Comments . . . . . . . . . . . . . . . . . . . . . . . . 413 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
.
.
Proteinr, Lipids, and Nucleic Acids in Cell Structure8 and Functions
BY ALBERT CLAUDE,The Rockefeller Znatitute for Medical Research, New York, New York I . Introduction . . . . . . . . . . . . . . . . I1. TheCell . . . . . . . . . . . . . . . . . . I11. The Nucleus . . . . . . . . . . . . . . . . IV The Cytoplasm . . . . . . . . . . . . . . . V Constitution and Duplication of Living Matter
. .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
423 425 426 428 433
ix
CONTENTS
VI . Phospholipids in Cell Structures and Functions . . . . . . . . . . . VII . Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . : . VIII . Nucleic Acids . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
.
434 436 437 439
AUTHOR INDEX. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
441
SLIBJECT INDEX . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
466
This Page Intentionally Left Blank
The Synthesis of Peptides
BY JOSEPH 5. FRUTON Department of Physiological Chemistry, Yale University, New Haven, Connectiml
CONTENTS
pooc
I. Introduction and Nomenclature.. . . . . . . . . . . . . . . . . . . . . . ............ 11. General Methods of Peptide Synthesis.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. Synthesis of Dipeptides by Partial Hydrolysis of Diketopiperasines 2. Condensation of Peptide Esters... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Synthesis of Peptide Derivatives by Means of Acylamino Acid Chlorides and Aeides. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Use of Amino Acid Chloridea in Peptide Synthesis.. . . . . . . . . . . . 5. Use of a-Halogen Acyl Halides in Peptide Synthesis. . . . . . . . . . . . . . . 6. “Adactone” Method for Syntheeb of Peptides 7. Condeneetion of Keto Acids and Amides.. . . . . 8. Use of N-Carbonic Acid Anhydrides of Amino Acids in Peptide Synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9. Use of Toluenesulfonylamino Acids in Peptide Synthesis. . . . . . . . . 10. Use of Aeidoacyl Chlorides in Peptide Synthesis.. . . . . . . . . . . . . . . 11. “Carbobenmxy” Method of Peptide Synthesis.. . . . . . . . . . . . . . . . . . . 12. Use of Phthalylamino Acids in Peptide Synthesis. . . . . . . . . . . . . . . . . . 13. Enzymatic Synthesis of Peptide Derivatives.. . . . . . . . . . . . . . . . . . . . . 111. Special Aspects of Peptide Synthesis.. . . . . . . . . . . . . . . . . . . . . . 1. Peptides of Glycine, Alanine, Valine, Leucine, and Isole 2. Peptides of Aspartic and Glutamic Acids.. . . . . . . . . . . . . . . . . . 3. Peptides of Phenylalanine and Tyrosine. . . . . . . . . . . . . . 4. Peptides of Cystine and Cysteine. . . . . . . . . . . . . . . . . . . . 5. Peptidesof Serine.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6. Peptides of Lysine and Other Diamino Acids. . . . . . . . . . . . . . . . . . . . . 7. Peptidesof Arginine.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8. Peptides of Histidine.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9. Peptides of Proline and Hydroxyproline.. . . . . . . . . . . . . . . . . . . . . . 10. Peptidesof Tryptophan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. Peptides of Methionine. . . . . . . . . . . . . . . . . . . . . . . . . ............. 12. Peptides of Other Amino Acids.. . . . . . . . . . . . . . . . . . . . . . . . . . . . References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ........................
1 5 5
6 9 11
21
25 32
33
66 68 62 66 67 72 73 73
I. INTRODUCTION AND NOMENCLATURE The objectives of the search for satisfactory methods of peptide eynthesis were clearly stated by Fischer and Fourneau (149) in the 1
2
JOSEPH
a.
FRUTON
memorable paper which initiated the systematic exploration of this field of study : “Der Gedanke, die aus den Proteinstoffen durch Hydrolyse entstehenden Aminoskuren durch Anhydridbildung wieder eu g r h e r e n Komplexen eu vereinigen, ist schon seit lkngerer &it von verschiedenen Forachern experimentell behandelt worden, W k emnern nur a n die Anhydride von Schaal (262), ihre Verwandlung einerseits in den kolloidalen Polyaaparaginharnstoff von Grimaux (199), anderseita in die Polyaspartakuren von H. Schiff (263), ferner an die Verauche von Schtitzenberger (270) nber die Vereinigung verschiedener Aminoshuren (Leucine und Leuceine) mit Harnstoff durch Erhiteen rnit Phosphorsiiureanhydrid, an die iihnlichen Beobachtungen Lilienfelds (238) tiber die Wirkung von Kaliumbisulfat, Formaldehyd und andereo Kondensationsmitteln auf ein Gemiach von Aminostiureestern und endlich an die Angaben von Balbiano und Trasciatti (39) tiber die Verwandlung dea Glycocolls in ein hornartigea Anhydrid durch Erhiteen mit Glycerin. Aber alle von ihnen beschriebenen Produkte sind amorphe, schwer characteresierbare Substanzen, tiber deren Struktur man ebensowenig wie tiber den Grad ihrer Verwandtschaft mit den natblichen Proteinstoffen etwas sagen kann. “Will man auf diesem achwierigen Gebiete xu sicheren Resultaten kommen, BO wird man zuerst eine Methode finden m h e n , welche es gestattet, aucceeaive und mit definierbaren Zwiachenstufen die Molektlle verachiedener Aminoakuren anhydridartig aneinander eu reihen.”
Since the enunciation of this view, it has become abundantly clear that one of the principal contributions of the organic chemist to the study of the structure and reactions of proteins has been indeed the development of several techniques for the synthesis of compounds in which amino acids are joined to one another “anhydridartig” by means of acid amide linkages. The services of peptide synthesis to protein chemistry have been many and various. The finding of synthetic peptides and peptide derivatives which are hydrolyzek by crystalline enzymes specifically adapted to the hydrolysis of proteins has buttressed the theory, first expressed by Hofmeister (217) and Fischer (126), that in proteins the peptide bond represents the most general type of linkage between the individual amino acid residues.* In addition, modern methods of peptide synthesis have made possible the preparation of special peptides of low molecular weight and known chemical structure In stressing the importance of the peptide bond in the structure of proteins, i t is not intended to exclude the possibility that other types of covalent linkage may also play a significant role in the architecture of the protein molecule (99,136). Of particular interest in connection with the search for labile bonds which may be involved in the phenomena associated with protein denaturation ia the recent suggeution of Linderatr0m-Lang and Jacobsen (239) that cysteine residues of proteins might participate in the formation of thiaeoline rings. Similar consideration might be given to the participation of serine (or threonine) residues in oxaeoline groupings (46). The recent synthesis, by Ehrentwhd and Davideeohn (119), of labile peptides with thioiminoether linkages is a further experimental approach in this direction.
SYNTHE8IB OF PEPTIDES
3
for use as models in the examination of several physical and chemical properties of proteins. Thus, the use of synthetic peptides has notably facilitated the interpretation of data on the acid-base relationships of proteins (101). Furthermore, in the study of the reactions of proteins with various chemical reagents, parallel experiments with peptides or peptide derivatives have frequently clarified the results observed with proteins. Of the numerous examples in the recent literature one may cite the study of the iodination of tyrosine-containing proteins (212), or the investigation of the action of mustard gas (250) and the nitrogen mustard gases (191) on proteins. Apart from their importance in providing simple models for study of the enzymatic degradation , physical properties, or chemical reactions of proteins, the newer techniques of peptide synthesis have been invaluable in the final proof of the chemical structure of several physiologically important substances, such as glutathione (202) and carnosine (272). The discovery that the antibiotics gramicidin, tyrocidine, and gramicidin S are peptides (219,282) and the recent reports that there is, in pancreatic hydrolyzates of several proteins, a factor (or factors) of peptide nature which promotes the growth of certain microorganisms and the rat (299,302) offer further opportunity for the fruitful use of the methods of peptide synthesis to establish the chemical structure and to study the physiological action of peptides of biological interest. The development, in recent years, of new methods for the separation of amino acids and peptides (103,242) has led to i renewed interest in the products of the partial hydrolysis of proteins (281). It is clear, however, that, whatever methods are used to separate and identify peptides obtained from proteins, the conclusive evidence for the identity of such peptides must come from the comparison of the isolated material with synthetic peptides of known structure. This is the procedure that allowed Fischer and Luniak (159) to establish definitely that the peptide isolated by Osborne and Clapp (253) from a gliadin hydrolyzate was indeed L-prolyl-L-phenylalanine. In a similar manner, Stein, Moore, and Bergmann (278) demonstrated the presence of glycyl-L-alanine and L-alanylglycine in partial hydrolyzates of silk fibroin. Enough has been said in the foregoing to justify the importance attached to peptide synthesis as a tool in the study of proteins. The purpose of this review is to survey the available methods for the synthesis of peptides. These methods will be evaluated, whenever possible, with regard to their relative difficulty, their adaptability to meet the many problems encountered in amino acid chemistry, and the yield and purity of the products of synthesis. Although the preparation of peptides containing nonprotein amino acids will occasionally be men-
4
JOSEPH 8. FRUTON
tioned, primary emphasis will be placed on the synthesis of peptides of amino acids definitely known to be formed upon protein hydrolysis. In what follows, the configuration of the amino acids will be given in accordance with the report of the Editorial Board of the Journal of Biological Chemielty (284). T h e amino acid residue6 will, in general, be designated by adding the suffix “yl” to the roots of the names of the free amino acids. “hue, for glycine, the term will be glycyl; for proline, prolyl; for tyrosine, tyrosyl; and 80 forth. Several departures from this rule will be noted, however. The peptides of cysteine will not be designated (‘cysteyl” but “cysteinyl” peptides, and for the cystine peptidee, the designation will be “cy& tinyl” rather than “cyetyl.” In the caw of peptides of glutamine, the amino acid reaidue will be termed “glutaminyl” to differentiate it from “glutamyl,” which refers to glutamic acid. Similarly, for asparagine peptides, the term will be “asparagiuyl” and, for aapartic acid peptides, it will be “mpartyl.” Jn addition, the term “tryptophyl” will be used to deeignate the amino acid residue of tryptophan. Since thia nomenclature impliee that an amino acid has been converted to an acyl group, it follows that the amino acid residues in a polypeptide should be listed in the sequence of substitution at the amino group of the adjacent amino acid. The tripeptide glycyl-calanyl-cleucine has, therefore, the following formula:
CHa NHrCH,CO---NdHCO--NH
tH9 HCOOH
A brief discussion of the configurational relationships of peptides appears necessary at this point since frequent mention will be found in the literature of racemic peptides which contain more than one optically active amino acid. It is clear that, in the case of a racemic dipeptide containing two optically active amino acids, i e . , Dcleucyl-Dcalanine, four isomers are possible: (a) D-leucyl-Palanine, (b) cleucyl-L-alanine, (c) cleucyl-D-alanine, and (d) mleucyl-calanine. Two racemates may be expected; one composed of forms a and b, and another composed of forms c and d. Separation of the two racemates may sometimes be achieved by taking advantage of their differences in solubility. It is then customary to designate the less soluble form by the letter “A” and the more soluble racemate by the letter “B.” Reports will be encountered in the literature of the synthesis of peptides containing more than one optically active amino acid and in which one amino acid residue is present in a single configuration, while another is given as the DL form, e.g., DL-alanylglycyl-L-glutamic acid. Such preparations are clearly mixtures of diastereoisomers. When attempts to separate the two forms are not successful, three possibilities may be envisaged: (a) the two isomers have very similar solubilities, thus making their separation by fractional crystallization difficult; (6)
5
SYNTHESIS O F PEPTIDES
they are isomorphic and form mixed crystals; or (c) they form addition compounds in stoichiometric proportions. Since i t cannot be predicted beforehand whether, after the synthesis of a mixture of such diastereoisomers, a pure peptide can be isolated, it appears preferable, whenever possible, to perform peptide syntheses with optically active amino acids under conditions where racemization is avoided. Furthermore, in the study of the specificity of proteolytic enzymes or of the properties of model compounds related t o proteins, it is usually desirable to have peptides containing optically active, rather than the racemic, forms of the amino acids. Greater interest attaches itself, therefore, to methods which permit the synthesis of such optically active peptides.
11. GENERALMETHODSOF PEPTIDE SYNTEIESIS 1. Synthesis of Dipeptides by Partial Hydrolysis of Diketopiperan'nes
In 1888 Abenius and Widmark (38) found that ditolyldiketopiperazine could be partially cleaved by acid hydrolysis : /CH1-Co CHaCsHdN
\
CO-CHI
\
/
CsH4CHa
I
NCsHdCHr -+ CH,C~HINHCH~CO-NCHICOOH
An analogous reaction of aliphatic compounds was not observed until 1901, when Fischer and Fourneau (149) hydrolyzed glycine anhydride by heating it briefly with concentrated hydrochloric acid and thus obtained glycylglycine, the simplest representative of the group of substances under discussion in this review. The method used by Fischer and Fourneau still is a convenient method for making this dipeptide, especially in view of the recent development of an excellent procedure for the synthesis of glycine anhydride from glycine (261). Fischer noted, however, that brief acid hydrolysis of DL-alanine anhydride and of obleucine anhydride did not yield the expected dipeptides (125). Fischer later found that brief hydrolysis with dilute sodium hydroxide a t room temperature would also yield dipeptides from diketopiperazines (132). In addition to glycylglycine, one of the two possible racemic forms of Dtalanyl-DL-alanine was prepared in this way. It soon became clear, however, that the method was not generally applicable to the synthesis of dipeptides containing optically active amino acids, for, when Fischer attempted to prepare L-alanyl-L-alanine from the corresponding anhydride, the product which resulted was partially racemic (153). Similarly, treatment of L-tyrosine anhydride with alkali caused appreciable racemiaation (166). As was shown by the later systematic studies of Levene (237) and Bergmann (82), among others, the racemization of optically
0
JOSEPH 8. FRUTON
active diketopiperazines is favored by alkali. I t is obvious, therefore, that this procedure is essentially limited to work with glycine anhydride. In fact, it is frequently convenient to treat glycine anhydride with sodium hydroxide and to use the resulting solution of the sodium salt of glycylglycine for condensation reactions involving the amino group of the dipeptide. The racemization of diketopiperazines, so pronounced in alkaline media, is notably less in acid. In some cases, it is possible, therefore, to achieve a satisfactory synthesis of optically active dipeptides by cautious treatment with hydrochloric acid. Thus, Greenstein (195) was able to make tcysteinyl-L-cysteine from L-cysteine anhydride by hydrolysis of the latter with concentrated hydrochloric acid a t room temperature. It must be added, however, that some diketopiperazines, such as histidine anhydride and tyrosine anhydride, are fairly'stable when they are heated with strong acid, while alanine anhydride and leucine anhydride require prolonged treatment with hot concentrated acid to effect cleavage of the ring. To the difficulties encountered in the partial cleavage of diketopiperazines containing like amino acid residues, must be added those met in the splitting of diketopiperazines derived from two different amino acids. In the latter case, two different dipeptides may result, as in the hydrolysis of glycyl-DL-leucine diketopiperazine, which yielded a mixture of glycyl-DL-leucine and DL-leucylglycine (166). The separation of such mixtures is usually difficult. If, however, the formation of only one of the two possible dipeptides is favored, this difficulty may be eluded. One of several instances in which a mixed dipeptide could be prepared in pure form by hydrolysis of a diketopiperazine was described by Bergmann and Tietzman (68), who obtained L-prolyl-L-phenylalaninefrom L-prolyltphenylalanine diketopiperazine. The dipeptide had the same rotation as that of the product obtained by Fischer and Luniak (159), who coupled L-prolyl chloride with L-phenylalanine ethyl ester, and then saponified the dipeptide ester. As the first available method for the synthesis of free peptides, the procedure discussed in this section has considerable historical interest. It is clear from the foregoing, however, that it has many limitations. Despite the occasional successes of this method, and its value as an adjunct to other synthetic procedures, it may be expected that, even for the synthesis of dipeptides, preference will be given to other methods. 2 . Condensation of Peptide Esters Curtius noted in 1883 (105) that glycine ethyl ester undergoes spontaneous transformation to yield glycine anhydride and a substance
7
SYNTHESIS OF PEPTIDES
which gives a positive biuret test. For this reason, the latter product was termed “biuret base.” Later studies by Curtius (107) showed that, if moisture is excluded, the formation of the anhydride is suppressed and the “ biuret base,” which he formulated as (triglycy1)glycine ethyl ester, is the chief product. In a similar manner, Fischer (134) converted (diglycy1)glycine methyl ester to (pentaglycy1)glycine methyl ester by heating the former at 100”. 2 NH&H&O-NHCH&O-NHCH&OOCH:
+
NH&H&O-(NHCHrCO)r-NHCH&OOCH~
+ CHtOH
More recently Pacsu and Wilson (254) and Frankel and Katchalski (176) have shown that, under suitable conditions, long-chain polycondensation products may be obtained from amino acid and peptide esters. The materials obtained represented mixtures of homologous peptide esters containing 20 to 100 amino acid units. It may be noted a t this point that, in general, amino acid esters and dipeptide esters readily yield diketopiperazines rather than polycondensation products. In fact, when a diketopiperazine composed of like amino acid residues is desired, it is most convenient to prepare it from the corresponding amino acid ester with ammonia in alcohol. In this manner, there have been synthesized a variety of diketopiperazines, such as histidine anhydride, lysine anhydride, and serine anhydride (172). Proline methyl ester, in particular, cyclizes with great ease (223). For the preparation of diketopiperazines derived from two different amino acids, treatment of the dipeptide ester with ammonia in alcohol usually leads to the desired product. Examples of this procedure are, among others, the synthesis of kleucyl-L-alanine diketopiperazine from L-leucylL-alanine methyl ester (136) and glycyl-L-valine diketopiperazine from glycyl-L-valine methyl ester (164). A modification of this method, employing a lithium hydroxide solution saturated with carbon dioxide, was useful for the preparation of L-glutamylglycine diketopiperazine from a-L-glutamylglycine ethyl ester (78). Fischer’s first attempts to develop a general method of peptide synthesis led him to prepare the carbethoxy derivative of glycylglycine ethyl ester by treatment of the dipeptide ester with ethyl chlorocarbonate (149). C rH ,OCOCl
+ NH 2CH $30-N
HCH zCOOC zH i -+ C~HsOCO-NHCH&O-NHCH&OOCrH‘
It was his hope, in this way, to introduce a substituent which would protect the reactive amino group from further attack in the course of condensation reactions and which could also be removed without hydrolysis
8
JOSEPH 6. FRUTON
of the peptide bonds of the synthetic product. When the carbethoxy peptide noted above was heated for 36 hours with m-leucine ethyl ester, ethyl alcohol was eliminated, and the resulting product was carbethoxyglycylglycylleucine ethyl ester (124). c4H1 I
CIHrOCO-NHCHICO-NHCH
lCOOCIH6
+ NHdHCOOC aH
L ---t
C4H9
I
C~H~OCO-NHCH~CO-NHCH~CO-NHC?HCOOC~H,
Fischer soon realized, however, that such condensation reactions are increasingly difficult as the peptide chain is lengthened and that even in the preparation of smaller peptides the yields are low. For this reason he turned his efforts t o the development of other methods of peptide synthesis. Of some interest in this connection is the behavior of carbethoxyglycylglycine ethyl ester on prolonged hydrolysis with alkali. Fischer (124) obtained a product which he formulated as glycylglycine carbamino acid : HOOC-NHCHaCO-NHCH&OOH
I t appeared, therefore, that the carbethoxy group could be removed from the acylated peptide ester without cleavage of the peptide bond. However, the complex nature of the reaction led Fischer t o abandon further study of its mechanism. It remained for Wessely (296) t o show that the product obtained by Fischer was actually carbonyl-bis-glycine, and the following sequence of reactions was suggested t o explain its formation:
It may be added that, in the same paper in which he described the hydrolysis of the carbethoxyglycylglycine ethyl ester, Fischer also reported the synthesis of carbonyl-bis-glycylglycine by the treatment of the dipeptide ester with phosgene (in toluene), followed by saponification : C1
/ co \
NH&H&O-NHCHiCOOCiHr
+ Cl
/
4
NH2CH1CO-NHCHaCOOCzHr
NHCH1CO-NHCHZCOOC~HK
co
‘NHCHICO--NHCH1COOC2Hr
9
SYNTHESIB OF PEPTIDES
Such compounds may also be prepared by the reaction of the sodium salts of amino acids or peptides with phosgene.
3 . Synthesis of Peptide Derivatives by Means of Acylamino Acid Chlorides and Azides While working on the synthesis of hippuric acid, Curtius found, in 1881 (104), that one of the products of the interaction of benzoyl chloride and glycine silver was the substance benzoylglycylglycine. As pointed out by Fischer (129) in the course of a polemic with Curtius, this reaction is rather complex in nature, and, although it may be considered to represent the first recorded synthesis of a well-defined peptide derivative, the method is not suitable for general application. Curtius’ studies, during the period 1890-1900, on the reactions of hydrazides and azides, led him to use these in the synthesis of peptide derivatives. After the report by Fischer and Fourneau (149) that dipeptides could be made by the partial hydrolysis of diketopiperazines, Curtius (106) described the use of azides of benzoylamino acids or peptides according t o the following reaction, illustrated for the case of the synthesis of benzoylglycylglycylglycine: C,HrCO-NHCH*CON,
+ NHsCHrCO-NHCHtCOOH
+
+ HNr
CoH‘CO-NHCH&O-NHCH&O-NHCH~COOH
In order to obtain the aside, the corresponding ester was treated with hydrazine hydrate, thus yielding a hydrazide, which was in turn converted to the azide by means of nitrous acid:
R I
C~H~CO-NH~HCON,
These reactions proceeded smoothly and, frequently, with excellent yield. Curtius and Levy (111) were able to make benzoyl(tetraglycy1)glycine ethyl ester by the condensation of benzoyl(diglycy1)glycine azide with glycylglycine ethyl ester, and with his collaborators, Curtius extended this method to the synthesis of benzoylated peptides containing alanine (110), aspartic acid (108), and aminobutyric acid (109). As was noted in the previous section of this review, Fischer’s first attempts to develop a general method of peptide synthesis led him to prepare the carbethoxy derivative of glycylglycine ethyl ester (144). When he conoluded that condensation of such esters with esters of amino acids was not feasible as a general procedure, he decided to convert the carbethoxyamino acids to the corresponding acid chlorides by means of
10
JOSEPH 8. FRUTON
thionyl chloride (127), a reaction which had been found by Meyer (246) to be suitable for the preparation of the acid chloride of pyridine carboxylic acid :
By warming carbethoxyglycine or carbethoxyglycylglycine with thionyl chloride at 35-10', noncrystalline products were obtained which were used directly for coupling in ethereal or in chloroform solution with amino acid or peptide esters. From the resulting carbethoxy peptide esters, the corresponding acids could be prepared by saponification. In several cases, these acids could be converted to acid chlorides with thionyl chloride, and the peptide chain lengthened by coupling with amino acid or peptide esters. In principle, the above methods of Curtius and Fischer provide the basis for the further development of the techniques of peptide synthesis. All the subsequent procedures for lengthening the peptide chain have involved the conversion of the carboxyl group of an amino acid into forms which permit reaction with the amino group of another amino acid. Of the various derivatives of carboxylic acids which have proved useful for this purpose, the aeides and chlorides have been of the most general value. Indeed, in many cases, the older azide method of Curtius is preferable to the chloride method, particularly in coupling reactions involving acyl peptides (cf. page 26). Although Fischer showed in 1905 (132) that it was possible to convert free amino acids to acid chlorides, it was realized that, in order to permit smooth coupling reactions, the amino group had to be blocked by acylation, or otherwise modified to avoid complicated side reactions in the course of the conversion of the carboxyl group to an acid chloride. In addition to the carbethoxy and beneoyl groups mentioned above, a variety of acyl substituents were introduced into peptide chemistry. In addition to others to be discussed later may be mentioned the naphthalenesulfonyl (144), phenylureido (170), benaenesulfonyl (145), and methanesulfonyl (207) groups. However, as long as it was necessary to remove an acyl substituent by hydrolysis, i t could not be used in the synthesis of free peptides, since attempts to eliminate the acyl group in thismanner invariably led to either partial or complete cleavage of the linkages between the amino acids. The essential problem of peptide synthesis thus became the development of methods which would obviate the necessity for the hydrolytic removal of an acyl substituent a t the end of a series of coupling reactions.
SYNTHESIS OF PEPTIDES
11
It should be added, however, that the methods diecussed in this section have been of considerable value in the preparation of benzoyl peptides, as well as amides of benzoylamino acids and benroyl peptides, which were required for studies of the specificity of proteolytic enzymes 4. Use of Amino Acid Chlorides in Peptide Syntheds
As noted earlier, Fischer, in 1905 (132), described a method of peptide synthesis which obviated the necessity for blocking the amino group of an amino acid prior to the conversion of the carboxyl group to an acid chloride. This method takes advantage of the fact that several amino acids, when shaken with phosphorus pentachloride and acetyl chloride, are readily converted into crystalline hydrochlorides of the amino acid chlorides. In this manner, Fischer prepared the acid chlorides of Dcleucine, Lalanine, DLphenylalanine, and cproline. Peptide synthesis was effected, in the case of balanylglycine (133), by the addition of calanyl chloride hydrochloride to a solution of glycine ethyl ester in dry chloroform, neutralization of the hydrochloric acid by the addition of sodium methylate, and the saponification of the dipeptide ester with alkali. The yields obtained by this procedure are not too satisfactory, since, as can readily be seen, one of the possible side reactions is the forester. Only isolated mation of a diketopiperazine from the -tide instances of its application can therefore be cited. Fischer and Luniak (159) used cprolyl chloride to make cprolyl-cphenylalanine and Abderhalden and Kempe (19) made ctryptophylglycine by means of Ltryptophyl chloride. Havestadt and Fricke (206) have reported the synthesis of Dcalanyl-thistidine through the coupling of Dcalanyl chloride with Ghistidine methyl ester but, as pointed out by Hunt and du Vigneaud (220), the identity of their product is open to some doubt. It is clear, therefore, that, except for a few cases, the amino acid chlorides are not suitable reagents in peptide synthesis. 5. Use of u-Halogen A w l Halides i n Peptide Synthesis
The greatest of Fischer’s many important contributions to protein chemistry may be said to have been the invention of the first of several methods now available to circumvent the difficulty encountered in the removal, by hydrolysis, of an acyl substituent of a peptide without cleavage of the peptide itself. The first report of this method came in 1903, when Fischer and Otto (160) described the synthesis of glycylglycylglycine. Chloroacetyl chloride and glycylglycine ethyl ester were coupled, and the product was saponified to yield chloroacetylglycylglycine. The next step was the introduction of the free amino group
12
JOSEPH 8. FRUTON
of the peptide by treatment of the chloroacetyl dipeptide with 25% ammonia at 100':
+ NHaCHaCO-NHCHsCOOCsHa -+ ClCHaCO-NHCHaCO-NHCHaCOOCsHr + HCl ClCHaCO-NHCHsCO-NHCHaCOOH + NH, ClCHaCOCl
NHaCH&O-NHCHrCO-NHCHaCOOH
It was later noted that the amination could be carried out more effectively by allowing the reaction mixture to stand at room temperature for two to three days. This ingenious method was applied in a similar manner to the condensation of a-bromopropionyl bromide with glycylglycine ethyl ester, thus leading to the synthesis of Dbalanylglycylglycine (128). In rapid succession, there followed a memorable series of papers which described the extension of this method to the synthesis of peptides of Dbleucine (132),Drcproline (169), Dbphenylahine (130), and other amino acids. In Table I are listed the various halogen acyl halides used in connection with these syntheses. TABLE I Halogen Acyl Halides Used in Peptide Sytthcsia
Amino acid residue
Halogen acyl halide
Glycyl . . . . . . . . . . . . . . . . . Chloroacetyl chloride Bromoacetyl bromide Alanyl. . . . . . . . . . . . . . . . . a-Bromopropionyl bromide (or ohloride) Valyl . . . . . . . . . . . . . . . . . . a-Bromoiaovaleryl chloride Leucyl . . . . . . . . . . . . . . . . . a-Bromoiaocaproyl ohloride Isoleucyl . . . . . . . . . . . . . . . a-Bromo-&methyl-mthylpropionyl ohloride Phenylalanyl . . . . . . . . . . . . a-Bromo-pphenylpropionyl chloride Prolyl . . . . . . . . . . . . . . . . . . a,6-Dibromovaleryl chloride
One of the principal difficulties of this method arose from the fact that those members of the group of a-halogen acyl halides in which the acyl group waa propionyl or larger could occur in at least two stereoisomeric forms (135). Thus, if the racemate were to be used in a reaction with an optically active amino acid or peptide, there would be produced a mixture of diastereoisomera. Except in rare instances, where the differences in the rates of reaction for the two isomeric acid halides are considerable, approximately equal amounts of the diastereoisomers would result, thus making the separation by fractional crystallization a laborious operation of dubious outcome. Accordingly, Fischer set about to prepare halogen acyl halides from optically active a-halogen acids. With Warburg (131) he prepared the optically active 2-bromopropionyl
13
SYNTHESIS OF PEPTIDES
chloride by treatment of levorotatory a-bromopropionic acid with thionyl chloride. This acid was obtained in two ways: either by resolution of the synthetic dl acid, or by treatment of talanine with nitrosyl bromide according to the method of Walden (293). Fischer assumed that a change of configuration (“ Walden inversion”) had taken place in the course of the latter reaction. Subsequent work (180) showed, however, that the 1-bromopropionic acid obtained from Galanine retains the configuration of the amino acid. The coupling of the l-bromopropionyl chloride with glycine ethyl ester, followed by saponification of the coupling product, gave a substance which was then subjected to amination to yield the peptide now recognized to be D-alanylglycine. Actually Walden inversion had occurred during the amination of the bromopropionylglycine rather than during the synthesis of the bromo acid, as Fischer had assumed (136, 252a). Despite this error, Fischer was, in general, correct in the conclusion that, in order to synthesize peptides containing amino acids of the L series (“natural” amino acids) it wm necessary, in the halogen acyl halide method, to prepare halogen acids from amino acids of the D-series (“unnatural” amino acids). These pamino acids could be obtained only by the resolution of synthetic racemic products, and Fischer and his collaborators devised a number of excellent, albeit time-consuming, procedures for achieving this end. The synthesis of D-alanylglycine from Galanine and glycine ethyl ester may be summarized in the following scheme: CHa NHIAHCOOH L-Alanine
+NOBr
CH8 BrhHCOCI
+ NH&H&OOC:H,
8“’
CH,
+sOCli
Br AHCOOH Z-Bromopropionic acid
Br HCO-NHCHgCOOH I-Bmmopropionylglycme
XH’ - 1”’ + Br
NHI
CHa
BrbHCOCI
HCO-NHCHICOOCXH,
NsOH
NHx HCO-NHCHXCOOH D- AhIlylgly Che
With the aid of several optically active a-halogen acyl halides, Fischer, and later Abderhalden, succeeded in synthesizing an impressive aeries of peptides,. in which all the components were either glycine or higher amino acids of the L or D series. The most notable achievements in this regard were the synthesis of the octadecapeptide bleucyl(triglycyl)-~-leucyl(triglycyl)-~-leucyl(octaglycyl)glycine which Fischer made in 1907 (138) and of a nonadecapeptide prepared by Abderhalden and Fodor in 1916 (10). I n the preparation of long peptides of this type, special care had to be taken in the conversion of halogen acyl peptides to
14
JOBEPH 8. FRUTON
the corresponding chlorides. Fischer effected this reaction by means of phosphorus pentachloride and acetyl chloride, as in the case of a-d-bromoisocapronylglycylglycylglycine. The resulting chloride could then be coupled with (pentaglycj 1)glycine in alkaline solution. The bromoisocapronylpeptide was then aminated with liquid ammonia, since aqueous ammonia was not effective, and the decapeptide was coupled with a-d-bromoisocapronylglycylglycylglycylchloride. Needless to say, as the chain length increased, the experimental difficulties became more serious. Furthermore, by coupling optically active halogen acyl halides with amino acids such as cphenylalanine, Ltyrosine, L-histidine, L-cystine, etc., or peptides of the L series such as glycyl-Ltyrosine, a large variety of interesting products were obtained by Fischer and Abderhalden, among others, and used for the study of the specificity of proteolytic enzymes. Despite the ingenuity and wide applicability of this method, its many difficulties greatly limit its general application. In addition to the labor involved in the synthesis of optically active peptides, the method has disadvantages in several other respects. First, the reaction of N-(halogen acyl) hydroxyamino acids with phosphorus pentachloride is complicated and, even if the hydroxyl group is blocked, as in the case of 0-carbomethoxy-N-chloracetyl-ctyrosine,chlorination of the carboxyl group may result in complete racemization (141). Surprising deviations from the expected result were also observed in. the course of the amination of several halogen acylamino acids. Thus, when Fischer attempted to prepare L-leucyl-cproline he found ,that, on treatment of a-d-bromoisocapronyl-tproline with ammonia, there resulted, not the expected dipeptide, but rather the substance a-hydroxyisocapronyl-L-prolinamide (162) : CHI
-
CHI
‘ck
lNH4
COOH
\ck \CHI
CONHI
AH*
AH-CH,
CHa
bH-cHa
SHa Br HCO-N
:1. CHI-
b
HO HCO-N Ha
< I CHI-
Hi
Later, it was found that bromoisocapronyl-N-phenylglycineundergoes a similar reaction on treatment with ammonia (151): CHI
‘ck
CHI
CHa \CH/
CH,
15
SYNTHESIS OF PEPTIDES
Another instance of anomalous behavior on amination was noted in the case of a-bromo-8-phenylpropionylglycine,from which, on amination, hydrogen bromide is eliminated to yield a derivative of cinnamic acid (146):
-
C6Hr
C6Hr
INHI
AH*
&
Br HCO-NHCHZCOOH
AH ISHCO--NHCH,COOH
Finally, it must be mentioned that the optical purity of many of the peptides prepared by the halogen acyl halide method is open to doubt. While the values for the rotation of the peptides described by Fischer have, in most cases, been confirmed by the application of newer methods of peptide synthesis, the synthetic products described from other laboratories have frequently been found to be partially racemic. One of the many examples to illustrate this is the recent report of Schott et al. (267) that the optical rotation of Gleucylglycylglycine is actually 12 to 15’ higher than that reported by Abderhalden and Fodor (lo),who prepared this peptide by amination of the product obtained by the reaction of ad-bromoisocapronyl chloride with glycylglycine.
6. “ Azlactone ” Method for Synthesis of Peptides In the classical methods of Curtius and Fischer, discussed in preceeding sections of this review, individual amino acids or peptides were linked to one another by first converting the carboxyl group of one of the reactants to an acid chloride or to an azide, which would then react readily with the amino group of the other reactant. The work of Mohr and Strohschein (249),in 1909,showed that, in place of such acylamino acid chlorides or asides, it was possible to use, for such reactions, the azlactones (also termed oxazolones) of benzoylamino acids. In particular, they prepared benzoyl-Dcalanine azlactone by the reaction of benzoylalanine with acetic anhydride, and coupled the product with glycine ethyl ester: CH:
C6H,CO--NH~HCOOH
-
CHS-CH-CO
(CH aC0) 2
0
A\c/ d I
+ NHICHICOOCIH~
__
F
CsHsCO-NH
A-
HCO-NHCH&OOCZH~
The wider application of this method to the synthesis of peptides did not develop until 1926, when Bergmann and his collaborators reported
16
JOSEPH 8. FRTlTON
eleqant procedures for the synthesis of phenylalanyl and tyrosyl peptides (60,67,90). In their work, use was made of the azlactones of a-ace& aminocinnamic acid (described by Erlenmeyer and Frllstiick, 121) and of a-acetamino-p-coumaric acid (the corresponding benzoyl compound was described by Erlenmeyer and Halsey, 122). For example, it was possible to synthesize cphenylalanyl-cglutamic acid and Pphenylalanyl-cglutamic acid (67) by the following series of operations:
co
C,HbCH=C-
rs\,/b +
COOH
&Ha
NHJXCOOH
AHr
COOH
+
AHt
CoH,
AH,
AH
bHt
CHaCO-NH&CO-NH
I
Ha __*
CJ%
bHt
&HI
AH1
catslyat
A
c:
HCOOH
COOH
COOH
-
CHaCO-NHAHCO-NH HCOOH Acetyl-cphenylalanyl-cglutamio acid Acetyl-D-phenylalanyl-~glutamic acid
mild hydrol.
I
CaHr
bHt
AHa
AH,
NHl&HCO-NH HCOOH cPhenylalany1-cglutamic acid D-Phenylalanyl-cglutamicacid
With regard to the first step, it may be mentioned that the azlactone can be coupled either with the amino acid ester in an organic solvent or with the sodium salt of the amino acid in aqueous solution. The product of the reaction is an acetyldehydrophenylalanylamino acid (or its ester). In many cases, the coupling also proceeds in good yield with the sodium salt of the amino acid in a mixture of acetone and water. If acetyldehydrophenylalanylglycine is treated with benzaldehyde and acetic anhydride, it is converted to the acetyldehydrophenylalanyldehydrophenylalanine aslactone (51) : CaHr AH CH,CO-NH!!CO-NHCHaCOOH
+ CaHrCHO + (CHaC0)iO
+
CHaCO-NH
CaHr CaHr I I
r8" C=N
CO
The coupling of this aslactone with an amino acid permits one to lengthen the peptide chain and, in this manner, a notable variety of long-chain acetyldehydrophenylalanyl peptides have been prepared by Doherty et d.(115). I n these studies, aalactones were also made by the action of
17
SYNTHESIS OF PEPTIDES
acetic anhydride on acetyldehydrophenylalanylpeptides in which trana8-phenyl-Dcsenne was the terminal amino acid :
CHSCO-NH
CcHi
C~HI
AH
bHoH
II CO-NH
c:
-
C d h C&
(CHC0)sO
HCOOH
AH
b
CHaCO-NH C=N
AH
eCO
Ld
In passing, mention may be made of two aspects of the chemistry of the acyldehydroamino acids and peptides. The first of these concerns the conversion of acetyldehydrophenylalanylamino acids, by mild acid hydrolysis, to the corresponding phenylpyruvylamino acids (189) : CeH6 AH CH,CO--NHIICO--NH
x
CeH, AH,
H c o o H -, bCO--NH
x
HCOOH
The second point of interest lies in the use of the chloroacetyl group as an acyl substituent in dehydroamino acids and peptides. Thus, from the azlactone of chloroacetyldehydrophenylalanine (prepared from chloroacetylphenylserine, 7), there may be obtained, on hydrolysis, an acylamino acid which, on amination by the method of Fischer, yields glycyldehydrophenylalanine (65). The last-named substance hasproved to be a substrate for the enzyme dehydropeptidase, present in animal tissues, which is specifically adapted to the hydrolysis of dehydropeptides (64,304). If, instead of hydrolyzing the chloroacetylazlactone, it is coupled with an amino acid such as eglutamic acid, tripeptides such as glycyldehydrophenylalanyl-Lglutamicacid may be prepared (65). Returning now to the reaction scheme for the synthesis of phenylalanylglutamic acid by the azlactone method, the acetyl dehydropeptide must next be converted to a saturated compound. This is readily effected by catalytic hydrogenation at low pressure with palladium black (298) or other related catalysts. In the case of acetyldehydrophenylalanyl-cglutamic acid, two diastereoisomers are formed. Under favorable circumstances, as in the example under discussion, the two isomers may be separated by fractional crystallization. Sometimes, however, such ready separation cannot be achieved. Thus, the diastereoisomeric beneoyl-D- and cphenylalanyl-carginines appear to have considerable affinity for one another, and all attempts to effect a resolution have hitherto been unsuccessful (186). This factor is a severe drawback of the azlactone method if optically active peptides or peptide derivatives are desired. The final step in the synthesis of phenylalanylglutamic acid, the
18
JOSEPH 8. FRUTON
removal of the acetyl group, is effected by hydrolysis with dilute acid. This operation entails considerable loss, however, owing to the formation of the diketopiperazine. Although the application of the azlactone method to peptide synthesis has involved primarily the use of the well-crystallized azlactones of acylaminocinnamic acid derivatives, the use of azlactones of other amino acids for this purpose has also been described. As noted earlier, Mohr and Strohschein (249) used the azlactones of benzoylamino acids, but it must be added that, in general, the azlactones of saturated acylamino acids are difficult to purify (236). Additional examples may be found in the work of Carter el al. (96), who prepared benzoylaminocrotonic acid azlactone by the treatment of benzoyl-Dballothreonine with benzoyl chloride and pyridine, or with two equivalents of acetic anhydride (98). This reaction is an extension to the aliphatic hydroxyamino acids of the reaction mentioned earlier for phenylserine. Another azlactone of an aliphatic acyldehydroamino acid is that of acetyldehydroleucine, which was prepared by Doherty et al. (115) from chloroacetyl-c leucine by treatment with acetic anhydride: CH,
CH,
CHI
‘Ck
CHa
‘Ck +
AH
CHIC=N/!CO
The synthesis of several acyldehydroamino acids from the corresponding halogen acylamino acids was described by Bergmann and coworkers in 1926 (58,66). Treatment of the halogen acids with acetic anhydride yielded azlactones from which there could be prepared, by hydrolysis, acetyldehydrophenylalanine,acetyldehydrotyrosine, and propionyldehydroaspartic acid. An excellent review of the chemistry of the azlactones has been provided recently by Carter (95), who has also made important contributions to the study of the stereochemistry of these substances (97).* Recently, there has been described (103s) the synthesis of an extensive series of thiazolones, the sulfur analogues of the azlactones, and it
* For a report of the extensive work on adactones studied in connection with the investigation of the chemistry of penicillin, see the chapter by J. W. Cornforth in ( l O a ) . Among the many important findings mentioned is the demonstration that some of the compounds previously thought to be a-acylamino acyl halides (e.g., hippuryl chloride) are actually hydrohalides of azlactones.
19
SYNTHESIS OF PEPTIDES
may be expected that these will prove to be valuable new reagents in peptide synthesis.
7. Condensation of Keto Acids and Amides Reactions that involve the condensation of keto acids with amides are of interest in peptide synthesis, not primarily because of their preparative value, but because of their possible relationship to the biological synthesis of peptide bonds (54). Bergmann and Grafe showed in 1930 (56) that pyruvic acid and acetamide readily react to form u,a-diacetaminopropionic acid, which, on treatment with acetic acid, is converted to a-acetaminoacrylic acid (acetyldehydroalanine). Hydrogenation of the acetyldehydroamino acid with palladium black as the catalyst gives acetyl-m-alanine.
CHaCOCOOH
+ 2 ClFaCONH*+ CHIlHCOCHa COOH
-
hCOCHI CHa CHICO-NHECOOH
2
It will be recalled that acyldehydroamino acids have also been prepared by the treatment of halogen acylamino acids with acetic anhydride (66). If, instead of acetamide, chloroacetamide is used for the condensation with pyruvic acid, a-chloroacetaminoacrylic acid is obtained, * and a subsequent amination leads to the synthesis of glycyldehydroalanine, which, on hydrogenation, gives glycyl-Dcalanine. Glycyldehydroalanine, like glycyldehydrophenylalanine, is hydrolyzed by the enzyme dehydropeptidase (64,198) : R
R AH NH&HaCO-NHbCOOH
-+
NHaCH,COOH
+ NHa +
A representative of an interesting group of substances was obtained when a,a-diacetaminopropionic acid was converted to an azlactone by treatment with acetic anhydride: A more satisfactory method for the synthesis of chloroacetyldehydroalanine has been described recently by Price and Greenstein (255b). The procedure involves the reaction of chloroacetonitrile with pyruvic acid in the presence of dry hydrogen chloride.
20
JOBEPH 8. FRUTON
This azlactone reacts rapidly with amino acids (e.g., glycine, alanine, phenylalanine) in alkaline solution to give diacetaminopropionylamino acids, which, upon mild hydrolysis, give pyruvylamino acids (57,271) : NHCOCH, R CHJCO-NH
I NHCOCHz
A
HCOOH + CHaCOCO-NH
1
HCOOH
Mention was made earlier in this review of the synthesis of phenylpyruvylamino acids by the hydrolysis of acetyldehydrophenylalanylamino acids (189). Shemin and Herbst (271) have shown that the oximes of the pyruvylamino acids may be reduced with Adams' platinum oxide catalyst to yield dipeptides, in analogy to the synthesis of amino acids from the oximes of keto acids (205) : RCH,LO-NH
r
NHz
HCOOH
2RCHAHCO-NH
x
HCOOH
In recent years, much interest has been evinced in the possibility that ketoacylamino acids (e.g., pyruvylalanine) might serve as acceptors of amino groups in enzyme-catalyzed transamination reactions similar to those discovered by Braunstein and Kritzman (93,209) in the case of pyruvic and a-ketoglutaric acids. Although this possibility has not received experimental support thus far, it should be mentioned that Herbst and Shemin (211) have shown that a transamination reaction can be effected by heating pyruvyl-DL-alanine with a-aminophenylacetic acid. The products of the reaction are both racemic Dtalanyl-DLalanines, benzaldehyde, and carbon dioxide. The last two products arise from the decomposition of phenylglyoxylic acid : CH,COCO-NH
XH'
HCOOH
NHa
+ CaH,AHCOOH --+
bH' X"'
+
CHI HCO-NH HCOOH CaHrCOCOOH -+ CaH'CHO + CO,
This reaction is analogous to the transamination between keto acids and aromatic amino acids studied by Herbst and Engel (208,210).
21
SYNTHESIS OF PEPTIDES
8. Use of N-Carbonic Acid Anhydrides of Amino Acids
It will be recalled that Fischer and Otto prepared several carbethoxyamino acids, which could be converted to acid chlorides by treatment with thionyl chloride (160). Coupling of the carbethoxyamino acid chlorides with esters of amino acids or peptides gave carbethoxy peptide esters (cf. page 10). Leuchs and collaborators (228-230) continued the study of the carbethoxyamino acid chlorides, and showed that these substances readily form N-carbonic acid anhydrides with the elimination of ethyl chloride. Thus, carbethoxyglycyl chloride gave rise to glycine N-carbonic acid anhydride: C*HrOCO-NHCH&OCI
+ CO-NHCHrCO
L
+ CIH~CI
A
In the presence of water, this anhydride, and those of other amino acids (112), rapidly opened with the elimination of carbon dioxide and the regeneration of the amino acid. In concentrated solution, the evolution of carbon dioxide was accompanied by the deposition of an insoluble polymer. Leuchs showed that this polycondensation could also be effected by warming a solution of an N-carbonic acid anhydride in organic solvents containing small amounts of water. On treatment with hydrogen chloride in alcohol, the carbonic acid anhydrides were converted into the hydrochlorides of the corresponding amino acid esters. These findings were confirmed and extended by Curtius (112), Fuchs (192), Wessely (295), and Woodward and Schramm (301), among others. The formation of the anhydrides has been effected with several carboalkyloxy derivatives of amino acids. Although the carbomethoxy derivatives (prepared with methyl chlorocarbonate) are especially suitable for this purpose, carbethoxy (from ethyl chlorocarbonate) and carbobenzoxy (from benzyl chlorocarbonate, 74) derivatives are converted readily into N-carbonic acid anhydrides. The N-carbonic acid anhydrides react not only with water and alcohol, as noted above, but also with amines and, in particular, the amino groups of amino acids and peptides. The formation, in moist organic solvents, of polymeric products is believed to be the result of a chain reaction in which a molecule of anhydride is opened by water and the amino group thus liberated is made available for reaction with another molecule of the anhydride. In this manner, there is formed a dipeptide which can react with a third molecule of anhydride to give a tripeptide, and so forth:
22
JOSEPH 8. FRUTON
R qO-NH
Lo-l
c:
HFO
+
+ COl
4
R +I8
L
O
I
1
(A)
(A) R
R
CO--NHbHCO P
NHICHCO-(NH
R
HCO).-NH
I:
HCOOH
+ n CO,
Polymers of glycine and sarcosine have been prepared by Wessely (297) and, more recently, Frankel and Katchalski (176a) have obtained polymers of t-carbobenzoxylysine (cf. page 59), which, upon hydrogenation (223a), gave polymers of lysine. In addition, Woodward and Bchramm (301) have made polymers of phenylalanine or leucine, as well as products arising from the interaction of the N-carbonic acid anhydrides of both these amino acids. In the presence of a n excess of an amino acid (sodium salt) in aqueous solution (273) (or an amino acid ester in a n organic solvent, 220), the N-carbonic acid anhydride will react to give a moderate yield of the expected dipeptide (or dipeptide ester). Sigmund and Wessely (273) made DL-phenylalanylglycinein this manner: CIH,
CSI
~ H Z
~ H s
CO-NH
HCO
+ NHiCHzCOOH
4
L
NH, HCO-NHCH1COOH
L 3 - J
With glycylglycine, the tripeptide DL-phenylalanylglycylglycine was obtained. * It will be clear from the foregoing, however, that, in the case of a reaction between a N-carbonic acid anhydride and an amino acid or peptide, there is considerable possibility of side reactions involving polymerization or hydrolysis. It may be questioned, therefore, whether it will be feasible to apply this method generally to the synthesis of peptides of precisely known structure, since the separation of the desired peptide from the polymeric product and the amino acid formed on hydrolysis may be attended with some difficulty. The usefulness of the N-carbonic acid anhydrides in several special aspects of peptide synthesis should be emphasized, however. For example, they have proved to be It has recently been reported (J. L. Bailey, International Biochemical Congress, Cambridge, 1949) that N-carbonic acid anhydrides may be used successfully for the synthesis of a variety of peptides by allowing an anhydride to react either with two equivalents of amino acid ester or with one equivalent of amino acid ester plus a tertiary base. At the present writing, details of this important extension of the Weeeely method are not available.
23
SYNTHESIS OF PEPTIDES
of considerable value in the synthesis of iysine and ornithine peptides (cf. page 59).
9. Uae of ToluenesuljonylaminoAcids in Peptide Synthesis It was noted earlier in this review that, prior to the introduction of the halogen acyl halides into peptide chemistry, Fischer had attempted to use acylamino acid chlorides for peptide synthesis in the hope that an acyl substituent might be found which could be selectively removed by mild hydrolysis. The carbethoxy group was found to be unsuitable for this purpose, and although the later work of Bergmann et al. (67) showed that some acetyl dipeptides could be deacetylated with dilute acid, other studies emphasized the fact that the selective hydrolysis of even the acetyl group could not be relied upon in all cases. There was need, therefore, for methods in which acyl substituents could be removed from peptides by procedures not involving hydrolysis with strong acids. * The first such method was proposed by Schoenheimer (266), who took advantage of the discovery of Fischer (142) that p-toluenesulfonylamino acids could be converted to the parent amino acid by reduction with a mixture of hydriodic acid and phosphonium iodide: R
c:
CHJCOHSOZ-NH HCOOH
HI
PHd
R
+
I
CHJCOH&~HNH&HCOOH
Schoenheimer showed that, under the conditions of this reduction, the hydriodic acid does not hydrolyze peptide linkages to an appreciable extent. By employing the azides of toluenesulfonylamino acids, he was able to synthesize several peptides by the following series of reactions, illustrated for the case of glycyl-Dbalanine: CH~C*H~SO-NHCH&ONJ
+ NHz
iH*-
CHJC~H~SOZ-NHCH&O-NH HCOOH
HI
PHJ
NHzCH*CO-NH
IHa
HCOOH
As was shown by du Vigneaud and Behrens (287), the toluenesulfonyl group may be removed from an amino acid derivative by reduction with sodium in liquid ammonia. Recently, Ehrensvlird (1 18s) has described the use of phenylthiocarbonyl chloride as a reagent for the protection of amino groups in peptide synthesis. The phenylthiocarbonyl group is stable to acids, but may readily be split off with dilute alkali in the preeence of lead hydroxide or lead carbonate. At the present writing, only a preliminary account of this method is available.
24
JOSIPPH 8. FEUTON
I n addition to the Curtius aeide method, the Fischer method for the conversion of acylamino acids to the corresponding acid chlorides (with thionyl chloride or phosphorus pentachloride) also serves satisfactorily for the toluenesulfonylamino acids and obviates the necessity for the preparation of the ester and the hydradde prior to the coupling. As noted earlier in the case of other acyl peptides, however, when toluenesulfonyl peptides are to be used for coupling reactions to lengthen the peptide chain, the azide method may be expected to give better yields. It is regrettable that insufficient information is available in the literature regarding the applicability of the Schoenheimer method to the synthesis of peptides containing some of the more complex amino acids. Woolley (303) has recently reported the synthesis of a mixture of the diastereoisomeric D and cserylglycyl-cglutamic acids by the use of this method. 10. Use of Azidoacyl Halides in Peptide Synthesis
An interesting modification of the halogen acyl halide method for peptide synthesis was developed by Bertho and Maier (91) and by Freudenberg, Eichel, and Leutert (179), who used a-azidoacyl halides for coupling reactions with amino acids. It had been shown by Forster and Fierz (175) that the reduction of a-azido acids with aluminum amalgam led to the formation of amino acids. In a similar manner, reduction of an azidoacylamino acid, e.g., dl-a-azidopropionylglycine gives DG alanylglycine. CHI N8bHcoa
+ NHlCHaCOOH
iH'
NI HCO-NHCH&OOH
Hi
+ CH8
NHIAHCO--NHCHlcOOH
The reduction may be effected not only with aluminum amalgam but also by catalytic hydrogenation in the presence of platinum or palladium (91). Since the azidoacyl halides are usually prepared by treatment of the corresponding chloro or bromo compounds with sodium ande, optically active halogen acids would be required for the synthesis of optically active peptides. Mention has been made previously of the difficulties which attend the preparation of such acids. It is not surprising, therefore, that the use of the azidoacyl halides has not assumed an important place among the methods for the synthesis of peptides. Although this method does not offer any decided advantages from a preparative standpoint, it is of interest in the historical development of the subject under
SYNTHESIS OF PEPTIDES
25
review because of the use of reduction, rather than hydrolysis or amination, as the final step in the peptide synthesis.
11.
“
Carbobenzoxy ” Method of Peptide Synthesis
Of the numerous contributions to the development of methods for the synthesis of ‘peptides, two are outstanding in their importance. The first of these was the use, by Fischer, of the halogen acyl halides (cf. page ll), and the second was the invention by Bergmann and Zervas (74) of the procedure which they termed the “carbobenzoxy” method. The latter has now assumed the pre-eminent place among the techniques of peptide synthesis. The potentialities of the carbobenzoxy method are indicated by its recent use for the synthesis of the pentapeptide valylornithylleucylphenylalanylproline, by Harris and Work (204a). In their search for acyl substituents which could be removed from peptides without resorting to hydrolysis, Bergmann and Zervas were led to consider the possibility of introducing the benzyl group into such an acyl moiety, for it had been shown by Rothemund and Zetzsche (259), Freudenberg et al. (178),and Fischer and Baer (173)that benzyl groups attached to oxygen or nitrogen atoms could be removed readily by catalytic hydrogenation. Accordingly, Bergmann and Zervas prepared the benzyl analog of ethyl chlorocarbonate (used by Fischer as a reagent for amino acids) by the reaction of benzyl alcohol with phosgene: C6HrCHsOH
+ ClCOCl + CaHrCHsOCOCl + HC1
The product, benzyl chlorocarbonate, was called carbobenzoxy chloride and, in what follows, it will be referred to by the latter term.* Carbobenzoxy chloride can be condensed with all amino acids and yields excellent crystalline derivatives with most of them. Exceptions are tleucine, L-isoleucine and L-proline,for which crystalline carbobenzoxy derivatives have not been described as yet. Of particular advantage is the fact that, during the reaction of carbobenzoxy chloride with the sodium salts of optically active amino acids, no appreciable racemization is observed, in contrast to the result noted with acylating agents such as bensoyl chloride (71),acetic anhydride (71,290),or ketene (94). For the conversion of the carbobenzoxyamino acids into compounds which will react with the amino group of an amino acid or peptide, the procedures of most general application are the Fischer acid chloride method (with phosphorus pentachloride or thionyl chloride) and the Curtius azide method. The choice between these two methods will be At a recent meeting of the American Chemical Society (September, 1947), Stevens and Mdne have reported the use of ally1 chlorocarbonate in place of the benzyl compound.
26
JOSEPH 8. FRUTON
determined, in large part, by the nature of the carbobenzoxy compound to be used. For certain carbobenzoxyamino acids, such as carbobenzoxyglycine (74) or carbobenzoxy-cphenylalanine (85), the conversion to the acid chloride, when conducted under proper conditions, proceeds smoothly and in good yield. On the other hand, the preparation of carbobenzoxybleucyl chloride is not entirely satisfactory, and for better yields, the conversion of carbobenzoxy-Lleucine methyl ester to the hydrazide and azide appears preferable (80). Clearly, in the case of carbobenzoxyserine (and presumably threonine) , treatment with phosphorus pentachloride would be complicated by the replacement of the hydroxyl group by chlorine, and here the azide method is the method of choice (181). Where a-peptides of glutamic acid or aspartic acid are desired, neither the acid chloride nor the azide method can be used, since this would yield, upon coupling with amino acids, disubstitution products of the dicarboxylic amino acids. For these amino acids, special methods are required, and these will be discussed in a later section (cf. page 41). As noted previously in the case of other acyl peptides, when carbobenzoxy peptides are to be coupled with amino acids or peptides t o lengthen the chain, the azide method is to be preferred, since the acid chloride procedure apparently is accompanied by side reactions of the halogenating agent with peptide bonds (254). An additional point of some importance relating to the carbobenzoxyamino acid chlorides is their tendency to form N-carbonic acid anhydrides, referred to previously (cf. page 21). This is a troublesome side reaction and, to be kept at a minimum, requires rapid operation under anhydrous conditions and strong chilling of the reaction mixture during the halogenation and subsequent coupling. In the case of acylamino acid azides, an unwelcome side reaction is the occasional tendency for the occurrence of the Curtius rearrangement to isocyanate derivatives, which then react with amino groups to form ureido compounds, or with hydroxy groups to form urethanes (77,181). In the synthesis of peptides by the carbobenzoxy method, the carbobenzoxyamino acid chloride or azide is preferably allowed t o react in an organic solvent (e.g., ethyl acetate, ether) with the appropriate amino acid or peptide ester: R
c:
C6HrCHtOCO-NH HCOCl
R'
+ NHsCJ HCOOCH, -+ CaH,CHsOCO-NH
!
HCO-NH
R'
c:
HCOOCHI
+ HCl
The free esters of amino acids or peptides may be prepared from the corresponding hydrochlorides by either of the two methods devised by
27
SYNTHESIS OF PEPTIDES
Fischer. The first of these involves the neutralization of the amino acid ester hydrochloride with concentrated alkali in the presence of solid potassium carbonate, and immediate extraction of the free base with ether or ethyl acetate (149). In the case of peptide ester hydrochlorides, the concentrated alkali usually is omitted, and only carbonate is used. The second method, which is preferable when the amino acid ester has an appreciable solubility in water (e.g. , histidine methyl ester) , entails the neutralization of an alcoholic solution of the hydrochloride with the calculated amount of sodium methylate (172). Since, in the coupling reaction of an acid chloride with a base, one molar equivalent of hydrogen chloride is formed, two molar equivalents of base are needed for the complete utilization of a carbobenzoxyamino acid chloride. When, however, the ester of a peptide or of a costly amino acid is used in the coupling reaction, the second molar equivalent of base may be provided by shaking the reaction mixture with an aqueous solution of potassium bicarbonate. In the case of amino acid esters which are not too expensive (Le., those of glycine, glutamic acid, tyrosine), the use of two molar equivalents of ester is to be recommended, since this obviates the possibility of the concomitant hydrolysis of the acid chloride by the aqueous bicarbonate. Furthermore, the ester hydrochloride formed during the coupling usually crystallizes out and may be recovered for future use. Although the Schotten-Baumann reaction of the carbobenzoxyamino acid chlorides or azides with the sodium salts of amino acids or peptides is occasionally successful, in many cases it yields products which are difficult to purify owing to the hydrolysis of appreciable quantities of the chloride (or azide) to the corresponding carbobenzoxyamino acid. The separation of such by-products from the carbobenzoxy peptides formed during the coupling reaction is frequently a laborious operation and markedly reduces the yield of the desired material. When coupling amino acid or peptide esters with the azides of carbobenzoxy peptides, only one molar equivalent of ester is required, since the hydrazoic acid liberated during the reaction passes off as a gas.
CIHICHSOCO-NH
R l
R’ I
CIHrCHrOCO-NHkHCO-NHCHCOOCHa
+ HNa
In contrast to the reaction between acid chlorides and amino acid esters, which proceeds quite rapidly (15-60 minutes are usually sufficient to
28
JOSEPH 8. FRUTON
ensure complete reaction), the azides react with amino acid esters much more slowly. For this reason, it is customary to allow the mixture to stand overnight a t room temperature. After the coupling reaction is completed, the solution of the carbobenzoxy peptide ester in ether or ethyl acetate is freed of any remaining free base and of carbobenzoxyamino acid (or carbobenzoxy peptide) arising as a result of hydrolysis, by successive extraction with dilute hydrochloric acid, dilute bicarbonate solution, and water. This procedure may require modification in special cases, however. For example, in the synthesis of carbobenzoxyglutamyl peptide esters in which the ycarboxyl is unsubstituted, extraction with bicarbonate must be avoided and the solution should be washed with hydrochloric acid, followed by several extractions with water. After the extraction of possible impurities has been achieved, the solution of the coupling product may be dried over sodium sulfate and concentrated under reduced pressure. In most cases, the carbobenzoxy peptide ester crystallizes out directly or may be induced to crystallize by the addition of petroleum ether. It is important that the purity of the carbobenzoxy peptide ester be established before proceeding to the next step in the synthesis, since, in the later stages of the carbobenzoxy method, the purification of the products is generally more difficult than in the case of the carbobenzoxy peptide esters. In order to convert the carbobenzoxy peptide esters to the corresponding carbobensoxy peptides, the former are saponified in acetone-water solution with slightly more than one molar equivalent of normal alkali. The product is then isolated by acidification of the solution and removal of the acetone under reduced pressure. Usually, crystalline products are obtained if the eater was pure; if an oily carbobenzoxy peptide results, however, it may be extracted with ethyl acetate, and the ethyl acetate solution then extracted with dilute bicarbonate. Acidification of the bicarbonate solution will precipitate the carbobenzoxy peptide and crystallization may sometimes be induced. Before proceeding to the discussion of the removal of the carbobenzoxy group from the acylated peptides, some comment may be inserted regarding the conversion of the carbobenzoxy peptide esters into products other than the corresponding carboxylic acids. I n particular, their reactions with hydrazine or ammonia have been used extensively in peptide chemistry. As shown by Curtius (cf. page 9), for the benzoyl peptide eaters, the reaction between carbobenzoxy peptide esters with hydrazine in absolute alcohol leads to the formation of well-crystallized hydrazides, which may, in turn, be converted to azides, thus permitting one to lengthen the peptide chain by means of further coupling reactions:
29
SYNTHES18 OF PEPTIDES
R'
R CaH,CH,OCO-NH
HCO-NH&HCOOCH~ R
C~HICHIOCO-NH
NHtNHi
R'
-
I:HCO-NH I:HCONHNH,R R CeH,CHaOCO-NH bHCO-NRI:HCON HNOI
I
The interaction of carbobensoxy peptide esters with dry ammonia in absolute methanol gives the corresponding amides, except in isolated cases, which will be discussed in a later section (cf. page 48). Such carbobensoxy peptide amides have proved to be useful where studies on the specificity of proteolytic enzymes have required substrates in which the terminal carboxyl group of the peptide chain is blocked: R C,H,CH,OCO-NH
R'
-
I:H C ~ N H A H C O O c H a
NHI
:
'R'
A
C~H~CHIOCO-NH H C S N H HCONH,
The final step in the carbobenzoxy method is the removal of the acyl substituent by catalytic hydrogenation with palladium black (298) or related catalysts a t pressures of somewhat more than one atmosphere. In the course of this hydrogenation, toluene and carbon dioxide are split off and the amino group is set free:
1
C~H,CH~OCO-NH HCO-NHR'2
CeH'CH.
+ COI + NH,
!
HCO-NHR'
The hydrogenation may be conducted in any of a variety of apparatus available commercially or which may be readily assembled in the laboratory. Whatever the form of the equipment, however, provision must be made for passing a stream of dry hydrogen through the system, and one must be able to determine from time to time whether carbon dioxide is still being evolved. As noted before, this operation represents the most important feature of the method, since the conditions are so mild as to preclude the scission of the peptide bonds. What is more, in the course of the development of the carbobensoxy method, substituents have been found for the polar side chain groups of several amino acids (e.g., the nitro group for the guanido group of arginine, the bensyl group for the sulfhydryl group of cysteine or for the imidasole group of histidine), which also may be removed by catalytic hydrogenation. Discussion of these special modifications of the carbobensoxy method will be found in subsequent sections of this review. A final attribute of the csrbobensoxy method which confers upon it
30
JOSEPH S. FRUTON
unusual advantages is the fact that, following the hydrogenation of a carbobenzoxy peptide, the solution contains only the free peptide and toluene. The catalyst is then removed by filtration, and the solvent (usually methanol plus a few drops of glacial acetic acid) and toluene may be removed completely by evaporation a t reduced pressure. The remaining product is the desired peptide uncontaminated with inorganic salts which may accompany peptides prepared by the halogen acyl halide method unless special treatment (addition of silver oxide, followed by hydrogen sulfide) is applied. If, in the course of the hydrogenation of a carbobenzoxy peptide in methanol, the free peptide separates, the addition of water prior to the removal of the catalyst will usually bring the peptide back into solution. It is obvious that the hydrogenation method may also be applied to carbobenzoxy peptide esters, and, in this case, it is necessary to add one molar equivalent of acid to neutralize the amino group which appears during the hydrogenation. In this manner, it is frequently possible to prepare peptide esters for coupling reactions. Thus, the compound carbobenzoxyglycyl-L-glutamyl-Ltyrosine ethyl ester was made by coupling carbobenzoxyglycyl chloride with cglutamyl-ctyrosine ethyl ester, the latter reactant having been obtained by the hydrogenation of the corresponding carbobenzoxy derivative (187) : CHiCOOH CsHdOH CaHrCH,OCO-NHCH&OCl
+
AH1 AH2 I I NHaCHCO-NHCHCOOCiHr-+ CHiCOOH CaHdOH
I"* I
CaHrCH*OCO-NHCHiCO-NH
HCO-NH
AHs HCOOCrHr
It may be added that the hydrogenation procedure has also been applied with good results to the preparation of amides of peptides, as well as amino acid amides. As in the case of the hydrogenation of the carbobenzoxy peptide esters, one molar equivalent of acid must be added to neutralize the amino group set free during the hydrogenation. It has been noted that glacial acetic acid is especially suitable for this purpose since the acetates of the amino acid or peptide amides crystallize more satisfactorily than do the corresponding hydrochlorides (188). Mention was made at the start of the discussion of the carbobenzoxy method that it is possible to remove, by catalytic hydrogenation, benzyl groups attached either to nitrogen or to oxygen atoms. It follows, therefore, that, if in a coupling reaction the benzyl ester of an amino acid ia used, the resulting carbobenzoxy peptide benzyl ester does not require
31
SYNTHESIS. OF PEPTIDES
prior saponification to yield the free peptide. The hydrogenation will remove both the N-carbobenzoxy and the 0-benzyl groups, thus eliminating one step in the peptide synthesis and frequently improving the over-all yield. An example is the coupling of carbobenzoxy-calanylglycyl azide with glycine benzyl ester, followed by the hydrogenation of the product to yield talanylglycylglycine (49). CHa CaH&HIOCO-NH
c:
+ NHnCH&OOCHsCaHr -+
HCO-NHCHzCONa
CaHbCH2OCO-NH
8"'
HCO-NHCHzCO-NHCHzCOOCH,CsH, CHa
HZ +
L
NHZ HCO-NHCHICO-NHCHiCOOH
One of the difficulties with the benzyl ester hydrochlorides of amino acids is that they cannot, in general, be prepared in high yield by the treatment of the amino acid with benzyl alcohol and hydrogen chloride, although examples of this procedure may be found in the literature (276). A better yield is usually obtained if the benzyl ester hydrochloride is made by the reaction of an amino acid N-carbonic acid anhydride with benzyl alcohoi and hydrogen chloride (86). Glycine benzyl ester hydrochloride has also been made from glycyl chloride hydrochloride and benzyl alcohol (203). It was noted in an earlier section that ptoluenesulfonylamino acids and peptides may be reduced by means of phosphonium iodide and hydriodic acid. Harington and Mead (202) showed that phosphonium iodide in acetic acid at, 45-50" would also eliminate the carbobenzoxy group, benzyl iodide being formed instead of toluene, as in the case of catalytic hydrogenation: R CtH,CHzOCO-NH
c:
HCOOH
PHd
CaHhCHJ
1
+ COz + NHI
HCOOH
The introduction of pbosphonium iodide as a reducing agent made it possible for Harington and Mead to adapt the carbobenzoxy method to the synthesis of glutathione since the metallic catalysts are readily inactivated by sulfur compounds. Of even greater importance for the synthesis of peptides of cystine, as well as of other amino acids, was the finding of Sifferd and du Vigneaud (272) that carbobenzoxy or benzyl groups could be removed smoothly by treatment with metallic sodium in liquid ammonia. The numerous fruitful applications of this reduction method will be discussed in later sections of this review. It has been noted (41) that, in some cases, carbobensoxy groups may
32
JOSBPH 8. FRUTON
be removed by treatment with absolute ethanol and dry HC1 a t 0'. Thus, carbobenzoxytyrosyltyrosine,when treated in this manner, gave an appreciable quantity of tyrosyltyrosine ethyl ester hydrochloride. This valuable observation has not received the further study it clearly merits. 12. Use of Phthalylamino Acida in Peptide Synlhe& Recently, Kidd and King have reported, in a preliminary communication (223b), a new method of peptide synthesis which takes advantage of the fact that the phthalyl group of phthalyl peptides may be removed by treatment with hydrazine, without scission of the peptide bonds. This ingenious technique is based on the observation of Ing and M a d e (221s) that N-alkyl phthalimides readily react with hydrarine to give phthalhydraride and the corresponding amine. Kidd and King have described the application of the new procedure to the synthesis of peptides of Lglutamic acid, using phthalyl-cglutamic acid anhydride as the acylating agent. At the present writing, the details of this work are not yet available. A more complete account of the phthalyl method has been published by Sheehan and Frank (270a), whose work was done independently of the English investigators, and who have applied it successfully to the eynthesis of glycyl-Dbphenylalanine, glycylglycine, glycyl-L-cysteine, and Dbphenyldanylglycylglycine. The salient features of the method are given in the following series of reactions for the synthesis of glycylphenylalanine.
The phthalyl method promises to be a valuable addition to the available techniques of peptide synthesis, and its further applications will be awaited with interest.
SYNTHESIS OF PEPTIDES
33
13. Enzymatic Synthesis of Peptide Derivatives
Recent experiments have given unequivocal proof for the expectation that, like other catalysts, the proteolytic enzymes may, under suitable conditions, cause the synthesis of peptide bonds aa well as their hydrolysis (62,M). Several examples of such enzymecatalyzed peptide synthesis are the following:
-
+ cleucinanilide papain benzoyl-cfeucyl-cleucinanilide(249) papain Benzoyl-cleucine + glycinadide benzoyl-cleucylglycinanilide(249) Obymotrypsin Benzoyl-ctyrosine + glycinanilide benzoyl-ctyrosylglycinanilide (40) papain Carbobenzoxy-cphenylalanylglycine+ ctyrosinamide Benzoyl-cleucine
r
carbobenzoxy-cphenyldanylglycyl-ctyrosie (101)
This type of peptide synthesis produces insoluble compounds, and , indeed, the insolubility of the products is an essential attribute of the method. This follows from the fact that the energy necessary for the synthesis of a peptide bond (about 3000 cal. per mole, 92) is provided by the removal, by crystallization, of the synthetic product from the solution. In the cases cited above, the solubility of the synthetic peptide derivative is less than its equilibrium concentration; consequently, in order to restore the balance of the equilibrium reaction, synthesis occurs, which in turn, causes more of the synthetic product to crystallize (54). At the present stage of its development, this method cannot be compared in usefulness for peptide synthesis with the procedures discussed earlier in this review. One of the principal difficulties lies in the removal of the substituent groups. While the carbobenzoxy group may be removed by hydrogenation, the scission of the amide or anilide linkage requires hydrolysis, with the attendant destruction of the peptide. Attempts have been made to effect enzymatic synthesis of peptide derivatives in which the terminal amino group is carbobenzoxylated and the terminal carboxyl group is linked with a benzyl group. While initial efforts (186)in this direction have not met with success, further exploration of such possibilities appears desirable. A further limitation of the enzymatic method becomes apparent if one considers the extreme specificity of action of the proteolytic enzymes. None of these enzymes acts a t peptide bonds indiscriminately, and each enzyme hydrolyzes or synthesizes only such peptide bonds as are present in the substrate in a certain structural setting. Extensive discussion of the specificity of proteolytic enzymes may be found in several recent review articles (47,52,182). One application of the synthetic capacity of the proteolytic enzymes,
34
JOSEPH 8. FRUTON
which takes advantage of the stereochemical specificity of enzyme action, has proved of considerable indirect value in peptide synthesis. As was noted in the Introduction, it is usually desirable t o employ, for the synthesis of peptides, either the t or D-amino acids, rather than the racemic forms. It has been shown that the stereochemical specificity of the proteolytic enzymes permits them to serve as agents in the resolution of Dtamino acids (190). This is illustrated below for the resolution of m-glutamic acid by activated papain: Carbobenroxy-Dcglutamic acid
+ aniline
Ipaprin
Carbobenroxy-ln-lllutamicacid Carbobenzoxy-Lglutsmic acid anilide Ohydrogenation D-
lutamic acid
1hydrol.
cGlutamic acid
T h e further exploitation of this method has provided procedures for the preparation of tamino acids from synthetic Dtamino acids, where the latter are less costly than are the amino acids isolated from protein hydrolyzates (cf. 112s). Furthermore, the enzymatic resolution represents an additional convenient method for, the preparation of Pamino acids.
111. SPECIAL ASPECTSOF PEPTIDE SYNTHESIS In what follows, there will be discussed some of the problems encountered in the synthesis of peptides containing particular protein amino acids, and attention will be given to the manner in which the general methods discussed earlier have been modified to meet these special problems. For the convenience of those who may wish t o discover rapidly whether a particular peptide or useful intermediate has been synthesized, tables have been prepared to accompany the sections on the individual amino acids.* One series of tables lists some of the more useful or interesting derivatives, their melting points, and references to papers in which their synthesis is described. In addition, another set of tables is devoted t o free peptides and their optical rotation (when available), since the majority of the peptides listed contain optically active amino acid residues. The bibliographic citations in the tables for the peptides give the literature sources for the values of the rotation listed in the tables. Where several references are cited for a single peptide, the rotation which is included in the table is, in general, the largest positive or negative value reported. In view of the large number of investigators whose data have been used in the preparation of these tablee, no assurance can be given of the accuracy of the melting points or The author ia greatly indebted to Dr. Sofia Simmonds for invaluable assistance in the preparation of the tablee and of the bibliography.
SYNTHESIB OF PEPTIDES
35
optical rotations as listed. The reader is urged to examine the original papers for details. It should be stressed that the material in the tables represents a selection from the extensive literature on peptide synthesis, and no attempt has been made to provide a complete listing of all peptides and peptide derivatives which have been described. Such a task, it was felt, did not accord with the purpose of this review. Although an effort has been made to list the peptides .and their derivatives according to some rational order, which will be readily apparent upon inspection of the tables, the compounds are grouped arbitrarily according to the particular amino acid residue which appears to be of primary interest. Thus, glycyl-btyrosine will be found in the table which lists the tyrosine peptides and not in the table for the glycine peptides. In order to conserve space, each peptide or peptide derivative is listed only once in the tables. In searching for a reference to a complex peptide or peptide derivative composed of several different amino acid residues, the reader is advised to examine more than one table. 1. Peplides of Glycine, Alanine, Vdine, Leucine, and Isoleucine
This group of amino acids may be treated as a unit since the side chains do not contain special reactive groups. The ester hydrochlorides of the optically active forms of all of them have been prepared and may readily be converted to the free base by treatment with alkali in the presence of an organic solvent (cf. page 27). The free esters may then be coupled with a variety of reactive amino acid derivatives. It is appropriate to mention a t this point that, although Galanine and Gleucine are currently readily available, this cannot be said for the corresponding optically active forms of valine and isoleucine. As has been noted elsewhere (185), however, the two last-named amino acids are now accessible in DL form, and the development of resolution methods based on the stereochemical specificity of proteolytic enzymes seems feasible (cf. page 33). The attachment of a glycyl residue to the amino group of an amino acid or peptide presents no especial difficulty. This can readily be achieved either by means of chloroacetyl chloride (or bromoacetyl bromide), followed by amination, or, preferably, by coupling with carbobenzoxyglycyl chloride, with subsequent hydrogenation. In general, if simple glycyl dipeptides are desired, the chloroacetyl chloride method will be found satisfactory. For the preparation of longer peptides or of derivatives with glycine a t the amino end of the peptide chain the use of carbobenzoxyglycyl chloride is to be preferred. The latter reagent may
36
.
JOSEPH 8 FBUTON
TABLE I1 Derivatives of cflyn'ne and Alanine Compound
M . p., "C. (ref.)
Glycine anhydride. . . . . . . . . . . . . . . . . . . . . . . . . . . ca . . 305 (261) Glycine benryl ester HC1 . . . . . . . . . . . . . . . . 139-140 (203,260) Glycine-N-carbonic acid anhydride. . . . . . . . - (228) Barcosine-N-carbonic acid anhydride. . . . . . . . . . . 99-100 (273) Glycylglycine ethyl ester HCI . . . . . . . . . . . . . . . . . . 182d (149) Glycyldehydroalanine. . . . . . . . . . . . 192-193 (56) Bencoylglycinamide . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 (132) Bensoylglycyl chloride . . . . . . . . . . . . . . . . . . . . . . . - (132) Beneoylglycylglycine . . . . . . . . . . . . . . . . . . . . . . . 208 (132) Carbobenzoxyglycine. . . . . . . . . . . . . . . . . . . . . 120 (74) Carbobenzoxyglycyl chloride . . . . . . . . . . . . . . . . . . 43 (74) Carbobenzoxyglycyl-a-aminoisobutyric acid . . . . . . 164 (80) Carbobensoxyglycylglycine. . . . . . . . . . . . . . . . . . . . 178. (74) Carbobenroxyglycylglycinamide. . . . . . . . . . . . 179-181 (188) Carbobenzoxyglycylglycinhydrazide. . . . . . . . . . . . 166 (258) Carbobenzoxy(diglycy1)glycine . . . . . . . . . . . . . . . . 196(78) Carbobenroxy(diglycy1)glycinamide. . . . . . . . . . . . . 220 ( M a ) Carbobenroxy(triglycy1)glycine . . . . . . . . . . . . . . . . 230 (78) Carbobenroxyglycylsarcosine . . . . . . . . . . . . . . . . . 102 (59) Carbobenzoxyglycyl-Lalanine. . . . . . . . . . . . . . . 135 (24, 49) Carbobensoxyglycyl-Galaninhydraride. . . . . . . . . 133 (77) Chloroacetylglycine. . . . . . . . . . . . . . . . . . . . . . . . . . . 98-100 (235) Chloroacetylglycylglycine. . . . . . . . . . . . . . . . . . . . 178-180 (136, 160) Cbloroacetylsarcosine. . . . . . . . . . . . . . . . . . . . . . . . 95-98 (235) Phenylpyruvylglycine . . . . . . . . . . . . . . . . . . . . . . . . . 167-168 (189) Pyruvylglycine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 (56) Toluenesulfonylglycinhydraride. . . . . . . . . . . . . . 155.5 (266) Toluenesulfonylglycylglycine. . . . . . . . . . . . . . . . . . 178.5 (266)
.
cAlanine anhydride. . . . . . . . . . . . . . . . . . . . . . . . . . 297 (134) D-Alanine-N-carbonic acid anhydride . . . . . . . . . . . 89 (220) Bencoyl-calaninamide . . . . . . . . . . . . . . . . . . . . . . 235-240 (201s) Bensoyl-Dcalaninamide . . . . . . . . . . . . . . . . . . . . 229-230 (244) Carbobenzoxy-calanine . . . . . . . . . . . . . . . . 84 (74) Carbobenmxy-obalanine . . . . . . . . . . . . . . . . . . 114-1 15 (74) 106 (272) Carbobenzoxy-Falanine. . . . . . . . . . . . . Carbobenaoxy-Saleninhydrazide. . . . . . . . . . . . . . . 143 (272) Carbobencoxy-bslanylglycine. . . . . . . . . . . . . . . . . 132-133 (278) Carbobeneoxy-balanylglycinhydrazide. . . . . . . 157 (49, 276) Carbobenaoxy-calanyl-Lalanine. . . . . . . . . . . . 152-153 (278) Chloroacetyl-Lalanine . . . . . . . . . . . . . . . . . . . 93.5-94.5 (6,167) Toluenmlfonyl-balanine . . . . . . . . . . . . . . . . . . 134-135 (158) Toluenmlfonyl-Dtalaninhydraeide . . . . . . . . . . . 171 (266) Toluenesulfonyl-DLalanylglycine. . . . . . . . . . . . . . . 147 (266)
37
SYNTHESIS OF PEPTIDES
TABLE I11 Dctivalivcs of Valine. W n e . and Isoleucine Compound Carbobenzoxy-cvaline . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chloroacetyl-cvaline.. . . . . . . . . . . . . . . . . . . . . . . . .
M. p., "C. (ref.) 64-65 (282a) 113-115 (164)
125-126 (43) chucinamide acetate. . . . . . . . . . . . . . . . . . . . . . . . . . . . chucinamide HCl. . . . . . . . ... . 230-237 (276a) chucine-N-carbonic acid anhydride. . . . . . . . . . . . . . . 76. Ei-78 (301) Dchuche-N-carbonic acid anhydride. . . . . . . . . . . . . . 48-50 (229) Acetyldehydroleucine. . . . . . . Acetyldehydroleucylglycine. . . . . . . . . . . . 130-132 (115) Bencoyl-cleucinamide. ..... Benzoyl-Dcleucinamide. . . . . . ............. 171 (244) Bencoyl-cleucinhydrazide. . Bencoyl-cleucyl-cleucine . . Carboethoxy-cleucyl-cleucine.. . . . . . . . . . . . . . . . . . . . 147-148 (136) Carbobenzoxy-cleucinamide.. . . . . . . . . . . . . . . . . . . . . . 122-123 (43) Carbobenzoxy-cleucinhydrazide.. . . . . . . . . . . . . . . . . . 121 (80) Carbobenzoxy-cleucylglycine. . . . . . . . . . . . . . . . . . . . . . 115 (78) Carbobenzoxy-cleucylglycylglycine. . . . . . . . . . . . . . . . 144 (78) Carbobenzoxyglycyl-calanyl-cleucinhydrazide ...... 186 (77) Carbobenzoxyglycyl-cleuci ne . . . . . . . . . . . . . . . . . . . . . . 141- 142 (277) Carbobenzoxyglycyl-cleacinamide. . . . . . . . . . . . . . . . . 123-124 (276s) .................. 136 (168) Chloroacetyl-cleucine. . . . . Toluenesulfonyl-Dbleucinhydrazide ........ 146 (266) Tolueneaultonyl-Dcleucylglycine.. . ........ 121.5 (266) Toluenesulfonyl-calanyl-cleucine. ................. 186 (266) Carbobenzoxyglycyl-cisoleucine. . . . . . . . . . . . . . . . . . . Chloroacetyl-cisoleucine . . . . . . . . . . . . . . . . . . . . . . . . . .
114-1 15 (277) 74-75 (16)
be obtained in crystalline form by the procedure described by Bergmann and Zervas (74), although some workers have encountered difficulty in crystallizing this substance without inoculation with seed crystals. The chloride is best used immediately after its preparation, a small sample being set aside under petroleum ether for the inoculation of the next preparation. With regard to the attachment of an alanyl residue to an amino acid or peptide, the bromopropionyl bromide (or chloride) method of Fischer is feasible only when Dcalanyl peptides are desired. The preparation of optically active a-bromopropionic acid, as was noted previously, is a tedious operation, and consequently, for the attachment of t or D-alanine residues, the carbobenzoxy method appears preferable. Unfortunately, carbobenzoxy-calanyl chloride is accessible only as an oil, and although
38
JOSEPH 9. FRUTON
TABLE IV Peptides of Cflyn'ne,Alanine, Valine, Leucine, and Ieoleucine P
Peptide Glycy lglycine ( Diglycy1)glycine (Triglycy1)glycine (Tetraglycy1)glycine (Pentaglycy1)glycine (Diglycyl)-D-leucine (Diglycy1)-Lleucine (Triglycy1)-~leucylglycine (Diglycy 1)-walanylglycine (Diglycy1)-Lleucylglycine Gly cyl-calanine Glycyl dehydroalanine Glycy I-~alanylglycine Glycyl (di-D-alanyl)glycine Glycy 1-talanyl-tleuciiie
...
... ...
... ...
II
4-27.5 -28.0 -28.4 +53.7 -43.2 -50
...
-64.3 4-104.8 -88.0
Temp., "C.
Concn. and solvent
...
... ... ...
...
...
... 26 26 24 20 24 20 *.. 20 20 22
...
D- Alanylglycine
LAlanylglycine D- Alany lglycy lglycine LAlanylglycylglycine LAlanyl (diglycy1)glycine LAlanyl(diglycy1)-balanylglycylglycine LAlanylglycy 1-Lalanine cAlanylglycy 1- leuc cine D- Alanyl-Lalanine LAlany I-D-alanine LAlanyl-halanine (Di-calanyl)-Lalsnine
-50.7 $50.2 -31.6 $32.4 $27.0 $13.2 -19.5 -11.2 -68.5 $68.9 -21.6 -72.2
20 20 25 20 21.5 26 24 20 20 20
20 20 20 20 23 20 25
190s 190a
2%, HzO 2%, Ha0 2.5%, HsO 3%, HzO 2.5%, HzO 4%, HzO
...
4.3%, HtO 2.9%, HIO 1.1%, HzO eq. HCI
Glycyl-Lalan yl-cleucyl-tiso-80.6 leucine Glycyl-Lvaline -19.7 Glycy 1-D-leucine 4-35.7 Glycy 1- leuc cine -35.1 Sarcosyl-Lleucine -30.4 Cly cy1-wleucylglycine $42.6 Glycyl-Lleucylglycine -41.2 Glycyl-~leucylglycyl-~leucinc -51 .O Glycyl-Lleucyl-talanine -59.0 Glycyl-~leucy1- leuc cine -67.0 Glycyl(di-Lleucyl)-Lleucine -78.6 Glycyl(tri-Lleucy1)-Lleucine -118.1 GIycyl-o-isoleucine $13.6 Glycyl-~isoleucine -14.7
149 128 129 129 134
...
...
Ref.
+1
78 233 78 24,30,167 66,255b 140 234 9,264
0.9%, N HC1
15
lo%,
164,231 275 168 275b 190a 78 34
Hi0
3 % ~HzO 4.2%, HsO 2.2%, HzO 2%, H i 0 2.5%, HtO 2.3%, 10% HCl 2.5%, Hz0 2.9%, EtOH 2%, N NaOH 2.5%, N NaOH 4.1%, HzO 4.1%, HzO
9 8 8 8
28,36 16,36
4.3%, HzO lo%, HtO lo%, HzO lo%, Hz0 3%, H i 0
80 133,140 32,49 49,136,140 17
8 % , Hz0
136 24 9 161 161 134,237,278 . . 12
3.7%, H20 2.3%, HzO 8.7%, HIO 7.4%, HzO 5%) HzO 3.5%, 2 N HCI
39
SYNTHESIS OF PEPTIDES
Temp.,
TABLE IV.-(Continued) Peptide (Tri-calany1)-Lalanine (Tetra-Lalany1)-talanine (Di-D-alanyl)glycine G Alanyl-cvaline L Alanyl-Lleucine G Alanyl-~leuc ylglycine ~ A l a n y l - ~ - l e yu1-~valine c D- Alany 1-D-leuc y 1-D-leucine GAlanyl-Lleuc yl-~isoleucine &Alanyl- leuc cine LAlanyl-~isoleucine G Alanyl-a-aminoisobutyric acid LAlanylsarcosy lglycine
[alD
-120.5 -136.4 +47.2 -5.9 -17.3 -30.4 -60.2 +62 .O -24.9 -31 .O +6.1
"C. 20 20 20 20 20 20 19 25 20 26 20
Concn. and solvent 2.2%, 2 N HCl 1.9%, 2 N HCl 3%, HzO
lo%, Ha0 5%, HzO 2.1%, HzO 4.4%, HzO 3.2%, N NaOH 4.975,N HCl 1.5%, H i 0 3.8%, N HC1
Ref. 12 12 234 164 138,266 9 30 237 15 201a
16
+34.5 +10.8
20 25
2%, H i 0 5%, H i 0
80 49
LValylglycine D-Valyl-Lvaline tValyl-Lvaline
+93.6 -74.0 -54
20 20 20
lo%, Ha0 lo%, H i 0
164 164 33
GLeucylgly cine GLeucylglycylglycine bLeucyl(diglycy1)glycine cLeucyl(pentaglycy1)glycine
+85.8 +57.7 +45.9 +5.2
24 25 20 20
GLeucyl (hexaglycy1)glycine
+6.3
20
LLeucyl (triglycy1)-Lleucine L-Leucylglycy1-Galanine L-Leucylglycyl-r.-lcucine L-Leucylgly cy 1-D-isoleucine ~ L e u c y l gycyl-~isoleucine l ~ L e uyl-D-alanine c L-Leucyl-Lalanine ~ L e uycl - ~ a l a ylglycine n LLeuc y 1-p-alanine xAeucyl-Lvaline D-Leucy 1-D-leucine D-Leucyl-Lleucine cLeucy 1-D-leucine GLeucyl-Gleucine (Di-D-leucyl )+-leucine (Di-L-leucy1)- leuc cine (Tri-cleucyl)-Lleucine ~ L e u c 1-D-isoleucine y LLeucyl-Lisoleucine
+21.3 +20.3 +6.0 +41.2 +26.5 +76.0 +23.5 -17.3 +28 .O +18.0 +13.9 -68.0 +68.9 -13.4 +46.0 -51.4 -90.0 +53.1 +25.7
20 20 20 20 20 22 20 20 26 20 25 20 20 20 25 20 20 20 20
LIsoleucylglycine
+33.6
20 -
4.3%, HzO 2,4%, HnO 5%, Ha0 9.6%, Hn0 3.7%, 0.1 N NaOH 6%, HzO 1 eq. NaOH 2.4%, HzO lo%, Hz0 3.6%, 10% HCl 2 % , 0.2 N HCl 3%, Hz0 2.5%, HIO 5%, MeOH 1.2%, HnO 5 % , HIO LO%, Hz0 1.5%, N NaOH 3.4%, N HC1 3.7%, N HCl 3.1%, N NaOH 3.5%, N NaOH 3.1%, N NaOH 7.6%, N NaOH 1.8%, HzO 3 % , N HCl
+
78,136 10,267 136 10 138 168 6,168 34
37 28,37 80 80,136 9 201s
30,164 154,237 154 154 136 237 B B
28 15,16
16
40
JOSEPH 8. FRUTON
it may be used in this form for peptide synthesis (53,220), the tendency of such noncrystalline carbobenzoxyamino acid chlorides to form N-carbonic acid anhydrides is appreciable. For this reason, the use of carbobenzoxy-Galanyl azide will, in many cases, be found to give somewhat better yields in coupling reactions. The considerations noted above in the cam of alanine apply with even greater force in the synthesis of leucyl peptides. The synthesis of L- and wleucyl derivatives, for which Fischer used optically active a-bromoisocapronyl chlorides, is now accomplished more readily by coupling reactions which involve the use of carbobenzoxy-c or D-leucyl azide. Carbobenzoxyleucyl chloride is an oil, and coupling reactions in which it is employed frequently do not proceed satisfactorily. The very large number of known synthetic peptides containing glycyl, alanyl, or leucyl residues attests to the relative ease with which these peptides may be prepared by means of currently available methods. Several of these peptides have proved invaluable as standard substrates for studies of proteolytic enzymes. Thus, glycylglycylglycine, calanylglycylglycine, and Gleucylglycylglycine have been used widely in the study of bacterial enzymes (243) and of the peptidases of intestinal mucosa (276), animal tissues and fluids (184). In addition, cleucinamide acetate has been recommended as a standard substrate for the manganese-activatable aminopeptidase of animal tissues (184). In this connection, it may be stressed once more that, for enzyme studies, the use of racemic peptides is to be avoided since there are several reports in the literature to the effect that the presence of the D or L antipode may markedly inhibit the rate of hydrolysis of an optically active peptide by proteolytic enzymes (117,277). In the synthesis of valyl and isoleucyl peptides, only the halogen acyl halide method has been used extensively, and insufficient information is available a t present regarding the application of the carbobenzoxy method to the synthesis of such peptides. There is good reason to believe, however, that, whenever peptides involving the optically active forms of valine or isoleucine are desired, this method will be found to be more suitable than the classical halogen acyl halide synthesis. Recently, Synge (282a) has reported the synthesis of a-(cvaly1)-Gornithine by the use of the carbobenzoxy method. It may be added that a large variety of peptides of nonprotein aliphatic amino acids have also been described. Thus, a-aminoisobutyryl peptides have been prepared both by the halogen acyl halide method (26) and the carbobenzoxy method (80), and a-amino-n-butyryl (14,27) as well as alloisoleucyl (36) peptides have been made by the halogen acyl halide procedure.
41
SYNTHESIS O F PEPTIDES
2. Peptides of Aapartic and Glutumic Acids
Although the coupling of the diesters (or the disodium salts) of these amino acids with N-substituted amino acid chlorides or asides, or with azlactones, presents no special problems, the synthesis of peptides involving the carboxyl groups of aspartic or glutamic acid frequently offers some difficulty. Fischer and Koenigs (155) described the synthesis of a Dcaspartylmonoglycine by the condensation of fumarylchloride with glycine ethyl ester. The resulting fumaryldiglycine ethyl ester was saponified and, upon aminatiQn of the fumaryldiglycine, the above dipeptide was obtained. It was not established, however, whether the product contained an a- or a 8-amide linkage.
Since L-aspartic acid may readily be prepared from Gasparagine by the method of Vickery and Pucher (2851, methods involving the use of the carbobenzoxyamino acid are preferable to the older procedure of Fischer. If disubstitution products are desired, carbobenzoxy-caspartic acid dichloride may be coupled with an amino acid ester in the usual manner (193). If monosubstitution products are sought, however, a different procedure must be employed. This special method depends on the fact that N-acylaspartic acid upon treatment with acetic anhydride gives an acid anhydride (cf. 67,203a). CHiCOOH
-
CHIC0
(CH8CO):O
RCO-NH
c:
HCOOH
In the case of the carbobenzoxy derivative, this reaction is not accompanied by racemization (74), unless sodium acetate also is present (186). The resulting carbobenzoxy-L-aspartic acid anhydride may be used for coupling reactions with alcohols to give monoesters and with amino acid esters to give carbobenzoxy-caspartyl peptide esters in which only one carboxyl group of aspartic acid is bound in peptide linkage.
42
JOSEPH 8. FFtUTON
CH&O
I
' 0
CsHrCHaOCO-NH
R
+ NHibHCOOCHa
-t
IHiCooHb
CeHrCHiOCO-NH H C O - N H HCOOCHa or
CHSCO-NHCHCOOCHa CaH'CHiOCO-NH
bHCOOH k
As indicated above, in the reaction with amino acid esters, carbobenzoxy-baspartic acid anhydride may give rise either to a- or to 8-peptide linkages. For example, in the coupling with glycine ethyl ester, the major product isolated is the a compound; thus, caspartyl-aglycine may readily be prepared (193). Similarly, in the reaction with whistidine methyl ester, there is formed the a-peptide derivative (197). On the other hand, the coupling with L-tyrosine ethyl ester leads to the isolation of the p-peptide derivative (88). A general method is available, however, for the synthesis of p-peptides of aspartic acid. This may be effected by taking advantage of the reaction of carbobenzoxy-baspartic acid anhydride with benzyl alcohol to give the a-benzyl ester (87). The 8-carboxyl may then be converted to an acid chloride with phosphorus pentachloride. L-Aspartyl-8-glycine (203), caspartyl-@-ctyrosine(88), and L-aspartyl-j3-L-histidine (288) may be prepared by coupling this acid chloride with the appropriate amino acid eater:
CsHrCHrOCO-NH
x
R
HCOOCH&sHr
+ NHabHCOOCH&aHr -+ R CHiCO-NH
CsHrCHsOCO-NH
bHCOOCH&sH'
bHCOOCHiCsHr
It may be added that carbobenzoxy-baspartic acid anhydride reacts with ammonia to give the a-amide, carbobenzoxyiso-L-asparagine (74). The synthesis of basparagine by the carbobenzoxy method has been effected through the reaction of carbobenzoxy-L-aspartyl-a-benzyl ester /%chloride with ammonia, followed by hydrogenation of the product (87). In the 'case of the synthesis of glutamyl peptides, the same general procedures may be used as for the aspartyl peptides. Carbobenzoxy-c glutamic acid readily gives the optically active anhydride on treatment with acetic anhydride, and this product reacts with glycine
43
8YNTHE818 OF PEPTIDES
TABLE V Derivatives of Aspartic and Glutamic Acids Compound
M. p., "C. (ref.)
a-chpartylglycine ethyl ester . . . . . . . . . . . . . . . . . . . 232 (78) Acetyl-caapartic acid anhydride . . . . . . . . . . . . . . . . . 141 (67) Carbobenzoxy-caspartic acid .................... 116 (74) Carbobeneoxy-casparagine...................... 165 (74) Carbobenroxy.cisoasparagine .................... 164 (74) Carbobeneoxy-caspartic acid anhydride ........... 84 (74) Carbobencoxy-a-csepartic acid benzyl ester ....... 84-85 (87) Carbobeneoxy-caspartic acid dichloride ........... 46 (193) Carbobeneoxy-a-baspartylglycine. . . . . . . . . . . . . . . . 171 (78, 193) Carbobenzoxy.8.caspartylglycine . . . . . . . . . . . . . . . . 154 (193,203) Carbobeneoxy.8.caspartyl.ctyrosine . . . . . . . . . . . . . 110 (74) Carbobeneoxy-pcaspartyl-8-benaylcysteinylglycine168-170 (248) Carbobeneoxy.a.casparty1.chistidine . . . . . . . . . . . . . 171 (197) Chloroacetyl-caspartic acid ..................... 142-143 (148) Chloroacetyl.basparagine....................... 148-149 (155) 130 (76) YcGlutamic acid ethyl ester .................... 151 (78) a-cGlutamylglycine ethyl ester .................. a-cGlutamyl.cleucinamide ...................... 175-177 (276a) a-cGlutamyl-ctyrosine ethyl ester ............... 144 (187) Carbobenaoxy-eglutamic acid ................... 120 (74) Carbobeneoxy.cglutamine ....................... 137 (74) Carbobenroxy-Lglutamic acid anhydride .......... 94 (74) Carbobeneoxy.eisog1utamine .................... 175 (74) Carbobenzoxy--y.cglutamic acid hydrazide ........ 178-179 (206a, 227a) Carbobene0xy.a-cglutamylglycine ................ 143 (193) Carbobenzoxy-7-cglutamylglycine . . . . . . . . . . . . . . . 159-161 (227a) Carbobeneoxy.a-cglutamylglycylglycine . . . . . . . . . . 142 (79) Carbobensoxy.a-cglutamyl.cleucinamide . . . . . . . . . 165-169 (276a) Carbobenroxy-a-cglutamyl-eglutamicacid ........ 176 (74) Carbobene0xy.a-cglutamyl.D.phenylalanine . . . . . . . 122 (187) Carbobencoxy.a-cglutamyl.cphenylalanine . . . . . . . 162 (187) Carbobeneoxy-a-cglutamyl-~rphenylalaninamide . . 185-187 (215) Carbobenzoxy.a-cglutamyl.btyrosine . . . . . . . . . . . . 185 (88) Csrbobeneoxy.a-cglutamy1.ctyrosinamide ........ 181 (187) Carbobeneoxy-a-bglutamyl-btyrosinhydraaide.... 194 (187) Carbobeneoxy-ycglutamyl-tcysteinylglycine ..... 163 (202) Carbobeneoxy-a-cglutaminyl-cphenylalanine..... 180 (187) Carbobenroxyglycyl-Lglutamicacid . . . . . . . . . . . . . . 160-162 (215) Carbobenzoxyglycyl.Licisoglutamine ............... 185 (78) Carbobensoxyglycyl.scglutamylg1ycine . . . . . . . . . . 98-100 (78) Carbobenzoxyglycyl.a.Lglutamylglycinamide ...... 175 (50) Chloroacetyl-cglutamic acid ..................... 143 (157) Chloroacetyl-cglutamine ........................ 130- 132 (283) Glycyl-cglutamic acid diketopiperazine ........... 240 (78) Phenylpyruvyl-cglutamic acid ................... 142-143 (189)
44
JOSEPH 8. FRTJTON
TABLE VI Pevtides t
-
Aspartic and Gluiumic Acids remr "C.
Peptide Glycyl-baspartic acid Glycyl-caaparagine Glycyl-baaparagin y 1-cleucine bleucylglycyl-baapartic acid cLeucyl-baspartic acid D-Leucyl-baaparagine cLeucy1-baaparagine a-chpartylglycine &cAapartylglycine a,B-cAapart yldiglycine &tAspartyl-btyroaine 8-c Aspartyl-Lcysteinylgly cine a-bAspartyl-D-histidine &cAspartyl-chistidine
+11.1 -6.4 -46.8 +55.3 +27.1 -53.8 +17.8 +36.7 +7.2 +33.8 +SO. 1 -29.0 -6 .O 4-38
Glycyl-r,-alanyl-cleucyl-~ ... glutamic acid Glycyl-calanyl-bglutamic acid -69.8 Glycyl-cglutamic acid -6.3 Glycyl-bglu tamine -2.4 Glycyl-a-bglutam ylglycine ... Glycyl-a-bglutamyl-ctyrosine ... Glycyl-a-cglu taminylglycine -28.4 D- Alanyl-eglutamine -20.1 cAlanyl-cglutamine +9.3 cleucyl-bglutamic acid +10.6 cleucyl-Lglutamine +12.6 a-cGlutamylglycine +80.3 y-cGlutam ylglycine +11.1 a-cGlutamyl-cglutamic acid +19.9 a-cGlutam yl-cphen ylalanine +27.0 a-LGlutamyl-ctyrosine +30.1 a-cGlutam yl-ccysteine 4-13 .65 a-cGlu tam yl-ccysteinylglycine +2.5 y-D-Glutamy 1-ccyeteinylglycine -34.6 y-LGlutamyl-ccysteinylglycine -21.3 Bia (y-bglutamyl)-bcystine - 120" bGlu taminylglycine 76 cGlutaminy1-cglutamic acid +I5 bGlu taminyl-tcysteine -9.8" Bia(bglutaminy1)-ccystine -119"
+
" A = 546 mp.
Concn. and solvent
20 20 20 20 20
D%, Hi0 7%, H i 0 4.8%, N HCl 5%, HIO B%, HtO
20
5%, Hi0 5.4%, Hz0 1.974, H 2 0 2.4%, H,O 2.2%, H,O 2.3%, HzO
20 23 22 23 18 25 30 27
... 20 20 19
...
... 19 16 18 20 18 22 14 18 25 19
... 25 16 27
... 18 18
... ...
I%, H i 0 I%, H s 0 I%, H i 0
+ 1eq. HC + 1eq. HC + 1eq. HC + 1eq. HC
Ref. 148 166 166 148 148 156 156 193 193 193 74 248 197 288 77 204 157 283 78 187 283 283 283 139,276a 283 193 227a 74 274
88 202 289 224 202,291 202 245 245 203 203
45
SYNTHESIS OF PEPTIDEB
ethyl ester to give the a rather than the y compound as the principal product. CHrCO &HI
I
C~H'CHXOCO-NHCHCO
'+
0 '
NHxCHXCOOCIH'+ CHrCOOH ~ H x CeHrCHaOCO-NH
HCO-NHCHaCOOCxH'
It is of interest that, in the reaction with Gtyrosine ethyl ester (88) Gglutamic acid diethyl ester (74), or Gcysteine benzyl ester (203), the major products are also derivatives of the a-peptides. It would appear, therefore, that the tendency to yield the a substitution product is more general with carbobenzoxy-cglutamic acid anhydride than it is with the aspartic acid analog (cf. also 245). The preparation of y-peptides of glutamic acid may be accomplished by a method similar to that noted above for the p-aspartyl peptides. Carbobenzoxy-cglutamic acid anhydride gives the a-benzyl ester with benzyl alcohol, and subsequent treatment of the y-carboxyl with phosphorus pentachloride yields the yacid chloride (87). This method has been used recently for the synthesis of derivatives of y-glutamylglutamic acid (91a,250a). With the aid of this chloride, there has been effected a synthesis of Gglutamine (87). Furthermore, Melville (245) has used the y-acid chlorides of carbobenzoxyglutamyl peptide esters to make several cglutaminyl peptides. Carbobenzoxyq-cglutamic acid hydraside has also been employed for the synthesis of y-glutamyl peptides (206aJ2!27a). The phthalyl method, developed recently, has also been applied to this purpose (22313). Recently, a synthesis of cglutamine has been described (183) in which carbobenzoxy-cglutamic acid diamide (prepared from the diethyl ester with ammonia) is selectively hydrolyzed at the a-amide linkage with cysteine-activated papain to yield carbobensoxy-cglutamine, from which L-glutamine may be obtained by catalytic hydrogenation. To date, the possibility of using this method for the synthesis of y-peptides of glutamic acid has not been explored experimentally. It may be added that, by the use of the carbobenzoxy method, Schneider (265) has prepared .several peptides of the dicarboxylic amino acid aminomalonic acid. Interest in the physiological role of peptides of glutamic acid has been heightened by the isolation of a polyglutamic acid peptide from the capsule of Bacillus anthracis (201,222), and of a glutamic acid peptide from sea weed of the genus Pelvetia (200), as well as the demonstration that glutsmic acid is bound in peptide linkage in folic acid (280).
46
JOSEPH 8. FRUTON
3. Peptides of Phenylalanine and Tyrosine
A large number of peptides containing one or both of these amino acids has been prepared, and by a variety of methods. Treatment of the esters of Lphenylalanine or Ltyrosine with halogen acyl halides has led to the synthesis, by Fischer and later workers, of numerous peptides such as glycyl-Lphenylalanine, glycyl->tyrosine, calanyl-c tyrosine, etc. Except for the glycyl dipeptides, peptides in which the amino group of phenylalanine or tyrosine is linked to another amino acid can be made more readily by means of the carbobenzoxy method. In general, the coupling of carbobenzoxyamino acid chlorides, azides, or anhydrides with the esters of phenylalanine or tyrosine proceeds without unusual difficulty. An indication of the advantages of the carbobenzoxy method for the synthesis of optically active peptides containing a tyrosine residue is provided by a comparison of the optical rotation of glycyl-L alanylglycyl-Ltyrosine as prepared by Abderhalden et al. (2) , who used this method, and as prepared by Fischer (140) by means of the halogen acyl halide method. The material obtained by the former workers had an [ a ]of~ 18.4', while Fischer's preparation had an [aIDof +4.0". For the synthesis of peptides in which the carboxyl group of the aromatic amino acid is involved in peptide linkage, at least three methods are available. Historically, the first of these was the use by Fischer (130) of a-bromo-8-phenylpropionyl chloride as the coupling agent; subsequent amination of the coupling product led to the synthesis of DGphenylalanyl peptides:
+
CHtCdh BrLHCOC1
+ NHsCHICO-NHCHSCOOH CHsCdIr
Br~HCO-NHCHtCO-NHCHtCOOH CHaCdh
NHa
NHS~HC~NHCHSCO-NHC~Scoo~
A comparable application of the halogen acyl halide method has not been described for the synthesis of Dbtyrosyl peptides. As a consequence of the search for more satisfactory procedures of peptide synthesis, a second method, involving the use of azlactones (cf. page 75), was introduced by Bergmann el al. (67). This method made possible the synthesis of peptides such as cphenylalanyl-Lglutamic acid (67) and mtyrosyl-carginine (90), among others. In the face of the potentialities of the carbobenzoxy method for the synthesis of phenylalanyl and tyrosyl peptides, however, both the
SYNTHESIS OF PEPTIDES
47
halogen acyl halide and azlactone methods must now be considered to be of subsidiary importance. The carbobenzoxy method is especially suitable for the synthesis of such peptides, not only because of its general advantages, but also because the carbobenzoxy derivatives of phenylalanine and tyrosine give excellent crystalline acid chlorides. In the case of the chlorination of carbobenzoxy-ctyrosine, the phenolic hydroxyl group must be protected; either an acetyl (88) or a carbobenzoxy (3) group may be used for this purpose. There are numerous examples of the use of carbobenzoxy-cphenylalanyl chloride (85) or of 0-acetyl-N-carbobenzoxy-L-tyrosylchloride (88) for peptide synthesis. Of particular interest is the synthesis by Barkdoll and Ross (41) of L-tyrosyl-ctyrosyl-L-tyrosine by successively building up the peptide chain through coupling reactions involving 0-aoetyl-N-carbobenzoxy-ctyrosyl chloride:
The fact that tyrosyl peptides may be made equally well by the use of carbobenzoxy-L-tyrosine azide is illustrated by the synthesis of N-carbobenzoxy-L-tyrosylglycylglycine ethyl ester (50) and of N-carbobenzoxy-Ltyrosyl-S-benzylcysteineethyl ester (204). Peptides of the aromatic amino acids have become of considerable interest in recent years because of the finding that some of them (or their derivatives) are hydrolyzed by crystalline pepsin (187,204) or by crystalline chymotrypsin (50). Thus, pepsin hydrolyzes carbobenzoxycglutamyl-ctyrosine (187) and, as shown by Harington and Pitt Rivers (204), even free peptides, notably ccysteinyl-ttyrosine. For the preparation of synthetic substrates for chymotrypsin, it has been frequently necessary to convert the terminal carboxyl group of a phenylalanine or tyrosine peptide to the corresponding amide. This usually proceeds smoothly by the treatment of the appropriate carbobenzoxy peptide ester with ammonia in methanol. In this manner,
48
J0SE)PH 8. FRUTON
carbobensoxy-ctyrosylglycine ethyl ester was tramformed into the amide (50). When carbobensoxy-Irphenylalanylglycine ethyl ester was treated in the Bame way, however, the reaction took an unexpected course, and the resulting product was not the carbobenzoxy peptide amide. Instead, there occurred the elimination of bensyl alcohol and the formation of the amide of 5-bensylhydantoin-3-aceticacid (188) :
-
CHaCdI, C:H,CHaOCO-NH
bHCO-NHCHaCOOCaH, NH CH~CIHI CO-NH bHCO-NCHaCONHa + C'H&HaOH a
U
Of a large number of other carbobenzoxy dipeptide esters subjected to treatment with ammonia in methanol, only two other cases of hydantoin formation have been reported thus far. In one of these, carbobenzoxyGleucylglycine ethyl ester yielded the amide of 5-isobutylhydaritoin-3acetic acid (112b):
NHi
CeH'CHaOCO-NH CH,
CH:
\Ck
CO-NHbHCO-NCHaCONHa
I
+ CeH'CHaOH
The other case involves an exactly analogous reaction involving carbobensoxy-cmethionylglycine ethyl ester (112b). It would appear that these reactions are closely related to that postulated in the conversion of carboethoxyglycylglycine ethyl ester to carbonyl-bis-glycine by means of alkali (cf. p. S), which has been explained by the assumption that there occurs the formation of hydantoin-3-acetic acid aa a transient intermediate (296). This is in accord with the observation that such hydantoin derivatives are readily hydrolyzed by alkali (1924. To avoid the difficulty presented by the occurrence of hydantoin formation, it is possible to couple carbobenzoxy-cphenylalanyl chloride directly with glycinamide to give the desired dipeptide amide (188).
49
SYNTHESIS OF PEPTIDES
TABLE VII Dcrivdives of Phyldcmins
M. p., "C. (ref.)
Compound
GPhenylalaninamide acetate. . . . . . . . . DcPhenylalanine-N-car~nicacid anh cPhenylalany1-bphenylalaninamide.. . . LPhenylslanyl-ctyroain~ide. . . . . . . Acetyl-wphenylalanyl-bleucine.. . . . . . . . . . . . . . . . . 183-184 (44) Acetyl-cphenylalanyl-bleucine. . . . . . . . . . . . . . . . . . 191-193 (44) Acetyl-bphenylalanyl-cglutamicacid. . . . . . . . . . . . . 140 (67) Acetyl-D-phenylalanyl-btyroaine. Acetyldehydrophenylalanine. . . . . a-Acetaminocinnamic acid arlactone. Acetyldehydrophenylalanyl-calanine.
. . . . . . . 196-196 (116)
Acetyldehydrophenylalanyl-btyroaine.. .....
217-218 (67,116)
Acetyldehydrophenylalanyl-carginine. . . . . . . . . . . . . 192193 (60) Acetyldehydrophenylalanyl-bproline... . . . . . . . . . . . 140-142 (44) Carbobenroxy-cphenylalanine.. . . . . . . . . . . . . . . . . . 126-128 (86) Carbobencoxy-Dcphenylalanine . . . . . . . . . . . . . 103 (74) Carbobencoxy-bphenylalanylglycine. . . . . . . . . . . . . . 161-162 (44,188) Carbobensoxy-bphenylanylglycinamide. Carbobenmxy-cphenylanyl-cgluta Carbobenzoxyglycyl-bphenylalanine.. Carbobenroxyglycyl-bpphenylalanina Carbobenzoxyglycyl-bphenylalanylg Carbobenroxy-balanyl-x.-phenylalanine. . . . . . . . . . . 68-68 (63) Carbobeneoxy-x.-leucyl-D-phenylaninhydrazide. . . 170 (2044 Chloroacetyl-cphenylalanine.................... 126 (165) F'yruvyl-Dbphenyhhnbe. . . . . . . . . . . . . . . . . . . . . . . 94 (66) Tolueneeulfonyl-L-pltenylalanine.. . . . . . . . . . . . . . . . . 184-166 (168)
In discutming the synthesis of peptides of the aromatic amino acids, it is appropriate to add that peptides of 3,bdiiodo-ctyrosine may be prepared either by the iodination of a tyrosine peptide with iodine-potassium iodide in alkaline solution or by coupling reactions involving diiodotyrosine (13,187). It may be added that the carbobenzoxy method has been applied to the synthesis of peptides of @-phenyl-8-alanine(116). 4. Peptides of Cystine and Cysleine
The classical methods of Fischer permitted the synthesis of cystine peptideg in which the amino groups of the cystine moiety are linked to
JOSEPH 8. FRUTON
TABLE VIII Den'vdaves of Turoa'nc
M. p., "C. (ref.)
Compound
N-Carbobenzoxy-ctyrosinhydrazide... N-Carbobensoxy-ctyrosylglycine.. . . . . . . . . . . . . . . . . 100 (50) N-Carbobenzoxy-btyrosylglycinamide.. . . . . . . . . . . . 116 (50) N-Carbobeneoxy-ctyrosylglycylglycine.. . . . . . . . . . . 213-215 (2) N-Carbobenzoxy-btyrosylglycylglycinamide... . . . . . 218 (50) N-Carbobencoxy-L-tyrosyl-Lcysteine . . . . . . . . . . . . . . 120 d. (204) N-Carbobenzoxy-btyrosyl-S-benzyl-Lcysteine. . . . . . 166 (204) Bis(N-carbobenzoxy-ctyrosy1)-Lcystine.... N-Carbobenzoxy-btyrosyl-ctyrosine.. . . . . . N-Carbobenzoxy-ctyrosyl-ctyrosinamide.. . . . . . . . . 187-189 (188) N-Carbobensoxy (tri-btyrosy1)-ctyrosine
... .....
224-225 (41) 107 (50)
Carbobenzoxy-calanyl+tyrosine. . . . . . . . . . . . . . . . . . 149-150 (53) Carbobenzoxy-calanylglycyl-btyrosine.. . . . . . . . . . . 128 (2) 0,N-Dicarbomethoxy-btyrosine. . . . . . . . . . . . . . . . . . . 97 (206) N-Chloroacetyl-btyrosine........................ 86-87 (129) Glycyl-Ltyrosine diketopiperazine . . . . . . . . . . . . . . . . . 295 (1 66) N-Toluenesulfonyl-btyrosine...
glycyl (171), alanyl, or leucyl (150) residues.* In order to prepare the corresponding cysteine peptides, Pirie (255) reduced the disulfide group with zinc and acid and isolated the sulfhydryl peptide as a cuprous mercaptide. The advent of the carbobenzoxy method wa8 decisive in the develop
* In naming peptidea in which both amino groups of cystine are substituted by similar amino acid residues, it appeara desirable to use the prefix "bis" as in bisglycylL-cystine. Similarly, where both carboxyl groups of cystine are linked to similar
61
SYNTHESIS OF PEPTIDES
TABLE IX Peptides of Phmylalanine and Tyosine reml "C.
Peptide Glycyl-D-phenylalanhe Glycyl-bphenylalanine Glycy1-D-phenylalanyl-cglutamic acid Glycyl-bphenylalanyl-bglutamii acid Glycyldehydrophenylalanine Glycyldehydrophenylalanylglycine DbAlanyldehydrophenylalanine cLeucyl-~-phenylalanine.2H~O LPhenylalanylglycine D-Phenylalanyl-D-leucine LPhenylalanyl-cglutamic acid Glycylglycyl-ctyrosine Glycyl-balanylglycyl-ctyrosine Glycyl-balenyl-Ltyrosine Glycyl-ctyrosine Glyc y l - ~yrosylglycine t Glycyl-diiodo-ctyrosine cAlanylglycy1-ctyrosine L-Alanyl-ctyrosine L- Alenyl-diiodo-ctyrosine bLeucy1-ctyroeine cTyroaylglycine >Tyros ylglycylglycine cTyroeyl-t-aspartic acid L-Tyrosyl-~tyromne cTyrosyl-Lcysteine Bis(ctyrosy1)-Lcystine D-Tyrosyl-carginine
-41.7 4-42,a -11.3
21 20 20
-4.7
20
... ...
.. . ...
...
...
274 165 65
... ...
65 65
...
...
17 20
...
2.4%, HAc 2.4%) HzO
+20.3
20
4.8%, HzO
+42.3 +18.4 -4.8 4-43.7 +24.1 +52.7 4-41.9 4-43.1 4-62.9 +10.4 4-83.5 4-42.8 4-20.4 4-30.1 +22.6 -70.8 -105.7
20 20 20 25 20 20 20 20 20 20 20 20 19 19 23 23 23
1.6%, HzO 2.6%) HYO 4.3%, HzO 2%, H20 1 eq. HCl 4.1%, Ha0 5%, N&OH 4.6%, Ha0 2%, HzO 7.8%, NHdOH 2%, HzO 2.2%, HzO 2.7%, 20% HCl 3.6%, HzO 1 eq. HC 4%, HzO 1 eq. HCl 5%, N HCl 5%, N HCl 6.4%, 0.2 N HCI
4-25 4-54.2
Ref.
Concn. and solvent
198a 2828 165 276 07
...
+
+
+
~
2 2,140 17 129,274 2 13 2,139 17 17 17 2,3,274 2 88 88 204
204 90
_ I
ment of new methods for the synthesis of cystine and cysteine peptides. The need for the synthesis of such peptides became urgent with the formulation of glutathione as 7-cglutamyl-tcysteinylglycine (218). Final proof of this structure was provided by Harington and Mead (202), who synthesized the tripeptide by means of the carbobenzoxy method, amino acid residues (e.~., glycine) the peptide should be named ccystinylbisglycine. In this manner, it may be possible to avoid the ambiguity inherent in names such a8 diglycyl-rccyetine, ccystinyldiglycine, or ccystinylglycine.
52
JOBBIPH 8. FaUTON
which they modified to eliminate the carbobenzoxy group by treatment with phosphonium iodide in acetic acid at 45-50'. This modification was necessitated by the poisoning effect which sulfur compounds exert on palladium or platinum catalysts for the hydrogenation reaction.
[f'+
3
HNH-OCOCHlC~Hr O-NHCHaCOOCaHr
I
NH-OCOCHLhH I
CHlSH
+C1COCHrCH1CBCOOCHI
~HNHI
b
ho--NHCHICOOCIHr NH4COCHsCeHr
I:Z:--COCHaCH*
c:
saponification
HCOOCHl
____)
reduction
bo-NHCH,COOC1Hr CHaSH AHNH-cOCHICH1
IH1
HCOOH
L-NHCH,COOH
Despite its considerable value in the initial synthesis of glutathione, this reduction method has given way to the procedure of Sifferd and du Vigneaud (272), who introduced the use of sodium in liquid ammonia at -60' (286) for the elimination of the carbobenzoxy group. The latter method gives better yields, and its use by du Vigneaud and collaborators, as well as by Harington, has led to the synthesis of a variety of interesting cystine and cysteine peptides. In the synthesis of cysteinyl peptides, it is frequently necessary to protect the sulfhydryl group in the course of the synthetic operations. This may be done by treatment of the cysteine moiety with benzyl chloride to give the S-benzyl thioether. S-benzylcysteine itself may be prepared by the reduction of cystine with sodium in liquid ammonia, followed by the discharge of the color due to the sodium ion with ammonium chloride and treatment with benzyl chloride (300). This procedure is also suitable for the preparation of S-beneylcysteinyl peptides (240) : CHT-S--5-CHI LNH1
hNH1
L N H R
hO-NHR
CHaSH
-
Na 2h N H I lis. NHI hO--NHR
benayl
CHaSCH&.H,
2 bHNHl
L---NHR
As may be expected, the S-benzyl group can be removed by reduction with sodium in liquid ammonia, so that this substituent is eliminated in the same operation which reduces the carbobenzoxy group. The application of the modified carbobenzoxy method for the synthesis of cysteine peptides is well illustrated by the sequence of reactiona
53
SYNTHESIS OF PEPTIDES
employed by du Vigneaud and Miller (291) for the synthesis of glutathione (cf. also 206s) : CHBCHxCaHt
k
NH4COCHiCeHr
~
~
+ ClCOCHaCHrbHCOOCH:
I
~
~
~
+
NH4COCHrCeHr
LHCOOCH: ~ H &
~
LO-NHCHaCOOCH:
H
~
isponification
r s
reduction
I:::-- lHx c o c H , c H I HCOOH
LO-NHCHxCOOH
By an analogous series of reactions, Kogl and M e r m a n (224) have synthesized 7-D-glutamyl-ccysteinylglycine, which these authors term epi-g-glutat hione. The aspartic acid analog of glutathione, 8-caspartyl-ccysteinylglycine (asparthione) has been synthesized by the coupling of carbobenzoxy-caspartyl-a-benzyl ester @-chloridewith S-benzyl-ccysteinylglycine methyl ester, followed by saponification and reduction of the coupling product (248). It is of interest that, when S-benzylcysteinylglycine was coupled with carbobenzoxy-cglutamic acid anhydride, the condensation product TABLE X Derivatives of cy8h*neand Cy8tine ~~
Compound
M.p., "C. (ref.)
S-Beneyl-ccysteine. . . . . . . . . . . . . . . . . . . . . . . . . . S-Bencyl-mysteine ethyl eater HC1. . . . . . . . . . . . S-Beneyl-mysteinylglycine . . . . . . . . . . . . . . . . . . . . Carbobeneoxy-ccysteinyl-L-tyroshe.. . . . . . . . . . . Carbobencoxy-S-benryl-ccystehe.. . . . . . . . . . . . . Carbobencoxy-S-benzyl-mystebhydraeide. . . . . . Carbobeneoxy-S-benzyl-ccysteinyl-ttyroshe. . .
216-218 d. (300) 156-157 (204) 166-167 (240) 160-162 (204) 93-95 (203) 133-134 (204) 198.200 (204)
Bis(carbobensoxyglycy1)-ccystbe. . . . . . . . . . . . . . 130-132 (222s) Biscarbobensoxy-ccystine. . . . . . . . . . . . . . . . . . . . . 123 (74) Biscarbobeneoxy-bcystinylbisglycine. . . . . . . . . . . 182-183 (240) 210 (196) Biscarbobeneoxy-Lcystinylhis(glycylg1ycine). . . . Bkarbobeneoxy-ccystinylbis-btyrosine.. . . . . . . 158 (204) Monochloroacetyl-mysthe. . . . . . . . . . . . . . . . . . . . 185-190 d. (35,150) Bbchloroacetyl-ccystbe... . . . . . . . . . . . . . . . . . . . 136 (171,255)
l
54
JOSEPH 8. FRUTON
represented the a-peptide derivative of glutamic acid, rather than the y-peptide derivative required for the synthesis of glutathione (cf. also page 42). du Vigneaud et al. (289) took advantage of this result to prepare a-r.,-glutamyl-ccysteinylglycine (isoglutathione) . TABLE XI Peptidea of Cvsteine and Cvstine Peptide Glycyl-bcysteine HCl r,-Cysteinyl-cglutamine L-Cyateinyl-btyrosine L-Cysteinyl-ccy steine Bie(glycy1-cleucy1)-ccyetine Bisglycyl-bcystine Bin(o-alanyl)-bcystine Bis(talany1)-kcystine Bie(bleucylglycy1)-ccystine Bis(bleucyl-qalanyl)-ccystine Bin (cleucyl)-ccystine bCystinylbisglycine bCystinylbk(glycylg1ycine) bCystinylbia(Ltyrosine) -~ a
I
Ref.
+2.4" +6.6@ +15.2 +35 -108.9 -104.3 -227.9 -137.4 -72.2 -115.3 -136.6 -86.0 -55 -50.8
X = 546 mp.
To the list of cysteinyl peptides prepared by means of the carbobenzoxy method there may be added ccysteinyl-cglutamine (203) and ccysteinyl-ctyrosine (204). It may be noted in this connection that either 8-benzyl-N-carbobenzoxy-ccysteinylchloride or the corresponding aside may be employed for the coupling reactions, although the use of the azide appears to give somewhat better yields. For the preparation of cystinyl peptides, the cysteinyl peptides may be oxidized by aeration (204). I n addition, Loring and du Vigneaud (240) made bcystinylbisglycine by treatment of its dicarbobenzoxy derivative with sodium and liquid ammonia and reoxidation by aeration of the sulfhydryl group which appeared during the reduction. In a similar manner, Greenstein (196) prepared ccystinylbis(glycylg1ycine). Although no reports of the synthesis of homocystine or homocysteine peptides have come to our attention, it may be expected that these can be made by the use of the methods described above. In addition, it should be of interest to explore further the possibilities of homocysteine thiolactone as a coupling agent in peptide synthesis in view of the finding
SYNTHESIS O F PEPTIDES
55
of du Vigneaud el d.(292) that this substance, on treatment with alkali, forms a diketopiperazine: S-CHI
It may be added that White (297a) has described the synthesis of glycyltaurine and glycylcysteic acid. 5. Peptides of Serine
Prior to the application of the carbobenzoxy method to the synthesis of serine peptides, two procedures had been employed for this purpose. Fischer and Roesner (163) made glycyl-Dcserine and optically inactive alanylserine from the corresponding chloracetyl and bromopropionyl derivatives. Fischer (137) obtained a partially racemized preparation of Gseryl-cserine by partial hydrolysis of Gserine anhydride, which had been isolated from a n acid hydrolyzate of wool. The hydrolysis of synthetic diketopiperazines containing seryl residues also was employed by Bergmann and Miekeley (62), who described the synthesis of optically inactive phenylalanylserine, and by Abderhalden and Bahn (4) , who reported the preparation of Dcserylglycine and optically inactive serylleucine. The latter workers have claimed that the treatment, with acid, of diketopiperazines composed of a serine residue and another amino acid residue yields seryl dipeptides, while treatment with alkali gives aminoacylserine dipeptides : acid
HN
/ \
CO-CH
R
alkali
Abderhalden and Bahn (4) have prepared tyrosylserine by the reaction of 0,N-dicarbobenzoxy-L-tyrosylchloride (3) with Dkserine, followed by hydrogenation of the coupling product. Esterification of this peptide, followed by treatment with ammonia in methanol gave the corresponding diketopiperazine, which, on partial hydrolysis with acid,
56
JOSEPH 8. FRUTON
yielded a noncrystalline mixture of the diastereoisomeric D- and cserylctyrosines. The use of the carbobenzoxy method has recently permitted the synthesis of a number of cseryl peptides (181). The procedure involved the coupling of carbobenzoxy-cserine azide with an amino acid ester, followed by saponification and hydrogenation: R
CHIOH
I
I
hHNH-OCOCHzCaH'
+ NHkHCOOCH: -+
In this manner, there have been prepared L-serylglycine, L-seryl-calanine, cseryl-cglutamic acid, and cseryl-cserine. Furthermore, Lserylglycyl-cglutamic acid has been made by the carbobenzoxy method as follows (186) : CHzCOOCzHs
CHIOH
AH*
t:
C~H'CH~OCO-NH H +
AO-NHCH;CON: CHzOH I
+
AHNH, AOOCZHI
CHzCOOCzHr I
60-NHCH~CO-NH~H I
C'OOH
An analogous series of reactions has led to the synthesis of L-seryl-b alanyl-cglutamic acid. It will be noted that, in the syntheses just cited, the azides have been used in preference to the acid chlorides. This was made necessary by the presence of the 8-hydroxyl group of serine, although it may be expected that the protection of this group by conversion to a benzyl
57
SYNTHESIS OF PEPTIDES
ether would make it possible to make the desired acid chloride for coupling reactions. In conducting the coupling reaction between carbobenzoxy-cserine azide and an amino acid ester, it is desirable to maintain the temperature below 25", since the azide, when warmed, may undergo an intramolecular condensation to form 4-carbobenzoxyaminooxazolidone-2 (181). CHiOH
CsHrCHzOCO-NH
A
CHIO
+
HCONj
I
CsHsCHzOCO-NH
A'
'co
HNH
A similar reaction had been observed by Schroeter (268) on heating the azide of 8-phenylhydracrylic acid. A recent report by Woolley (303) describes the use of p-toluenesulfonyl-m-serine azide in coupling reactions with amino acid esters. Attention may be called to the excellent method for the isolation of relatively large quantities of L-serine from hydrolyzates of silk fibroin (279). This obviates the need for the resolution of synthetic serine according to the method of Fischer and Jacobs (152). The value of optically pure Lserine and its derivatives in the study of protein metabolism is brought out by the recent findings of Fishman and Artom (174) and of Dent (113). Little can be said a t present concerning the synthesis of peptides of D- or Lthreonine or of D- or Lallothreonine, except that it is to be anticipated that the knowledge gained in the synthesis of serine peptides will TABLE XI1
-
Derivatives of Serine
Compound
M. p., "C. (ref.)
LSerine anhydride. ..................... 0-Benzoyl-Dbserhe ........... Carbobenzoxy-Lserine . . . . . . . . Carbobeneoxy-tserinamide. . . . .
. . . . 132-133 (181)
Carbobenaoxy-bser ylglycine . . . . . . . . . . . . . . . . . . . 131 (181) Carbobenzoxy-Lserylglycinhydrazide. . . . . . . . . . . . . . . . 181-182 (186) Carbobeneoxy-tseryl-Lalanine . . . . . . . . . . . . . . 161-162 (181) Carbobenaoxy-bser yl-Gglu tarn . . . . . . . . . . . . . . . 152-153 (181)
Toluenesulfonyl-Dr,serinhydrazide. . . . . . . . . . . . . . Toluenesulfonyl-Dbserylglycinhydrazide.. . . . . . . . . . . . . 215-216 (303)
58
JOSEPH 8. FRUTON
prove useful in their preparation. It may be hoped that this gap will be filled in the not too distant future, and that threonine peptides and related derivatives will become available for studies of enzyme specificity and intermediary metabolism. TABLE XI11 Peptides of Serine Concn. and solvent
Pcptide
+30.2 -3 0 . 4
~ S eylglycine r L-Seryl-talanine tSeryl-tglutamic acid L-Seryl-Lserine
-9.4 f14.2
26 26 25
25
1
6%, N HCl 6%, N H C I 6%, N HCI 7%, N HCl
Ref. 181 181
181 181
6. Peptides of Lysine and Other Diamino Acids
As in the case of several other amino acids with polar side chain groups, only isolated cases of the synthesis of lysine peptides were described prior to the invention of the carbobenzoxy method. Fischer and Suzuki (172) prepared optically inactive lysyllysine (as a picrate) by partial hydrolysis of the diketopiperaaine. Abderhalden and Sickel (31) obtained peptides of Dclysine, in which both the a- and c-amino groups were involved in peptide linkage, by the treatment of the amino acid with halogen acyl halides, followed by amination of the coupling products. An interesting application of the Schoenheimer method (cf. page 23) was the preparation, by Enger (120),of eglycyl-Dclysine by the conversion of ebenzoyllysine to the a-toluenesulfonyl derivative, removal of the e-beneoyl group by partial hydrolysis with alkali, and treatment of the resulting product with toluenesulfonylglycyl chloride. The toluenesulfonyl groups were then removed by reduction with phosphonium iodide and hydriodic acid and the dipeptide, which was difficult to purify, was obtained as a crystalline picrolonate:
lH1
NH-COC6H'
c:
( Ha14
( Hl)4
+CHaC~H~SOi-NHCH~COCI
__ ____)
AHNH-SO~C~H~CHI ~ H N H - S O ~ C ~ . H & H I
AOOH
AOOH NH-COCH2NH-SO2CaHdCHa
AI
(
Hi),
CHNH-SOIC~H~CHI
AOOH
NH-COCH~NHI PHiI (
+ HI
A
H1)4
LNH2 LOOH
59
BYNTHEBIB OF PEPTIDEB
Turning now to the application of the carbobenzoxy method to the synthesis of lysine peptides, it may be well to discuss first the preparation of clysyl peptides. This presents no special difficulties, since cqa-dicarbobenzoxy-clysine azide may readily be coupled with amino acid or peptide esters in the usual manner: C~H'CHZOCO-NH
CoHsCHnOCONH
I
I
R I
60N:
60-NHbHCOOCH:
For example, Bergmann et al. (813 4 ) made clysylglycine, clysyl-c glutamic acid, clysyl-caspartic acid, and clysyl-ehistidine by the saponification of the appropriate coupling product, followed by hydrogenation in the presence of palladium. I n view of the basic character of clysylglycine and clysyl-chistidine, these dipeptides were isolated as sulfates. Prelog and Wieland (255a) made clysylglycylglycyl-c glutamic acid by an analogous procedure. For the synthesis, by the carbobenzoxy method, of peptides in which both of the amino groups of lysine participate in peptide linkage, special methods are not required, since the reaction of a carbobenzoxyamino acid chloride (or azide) with Glysine methyl ester will lead to the synthesis of ale-lysine peptides. When, however, peptides are desired in which the a-amino group of lysine is bound in peptide linkage and the eamino group is free, the modification introduced by Bergmann et al. (86) must be used. This procedure takes advantage of the property of carbobenzoxyamino acid chlorides to eliminate bensyl chloride and to form the corresponding N-carbonic acid anhydrides (cf. page 27). Thus, when dicarbobenzoxy-clysyl chloride is heated to 50-60°, an anhydride is formed which, on treatment with water, yields ecarbobensoxy-clysine. On treatment with methanol and hydrochloric acid, the anhydride gives a-carbobenzoxy-clysine methyl ester hydrochloride : CsHICH:OCO-YH CeHBCHxOCO-NH
L
( Hi14
CeH'CHxOCO-NH
A
H
bOCl
OC
'I
'0-
0
60
JOSEPH 5. FRUTON
The monocarbobenzoxy ester may be coupled with the chloride or azide of a carbobenzoxyamino acid or peptide to give a dicarbobenzoxy peptide ester which, in turn, may be converted to a hydrazide and azide, thus making it possible to lengthen the peptide chain by a further coupling reaction with an amino acid or peptide ester. After saponification of the coupling product, both carbobenzoxy groups may be removed by catalytic hydrogenation to yield the free peptide: C,H,CH*OCO-NH
+
AHR
Aoa
AOOCHa
LOOH
Bergmann et al. (86) have used this method to prepare a series of lysine derivatives, some of which were later found to be substrates for crystalline pancreatic trypsin (214). For example, a-hippuryl-Llysinamide hydrochloride was made as follows (63,214) :
+ CsH,CO-NHCHzCOCl+
A practical note of some importance in carrying through the conversion of dicarbobenzoxy-Llysine to the chloride (preparatory to the synthesis of the N-carbonic acid anhydride) is the desirability of using the syrupy dicarbobenzoxyamino acid rather than the crystalline derivative. The latter, because of its insolubility in ether, reacts with phosphorus pentachloride extremely slowly. As shown by Synge (282a), the methods outlined above for the synthesis of lysine peptides are applicable t o bornithine a8 well. This
61
SYNTHESIS O F PEPTIDES
TABLE XIV Derivatives of Lysine, Ornithine, Arginine, and Hietidine Compound
M. p., "C. (ref.)
200-202 (214) a-Benzoyl-Llysinamide HCl . . . . . . . . . . . . . . . . . . . . . . t-Carbobenzoxy-Llysine methyl ester HCl cCarbobenzoxy-Llysine-N-carbonicacid a a,cDicarbobenzoxy-Llysine. ...................... 150 (86) a,cDicarbobenzoxy-Llysinamide... . . . . . . . . . . 155 (86) a,t-Dicarbobenzoxy-~lysinhydrazide ... . . . . . . . 159 (81) a,tDicarbobenzoxy-L-lysylglycine. . . . . . . . . . . . . . . . . 158-159 (84) a,cDicarbobenzoxy-Llysyl-cglutamic acid.. . . . . . . . 130 (81) a,t -Carbobenzoxygl ycyl-t-carbobenzoxy-~lysinhydrazide.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 (86) a-Glycyl-Llysine methyl ester 2HC1. ...... pHippury1-clysinamide HCl . . . . . . . . . . . . . a-Hippuryl-ccarbobensoxy-Irlysine.. . . . . . . . . . . . . . . 148-149 (214)
d-Carbobenzoxy-cornithine - (282a) d-Carbobenzoxy-Lornithine-N-carbonic acid anhy86-88 (282a) dride. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . a,&Dicarbobenzoxy-~ornithine.. . . . . . . . . . . . . . . . . . 112-114 (282a) ~ , b D i c a r b o b e n ~ o x y - ~ o m i t h y l - ~ l e u c i n e ~ 2..... H ~ O .63-64 . (282s) a-(Carbobenzoxy-Lvaly1)-bcarbobenzoxy-L ornithine . . . . . . . . . . . . . . 193-195 (282s) LArgininamide 2HC1. ............................ Triacetylan h ydro-D~arginine. . . . . . . . . . . . . . . . . . Benzoyl-Largininamide HCl . . . . . . . . . . . . . . . . . . . . . . Monocarbobenzoxy-Larginirie. . . . . . . . . . . . . Carbobenzoxynitro-~arginine. . . . . . . . . . . . . Carbobenzoxyglyc ylnitro-~arginine Hippurylnitro-~arginine. . . . . . . . . . . . . . . . Hippurylnitro-Largininamidc.. . . . . . . . . . . . . . . . . . . .
148-152 (216)
GHistidine anhydride. . . . . . Bensoyl-Lhistidinamide. . . . Carbobenzoxy-Lhistidine. . . . . . . . . . . . . . 209 (74) Carbobensox y gly cy 1-thistid Carbobenzoxy-Lalanyl-&his Carbobenzoxy-j3-alanyl-Lhistidine.. . . . . . . . . . . . . . . . 160-161 (272)
investigator has also found that 6-carbobenzoxy-cornithine may be prepared conveniently by the reaction of carbobenzoxy chloride with the copper derivative of ornithine. Bergmann and Koster (60) showed that phenylalanylarginine (cf. page 17), on treatment with acetic anhydride, yields phenylalanylornithine :
62
NH~~HCO-NHC~HCOOH ThiB method has many disadvantages, however, not the least of which is the extensive racemization caused by the acetic anhydride. It may be added that the methods developed by nergmann et al. (86) for the synthesis of lysine peptides have been applied with success by Schneider (265) in the synthesis of peptides of a,p-diaminopropionic acid. TABLE XV Peptides of Lysine, Ornithine, Arginine, and Histidine
-
GLysylglycine sulfate ~ L y s y(diglycy1)-bglutamic l acid cLysyl-bglutamic acid cLysyl-baspartic acid cLysyl-bhistidine sulfate a-(cValyl)-bornit.hine HCl.%H,O cOmithyl-cleucine HCI.jf;HaO
Concn. and solvent
22 20 19 20 20
1.5%, H20 1%, HzO 3.1%, HzO 1.3%, Hz0 1.3%, H20
84 255a 81 84 81
+2
19 18
2%, Hz0 2.1%, H20
282a 282a
...
...
...
...
...
+25 -2.5 +27.0 -20.4 +20.5 +12.3 +32.1
26 24 27 28 25 30 18
+
30 +33.5 +22.9 +23 +35.3
+19
Glycyl-barginine sulfate Glycylglycyl-bhistidine sulfate Glycyl-bhistidine HCI D-Alanyl-bhistidine GAlany 1-bhistidine p-Alanyl-D-histidine p- Alany 1-bhistidine p-Alanyl-1-methyl-bhietidine L-Leucyl-thistidine
I
Temp. "C.
Peptide
-
... 1%, HzO 1%, Hz0 I%, HtO 2%, H2O 2%, HZ0 5%, Hz0 5%, Hz0
1
Ref.
~
.83 258 221 220 220 240 272 45 147
P
7. Peptides of Arginine Of the protein amino acids, arginine presents perhaps the greatest difliculties in peptide synthesis. This is due, in large part, to the strongly
SYNTHESIS OF PEPTIDES
63
basic character of the guanido group. Because of the physiological importance of arginine and its derivatives, numerous attempts have been made to overcome these difficulties and to solve the problems encouutered in the synthesis of arginine peptides. Fitxher and Suzuki (172) obtained, by the autocondensation of Garginine methyl ester, a substance which they designated tentatively as “arginylarginine.” Experiments by these and later workers (118, 226), however, cast considerable doubt on this formulation of the reaction product, but it remained for Zervas and Bergmann (305) to show that the material actually was the anhydride of a,6-bisguanido-n-valeric acid. What is more, it was found that, in the course of the reaction, ornithine methyl ester is formed. It became clear, therefore, that the autocondensation of Garginine methyl ester leads to a dismutation of arginine into ornithine and bisguanidovaleric acid : NHz
NH,
A=NH
L=NH
ILH
AH
AI
( Hz)a
CHNHl COOCHI I
NHz
I
(CHI):
‘ b Hi): NH + L N H Z (
LHNHJNH
b0-
LOOCHI
I
Some progrem in the synthesis of arginine peptides came with the demonstration by Bergmann and Koster (60) that acetylaminocinnamic acid azlactone reacts readily with carginine in alkaline solution to give acetyldehydrophenylalanyl-carginhe. On hydrogenation in the presence of palladium black, there resulted a mixture of diastereoisomers which were hydrolyred with dilute hydrochloric acid to remove the acetyl groups. By the use of salicylaldehyde (an excellent precipitant for arginine and other amino acids, 70), these workers succeeded in isolating D-phenylalanyl-carginine, but were unable to obtain the isomeric cphenylalanyl-carginine in pure form. An analogous sequence of reactions, using acetylamino-p-acetoxycinnamic acid azlactone, led to the synthesis of D-tyrosyl-1,-arginine (90). The limited possibilities of the azlactone method for the synthesis of arginine peptides led to the application of the carbobenzoxy method when it became available. The principal problem was to find a substituent for the guanido group which would mask its basic properties and which also could be removed by catalytic hydrogenation at the end of the peptide synthesis. Bergmann et al. (83) showed that the nitro group was suitable for this purpose, since nitro-carginine (225) could be. con-
64
JOSEPH 8. FRUTON
verted to carginine by reduction with hydrogen in the presence of palladium black. It was noted, however, that, at normal temperature and pressure, the hydrogenation is slow. Nitro-L-arginine was used by Bergmann et al. for the synthesis of glycyl-barginine sulfate.
dOOH
In later studies by Bergmann and collaborators, it was shown that crystalline pancreatic trypsin hydrolyzes certain arginine derivatives (55), some of which were prepared by the method outlined above. Thus, hippuryl-L-argininamide hydrochloride was synthesized by the coupling of hippuryl chloride with nitro-carginine methyl ester, followed by treatment with ammonia and hydrogenation of the nitro group (216). It may be added that a typical substrate for crystalline trypsin, benzoylcargininamide hydrochloride, may be prepared from the corresponding methyl ester by treatment with ammonia (55,114). Thus far, no well-defined peptides have been synthesized in which the carboxyl group of arginine is linked to another amino acid. The substituents employed thus far to block the guanido group are not sufficient to mask its reactivity, and, as a result, attempts to convert the carboxyl group of carbobenzoxynitro-carginine or dibenzoyl-carginine to an acid chloride or an aside have not been successful. Nor has it been possible to prepare dicarbobenzoxy-carginine (186) by the methods which yield other disubstitution products of arginine, such as dibenzoylcarginine (123) or dibenzenesulfonyl-L-arginine (100). Another possible approach to the synthesis of arginyl peptides, which remains to be explored, may be the prior synthesis of the corresponding ornithyl peptides by the method discussed previously for lysine peptides, followed by the conversion of the &amino group to a guanido group by treatment with cyanamide (42), guanidine (256), S-methylisothiourea or O-methylisourea (196a,269).
65
SYNTHESIS OF PEPTIDES
8. Peptides of Histidine The preparation of peptides in which the amino group of Ghistidine is linked to another amino acid may readily be achieved by the coupling of histidine methyl ester with acid chlorides or azides. Histidine methyl ester is best prepared from the dihydrochloride by neutralization with sodium methylate according to the directions of Fischer and Cone (147). The reaction of d-a-bromoisocapronyl chloride with L-histidine methyl ester yields a crystalline product which, upon saponification and amination, gives Lleucyl-chistidine (147). In a similar manner, Abderhalden and Geidel (11) prepared glycyl-chistidine and related dipeptides of histidine. I n general, the yields in these syntheses were unsatisfactory. The introduction of the cnrbobenzoxy method made it possible to synthesize a large variety of histidine peptides in good yield. du Vigneaud and collaborators have used this method with marked success in the syntliesis of glycyl-L-histidine (221), calanyl-L-histidine (220), 8 - e aspartyl-L-histidine (288), as well as several other similar peptides. The procedure usually employed was to allow the appropriate carbobenzoxyamino acid chloride to react with Ghistidine methyl ester, and the carbobenzoxy peptide ester was saponified and hydrogenated. In the synthesis of a-caspartyl-D-histidine, Greenstein and Klemperer (197) used carbobenzoxy-Laspartic acid anhydride for the coupling reaction. In the synthesis of p-alanyl-chistidine (carnosine) or 8-alanyl-1-methyl-c histidine (anserine), du Vigneaud and associates condensed carbobenzoxy8-alanyl aside with the appropriate ester (45,272):
CaH,CH*OCO-NHCH,CH&ONa
+ NHn
I
I
N \Ck
NH--OCOCH&sHr AH*
NH
ssDonification
NHzCH&H&O-NH
COOH
bHCHzC===C
H
A practical detail which may be noted in connection with syntheses of the type mentioned in the preceding paragraph is that, because of the basic character of the imino group of the imidazole ring, acid is to be
66
JOSEPH 8. FRUTON
omitted in the extraction of the reaction mixture (cf. page 28). Instead, several extractions with water are substituted. In the course of their studies on the synthesis of amino-N-methy1-Lhistidine, du Vigneaud and Behrens (287) found that the imino group of the imidazole ring, after treatment with sodium in liquid ammonia, would react with benzyl chloride to give a benzyl derivative. The benzyl group may be removed by reduction with sodium in liquid ammonia, as in the case of carbobenzoxy and S-benzyl groups. In this manner, it was possible to block the imino group while methylating the a-toluenesulfonylamino group according to the method of Fischer and Bergmann (145). It may be added that du Vigneaud and Behrens also found that the toluenesulfonyl group may be removed by reduction with sodium and liquid ammonia. There are relatively few examples of the synthesis of histidyl pepti,des. The first of these was the preparation of histidylhistidine from L-histidine anhydride (22,172); no data are available, however, to indicate the extent of racemization of the dipeptide. Another method for the synthesis of Dthistidyl peptides was described by Bergmann and Zervas (73), who showed that the treatment of acetyl-Dthistidine with acetic anhydride gives rise to a product (not isolated, but assumed to have the azlactone structure indicated below) which reacts wkh glycine ethyl ester to give acetyl-Dthistidylglycine ethyl ester. Brief hydrolysis with dilute hydrochloric acid gave the desired peptide, although the yield in this method is an extremely modest one. CH==CCH&HCOOH
I
NH
I
N
\C<
I
NH
( C H K O ) rO
--+
(!!OCHI CHaCO-N
CH==CCH&H-cO
I
I
\C
CH==CCH&HCO-NHCH&OOC1H,
I
NH \C&
I
N
I
NH LOCHS
I
A-
+NH1CHtCOOCtRfi r
N\C/
(!Ha CH==CCH&HCO-NHCH*COOH +AH
NI
NHx I
\Cif
In the course of these studies, Bergmann and Zervas (72) also found that the imidazole ring of beneoyl-chistidine methyl ester could be subetituted with a hippuryl residue by treatment with hippuryl chloride. The resulting hippurylbenzoyl-thistidine methyl ester reacts with
67
SYNTHESIS OF PEPTIDES
glycine to give benzoylglycylglycine and to regenerate the benzoyl-L histidine methyl ester: CH=CCH&HCOOCHI
I
NH
I
CHcCCHrCHCOOCHa
I
N
CaH,CO-NHCH*COCI
NH
I
1
N
NH
\C<
A\
\cg
I
N
AOCaH,
OCHtNH-COCaH, NHaCHaCOOH
CaH~CO-NHCH*CO-NHCH&OOH
9. Peptides of Proline and Hydroxyproline Interest in the synthesis of proline peptides was stimulated in the early part of this century by the finding of prolylglycine diketopiperazine in tryptic digests of gelatin (232) and of Lprolyl-kphenylalanine in acid hydrolyzates of gliadin (253). The first systematic efforts to make proline peptides were those of Fischer, who applied the halogen acyl halide method. He found, however, that the treatment of bromoisocapronyl-Dcproline with ammonia does not lead to the formation of the expected leucylproline, as had been originally thought (143), but rather results in the formation of oxyisocapronylprolinamide (162) : CH,
CHI
CH, COOH
‘Ck
I
CHn
I
BrCHCO-N
&H-cHr
’ 1
\
CHI-
NH:
\C$ I
CHI CONHz
I
HI
Abderhalden and Merkel (23) have reported, however, that the reaction of halogen acyl halides with prolyl peptides yields products which may be converted to longer peptides by amination. They describe the synthesis of peptides such as glycyl-L-prolyl-cphenylalanine and L-leucyl-c prolyl-ctyrosine. The halogen acyl halide method was also applied to the synthesis of prolyl peptides, and in this case better success was achieved. It was found that the condensation of a,d-dibromovaleryl chloride (169,177) with amino acids gave crystalline pr0duct.s which, on treatment with ammonia, yielded the desired DL-prolyl peptides. The sequence of reactions is i!lustrated below for the synthesis of Dcprolylglycine (194).
68
JOSEPH 8. FRUTON
x'
Br(CH2)J HCOCl
+ NH&H&OOH
-+
CH-CHz
Br
Br(CHZ)~AHCO-NHCHzCOOH
2b H z
\ /
hHCO-NHCHzCOOH
NH
In a similar manner, Grassmann et al. (194) made D~prolylglycylglycine and Abderhalden and Nienburg (25) prepared D- and L-prolyl-L-leucine. A second method devised by Fischer for the synthesis of prolyl peptides involved the use of L-prolyl chloride, prepared by the treatment of cproline with phosphorus pentachloride in acetyl chloride (162). By coupling this amino acid chloride with D- or L-phenylalanine ethyl ester, Fischer and Luniak (159) obtained products which, after saponification, yielded L-prolyl-D-phenylalanine or L-prolyl-L-phenylalanine. These authors also demonstrated the identity of the second of these two isomers with the peptide isolated by Osborne and Clapp from gliadin hydrolyzates (253). As in the case of the synthesis of peptides of other amino acids, the introduction of the carbobenzoxy method notably widened the range of possibilities for the synthesis of proline peptides. Bergmann et al. (89) prepared glycyl-L-proline and L-alanyl-L-proline by the reaction of the appropriate carbobenzoxyaminoacyl chloride with cproline, followed by the hydrogenation of the carbobenzoxy dipeptides:
g-K"
HCOOH \Nk
+ CsHsCHzOCO-NHCHzCOCl-+ CHI-CHz AHz
\ /
AHCOOH
N-COCHZNH-OCOCHzCsHs
CH 1 - C H 2 2 b H ~ &HCOOH
\ /
N-COCHzNHz
Abderhalden and Nienburg (25) also used the carbobenzoxy method to make L-prolyl peptides by the coupling of carbobenzoxy-L-prolyl chloride with various amino acids or their esters. In this manner, L-prolylglycine, L-prolyl-L-tyrosine, and L-prolyl-L-leucine were synthesized (cf. also 217). Other workers have described the synthesis of L-prolyl-L-vnlinc (282a) and of L-prolyl-L-glutamic acid (273a) by this method. I t was noted that, after the hydrogenation of carbobenzoxy-L-prolyl-L-proline, the product was not the expected dipeptide, but rather cproline anhydride. This result focuses attention on the pronounced tendency of peptides of proline to undergo cyclization to diketopiperazines, a reaction for which Smith and Bergmann (276) have provided additional experi-
69
SYNTHESIS OF PEPTIDES
mental evidence. These workers showed that, in the course of the hydrogenation of carhobenzoxyglycyl-L-prolinamide, there resulted glycyl-L-proline diket opiperazine : CO-N
CHz--- CHz AHz
\"
I
CsHsCHZOCO-NHCH?CO
AHCONHz -+
,CH*-[H*
/
\
\-
/
CHa NH-CO
CH-
H*
Smith and Bergmann also prepared glycylglycyl-L-proline by the coupling of carbobenzoxyglycylglycyl azide with tproline benzyl ester, followed by the hydrogenation of the reaction product. A recent paper of Bergmann and Tietzman (68) provides a n additional method for the preparation of proline peptides. The reaction of acetaminocinnamic acid azlactone with L-proline yields a crystalline acetyldehydrophenylalanyl-L-proline (102), which, on treatment with acetic anhydride, is converted to N-acetyldehydrophenylalanyl-L-proline diket,opiperazine. Whereas diketopiperazines are, in general, rather stable, this compound is quite reactive for, when it is treated with L-leucine, there are formed acetyl-L-leucine and Dtprolyldehydrophenylalanine. On the other hand, if the N-acetyldiketopiperazine is treated with glycine, there results dehydrophenylalanyl-Gproline diketopiperazine. TABLE XVI Derivatives of Proline and Hydrozyproline Compound M. p., "C. (ref.) 173-175 (276) cProlinamide HCl. . . . . . . . . . . . . . . . . . . . . . . . . . . . . Carbobenzoxy-L-prolyl-bvaline. . . . . . . . . . . . . . . . . . . 134-135 (282a) Carbobenzoxy-bprolyl-rrproIine.. . . . . . . . . . . . . . . . . 186-187 (25) Carbobenzoxyglycyl-bprolirie . . . . . . . . . . . . . . . . . . . . 156 (89) Carbobenzoxyglycyl-bprolylglycine. . . . . . . . . . . . . . . 136-137 (273a) Carbobenzoxyglyryl-Lproliriamide. . . . . . . . . . . . . . . . 15C151 (276) Glycyl-bproline diketopiperazine . . . . . . . . . . . . . . . . . 213 (162,276) Hydroxy-~prolinamide. . . . . . . . . . . . . . . . . . . . . . . . . 139 (276) Hydroxy-Lproline benzyl ester HCI. . . . . . . . . . . . . . 147-150 (276) Hydroxy-tproline methyl ester HCl . . . . . . . . . . . . . . 162-164 (276) Carbobenzoxyhydroxy-bprolinhydrazide.. . . . . . . . . 149.5 (276) Carbobenzoxyglycyl-hydroxy- prol line . . . . . . . . . . . . 124.5(276) Carbobenzoxyglycyl-hydroxy-bprolinamide.. . . . . . . 208 (276) Chloroacetyl-hydroxy-Lproline.. . . . . . . . . . . . . . . . . . 160 (20)
70
JOSEPH 8. FRUTON
After the hydrogenation of the last-named diketopiperazine, it is possible to isolate tphenylalanpl-L-proline diketopiperazine, which, on partial hydrolysis with acid, yields tprolyl-tphenylalanine. This dipeptide is ident,ical with the substance isolated by Osborne and Clapp (253) from a gliadin hydrolyzate and synthesized by Fischer and Luniak (159). In addition to the two prolyl peptides mentioned above, Bergmann and TABLE XVII Peptides of Proline and 1 droxy..ptoline
-
-
Peptidc
Temp.,
Ref.
blD
"C.
Glycylgl ycyl- prol line GIycyl- prol line Glycy l - ~ p r oylgl l ycine LAlanyl- prol line &Alanyl-Lproline LLeucyl-r,prol yl-ttyrosinc ~ P r oylgl l ycine LProlyl-Lalanine r,Prolyl-Lvaline.HzO t P r o l y I- leuc cine LProlyl-Lglutnmic acid ~ P r o l y l - i p h e ylalanine n LProlyl-Ltyrosine LProlyl-tserine
-101.5 -113.8 -109.2 -114.4 -93.3 -7.5 -22.8 -66.3 -61 -73.5 -49.0 -40.9 -7.8 -47.9
26 20 21 20 26 25 20 20 20 25 20 20 20
276 89 273a 89 20 la 23 23,25,247 23 282a 25 273a 23,68,159 25,247 23
Glycylgl ycyl-h ydroxy-r-proline Glycyl-hydroxy-r,-proline Hydroxy-Lprolylglycine
-97.7 -128.4 -22.4
26 26 26
276 276 276
20
P
71
SYNTHESIS OF PEPTIDES
Tietzman (68) obtained several other proline derivatives as a result of the extremely interesting reactions of the N-acetyl diketopiperazine. Despite its obvious limitations for peptide synthesis, this subst*ance,and others related to it, deserves further study. It. may be added that the formation of acetyl-cleucine during the interaction of the acylated diketopiperaxine with leucine is yet another case of acyl migration, similar to that noted by Bergmann et al. (69) in the reaction of diacetylglycine anhydride wi t,h glycine or carginine. With regard to the synthesis of peptides of hydroxy-L-proline, it may suffice to indicate that the carbobenzoxy method has permitted the preparation of glycylhydroxy- proli line, glycylglycylhydroxy-L-proline, and hydroxy-L-prolylglycine (276) by procedures analogous to those employed for the corresponding proline peptides. As in the case of
-
TABLE XVIII Derivatives of Tryptophan and Methionine i
hl. p., "C. (ref.)
Compound
-
..... 90 (18,186) LTryptophan methyl ester.. Carbobenzoxy-~tryptophnn. . . . . . . . . . . . . . . . . . . . . . 126 (275a) Carbobenzoxy-DLtryptoplian . . . . . . . 169-170 (275a) Carbobenzoxy-ctryptophyl chloride. . . . . . . . . . , . . . . 75 d. (275a) ..... Carbobenzoxy-~tryptophylglycine. . . . . . 156 (275a) Carbobenzoxy- try ptoph y I-Lalanine . . . . . . . . . , . . . . 165 (275a) Carbobenzoxy-~tryptoph yl- tryptophan . . . . . . . . . . . 207 (275a) Carbobenzoxyglycyl-L-tryptophan .. , . . . . . . . . . . . . . . 142 (275a) Carbobenzoxyglycyl-~tryptophanamide. . . . . . . . . . , . 145 (275a) n-kucyl-Ltryptophan dikctopiperazine.. . . . . . . . . . , 218-219 (185a) LMethioninamide acctatc. . . . . . . . . . . . . . . . . . . . . . . . Carbobenzoxy-D-methionine . . . . . . . . . . . . . . . . . . . .... .. . . Carbobenzoxy-Di,methionine. . Carbohenzoxy-~methioniilaniitle . .. . . . . . . . . . ... . . . Carbobenzoxy-D~methiorlinnmidc. , , . . . . . . . . . . Carbobenzoxy-~methioninhydrazide , , . . . . . . . . . . . . Carbobenzoxy-Dbmethioninh ydrazide . . . . . .. . . . . . . . Carbobenzoxy-~methionylglycine. . . . . . . . . . . . . . . , . Carbobenzoxy-D~methionylglycinc . . . . . . . . . . . Carbobenzoxy-Lmcthiony lglycinamide . . . . . . . . . . . . . Carbobenzoxy-~methionylglycyIgl yrinc . . . . . . . . . . . . Carbobenzoxy-lrmethionyI- tyros sine . . . . . . . . . . . . . . Carbobenzoxy-Lmethiony I-tmethioninc . . . . . . , . . , . Carbobenzoxy-rrmethionyl-r,methioninamidc. . . . . . . Carbobenzoxyglycyl-~methionine .. . . . . . . . . . . . . . , . Carbobenzoxygl ycyl-~methioninamide Carbobenzoxy-a-~glutamyI-L-methionine . . . . . . . . . .
.
103 (112b) 69-70 (112a) 112 (112a) 125 (112b) 110-121 (112b) 110-112 (112b) 107-108 (112b) 130-131 (112b) 141-143 (112b) 120-122 (112b) 137-138 (112b) 137-138 5 (112b) 118-120 (112b) 196 (112b) 110-111 (112b) 131-134 (112b) 135-136 (112b)
72
JOBEPH 8. FRUTON
glycyl-L-proline, glycylhydroxy-L-proline tends to cyclize in alkaline solution, and, furthermore, the hydrogenation of carbobenzoxyglycylhydroxy-L-prolinamide gives rise to glycylhydroxy-L-proline diketopiperazine. It is clear, therefore, that peptides of hydroxy-L-proline share the tendency, exhibited by proline peptides, of diketopiperazine formation. 10. Peptides of Tryptophan Despite the considerable physiological importance of L-tryptophan, few peptides of this amino acid have been described in the literature. Little has been added to the work of Abderhalden and Kempe (19), who made L-tryptophylglycine, albeit with great difficulty, by coupling L-tryptophyl chloride (18) with glycine. The same authors also described the preparation of glycyl-L-tryptophan, L-alanyl-L-tryptophan, and cleucyl-&tryptophan by the reaction of the appropriate halogen acyl halide with L-tryptophan, followed by the amination of the coupling product. Recent work has shown th at the carbobenzoxy method is readily applicable to the synthesis of tryptophan peptides. This method has been used for the preparation of D-leucyl-L-tryptophan diketopiperazine (185a) and Smith (275a) has reported the synthesis of a number of carbobenzoxy dipeptides containing Gtryptophan residues. It, may be expected that, in view of the increased availability of tryptophan a t TABLE XIX Peptides of Tryptophan and Methionine [a]D
Temp., "C.
Glycy 1-&tryptophan 1,Alanyl-Ltryptophan bLeucylgly cyl-Ltryptophan L-Leucyl-rrtryptophan L-Leucyl-~tryptoph y I-L-glutamic acid LTryptoph ylglycine LTryptophyl-Lglutamic acid
$21.6 +18.7 +32.3 +4.5
20 20 20 20
5,19 19,29 19 19
+17.4 +78.7 +34.4
20 20 20
1 19 1
Gly cyl-D-methionine Glycyl-tmethionine a-LGlutamyl-Lmethionine D-Methionylglycine LMethionylglycine LMethionylglycylglyche LMethionyl-Ltyrosine 1,MethionyLLmethionine
+9.1 -10.0 +18.8 -88.0 +86.8 +73.1 +18.6 -+26.5
25 22 25 23 21 23 22 25
112b 112h,213 112b 112b 112b 112b 112b I12b
Peptide
-
Concn. and solvell t
SYNTHESIS O F PEPTIDES
73
relatively low cost, further progress in the synthesis of peptides of this amino acid may be forthcoming. The properties of such peptides and their derivatives should be of interest in connection with the study of the antibacterial agent gramicidin, which is particularly rich in tryptophan content. 11. Peptides of Methionine
Until recently, the only methionine peptide mentioned in the literature was glycyl-L-methionine prepared by Hess and Sullivan (213) by the amination of the chloroacetyl derivative. The use of the carbobenzoxy method has given a preparation of this peptide melting about 55" higher than the product described in (213), and has also permitted the synthesis of an extensive series of peptides such as L-methionjlglycine and c m e t h ionyl-L-tyrosine by the coupling of carbobenzoxy-L-methionine azide with the appropriate amino acid ester, followed by saponification and catalytic hydrogenation with palladium black (112b). In some cases, however, the removal of the carbobenzoxy group could not be accomplished readily with palladium black as the catalyst, and here reduction by means of sodium and liquid ammonia proved effective. In view of the possible demethylation of the methionine side chain to give derivatives of homocysteine (279a), methyl iodide was added a t the end of the reaction so as to ensure remethylation of the sulfur. 12. Peptides of Other Amino Acids In the preceding discussion of the special aspects of peptide synthesis, primary attention has been given t o peptides related to amino acids which are definitely known to arise on the hydrolysis of proteins. In what follows, there are added some comments concerning the synthesis of peptides containing amino acid residues not shown to occur in proteins but which nevertheless have some physiological interest. Among the nonprotein amino acids, p-aminobenzoic acid has recently assumed increased interest because of the recognition of its role as a vitamin and also because of its identification as a component, of folic acid (380). The fact that, in folic acid, the p-aminobenzoyl residue is linked to glutamic acid by a peptide bond confers upon peptides composed of these two amino acids considerable importance. I n this connection, it may be noted th at Balls and Kohler (40) have synthesized glycyl-paminobenzoic acid by the halogen acyl halide method. I t may be expected that the use of the carbobenzoxy method will make it possible to prepare more complex members of this group of peptides.
74
JOSEPH 8. FRUTON
Another nonprotein amino acid which has been used for peptide synthesis is D,-glucosaminic acid as well as its enantiomorph D,-epiglucosaminic acid. I n the case of the latter, coupling with carbobenzoxyglycyl chloride gave a crystalline product, which, upon hydrogenation, yielded glycyl-D,-epigliicosaminic acid (85). The fact that this dipeptide was hydrolyzed by peptidases present in extracts of intestinal mucosa was taken to support the view that the configuration of carbon atom 2 of epiglucosaminic acid is the same as that of the asymmetric carbon atom of L-alanine. I n the course of this investigation, there was also synthesized L,-phenylalanyl-D,-glucosaminic acid by the condensation of carbobenzoxy-L-phenylalanylchloride with benzal-D,-glucosaminic acid ethyl ester, followed by saponification of the coupling product and removal of the carbobenzoxy and benzal groups by catalytic hydrogenation :
lHxCsH8 C~H~CH1OCO-NII HCO-NH
rCzH6 aaponifioation
H-CH-CH-CH-CHI AH AH
d
d
reduction
\ck
AH AH AH
In this connection, reference may be made to the application of the carbobenzoxy method to the synthesis of compounds in which the amino group of glucosumine is linked to an amino acid. Bergmann and Zervas (75) have described the synthesis of glycyl-D,-glucosamine and L,-alanylD,-glucosamine by the reaction of the appropriate carbobenzoxyamino acid chloride with tetraacetyl-D,-glucosamine, saponification of the coupling product with sodium methylate, and finally, hydrogenation to remove the carbobenzoxy group. A nonprotein amino acid which has recently attracted considerable interest is /3,/3-dimethyl-D-cysteine, shown to be a component of the thiazolidine ring of the various penicillins (102). At the present writing, the details of the synthetic work conducted in the United States and in
SYNTHESIS OF PEPTIDES
75
England are not avaiIahIe. * Miescher and collaborators, however, have recently described some synthetic reactions which involve the coupling of hippuryl chloride and of glycyl chloride (133) with thiazolidine-4carboxylic acid methyl ester (257) and with 5,5-dimethylthiazolidine-4carboxylic acid methyl ester (251,252,294). REFERENCES 1. Abderhalden, E. (1909). Ber. 42, 2331. 2. Abderhalden, E., Abderhalden, R., Weidle, H., Baertich, E., and Morneweg, (1938). Fermenfforschung 16, 98. 3. Abderhalden, E., and Bahn, A. (1933). 2.physiol. Chem. 219, 72. 4. Abderhalden, E., and Bahn, A. (1935). 2. physiol. Chem. 234, 181. 5. Abderhalden, E., and Baumann, L. (1908). Bet. 41, 2857. 6. Abderhalden, E., and Brockmann, H. (1928). Fermenlforschung 9, 446. 7. Abderhalden, E., and Buadze, S. (1926). Fermentforschung 8, 489. 8. Abderhalden, E., and Fleischmann, R. (1928). Fermentforschung 9, 524. 9. Abderhalden, E., and Fodor, A. (1912). 2. physiol. Chem. 81, 1. 10. Abderhalden, E., and Fodor, A. (1916). Ber. 49, 561. 11. Abderhalden, E., and Geidel, W. (1931). Fermentforschung 12, 518. 12. Abderhalden, E., and Gohdes, W. (19319. Fermentforschung 13, 52. 13. Abderhalden, E., and Guggenheim, M. (1908). Ber. 41,1237. 14. Abderhalden, E., and Haase, E. (1932). Fermentforschung 13, 303. 15. Abderhalden, E., and Hirsch, P. (1910). Ber. 43, 2435. 16. Abderhalden, E., Hirsch, P., and Schuler, J. (1909). Ber. 42, 3394. 17. Abderhalden, E., and Hirszowski, A. (1908). Ber. 41, 2840. 18. Abderhalden, E., and Kempe, M. (1907). 2.physiol. Chem. 62, 207. 19. Abderhslden, E., and Kempe, M. (1907). Ber. 40, 2737. 20. Abderhalden, E., and Koppel, W. (1928). Fermentforschung 9, 439. 21. Abderhalden, E., and Koppel, W. (1928). Fermenfforschung 9, 516. 22. Abderhalden, E., and Leinert, F. (1937). Fermenlforschung 16, 324. 23. Abderhalden, E., and Merkel, R. (1936). Fermenlforschung 16, 1. 24. Abderhalden, E., and Neumann, A. (1934). Fermentforschung 14, 133. 25. Abderhalden, E., and Nienburg, H. (1933). Fermentforschung 13, 573. 26. Abderhalden, E., and Rossner. E. (1927). 2. physiol. Chem. 163, 149. 27. Abderhalden, E., and Saito, M. (1930). Fennentforschung 11, 539. 28. Abderhalden, E., and Schuler, J. (1910). Bet. 43, 907. 29. Abderhalden, E., and Sickel, 11. (1927). 2. physiol. Chem. 171, 93. 30. Abderhalden, E., and Sickel, I€. (1928). Fermentforschung 9, 462. 31. Abderhalden, E., and Sickel, II. (1928). Fermentforschung 10, 302. 32. Abderhalden, E., and Singer, W. (1926). Fermenlforschung 8, 187. 33. Abderhalden, E., and Vlassopoulos, V. (1929). Fermentforschung 10, 365. 34. Abderhalden, E., and Weber, L. E. (1910). Ber. 43, 2429. 35. Abderhalden, E., and Wybert, E. (1916). Ber. 49, 2449, 2838. 36. Abderhalden, E., and Zeisset, W. (1931). 2. physiol. Chem. 200, 179.
W.
While this review was in the press, there appeared the monograph summarizing the wartime research on the chemistry of penicillin (100a). Regrettably, it has not been possiblr to include in the present article the many important contributions to peptide chemistry reported in the monograph.
76 37. 38. 39. 40. 41. 42. 43. 44.
JOSEPH 8. FRUTON
Abderhalden, E., and Zeisset, W. (1932). Fermenlforachung 13, 330. Abenius, P. W., and Widmark, 0. (1888). Ber. 21, 1662. Balbiano, L. (1901). Ber. 34, 1501. Balls, A. K., and Kohler, F. (1931). Ber. 64, 34. Barkdoll, A. E., and Ross, W. F. (1944). J . Am. Chem. SOC.88, 951. Baumann, E. (1873). Ann. 187, 83. Behrens, 0. K., and Bergmann, M. (1939). J. Biol. Chem. 129, 587. Behrens, 0. K., Doherty, D. G., and Bergmann, M. (1940). J. Biol. Chem. 138, 61.
45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65.
Behrens, 0. K., and du Vigneaud, V. (1937). J . B i d . Chem. 120, 517. Bergmann, M. (1924). Naturwissenschaften 12, 1158. Bergmann, M. (1942). Advances i n Enzymol. 2, 49. Bergmann, M., and Fraenkel-Conrat, H. (1938). J Biol. Chem. 124, 1. Bergmann, M., and Fruton, J. S. (1937). J . Biol. Chem. 117, 189. Bergmann, M., and Fruton, J. S. (1937). J . Biol. Chem. 118, 405 Bergmann, M., and Fruton, J. 5. (1938). J . Biol. Chem. 124, 321. Bergmann, M., and Fruton, J. S. (1941). Advances in Eneymol. 1, 63. Bergmann, M., and Fruton, J. S. (1942). J . Biol. Chem. 146, 247. Bergmann, M., and Fruton, J. S. (1944). Ann. N . Y . Acad. Sci. 46, 409. Bergmann, M., Fruton, J. S., and Pollok, H. (1939). J . Biol. Chem. 127, 643. Bergmann, M., and Grafe, K. (1930). 2. phyaiol. Chem. 187, 187. Bergmann, M., and Grafe, K. (1930). 2. physiol. Chem. 187, 196. Bergmann, M., Kann, E., and Miekeley, A. (1926). Ann. 449, 135. Bergmann, M., and Koster, H. (1926). 2. phyaiol. Chem. 169, 179. Bergmann, M., and Koster, H. (1927). 2.phyaiol. Chem. 187, 91. Bergmann, M., and Miekeley, A. (1924). 2.phyaiol. Chem. 140, 128. Bergmann, M., and Miekeley, A. (1927). Ann. 468, 40. Bergmann, M., and Ross, W. F. (1936). J . Am. Chem. SOC.68, 1503. Bergmann, M., and Schleich, H. (1932). 2. physiol. Chem. 206, 65. Bergmann, M., Schmitt, V., and Miekeley, A. (1930). 2. physiol. Chem. 187, 264.
Bergmann, M., and Stern, F. (1926). Ann. 448, 20. Bergmann, M., Stern, F., and With, C. (1926). Ann. 449, 277. Bergmann, M., and Tietzman, J. E. (1944). J . Biol. Chem. 166, 535. Bergmann, M., du Vigneaud, V., and Zervas, L. (1929). Ber. 82, 1909. Bergmann, M., and Zervas, L. (1926). 2. phyaiol. Chem. 162, 282. Bergmann, M., and Zervas, L. (1928). Bimhem. 2. 303, 280. Bergmann, M., and Zervas, L. (1928). 2. phyaiol. Chem. 176, 145. Bergmann, M., and Zervas, L. (1928). 2.phyaiol. Chem. 176, 154. Bergmann, M., and Zervas, L. (1932). Ber. 86, 1192. Bergmann, M., and Zervas, L. (1932). Ber. 86, 1201. Bergmann, M., and Zervas, L. (1933). 2.phyaiol. Chem. 221, 51. Bergmann, M., and Zervas, L. (1936). J . B i d . Chem. 113, 341. Bergmann, M., Zervas, L., and Fruton, J. S. (1935). J . Biol. Chem. 111, 225. Bergmann, M., Zervas, L., and Fruton, J. S. (1936). J . Biol. Chem. 116,593. Bergmann, M., Zervas, L., Fruton, J. S., Schneider, F., and Schleich, H. (1935). J . Biol. Chem. 109, 325. 81. Bergmann, M., Zervas, L., and Greenstein, J. P. (1932). Ber. 86, 1692. 82. Bergmann, M., Zervas, L., and Koster, H. (1929). Ber. 62, 1901. 83. Bergmann, M., Zervaa, L., and Rinke, H. (1934). 2. physid. Chem. 224, 40.
66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80.
SYNTHESIS OF PEPTIDES
77
84. Bergmann, M., Zervas, L., Rinke, H., and Schleich, H. (1934). Z. physiol. Chem. 224, 26. 85. Bergmann, M., Zervas, L., ltinke, H., and Schleich, H. (1934). 2. physiol. Chem. 224, 33. 86. Bergmann, M., Zervas, L., and RQSS, W. F. (1935). J . Biol. Chem. 111, 245. 87. Bergmann, M., Zervas, L., and Salzmann, L. (1933). Ber. 66, 1288. 88. Bergmann, M., Zervas, L., Salzmann, L., and Schleich, H. (1934). Z. physiol. Chem. 224, 17. 89. Bergmann, M., Zervas, L., Schleich, H., and Leinert, F. (1934). Z . physiol. Chem. 212, 72. 90. Bergmann, M., Zervas, I,., and du Vigneaud, V. (1929). Ber. 62, 1905. 91. Bertho, A., and Maier, J. (1!432). Ann. 496, 113. 9la. Boothe, J. H. et al. (1948). J . A m . Chem. SOC.70, 1099. 92. Borsook, H., and Huffman, H. M. (1938). In Chemistry of Amino Acids and Proteins, p. 865. C. L. A. Schmidt., ed., C. C Thomas, Springfield, Ill. (Cf. also errata in Addendum, 1943.) 93. Braunstein, A. E. (1947). Advances i n Protein Chem. 3, 1. 94. Cahill, W. M., and Burton, I . F. (1940). J . Biol. Chem. 132, 161. 95. Carter, H. E. (1946). Organic Reactions 3, 198. 96. Carter, H. E., Handler, P., and Melville, D. B. (1938). J. Biol. Chem. 129, 359. 97. Carter, H. E., and Risser, W. C. (1941). J . Biol. Chem. 139, 225. 98. Carter, H. E., and Stevens, C. M. (1940). J . Biol. Chem. 133, 117. 99. Chibnall, A. C. (1942). Proc. Roy. Soc. London B131, 136. 100. Clarke, H. T., and Gillespie, 11. B. (1932). J . Am. Chem. SOC.64, 1964. 1OOa. Clarke, H. T., Johnson, J. It., and Robinson, R. (1949). The Chemistry of Penicillin. Princeton Univ. Press. 101. Cohn, E. J., and Edsall, J. T. (1943). Proteins, Amino Acids, and Peptides. Reinhold Pub. Corp., N. Y. 102. Committee on Medical Research and Medical Research Council (1945). Science 102, 627. 103. Consden, R., Gordon, A. H., and Martin, A. J. P. (1944). Biochem. J . 38,224. 103a. Cook, A. H., Harris, G., Heilbron, I. M., and Shaw, G . (1948). J. Chem. Soc. 1056. 104. Curtius, T. (1881). J. prakt. Chem. 24, 239. 105. Curtius, T. (1883). Ber. 16, 755. 106. Curtius, T. (1902). Ber. 36, 3226. 107. Curtius, T. (1904). Ber. 37, 1284. 108. Curtius, T., and Curtius, H. (1904). J . prakt. Chem. 70, 158. 109. Curtius, T., and Gumlich, 0. (1904). J . prakt. Chem. 70, 195. 110. Curtius, T., and Lambotte, E;. (1004). J . prakt. Chem. 70, 109. 111. Curtius, T., and Levy, L. (1904). J . prakt. Chem. 70, 89. 112. Curtius, T., and Sieber, W. (1922). Ber. 66, 1543. 112a. Dekker, C. A., and Fruton, .I. S. (1948). J . Biol. Chem. 173, 471. 112b. Dekker, C. A,, Taylor, S. P., and Fruton, J. S. (1949). J. Biol. Chem. 180, 155. 113. Dent, C. E. (1347). Biochem. J. 41, 240. 114. Dirr, K., and Spiith, H. (1935). 2. physiol. Chem. 237, 121. 115. Doherty, D. G., Tietzman, J. E., and Bergmann, M. (1943). J. Biol. Chem. 147, 617.
78
JOSEPH S. FRUTON
116. Dyer, E. (1941). J . Am. Chem. SOC.63,265. 117. Edlbacher, S., and Baur, H. (1941). 2.physiol. Chem. 270, 176. 118. Edlbacher, S., and Bonem, P. (1925). 2. physiol. Chem. 146, 69. 118a. Ehrensvhrd, G. (1947). Nature 169,500. 119. Ehrensviird, G., and Davidssohn, B. (1946). Arkiv. Kemi, hfineral. Geol. 24A, No. 6. 120. Enger, R. (1930). 2. physiol. Chem. 191,97. 121. Erlenmeyer, E., and Frtisttick, E. (1895). Ann. 284, 48. 122. Erlenmeyer, E., and Halsey, J. T. (1899). Ann. 307, 139. 123. Felix, K., and Dirr, K. (1928). 2. physiol. Chem. 176,29. 124. Fischer, E. (1902). Ber. 36, 1095. 125. Fischer, E. (1902). Ber. 36, 1103. 126. Fischer, E. (1902). Chem. Zt. 26, 939. 127. Fischer, E. (1903). Ber. 36,2094. 128. Fischer, E. (1903). Ber. 36,2982. 129. Fischer, E. (1904). Ber. 37,2486. 130. Fischer, E. (1904). Ber. 37,3062. 131. Fischer, E. (1905). Ann. 340, 123. 132. Fischer, E. (1905). Ber. 38,605. 133. Fischer, E. (1905). Ber. 38,2914. 134. Fischer, E. (1906). Ber. 39,453. 135. Fischer, E. (1906). Ber. 39,530. 136. Fischer, E. (1906). Ber. 39,2893. 137. Fischer, E. (1907). Ber. 40, 1501. 138. Fischer, E. (1907). Ber. 40, 1754. 139. Fischer, E. (1907). Ber. 40, 3704. 140. Fischer, E. (1908). Ber. 41,850. 141. Fischer, E. (1908). Ber. 41,2860. 142. Fischer, E. (1915). Ber. 48, 93. 143. Fischer, E., and Abderhalden, E. (1904). Ber. 37, 3071. 144. Fischer, E., and Bergell, P. (1903). Ber. 26, 2592. 145. Fischer, E., and Bergmann, M. (1913). Ann. 3Q8,96. 146. Fischer, E., and Blank, P. (1907). Ann. 364, 1. 147. Fischer, E., and Cone, L. H. (1908). Ann. 363, 107. 148. Fischer, E., and Fiedler, A. (1910). Ann. 376, 181. 149. Fischer, E., and Fourneau, E. (1901). Ber. 34,2868. 150. Fischer, E., and Gerngross, 0. (1909). Ber. 42, 1485. 151. Fischer, E., and Gluud, W. (1909). Ann. 369, 247. 152. Fischer, E., and Jacobs, W. A. (1906). Ber. 39,2942. 153. Fischer, E., and Kautzsch, K. (1905). Ber. 38, 2375. 154. Fischer, E., and Koelker, A. H. (1907). Ann. 364, 39. 155. Fischer, E., and Koenigs, E. (1904). Ber. 37,4585. 156. Fischer, E., and Koenigs, E. (1907). Ber. 40, 2048. 157. Fischer, E., Kropp, W., and Stahlschmidt, A. (1909). Ann. 366, 181. 158. Fischer, E., and Lipschitz, E. (1915). Ber. 48,360. 159. Fischer, E., and Luniak, A. (1909). Ber. 42, 4752. 160. Fischer, E., and Otto, E. (1903). Ber. 36,2106. 161. Fischer, E., and Raske, K. (1906). Ber. 39, 3981. 162. Fischer, E., and Reif, G. (1908). Ann. 366, 118. 163. Fischer, E., and Roesner, H. (1910). Ann. 376, 199.
SYNTHESIS OF PEPTIDES
79
164. Fischer, E., and Scheibler, €1. (1908). Ann. 363, 136. 165. Fischer, E., and Schoeller, W. (1907). Ann. 367, 1. 166. Fischer, E., and Schrauth, W. (1907). Ann. 364, 21. 167. Fischer, E., and Schulee, A. (1907). Ber. 40, 943. 168. Fischer, E., and Steingroever, J. (1909). Ann. 366, 167. 169. Fkcher, E., and Suzuki, U. (1904). Ber. 37, 2842. 170. Fischer, E., and Suzuki, U. (1904). Ber. 37, 3306. 171. Fischer, E., and Sueuki, U. (1904). Ber. 37,4575. 172. Fischer, E., and Suzuki, U. (1905). Ber. 38, 4173. 173. Fischer, H. 0. L., and Baer, E. (1932). Ber. 66, 337, 345. 174. Fishman, W. H., and Artom, C. (1944). Proc. SOC.Ezptl. Biol. Med. 66, 59. 175. Forster, M. O., and Fierz, H. E. (1908). J . Chem. SOC.93, 1862. 176. Frankel, M., and Katchalski, E. (1942). J . Am. Chem. SOC.64, 2264, 2268. 176a. Frankel, M., and Katchalski, E. (1944). Scientific Papers Presented to C. Weizmann, p. 24, Palestine Chemists Organisation Jerusalem. 177. Frankel, M., and Kuk, 5. (1930). Biochem. 2.226, 221. 178. Freudenberg, K., Durr, W., and Hochstetter, H. von. (1928). Ber. 61, 1735. 179. Freudenberg, K., Eichel, H., and Leutert, F. (1932). Ber. 66, 1183. 180. Freudenberg, K., and Miirkcrt, L. (1927). Ber. 60, 2447. 181. Fruton, J. S. (1942). J. Biol. Chem. 146, 463. 182. Fruton, J. S. (1946). In Currents in Biochemical Research, p. 123. D. E.
Green, ed., Intersci. Pub., N. Y. 183. Fruton, J. S. (1946). J. Biol. Chem. 166, 333. 184. Fruton, J. 5. (1946). J . Biol. Chem. 166,721. 185. Fruton, J. S. (1947). Yale J . B i d . Med. 19, 999. 185s. Fruton, J. S. (1948). J . Am. Chem. SOC.70, 1280. 186. Fruton, J. S. (Unpublished experiments.) 187. Fruton, J. S., and Bergmann, M. (1939). J . Biol. Chem. 127, 627. 188. Fruton, J. S., and Bergmann, M. (1942). J . Biol. Chem. 146, 253. 189. Fruton, J. S., and Bergmann, M. (1946). J . Biol. Chem. 166, 449. 190. Fruton, J. S., Irving, G. W., Jr., and Bergmann, M. (1940). J . Biol. Chem. 133, 703. 19Oa. Fruton, J. S., Smith, V. A., and Driscoll, P. E. (1948). J . Biol. Chem. 173, 457. 191. Fruton, J. S., Stein, W. H., and Bergmann, M. (1946). J . Org. Chem. 11, 559. 192. Fuchs, F. (1922). Ber. 66,2943. 192a. Griinacher, C., and Landolt, H. (1927). Helu. Chim. A d a 10, 799. 193. Grassmann, W., and Schneider, F. (1934). Biochem. 2. 273, 452. 194. Grassmann, W., Schoenebeck, O., and Auerbach, G. (1932). 2. physiol. Chem. 210, 1. 196. Greenstein, J. P. (1937). J. Biol. Chem. 121, 9. 196. Greenstein, J. P. (1938). J. Biol. Chem. 124, 255. 196a. Greenstein, J. P. (1938). J. Org. Chem. 2, 480. 197. Greenstein, J. P., and Klemperer, F. W. (1939). J . Biol. Chem. 128, 245. 198. Greenstein, J. P., and Lcuthardt, F. M. (1946). J . Natl. Cancer Znat. 6, 197. 198a. Greenstein, J. P., Price, V. E., a n d h u t h a r d t , F. M. (1948). J . Biol. Chem. 176, 953. 199. Grimaux, E. (1882). Bull. soc. chim. 38, 64. 200. Haas, P., and Hill, T. G. (1931). Biochem. J . 26, 1472. 201. Hanby, W. E., and Rydon, H. N. (1946). Biochem. J . 40,xxi.
80
JOSEPH 8. FRUTON
201a. Hanson, H. T., and Smith, E. L. (1948). J. B i d . Chem. 176, 833. 202. Harington, C. It., and Mead, T. H. (1935). Biochem. J . 29, 1602. 203. Harington, C. R., and Mead, T. H. (1936). Biochem. J. 30, 1598. 203s. Harington, C. R., and Overhoff, J. (1933). Biochem. J. 27, 338. 204. Harington, C. R., and Pitt Rivers, R. V. (1944). Biochem. J. 38, 417. 204a. Harris, J. I., and Work, T. S. (1948). Nature 161, 804. 205. Hartung, W. H. (1928). J. A m . Chem. SOC.60, 3310. 206. Havestadt, I,., and Fricke, R. (1924). Ber. 67, 2048. 206a. Hegediis, B. (1948). Helv. Chim. Acta 31, 737. 207. Helferich, B., and Griinert, H. (1940). Ann. 646, 178. 208. Herbst, R. M. (1936). J. A m . Chem. SOC.68,2239. 209. Herhst, R. M. (1944). Advances i n Enzymol. 4, 75. 210. Herbst, R. M., and Engel, L. L. (1934). J . B i d . Chem. 107, 505. 211. Herhst, lt. M., and Shemin, D. (1943). J . B i d . Chem. 147, 541. 212. Herriott, R. M. (1937). J. Gen. Physiol. 20, 335. 213. Mess, W. C., and Sullivan, M. X. (1941). J . A m . Chem. Soc. 63, 881. 214. Hofmann, K . , and Bergmann, M. (1939). J . Biol. Chem. 130, 81. 215. Hofmann, K., and Bergmann, M. (1940). J . Biol. Chem. 134, 225. 216. Hofmann, K., and Bergmann, M. (1941). J . B i d . Chem. 138, 243. 217. Hofmeister, F. (1902). Ergeb. Physiol. 1, 159. 218. Hopkins, F. G. (1929). J . Biol. Chem. 84, 269. 219. Hotchkiss, R. D. (1944). Advances in Enzymol. 4, 153. 220. Hunt, M., and du Vigneaud, V. (1938). J . Biol. Chem. 124, 669. 221. Hunt, M., and du Vigneaud, V. (1939). J . Biol. Chem. 127,43. 221a. Ing, H. It., and Manske, R. H. F. (1926). J. Chem. SOC.2348. 222. Ivanovics, G., and Briickner, V. (1937). z. ~mmunil(itsfor8ch.90, 304. 222a. Jacobsen, C. F. (1947). Compt. rend. trav. lab. Carlsberg. S k . chim. 26, NO. 1-2. 223. Kapfhammer, J., and Matthes, K. (1934). 2. physiol. Chem. 223, 43 223a. Katchalski, E., Grossfeld, I., and Frankel, M. (1947). J. A m . Chem. Soc. 69, 2564. 223b. Kidd, D. E., and King, F. E. (1948). Nature 162, 776. 224. Kogl, F., and Akkerman, A. M. (1946). Rec. trav. chim. 66, 216. 225. Kossel, A., and Kennaway, E. L. (1911). Z. phyeiol. Chem. 72, 489. 226.,,Kossel, A., and Staudt, W. (1927). Z . physiol. Chem. 170, 91. 227. Kraft, W. M., and Herhst, R. M. (1945). J . Org. Chem. 10, 483. 227a. Le Quesne, W. J., and Young, G. T. (1949). Nature 163, 604. 228. Leuchs, H. (1906). Ber. 39, 857. 229. Leuchs, N.,and Geiger, W. (1908). Ber. 41, 1721. 230. Leuchs, H., and Manasse, W. (1907). Ber. 40, 3243. 231. Levene, P. A., Bass, L. A., and Steiger, R. E. (1929). J . Biol. Chem. 81, 221. 232. Levene, P. A., and Beatty, W. A. (1906). Ber. 39, 2060. 233. Levene, P. A., and Pfaltz, M. H. (1926). J . Biol. Chem. 68, 277. 234. Levene, P. A., and Pfaltz, M. H. (1926). J . B i d . Chem. 70, 219. 235. Levene, P. A., Simms, H. S., and Pfaltz, M. H. (1924). J. Biol. Chem. 61, 445. 236. Levene, P. A., and Steiger, R. E. (1931). J . Biol. Chem. 93, 581. 237. Levene, P. A., Steiger, R. E., and Marker, R. E. (1931). J. Biol. Chem. 93, 605. 238. Lilienfeld, L. (1894). Arch. Physiol. 383, 555.
SYNTHESIS OF PEPTIDES
81
239. Linderstrgm-Lang, K., and Jacobsen, C. F. (1941). J . Biol. Chem. 137, 443. 240. b r i n g , H. S., and du Vigneaud, V. (1935). J . Biol. Chem. 111, 385. 240a. MacAllister, R. V., Harmon, K. M., and Niemann, C. (1949). J . Biol. Chem. 177, 767. 241. Martell, A. E., and Herbst, R. M. (1941). J . Org. Chem. 6 , 878. 242. Martin, A. J. P., and Synge, It. L. M. (1945). Advances in Protein Chem. 2 , 1. 243. Maachmann, E. (1940). Biochem. 2. 307, 1. 244. Max, J. (1909). Ann. 369, 276. 245. Melville, J. (1935). Biochem. J. 29, 179. 246. Meyer, H. (1901). ibfonat8h. 22, 11. 247. Migliardi, C. (1941). Arch. 8 1 5 . biol. Italy 27, 327. 248. Miller, G. L., Behrens, 0. K., and du Vigneaud, V. (1941). J . Biol. Chem. 140, 411. 249. Mohr, E., and Strohschein, F. (1909). B ~ Y42, . 2521. 250. Moore, S., Stein, W. H., and Fruton, J. S. (1946). J . Org. Chem. 11, 675. 25%. Mowat, J. H. el al. (1948). J . Am. Chem. Soc. 70, 1096. 251. Neher, R., Spielman, M., Werner, L. H., Wettstein, A., and Miescher, K. (1946). Helv. Chim. Acta 19, 1874. 252. Neher, R.,Wettstein, A., and Miescher, K. (1946). Helv. Chim. Acta 29, 1815. 252a. Neuberger, A. (1948). Advance8 in Protein Chem. 4, 330. 253. Osborne, T. B., and Clapp, 5.H. (1907). Am. J . Physiol. 18, 123. 254. Pacsu, E., and Wilson, E. J. (1942). J . Org. Chem. 7 , 117, 126. 255. Pirie, N. qV. (1931). Biochem. J . 26, 614. 255a. Prelog, V., and Wieland, P. (1946). Helv. Chim. Acta 29, 1128. 255b. Price, V. E., and Greenstein, J. P. (1947). Arch. Biochem. 14, 249. 256. Ramsay, H. (1908). Ber. 41,4385. 257. Ratner, S., and Clarke, H. T. (1937). J . Am. Chem. SOC.69, 203. 258. Rinke, H. (1934). Thesis, Leipzig. 259. Rothemund, K. W., and Zetzsche, F. (1921). Ber. 64, 2038. 260. Ruggli, P., Ratti, R., and Henze, E. (1929). Helv. Chim. Acta 12, 361. 261. Sanni4, C. (1942). Bull. 8oc. chim. 9, 487. 262. Schaal, E. (1871). Ann. 167, 24. 263. Schiff, H. (1899). Ann. 307, 231. 264. Schneider, F. (1935). Thesis, Munich. 265. Schneider, F. (1937). Biochem. 2. 291,329. 266. Schonheimer, R. (1926). 2.physiol. Chem. 164, 203. 267. Schott, H. F., Larkin, J. B., Rockland, L. B., and Dunn, M.S. (1947). J . Org. Chem. 12, 490. 268. Schroeter, G. (1908). 2. angew. Chem. 41, 2115. 269. Schtitte, E. (1943). 2. physiol. Chem. 279, 52. 270. Schutzenberger, P. (1891). Compt. rend. 112, 198. 27Oa. Sheehan, J. C., and Frank, V. S. (1949). J . Am. Chem. Soc. 71, 1856. 271. Shemin, D., and Herbst, R. M. (1938). J . Am. Chem. SOC.60, 1951. 272. Sifferd, R. H., and du Vigneaud, V. (1935). J . Biol. Chem. 108, 753. 273. Sigmund, F., and Wessely, F. (1926). 2. physiol. Chem. 167, 91. 273s. Simmonds, S., and Fruton, J. S. (1948). J . Biol. Chem. 174, 705. 274. Simmonds, S., Tatum, E. L., and Fruton, J. S. (1947). J . Biol. Chem. 169,91. 275. Smith, C. S., and Brown, A. E. (1941). J . A m . Chem. Soc. 63, 2605. 275s. Smith, E. L. (1948). J . Biol. Chem. 176, 39. 275b. Smith, E. L. (1948). J . Biol. Chem. 176, 21.
82
JOSEPH 8. FRUTON
276. Smith, E. L., and Bergmann, M. (1944). J . Biol. Chem. 163, 627. 276s. Smith, E. L., and Slonim, N. B. (1948). J . Biol. Chem. 176, 835. 277. Stahmann, M. A., Fruton, J. S., and Bergmann, M. (1946). J . Biol. Chem. 164,753. 278. Stein, W. H., Moore, S., and Bergmann, M. (1944). J . Biol. Chem. 164, 191. 279. Stein, W. H., Moore, S., Stamm, G., Chou, C., and Bergmann, M. (1942). J . Biol. Chem. 143, 121. 279s. Stekol, J. A. (1941). J. Biol. Chem. 140, 827. 280. SubbaRow, Y. el al. (1946). Ann. N . Y . Acad. Sci. 48, 255. 281. Synge, R. L. M. (1943). Chem. Revs. 82, 135. 282. Synge, R. L. M. (1945). Biochem. J. 39, 363. 282a. Synge, R. L. M. (1948). Biochem. J. 42, 99. 283. Thierfelder, H., and von Cramm, E. (1919). Z . phyuiol. Chem. 106, 58. 284. Vickery, H. B. (1947). J . Biol. Chem. 169, 237. 285. Vickery, H. B., and Pucher, G. W. (1943). J . Biol. Chem. 160, 197. 286. du Vigneaud, V., Audrieth, L. F., and Loring, H. S. (1930). J . Am. Chem. Soc. 62,4500. 287. du Vigneaud, V., and Behrens, 0. K. (1937). J . Biol. Chem. 117, 27. 288. du Vigneaud, V., and Hunt, M. (1938). J . Biol. Chem. 126, 269. 289. du Vigneaud, V., Loring, H. S., and Miller, G. L. (1937). J . Biol. Chem. 118, 391. 290. du Vigneaud, V., and Meyer, C. E. (1932). J . B i d . Chem. 98, 295. 291. d u Vigneaud, V., and Miller, G. L. (1936). J . Biol. Chem. 116, 469. 292. du Vigneaud, V., Patterson, W. I., and Hunt, M. (1938). J . Biol. Chem. 126, 217. 293. Walden, P. (1896). Ber. 29, 133. 294. Werner, L. H., Wettstein, A., and Mieacher, K. (1947). Helv. Chim. A d a SO, 432. 295. Wessely, F. (1925). 2. phyeid. Chem. 146, 72. 296. WMely, F., Kemm, E., and Mayer, J. (1929). Z. phyuiol. Chem. 180, 64. 297. Wessely, F., and Sigmund, F. (1926). Z. phyaiol. Chem. 169, 102. 297a. White, J. (1933). J. Biol. Chem. 102, 249. 298. Willsttitter, R., and Waldschmidt-Leita, E. (1921). Ber. 64, 128. 299. Womack, M., and Rose, W. C. (1946). J. B i d . Chem. 162, 735. 300. Wood, J. L., and d u Vigneaud, V. (1939). J . Biol. Chem. 130, 109. 301. Woodward, R. B., and Schramm, C. H. (1947). J . Am. Chem. SOC.69, 1551. 302. Woolley, D. W. (1945). J. Biol. Chem. 169, 753. 303. Woolley, D. W. (1948). J . Biol. Chem. 172, 71. 304. Yudkin, W. H., and Fruton, J. S. (1947). J . Biol. Chem. 169, 521. 305. Zervas, L., and Bergmann, M. (1928). Em. 61, 1195.
Amino Acid Composition of Purified Proteins
BY G. R. TRISTRAM Sir William Dunn Institute of Biochemistry, University of Cambridge, Cambridge, England
CONTENTS Page ...................... I. Introduction.. .............................. 84 11. Purposes of Amino Acid Analysis.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 ...................... [II. Methods of AnalysiR.. ....................... 86 1. Some General Analytical Determinations.. . . . . . . . . . . . . . . . . . . . . . . . 87 87 a. Total Protein Nitrogen., . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . b. Moisture Content.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 c. Amide Nitrogen.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 2. Methods of Amino Acid Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 a. Periodate Oxidation of Hydroxyamino, Acids.. . . . . . . . . . . . . . . . 88 b. Electrodialysis of Basic Amino Acids. ....................... 89 go c. Enzymatic Methods.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . d. Microbiological Assay.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 e. Isotope Dilution.. ......... I. Partition Chroma g. Sulfur-containing Acids. . . . . . . . . . . . . . . . . h. Dicarboxylic Acid i. Additional Methods. .............................. 3. Determination of Free IV. Establishment of Accuracy and Specificity of Methods of Analysis.. , 1. Purity of Amino Acids a. Racemic Acids .......................................... 100 b. Optically Active Acids.. ....................... . . loo 2. Choice of Protein8 for Testing Methods of Amino Acid Analysis. . . . 100 3. Criteria for Establishment of Accuracy and Specificity.. . . . . . . . . . . . 102 a . Primary Standard Methods ................................ 102 b. Routine Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V. Comparison of Analytical Mcthods.. . . . . . . . . . . . 1. Reporting Analytical D a t a . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. The Data of Tables IV-XXI.. . . . . . . . . . . . . . . . . . . . . 107 a. Hydroxyproline.. . . . . . . . . . . . . . . . . . . . b. Glycine.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 c. Proline ...................... . . . . . . . . . 110 d. Methionine ....... . . . . . . . . . . . . . . 111 e. Cystine-Cysteine. . ............................ 111 f . Dicarboxylic Acids ............................ 114
83
84
Q.
R. TRIBTRAM
g. Serine and Threonine..
Page . . . . . . . . . . . . . . . . . . 115
h. Nonpolar Amino Acids.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i . Tryptophan and Tyrosine., . j.Basic Amino Acids.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Degree of Accuracy of Various Methods.. . . . . . . . . . . . . . . . . . . . . . . . 4. Modification and Destruction of Amino Acids during Hydrolysis., , , VI. Structure of Proteins as Revealed by Amino Acid Analysis.. . . . . . . . . . . . . 1. Definition of Purity.. . . . . . . . . . . . . .......... 2. Determination of Purity of Soluble 3. Purity of Protein Preparations Used in Present-Day Analytical Work 4. Reporting Total Amino Acid Analyses.. . . . . . . . . . . . . . . . . . . . . . . . . . 5. Molecular Weights. . . . . . . . . . . . . . . . . . . ................ VII. Amino Acid Composition of Certain Proteins ................ 1. Salmine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Insulin.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Edestin . . . . . . . . . . . . . . . . . . .
115 121 121 123 125 125 126 127 128 129 129 130
.........
6. Ovalbumin. ...
VIII. General Discussion. . . . . . . . . . . . . . . . 3. Ionic Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
144
IX. Conclusion and Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
148
I. INTRODUCTION Amino acid analysis has been a subject of research for many decades, and its numerous and formidable difficulties are attested by the vast literature on the subject (cf. Vickery, 1946). Progress was delayed for many years because methods of analysis were, in the main, based upon the gravimetric principles of elementary analysis, and analysts only slowly realized the difficulties which attend the quantitative isolation of a single constituent or any of its derivatives from biological materials. I n recent years the development of new methods, micro in scale, based upon many novel principles (e.g., isotope dilution, microbiological assay, partition chromatography) has revolutionized amino acid analysis, which, it is now reasonable to hope, will shortly become a routine procedure, and that valuable information will be obtained upon the amino acid stoichiometry of proteins. Modern methods of amino acid analysis have been the subjects of several critical reviews (General : Martin and Synge, 1945; microbiological
AMINO ACID COMPOSITION OF PURIFIED PROTEINS
85
assay: Snell, 1945, 1946; isotope dilution: Shemin and Foster, 1946; chromatography: Cannan, 1946; enzymatic methods: Archibald, 1946). For this reason the discussion of analytical procedures will be limited to consideration of their advantages and weaknesses and to a discussion of the precautions necessary to ensure that, in the investigation of amino acid stoichiometry, the highest degree of accuracy has been achieved. I n view of the confusion which has been caused by ( a ) the use of proteins of indeterminate or doubtful purity for the purpose of testing methods of analysis, and ( b ) the lack of uniformity in technique, it is becoming increasingly apparent th at analytical procedure should conform t o certain criteria and t,liat the use of such proteins as casein for the testing of analytical methods should be avoided (cj. Chibnall, 1946; Dunn and Rockland, 1946). Furthermore, in order to facilitate the comparison of the data of various workbrs some standard system of reporting analytical data must be adopted (cj. Stein, 1946; Chibnall, 1946; Brand and Edsall, 1947). Although modern developments concerning the inhomogeneity of proteins which were hitherto regarded as examples of a single protein molecule may subsequently render invalid many of the points raised, the stoichiometric and structural interpretations which may be obtained from amino acid composition will be discussed for a restricted number of proteins. A more general list of analyses will be appended with little or no discussion.
11. PURPOSES OF AMINOACID ANALYSIS Amino acid analyses reported in the literature may be classified into three sections: ( I ) Those carried out for the assessment of the nutritive value of a protein and/or foodstuff. In this case it is common practice to limit analysis to the essential amino acids, though it is by no means certain that only the essential amino acids listed by Rose play any significant part in protein metabolism. This limitation is also unfortunate from the more fundamental viewpoint since it restricts the analysis of the so-called nonessential amino acids to the efforts of a small number of workers, and reduces the scope and value of the comparative method which is so valuable at the present stage of amino acid chemistry ( c j . Chibnall, 1946; Tristram, 1946). (2) Those carried out for the purpose of controlling protein fractionation, in which case comparative rather than absolute values are essential. Analysis may be limited, in the first instance at any rate, to the estimation of a few amino acids, particularly those with specific groupings (e.g., tyrosine, arginine, and cystine).
86
0. R. TRISTRAM
(3) Finally the amino acid composition of a protein molecule has, with the development of comprehensive and accurate methods of analysis, become of more fundamental value, and protein analysis is now undertaken .with the object of investigating protein stoichiometry and of correlating physical and chemical properties with amino acid composition. Certain of the fundamental aims of analysis are discussed below. Amino Acid Distribution and Determination of Minimum Molecular Weight. From a complete amino acid analysis of a protein of known molecular weight it is possible to calculate the number of residues of each amino acid within the molecule. As with elementary analysis, it is also possible to calculate the minimum molecular weight from the amounts of individual amino acids in the protein. However, the values so obtained bear a simple relationship with the known molecular weight only when the number of residues of the particular amino acid approaches unity. I n all cases, however, the relationship between the observed molecular weight (M,) and minimum molecular weight (M,) is that M. = M , X n, where n is the number of residues of an amino acid in the protein molecule. Sequence o j Amino Acids in Protein Molecule. Although amino acid data cannot reveal the sequence of amino acids within the protein molecule, a precise knowledge of amino acid composition is a vital preliminary to any studies upon the detailed architecture of protein molecules’ Distribution of Specific Groupings. (See Table 111.) The proportion of any individual amino acid in a protein is at present, almost without exception, of little significance. An accurate knowledge of the distribution of certain specific groups ( e . g . , ionic, polar but uncharged, and nonpolar groups) is of great importance in the interpretation of chemical, physical, and biological properties, although thesb properties, particularly the last-named, must be influenced by the relative distribution of the various groups in the protein molecule. Establishment o j Protein Purity. Precise amino acid analysis must ultimately take its place as one of the criteria for the establishment of protein purity. In this respect analysis might serve as a check upon the determination of purity by phase rule solubility. If a protein were pure, the amino acid composition of the protein in the solid phase and of that in the supernatant solution with which the solid is in equilibrium should be identical.
111. METHODSOF ANALYSIS While there is little doubt that the newer methods can attain a moderately high precision in capable hands, there appears t o be a lack of understanding of the factors which control and limit such precision. As is perfectly obvious, such limitations (ignoring the human element) deter-
AMINO ACID COMPOSITION OF PURIFIED PROTEINS
87
mine to what extent the results of analysis represent the amino acid composition of protein hydrolyzates. Several of the more recent analytical procedures will be discussed from this critical viewpoint. Of the modern methods the author has had personal experience with partition chromatography, electrodialysis, specific decarboxylases, and periodate oxidation. Consideration of these methods may therefore be more intimate than it can be for those methods (e.g., microbiological assay and isotope dilution) in which personal practical experience is lacking. 1. Some General Analytical Determinations
a. Total Protein Nitrogen. While, strictly speaking, the determination of total protein nitrogen cannot be regarded as amino acid analysis, its accurate estimation is the foundation upon which depends the calculation of analytical data, irrespective of the system adopted for the expression of the results. The accurate determination of total nitrogen has long been a subject of investigation and controversy, and is all too frequently overlooked. It would greatly assist the comparison of the results of analyses (particularly when these are partial analyses) if the total nitrogen of the proteins examined were reported and determined, preferably by some standard procedure such as that suggested by Chibnall et al. (1943a). These workers showed that the micro-Kjeldahl method was capable of a higher precision than generally supposed, and were of the opinion that, with reasonable care, duplicates agreeing to 0.1 % could be obtained on glutamic acid and other amino acids. It is now clear that the accuracy of estimation of total nitrogen is closely related to the amino acid composition and, in order to ensure the complete digestion of proteins rich in lysine etc., the most rigorous conditions of heating are required. Although theoretical recoveries could be obtained in 4 hours by the use of a mercury catalyst (Chibnall et al., 1943a) such a catalyst was not favored by these workers because it involved the use of sodium thiosulfate. b. Moisture Content. The determination of the dry weight of proteins has long been a source of difficulty since the hydrophilic nature of proteins is not always recognized and there is some doubt of the extent to which water is an integral part of the protein molecule. Brand and Kassell (1942b) showed, with the aid of water containing 01*,that the moisture content may be accurately determined by drying a t 110" in vacuo or in air. The standard procedure of Chibnall and coworkers has been to dry at 105"in air for 12 hours and then to cool for 1 hour in a vacuum desiccator after which the tightly stoppered sample is allowed to equilibrate in a small desiccator kept permanently in the balance case.
88
Q. R. TRISTRAM
c. Amide Nitrogen. (See Table IV.) The determination of amide nitrogen is a t present carried out by one of three methods: ( a ) the acid hydrolysis technique of Gordon et al. (1941) (see Rees, 1946, for the protocols of this procedure), ( b ) the two hour hydrolysis with 5% hydrochloric acid (Shore et al., 1936), and (c) the alkaline hydrolysis procedure of Warner and Cannan (1942) extensively used by Brand and coworkers. The reviewer knows of no critical work on the relative merits of these methods although such work seems to be essential in view of the wide divergence in the values for the amide nitrogen of insulin (Table IV) reported by various workers. It is the experience of Rees that the first two methods invariably give the same value for amide nitrogen. It should be borne in mind that since the work of Warner and Cannan (1912) it has been shown that there are in proteins large amounts of hydroxy acids which are known to be unstable to alkali (Nicolet and Shinn, 1941~). 2. Methods of Amino Acid Analysis
a. Periodate Oxidation of Hydroxyamino Acids. The method of Nicolet and Shinn (1941a)b) has been improved by Rees (1946)) whose critical work suggests that in order to achieve the highest possible accuracy the following precautions should be observed and possible interference of certain amino acids always borne in mind. (1) Although Martin and Synge (1941) found no evidence for the existence of higher homologs of threonine in castor seed globulin, edestin, gluten, arachin, and zein, the existence of such homologs cannot be excluded. Since the estimation of acetaldehyde by titration is not specific it is necessary to confirm the absence of any homolog by the preparation of the dimedone derivative of the aldehyde, obtained by periodate oxidation, from each protein investigated. (6) The formaldehyde obtained by periodate oxidation can only be ascribed to serine if the absence of hydroxylysine has been confirmed. In the few proteins in which hydroxylysine is known to be present ( e . g . , gelatin and collagen) the serine must carry the appropriate correction (cf. Macphervon; 1946 Rees 1946). The interference of sugars was also discussed by Rees (1946) and he pointed out the difficulty of giving more than an approximate value for the serine content of ovalbumin. (3) Owing to variations in the recovery of the aldehydes it is necessary to include a control experiment (using pure serine and threonine) in each series of experiments. (4) Interference by certain substances influences the reliability of this method. (a) Tryptophan: Although formaldehyde condenses readily
AMINO ACID COMPOSITION OF PURIFIED PROTEINS
89
with tryptophan this amino acid is largely eliminated by the acid hydrolysis of proteins. Interference is therefore largely potential, but should be borne in mind if methods of hydrolysis are used which are likely to have a sparing action upon tryptophan. (The presence of heavy metals in hydrochloric acid is thought to increase humin formation, LinderstrZm-Lang, unpublished work.) ( b ) Histidine : Under certain conditions formaldehyde condenses with histidine (Neuberger, 1944). Although recovery experiments by Rees (1946) on horse globin (histidine 8.7%, serine 5.7%) show no abnormal loss, the possible interference of this amino acid cannot be ignored when investigating proteins rich in histidine. Recent analyses of horse myoglobin by Rees (unpublished) have shown, for example, that the recovery of formaldehyde is influenced by the presence of considerable amounts of histidine. (Myoglobin contains 8.8% of this amino acid.) ( c ) Glycine: This amino acid interferes though only when present in large amount ( e . g . , 50% of the total nitrogen). It should therefore not be used to replace alanine in the standard recovery experiments. ( d ) Nicolet (1943) has suggested th at cystine might give rise to high values for periodate ammonia but analysis of mixtures of amino acids simulating insulin (Rees, 1946) shows that, under normal circumstances, no appreciable error is involved. There is, however, a suggestion that the presence of relatively large amounts of cystine and/or tyrosine during hydrolysis has reduced the recovery of serine as formaldehyde by 2-4%. ( e ) Carter and Neville (1947) suggested th a t hydroxyproline yielded formaldehyde on treatment with periodate but this has now been shown t o be incorrect (Rees, unpublished experiments; Carter and Loo 1948). Rees (1946) suggests the application of standard correction factors of 100/94.7 and 100/89.5 to allow for the decomposition, during hydrolysis, of threonine and serine, respectively. b. Electrodialysis of Basic Amino Acids. This method was applied to the analyRis of protein hydrolyzates by Albanese (1940) and Theorell and Wkeson (1942). In a rec,ent article Macpherson (1946) indicated that the procedure of these workers was not fully satisfactory due presumably to simple diffusion and to the incomplete removal of monoamino acids. He showed that (a) redialysis of the cathode solution at p H 5.8 was necessary to remove monoamino acids, particularly tyrosine, which interferes with the colorimetric estimation of histidine, ( b ) ammonia should be removed by concentration of the final catholyte in vacuo, and ( c ) the complete transport of histidine into the cathode compartment was ensured by efficient stirring. Under these conditions the colorimetric estimations of arginine (modified Sakaguchi) and histidine (modified Pauli) are extremely precise. Macpherson reported that the estimation of arginine is free from interference unless the histidine: arginine ratio
90
0.
R. TRIBTRAM
exceeds 3 :1, and that tyrosine is the only possible source of error in the estimation of histidine. c. Enzymatic Methods. Concurrent with the development of other methods in the Biochemical Laboratory, Cambridge, Gale (1945, 1946) isolated, and successfully applied to the estimation of the amino acids concerned, specific decarboxylases for histidine, arginine, glutamic acid , tyrosine, and ornithine. The procedure, being gasometric, is capable of precision control and does not demand any elaborate bacteriological equipment other than means of growing the various organisms. The source and stability of the various enzymes are shown in Table I. TABLE I Source and Ueeful Life of Preparations of Specific Decarbozylases Amino acid bargininc blysine L-histidine tglutamic acid ktyrosine Lornithine
8
Source of enzyme" Escherichia coli 702( Bacterium cadaveris 6578 Clostridium welchii (BW21)6785 C. welchii (SRl2) 6784 Streptococcus faecalis 6783 C. septicurn (I' 111) 547
Form in which enzyme used
Useful life
Acetone powder 3-4 weeks Acetone powder 3-4 weeks
Method of preservation Ice chest Ice chest
Acetone powder 2-3 months Desiccator Fresh cells
3 days
Acetone powder 3-4 weeks Fresh cells
HP. 25°C. Desiccator
24 hours
Exact strains of the Organisms must be used (Gale, personal communication)
The estimation of lysine, in the basic amino acid solution resulting from electrodialysis, by the specific decarboxylase has now been incorporated into the standard procedure of Macpherson (personal communication). The enzymatic method of Gale has a potential source of error, common to all biological methods, in that it estimates the L isomers specifically. The presence of D isomers in protein hydrolyzates is by no means clearly established, but all values obtained by biological methods (enzymatic or microbial assay) must a t present be regarded as minimal only. This limitation also applies to isotope dilution as it is a t present being applied (Rittenberg and Foster, 1940; Foster, 1945; Shemin and Foster, 1946). d. Microbiological Assay. This method depends upon the observation that, in the presence of limiting amounts of an essential nutrilite, the amount of growth obtained is a function of the concentration of that essential nutrilite and increaSes to a maximum as the concentration of the nutrilite is increased (Snell, 1946). The method has in the past few years
AMINO ACID COMPOBITION O F PURIFIED PROTEINS
91
found extensive use. Many conflicting results have been obtained and the reasons for the disagreement and the conditions under which concordant and apparently accurate values may be obtained are only now being fully realized. They have been discussed in detail in the excellent articles by Snell (1945, 1946) and a brief outline will serve the purpose of this review. It is known that the acid-producing organisms require certain vitamins and amino acids for growth, and that the actual requirements of vitamins and amino acids are closely interdependent. For example, Stokes and Gunness (1945s) found that Lactobacillus delbriickii, L. casei, and L. arabinosus required lysine, threonine, and alanine if grown on a medium containing pyridoxine, whereas if an equal amount of pyridoxamine or pyridoxal were used these amino acids were not required ( b e l l and Guirard, 1943). This is important because pyridoxine is partially converted to pyridoxamine or a substance of similar activity on autoclaving with amino acids (Snell and Rannefeld, 1945). Another example of the method which may be cited is the finding of Shankman (1943) that with a certain medium phenylalanine and tyrosine were essential for the growth of L. arabinosus, but on increasing the concentration of the amino acids in the basal medium these amino acids became accessory only. Snell (1945) summarized the importance of a complete knowledge of the nutritive requirements of the various micro-organisms: “An ideal medium for use in the microbiological assay of any essential nutrilite would be one which contained every other substance either essential for growth or stimulatory to the growth of the assay organism, but was completely free of the nutrilite concerned. This ideal is approached only more or less closely by assay methods now in use. Intimate knowledge of the nutritional requirements of the test organkm is thus fundamental to the development of an assay method, and underlies all future improvement in these methods.”
Pending the development of such metabolic studies all analyses in which the growth of micro-organisms is regarded as a function of the dosage of a particular amino acid must be treated with reserve and regarded as first approximations only. It has been variously suggested (Snell, 1945, 1946; Dunn and Rockland, 1946) that certain criteria should be fulfilled in order to guarantee that results are both reproducible and accurate. The following are the more important practical recommendations: (1) The method should be efficient in the hands of a capable operator (Snell, 1945). (2) The aminoacid must not be synthesized by the organism under the conditions of the experiment (Dunn and Rockland, 1946).
(3) The micro-organism should be cultured under such conditions that biological
92
a.
R. TRISTRAM
variation is prevented or minimized (Dunn and h c k l a n d , 1946). (Hadley, 1927, 1937, stated t h a t incitants to variation include temperature, oxygen tension, food substances, antiseptics, and metabolic products.) (4) A precise procedural discipline should be adopted (Dunn and Rockland, 1946). (5) The composition of the basal medium should be such t h a t the influence of any substance (whether it is a stimulant or an inhibitor) other than the test amino acid is almost completely nullified (Dunn and Rockland, 1946). (6) The results obtained should be comparable to those obtained by other methods. (7) Control experiments, using test mixtures and the addition of the amino acid to the hydrolyzate, should give quantitative recoveries. (8) Repeated assay should give concordant results. (9) There should be agreement a t various levels of assay. (10) There should be agreement between the values obtained with various organisms, although such a criterion may not be valid (see p. 104). (11) The amino acid should be the only substance which supports growth in deficient medium (Snell, 1945, 1946). (12) There is, of course, one final criterion that, in view of the many possible variants discuesed above, any organism should first have been the subject of extensive trial.
We may conclude, in general, that : Although microbiological assay is yielding many results, indeed the bulk of the values in the literature, which compare favorably with those obtained by other methods (e.g., isotope dilution) it should be borne in mind that the method is still somewhat empirical and the attainment of accuracy requires the exercise of a strict procedural discipline, and the use of amino acids of a high purity (cf. Smith and Greene, 1948). Many of the earlier values, obtained before the requirements and the variations in requirements were understood, are of doubtful accuracy. Schweigert and Snell (1947) have summarized the advantages and disadvantages of microbiological methods. The advantages are (1 ) They are highly specific and highly sensitive (Snell, 1945). (2) They eliminate many of the laborious separations previously necessary in protein analysis. (3) They are applicable in a number of instances to amino acids for which no other method suitable for routine application is available. (4) They demand little expensive equipment and are similar for all amino acids. (This view does not take into account the obvious fact that microbiological assay can only attain full efficiency when the equipment, technique, and the general background of a bacteriological laboratory is available.) (5) They are admirably adapted to routine work; several estimations are little more trouble than a singlc one. Their present disadvantages include : (1) Unfamiliarity to most chemists of an analytical technique which requires control of a multitude of factors, some of which are imperfectly understood. (2) Danger that
AMINO ACID COMPOSITION OF PURIFIED PROTEINS
93
the lack of activity of the unnatural isomers may cause these to be overlooked (Chibnall, 1946). (3) Possibility that unrecognized factors may combine to alter the specificity of the test organism in a given instance. (This risk is always present.) (4) Absence, so far as is as yet known, of any organism which requires certain of the amino acids, e.g., hydroxyproline. (5) (Added by author.) Inability to indicate the presence of amino acids as yet undiscovered, but which if present may modify the requirements and response of the organism(s) to certain of the essential acids. (6) The degree of utilization of D-amino acids depends upon the form of certain vitamins present (e.g., pyridoxamine or pyridoxine-see Lyman and Kuiken, 1948). This may invalidate the use of Dbamino acids either in media or standard solutions. Dunn, Camien, Shankman, and Block (1947) have claimed that certain organisms may be used for the estimation of D-amino acids, but since the view has long been held that only L-amino acids are utilized by living matter, it is clear that further systematic study is necessary so that the exact role of the unnatural isomers in the metabolism of micro-organisms may be established. Isotope Dilution. It is now generally accepted (Brand and Edsall, 1947) that isotope dilution is the one method of amino acid analysis which is capable of giving approximately theoretical values, and therefore is more likely to be used to provide a few standard values for comparative purposes, rather than for general amino acid analysis. The great advantage of the method is that ( a ) the isolation of a pure specimen of amino acid or derivative need not be quantitative, and ( b ) i t is applicable to any substance which can be labeled and isolated. Shemin and Foster (1946) have outlined the procedure and have indicated that the errors, which amount to about 1-2%, and which are calculable, may be classified into two groups: (1) Variations and uncertainty in the purity of the compound added. The purity is relatively easily controlled since the amino acids are almost invariably synthetic. (2) The purity of the compound isolated. The effects of impurities are threefold. First, nonnitrogenous contaminants introduce no error. Second, 1 % of nonisotopic atoms introduce an error of about 1.25% when the Co/C = 5 (where Co is the concentration of the marked element (NL6) in the added compound and C the concentration in the isolated material). With increasing values of Co/C the error decreases to about 1%. Third, the presence of the D isomer containing N16 will introduce a large error. Thus when the isotopic DL mixture is added (Co/C = lo), if the L isomer is contaminated with 1 % of the D , an error of 9% will be introduced. The exponents of this technique however claim that the L isomer can be freed from the D component by means of fractional crystallization (cf. Keston et. al., 1949).
94
Q. R. TRISTRAM
The method appears to possess every advantage, but so far it has not found extensive use and it is therefore difficult to assess its merits from the practical standpoint. Certain disadvantages, probably minor in character, exist a t present: (1) The criteria for judging the purity of the synthetic amino acids (carbon, hydrogen, and nitrogen analyses) are probably insufficient and could probably be improved by the use of some chromatographic test. (2) Little or no attention has been paid to the possible existence of D-amino acids, although, potentially, the method should be capable of estimating these isomers. (3) As a t present practised (Foster, 1945; Shemin and Foster, 1946) the method requires fairly large amounts of amino acids. (4) The synthesis of many of the amino acids is laborious. (5) At present samples of amino acid are isolated qualitatively by methods involving partial crystallization and solubility differences. I t is unfortunate that several of the newer and more physical methods have not been adapted to the separation of the amino acids (e.g., partition chromatography, electrodialysis, and the displacement technique of Tiselius, (1942) since it is evident that their use would remove the doubts created by the use of isolation procedures, and would make the method applicable to all the amino acids rather than to those few acids which crystallize easily or which form crystalline derivatives. The use of isolation procedures has been shown t o be open to errors, which may be serious in the measurement of isotopic dilution. Keston, Udenfriend, and Cannan (1949) have shown that the coprecipitation of a homologous series, such as the amino acids, is a common phenomenon, and great care is necessary in the establishment of constancy of isotope content. j . Partition Chromatography of Acetylated Amino Acids. This method which was introduced by Martin and Synge (1941), was probably the first which made possible the serial microestimation of certain of the monoamino acids, and has become the basis of many modern methods and procedures (e.g., partition chromatography on paper and starch, the partition chromatography of penicillins and streptomycin, etc.). It was further investigated by the author (Tristram. 1946) and criteria were laid down for its most efficient application. Advantages of the method are: (1) The method estimates the amino acids irrespective of their isomeric form and in this respect is superior to biological methods. (2) The fractions obtained are available for further investigation, and the method should be of great value in the further development of isotope dilution. (3) It is particularly suited to the separation of substances which are alike in their reactive groups but which differ in their nonpolar characteristics (Cannan, 1946). Disadvantages of the method are: (1) At present the values obtained
AMINO ACID COMPOSITION OF PURIFIED PROTEINS
95
tend to show, in the initial experiments at any rate, somewhat wide variations. In the writer's opinion this is, in the main, due to differences in the degree of acetylation, probably due to anhydride formation during the process of acetylation as a result of a pH fall to about 6 . Recent experiments by Macpherson, Rees, and Tristram (unpublished work) indicate that most of the uncertainties disappear when the basic and dicarboxylic amino acids are removed prior to the acetylation of the monoamino acid fraction. (2) The recovery of acetylamino acids is occasionally influenced by the amino acid mixture and preliminary experiments are usually necessary to decide upon the choice of solvents. I t is unfortunate that this method has not been more widely used because it is intrinsically of great value and, being based upon a physical property (i.e. partition coefficient), it should be possible to calculate the errors of estimation once the properties of the gel and the process of acetylation are fully understood. A more recent developmeat of chromatography in which starch is the carrier of the stationary phase has been developed by Moore and Stein (1948) (cf. earlier work by Synge, 1944 and Elsden and Synge, 1944). According to Brand and Edsall (1947) this method promises to be of great value. In a personal communication, Stein has reported certain analyses on the proteins B-lactoglobulin and bovine serum albumin. These are in fair agreement with the values obtained by other methods except that the isoleucine contents (the leucine isomers are capable of being separated on starch) are somewhat lower than those reported by microbiological assay (see Table 11) ( c j . Smith and Greene, 1947a, 1948a). The details of starch chromatography have now been published (Stein, Moore, 1948, 1949, J . Biol. Chem., 176,337, 367; 178, 79) and it is clear that these workers have developed a method of the first importance. Using only 3-4 mg. of protein, it is possible t o estimate all the known amino acids. The degree of accuracy revealed by the analysis of ad ZQC mixtures is 100 f 3% except in the case of the dicarboxylic amino acids which undergo a small degree of esterification in the presence of alcohol-hydrochloric acid systems. In the latter instance the accuracy is 100 f 6%. So far complete analytical data for p-lactoglobulin and bovine serum albumin have been reported and the values are in excellent agreement with those found by other modern methods. Without wishing to detract from the value of this excellent method the author is of the opinion that the smoothest operation of the method can only be attained when considerable technical assistance is available. g. Sulfur-Containing Acids. Many methods have been suggested for the estimation of the sulfur-containing acids. There is little doubt
96
Q. R. TRISTRAM
TABLE I1 Comparison of Starch Chromatography _ . .a d Other Methods of Analysis Bovine serum albumin Amino acid
Phenylalanine Leucine Isoleucine Methionine Tyrosine Valine Proline Glutamic acid Aspartic acid Alanine Threonine" SerineO Glycine Arginine Lysine Histidine a
Starch chromatography 6.59 12.27 2.61
-
5.06 5.92 4.75 16.5 10.95 6.25 5.83 4.23 1.82 5.90 12.82 4.00
Microbial assay 6.05 t o 6.48 11.8 to 13.7 2.7 to 2.97 5 . 2 to 5.5 5.4 to 6.6 5.1 to 5.6 16.6 to 18.0 10.25 to 11.1 6.2, 6.3 4.9, 4.5 1.9to2.0 5.9, 6.1 10.3 to 12.4 3.8 t o 4.1
&Lactoglobulin Starch chromatography
Silica chromatography
3.78
4.2
5.86
-
3.64 5.62 5.14 19.08 11.52 7.09 4.92 3.96 1.39 2.91 12.58 1.63
20.1
3.9 5.74 5.5 6.1 -
Microbial assay 3.5 to 4 . 3 15.4 6.1 to 8.4
-
3.6 5.5 to 5 . 9 4.1 19.5, 19.1 11.2, 11.4 6.2 4.6, 4 . 8 -5.0, 3 . 2 1.4 2.87, 2.91 11.4 1.50, 1.62
Carry the corrections recommended by Rees (1946).
that methionine and the total cystine-cysteine may be satisfactorily estimated when the protein has been hydrolyzed under carefully controlled conditions (see Baernstein, 1935a,b,c, Kuhn, Birkhofer and Quackenbush, 1939) or, in the case of the method of Blumenthal and Clarke (1935) (see Bailey, 1937, on sulfur distribution), when oxidation has been rigorously controlled. Attempts to differentiate between cystine and cysteine (Hess and Sullivan, 1943; Brand and Kassell, 1941) by suppressing the oxidation of cysteine during hydrolysis with such reagents as titanous chloride etc. are probably vitiated by the interaction of cystine and cysteine with other amino acids or their breakdown products (Olcott and Fraenkel-Conrat, 1947). It is the opinion of Bailey (1937) that the most reliable estimation of methionine and cyst(e)ine is a t present given by the differential oxidation method of Blumenthal and Clarke (1935) but thiv procedure does not differentiate between sulfhydryl and disulfide and the allimportant information on the number of dieulfide groups cannot, be obtained. Determination of Cysteine in 1 nlact Protein. Numerous reagents (porphyrindin, iodosobenzoate, mercuribenzoate, iodine in glacial
AMINO ACID COMPOSITION OF P U R I F I E D P R O T E I N S
97
acetic acid, iodoacetate, etc.) have been suggested for the titration of sulfhydryl groups in the intact protein which has been denatured by reagents such as guanidine, urea, etc. The interpretation of the results of such titrations is by no means easy, because the use of different denaturation reagents gives different values for sulfhydryl (see Neurath et al., 1944; Greenstein, 1938, 1939) and it cannot, be assumed that even the highest value represents the actual sulfhydryl content of the protein. In addition many of the oxidation reagents involve the reoxidation of sulfhydryl to disulfide and it has been the experience of Rees (personal communication) that, when the end point of the titration, under recommended conditions, with o-iodosobenzoate has been reached, the solution of suspension of protein still gives the nitroprusside test for sulfhydryl. The validity of any of the titrations on the intact protein can only be tested by control experiments on synthetic peptides. Until these are available the cysteine content of proteins has been assumed to be that found by titration of the protein denatured by guanidine. h. Dicarboxylic Acids. Many methods have been suggested for the estimation of the dicarboxylic acids (for details see Block and Bolling, 1945) but being in the main based upon gravimetric principles they are now known to yield minimal values. The most important method, based upon the original work of Foreman (1914a,b) was developed by Chibnall and coworkers (Bailey et al., 1943). This gave values for glutamic acid which were close approximations to those found by modern procedures (e.g., isotope dilution and microbial assay). It is now clear however, that the aspartic acid values were somewhat low due probably to losses during the removal of cystine as its cuprous mercaptide. The suspected presence of D-glutamic acid (Chibnall, 1946), bearing in mind the inability of certain of the modern methods to estimate the antipode, means that a true estimate of these acids may be obtained only by the use of a method which is not stereospecific. Consden, Gordon, and Martin (1948) suggested the use of synthetic anion exchange resins for the separation and estimation of the dicarboxylic acids but as used by these workers the method did not yield values of high reproducibility. Wiltshire (in Chibnall’s laboratory) has modified and improved the method so that it is now possible t o estimate the two dicarboxylic acids with fair precision. This worker has carried out numerous analyses and has obtained values for glutamic acid (Table X) which approximate to those of Bailey el al. (1943). His aspartic acid values are equal to, and, in some instances exceed those found by microbial assay (Table IX). Confidence in the work of Wiltshire is increased by the satisfactory recoveries he has obtained, for glutamic and aspartic acids, from ad hoc mixtures of amino acids representing such diverse proteins M gliadin
98
0. R. TRISTRAM
(recoveries 98.4% and 101.7%) and horse globin (recoveries 101.4%
and 96.9%). i . Additional Methods. Colorimetric Techniques. Many attempts have been made in the past to use the color reactions of various amino acids for their estimation. Such procedures, while specific for the pure amino acid, have been handicapped by the variable interference of different mixtures of amino acids and by the absence of other methods for comparing the results of analysis. Now that pure amino acids are available for control mixtures, and since it is now possible to fractionate amino acids either individually or into allied groups, colorimetric procedures may yet become important (e.g., the use of color methods for the estimation of arginine and histidine after their separation by electrodialysis; 8ee Macpherson, 1946). Certain of the few classical color methods have provided valuable information. Thus the method of Folin for tyrosine and tryptophan was developed by Lugg (1937, 1938) and although reinvestigation with the aid of full control mixtures (not available in 1937) may modify the accuracy claimed by Lugg, the method still remains one of the most accurate for these amino acids. TABLE I11 Terminal Residues of Proteina
-
P
Protein
Hemoglobin (globin fraction)O Horse, donkey Human, adult Human, fetal COW
Sheep Goat Myoglobin, home. Insulin, ox, pig, sheep*
M.W.
64 ,000 64 ,OOO 64 ,OOO 64 ,000 64 ,OOO 64 ,OOO 17,000 " 12 ,000"
-
&Lactoglobulin* Edestind
50,000
Ovalburnin' Salminen
45 ,Ooo 8,000
Porter,&Sanger (1948). Elanger (1945), Porter & Sanger (1948). e Porter (1948). ,I hnger, Porter (1947). .Porter (1949). (In the press.)
Terminal residue
No. of terminal
residues pcr molecule
Valine Valine Valine
6 5 2.7
Valine Methionine
2 2
Glycine Glycine, phenylelanine Leucine Glycine (leucine)
1 2
-
Proline
3 1 (0.33) nil 1
AMINO ACID COMPOSITION OF PURIFIED PROTEINS
99
Older Gravimetric Procedures. As indicated in the introduction most of the classical gravimetric procedures were influenced to a varying degree by the effect of amino acid mixtures upon the solubility of any one acid or its derivative, and the values obtained by such methods must now be regarded as minimal. 3. Determination of Free Amino Groups in Proteins While not, strictly speaking, a method of amino acid analysis, the finding (Sanger, 1945) that the terminal amino groups could be acylated and estimated as 2,4-dinitrophenyl derivatives provided the first means for the investigation of intramolecular structure and it would now appear to be an essential analysis in the investigation of the amino acid structure of proteins. While the method of Sanger (1945) does not claim or achieve great accuracy it is within the limits required for such a procedure. For instance Sanger found that insulin contained four terminal amino groups in a molecule of 12,000. To demonstrate that this number was four and not five or three would require an accuracy of only 20 %. Sanger and Porter (1947) and Porter and Sanger (1948) have extended the original observations of Sanger (1945) and their results are recorded in Table 111.
IV. ESTABLISHMENT OF THE ACCURACY AND SPECIFICITY OF METHODS OF ANALYSIS
Now that such a variety of methods are in existence it is at last possible to estimate all the known constituents of protein hydrolyzates, with the exception of hydroxyproline. Before the results of analysis can inspire confidence, however, the accuracy and specificity of the various methods of analysis must be established, and it is clear that this can only be done if amino acids of high purity and proteins of standard composition are available. 1. Purity of Amino Acids Used as Standards and in Media
It is of obvious importance that the amino acids used, during the course of protein analysis, either as standard materials or for the preparation of media, should conform to a very high standard of purity. In general the question of the purity of the amino acids has not been given the necessary attention and there are numerous examples of the use of amino acids (including those of synthetic origin) containing considerable impurity. For example, Smith and Greene (1948) reported that- the Dtisoleucine used by them (1947) had contained Dtalloisoleucine; Hegsted (1944) reported the presence of lO-ZO% Dtisoleucine in synthetic DL-leucine; the author has had a sample of hydroxyproline contain-
a. R.
100
TRIBTRAM
ing 34% glycine and a sample of valine (from natural sources) which contained valine, alanine, and leucine in the proportions required to give a total nitrogen of 11.9%. It is therefore imperative that before use every amino acid should be tested for purity by suitable means. It is suggested that purity should be established, within acceptable limits, by means of the following criteria (cf. Dunn and Rockland, 1947, for similar suggestions). TABLE IV Comparison of Nitrogen Ammonia Values Yielded b y Hydrolysis of Proteins under Various Conditions" Hydrolysis systems Casein Edestin Hemoglobin Insulin &Lactoglobdin Ovalbumin
Conc. HCI, 37OC.b 1 .71b 2 . 14b O.93b 1 .6gb 1 .30b 1.26b
N HCI, 100"C.C
2.32,
-
1.66' 1.311 1.19d.'
Alkaline 35" etc.,
20% HCI, 24 hr. 5 p t . i .
2.27,
1 .93b 2.33b 1.13b 2.02,' 1 .8Sb
2.15~ 1.32,f 1.310 1.19,
(Values quoted as ammonia per 100 g. protein.) See p. 88. Data from Rees (1946). Amide nitrogen. *Shore el al. (1935-6) (see Warner, Cannan, 1942). Amide nitrogen. 'Shore et al. (1935-6). Harington, Mead (1936). Warner, Cannan (1942). Arnide nitrogen. 0 Brand et al. (1945). Brand (1946). Linderstr@m-Lang,Jacobsen (1940). j Reee (1946). Amide. b
a. Racemic Acids. (1) Freedom from ammonia and inorganic material. (2) Total nitrogen of the dry material. (3) Amino nitrogen. (4) Paper chromatography (which will indicate amino acid impurities down to a limit of 0.5 to 1.0%). (5) Microbiological assay, if the acid is to be used with micro-organisms, for amino acids likely to be present as impurities. (6) Differential solubility (phase rule). b. Optically Active Acids. The criteria listed for the racemic acids, as well as (7) Optical rotation. I t is good practice to prepare both the D and L isomers and to show that the rotations are equal and opposite in sign.
2. Choice of Proteins for Testing Methods of Amino Acid Analysis In the development of methods of amino acid analysis it is not necessary to use proteins which are considered to be homogeneous, rather
AMINO ACID COMPOSITION OF PURIFIED PROTEINS
101
is it necessary to use easily obtained proteins the preparation of which may be duplicated in any laboratory. Indeed Stein (1946) and Snell (1946) have suggested that some readily obtained protein should be prepared in standard form and made available for analytical purposes. The comparison of analytical data is seldom straightforward because of the difficulty of establishing that the proteins used by the various TABLE V Comparison o j Methode for Eetimution of~-Glycine Microbiological Partition chromatography a-Y
Protein Serum albumin, bovine Human Casein Edestin Fibrin, bovine Horse hemoglobin Insulin &Lactoglobin Ovalburnin Salrnine Silk fibroin -pGlobulin, Human Tobacco mosaic virus
-
5.12' 5.6" 1.5,"1 .55d -
5.8'
-
4.51h4 . 6 " 1.4" 3.0' (3.1'1 43. 6h 4.2' 1.9f
-
Shemin (1945). Rittcnberg, Foster (1940). c Foster (1945). * Keston et al. (1946). * Sanger (1945). J Moore, Stein (1943). 0 Stein, Moore (1949). A Shankrnan el al. (1947). b
2.0,b 1.9' 1.6l 2.7,[ l . Q h
1.95" 1.600 -
R. J. Block, private communication, ninhydrin method. i Knight (1947). Velick, Ronroni (1948). I Brand, Saidel (unpublished). Brand (1946). Brand (unpublished expts.). 0 Tristram (1947b). p Tristrarn (1946). I)
workers concerned are indeed of identical composition. For this reason it should be borne in mind that many of the differences between the data recorded in Tables I V to XXI may indicate that the proteins are unidentical rather than that the various methods have differing accuracies. Many workers have used casein as a test protein and, in spite of the claim to the contrary (Dunn and Rockland, 1946), it is evident (see Chibnall, 1946) that this material is by no means of that constant composition which warrants its use as a test protein.
102
Q. R. TRISTRAM
3. Criteria for the Establishment of Accuracy and SpecijEcity The use of numerous methods based upon entirely different principles means that no universal criteria for estimating accuracy and specificity may be formulated but certain procedures have been suggested (cf. Stein, 1946; Snell, 1946; Tristram, 1947a) whereby the accuracy and specificity may, to a great extent, be controlled. These will be discussed with a TABLE VI Comparison of Methods for Estimation of ProlineProtein Bovine serum albumin Human serum albumin Casein Edestin Fibrin (ox) Hemoglobin (horse) Insulin @-Lactoglobulin Ovalbumin Salmine Silk fibroin Tobacco mosaic virus Human 7-globulin Colostrum pseudoglobulin
Partition chromatography
Microbiological assay
4.75p 10.5b 3 . 9gb 3.85b 3.96 2.56' 5.4b, 5.14p 3.6b 5.78. 0 .74e -
5.5,f 5.6,0 5.1' 5.10 11.2,' 10.9-12.3* 4.4,'4.6i 2.9,02.9/ 4.1k 4.1' 7.9'
-
(Values in g. amino acid per 100 g. protein .) b Tristram (1946). Tristram (unpublished). 6 Tristram (1949). Tristram (1947b). f Velick, Ronzoni (1948). 0 Brand (1946). h Henderson, Snell (1948). 0
-.! . . . .L-
view
*l--:-
-J--L:..-
GW I~IWII-~LUU~GIUII
^ ^ as
-A .
p
Color
-
5.8'" 8.10 10.0"
i Brand,
Saidel (unpublished) Smith, Greene (1948). F. J. Ryan, E. Brand (unpublished). I Block, Bolling (1945a). Knight (1947). Hansen el al. (1947). 0 Ross (1941). Stein, Moore (1949). i
.....-..
-P CL.. "-1 -,...4:-... ,c , . : , ,.,.:A ,.-.,.l-. i ~rUI UIC ~ C I I C I ~ I UI U M I I G UI ~ I U I I I Ua v i u ~ ~ i i a i y -
sis. Stein (1946) made the suggestion that modern methods may be divided into the following two groups: a . Primary Standard Methods. Such methods are based upon a firm and valid theoretical foundation which permits the precision calculation of the limits of accuracy. Such methods need not be rapid nor even micro in scale, but should have accuracy as their primary aim so that the results obtained on a few selected proteins may act as standards against
AMINO ACID COMPOSITION O F P U R I F I E D P R O T E I N S
103
which the values obtained by the second group of methods could be compared. Stein was of the opinion that isotope dilution was of this class. but while this may be true, the use of the isotopic method has been so restricted that it is impossible to say whether the theoretical accuracy is borne out in practice. While this and certain other modern methods of analysis are fundamentally precise (e.g., partition chromatography, electrodialysis, and enzymatic methods) uncertainties concerning the breakdown of amino acids during hydrolysis makes it necessary to place all methods in the second group (Stein, 1946). TABLE VII Comparison of Methods for Eelination of Methionine. Protein Serum albumin, bovine Human Casein Edestin Hemoglobin horse &Lactoglobulin Ovalbumin Pepsin Silk fibroin Tobacco mosaic virus Human 7-globulin Fibrin, horse Colostrum pseudoglobulin
Iodometric 0.45: 0 .8lC 1.28' 2 . 72,d 3.1,' 2.83' 3.1,' 3.5,h 2.87' 2.34,f 2.18,b2.39)1 2 .38,' 2.12,' 2.46 2.39 1 .oi 3 .2,k3 .as 5.1: 5.24,O 5.07' 4.49,' 4.58,' 3.75d 5.28" -
2 .O' 1.06," 2.5P
Values in g. per 100 g. pi:otein. Hess, Sullivan (1943). Brand (1946). Csonka, Denton (1946). Bacrnstein (1936b). f Lugg (1938a,b). 0 Bailey (1937). Baernstein (193Aa). Lavine (1943). i Kuhn et al. (1939). Brand, Kasscl (1942b). Knight (1943). Smith, Greene (1947b).
4
3ulfur distribution
0.84" 0 . 8 , 0.86,8 ~ 0.80~ 1.28' 3.2,c 2 . 5 a 3 . 1 , 8 2 . 6 9 r 2 . 6 , 3.0,. 2.9,q 2.22," 2.35,l 2.07* 2.44'
3.22= -
-
-
Microbiological assay
1.53Q 2.5,' 2.93u 4 . 5 , ~4 . l p 3.28 1.45' 0.14,. 0.15p 0.06,~ 01.12" 2.16,g 2.37' 1.1,r1.3w
Brand, Kassel (1941). Blumenthal, Clarke (1935). * Stokes et al. (1945). Lyman ct al. (1946). Henderson, Snell (1948). Dunn, Rockland (1946), Dunn et al. (1946a,b). Riesen el al. (1946). "Smith, Greene (1947a). * Knight (1947). " Hansen et al. (1947). * Bolling, Block (1943). Velick, Roneoni (1948). 0
104
0.
R. TRISTRAM
b. Routine Methods. This group of methods is based upon empirical rather than fundamental grounds and the establishment of their accuracy and specificity will now be discussed in detail. Stein (1946) suggested the following criteria: (1) The analysis of ad hoc mixtures of amino acids of high purity simulating the protein hydrolyzate under examination. (2) The analysis of the proteins previously analyzed by a primary method. The second criterion of Stein is largely inapplicable owing to the absence of an adequate number of analyses by primary methods, if indeed any of the modern methods, as at present used, do attain the theoretical requirements of a primary method. Chibnall (1946) and coworkers (Macpherson, 1946; Rees, 1946; Tristram, 1946) have adopted the following criteria in an attempt to overcome this difficulty: (1) The analysis of ad hoc mixtures of amino acids simulating the protein hydrolyzate under examination. ( 2 ) The comparison of the values obtained by the various methods of analysis. It seems reasonable to assume that, when two methods estimate an amino acid by making use of different properties of the molecule (e.g., partition chromatography and microbiological assay make use of distribution coefficient and growth-promoting effects, respectively) errors inherent in the two different methods will be completely unrelated. If the values obtained are in agreement, there are good grounds for regarding them as correct. Snell (1946) suggested that it was sufficient to obtain agreement between values obtained by assay with two or more micro-organisms. This may be so if one of the organisms is of the acid-producing type (lactobacilli, Streptococcus faecalis, Leuconostoc mesenleroides, etc.) and the other is a mutant of Neurospora crassa, but may not be in the case if both organisms are acid-producing. Perusal of the basic media and essentials for growth of the micro-organisms (see Snell, 1945, Tables 2 and 3) in common use suggest that they are metabolically very closely related, and it seems reasonable to assume, pending a fuller knowledge, that any errors in the assay of amino acids due to metabolic variations may be common to the whole group (see p. 92). There is a third criterion which it might be advisable to include. In all considerations of the amino acid composition of proteins it is tacitly assumed that all the amino acid constituents of proteins are known. This map well be so, but until there is adequate proof, the existence of additional amino acids cannot be excluded. For this reason it is suggested that any method of analysis should be capable of indicating the presence of unknown amino acids or, alternatively, that, prior to the analysis of any protein, the amino acid constituents should be characterized by some general qualitative method such as paper chromatography.
TABLE V I I I Comparison of the Methods f o r the Estimation of. Cystine-C . Protein
Serum albumin, bovine Casein Edestin Fibrin Hemoglobin, horse Insulin p-Lactoglobin Ovalbumin y-G lobulin , human Myosin Colostrum pseudoglobulin Tobacco mosaic virus
Iodometric
0.23,b0.29,' 0.3d 1.43," 1.4,c 1.5' 1.3,f l.1=
Photometric
Colorimetric
6.5, 5.4 (1.1)* 0.34 (0)' 1.2 (0)e
5 . 8 (0.3)i 0.32= 1.3 (0.32); 1.18," 1.2,' 1.141 2.1k
-
1. O l g (globin) 12.0,p 12.2' 1.8,g 2.03,' 1.72. 1.4f
-
a Values in g. per 100 g. protein Cysteine values in parentheses. *Lugg (1938a,b). Baernstein (1936b). Beach et al. (1939). Kassel, Brand (1938a,b). f Bailey (1937). g Kuhn el al. (1939). * Brand (1946).
(0.42)' 11.6 (0.6)& 12.5,' 11.7. 3.39 (1.11)h
-
-
12.8,- 8.4,- 10.6,' 11.6-12.6' 3.5" 2.40 (1.37);
-
2.55,'2.37 (0.7)
3.04'
-
0.68 (0.57)f
Smith, Greene (194%). Hess, Sullivan (1943). Block, Bolling (1945). I Sullivan, Hesa (1939). Bolling, Block (1943). * Blumenthal, Clarke (1935). 0 Greenstein (1939) p Greenstein (unpublished). q Greenstein (1938). i
kineo Microbio- Differlogical ential midation assay
Titration of SH on denatured protein
-
0.34" 1. 4gb
( 0 . 3 4 ) , u (0.32)'
-
-
-
-
12.5'
-
-
-
-
(1.28),@(1.41)j
-
-
-
(0.49)" (0.52)i
-
(0.56)o (globin) (0)p
-
(1,17),' (1.25)i
-
(0.69)'
Greenstein, Edsall (1940).
v
* Knight (1947).
3
Vickery, White (1932). Miller, d u Vigneaud (1937). ~Velick,Ronzoni (1948). du Vigneaud et al. (1927). Vassel (1941). Greenstein (1940). ' Chibnall (1946).
E!
3
3
+
0
m
106
CI. R. TRISTRAM
V. COMPARISON OF ANALYTICAL METHODS Values for amino acids, in cert,ain proteins, obtained by various methods are contained in Tables IV to XXI. Selection of these values has been restricted to those obtained by analysis conforming, as far as possible, to the criteria outlined above. In collecting the data it became apparent that many workers were rendering a great disservice to amino TABLE IX Comparison of Methods for Estimation of Asparlie Acida
-1sotOpi Ionic diluextion change
Protein
Serum albumin, bovine Human Casein Edestin Fibrin, bovine Hemoglobin, horse Insulin pLactoblobulin Ovalbumin Silk fibroin Tobacco mosaic virus Human 7-Globulin Colostrum pseudoglobulin
P
Isolation
10.28 9 .85b
-
-
6.7' 12.0,f 11.8,0 11.2h
11.2* 10.4c
-
-
-
-
Values in g. protein. bshemin, Foster (1946). Foster (1945). d G. Wiltshire (unpublished). Bailey el at. (1943). f Chibnall et d.(1943). Cannan (1944). Kibrick (1944).
10.6,"1 1 . 1 , ' 10.45% 10.4" 7 . 2 , i 6 .1,p7.3,n7.4i 13.4' 13.0" 10.V
5.680 6.8" 11.5,i 9 . 3 , " 11.4" 9.88,f 9.5,g 10.1' 8.13,f 7.45,@7 . 85h 9.3,' 8 . O p 2.W 13.9 8.8"
11.20
amino acid Per
Microbiological assay
9.3,'9.4'
g.
Henderson, Snell (1948). Hac, Snell (1945). Knight (1947). Hansen el al. (1947). Velick, Ronzoni (1948). "Brand (1946). Chibnall (1946). Stokes, Gunness (1945a). f
acid chemistry by their lack of appreciation of the need for control experiments and that the comparison of results was, in many cases, impossible because the proteins used in the analyses were not only heterogeneous but grossly impure. On occasions comparison was made difficult because authors had failed to indicate whether or not the values were based upon the dry, ash-free protein. Before proceeding with a discussion of the data contained in these Tables, it may be advantageous to discuss the systems for reporting analytical data.
AMINO ACID COMPOSITION OF PURIFIED PROTEINS
107
1. Reporting Analytical Data
The reporting of analytical data has been discussed by Brand and Edsall (1947)) who recommended that data should be recorded as the weight of amino acid in 100 g. protein (and also as grams amino acid residue per 100 g.). They criticized the use, by Chibnall and coworkers, of the system whereby data was expressed as grams amino acid nitrogen in 100 g. protein nitrogen although the reasons for the criticism were not made clear. It must be emphasized that this system was suggested as a means of overcoming the difficulty of comparing the results of analysis created by the inability of many workers to express data in a form which made comparison possible. In the writer’s view the reporting of analytical data for the purpose of comparison (this is distinct from the reporting of complete analyses of proteins, which will be discussed later) should be done in such a way as to indicate the following (see pages 102 and 104): (1) That the accuracy and specificity have been checked by adequate control analyses. (2) That the values are comparable to those obtained by methods based upon entirely different principles. (3) That it be clearly indicated whether the values are corrected for ash and moisture content of the protein. In addition the total protein nitrogen and moisture content of the protein should be included as an indication of the state of the protein and to enable the values to be used for the calculation of molecular data. If these conditions are observed, the data may be expressed in the form preferred by the individual worker. I n Tables IV to XXI all data have been expressed as grams amino acid per 100 g. protein to conform with the usage of the majority of workers. 2. The Data of Tables IV to
XXI
While there are still differences in the results of analysis the degree of agreement achieved is remarkable and, for the first time in the history of amino acid analysis, it is possible to make use of amino acid data with justifiable confidence. A brief discussion of the data will indicate probable reasons for certain of the differences and also suggest where further work is necessary. Before discussing the amino acids individually one may point out that the majority of analyses have been carried out for those amino acids which are classed as essential for the growth of the young animal or which contain easily estimated polar groups. The reason for this, as has been indicated earlier, is that by far the greater number of analyses are carried out during nutritional studies or by those who have only a transient
TABLE X
-
Kethoda for Estimation of Glutamic Acid. Protein
Isotope dilution (L isomer)
Serum albumin, bovine HUm2lIl CaSein
16.98 17.03* -
Eklestin Fiben, bovine Hemoglobin, horse
13.W 8.5d
Ionic exchange
-
0 Values in g. amino acid per 100 g. protein. * Shemin (1945). Shemin, Foster (1946). Foster (1945). G. Wiltshire (unpublished). Bailey d al. (1943). Chibnall d al. (1943b).
22.01' 20.7,e 16.9,s21.2h
Insulin &Lactoglobulin Ovalbumin Pepsin Silk fibroin Tobacco mosaic virus Human 7-globulin Colostrum pseudoglobulin
Isolation (DL isomer)
-
18.6. 21.5,g 18.5,' 21. 2h 14.3,- 16.2,p 16.9*
19. I d
-
-
-
* Kibrick
(1944). Snell, Guirard (1943). i Velick, Ronzoni (1946). Hier et d. (1945). Henderson, Snell (1948). Brand (1946). a Lyman et d.(1945). 0 Dunn et al. (1944s). f
Microbiological assay (L isomer) 17.0,i 18.0,t 16.6,'16.9" 17.4'" 22.4," 22.5," 21 .7,q 19.7,' 21 .4,' 22.3," 21 .7' 21.3,q 19.1,r 19.4' 12.4,' 15.01L14.3' 10.8q 17.5,' 20.2," 19.9' 18.9,q 19.5,' 1 8 . 7 15.0,q 13.7,' 14.3" 11.5' 2.16,o 2.2Q 12.5,' 11.3" 11.8" 10.7,' 12.3'
Gale (1945). Hac el al. (1945). Lewis, Olcott (1945). 8 Knight (1947). Hansen et al. (1947). y Cannan (1944). * Chibnall (1946). P Q
Specific decarboxyl (L isomer)
AMINO ACID COMPOSITION OF PURIFIED PROTEINB
109
interest in protein analysis. The remaining amino acids, such as glycine, alanine, proline, and hydroxyproline, have been neglected, to the disadvantage of those who are interested in the stoichiometry of the amino acids and who may not possess the facilities to extend the few analyses reported for these amino acids. a. Hydroxyproline. A t present there is no satisfactory method for the estimation of hydroxyproline and, apart from a few special cases (collagen, gelatin, etc.) it is probably true that there is no evidence that TABLE X I Comparison of Methods for Estimation of Serineo Protein Serum albumin, bovine Human Casein Edestin Hemoglobin, horse Insulin &Lac toglobulin Ovalbumin Salmine (free base) Silk fibroin Tobacco mosaic virus Human y-globulin Colostrum pseudoglobulin
Microbiological assay
-
4.9,j 4.5k 3.7k 6.41
5.0,' 5.87,c5.5d 6.3" 5.8,' 5.56d 5.2" 4.08,c 5.0' 8.15," 7 . 4d 9.15," 7.44,d 6.94' 1 6 . 3 , ~12.6r 6.4h 11.4'
0 Values in g. amino acid per 100 g. protein. Nicolet, Shinn (1941a). Reee (1946). Boyd, Logan (1942). E. Brand, B. Kaasell (unpublished). Block, Bolling (1945a). Coleman, Hewitt (1946). @
Periodate oxidation
-
5.8,k6.6i 5.0,k 3.2' 10.3;
-
14.5' 7.2"
-
*Ross (1941). Brand, Saidel (unpublished). i Velick, Ronzoni (1948). Brand (1946). Stokes, Gunness (1945a). Knight (1947). Tristram (1947b).
hydroxyproline is a general protein constituent. Absence of hydroxyproline values however is a serious drawback in assessing the completeness of any protein analysis, since it is impossible to decide whether, when 97-99% of the total protein nitrogen has been recovered, the missing nitrogen represents hydroxyproline nitrogen or is an indication of the overall accuracy achieved. Paper chromatography may clear up the doubt about the general distribution of hydroxyproline in proteins. Indeed it has been shown to be absent from insulin (A. H. Gordon, personal communication), 8-lactoglobulin, ovalbumin, and horse myoglobin (unpublished experiments, Biochemical Laboratory, Cambridge).
110
0 . R . TRISTRAM
b. Glycine. It is now possible to estimate this amino acid by isotope dilution or microbiological assay and (although there is good agreement in the values obtained for glycine many proteins remain to be assayed). Recent experiments (Macpherson, Rees, and Tristram, unpublished work) have indicated that it is possible to estimate glycine by means of TABLE XI1 Comparison of Methods for Estimation of ThreonineO ~
Periodate oxidation Serum albumin, bovine Human Casein
6 . 5,c 7 . 5.0' 4.5,'3.5,'4.39
Edestin Hemoglobin, horse Insulin phctoglobulin
3.85' 4.36' 2.08,'3.2,C3.5'nd 5 .85,15.15,' 5.8* 5.35a 4.05' 1.57,"1.5' 5.3i 8.4,C
Ovalbumin Silk fibroin Tobacco mosaic virus Human 7-globulin Horse ?-globulin Colostrum pseudoglobulin
lbsd
Values in g. amino acid per 100 g. protein. * 10 % correction added; Reee (1946) suggested 5 % would be sufficient. e Brand (1946). d Velick, Ronzoni (1948). 1 Rees (1946). Shinn, Nicolet (1941). Winnick (1941). * Bolling, Block (1943). Coleman, Hewitt (1916). 0
-
~~
~
~
Microbiological away 6 . 3 , ' 6.2"
-
4.3,"4 . 2 , 0 4 . 1 , 1 4 . 8 , ~ 4.3" 3 . 1 4 , ~4 . 3 4 , ~ 3.7"
-
4.6,O 4 . 8 ~ 3.6,O 4.48p 1 .2," 1 .36O 8.7,09.9' 8.P
11.1* 9 . 0 , ' 10.2," 10.3&
-
Ross (1941). &Smith,Grcenc (1947b). 1 Hier el al. (1945). Henderson, Snell (1948). I) Dunn el ~ l (1946a,b). . Stokes et ul. (1945). Horn el al. (1947). Smith, Greene (1947a). Knight (1947). 8 Hansen el al. (1947). E. Brand, B. Kassell (unpublished). i
Q
partition chromatography after the basic and dicarboxylic acids have been removed, and various proteins are now being analyzed. c. Proline. Proline may now be determined by partition chromatography (Tristram, 1946; Gordon, Martin, and Synge, 1943) and microbiological assay. Values by the latter method are mainly obtained (Ryan and Brand, 1944) by the use of a proline-requiring mutant of Neurospora crassa. The microbiological values, which are based on unpublished experiments, tend to be higher than those obtained by
AMINO ACID COMPOSITION OF PURIFIED PROTEINS
111
partition chromatography and a final assessment of the proline content of protein hydrolyzates must await further analyses and the development of additional methods for the assay of this acid. d . Methionine. While there is reasonable agreement between the values obtained by the chemical methods (iodometric and sulfur distribution) recent determinations by microbiological assay have tended to give lower values (e.g., 8-lactoglobulin, ovalbumin). It is impossible to say without further work whether this is due to errors in the chemical or microbiological methods. (See p. 95.) e. Cystine-Cysteine. In discussions of the methods for the estimation of these amino acids it has been suggested that none of the methods at present in use inspired confidence. Reasons have been given for thinking that only the total cystine-cysteine could be estimated after hydrolysis, or, preferably, by differential oxidation, and that the cysteine content should bP estimated by titration of the protein denatured in guanidine TABLE XI11 Comparison of Methods f o r Estimation of Phenylalaninea Protein
Serum albumin, bovine Human Caeein Edestin Hemoglobin, horse Insulin &Lactoglobulin Ovalbumin Silk fibroin Tobacco mosaic virus Human r-globulin Colostrum pseudoglobulin Values in g. per 100 g. protein. *Foster (1945). Velick, Ronroni (1948). Brand (1946). Hier et al. (1945). Henderson, Snell (1948). 0 Dunn el al. (1945). *Stokes el al. (1945a). Hegsted (1944). i Smith el al. (1946a,b).
a
Microbiological assay 6 . 0 , 6~. 2 , d6.4; 6.51 7.8d 4 . 9,n 5.2,9 5.9,h 3 . 7 ' 5.3,' 5.41 5 . 4 5 , f4.2,' 5.221 7.5* 7.9,d 7.95" 4.3,A 3.541' 7.9h I .28,9 1 .4gA 6.8,h 8 . 4 ' 4 . 6 , d4 . 6 i 4.25,j 3.7,'"3 . 7 i
Partition chromatography 6.57"
-
6.680
5.71" 7.9p 8.10
4.2,o 3.78" 7.42O 3.36~ -
-
-
Brand el al. (unpublished expts.). Knight (1947). Hansen el al. (1947). nS. Moore, W. H. Stein (private communication). Tristram (1946). Tristram (unpublished expts.). Hess, Sullivan (1944). Bolling, Block (1943). Ross (1941). I
a.
112
R. TRISTRAM
TABLE XIV Comparison of Methods for Estimation of Leueine Isomersa Protein
Partition chromatography
Micro biological assay
Fibrin, human
14.86 (2. 56)c 15.9 (2.7)' 1 6 . 6 (2.9),1 14.8 (3.0)& 1 4 . 6 (3.3)' 1 3 . 6 (1.7)i 15.82 (5.1)d*' 17.85 (7.6)k 16.0 (6.2),"'8" 15.5 (5.6)O 15.32 (6.05),p (9.8)'" 1 5 . 5 (6.0)' 12.40 (2. 2)d*r 12.1 (4.7),q-* [7.411 13.95 (6.47)k 1 2 . 8 (5.9),' 1 2 . 3 (5.2)"J 11.7,
Hemoglobin, horse Insulin
15.7 (0)"*" 16.00 1 5 . 7 (2. 8)d*r 16.3 (2.9),i 1 5 . 8 (2.8)'
,%Lactoglobulin
20. I* 21.36 (5. 9)c 15.94d 3.10 1.61*
Serum albumin, bovine Human Casein
Edestin
Ovalbumin Silk fibroin Salmine Tobacco mosaic virus a-Globulin, human Colostrum pseudoglobulin
-
2 2 . 3 (7.0)," 2 3 . 8 (8.4)l (6.1),be123.4 (6.4)" 16.2 (7.0)O 2 . 1 (1.1),4 ( 0 . 8 ) ~ (1.44)' 13.2 (5.7)," 1 5 . 9 ( 6 . 6 P 12.0 (2.7),i 12.2 (2.8)g~' 1 2 . 4 (4.7),k1 2 . 7 ( 4 . 2 ) , * 11.73 (3.2)q.'
Figures are for total leucines; isoa Values in g. amino acid per 100 g. protein. leucine values, where known, are given in parentheses; leucine only in brackets [ 1. Old values (see Smith el al., 1946a,b, 1947a) affected as b New isoleucine values. presence of DL-alloisoleucine in standard has now been demonstrated (Smith and Greene, 1948). c S. Moore, W. H. Stein (private communication). Stokes et al. (19458). Tristram (1946). p Kuiken el al. (1943). Darmon el al. (1948). Smith el al. (1946a,b). J Tristram (see Bailey, 1944). Tristram (unpublished). 'Smith, Greene (1947a). * Smith, Greene (1948). * Tristram (194713). Brand el al. (1945). Velick, Ronzoni (1948). " Bolling, Block (1943). i Brand (1946). 0 Block, Bolling (1945a). k Henderson, Snell (1948). Knight (1947). Hier el al. (1945). * Hansen et al. (1947). Ryan, Brand (1944). v Foster (1945). m Brand, Saidel (unpublished). 0
0
AMINO ACID COMPOSITION OF PURIFIED PROTEINS
113
solution. In general titration of sulfhydryl in the intact protein yields fairly concordant results, provided the conditions are properly controlled. The importance of an accurate knowledge of disulfide and sulfhydryl in the protein molecule has been made clear by the work of Sanger (1945) and Porter and Sanger (1948) on the terminal groups of proteins, and TABLE XV Comparison of Methods for Estimation of Valinea Protein
Partition
Serum albumin, bovine Human Casein
5 . 84b
Edestin Fibrin, bovine Hemoglobin, horse Insulin @-Lactoglobulin Ovalbumin Silk fibroin Salmine Tobacco mosaic virus 7-Globulin, human Colostrum pseudoglobulin
4.75c
-
6 .02c
8.4d 7.49" 5.74,e 5 . 62b 4.62= 3.V 3.1.
Values in g. amino acid per 100 g. protein. b S . Moore, W. H. Stein (private communication). c Tristram (1946). d Tristram (unpublished). Tristram (1947b). f Hier el al. (1945). Henderson, Snell (1948). A Velick, Ronzoni (1948). i Brand (1946). McMahan, Snell (1944). 0
Microbiological assay 5.4,' 6.6,s 6 . 6 , A6 5' 7.7' 6.7,k 6.8,h 6.25," 7.150 6 . 7," 5 . 2 , "7 . 1 , j 6 . 9p 6 . 6 , 4.7,O ~ 6.6r 5.5' 10.7,'8.8,L 9 . 4 ~ 8.8,' 9.1A 5.83,. 5.5," 5 . 8,k 5.99 6.8,k 7.0," 7 . 3 6 ~ 3.5," 3.4,p 3.2k 4.1' 7.0," 9.2I 9.7,' 9 . 7 s 9.3,c 8.7,' 9 . 1 U
B. M. Guirard, E. E. Snell (unpublished). Kuiken el al. (1943). Stokes el al. (1945). * Hegsted (1944). 9 Guirard el al. (1916). Brand el al. (unpublished expts.). Block, Bolling (1945a). Knight (1947). Smith el al. (1946a,b). * Hansen el al. (1947).
this information would best be achieved by a systematic study of the estimation of cystine, cysteine (in the intact protein), and methionine, which should include an investigation of the effect of hydrolysis conditions upon the modification and breakdown of these acids. Once the sulfur distribution in proteins is satisfactorily established it will be possible to decide whether disulfide linkages are the sole means of cross-linking the intramolecular chains of the protein or whether there may be additional bonds such as those envisaged by Chibnall (1942).
114
0 . R. TRISTRAM
j . Dicarboxylic Acids. Aspartic Acid. It now seems certain that the values reported by Chibnall and coworkers using the classical isolation methods are somewhat low. Those obtained by microbiological assay and isotope dilution are in good agreement but since both are specific for the L isomer further analyses by a nonselective procedure would seem to be essential. Gluhmic Acid. Chibnall and coworkers (1939) have indicated that wglutamic acid may be present in certain protein hydrolyzates and consider that the inability of microbiological assay and, as used at present, isotope dilution to estimate the D acid accounts for the lower results obtained by these methods. Shemin and Foster (1946) have suggested that the amount of D-glutamic acid in protein hydrolyaates does not exceed 1%, and infer that cystine contamination is the cause of the low specific rotations observed in the glutamic acid fractions isolated by the Cambridge group. For instance, they state that by successive recrystallization to constant isotope content and 7.7% total nitrogen the rotation of a glutamic acid hydrochloride fraction from human serum albumin was -4.3" instead of +31". If, as they state, this low rotation is due to the presence of cystine hydrochloride, a nitrogen content of 7.7% would be obtained with a mixture containing 5% cystine and having aD 17.5", whereas a mixture having aD - 4' would contain 15% cystine (total nitrogen 7.82%). With the development of ionic exchange for the
+
TABLE XVI Comparison of Methods for Estimation of Alaninea Protein
Casein Edestin Hemoglobin, horse Insulin phctoglobulin Ovalbumin Salmine Silk fibroin Tobacco mosaic virus
MicroPartition hromatog biological assay raPhY
Oxidation
Isotope. dilution
3 . 19' 4.31' 7.4" 4.46 6.64* 6.726 1.1* 29.7'
-
a Values in g. amino acid per 100 g. protein. b Tristram (1946). * Tristram (unpublished). d Tristram (1947b). Brand (1946).
Knight (1947). Block, Bolling (1945s). * R. J. Block (private cornmunication). Keston et al. (1946). f Coleman, Hewitt (1946).
AMINO ACID COMPOSITION OF PURIFIED PROTEINS
115
estimation of these acids (WiItshire, see page 97) a method is now available which is not stereospecific. The values so far obtained by this procedure tend to confirm those of Bailey et al. (1943) for glutamic acid (see Table X) and those obtained by microbial assay for aspartic acid (see Table IX). The probable existence of D-glutamic acid in protein hydrolyzates raises the possibility that other amino acids exist in the D form. While the D-amino acids in general constitute only a small fraction of the total, the widespread use of methods of analysis which are specific for the L isomer means that the D isomer, if present, will be unestimated. Although this point is largely ignored a t the present time it should be borne in mind that not only may the values obtained by the use of stereospecific methods be minimal, but the presence of the D isomer may influence the activity of the micro-organisms (cf. Lyman and Kuiken, 1948). g. Serine and Threonine. These acids are estimated by periodate oxidation or by microbiological assay. There are differences between certain periodate and microbiological values and also between independent periodate values. These are particularly important in the cases of insulin and 8-lactoglobulin, involving a difference in the number of residues of serine and threonine in the molecules of these proteins. It is therefore of interest to discuss the methods employed by Rees (1946) and Brand (1946) in order to assess the more probable values. The precautions taken by Rees (1946) have been discussed above. This worker found the procedure of Nicolet and Shinn (1941a,b,c) somewhat unsatisactory and adopted various modifications including the estimation of total periodate nitrogen as a check on the sum of serine nitrogen plus threonine nitrogen. Rees (private communication) is also of the opinion that each protein should be regarded as a separate problem, and a strictly routine procedure avoided. Brand (1946) used the procedure of Nicolet and Shinn, but has not published any details of his experiments. Confidence in the values reported by Rees is enhanced by his extensive analysis of ad hoc mixtures of amino acids representing various protein hydrolyzates (Table XXII). (See p. 88.) The agreement between total hydroxyamino nitrogen and the ammonia nitrogen produced by periodate oxidation (in carbohydrate-free proteins) is also illustrated in Table XXII. h. Nonpolar Amino Acids. I t had long been recognized by protein analysts that this group of amino acids (phenylalanine, leucine, isoleucine, valine, and alanine) was the,greatest obstacle to attaining a complete amino acid analysis of any protein, since the methods available before 1941 were little better than detection procedures. Consequently
c c Q,
TABLE XVII Comparison of Methods for Estimalion of Tyrosine. Protein Serum albumin, bovine Human CaEein Edestin Fibrin, bovine Hemoglobin, horse Insulin @-Lactoglobin Ovalbumin Silk fibroin Tobacco mosaic virus 7-Globulin, human a Values in g. amino acid per 100 g. protein. Acid hydrolyzates. Thomas (1944). Lugg (1938a,b). Chibnall (1946). f Bolling, Block (1943). 0 Coleman, Hewitt (1946). * Ross (1941).
Colorimetric
Photometric
-
5.6,'5.49i 4.741 6.1"
-
6.0,c 6.35d 3.98,' 4.34d -
13.0. 4.2,
-
10.69 3.9,*3.6,c 3 . P
-
-
5.9"
-
12.2,' 12.323.78' 4.0"
-
6. 75m
Velick, Ronzoni (1948). Brand et al. (1944). Knight (1947). Brand et al. (1945s). Brand (1946). Shemin (1945). 0 Foster (1945). p Henderson, Snell (1948). * Gunness et al. (1946). j
Isotope dilution
Microbiological assay
5.53" 4.75" 3.w
4 . 3 7 5.3,q 5.2'
-
-
-
-
5.6,q 6.5,' 6 . D 4.29,*3.7p 5.66,s 6.0' 3.06'
-
3.69 3.4q 3.4q
-
Partition chromatographyb
I
5.0'
-
5.76" 4.49"
-
12.05' 3.9," 3.64' 3.87" 12.5' -
-
Hier et al. (1945). Gale (1945). Decarboxylase. 'S. Moore, W. H. Stein (private communication). Tristram (1946). * Tristram (unpublished experiment). Brand, Kassell (1939).
GJ !a
e l
i?
m
el
!a
2!
AMINO ACID COMPOSITION OF PURIFIED PROTEINS
117
the achievements of recent years have been remarkable, and it is now possible to conduct analyses for each of the amino acids with a total expenditure of less than 50 mg. protein. The values obtained by the application of partition chromatography and microbiological assay, and TABLE XVIII Comparison of Methods for Estimation of Trypto Iana
Serum albumin, bovine Human Casein Edestin Fibrin Hemoglobin, horse &Lactoglobulin Ovalbumin Tobacco mosaic virus y Globulin, human Colostrum pseudoglobulin
Colorimetric
Photometric
-
0.58,g 0.58*
0 . 17h 1.33: 1 . 2 ~ 1.2k 1 . 4Sb
-
1.9d 1 .23c 2.0' 2.731 2.33,
Values in g. amino acid per 100 g. protein. bLugg (1938a,b). c Sullivan, Hess (1944). d Bolling, Block (1943). Ross (1941). I Smith et al. (1946a,b). Velick, Ronsoni (1948). 0
@
-
3.5"
1.94* 1.19h
-
2.86' -
* Brand,
Microbiological assay 0.45i
1.07,' 1 ,261 1.28,' 1.02i 2.1' 1.18," 1.41' 2.3,' 2 . 1 . 2,631 2.27-3.2,i 3.2,. 2,201
-
Kassell (1939). Cf.Brand, Saidel (1943). i Brand (1946). f Henderson, Snell (1948). Ryan, Brand (1944). 8 Stokes et al. (1945). Greene, Black (1943). Knight (1947). Hansen el al. (1947). f
0
to a lesser extent isotope dilution, are in reasonably good agreement, although certain values require reinvestigation. Two examples are the isoleucine content of 8-lactoglobulin and edestin (Table XIV). The isoleucine content of 8-lactoglobulin has been variously reported as 7.0, 8.4, 8.7, 6.1, 6.4 (microbiological assay) and 5.9 (starch chromatography). The inconsistent values obtained by microbial assay may probably be explained by a recent observation by Smith and Greene (1948) (see page 99). Similarly the isoleucine content of edestin was reported as 4.7, 6.2, and 6.5 (microbiological assay). Smith and Greene (1947a) have now modified their earlier value (6.2%)to 4.7%. The distribution of the leucine isomers was estimated by Darmon, Sutherland, and Tristram (1948) by infrared absorption of the acetyl derivatives. These workers reported that the ratio of isoleucine to leucine was 20:80 compared to a
TABLE X I X Comparison of Methods j o t Estimation of Arginine" ~~
Protein Serum albumin, bovine Human Casein
Microphotometric 6.2b 6. 15b
Dialysis, photometric -
-
-
4.099
Edestin
-
16.659
Fibrin Hemoglobin, horse
-
3.630
3.47b 2. 87b
3.070 2.799
-
5.739 1.19 85.23.h 10.59
Insulin &Lactoglo bulii Ovalbumin Silk fibroin Salmine Tobacco mosaic virus Human 7-globulin Colwtnim pwd7globulin
-
-
-
9 . l l C9.2d 4.V
-
-
a Values in g. amino acid per 100 g. protein. * Brand, Kassell (19424. "Ross (1941). Knight (1942). ' Smith, Greene (194%). Brand (1946). 0 Macpherson (1946). * Tristram (194%).
Enzymatic
-
Gale (1946). Velick, Ronzoni (1948). Hier et al. (1945). Henderson, Snell (1948). " McMahan, Snell (1944). ,, Guirard et al. (1946). Stokes el al. (1945a). *Smith, Greene (1947a). Knight (1947). j
Microbiological
=Y 5.9,j 6.1,k 6.2' 3.71,"' 3.7,a 3.9" 3.6,k 3.8' 1 6 . 7 , ~17.4' 7.2k 2.8," 3.5" 3.4i 2.91," 2.88" 2.8,'2.95p 5.6,=5.7," 5.9O 1.1,O 0.97," 0.99"
-
8.9," 9.89 5.1' 5.6,'5.2,'3.64"
Isolation
3.72; 3.216.76,' 16.7; 14.16' 7.7,' 6.6y 2.9,' 3.62r 3.82; 3.04. 2.89,; 3 . B 5.66,' 5.73. 0.76' 86.0"
-
-
-
Hansen et al. (1947). Vickery (1940). Chibnall et al. (1943b). * Albanese (1940). * Chibnall (1942). Block, Bolling (1945s). Foster (1945). * Bolling, Block (1943).
AMINO ACID COMPOSITION OF PURIFIED PROTEINS
119
ratio of 40:70 by the more recent microbiological determination. It is clear that further work is necessary before the isoleucine content of this protein is accurately known. i. Tryptophan an2 Tyrosine. As a result of the careful investigations of Lugg (1938) and others the estimation of these amino acids is fairly satisfactory, although in view of the unstable nature of tryptophan e.g., the recent finding of the effect of tryptophan upon cystine (Olcott and Fraenkel-Conrat, 1947) it is possible that all tryptophan values so far TABLE XX Comparison of Methods for Estimation of Histidine" ~
Protein Serum albumin, bovine Human
Casein
Edestiri
f Iemoglobin, horse Insulin 8-Lactoglobulin Ovalbumin Silk fibroin -pGlobulin, human Colostrum pseudoglobulin Tobacco mosaic virus
Isolation 3.35b
Dialysis, photometric 3 . 83h
3 . 22b 3 . 52h 2.72,c 2 . 5 2 3.17' 2.84,; 2 . 8ah
2.72,c 2 . 4 d 2.52,b 3.04 7 . 17,c8 . 7 8 7.66. 1.8,1 1.54d 1.45d D.470 -
2.89,' 3.02h 8.39' 8.79h 4.9,'5.3* 1.64,' 1. 58h 2.38,' 2 . 27h 0.36' 2 5' -
-
Eney-
matic
Microbiological assay 3.62,' 4 . 1,n 3.80 3.38,p 4.0q 2.8,'2.96,r 3.1. 3 . 0 , ' 3 . 0 , ~3.0' 2.8,=2.6," 3.2= 2.950 2 . 6 2 , ~2.61
5.3g 1.51,= 1.62w 2.3= 0.34,' 0.34,: 0.41* 2.01v 2.22,' 2.3,' 2 . 1 3 ~ 0 . 0 , e 0,02=
_=__
a Vnlurs i n g. amino acid per 100 g. protein. b Vickcry, Winternite (1944). c Albanese (1940). Chibnall el al. (1043b). Vickery (1942). Bolling, Block (1943). a Coleman, Hewitt (1946). Value corrected from, 0.33 to 0.47 % for losses. Brand, Saidel (1943). Macpherson (1946). Ilanke, Koessler (1920). Brand (1946). Henderson, Sw11 (1048). Gale (1945). I
IIier el al. (1945). Brand, Saidel (unpublished experiments). ~ 1 , y m a net al. (1947). q Velick, Ronzoni (1948). Dunn, Rockland (1946). Dunn et al. (1945). Dunn et al. (194513). Y B . M. Guirard, E. E. Snell (unpublished). 0 Knight (1947). I0 Smith, Greene (1947a). * Stokes el al. (1945). Y Smith, Greene (1947b). * IIansen el al. (1947).
c. h) 0
TABLE XXI Comparison of Methods for Estimation of Lqsine.
Protein Serum albumin, bovine
Lra1ues in g- . amino acid per 100 g . protein. b Shemin, Foster (1946). Foster (1945). Albanese (1940). ' Chibnall et al. (1943b). Chibnall (1946). Bolling, Block (1943). * Zitfle, Nelson (1944).
Electrodialysis (difference)
-
-
-
7.35d 2.37; 2 . 91d 9.ld 7.9d 2.43 9.75,' 9.90 5.1. -
Edestin Fibrin, bovine Hemoglobin, horse Insulin &Lactoglobulin Ovalbumin Silk fibroin Tobacco mosaic virus -pGlobulin, human Colostrum pseudoglobulin 0
Enzymatic
-
Human Casein
L
Isolation
-
' Gale (1945).
-
-
8.1.* 8.25'
8.34'
2.38' 9.1' 8.2,i9j 9 . 15' 2.47' 11.4,k 11.2' 6.13'
2.24'
-
-
-
-
Catholyte of Macpherson (1946). Brand et al. (1945a). Macpherson (1946). Brand, Saidel (unpublished). Hier el al. (1915). Henderson, Snell (1948). p Dunn el al. (1944). q Stokes et al. (1945a). j
-
-
8.59' 2.51' 11.1' 6.14' 0.7' 1.03' -
Microbiological
12.4," 12.3" 10.3,O 11.9" 12.3" 8.3,p 7.7,q 7 . 6 n 7.6,'8.2,' 8.07" 2.49,'3.23,.2.09O 8.5"
8.7,' 8.45' 2.61," 2.411.1,q 10.4,r11.5r 6.6,q 6.32' 0.6,~0.7q 1.34,q 1.47" 8.1,'"7 . 2 w 6.13,06.1,= 7.2-
-
'B. M. Guirard, E. E. Snell (unpublished). * Horn el al. (1947). Smith, Greene (1947a). " Velick, Ronzoni (1948). *Knight (1947). Smith, Greene (1917b). Hansen et al. (1947).
0
; 4
0
k
AMINO ACID COMPOSITION OF P U R I F I E D P R O T E I N S
121
obtained are first approximations only. Although Olcott and FraenkelConrat (1947) worked with acid solutions, in which tryptophan is known to be unstable, recent experiments by Rees and Schmid (personal communication) suggest that the recovery of tryptophan, after alkaline hydrolysis, may be influenced by the presence of cystine. Using the method of Graham et al. (1947) in which tryptophan was estimated in the form of the colored compound formed with p-dimethylaminobenzaldehyde, it was found that, with increasing proportions of cystine or cysteine (but not methionine), the color produced by a known amount of tryptophan fell to 50% of that produced by tryptophan alone. The color produced in the presence of cystine developed only after the lapse of several hours rather than the usual 30 minutes. When the equivalent of 1.2% of tryptophan was added to insulin and the mixture subjected to alkaline hydrolysis, the color did not appear for several hours and was finally equivalent, after 17 hours, to a recovery of 42%. At present, however, these workers have not confirmed that these observations are evidence of the destruction of tryptophan rather than of the interference of cystine in the formation of the tryptophan-p-dimethylaminobenzaldehyde complex. Although Lugg (1938) carried out exhaustive control experiments on solutions containing tryptophan, he was not able to investigate to what extent the lability of tryptophan was increased when in peptide linkage. Because of this uncertainty the use by Brand (1946) of tryptophan as the basis for calculating the distribution of residues in 8-lactoglobulin and its molecular weight is not seund even though the estimation of the tryptophan remaining after hydrolysis is probably very accurate. j. Basic Amino Acids. A variety of methods are now available for the determination of these amino acids and there is over-all agreement between the values reported for the various proteins. 3. Degree of Accuracy of Various Methods The over-all agreement among the various methods of analysis suggests that they are specific for the amino acids concerned. Before undertaking the discussion of protein stoichiometry it is necessary to consider the degree of accuracy of the various methods of analysis since on this will depend the certainty with which the amino acid composition of proteins, as found by the analysis of their hydrolyzates, may be placed upon a stoichiometric basis. It is by no means easy to establish accuracy, for, while the recovery of amino acids from ad hoc mixtures may be assumed to be a measure of the accuracy of the recovery of amino acids from protein hydrolyzates, no means are available for evaluating the modification which may occur
122
0 . R . TRISTRAM
TABLE X X I I Estimation of Serine and Threoninea Threonine Nb
Total hydrox- Periodate amino NHa-Nb Nb.c
Serine Nd
Threonine Nd
7
Edestin Control Horse hemoglobin Control Insulin Control pLactoglobulin Control
4.04 4.03 4.12 4.03
2.30 2.26 2.89 2.92
4.01
1.47
3.85 3.18 2.98
1.47 3.66 3.62
6.34 6.29 7.01 6.95 5.49 6.32 6.78 6.60
6.28 6.14 7.08 7.01 5.50 5.40 6.74 6.52
Uncorrected values, aa per cent total nitrogen. Rees (1946). c Columns 2 and 3. d Brand (1946). a
during the hydrolysis of the peptide bond. Thus the degree of accuracy suggested by the analysis of control mixtures is without doubt greater than that which is actually attained during the analysis of protein hydroly zates. Table XXIII illustrates what is generally regarded as the probable accuracy of the more bopular modern methods. The accuracy claimed *
TABLE X X I I I Accuracy of Modern Methods of Amino Acid Analysis as Suggested by Analysis of Control Miztures Method
Approx. accuracy ( f )
~~~
Electrodialysis: arginine, histidine, lysine (Macpherson, 1946) 3 Enzymatic methods (Gale, 1945, 1946). . . . . . . . . . . . . . . . . . . . 3 Periodate oxidation: serine, threonine (Rees, 1946). . . . . . . . . . 3 Chromatography (Tristram, 1946; Moore, Stein, 1948, 1949). 3-8, 3-5 Isotope dilution (Shemin, Foster, 1946). . . . . . . . . . . . . . . . . . . . 1-2 Microbiological assay a. various authors.. .................. 4-8 b. Stokes, Gunness (1945a). . . . . . . . . . . . 15 Ionic exchange (Wiltshire; unpub.). . . . . . . . . . . . . . . . . . . . . . . . 5
for the methods quoted in the table is based upon the analysis of ad hoc mixtures of amino acids, or, in the case of microbiological assay, upon the recovery of amino acids added to protein hydrolyzates and for isotope
AMINO ACID COMPOBITION OF PURIFIED PROTEIN8
123
dilution upon theoretical considerations. With regard to electrodialysis, periodate oxidation, ionic exchange, and partition chromatography, the accuracies are based upon long experience and are the result of a large number of estimations. As was stated above this degree of accuracy may not be attained in the analysis of hydrolyzates because it is impossible to carry out control experiments which simulate the liberation of the amino acid from peptide linkage. This means that, owing to the possibility of modification or destruction, it is as yet impossible to establish with certainty the exact relationship between the amino acid content of a hydrolyzate, as revealed by a controlled analysis, and the amino acid composition of the protein from which the hydrolyzate was prepared. It is my own experience that the analysis of various samples of a protein produces occasional values, reproducible in themselves, which do not conform with the generally accepted results. It has been usual to attribute such observations to untraceable technical errors, but it is rarely found that control analyses are subject to the same “error.” These abnormal errors may be due either to variations in the degree of modifications during hydrolysis or to differences in the protein preparations, many of which are now being shown to be electrophoretically polydisperse. 4. Modijication and Destruction of Amino Acids during Hydrolysis
Martin and Synge (1945) discussed the modification and destruction of amino acids under the conditions of hydrolysis and suggested that while certain species of amino acid are subject to greater modification than others it could not be assumed that any are immune. In general it is probably true to say that (1) only amino acids with reactive polar groups in the side chain will undergo decomposition, (2) modification of nonpolar acids will be limited to racemization or in extreme cases to deamination (modification of the a-carbon will no doubt occur more readily at the time of hydrolysis of the peptide bond), and (3) the nature and extent of the modification will depend upon the particular amino acid mixture, and, in the first instance, the reactivity of polar side chains may be influenced by their distribution in the peptide chain. There is therefore likely to be uncertainty as to the number of residues of any particular amino acid in a protein molecule when the number is at all large (Martin and Synge, 1945). These workers suggested that “the difficulty can be to some extent overcome by analyzing protein hydrolyzates prepared in different ways, and by employing a number of analytical procedures for the same amino acid. A number of concordant results obtained in this way gives confidence in the significance of the figure arrived at.” While such a procedure increases the confidence with which
124
0. R . TRISTRAM
amino acid data may be interprete,d, the over-all accuracy of analytical methods (see Table XXIII) does set an upper limit to the number of residues which may be established. If any protein contains n residues of a particular amino acid, it is clear that the accuracy required to state with certainty that there are n residues and not n k 1 will vary inversely with the number of residues (see Tristram, 1946). If it is assumed that in determining the number of residues the permitted variation was n k 0.4 residues, the amino acid must be estimated with an accuracy of f0.4/n. I n Table XXIV 0.4/n has been calculated for certain numbers of residues. From the data in this table it follows that only those proteins which contain a maximum of 24-32 residues of any one amino acid may be analyzed with that degree of accuracy required to establish the amino acid composition to the nearest unit. In general this limits analysis, the object of which is the establishment of amino acid stoichiometry, to those proteins with molecular weights of 40,000 or less. It does not follow that the analysis of proteins of high molecular weight is valueless, TABLE XXIV A m r a c y Required in Amino Acid Analysis when Estimating Number of Residues in Protein (Sub) Molecule
a In calculation of accuracy required for precise estimation of number of residues (i,e., n and not n - 1 or n l ) , it has been assumed that permitted variation would be n f 0.4, and accuracy therefore f 0 . 4 / n .
+
TABLE XXV Amino Acid Content of Crystalline Clostridium botulinum Type A Toxin"
-
Weight, g./lOO g. protein
Amino acid
7.74 1.17 1.86
Lysine Phenylalanine Tryptophan Leucine Glutamic acid Aspartic acidb Threonine Cysteine Cystine
10.3
15.57 20.10 8.49 0.27 0.53 I
a
b
Buehler et al. (1947). Quoted as asparagine.
I
Moles/9 X 106
Inferred accuracy (0.4/4
477 64 82 708 953 1370 642 20 40
0.08 0.6 0.5
0.06 0.04
0.03 0.06 2 1
AMINO ACID COMPOSITION O F PURIFIED PROTEIN6
125
but it does mean that such analyses have only a comparative value. Failure to appreciate this limitation may result in the inference of a far greater accuracy than that of which the methods are capable. This is demonstrated in a recent publication by Buehler et al. (1947), who carried out a very careful amino acid analysis of crystalline Clostridium botulinum type A toxin (M.W. 900,000) which they showed to be homogeneous electrophoretically, ultracentrifugally, and serologically. Certain of the results of the analysis are reported in Table XXV. In column 4 the inferred accuracy has been calculated using the convention suggested above (ie., +0.4/n). It is obvious that the methods used for analysis do not approach the inferred degree of accuracy.
VI. STRUCTURE OF PROTEINS AS REVEALED BY AMINO ACID ANALYSIS Attention has been drawn to various factors which limit and may reduce the precision of amino acid analysis. It is clear that amino acid analysis has fundamental value only when there are adequate grounds for assuming that the proteins concerned are single chemical substances. 1. DeJinition of Purity
No universal or dogmatic definition of protein purity is possible owing to the diverse nature and properties of proteins. Pirie (1940) suggested that an ideal definition would be that the preparation should contain particles which were identical in size, chemical composition, and physical properties; and that (where the protien has enzymatic, antigenic, or other properties) each particle should carry the full unmodified activity of the starting material. (See also Herriott, 1942; Shedlovsky, 1943.) Such a definition would have to be modified to include at least two amendments: (a) that the biological activity may or may not survive the isolation of the dried crystalline preparation, and ( b ) that the protein may exist in solution as an equilibrium mixture of particles of more than one size (cj. insulin). Further the above definition of protein purity is clearly limited to those proteins which exist in soluble form and does not include that wide and important group of proteins such as silk fibroin, keratins, and myosin, the last of which is known to be polydisperse. So far it is not possible to define criteria for the establishment of the purity of such proteins. 2. Determivnation of Purity of Soluble Proteins Pirie (1940), Shedlovsky (1943), and Chow (1944) stated that the most reliable means at present available for the establishment of protein purity are electrophoretic and ultracentrifugal studies and the estimation
126
Q. R. TRISTRAM
of solubility under conditions based upon the phase rule (see Kunitz and Northrop, 1930, 1938; Herriott, 1942). Recent experiments involving electrophoretic studies under different conditions of pH have suggested that certain proteins hitherto regarded as single substances may be polydisperse (cf. Longsworth et aZ., 194O-egg albumin; Li, 1946; Grijnwall, 1942; McMeekin et al., 1948; C . F. Jacobsen, personal communication, j3-lactoglobulin). These observations are of fundamental importance since they suggest that either the proteins are mixtures of chemical individuals or that electrophoretic polydispersity may be no more than an indication of small differences in the state of the polar groups of various molecules. For instance it is conceivable that in preparations of 0-lactoglobulin a certain fraction of the molecules contain fewer amidiaed carboxyl groups. At pH 8.5, when the protein carries a strong negative charge, such differences would be masked, but a t pH 4.8, a point of small charge, the differences should be significant and observable. The criteria suggested above for the establishment of protein purity do not include that of constant amino acid composition, but, with the development of accurate and specific methods of amino acid analysis it may soon be possible to include such a criterion, the adoption of which would presuppose that proteins were governed by the laws of organic chemistry. While it may be true that the physiological activities of proteins demand the presence of specific groupings, there is no evidence which would support the thesis that every unit is unchangeable in composition (i.e., alanine for glycine, etc.) or position. It seems somewhat previous therefore to attempt to define the criteria for the establishment of protein purity until many problems of structure have been solved, certain of which are listed below: (1) The extent to which all molecules in any protein preparation are replicas of one another. (a) The effect of changes in composition ( e . g . , alltnine for leucine, amide content, and glutamic acid for aspartic acid) upon the physical properties of the protein. And (3) the effect of a change in juxtaposition of polar and nonpolar groups upon the physical properties. 3. Purity of Protein Preparations Used i n Present-Day Analytical Work
In collecting data for this review it became apparent that the estimation of the purity of the protein preparations was not invariably checked by all, if any, of the criteria (no matter how inadequate these may be) described above. In general the purity of the protein preparations was defined in the following ways : (1) Electrophoretic and ultracentrifugal studies on a crystalline protein. ( 2 ) Total nitrogen, ash, and moisture
AMINO ACID COMPOSITION OF PURIFIED PROTEINS
127
content, generally quoted without any other data. And (3) preparation of a crystalline protein by means of a well-known method (e.g., b-lactoglobulin by the method of Palmer, 1934). It is perfectly obvious that the total nitrogen content etc. of a preparation is no guide to its purity; and it is now accepted that this is also true of crystallinity, although many workers still attach importance to the crystalline state. (See Pirie, 1940, for a discussion of several instances in which crystalline preparations have been shown to be contaminated with other proteins.) It has often been assumed that a protein, crystallized by a well-known method, and with the same nitrogen content reported by earlier authors, is similar in every detail to the original preparation. Thus Palmer (1934) prepared from milk the protein now known as @-lactoglobulin. This mode of preparation became standard, and @-lactoglobulin came to be regarded as an excellent example of a protein, pure by the criteria of electrophoretic mobility, sedimentation contant, and solubility. * Recent electrophoretic studies (Li, 1946; McMeekin et al., 1948) have indicated that, whereas 8-lactoglobulin appears t o be homogeneous a t pH 8, a t pH 4.8 there are a t least two distinct fractions (Li, 1946, reported the presence of three fractions a t p H 4.8). McMeekin et al. (1948) also showed that there was definite evidence of a fractionation of the two components by (a) fractional crystallization of the standard preparations from acetate buffers of varying pH, and (b) fractional crystallization of the mother liquors from which a subnormal amount of “standard” crystals had been taken. In the light of this evidence it is difficult to see how i t can be assuined that the components, of which there are a t least two, are always in the same proportion in milk, or that the use of an approximately standard method of preparation will yield a mixture of constant proportions. Although there is no evidence that the two components of @-lactoglobulin are chemically distinct the practice of using a well-known method of preparation without adequate control is far from satisfactory. 4. Reporting Total Amino Acid Analyses
In the investigation of the stoichiometric relationships of amino acids in a protein the reporting of data is somewhat more complex than that outlined previously (page 107). Firstly, the accuracy of the analyses must be indicated ( a ) by control experiments and (b) by comparison with the data reported by other workers. Secondly, it must be shown that the recovery of the constituents of the protein is complete. Thirdly, the
* Editors’ footnote-Concerning other methods of preparing 8-lactoglobulin, and for a further discussion of the points raised here, see the article on “ Milk Proteins” elsewhere in this volume.
128
a.
R. TRISTRAM
essential calculated data must be included to indicate the molecular structure and the minimum molecular weight of the protein. It is suggested that the data should be calculated in the following forms (cf. Brand and Edsall, 1947). (I) Amino acid nitrogen as per cent of protein nitrogen to demonstrate the completeness of the analysis. ((2) Weight of amino acid in 100 g. protein to form the basis of the calculations. (9)Gram moles of amino acid per lo4 g. or 106 g. protein. ( 4 ) Minimum molecular weight. (6) Residues of each amino acid per molecule. The following essential data should also accompany analyses: (I) Protein nitrogen. ( 2 ) Moisture, ash, and, where possible, the total sulfur. (3) Results indicating specificity of methods used. Control analyses, preferably carried out on ad hoc mixtures of amino acids. (4) Protein analyses-with indications of the corrections applied, if any, as a result of the control analyses. Certain proteins, salmine, insulin, P-lactoglobulin, horse hemoglobin, ovalbumin, and edestin, will now be discussed in detail, assuming for the purpose of discussion that each is chemically a single species.
5 . Molecular Weights
In the derivation of the molecular formulas a knowledge of the molecular weight is necessary and should be obtained by methods independent of the amino acid data. There is considerable disagreement in the TABLE XXVI Assumed Molecular Weights of Certain Proleans Protein
References Gutfreund, 1948b; Oncley, Ellenbogen'
Insulin Salmine @-Lactoglobulin
37,000
Ovalbumin Hemoglobin, horse Edestin
46 ,000 68,000 50 ,oow
-
Gutfreund, 1945; Ogston, 1948; Crowfoot, Riley, 1938. Taylor et al., 1932; Bull, 1941; Gutfreund, 1944. Adair, Adair (1934a,b);Adair (1927). Burk, Greenberg (1930).
In dilute solution at an acid pH . In acid solution. c In 6.6 A l urea. d Chemical data. 0
b
AMINO ACID COMPOSITION OF P U R I F I E D PROTEINS
129
values recorded for the molecular weights of the proteins to be discussed and, when there is agreement on the order of the molecular weight, a considerable degree of scatter from a mean value. The absence of absolute values for molecular weight further limits the precision with which the number of residues of an amino acid per molecule may be established. This limitation becomes more serious with increasing numbers of residues (see page 124). For instance, if &lactoglobulin is assumed to have a molecular weight of 37,000, the protein contains 29 residues of lysine and four of histidine. If a molecular weight of 39,000 is assumed, the number of residues is 30 and 4, respectively. For the purposes of discussion therefore the following molecular weights in Table XXVI have been assumed.
VII. AMINOACID COMPOSITION OF CERTAIN PROTEINS 1. Salmine
This interesting protamine with its singular composition (Table XXVII), containing as it does some 86% of arginine, has recently been submitted to complete amino acid analysis by Block and Bolling (1945a) and Tristram (194713). The two analyses are in close agreement and Block (personal communication) has re-estimated glycine and alanine, obtaining similar values to those quoted in Table XXVII. The amino acid data give no information on the distribution of the arginine and monoamino acids, but end group assay (Porter and Sanger, 1948) has indicated that the imino group of proline occupies a terminal position. This suggests that salmine is a normal peptide of some 57 residues. This is in direct contrast to the findings of Fraenkel-Conrat and Olcott (1947), who suggested, from titration and esterification experiments, that salmine was a cyclic peptide. Fraenkel-Conrat (personal communication) has stated that, in further experiments, the formol titration of salmine sulfate gave only its guanidyl content although with a mixture of methylguanidine and proline it was possible to titrate 80% of the proline. He suggested that the reaction mixture used by Porter and Sanger (1948) (ie., dinitrofluorobenzene in sodium bicarbonate, pH 8.5, at 20°C.) probably caused the hydrolysis of a labile prolinearginine bond. It seems probable that, if such a hydrolysis did occur at pH 8.5, it would also occur during the formol titration a t pH 9.0. Experiments by the writer using the formol titration have indicated that the salmine sulfate used by Tristram (1947) contains one amino or imino group per molecule of 8000. One possible explanation is that the salmine sulfate used by Fraenkel-Conrat (Columbia River salmon) and Tristram (1947) (Spring or Chum salmon) have a somewhat different molecular structure.
130
0.
R. TRISTRAM
2. Insulin
The amino acid composition of insulin (Table XXVIII) has been determined largely by two groups of workers, namely, Brand (1946) and coworkers, and the group working in the laboratory of Chibnall (Macpherson, 1946; Rees, 1946; Tristram, 1946). (Velick and Ronzini, 1948, have also published an analysis of this protein.) TABLE XXVII Amino Acid Composition of Salminea ~~
Amino acid Arginine Leucine Valine hline Glycine Alanine Brine Total
G. per 100 g. protein
G . moles per 1O.g. protein
Residues per molecule*
85.2 1.64 3.14 5.80 2.94 1.12 9.1 108.94O
49.0 1.25 2.69 5.05 3.96 1.26 8.63 71.92
40 1 2 4 3 1 7
6gd ~
~-
Derived Data Mean residue weight N distribution!. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137.5 analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134.5 Total number residues#. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 Polar groups, %. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81.3 (70 % arginine) 20.0 Non polar groups, %. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hydroxylgroups % . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.3 End groups (Porter, Sanger, 1948). . . . . . . . . . . . . . . . . . . . . . . . . > 1 (Proline) H80,content of protamine sulfate (19.85%). . . . . . . . . . . . . . . 40 groups Spring salmon. Total N = 30.7 % (free base). Tristram (1947b). Assumed. 99.92 % of total protamine N. Apparent M.W. from amino acid composition 7870 (7460-8260). Block and Bolling (1946) reported that salmine from Columbia River salmon contains isoleucine only. Chibnall (1942). M.W./mean residue weight. 0
b
-
Certain differences exist between the two groups of analyses (see Table XXIX). While reinvestigation of certain values may be necessary there aeem to be good grounds for accepting the following values: arginine 3.07 % (Macpherson, 1946; Gale, 1946; Chibnall, unpublished experi-
131
AMINO ACID COMPOSITION OF PURIFIED PROTEINS
TABLE XXVIII Amino Acid Composition of Insulin, Edestin, and fl-Ldctoplobulin* Amino acid
--
Glycine Alanine Valine Leucine Isoleucine Proline Phen ylalanine Half-cystine Cysteine Arginine Hietidine Lysine Aspartic acid. Glutamic' acid Amide NHI
4.3 4 5 7 75 13 2 2 77 2 53 8 14 12 5
-
3 4 2 6 18 1 5 2 13
Berine
Threonine Tyronine Tryptophan Methionine - .
l 2
07 91 51 8 60 69 23 08 03
-
57.4 505 66 2 100 9 21 1 22 0 49 3 104 0
-
-
17 6 31 7 17 2 511 125 1 (96 8 49 9 17 5 72 0
2 4 2 6 15 (12) 6 2 9
-
-
--
11s 601 853 6
Tofa2
6 8 12 3 3 6 12
'040
-
Edestin'
3 1 -1 7 _ .
__
2
_
4.31 48.4 5.7 48.8 36.0 4.7 57.3 7.5 4.25 37.0 33.0 5.45 7.75 0.93 4.15 0.50 96.0 16.7 2.9 18.7 16.45 2.4 12.0 90.2 140.9 20.7 2.15 (126.4) 60.0 6.3 32.4 3.85 24.0 4.34 1.48 7.25 2.4
1 2 -
3
-
24 24 18 29 18 16 4 2 48 9 8 45 70 (63) 30 16 12 3
-
20.00 72.00 48.90 118.00 45.00 46.9 23.90 19.10 9.16 16.45 10.20 77.50 85.90 146.50 (76.5) 38.70 43.40 20.60 9.32 21.50
1__ 16.75' 866.05
108.6h
--
1.50 6.42 5.72 15.46 5.90 5.4 3.95 2.29 1.10 2.86 1.58 11.30 11.40 21.50 1.30 4.07 5.15 3.73 1.90 3.20
Column No. 1 provisional value. g. amino acid per 100 g. protein. Column No. 2 = amino mid per 10' g. protein. Column No. 3 apparent residues per molecule. 15.54 %. M.W. (submolecule) 12.000. Total 8 Total N 3.33 %. Total N 18.55%. M.W. 50.000 units. 8 0.88%. dTotal N 15.60%. M.W. 37,000. 'For Iactodobulin, see Table XXXII. 97.8 % of total protein N. s Apparent M.W. of aubrnolecule from amino aoid oompoaition 12.200 (11,300-13,600). * 94 % of t o b l protein N. 98 % of total protein N. f Apparent M.W. from amino acid data 37.200 (35,000-38,800). a
--
--
-
3
~
7 27 18 44 17 (17) 9 7 3.5 6 4 29 32 55 (28) 14 16 8 3.5 8 3161 _3
g.
moles
-
Derived data Mean renidue weight N diatribution (Chihnall, 1942) Analytical data Total number reaiduee' Groups, % of total groupe Polar groups Cationic Free Anionic Told Ionic Amide Phenolic Hydroxyl Tdal Nonpolar groupe LipophilicTold End groupen
-
Inaulin
&Lactonlobulin
Edsstin
114 I12 I06
117.4 119.6 422
113.6 116.4 328
7.8 9.3 17.1 11.3 8.4 7.9 44.7
17.0 13.5 50.6 16.3 3.1 11.9 65.6'
12.3 18.5
48.5
36.8
...
66.5
-
L (2 glycine. 2 phenylalanine)
M.W./mean residue weight. Including cysteine and tryptophan. AU nonpolar hydrooarbon aide chainn. " See Table 111.
1 (glycine)
P
S0.8
9.0 2.4 9.7 64.2' 45.6 qS.0
-
3 (leucine)
-
182
0. R. TBISTRAM
ments-see Table XIX), cysteine 0, aspartic acid 6.8% (Brand, 1946) (Wiltshire, personal communication, found 7.5 % by ionic exchange), glutamic acid 18.6% (Chibnall, 1946), amide nitrogen 1.69% (Rees, 1946; Harington and Mead, 1936), serine and threonine, 5.23% and 2.08% (Rees, 1946). The higher amide ammonia content reported by TABLE XXIX Variations in Amino Acid Composition of Insulin as Found by Various Workers Brand (1946) Cambridge group
I
I
Amino acid
Arginine Cysteine Aspartic acid Glutamic acid Amide NH, Serine Threonine Total protein N a
G./100 g. protein
Itesidues per submolecule (12,000)
3 0.5 6 16 15 7 3
3.5 0.6 6.8 20.2 2.15 5.8 3.2 1( 04
G.’lOOg’ protein
Residues per submolecule (12,000)
3.07 0 7.5”
2
7 15 12 6 2
18.6 1.69 5.23 2.08 I .54
G . Wiltshire, unpublished.
Brand is difficult to explain since it is in excess of the values reported by Rees (1946) and by Harington and Mead (1936), and also of the total ammonia found by Rees (1946) and Linderstrom-Lang and Jacobsen (1940) after hydrolysis with 6 N hydrochloric acid for 24 hours a t 105OC. IL seems possible th at the higher amide and total protein nitrogen are correlated (d.Scott and Fisher, 1935). Brand used a sample of insulin prepared by du Vigneaud and the solvent used for crystallization is not known. The work of Scott and Fisher indicated that insulin recrystallized from ammonium acetate did appear to have a somewhat higher total nitrogen. Structure of Insulin. Evidencc has been presented (Gutfreund, 194813; Oncley and Ellenbogen, unpublished dat,a) which suggests that the basic molecule of insulin has u molecular weight of 12,000 and that the higher molecular weiglits which have been recorded are an indication of the degree of association of the protein (a function of pH and concentration), the maximum polymer, according to Gutfreund, being one which contains four “submoleculcs.” Wrinch suggested th at the insulin molecule was a single polyhedron,
AMINO ACID COMPOSITION OF PURIFIED PROTEINS
133
a view not supported by the work of Crowfoot (1938), which was interpreted (Bernal 1939) as indicating the presence of 12, 18, or 24 peaks, depending upon the molecular weight of the protein. The protein may therefore contain 12, 18, or 24 chains within its molecule, each being linked through R groups (e.g. cystine). Crowfoot (1938) also suggested that the amino acid residues in each subgroup are so linked in the chain that they cover the surface of a cube, and, from unit cell measurements, gave the molecular weight as 36,000. This may well be due t o a n association of three molecules of 12,000 to form the crystal which has trigonal symmetry. Amino acid analysis gives a mean residue weight of 114, SO that there are 106 or 318 residues in molecules of weight 12,000 or 36,000, respectively. If therefore, the interpretation of Bernal is correct, and molecular weight evidence certainly is in agreement with his interpretation, a molecule of weight 12,000 might be expected to contain four chains with a n average of 25-27 residues each. Chibnall (1946) showed that the amino nitrogen of the intact protein (van Slyke determination) was equivalent to seven amino groups in 12,000 (Brand, 1946, reported 21 terminal amino groups in 46,000 or 7.6 per 12,000), two being due to c lysine amino. Sanger (1945) showed (Table 111) that the true figure is six (dinitrophenyl derivatives), two being glycine and two phenylalanine, with two due to e-amino group of lysine. The presence of glycine in a terminal amino position explains the higher van Slyke values, since Schmidt (1929) showed that glycylglycine gives 135% of the true amino nitrogen in 10 minutes. The findings of Sanger (1945) are in agreement with the titration data of Harington and Neuberger (1936), LinderstrZm-Lang and Jacobsen (1940), and Cohn (see Edsall, 1946), which also indicate the presence of four carboxyl groups (per 12,000) in excess of the dicarboxylic acids found by analysis. Chibnall (1946) suggested that the presence of four chains in the molecule (12,000) was a reasonable working hypothesis. It can be seen (Table XXVIII) that the four chains must be dissimilar because there are only three prolyl, two argininyl, two lysyl, two threonyl and (probably) three isoleucyl residues; and few of the remaining residues are multiples of four. The over-all asymmetry of the chains has now been demonstrated by Sanger (1947, 1948), who has isolated a fragment containing a terminal glycyl amino group containing about 25 residues. The mode by which the chains are linked together is as yet unknown. There are twelve residues of half cystine, since insulin contains no sulfhydryl groups (Bailey, unpublished; Olcott, and Fraenkel-Conrat, 1947) so that the molecule may be made up of four chains linked b y dsulfide bridges (although there need not necessarily be a n even distribution of disulfide groups).
134
0 . R. TRISTRAM
3. Edeetin
Edestin (see Table XXVIII) has a molecular weight in native form of 309,000 (Svedberg and Pedersen, 1940). The molecule undergoes depolymerization in urea into units with a molecular weight of 50,000 (Burk and Greenberg, 1930). Although the submolecule has too great a molecular weight to allow all the residues to be estimated with the required precision (see Table XXIV) the analyses reveal several interesting features especially when the denaturation of edestin into edestan (Osborne, 1902; Adair and Adair, 1934; Bailey, 1942b) and the end group assay of Porter and Sanger (1947) are taken into account. Osborne found that the addition of acid to salt solution of edestin caused an irreversible change into a derivative which he called edestan. Adair and Adair (1934) showed that the mean molecular weight of this compound was 17,000. The mechanism of the formatiou of edestan has been investigated by Bailey (1942b), who showed that the change was accompanied by the production of small and variable amounts of nonprotein nitrogen which contained tryptophan and ammonia and other unidentified compounds. The formation of edestan was also accompanied by the liberation of sulfhydryl groups in much the same amounts as when edestin is denatured in urea (Greenstein, 1939) indicating fission of the peptide chain. Chibnall (1942) suggested that, the relationship between analytical and titration data indicated that the submolecule (50,000) consisted of one long peptide chain, and that the presence of about four a-amino groups, demonstrated by the Van Slyke determination on the intact protein, was in keeping with the formation of edestan, under the conditions of the Van Slyke determination, and the liberation of two additional amino groups. Sanger and Porter (1947) have now shown that the edestin submolecule has glycine as the sole terminal group. A small amount of leucine remains to be explained but it may be an indication of the presence of free amino acids adsorbed on the protein. The finding of free amino acids by Bailey (1942b) may be due to their adsorption on the protein, and liberation when edestin is converted to edestan, rather than their production on fission of the larger molecule. A reinvestigation of this subject using the modern methods for the microestimation and detection of amino acids would provide valuable additional information. Consideration of the analytical data suggests ( a ) that the fragments of molecular weight 50,000 are not identical since there are less than six residues of tryptophan (two) and many of the others are far from being multiples of six, and ( b ) that the fragments which make up edestan must also be dissimilar since few of the residues are multiples of eighteen.
AMINO ACID COMPOSITION OF P U R I F I E D PROTEINS
135
The data submitted by Chibnall (1942) and Sanger and Porter (1947) seem t o offer substantial proof of the single chain structure of the edestin submolecule. Analysis suggests that there is a maximum of two groups of sulfhydryl per molecule (50,000), and four groups of half cystine (two disulfide groups). If disulfide actually pre-esists in the protein molecule, in edestin it must be concerned in intramolecular linkages uniting various sections of the chain. 4. 8-Lactoglobulin Kekwick and MacFarlane (1943) considered that this protein “approaches most closely to the ideal protein, since its solubility is almost independent of solid phase (SZrensen and Palmer, 1938) and it is homogeneous in the ultracentrifuge and electrophoresis apparatus (Pedersen, 1936).” Modern views on this protein (based upon the findings of Gronwall; 1942, Li, 1946; McMeekin el al., 1948; and Jacobsen, personal communication) are not in accord with the above view, although the recent findings have not demonstrated that 8-lactoglobulin is polydisperse in the sense that the components are chemically distinct. They do complicate the interpretation of analytical data because it is difficult to decide whether the minor variations in analytical data are due to variations in accuracy or t o varying proportions of chemically distinct components in the samples of 8-lactoglobulin used for analysis. The analytical data do illustrate the over-all structure of this protein, although uncertainty about the cystine-cysteine distribution and the free carboxyl groups renders the intramolecular structure somewhat obscure. a. Cysline-Cysleine. Brand and Kassell (1942b) found by analysis of an acid hydrolyzate 1.10%cysteine and 2.29% cystine. These values are equivalent to 3.5 residues each of sulfhydryl and disulfide per molecule of molecular weight 37,000. They may be affected by the recent findings of Olcott and Fraenkel-Conrat (1947), and so far as the author is aware, no estimation of sulfhydryl in the intact protein has been carried out. b. Free Carboxyl Groups and Type of Molecule. Chibnall (1942) suggested that the protein might contain eight or nine chains by virtue of the differences between the basic groups found by analysis of the hydroly?;ate and titration of the intact protein, and by the determination of the free amino nitrogen in the intact protein. More recent analyses have shown that the lysine content is 11.4% (cf. 9.75%) or 29 instead of 25 groups per molecule, so that there are 39 basic groups. The titration data of Cannan et al. (1942) indicate the presence of 42-43 titratable cationic groups in the intact protein, suggesting that there are three or four terminal amino groups within the molecule of P-lactoglobulin.
136
G. R. TRISTRAM
Porter (1948) has found, by means of the fluorodinitrobenzene reagent of Sanger (1945), th at p-lactoglobulin contains three terminal leucine groups (Table 111). I n the case of insulin i t was shown that there were good grounds for assuming the presence of terminal carboxyl groups equivalent in number t o the terminal amino groups found by Sanger (1945). With P-lactoTABLE XXX Anionic Groups of &Lactoglobulinn (Equivalents per molecule) Acid
Worker
Method
Cannan et al. (1942)
Total Aspartic acid
Titration curve of protein Isotope dilution, microbial assay Ionic exchange Isotope dilution, microbial assay Isolation Ionic exrhmgc
Aspartic acid Wiltshire (unpub.) Glut,amic acid Gliitamic acid Bailey et al. (1943) Glutamir acid Wiltshire (unpub.) Itees (1946), Brand (1946) Amide 1
a
.~
'
-
Equivalents per molecule. Possible free anionic groups: A1 + G I - A m = 52, A ,
~
+G2
-
Am
=
5!1
globulin however, this cannot be assumed because there is considerable doubt about the dicarboxylic acids found by analysis and titration. In Table XXX all the known data are presented as the number of residues of aspartic acid, glutamic acid, and amide nitrogen found by analysis, and the free carboxyl groups found by titration of the intact protein. Although the evidence of Porter (1948) is that 8-lactoglobulin contains three intramolecular chains it is by no means certain that the terminal amino groups are balanced by a corresponding number of carboxyl groups. Analytical evidence for the glutamic acid content is based upon microbiological assay (48 groups), which is known to be specific for the L isomer, and on ionic exchange (Wiltshire, using a n elaboration of the method of Consden, Gordon, and Martin, 1948) and isolation (55 groups). The latter methods are not specific for the L isomer and thus may include any D-amino acid. (Chibnall, 1946, suggested that acid hydrolyzates of &lactoglobulin contain D-glutamic acid, although this has not been shown to be a constituent of the native protein.) If the higher value of 59 free anionic groups is correct it would suggest th a t the titration of free carboxyl groups was incomplete. It seems inconceivable that as
137
AMINO ACID COMPOSITION OF PURIFIED PROTEINS
TABLE XXXI Amino Acid Composition of Ovalbumin, Horae Hemoglobin, and Horse Myoglobina
-
Ovalbuminb Amino acid 1
-
-
2
3
-
Horse myoglobind
Horae hemoglobin. 1
___
3
2
5.60 74.8 48 40.6 19 3.05 64 25 6.72 75.5 7.40 83.2 9.10 77.8 28 7.05 60.2 50 70.2 32 9.2 76 15.40 117.6 0 0 0 ) 53.5 25 7.0 33.9 3.0 22 31.3 14 3.60 30 21 7.66 46.4 7.70 46.6 2 3.75 0.45 0.51 4.25 4.64 1.35 11.2 0.56 6 32.9 3.65 21.0 14 5.72 16 15.15 2.36 8.71 56.2 36 7 6.30 43.2 38 8.51 68.3 20 70.0 10.6 79.9 51 0.30 32 16.50 112.2 8.5 57.0 38 52 1.23 (72.3) (33) 1.1 (56.0) (36) 33.0 5.8 65.0 36 8.16 36 4.03 34.0 4.36 36.6 24 16 3.68 20.3 3.03 11 16.7 9 1.20 1.70 2.1 8.3 6 5.80 5.20 35.0 1.0 6.7 4.5 16 108.8' 796.7 106.87 898.8 977 643
Glycine A1anin e Valine Leucine Iaoleucine Proline Phenylalanine Half-cyatine Cyateine Arginine Histidine Lysine Aspartio m i d Glutamio soid Amide NHi Serine Threonine Tyrosine Tryptophan Methionine Total
i'51
-
- -amino acid per 100 provisional value, 7
1
2
-
78.0 89.4 34.9
13 16
16.80 128.2
22
29.0 30.8
5 6
-
0
6.86 7.95 4.09 3 34 5.09 0
--
-
Derived dutn
I
*
M.W./mean residue weight. Including cyateine and tryptophan. Including tryptophan. C/. Table 111.
6
2 9 18 10 19 (8) 6
7 2 2 2
- __ Column No. 2 108.481 846.3 ~.
148
= g. moles
-
-
Mean residue weight N distribution Analytical data Total number of residues* Groupa aa percentage of total groupa Polar groups Cationic Free Anionic Total i o n i c Amide Phenolic Hydroxyl Total Nonpolar groups Lipophilic Total End groups'
3
2.2 12.65 54.90 8.5 15.5 106.2 8.2 61.7 16.48 112.0 0.8 47.2 3.46 33.0 38.3 4.56 13.3 2.40 11.5 2.34 1.71 11.5
g. g. protein. Column No. = amino aeid per 10' g. protein. Column No. 3 = apparent reaidues per molecule. b Total N 15.76 %: 46.000. .. M.W. 0 Total N 16.8%; M.W. 68,000;total 5 0.0140%. 4 Total N 16.90%; M.W. ca. 17,000;total 5 = 0.37%. 96.8 % of total protein N. I07.3 % of total protein N (including heme N). 0 98.02% of total protein N (including heme N). a
-
Horae myoglobin
Ovalbumin I
114.2 116.0
112.6 108.1
toa
680
11.6 13.9 SS.4 9.1 2.6 8.5 47.6'
16.2 0.5 86.7 6.7 2.0 10.8 46.8'
47.6
43.0
66.6 nil
69.1 6 Valine
112.0 110.0 146 10.8 14.4 94.6
5.5 1.4 8.9 61.4i
39.0 48.0 1 Glycine
138
G. R. TRISTRAM
many as seven groups would remain untitrated although the carboxyl groups may be unreactive in much the same may as the amino groups are unreactive with certain reagents (Porter, 1948). It is clear therefore that a reinvestigation of the carboxyl groups in the intact protein and of the dicarboxylic acid content of the protein is necessary, and that such a study must be correlated with an investigation of anionic groups in the components indicated by electrophoretic studies. 5. Oualbumin
It has been shown by electrophoretic studies (Longsworth, Cannan, conand MacInnes, 1940) that this crystalline protein (Table =XI) tains two fractions with distinct electrophoretic mobilities. Cannan (private communication to Chibnall) estimated that the amount of the second component varied widely from sample to sample (10-77%). It has not been possible to achieve any fractionation by repeated crystallization or by partial acid or heat denaturation, and it would appear that the two components are very similar in composition. Amino Acid Data. Ionic Groups. Cannan et al. (1941) (see also Kekwick and Cannan, 1936) found, by titrationof the intact protein about 42 cationic and 50-51 anionic groups. The most recent analytical data suggest the presence of 42 cationic groups. There is considerable difference in the number of anionic groups found by various workers and further work would seem to be necessary (Table XXXII). TABLE XXXII Dicarboxylic Acids of Ovalbumin
Acid
Aspartic acid Aspartic acid Glutamic acid Glutamic acid Glutamic acid Glutamic acid Glutamic acid Glutamic acid Amide nitrogen
Worker
Wiltshire (unpub.) Smith, Greene (1947) Hac, Snell (1945) Chibnall el al. (1943b) Kibrick, 1944 Hac, Snell (1945) Lewis, Olcott (1945) Lyman et al. (1945) Wiltshire (unpub.) Various
Possible free anionic groups: AS +Gn A l + G , - A m = 53.
Method
Ionic exchange Microbial assay Microbial assay Isolation, ionic exchange Isolation, ionic exchange Microbial assay Ionic exchange Various
- Am
= 44; A2
+GI
- Am = 51;
AMINO ACID COMPOSITION OF PURIFIED PROTEINS
139
From the above data it is evident that the ovalbumin molecule cannot contain more than a single chain. The presence of the carbohydrate moiety (Neuberger, 1938) part of which is an acetylated amino sugar, makes the interpretation of the titration data somewhat complex. That conditions may be unique was suggested by Porter (in press), who could find no evidence for the presence of any terminal amino groups (cf. Sanger, 1945) in this protein. peptide chain CH-NH-CO-CH-NH ,,
I
I I
I I
peptide chain NH-CH-CO
'
6. Horse Hernoglobulin
This protein (Table XXXI), although the molecule is somewhat too large to permit accurate stoichiometric analysis (see Table XXIV), has been the subject of extensive physical and analytical investigation. The amino acid composition reveals the basic nature of the globin fraction. This protein is remarkable in that it apparently contains no isoleucine. According to Brand and Grantham (1946) the same is true also of adult human and bovine hemoglobin, although they find dog hemoglobin to contain seven isoleucine residues per molecule of protein, and fetal bovine hemoglobin to contain approximately three. They have also found marked species variations in methionine content, the extremes being two residues per molecule in dog and eight in adult bovine hemoglobin. The distribution of cystine-cysteine in horse hemoglobin is still somewhat uncertain. Kuhn et al. (1939) found 1.01% of cysteinecystine in the globin portion of the protein ; Rees (unpublished experiments) found 0.72% cyst(e)ine by differential oxidation. Direct titration of cysteine in the denatured globin was carried out by Mirsky and Anson (1936) and Greenstein (1939), who found 0.42% and 0.56% of cysteine, respectively. Construction of Molecule (&I. W . 66,000). Chibnall (1942) computed from a determination of a-amino nitrogen (Van Slyke) on the intact protein an excess of about sixteen terminal a-amino-groups. This was based upon a lysine value of 37 residues, compared to a more recent value of 38 residues. Porter and Sanger (1948) have shown that horse hemoglobin contains six valine residues per molecule in the terminal amino position. If it is assumed that all the nonmethionine sulfur is in the form of cystine, there may be a maximum of three disulfide groups
140
Q. R. TRISTRAM
TABLE XXXIII Amino Acid Distribution of Some Muscle Proteinsa I
Myosinb
Tropomyosinb
Substance 1
2
1
2111p
1
2
---
-- - Total N Total S Arginine Histidine Lysine Glutamic acid Aapartic acid Amide NH, Glycine A 1anine Valine Leucine Isoleucine Phenylalanine Tryptophan Tryosine Prolins Serine Threonine Half oyetine Cysteine Methionine Total
-
16.7 1.10 7.36 42.3 2.41 16.6 11.92 81.4 22.1 150.3 66.9 8.9 1.45 (86.7 25.3 1.9 73.0 6.5 22.1 2.6
-
16.7
-
7.80 0.86 15.70 32.90 9.10 1.1 '
44.8 5.5 07.4 23.6 68.4 63.6)
(0.4) 8.8 3.13
98.8 26.7
-
16.8
-
-
{
16.60
19.0
26.2 3.9 18.8 16.7 41.2 42.9 11.7
4.60
27.9 0 17.2 11.3 41.7 24.4 6.3
3.4
22.8
2.80
18.8
1.17
109. 08 780.0
19. 78
41.8
16.60
-
-
1
119.0
3.10 1.30 4.38 2.90 0.76
-
16.4
6.33 36.4 6.23 30.1 4.21 27.2 5.01 32.3 9.54 65.4 9.42 64.6 11.40 77.7 4.78 32.6 9.70 7 3 . 0 12.40 93.4 (1.10 (64.7) (1.15) (67.7) 6.61 74.9 6.09 81.1 8.66 96.3 6.72 75.4 7.40 6 3 . 2 12.0 102.7 11.50 ji48.o{ ::::)lizi.o 7.90 18.5 6.55 33.7 3.06 11.3 2.31 2.05 10.0 29.3 5.31 4.57 26.2 49.7 5.71 3.67 32.0 69.6 7.30 7.47. 71.1 62.9 7.47' 7.26' 61.1 9.32 1.12 1.09 9.09
4.3 0.8 3.4 1.9 4.33 6.11 1.4
115.6
-
7.86
o€.o.s
2.70 111.8
18.10 809.4
Derived data M.W. Average residue weight N distribution Analysis Total reniduea/lOI g. Group distribution. % Polar groups Cationic Free Anionic Total ionic Amide Phenolic Hydroxyl Total Non polar groups Lipophilic Total
-
101 (in urea)
80,000
116.8 115.3 866
115.8 117.2 859
107.5 912
16.1 18.W
18.4 26.6
14.2 9.4
14.0 6.4
94.1
46.0
C3.6
C0.4
9.9 2.2 9.7 67.7
7.4 2.0 7.7
7.1 3.2 14.5
7.6 2.8 14.6
8C. 8
60.6
47.4
32.4
35.2
42.0
42.3
60.4
61.1
$6.)
-
140.000
160,000
906
Column No. 1-g. amino acid per 100 g. protein. Column No. 2-g. moled per 10' g. protein. Bailey (1948). a Velick. Ronsoni (1948). 4 Found by Duboisson, Hamoir (1943) from titration data. Corrected for hydrolysis losses (el. Reea, 1946). b
AMINO ACID COMPOSITION OF PURIFIED PROTEINS
141
per molecule; that this figure is too high is suggested by the fact that sulfhydryl groups are known to be present (Mirsky and Anson, 1936; Greenstein, 1939). It is clear that a complete reinvestigation of the sulfur (disulfide and sulfhydryl) content of hemoglobin is necessary because the six chains, in two sets of three chains (2 X 34,000), must be linked by covalent linkages, otherwise denaturation might be expected to yield smaller fragments than those of 34,000. The site of attachment of the four heme molecules is not yet known with certainty, although the work of Porter and Sanger (1948) suggests that they are not attached through any of the CY- or c-amino groups. Wyman (1948) has reviewed the evidence on this point in detail, and favors the view that the imidazole groups of certain histidine residues are involved. Certain collected analyses are reported in Tables XXXIII-XXXVI without discussion since most of the proteins concerned are of high molecular weight and it is not possible to establish their purity. VIII. GENERAL DISCUSSION Vickery (1946) stated that: “ I n recent years there has arisen the conviction that not only the chemical properties of the proteins, but also many, if not all, of the physical properties can be assigned a rational explanation in terms of the amino acid composition, if this is sufficiently well-known and can be adequately interpreted. In addition when a protein has enzymological, immunological, or hormonal properties, the amino acid composition may provide the basis of the function.” Amino acid analyses may now be regarded as sufficiently accurate and specific to provide an over-all picture of the structure of simple proteins with molecular weights not exceeding 35,000-50,000 and containing not more than thirty residues of any one amino acid. As yet, however, the amount of any one amino acid has little significance in the interpretation of structure and biogenetic relationships (Bailey, 1948); and it is probably more profitable, therefore (Bailey, 1944; Astbury, 1942; see Bailey, 1948), to compare whole groups of amino acids expressed as a percentage of the total residues in order t o characterize the protein in terms of types of side chains. The amino acid analysis of proteins of large molecular weight is still insufficiently accurate to give precise figures for the number of amino acid residues of each type in a protein molecule. However, the over-all proportions of the residues of different classes-cations, free anions, amide, hydroxyl, and other polar groups, and nonpolar residues-differs strikingly from one type of protein to another. An over-all picture of the general character of the molecule in terms of the distribution of these
142
Q. R. TRISTRAM
TABLE XXXJV Amino Acad Composition of Human Plasma Proteins" Substance Total N Amide NHS Glycine Alanine Valine Leucine Isoleucine Proline Phenylalanine Cysteine Half-cystine Methionine Tryptophan Arginine Histidine Lysine Aspartic acid Glutamic acid Serine Threonine Tyrosine Total Average residue weight N distribution Analytical data Residues/106 g. proteind Group distribution, % Polar groups Cationic Free Anionic Total ionic Amide Phenolic H ydroxyl Total. Nonpolar groups Lipophilic Total
-Column-No.
E ibri gen*mo
Albr iinb 2 1 1 2 17.1 16.9 15.95 (94.1) 1.07 (63) 74.7 21.4 1.6 5.6 41.5 3.7 35.4 4 1 66.0 7.7 54.2 7.1 11.0 84.0 36.7 13.0 4.8 1.7 49.5 44.4 5.1 5.7 27.9 4.6 7.8 47.3 3.3 0.4 0.7 5.8 19.0 2.3 5.6 46.4 2.6 17.1 1.3 8.7 16.0 0.2 3.3 1.o 44.8 6.2 7.8 35.7 17.1 22.6 2.6 3.5 84.3 12.3 63.0 9.2 13.1 10.4 98.5 78.2 17.4 98.7 14 5 118.6 66.3 7.0 35.3 3.7 51.7 6.10 42.0 5.0 26.0 30.2 5.5 4.7 780,7 846.6 106.9 110.0 Derived da 115.9 112.5 875
14.3 11 8 26 1 10 8 3 5 13 5 60 1 32.2 40.7
119.2 117.5 846
~~
y-Globulinb 1 2 16.03 1 . 3 5 (79.5) 4.2 56.0
-
-
9.7 9.3 2.7 8.1 4.6 0.7 2.4 1.1 2.9 4.8 2.5 8.1 8.8 11.8 11.4 8.4 6.8 108.3
83.0 71.0 20.6 70.5 27.9 5.8 19.9 7.39 14.2 27.6 16.1 55.5 66.2 80.4 108.8 70.6 37.6 839.1
113.5 110.8 890
16.9 15.8 32.7 7.45 3.07 9.1 63.0
11.1 7.5 18.6 8.9 4.2 20.2 64.2
36.6
30.6 40 0
39.8
P P
4 1 = g. amino acid per 100 g. protein. Column No. 2 = g. moles per 1 0 6 g. protein. From Brand, Edsall (1947), (see also Brand, 1946). From Bailey (1944). d Approximate. * Including tryptophan and cysteine.
TABLE XXXV Amino Acid Distribution a Miscellaneous Proteins-
-
Amino Acid
-
18.7
2
1
Total N Arginine Hiatidine Lysine Glutamic acid Aspartic acid Amide NHs Glycine Alanine Valine Leucine Isoleucine Phenylalanine Proline Tryptophan Tyrosine brine Threonine Cystine Cyteine Hydrox yproline Hydroxylysine Methionine Total Average residue Weirht ..__OI_
N dmtribution Analysis ReaidueajlO'g. Group dmtnbution Polar Erollnl .._ .
CatLi7 Free anionic T d Amide Phenolic Hydroxyl Totul Nonpolar groups Lipophilic Told
Wool keratind
Silk fibroinc
-
1.1 0.36 0.68 2.16 2.76
6.33 2.32 4.66 14.7 20.75
43.6 29.7 3.6 0.91
581.0 334.0 30.8 6.95 8.40 20.4 6.43
-
-
1.1
3.36 0.74
-
-
70.7 154.3 13.45
12.8 16.2 1.6
...... ...... ......
.......
-
-
1176 d
}
....... .......
......
110.6
I
1
;:7";
3.8 0.0 5.55 13.17 22.50 32.0 77.6
59 6 18 95 49
I
1
70 79
9 4 (83 2) 87 0 46 4 39 7 863 22 1 82 6 8 8 25 7 95 4 53 9 98 9
I -
-
116 107 890
1
49.4 4.77 30.7 77.0 47.3 (7.15) 363.0 106.9 29.1 42.8 15.2 131.3
{
.......
......
.......
......
.......
110.8 7.4
14.0 1.1 0.8
107 6.8 5.4 1073.47
14.5 1.2 0.9
5.5 32.1 19.2
1.0 3.37 2.28
117.3
1st-
90.5 90.5 1100
91.2
9.5 2.9 17.0 47.3
7.75 10.3 18.06 0.5 0.52 14.5 33.6
43.0 61.7
29.4 61.0
;:; 16.9
-
......
5.5 30.3 18.5
7 7 ... 10.7 18.4 0.69 0.53 14.4 34.0
-
2^
1
.......
882
I
18.6 8.59 0.74 4.47 11.3 6.3 0.1 27.2 9.5 3.4 5.6 2.5 15.1
16.29
2
1 -
Gliadinf
ZeinJ
Collagen'J
......
1.0 3.18 2.20
4 7'
2
49.2 4.71 31.5 76.1 42.2 (5.3) 359.0 104.6 28.2 40.0 15.5 128.9
8.55 0.73 4.60 11.2 5.6 0.09 26.9 9.3 3.3 5.23 2.55 14.80
90
......
0.7 111.5
UI. 1300
1
- 2 1 18.0
......
.......
76.6 76.5
16.3 10.39 1.05 2.76 14.1 6.57 1.42 6.53 4.14 4.64 11.3 3.65 9.5 1.80 4.65 10.01 6.42 11.88
P
Gelatin!
394 4.06 708 920
64 203 82 672 374
642 20 40
..........
1
..........
1.06 116.75
7754 64
116.9
17.7 8.7
1S.l 63.0
-
2 -17.66 1
16 2 1.71 1.32 0 26.9 4.61 3.61
9.85 8.52 0 183.0 34.7 12121
2.74 1.82 0.65 45.7 1.34 5.45
10.5 3.52 122.6 5.91 10.53 0.12 5.25 7.05 3.45 0.83
118.0 30.1 172.0 35.80 91.5 0.55 29.0 67.1 29.0 6.9
2.13 2.66 11.90
...... ......
...... ......
............
16.2 85s.1
1.69 103.8
239 4.62 1.03 60 477 7.74 953 15.57 1370 20.10 12.131 1.38 113701 166 3.92 5.29 10.3 11.94 1.17 2.60 1.86 13.5 4.36 8.49 0.27 0.53
2
1
-
1
2.41 106.61
-
111.8 110.0
1~
-
6.44 13.35 0.6 3.2 4.9 2.1
I
-
23.9 22.7 91.0 39.0 116.0 2.94 17.7 46.7 17.68
>
sU
11.35 769.6
116.2 118.0 855
2.04 0.63
3.74
3.23 40.2
3.75 2.06 7.6 51.1
52.2
38.0
10.7
n
s.74
63.0
amino acid in 10' g. protein. b Including indole. sulfhydryl. Buehler d d.(1947). A Reaiduee per mol. (900,000). 1
15.75 11.75 4.45 311.0 10.1 [321.01
r
Txiatram
c
cp
w
144
0. R. TRIBTRAM
residues is illuminating for many purposes. The available data for more than twenty differentproteins are listed in Table XXXVII. The proteins are arranged in descending order with respect to the percentage of total ionic groups. It should be noted that the percentage of glycine residues in the molecule is given by the difference between the figure for lipophilic nonpolar groups and the total nonpolar groups.* An approach of this nature cannot provide an explanation of all protein functions, because many of these must result from the relative positions of the amino acids along the peptide chains with the further complication produced by the folding of the protein chain and the proximity of groups in adjacent chains. The data in Table XXXVII bring out certain interesting points. 1. Phenolic and Hydroxyl Groups
Except for special cases (e.g., insulin and casein) the proportion of tyrosine in all these proteins is remarkably constant. This relative constancy, with notable exceptions, e.g., wool keratin, silk fibroin, and fibrinogen, is also shown by aliphatic hydroxyl groups. This would suggest that these groups have a special part to play in the functioning of these proteins. For instance the high hydroxyl OH content of wool keratin, silk fibroin, and fibrinogen, all of fibrous nature, suggests that these groups may t
These proteins have molecular weights of the same order of magnitude, and the comparison of groups is astonishingly close, although the p-lactoglobulin has a higher proportion of anionic groups. Of the two proteins ovalbumin is soluble in water, whereas 8-lactoglobulin requires the presence of salt. This would suggest, in keeping with physicochemical data, that the solubility characteristics of a protein are functions not only of the number of polar groups, but of their disposition along the folded peptide chain. The dipole moment of p-lactoglobulin is much higher than that of egg albumin, indicating a greater asymmetry in the distribution of anion and cation charges (Oncley, in Cohn and Edsall, 1943).
3. Ionic Groups The proportion of total ionic groups, upon which depend many of the physical properties, shows enormous variation from 2.6% in zein to 45% in the case of tropomyosin, the latter containing the highest pro* The importance of glycine residues for the flexibility of polypeptide chains and for freedom of internal rotation has been particularly stressed by Neurath (1943).
145
A M I N O ACID COMPOSITION O F P U R I F I E D P R O T E I N 8
TABLE XXXVI The Amino-Acid Composition of Certain Crystalline Enzymea 1. Weight of amino acid in 100 g. protein 2. Gram-moles of amino acid in 106 g. protein 3. Residues per molecule
-
Constituent
Glycine Alanine Valine Leucine Isoleucine Proline Phenylalanine Cysteine Cyatine/2 Methionine Tryptophan Arginine Hiatidine Lyaine Aspartic acid Glutamic aeid Amide NAa Serine Threonine Tyroaine c
Total
' Includea
I
Chymotrypainogen Pepsin T.N. 14.85% T.N. 18.18% 2.07 % a 0.84 % L I34,4M. 3e.m - -- u v1 3 1 3 2 2 -__70.8 5.3 6.4 2e 29 85.3
-
-
-
7.1 10.4 10.8 5.0 13.4 0.6 1.84 1.7 2.38 1.0 0.9 0.8 18.0 11.9 1.8 12.2 0.0 8.5
80.8
21 27 28 16 13 2 4 4 4 2 2 2 41 28 (32) 40 28
-
79.3 82.3 43.4 38.7 4.1 13.8 11.4 11.8 6.7 6.8 8.2
120.2 80.9 (93.9) iie.1 80.8 48.9 892.7
ie
-
10.1 10.4 5.7 5.9 3.8 1.29 3.3 1.22 5.57 2.82 1.23 8.0 11.3 9.0 1.88 11.4 11.4 2.98
-__
308
112.38
I
-
-
88.2 78.3 43.6 61.3 21.8 10.7 27.5 8.2 27.3 18.2 7.9 54.7 84.8 81.2 (109.3: 108.6 95.7 18.3
32 29
I
1 3 -1.3 1i.3 3 -
7.3 0 3.1 3.8 3.8
ie 19 8 4 10 3 10
0.8' 8.61
4.43 0
8
6 . ie
3 20 31 22 (40) 40 35 8
4.22 10.4 14.2 13.0 2.49 12.0 9.0 7.93
871.7 __ 320 -
108.84
v
0.26% Hap04 (1 residue). Includes 0.82 % Sulfate-9. Includes 0.94 % Sulfate-9. Values corrected for 2.82 % Sulfate and 0.2 % anb. * May be artifact of hvdrolvnia . - (el. . . Oloott and Fraenkel-Conrat, 1947).
'
F_-
Derived data
-
Pepsin
-
-
Ribonuclease' T.N. 18.5% 3.88 % c M' ca. 15,000 -
Zhymotrypeinogen
82.3 0 23.8 31.3 21.8 6.0 54.2 29.7
0 29.8 27.2 71.1 100.4 88.3 (1413.2) 114.2 76.8 43.8
9 0 4 5 3 0.7 8 6 0 5 4
11
ie 13 (22)
17 11 7
--
801.4 122 -
-Ribonuclease-
1. M.R.W.
a. N distribution b. Analytical data 2. Total No. of residues (MW i. MRW) 3. Groupa aa percentage of total groups a. Polar groups Cationio Free anionie Total ionie Amide Phenolic Hydroryl Indole Total (including cysteinc and tryptophan) b. Non-polar group8 Hydrocarbon side chains Total, 4. End-groups (D.N.P.)
110.2 110.0 310
110 108 331
122.3 115 124
1.9 12.0 13.8 10.3 5.2 22.0 1.2 61.4
8.8 3.9 12.7 12. 1 1.8 22.7 3.0 49.3
15.7 e.0 21.7 17.8 5.4 23.2 0 88.1
36.8 48.2
38.6 44.4
28.0 30.1
L
(Data of E. Brand. See J. H. Northrop, M. Kunit. and * Columbia University Presa, New York (1848) p. 28.) f Glyoine reaidur Total - hydrooarbon Bide e m .
-
R. M. Herriott. "Crystalline Enzymes,"
-
Protein
TABLE XXXVII Comparative Distribution of Polar and Nonpolar Groups i n Proteing
Mean eaidue wight
Resiue p e r
-
nolecule -
Salmine. spring salmon Tropomyosin Myosin Home myoglobin Serum albumin. human &Lactoglobulin Edeatin Fibrinogen. human Hemoglobin. horse Ovalbumin Type A Toxin Clostridium botulinum Aldolase (myogen A) Triosephoaphate dehydrogenase Gelatin Collagen Serum 7-globulin Insulin
137.5 116.5 115.5 112.0 118.2 114.9 118.5 114.3 112 115
58 859b 866b 146’ 864b 322 422 875’ 582 400
116.9 109.6 110.5 91.2 90.5 112.3 113
Wool keratin
116
Silk fibroin
76.6 117 111
Gliadin Zein
-
Pc :atiosu
Free rnioni3
rgrOU
rotal ionic TOUPS
-70
-
onpo r groups
-
Lmide
Phenolic
-
-
HYiroxyl
10.3 7.7 9.7 8.9 9.1 9.7 11.9 13.5 10.8 8.5
ulfhy dry)
-
lndole
-
Tots
Amino
Liporotal ?hiliC‘
end groupd
- - 14.5 35.2 32.4 39.0 33.6 45.6 38.8 32.2 43.0 47.5
20.5 Proline
0.7 1.1 0.5 0.4 0.55 2.8
80.3 62.8 0.5 57.7 1.4 51.4 0.1 53.0 1.1 54.2 0.85 63.5 56.1 1.8 46.8 1.0 47.6 1.5
0.25
1 .o
34.8 42.0 42.3 29.4 30.0 30.6 48.5
37.0 50.4 51.1 64.3 63.0 40.0 55.3 2 phenylalanine glycine.
-
-
-
18.4 16.1 19.8 16.9 12.3 17.0 14.3 16.2 11.5
26.6 18.05 14.4 15.8 18.5 13.5 11.8 9.5 13.9
k5.0 14.1 14.2 12.7 10.8 10.5 36.1 25.7 25.4
7.4 9.9 5.5 7.45 9.0 16.3 10.8 6.7 9.1
2.0 2.2 1.40 3.1 2.4 3.1 3.5 2.0 2.5
776 912b 906‘ 1100 1100 890b 106
10.0 14.2 14.0 7.75 7.7
22.3 23.6 20.4 18.05 18.4 18.6 17.1
17.7 7.1 7.5 0.5 0.7 8.9 11.3
8.7 3.2 2.8 0.5 0.5 4.2 8.4
14.4
-
7.8
12.3 9.4 6.4 10.3 10.7 7.5 9.3
20.2 7.9
0.65 0
1.6 0
63.0 50.6 47.4 33.6 34.0 54.0 44.7
890
9.7
7.2
16.9
9.5
2.9
17.0
-
1.0
47.3 43.0 52.7 ? glycine,
1300 855b 900s
1.05 3.74 2.04
2.78 0 0.63
-
-
11.1
3.8 0.0 3.74 37.5 2.67 23.6
-- - - - - -
13.1 14.5 14.6 14.5
13.2 5.5 2.06 7.5 3.23 10.7
0
-
1.2 1.1
-
1 glycine
40.7 53.1 6 valine 52.5 nil
p henylalanine
22.5 32.0 77.5 0.35 51.1 38.0 2 40.2 52.2
---
I
35.2 48.0 Glycine 39.2 48.0 3 leucine
7
_y
_. -
9 Calculated an per cent of total protein. The proportion of ionic groups wiU be affected by the numbers of chains in the molecule. assume a single chain. b Residues per lo* g. protein. a Found by Dubuimon. Hamoir (1943)from titration data. d All nonpolar groups with hydrocarbon aide chains. Glycine content not known. I End groups per molecule (method of Sanger, 1945).
-
-
The figures given
AMINO ACID COMPOSITION OF PURIFIED PROTEINS
147
portion of ionic groups so far found in any protein, with the exception of the protamines. It is clear that the solubility of a protein bears no simple relationship to the number of ionic groups, because the proteins listed in Table XXXVII, arranged as they are in order of decreasing valency, show no regular gradation in solubility, although a low valency (silk fibroin, zein, and gliadin) is indicative of insolubility. The high insolubility of wool keratin in spite of the high proportion of ionic and hydroxylic groups, and the relatively low solubility of salmine, give a further indication of the dependence of properties upon the disposition of groups along a folded peptide chain. 4. Lipophilic Groups
The term lipophilic has been applied to all those amino acids with hydrocarbon side chains. Such groups, which are present to a considerable extent (30-660/,), have no known function, but undoubtedly are of considerable importance in controlling ,the interaction of lipide and protein a t cell surfaces. Comparison of the proteins isolated from rabbit muscle, myosin and tropomyosin (Bailey, 1948), aldolase (myogen A), and triosephosphate dehydrogenase (Velick and Ronzoni, 1948), clearly distinguishes the structural proteins from those of the myogen fraction (Table XX X III).
5. Intramolecular Linkages With the observation that insulin is made u p of several intramolecular chains (Chibnall, 1942, 1946; Sanger, 1945), an observation now extended to many proteins (Sanger, 1947; Porter and Sanger, 1948), attention must, of necessity, be drawn to the means whereby such chains are held together. The linkages must be covalent, otherwise denaturation might be expected to yield a product with a molecular weight equal to the mean molecular weights of the constituent chains. I t has been generally supposed that cystine provides the main if not the only covalent linkage, although the evidence on which such a view is based is only circumstantial. I n the case of insulin there seems little reason to doubt that the four chains per submolecule (12,000) are linked through disulfde groups. In horse hemoglobin, however, it is difficult on the assumption that the chains are linked through disulfide groups, to reconcile the evidence provided by end group assay, which suggests the presence of six chains, with the apparent presence of less than three disulfide groups. This may suggest the presence of other types of linkages, because disulfide can only provide the required number of groups if all the cystine-cysteine
148
a.
R. TRISTRAM
(0.7 t o 1.0%) in the protein is present as disulfide, although sulfhydryl is known to be present after denaturation (Mirsky and Anson, 1936). Certain proteins (e.g., edestin) are composed of but a single chain* although i t is suggested th at a t least a portion of the sulfur is present as disulfide. I n such instances it must be presumed th a t the disulfide groups are concerned in intramolecular linkages. Other proteins such as myoglobin contain no cystine and are therefore made u p of a simple, folded polypeptide chain.
IX. CONCLUSION AND SUMMARY I n the last few years protein chemistry has made great progress, and amino acid analysis is now among the more accurate of the methods used in the investigation of the properties and structure of proteins. Analysis, and the work of Sanger (1945) and Porter and Sanger (1946), have revealed that the stoichiometry of proteins follows no simple rule (such, for instance, as that propounded by Bergmann and Niemann). It is clear that the over-all amino acid pattern cannot be used for the explanation of the general principles underlying protein structure unless the component chairis are chemically identical; if they are not, as seems to be the case with insulin (Sanger, 1947, 1948) we must await further investigations on the composition of individual chains. The author wishes to thank those colleagues and others who provided him with unpublished data. He is grateful to Dr. K. Bailey for valuable criticisms. In particular he extends his thanks to Professor A. C. Chibnall for the advice and encouragement which has been given so freely for many years.
REFERENCES Adair, G. S. (1927). J . Am. Chem. SOC.49, 2524. Adair, G. S., and Adair, M. E. (1934a). Biochem. J. 26, 199. Adair, G. S., and Adair, M. E. (1934b). Biochem. J. 28, 1230. Albanese, A. A. (1940). J. Biol. Chem. 134, 467. Archibald, R. M. (1946). Ann. N . Y . Acad. Sci. 47, 181. Astbury, W. T. (1942). J . Chem. SOC.,337. Baernstein, H. D. (1936a). J. Biol. Chem. 106, 451. Baernstein, H. D. (1936b). J. Bid. Chem. 116, 25. , Baernstein, H. D. (1936~). J. Biol. Chem. 116, 33. Bailey, K. (1937). Biochem. J. 31, 1396. Bailey, K. (1942b). Biochem. J. 36, 140. Bailey, K. (1944). Advances in Protein Chem. 1, 289. Bailey, K. (1948). Biochem. J . 43, 271. Bailey, K., Chibnall, A. C., Rees, M. W., and Williams, E. F. (1943). 37, 360.
Biochem. J .
* Ovalbumin does not appear to contain any terminal amino groups (Porter, 1949, in press) but the presence of the carbohydrate moiety complicates the interpretation of these findings).
AMINO ACID COMPOSITION OF PURIFIED PROTEINS
149
Beach, E. F., Bernstein, S. S., Hummcl, F. C., Williams, H. H., and Macy, I. G. (1939). J . Biol. Ch2m. 130, 115, 733. Bernal, J. D. (1939). Proc. Roy. SOC.London Ai70, 75. Block, R. J., and Bolling, D. (1945a). Arch. Biochem. 6, 419. Block, R. J., and Bolling, D. (1945b). The aminoacids of proteins and foodstuffs. Thomas, Springfield, Ill. Blumenthal, D., and Clarke, H. T. (1935). J . Biol. Chem. 110, 343. Bolling, D., and Block, R. J. (1943). Arch. Biochem. 2, 93. Bbwes, J. H., and Kenten, R. H. (1948). Biochem. J . 43, 358. Boyd, M. J., and Logan, M. A. (1942). J . B i d . Chem. 146, 279. Brand, E. (1946). Ann. N . Y . Acad. Sci. 47, 187. Brand, E., and Edsall, J. T. (1947). Ann. Reo. Biochem. 16,223. Brand, E., and Grantham, J. (1946). J . Am. Chem. SOC.68,724. Brand, E., and Kassell, B. (1939). J. Biol. Chem. 131,489. Brand, E., and Kassell, B. (1941). J . Gen. Physiol. 26, 167. Brand, E., and Kassell, B. (1942a). J . Biol.Chem. 146, 359. Brand, E., and Kaesell, B. (194213). J . Biol. Chem. 146,365. Brand, E., Kassell, B., and Saidel, L. J. (1944). J . Clin. Invest. 23, 437. Brand, E., Ryan, F. J., and Diskant, E. M. (1945). J . Am. Chem. SOC.67, 1532. Brand, E., and Saidel, L. J. (1943). A m . Chem. SOC.Meeting Abs. p. B-22. Brand, E., and Saidel, L. J. (1944). A m . Chem. Soc. Meeting Abs. p. B-26. Brand, E., Saidel, L. J., Goldwater, W. H., Kassell, B., and Ryan, F. J. (1945). J . A m . Chem. SOC.67, 1524. Buehler, H. J., Schauta, E. J., and Lamanna, C. (1947). J . Biol. Chem. 169, 295. Bull, H. B. (1941). J. Biol. Chem. 137,'143. Burk, N. F., and Greenberg, D. M. (1930). J . Biol. Chem. 87, 197. Cannan, R. K. (1944). J . Biol. Chem. 162,401. Cannan, R. K. (1946). Ann. N . Y . Acad. Sci. 47, 135. Cannan, R. K., Kibrick, A. C., and Palmer, A. H. (1941). Ann. N . Y . Acad. Sci. 41, 243. Cannan, R. K., Palmer, A. H., and Kibrick, A. C. (1942). J . Biol.Chem. 142,803. Carter, H. E., and Neville, H. E. (1947). J . Biol. Chem. 170, 301. Carter, H. E., and Loo, Y. H. (1948). J . Biol. Chem. 174, 723. Chibnall, A. C. (1942). Proc. Roy. SOC.London B131, 136. Chibnall, A. C. (1946). J . Intern. SOC.Leather Traded Chemiete 30, 1. Chibnall, A. C., Rees, M. W., Tristram, G. R., and Williams, E. F. (1939). Nature 144, 71.
Chibnall, A. C., Rees, M. W., and Williams, E. F.. (1943a). Biochem. J . 37, 354. Chibnall, A. C., Rees, M. W., and Williams, E. F. (1943b). Biochem. J . 37, 372. Chow, B. F. (1944). Advances in Protein Chem. 1, 153. Cohn, E. J., and Hendry, J. L. (1923). J . Gen. PhySiOl. 6 , 521. Cohn, E. J., and Edsall, J. T. (1943). Proteins, Amino Acids and Peptides, Reinhold, New York. Coleman, D., and Hewitt, F. 0. (1946). J . Dyer8 and Colourisls Soc. 144. Consden, R., Gordon, A. H., and Martin, A. J. P. (1948). Biochem. J . 42, 443. Crowfoot, D. (1938). Proc. Roy. SOC.London A164, 580. Crowfoot, D., and Riley, D. (1938). Nature 141, 521. Caonka, F. A,, and Denton, C. A. (1946). J . Biol. Chem. 163, 329. Darmon, 8. E., Sutherland, C. B. B. M., and Tristram, G. R. (1948). Biochem. J .
42, 508. Dubuisson, M., and Hamoir, G . (1943). Arch. intern. Phyeiol. 63, 308.
150
Q. R. TRISTRAM
Dunn, M. S., Camien, M. N., Shankman, S., Rockland, L. B., and Goldberg, S. C. (1944a). J . Biol. Chem. 166, 591. Dunn, M. S., Camien, M. N., Fraenkl, W., and Rockland, L. B. (1944). J . Biol. Chem. 166, 715. Dunn, M. S., Camien, M. N., Shankman, S., and Block, H . (1946a,b). J . Biol. Chem. 163, 577, 589. Dunn, M. S., Camien, M. N., Shankman, S., and Block, H. (1947). J . Biol. Chem., 168, 43. Dunn, M. S., Camien, M. N., Shankman, S., and Rockland, L. B. (1945). J. Biol. Chem. 169, 653. Dunn, M. S., and Rockland, L. B. (1946). Arch. Biochem. 11, 89. Dunn, M. S., and Rockland, L. B. (1947). Advances i n Protein Chem. 3, 296. Dunn, M. S., Shankman, S., and Camien, M. N. (1945a). J . Biol. Chem. 161, 643. Dunn, M. S., Shankman, S., and Camien, M. N. (1945b). J . Biol. Chem. 161, 669. Edsall, J. T. (1946). A n n . N . Y . Acad. Sci. 47, 229. Elsden, S. R., and Synge, R. L. M. (1944). Biochem. J. 38, ix. Foreman, F. W. (1914a,b). Biochem. J . 8, 463, 481. Foster, G. L. (1945). J . Biol. Chem. 169, 431. Gale, E. F. (1945). Biochem. J. 39, 46. Gale, E. F. (1946). Nature 167, 265. Gordon, A. H., Martin, A. J. P., and Synge, R. L. M. (1941). Biochem. J. 36, 1369. Gordon, A. H., Martin, A. J. P., and Synge, R. L. M. (1943). Biochem. J . 37, 79. Graham, C. E., Smith, E:. P., Hier, S. W., and Klein, D. (1947). J . Biol. Chem. 168, 711. Greene, R. D., and Black, A. (1943). Proc. SOC.Exptl. Biol. Med. 64, 322. Greenstein, J. P. (1938). J . Biol. Chem. 186, 501. Greenstein, J . P. (1939). J . Biol. Chem. 128, 233. Greenstein, J. P. (1940). J . Biol. Chem. 136, 795. Greenstein, J. P., and Edsall, J. T. (1940). J . Biol. Chem. 133, 397. Gronwall, A. (1942). Compt. rend. trac. lab. Carlsberg. 567. chim. 24, 185. Guirard, B. M., Snell, E. E., and Williams, R. J. (1946). Proc. SOC.Exptl. Biol. Med. 61, 158. Gunness, M., Dwyer, I. M., and Stokes, J. L. (1946). J . Biol. Chem. 163, 159. Gutfreund, H. (1944). Nature 163, 406. Gutfreund, H. (1945). Nature, London, 166, 237. Gutfreund, H. (1948a). Biochem. J. 42, 156. Gutfrepnd, H. (1948b). Biochem. J. 42, 544. Gutfreund, H., and Ogston, A. H. (1946). Biochem. J. 40, 432. Hac, L. R., and Snell, E. E. (1945). J . Biol. Chem. 169, 291. Hac, L. R., Snell, E. E., and Williams, R. J. (1945). J . Biol. Chem. 169, 273. Hadley, P. (1927). J. Infectious Diseases 40, 1. Hadley, P. (1937). J . Infectious Diseases 60, 129. Hanke, M. T., and Koessler, K. K. (1920). J . B i d . Chem. 43, 527. Hansen, R. G., Potter, R. L., and Phillips, P. H. (1947). J . Biol. Chem. 171, 229. Harington, C. R., and Mead, T. H. (1936). Biochem. J . 30, 1598. Harington, C. R., and Neuberger, A. (1936). Biochem. J . 30, 809. Hegsted, D. M. (1944). J. Biol. Chem. 162, 193. Henderson, L. M., and Snell, E. E. (1948). J . Biol. Chem. 172, 15. Herriott, R. M. (1942). Chem. Rev. 30, 413. Hem, W. C., and Sullivan, M. X. (1943). J . Biol. Chem. 161, 635.
151
AMINO ACID COMPOBITION OF PURIFIED PROTEINS
Hesa, W. C., and Sullivan, M. X. (1944). Arch. Biochem. 6, 165. Hier, 8. W.,Graham, C. E., Friedes, R., and Klein, D. (1945). J . Biol. Chem. 161,705. Horn, M. J., Jones, D. B., and Blum, A. E. (1947a,b). J . Biol. Chem. 169,71, 739. Kassell, B., and Brand, E. (193&,b). J . Biol. Chem. 126, 145, 435. Kekwick, R. A., and Cannan, R. K. (1936). Biochem. J . 30, 227. Kekwick, R. A., and MacFarlane, A. S. (1943). Ann. Rev. Biochem. 12, 93. Keston, A. S., Udenfriend, S., and Cannan, R. K. (1948). J . A m . Chem. Soc. 68, 1390.
Keston, A. S., Udenfriend, S., and Cannan, R. K.
(1949).
J . Am. Chem. Soc. 71,
249.
Kibrick, A. C. (1944). J. Biol. Chem. 169,411. Knight, C. A. (1942). J . Am. Chem. Soc. 64,2734. Knight, C. A. (1943). J . Biol. Chem. 147, 663. Knight, C. A. (1947). J . Biol. Chem. 171, 297. Kuhn, R., Birkhofer, L., and Qflackenbush, F. W. (1939). Ber. 72,407. (1943). Kuiken, K. A., Norman, W. H., Lyman, C. M., Hale, F., and Blotter, J . Biol. Chem. 161, 615. Kunitz, M., and Northrop, J. H. (1938). Compt. rend. lab. trao. Carlsberg. Sdr. chim. 22, 288. Lavine, T. F. (1943). J . Biol. Chem. 161, 281. Lewis, J. C., and Olcott, II. S. (1945). J . Biol. Chem. 167, 265. I i , C. H. (1946). J. A m . Chem. Soc. 68,2746. Linderstr~m-Lang,K., and Jacobsen, C. F. (1940). Enzymologia 8, 64. IAongsworth,L. G., Cannan, It. K., and MacInnes, D. A. (1940). J . Am. Chem. SOC.
+.
62, 2580.
Lugg, J. W. H. (1937). Biochem. J . 91, 1422. Lugg, J. W. H. (193&,b). Biochem. J. 32, 775, 2123. Lyman, C. M., and Kuiken, K. A. (1948). Federation Proc. 7, 170. Lyman, C. M., Kuiken, K. A,, Blotter, L., and Hale, F. (1945). J . Biol. Chem. 167, 395.
Lyman, C. M., Kuiken, K. A., Blotter, L., and Hale, F.
(1947).
J . Biol. Chem. 171,
233.
Lyman, C. M., Moseley, O., Butler, B., Wood, S , and Hale, F. (1946). J . B i d . Chem 166, 181. McMahan, J. R., and Snell, E. E. (1944). J . Biol. Chem. 162, 83. McMeekin, T. L., Polis, B. D., DellaMonica, E. S., and Custer, J. €1. (1948). J . A m . Chem. Soc. 70, 881. Macpherson, H. T. (1946). Biochem. J . 40, 470. Martin, A. J. P., and Synge, R. L. M. (1941). Biochem. J . 36, 294. Martin, A. J. P., and Synge, R. L. M. (1945). Advances in Protein Chem. 2 , 1. Miller, C . L., and du Vigneaud, V. (1937). J . Biol. Chem. 118, 101. Mirsky, A. E., and Anson, M. L. (1936). J . Gen. Physiol. 19,439. Moore, S., and Stein, W. H. (1943). J . Biol. Chem. 160, 113. Moore, S., and Stein, W. H. (1048). Ann. N . Y . Acad. Sci. 49, 265. Moore, S., and Stein, W. II. (1949). J . Biol. Chem. 178, 79. Neubergcr, A. (1938). Biochem. J . 32, 1435. Neuberger, A. (1944). Biochem. J . 38, 309. Neurath, H. (1943). J . Am. Chem. SOC.66, 2039. Neurath, H., Greenutcin, J. P., Putnam, F. W., and Erickson, J. 0. (1944). Chem. Revs. 34, 157.
152
0. R. TRISTRAM
Nicolet, B. H. (1943). Ann. Rev. Biochem. 12,276. Nicolet, B. H., and Shinn, L. A. (1941a). J . Biol. Cheni. 139, 687. Nicolet, B. H., and Shinn, L. A . (1941b). J. A m . Chem. SOC.63, 1486. Nicolet, B. H., and Shinn, L. A . (1941~). J . Biol. Chcm. 140, 685. Nicolet, B. H., and Shinn, L. A. (1942). J . Biol. Chem. 142, 139. Northrop, J. H., and Kunits, M. (1930). J . Gen. Physiol. 13,781. Ogston, A. G. (1948). Biochem. J. 44, 33. Olcott, H.S., and Fraerikel-Conrat, H . (1947). J . B i d . Chem. 171, 583. Osborne, T. B. (1902a,b,c). J . A m . Chern. SOC.24, 28, 39. Palmer, A. H. (1934). J. Bio2. Chem. 104,359. Pederscn, K.0. (1936). Biochem. J. 30, 961. Pirie, N. W. (1940). B i d . Revs. Cambridge Phil. soc. 16, 377. Porter, It. R. (1948). Biochem. el Biophys. Acla 2, 105. Porter, R. It., and Sanger, F. (1948). Biochem. J. 42,287. Rees, M. W. (1946). Biochem. J. 40, 632. Riesen, W. H., Schweigert, B. S., and Elvehjem, C. A. (1946). J . Biol. Chem. 166, 347.
Rittenberg, D., and Foster, G. L. (1940). J . Biol. Chem. 133, 737. Ross, A. F. (1941). J. ljiol. Chem. 138,741. Ryan, F.J., and Brand, E. (1944). J . Biol. Chem. 164, 161. Sanger, F. (1945). Biochem. J . 39,507. Sanger, F. (1947). Nature 160,295. Sanger, F. (1948). Biochem. J . 44, 126. Sanger, F.,and Porter, R. R. (1947). XZlh Intern. Congr. Pure Applied Chem. 11th Congr. London. Schmidt, G . L. A. (1929). J. Biol. Chem. 82, 587. Schweigert, B. S., McJntyre, J. M., Elvehjem, C. A , , and Strong, F. hl. (1944). J . Biol. Chem. 166, 183. Schweigert, B. S. and Snell, E. E. (1947). Nutrition Abstracts & Revs. 16, 497. Scott, D. A., and Fisher, A. M. (1935). Biochem. J . 29, 1048. Shankman, S. (1943). J. Biol. Chem. 160,305. Shankman, S., Camien, M. N., and Dunn, M. S. (1947). J . Biol. Chem. 168,51. Shedlovsky, T. (1943). Ann. N. Y. Acad. Sci. 43,253. Shemin, D. (1945). J . Biol. Chem. 169,439. Shemin, D., and Foster, G. L. (1946). Ann. N . Y . Acad. Sci. 47, 119. Shinn, L. A., and Nicolet, B. H. (1941). J . Biol. Chem. 138,91. Shore, A., Wilson, H., and Stueck, G. J. (1935-6). J . Bzol. Chem. 112,407. Smith, E. L., and Greene, R. D. (1947a). J . Biol. Chem. 167,833. Smith, E. L., and Greene, R. D. (1947b). J . Biol. Chem. 171, 355. Smith, E. L., and Greene, R. D. (1948). J . Biol. Chem. 172, 111. Smith, E. I,., Greene, R. D., and Bartner, E. (1946a,b). J. Biol. Chpm. 164, 159, 359.
Snell, E. E. (1945). Advances i n Protein Chem. 2, 85. Snell, E. E. (1946). Ann. N . Y . Acad. Sci. 47, 161. Snell, E. E., and Guirard, B. M. (1943). PTOC.Null. Acad. S c i . U . S. 29, 66. Snell, E. E., and Rannefeld, A. N. (1945). J . Bzol. Chem. 167, 475. Sflrensen, M., and Palmer, A . H. (1938). Compt. relid. trau. lub. Carlsberg. SCT.chim.
21, 283. Stanley, W. M., and Lauffer, M. A. (1939). Science 89, 345. Stein, W. H. (1946). Ann. N . Y . Acad. Sci. 47, 59.
AMINO ACID COMPOSITION OF PURIFIED PROTEINS
153
Stokes, J . L., and Gunnese, M. (1945a). J . Biol. Chem. 167, 851. Stokes, J. L., and Gunness, M. (1945b). Science 101,43. Stokes, J . L., Gunness, M., Dwyer, I. M., and Caswell, M. C. (1945). J . Biol. Chem. 160,35. Sullivan, M. X., and Hem, W. C. (1939). J . Biol. Chem. 1S0, 745. Sullivan, M. X., and Hem, W. C. (1944). J . Biol. Chem. 166, 441. Svedberg, T., and Pedersen, K. 0. (1940). The Ultracentrifuge. Clarendon Preas, Oxford. Synge, R. L. M. (1944). Riochem. J . 38, 285. Taylor, G. L.,Adair, G . S., and Adair, M. E. (1932). J . Hyg. 32, 340. Theorell, H., and Akeaon, A. (1942). Arkiv. Kemi, Mineral. Geol. 16A, No.8. Thomas, L. E. (1944). Arch. Biochem. 6, 175. Tiselius, A. (1943). Kol2oid.-2. 106, 101. Tiselius, A. (1942). Advances in Colloid Sci. 1, 181. Tristram, G. R. (1946). Riochem. J . 40,721. Tristram, G.R. (1947a). Nature 169, 581. Tristram, G. R. (1947b). Nature, London 160, 637. Tristram, G. R. (1949). Haemoglobin. Butterworths Scientific Publicstione, London. Vassel, B. (1941). J. Biol. Chem. 140, 323. Velick, S. F.,and Itonzoni, E. (1948). J . Biol. Chem. 173, 627. Vickery, 11. B. (1940). J . Biol. Chem. 132, 325. Vickery, H. B. (1942). J. Biol. Chem. 144, 719. Vickery, H. B. (1946). Ann. N . Y . Acad. Sci. 47,63. Vickery, H. B., and White, A. (1932). J . Biol. Chem. 99, 701. Vickery, H. B., and Winternitz, 0. (1944). J . Biol. Chem. 166,211. du Vigneaud, V., Jensen, H., and Wintersteiner, 0. (1927). J . Pharmacol. Ezpll. Therap. 32, 367. Warner, R. C., and Cannan, R. K. (1942). J . Biol. Chem. 142,725. Winnick, T. (1941). J . Biol. Chem. 142,461. Wyman, J. (1948). Advances i n Prolein Chem. 4,407. Zittle, C. A., and Nelson, R. E. (1944). J . Biol. Chem. 166, 401.
This Page Intentionally Left Blank
Biological Evaluation of Proteins
BY JAMES B. ALLISON Rutgers University, New Brunswick, New Jersey
CONTENTS Page I. Introduction, . . . . . . ............................................ 11. Evaluation through Nitrogen Balance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. Protein Minima for Nitrogen Equilibrium.. . . . . . . . . . . . . . . . . . . . . . . . 2. Nitrogen Balance Index of Nitrogen Intakes.. . . . . . . . . . . . . . . . . . .
155 157 158 . 161
1. Growth and Nitrogen Retention.. . . . . . . . . . . . 2. Protein Efficiency Ratios.. . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Protein Efficiency and Nitrogen Retention. . . . . . . . . . . . . 1. Depletion in Protein
....................................
Potency Ratio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Plasma Albumin Regeneration.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Repletion Areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Repletion in Liver Protein.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8. Repletion in Body Weight ............................... V. Evaluation through Amino Acid Analysis.. . . . . . . . . . . . . . . . . . . . . . . . . . . 1. Correlation between Amino Acid Composition and Biological Tests. . . 2. Some Causes of Disagreement.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI. Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References. . . . . . .......................................... 4. 5. 6. 7.
181
185 186 186 187 .189 192 192 193 195 196
I. INTRODUCTION Growt,h, reproduction, repair, maintenance, resistance to disease, and all of the faculties of the living system are correlated with the intake and utilization of foods. The foods are the raw materials which are needed to supply energy to the living machine and to construct and t o repair it, Dietary proteins, through digestion in the gastrointestinal 155
156
JAMES B . ALLISON
tract, furnish amino acids, possibly polypeptides too, which are the raw materials needed t o build the body proteins of animals. These body proteins form the matrix of the living system; they are the catalysts, the centers around which the dynamic equilibria of life develop; they are the protein stores of the body. These stores are the proteins of the gut, the liver, the plasma, and other tissues of the body. There are no protein reserves in the sense that there are f a t or carbohydrate reserves but the body utilizes tissue proteins to maintain the nitrogen integrity of essential structures. The studies of Whipple and associates (1946) and Schoenheimer and Rittenberg (1940) emphasize the dynamic equilibria which exist within these stores so that one type of tissue protein contributes to the construction of others and food nitrogen entering into these equilibria loses its identity and becomes part of body nitrogen. The primary purpose, therefore, of dietary.proteins is to fill the protein stores of the body and their biological evaluation is the study of this purpose. Indeed, the biological value of a protein was defined by Thomas (1907) as the amount of nitrogen retained in the body of an animal, a concept which has been so ably developed by Mitchell (1944) and associates in their many excellent papers and reviews. These reviews form a background for this presentation of the biological evaluation of proteins, a presentation which will attempt to expand and supplement the previous surveys. A new phase of research on the nutritive value of proteins began when mixtures of amino acids replaced proteins in the diets of animals. Such studies were inaugurated by Rose in 1930, who demonstrated th a t a mixture of ten of the amino acids, namely valine, methionine, threonine, leucine, isoleucine, tryptophan, lysine, histidine, and arginine, must be included in optimum quantities in the diet for normal growth in the rat. An essential amino acid was defined, therefore, as one which could not be synthesized by the animal in sufficient quantities to meet the needs for growth (Rose, 1938). An amino acid may be essential for growth, however, but not be essential for all the faculties of the living system. It is now clear that the pattern of essential amino acids varies with the physiological state of the animal and with the species. The retention of nitrogen in the animal is a function of this pattern. Consideration will be given in this review, therefore, to the correlation between the amino acid composition of a protein and its nutritive value. The purposes of protein nutrition are filled, however, only if all the raw materials are present in optimum amounts. If carbohydrate and fat are missing, for example, amino acids of the diet are utiliqed to supply energy so t ha t the construction and repair of the living machine is reduced or interrupted entirely. Certain of the vitamins are active spots in the protein catalysts of intermediary metabolism. Absence of these vitamins
BIOLOGICAL EVALUATION OF PROTEINB
157
interrupts the formation of the catalysts and the machine cannot function, Thus the biological evaluation of any food such as proteins must be done in the presence of all other foods, with the full realization of their interdependence. Unless otherwise mentioned attempts have been made to make protein the only dietary variable in the researches which will be discussed in the following pages. The simplest method used to evaluate nitrogen retention in a n animal is to determine the difference between nitrogen intake and nitrogen excreted. This difference, called nitrogen balance, shows whether a n animal is maintaining, losing, or gaining nitrogen. Evaluation through nitrogen balance will be elaborated in this review into its quantitative forms. These forms become most complicated when applied to the growing animal so that the concept of growth as a function of nitrogen retention will be examined. Growth of tissues can take place also in a n adult animal depleted in protein stores. These stores are reduced in malnutrition, and in their reduction the body loses the capacity to repair damage, to maintain barriers to destructive forces. Repleting the protein stores of the body is a form of therapy very important to medical practice. Dietary proteins may replete tissues differently than they support growth in young animals or maintain the nitrogen integrity of a n adult. This review will consider, therefore, the evaluation of dietary proteins through tissue regeneration.
11. EVALUATION THROUGH NITROGEN BALANCE Nitrogen balance is the difference between dietary nitrogen intake and nitrogen excreted in the urine and in the feces. If the nitrogen intake equals the total nitrogen excreted, nitrogen balance is zero and the animal is said to be in nitrogen equilibrium. If the nitrogen intake is greater than the nitrogen excreted, the animal has gained in nitrogen and is said to be in positive nitrogen balance. If the nitrogen intake is less than the nitrogen output, the animal is losing nitrogen from the body stores and is said to be in negative balance. Thus the term nitrogen balance ( B ) is used for the over-all accounting of nitrogen assimilation. This term is defined mathematically by the following equation:
B=I-(F+U)
(1)
where Z is nitrogen intake, F is fecal nitrogen, and U is urinary nitrogen. Positive nitrogen balances can be maintained on a mixture of so-called essential amino acids, demonstrating that the amino acids are the sources for proteins and other nitrogenous constituents in the body of an animal. If one of the amino acids, essential to maintenance of nitrogen balance, is eliminated from the diet the animal will go into negative balance and
158
JAMES B . ALLISON
lose nitrogen from body protein stores. The expression “essential for maintenance l 1 is used advisedly because the terms “essential ” or “indispensable” were applied originally to amino acids not synthesized by the animal organism at a speed necessary to meet the demands for normal growth. Rose and coworkers (Borman et al., 1946) and Albanese (1947) point out, however, that this concept of an essential amino acid need not apply to other functions such as reproduction, detoxication, or even over-all maintenance of nitrogen balance in an adult organism. Lysine, for example, is an indispensable amino acid for growth in the young rat, but Mitchell (1947) has demonstrated that this acid is not essential for the maintenance of nitrogen equilibrium in the normal mature rat. Lysine, on the other hand, is required for maintenance of nitrogen equilibrium in the dog (Rose and Rice, 1939; Allison, Anderson and White, 1949) and in protein nutrition in adult man (Rose, 1938, 1947; Albanese, 1947). Mitchell suggests that synthetic reactions leading to the formation of lysine in a rapidly growing animal such as the rat may be inadequate for growth but adequate for maintenance. Frazier et al. (1947) have demonstrated lysine to be essential t o the proteindepleted rat when regeneration of new tissue takes place. The nitrogen balance method has been used recently with particular success by Rose (1947) in the determination of the amino acids essential for maintenance of nitrogen equilibrium in man. His results demonstrated that valine, methionine, threonine, leucine, isoleucine, phenylalanine, tryptophan, and lysine are essential to this maintenance. The exclusion of any one of these amino acids from the food was followed by It a pronounced negative nitrogen balance, a profound failure in appetite, a sensation of extreme fatigue, and amarked increase in nervousinstability.” The effect of omission of one of these acids is so immediate that there can be little, if any, stores of free amino acids in the body to overcome dietary deficiencies. Dietary deficiencies in essential amino acids are reflected rapidly, therefore, in the reduction of the nitrogen balance. The nitrogen balance method of evaluating amino acid mixtures has been developed by many workers into a measure of the minimum amount of protein necessary to maintain nitrogen equilibrium in animals or in man. 1. Protein Minima for Nitrogen Equilibrium
Melnick and Cowgill (1937) gave careful consideration t o the quantitative determination of the minimum amount of dietary protein nitrogen necessary t o maintain nitrogen equilibrium. They determined this amount of nitrogen in the dog by plotting nitrogen intake against nitrogen balance in the region of nitrogen equilibrium, interpolating to the point of exact equilibrium. Their determinations demonstrated quantitative
BIOLOGICAL EVALUATION OF PROTEINS
159
variations in the ability of dietary proteins to maintain nitrogen equilibrium, variations associated with the presence and availability of essential amino acids. An acid hydrolyzate of casein, for example, in which the tryptophan was destroyed by the acid was shown to be incapable of maintaining nitrogen equilibrium in the dog no mattcr how much hydrolyzate nitrogen was included in the diet. When an optimum amount of tryptophan was added to the hydrolyzate, nitrogen equilibrium was maintained by feeding 150 mg. hydrolyzate nitrogen/day/kg. body weight. When an insufficient amount of tryptophan was added, 300 mg. nitrogen/day/kg. body weight was required to maintain the nitrogen equilibrium in the same dog. The minimum amount of nitrogen necessary to maintain equilibrium increases, therefore, as the content of an essential amino acid or acids decreases below optimum quantities. It is possible that the protein minima for nitrogen equilibrium arc always determined by the essential amino acid present in least quantity in the protein or made available in the smallest amounts to the animal organism. Risser (1946) , using Melnick’s and Cowgill’s techniques, measured the minimum amount of casein and fibrin required to maintain nitrogen balance in dogs, an amount which was labeled MPN. He found that fortification of casein with cystine reduced the M P N value and that addition of both cystine and methionine reduced the M P N value still further. Frost and Rivser (1946) have applied these nitrogen balance techniques in dogs for the evaluation of casein and fibrin hydrolyzates, variously supplemented by essential amino acids. These authors and others have checked their determinations of protein minima for nitrogen equilibria by feeding this quantity of nitrogen t o animals for many days. Kade et al. (1946) maintained nitrogen equilibrium for 21 to 25 days in dogs fed an acid hydrolyzate of casein fortified with tryptophan. Their data emphasizes the fact that dogs can be standnrdized to retain larger or smaller amounts of nitrogen. Cox and Mueller (1946) compared the relative efficiency of an enzymatic casein hydrolyzate and two mixtures of crystal amino acids in maintaining nitrogen equilibrium in protein-depleted dogs. At levels just sufficient to maintain nitrogen equilibrium, a mixture of ten essential amino acids and glycine promoted nitrogen retention somewhat more effectively than did the hydrolyzate. At a higher level the hydrolyzate was much more effective than the dmino acid mixture. The objection to the determination of protein minima for nitrogen equilibrium as a method of evaluating proteins is the great variation in minima from animal to animal. Melnick and Cowgill (1937) pointed to this variation, emphasizing the need to select animals of similar nitrogen needs if comparisons between one dietary protein and another are to be
160
JAMES B. ALLISON
made. Their work was extended b y Allison and Anderson (1945) and Allison, Anderson, and Seeley (1946) to demonstrate that a linear relationship between nitrogen balance and nitrogen intake in the dog exists over the whole range of negative nitrogen balance, extending into the region of positive balance but eventually becoming curvilinear in that region. This relationship is idealized in Fig. 1 for data obtained by
NiTROGEN INTAKE
g./day/m2
FIG.1. Curves illustrating the relationshipbetween nitrogenbalance and nitrogen intake in the diet. They are produced from data obtained by feeding dogs one type of protein such as defatted, dried, whole egg. Curve a represents data obtained on a protein-depleted dog while curves b, c, and d illustrate data obtained on normal dogs with differing protein stores.
feeding one type of protein to dogs. Curve d illustrates data secured on a dog with saturated protein stores, the excretion of nitrogen on a protein-free diet being high, 4 g./day/m.2 body surface area. Approximately 5 g. dietary nitrogen/day/m.2 was required to maintain nitrogen equilibrium in this dog. Such an animal, with high protein stores, can be prepared by feeding a high-protein diet, the proteins being of good quality such as those found in whole egg or milk. Curves b and c represent data obtained on normal animals in which the protein stores are not quite so full, less nitrogen being required t o maintain equilibrium. These data could be obtained from experiments on dogs which had been prepared by feeding ordinary mixed diets. Curve a illustrates data obtained on a dog which had been depleted markedly in proteins by feeding the
BIOLOGICAL EVALUATION O F PROTEINS
161
animal a protein-free diet for a period of time. This depleted animal required only 1 g. dietary nitrogen/day/m.2 body surface area to maintain equilibrium. The magnitude of the protein stores is reflected, therefore, by the amount of nitrogen excreted from the body tissues when the animal is receiving a nitrogen-free diet, being high when the stores are high and low when the stores are low. The excretion of nitrogen drops when an animal is placed on a protein-free diet, rapidly at first and then more slowly, reaching a relatively constant value in the depleted state. Thus the passage of an animal from a state of high protein stores t o low can be followed roughly by the decrease in excretion in nitrogen. The excretion of nitrogen on a protein-free diet reflects more strictly, however, the metabolism within the protein stores of the body and a change in the excretion may be correlated with a shift in this metabolism, which may or may not in turn be associated with markedly altered stores. Under any circumstances, however, the nutritive value of one protein cannot be compared to another by determining protein minima for nitrogen equilibrium unless the physiological states of the animals are known. The linear relationship illustrated in Fig. 1 suggests that there is a minimum excretion of nitrogen which must be met by dietary nitrogen to maintain nitrogen equilibria. These curves are described by the following equation:
B
=
K'I - Eo
(2)
where B is nitrogen balance, I is dietary nitrogen intake, and E e is the excretion of nitrogen during zero nitrogen intake. E,, the sum of the excretion of feces nitrogen, F,, and of urine nitrogen, U,, may represent a form of minimum nitrogen excretion. The slopes ( K ' ) are identical for lines b, c, and d, in the region of negative nitrogen balance, and are independent of the magnitude of nitrogen excretion or intake. These curves represent data obtained on so-called normal dogs. Under these conditions protein minima may vary but the slopes remain constant. 2. Nitrogen Balance Index of Nitrogen Intake It has been shown by Allison, Anderson, and Seeley (1946) that the slopes of lines such as those plotted in Fig. 1 decrease as the nutritive value of the dietary nitrogen decreases, being a function of the amount of nitrogen retained in the body of the animal. The slopes ( K 1 )have been called the nitrogen balance indexes of dietary nitrogen intake by these authors, since they measure the rate at which dietary nitrogen fills the protein stores of the body. These indexes do not vary in normal dogs even though there may be differences for nitrogen equilibria. The
162
JAMES B. ALLISON
indexes do, however, increase in the depleted dog as illustrated by the increased slope of curve c (see Allison, Seeley, Brown, and Ferguson, 1947). As the protein stores of the animal are reduced the position of the curve becomes higher on the y axis and the linear relationship between nitrogen balance and nitrogen intake extends further and further into the region of positive balance. Thus the capacity to grow in nitrogen increases as the stores are depleted (Allison, 1948).
3. Digestibility and Absorbed Nitrogen It is, in many respects, more meaningful to plot nitrogen absorbed from the intestinal tract instead of dietary nitrogen intake so that the nitrogen balance produced is correlated with the dietary nitrogen entering the blood stream rather than the gut. The fraction of dietary nitrogen absorbed from the intestinal tract into the blood stream is called the true digestibility. Thus digestibility (D)is defined as:
D = -A
(3)
I
where A is absorbed nitrogen and Z is dietary nitrogen intake. The nitrogen absorbed from the intestinal tract ( A ) is calculated according t o the following equation:
A
=
Z
-
(F - F,)
=
N intake
-
fecal food N
(4)
where Z is nitrogen intake, F is total feces nitrogen, and F , is the excretion of nitrogen in the feces which comes from body, not food, sources, so-called “metabolic fecal nitrogen.” The quantity F , is usually determined by measuring the amount of nitrogen excreted on a nitrogen-free diet. The assumption is made that the excretion of body nitrogen is constant and does not vary with nitrogen intake. This assumption has been tested in the mouse by Bosshardt and Barnes (1946)) who found that there is a linear relationship between fecal nitrogen and nitrogen intake. Body fecal nitrogen (F,) could be determined, therefore, by extrapolation to zero nitrogen intake. Blaxter and Mitchell (1948) have determined metabolic fecal nitrogen in a similar way in sheep and other animals. The data of Bosshardt and Barnes indicate that F , is different on protein-free or low-nitrogen diets than under conditions of protein feeding. Thus incorrect values for absorbed nitrogen would be calculated from equation 4 if F , was measured on a protein-free diet. The excretion of fecal nitrogen on a protein-free diet is designated F , since it may differ from F,. The value F o is a variable, reflecting the composition of the
BIOLOGICAL EVALUATION O F P R O T E I N S
163
diet and physiological state of the animal. Depletion in proteins reduces, for example, the excretion of feces nitrogen, a reduction which is associated with decreasing protein stores. The value for F , is quite constant, however, under constant experimental conditions and for a given physiological state, and it is approximately equal to F , a t low nitrogen intakes in the dog. It is believed, therefore, that absorbed nitrogen can be calculated with some accuracy from equation 4 in the dog. This has been done and absorbed nitrogen instead of nitrogen intake has been plotted against nitrogen balance. A linear relationship between nitrogen balance and absorbed nitrogen was found in the region of negative nitrogen bal-
0
5 10 ABSORBED NITROGEN g./day/rn?
FIG.2. Curves illustrating the relationship between nitrogen balance and absorbed nitrogen in protein depleted dogs.
ance becoming curvilinear in the region of positive balance, curves such as those shown in Fig. 1 being obtained (Allison and Anderson, 1945). Similar curves were presented by Bricker, Mitchell, and Kinsman (1945) for data obtained in man and by Bricker and Mitchell (1947) for the rat. The data plotted in Fig. 2 illustrates the linear relationship between nitrogen balance and absorbed nitrogen in a group of protein-depleted dogs. Data from depleted animals were selected to illustrate the marked extension of the linear relationship into the region of positive nitrogen balance. The greater the degree of depletion the greater the positive nitrogen balance which can be produced by any one kind of protein. Theoretically, if all the protein stores of the body were filled, the animal
164
J A M E S B . ALLISON
could not be put into positive nitrogen balance no matter how much protein was included in the diet. The circles in Fig. 2 describe data obtained while feeding wheat gluten t o the dogs and the angular symbols illustrate similar data obtained while feeding defatted, dried, whole egg protein. These data demonstrate th at less whole egg than wheat gluten nitrogen is required t o produco nitrogen equilibrium or fill the protein stores of the body. The slope of the line which is a measure of the rate a t which the protein stores are being filled is greater for whole egg than for wheat gluten, illustrating the correlation between the slope and biological utilization of the protein. Furthermore, each dietary protein has a n upper limit for utilization. Feeding more than th a t amount is inefficient and may put an excess burden on the organism to form and excrete organic compounds containing waste nitrogen. The mathematical expressions for the linear portion of the curve in Fig. 2 is:
B
=
K A - Eo
(5)
where B is nitrogen balance, A is absorbed nitrogen, and E , is the excretion of nitrogen when the nitrogen intake is zero. E o is the y intercept and K is the slope of the line. 4. Nitrogen Balance Index of Absorbed Nitrogen
The slope ( K ) of the,lines in Fig. 2 has been called the nitrogen balance index of absorbed nitrogen by Allison, Anderson, and Seeley (1915), because this slope is the rate of change of nitrogen balance with respect to absorbed nitrogen, a measure of the rate a t which the protein stores of the body are being filled. These authors defined the index as the tangent to any point on the curve giving the rate a t th a t point of absorbed nitrogen. I n the regions where the curve is linear the index measures t,he rate of greatest change in protein stores and, unless otherwise specified, will be calculated only for these regions. This calculation of the index for absorbed nitrogen, however, is not new. All determinations for biological values, where the fraction of nitrogen rctained in the body of the animal was being sought, have assumed 3 linear relationship between nitrogen balance and absorbed nitrogen. It will be shown later in this review that, if the amount of so-called body nitrogen excreted in the feces and in the urine is constant and separate from food nitrogen, the slopes of these lines are the fractions of nitrogen rrt:iined in the body of the animal; they are the biological values of the protcins as defined by Thomas and developed by Mitchell. It will be demonstrated, however, that K may be a function of rather than equal to the biological value.
165
BIOLOGICAL EVALUATION OF PROTEINS
For that reason the term nitrogen balance index of absorbed nitrogen is retained. The data recorded in Table I summarizes some of the digestibilities and nitrogen balance indexes for absorbed nitrogen which have been obtained for protein sources fed to adult dogs. The “net protein values” recorded in the last column were calculated by multiplying nitrogen intake (I) by digestibility ( D ) , by nitrogen balance index of absorbed nitrogen ( K ) and by the per cent protein in the source. Net protein TABLE I Digeatibilitiea, Nitrogen Balance Indezes of Absorbed Nitrogen, and Net Protein Values Obtained by Feeding Diflerent Proteins to Normal Adult Dogs --
Protein source
+
Casein methionine Fibrin hydrolyzate methionine Egg white Lactalbumin Lactalbumin hydrolyzate Bovine round roast Bovine rib roast Casein Casein hydrolyzate Wheat gluten lysine Chicken entrails Flounder entrails Flounder heads Wheat gluten a-Protein
+
+
--
Digestibility
0.95 0.97 0.95 0.97 0.98 0.99 0.99 0.95 0.99 0.95 0.95 0.96 0.83 0.95 0.82
Nitrogen balance index
Net protein valuc?
1.5
1.1 1. o 0.86 0.74 0.80 0.14 0.14 0.65 0.64 0.64 0.09 0.11 0.06 0.31 0.28
I .2 1.1 1 .o 1 .o 0.85 0.83 0.80 0.80 0.82 0.77 0.77 0.52 0.40 0.39
d
values, the use of which was suggested by Mitchell and Carman (1924), are functions of the retention of nitrogen by the animal, digestibility, and the concentration of protein in the source. These net values measure the rate of filling of the protein stores of the animal by the protein source; they do not evaluate the protein as such in the source. The low net protein value of bovine roast, compared to casein, results from the fact that the meat contains more water and other nonprotein constituents than the casein. Actually the meat protein is a bit superior t o the casein, as indicated by its higher nitrogen balance index for absorbed nitrogen. The effects on the index of improving the pattern of amino acid by supplementation is illustrated by its increase when lysine is added to wheat gluten or methionine to fibrin or casein. Nitrogen balance indexes increase as the animal is depleted in proteins.
166
J A M E S B . ALLISON
This increase is illustrated by the data recorded in Table 11, the effect of depletion being more marked on the indexes of the poorer proteins. Thus, when there is greater need, more nitrogen is retained and conserved in the body of the animal. I n this connection, Silber, Seeler, and Howe TABLE I1 Nitrogen Balance Indexes i n Normal and Protein-Depleted Dogs
-
-
Nitrogen balance index Protein source
Normal Egg white Casein Casein hydrolyzate Wheat gluten a-Protein
-
0.96 0.80 0.80 0.44 0.39
~
Depleted 1.2
0.93 0.92
0.70 0.73
-
(1946) found that depletion in proteins reduced the excretion of amino nitrogen following intravenous administration of amino acid mixtures. The variable nitrogen balance (B) has in it the variable absorbed nitrogen ( A ) , so th at a less complicated equation can be obtained by simplifying equation 5 to the following:
u = (1 - K ) A + u,
(6)
where U is urinary nitrogen excretion, K is the nitrogen balance index of absorbed nitrogen, and U o is the excretion of urinary nitrogen on a protein-free diet. This form of the equation is most useful in calculation of the indexes. It has been used, too, for the calculation of indexes of hydrolyzates fed intravenously into dogs (Allison, Seeley, and Ferguson, 1947). The value C', can be determined by feeding orally a source of protein with a nitrogen balance index of 1. Under these conditions, if the amount of protein nitrogen fed is sufficient to bring the dog into nitrogen equilibrium, the urinary nitrogen excreted will be equal to U , and the animal will not be depleted in proteins. Proteins, with indexes of 1, which have been used for this purpose are found in whole egg, lactalbumin, and egg white. Whole egg was used with particularly good success, the dog being fed this protein source once a week to determine U,, the remainder of the week being devoted to the determination of indexes for hydrolyzates fed intravenously. If a t any time U , dropped below 60 mg./day/kg. body weight the index tended to increase, becoming larger as U , decreased, reflecting the depletion in the protein stores of the animal (Allison, Seeley, and Ferguson, 1947).
BIOLOGICAL EVALUATION O F P R O T E I N S
167
5 . Nitrogen Balance Index a Function of Nitrogen Retention
The nitrogen balance index of a dietary protein has been defined as a function of the rate of filling of the protein stores of a n animal at any given nitrogen intake. This index can be expressed also a s a function of the amount of nitrogen retained in the body of the animal-nitrogen which is used t o construct and repair the living system. The following derivation illustrates the correlation between the index and nitrogen retention. The excretion of nitrogen on a nitrogen-free diet (En) is expressed as follows:
Eo = (Fo
+ Uo)
(7)
where F o is the excretion in the feces and U o the excretion in the urine. Equation 5 can be written, therefore, as:
KA = B
+ F , + Un
(8)
Since by definition:
B=I-F-U then :
KA
=
I
-
F
-
U
(9)
+ Fo + U o
(10)
Absorbed nitrogen is calculated as follows: A = I - ( F - Fo) (11) By substituting equation 11 in equation 10 and solving for K , we obtain the following relationship:
K =
A
-
(I'
-
Uo)
A
Equation 12 states that K is a function of the fraction of nitrogen retained in the body of the animal. K is equal to this fraction when U o represents the excretion of body nitrogen a t all levels of absorbed nitrogen. Under these conditions, U - U o would be equal to the excretion of urinary nitrogen derived from the food. Data are accumulating to demonstrate, however, that body nitrogen excretion is not the same at all levels of nitrogen intake. The conservation of body nitrogen through feeding nitrogen, for example, is illustrated by curve I11 plotted in Fig. 3 (Allison, Seeley, Brown, and Anderson, 1946). The open circles record data obtained by feeding the protein-free diet, the black circles, by feeding protein nitrogen. Cuscin was fed to secure the data plotted in curve I. Since the nitrogen balance index ( K ) for casein is less than 1 (0.8), the excretion of urinary nitrogen increased during the nitrogen-feeding period. Curve I1 illustrates data obtained by feeding a dog egg white protein. The index for
168
JAMES B. ALLISON
this protein was 1, so that the excretion of nitrogen was not altered by feeding nitrogen. Curve I11 resulted from feeding egg white to a protein-depleted dog. Under these conditions the nitrogen balance was greater than 1, the urinary nitrogen excreted during the nitrogenfeeding period being less than during the protein-free feeding period. Thus, in the depleted dog, the excretion of nitrogen from the protein stores was less while the animals were receiving nitrogen than during the protein-free feeding period. It will be demonstrated later that proteindepleted animals are not always repleted uniformly in their protein stores, I
z
g 0
1.0
-c--LLQ* 0-* 0- -0ea-
a
I
t 2 >-
1.5
- (=Q,
a 4
z_
a: 3
1.0-
I
I
1
P
m
'*** I
I
FIG.3. Urinary nitrogen (g./day/m.* body surface area) excreted by dogs on a protein-free diet ( 0 )or a diet containing protein nitrogen (0). (Allison, Seeley, Brown, and Anderson, 1946.)
some stores being repleted more rapidly than others. It is not unreasonable, therefore, to find that during the feeding of protein the stores are filled in such a way that body nitrogen is conserved. Some of the protein stores, for example, which contribute to the excretion of nitrogen on a protein-free diet could mix with dietary nitrogen to build protein tissues of the body, thus reducing the excretion of body nitrogen, a form of internal supplementatiop. Dietary proteins such as casein to which an optimum amount of methionine has been added build u p protein stores in the normal dog in such a way that nitrogen balance indexes ( K ) greater than 1 are obtained. The only way indexes greater than 1 can be obtained is for dietary nitrogen to reduce the excretion of body nitrogen. When dogs
BIOLOGICAL EVALUATION OF PROTEINS
169
are receiving a protein-free diet they may be forced to draw upon tissue proteins to supply sufficient methionine for metabolic needs. The breakdown of tissue to supply the methionine on a protein-free diet could be reflected by a high excretion of urinary body nitrogen. If sufficient methionine is supplied in the diet this drain on the tissue may be relieved reducing the excretion of nitrogen. Much larger than optimum amounts of methionine, however, cause marked reduction in certain body protein stores in animals. Brown and Allison (1948), for example, fed a relatively large excess of methionine in a diet containing casein to rats. The animals lost nitrogen from their bodies but the proteins of the liver, kidneys, and plasma globulins were increased. Thus some tissues were being built up while others were torn down under the influence of the abnormal pattern of amino acids in the diet, the abnormal pattern increasing the excretion of body nitrogen. Addition of arginine with the methionine to the casein corrected, in whole or in part, these abnormal shifts of nitrogen in stores of the rat. These experiments and others suggest that the separation of food from body nitrogen is a difficult task. 6. Biological Value
The separation of food from body nitrogen is a necessary task, however, to calculate “ biological values ’’ of dietary protein as defined by Thomas and Mitchell. These values have been defined by them as the fraction of absorbed food nitrogen retained in the body of the animal, a fraction which is often calculated using a relationship such as equation 12. To make this calculation it is assumed that U , is equal to the excretion of body nitrogen a t all nitrogen intakes as well as at zero levels. The reality of the separation of body from food nitrogen, has been re-emphasized by Block and Mitchell (1946-1947). It is possible, as they point out, that certain reactions are irreversible and free from disturbances of dietary amino acids. Indeed, if all the protein stores of the body are filled or repleted uniformly during protein feeding it seems logical to expect a relatively constant minimum body nitrogen excretion. This excretion may shift, however, during experimental periods because of alterations in the equilibria between the protein stores of the body. The ideal experiment to determine “biological value” would be set up so that the physiological state of the animal is not altered. Various methods for measuring the excretion of body nitrogen without altering the physiological state have been discussed by Mitchell. Most of these methods involve a short-time measurement of nitrogen excretion on a protein-free or low-nitrogen diet containing a protein of good quality.
170
JAMES B. ALLISON
The low-protein diet often used to measure the excretion of body nitrogen contains proteins of excellent quality and digestibility such as those in whole egg. If the nitrogen of the whole egg is completely retained to maintain the nitrogen integrity of the animal, the excretion of urinary nitrogen represents the minimum protein requirement for maintenance, a requirement that is taken as a measure of the excretion of body nitrogen. Recently, Murlin et al. (1946a) have measured the excretion of nitrogen in man fed a protein-free diet with the aim to calculate “biological values” for dietary proteins. They found that urinary nitrogen U , varied in man with: “(a) level of protein in the pre-experimental diets; (b) the position of the no-protein period in the series of periods; (c) the nature of the protein, called here ‘supporting protein,’ immediately preceding the no-protein, and its level of intake; and (d) conditions antecedent t o the supporting proteins which could affect the accrued nitrogen deficit to the beginning of the no-protein period,” the same factors which cause variations in experimental animals a n d emphasize the susceptibility of this measurement to changing nutrition and physiological state of the animal. Murlin and coworkers, however, were able to measure a value for U,in man by using short-term feeding experiments. They demonstrated a mathematical relationship between body weight and urinary nitrogen ( U o ) . Thus even though the excretions of nitrogen on a protein-free diet are variable, values which satisfy equation 12 for a given physiological state were determined. Whether or not the ‘‘ biological values” so calculated are the actual amounts of nitrogen retained in the body they are functions of th at amount; they are the rate a t which the protein stores are filled. These authors emphasized the necessity for performing the experiments in a definite order if comparisons were to be made between nutritive values of different proteins and amino acid mixtures. They pointed out, for example, th a t the depleting effects of a protein-free period promoted retention of a protein consumed immediately afterward, results similar to those found in animals (Allison ct al., 1946). They emphasized too, that values calculated from data obtained in the region of positive nitrogen balance may be low because of the curvilinear nature of the relationship between nitrogen intake and nitrogen balance in this region. Murlin and associates found that the “biological values’’ of mixtures of amino acids were always lower than the natural protein by 10 to 40%. They explained this lower value as due, in whole or in part, to the presence of unnatural isomers in the mixture of amino acids. Their data support the conclusion that “ th e unnatural isomers in these experiments, insofar as they escape deamination, belong t o this class of dispensable compounds; while insofar a s they are deaminated and can be recognized as contributingextra nitrogen to the
BIOLOGICAL EVALUATION O F PROTEINS
171
urea and ammonia fractions of the urine, they are in the same class as non-essential amino acids” (Murlin et al., 194613).
7. Biological Eficiency Swanson and coworkers (Willman e l al., 1947; Brush et al., 1947) used a standardized procedure for the determination of ‘(biological value” in rats, a procedure based on the classical nitrogen balance test as developed by Mitchell (1924) and which was designed to satisfy the requirements of equation 12. They obtained values, however, greater than 1, which they also recognized could not, by definition, be “biological values.” Thus they called their determination a measure of biological “efficiency ” rather than of biological “value.” They recognized that using these procedures the feeding of nitrogen reduced the excretion of nitrogen from the protein stores of the animal. They measured body nitrogenexcretion dy placing the rats on a nitrogen-low diet until they were somewhat depleted in proteins. This was done so that a constant value for excretion of nitrogen on a protein-free diet could be obtained for substitution in equation 12. It has been pointed out previously that this excretion decreases rapidly a t first and then more gradually until the excretion is essentially constant in a protein-depleted animal. This procedure undoubtedly depleted their rats in proteins so that during repletion the protein stores were shifted so as to reduce the excretion of nitrogen from the animal. Their methods of calculating and illustrating their results are shown graphically in Fig. 4. The body nitrogen spared by methionine, illustrated in Fig. 4, demonstrates that conservation can result from feeding single amino acids. Miller (1944) and Allison, Anderson, and Seeley (1946, 1947) observed a reduction in the excretion of urinary nitrogen when methionine was added to a protein-free diet fed to dogs, a reduction due primarily to decreased excretion of urea nitrogen. Johnson, Deuel, Morehouse, and Mehl (1947), on the other hand, found that the addition of DL-methionhe to low-protein diets of man did not decrease the excretion of urinary nitrogen. The addition of methionine provided a n excess amount in intermediary metabolism, the excess appearing in the urine as sulfate or as unchanged methionine. Similarly, Schwimmer and McGavack (1948) have reported that methionine did not conserve body nitrogen in man receiving a protein-free diet. The work of Cox et al. (1947) supplements these experiments demonstrating that the addition of methionine to casein hydrolyzate did not increase nitrogen retention in man while it did so in the rat and in the dog. Excess methionine, however, reduced the utilization of nitrogen in the rat (Swanson et al., 1946) and in the dog (Brown and Allison, 1947). These
.
172
J A M E S B . ALLISON
and other unpublished data indicate that the rat and the dog are more easily depleted in methionine than is man. Brush et al. (1947) demonstrated further, in the rat, that cystine, choline, and all the essential amino acids, except phenylalanine, valine, and tryptophan, exert some nitrogen-sparing action. The fact that single amino acids can reduce the I. NITROGEN-LOW.
3. THE I0 ESSENTIAL l Y l N O ACIDS
2. EGG PROTLINS.
4 , METHlONlNE
5. ALL ESSENTIAL AM1
!
I
EXCEPT YETHIONIM
I c 140
0
m 40-
2 120-
0
0 30-
i
5 20-
0
L
n
$100-
0
B t z
5 z
u 1
2
3
60-
4
-o’q0
0
5
k
I 40-
a
,o
EO-
a
10-
1
10-
”
20
d
-
20;
!30t
2ot
I
A. NITROGEN I N URINE.
0. 000V NITROGEN SPARE0
FIQ.4. Nitrogen in urine and body.nitrogen spared when rats were fed a basal nitrogen-low ration (l),the ration supplemented with egg protein (2), the ten essential amino acids (3), methionine (4), or the essential amino acids except methionine (5) (Brush, Willman, and Swanson, 1947).
excretion of nitrogen supports the belief that these acids alter the metabolism of the protein stores. These acids may be in such deficiency while the animals are on a protein-free diet that protein reserves are utilized to supply them, the excess catabolism thereby contributing to the excretion of nitrogen. Conservation of this nitrogen could then take place when these acids are included in the diet.
8. Replacement Value Murlin and coworkers (1938) have introduced a method of interpreting nitrogen balances that utilizes the concept of a reference protein of
BIOLOGICAL EVALUATION O F PROTEINS
173
high nutritive value. Such a method is particularly useful in human nutrition where low-protein diets are unpalatable. In its original form this technique was simply a comparison of nitrogen balances of adults fed a given quantity of reference protein nitrogen of high nutritive value, such as the proteins of milk or egg, with the balances produced while receiving the same intake of nitrogen from the food under test. Mitchell (1944) has applied this replacement test successfully to rats using the following formula for calculation of the value ( V R ) . v n = 100
- Bz - Bi ___ I
x
100
where B1 is the nitrogen balance of the rat subsisting on the reference protein diet, Bz is the nitrogen balance of its pair mate subsisting on the protein under test, and I is the average of the nitrogen intakes of the two rats in the pair, intakes which should be nearly the same. Mitchell used this technique to study the effect of food processing on the nutritive value of the protein. In these studies B1 was the unprocessed while B* was the processed proteins. Any change from 100% indicated, therefore, the effect of processing on the protein.
111. EVALUATION THROUGH GROWTH Growth has in it most of the phases of metabolism which contribute toward the retention of nitrogen by the animal. Evaluation of proteins or amino acids through growth is, therefore, one of the most rigorous of all methods, integrating most of the functions of proteins into one measurement. The very breadth of this approach to evaluation, however, makes it difficult to standardize and to interpret (see reviews by Mitchell, 1944, and Barnes and Bosshardt, 1946). 1. Growth and Nitrogen Retention
The concept of nitrogen retention (biological value) was applied originally to the protein requirements for maintenance in adult animals but Mitchell and coworkers (Mitchell (1944) have extended this measurement to growing animals. The nitrogen retained in growing animals is the sum of the fractions of nitrogen retained for growth and for maintenance. Barnes, Bates, and Maack (1946) have shown that the relative proportion of dietary nitrogen entering into growth and maintenance varies markedly depending upon the amount and the nutritive quality of dietary proteins. They found that there was an increase in consumption of dietary proteins by the young rats as the amount of protein source in the diet was raised. The increased consumption of protein increased the rate of growth until a maximum was obtained. They used four
174
JAMES I). ALLISON
protein sources in their study, whole egg, soy flour No. I, soy flour No. 2, and wheat gluten. Soy flour No. 1 was processed to develop a highquality protein; soy flour No. 2 was processed to produce a n intermediatequality protein. T o determine the minimum maintenance requirements of young 60-g. rats, grams of protein gained was plotted against grams of protein absorbed, and the absorbed nitrogen a t zero body protein gain measured by extrapolation. The protein minima for nitrogen equilibrium were
80
-
60
-
z
0
G
N
4
! i z I-
y
B
40
-
20
-
WASTE ENERGI
BlOL VALUE
PROTElN G U N
MAINTENANCE
~.
0
20
40 60 80 GRAMS PROTEIN ABSORBEC
100
FIG.5 . Comparative utilization of absorbed protein (well heat-treated soy flour) in rats for maintenance and growth (Barnes and Bosshardt, 1946).
determined also in adult rats using the techniques of hlelnick and Cowgill (1937), described previously in this review. These minimum requirements expressed on the basis of body surface area were essentially the same for young and adult rats, providing a basis for the calculation of the amount of nitrogen utilized for maintenance during the 42-day growth period used in this study. The average amount of protein utilized for maintenance was calculated assuming that whole egg was 100% utilized for maintenance a t low levels of intake, an assumption that is supported by the work of Mitchell and Carman (1926), Murlin et al. (1938), and by the nitrogen balance studies in the dog. The per cent utilization of soybean protein for maintenance and growth is illustrated in Fig. 5 (Barnes et al., 1946). As protein consumption increases the fraction of absorbed proteins utilized for maintenance decreases, during which the fraction utilized for growth rises to a maximum and then falls. Thus there is a decrease in biological value as the ’ maintenance food consumption increases. The biological value at 100%
175
BIOLOGICAL EVALUATION O F P R O T E I N S
corresponds in magnitude to measurements made on adult animals in the region of negative or low-positive nitrogen balance. The relative participation of growth and maintenance in the establishment of the biological value is recorded in Table 111. Since the protein requirements TABLE 111 lielalive Participation of Growth and Maintenance i n Establishment of Biologicd Value with Diets of about 10% Protein”
1 Protein sourre
Whole egg Soyflour No. 1 Soyflour No. 2 Wheat gluten
Protein apparently absorbed
36 27 20 20
1 Biological value, %
Relative Relative participation participation of growth, of maintenance, yo %
99 67 64
77
35
26
65 59
23 35 41 74 _i__l_
*Barnes and Bosshardt (1946).
for growth and maintenance are different, these two factors should be measured independently and not in combination (Barnes et al., 1946).
2. Protein Eficiency Ratios The methods used most extensively for the determination of nutritive quality of proteins in growing animals are derived from the work of Osborne, Mendel, and Ferry (1919). These authors expressed the nutritive quality of a protein as a protein efficiency ratio (grams gain in body weight per gram protein eaten). The original method of Osborne el al. required that the maximum ratio be established by feeding varying amounts of protein, this maximum being taken a s the index to the nutritive value of the protein for growth. Few laboratories have followed this requirement to determine the maximum ratio but have fed a constant, usually a lO%-protein, diet. Barnes and Bosshardt (1946) have reexamined this method in rats, applying it also to mice. They concluded that “the common practice of employing a 10 per cent protein diet, regardless of the nature of the protein, will result in a considerable distortion of nutritive values, and the magnitude of the error will increase as the nutritive quality of the protein decreases.” Paired feeding is often used in place of ad libitum feeding to equalize the ingestion of the test proteins under comparative assay. These authors have pointed out, however, that if the restriction is severe in paired feeding some of the better protein may be wasted as fuel. Zucker et al. (1941) emphasized:
176
JAMES B . ALLISON
“the absolute daily intake of any factor fed a t a constant level in a diet supplied ad libitum increases steadily with growth, and the absolute requirement may increase more slowly than the food intake, or actually decrease.” These are variables which often have not been considered adequately in evaluating proteins in growing tests. Hartr, Travers, and Sarich (1947) have made a comparison of ( a ) litter mates us. randomly selected males, and ( b ) moderate restriction of food intake us. ad Zzbitum feeding in the determination of protein eficiencirs in rats. They concluded that preconditioning of the population of rats by feeding 10%casein ration for one week improves subsequent tests somewhat. Partial restriction of food intake (10 g. of casein ration daily) reduced the growth response on 10% casein by approximately 20% below the mean for ad libitum fed animals. The variance, however, for the animals restricted in quantity of diet was only about one-eighth to one-tenth that observed for animals fed ad Izbitum. Thus, Cood restriction improved the discriminatory capacity of a protein efficiency assay on rats. These authors, emphasized, however, the need to study further the physiological implication of partially restricting the food intake.
3. Protein Eficiency and Nitrogen Retention Mitchell (1942, 1944) has discussed thoroughly the shortcomings of the protein efficiency method for comparative assay of dietary proteins, shortcomings such :is those recorded above. He emphasized th a t the method assumes (1) that there is no requirement for maintenance and (2) that the protein content of the gains in body weight of growing animals is constant. Similarly, Albanese (1947), in a review of the amino acid requirements of man, pointed out: “ th a t composition of weight gain differs in quality, sometimes it may predominate in fluids and other times in fat or protoplasmic tissues. This discrepancy also raises questions as to the function and metabolic fatc of the retained nitrogen which fails t o appear as body tissue.” Probably one of the best examples of “the disturbing effect of a variable cornposition body weight gains in the interpretation of a protein nutrition experiment” (Mitchell, 1944) is illustrated by a n experiment reported from Mitchell’s laboratory (Be‘adles et al., 1933). These authors demonstrated that in paired-feeding tests, the protein of Limburger cheese is equal to th at of fresh milk curd in growth-promoting value, although it is less digestible and has a lower biological value. The explanation lay in the composition of the body weight gains, the gain achieved by the Limburger cheese ration being definitely lower in protein and higher in fat content than the gains produced on the milk curd ration. The following experiments on the growth of puppies illustrates also
177
BIOLOQICAL EVALUATION OF PROTEINS
the effect that the quality of protein can have on the type of body weight gain. Six 12-week-old beagle puppies were put on a protein-free diet (Allison and Anderson, 1945) to which 25% wheat gluten had been added. These puppies came from two litters, four from one and two from another. The daily caloric intake/kg. body weight over a period of 70 days are plotted in Fig. 6. The caloric intakes gradually decreased from a n initial value of approximately 225 cal./day/kg. to around 125 cal./day/kg. a t the end of 70 days of wheat gluten feeding. The dogs grew well during
200-
1
''$\g b'---F
$ 150-
2 \
2-
43'
Ann '
n -0
Q
loo-
a
ss
50-
@ 1
I
I
I
1
I
DAYS
FIG.6. Cal./day/kg. body wcight catcn by litter mate puppies fed diet containing 25 % wheat gluten. After 40 days of feeding wheat gluten, two litter mates wore fed equivalent amount whale egg pratcin nitrogen.
this period and appeared to be in good physical condition but a close examination demonstrated that they were soft and fat. Three of the litter mates were removed from the wheat gluten diet a t the end of 40 days and placed on an equivalent amount of defatted whole egg nitrogen. The caloric intake of the dogs fed whole egg protein immediately decreased below the intake of those on wheat gluten. The circles in the rectangle in Fig. 6 illustrate the caloric intakes of these dogs. The growth of the dogs receiving wheat gluten is illustrated in Fig. 7. All dogs received wheat gluten for 40 days, a t which time dogs 90 and 89 were placed on the diet containing an equivalent amount of whole egg nitrogen (indicated by circles). The loss in weight which accompanied the transfer to whole egg is illustrated by these data. This loss in weight, however, is not the result of a decrease in nitrogen. The data in Table I V demonstrate that the dogs fed whole egg were gaining nitrogen better than those fed wheat gluten. These data were obtained during two
178
JAMES B. ALLISON
TABLE IV Nitrogen Source, Intake, Excretion, and Gain during Two Weekly Periods of Growth in Dogs'
I
Period I Dog No.
Nitrogen ~ource
NitroNitrogen gen intake excreted
__
86 87 89 90
-
_
Wheat gluten Wheat gluten Whole egg Whole egg
60.9 42.4 26 .O 32.0
-
Nitrogen gained
Period 11 Nitrogen intake
_
53.7 29.0 11.3 10.7
NitroBody nitrogen gen excreted gained
_
7.2 13.4 14.7 21.3
~
59.7 44.2 33.6 26.2
52.9 40.1 15.2 10.7
~
6.8 4.1 18.4 15.5
See Fig. 7. Valuefi given in g./week.
successive weekly collection periods, beginning on the forty-third day, 3 days after two of the four dogs had been changed to the whole egg diet. The average gain in nitrogen of the wheat gluten dogs during those collection periods was 7.9 g. The average gain in nitrogen in the dogs receiving whole egg over those same periods was 17.5 g.
2 10
I
Dog
*
I-
I
2
s
> 005 m 10
~CIG.
7.
20
30
40 -..,,.
50
60
70
Gain in body weight of litter mates fed diet containing 25 % ' wheat gluten for 40 days.
In a second experiment a group of three litter mates were fed the whole-egg diet. Two from this same litter and one beagle puppy of the same age as the others (12 weeks) were placed on the wheat gluten diet. The grams body weight gained per gram nitrogen intake was measured over a 3-week period for each puppy. These values recorded in the third column of Table V are essentially the same whether the dogs were fed wheat gluten or whole egg. The nitrogen intake, nitrogen excreted, and nitrogen gained the last 2 weeks of the experiment are also listed in Table V. These data demonstrated that the body nitrogen gained per
179
BIOLOGICAL EVALUATION OF PROTEINS
TABLE V Nitrogen Intake, Excretion, and Gain by Dogs on dl-Day Diet of Wheat Gluten or Whole Egg
-
Wheat gluten Wheat gluten Wheat gluten Whole egg Whole egg Whole egg
-
G Z
62.8 61.3 50.2 66.9 45.3 48.1
51.1 47.8 43.6 31.8 22.4 25.1
11.7 13.5 6.6 35.1 22.9 23.0
0.19 0.22 0.13 0.52 0.51 0.48
Nitrogen
G . gain/g. intake, I ,
Nitrogen source
91 93 94 95 96 97
g./week
Nitrogen excreted, g./week
Body nitrogen gained, G, g./week
nitrogen
,
9.6 8.0 8.6 10.6 8.3 8.0
gram nitrogen intake is much greater for whole egg than for wheat gluten, emphasizing the error that may be encountered in using growth curves to evaluate proteins in dogs. Hegsted and Worcester (1947), however, found a very high correlation between gain in weight and protein efficiency measurements in rats. They found too that protein efficiency is a function of weight gain rather than a characteristic of the dietary protein. They concluded that little additional information is gained in these growth methods by calculating protein efficiencies. In some studies on the utilization of proteins of white and whole-wheat flour, Chick, Copping, and Slack (1946) found that protein efficiency ratios reflected the retention of nitrogen in the body of animals as determined by cwcass analysis. Both good and poor correlation between growth and nitrogen retained are obtained, therefore, depending upon the nature of the growth of the animal. One of the fundamental questions which a growth method presents is: “What is normal growth?” This question has been answered, in part, by studies of Zucker et al. (1941, 1942) on the growth of the rat in relation to the diet. Their analysis emphasizes the fact that the shape of the growth curve has great significance in interpreting the response of the animal to different nutritional states. They found that the growth of rats on good stock diets is characterized by a “progressive decrease in gain for each successive time interval after weaning.” The following equation describes their data: log
w
=
-(k/t)
+ log c
(14)
where W is the weight of the rat and t is time. “Neither natural variation in size of the animal nor artificial stimulation of growth rate causes a deviation from the empirical formula of growth. While inherent size
180
JAMES B. ALLISON
of rats varies considerably, the slope of the plot (k of the formula) varies but little for each sex in albino rats. I n all cases where data for males and females are available, the ratio of k for males and k for females is constant within 3%.” Thus B norm is established for rats in which there are few if any restrictions to growth. This norm furnishes a basis for studies on deviations in animals fed diets suboptimal for growth. B y this knowledge (Zucker and Zucker, 1943, 1944; Zucker et al., 1941, discovered that there is a factor necessary for optimal growth associated with liver and certain proteins of good quality. Standard procedures are needed in growth methods so that results from different laboratories can be compared. Barnes, Rosshardt, and associates are developing such procedures applying them particularly to the mouse (Bosshardt et al., 1946). The mouse has advantages over the rat, such as smaller size, lower food consumption, and shorter test periods. These authors demonstrated the importance of a standard pretest treatment of the animals. They found that a t dietary protein levels giving maximum protein efficiency ratios, body weight gain is a true index of comparative body protein gains. They pointed out, however, that growth cannot be used as an absolute index of utilization of absorbed protein for body protein gain. They determined, therefore, the true dietary levels in the mouse for maximal utilization for growth in body protein, finding it to be approximately 6.5% of extracted whole egg, 8.0% of casein, and 25% of wheat gluten, corresponding to absorption of 3.5 g., 4.5 g., and 12 g., respectively. The calculation of absorbed proteins was made possible by the measurement of “metabolic fecal nitrogen” using the method described by Bosshardt and Barnes (1945).
IV. EVALUATION THROUGH TISSUE REGENERATION Animals do not possess adequate stores of free amino acids. For th a t reason, when they are placed on diets low in proteins there is a n immediate loss of body protein stores, a loss which reveals the dynamic equilibrium that exists between the proteins of the various tissues of the body. This concept of a dynamic equilibrium, developed by Whipple and associates a t Rochester, is described by them briefly in one of their recent papers as follows: “Food proteins yield the amino acids absorbed from the intestinal tract, and the amino acids are synthesized, in the liver cells (and elsewhere) into plasma proteins. These plasma proteins, (and amino acids) supply the protein requirements of the body cells. Normally, there is a considerable reserve of plasma protein-forming material (1 to 5 times the circulating mass), which reserve may be reduced by fasting, low protein diet, or plasma depletion. This depletion of protein reserves lowers the body resistance t o infection and intoxication. Thus, body
181
BIOLOGICAL EVALUATION OF PROTEINS
protein stores, protein production, and protein wear and tear are in a nicely balanced or steady state, a dynamic equilibrium.” 1. Depletion in Proteins
Depletion in proteins reduces the body protein stores, causes a shift in water balance, and impairs the functions of organs. Depletion in protein stores of the body is studied most often through changes which take place in the plasma proteins. Chemical fractionation of the plasma proteins from animals in the protein-depleted state proved that albumin was reduced markedly below normal while the globulins were reduced less or not at all (see Fig. 8). Zeldis et al. (1945) studied the effects of
CONTROL
BEFfflE
FEEDING
l m l
5MY5
*FmFEEDING
D M Y S AFTER
FEEDING
A A ALL A*,
20 MY5 N T E R FEEDlffi
28 MY5 AFTER FEEDING
%DAYS AfTER FEEDING
35 M Y S M T E R FEEDlffi
FIG.8. Descending electrophoretic patterns of plasma of a dog following depletion in proteins and repletion with casein hydrolyzate (Chow, Seeley, Allison, and Cole, 1948).
low-protein feeding on the plasma proteins as revealed by electrophoretis analysis. Their findings agree with those of chemical analysis that the albumin fraction is markedly reduced. They point out, however, that the degree of depletion of electrophoretic albumin is considerably greater than that of chemical albumin. Allison, Anderson, and Seeley (1946) presented data also to show that albumin :globulin ratios, determined by salt fractionation, are always greater than ratios determined through electropheresis. Chow, Allison, Cole, and Seeley (1945) found a decrease in electrophoretic plasma albumin and an increase in a-globulin in depleted dogs. Similar changes in plasma proteins were found to be associated with malnutrition, tuberculosis, or cancer in man (Chow, 1946). Chow et al. (1945, 1948) reported also that the yglobulin fraction is reduced below normal in depleted dogs,?a reduction accompanied by a n increased susceptibility to infection.
182
J A M E S B . ALLISON
The increase in the a-globulin fraction during depletion is largely the result of a fall in plasma volume and is not a real increase in total circulating globulin. The plasma volume decreases and the extracellular fluid increases as the plasma albumin falls in the depleted animals (Allison et al., 1947). As pointed out by Zeldis and coworkers, it is difficult to deplete the globulin fraction of the blood, the globulin being more like essential organ tissue protein and less like labile protein reserves. Indeed, when a dog is depleted so that the globulin fraction is markedly decreased, repletion is difficult, the animal often dying in the depleted state. Accompanying an increased lipide fraction in the blood of the depleted dog is impairment of liver function and the production of a fatty liver. I,i and Freeman (1946a, 1946b), for example, demonstrated the production of a fatty liver in dogs fed a 33%-fat, protein-deficient diet for 10 to 16 weeks. Protein depletion impaired hepatic dye clearance and caused an elevation of serum phosphate in dogs (Seeley, 1947). Kosterlitz (1946) demonstrated that liver cytoplasm is lost during fasting and in protein deficiency. The quantity of protein in the liver is rapidly reduced when rats are placed upon low-protein or protein-free diet (Addis, Pool and Lew, 1936a, 1936b; Kosterlitz and Campbell, 1945, 1945-1946, 1946; Harrison and Long, 1945; Brown and Allison, 1947). Other tissues are also effected by depletion in proteins. Li and Freeman (194Gc) found that the incidence of “peptic” ulcers was high in protein-deficient dogs. Armstrong and Haydee (1947) demonstrated that bone atrophy will develop in mature rats fed a protein-deficient diet. Corneal vascularization in rats resulted also from the absence of proteins or different amino acids in the diet (Hall et al., 1946; Niven et al., 1946; Berg et al., 1947; Sydenstricker et al., 1947). Thus corneal tissue like all other tissues of the body require a daily complement of essential amino acids to maintain their integrity. 2. Repletion in Plasma Proteins
The first experimental evidence that the repletion in plasma proteins was influenced by diet was presented by Kerr, Hurwitz, and Whipple (1918). These investigators found that depleted dogs regenerated plasma proteins more rapidly on a mixed protein diet than during fasting. Plasmapheresis was used to deplete the dogs. Their results were substantiated by Smith, Belt, and Whipple (1930), who concluded that part of the increase in plasma proteins, following plasmapheresis, was the result of replacement from a reserve supply of protein held in the body cells. The effect of diet, therefore, upon the plasma protein regeneration was superimposed upon this exchange. IIolman, Mahoney, and Whipple (1934) developed a more quantitative method for the study of the influ-
BIOLOQICAL EVALUATION OF PROTEINS
183
ence of plasma protein formation in dogs. They reduced the plasma protein concentration by plasmapheresis each day, t o from 3.6 to 3.9 grams per cent. The bleeding was Continued over a period of 4 to 6 weeks until a relatively small and constant amount of plasma was removed each day to decrease the plasma protein concentration to a constant low level, the small daily regeneration of protein being attributed to the 7% of vegetable protein in the diet. Test proteins were fed for 7 days to these depleted dogs and plasmapheresis continued daily to reduce the plasma protein concentration to the standard hypopr0,teinemic level. Repletion was expressed a s the “grams of new plasma protein resulting from the feeding of 100 grams of test protein.” Pommerenke, Slavin, Kariher, and Whipple (1935) improved the determination by introducing weekly plasma volume determinations so that the daily removal of plasma could be calculated more accurately. Weekly nitrogen balances were determined also. McNaught, Scott, Woods, and Whipple (193G) found that dogs tolerated a plasma protein concentration of 4.0 to 4.2 better than 3.6 to 3.9 grams per cent, and they maintained their hypoproteinemic dogs a t this level. This method has been developed into one of the best tools for evaluation of proteins in animals (Madden and Whipple, 1940; Madden et al., 1943). 3. Production Ratio In their recent studies the Rochester group have used the doubly depleted dog. Double depletion is produced by sustained bleeding of dogs fed a protein-free or low-protein diet with adequate iron. Thus the reserve stores of blood-protein-producing materials are depleted, levels of 6 to 8 grams per cent of hemoglobin, and 4 to 5 grams per cent of plasma being maintained. New hemoglobin in these dogs may be derived in part from plasma protein. Further studies by Whipple, Miller, and Itobscheit-Robbins (1947) have shown that doubly depleted dogs will continue to produce much plasma and hemoglobin for many weeks while being fed a low-protein or protein-free diet with abundant iron. Thus the blood proteins take priority over other tissue proteins, an example of the “ebb and flow” between tissue, organ, and blood proteins. Body weight was lost as tissue and organ proteins were transformed into blood proteins, the average dog tolerating this raiding of body tissue proteins for from 7 to 11 weeks. For every kilogram of weight loss there was 50 to 140 g. blood proteins formed, the weekly blood protein production ranging from 40 to 66 g. This heavy demand on body protein did not bring about a “premortal rise” in urinary nitrogen, however, the excretion remaining low, conserving body nitrogen. Whipple el al suggest that “ premortal rise ” in many experiments may be
184
JAMES B. ALLISON
associated with a terminal infection leading to catabolism of tissue nitrogen. Whipple, Robscheit-Robbins, and Miller (1946) found that relatively incomplete protein like globin can contribute effectively to the protein stores of the doubly depleted dog and they suggest that “it must obviously be supplemented by the amino acids inadequately represented in globin, drawn from some reserve store, to produce cell proteins, or plasma proteins.” They pointed out that these depleted animals will use efficiently a variety Qf proteins, digests, and the growth mixture (Rose) of pure amino acids. The capacity of dietary nitrogen to replete in these experiments is measured in terms of a production ratio, which is the ratio between protein output and intake. Relative production of hemoglobin with respect to plasma protein is expressed as a ratio between plasma protein and hemoglobin. Hemoglobin (by vein) , dog hemoglobin (tryptic digest by veins), and horse hemoglobin (oral) contribute to the production of blood proteins in these depleted dogs. The utilization of dog hemoglobin is improved by methionine supplementation but not isoleucine (Robscheit-Robbins et al. , 1946a). The addition of isoleucine was found essential, on the other hand, to support normal hematopoiesis with human or beef globin in the rat (Orten, Bourque, Underhill, and Orten, 1945). Further observations by Miller and Alling (1947) demonstrated that when Dbisoleucine is “added to a fed supplement of methionine or methionine and cystine, the utilization of parenterally given hemoglobin nitrogen is even better than the sulfur-containing amino acids alone.” These data emphasize the difficulty of determining the “ideal” amino acid pattern by feeding experiments alone. These authors suggest that the ideal pattern be defined in terms of the total amino acid pattern of each animal. The response in blood protein output and urinary nitrogen excretion to mixtures of essential amino acids has been studied by RobscheitRobbins, Miller, and Whipple (1947). The average production ratio for the ten essential amino acids (Rose) was 19. When one of the essential amino acids was removed the ratio rose to 25. On a good dietary protein the ratio was 15. The authors suggest that good dietary proteins like egg and lactalbumin are utilized to replete protein stores in organs and tissues, this repletion being reflected in a moderate way in the blood proteins. The amino acid mixtures, on the other hand, with or without one of the ten essential acids, cause weight loss, and from this loss materials were derived which accelerated the formation of blood proteins more than when the animal was receiving a good dietary protein. They found that methionine, threonine, phenylalanine, and tryptophan, when eliminated singly from the growth mixtures of amino acids, resulted in a
BIOLOGICAL EVALUATION OF PROTEINS
185
sharp rise in urinary nitrogen, a rise which was corrected when the amino acid was replaced in the mixture. “Histidine, lysine, and valine have a moderate influence upon urinary nitrogen balance toward nitrogen conservation. Leucine, isoleucine, and arginine have minimal or no effect upon urinary nitrogen balance when those individual amino acids are deleted from the complete growth mixture of amino acids during 3 t o 4 week periods. Tryptophane and to a less extent, phenylalanine and threonine, when returned to the amino acid mixture are associated with a conspicuous preponderance of plasma protein output over the hemoglobin output. Arginine, lysine, and histidine when returned to the amino acid mixture are associated with a large preponderance of hemoglobin output.” None of the amino acid mixtures, whether they contain the ten essential acids or not, will prevent weight loss in the doubly depleted dog. A good dietary protein such as casein, lactalbumin, whole egg, or liver protein in amounts of 150 to 250 g. protein/week will produce positive nitrogen balance, maintain body weight, and produce considerable amounts of hemoglobin and plasma proteins. Under comparable conditions, mixtures of essential amino acids will produce positive nitrogen balance and large amounts of hemoglobin and plasma proteins, but such mixtures will not maintain body weight. Miller, RobscheitRobbins, and Whipple (1947) suggest that some unidentified substance is present in the dietary proteins and absent in the mixture of amino acids which may be responsible for maintenance of weight in the doubly depleted dog. This unidentified factor may be similar to that reported by Womack and Rose (1946), who pointed out that protein may contain unidentified substances which, like arginine, are required for maximum increases in weight. 4. Potency Ratio Melnick, Cowgill, and Burack (1936) altered the plasmapheresis method of Holman, Mahoney, and Whipple (1934) in an attempt to increase the significance of regeneration of plasma proteins in proteindepleted dogs. They fed, for example, a protein-free basal diet instead of a low-protein diet so that the basal diet could not stimulate the formation of plasma protein nor act as a supplement t o any test protein added to it. They added the test proteins to the protein-free diet in “equal absolute amounts above the minimum each required t o establish nitrogen equilibrium.” They did this so that an equivalent amount of each test protein, over and above that needed to maintain nitrogen equilibrium, would be available for repletion of plasma and tissue proteins. Seeley (1945) and Allison, Seeley, Brown, and Anderson (1946) have demonstrated that the repletion of plasma proteins in proteindepleted dogs does not take place until the animal is in positive nitrogen
186
JAMES B. ALLISON
balance and that the magnitude of repletion increases with the increase in positive balance. The potency ratio of a protein was expressed by Melnick el al. as the “ratio of (a) the amount of serum protein per week removed by bleeding, above that regenerated by the animal when eating the basal protein-free diet, to (b) the dietary protein increments.”
5. Plasma Albumin Regeneration Weech el al. (1935, 1938) developed a method for studying plasma protein regeneration in dogs, a method which did not involve plasmapheresis. They believed that the maintenance of hypoproteinemia in dogs by repeated plasmapheresis could invoke a repletion in plasma proteins which would be different than that associated with repletion from a lowered concentration toward normal. They depleted dogs mildly, therefore, for 3 weeks by feeding a low-protein diet, followed by 1 week of feeding of the basal diet plus the test protein. The total rise in plasma albumin was measured during the week of protein feeding and this value was used as a measure of the potency value of the protein. By following the increase in plasma albumin they measured the plasma protein, which reflects best the magnitude of protein stores of the animal. 6. Repletion Areas
None of these methods for the determination of repletion in plasma proteins lend themselves easily to a correlation between repletion of the various plasma proteins and the nitrogen balance produced. Seeley (1945) developed a technique so that: “the nitrogen excretion before, during, and after regeneration could be measured, utilization of the protein calculated, regeneration of plasma broteins determined and nitrogen balances established.” The dogs were depleted by feeding a proteinfree diet until the plasma protein concentration was between 4.0 and 4.5 g./100 ml. plasma, and essentially constant. After the plasma protein had been reduced, the dogs were fed sufficient protein to develop a positive nitrogen balance for 5 days, following which they were returned to the protein-free diet. Thus, the dogs were taken from a depleted state through a regeneration period, back to the depleted state. Since the plasma volume was not altered significantly the repletion in proteins could be expressed in terms of concentration changes. This was done by measuring the areas under curves during repletion. Seeley found that repletyon in plasma proteins in hypoproteinemic dogs increased with increasing positive nitrogen balance. Under the experimental conditions used, beef serum protein favored repletion of plasma albumin while casein favored the formation of both albumin and globulin. Similar results were reported using the plasmapheresis technique, by Holman
BIOLOGICAL EVALUATION O F PROTEINS
187
et al. (1934) and by Madden and Whipple (1940). Thus, different patterns of amino acids produce different types of repletion in the hypoproteinemic animal. To study repletion of different types of plasma proteins more adequately and also to measure the total nitrogen utilized by the animal during the repletion process, the techniques of Seeley were altered to embrace a 30-day rather than a 5-day repletion period (Allison, Anderson, and Seeley, 1946). The effects of depletion and repletion on the plasma protein fractions are illustrated in Fig. 8. Using these techniques Chow et al. (1948) demonstrated that a casein hydrolyzate brought about an increase in total circulating albumin and globulins, the globulins being raised to above control values. Lactalbumin hydrolyzate, on the other hand, did not bring about an increase in any of the plasma globulins except the y fraction; most of the repletion being associated with the albumin fraction. The data recorded in Table VI demonstrate that repletion with whole egg, casein, and wheat gluten are correlated well with the nutritive value of these proteins. The gains in plasma protein nitrogen and in body nitrogen are approximately in the ratio of 1:30 previously reported by Elman (1947). AUo in the last column of this table is the difference between protein-free urinary nitrogen excretion in the repleted and depleted states. Since the urinary nitrogen excretion on a protein-free diet increases with an increase in certain labile protein stores of the body, a positive value for A U Omeans an increase in these stores. Values for AUo demonstrate that these labile stores were not increased until the body nitrogen gained was greater than approximately 80 g./m.* of body surface area. Wheat gluten did not replete these stores even after 30 days of feeding (see Fig. 2).
7. Repletion in Liver Protein Liver protein is lost rapidly during short periods of fasting or feeding of diets low in protein (Addis, Poo, and Lew, 1936a, 193613). Much of the background for studies on depletion and repletion in liver prdteins has been prepared by Kosterlitz (1944) and Kosterlitz and Campbell (1945, 1945-1946, 1946), who have presented data to show that the protein lost or gained represents actual liver cytoplasm, both particulate and interparticulate matter taking part. These authofs have called the fractions of liver cytoplasm which are easily lost “labile liver cytoplasm.” They found that the amount of labile cytoplasm present in the liver is directly proportional to the logarithm of the protein intake. They developed two methods for the assay of proteins. The first method measured the ability of the dietary protein t o maintain the labile liver cytoplasm when
188
JAMES B. ALLISON
TABLE VI Data Obtained during 30 Days of Repletion i n Different D(
m
P
N intake
Fecal N
Urinary N
Body N gained
Plasma protein N gained
Avob
Whole Egg
174 175 189 192 336
25.7 26.1 27.7 25.7 38.6
55.3 62.0 60.8 55.3 161.3
93 .O 86.9 100.5 93 .O 136.1
2.9 2.1 3.1 2.9 5.2
+0.7 +0.3 +0.2 +O.l +0.6
75.8 96 .O 95.9 113.7
1.7 2.5 3.9 3 .O
+0.8 +0.6 +0.3
1.7 3.3 2.2 3.4
-0.4 -0.5 -0.2 -0.3
Casein
182 195 196 228
16.3 9.6 15.1 18.9
89.9 88.4 85.0 95.4
-
Wheat Gluten
136 298 333 352
5.1 12.0 21.2 16.9
98.5 215.3 249.1 256.4
32.4 70.7 62.7 78.7
-
Allison, Anderson, and White (1949). Values in g./m.* Difference between protein-free urinary N excretion in repleted and depleted states, g./day /m .a 6
rats were transferred from a stock diet to the test diet. The second method measured ability of the dietary protein to form labile liver cytoplasm after the rats had been on a protein-free diet for 4 days. Similarly, Harrison and Long (1945) developed a method for assay of dietary protein based on the regeneration of liver protein in the rat. The rats were standardized'for 1 week on a diet containing 20% casein and then fasted for 48 hours, after which they were fed for 4 days synthetic diets containing the'protein to be tested. Their data demonstrate that the regeneration of liver protein in the rat, following a fasting period, can be used as a method of assay. Gliadin was found to be superior to rein, both proteins promoting the regeneration of liver proteins better than gelatin. Casein and lactalbumin were essentially the same in ability to stimulate liver protein regeneration, but were both much
BIOLOQICAL EVALUATION O F PROTEINS
189
superior to zein or gliadin. They found too, that the addition of methionine or cystine to a diet inadequate in casein increased the regeneration of liver protein. Recently Gurd, Vars, and Ravdin (1947) reported on another approach to the relation of nitrogen metabolism to the regeneration of liver protein. They achieved a large reduction in liver protein in rats by feeding a nonprotein diet for 14 days followed by hepatectomy. They determined nitrogen balances and new liver protein during a 14-day postoperative period while the animals were being fed the test proteins. They concluded that “ a close correlation appeared to exist between the amount of new liver protein formed and the amount of nitrogen saved or spared, irrespective of whether the rats were in positive or negative nitrogen balance. It is suggested that any factor, or summation of factors, which causes an over-all change in nitrogen metabolifim will affect the liver protein.” Vars and Gurd (1947) showed that “the rate of liver protein production was most rapid in the first two post-operative days, and more rapid in the protein-starved than in casein-fed rats during this period.” The degree of regeneration of liver protein, however, which occurred in protein-starved rats was greatly enhanced by feeding protein postoperatively for 14 days. This regeneration was proportional to the level and nutritive value of the protein fed. These studies on regeneration of liver protein emphasize the unique role of the liver in protein metabolism. The rapidity with which the so-called labile cytoplasm can be decreased or increased according to the demands of the body recalls the concept of a dynamic equilibrium between blood and tissue proteins which has been used so successfully by Whipple and associates. Indeed, liver protein may be increased while skeletal muscle protein is decreased. Brown and Allison (1947), for example, have shown that a large excess of methionine in a diet containing casein is associated with an increase in plasma globulin, and liver and kidney nitrogen, while there is a loss in skeletal-muscle nitrogen. Whipple, Robscheit-Robbins, and Miller (1946) found that relatively incomplete protein and amino acid mixtures have high production ratios for plasma proteins. They suggest that these incomplete amino acid mixtures are supplemented internally. Thus regeneration of any tissue protein of the body will be a function of shifts in the dynamic equilibrium in which they are a part, as well as a function of the pattern of amino acids introduced into the animal body. 8. Repletion in Body Weight Frazier, Wider, Steffee, Woolridge, and Cannon (1947) studied the dietary utilization of mixtures of purified amino acids in protein-depleted
190
JAMES B. ALLISON
adult rats. The rats were placed on a low-protein depletion diet for approximately 3 months. Animals were chosen for repletion using uniformity in initial weights, percentages of weight loss (25-33%), and concentrations of serum proteins (4.064.85%) and hemoglobin (1 1.215.3%) as criteria of depletion (Wissler, Woolridge, Steffee, and Cannon, 1946). The nitrogen source fed t o these rats, other than traces in cornstarch and vitamins, consisted of either sixteen (ration A), ten (ration B), or nine (ration C) purified amino acids. The influence of these rations upon (a) weight and recovery, ( b ) food consumption, and ( c ) plasma and red-cell volumes and regeneration of serum proteins and erythrocytes were studied. The mixtures with sixteen amino acids were patterned after the amino acid composition of casein, the mixture with ten amino acids contained the essential amino acids for growth, and the mixture with nine amino acids lacked one of the essentials. These studies demonstrated that the protein-depleted rats gained weight rapidly and maintained good appetites when fed ration A, B, or the nine essential amino acids, histidine, lysine, tryptophan, phenylalanine, methionine, threonine, leucine, isoleucine, and valine. Arginine was not indispensable. Thus, the same nine amino acids essential for growth are also essential for gain in weight in these depleted rats. Omission of any one of the nine essential acids from the diet led to marked loss of weight and of appetite, the daily food consumption being reduced one-third to one-half. The effects of omission of tryptophan from ration A on weight and appetite are illustrated in Figs. 9 and 10. The weight loss in most cases was more marked than when the rats received a comparable low-protein level ration, a result which suggests that the unbalanced amino acid mixture has abnormal effects in nitrogen metabolism in the rat as in the dog. These authors believe that the omission of any one of the nine indispensable amino acids will produce metabolic disturbances altering the sense of well-being to such an extent that appetite fails. Good appetite and weight gains were maintained when one amino acid (lysine) was administered pnrenterally and others consumed orally. When, however, the lysine was replaced by a salt solution both the weights and appetites of the animals declined. There was no evidences in these studies for accessory polypeptide or other faqtors necessary for normal utilization of amino acids. Benditt, Humphreys, Straube, Wissler, and Steffee (1947) studied the effect of these diets on the synthesis of plasma and erythrocytes in protein-depleted rats. They found that the nine amino acids essential for appetite maintenance and weight gain are also indispensable in the hypoproteinemic rat for construction of serum protein and erythrocytes. A rat bioassay for proteins and protein digests has been developed by
BIOLOGICAL EVALUATION O F PROTEINS
0
191
-“* \
5
10
15
20
DAYS
FIQ. 9. CAmparison of weight changes in protein-depleted rats fed ration A, containing sixteen amino acids, or ration A from which tryptophan had been omitted (Frazier, Wissler, Steffee, Woolridge, and Cannon, 1947).
DAYS
FIQ. 10. “Food consumption areas” in a rat fed ration A, containing sixteen amino acids, followed by ration A from which tryptophan had been omitted (Frazier, Wissler, Steffee, Woolridge, and Cannon, 1947).
Tomarelli and Bernhart (1947) which takes into consideration tissue regeneration after a short period of negative nitrogen balance. They compared the daily amount of casein nitrogen with t,he daily amount of other forms of dietary nitrogen required to maintain the weight of an adult rat for 1 week, after a preliminary period of 1 week on a protein-free
192
JAMES B. ALLISON
diet. They conducted experiments to demonstrate that weight changes are valid criteria of comparative retention of nitrogen in the body of the rat.
V. EVALUATION THROUGH AMINO ACIDANALYSIS The preceding discussions on evaluation emphasize the importance of the presence of essential amino acids in the diet. Recently Cannon (1947, 1948) has emphasized the need to present the essential amino acids to the animal in the proper proportions and in the proper amounts, these proportions and amounts varying with the state of depletion of the various tissues in the body. To quote: ‘If it is true that essential amino acids are not stored individually in the tissues, even temporarily, to be utilized later for purposes of tissue synthesis, and if, moreover, the synthesizing mechanisms which fabricate complete tissue proteins must have available all necessary building stones which make up the particular protein to be synthesized, it would seem likely that the synthesizing mechanisms operate on a ‘perfectionistic’ or ‘all or none’ principle to the extent tha t if they cannot build a complete protein when it is required they will build none a t all. This would suggest also, t h a t the process of synthesis must be total rather than partial, and to be effective the synthesizing mechanism must have available and practically simultaneously all the essential amino acids in adequate proportions and amounts.”
He emphasized too, that this pattern of amino acids and its utilization may vary with the physiological state of animals. A protein-depleted rat, for example, will utilize two to five times the amounts of essential amino acids utilized by a control animal. Moreover the utilization of lysine and leucine are especially elevated, probably because of the greater need for these essential amino acids in the regeneration of skeletalmuscle tissue. 1. Correlation between Amino Acid Composition and Biological Tests The obvious relationship between retention of nitrogen of a dietary protein and the presence of these essential amino acids in the protein have given impetus to correlations between amino acid composition and nitrogen retention. Harte and Travers (1947), for example, in their analysis of human amino acid requirements for nitrogen balance, point out that the nutritive values of proteins are determined largely by the essential amino acid fraction. Mitchell and Block (1946) and Block and Mitchell (1946-1917) have made an extensive study of the correlation of the amino acid composition of proteins with their nutritive value. They showed that the amino acid composition of the protein reveals much concerning the nutritive value of a protein. A correlation between amino acid composition of proteins and biological tests was established by expressing as: “percentage deviations of the contents of the indis-
BIOLOGICAL EVALUATION OF PROTEINS
193
pensable amino acids from the corresponding contents of a protein mixture, that of whole egg which has been shown to be almost completely utilized in the nutrition of the growing rat. With the results expressed in this way, it is possible to detect the limiting essential amino acid in each food protein, and even to assign to each food protein a chemical score that has a high degree of correlation with both the biological value and the protein efficiency ratio determined with growing rats.” This method of correlation gives promise in adult human nutrition but fails in poultry nutrition. 2. Some Causes of Disagreement
The causes of disagreement between any chemical and biological estimations are discussed by Block and Mitchell. It is obvious that, if amino acids are not made available to the organism through digestion or absorption, correlation between amino acid compositions of the protein and nutritive value must fail. Russell and coworkers (1946), for example, found that the availability of methionine in legumes changed with the variety and that the nutritive value of the protein could not be predicted from an analysis for this amino acid. In this connection, Melnick, Oser, and Weiss (1946) made an interesting study of the rate of enzymic digestion of proteins as a factor in nutrition. They suggested that: “for optimum utilization of food proteins all essential amino acids must not only be available for absorption but must also be liberated during digestion in vivo at rates permitting mutual supplementation.” They believe that differences in nutritive value of proteins can be caused by differences in the rate of release of individual amino acids in the intestinal tract. Their data indicate that, during digestion, methionine is liberated at a slower rate than leucine, or lysine in soy protein and that heat processing increases the liberation of methionine, improving the nutritive value of the protein. Recently, the effect of heat treatment of soybean oil meals on digestion and nutritive value of the proteins observed first by Osborne and Mendel (1917) has been investigated extensively. Evans and McGinnis (1946) and Evans, McGinnis, and St. John (1947) demonstrated that moderate heating of raw soybean oil meal increased its nutritive value for growing chicks while autoclaving a t 130°C. for 30 to 60 minutea decreased the value. They found that moderately heated soybean oil meals were more completely digested by the chick or by trypsin and erepsin in udtro than the raw or overheated meals. Similarly, Clandinin, Cravens, Elvehjem, and Halpin (1916) have shown that heating solventextracted soybean flakes in an autoclave a t 15-lb. pressure for 4 minutes resulted in a meal of high nutritive value in chicks. Continued heating
194
JAMES B. ALLISON
for 4 hours reduced the nutritive value. Riesen, Clandinin, Elvehjem, and Cravens (1947) have studied the liberation of the essential amino acids by acid, alkaline, and pancreatic hydrolysis of raw, properly heated, and overheated soybean oil meal. Prolonged heat resulted in a decreased liberation of lysine, arginine, and tryptophan from acid hydrolysis. Proper heat treatment increased, while excessive heat treatment decreased the liberation of essential amino acids by pancreatic hydrolysis. Their data indicate that trypsin inhibitor is not the only factor involved in the improvement of the nutritive value of soybean meal by heat. They point out that: “the amino acid composition of proteins based upon acid hydrolysis does not necessarily indicate the extent of liberation of each amino acid from the protein by enzymatic digestion.’’ Mitchell and Block (1946) have reported a marked reduction in the digestibility and nutritive value of a cereal breakfast food submitted to puffing process. Eldred and Rodney (1946) found that heating casein could bring about changes which reduced the amount of free lysine released by enzymes although the total amount present in the acidhydrolyzed protein was the same in raw (7.5%) as in heated casein (7.7%). Block el al. (1946) found that baking and drying a protein food reduced the availability of its lysine to rats. Recognizing the need for proteins of high nutritive value in a palatable form to supply a highprotein diet to presurgical and postsurgical patients, they prepared a food in which proteins were mutually supplementary. This food consisted of white flour, sugar, egg white, lactalbumin, hydrogenated vegetable oil, dried yeast, molasses, and salt with a distribution of essential amino acids similar to whole-egg pr.,t?ins. The raw cake had a very high protein efficiency (3.3-3.8). Baking snd drying of the cake reduced the protein efficiency to values as low as 1.0. Toasting slices of the cake (100-130”) reduced the values to less than 0.7. Addition of lysine restored the nutritive value (3.2). It is suggested that heat may cause the free carboxyl groups of the dicarboxylic amino acids to react with c-amino groups of lysine to form a new peptide linkage which is resistant to enzymic digestion but not acid hydrolysis. An amino acid analysis of a dietary protein would fail to identify polypeptides or unknown substances other than amino acid combinations which may be essential for maximum utilization of the protein. Woolley (1941, 1946), for example, suggested that streptogenin may be involved in the nutrition of the mouse. Womack and Rose (1946) and Rose and Rice (1939) gave evidence for an unidentified substance in proteins which is required for maximum increases in weight in rats. Miller, RobscheitRobbins, and Whipple (1947) suggested that whole protein may contain
BIOLOGICAL EVALUATION OF PROTEINS
195
something, not found in amino acid mixtures, which is responsible for the increased conservation of nitrogen and gain in weight in doubly depleted dogs. Correlations between amino acid composition and biological evaluations could fail, therefore, because of substances, other than amino acids, present in the protein source which would effect the retention of nitrogen by the animal.
VI. SUMMARY The following summary is not a complete survey of this review on the biological evaluation of proteins, but is an attempt to present briefly some of the more salient features. (1) The protein stores of an animal are the proteins of the tissues and body fluids; these stores are drawn upon to maintain the nitrogen integrity of tissues. ( 2 ) The protein stores can be maintained if the proper complement of amino acids is made available in the diet. Essential amino acids are those which cannot be synthesized a t sufficient rates to meet the needs of the body and must, therefore, be included in the daily diet. The amount of dietary protein needed to maintain the stores of the body is a function of the pattern of essential amino acids and also of the physiological state of the animal. The amount of nitrogen excreted from the protein stores decreases, for example, as the stores are depleted so that less dietary nitrogen is needed to maintain nitrogen equilibrium in a depleted than in a normal animal. (3) The relationship between nitrogen balance and either dietary nitrogen intake or absorbed nitrogen is linear in the region of negative nitrogen balance, becoming curvilinear in the region of positive balance. The slopes of the lines are functions of the rate a t which the protein stores are filled, being high when the nutritive value of the dietary protein is high and low when this value is low. These slopes have been called the nitrogen balance indexes of nitrogen intake and absorbed nitrogen, respectively. They are constant for any one protein in normal animals but they increase in value when the animal is depleted in proteins. ( 4 ) The linear relationship between nitrogen balance and nitrogen intake extends farther and farther into the region of positive balance as the animal is depleted in proteins showing a greater and greater capacity t o grow in nitrogen as the protein stores are reduced. Each protein, representing a different pattern of amino acids, produces a characteristic maximum positive nitrogen balance, the magnitude of which decreases as the nutritive quality decreases. (6) The nitrogen balance index for absorbed nitrogen can be equal to the fraction of dietary nitrogen retained in the body of the animal if the
196
JAMES B. ALLISON
excretion of secalled body nitrogen is constant. The fraction of nitrogen retained in the body has been defined as the “biological value’’ of the dietary protein. Evidence is presented, however, to show that dietary nitrogen can alter the excretion of body nitrogen from the protein stores making the measurement of “biological value” a difficult task. (6) The relative proportion of dietary nitrogen entering into growth and maintenance of animals varies markedly depending upon the amount fed and nutritive quality of proteins. (7) The grams gained in body weight per gram of protein eaten (protein efficiency ratio) under constant experimental condition may be correlated with the gain in nitrogen in the body of the animal. Poor correlations occur, however, when the gain in weight is associated more or less with increases in fat, water, or tissue constituents other than protein. (8) Depletion in proteins results in the reduction of protein stores of the body such as the proteins of the gut, of the liver, plasma albumin, globulin, y-globulin, and proteins in other tissues. Thirty times as much nitrogen may be lost from body tissues as from plasma during depletion in proteins. The stores are not depleted uniformly so that a depleted animal needs a different pattern of essential amino acids in thc diet than the normal animal requires. (9) During repletion in proteins restoration of nitrogen is a function of the patterns of amino acids fed, different patterns producing various rates of repletion of the protein. An imbalance of amino acids can cause one type of tissue to be built up at the expense of others. (10) Lack of any one of the essential amino acids in the diet produces negative nitrogen balance, stops repletion, and causes the animal to lose appetite and become sick. (11) Correlations between the nutritive value of a protein and the amino acid composition may be excellent when the composition represents the pattern of amino acids which are absorbed into the animal. Digestive and other processes which prepare the dietary protein for absorption do not, however, always make available the pattern of amino acids revealed by chemical analysis. REFERENCES Addis, T., Poo, L. J., and Lew, W. (1936a). J . Biol. Chem. 116, 111. Addis, T.,Poo, L. J., and Lew,W. (1936b). J . Biol. Chem. 116, 117. Albanese, A. A. (1947). Advances in Protein Chem. 3, 227. Allison, J. B. (1948). Am. J . Med. 6, 419. Allison, J. B.,and Anderson, J. A. (1945). J . Nutrition 29, 413. Allison, J. B., Anderson, J. A,, and Seeley, R. D. (1946). Ann. N . Y . Acad. Sci. 47,245.
BIOLOGICAL EVALUATION O F PROTEINS
197
Allison, J. B., Anderson, J. A., and Seeley, R. D. (1947). J . Nutrition 33, 361. Allison, J. B., Anderson, J. A., and White, J. I. (1949). Trans. Am. Assoc. Cereal Chem. 7 , 29. Allison, J. B., Seeley, R. D., Brown, J. H., and Anderson, J. A. (1946). J . Nutrition 31, 237. Allison, J. B., Seeley, R. D., Brown, J. H., and Ferguson, F. P. (1946). Proc. SOC. Exptl. Biol. Med. 63, 214. Allison, J. B., Seeley, R. D., and Ferguson, F. P. (1947). J . Biol. Chem. 171, 91. Armstrong, W. D., and Haydee, E. (1947). Federation Proc. 6, 235. Barnes, R. H., Bates, M. J., and Maack, J. E. (1946). J . Nutrition 32, 535. Barnes, R. H., and Bosshardt, D. K. (1946). Ann. N . Y . Acad. Sci. 47, 273. Beadles, J. R., Quisenberry, J. H., Nakamura, F. L., and Mitchell, H. H. (1933). J . Agr. Research 47, 947. Benditt, E. P., Humphreys, E. M., Straube, R. L., Wissler, R. W., and Steffee, C. H. (1947). J. Nutrition 33, 85. Benditt, E. P., Humphreys, E. M., Wissler, R. W., Steffee, C. H. Jr., Frazier, L. E., and Cannon, P. R. (1948). J . Lab. Clin. Med. 33, 257. Benditt, E. P., Woolridge, R. L., Stepto, R. L., Elttxter, K. L., and Mitchell, H. H. (1948). J . Animal Sci. 7, 351. Berg, J. L., Pund, E. R., Sydenstricker, V. P., Hall, W. K., Bowles, L. L., and Hock, C. W. (1947). J. Nutrition 33, 271. Block, R. J., Cannon, P. R., Wissler, R. W., Steffee, C. H., Jr., Straube, R. L., Frazier, L. E., and Woolridge, R. L. (1946). Arch. Biochem. 10, 295. Block, R. J., and Mitchell, H. H. (1946-1947). Nutrition Abstracts & Revs. 16,249. Borman, A., Wood, T. R., Black, €I. C., Anderson, E. C., Oesterling, M. J., Womack, M., and Rose, W. C. (1946). J . Biol. Chem. 166, 585. Bosshardt, D. K., and Barnes, R. H. (1945). J . Nutrition 31, 13. Bosshardt, D. K., Paul, W., O’Doherty, K., and Barnes, R. H. (1947). Federation Proc. 6 , 403. Bosshardt, D. K., Paul, W., O’Doherty, K., and Barnes, R. H. (1946a). J . Nutrilion 32, 641. Bosshardt, D. K., Ydse, L. C., Ayres, M. M., and Barnes, R. H. (1946). J . Nutrition 31, 23. Bricker, M. L., and Mitchell, H. H. (1947). J . Nutrition 34, 491. Bricker, M. I,., Mitchell, H. H., and Kinsman, C. M. (1945). J . Nutrition 30, 269. Brown, J. H., and Allison, J. B. (1947). Am. Chem. Soc., New York meeting, Abstracts p. 51c. Brown, J. H., and Allison, J. B. (1948). Proc. Soc. Exptl. Biol. Med. 69, 196. Brush, M. K., Willman, W., and Swanson, P. P. (1947). J . Nutrition 33, 389. Campbell, R. M., and Kosterlita, 1%.W. (1946). J . Physiol. 106, 33. Cannon, P. R. (1948). Proc. Znst. Med. Chicago 17, 1. Cannon, P. R. (1948). Federation Proc. 7, 391. Chick, H. (1942). Lancet 242, 405. Chick, H., Copping, A. M., and Slack, E. B. (1946). Lancet 260, 196. Chow, B. F. (1946). A m . N . Y . Acad. Sci. 47, 297. Chow, B. F., Allison, J. B., Cole, W. H., and Seeley, R. D. (1945). Proc. SOC. Exptl. Biol. Med. 60, 14. Chow, B. F., Seeley, R. D., Allison, J. B., and Cole, W. H. (1948). Arch. Biochem. 16, 69.
Clandinin, D. R., Cravens, W. W., Elvehjem, C. A., and Halpin, J. G. Poultry Sci. 26, 399.
(1946).
198
JAMES B. ALLISON
Cox, W. M., Jr., and MueUer, A. J. (1946). J . Nutrition 81, 581. Cax, W. M., Jr., Mueller, A. J., Elman, R., Albanese, A. A., Kemmerer, K. S., Barton,
R. W., and Holt, L. E., Jr. (1947). J . Nutrition 38, 437. FJdred, N . R., and Rodney, G. (1946). J . Biol. Chem. 162, 261. Elman, R. (1947). Parenteral Alimentation in Surgery. Hoeber, New York. Evans, R. J. (1946). Arch. Biochem. 11, 15. Evans, R.,I., and McGinnis, J. (1946). J . Nutrition 31, 449. Evans, R. J., McCinnis, J., and St. John, J. L. (1947). J . Nutrition 33, 661. Frazier, I,. E., Wissler, R. W., Stcffee, C. H., Woolridge, R. I,., and Cannon, P. R. (1947). J . Nutrition 33, 65. Frost, D. V., Heinsen, J., and Olsen, R. T. (1946). Arch. Biochem. 10, 215. Frost, D. V., and Risser, W. C. (1946). J . Nutrition 32, 361. Goyco, J. A,, and Asenjo, C. F. (1947). J . Nutrition 33, 593. Gurd, F. N., Vars, H. M., and Ravdin, I. S. (1947). Federation Proc. 6, 257. Hall, W. K., Sydenstricker, V. P., Hock, C. W., and Bowles, L. L. (1946). J . Nutrition 32, 509. Harrison, €1. C., and Iang, C. N. H. (1945). J . Biol. Chem. 161, 545. Harte, R. A., and Travers, J. J. (1947). Science 106, 15. Harte, R. A,, Travers, J. J., and Sarich, P. (1947). J . Nutrition 34, 363. Hawley, E. E., Edwards, 1,. E., Clark, L. C., and Murlin, J. R. (1946). J . Nutrition 32, 613. Hegsted, D. M., and Worcester, J. (1947). J . Nutrition 33, 685. Holman, R. L., hfahoney, E. B., and Whipple, G. H. (1934). J . Ezptl. Med. 18,251. Johnson, R. M., Deuel, H. J., Jr., Morehouse, M. G., and Mehl, J. W. (1947). J . Nutrition 33, 371. Kade, C. F., Jr., Houston, J., Krauel, K., and Sahyun, M. (1946). J . Biol. Chem. 163, 185. Kerr, W. J., Hurwitz, S. H., and Whipple, G. H. (1918). A m . J . Physio?. 47, 356. Kosterlitz, H. W. (1944). Biochem. J . 38, 14. Kosterlitz, H.W. (1944). Nature 164, 207. Kosterlitz, H. W. (1946). J. Physiol. 106, 11. Kosterlitz, H. W., and Campbell, R. M. (1945). J . Physiol. 104, 16. Kosterlitz, H . W., and Campbell, R. M. (1946). Nature 167, 628. Kosterlitz, H. W., and Campbell, R. M. (1945-1946). Nutrition Abstracts & Revs. 16, 1. Li, Tsan-Wen, and Freeman, S. (1946a). A m . J . Physiol. 146, 646. Id, Tsan-Wen, and Freeman, S. (1946b). A m . J . Physiol. 146, 660. Li, Tsan-Wen, and Freeman, S. (1946~).Gastroentenology 6, 140. McNaught, J. B.,Scott, V. C., Woods, F. M., and Whipple, G. H. (1936). J . Expll. Med. 63, 277. Madden, S. C . , Carter, J. R., Kattus, A. A., Jr., Miller, L. I,., and Whipple, G. H. (1943). J . Exptl. Med. 77, 277. Madden, S. C., and Whipple, C.H. (1940). Physiol. Revs. 20, 194. Melnick, D., and Cowgill, G . R. (1937). J . Nutrition 13, 401. Melnick, D., Cowgill, G. R., and Burack, E. (1936). J . ExpU. Med. 64, 877. Mclnick, D., Oscr, B. L., and Weiss, S. (1946). Science 103, 326. Miller, L.I,. (1044). J . Biol. Chem. 162, 603. Miller, L. L.,and Ailing, E. T,. (1947). J . Exptl. Med. 86, 55. Miller, L. L.,Robscheit-Robbins, F. S., and Whipple, G. H. (1947). J . Exptl. Med. 66, 267.
BIOLOGICAL EVALUATION OF PROTEINS
199
Mitchell, H. H. (1924). J . Biol. Chem. 68, 873. Mitchell, H. H. (1942). J . Animal Sci. 2, 263. Mitchell, H. H. (1944). Ind. Eng., Chem. Anal. Ed. 16, 696. Mitchell, H. H. (1947). Arch. Biochem. 12, 293. Mitchell, H. H., and Block, R. J. (1946). J . Biol. Chem. 163, 599. Mitchell, H. H., and Carman, G. G. (1924). J . Biol. Chem. 60, 613. Mitchell, H. H., and Carman, G. G. (1926). J . Biol. Chem. 68, 183. Mitchell, H. €I., Hamilton, T. S., Beadles, J. R. and Simpson, F. (1945). J . Nutrition 29, 13. Murlin, J. R., Edwards, L. E., Fried, S., and Szymanski, T. A. (1946). J . Nutrition 31, 715. Murlin, J. R., Edwards, L. E., Hawley, E. E., and Clark, L. C. (1946a). J . Nutrition 31, 533. Murlin, J. R., Edwards, L. E., Hawley, E. E., and Clark, L. C. (1946b). J . Nutrition 31, 555. Murlin, J. R., Nasset, E. S., and March, M. E. (1938). J . Nutrition 16, 249. Newell, G. W., and Elvehjem, C. A. (1947). J . Nutrition 33, 673. Niven, C. F., Washburn, M. R., and Sperling, G. A. (1946). Proc. SOC.Exptl. Biol. Med. 63, 106. Orten, J. M., Bourque, J. E., and Orten, A. U. (1945). J . Biol. Chern. 160, 435. Osborne, T. B., and Mendel, L. B. (1917). J . Biol. Chem. 32, 369. Osborne, T. B., Mendel, L. B., and Ferry, E. L. (1919). J . Biol. Chem. 32, 369. Pommerenkc, W. T., Slavin, H. B., Karihcr, D. H., and Whipple, G . H. (1935); J . Ezptl. Med. 61, 261. Riesen, W. H., Clandinin, D. R., Elvehjem, C. A., and Cravens, W. W. (1947). J . Biol. Chem. 167, 143. Risscr, 1%’. C. (1946). J . Nutrition 32, 485. Risser, W. C., Schenck, J. R., and Frost, D. V. (1946a). J . Nutrition 32, 499. Risser, W. C., Schenck, J. R., and Frost, D. V. (1946b). Science 103, 362. Rabscheit-Robbins, F. S., Miller, L. L., Alling, E. L., and Whipple, C . H. (1946s). J . Ezptl. Med. 83, 355. Robscheit-Tbbbins, F. S., Miller, L. L., and Whipple, G. H. (1946). J . Ezptl. Med. 83, 463. Robscheit-Robbins, F. S., Miller, L. L., and Whipple, G. H. (1947). J . Ezptl. Med. 86, 243. Rose, W. C. (1938). Physiol. Revs. 18, 109. Rase, W. C. (1947). Proc. A m . Phil. SOC.91, 1 . Rose, W. C., and Rice, E. E. (1939). Science 90, 186. Russell, W. C., Taylor, M. W., Mehrhof, T. G., and Hirsch, R. R. (1946). J . Nutrition 32, 313. Schoenheimer, R., and Rittcnberg, D. (1940). Physiol. Revs. 20, 218. Schwimmer, D., and McGavack, T. H. (1948). N . Y . Slate:J. Med.-48, 1797. Seeleg, R. D. (1945). A m . J . Physiol. 144, 369. Seeley, R. D. (1947). Personal communication. Silber, R. H., Seeler, A. O., and Howe, E. E. (1946). J . Biol.Chern. 164, 639. Smith, H. P., Belt, A . E., and Whipple, G. H. (1920). Am. J . Physiol. 62, 54. Swanson, P. P., Everson, G. ,J., and Stewart, G. F. (1946). ZowaStale College Report on Agricultural Research, p. 238. Sydenstricker, V. P., Schmidt, H. L., Jr., and Hall, W. I<. (1947). Proc. SOC. Exptl. Biol. Med. 64, 59.
200
J A M E S B. ALLISON
Thomas, K. (1909). Arch. Anat. u. Physiol. Anat. Abt., 219. Tomarelli, R. M., and Bernhart, F. W. (1947). J . Nutrition 34, 263. Vara, H. M., and Gurd, F. N. (1947). Federation Proc. 6, 299. Weech, A. A., Goettsch, E. (1938). Bull. Johns Hophins Hosp. 63, 154. Weech, A. A., Goettsch, E., and Reeves, E. B. (1935). J . Exptl. M e d . 61, 299. Whipple, G. H., Miller, L. L., and Robscheit-Robbins, F. (1947). J . Ezptl. Med. 86, 277. Whipple, G. H., Robscheit-Robbins, F. S., and Miller, L. 1J.u (1946). A n n . N . Y . Acad. Sci. 47, 317. Willman, W., Brush, M. K., Clark, H., and Swanson, P. P. (1947). FederationProc. 6, 423,
Wissler, R. W., Woolridge, R. L., Steffec, C. H., and Cannon, P. It. (1946). J. Immunol. 62, 267. Womack, M., and Rose, W. C. (1946). J . Biol. Chem. 162, 735. Woolley, D. W. (1941). J. Exptl. Med. 73, 487. Woolley, D. W. (1946). J . Biol. Chem. 162, 383. Zeldis, L. J., Alling, E. L., McCoord, A. B., and Kulka, J. P. (1945). J . Ezptl. Med. 82, 157. Zucker, I,., Hall, L., Young, M., and Zucker, T . F. (1941). Growth 6, 399. Zucker, L., and Zucker, T. F. (1942). J . Gen. Physiol. 26, 445. Zucker, T. F., and Zucker, L. (1943). Znd. Eng. Chem. 36, 868. Zucker, T. F., and Zucker, L. (1944). Proc. Soc. Ezptl. Biol. M e d . 66, 136.
Milk Proteins BY THOMAS L. McMEEKIN A N D B. DAVID POLIS Eastern Regional Research Laboratory, * Philadelphia, Pennsylvania
CONTENTS I. Introduction ....................................................... 11. Protein Distribution in Milk 1. Casein. . . . .
..................
Page 202
. . . . . . . . . . 204
d . Properties and Composition of a- and 8-Caseins.. . . . . . . . e. Physical Properties of Cascin.. . . . . . . . . . . . . . . . . . . g. Rennet Casein.. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . 209
a. Colostrum Globulin. ....................................... 2. “Albumin” Fraction; @-Lactoglobulin. .................... a. Preparation of 8-La .................... b. Properties of @-Lactoglobulin ...................... V. Amino Acid Composition of R .................... VI. Enzymes in Milk.. . . . . . . . . . . ............................ 1. Carbohydrases.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . a. Amylase.. . . . . . . . . . . . . . b. Lactase.. . . . . . . . . . ...................... 2. Dchydrogenase.. . . . . . . . . . . . . a. Xanthine oxidase. . . . . . . . 3. Esterases.. . . . . . . . . . . . . . . . . a . Lipase.. . . . . . . . . . . . . . . . b. Phosphatase.. . . . . . .................................... 4. Protease ....................
210 214 214 215 219 219
221
......................................... 222 ........................................ 222 b. Lactoperoxidase.. . . . . . . . . . . . . . . . . . . . . . . 222 VII. Relationship of Milk Proteins t . . . . . . . . . . . . . . . . . . . 223
....................
* One of the laborntorics of the Bureau of Agricultural and Industrial Chemistry, Agricultural Research Administration, United States Department of Agriculture. 201
202
THOMAS L. McMEEKIN A N D B . DAVID POLIS
I. INTRODUCTION The importance of milk as a food and its availability have made milk proteins a favorite subject for investigation. This has been particularly true of casein, which for a long time was considered a pure protein (46). Primarily because it is so easily obtained, cow’s milk has been investigated more frequently than any other milk. Unless otherwise stated in this review, the discussion refers to the proteins from cow’s milk, Information on human milk proteins is included when the data are available. Although in recent years there have been notable advances in techniques for separating and investigating proteins, the new results have largely confirmed and extended the thorough investigations of Crowther and Raistriclc (25), of Howe (58), and of Wells and Osborne (142) with regard to the individual proteins of milk and their relationship to the proteins of blood serum.
11. PROTEIN DISTRIBUTION IN MILK The amount and type of protein in milk have been estimated by several supplementary procedures. The method of determining nitrogen distribution on the basis of the amount of protein separated by isoelectric precipitation, salt fractionation, and heat coagulation is relatively simple and has given comparative information concerning the proteins in milk (8,89,109). Results obtained by this method (Table I) show, for example, that casein accounts for about 80% of the protein nitrogen of cow’s milk and only 30% of the protein nitrogen of human milk. The electrophoretic method of determining the protein composition of milk gives more accurate information concerning the number and relat,ive amounts of the components than does the nitrogen distribution method. Since approximately 80% of the protein of milk is casein, and casein is electrophoretically heterogeneous (see casein section), it is advantageous to separate the casein from whey by acidification before determining the protein composition of milk by the electrophoretic method. The protein components of whey have been determined by Smith (118), and by Deutsch (28) by the electrophoretic method. Smith’s data on the number of components of whey and their concentrations and mobilities are given in Table 11. A further method for evaluating the protein composition of milk is determining size distribution by means of the ultracentrifuge. Pedersen (99) has found the casein fraction to be heterogeneous with respect to size. Milk serum or whey contained three main boundaries. The a-component with sedimentation constant Szo= 1.8 he identified as
203
MILK PROTEINS
Cow’s milk (89)
Human milk (8)
Nitrogen
-
Mg./100 ml. % of total N Mg./100 ml. % of total N
...
540
Total NonproteinCaseinb Whey protein. (total) Globulind Album in* Proteosej
30 430 80 19 43 18
-
5.5 79.5
162 36 49
15.0 3.5 8.0 3.0
... ... ...
... 22 30
77
48
... ...
...
Not precipitated by 15% trichloroacetic acid. precipitated by acetic acid, p H 4.7. Nitrogen value of the casein filtrate (whey) minus nonprotein nitrogen. Precipitated from casein filtrate by saturation with magnesium sulfate, p H 7.0. Obtained by subtracting globulin nitrogen plus nonprotein nitrogen from the total nitrogen of the casein filtrate. Nitrogen content of filtrate from heat-coagulated milk minus nonprotein nitrogen. TABLE I1 Composition of Whey from Cow’s Milk as Shown by Electrophoresis-
-
P
Component ( a ) Euglobulinb ( b ) Pseudoglobulin’ (c) Component ( d ) Component ( e ) @Lactoglobulin (f) Component
-
Conc., %
-
6 4 18 12 55 5
Mobility
x
10-6
-1.7 -2.5 -3.6 -4.5 -5.1 -6.4
-
Determined in Verona1 buffer, pH 8.4, ionic strength 0.1, with protein concentration 1.2 %. b Insoluble in 0.3 saturation with ammonium sulfate and insoluble in water a t isoelectric point in absence of salt. *Insoluble in 0.3 saturation with ammonium sulfate but soluble in water a t isoelectric point in absence of salt. 0
Kekwick’s lactalbumin; the 8-component with SZO= 3.0, as Palmer’s O 7.0, as classical globulin. @-lactoglobulin;and the y-component with S ~ =
111. SEPARATION AND PROPERTIES OF MILKPROTEINS Analytical methods are useful in revealing the amount and the approximate type of protein in milk and are valuable in following chemical
204
THOMAS L. McMEEKIN AND B . DAVID POLIS
separations. For complete separation and characterization of individual proteins of milk, however, it is necessary to fractionate by chemical or other means. 1. Casein
The ease of separating casein from the other constituents of milk by acidifying to pH 4.7 has made it one of the most extensively studied proteins. Usually casein is further purified by dissolving i t in milk alkali and reprecipitating at p H 4.7 one or more times before characterization or fractionation. Since casein is a mixture of proteins, is labile to alkali (22), and contains a proteolytic enzyme (141), the product obtained is somewhat determined by the method of purification. However, the electrophoretic patterns obtained by Warner (140) on reprecipitated casein were similar to the patterns reported by Mellander (87) on casein which had not been reprecipitated. a. Comparison o j Human and Cow’s Casein. Mellander (87) has recently published an investigation and review of the chemical differences between casein from human milk and cow’s milk. Human milk yielded only from 0.3 to 0.6% casein, as compared with 3.0 to 3.5% casein from cow’s milk. Human casein was much more difficult to separate than cow’s casein. In fact, the wide differences found in the sulfur content of human casein by different methods indicate that the separation of casein from the other proteins of human milk is uncertain. Mellander found that both human and cow’s casein contained three components, as shown by electrophoresis. Human a-casein had a mobility of only 4.5 X as compared with 7.4 X lo+ for COW’S a-casein in phosphate buffer of ionic strength 0.15 and pH 7.6. The greater mobility of cow’s a-casein may be associated with its higher phosphorus content. Caseins from human milk and cowJs milk differ markedly in the influence of p H on the electrophoretic pattern. Thus in cow’s casein the electrophoretic pattern showed only one component a t a pH acid to the isoelectric point, whereas the human casein electrophoretic pattern showed much less change in the distribution of components with change in pH. b. Heterogeneity o j Casein. The solubility studies of LinderstrplmLang and Kodama (76) demonstrated that casein is a mixture. These results have led to numerous attempts to fractionate casein into its components. Fractions have been reported which differed from the original casein in solubility, phosphorus content, and reactions with rennet. Thus Linderstrplm-Lang (74) reported the separation of casein into fractions by means of alcohol containing a small quantity of hydrochloric acid. In general, his purified fractions were characterized by
MILK PROTEINS
205
differences in phosphorus content. Original casein containing 0.79 % phosphorus was divided into fractions containing 0.10, 0.52, and 0.96% phosphorus, respectively, and accounting for 3, 25, and 68% of the total casein. Mellander (86) found that casein was composed of three electrophoretic components, which he designated a, 8, and y in the order of decreasing mobility. The a- and y-casein fractions were isolated in the electrophoresis apparatus. On the assumption that the caseins had 15.6% nitrogen, the phosphorus contents of a-casein and y-casein were calculated as 0.96 and 0.05%. y-casein had a low mobility at p H 6.98 in phosphate buffer. This fact later suggested to Mellander (87) that the electrophoretic diagram attributed to y-casein was due to boundary anomalies. The electrophoretic experiments reported by Warner (140) showed a slow-moving component in the impure 0-casein fraction. It seems likely that the y-casein of Mellander is similar to the alcoholsoluble, low-phosphorus-contiiining casein isolated by Osborne and Wakeman (97), and by LinderstrZm-Lang (74). The amount of lowphosphorus protein in unfractionated casein appears not to be more than 5 % of the total. Electrophoretic analysis of the low-phosphorus casein obtained by alcohol-water extraction is necessary to prove the identity of these proteins. c. Separation of a- and 0-Caseins. Warner (140) has applied electrophoretic analysis to the chemical fractionation of casein. He was unable to demonstrate a separation of the components by precipitating casein in the presence of salt followed by the use of acetone, as suggested in the work of Cherbuliez and Meyer (19), and Cherbuliez and Schneider (20). The method of Groh et al. (42)of precipitating from urea solution with alcohol produced some fractionation, though some of the fractions showed patterns not found in the electrophoretic pattern of the original casein. Some separation of the components was obtained by LinderstrZm-Lang's (74) method of fractionation with sodium chloride on the acid side of the isoelectric point. This procedure was inadequate for a complete separation with a reasonable number of precipitations. Warner separated a- from 0-casein by taking advantage of the higher solubility of @-casein a t pH 4.2. Preparation of a-Casein. Warner prepared a-casein by dissolving undried, purified, acid-precipi1,ated casein in sufficient sodium hydroxide to give a 1% solution with a p H of 6.5. The solution was made to p H 3.5 by adding dilute hydrochloric acid. After the solution was chilled to 2", it was diluted to a concentration of 0.2 to 0.3% protein, and 0.01 N sodium hydroxide was added until p H 4.2 was reached. At this stage a precipitate formed, and the atldition of sodium hydroxide was continued
206
THOMAS L. McMEEKIN AND B. DAVID POLIS
until a clear supernatant was obtained on centrifugation. Usually this occurred a t a pH 4.4. The precipitate was dissolved, and the entire procedure repeated a t least six times. The isolated protein then showed an electrophoretic pattern of a-casein free of @-casein. Mellander (87) has used this method of separating a-casein from cow's casein, but he was unable to separate the a-component from human casein by this procedure. Preparation of 8-Casein. Warner obtained 8-casein from the filtrates of wcasein fractions by warming to room temperature and adjusting the pH to 4.9. The protein precipitated under these conditions was largely &casein. It was further purified by dissolving in sodium hydroxide to give a 0.2% solution. The Bolution was chilled to 2", and 0.01 N hydroTABLE I11 Comvoeition of Casein Frartions Constituent Phosphorus Total N Amino N Arginine Histidine Lysine Tyrosine Tryptophan 4
b
a-Casein,a % ' 0.9gb (140) 15.5(140) 0.99 (37) 4.3 (37) 2.9 (37) 8.9 (37) 8.1 (37) 1.6(37)
,%Casein; %
Alcohol-soluble casein, % (97)
0.60(140) 15.4(140) 0.73 (37) 3.4 (37) 3.1 (37) 6 . 6 (37) 3.2 (37) 0 . 6 (37)
0.06 15.6 ...
2.9 2.3 4.0 2.5 ...
Reference numbers are given in parentheses. 0.97% phosphorus for a-cascin separated by electrophoresis (86).
chloric acid was added until pH 4.9 was reached. The precipitate was filtered off and dissolved with sodium hydroxide to give a 0.03% protein solution. This solution was made to pH 4.5 a t 2". The precipitate formed (chiefly a-casein) was removed, and the filtrate was warmed to room temperature. The precipitate then formed was largely /.%casein. A final purification was accomplished in the same manner by precipitating a 0.5% solution of the 8-casein at 2" and pH 4.9. d. Properties and Composition of a- and @Caseins. These two fractions separated by Warner were not homogeneous over the entire p H range but were distinct fractions, neither one of which was contaminated with the other. The mobility of the purified a-casein in Verona1 buffer a t an ionic strength of 0.1 and a p H of 7.78 was 6.98 (cm.l volt-' set.-* X 10-6)). However, in the presence of 20% 8-casein, the mobility of the a-casein was reduced to 6.30 X indicating interaction of these two fractions. The mobility of purified 8-casein was 3.27 X 10-6, and its
MILK PROTEINS
207
mobility was only slightly changed by the presence of a-casein. Table I11 shows the composition of purified casein fractions. e. Physical Properties of Casein. Since electrophoretic data indicate that casein is heterogeneous, it is not surprising that the reported molecular weights of casein should be unsatisfactory. It has been reported (133) that small amounts of calcium produce an extensive change in the sedimentation constant of casein. This may help explain the wide differences in the values reported for the molecular weight of this substance. Svedberg, Carpenter, and Carpenter (131) reported that casein prepared by the Van Slyke and Baker method is polydisperse, the main component having a molecular weight of 75,000 to 100,000 and a sedimentation constant of Szo = 5.6 x Similar results were obtained with casein prepared by the Hammarsten method. The sedimentation constant of casein prepared in different ways was reported by Pedersen (99) in phosphate buffers. In the experiments of Svedberg, to be 12 X Carpenter, and Carpenter, the salt concentration was only 0.017 M , whereas in Pedersen’s experiments, the salt concentration was 0.25 M . This may explain the wide variations for the sedimentation constant of casein reported in these investigations. Burk and Greenberg (16) have reported a molecular weight of 33,000 for casein in 6.66 M solutions of urea. The combination of calcium ions with casein has been investigated by means of the ultracentrifuge (18). The law of mass action accurately described the data, yielding the dissociation constant of calcium caseinate. The dissociation constant is independent of the concentration of protein and the pH from 6.3 to 8.5. Equations have been reported by Klotz (67) for describing the combination of casein with calcium ions in terms of sixteen dissociation constants rather than one. Studies on flow birefringence and viscometry of sodium caseinate solutions by Nitschmann and Guggisberg (93) led to the conclusion that casein particles are rod-shaped, with an axial ratio of a / b = 8.7 and a particle length of 290 d units, depending somewhat on the method of calculating. Data were also reported for streaming birefringence of sodium caseinate in 1.6 M sodium sulfate. Apparently a n aggregation of casein molecules takes place in this salt, solution, since Edsall (30) has calculated a particle length of 2200 d units from these data. I n the separation of the components of casein, it is important to know whether decomposition has occ.urred during fractionation. It was noted by Cohn and Berggren (22) th at casein solutions hydrolyze even in weakly alkaline solutions, with an increase in base-combining capacity and in water-soluble protein. It has often been observed that the viscosity of casein solutions decreases with time. Th at the change in viscosity is
208
THOMAS L. McMEEKIN AND
€3.
DAVID POLIS
enzymatic and is associated with an increase in nonprotein nitrogen was clearly demonstrated by Warner and Polis (141). No method for removing the proteolytic component from the bulk of the casein was found, though it was destroyed by heating casein solutions t o 80°C. for 10 minutes a t pH 8.6. To prove that the isolated components of casein were not the result of decomposition, Linderstrgm-Lang (74) and Warner (140) have compared the properties of the reconstituted casein with the unfractionated material. The reproduction of the electrophoretic pattern of the original casein by reconstitution with the separated components is strong evidence that the fractions were not changed during separation. f . Phosphopeplones from Casein. Posternak (103) isolated phosphopeptone from tryptic digests of cow’s casein. His first compound contained 5.9% phosphorus and 11.9% nitrogen and showed the presence of glutamic and aspartic acids, serine, and isoleucine. Since serine was the only hydroxy acid present, he considered it likely that phosphoric acid was bound to the hydroxy group of serine. Lipmann (77) isolated phosphoserine from casein, and Levene and Hill (71) isolated phosphoserylglutamic acid from phosphopeptone prepared according to Posternak. Independently, Rimington and Kay (107) described the preparation of a phosphopeptone by tryptic digestion of casein solutions. Their preparation contained 7.05 % phosphorus and 10.13 % nitrogen. Damodaran and Ramachandran (26) digested casein with pepsin followed by trypsin and isolated the barium salt of a phosphopeptone which contained 4.34% phosphorus and 6.46 % nitrogen. This phosphopeptone was reported to be composed of three glutamic, three isoleucine, and four serine residues. Mellander (87), using the method of Dumodaran and Ramachandran, has prepared phosphopeptone from human casein and cowls casein. He found approximately 50% of the total phosphorus of casein in the phosphopeptone. The presence of alkaline phosphatase in commercial trypsin reduced the yield of phosphopeptone. The amount of phosphatase in the trypsin was reduced by heating its solutions to 50°C. for 30 minutes a t p H 2.0. As prepared by Mellander, the barium salt of the phosphopeptone from human casein contained approximately 6.3 % nitrogen and 4.9 % phosphorus, while that from cow’s casein contained about 7.0% nitrogen 5.3% phosphorus. The large variations in composition among individual preparations indicated that the product probably was a mixture or that each preparation differed in structure. After a second treatment with trypsin and fractional precipitation, Nicolet and Shinn (92) isolated a phosphopeptone from casein containing eight amino acids. The peptide contained one molecule of serine, two of phosphoserine, two of isoleucine, and two of glutamic acid. The dipeptide phosphoserylglutamic acid was isolated and its structure determined.
MILK PROTEINS
209
g . Rennet Casein. Although the clotting of milk or casein by rennet has been much studied, the nature of the change produced in casein by this enzyme is unknown. It has not been possible to distinguish calciumfree preparations of rennet casein from acid casein by conventional analytical procedures. Since rennet casein cannot be clotted a second time, it would appear that there is a structural difference between acid and rennet casein. Many investigators have been concerned with the removal of the proteolytic action from rennet preparations. Berridge (9) has crystallized rennet and found that the pure enzyme has proteolytic activity but that the optimum pH for rennet proteolysis is more alkaline than the optimum for pepsin. Holter (53) considered the proteolytic property to be unimportant in coagulation of casein by rennet. Nitschmann and Lehmann (94) have recently reported that the action of rennet on casein solutions produces a change in the electrophoretic pattern of the casein. They incubated a 6% solution of sodium caseinate with rennet a t 35°C. for 30 minutes at pH 6.6. After equilibrating with Verona1 buffer at pH 7.4, which presumably destroyed rennet activity, an electrophoretic analysis of the solution was made. It was found that the electrophoretic pattern of rennet-treated casein differed from that given by the same casein solution treated with boiled rennet, in that the a-casein was split into two peaks. The mobilities of the a;-and az-comand 8.83 X lod6,respectively, as ponents were reported to be 9.55 X compared with acid a-casein of 9.30 X The descending boundary of rennet-treated casein was not split into two peaks. Warner (140) found that the electrophoretic pattern of a-casein splits into two peaks in some preparations. In view of this, Nitschmann and Lehmann consider the action of rennet on a-casein as an acceleration of the tendency to split into two components. The possibility that other components of casein stabilize a-casein was also considered. These authors (95) have also compared the behavior of mixtures of rennet and acid casein with calcium chloride solutions. One of the two caseins was dyed by coupling with a diazonium salt, and the amount precipitated by calcium chloride from mixtures was determined colorimetrically. I n this manner the relative amounts of acid and rennet caseins present were estimated. It was shown that from mixed solutions of rennet and acid caseins about equal amounts of rennet and acid caseins were precipitated by adding calcium chloride. This result is in agreement with the theory of LinderstrZm-lang (74) that the stability of the calcium caseinate system is due to the protective colloid action of one of the components. Nitschmann and Lehmann attempted to estimate the percentage of the protective colloid fraction in acid casein and concluded that it was less than 20% of the total casein.
210
THOMAS L. McMEEKIN AND B. DAVID POLIS
IV. PROTEINS OF WHEY After casein is removed by isoelectric precipitation at p H 4.6, the whey contains between 0.6 and 0.7% protein or about 20% of the total protein of skim milk. The number of components and relative amounts of each, as shown by electrophoresis, are given in Table 11. I t has been recognized for a long time that whey contains a globulin f,raction and an albumin fraction (113), as characterized by salt fractionation methods. 1. Globulin Fraction
Since the work of Sebelien (1 13) was reported, the classical globulin of milk has been prepared by saturating whey with magnesium sulfate. Crowther and ltaistrick (25) used anhydrous magnesium sulfate for precipitating milk globulin. They observed that, like serum globulin, milk globulin could be separated into euglobulin and pseudoglobulin by dialysis. Smith (118) has found that globulin prepared from whey by repeated precipitation with saturated magnesium sulfate gave preparations which show complex electrophoretic patterns. He devised a method of preparing an electrophoretically homogeneous globulin with a mobility of 1.7 to 2.5 u by fractionation with ammonium sulfate. The crude globulin fraction was precipitated by 0.5 saturation with ammonium sulfate. The precipitate was dissolved to about 3% protein, the pII adjusted to 4.6, and ammonium sulfate added to 0.25 saturation. The precipitate was removed and discarded. The immune proteins were precipitated from the supernatant a t 0.4 saturation with ammonium sulfate a t pH G.O. The precipitate was reworked by dissolving it in water a t lo, adjusting to p H 4.5 and removing a n insoluble residue th a t formed. The supernatant was precipitated a t 0.3 saturation with ammonium sulfate. The filtrate was made to p H 6.0 and ammonium sulfate added to 0.4 saturation. Both fractions appeared to be electrophoretically homogencous. On dialysis, however, the euglobulin and pseudoglobulin separated from the final precipitate a t 0.4 saturation appeared to be more homogeneous by electrophoresis than the corresponding fractions obtained a t 0.3 saturation. As shown in Table 11, the electrophoretic component of whey with a mobility of u = -3.6, which amounted.to 18% of the protein of whey, appears not to have been separated in the pure form. a. Colostrum Globulin. The marked difference in the physical and biological properties of colostral milk or colostrum obtained for the first few days after parturition can be ascribed primarily to its high protein content and atypical protein distribution. The total protein concen-
MILK PROTEINS
21 1
tration of cow’s colostrum at birth may be about 17.5% (32). Of this, 5 % is casein and 11.3% albumin and globulin. This high protein level falls rapidly in the first 24 hours and then gradually approaches the normal composition of milk, that is 3.7% total protein, of which 3.0% is casein and 0.7% is albumin and globulin. The most striking difference between colostrum and normal milk is the high globulin content of the colostrum. By salt fractionation studies Howe (54-58) demonstrated the absence of euglobulin and pseudoglobulin I in the blood of a newborn calf. When colostrum was fed before the calf was 21 hours old, these globulins always appeared in the blood. When the calf was fed only milk, euglobulin and pseudoglobulin were absent from the blood for some time. Howe demonstrated the presence of euglobulin, pseudoglobulin I, and pseudoglobulin I1 in colostrum. He then showed that the high concentration of euglobulin and pseudoglobulin appearing in the blood of a calf after colostrum ingestion was transient. After the first day the concentration of these two proteins decreased, and the high levels characteristic of the adult animal did not appear until the age of 12 or 14 months. Associated with the feeding of colostrum (125), there was invariably an albuminuria. Howe (56) identified the excreted proteins as euglobulin and pseudoglobulin. Similar studies on the human subject were made by Lewis and Wells (72) and by Boyd (11). They found that the blood of a n infant had practically no euglobulin, although it contained the same concentrations of pseudoglobulin I and I1 as the adult. This fraction gradually appeared in infants even when they were given no colostrum. It appeared much more rapidly, however, when colostrum was fed. It is apparent that the colostrum contains proteins with antibody activity th a t may be absorbed unaltered from the digestive tract of the young animal. The importance of colostrum from the calf was indicated more directly by Smith and Little (124), who found that more calves fed colostrum survived than those deprived of colostrum and fed milk. They concluded that colostrum protects the young animal against organisms which are harmless to it later when its own anti-infective facilities are in operation. About the same time Orcutt and Howe (96) showed the passage of natural agglutinins of Brucella abortus from a cow to her calf by the globulins of the colostrum. Timmerman (137) reported the presence of agglutinins for typhoid bacillus in human colostrum during the first five days of suckling. Kuttner and Ratner (70), however, indicated th a t colostral antibodies do not have the same importance for humans as for animals. They found the blood of infants born from mothers who were not immune to diphtheria to contain no antitoxin. The blood of infants born t o immune mothers had as much antitoxin as the mother’s blood
212
THOMAS L. McMEEKIN AND B. DAVID POLIS
even before they received colostrum. They infer, therefore, that the important mode of transfer of antibody from the human mother to the infant is by way of the placenta. Jameson et al. (60)found that feeding cow’s colostrum to man and to the adult rat produced an additional globulin component in the serum. The a-fraction was greatly increased, and the y-fraction was either split or a new fraction appeared. Highly immune rat serum showed similar changes. The feeding of other proteins like casein, serum albumin, serum globulin, and liver and kidney protein did not have this effect. Early work on the chemical composition of colostrum showed the presence of the three main protein components of milk. Crowther and Raistrick (25) did a comparative study of the proteins of the colostrum and milk of the cow and their relations t o serum proteins. With the methods available to them a t the time (1916) they arrived a t the conclusions: (a) that casein, total lactoglobulin, and lactalbumin are distinct proteins and have the same composition whether prepared from colostrum or normal milk; (b) the globulin fractions obtained from colostrum and milk, although occurring in small amounts in milk, are alike and closely allied to or identical with the serum globulin from ox blood; ( c ) the eulactoglobulin and pseudolactoglobulin are identical insofar as composition is concerned; and (d) the lactalbumin from colostrum or milk is different from blood serum albumin. Woodman (145) by the method of protein racemization, with which he claimed the ability to establish the identity or nonidentity of related proteins, arrived at similar conclusions. An outstanding and definitive study of the proteins of colostrum was made recently by Smith el al. (117,119-123). Whole colostrum and the various fractions isolated by conventional precipitation procedures with ammonium sulfate were studied electrophoretically. Table IV, taken from Smith’s data, illustrates in concise manner the colostrum protein picture. Fraction A, the casein, was obtained by isoclectric precipitation a t p H 4.5. It is obvious that this complex protein fraction does not have the distribution found in normal milk. This has been proved in greater detail by other studies (80) on the variation of the casein composition with the stage of lactation. Fraction B was obtained between 0 and 0.3 saturation, with ammonium sulfate, and fraction C was obtained between 0.3 and 0.5 saturation. These two fractions contain most of the colostral protein. By precipitation a t 0.4 saturation with ammonium sulfate at pH 6.0,followed by solution and reprecipitation, Smith was able to isolate an electrophoretically pure globulin from colostrum in high yield. Apparently this procedure is not applicable to the isolation of the globulin from normal whey. Probably because of the preponderance of this protein in colostrum, Smith was able to obtain it, quant,itatively free of other pro-
MILK PROTELNB
213
teins, when determined electrophoretically. This protein fraction completely accounts for all the immune properties of the colostrum. The immune lactoglobulin isolated in electrophoretically pure form was subjected t o more extensive investigation of its properties. By exhaustive dialysis, the immune lactoglobulin could be separated into water-insoluble euglobulin and water-soluble pseudoglobulin fractions. The euglobulin migrated in Verona1 buffer, pH 8.5, with a velocity of - 1.9 u, as compared with -2.2 u for the pseudoglobulin. Both fractions TABLE IV Electrophoretic Composition of Colostrum and Fractions from Colostrum"
-
Frartion
Whole colostrum (A) Casein (B) Immune lactoglobulin (C) Immune lactoglobulin (U) ,%Lactoglobulin
=_
Dry wt. isolated, g. 355 60
u
:
___
... -1.3
101
...
113
...
17
.
I
.
;
From the study of Smith (117). Electrophoresis patterns were obtained with Tiselius apparatus a t 1°C. in Verona1 buffer, pH 8.3 to 8.4,ionic strength 0.1. b Immune lactoglobulin. 0
had immune activity. The total colostral globulin had a n isoelectric point a t pH 5.85. The euglobulin from another animal was isoelectric a t pH 6.2, the pseudoglobulin a t pH 6.0. With respect to the variations of mobility with pH, the colostrum globulin resembled the plasma T-globulin more than the y-globulin. The immune globulin from colostrum always contained carbohydrate. Smith el al. (123) reported 2.65% of the carbohydrate as hexose and 1.48% as hexosamine. The molecular weight of immune lactoglobulin was reported a t about 160,000 to 190,000. The diffusion constant Dzo was given as 3.G X lo-' (117). Amino acid values for colostrum globulin as reported by Smith and Greene (117,121) and by Hansen et al. (48) are given in Table V (see p. 218). The colostrum immune globulin and plasma y-globulin were quantitatively equivalent in producing. anaphylaxis in guinea pigs. The immune activity of bovine plasma is present in both T- and y-components. On the basis of comparison of the elementary composition, isoelectric points, diffusion constants, and amino acid analyses of colostrum globulin
214
THOMAS L. McMEEKIN AND B. DAVID POLIS
with those of T- and y-globulin, Smith (121) concludes that, although these three proteins, which are associated with immune activity in the cow, are closely related, they are not identical.
2. “Albumin” Fraction; P-Lactoglobulin The portion of whey soluble in saturated magnesium sulfate or 0.5 saturated ammonium sulfate is commonly designated the albumin fraction. This fraction contains a variety of proteins, some of which have enzymatic properties. Wichmann in 1899 (143) crystallized lactitlbumin from this fraction with ammonium sulfate. Later Sjdgren and Svedberg (116) also obtained lactalbumin in crystalline form from whey with ammonium sulfate and dilute sulfuric acid. Palmer (!B) was unable to obtain crystalline lactalbumin by this method. Sprrensen and Sprrcnsen (127) also had difficulty in preparing lactalbumin. Palmer (98), however, prepared a crystalline protein from the albumin fraction of whey in good yield by prolonged dialysis of the proteins of this fraction a t p H 5.2 in the abscnce of salt. This crystalline protein has been named 8-luctoglohulin (17), since it was identified with the 8-component of milk serum, as shown by ultracentrifugal studies (99). Svedberg and Pedersen (132) arc of the opinion that the previously described crystalline lactalbumin represents an impure 0-lactoglobulin. The view that cryst,alline 0-lactoglobulin is essentially the same as crystalline lactalbumin appears probable, since Sprrensen and Sprrensen (127) had no difiiculty in recrystallizing 0-lactoglobulin with ammonium sulfate a t pII (i-7. They tilso described (127) a crude crystalline lactalbumin preparation which was not sufficiently characterized to determine whether it differed from crystalline 0-lactoglobulin. In addition, they described the separation of lactomucin, a green mucoid fraction, a red fraction, and a crystalline protcin insoluble in water and dilute salt solutions. The latter protein was crystallized from aqueous solutions a t p H 6.5 to 7.0 by the addition of ammonium sulfate. Pedersen (99) has attributed a component of wliey with a sedimentation constant S20 = 1.9 x to a lactalbumin isolated by Kekwick (unpublished). a. Preparation of 0-Lactoglobulin. In Palmer’s (98) first method for preparing P-lactoglobulin, casein was precipitated by adding hydrochloric acid to pH 4.6. The pH of tfhe whey was then made 6.0, and the solution half saturated with ammonium sulfate to remove the globulin. The filtrate then was saturated with ammonium sulfate, and the precipitated albumin redissolvcd in water. On long dialysis a t p H 5.2, a n oil accumulated, which gradually changed into large crystals. Palmer also described a modification of this method which consists in concentration of whey by freezing, followed by fractionation with sodium sulfate. He obtained a
MILK PROTEINS
215
yield of 1.8 g. of crystalline 8-lactoglobulin per liter of whey or 60% of the total protein in the albumin fraction. The first method of Palmer, which involves the use of ammonium sulfate, is convenient when large amounts of 8-lactoglobulin are to be prepared. SIdrensen and SZrensen’s (127) method consists in precipitation of casein and the globulin fraction with ammonium sulfate. p-Lactoglobulin is obtained from the portion of whey soluble in 2.3 M but insoluble in 3.3 M ammonium sulfate. In recrystallizing 8-lactoglobulin, Palmer used sodium chloride or dilute ammonium hydroxide for dissolving the p-lactoglobulin. It has been reported (15) that dissolving p-lactoglobulin in dilute sodium hydroxide produces a change in the molecular weight, as determined by osmotic pressure. Separating 8-lactoglobulin by means of alcohol at low temperatures has been suggested (4). b. Properties of /3-LactogZobulin. Palmer (98) reported the solubility of p-lactoglobulin in water, dilute sodium chloride, and concentrated sodium sulfate solutions. He observed th at a fourfold increase in the amount of solid phase did not change the solubility, indicating the relative purity of this protein. The effect of varying the amount of solid phase on its solubility in dilute ammonium chloride solutions was extensively investigated by Sgrensen and Palmer (12G). They found that the change in solubility with a threefold variation in the amount of solid phase was not more than 1 %. These results on its solubility as well as the elegance of the crystal form and other properties of P-lactoglobulin have made it a favorite protein for investigation. Pedersen (100) calculated the molecular weight to be 39,000, and found i t to be stable from p H 1 to 9. From the manner in which the sedimentation and diffusion constants varied with the pH of the solution, he assumed that the molecular weight remained constant and that the shape of the molecule changed. He found the isoelectric point to be p N 5.19 in acetate buffers by the electrophoretic method. Crowfoot (24) determined its crystal structure and unit cell by means of X-ray measurements. Assuming 8 molecules per unit cell, she calculated a molecular weight of 40,000 for air-dried P-lactoglobulin (33). A value of 35,800 for the anhydrous molecular weight of p-lactoglobulin was deduced from Crowfoot’s data on the assumption that the air-dried crystals contained 9.78% moisture (83). A recent X-ray investigation of this protein has given a value of 35,000 for the anhydrous molecular weight (114). A value of 35,050 for the molecular weight of p-lactoglobulin has been found by osmotic pressure measurements in 0.5 M sodium chloride (15). MchiIeekin and Warner (83) determined directly the amount of water in a single p-lactoglobulin crystal and found i t to be 46% of the total weight. It was also shown that ammonium sulfate penetrated the crystal and, on the assumption that the
216
THOMAS L. McMEEKIN AND B. DAVID POLIS
salt dissolved in the water of crystallization, the concentration of salt in the water of crystallization reached 82% of the salt concentration in the surrounding liquid. The dissociation curve of /3-lactoglobulin has been determined under a variety of conditions (17). The effects of temperature, concentration of potassium chloride, concentration of protein, and the addition of formaldehyde on the dissociation curve were evaluated. The results were consistent with the presence of 58 carboxyl, 34 amino, 6 imidazole, and 6 guanidino groups in 1 mole (40,000 grams) of P-lactoglobulin. The ratio of the net charge to the electrophoretic mobility was constant from pH 4 to 9. A complete amino acid analysis of ,&lactoglobulin has been reported by Brand et ul. (13). Their results are tabulated in Table V with minor modifications. The amino acid composition of fl-lactoglobulin differs markedly from the amino acid composition of serum albumin. This is consistent with the differences noted by earlier workers between the composition of lactalbumin and serum albumin. Gronwall (41) has reported extensive solubility studies on 0-lactoglobulin under a variety of conditions. Certain aspects of this work have been reviewed by Edsall (31). In agreement with Palmer’s results, it was found that the solubility did not increase more than 1% when the protein nitrogen in the solid phase was increased several times. As a test for purity, these results as well a s those of Palmer are deficient in two respects, namely, ( u ) the amount of solid phase is much too great to demonstrate purity, being from five to ten times greater than the solubility, and ( b ) there is no indication that the values for solubilit,y are equilibrium values, since the solubility was determined after a pcriod of 24 hours only. A value of 0.16 mg. of nitrogen per milliliter was found for the solubility of p-lactoglobulin in water, which is somewhat lower than the 0.19 reported by Palmer. The value for the solubility in water obtained by extrapolating the salt solubility curve to zero salt concentration was 0.14 mg. of nitrogen per milliliter. Gronwall found that the solubility in sodium chloride solutions varied from one preparation to another, even though the preparations were made by identical methods. His results for the solubility in dilute sodium chloride solutions were twice as greut as those found by Palmer. These solubility results, as well as the electrophoretic results of Li, indicate that 0-lactoglobulin is not a homogeneous protein. Li (73) found that 0-lactoglobulin contained three components as determined by electrophoresis a t p1-I 4.8 and 6.5. Pedersen (100) had previously reported P-lactoglobulin to be electrophoretically homogeneous when a 0.2% protein solution was used. Bosshardt, Moore, and Brand (10) found freshly crystallized P-lactoglobulin to be electrophoretically homogeneous at pH 4.0, 7.4, and 8.6. By implication, this report suggests that the heterogeneity of p-lactoglobulin is produced by recrystal-
MILK PROTEINS
217
lizatiofi. Variations in the solubility of 8-lactoglobulin in water and sodium chloride solutions have been correlated (82) with variations in its electrophoretic pattern (79). p-lactoglobulin preparations with different electrophoretic composition were obtained by fractionation with acetate buffers and also with alcohol. These preparations gave different solubilities in water and sodium chloride solutions. Further p-lactoglobulin fractions of different properties were separated from whey proteins after the partial removal of 8-lactoglobulin by the method of Palmer. These preparations varied in solubility and in electrophoretic composition at p H 4.8, but were homogeneous a t p H 8.4. Since the dielectric increment is large--1.5 units per gram per liter (34)-it is to be expected that the solubility of p-lactoglobulin will be markedly influenced by ions and dipolar ions, making complete separation from other proteins difficult. Gronwall has demonstrated that the solubility of p-lactoglobulin is increased in solutions of glycine and glycine peptides proportionally to the dielectric constant of the solvent. The effect of ovalbumin in increasing the solubility of p-lactoglobulin was of the order t o be expected from the influence of ovalbumin on the dielectric constant of the solvent. Davis and Dubos (27) have reported th at 8-lactoglobulin binds fatty acids in a manner similar to the binding by serum albumin. A crystalline compound of P-lactoglobulin with two equivalents of dodecyl sulfate has been described (81). This derivative is apparently undissociated, since the dodecyl sulfate was not removed by barium ion. The mobility of the derivative was about 8 % greater than that of 8-lactoglobulin at pH 8.4, and the apparent isoelectric point was slightly more acid than that of 8-lactoglobulin. The solubility of the derivative is about one-half as large in water and one-third as large in dilute salt solution as th a t of 8-lactoglobulin. The derivative has a greater alkali-combining capacity than 8-lactoglobulin, but has essentially the same acid-combining capacity. Solutions of the derivative require a higher temperature for heat coagulation than do 0-lactoglobulin solutions. This effect is similar to the influence of dodecyl sulfate on the heat coagulation of albumin solutions (6). Several studies have been made on the denaturation of p-lactoglobulin (14,59,75). Briggs and Hull (14) demonstrated th a t the heat denaturation of 8-lactoglobulin a t p H 7.0 in a buffer solution of 0.1 ionic strength involves a t least two reactions which could be followed by electrophoretic mobilities. The first process, initiated only above 65", was accompanied by a fourfold increase in particle weight and a n increase in frictional ratio. The second process, which takes place only after the first process has occurred, was accompanied by a marked increase in electrophoretic
218
THOMAS L. McMEEKIN A N D B. DAVID POLIS
TABLE V Amino Acid Composition of Milk Proteins0J
-
Constituent
Human casein
cow casein
Cow p-lactoglobulin
cow colostrum globulin
Immune globulin of cow milk
5.63 (117 1.1 (48)
5.67' (118) 1. o o c (121)
(13) (66) (13) (13) (122) (13) (13)
0 . 2 (117) 8 . 9 (117) 2 . 7 7 (121 0.0 (48) 3 . 6 (117)
9 . 6 (123) 9 . 6 (123) 3 . 0 3 (121)
1.11 (13) 2 . 2 9 (13) 3.22 (13)
3 . 2 6 (121 0 . 8 9 (121
3 . 1 5 ~(121) 0.9OC (121)
1.94 (13) 2.88 (13) 1 . 5 8 (13) 11.4 (13)
2.74 (117 4.36 (121 1.99 (121 6 . 3 (121)
2.70 (123) 4.05" (121) 2.05" (121) 6.8' (121)
1 1 . 4 (13) 19.5 (13)
9 . 4 (48) 2 . 3 (48)
Total h' Total S Amino N Amide N
5 1 (87) 0 78(102
15.65 (84) 0 . 7 8 (61) 0.93 (88) 1 . 4 2 (104
15.60 (13) 1.60 (13) 1.24 (13) 1.07 (13)
Glycine Alanine Valirie Lcucirie Isoleucinc Pro li ne Pheny lalnnine
0 0 0 2 6 3 8 9 5 8
1.9 3.5 7.2 10.3 7.6 11.6 5.5
1.4 7.4 5.8 15.6 6.1 4.1 3.5
Cystcine Half-Cystine Methionine
0 6 (144) 2 70(102
0 . 3 4 (61) 3 . 1 (61)
Tryptophan Arginine Histidine Lysine
1 05 (102) 3 58 (102) 2 0 (144) 5 6 (144)
1.2 4.0 3.2 8.2
Aspartic Acid Clutamic Acid
4 6 (144) 0 9 (144)
Serine Threonine Tyrosine
5 4 (144) 4 5 (144) 6 11 (8)
Totals
0 2 5 2
(144) (144) (144) (144) (144) (144) (144)
7 64
(115) (138) (51) (51) (51) (51) (51)
(129) (84) (84) (35)
7 . 2 (45) 22.0 (3) 5 . 9 (104) 4 . 5 (104) 6 . 1 (78)
IS 34
5 . 0 (13) 5.85 (13) 3.78 (13)
13.86
3 . 8 8 (123)
9 . 4 (121) 6 . 7 (48)
0.5" (121)
6.81
6.66 -.
-
Grams per 100 grams protein. Reference numbers are given in parentheses. Avcrnge of milk ruglohulin and milk pseudoglobulin.
mobility and particle weight and followed the concentration time characteristics of a sccond-order reaction. The second reaction was repressed at 75°C. ancl did not take place a t 99". The denaturation of P-lactoglobulin in urea solutions has been investigated by Jacobsen and Christensen (59). Denatured p-lactoglobulin was precipitated by the addition of 10 volumes of a mixture consisting of 0.8 M acetic acid, 0.4 M sodium acetate, and 0.5 M mag-
MILK PROTEINS
219
nesium sulfate. It was found that denaturation was fast a t 0" and much slower a t 37.4". The protein denatured a t 0" was rapidly reversed when the temperature was increased to 34.4'. The presumably denatured and reversed protein was recrystallized by adding saturated ammonium sulfate solution to two-thirds saturation. V. AMINOACID COMPOSITION OF MILK PROTEINS
The methods available for separating the electrophoretic components of casein are still inadequate. As a consequence, unfractionated acidprecipitated casein is a much studied and utilized protein. The results of recent amino acid analyses on both cow's and human unfractionated caseins are included in Table V with amino acid analyses of other milk proteins. The amino acid composition of human and cow's caseins appears to be much the same. This is particularly true when the results of a single investigation are compared (144). The phosphorus content of human casein is much lower than that of cow's casein. The sulfur content of human casein appears to vary widely, depending on the method of separation (87), but 0.78% has been selected for Table V. VI. ENZYMES IN MILK The presence of an enzyme in milk was first demonstrated by Arnold (2) in 1881. Since then, although a considerable number of oxidative and hydrolytic enzymes of milk have been reported, only two, xanthine oxidase (5) and lactoperoxidase (135)' have been isolated and purified extensively. Any enzyme obtained from milk is not necessarily a true milk enzyme but may be derived from bacteria or leucocytes in the milk. The origin of milk enzymes is at best poorly defined. That they may be infiltrated plasma enzymes or secreted mammary gland enzymes are subjects more for a speculative than a factual report. The collection and isolation of the enzyme in the presence of bacteriostatic agents such as chloroform or formaldehyde and the isolation of the milk enzyme from the mammary gland itself have been cited as criteria for true milk enzymes. 1. Carbohydrases a. Amylase. Of all the enzymes in normal milk, amylase is probably the least variable in quantity. Koning (68) reported the decomposition of 22.5 mg. of soluble starch in half an hour by 100 ml. of milk from healthy cows. Colostrum and milk from diseased udders show greater amylase activity than normal milk (21). Human milk has a higher amylase content than that of the cow or of other species (52,91,112,146). Cream has a higher amylase activity than skim milk. Most of the amylase is precipitated with the casein fraction of the milk proteins,
220
THOMAS L. McMEEKIN A N D B . DAVID POLIS
although some remains in the whey. Milk amylase shows maximum activity a t p H 5.8 to 6.2 at 30°C. It is inactivated by heating for 1 hour at 60 to 65°C. (38,50). Because of the low amylase content of old milk, G i f i o r n (36) proposed the use of amylase activity as a criterion of the quality of milk. b. Lactuse, This enzyme acts on lactose in milk splitting the molecule into glucose and galactose. Stoklasa (128) and Vandevelde (139) have reported that it is present in normal milk. This has been denied by Svanberg (130).
2. Dehydrogenase a. Xunlhine Oxidase. (This enzyme is also known as aldehydrase and Schardinger enzyme.) In 1902, Schardinger (111) reported that the addition of a n aldehyde and methylene blue to fresh milk resulted in the disappearance of the blue color in the absence of oxygen. Later Morgan, Stewart, and Hopkins (90) and Dixon and Thurlow (29) demonstrated that milk contains an enzyme which oxidizes xanthine to uric acid. As shown by Ball (5), the enzyme is adsorbed on the fat globules in milk. It can be readily prepared by extracting the cream with disodium acid phosphate a t 38°C. Purification is accomplished by digestion with commercial lipase, clarification with calcium chloride, and precipitation by saturation with ammonium sulfate to 60% a t 0°C. The enzyme can be further purified by precipitation with ammonium sulfate to 33% saturation. Corran et at. (23) have isolated a flavoprotein from milk which exhibits both xanthine aldehyde oxidase and dihydrocoenzyme I oxidase (diaphorase) activity. Their method of preparation involved fractionation of milk with ammonium sulfate, precipitation of the active fraction with alcohol, adsorption and elution of the enzyme from alumina gel and further concentration and fractionation with ammonium sulfate. These milk flavoproteins catalyze the oxidation of xanthine, aldehyde, and dihydrocoenzyd I. Xanthine oxidase is not specific toward any one purine. It oxidize/ a variety of aliphatic and aromatic aldehydes. When reacting with purines or aldehydes, i t also can reduce nitrates to nitrites (29,44). Keilin and Hartree (63) have shown that if ethyl alcohol and catalase are added to a digest containing xanthine oxidase and hypoxanthine, the ethyl alcohol is oxidized to acetaldehyde. The flavin moiety of milk xanthine oxidase is similar to, if not identical with, flavin adenine dinucleotide (23). Ball (5) succeeded in the reversible resolution of milk flavoprotein by prolonged dialysis against distilled water a t 0°C. Corran and coworkers (23) were unable to split the flavin reversibly from their preparation. Preparations of xanthine oxidase
MILK PROTEINS
221
dried or treated with cyanide are irreversibly inactivated. The dihydrocoenzyme I oxidase activity, however, is unaffected. Heating for 10 minutes at 80°C. denatures the protein and liberates the coenzyme (23). Xanthine oxidase has its isoelectric point a t pH 6.2. Ball has calculated its maximum molecular weight to be 74,000. The absence of xanthine oxidase in human milk has been made the basis of a test for the distinction between human and cow’s milk (108). 3. Esterases a . Lipase. Employing high-fat creams, saturated with sucrose as a preservative, for a substrate, Rice and Markley (106) indicated definitely that a true lipase resembling pancreatic lipase is present in cow’s milk. Hippius (52) and Moro (91) showed that human milk has a higher lipase content. The enzyme is inhibited by high acidity. It is destroyed by heat (20 minutes a t 63°C.) and ultraviolet light (62), and is inactivated by traces of heavy metals, 0.2% fluoride, and 0.1% hydrogen peroxide. Homogenation of milk activates the lipase. There is a definite relationship between the oestrus cycle and lipase activity (65). The amount of enzyme usually increases with advanced lactation and with any abnormality of the udder. The addition of pitocin has been shown to activate the tributyrinase activity of milk (64). b. Phosphatase. Raw milk contains a phosphatase that has the usual characteristics of mammalian phosphatase in its specific activation by magnesium ions and in its pH activity relations (39). The activity ranges from pH 6-10, with the optimum at pH 9. Massart and Vandendriessche (85) have shown that the phosphatase of milk is not inhibited by sodium fluoride. Inhibition is obtained by potassium cyanide and cysteine. The enzyme is activated both by zinc and magnesium. On the basis of inhibition and acceleration experiments on the activity of the enzyme, these investigators conclude that milk phosphatase is a metalloprotein and that zinc is the activating metal. Guiltonneau and coworkers (43) have reported the presence of two phosphatases in cow’s milk, as determined in butter and buttermilk. The enzymes are present in weak concentration in whole milk, are absent in completely skimmed milk, and are present in large amounts in buttermilk and the aqueous portion of butter Because they find two zones of maximal activity, pH 4.2 and 7.6 to 7.8, the former maintaining activity after heating to 73°C. for 50 minutes while, in the latter zone, activity is destroyed after heating for 20 minutes at 63”, they conclude that two enzymes in milk will convert pyrophosphate to orthophosphate. Milk phosphatase is destroyed by heat but more slowly than bacterial organisms so that it is used as a test for pasteurization (85).
222
THOMAS L. McMEEKIN A N D B . DAVID POLIS
4 . Protease
The presence of a presumably nonbacterial proteolytic enzyme in normal milk has been demonstrated by numerous investigators, notably by Thatcher and Dahlberg (134). The enzyme described is tryptic in nature; the proteins are broken down below the peptone stage Almost all the proteolytic activity in milk is precipitated with the casein when milk is acidified. Warner and Polis (141) showed the presence of a proteolytic enzyme in commercial and purified laboratory preparations of casein. The enzyme they studied had optimal activity a t pH 9.2 in borax buffer, and Warner obtained a 150-fold increase in its activity by precipitation of a dialyzed casein enzyme digest a t pH 4.5 and fractionation of the filtrate with ammonium sulfate. His concentrated enzyme showed four boundaries on electrophoresis, the proteolytic activity migrating with one of the two slower boundaries. 5. Oxidases a. Catalase. That catalase in milk is a secreted enzyme has been shown by Grimmer (40) and Harden and Lane-Claypon (49). The catalase content of milk varies with breed (69), individual animal, time of milking, and feed. Colostral milk has a higher catalase content than normal milk (105). The enzyme is not uniformly present in human milk. Cream and separator slime contain higher proportions of the enzyme than the milk. Increased bacterial or leucocyte counts are invariably followed by higher catalase contents. The catalase activity is precipitated with the casein. The maximum activity of the enzyme at 0°C:. occurs at p H 0.8 to 7.0 (7). The catalase activity of milk is completely destroyed by heating for 30 minutes at 65" to 70°C. b. Lactoperoxidase. This enzyme has the distinction of being not only the first enzyme demonstrated in milk (2) but also the only milk enzyme reported isolated in a crystalline state (135). This was accomplished by fractionation with ammonium sulfate to remove the casein, heating to 70°C. for 15 minutes to remove inert protein, precipitation of other impurities with basic lead acetate, removal of red impuritiea with acetone, and final purification by electrophoresis. The yield does not exceed 0.2 g. from 100 liters of milk or about 2% of the total amount present in milk. The enzyme is a hemin protein that has the properties of an albumin. The lactoperoxidase shows distinct differences, as compared with Agner's (1) verdoperoxidase isolated from leucoeytes. Recrystallized lactoperoxidase was found to be homogeneous in a physicochemical respect by Theorell and Pedersen (136). These investigators reported the molecular weight to be 93,000, the sedimentation
MILK PROTEINS
223
constant SZO= 5.37, the diffusion constant D ~ = o 5.95 X lo-’, the partial specific volume V = 0.764, the frictional ratio f/fo = 1.18. The iron content, 0.07%was slightly higher than 1 mole of iron per mole of enzyme. Light absorption coefficients reported in the region 2400-6900 showed peaks at 600, 500, 410, and 280 mp.
VII. RELATIONSHIP OF MILK PROTEINS TO SERUM PROTEINS The proteins of milk have often been compared with the proteins of blood serum (25,47,48,118,142). The purity of the proteins has always been a complicating factor in such comparisons. It was early recognized that casein differed chemically and biologically from the proteins of serum (127,142). The immunological evidence of Wells and Osborne indicated that whey proteins are related to the serum proteins of the same animal. Their results demonstrated, however, that lactalbumin differs from serum albumin. This finding was in agreement with the dissimilar composition of these two substances (25). The relatively complete amino acid analysis of Brand el al. (12) on 0-lactoglobulin and serum albumin also show that these proteins differ in composition. That the proteins of milk whey and of blood serum are characteristic for the species has been demonstrated by electrophoretic studies (28). Smith’s (119) comparison of the amino acid composition of the immune globulin of milk with that of the immune globulins of serum revealed a remarkably similar amino acid pattern, though individual differences were found. His mobility studies indicated a close similarity of the immune globulin of milk with the T-component of serum rather than with y-globulin. The component of whey in Table I1 with a mobility of -6.4 u, amounting to 5% of the protein of whey, has a mobility close to that of serum albumin (110). Based on immunological results, Peskett (101) has reported that normal milk contains a small amount of serum albumin and that it increases during the secretion of abnormal milk. ADDENDUM
Gordon et al. (147) have recently reported the results of complete amino acid analyses on whole casein, a-casein and &casein. Their results are given in Table VI. These values for the amino acid contents of the caseins supplement the previously reported values in Table 111. The apparent specific volumes of whole casein, a-casein and /%casein, calculated from densities (148), are in excellent agreement with the specific volumes calculated from the amino acid residues as reported by Gordon et al.
224
THOMAS L, McMEEKIN AND B . DAVID POLIS
TABLE V I Amino Acid Composition of Caseins g./lOO g. protein Whole Casein
-
a-Casein
,%Casein
Total N Total P Amino N
15.63 0.86 0.93
15.53 0.99 0.99
15.33 0.61 0.72
Glycine Alanine Valine Leucine Isoleucine Proline Phenylalanine Cystine bfethionine Tryptophan Arginine Histidine Lysine Aspartic acid Glutamic acid Amide N Seine Threonine Tyrosine
2.7 3.0" 7.2 9.2 6.1 11.3 5 .O 0.34 2.8 1.2 4.1 3.1 8.2 7.1 22.4 1.6 6.3 4.9 6.3
2.8 3.7" 6.3 7.9 0.4 8.2 4.6 0.43 2.5 1.6 4.3 2.9 8.9 8.4 22.5 1.6 6.3 4.9 8.1
2.4 1.7" 10.2 11.6 5.5 16.0 5.8 0.0-0.1 3.4 0.65 3.4 3.1 6.5 4.9 23.2 1.6 6.8 5.1 3.2
Total a
115.6*
115 7b
117.4*
These values are provisional.
* Total includes amino acids, amide
N calculated as ammonia, and phosphorus
calculated as phosphoric acid.
TABLE V l r Apparent Specijic Volumes of Caseins
Whole Casein a-Casein &Casein
Apparent specific volume calculated from amino acid residues
Apparent specific volume from densities a t 25°C.
0.731 0.725 0.743
0.731 0.728 0.741
MILK PROTEINS
225
Table VII shows the results obtained by the two methods for calculating the apparent specific volumes. The excellent agreement of the values by the two methods indicates that the apparent specific volume of a protein is essentially determined by the volume of its amino acid residues.
REFERENCES 1. Agner, K. (1941). A& Physiol. S a n d . 11, Suppl. VIII. 2. Arnold, C. (1881). Arch. Pharm. 219, 41. 3. Bailey, K., Chibnall, A. C., Rees, M. W., and Williams, E. F. (1943). Biochem. J . 37, 360. 4. Bain, J. A., and Deutsch, H. F. (1948). Arch. Biochem. 16, 221. 5. Ball, E. G. (1939). J. Biol. Chem. 128, 51. 6. Ballou, G. A., Boyer, P. D., Luck, J. M., and Lum, F. G. (1944). J . Biol. Chem. 165, 589. 7. Balls, A. K., and Hale, W. 8. (1932). J . Assoc. O&. Agr. Chemists 16, 483. 8. Beach, E. F., Bernstein, S. S., Hoffman, 0. D., Teague, D. M., and Macy, I. G. (1941). J . Biol. Chem. 139, 57. 9. Berridge, N. J. (1945). Biochem. J . 39, 179. 10. Bosshardt, D. K., Moore, D. H., and Brand, E. (1947). In E. Brand and J. T. Edsall, Ann. Rev. Biochem. 16, 256. 11. Boyd, G. L. (1922). Can. Med. Assoc. J . 12, 724. 12. Brand, E. (1946). Ann. N . Y . Acad. 81%47, 187. 13. Brand, E., Saidel, L. J., Goldwater, W. H., Kassell, B., and Ryan, F. J. (1945). J. Am, Chem. SOC.87, 1524. 14. Briggs, D. R., and Hull, R. (1945). J . Am. Chem. Soe. 67, 2007. 15. Bull, H. B., and Currie, B. T. (1946). J . Am. Chem. SOC.68, 742. 16. Burk, N. F., and Greenberg, D. M. (1930). J . Biol. Chem. 87, 197. 17. Cannan, R. K., Palmer, A. H., and Kibrick, A. C. (1942). J . Biol. Chem. 142, 803. 18. Chanutin, A., Ludewig, S., and Masket, A. V. (1942). J . Biol. Chem. 143,737. 19. Cherbuliea, E., and Meyer, F. (1933). Helv. Chim. Acta 16, 600. 20. Cherbuliea, E., and Schneider, M. L. (1932). Helv. Chim. Acta 16, '697. 21. Chrzaazcz, T., and Goralowna, C. (1925). Biochem. 2.166, 172. 22. Cohn, E. J., and Berggren, It. E. L. (1924). J . Gen. Physiol. 7, 45. 23. Corran, H. S., and Green, D. E. (1938). Biochem. J. 32,2231. Corran, H. S., Dewan, J. G., Gordon, A. H., and Green, D. E. (1939). Biochem. J . 33, 1694. 24. Crowfoot, D. (1941). Chem. Revs. 28, 215. 25. Crowther, C., and Raistrick, H. (1916). Biochem. J . 10, 434. 26. Damodaran, M., and Ramachandran, B. V. (1941). Biochem. J . 36, 122. 27. Davis, B. D., and Dubos, R. J. (1947). J . Ezptl. Med. 86, 215. 28. Deutsch, H. F. (1947). J. Biol. Chem. 169, 437. Deutsch, H. F., and Goodloe, M. B. (1945). J . Biol. Chem. 161, 1. 29. Dixon, M., and Thurlow, 8. (1924). Biochem. J . 18, 971, 976, 989. 30. Edsall, J. T. (1942). Advances i n Colloid Sn'. 1, 269. 31. Edsall, J. T. (1947). Advances i n Protein Chem. 3, 383. 32. Engel, H., and Schlag, H. (1924). Milchw. Forsch. 2, 1. 33. Fankuchen, I. (1943). In Proteins, Amino Acids and Peptides. E. J. Cohn and J. T. Edaall, eds., Reinhold, New York, p. 328. 34. Ferry, J. D., and Oncley, J. L. (1941). J . Am. Chem. SOC.63, 272.
226
THOMAS L. McMEEKIN AND B . DAVID POLIS
35. Gale, E. F. (1945). Biochem. J . 39, 46. 36. Giffhorn, A. (1906). Milchw Zentr. 2, 236. 37. Gordon, W. G., Semmett, W. F., Cable, R. S., and Doherty, D. G. (1947). Federation Proc. 6, No. 1, 255. 38. Could, B. S. (1932). J. Dairy Sci. 16, 230. 39. Graham, W. R., and Kay, H. D. (1933). J . Dairy Research 6, 54, 63. 40. Grimmer, W. (1913). Biochem. 2. 63, 429. 41. Gronwall, A. (1942). Compl. rend. lrau. lab. Carlsberg. Sbr. chim. 24, Nos. 8-11. 42. Groh, J., Kardoe, E., Denes, K., and Serenyi, V. (1934). 2. physiol. Chem, 416,32. 43. Guiltonneau, G . , Chevalier, R., and Jarrouse, H. (1944). Compt. rend. 218, 1006. (Chem. Abstracts, 1946. 40, 3836'). 44. Haas, P., and Hill, T. G. (1925). Biochem. J . 17, 671. 46. Hac, L. R., and Snell, E. E. (1945). J . Biol. Chem. 169,291. 46. Hammarsten, 0. (1883). 2.physio2. Chem. 7, 227; (1885). Ibid. 9,273. 47. Hansen, R. G., and Phillips, P. H. (1947). J . Biol. Chem. 171,223. 48. Hansen, R. G., Potter, R. L., and Phillips, P. H. (1947). J . Biol. Chem. 171, 229. 49. Harden, A., and Lane-Claypon, J. E. (1912). J . Hyg. 12, 144. 50. Heiduschka, A., and Komm, E. (1931). 2.physiot. Chem. 196, 187. 51. Henderson, L. M., and Snell, E. E. (1948). J . Biol. Chem. 172, 15. 52. Hippius, A. (1905). Jahrb. Kinderheilk. 61, 365. 53. Holter, H. (1932). Biochem. 2.266, 160. 64. Howe, P. E. (1921). J. Biol. Chem. 49,93. 65. Howe, P. E. (1921). J . Biol. Chem. 49, 115. 66. Howe, P. E. (1922). J . Biol. Chem. 62, 51. 57. Howe, P. E. (1922). J. Biol. Chem. 63, 479. 68. Howe, P. E. (1924). J. Ezpll. Med. 39, 313. 59. Jacobsen, C. F., and Christensen, L. K. (1948). Nature 161,30. 60. Jameson, E., and Alvarez-Tostado, C. (1939). J . Phys. Chem. 43, 1165. 61. Kassell, B., and Brand, E. (1938). J . Biol. Chem. 126, 435. 62. Kay, H. D. (1946). Nature 167, 511. 63. Keilln, D., and Hartree, E. F. (1936). Proc. Roy. SOC.London B119, 114. 64. Kelly, P. L. (1945). J. Dairy Sci. 88, 793. 65. Kelly, P. L. (1945). J. Dairy Sci.88, 803. 66. Keeton, A. S.,Udenfriend, S., and Cannan, R. K. (1946). J . A m . Chem. Snc. 68, 1390. 67. Klotz, I. M. (1946). Arch. Biochem. 9, 109. 68. Koning, C. J. (1908). Milchw. Zentr. 4, 156. 69. Kooper, W. D. (1911). Milchw. Zentr. 7, 264. 70. Kuttner, A., and Ratner, B. (1923). A m . J . DiSeaSeS Children 26, 413. 71. Lcvene, P. A., and Hill, D. W. (1933). J . Biol. Chem. 101, 711. 72. Lewis, J. H., and Wells, H. G. (1922). J . A m . Med. Assoc. 98, 863. 73. Li, C. H. (1946). J . A m . Chem. SOC.68, 2746. 74. Linderstrdm-Lang, K. (1929). Compl. rend. trau. lab. Cadsberg. Sbr. chin. 17, No. 9. 75. Linderstrgm-Lang, K., Hotchkiss, R. D., and Johansen, G. (1938). Nature 142, 996.
MILK PROTEINS
227
76. Linderstr$m-Lang, K., and Kodama, S. (1925). Compt. rend. trau. lab. Carlsberg. SLr. chin. 16, No. 1. 77. Lipmann, F. (1933). Biochem. Z.262, 3. 78. Lugg, J. W.H. (1938). Biochem. J. 32, 2123. 79. McMeekin, T. L., DellaMonica, E. S., and Custer, J. H. (1947). Federation Proc. 6, No. 1, 277. 80. McMeekin, T.L., DellaMonica, E. S., and Custer, J. H. (Sept. 1947). Abstracts of Papers, 112th meeting, Am. Chem. SOC.,New York, p. 27A. 81. McMeekin, T.L., Polis, B. D., DellaMonica, E. S., and Custer, J. H. (1948). Federation Proc. 7 , No. 1, 172. 82. McMeekin, T.L.,Polis, B. D., DellaMonica, E. S., and Custer, J. H. (1948). J . Am. Chem. SOC.70, 881. 83. McMeekin, T.L., and Warner, R. C. (1942). J . Am. Chem. Soc. 64,2393. 84. Macpherson, H. T. (1946). Biochem. J . 40, 470. 85. Massart, L.,and Vandendriessche, L. (1945). Enzymologia 11, 261. 86. Mellander, 0. (1939). Btochem. 2.300, 240. 87. Mellander, 0. (1947). Upsnla Ldkareforen. Forh. 62, 107. 88. Mellon, E. F.,Korn, A. H., and Hoover, S. R. (1947). J . A m . Chem. SOC. 69,827. 89. Menefee, S. G., Overman, 0. R., and Tracy, P. H. (1941). J . Dairy Sci. 24, 953. 90. Morgan, E. J., Stewart, C. P., and Hopkins, F. G. (1922). Proc. Roy. SOC. London B94, 109. 91. Moro, E. (1902). Jahrb. Kinderheilk. 66, 361. 92. Nicolet, B. H., and Shinn, L. A. (1946). Abstracts of Papers, 110th meeting, Am. Chem. SOC.,Chicago, p. 20B. 93. Nitschmann, H. (1938). Helv. Chim. A c h 21, 315. Nitschmann, H., and Guggisberg, H. (1941). Zbid. 24,434, 574. 94. Nitsohmann, H., and Lehmann, W. (1947). Ezperientia 3, 153. 95. Nitschmann, H., and Lehmann, W. (1947). Helv. Chim. Acta 30, 804. 96. Orcutt, M. L.,and Howe, P. E. (1922). J. Exptl. Med. 36, 291. 97. Osborne, T.B., and Wakeman, A. J. (1918). J . Biol. Chem. 33, 243. 98. Palmer, A. H. (1934). J. Biol. Chem. 104, 359. 99. Pedersen, K. 0. (1936). Biochem. J . 30, 948. 100. Pedersen, K.0. (1936). Biochem. J . 30, 961. 101. Peskett, G. R. (1932). Ann. Rept. Natl. Znst. Research Dairying, Reading, England, p. 52. 102. Plimmer, R.H. A,,and Lowndes, J. (1937). Biochem. J . 31, 1751. 103. Posternak, 8. (1927). Biochem. J . 21, 289. 104. Rees, M. W. (1946). Biochem. J . 40,632. 105. Reid, K. (1931). Milchw. Forsch. 11, 590. 106. Rice, F. E.,and Markley, A. L. (1922). J . Dairy Sn'. 6, 64. 107. Rimington, C., and Kay, H. I). (1926). Biochem. J . 20,777. 108. Rodkey, F. L., and Ball, E. G. (1946). J . Lab. Clin. Med.31, 354. 109. Rowland, S.J. (1937). J . Dairy Research 8, 1, 6. (1938). J . Dairy Research 9,42. 110. San Clemente, C. L., and Huddleson, I. F. (1943). Mich. Slate Coll. Agt. Expt. &a. Tech. Bull. 182, 3. 111. Schardinger, F. (1902). 2. Untersuch. Nahr. u. Genussm. 6 , 1113.
228
THOMAS L. McMEEKIN A N D B . DAVID POLIS
112. 113. 114. 115.
Schenk, R. (1928). Arch. wiss. u. prakt. Tierheilk. 68, 375. Sebelien, J. (1885). 2. physiol. Chem. 9, 445. Senti, F. R., and Warner, R. C. (1948). J. A m . Chem. SOC.70, 3318. Shankman, S., Camien, M. N., and Dunn, M. S. (1947). J . Biol. Chem. 168,
116. 117. 118. 119. 120. 121. 122. 123. 124.
Sjogren, B., and Svedberg, T . (1930). J . A m . Chem. SOC.62, 3650. Smith, E. L. (1946). J . Biol. Chem. 164, 345. Smith, E. L. (1946). J . Biol. Chem. 166, 665. Smith, E. L. (1948). J . Dairy Sci. 31, 127. Smith, E. L., and Coy, N. H. (1946). J . Biol. Chem. 164, 367. Smith, E. L., and Greene, R. D. (1947). J . Biol. Chem. 171, 355. Smith, E. L., and Greene, R. D. (1948). J . Biol. Chem. 172, 111. Smith, E. L., Greene, R. D., and Bartner, E. (1946). J. Biol. Chem. 164, 359. Smith, T., and Little, R. B. (1922). J. Exptl. Med. 36, 181, 453. (1923). J . Expll. Med. 37, 671. Smith, T., and Little, R. B. (194). J . Expll. Med. 39, 303. wrensen, M., and Palmer, A. H. (1938). Compt. rend. trau. lab. Carlsberg. S6r. chim. $1, No. 21. wrensen, M., and Sdrensen, S. P. L. (1939). Compt. rend. lrau. lab. Carlsberg 23, No. 7. Stoklasa, J. (1904). z. landw. Versuchsw. Deut. Oesterr. 7 , 755. Sullivan, M. X., and Hess, W. C. (1944). J . Biol. Chem. 166, 441. Svanberg, 0. (1930). Z. physiol. Chem. 188, 207. Svedberg, T., Carpenter, L. M., and Carpenter, D. C. (1930). J . A m . Chem. SOC.61, 241; 62, 701. Svedberg, T., and Pedersen, K. 0. (1940). The Ultracentrifuge. Oxford Univ. Press, London, p. 379. Svedberg, T., and Pedersen, K. 0. (1940). The Ultracentrifuge. Oxford Univ. Press, London, p. 408. Thatcher, R. W., and Dahlberg, A. C. (1917). J . Agr. Research 11, 437. Theorell, H., and h e s o n , A. (1943). Ark. Kemi., Mzneral. Geol. 17B,No. 7, 1. Theorell, H., and Pedersen, K. 0. (1944). The Suedberg Almqoist Wiksells, Upsala, p. 523. Timmerman, W. A. (1930). Brit. J . Expll. Path. 11, 447. (1931). Z. Zmmunik%lejorech. 7, 388. Tristram, G. R. (1946). Biochem. J. 40, 721. Vandevelde, A. J. J. (1908). Biochem. Z. 11, 61. Warner, R. C. (1944). J . A m . Chem. SOC.66, 1725. Warner, R. C., and Polis, E. (1945). J. A m . Chem. SOC.67, 529. Wells, H. G., and Osborne, T. B. (1921). J . Infectious Diseases 29, 200. Wichmann, A. (1899). 2. physiol. Chem. 27, 575. Williamson, M. B. (1944). J . Biol. Chem. 166, 47. Woodman, H. E. (1921). Biochem. J . 16, 187. Zaitschek, A. (1904). Arch. ges. Physiol. (PjEQgers) 104, 539. Gordon, W. G., Semmett, W. F., Cable, R. S., and Morris, hl. (1949). J . A m . Chem. SOC.71, 3293. McMcekin, T. L., Groves, M. L., and Hipp, N. J. (1949). J . A m . Chem. SOC. 71, 3298.
51.
125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148.
Plant Proteins
BY J. W. H. LUGG* Department of Biochemistry, University of Melbourne, Melbourne, Victoria
CONTENTS Page 230 . . . . . . . i . . . . . . . . . . . . . . . . . . . . 232 ......................... 232 ......................
1. Some General Remar
...................
a. Preparation and Properties.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . b. Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Applicability of Analytical Methods ......... Results Obtained with Preparations lants. . . . . . Nutritive Value. . . . . . . . . . . . . . . . . . . . . ..........
d. Granular and Fluid Parts of Cell Contents..
233 235 235 236 236
. . . . . . . . . 239
e. Pollen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ......................................
...................................... 5 . Pteridophytes
.................
d. Brief Assessment of Present Position.. . . . . . . . . . . . . . . . . . . . . . . . 254 10. Phytopathogenic Viruses ........... . . . . . . . . . . 254 11. Some Phylogenetic Considerations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
* Present address: Department of Biochemistry, University of Malaya, Singapore, Malaya. 229
230
J. W. H. LUQQ
Page 111. "Individual" Proteins of Plants.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259 1. Spermatophytes-Leaves of Angiosperms, Tubers. . . . . . . . . . . . . . . . . . . 260 2. Spermatophytes-Seeds of Angiosperms. . . . . . . . . . . . . . . . . . . . . . . . . . . 260
. . . . . . . . . . . . . . 263 3. Thallophytes (Fungi and Bacteria). . . . . . . . . . . . . 1. Nriclei . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3. Chloroplasts.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . 267
5. Bacteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
V. Protein Metabolism in Plants.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
268 269
1. Extent of the Field.. . . . . . . . . . . . . . . . . . . . . . 2. Certain Anabolic Aspects. . . . . . . . a. The Green Plant.. . . . . . . . . . Assimilation of Nitrate. .. Utilization of Ammonia. . . . . . . . . . . . . . . . Brief Summary of Present b. Nitrogen Fixation. ......... c. Sundry Plant Organisms.. .. d. Theories of Protein Synthesis 3. Protein Metabolism a. Earlier Work..
280 e. Brief Summary of Main Implications., . . . . . . . . . . . . . . . . . . . . . . . 281 4. Protein Metabolism in Starving Leaves.. . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Plant Protein Metabolism Generally-Regulation . . . . . . . . . . . . . . . . . . 282 a. Anabolism in Relation to Catabolism., . . . . . . . . . . . . . . . . . . . . . . . 282 h. Hormonic Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283 ion to Carbohydrate, Rrspiration, c. Protein Metabolism in and Other Factors.. . . . ................................ 284 Protein and Amino A ith Amide) Levels.. . . . . . . . . . . . . 284 Protein and Water Contents.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284 Protein and Carbohydrate Contents. . . . . . . . . . . . . . . . . 285 Protein Metabolism in Relation to Respiration. . . . . . . . . . . . . . 286 6. Questioned Exclusive Validity of Conflicting Theories. . . . . . . . . . . . . . . 288 Findings-Possible Significance of the Protein .......................................... 290 Protein-Virus Relationships, . . . . . . . . . . . . . . . . 282 VI. Conclusion ...................................... References. . . . . . . . . . . . . . . . . . . . . . . . . . . ............................ 295
I.
INTRODUCTION
The nature of a subject and the existing state of knowledgc about it may well prohibit the development of a thoroughly consistent and inter-
PLANT PROTEINS
23 1
locking scheme of subdivision in a review. As will be seen from the table of contents and from the text itself, phylogenetic considerations, derived essentially from the classification of Engler and Prantl, have provided the broad basis of the scheme adopted here, a t least in the earlier sections. Where, for one reason or another, it has seemed desirable to make departures I have done so, rather than adhere to a scheme which, in any event, is arbitrary. A survey of the literature* reveals that proteins belonging to the vast spermatophyte (flowering plants) division of the plant world have received far more attention than have those of all the other divisions together. Furthermore, within the spermatophyte division itself, the gymnosperms (conifers, etc.) have been almost entirely ignored. Very little work has been done on proteins belonging to the pteridophyte (club mosses, horsetails, and ferns) and bryophyte (mosses and liverworts) divisions. Those belonging t o the thallophyte division (in which some would include the viruses with the algae, fungi, actinomycetes, and bacteria), however, have received a considerable measure of attention. Man’s great, needs, as an animal, are sources of foodstuffs and freedom from disease; and the significance of the distribution of work on the plant proteins is to be seen in the light of these facts. It would be a difficult and lengthy undertaking to review adequately our knowledge of the plant proteins in all its aspects. It is not attempted here. Most of the space has been allotted to the sections dealing with the bulk proteins of plants and with protein metabolism. Relatively little space has been devoted to consideration of the “individual” proteins of seeds because, although an enormous amount of work has been done on these substances, it has provided material for many textbooks and much of the newer work is thus frequently reviewed in successive editions and in new texts. During recent years much information about plant enzymes has been accumulated and would, of itself, provide sufficient material for a fairly lengthy review. I have been obliged, therefore, to treat the enzymes mainly in incidental fashion. The bulk proteins have been considered chiefly with regard to amino acid composition and preparative methods. These are tending to become active fields of investigation and much remains to be done in them. Nutritive values for animals have been discussed briefly, and mainly insofar as feeding trials have yielded information which tends to confirm or contradict available analytical data when assessed on the bases of ( a ) what little is known about requirements and ( b ) the assumption, a t * The review has been based essentially on reports (some in abstract only) available in Australia by the end of 1947. Information obtainable in England up to March, 1948, has been incorporated subsequently.
232
J. W. H. LUQQ
present in dispute, that nutritive value is independent of the order of peptide linkages and is determined solely by amino acid composition. For reasons given in Section II,1 the physical properties of the bulk proteins can usually be described in only the broadest terms. I n the sections dealing with protein metabolism attention has been given to possible routes and modes of both synthesis and breakdown, and to the influence upon them of factors such as phylogeny, ontogeny, and aspects of plant nutritional status. Some consideration has been given to the question of the origin and significance of the protein cycle within living tissues. The term “protein” has been used in this review variously to designate both individual substances and mixtures of them. The terms “purity” and “pure” as applied to proteins have been used generally to designate freedom from nonprotein contaminants. The phrase “amino acid composition” has been used to cover amino acid, imino acid, and amide group contents. The phrase “cystine plus cysteine” has been condensed to “cyst(e)ine.” “Caloric ” is the normal (15’) calorie. 11. BULKPROTEINS OF PLANTS, PLANT ORQANS,ETC. 1. Some General Remarks For many years the only substantial information about plant proteins as substances pertained to those proteins which had been extracted in almost pure, and sometimes almost individual, state from seeds. But during the past decade much information has been gained concerning the amino acid compositions of the bulk proteins of plant tissues. The preparations normally contain several proteins and are more or less impure, and thus only the broadest conclusions can be drawn concerning physical properties. The composition data have long been required for use in dietetics and nutrition studies, and while the tardiness of their appearance is attributable to several causes one of these causes is of particular importance. For it is only during recent years that methods have been evolved which, under suitable conditions, permit the estimation of certain amino acids in impure preparations with a reasonable belief that some reliance can be placed in the results. Shortcomings of some of the newer methods, even when known, are not aiways appreciated by those who use the techniques. Hence the results of some of even the most recent analyses should be accepted with reserve. Preparations of the bulk proteins of plants have generally been obtained in a denatured condition. It is not known whether the types of denaturation undergone have appreciably changed the amino acid
PLANT PROTEINS
233
compositions of the proteins biit the possibility cannot be ignored. The unmasking or freeing of sulfhydryl groups by denaturation is a common phenomenon, but liberation of amino acid by hydrolytic cleavage (as shown by Bailey, 1942, to occur in the conversion of edestin into edestan) may not occur in heat or alcohol denaturation at, or near, the isoelectric points of the proteins. Values shown in the tables on amino acid composition have been reported on a nitrogen basis ( i e . , as percentages of the protein nitrogen occurring in amino and imino acids and amide groups). This facilitates making comparisons of composition between proteins contaminated to varying degrees with impurities (Lugg, 1938a). In this connection see Vickery and Clarke (1945). A brief summary of some of the more important conclusions t o be drawn from the work reviewed in Section I1 is embodied in the discussion of phylogenetic considerations (see Sect. II,11). (See also the conclusion, Sect. VI.) 2. Spemtatophytes-Leaves of Angiosperms a. Preparation and Properties. It is noteworthy that the methods used by Rouelle (1773a,b), Fourcroy (1789), and other early workers (references are given by Osborne, 1924) have provided the broad basis of methods employed by Subsequent investigators. Serious work of modern nature began about 1920, when Osborne and Wakeman (1920) and Chibnall and Schryver (1920) separated crude protein preparations from spinach and cabbage leaves, respectively. Osborne, Wakeman, and Leavenworth (1921) macerated leaves with water and employed, as subsequent solvents, dilute aqueous and aqueousalcoholic sodium hydroxide solutions in an attempt to extract all the protein. Chibnall and Schryver (1921) cytolysed leaves with etherwater before macerating them. Subsequently Chibnall and collaborators (Chibnall, 1923; Miller and Chibnall, 1932; Chibnall, Miller, Hall, and Westall, 1933) freed the cytolyzed leaves from vacuole contents by pressure before macerating them to extract protein. “ Used ” ether-water (i.e., ether-water already used a t least once) was found to afford better yields of protein than did fresh ether-water or ether. During this period Chibnall and colleagues were carefully filtering the leaf extracts in order to obtain purer protein on flocculation. Chibnall (1939) has attributed the improved yields with “used” ether-water to two factors: ( I ) diminished impairment of the solubility of the protein in the fluid cytoplasm, and (2) liberation of protein from the chloroplasts. Vickery (1945) has suggested that the proteins were probably more soluble a t the higher salt concentrations associated with the employment
234
J. W. H.
LUQQ
of “used” ether-water. However, Lugg (1938a) encountered diminished extraction in the presence of moderate amounts of sodium chloride, and Crook (1946) found Vickery’s suggestion unacceptable. When plant leaves are macerated in the presence of water or a mildly alkaline buffer, the chloroplasts may be mostly fragmented (see Granick, 1938a,b; IIanson, 1941). Even so, from comparison of results obtained by Granick, Hanson, and Menke (1938), Chibnall (1939), and Lugg (1939a), the chloroplasts and their fragments may be expected to retain almost all of their protein (about 40% of that in the leaf cells). Lugg (1939a) observed that most of the protein in what he termed the “granule” portion (whole and fragmented chloroplasts, nuclei, etc.) of the protoplasm could be released therefrom to mildly alkaline buffers, in the presence of ethanol and ether a t suitable concentrations. This observation provided the basis for the alkaline buffer-lipide solvent methods (Lugg, 1939a; Lugg and Weller, 194413) of obtaining from leaves protein samples of reasonable representativeness and purity. Although there is no indication that leaf proteins suffer damage a t p H 9.2 (Lugg, 1939a), Crook (1946) preferred extraction a t pH 8.0. He did not prepare protein samples, however. Crude “whole ” protein preparations, made by removing nonprotein nitrogenous substances from leaves, are amenable only to restricted analytical work but have been of value in permitting some examination of the represcntativeness of extracted protein preparations to be made. From materials such as (and similar to) these, some workers have extracted protein in altered form for restricted analytical work. Miller (1935) used hot dilute hydrochloric acid; Tristram (1939) employed hot dilute sulfuric acid; Wilkins (1937) and Albanese (1944) used hot 90% formic acid, and Hamilton, Nevens, and Grindley (1921) used alkali. I n the course of preparation, protein samples have been extracted with aqueous solvents, with ethanol and ether, and sometimes with acetone. They are then found to be almost free of lipide and are in a denatured condition. They contain varying amounts of polysaccharide, polyuronide, or polysaccharide-polyuronide complexes, but none, or very little, of these substances appears to be held as prosthetic groups. The living cells unquestionably contain conjugated protein (see Sect. IV) but much of what we know a t present about the compositions and physical properties of leaf proteins pertains to the simple proteins and to the simple protein moieties of the conjugated proteins. Rouelle fractionally flocculated the green plant juice by heating. Menke (1938) fractionally flocculated leaf proteins from solution with the aid of ammonium sulfate, ethanol, and acid. Chibnall, Lugg, and colleagues have extensively employed acid as a flocculating agent,
PLANT PROTEINS
235
and Chibnall’s earlier observation that the proteins are of minimum solubility in the pH range 4-5 has been confirmed for many species of plants. It has been observed repeatedly that the proteins are more freely soluble on the alkaline than on the acid side of this range. Lugg (1939a) found the solubilities to be depressed still further in the presence of ethanol, and noticed that the solubility relationships of the “granule” and “ nongranule ” (essentially fluid cytoplasmic) proteins were similar. Inasmuch as the fluid cytoplasm itself probably contains several major protein components (Frampton and Takahashi, 1944, 1946), it would seem that all the chief individual proteins of the leaves (at least when freed from prosthetic groups) are of low solubility, and presumably have isoelectric regions, within the same pH range. The protein in leaves dried even at fairly low temperature is found to be almost insoluble in water or mildly alkaline buffers. Again, acidflocculated leaf protein is readily coagulated (denatured) if heated above about 50°C. At about pH 5-4the leaf proteins rather readily undergo surface denaturation. In these facts is illustrated the lability of the leaf proteins. In the reviewer’s opinion the bulk proteins of leaves cannot be placed satisfactorily in any of the classes of the well-established classifications of proteins, and need a place of their own. The prospect of utilizing leaf protein concentrates in human nutrition has been discussed by Pirie (1942). Slade, Branscombe, and McGowan (1945) have considered techniques for the preparation of protein concentrates from fresh pasturage, and Sullivan (1943) has suggested the manufacture of protein concentrates from dried pasturage, the first stage being the dissolution of the protein in 0.25 N sodium hydroxide. b. Composition APPLICABILITY OF ANALYTICAL METHODS. Methods of studying the amino acid composition of plant proteins which are now known to be unreliable were used in the earlier analytical work (e.g., Osborne, Wakeman, and Leavenworth, 1921; and references given by Chibnall, 1939). It is only during the past decade that amino acid analyses, in any way reliable, have been made. The effects of impurities upon analyses (see Chibnall, 1939; Lugg, 1939b; Tristram, 1939; Kuiken, Norman, Lyman, Hale, and Blotter, 1943) have tended to enforce upon workers modifications of existing methods and the development of new ones, but some doubts remain as to the validity of results obtained. I n the first place the precise nature of the major organic contaminants of extracted protein preparations (believed by Lugg, 1939b, to be pentosan, pectin, or mucilage) is unknown. In the second place, some of the methods designed to overcome possible effects of such substances are much less specific than those which could be used with pure, simple proteins.
236
J. W. H. LUGQ
RESULTS OBTAINED WITH PREPARATIONS FROM MATURE PLANTS. Most of the work has been done on preparations from leaves (often with attached petioles) of plants which were neither extremely young nor senescent, but many of the preparations have been of uncertain representativeness. The nitrogen contents of the preparations from leaves have usually been within the range 13-15%. In Table I are shown ranges of values of amino acid contents of such preparations from several families of spermatophytes. Results obtained by methods now known to be unreliable have been excluded. Points which have emerged from this work are ( I ) that the composition of the bulk protein of the leaf for any one species probably varies but very little with manurial treatment, climate, and age of tissue (within the age limits already mentioned), (2) that variations in compositions among plant families are probably slight, but may be definite-e.g., the relatively low cyst(e)ine and methionine values for the Leguminosae compared with those for the other plant families. NUTRITIVE VALUE. Computations by Mitchell and Block (1946) suggest that in the leaf protein of alfalfa the isoleucine level may constitute a first limiting factor in animal nutrition. The feeding trials of Haag (1931) and of Marais and Smuts (1939) suggest instead that the level of the sulfur-containing amino acids may be limiting. From what is known of the nutritional requirements of animals, bulk proteins of the leaf of mature spermatophytes (even the legumes) can be considered of fairly high “biological value.” In ruminant nutrition the matter is complicated by the possibility that the metabolic activity of the rumen flora, especially in regard to synthesis of protein, may modify the over-all nutritional worth of the total nitrogenous substances present in the fodder; but there is evidence (see Sect. II,4,b) that in some instances, a t least, the protein composition of fodder may nevertheless be correlated with body weight increase and wool growth of sheep. The utilization of nonprotein nitrogen by ruminants has been discussed recently by Owen (1947). EFFECTS OF EXTREMES OF PLANT AGE AND OF HARVESTING HISTORY.
Little work has been done with leaf proteins belonging to the extremes of plant age. Lugg and Weller (1941) found the protein in axial organs of etiolated Lupinus anguslijolius (New Zealand blue lupine) seedlings (most of it was contained in the plumules and stalks) to be of low tryptophan and methionine content in comparison with the leaf proteins of other, mature legumes. They found also (Lugg and Weller, 1948b) that the protein in a batch of senescent leaves of Trijolium subterraneum (subterranean clover) was of distinctly lower methionine and distinctly higher cyst(e)ine content than that in leaves of the mature plant.
237
PLANT PROTEINS
TABLE I Amino Acid Compositionsa of Bulk Proteins in Leaves of Various Spermutophytes, Arranged i n Families P
Amino acid Glycine Alanine Valine Leucine Isoleucine Phenylalanin c Cyst (e)ine Methionine Tymsine Tryptophan Arginine Histidine Lysine Aspartic acid Glutamic acid Profine Threonine Amide
Solanaceae
1.3-1.4 1.4-1.5 2.6 1.8-1.9
-
-
-
-
5.3-5.6
-
T biaceae
-- I I
-
2.6 1.7 18.9
-
6.6 6.6 6.6
-
minosae
-
-
4.6 7.3 3.6
8.4 1.1-1.3 1.2-1.4 2.3-2.6 1.6-1.9 13.0-14.0 3.8-4.0 6.4-6.5 4.7-6.4 6.4-6.7 -
4.0 5.1-5.3
Cruciferae
Chenopodiaceae
-
1.4-1.6 1.4-1.5 2.6-2.7 1.4-1.5 18..4-14.1
-
6.8-6.8
-
6.0-6.0
Gramineae 0.4b 4.4"5.1 3.3-4.8' ~7.1-8.8 8.6"2.6 1.3-1.5 1.4-1.6 2.3-2.5 1.8-2.1 13.7-14.3 3.6-3.7 6.3-6.6 4.9-6.4 6.6-7.8 3.1 3.0
4.7-5.3
As per cent protein nitrogen. Compiled from the data of Chibnall(1939), Lugg (1938a,b, 1939a, 1943, and unpublished work), Tristram (1939), Martin and Synge (1941), Lugg and Weller (1944b, 1948a), Lugg and Best (1945), and Block and Bolling (1945). Italicized values are for preparations of untested representativeness. * These valuee, obtained by an ester distillation method, were formerly assumed to be underestimated. From comparison to the values for a different preparation, obtained by a partition chromatographic method, it seems probable that they were, however, not underestimated. Di-iodotyrosine could not be detected in those preparations which were submitted to the necessary tests. Members of plant families variously represented in the table are: Solanaceae: Nicotiana tabacum (tobacco). Euphorbiaceae: Ricinus communis (castor bean). Leguminosae: Medicago saliva (lucerne, alfalfa), M . denticulata (burr medic), Trifolium repens (wild white clover), T . pralense (red clover), T . subterraneum (subterranean clover), Phaseolus vulgaris (runner bean), P . mulliflorus (scarlet runner bean). Cruciferae : Cochlearia amoracia (horse-radish), Crambe cordifolia (kale). Chenopodiaceae: Atriplex nummularium (old-man saltbush), Spinacea oleracea (spinach), Beta cicla (beet spinach). Gramineae: Dactylis glomerata (cocksfoot), Lolium perenne (perennial rye grass), L. italicum (Italian rye grass), Cynosurus crystatus (crested dog's tail), Festuca rubra var. (Chewing's fescue), F . duriuscula (hard fescue), Poa trivialis (rough-stalked meadow grass), Phleum pratense (timothy), Phalaris tuberosa (Toowoomba canary grass), Hordeum murinum (barley grass), Zea maye (maize, Indian corn). a
238
J. W. H. LUQQ
Smith and Wang (1941), from a n extensive study of the cyst(e)ine and methionine contents of the leaf proteins of several pasture species, concluded that the protein composition was affected by treatments such as early and late harvesting. I t is unfortunate that results obtained by the methods of estimation they employed are subject to serious systematic errors (Lugg, 1938a), and th at these can be so capricious that even a careful statistical analysis could fail to disentangle their effects and assure the drawing of valid conclusions from the data. In view of the findings of Lugg and Weller for senescent leaves, i t must be conceded th a t the treatment effects claimed by Smith and Wang may be real. Further investigation is desirable, particularly in view of the fact that hays are usually made from senescent plants and from plants which have died and dried out in the field. NITROGEN CONTENTS OF “PURE ” PREPARATIONS. Estimates of the nitrogen contents of the hypothetically “p u re” bulk proteins of leaves have been sought for the following reasons: ( 1 ) in order th a t a calculated allowance might be made for the effects of the impurities upon analytical procedures, ( 2 ) so that the actual mass of protein in a pasture might be calculated from the protein nitrogen content, ( 3 ) in order to correct heats of combustion of the protein preparations to a “pure” protein tlasis, and ( 4 ) to satisfy simple curiosity concerning magnitudes of nitrogen contents. The estimates have been based upon partial amino acid analyses (Miller, 1936: 16.0% nitrogen; Lugg, 1939b: 16.5% nitrogen) and upon the amounts of humin yielded by acid hydrolysis (Lugg, 193913: 16.7 and 16.8% nitrogen for two preparations from one species; Lugg and Stanley, 1948: 16.1-16.7% nitrogen for preparations from four different species). Lugg and Stanley considered the variation among their preparations to be fortuitous. As bulk protein preparations from leaves seem to be of very similar amino acid composition, they gave weight to the value, 16.5y0 nitrogen, computed by Lugg (1939b), and for calculation purposes used a common value, 16.4% nitrogen, for their four preparations. Chibnall (1939) used the value 17% nitrogen for purposes of calculation. c. Heats of Combustion. Hitherto there has been no information available for use in metabolism studies, concerning the heats of combustion of leaf proteins. T o fill the gap Lugg and Stanley (1948) have recently measured the heats of combustion of reasonably representative preparations obtained from four species of pasture plants. The preparations were impure and corrections, believed to be appropriate, were made for the heat contributions of the impurities. The values, expressed as calories per gram of hypothetically pure protein (supposedly containing 16.4 % nitrogen) were : Phalaris luberosa, 5839; Hordeum murinum, 5882;
PLANT PROTEINS
239
Medicago sativa, 5888; and Medicago denticulata, 5899. The fact that these values are so similar is in harmony with the indications th a t the amino acid compositions of the proteins differ but little from one another. The values lie close to the upper limit of the range of values for seed proteins (Sect. I I I , Z , e ) . The leaf proteins are known t o contain, as residues, moderately large amounts of arginine, lysine, and the leucines, and not inconsiderable amounts of valine, tyrosine, and phenylalanine. From what is known and what may be inferred these residues would have relatively large heats of combustion per gram. T o these facts the rather large heats of combustion of the leaf proteins are probably attributable. d. Granular and Fluid Purts of Cell Contents. Following upon the work of Noack (1927), Menke (1938) separated by centrifugation nuclei, chloroplasts, and fluid cell contents (remaining cytoplasm plus vacuolar fluid) from the juice obtained by macerating plant leaves. Fractionations were also effected by differential precipitation with acid, ethanol, and ammonium sulfate. The fates of the other protoplasmic inclusions, such as mitochondria and microsomes, probably varied with the conditions. In his earlier work, using ether as a cytolyzing agent, Chibnall separated leaf cell contents into vacuolar and protoplasmic fractions. By filtration the latter was divided into what he called “chloroplastic” and (L cytoplasmic ” fractions (Chibnall, 1939). Partial amino acid analyses (Chibnall, 1939) of protein samples obtained from the vacuolar, protoplasmic, and ((cytoplasmic”material, indicated that the vacuolar protein was a little higher and the “chloroplastic” protein a little lower in lysine content than the “cytoplasmic” protein, and that the vacuolar protein was of somewhat lower tryptophan content than the other two fractions. The reported histidine contents must be considered unreliable (see Lugg and Weller, 1948a). Chibnall’s vacuolar protein represented only a few per cent of the total protein and may have consisted, in part, of “cytoplasmic ” protein washed through ruptured cell walls. Noack and Timm (1942) (see also Timm, 1942) found the “cytoplasmic” protein from spinach leaves to be richer in lysine and glutamic acid than the “chloroplastic” protein, but considered the composition differences (values for other amino acids were also compared) to be small. Lugg (1939a) encountered fairly small composition differences between preparations of leaf proteins made in various ways, and attributed them to variation in the extents of inclusion of the granule fraction. The existing evidence, then, is that the bulk proteins of the granular and fluid portions of the protoplasm (or, indeed, of the entire cell contents) are of fairly similar amino acid composition. For reference t o the
240
J. W. H. LUQG
physical properties of extracted “granule” protein and modes of occurrence see Sections II,2,a and IV,3. 3. Spermatophytes-Miscellaneous
(Roots, Tubers, (‘Fruits,” Etc.)
I n view of the importance of root crops in human and livestock nutrition, the root crop proteins merit more attention than they have thus far received. The only extensive analytical work appears t o be that of Davies (1926; 1927a,b), who prepared protein samples from turnip, mangoid, carrot, and parsnip roots and submitted them to partial amino acid analysis by procedures now known to be unreliable. a. Turnips. An earlier study of a (‘soluble” protein obtained from Swede turnips was made by Williams (1916-1917). I t is questionable that the preparation was other than reasonably representative of the protein in the cells that had been ruptured. Of the protein nitrogen he found 0.3% as glycine, 3.6% as alanine, 7.5% as valine, 6.1% as leucine(s), 2.4% as phenylalanine, 1.4% as tyrosine, 3.2% as proline, 4.7% as aspartic acid, 1.9% a8 glutamic acid, 6.4% as arginine, 5.2% as histidine, and 5.3% as lysine. The values for tyrosine, glutamic acid, and arginine are considerably lower than corresponding values for leaf bulk proteins. But a sample of turnip protein, prepared by Williams, was found by Lugg (1938b) to have cyst(e)ine and methionine contents similar to those of bulk proteins of spermatophyte leaf. If it be assumed that Williams’ tyrosine, glutamic acid, and arginine values were depressed by unexpectedly large losses (he used isolation methods), it would seem that the bulk proteins of roots may be of rather similar composition to those of leaves. b. Potatoes. The potato tuber is one of the most important foodstuffs in human nutrition, and the quality of the protein has long been the subject of feeding investigations. Among the more recent are those of Groot (1942), Lintzel (1942), Hutchinson, Bacon, Macrae, and Worden (1943), and Chick and Cutting (1943); the last-named authors paid particular attention to the nutritive value of the nonprotein nitrogenous constituents of the tuber (see Sect. V,2,a). In general, the feeding trials indicate that the total nitrogenous constituents of the tuber are of high nutritive value, and that the protein itself is of fairly high value. Kiesel, Belozersky, Agatov, Bivshich, and Pavlova (1934) made partial amino acid analyses of potato tuber protein and potato-leaf protein preparations and obtained similar values for them. Their preparations seem to have been of fair purity, and although their analytical methods may not have yielded reliable values the conclusion to be drawn from the comparison may be valid.
PLANT PROTEINS
24 1
Levitt (1946) has extracted samples of protoplasmic protein from potato tubers and estimated the average molecular weight to be 40,00050,000. c. “Fruits.” The juices of many “fruits” (used here in the popular, not botanical, sense) are distinctly acid. When the fruit is macerated the protein may be flocculated and then, if denaturation follows, it cannot be dissolved from the residue without the aid of reagents which may cause some destruction. Thus Smith (1925) and Sinclair, Bartholomew, and Nedvidek (1935), who used alkali to dissolve the protein, were almost certainly examining altered material. The difficulty might be overcome by macerating such fruits in the presence of mildly alkaline buffers of high buffer capacity. Smith argued that, since the orange protein could be precipitated from solution a t a pH value (4.7) greater than that of the orange juice (4.3), it must exist in insoluble form in the fruit. This does not follow. In the first place he was dealing with (probably) denatured and further modified protein, and in the second place the cell protoplasm is not necessarily of the same pH value as the vacuole contents. By the now largely abandoned Van Slyke nitrogen distribution method of analysis, Sinclair, Bartholomew, and Nedvidek found about 6% amide nitrogen, 12.5% arginine nitrogen, 2.6% histidine nitrogen and 10.6% lysine nitrogen in orange protein; these values are rather similar to those which may be obtained for leaf proteins by the same method (Miller and Chibnall, 1932). With the aid of alkaline buffers, Hulme (1946a,b) obtained crude protein preparations from apple pulp. They contained oxidase and peroxidase. d . Latex. Bondy and Freundlich (1938) obtained two protein fractions from ammonia-preserved rubber latex, and Kemp and Straitiff (1940) asserted that three proteins could be separated from the latex. Tristram (1940, 1941) was unable to isolate individual proteins from the dried latex of Hevea brasiliensis or from crepe rubber. He was of the opinion that his preparations were reasonably representative of the total protein present and, from partial amino acid analyses, concluded that they were of similar composition to those of the bulk proteins of spermatophyte leaves. However, Altman (1939) reported a high proline content for the protein in Hevea latex. e. Pollen. Little work has been done on pollen proteins except in regard to their relationships with allergies. Vivino and Palmer (1944) found that ether-extracted pollens from the dandelion, clover, and other flowers probably contained insufficient tryptophan and methionine (in conjunction with the cyst(e)ine present) for normal rat growth.
242
J. W. H. LUGG
4. Spermalophytes-Seeds
of Angiosperms
The term “seed” is used in its popular sense in this review. Thus it is meant to embrace true seeds, such as the soybean, and certain fruits, such as the so-called “seed” of the maize (corn) plant. As Vickery (1945) has pointed out, the simple plant seed is one of the most important organisms with which we are concerned, since i t is the basis, directly and indirectly, of almost all our foodstuffs. We cannot afford to confine our attention entirely to the protein content when we regard the seed as a source of the essential amino acids, because some of these acids may exist in seeds in the free state or combined in simple peptides. Before work on the compositions of the proteins is undertaken some preliminary prcparative work must be performed. a. Preparation. Considered as crude protein preparations the powdered seeds of plants are crude indeed. They contain in addition to the protein amino acids, fats, carbohydrates, polyuronides, salts, nonprotein nitrogenous substances, and sundry other substances, the nature and amounts of which depend greatly upon the particular type of seed under consideration. When oil-bearing seeds are extracted with lipide solvents the residues may consist very largely of protein, and in recent years estimates of the amino acid compositions of such residues have been reported, but the values (even on a nitrogen basis) cannot safely be regarded as giving the amino acid Compositions of the seed proteins. The physical properties of the various individual proteins in most seeds are usually so diverse that no mild single-extraction method can be expected to yield a sample of reasonable purity and reasonable representativeness. Wormell’s (194G) process, which involves the use of alkali, could be effective in bringing most of the protein into solution, but only a t the expense of some destruction. The same criticism applies to the method of Hamilton, Nevens, and Grindley (1921). If a succession of solvents (cach suitable for the extraction of a particular type of protein) is employed there is still a risk that the total protein extracted will be lacking in representativeness. For these reasons, and because there is evidence that certain analytical methods may be applied to protein preparations grossly contaminated with polysaccharide and polyuronide irnpuritics, some workers have first directed their cfforts to the removal of nonprotein nitrogenous and certain other impurities from the seeds, and have then analyzed the residues. Other workers have at,temptcd to extract all the protein, or nearly all of it, from sccds (but often in a greatly modified condition) by procedures
PLANT PROTEINS
243
which are designed to leave most of the polysaccharide and polyuronide behind, and have analyzed the extracts. Lugg and Weller (1944a) removed nonprotein nitrogenous and certain other substances from legume seeds. I n their procedure water-soluble nonprotein nitrogenous substances are removed by ethanolic-aqueous extractants, and lipides (including those which contain nitrogen) are removed by ethanol and ether extractions. Csorika (1039) used 21 % hydrochloric acid solution to dissolve starch from samples of pulverized maize. Addition of ethanol to the extract precipitated the starch, leaving presumably modified protein in solution. The residue from the 21 % hydrochloric acid solution extraction was treated successively with saline and with diluted ethanol for the extraction of protein fractions. He was successful in separating most of the starch from the protein. Block (1945) (see also Block and Bolling, 1945) following the work of Doty (1941) found it possible to remove most of the starch from defatted, pulverized cereal seeds by first treating with hot water and then, after cooling, with salivary amylase. The residues contained about 90% of the original nitrogen and it seems, therefore, th a t very little protein could have been lost. Albanese (1944) found the extraction of protein from seeds with hot 90% formic acid to be unsatisfnctory. b. Composition. As amino acid losses can be large in the acid hydrolysis of protein preparations contaminated with polysaccharide or polyuronide, much of the voluminous data which have appeared during the past few years on seed protein compositions should be treated with some caution. The literature up to 1945 has been reviewed by Block (1945) and by Block and Bolling (1945), who have compiled some valuable tables. They have themselves provided much of the more recent data. The variability among types of seeds is rather high, but, since the compilations of these authors are readily available, no useful purpose would be served in reproducing their tables in extensive fashion. I n Table 11, however, are shown values (on a nitrogen basis), calculated from their data and from those of other workers, for the bulk proteins of two monocotyledonous and two dicotyledonous seeds. The least reliable available values have been excluded. Block (1945) has drawn attention to the rather low lysine contents of cereal seed proteins. His tables indicate that the proteins of oat and rice seeds are superior to those of wheat and maize seeds, from an animal nutrition point of view. Other important conclusions drawn by Block are that the proteins of linseed and sesame seed are relatively rich in
244
J. W. H. LUOQ
tryptophan and that the proteins of the peanut are poor in lysine, tryptophan, methionine, and threonine. Recent reviews of the nutritive values of legume seed proteins are those of Payne and Stuart (1944) for the soybean, and Jones (1944) for the soybean and the peanut. Jones' assessment of the nutritive value of peanut protein is a t variance with Block's. Attention has been drawn (Lugg, 1945,1946; see also Johanson and Lugg, 1946) to the fact that, in general, legume seeds and their bulk proteins seem to be of very low or rather low methionine content (highest for soyTABLE I1 __-___
Amino Acid Compositions" of Bulk Proteins of Certain Seeds Amino acid
I
Sunflower
Soybean
~
Maize
-
Wheat
3.2 2.7 4.4 3.8 3.1 2.2 2.7 3.0 3 0 0.9b 1.3b 1.lb 1.0 2.w 0.9 1.7 2.0 2.7 2.1 1.1 0.7 1.0 1.0 16.5" 9.76 10.8c 7.6" 2.9 3.6 4.2 3.9 4.6 6.5 3.2 2.4 3.0 3.0 2.4 2.7 I _____ __ __ -~ -~ Compiled from the data of Block (1945), Block a As per cent protein nitrogen. and Bolling (1945), Kuiken, Norman, Lyman, Hale, and Blottcr (1043), Greene and Black (1944), and Johanson and Lugg (1946). b Values of doubtful reliability. e Values likely to be low.
Valine Leucine Ideucine Phenylalanine Cyst (e)ine Methionine Tyrosine Tryptophan Arginine Histidine Lysine Threonine
-
3.9 4.1 3.5 2.9
-
~
--
~
bean); the cyst(e)ine values for the proteins range from very low to moderately high, and the tryptophan values are rather low. The results ~of feeding trials (Hayward and Hafner, 1941 ; Almquist, Mecchi, Kratzer, and Grau, 1942; Woods, Beeson, and Bolin, 1943; Peterson, Lampman, Bolin, and Stamberg, 1944) indicate that the methioninc contents of legume seeds may constitute a first limiting factor in the nutrition of the rat and the chick. Furthermore, a study of body weight increase and wool growth of sheep fed various legume seeds as the main sources of dietary protein indicates that those seeds in which the bulk proteins have notably low cyst(e)ine and methionine content are of decidcdly low nutritive value (Stewart and Pvloir, 1947). The analytical values (obtained, as they have been, with the aid of methods of low specificity) thus receive a degree of confirmation from the results of the feeding trials,
~
PLANT PROTEINS
245
as the cyst(e)ine and methionine contents of a seed could be greater, but not less, than those of the bulk proteins in it. Everson, Steenbock, Cederquist, and Parsons (1944) found that freshly germinated and immature soybeans contained protein of higher nutritive value than those of raw mature beans. Similar findings were reported for immature and mature soybeans and cowpeas by Sherman (1941). Possible changes in the amounts of protease inhibitors (see below) may lie behind these observations. 1,ugg and Weller (1944) found but little difference in cyst(e)ine and methionine contents of the proteins in mature and immature peas, and their work on New Zealand blue lupine seeds resulted in similar findings. The well-known apparent improvement in the nutritive value of soybean protein, which is brought about by heating the seeds, is now accepted as being due to the destruction of one or more protease inhibitors by the heat. The nature of the main inhibitor is discussed in Section 111,2,c. In the literature there is a vast amount of additional information (which cannot be reviewed hcre) concerning the digestibilities, nutritive values, and supplementary relationships of proteins in various seeds. c. Parts of Seeds. To the embryo is attached, in the case of the monocotyledonous seed, the endosperm, which is generally of high starch content. I n the dicotyledonous seed the embryo lies between the two cotyledons and is attached to them. In fact, the cotyledons are sometimes regarded as constituting part of the embryo. Certain dicotyledonous seeds are provided also with an endosperm. The endosperm and cotyledons contain the bulk of the food reserve required by the growing seedling. Inasmuch as the embryo (in its narrowest sense) becomes the rapidly developing part on germination, it is presumed that it contains a large number of enzymes and considerable amounts of protein appropriate to nuclear function. Most of the seed protein is contained in the cotyledons of those dicotyledonous seeds which lack endosperm, and in the endosperms of cereal seeds. The aleurone layer in the cereal seed, though rich in protein and almost free from starch, is too small in bulk to contain much protein, most of which is found in the starch cells. The protein in the cotyledons of dicotyledonous seeds is frequently seen in crystalline form in the aleurone grains and these usually occupy special cells, other cells being reserved for the storage of starch or fat. Perhaps the most impressive data on the compositions of bulk proteins from individual parts of seeds are those of Block and Bolling (1945) for the embryos of wheat ant1 maize. In comparison with the proteins of the entire seeds, the embryo proteins seem to be relatively rich in
246
J . W. H . LUGG
arginine and lysine. The nutritive values of the proteins of wheat and maize embryos have been discussed by Stare and Hegsted (1944). The proteins of bran and flour and other cereal products have also been investigated and tables of amino acid contents have been compiled by Block and Bolling (1945) and by Block (1946). 5. Pteridophytes Little study has been made of the proteins of pteridophytes. Mazur and Clarke (1938) worked with the fern Osmunda claytoniana; although it is not clear from their paper, it is probable that the material examined by them contained other tissue as well as frond. These workers extracted their material with ethanol; it is possible th at the residual crude protein preparation contained nonprotein nitrogenous substances in addition to protein. They then extracted most of the nitrogenous substances by heating the preparation with 90% formic acid and, after evaporating the formic acid from the extract, hydrolyzed with mineral acid. No mention was made of the amount of nitrogen found in the “humin,” but from consideration of their work on other plants it was probably very large. Consequently, since massive humin formation appears to be due to the presence of polysaccharide or polyuronide, the results obtaincd may have heen vitiated seriously. Lugg worked with the bracken fern Pteridium aquilinum and thc club moss Selaginella uncinata (see Lugg, 1913). Extracted protein preparations were made from pinnae of the former plant and wcre tested for representativcness. The methods of extraction were those (I,ugg, 1039a) which had been developed for use with leaves of the spermatophytes, and thc extracted protein seemed to behave physically in a manner entirely similar to that of spermatophyte leaf protein. For the latter plant, analytical work was confined t o a crude “whole” protein preparation made by extracting nonprotein nitrogenous and ccrtain other substances from the leaves and stalks. The tissue proteins could be rendered virtually insoluble in water by drying a t about 70°C. The results of these studics are shown in Table 111. 6. Bryoph ytes An estrcmely limitcd study has been made of the protein in thalli of the liverwort Lunularia cruciata (see Lugg, 1943) Thc protein is rendered virtually insoluble in water by drying the thalli at, about 70°C. and i t is probable that the general physical properties of thc (prcsumed) protein mixture are similar to those in the leaves of spermatophytes. Of the thallus protein nitrogen, Lugg found 1.3% occurring as cyst(e)-
247
PLANT PROTEINS
ine, 1.4% as methionine, 2.4 % as tyrosine, 0.0 % as diiodotyrosine, 1.7 % as tryptophan, and 5.5% as amide. TABLE I11 Amino Acid Compositions" of Bulk Proteins of Pteridophvte Tissues Amino acid Cyst (e)ine Methionine Tyrosine Tryptophan Arginine Lysine Amide
Pteridium aquilinum
Osmunda claytoniana
Selaginella uncinata
1.1-1.2 1.6-1.7 2.1-2.2 1.1-1.2 15.3 6.4 4.9-5.1
0.6 1.11
1.2 1.5 2.6 1.4
-
7.6 3.2
-
-
5.3 I_
As per cent protein nitrogen. Compiled from the data of Mazur and Clarke (1938), and of Lugg (1943). Diiodotyrosine could not be detected in the proteins of P. aquilinum and S . uncinata. 0
7 . l'hallophytes (Algae) A considerable body of data concerning the chemical compositions of marine and fresh water algae has been published by Mazur and Clarke (1938, 1942). As in their work on the pteridophyte Osmunda claytoniana, they prepared crude protein samples by extracting the tissues with ethanol. The preparations were then submitted to two types of hydrolytic treatment. For the estimations of tyrosint: and tryptophan they hydrolyzed with alkali and applied the method of Folin and Marenzi to the hydrolyzates. This method would probably yield highly unreliable values in such circumstances (see Lugg 1939b). For the estimations of other amino acids they extracted most of the nitrogenous substances from the preparations with hot 90% formic acid, removed the formic acid, and hydrolyzed the extracted material with mineral acid. Cystine was estimated by the labile sulfur method of Blumenthal and Clarke (1935), and methionine from the total organic sulfur minus labile sulfur. But the insoluble humin nitrogen values ranged from 5.0 to 20.1 % of the extracted nitrogen, and very large fractions of the cystine and methionine could have been associated with the insoluble and soluble humin (see Bailey, 1937s). Arginine, histidhe, and lysine were estimated by Block's (1934) method and by modifications, and here again the potentialities for interference by protein contaminants may be serious (see Tristram, 1939). Aspartic and glutamic acids were estimated by the method of Jones and Moeller (1928) and the values may be fairly reliable, but the method is not now
248
J. W. H. LUQQ
considered to be a good one. Sundry monoaminomonocarboxylic acids were estimated by Brazier’s (1930) method, but the authors considered the values to be of a low order of accuracy. Glycine, proline, and hydroxyproline were estimated by precipitation methods, but insufficient is known about the effects of protein contaminants upon the estimations of proline and hydroxyproline, in particular, for the reviewer to be able to assess the reliability of the values obtained. The qualitative findings of Mazur and Clarke may be summed up as follows: From their work (1938) on the fresh water alga Phormidium valderianum and the marine algae Chondrus crispans, Sargassum jluitans, S. natans, Laminaria sp., Ulva lactuca and Ulva sp., they concluded that, of the amino acids sought, the Ulva proteins lacked methionine, tyrosine, and lysine, Laminaria lacked methionine and lysine, and Sargassum and Chondrus lacked methionine. Phormidium, on the other hand, lacked cystine and lysine. From their later work (1942) on Fucus jurcatus, Cystoseira osmundaceae, Egregria menziesii, Macrocystis pyrifera, Lessoniopsis littoralis, Caulerpa racemosa, Codium fragile, Gloeotrichia echinuzata, and diatoms (presumably mixed), they had to conclude that there was less regularity among the results for the marine algae than the earlier work had suggested. A point of interest was their inability to find definite evidence for the presence of arginine in the proteins of the brown algae FUCUS,Cystoseira, and Egregria. Of the protein nitrogen in Chondrus, Sargassum, Laminaria, Ulva, and Phormidium, respectively, they found 2.1, 0.5, 2.7, 0.8, and 1.6% as glycine, 3.7, 4.8, 6.4, 6.5, and 5.2% as alanine, 2.8, 6.4, 5.1, 5.2, and 6.7% as valine, 5.3, 0.3, 2.5, 5.2, and 2.1% as leucine, 1.5, 0.3, 1.0, 2.3, and 1.1% as phenylalanine, 2.5, 5.9, 1.9, 4.1, and 0.9% as aspartic acid, 8.2, 2.9, 7.3, 7.6, and 4.4% as glutamic acid, 7.1, 7.3, 7.6, 7.0, and 7.0% as proline, and 1.0, 2.6, 1.8, 3.5, and 2.3% as hydroxyproline. Rather profound differences for some of the amino acids are revealed in this list. For the rest of the quantitative data the original papers must be consulted. Insofar as their analytical results may be valid, the findings of Mazur and Clarke in these two studies are of the utmost phytogenetic importance. Stokes and Gunness (1946) argued that the profound qualitative differences among the algae might well be repeated among the less highly developed thallophytes, but their work (see Sects. II,8 and II,9) failed to confirm this expectation. Furthermore, Lugg (1943) obtained values for Ulva Zactuca protein (5.6% of the protein nitrogen as amide, 1.3% as tyrosine, 0.0% as diiodotyrosine, 1.0% as tryptophan, 2.2% as cyst(e)ine, and 1.2% as methionine) which, for the most part, are not in harmony with those of Mazur and Clarke for this alga. For these reasons, and in light of the foregoing criticism of methods, the reviewer is disinclined to
PLANT PROTEIN8
249
accept many of the values reported by Mazur and Clarke together with the phylogenetic implications. In its intention and scope, however, their work remains a worthy model. It should stimulate the prosecution of new investigations of the compositions of algal proteins by other methods.
8. Thallophytes (Fungi) The proteins of yeasts have been under investigation for a considerable time because of the importance of these plants in bread making and in the fermentation industries. The growing importance of the penicillins in medical practice has called attention to the metabolism (see, for example, Wolf, 1948) of the molds which produce them, and thus to the mold proteins. In regard to time, this development has coincided, roughly, with the development of the microbiological methods for estimating amino acids ; and incidentally, the rise of the microbiological methods has called attention to the metabolism and to the proteins of the fungi and bacteria employed in these methods. But fermentation residues, apart from questions of direct academic interest, may be of economic importance as feedstuffs or as fertilizers. Especially when viewed as potential feedstuffs the questions of protein content and protein composition become important (see Tanner, Pfeiffer, and Van Lanen, 1945). The nutritive value of yeast protein has been discussed by Carter and Phillips (1944). There is direct experimental evidence (supported by the analytical data in Table JV) that the methionine content (brewers’ and ToruZu yeast) is rather low for satisfactory nutrition (Klose and Fevold, 1944). Csonka (1935) separated water-, saline-, and alkali-soluble protein fractions from ether-cytolyzed yeast cells, and succeeded thereby in extracting most of the protein. The protein extracted with the aid of alkali, however, must be presumed to have undergone some decomposition. In some instances partial amino acid analyses have been performed on untreated yeast cells. Stokes and Gunness (1946), in the most extensive study yet made of fungal protein composition, analyzed samples which almost certainly contained some nonprotein nitrogenous substances. They merely washed the materials with water and dried them at 105°C. before powdering them. The wash water was intended primarily to remove any of the adherent culture media, but it may have removed also some of the nonprotein nitrogenous substances from the cells and mycelia. Stokes and Gunness estimated amino acids in acid and alkali hydrolyzates of their preparations by microbiological methods. The nitrogen contents of the preparations ranged from 5.21 to 8.95%, values suggesting
250
J. W . H. LUGG
TABLE IV Amino Acid Compositions' of Bulk Proteins of Fun@ Penicillium Aspergillus Rhodotmula Saccharonotalumb nigerb rubrab myces spp. Valine Leucine Isoleucine Phenylalanine Methionine Tyrosine Tryptophan Arginine Histidine Lysine Threonine Glutamic acid 4
2.9 3.7 2.1 1.6 0.6
2.5 3.0 1.8 1.4 0.4
3.0 3.9 2.5 1.6 0.6
1.0 4.9 7.4 4.9 2.6
0.7 6.4 4.7 3.9 2.5
0.7 13.3 6.0 6.4 2.4
-
-
-
-
-
-
3.7b-4.0 4.56-4.9 3.P4.0 1.9-2.2 0.76-0.9 1.1-2.3
-
7.8-8.7 1.44.4 6.1-9.1 3.7 4.9-5.8
Rhizopus nigricanub
2.2 2.7 1.1 1.1
0.5
-
0.6 6.7 4.6 5.3 2.0
-
-
As per cent protein nitrogen. Compiled from the data of Stokes and Gunncss
(1946)and Block and Bolling (1945). b Values most likely to be affectedby preaenceof nonprotein nitrogenoussubstances.
the presence of large amounts of impurities which would probably include polysaccharide and/or poly uronide. Their methionine and arginine values, in particular, are consequently likely to be low, acid hydrolysis having been employed as a preliminary step in their estimation. Data for yeast, reported by Fink and Just (1942), are probably less reliable than those of the more recent studies, and have not been used in compiling Table IV. 9. Thallophytes (Ascomycetes and Bacteria) Leach (1905-1906) seems to have been the first to attempt the estimation of amino acids (free and combined) in bacteria. Tamura (1913, and subsequent papers) made a far more extensive study. From 1914 until 1922 little appears to have been done in this field. But from 1922 until 1935 there was a fairly steady outflow of work, with Mycobacterium tuberculosis under frequent investigation. To this period belong papers by Johnson and Brown (1922), Coghill (1926a,b), Hetler (1927), Eckstein and Soule (1931), and Greene (1935), among others. The methods of estimation used in these earlier studies would now, for the most part, be considered unsuitable, and the work is thus mainly of historical interest. More recent work is that of Burris (1942), Camien, Salle, and Dunn (1945), Stokes and Gunness (1946), and Freeland and Gale (1947).
PLANT PROTEINS
25 1
a. Problems in Interpretation o j Analytical Results. No attempt seems to have been made to separate reasonably pure and representative protein samples. In many of the studies no very serious attempt has been made to remove nonprotein nitrogenous substances; in some instances it has been the deliberate policy of the investigators to estimate the total amino acid (free and combined) contents of the organisms (e.g., Burris, 1942). For the most part, therefore, the results cannot readily be interpreted as amino acid analyses of the proteins. The authors themselves, however, have sometimes assumed that this interpretation is valid. Even when attempts have been made to extract the free amino acids it may be found that lipides (and hence probably lipide nitrogen) have been left as contaminants of the protein, A further potential hazard arises from the fact that chitin may form part of the cell walls of the bacteria, and a still graver hazard (so far as the interpretation of results is concerned) is contained in the possibility that the integuments are comprised wholly, or in part, of a substance akin to that obtained from the capsule of Bacillus anthracis (see Sect. 111,3). If this substance, a D( - )-glutamic acid polypeptide, contaminates the bacterial protein preparations we have to decide whether it should be regarded as part of the bacterial bulk protein or not. The decision will necessarily depend upon the object we have in view. These uncertainties must be kept in mind when appraising the significance of the values shown in Table V, which has been drawn up from data recently reported. Camien, Salle, and Dunn (1945)washed their cultures of organisms several times with normal saline at low temperature, then with 50 and 95% ethanol, and finally with ether. They remarked that the products appeared to consist of intact cells. For the estimation of tryptophan, alkali hydrolysis was used. Acid hydrolysis was used for the estimation of the other amino acids. Amino acid determinations were made by microbiological methods. There is a possibility that the products (the nitrogen contents ranged from 7.74 to 13.98%) contained some nonprotein nitrogenous substances. Stokes and Gunness (1946)simply washed their cultures with water, before drying and powdering them. The products almost certainly contained nonprotein nitrogenous substances. Their nitrogen contents ranged from 9.09% (Streptomyces griseus) to 13.19%. A microbiological method of assay was applied to hydrolyzates of the preparations. Freeland and Gale (1947)investigated the cell proteins indirectly, by analyzing ( a )hydrolyzates of the intact cells and ( b ) the fluid obtained by heating cell suspensions in water a t about 100°C. for 20 minutes. The fluid was believed to contain not only the amino acids adhering to the cell
ta ta
cn TABLE V Amino Acid Compositionsa of Bulk Proteins of an A d i m y c e t e and of Schiromyeetcs (Badcria)b ~~
Actinomycete
Gram-positive bacteria
Amino acid
Staph@CoCLwI
aureus
Valine Leucine hleucine Phenylalanine Methionine Tyrosine Tryptophan Arginine Histidine Lysine Threonine Glutamic acid a
4.6 4.5 1.9 1.5 0.6
3.5-5.1 3.4-5.0 3.3-4.7 1.4-2.2
-
0.6-0.8
1.0 10.3 2.5 4.6 3.1
0.34.5 6.2-9.7 2.6-4.1 5.7-9.2 2.4-3.6 5.Sd-6. gd
-
-
-
0.8 -
-
-
1.0-1.8
-
5.4 2.0 7.6
4.8-5.0 1,2-1.3 9.3-10.6
4.4
4.04.9
-
-
Gram-negative bacteria
Bacillus Escherichia subtilia
4.2" 5.lc 3.2c 1.9 1.05 1.5 0.5' 7.6 2.5 7.5 2.6c 5.3
Organism Bacillus no. 6578 brcm's
coli
4.6 5.0
4.2 1.7' 1.2" 1.4 0.6 10.4 3.3 7.4 2.8' 5.6
-
-
I
I
I
-
-
1.7 -
2.2 -
12.2 2.0 8.2
10.0 2.5 7.9
-
-
-
7.3 As per cent protein nitrogen. Compiled from the data of Camien, Salle, and Dunn (1945),Stokes and Gunness (1946),and 5.6
4.9
6.4
Freeland and Gale (1947). By the "ditferential oxidation" method, Johanson, Moir, and Underwood (1948)have obtained values (1.5to 1.6 and 1.2 to 1.5, respectively) for the cyst(e)ine and methionine contents of the bulk proteins of rumen bacteria. c Values most likely to be vitiated by the presence of nonprotein nitrogenous substances. Including D( -)- aa well as L( +)-glutamic acid.
a W I?
2 Q
PLANT PROTEINS
253
surfaces but also any amino acids existing in the free state within the original cells. The amino acid compositions of the cell proteins were calculated by difference. The methods of estimation employed by these workers were those using the specific decarboxylases (see Gale, 1945, 1946). b. Selection of Values. I n drawing up Table V the reviewer has excluded values reported by Stokes and Gunness when alternative values exist, because of the probability that their values are the least valid for the proteins. Such values of theirs as have been included have been followed by a superscript, and attention has been drawn to the fact that they may be vitiated more seriously than the other values. Because of the likely presence of polysaccharide and/or polyuronide in the materials, the tyrosine, methionine, and arginine values, in particular, are likely to be low, as they were determined after a preliminary acid hydrolysis. The glutamic acid values for the Lactobacillus species include the D( -) with the L ( + ) enantiomorph values. Camien, Salle, and Dunn found the D ( - ) : L ( + ) ratio in the various hydrolyzates to vary from about 1 :7 to about 1 :2. The higher ratios could not conceivably be accounted for by racemization of the L ( + ) enantiomorph during hydrolysis (see Chibnall, Rees, Williams, and Boyland, 1940) and the question arises whether a considerable amount of the D( -) form could exist as a residue in the normal “living” proteins of the cells. The presence of this enantiomorph in the proteins of certain tissues was claimed some years ago by Kogl and Erxleben (1939) but Chibnall, Rees, Williams, and Boyland refuted the claim. Assuming the validity of the values of Camien, Salle, and Dunn, the presence of so much of the D(-) form in hydrolyzates of bacterial protein preparations suggests that the bacterial integuments may have contained the D( - )-glutamic acid polypeptide referred to earlier in this section. Lactobacillus species are not regarded as capsulated organisms, but the possibility that a D( -)-glutamic acid polypeptide exists in the cell wall is not excluded. If a function of this material in the capsule of Bacillus anthracis is to protect the organisms from enzyme attack on the part of the host, then it could well serve the same function in the cell wall of Lactobacillus. It it is present, and if our attention is to be directed to the proteins involved in the living processes of the cells, then the glutamic acid values shown in Table V may, in general, need to be reduced considerably while the values as a whole, for all amino acids, may need to be increased slightly. c. Significance of Findings. The data of Stokes and Gunness indicate that the amino acid composition is of fairly high uniformity for a microorganism cultured under uniform conditions. Even when conditions
254
J . W. H. LUOO
are far from uniform, however, the composition seems to be almost unaffected (Camien, Salle, and Dunn; Freeland and Gale). Thus an early hypothesis, based upon relatively unreliable data, that bacterial bulk protein composition is virtually independent of the conditions of growth of the organisms, seems to have been substantiated fairly well. Stokes and Gunness as well as Camien, Salle, and Dunn considered that their data emphasized similarities in composition rather than differences. Freeland and Gale, however, drew attention to the fact that their data showed the proteins of the Gram-positive organisms (especially the cocci) to be of considerably lower arginine content than were those of the Gram-negative organisms. Their analyses were restricted to relatively few amino acids, but the more extensive compilation of values represented by Table V affords no further generalizations of similar nature. The extraordinarily high arginine and histidine contents of hydrolyzates of unwashed Azotobacter vinelandii cells, reported by Burris (1942), may have been due to the presence of large amounts of these amino acids in the free state. d . Brief Assessment of Present Position. Our knowledge of the amino acid compositions of the bulk proteins of ascomycetes and bacteria has been broadened endrmously in recent years, but cannot ;be regarded as extremely satisfactory. Apart from the fact that the presence of capsular material containing D( - )-glutamic acid residues introduces new problems in purification and/or interpretation, the courses followed by most investigators leave a strong probability that their materials were contaminated by nonprotein nitrogenous substances. Again, investigators have tended to pay too little attention to the influence of nonnitrogenous contaminants upon the analyses. The important conclusions ( a ) that the amino acid compositions of the proteins are virtually independent of the conditions of growth and ( b ) that the arginine contents of the proteins of Gram-positive organisms are lower than those of Gram-negative organisms, appear, however, to be justified.
10. Phytopathogenic Viruses Since the demonstration by Stanley (1935), Best (1936), and Bawden, Pirie, Bernal, and Fankuchen (1936) of the protein nature of tobacco mosaic virus, many other phytopathogenic viruses have been shown to be proteins. So far as is known, each virus is represented by a single, distinctive nucleoprotein which may or may not be associated with other substances. It seems proper to regard the phytopathogenic viruses (including the bacteriophages) as being composed (at least in part) of
PLANT PROTEIN 8
255
plant proteins though not necessarily of host plant proteins, because they normally occur only in plants; and in this review they are regarded, for convenience, as living organisms with a place at the lower extreme of the evolutionary series, and are treated as bulk proteins in spite of the fact that they seem to occur as individual substances. Existing evidence suggests that they are simple nucleoproteins, but Pirie (1946) has expressed the opinion that there is little to justify the assumption that these viruses are all of the same, relatively simple constitution. From the infectious juices of spermatophytes, the viruses may be Yeparated by differential precipitation a t appropriate p H values, by differential precipitation with salts or ethanol, and by differential sedimentation in centrifugal fields. By suitable purification procedures some have been obtained in crystalline or paracrystalline form. Among phytopathogenic viruses which have been obtained recently in crystalline or otherwise fairly pure form are Southern bean mosaic (Miller and Price, 1946), oat mosaic (Sukliov, Vovk, and Alekseeva, 1943), potato Y (Melchers, 1943), wheat mosaic (Sukhov, 1943), and tobacco necrosis (Bawden and Pirie, 1945). In general, they appear to be high molecular weight (of the order of millions) nucleoproteins, the nucleic acid being of the ribose type. Particle size, shape, and interaction vary greatly from virus t o virus. ‘These and other aspects of our knowledge of these substances have been reviewed in detail by Bawden (1943), Stanley (1943), Greenstein (1944), Pirie (1945, 1946), and by other active workers in this field. Many excellent electron micrographs of crystalline phytopathogenic viruses have recently been published by Price, Williams, and Wyckoff (1946). See also Crook and Sheffield (1946), and Sigurgeirsson and Stanley (1947). The recent work of Knight (1947) has contributed much to our knowledge of the amino acid compositions of phytopathogenic viruses. Previously, tobacco mosaic virus alone had been submitted to extensive analytical study. Table VI contains values for the amino acid contents of several phytopathogenic viruses. Tryptophan estimations made by applying color reactions to unhydrolyzed proteins and comparing the colorations yielded thereby with those yielded by free tryptophan or by “standard” proteins, are considered (see Best and Lugg, 1944) entirely’ unreliable. Such values have been excluded. In some instances the preparations analyzed were of unquoted nitrogen content, and it has been necessary to assume reasonable values (as indicated) in compiling the table. The phosphorus contents of these viruses fall within the range 0.52 t o 0.56%. The viruses represented here are customarily regarded as closely related. Even so, Holmed ribgrass virus alone appears to contain
256
J. W. H. LUGG
TABLE VI
-
-
--
.1 rnino Acid Cotnposilionaa of Some Phytopathogetiic Virusek
Amino acid
I
Tololmes Yellow Green iolme: tmco ribJI4L)l tuciiba iucuba naskec mosaic grass
-
Glycine Alanine Valine Leucine Isoleucine Phenylalanine Cyst (e)ine Methionine Tyrosine Tryptophan Arginine Histidine Lysine Aspartic acid Glutamic acid Proline Hydroxy proline hrine Threonine Amide
2.1 4.8 6.6 6.0 4.2 4.3 0.5 0.0 1.7 1.7 18.3 0.0 1.6 8.6 6.5 4.3 +? 5.8 6.4 8.0
2 .o 4.8 6.6 6 .O 3.7 4.3 0.4 0 1.7 1.7 20.5 0.0 1.7 8.8 6.5 4.2 5.7 6.5 -
--
2.1 4.8 6.3 5.9 3.7 4.2 0.4 0 1.7 1.7 20.4 0 .o
1.7 8.7 6.6 4.3 -
5.6 6.7 -
1.5 6.1 4.5 5.8 3.8 2.8 0.5 1.2 3.1 1.2 18.5 1.2 1.7 8.0 8.9 4.0
-
4.6 5.3 -
1.9 4.9 6.5 6.0 4.3 4.3 0.5 0 1.8 1.8 18.3 0.0 1.7 8.6 6.6 4.3 5.6 6.5 -
2.1 4.5 6.4 6.0 4.2 4.3 0.4 0 1.8 1.8 18.6 0.0 2.3 8.5 6.0 4.0
CU:uinbe 3
1.4 6.3 6.0 3.5 5.1 0 0 1 8 0.4
17.5 0.0 2.9 3.7 ~-
-
-
5.5 6.5 -
7. 5 4.5 -
1.7 5.8 6.4 6.0 3.0 5.0 0 0 1. Y
0.4 17.6 0 .o 2.8 8.3 3.7 4.2 7.5 4.5 -
-
As per cent of protein nitrogen. Compiled froni the data of Ross (1940, 1941, 1942), Knight (1942n,b, 1943, 1947), Knight and Stanley (1941), Heus, Sullivan, and Palmes (1941), Stanley (1943), Best and Lugg (1944), Stokes, Gunness, Dwyer, and Caswell (1945), and Caw and Stanley (1947). b For purposes of calculation, a value (16.6%) has been assumed for the nitrogen contents OF preparations when authors have failed to state them or otherwise provide information concerning them. a
histidine and methionine. The publication of values for the amino acid contents of viruses less closely related is awaited with interest. 11. Some Phylogenetic Considerations
Although, or perhaps because, Darwinian theory has, as Spengler (1918) pointed out, a distinctively Western metaphysical basis, it has remained an integral part of Western scientific doctrine. Accordingly, within the broad confines of this theory and its modifications, Western investigators have sought to base phylogenetic relationships among organisms upon chemical and other factors. Of particular interest here are the phylogenetic relationships within the plant world demonstrttt,ed
PLANT PROTEINS
257
serologically by Mez and others (for brief review and references see Gortner, 1938). Bailey (1937b) showed that the “myosins” (presumably actomyosins) of mammalian, avian, fish, and crustacean skeletal muscles are probably of very similar amino acid composition. Greenstein and Leuthardt (1944) showed that, while the compositions of the bulk proteins of homologous tissues of various animals are similar, there may be characteristic differences (as, for example, among the fractions of cyst(e)ine occurring as cysteine). Lugg (1938a,b; 1939a), Chibnall (1939), and Tristram (1939) drew attention to the similarities in compositions of the angiosperm leaf proteins. Mazur and Clarke (1938, 1942), in their studies of so-called “lower ’’ plants, claimed that highly distinctive composition differences existed among the bulk proteins. However their analytical methods (see Sects. II,5 and II,7) may have been at fault. Lugg (1943) showed that the compositions of the bulk proteins of the main somatic photosynthesizing tissues of spermatophytes, pteridophytes, and bryophytes are probably similar, and drew attention to the fact that the tissues of the bryophyte with which he worked belonged to the gametophyte generation of the plant. Unfortunately, there exists little acceptable information concerning the compositions of algal bulk proteins. Moving further down the accepted phylogenetic scale into the thallophyte division, we have now a considerable volume of information about the compositions of fungal and bacterial bulk proteins. Compositions and ranges of composition for various species in the main phytological divisions (the thallophyte division has been subdivided) are shown in Table VII. The questions whether the phytopathogenic viruses are living organisms, and, if SO, what place they should occupy in the phylogenetic scale, have been decided arbitrarily by assigning them a position at the lower end of the scale (i.e. right hand side of the Table). The data of Mazur and Clarke (1938, 1942) for a pteridophyte and numerous algae, have not been included. Proceeding from the spermatophytes down the scale it will be seen that the first appreciable departure in composition occurs with the alga, although it (Ulva lactuca) might well be held to lie in the main line of evolutionary development of the land plants (see, for example, Bower, 1929). The departure in coniposition may be associated with the fact that the tissues of this plant, unlike those of the plants above it in the scale, exist in a water instead of an air environment. The extents of the composition ranges among members of the individual subdivisions of the thallophyte division are so large as to mask such differences as may exist between the subdivisions themselves. The large ranges may be a result of deficiency of methods. But it is with the phytopathogenic
258
J. W. H. LUGQ
viruses that the marked (qualitative) composition differences are encountered. Even within the group itself there are interesting qualitative differences. The close similarities in composition of the bulk proteins in the photosynthesizing tissues of plants belonging to the first three divisions has evoked a somewhat vague speculation concerning protein function and TABLE VII Amino Acid Composiliona~ of Bulk Proteins i n Spermatophytes, Pteridophyles, a Bryophyle, an Alga, Fungi, an Actinmycele, Bacteria, and Phylopathogenic Viruses
-
Thallophytea Amino acid
Spermatophytesb
Pteridophyted Alga'
Fungi
ActinoBacterii mycete
Phytopathogenic Viruses
-- -Glycine Alanine Valine Leuaine Isoleucine Phenylalanine Cyst(e)ine Methionine Tyrosine Tryptophan Arginine Histidine Lyaine Aapartic acid Glutamic acid Proline Hydroxyproline Berine Threonine Amide
0.4 4.4-6.1 3.3-4.6
} 7.11;: 2.4-2.6 1.1-1.6 1 .2-1.6 2.3-2.7 1 .4-1 . 9 18.4-14 . O 3.6-4.0 6.0-8.8 4.7-6.4 6.4-7.8 3.1
-
3.0-4.0 4.7-6.0
1.1-1.2 1.6-1.7 2.1-2.6 1.l-1.4 15.3
6.4 -
4.9-5.3
-
2.2-4.0 2.7-4.9 1.1-4.0 1 .l-1.9
-
2.2 1.2 0.4-0.9 1 . 3 1.1-2.3 1 . 0 0.6-1.0 4.9-13. 1.4-7.4 3.9-9.1
-
-
6.6
2.0-3.7 4.9-6.8
-
-
-
1.0 10.3 2.6 4.6
3.66.1 3.4-6.1 3.3-4.7 1.4-2.2 1.61.6 0.6-1.6 0.8-2.2 0.34.6 4.8-12.: 1.2-4.4 6.7-10.1
-
4.0-7.3
4.6 4.5 1.9 1.5
-
0.6
-
-
3.1
-
2.4-3.6
- I -
1.4-2.1 4.66.1 4.6-6.6 6.8-6.0 3.0-4.3 2.8-6.1 04.5 0.0-1.2 1.7-3.1 0.4-1.8 17.6-20.6 0.0-1.2 1.6-2.9 8.0-8.8 3.7-8. Q 4.0-4.3 +? 4.6-7.6 4.66.7 8.0
Aa per cent protein nitrogen. Comp d from tables and data already given.
' Main photosyntbesiring tisaues.
composition in these tissues and in seeds (Lugg, 1943). There has been little speculation yet concerning the possible significance of composition similarities and differences in the proteins of plants in the thallophyte division, and of their relationships to the bulk proteins of plants in the other plant divisions. Stokes and Gunness (1946) were inclined to attribute such variation as might be encountered in bacterial bulk protein composition with variation in nature of the culture medium to changes in the amounts and natures of the cellular enzymes required to deal with the different nutrients, rather than to changes in the structural proteins. It is well known that adaptive enzyme activities (and, presumably, quan-
,
PLANT PROTEIN6
259
tities), in particular, can be modified greatly by the nature of the medium, and the view of Stokes and Gunness is in harmony with the belief (see, for example, Virtanen and De Ley, 1948) that nearly all the protein in the cells of micro-organisms is enzyme protein. The belief itself cannot be held to derive much support from the existing analytical data, because the observed variations in composition can readily be attributed to analytical deficiencies, but it may ultimately prove to be justified. However, the fraction of cell protein which can be regarded as enzyme (with a definition of “enzyme sufficiently narrow to exclude nuclear genes, cytogenes, plasmagenes, etc.) may not be overwhelmingly large. Incidentally, the term “structural,” as used by Stokes and Gunness and by others, is becoming highly ambiguous. In a review relating existing knowledge of the proteins with evolutionary theory, Synge (1945) has drawn attention to the fact that one of the characteristics which distinguishes the “lower’’ from the “higher l 1 organisms is their capacity to bring into peptide combination amino acids of the dextro type, as in the synthesis of gramicidin, tyrocidine, and the capsular substance of Bacillus anthracis. He is inclined t o regard this capacity as a survival of a chemically complicated stage, in the evolution from which the “higher” organisms have developed by a species of “rationalisation” in which (in particular) the amino acids going to form proteinaceous substances have become standardized as lev0 types and largely standardized with respect to variety. Discussing the implications of his recent analytical work on mutant strains of tobacco mosaic virus, Knight (1947) concludes that relatively small changes in amino acid composition may be associated with the development of strains which are highly lethal to the host plant.
111. “INDIVIDUAL” PROTEINS OF PLANTS Many of the proteins which, in a longer review, would be referred to in this section cannot be regarded as individual substances when judged by compliance with the requirements of various tests: crystallinity, electrophoretic homogeneity, definiteness of sedimentation and diffusion constants, and solubility independent of the amount of solid phase. Some have been found to pass one test and fail with another. (Incidentally, failure in the last-mentioned-“ phase rule ”-test may simply mean-cf. tobacco mosaic virus-that the degree of association of identical protein units has varied with the amount present.) Others have not been submitted to critical tests. Others again are known to consist of mixtures but are treated, by common usage, as if they were individual proteins.
260
J. W. R . LUQQ
1. Spermatophytes-Leaves of Angiosperms, Tubers
Brief reference has already been made (Sect. II,2,a) to the work of Frampton and Takahashi (1944, 1946). From the manner in which their extracts were prepared it is unlikely that they were dealing with other than the proteins in the fluid cytoplasm of the leaf cells of tobacco and pea bean plants. Their extracts from potato tubers may have contained protein fairly representative of the total protein present. In their electrophoretic studies they obtained evidence of the occurrence of three main protein components in the tobacco leaf extracts and of two in the extracts of leaves of the pea bean. Several components were present in the potato tuber extracts, but no attempt was made to identify any of them with tuberin. I n every case each of the main “components” could have been a mixture of closely associated proteins. It is extremely unlikely that the individual proteins of leaf cells are restricted to two or three in number, apart from the possibility that any single, simple protein is associated with different prosthetic groups to furnish a variety of conjugated proteins. Reference must be made here to an electrophoretic separation (Wildman and Bonner, 1946, 1947) of spinach leaf “ cytoplasmic” proteins into two fractions, one of which contained numerous enzymes and could be further fractionated. The other (large) main fraction was associated with auxin and with phosphatase activity. It appeared t o be homogeneous, with a molecular weight of 185,000. Little is known about the physical properties of the individual proteins, and until they can be isolated in quantity it is unlikely that anything will be learned about their compositions. As the leaf bulk proteins of different plant species appear to be of similar amino acid composition the question arises whether there is in the leaves of each species a similar assemblage of individual proteins differing considerably in composition among themselves, or a dissimilar assemblage of individual proteins each of similar amino acid composition, or similarity with regard to both assemblage and composition. The first and last possibilities seem to be discounted somewhat by the work of Frampton and Takahashi. Janisch (1942) and Kohler (1942) have reported the occurrence of protein crystals in apparently healthy potato tubers. 2. Spermatophytes-Seeds of Angiosperms
a. Preparation and Properties. From microscopic examination of certain seeds Hartig (1855) formed the opinion that much of the protein in them must be of crystalline structure, and three years later Masche (1858) artificially crystallized a seed protein. But Osborne’s work on the
PLANT PROTEINS
261
crystallization of seed proteins (see Osborne, 1890-1901) was more extensive than that of any other worker, before and since his time. Until recently the impression had persisted that the crystallized seed proteins were individual substances. This is now known to be unsound. Osborne depended largely upon the solubility relationships and upon estimation of the elementary compositions of his crystalline preparations as means of ch&racterizing them. Other workers have made what are believed to be identical preparations from the appropriate seeds and by the same techniques. Many of these preparations seem to be reasonably reproducible, even if they are not individual proteins. Much of the protein in dicotyledonous seeds is usually found to belong to the globulin class; the globulins are the proteins which Osborne succeeded so frequently in crystallizing. In general, the oil-bearing dicotyledonous seeds have most readily yielded crystalline globulins. They are fairly stable substances, usually undergoing only slight denaturation when dried a t fairly low temperatures. The cereal seeds are notable for their large contents of what were formerly regarded as highly distinctive and characterizable proteins, namely the gliadins and the glutelins. These proteins appear to be confined to the endosperm. However, if gliadins are defined as “proteins fairly soluble in slightly diluted ethanol,” they are not confined exclusively to cereal seeds nor even to the plant. world. Osborne’s papers abound with references t o the albumins and proteoses of seeds. These, especially the proteoses, have received little attention in comparison with the other types. Osborne frequently extracted defatted seed meal with water in the first instance, thereby removing proteose, albumin, and such globulin as might be dissolved by the small amounts of salts in the seed. This readily soluble globulin could then be precipitated by dialyzing the salt from the extract. Traces of acid in the seed might prevent the dissolution of protein ordinarily soluble in water (as Osborne believed to be the case with the seeds of many legumes). Following the extraction with water he extracted the material with dilute salt solutions to dissolve the globulins. Alternatively, he extracted the defatted seed meal directly with dilute salt solution and proceeded then to separate the constituent proteins in the extract. Of the protein not extracted by these operations some will consist of nucleoprotein, soluble a t mild alkalinity. The rest might consist, in part, of denatured protein and, in part, of native protein still retained in unruptured cells. Testas are particularly resistant to grinding operations but appear to contain very litttle of the total nitrogen of the seed. The alkalinity of solutions sometimes used by Osborne and by other workers to dissolve out appreciable quantities of the retained protein could
262
J. W. H. LUGG
scarcely be expected (see Blish and Sandstedt, 1929-1930; Neglia, Hess, and Sullivan, 1938) not to cause some decomposition. b. Individuality of Preparations. Questions concerning protein individuality and classification, as determined by solubility and precipitability, were raised sharply by Gortner, Hoffman, and Sinclair (1928). They found that the amounts of protein extracted from wheat flour by dilute solutions of potassium fluoride, chloride, bromide, and iodide a t equivalent concentrations varied enormously. They varied, too, with the concentration of the salt. The hazard of defining the globulin as “that fraction of the bulk protein which is soluble in dilute salt solutions” is obvious. They pointed out that, if the system of proteins present is viewed as a colloid system, and if solubility is taken to be synomymous with ‘(peptisation,” the results obtained were, perhaps, only to be expected. By the same token, the protein fractions obtained by following a prescribed technique need not necessarily represent definite chemical entities; they may be merely “peptized” fractions of the protein present. That it is virtually impossible to regard the so-called ‘(main globulin’’ (glycinin) of soybeans as a chemical entity is clear from extraction experiments described by Smith and Circle (1938), from others detailed by Vickery (1945), and from the work of Csonka and Jones (1933), who worked with several varieties of soybeans. Irving, Fontaine, and Warner (1945) and Fontaine, Irving, and Warner (1945) demonstrated electrophoretically the presence of a t least three well-defined components in major protein fractions obtained from peanuts. The arachin and conarachin of Johns and Jones (1916) were found each to contain two of these components and thus cannot be regarded as individual proteins. Much work has been done on the fractionation of wheat gluten. Sandstedt and Blish (1933) described the isolation of glutenin, mesonin, and gliadin fractions by thermal fractionation of warm alcoholic-acetic acid dispersions. Although Stockelback and Bailey (1938) were of the opinion that the glutenin and mesonin of Sandstedt and Blish were chemical entities (Sandstedt and Blish themselves were dubious), Harris and Johnson (1945a,b) were unable to agree that definite entities could be obtained by such a method. Spencer and McCalla (1938) concluded from their study of the fractional solubility of gluten in solutions of sodium salicylate that the protein consists of a complex that can be progressively fractionated. Kuhlmann (1937) considered gliadin to be an adsorption complex of proteins. Blish (1945) has reviewed evidence of the complexity of gluten fractions as revealed by osmotic pressure, sedimentation, and electrophoretic studies. Barmore (1947) has expressed the conviction that gliadin and glutenin are fractions of a complex system, differing systematically in physical and chemical properties but showing
PLANT PROTEIN0
263
no sharp distinction. Rich (1936), from a study of wheat flour proteins, went so far as to suggest that the protein as a whole should be regarded as a single substance which could be split into arbitrary fractions. It seems improbable that any gliadin (e.g., zein, hordein, wheat gliadin) or any glutelin (e.g., zeinin, hordenin, glutenin, avenin) preparation thus far obtained has been, at best, other than a mixture of closely related individual proteins. c. Toxic Proteins, T r y p s i n Inhibitors, and Enzymes. The earlier investigations concerning seed proteins toxic to animals have been reviewed by Osborne (1924). Legume seed proteins with toxic properties have been discussed in a recent review by Rotht5a and BruBre (1944). According to Kabat, Heidelberger, and Bezer (1947), both Kunitz and Cannan (personal communications) succeeded in crystallizing ricin, but they themselves contend that crude extracts of the castor bean contain crystallizable and noncrystallizable forms of ricin, which differ in toxicity but not immunochemically, electrophoretically, or ultracentrifugally. The molecular weight of ricin is probably about 80,000; the isoelectric point is a t about pH 5.4. The chief trypsin inhibitor (antienzyme) of the soybean has been crystallized by Kunitz (1946, 1947a) and shown to possess the properties of a globulin of molecular woight about 24,000 and with a n isoelectric point a t about pH 4.5. Kunitz (1947b) has also crystallized the inhibitor-trypsin compound. See Bowman (1948) for reference to sundry bean trypsin-inhibiting factors. A “chronology of enzymc crystallization,” commencing with Sumner’s (1926) classical work on urease, and proceeding to 1946, has been published by Sumner and Somers (1947). These authors have presented an excellent and precise account of enzymes of plant and animal origin. d . Compilations of Physical Properties. Molecular weight, sedimentation constant, diffusion constant, isoelectric point, pH stability range, and other physical data for seed proteins have been collected by Svedberg and Pedersen (1940), Cohn and Edsall (1943), and Schmidt (1945). It is noteworthy that the globulins amandin, excelsin, and edestin have molecular weights of the same order of magnitude (about 300,000). e. Composition, Heats of Combustion. Many of the earlier estimations of the amino acid contents of seed proteins have been tabulated by Mitchell and Hamilton (1929) and by Onslow (1931). Much of the later work has been collected and tabulated by Block and Bolling (1945) and by Block (1945). Among recently reported values are lysine contents (Horn, Jones, and Blum, 1917), histidine contents (Horn, Jones, and Blum, 1948), leucine, valine, phenylalanine and tryptophan contents (Smith, Greene, and Bartner, 1946), arginine, histidine, lysine, threonine,
264
J. W. H. LUGG
leucine, isoleucine, valine, tyrosine, tryptophan, phenylalanine, methianine, and cystine contents (Smith and Greene, 1917), “corrected” isoleucine contents (Smith and Greene, 1948),aspartic acid and glutamic acid contents (Hac, Snell, and Williams, 1945), and leucine and valine contents (Roche and Mourgue, 1943). The most determined effort to acquire knowledge of the complete amino acid composition of a seed protein (edestin) was made by Chibnall and colleagues (see Chibnall, 1946,who also quotes well-considered values for zein and wheat gliadin). While there is considerable variation in the compositions of the seed globulins, similarities are sometimes marked within a plant family (e.g., Cucurbitaceae-see Smith and Greene, 1947). I n general, the seed globulins appear to be of high arginine content. The gliadins, in general, are of high proline, glutamic acid, and amide contents; these characteristics have given rise to the alternative name “prolamins,” but the contents of other amino acids may differ markedly. A search of the literature up to 1948 reveals that little work has been done on heats of combustion of proteins apart from Benedict and Osborne’s (1907)study of the heats of combustion of about twenty seed proteins belonging variously to the albumin, globulin, gliadin, and glutelin classes. These authors found hordein to have the largest value (5908cal. per gram) and wheat albumin the lowest (5351 cal. per gram). Some intermediate values (in cal. per gram) were 5467 for conglutin, 5627 for edcstin, 5668 for legumelin, and 5718 for phaseolin. These are recalculation values based upon a newer value (6313.3cal. per gram) for benzoic acid as standard substance. Benedict and Osborne were able roughly to correlate higher values with higher carbon and/or lower oxygen contents of their highly purified preparations. Aberrations were attributed to possibly unusual amino acid compositions.
3. I’hallophytes (Fungiand Bacteria) Fractionation of yeast proteins has been described by Stern, Schein, and Wallerstein (1946). Kunitz and McDonald (19iG)have succeeded in crystallizing four proteins present in yeast, one of them being hexokinase (see also Berger, Slein, Colowick, and Cori, 1946).* * Several crystalline enzymes from yeast have also been prepared in Warburg’s laboratory in Berlin, for example, the original yellow ferment (Warburg, O., and Christian, W., 1933. Biochern. 2. 266, 377), acetaldehyde reductase (Negelein, E., and Wulff, H. J., 1937. Zbid. 293, 351), the “new yellow enzyme” (Hass, E., 1938. Zbid. 298, 378), the “oxidative enzyme of fermentation” (Warburg, O., and Christian, W., 1939. Zbid. 303,40), enolase (Warburg, O., and Christian, W., 1941. Zbid. 310, 384). Most of these papers have been published in a recent book by Warburg, Wase e rst o ~~b e rl ra g en dFermenle e (Verlag Dr. Werner Saenger, G.M.B.H., Berlin, 1948).-Edilor’s footnote.
PLANT PROTEINS
265
Certain individual (or apparently individual) proteins of acid-fast bacilli have been discussed by Seibert (1941), who, incidentally, succeeded in crystallizing tuberculin protein of high tuberculin activity (Seibert, 1926, 1928). The capsular substance of Bacillus anthracis appears to consist essentially of a D( -)-glutamic acid polypeptide (of molecular weight exceeding 50,000) in which some terminal carboxyl groups have entered into peptide bond formation (see Hanby and Rydon, 1946).
IV. MODESOF OCCURRENCE OF PROTEIN IN PLANTS It is usually tacitly assumed that a relationship exists between simple protein type and function. A clear-cut example in the animal world is seen in the obvious functions of the epidermal structures (wool, hair, nails, etc.) and the relatively weakly hydrophilic nature of the proteins which comprise the bulk of these structures. There is an increasing body of evidence, however, to support the view that in most plant and animal tissues the protein is, a t least to a large degree, conjugated with other substances. When the resulting complex proteins are of high stability and are easily purified it is sometimes a simple matter to relate function and type of complex (e.g., animal hemoglobin). Relationships between function and protein type in the plant world are still largely matters of speciilation, and will remain so until much more is known about the properties of the simple proteins, the extent to which they occur as such or as conjugates, the properties of the conjugates themselves, and the distributions of the proteins in the tissues and within the cells. The forms in which protein occurs (or is presumed to occur) in plants, and information about the distribution of it are discussed below. 1. Nuclei
Feulgen and Rossenbeck (1924), using the now well-known color reaction, demonstrated the presence of I ‘ thymo ”-nucleic acid (desoxyribonucleic acid) in the nuclei of cells of cereal embryos, yeast, and plant stems. They suggested that “yeast ” nucleic acid (ribonucleic acid) is located in the cytoplasm. Subsequent work (e.g., Feulgen, Behrens, and Mahdihassan, 1937; Delaporte, 1939) with plant cells has largely confirmed the earlier observations and conclusions. Belozersky (1939b) has drawn attention to the varying ratio of the two nucleic acids in plant cells. There is a growing acceptance of the belief that animal and plant cells contain desoxyribonucleic acid only in the nuclei and that the cytoplasm contains ribonucleic acid but not desoxyribonucleic acid. Insofar as plant cells may be held to resemble animal cells, the nucleoli (see Caspersson and Schultz, 1940) might be expected to contain ribonucleic acid.
266
J. W. H. LUQQ
There is evidence that the chromosomes, to which the desoxyribonucleic acid appears to be confined, also contain ribonucleic acid (see reviews by Greenstein, 1944; and Davidson, 1945). It is reasonable to suppose that the desoxyribonucleic acid of plant cell nuclei is at least largely combined with one or more proteins, and that any ribonucleic acid present is similarly combined. The presence of chromosomin (Stedman and Stedman, 1943; see also Stedman, 1945) in plant cell nuclei appears not yet to have been demonstrated.* 2. Mitochondria, Microsomes, ((Dissolved” Protein, Etc.
Little work has been done on the small inclusions of plant cell cytoplasm but there has been considerable study in recent years of corresponding bodies occurring in animal cells. They appear to consist essentially of protein, ribonucleic acid, and lipide, and it is highly probable that most, if not all, of the protein they contain exists in combination with the other constituents. This work, up to 1944, has been reviewed by Schmitt (1944). There is basis for the belief that some of them may be centers of enzyme activity, the enzymes being either integral parts of the bodies or adsorbed a t the surfaces. The evidence that mitochondria in the cells of plant and animal tissues are of similar structure and may have, a t least in some respects, similar functions to perform, has been reviewed by Newcomer (1940) (see also Guillermond, 1941). The microsomes may contain most of the protein in the “hyaloplasm” of liver cells. Existing evidence suggests that the particulate inclusions of protoplasm, whatever their size, are not free to be moved far at random by molecular impact (diffusion) forces. Observations with the aid of the ultramicroscope and polarized light support the view that protoplasm normally possesses a lattice structure, and it may well be that protein lying exterior to the inclusions is responsible, a t least in part, for the structure. The protein may be present in the form of protein-lipide conjugates of great complexity (see Schmitt, 1944). Needham (1942) and others have sought, in terms of “fibriaation” and in terms of the paracrystalline state of proteins and other large molecule species, to provide a tentative basis for morphogenetic movements. There appears to be no reliable, quantitative information, though, about the fraction * It should be noted that many papers in the literature refer to nucleic acids as ribo- or desoxyrihonucleic acids, although it has not been established, in most cases, that the sugar of the nucleic acid is actually ribose. Chargaff and Viacher (1948), in their recent review (Ann. Rev. Biochem. 17, 2011, prefer to designate these two main groups as pentose- and desoxypentosenucleic acids, respectively, except in cases in which the sugar component has been clearly proved to be ribose.-~dilors’ footnote.
PLANT PROTEINS
267
of the protein in any plant cell that lies exterior to the inclusions. Wildman and Bonner’s (1946, 1947) work (see Sect. II1,l) possibly suggests that it may be large in leaf cells. 3. Chloroplasls Menke (1938) and Chibnall (1939) have given values for the contents of protein, lipide, ash, and other constituents of disrupted chloroplasts. I n more recent work with chloroplast grana from the leaves of two plant species, Bot (1942) found the composition to vary somewhat with the season and the age of the plant. For neither plant were the protein :chlorophyll and lipide :chlorophyll ratios constant. The protein: chlorophyll ratios were estimated to have varied from about 14: 1 to 7: 1. Bot considered the chlorophyll to be present in two phases, (a) dispersed in the lipide, and ( b ) combined with the protein. Hanson (1941) (cf. Granick, 1938a,b), who took care to avoid disrupting the chloroplasts while isolating them from other cell constituents, believed the protein and chlorophyll to be combined in the ratio of about 5: 1 by weight. The work of Smith (1941a,b) and of Smith and Pickels (1941) on the effects of detergents has been of particular interest. There is no doubt that some, at least, of the chloroplast protein must exist in combination with chlorophyll, and the present evidence is that the magnesium of the chlorophyll is not involved in the bond with the protein, which appeared (the source was spinach) to have a molecular weight of a t least 265,000. This work has been reviewed by Smith (1942) and by Vickery (1945). Fishman and Moyer (1942) and Moyer and Fishman (1943) found the electrophoretic mobility-pH curves for the protein-chlorophyll complexes obtained from closely related species to be similar, but to differ markedly from that for the complex derived from an “unrelated” species. From the manner of preparation, however, it seems anything but reasonable to assume that simple protein-chlorophyll (a and b ) conjugates were under investigation. Wildman and Gordon (1942) found auxin to be associated with the protein in chloroplasts from spinach. There is evidence of the presence of nucleoprotein and of enzymes such as catalase and carbonic anhydrase in chloroplasts, but Vickery (1945) is of the o,pinion that most of the protein is probably combined with lipide and that the resulting lipoprotein is combined in turn with chlorophyll and, possibly, carotenoids. 4. Reserve Protein of Seeds Undoubtedly crystalline protein (globulin) can be seen in the cells of some oil-bearing seeds under the microscope, and protein crystals of
268
J. W. H. LUGG
similar appearance can be prepared from saline extracts. Hence these proteins, presumably, occur in the simple form in the cells. Amorphous material, which appears to be protein, can be seen under the microscope in the aleurone layer cells in certain seeds, and it is likely that such protein also occurs in the unconjugated state for the most part. As is wellknown, cereal seed glutens (gliadins plus glutelins, or gliadin-glutelin complexes) may be obtained from the flours by washing the starch grains out with water. There is little to suggest that these occur within the cells associated with prosthetic groups to any considerable extent. In this connection, however, it might be mentioned that Gordon (1946) found auxins to be associated with the endosperm proteins as well as with those of the embryos of wheat grains.
5 . Bacteria The presence of definite structures within the bacterial cell, long disputed, seems to be well-established now. Nuclei probably do exist as fairly discrete units, in some bacteria at least. I n others a diffuse type of nucleus may be present. It is possible that chromosomes occur free in the cytoplasm in some instances. There is reason to believe also that the cytoplasm in bacteria is not lacking organization and structure. These matters have recently been reviewed and excellent photographs have been provided by Robinow (1945). See also Pringsheim and Robinow (1947). Whether the chromatinlike materials contain chromosomin is not known. It seems likely that they consist, at least in part, of protein-desoxyribonucleic acid conjugates. Belozersky (1939a, 1940) reported the occurrence of very large quantities of nucleoprotein in many different species of bacteria. Petrik (1944) claimed that the nucleic acid in the nucleoproteins of certain bacilli is largely of the desoxyribose type. On the other hand, Boivin and Vendrely (1943) and Vendrely and Lehoult (1946) stated that desoxyribonucleic acid enters into the constitution of the smaller fraction of the total nucleoprotein of bacterial cells. There are probably great variations in both the absolute and relative amounts of the two types of nucleic acid in bacterial cells-Boivin and Vendrely found pronounced differences in the nucleoprotein contents of different strains of the same species. Chargaff and Zamenhof (1948), drawing attention to the relatively minute amounts of desoxyribonucleic acid they were able to isolate from yeast, suggest that all molecules of the substance may not have the same biological function in unicellular organisms. Stockinger, Ackerman, and Carpenter (1944) have described a nucleoprotein fraction associated with lipide, obtained from Neisseria gonorrheae. Reference has already been made to the occurrence of a high molecular
PLANT PROTEINS
269
weight polypeptide of D( - )-glutamic acid in Bacillus anthracis capsules (Sects. II,9 and 111,3).
METABOLISM IN PLANTS V. PROTEIN 1. Extent of the Field Viewed in its widest aspects protein metabolism in plants is found t o have an important place in almost every part of the nitrogen cycle in organic nature. Most of the nitrogen in living plant tissues is in the form of protein. Placing the protein a t the top of the cycle and nitrate at the bottom, we may imagine the exothermic, catabolic processes proceeding downward around one-half of the cycle and the endothermic, anabolic processes proceeding upward around the other half t o complete it. I n the first half of the cycle we have, in decaying plant matter, the liberation of amino acids and the production of ammonia therefrom by processes occurring within the cells and by the activity of micro-organisms such as Pseudomonas. The first decomposition products to accumulate are probably mostly those shown in equations ( I ) , (a), and (S),more particularly those shown in equation ( I ) :
An alternative path to the ammonia stage is via urea (readily hydrolyzed to ammonia and carbon dioxide) and other nitrogenous substances excreted from the animal body. In the soil, organisms such as Nitrosomonas and Nitrococcus oxidize the ammonia t o nitrite (equation 4 ) and the cycle is continued to the nitrate stage by Nitrobacter and other organisms (equation 6 ) : NH,
+30 --+
+O
HNOi
+ H20
HNOt ---t HNOa
In the second half of the cycle there are two important alternative routes. The first is through the green plant which can absorb nitrate and, by a succession of anabolic processes, incorporate the nitrogen into its tissue proteins. The second route is (with the aid of anaerobic organisms) probably via nitrite (equation 6 in reverse) and ammonia (equation 4 in reverse, or produced as in equation I ) by the reaction shown in equation (6) to atmospheric nitrogen, and thence to protein by the anaerobe Clostridium pasteurianum (butyricum), the aerobe Azotobacter,
270
J. W. H. LUGQ
the symbiotic root nodule bacteria (Rhizobium spp.) in association with the leguminous plants, and by sundry algae and fungi. HNOi
+ NHI -+
Ns
+ 2 HIO
(6)
It would be impossible in a review of this length to cover adequately all these aspects and others besides. It has been necessary rather to devote attention to the best developed and the most progressive fields. Reviews of fairly general nature are those of Chibnall (1939), Wood (1945), and Steward and Street (1947). 2. Certain Anabolic Aspects
It is convenient here to consider some aspects of protein synthesis as if they were truly separable from catabolic processes. It will be necessary subsequently to bring anabolic and catabolic processes together in a general consideration of protein metabolism in the plant. a. The Green Plant ASSIMILATION OF NITRATE. Astonishingly little is known about the early fate of nitrate nitrogen taken in by the plant root from the soil. Whether the roots or the leaves of a plant play the more important role in its assimilation seems to depend very much upon the plant concerned (for references see Street, Kenyon, and Watson, 1946). Leeper (1941), following Burstrom’s (1939) study of manganese function in plants, has shown that nitrate tends to accumulate in the leaves of some manganesedeficient gramineous plants, whereas it is fairly rapidly assimilated by normal plants if the leaves are in sunlight. It is concluded (Burstrom, 1943, 1946) that, while assimilation in leaves requires simultaneous photosynthesis, the respiration of preformed metabolites suffiles in roots. The first stage of assimilation is probably by reduction to nitrite (Eckerson, 1924), but our knowledge of the subsequent stages to ammonia (if this is the course followed) is largely speculative. Chibnall (1939) based his suggestion that these stages may be represented by: nitrite
-+
hyponitrous acid -+ hydroxylamine
-* ammonia
upon observations by Lemoigne, Monguillon, and Desveaux (1937) that extracts from leaves of the lilac will reduce nitrite to hydroxylamine, and by Corbet (1935) that ammonia may be oxidized t o hyponitrous acid by soil micro-organisms. The stages are so speculative, however, that older theories due to Baudisch (1911, 1913) and developed by Baly, Heilbron, and Hudson (1922)-that the nitrite reacts with an early product of carbohydrate synthesis (supposedly an activated form of formaldehyde) -cannot be ignored entirely. On the other hand, in inconclusive support
PLANT PROTEINS
27 1
of the view that production of ammonia is a necessary stage in protein synthesis from nitrate, are the observations of Warburg and Negelein (1920) that the alga ChZoreZZa in the dark can reduce nitrate to ammonia, of Eckerson (1924) that within the cells of tomato plants ammonia is produced from nitrate, and of Dittrich (1930-1931) that leaf and root saps can reduce nitrate to ammonia. It might be expected that in normally growing plants ammonia nitrogen taken in by the roots would be assimilated more rapidly than nitrate nitrogen, as the latter would need t o be reduced first (Tiedjens and Blake, 1932). Confirmation of this is found in the work of Sideris, Krauss, and Young (1937, 1938). I t is necessary a t this stage to refer to the work of Virtanen and Arhimo (1939), who found that the roots of growing pea and oat plants converted absorbed nitrate into nitrite and hydroxylamine, but were unable to detect hydroxylamine if the roots were absorbing ammonium sulfate instead of potassium nitrate-a finding which supports the chain of reactions favored by Chibnall. They were of the opinion, however, that the hydroxylamine is reduced to ammonia only if a-keto acids (oxalacetic and a-ketoglutaric) are virtually absent. In their presence the hydroxylamine was believed to form the corresponding oximes, which, by reduction, would yield aspartic and glutamic acids, respectively. In this scheme of processes, prior reduction of nitrate to ammonia is not always necessary for amino acid synthesis; but it may be the more general path. Chibnall (1939) suggested that reductive UTILIZATION OF AMMONIA. amination (von Euler, Adler, Gunther, and Das, 1938) and transamination (Braunstein and Kritzmann, 1937) reactions were probably involved in the synthesis of amino acids from ammonia in plant tissues. There are many indications that amino acids are indeed synthesized from ammonia in plant tissues, and fairly definite relationships appear to have been established. For example, Wood and Petrie (1938) have observed relationships between the ammonia and amino acid contents of leaves which are consistent with the hypothesis that the carbon chains (now believed to be a-keto acids) needed for synthesis constituted the limiting factor. There is evidence of the existence of many free amino acids in normal leaf tissue. Vickery (1925) found in the deproteinized sap of alfalfa leaves arginine, lysine, tyrosine, phenylalanine, swine, leucine, valine, and alanine. Large amounts of arginine, lysine, and aspartic acid were present as peptides or in other combinations. This does not constitute proof, however, that they represent stages in protein synthesis. Dent, Roepke, and Steward (1947) have demonstrated the presence in potato
272
J.
W.
H. LUQG
tubers of free glycine, alanine, valine, isoleucine, proline, a-amino-nbutyric acid, serine, threonine, arginine, histidine, lysine, aspartic acid, glutamic acid, asparagine, glutamine, phenylalanine, tyrosine, tryptophan, cystine, methionine, and methionine sulfoxide. More recently Dent (1948) has found in these tissues free leucine and (probably) both 8-alanine and 7-aminobutyric acid. Once again, irrespective of the sites of origin (leaf or root system) of these amino acids, their mere presence in the tubers is no proof that they represent stages in the synthesis of protein. From detailed consideration of the information yielded by the experiments of many investigators, Chibnall (1939) summarized the probable steps of protein synthesis from ammonia as follows. The ammonia has three routes of combination with carbon skeletons to yield amino acids (these in turn condensing to form protein) : (1) directly, via the appropriate a-keto acids (equation 1 in reverse) to yield the individual amino acids, ( 2 ) via oxalacetic acid to yield aspartic acid and asparagine, which, by ammonia donation, would regenerate the oxalacetic acid and provide for the synthesis of the individual amino acids from the appropriate a-keto acids, (3) via a-ketoglutaric acid as in route (a), glutamic acid and glutamine taking the places of aspartic acid and asparagine, respectively. The appropriate a-keto acids could arise from the metabolism of carbohydrates, fats, and proteins. Chibnall (1939) was unable to accept the conclusions drawn successively by Bjorksth (1930) Mothes (1933), and Schwab (1936) from infiltration studies, that wheat seedlings could synthesize protein from administered urea and other nitrogenous substances, that the organic acid parts of administered ammonium salts were metabolized in leaves to yield oxalacetic acid or fumaric acid, which then reacted with the liberated ammonia to give asparagine, and that amide formation from ammonia probably depended only upon the carbohydrate level and not upon the C4-dicarboxylic acids. From his own infiltration experiments, however, he concluded that, while administered ammonium glutarate, gluconate, and glutamate may have provided both the nitrogen and carbon chain for glutamine synthesis, it was far more likely that the chain already existed in the leaves in the form of a-ketoglutaric acid. The broad validity of Chibnall’s summary has not been subject to any serious refutation in work published subsequently. On the contrary, various experiments have provided degrees of confirmation. To give one recent example: Viets, Moxon, and Whitehead (1946) compared the rates of accumulation of various nitrogenous substances in the roots and tops of Zea mays (maize) plants supplied with ammonium salts in nutrient
PLANT PROTEINS
273
solution, their experiments indicating that the absorbed ammonia entered into the synthesis of asparagine, glutamine, and a-amino acids, largely in the roots. In particular respects, however, the summary may be a t fault. Rautanen (1946) (see also Virtanen and Laine, 1938a) studied the transamination reaction of Braunstein and Kritzmann (1937) in green plants. He used oxalacetic and a-ketoglutaric acids and the corresponding amino acids (aspartic and glutamic acids) as the dicarboxylic acids, and noted that, whereas certain aliphatic amino acids (notably alanine) might be synthesized from the corresponding a-keto acids, tyrosine and probably valine could not. He suggested that the synthesis of aromatic amino acids is probably by some other mechanism. Many amino acids may be regarded as alanine derivatives, and the formation of some of them might conceivably be by substitution in preformed alanine. Work with normal and mutant strains of fungi and bacteria suggests possible routes of synthesis of sundry amino acids in plant tissues, generally. Thus, for example, tryptophan may be synthesized from serine and indole (Tatum and Bonner, 1943), arginine from ornithine (Srb and Horowitz, 1944), and cysteine and methionine (Lampen, Roepke, and Jones, 1947) in the steps: sulfate .--t sulfite -+ sulfide .--t cysteine --t homocysteine + methionine
Again, it seems likely, as was suggested by Vickery, Pucher, Wakeman, and Leavenworth (1937)) that glutamine and asparagine play dissimilar roles. Steward and Street (1947) have reviewed the relevant evidence and point out that it is consistent with the hypothesis that glutamine is more intimately associated than asparagine with protein synthesis, while asparagine is more directly concerned with ammonia storage. The amides, however, must be deamidated before the amino groups themselves can enter into transamination reactions (Virtanen and Laine, 1941). The isolation by Green, Leloir, and Nocito (1945) from pig heart muscle of two specific transaminises, one of which catalyzes transamination between glutamic and pyruvic acids, and the other transamination between glutamic and oxalacetic acids, and the observation of O’Kane and Gunsalus (1947) that transamination between aspartic and pyruvic acids in certain animal tissues is mediated by glutamic acid are of special interest in these considerations of transamination in plants. Braunstein (1947) has discussed the mechanism of transamination reactions and has emphasized the importance of the dicarboxylic acids in the nitrogen metabolism of plants and animals. BRIEF SUMMARY OF PRESENT POSITION. The most usual route of synthesis of protein from ammonia in the green plant is probably via the
274
J . W. H. LUG0
amino acids. The evidence is not by any means conclusive but is strongly suggestive of this. The route of synthesis from nitrate taken in by the roots is probably via ammonia in most instances, but there is evidence that in some circumstances the nitrogen may become attached to carbon chains at the (postulated) hydroxylamine stage of its reduction to ammonia. b. Nitrogen Fixation. By reactions such as that shown in equation (6), whether they occur in plant tissues or in the soil solution, the soil tends to lose nitrogen to the atmosphere, but the tendency is normally countered by the capacity of certain organisms t o bring atmospheric nitrogen into compound formation. It is often deliberate agricultural policy to improve the nitrogenous manurial status of the soil by fostering the growth of such organisms and of associations of them. There is still much speculation about the steps involved in the process of the fixation of atmospheric nitrogen. Recent reviews are those of Burris and Wilson (1945) and Wilson and Burris (1947). Reference is made to them later. Allison, Hoover, and Morris (1937), Fritsch and De (1938), and Burris, Eppling, Wahling, and Wilson (1943) have shown that algae are capable of assimilating atmospheric nitrogen. The experiments of Fogg (1942) suggest that nitrogen fixation by these organisms may cease when they are supplied with nitrate or ammonia. Schanderl (1942) has reported nitrogen fixation by various fungi. Horner, Burk, Allison, and Sherman (1942) have shown that traces of molybdenum or vanadium can markedly increase nitrogen fixation by various strains of Azotobacter croococcum, A . vinelandii and A . agile. Although the hypotheses that the nitrogen is reduced either to hydroxylamine (Virtanen, ,1938, 1939) or ammonia are generally favored, the hypothesis that the fixation is oxidative rather than reductive is not without support. Burris and Wilson (1945) and Wilson and Burris (1947) have reviewed the evidence in favor of the different hypotheses. Burris and Wilson (1946) supplied actively growing cultures of A . uineZandii with I6N-enriched ammonium sulfate and discovered that, of the amino acids obtained by hydrolyzing the cultures, glutamic acid contained the greatest enrichment of 16N, just as Burris (1942) had found for the organisms utilizing 16N-enriched molecular nitrogen. It is pointed out by Wilson and Burris (1947) that nitrogen fixation ceases when the organisms are supplied with ammonium salts (cf. Fogg’s experiments with algae), and they consider this and other evidence to support very strongly the hypothesis that the fixation is reductive and via ammonia. If fixation by Azotobacter is unlikely t o be oxidative, it is perhaps still less likely (Blom, 1931) to be so for the anerobe Clostridium pasteurianum.
PLANT PROTEINS
275
The work of Virtanen and colleagues (e.g., Virtanen and Laine, 1937, 1938a,bi Virtanen, 1938, 1939) led them to favor the hypothesis that in the fixation of nitrogen by legume root nodule bacteria (Rhizobiumspp.) it is hydroxylamine, not ammonia, which enters into subsequent reactions. In their opinion hydroxylamine reacts with an a-ketodicarboxylic acid to yield the corresponding oxamino acid, which in turn is reduced to the corresponding dicarboxylic amino acid. Thence, by transamination reactions, various amino acids could be produced, and from these protein could be synthesized. Their earlier work indicated that aspartic acid (in particular), 8-alanine, and oximinosuccinic acid were the nitrogenous substances furnished to the legume by the nodule bacteria. Glutamic acid has been added to these recently (Virtanen, Linkola, Hakala, and Rautanen, 1946). Schanderl’s (1939) suggestion that symbiotic fixation of nitrogen takes place in the aerial rather than in the subterranean parts of plants has lacked support. Following the discovery of the widespread occurrence of hemoglobin in the functioning root nodules, Virtanen and Laine (1945) suggested that the hydroxylamine required in their formulation of nitrogen fixation could arise from a reaction: N2
+ 2 MetHb(Fe+++)F! 2 NH2OH + 2 Hb(Fe++)
where H b denotes hemoglobin and MetHb, methemoglobin. The validity of this reaction has been questioned by Wilson and Burris (1947) on the ground that the two products on the right hand side are both in states of greater reduction than the reactants on the left. Although Keilin and Smith (1947) believe that the hemoglobin is in some way intimately connected with the mechanism of the fixation, they have been unable to detect methemoglobin in the nodules and therefore reject the suggestion of Virtanen and Laine. In their review of biological nitrogen fixation Wilson and Burris (1947) favor ammonia as the most likely key intermediate compound of nitrogen. Their conclusion is based partly upon the experimental evidence furnished by studies with 16N and upon examination of the properties of a hydrogenase which might play a role in the fixation process. Although the evidence has been drawn largely from work with Azotobacter, they consider that the hydroxylamine hypothesis has little to commend it in any connection. Wood (1947), however, finds that some oximino acids (especially oximinosuccinic and a-oximinoglutaric) are assimilated by Azotobacter from very dilute solution and suggests that they are possible intermediates in nitrogen fixation. c. Sundry Plant Organisms. The volume of literature relevant to protein synthesis by fungi and bacteria is far too vast to be reviewed
276
J. W. H. LUGO
adequately here. Attention may be drawn, however, to the fact that these organisms differ enormously with regard t o the nitrogen sources upon which they can subsist and multiply (see Gale, 1947; Taylor, 1947). Lactobacillus spp. and Neurospora mutants show varying degrees of dependence upon preformed amino acids in their nutrient media, but an extreme of dependence seems to be reached by Leuconostoc mesenteroides (see Snell, 1945). Steward and Street (1947) have properly insisted that facts such as these do not constitute proof that the cell proteins are built up by simple condensation of the amino acids, however. The cells of many species of bacteria and fungi probably contain complex assemblages of free amino acids, though these may be present in undetectably small quantities, especially in Gram-negative organisms (see Freeland and Gale, 1947; Taylor, 1947). If it is reasonable to presume that one stage in the synthesis of protein in green plants is represented by the amino acids, it seems equally reasonable to suggest that this holds, too, for the other plants including the fungi and bacteria. d. Theories of Protein Synthesis. We may regard as the classical theory of biological synthesis of peptides and proteins that in which various amino acids are pictured as condensing together by loss of water between the a-amino group of one acid and the a-carboxyl group of the next. While the validity of the theory has been tacitly assumed in much of the foregoing discussion, proof is lacking and doubts have been expressed. Herbst and Shemin (1943) have suggested that peptide chains are formed by repeated amination and acylation of keto peptides, the amination being brought about by transamination reactions, imine formation followed by reduction, or oxime formation followed by reduction. Albaum and Cohen (1943) found that transawinase activity in germinating oat seedlings (homogenates of seedling tissue) increased more rapidly than did protein synthesis. They felt that this and other considerations were in keeping with the theory of Herbst and Shemin and with the more radical theory of Linderstrgm-Lang (1939), who suggested, merely by way of example, that peptide synthesis might occur by condensation of a-keto aldehydes with amino compounds followed by dehydrogenation and transamination. These alternative theories and others (such as that of Alcock, 1936) seem to have relatively little to support them (see, for example, Wood and Petrie, 1942). But energy must be supplied to effect the condensation of amino acids, and it is possible that they are first phosphorylated. This suggested modification of the classical theory is discussed briefly in Section V,5,c. Ideas of the mechanism by which specific proteins are elaborated are
PLANT PROTEINS
277
highly speculative. It has been suggested (see Bergmann and Fruton, 1941) that specificity of proteinase action might be the directing agency which determines the pattern of synthesized protein. There can be no provision in this hypothesis for prior phosphorylation of the amino acids. The “protein template’’ concept has been discussed by Petrie (1943) and will not be dealt with here. It has been invoked to account for virus (e.g., Stanley, 1938) and gene (e.g., Gulick, 1938) multiplication. For a discussion of template concepts in relation t o genic control of enzyme synthesis and gene-enzyme-substrate relationships, see contributions to a conference on “ Genic Action in Micro-organisms” (Conference, 1945) and an article by Spiegelman and Kamen (1946). I n passing it might be pointed out that in the light of Rothen’s (1946, 1947a) work i t is perhaps unnecessary to assume that gene replicas have to be formed within the nuclear membrane and migrate through i t to reach the cytoplasm. I n the opinion of Caspersson and colleagues (for brief review see Davidson, 1945), that part of the chromatin which is believed to contain ribonucleic acid (heterochromatin) controls the synthesis of histones which, diffusing from the nucleoli toward the nuclear membrane, induce the production of cytoplasmic ribonucleic acid and this, in turn, controls the synthesis of cytoplasmic protein.
3. Protein Metabolism in Seedlings A brief summary of the main conclusions and speculations to which the work on protein metabolism in seedlings has led is given in Section v,3,c. Most of the early work on protein metabolism in plants was done in connection with the broad problem of protein regeneration in seedlings. Chibnall (1939) has given an excellent review of this work, particularly from the historical and polemic angles and has linked it and the ideas which it engendered with the more recent work on protein metabolism in the plant generally. His book should be consulted for references to the relevant voluminous literature. It will be found that modern investigators turn again and again to the formulations of the earlier workers in seeking to understand various aspects of protein metabolism in seedlings and other plant organisms. It is accordingly desirable to outline the salient points of this earlier work very briefly. a. Earlier Work. Piria (1844) concluded that the large amounts of asparagine produced by the germination of vetch seeds arose from nitrogenous reserves in the seeds (in effect he suggested protein). Subsequently, seedlings of other members of the Papilionaceae were found to produce much asparagine. Boussingault (1868) drew attention to the fact that the asparagine in etiolated seedlings tended to disappear
278
J . W.
H. LUG0
rapidly when the plants were subsequently allowed to grow in the light. In Pfeffer’s (1872) view asparagine was a direct primary decomposition product of the protein in seeds of the Papilionaceae. He believed that the asparagine would diffuse from the cotyledons to the axial organs and there be regenerated into a different type of protein if suitable nitrogenpoor or nitrogen-free substances (e.g., carbohydrate) were available for the purpose. In the protein metabolism of other seedlings the role played by asparagine seemed to be far less important. Schulze (see Schulze and Umlauft, 1876) asserted that the presence of available carbohydrate did not necessarily lead to protein regeneration from asparagine. Certain experiments suggested that both protein and asparagine might be produced from other nitrogenous substances which were known to include amino acids, and it seemed that, if asparagine were utilized in protein regeneration, it was more probable that it was first broken down to ammonia and to acids like malic or succinic, as had been suggested by Mercadante (1875). Borodin (1876, 1878) was able to show that asparagine probably played a role in protein metabolism in different parts of many (if not all) plants. He believed that his own, Pfeffer’s, and Schulze’s findings could be reconciled by supposing ( a ) that very few nonnitrogenous substances (perhaps only glucose) could be utilized in protein regeneration from asparagine, ( b ) that even glucose could not be so used if it were required for respiration or other purposes, ( c ) that the breakdown of protein to give asparagine occurred in the living parts of all plants, and ( d ) that there was probably breakdown and regeneration of protein in all living plant organs in connection with protoplasmic activities such as respiration; the protein breakdown would produce asparagine, which would accumulate if carbohydrate were not available for prompt regeneration of protein. Borodin was thus formulating a general scheme for protein metabolism in the plant. His views markedly influenced the subsequent theorizing of others. Schulze so modified his opinions as to represent the stages of protein regeneration by: protein (in cotyledons) -+ amino acids --t nitrogen-rich residues (perhaps ammonia) -+ amides (asparagine and glutamine) -+ protein (in axial organs). I n this scheme the amides represent the first stage in regeneration, whereas formerly Schulze had held that the cotyledon protein amino acids were more suitable; he and colleagues, following upon the work of von Gorup-Besanez (1874), had detected the presence of many amino acids in seedlings. Meanwhile Prianishnikov obtained evidence which tended to support Schulze’s earlier views. The former concluded (1899a,b) that asparagine and glutamine, insofar as they arose from metabolism of amino acids which were not being utilized in protein synthesis, should be regarded as
PLANT PROTEINS
279
nitrogen storage substances utilizable under favorable circumstances. Boussingault (1868) had suggested that asparagine formation in the plant was probably a n ammonia detoxication process. This view, which had been unacceptable to some in the intervening years, was quite acceptable to Prianishnikov. b. Some Later Work. It was apparent to Schulze and others, from the admittedly poor analyses then available, th a t the dicarboxylic amino acids were present in far too small quantities in the seed proteins to provide directly for the total amide which might be formed during germination. Subsequent and more thorough analytical work has amply confirmed this view. A recent study of amide production in etiolated seedlings of Lupinus angustijolius, Vicia atropurpurea, and Cucurbita pep0 by Vickery and Pucher (1943) is of interest. Seedlings of C . pep0 differed from the others in th at they produced more glutamine than asparagine. Schwab (1936), who studied the production of both amides in seedlings of various plants and was unable to confirm Schulze’s belief that high production of glutamine was associated with fat metabolism (oil-bearing seeds), found about equal amounts of the two amides in seedlings of C. pepo. The possibility th at these two amides play different roles is discussed in Section V12,a. Yet, if amide formation in some instances arises largely from detoxication of ammonia, certain plants can nevertheless store considerable amounts of ammonia as ammonium salts. These plants might be classed as “ammonia plants,” whereas those which tend to accumulate amides might be classed as “amide plants” (see Ruhland and Wetzel, 1929, who found that the former type of plant was characterized by possessing a very acid sap), but the distinction is not sharp (Schwab, 1936). Prianishnikov (1922) maintained that seedlings which possessed considerable carbohydrate reserves would be expected to bring ammonia most readily into amide formation. Arenz (1942), however, considers that the ammonia detoxication powers of legume seedlings are greater than those of cereal seedlings, in spite of the large carbohydrate reserves of the latter. The level of oxalacetic and a-ketoglutaric acids, rather than the carbohydrate level, may be the important factor. Chibnall (1939) was strongly in favor of the views of Prianishnikov but felt that they could be harmonized with the later ones of Schulze if protein regeneration were thought of in terms of amino acid requirements as well a s in terms of amino acid availability. From comparisons of partial amino acid analyses, Lugg and Weller (1941) believed th a t the low methionine content of Lupinus angustifolius seed protein might be limiting protein regeneration in etiolated seedlings. They argued that if this were so the etiolated seedlings probably lacked certain synthetic mechan-
280
J. W. H. LUGG
isms, as the seed protein is relatively rich in cyst(e)ine. The tentative conclusion (supporting Prianishnikov’s views) was that amino acids arising directly from the breakdown of cotyledon protein are utilized in protein regeneration, and that the etiolated plant may not be able to synthesize necessary amino acids from asparagine (here present in abundance). But these seedlings must have been seriously lacking in available carbohydrate, and it is possible that tissue development and protein regeneration may have come to a halt for this reason. The work should be repeated with seedlings cultured on glucose solution instead of water. Using the nitroprusside reaction and iodine titration, Soutoulov (1946) has found that the very early stages of germination of Lupinus luteus seeds in the dark are characterized by the appearance of sulfhydryl groups, though peptide cleav.age appears to be extremely small. He suggests that protein breakdown begins with the reduction of disulfide bonds. However, the appearance of sulfhydryl groups could be a consequence not of reduction of disulfide bonds but of protein denaturation, and denaturation may be a necessary preliminary step (see LinderstrflmLang, Hotchkiss, and Johansen, 1938) to enzymatic proteolysis. I n a recent investigation Damodaran, Ramaswamy, Venkatesan, Mahadevan, and Ramdas (1946) studied the amide and amino acid changes in germinating seedlings of Dolichos biflorus (horse gram) , Phaseolus mungo (black gram) and Cicer arcetinum (Bengal gram). The seed proteins were rapidly “solubilized,” up to 90% of the total nitrogen being in soluble form by 24 hours. Of the amino acids, aspartic and glutrtmic acids appeared not to decrease in amount during the first few weeks of growth, whereas the others did. Asparagine and glutamine increased in quantity, and evidence was obtained pointing to the presence of a third amide. From the rapid disappearance of arginine the authors concluded that it might be of importance as an amide precursor. A little work has been done on protein metabolism in gymnosperm seedlings (see Klein and Taubock, 1932). Arginine appears to play an important role in nitrogen translocation. c. Brief Summary of Main Implications. When a seed is germinated the transfer of nitrogen to the growing parts involves the breakdown (at least in part) of the seed protein. This is clear not only from analytical evidence represented by ( a ) the differences in composition between seed and seedling protein and ( b ) the fact that asparagine and glutamine often appear in large quantities, but from cytological considerations, namely, the improbability that protein, as such, migrates across cell boundarics. There is evidence that the seed protein nitrogen may be translocated in the form of amino acids, and that breakdown of protein need not
PLANT PROTEINS
28 1
proceed beyond the amino acid stage except for the purpose of supplying the seedling with respirable material, but it is not conclusive. If breakdown does (for any reason) proceed beyond this stage, liberated ammonia is usually stored as amide (asparagine and/or glutamine) but may be stored, a t least to some extent, as ammonium salts. Whether amide formation should be regarded primarily as a detoxication requirement or as a necessary phase in protein regeneration is uncertain. The older view that asparagine and glutamine are of equal significance in ammonia storage is yielding to the belief that the former is more intimately associated with storage and the latter with protein synthesis. There is evidence that in some instances arginine may play important roles, ( a ) as a precursor of amide, and ( b ) as an actual vehicle for nitrogen translocation. 4. Protein Metabolism in Starving Leaves
From consideration of a large volume of work th a t has been done on the protein metabolism in air of starving leaves kept in darkness (e.g., Yemm, 1937; Vickery and Pucher, 1939; Vickery, Pucher, Wakeman, and Leavenworth, 1937, 1939; Wood, Cruickshank, and Kuchel, 1943; Wood and Cruickshank, 1944; Wood, Mercer, and Pedlow, 1944), Wood (1945) concluded that, except when the carbohydrate content is very high, the general course of events is the same in all cases. Protein hydrolysis commences promptly and continues smoothly until the chloroplasts begin to disintegrate and, as Michael (1935), Chibnall, and colleagues (see Chibnall, 1939), and Wood, Cruickshank, and Kuchel (1943) have shown, the proteins of the granular (mainly chloroplast) and fluid portions of the’ protoplasm are broken down a t much the same rate. During this time the amino acid and amide contents successively reach maximum values and then decline, and the ammonia content rises slowly from a low to a high value. These events, then, occur during the period of decline in the life of the plant tissue due to starvation, and possibly also in the period of decline with advancing age. Questions of the protein cycle (see Section V,5,a) apart, we are perhaps justified in regarding their general, if not precise, sequence as the reverse of that obtaining in the period of active growth and development. They give point to Prianishnikov’s dictum, recently re-emphasized (Prianishnikov, 1945), th at ammonia may be regarded as the alpha and omega of nitrogen metabolism. Vickery (1941) has given attention to other “end products” of nitrogen metabolism in plants. The respiration rate during starvation of the leaves falls appreciably from its initial value and later undergoes a “climacteric” increase, which
282
J. W. H. LUQQ
appears to be associated with the rapid catabolism of amino acids and with the onset of protoplasmic disorganization. Protein breakdown is greatly retarded, however, when leaves are starved in an atmosphere of nitrogen, though the carbohydrate loss may be greater than it is in air; and Wood and Cruickshank (1 944) have suggested that the protoplasm may be in a more static condition, the proteases thereby being prevented from coming readily into contact with the proteins. 5. Plant Protein Metabolism Generally-Regulation
Aspects of plant protein metabolism discussed in Section V,5 include the relationships between anabolic and catabolic processes, the nature of these processes, and their dependence upon environmental and other factors (e.g., amino acid, water, and carbohydrate contents of a tissue, respiration rate, etc.). a. Anabolism in Relation to Catabolism. During the life history of a plant tissue or entire organism phases of growth are normally distinguishable, although the dividing lines may be rather arbitrary. Periods such as growth, maturity, and senescence may be differentiated on the bases of various, and sometimes conflicting, criteria (e.g., changes in actual mass of living tissues, in the quantity of protein in it, in biological function, etc.). It is not the purpose here to discuss the adequacy of the criteria which might be adopted in each circumstance. But with regard to protein metabolism it is permissible to select a particular protein and define periods of over-all synthesis and over-all breakdown of it as those periods during which it is increasing and decreasing in quantity, respectively. The quantity may be in reference to a single cell, to a tissue or organ, or to the plant as a whole. The defined periods become physically meaningful provided they are long enough for the change, in whichever direction it may be, to be detected. I n an arbitrary way it is permissible similarly to define periods of over-all synthesis and over-all breakdown for the assemblage of proteins in the cell, tissue, or entire plant. These periods may or may not coincide with periods during which respirable reserves accumulate and decline. I n any event, although these reserves may be accumulating faster than they are being depleted, they are probably called upon to provide the energy for protein synthesis whether it be over-all synthesis or not (see below and Sect. V,7). Vickery, Pucher, Schoenheimer, and Rittenberg (1938, 1940), using IbN-enriched ammonium-salt nutrient solutions, showed that this nitrogen isotope was taken into the proteins of tobacco and buckwheat plants in
PLANT PROTEINS
283
excess of the proportional increase in over-all synthesis of protein for the period of treatment. Hevesy, LinderstrGm-Lang, Keston, and Olsen (1940) obtained similar results with the sunflower plant. The conclusion is obvious that the tissue proteins, once formed, are not static substances even during the stages of over-all synthesis of protein, that they are constantly being broken down and built up during these stages, and that in all probability the cyclic process is in operation during the stages of over-all breakdown too. It is nonetheless conceivable that over extremely short periods of time protein synthesis or breakdown, alone, may be proceeding in (say) a single cell. During each such period there would be no cyclic process, but an alternating succession of them could be held t o represent a protein cycle of a type rather different from that usually envisaged. The anabolic processes in this protein cycle (whatever conception of it may be valid in particular circumstances) may be merely the reverse of the catabolic ones. There are two ways, however, in which protein synthesis might not, for the most part, occur by direct recondensation of liberated amino acids, these being (a) synthesis by an alternative mechanism (e.g., prior phosphorylation-see Sect. V15,c), and ( b ) preliminary catabolism of the amino acids, followed either by resynthesis of them and their condensation by some mechanism, or by protein synthesis along another route altogether (see references to alternative theories of protein synthesis in Sect. V 1 2 , d ) . The second of these broad possibilities (b) is envisaged in the hypothesis of Gregory and Sen (1937). With reference to results obtained by various investigators including themselves, they argued that the law of mass action could be invoked to test whether protein synthesis and degradation were reverse processes, and concluded that they could not be so. They postulated that a cyclic process of protein synthesis and degradation is in operation the whole time, that the nitrogen-free amino acid residues are respired and that the synthesis of new amino acids requires the provision of new carbon chains from metabolized carbohydrate. The influence of the views of Borodin, Pfeffer, and Schulze is apparent in this hypothesis. Chibnall (1939) indicated that the hypothesis was in harmony with deductions he had drawn from studies of the metabolism of detached leaves. The tendency to accept the hypothesis in its entirety has been strengthened by the demonstration (with the aid of lSN) of the existence of a protein cycle of some sort. b. Hormonic Control. Mothes’ (1931) claim that at maturity leaves are unable to synthesize (ie., engage in over-all synthesis of) protein cannot be sustained. After considering the available evidence, Wood (1945) concluded that the mature leaves of some plants can engage in
284
J. W. H. LUGG
over-all synthesis of protein whether they are attached to, or detached from, the plant, and pointed out that protein synthesis in the latter instances cannot be determined by factors originating elsewhere in the plant. In this he opposed Chibnall’s (1939) suggestion that protein metabolism in leaves might be regulated by influences (possibly hormone) from the root system. But the capacity for protein synthesis appears to decrease with age of leaf (see Walkely and Petrie, 1941) and Wood mentions that the effect of age may be connected with changes in the enzyme systems involved. It is conceivable, however, that the enzyme systems themselves are under the control of hormones coming, at least in part, from distant tissue. Vickery, Pucher, Wakeman, and Leavenworth (1946) are convinced that controls do have their origins in tissues other than the leaves themselves. That such controls must operate (at least, during certain phases of the plant’s life history) seems evident from the well-known fact that leaf protein is called upon for the formation of seed. Tincker (1947) and Skoog (1947) have recently reviewed plant hormones. c. Protein Metabolism i n Relation to Carbohydrate, Respiration, and Other Factors. PROTEIN AND AMINO ACID (WITHAMIDE) LEVELS. A relationship between amino acid and protein levels was established by Petrie and Wood (1938a,b) for leaves of gramineous plants. Under equilibrium conditions, assuming mass action relationships to hold, the curve relating these variables should be convex to the amino acid concentration axis, provided the concentrations of all the amino acids vary in strict proportionality, but Petrie and Wood found the curve to be concave to this axis, and suggested that as the total amino acid concentration increased some amino acids may have increased disproportionately slowly, or even decreased, in concentration. Cystine was believed to be one such amino acid. A rather similar type of curve was found by these authors to express the relationship between protein concentration on the one hand and amino acid plus amide concentration on the other. More recent work by Phillis and Mason (1942) and by Mason and Phillis (1943) on leaves of the cotton plant tends to confirm these observations, insofar as “crystalloid” nitrogen might be regarded as representing amino acid plus amide nitrogen. Both groups of workers found that the maintenance of a given protein level required a higher level of the other variable with increasing age of leaf. Petrie and Wood (1938a,b), followPROTEIN AND WATER CONTENTS. ing observations by Mothes (1931), also discerned a relationship between water and protein contents in leaves of gramineous plants. While the amino acid and amide contents appeared to be independent of increase in the water content at any given ammonia content, the protein content
PLANT PROTEINS
285
increased. Petrie (1943) has discussed possible (but not entirely convincing) explanations of the phenomenon. PROTEIN AND CARBOHYDRATE CONTENTS. Dependence of protein synthesis upon carbohydrate level and carbohydrate metabolism was suggested by much of the early work on seedlings. Paech (1935) (see Pfeffer, 1872) expressed the relationship between protein, “active ” carbohydrate (monoses), and “active ” nitrogenous substances (e.g., ammonia) in plant cells in thc form of a law of mass action. According to this formulation, active carbohydrate and active nitrogenous substances could not coexist in high concentrations in the cell because there is a limit to the cell’s capacity for holding protein; neither could they coexist a t very low concentrations because there is a limiting, small amount of protein required to keep the cell alive. Although the formulation may seem to have applicability to many tissues (e.g., ripening gramineous seeds, starving leaves, etc.) it has received trenchant criticism on various grounds (e.g., Chibnall, 1939; Wood and Petrie, 1942). Chibnall (1939) believed that Paech’s ideas might prove more useful if the term “active carbohydrate” were replaced by “necessary series of a-keto acids.” He was of the opinion, furthermore, that the a-keto acids could arise in organic acid cycles, such as that of Krebs and Johnson (1937), connected with carbohydrate metabolism. While complete cycles of this type have not been demonstrated in plant tissues, the presumptive evidence that they do exist is very strong. Of spermatophyte organs, leaves and fruits frequently contain appreciable amounts of one or more of the acids involved in the modified tricarboxylic acid cycle (for a concise discussion of the modifications necessitated by the work of Wood, Werkman, Hemingway, and Nier, 1941, and others, see Baldwin, 1947), for example, malic, succinic, isocitric, aconitic, oxalacetic, and a-ketoglutaric acids. Citric acid, which has access t o the cycle by a side reaction and may be concerned in fat metabolism (see Breusch, 1943), is very commonly encountered in quantity, as is malic acid. It is conceivable that several organic acid cycles-may operate at the same time (one perhaps predominantly, depending upon the plant and the circumstances) in the cells of plants of all the phytological divisions. Such cycles, linked with the processes involved in nitrogen metabolism, could provide nitrogen-free substances and energy, required in protein synthesis. The reduction of nitrogen to ammonia involves a large increase in free energy. A still larger increase is involved in the reduction of nitrate to ammonia. On the other hand, relatively small increases in free energy are involved in the formation of amino acids from ammonia and a-keto acids, and in the formation of the peptide bond. It has been
286
J . W. H. LUQQ
suggested that peptide bond formation, which could occur in a single step, is coupled with energy-producing oxidations through energy-rich phosphorylated intermediates (see Lipmann, 1941 ; Cohen, 1945). Thus the amino or #carboxyl groups of the amino acids might be phosphorylated prior to peptide bond formation. Similar preliminary phosphorylations can be incorporated in alternative theories of protein synthesis referred to in Section V,2,d. PROTEIN METABOLISM IN RELATION TO RESPIRATION. Consideration of the problems apparent in the last three paragraphs poses the question of the relationships between plant respiration and protein metabolism. Attention has been drawn (e.g., Petrie, 1943) to the fact that, as the ratio of protein level to amino acid level is high in living tissue, the maintenance of any specified protein level in the living cell almost certainly requires an appropriate constant dissipation of energy for that purpose. One cannot speak of thermodynamic equilibrium in such a system, but the concept of the “steady state,” in which concentrations of reactants and products and the rate of dissipation of energy are all constant, is a useful one. Thus it is not only with regard to over-all synthesis of protein but with regard to the maintenance of the steady state (where this exists) that the protein metabolism must be linked directly or indirectly with the catabolism (respiration) of replenishable material (carbohydrate, fat, organic acids). If the supply of such materials fails, as it does in starving leaves (discussed earlier), respiration can continue only a t the expense of catabolizable protein. The respiration rates and protein contents of the leaves of gramineous plants a t various stages of growth were found by Petrie and Williams (1938) and Wood (1941) to be correlated positively, and it was suggested that this relationship might be expected if the protein content were assumed to provide a measure of the number of respiratory “seats,” other factors being maintained reasonably constant. Conditions under which the respiration rate appeared to be linked very closely with protein content were met by Richards (1938) in growing barley plants. In an entirely different type of material (the apple) Hulme and Smith (1939) found the respiration rate t o be approximately proportional to the protein level. Richards was led to suggest that a given rate of carbon dioxide production could maintain only a definite amount of protein, this amount probably being maximal in the normal plant growing upder conditions of balanced nutrition. Wood and Cruickshank (1944) have argued that the maintenance of a definite protein level depends upon whether syntheses (from carbohydrate and nitrogenous substances) of the most readily oxidized amino acids occur at rates at least as great as their rates of oxidation.
PLANT PROTEINS
287
Gregory and Sen (1937) found that in normal barley plants and also in nitrogen-deficient ones (in which the carbohydrate content was high) the respiration rate was correlated positively with both the amino acid and protein contents, whereas in potassium-deficient plants (in which the carbohydrate content was low) the respiration rate was positively correlated with both protein and carbohydrate contents. They argued that in the potassium-deficient leaves (the amino acid content was high) the protein cycle, as conceived by them, was in rapid operation and that the high carbon dioxide output was therefore to be expected, whereas in the nitrogen-deficient leaves (both protein and amino acid contents were low) the low carbon dioxide output was related to the low rate of operation of the protein cycle. Wood and Cruickshank (1944) have drawn attention to the fact that in these studies and in those of Richards (1938), who worked with phosphorus-deficient barley plants, positive correlations were obtained between respiration rate and amino acid content when the carbohydrate content was high. A further instance, perhaps, is to be found in the work of Brown (1946) on the relationship between respiration rate and amino acid level in barley seedlings. It may be hazardous, however, t o attempt to draw conclusions of general validity from experiments with mineral-deficient plants. According to Richards and Templeman (1936) the ratio of amide nitrogen to amino acid riitrogen increases in phosphorus-deficient plants. This might be attributed t o a decreased rate of synthesis of amino acids from amides or to a n increased rate of catabolism of amino acids. On the other hand, ,water-soluble nitrogenous substances, in general, tend to accumulate in phosphorus-deficient plants (e.g. , Williams, 1938) and this effect might be attributed to a decreased rate of synthesis of protein, particularly as it has been found that the rate of synthesis of protein in aerated potato tuber discs (Steward and Preston, 1941b) increases with addition of phosphate. But the high amino acid contents of potassium-deficient plants have been attributed by Richards and Templeman to a disorder of the protoplasmic organization rather than to an uncomplicated decrease in the rate of synthesis of protein, and it is possible that some degree of protoplasmic disorganization may accompany any gross mineral deficiency. Steward and Street (1947) have reviewed the work of Steward and Preston (1940, 1941a,b), Steward, Stout, and Preston (1940) , and Steward, Preston, and Ramamurti (1943) on respiration and protein synthesis by potato tuber discs in aerated solutions. They accept the Gregory and Sen hypothesis (Sect. V,5,a) and endeavor to account for the high rates of protein synthesis and respiration which accompanied vigorous aeration in halide and phosphate (but not carbonate) solutions
288
J . W. H. LUGG
in the following way: Vigorous aeration removes carbon dioxide and introduces a block in the carboxylic acid cycle at the point of carbon dioxide addition ; a-ketoglutaric acid (originating from glutamine and glutamic acid, these being replenishable by metabolism of stored amino acids) now enters the cycle, and the liberated ammonia reacts with suitable products of carbohydrate metabolism to form (ultimately) protein. If a-ketoglutaric acid is required to keep the cycle in operation and if the source of it is as suggested, the mechanisms for replenishing the glutamine (in particular) and glutamic acid must operate with speed and efficiency, for Dent and Steward (1948) have recently found that protein synthesis is accomplished with very small losses of glutamine, asparagine, and 7-aminobutyric acid from the discs, moderate losses of glutamic acid, aspartic acid, glycine, alanine and valine, but very large losses indeed of the other amino acids initially present (see Sect. V , 2 , a ) .
6 . Questioned Exclusive Validity of Conjlicting Theories Consider briefly the exothermic part of the nitrogen cycle (Sect. V , l ) . There is no reason to believe that the catabolic processes by which ammonia is produced (equations 1 , 2 , S ) from amino acids by Pseudomonas are essentially different from the processes by which it is produced in starving leaves. Again, we are not obliged to believe that the processes by which nitrite is produced from ammonia (equation 4 ) by Nitrococcus and Nitrosomonas, and nitrate from nitrite (equation 6) by Nilrobacter differ essentially from the processes by which nitrite and nitrate are produced in Azotobacter, for, while their presence in cultures of the latter organism has given support to the hypothesis that nitrogen fixation is oxidative rather than reductive, they could conceivably arise from oxidation of nitrogen first fixed as ammonia (equations 6 and 4, in reverse). In this connection starving, detached spermatophyte leaves have been found to increase their nitrate contents in some instances (Vickery, Pucher, Wakeman, and Leavenworth, 1937, 1939). See also Pearsall and Billimoria (1936, 1937) for evidence, furthermore, of nitrite production in, and actual loss of nitrogen from, starving leaves (equation 4 in reverse?, equation 6 1). The general picture which emerges from these and other considerations is that many unicellular and multicellular plant organisms may be capable of using, either by constitution or adaptation, the same or closely similar enzyme systems in processes connected with their protein metabolism. This also appears to be true for carbohydrate and fat metabolism. Indeed it has become apparent in recent years that the same, or very similar, enzymes are of widespread occurrence in plants and in animals. But with reference to plant protein metabolism, it seems likely that a
PLANT PROTEINS
289
process which may predominate in one plant tissue may be of negligible consequence in another wherein an alternative process predominates. Both, and even additional, alternative processes may be operating simultaneously in each case, but the organism may or may not exhibit great powers of adaptation with respect to making predominant and effective the process which best suits the circumstances in which it is placed. Thus conflicting theories, each derived from consideration of particular sets of data, are not necessarily mutually exclusive. For example, the course of protein synthesis from existing amino acids, as visualized by Prianishnikov (Sect. V,3,a), may be followed extensively in the regeneration of protein in seedlings and in the operation of the protein cycle in leaves respiring anaerobically (see remarks in Sect. V,7), whereas the course visualized by Gregory and Sen (Sect. V15,a) may be followed largely in mature leaves respiring aerobically and in aerated potato tuber discs. In this connection, work on the administration of 16N-enriched amino acids to animals (reviewed by Rittenberg and Shemin, 1946) suggests that protein synthesis (in animals) may occur without much preliminary catabolism of the amino acids, but it does not invalidate the view that the other course of synthesis may be followed to a minor degree. Gregory and Sen may have reached valid conclusions from questionable premises. Their use of the law of mass action to show that the paths of protein breakdown and synthesis must be dissimilar has been adversely criticized by Wood (1945) on the ground that the law should not be applied unless the contents of the various free amino acids are in a fixed proportionality, and Wood and Cruickshank have shown that the proportionality may be subject to marked variations. However, Petrie (1943) has pointed out that the cell protein level might well be independent of the ratios existing among the amounts of the free amino acids if the centers of protein synthesis and transamination are closely associated. But then again, as Paech (1935) realized, the classical mass action principle cannot be expected to apply rigorously to a complete, complex polyphase system which, in any event, never approaches thermodynamic equilibrium. This, of course, does not imply that the principle would necessarily lack rigorous applicability within one phase or part of a phase. We know little about the distribution of intracellular enzymes, but experimental evidence supports the view that a t least some of them occupy ordered positions. Oparin (1934, 1937) contended that the one enzyme might exercise a synthetic or lytic function according to whether it were adsorbed at a surface or free in the cytoplasm. With regard to reversibility of enzyme action a t the various centers, mechanisms may exist for the effective immobilization of substrates and products. It is well to remember, in this Connection, that substrates may be susceptible
290
J. W. H . LUQQ
to attack without coming into direct contact with enzymes (Rothen, 1946, 1947a,b) and that relatively slight changes in the properties of intervening membranes might convert them from being effective mediators of enzyme action into effective barriers. The implications of Rothen’s work, incidentally, are so far-reaching and important that experimental confirmation of his findings is urgently required. The call for elaborate mechanisms for separating substrates and products is replaced by a simple requirement that some, at least, of the relevant enzymes and/or template surfaces be localized, if we turn to the hypot,hesis that amino acids undergo phosphorylation prior to condensation. Acceptance of this hypothesis permits the visualizing of simple circumstances in which high levels of protein could coexist in ( I steady state ” relationships with low levels of phosphorylated amino acids locally, and still relatively low levels of free amino acids in the cell as a whole. Nonetheless, both mechanisms may be operative. Steward and Street (1947) suggest a template basis for much of the protein synthesis in growing cells, but argue that proteinase-catalyzed synthesis may be of importance in others.
7 . Coordination of Some Findings-Possible Significance o j the Protein Cycle I n an attempt to elucidate and coordinate some of the experimental findings already mentioned and to provide a non-“vitalistic ” basis for the protein cycle, the following speculative views are tentatively advanced. Native protein is probably required in the cell to provide enzymes and active surfaces and for the development and maintenance of protoplasmic structures and t.he biological “field” as a whole. In consequence of their chemical nature the proteins, whatever their functions may be, are subject to minimum, * fairly slow, but inevitable degradations, which we may refer to as denaturation, and the cell is equipped with proteases to break down the denatured protein. Some, if not all, of the heat of hydrolysis will be dissipated. Most of the breakdown products will be amino acids * It has been suggested to me by Dr. T. 8. Work that the rate of denaturation of an apoenzyme, or of an enzyme of the unconjugated, simple protein type, might be increased by participation in catalytic reactions. It might, in fact, be suggested that the protein of active surfaces, generally, may suffer an increase in rate of denaturation with increase in the degree of participation in reactions. It is of interest to note that Morel (1941) has estimated that an “accident” overtaking molecules of coenzymes I and I1 on an average of once in 240,000 catalytic reactions will account for the loss of dehydrogenase activity in certain micro-organisms. The experiments upon which this estimate was based are not such as will permit an estimate to be made of the rat@ of denaturation (and, presumably, regeneration) of the apoenzymea.
PLANT PROTEINS
29 1
suitable for the resynthesis of protein, probably, for the most part, by a different mechanism (prior phosphorylation?). I n any case energy is required for the synthesis and will be provided ultimately, in most circumstances, chiefly by the catabolism of available carbohydrate under both aerobic and anaerobic conditions. The immediate fates of the liberated amino acids, and the protein and amino acid levels that are set up will depend upon various factors, notably perhaps the degree of aeration. With vigorous aeration and rapid removal of carbon dioxide, the carboxylic acid cycle(s) can probably be kept in operation only with the aid of acids furnished by the deamination of amino acids in various ways (for example, in the modified tricarboxylic acid cycle, with the aid of acids such as a-ketoglutaric, oxalacetic, and fumaric), and under these extreme conditions it is possible that the general metabolic sequence, temporarily (at least) out of normal control, regulates the protein metabolism and may (as in the aerated potato tuber discs) lead indirectly to a marked increase in the ratio of protein nitrogen to amino-nitrogen. Under anaerobic conditions, at the other extreme, the energy required for resynthesis of protein can be derived from the catabolism of available carbohydrate but only through anaerobic stages, and provision of the deaminated residues of amino acids would not bring into operation organic acid cycles which have failed for lack of hydrogen acceptors. Even fortuitous oxidative deamination probably does not proceed, and there will be little (if any) tendency for the tissues to accumulate amides as ammonia storage substances. Most of the circumstances encountered would be intermediate between these extremes, varying amounts of the amino acids liberated by the hydrolysis of denatured protein being utilized directly for protein resynthesis or catabolized and resynthesized first. The general metabolism of mature cells which are not engaged in the synthesis of substances for use elsewhere might be regarded as serving, more or less directly and more or less efficiently, the purpose of regenerating native from denatured protein. It has been assumed that the proteins ultimately are responsible for protoplasmic organization, and that in a mature cell of the type considered above this might be maintained indefinitely (with a n implied exclusion of fortuitous zymolysis and no expenditure of energy) but for denaturation. But the manifestations of life are not associated with static situations, and even if it were merely in consequence of a partial carry-over” of systems required in the early growth of the cell, protoplasmic organization in the mature cell of the type considered clearly involves more than the maintenance of a structure (protein or otherwise) within it. The simplified picture presented, which is thus scarcely ((
292
J. W. H. LUG0
adequate even for this type of cell, necessarily requires modification if there is active growth and a concomitant demand for over-all synthesis of protein, or if the cell is engaged (as so many are) in the synthesis of materials for use elsewhere. It again needs to be modified if the usual reserves have become depleted and protein is required as a source of energy. However, within its broad framework many experimental findings can be fitted readily, e.g., the increase in respiration rate with rise in temperature, the correlation of respiration rate with protein content in some instances, with amino acid content in others, and with both in yet other instances, and the small over-all breakdown of protein in leaves starved in an atmosphere of nitrogen. The effects of phosphorus deficiency are most readily comprehended (see, for example, Steward and Preston, 1941) as arising from impairment of the rates of phosphorylations indirectly in carbohydrate metabolism, and perhaps more directly with regard to the actual mechanism of synthesis of protein from amino acids.
8. Spermatophyte Tissue Prolein-Virus Relationships Martin, Balls, and McKinney (1939) found that in the early stages of infection of tobacco plants with mosaic virus the virus multiplied a t the expense of the normal tissue proteins. Later, with the appearance of visible symptoms, there was a rapid increase in virus multiplication and in total protein production. They indicated two possible explanations, ( 1 ) that there could be direct conversion of normal protein into virus and (6) that there might be competition for available nitrogen between the synthetic processes leading to formation of normal proteins on the one h n d and virus on the other, accompanied by acceleration of nitrogen assimilation and protein synthesis. Earlier than this Stanley (1937) had come to the conclusion that, whereas tobacco mosaic and aucuba viruses stimulated protein metabolism in the host plant, other viruses brought about a decrease in protein synthesis. Vorob’eva (1939) found both stimulation and retardation of protein synthesis in tobacco plants infected by mosaic virus, the effect depending upon the type of plant. See also a review by Wynd (1943), dealing inter alia with protein metabolism in virus-infected plants and citing numerous instances of acceleration and retardation of protein synthesis associated with infection. That the general level of nitrogen nutrition may have an important bearing upon the host plant protein-virus relationships was shown by Spencer (1941). Infected tobacco plants in severe nitrogen deficiency were found to be incapable of breaking down the already formed mosaic virus for the synthesis of their own proteins. Similarly, the virus could not multiply at the expense of already formed tissue protein. These
PLANT PROTEINS
293
observations, together with the fact that the tissue bulk protein and virus are of widely different composition, tend to invalidate explanation ( 1 ) advanced by Martin, Balls, and McKinney. Consistent with the observations of Spencer are those of Brierley and Stuart (1946), who found a low expression of symptoms and of degree of infection when onion plants growing at a low level of nitrogen nutrition were inoculated with onion yellow dwarf virus. Whereas Frampton and Takahashi (1944) found three main protein components in clarified juice obtained from healthy leaves of the tobacco plant, they found four components in juice from mosaic disease-infected plants. The fourth band in their electrophoresis scanning patterns was attributed to the virus itself. They were of the opinion that the virus infection did not alter the physical properties of the original three main components or the relative amounts of them in the tissues. On the other hand Lugg and Best (1945) found the solubility of the leaf bulk proteins in borate buffer-ethanol-ether solutions to be depressed if tobacco plants had been infected with the disease, the effect being more marked the longer the period of infection. The change in physical properties appeared not to be associated with any marked change in amino acid composition. The virus and normal bulk protein differ greatly in composition, and the small differences found by Lugg and Best between the partial compositions of bulk proteins in infected and healthy leaves could be attributed to admixture of protein of composition similar to that of normal bulk protein with only a small amount of virus. This interpretation of.the analytical data was in harmony with the results of the older isolation and infectivity tests (e.g., Best, 1939). More recent work, however, suggests (see Bawden and Pirie, 1946) that the leaf residues may contain more virus than the expressed juice itself and that the virus content of infected leaves may be much,higher than formerly believed. Bawden and Pirie submitted the leaf residues variously to fine grinding, treatment with trypsin, and treatment with snail crop fluid, and separated with the aid of the ultracentrifuge material deemed to be virus, but its infectivity was much lower, weight for weight, than that of virus separated from the juice. Whether this material obtained from the leaf residues consisted entirely of virus of low infectivity or of physically modified leaf protein plus some virus is not absolutely determined. If the former is true, it must be admitted that in some instances the virus may represent as much as a third of the total protein in infected leaves. Bawden and Pirie’s (1946) suggestion that the virus infection probably depresses specifically the synthesis of components of the normal protein mixture that are richest in tryptophan is based upon a value for the tryptophan content of the virus (2.9%, on a nitrogen basis) which
294
J.
W. H. LUQQ
is not acceptable to the reviewer (see Table VI). Agatov (1941), incidentally, has reported slight differences between the amino acid compositions of the proteins in healthy and tobacco mosaic disease-infected tomatoes. I n their later publication, Frampton and Takahashi (1946), apparently unaware of the work of Lugg and Best, restated their earlier findings for the mosaic disease-infected tobacco plants. But they found that extracts from tobacco plants infected with potato X virus gave scanning patterns which indicated that a change had occurred in the ratios of the components that are encountered in healthy tissue. The indications, furthermore, were that a change in the physical properties of one of the components had resulted from infection of the tissues. They tentatively attributed the change in properties t o combination of the component with the virus. It is highly probable that, if they had investigated tobacco plants which had been infected with the mosaic virus for longer periods, they would have detected changes in the electrophoretic behavior of one or more of the original components. Whether the effect is a fairly general one with virus infection cannot be stated, but it is possible that by modifying the original proteins in such a way as to render them incapable of fulfilling their normal functions, the virus might cause injury to the host plant. Possibly, too, the stimulation of protein synthesis on the part of the host plant is a compensatory measure designed to maintain the normal (i.e., unmodified) proteins at or above the minimum level for adequate physiological activity, and can be expected to take place if the synthetic mechanism is not over-disturbed by the infection. I n this connection the observation of Rischkov and Gromyko (1941) (see also Rischkov, 1943) that in tomato plants infected with mosaic virus an inert, nonvirus protein can accumulate is highly pertinent. Woods and DuBuy (1941) found that hydrogen cyanide blocks a respiratory enzyme system the functioning of which appears to be necessary for mosaic virus multiplication and tissue protein synthesis. Their experiments with detached leaves (infected and healthy) of tobacco plants suggested that the virus multiplied at the expense of the proteinchlorophyll complex. They contended that the virus and tissue chromoprotein (ie., the protein-chlorophyll complex) are synthesized in closely parallel ways, and suggested that the virus may be an aberrant chondriosomal or chromoprotein derivative. Dufrdnoy (1944) has reviewed theories of such relationships. VI. CONCLUSION Among the points which have been brought out are: (1) that except for some rather wide departures among plants in the thallophyte division
PLANT PROTEINS
295
(possibly due, in part, to analytical deficiencies), the bulk proteins of the main tissues of plants (excluding phytopathogenic viruses) appear to be of somewhat similar amino acid composition, the similarities being pronounced among the spermatophytes but extending strongly into the pteridophyte and, perhaps, bryophyte divisions; (2) that the bulk proteins of seeds, in comparison with the uniformity of those of the main, somatic, photosynthesizing tissues of spermatophytes, are of distinctly dissimilar composition; (3) th at the individuality of “individual ” proteins of seeds is often in question, and that even when the “individual” proteins belong to the same class, as judged by physical properties, they may be of dissimilar composition with regard to some amino acids and yet reveal class similarities with regard to other amino acids; and (4) that in many instances such differences as may exist in the courses of protein metabolism in the cells and tissues of plants may be of degree (in the sense of predominance of one or more of a number of alternative routes and/or mechanisms) rather than of kind, exclusively. If point (4) has been unduly stressed in the text it is because the reviewer feels that i t tends to be overlooked. Of conflicting theories even one is not necessarily valid; neither does the “truth” necessarily lie in a compromise between them. Each, however, may be appropriate to the principal series of events occurring in particular instances, and in the individual instances themselves each theory may be appropriate to a particular series of events.
Acknowledgments Grateful thanks are due to Professor V. M. Trikojus for helpful consideration and valued criticism, to Professor J. S. Turner and Associate Professor E. McLennan for assistance with taxonomic and phylogenetic questions, and to Dr. Ernest Baldwin for a criticism of the views advanced in Section V,7. Special thanks are due to F. J. R. Hird and A. hl. Gallacher for assistance with literature surveys.
REFERENCES Agatov, P. (1941). Biokhimiya 6, 37. Albanese, A. A. (1944). Ind. Eng. Chem. Anal. Ed. 16, 609. Albaum, H. G., and Cohen, P. P. (1943). J . Biol. Chem. 149, 19. AIcock, R. S. (1936). Physiol. Revs. 16, 1. Allison, F. E., Hoover, S. R.,and Morris, H. J. (1937). Bolan. Gaz. 98, 433. Almquist, H. J., Mecchi, E., Krateer, F. H., and Grau, C. R. (1942). J . Nutrition 24, 385. Altman, R. F. A. (1939). Arch. Rubbercull. Neder1and.-Indits 23, 239. Arenz, B. (1942). Bodenkunde u . Pjlanzenerndhr. 27, 352. Bailey, K. (1937a). Biochem. J . 31, 1396. Bailey, K. (1937b). Biochem. J . 31, 1406.
296
J. W.
H. LUGG
Bailey, K. (1942). Biochem. J . 36, 140. Baldwin, E. (1947). Dynamic Aspects of Biochemistry, Cambridge Univ. Press, Cambridge. Baly, E. C. C., Heilbron, I. M., and Hudson, D. P. (1922). J . Chem. SOC.121,1078. Barmore, M . A. (1947). Cereal Chem. 24, 49. Baudisch, 0. (1911). Ber. 44, 1009. Baudisch, 0. (1913). 2. angew. Chem. 26, 012. Bawden, F. C. (1943). Plant viruses and Virus Diseases. Chronica Botanica, Waltham, Mass. Bawden, F. C., and Pirie, N. W. (1945). Brit. J . ezptl. Path. 26, 277. Bawden, F. C., and Pirie, N. W. (1946). Brit. J . ezptl. Path. 27, 81. Bawden, F. C., Pirie, N. W., Bernal, J. D., and Fankuchen, I. (1936). Nature 138, 1051.
Belozersky, A. N. (1939a). Microbiology U.S. S. R. 8, 504. Belozersky, A. N. (193913). Compl. rend. acad. sci. U. R. S. S. 26, 751. Belozersky, A. N. (1940). Microbiology U.S. S. R. 9, 107. Benedict, F. G., and Osborne, T. B. (1907). J . Biol. Chem. 3, 119. Berger, L., Slein, M. W., Colowick, S. P., and Cori, C. F. (1946). J . Gen. Physiol.
2Q, 141. Bergmann, M., and Fruton, J. S. (1941). Advances i n Enzymol. 1, 63. Best, R. J. (1936). Australian J . Ezptl. Biol. Med. Sci. 14, 1. Best, R. J. (1939). J . Australian Znst. Agr. Sci. 6, 162. Best, R. J., and Lugg, J. W. H. (1944). Australian J . Ezptl. Biol. Med. Sci. 99,247. BjBrkstBn, J. (1930). Biochem. 2. 226, 1. Blish, M. J. (1945). Advances in Protein Chem. 2, 337. Blish, M. J., and Sandstedt, R. M. (1929-1930). J . Biol. Chem. 86, 195. Block, R. J. (1934). J . biol. Chem. 106, 457. Block, R. J. (1945). Advances in Protein Chem. 2, 119. Block, R. J. (1946). Arch. Biochem. 11, 235. Block, R. J., and Bolling, D. (1945). Amino Acid Composition of Proteins and Foods, C. C. Thomas, Springfield, Ill. Blom, J. (1931). Zentr. Bakt. Parasitenk. Abt. ZZ, 84, 60. Blumenthal, D., and Clarke, H. T. (1935). J . Biol. Chem. 110, 343. Boivin, A., and Vendrely, R. (1943). Compt. rend. SOC. biol. 137, 432. Bondy, C., and Freundlich, H. (1938). Compt. rend. trau. Lab. Carlsberg.S8r. chim. 22, 89.
Borodin, J. (1876). Botanische Jahresberichte 4, 919. Borodin, J. (1878). Botanisches Zeitung 36, 802 Bot, G. M. (1942). Chronica Botan. 7, 66. Bouseingault, J. B. (1868). Agron. chem. agr. el physiol. 4, 245. Bower, F. 0. (1929). Huxley Memorial Lecture. Macmillan, London. Bowman, D. E. (1948). Arch. Biochem. 16, 109. Braunstein, A. E. (1947). Advances i n Protein Chem. 3, 1. Braunstein, A. E., and Kritzmann, M. G. (1937). Enzymologia 2, 120. Brazier, M. A. B. (1930). Biochem. J . 24, 1188. Breusch, F. L. (1943). Science 97, 490. Brierley, P., and Stuart, N. (1946). Phytopathology 36, 297. Brown, R. (1946). Ann. Botany 10, 73. Burris, R. H. (1942). J . Biol. Chem. 143, 509. Burris, R. H., Eppling, F. J., Wahlin, H. B., and Wilson, P. W. (1943). J . Biol. Chem. 148, 349.
PLANT PROTEINS
297
Burris, R. H., and Wilson, P. W. (1945). Ann. Rev. Biochem. 14, 685. Burris, R. H., and Wilson, P. W. (1946). J . Biol. Chem. 166, 595. Burstr6m, H. (1939). Planta 29, 292. Buratrom, H. (1943). Arkiu Botan. 30B,No.8,7. Burstr6m, H. (1946). Kgl. Lantbruks-HBgskol. Ann. 13, 1. Camien, M. N., Salle, A. J., and Dunn, M. S. (1945). Arch. Biochem. 8, 67. Carter, H. E., and Phillips, G. E. (1944). Federation Proc. 3, 123. Caspersson, T., and Schultr, J. (1940). Proc. Natl. Acad. Sn'. U . S.26, 507. Chargaff, E., and Zamenhof, S. (1948). J . Biol. Chem. 173, 327. Chibnall, A. C. (1923). J. biol. Chem. 66, 333. Chibnall, A. C. (1939). Protein Metabolism in the Plant. Yale Univ. Press, New Haven, Conn. Chibnall, A. C. (1946). J . Intern. SOC.Leather Trades' Chemists 30, 1. Chibnall, A. C., Miller, E. J., Hall, D. H., and Westall, R. G. (1933). Biochem. J . 27, 1879. Chibnall, A. C., Rees, M. W., Williams, E. F., and Boyland, E. (1940). Biochem. J . 34,285. Chibnall, A. C., and Schryver, S. B. (1920). J . Physiol. 64, Proc. July. Chibnall, A. C., and Schryver, S. B. (1921). Biochem. J . 16, 60. Chick, H., and Cutting, M. E. M. (1943). Lancet 246, 667. Coghill, R. D. (1926a). J . Biol. Chem. 70, 439. Coghill, R. D. (1926b). J. Bio2. C h m . 70, 449. Cohen, P.P. (1945). Ann. Reu. Biochem. 14, 357. Cohn, E. J., and Edsall, J. T. (1943). Proteins, Amino Acids and Peptides. Reinhold, New York. Conference (1945). Ann. Missouri Botan. Garden, 32, 107. Corbet, A. S. (1935). Biochem. J . 29, 1086. Crook, E. M. (1946). Biochem. J . 40, 197. Crook, E. M.,and Shefield, F. M. L. (1946). Brit. J . Ezptl. Path. 27, 328. Csonka, F. A. (1935). J. Biol. Chem. 109, 703. Csonka, F. A. (1939). J . Agr. Research 69,765. Csonka, F. A.,and Jones, D. B. (1933). J . Agr. Research 46, 51. Damodaran, M., Ramaswamy, R., Venkatesan, T. R., Mahadevan, S., and Ramdas, K. (1946). Proc. Indian Acad. Sci. 23B, 86. Davidson, J. N. (1945). Biochem. J. 39, lix. Davies, W. L. (1926). J. Agr. Sci. 16, 293. Davies, W. L. (1927a). J. Agr. Sci. 17, 33. Davies, W. L. (192713). J. Agr. Sci. 17, 41. Delaporte, B. (1939). Rev. gbn bolan. 61, 449. Dent, C. E. (1948). Personal communication. Dent, C. E., Stepka, W., and Steward, F. C. (1947). Nature 160, 682. Dent, C. E., and Steward, F. C. (1948). Personal communication from Dr. Dent. Dittrich, W. (1930-31). Planla 12, 69. Doty, D.M. (1941). Znd. Eng. Chem. Anal. Ed. 13, 169. Dufrhoy, J. (1944). Rev. Applied Mycol. 23, 472. Eckerson, S. H. (1924). Botan. Guz. 77, 377. Eckstein, H.C.,and Soule, M. H. (1931). J . Biol. Chem. 91,395. Euler, H. von, Adler, E., Gunther, G., and Das, N. B. (1938). 2. physiol. Chem. 264, 61. Everson, G. J., Steenbock, H., Cederquist, D. C., and Parsons, H. T. (1944). J . Nutrition 27, 225.
298
J . W. H. LUQG
Feulgen, R., Behrens, M., and Mahdihassan, 9. (1937). 2.physiol. Chem.246, 203. Feulgen, R., and Rossenbeck, H. (1924). 2. physiol. Chem. 136, 203. Fink, H., and Just, F. (1942). Biochem. 2. 312, 390. Fishmann, M. M., and Moyer, L. S. (1942). Science 96, 128. Fogg, G. E. (1942). J . Ezptl. Biol:19, 78. Fontaine, T. D., Irving, G . W., and Warner, R. C. (1945). Arch. Biochem. 8, 239. Fourcroy, A. F. (1789). Ann. chim. 3, 252. Frampton, V. L., and Takahashi, W. N. (1944). Arch. Biochem. 4, 249. Frampton, V. L., and Takahashi, W. N. (1946). Phytopathology 36, 129. Freeland, J. C., and Gale, E. F. (1947). Biochem. J . 41, 135. Fritsch, F. E., and Do, P. K. (1938). Nature 142, 878. Gale, E. F. (1945). Biochem. J. 39, 46. Gale, E. F. (1946). Nature 167, 265. Gale, E. F. (1947). J. Cen. Microbiol. 1, 53. Gaw, H. Z., and Stanloy, W. M. (1947). J . Biol. Chem. 167, 765. Gordon, S. A. (1946). Am. J . Botany 33, 160. Gortner, R. A. (1938). Outlines of Biochemistry, 2nd ed. Wiley, New York. Gortner, R. A., Hoffman, W. F., and Sinclair, W. B. (1928). Colloid Symposium Monograph 5, 179. Gorup-Besanez, F. von. (1874). Ber. 7, 146. Granick, S. (1938a). Am. J . Bohny 26, 558. Granick, S. (1938b). Am. J . Botany 26, 561. Green, D. E., Leloir, L. F., and Nocito, V. (1945). J . Biol. Chem. 161, 559. Greene, R. A. (1935). Soil. Sci. 39, 327. Greene, R. D., and Black, A. (1944). J. Biot. Chem. 166, 1. Greenstein, J. P. (1944). Advances i n Profein Chem. 1, 209. Greenstein, J. P., and Leuthardt, F. M. (1944). J . Biol. Chem. 166, 349. Gregory, F. G., and Sen, P. K. (1937). Ann. Botany N.S. 1, 521. Groot, E. H. (1042). 'Arch. nberland. physiol. 26, 472. Guillermond, A. (1941). The Cytoplasm of the Plant Cell. Chronica Botanica, Waltham, Mass. Gulick, A. (1938). Quart. Rev. Biol. 13, 140. Haag, J. R. (1931). J. Nutrition 4, 363. Hac, L. R., Snell, E. E., and Williams, R. J. (1945). J . Biol. Chem. 169, 273. Hamilton, T. S., Nevens, W. B., and Grindley, H. S. (1921). J . Biol. Chem. 48, 249. Hanby, W. E., and Rydon, H. N. (1946). Biochem. J . 40, 297. Hanson, E. A. (1941). A u s l r a h n J . Ezptl. Biol. Med. Sci. 19, 157. Harris, R. H., and Johnson, M. (1945a). Arch. Biochem. 6, 231. Harris, R. H., and Johnson, M. (1945b). Arch. Biochem. 8, 135. Hartig, T. (1855). Bohnisches Zeitung 13, 881. Hayward, J. W., and Hafner, F. H. (1941). Poultry Sci. 20, 139. Herbat, R. M., and Shemin, D. (1943). J . Biol. Chem. 147, 541. Hess, W. C., Sullivan, M. X., and Palmes, E. D. (1941). Proc. SOC.Ezptl. Biol. Med. 48, 353. Hetler, D. M. (1927). J . Biol. Chem. 72, 573. Hevesy, G., Linderstrgm-Lang, K., Keston, A. S., and Olsen, C. (1940). Compt. rend. Irav. lab. Carlsberg. Sdr. chim. 23, 213. Horn, M. J., Jones, D. B., and Blum, A. E. (1947). J . Biol. Chem. 169, 71. Horn, M. J., Jonee, D. B., and Blum, A. E. (1948). J . Biol. Chem. 173, 149. IIorner, C. K., Burk, D., Allison, F. E., and Sherman, M. S. (1942). J . Agr. Research 66, 173.
PLANT PROTEINS
299
Hulme, A. C. (1946a). Nature 168, 58. Hulme, A. C. (1946b). Nature 168, 588. Hulme, A. C., and Smith, W. H. (1939). Dept. 815.Znd. Research Brit. Rept. Food Invest. Board 1938, 127. Hutchinson, J. C. D., Bacon, J. S. D., Macrae, T. F., and Worden, A. N. (1943). Biochem. J . 37,550. Irving, G. W., Fontaine, T. D., and Warner, R. C. (1945). Arch. Biochem. 7, 475. Janisch, R. (1942). Nachrbl. deut. Pflanzenschutzdienst 22, 56. Johns, C. O., and Jones, D. B. (1916). J. Biol. Chem. 28, 77. Johnson, T. B., and Brown, E. B. (1922). J . Biol. Chem. 64, 721. Jones, D. B. (1944). Federation Proc. 3, 116. Jones, D. B., and Moeller, 0. (1928). J . Biol. Chem. 79, 429. Johanson, R., and Lugg, J. W. H. (1946). Australian J . Ezptl. Biol. Med. Sci. 24, 321. Johanson, R., Moir, F., and Underwood, E. J. (1948). I n press (from Prof. Underwood). Kabat, E. A., Heidelberger, M., and Bezer, A. E. (1947). J. Biol. Chem. 168, 629. Keilin, D., and Smith, J. D. (1947). Nature 169, 692. Kemp, A. It., and Straitiff, W. G. (1940). J . Phys. Chem. 44, 788. Kiesel, A,, Belozersky, A , , Agatov, P., Bivshich, N., and Pavlova, M. (1934), Z . physiol. Chem. 226, 73. Klein, G., and T a u b k k , K. (1932). Biochem. 2.261, 10. Klose, A. A., and Fevold, H. L. (1944). Proc. SOC.Ezptl. Biol. Med. 66, 98. Knight, C. A. (1942a). J. Biol. Chem. 146, 11. Knight, C. A. (1942b). J. Am. Chem. SOC.64, 2734. Knight, C . A. (1943). J. Biol. Chem. 147, 663. Knight, C. A. (1947). J . Biol. Chem. 171, 297. Knight, C. A,, and Stanley, W. M. (1941). J . Biot. Chem. 141, 39. Kogl, F., and Erxleben, H. (1939). Z . physiol. Chem. 268, 57. Kohler, E. (1942). Nachrbl. deut. Pflanzenschutzdienst 22, 59. Krebs, H. A., and Johnson, W. A. (1937). Enzymologia 4, 148. Kuhlmann, A. G. (1937). Nature 140, 119. Kuiken, K . A., Norman, W. H., Lyman, C. M., Hale, F., and Blotter, L. (1943). J . Biol. Chem. 161, 615. Kunitz, M. (1946). J. Gen. Physiol. 29, 149. Kunitz, M. (1947a). J . Gen. Physiol. SO, 291. Kunitz, M. (1947b). J. Gen. Physiol. 30, 311. Kunitz, M., and McDonald, M. R. (1946). J . Gen. Physiol. 29, 143. Lampen, J. O., Roepke, R. R., and Jones, M. J. (1947). Arch. Biochem. 13, 55. Leach, M. F. (1905-1906). J. Biol. Chem. 1, 463. Leeper, G. W. (1941). J . Australian Znst. Agr. Sci. 7 , 161. Lemoigne, M., Monguillon, P., and Desveaux, R. (1937). Bull. soc. chim. biol. 19, 671. Levitt, J. (1946). Plant physiol. 21, 562. Linderstrgim-Lang, K. (1939). Ann. Rev. Biochem. 8, 37. LinderstrBm-Lang, K., Hotchkiss, R. D., and Johansen, D. (1938). Nature 142, 996. Lintzel, W. (1942). Forschungsdienst Sonderheft 16, 749. Lipmann, F. (1941). Advances i n Enzymol. 1, 99. Lugg, J. W. H. (1938a). Biochem. J . 32, 2114. Lugg, J. W. H. (1938b). Biochem. J. 32, 2123.
300
J. W. H. LUGG
Lugg, J. W. H. (1939a). Biochem. J . 33, 110. Lugg, J. W. H. (1939b). Protein Metabolism in the Plant, Appendix 1. A. C. Chibnall, ed., Yale Univ. Press, New Haven, Conn. Lugg, J. W. H. (1943). Biochem. J . 37, 132. Lugg, J. W. H. (1945). Ann. Reu. Biochem. 14, 263. Lugg, J. W. H. (1946). Australian Chem. Inst. J . & Proc. 13, 88. Lugg, J. W. H., and Best, R. J. (1945). Australian J . Ezptl. Biol. Med. Sci. 23, 235. Lugg, J. W. H., and Stanley, P. (1948). Brit. J . Nutrition 2, 21. Lugg, J. W. H., and Weller, R. A. (1941) Biochem. J . 36, 1099. Lugg, J. W. H., and Weller, R. A. (1944a). Australian J . Ezptl. Biol. Med. Sci. 22, 149. Lugg, J. W. H., and Weller, R. A. (1944b). Biochem. J . 38, 408. Lugg, J. W. H., and Weller, R. A. (1948s). Biochem. J . 42, 408. Lugg, J. W. H., and Weller, R. A. (1948b). Biochem. J . 42, 412. MaraiB, J. S. C., and Smuts, D. B. (1939). Onderstepoort J . Vet. Sci. Animal Znd. 12, 369. Martin, A. J. P., and Synge, R. L. M. (1941). Biochem. J . 36, 294. Martin, L. F., Balls, A. K., and McKinney, H. H. (1939). J . Bid. Chem. 130, 687. Maschke, 0. (1858). J. prakt. Chem. 74,436. Mason, T. G.,and Phillis, E. (1943). Ann. Botany 7, 399. Maeur, A., and Clarke, H. T. (1938). J . Biol. Chem. 123, 729. Maeur, A., and Clarke, H. T. (1942). J . Biol. Chem. 143, 39. Melchers, G. (1943). Ber. deut. botan. Ges. 61, 89. Menke, W. (1938). 2.Bot. 32, 273. Mercadante, M. (1875). Gaz. Chim. itul. 6 , 187. Michael, G. (1935). 2.Bolan. 29, 385. Miller, E.J. (1935). Biochem. J . 29, 2344. Miller, E.J. (1936). Biochem. J . 30,273. Miller, E.J., and Chibnall, A. C. (1932). Biochem. J . 26, 392. Miller, G., and Price, W. C. (1946). Arch. Biochem. 11, 329. Mitchell, H. H., and Block, R.J. (1946). J . Biol. Chem. 163, 599. Mitchell, H. H., and Hamilton, T. S. (1929). Biochemistry of the Amino Acids. Chemical Catalog Co., New York. Morel, M. (1941). Ann. inst. Pasteur 67, 285. Mothes, K. (1931). Planta 12, 686. Mothes, K. (1933). Ber. deut. Bolan. Ges. 61, (31). Moyer, L. S., and Fishmann, M. M. (1943). Botun. Gaz. 104, 449. Needham, J. (1942). Biochemistry and Morphogenesis, Cambridge Univ. Press, Cambridge. Neglia, F. J., Hess, W. C., and Sullivan, M. X. (1938). J . Biol. Chem. 126, 183. Newcomer, E. H. (1940). Bolan. Rev. 6, 85. Noack, K. (1927). Biochem. 2. 183, 135. Noack, K.,and Timm, E. (1942). Naturwissenschaften 30, 453. O’Kane, D.E.,and Gunsalus, I. C. (1947). J . Biol. Chem. 170, 433. Onslow, M. W. (1931). The Principles of Plant Biochemistry, Part 1. Cambridge Univ. Press, Cambridge. Oparin, A. I. (1934). Ergeb. Enzymforsch. 3, 57. Oparin, A. I. (1937). Enzymologia 4, 13. Osborne, T.B. (1890-1901). Proteid Constituents of Several Seeds (A Collection of Papers). Dr. Thomas B. Ostorne, New Haven, Conn.
PLANT PROTEINS
301
Osborne, T. B. (1924). The Vegetable Proteins. Longmans, Green, London. Osborne, T. B., and Wakeman, A. J. (1920). J . Biol. Chem. 42, 1. Osborne, T. B., Wakeman, A. J., and Leavenworth, C. S. (1921). J . Biol. Chem. 49, 63. Owen, E. C. (1947). J. Dairy Research 16, 142. Paech, K. (1935). Planta 24, 78. Payne, D. S., and Stuart, L. S. (1944). Advances i n Protein Chem. 1, 187. Pearsall, W. H., and Billimoria, M. C. (1936). Nature 138, 801. Pearsall, W. H., and Billimoria, M. C. (1937). Biochem. J . 31, 1743. Peterson, C. F., Lampman, C. E., Holin, D. W., and Stamberg, 0. E. (1944). Poult r y Sci. 23, 287. Petrie, A. H. K. (1943). Biol. Revs. Cambridge Phil. Soe. 18, 105. Petrie, A. H. K., and Williams, R. F. (1938). Australian J . Ezptl. Biol. Med. Sn'. 16, 347. Petrie, A. H. K., and Wood, J. G. (1938a). Ann. Botany N . S. 2, 33. Petrie, A. H. K., and Wood, J. G. (1938b). Ann. Botany N . S. 2, 887. Petrik, F. G. (1944). J . Bact. 48, 347. Pfeffer, W. (1872). Jahrb. wiss. Botan. 8, 429. Phillis, E., and Mason, T. G. (1942). Ann. Botany N . S. 6, 469. Piria, R. (1844). Compt. rend. 19, 575. Pirie, N. W. (1942). Chemistry & Industry, 45. Pirie, N . W. (1945). Advances in Enzymol. 6, 1. Pirie, N. W. (1946). Ann. Rev. Biochem. 16, 573. Prianishnikov, D. (1899a). Landw. Vers. Sta. 62, 137. Prianishnikov, D. (1899b). Landw. Vers. Sta. 62, 347. Prianishnikov, D. (1922). Ber. dent. botan. Ges. 40, 242. Prianishnikov, D. N. (1945). Bull. acad. sci. U. R. S. S. Sdr. biol., 147. Price, W. C., Williams, R. C., and Wyckoff, R. W. G. (1946). Arch. Biochem. 9, 175.
Pringsheim, E. G., and Robinow, C. F. (1947). J . Gen. Microbiol. 1, 267. Rautanen, N. (1946). J. Biol. Chem. 163, 687. Rich, C. E. (1936). Cereal Chem. 13, 522. Richards, F. J. (1938). Ann. Botany N . S. 2, 491. Richards, F. J., and Templeman, W. G. (1936). Ann. Botany N . S. 2, 60, 367. Rischkov, V. L. (1943). Phytopathology 33, 950. Rischkov, V. L., and Gromyko, E. (1941). Microbiology U. S. S. R. 10, 7. Rittenberg, D., and Shemin, D. (1946). Ann. Rev. Biochem. 16, 247. Robinow, C. F. (1945). The Bacterial Cell, Addendum. R. J. Dubos, ed., Harvard Univ. Press, Cambridge, Mass. Roche, J., and Mourgue, M. (1943). Compt. rend. SOC. biol. 137, 766. Ross, A. F (1940). J . Biol. Chem. 136, 119. Ross, A. F. (1941). J . Biol. Chem. 138, 741. Ross, A. F. (1942). J. Biol. Chem. 143, 685. Rothba, F., and BruBre, P. (1944). Ann. pharm. franc. 2, 104. Rothen, A. (1946). J . Biol. Chem. 163, 345. Rothen, A. (1947a). J. Biol. Chem. 167, 299. Rothen, A. (1947b). Advances in Protein Chem. 3, 123. Rouelle, M. (1773a). J. mbd. chir. pharm. 39, 250. Rouelle, M. (177313). J. mld. char. pharm. 40, 59. Ruhland, W., and Wetzel, K. (1929). Planta 7, 503. Sandstedt, R. M., and Blish, M. J. (1933). Cereal Chem. 10, 359.
.
302
J. W. H. LUGG
Schanderl, H. (1939). Gartenbauwiss. 13, 406. Schanderl, H. (1942). Wochschr. Brau. 69, 59. Schmidt, C. L. A. (19451. Chemistry of the Amino Acids and Proteins. C. C. Thomas, Springfield, Ill. Schmitt, F. 0. (1944). Advances in Protein Chem. 1, 25. Schulze, E.,and Umlauft, W. (1876). Landw. Jahrb. 6, 819. Schwab, G . (1936). Planta 26, 579. Seibert, F. B. (1926). Science 63, 619. Seibert, F. B. (1928). Am. Rev. Tuberc. 17, 402. Seibert, F. B. (1941). Bact. Revs. 6, 69. Sherman, W. C. (1941). Agr. Ezpt. Sta.,Alabama 61st Ann. Repf.,25. Sideris, C. P., Krauss, B. H., and Young, H. Y. (1937). Plant Physiol. 12, 899. Sideris, C. P., Krauss, B. H., and Young, H. v. (1938). Plant Physiol. 13, 489. Sigurgiersson, T., and Stanley, W. M. (1947). Phytopafhology 37, 26. Sinclair, W. B., Bartholomew, E. T., and Nedvidek, R. D. (1935). J. Agr. Research 60, 173. Skoog, F. (1947). Ann. Rev. Biochem. 16, 529. Slade, R.E.,Branscombe, D. J., and McGowan, J. C. (1945). Chemistry & Zndustry, 194. Smith, A. H. (1925). J. Biol. Chem. 63, 71. Smith, A. K.,and Circle, S. J. (1938). Znd. Eng. Chem. 30, 1414. Smith, A. M., and Wang, T. (1941). Biochem. J . 36, 404. Smith, E. L. (1941a). J . Gen. Physiol. 24, 565. Smith, E.L. (1941b). J . Gen. Physiol. 24, 583. Smith, E.L. (1942). Chronica Botan. 7, 148. Smith, E. L., and Greene, R. D. (1947). J . Biol. Chem. 167, 833. Smith, E. L., and Greene, R. D. (1948). J . Biol. Chem. 172, 111. Smith, E. L., Greene, R. D., and Bartner, E. (1946). J . Biol. Chem. 164, 159. Smith, E. L., and Pickels, E. G . (1941). J . Gen. Physiol. 24, 753. Snell, E. E. (1945). Advances i n Protein Chem. 2, 85. Soutoulov, A. N. (1946). Compt. rend. acad. sci. U . R . S.S.63,331. Spencer, E. L. (1941). Plant Physiol. 16, 227. Spencer, E. Y.,and McCalla, A. G. (1938). Canadian. J . Research 16C, 483. Spengler, 0. (1918). The Decline of the West (Atkinson's translation, 1926). George Allen and Unwin. Ltd., London. Spiegelman, S., and Kamen, M. D. (1946). Science 101, 581. Srb, A. M.,and Horowitz, N. H. (1944). J . Biol. Chem. 164, 129. Stanley, W. M. (1935). Science 81, 644. Stanley, W. M. (1937). Phytopathology 27, 1152. Stanley, W. M. (1938). Am. Naturalist 72, 110. Stanley, W. M. (1943). Virus Diseases. T. M. Rivers, ed., Cornell Univ. Press, Ithaca, N. Y. Stare, F. J., and Hegsted, D. M. (1944). Federation Proc. 3, 120. Stedman, E. (1945). Biochem. J . 39, lviii. Stedman, E.,and Stedman, E. (1943). Nature 162,267. Stern, K. G., Schein, A. IT., and Wallerstein, J. S. (1946). J . Biol. Chem. 166, 59. Steward, F. C., and Preston, C. (1940). Plc n! Physiol. 16, 23. Steward, F. C., and Preston, C. (1941a). Plant Physiol. 16, 85. Steward, F. C., and Preston, C. (194lb). Plant Physiol. 16, 481. Steward, F. C., Preston, C., and Ramsmurti, T. K. (1943). Ann. Botany N. S. 7 , 244.
PLANT PROTEINS
303
Steward, F. C., Stout, P. R., and Preston, C. (1940). Plant Physiol. 16, 409. Steward, F. C., and Street, H. E. (1947). Ann. Rev. Biochem. 16, 471. Stewart, A. M., and Moir, R. J. (1947). Persona1 communication Stockelback, L. S., and Bailey, C. H. (1938). Cereal Chem. 16, 801. Stockinger, H. E., Ackerman, H., and Carpenter, C. M. (1944). J. Bact. 47, 129. Stokes, J. L., and Gunness, M. (1946). J . Bact. 62, 195. Stokes, J. L., Gunness, M., Dwyer, I. M.,and Caswell, M. C. (1945). J.Biol. Chem. 160, 35. Street, H.E.,Kenyon, A. E., and Watson, G. M. (1946). Ann. Applied Biol. 33, 369. Sukhov, K. S. (1943). Compt. rend. acad. sci. U.R. S. S.39, 73. Sukhov, K. S., Vovk, A. M., and Alekseeva, T. S. (1943). Doklady Akad. Nauk. S.S.S.R. 41, 358. Sullivan, J. T. (1943). Science 98, 363. Sumner, J. B. (1926). J . Biol. Chem. 69, 435. Sumner, J. B.,and Somers, G. F. (1947). Chemistry and Methods of Enzymes. Academic Press, New York. Svedberg, T., and Pedersen, K. 0. (1940). The Ultracentrifuge. Clarendon Press, Oxford. Synge, R. L. M. (1045). Biokhiniiya 10, 179. Tamura, 8. (1913) Z. physiol. Chem. 87, 85. Tanner, F.W., Pfeiffer, S. E., and Van Lanen, J. M. (1945). Arch. Biochem. 8,29. Tatum, E. L., and Bonner, D. M. (1943). J. Biol. Chem. 161,349. Taylor, E. S. (1947). J.Gen. Microbiology 1, 86. Tiedjens, V. A., and Blake, M. A. (1932). Agr. Expt. Sta. New Jersey Bull. 647, 3. Timm, E. (1942). Z. Botan. 38, 1. Tincker, M. A. H. (1947). Research 1, 11. Tristram, G. R. (1939). Biochem. J . 33, 1271. Tristram, G. R. (1940). Biochem. J . 34, 301. Tristram, G. R. (1941). Biochem. J. 36,413. Vendrely, R.,axid Lehoult, Y. (1946). Compt. rend. 222, 1357. Vickery, H.B. (1925). Carnegie Inst. Wash. Year Book 24, 349. Vickery, H. B. (1941). Biol. Symposia 6, 3. Vickery, H.B. (1945). Physiol. Revs. 26, 347. Vickery, H.B.,and Clarke, H. T. (1945). Science 102,454. Vickery, H.B.,and Pucher, G. W. (1939). J. Biol. Chem. 128, 685. Vickery, H. B., and Pucher, G. W. (1943). J . Biol. Chem. 160, 197. Vickery, H. B., Pucher, G. W., Schoenheimer, R., and Rittenberg, D. (1939). J. Biol. Chem. 129, 791. Vickery, H. B., Pucher, G. W., Schoenheimer, R., and Rittenberg, D. (1940). J. Biol. Chem. 136, 531. Vickery, H.B., Pucher, G. W., Wakeman, A. J., and Leavenworth, C. S. (1937). Agr. Expt. Sla. Connecticut Bull. 399, 757. Vickery, H.B., Pucher, G. W., Wakeman, A. J., and Leavenworth, C. S. (1939). Agr. Expt. Sta. Connecticut Bull. 424,7. Vickery, H.B.,Pucher, G. W., Wnkeman, A. J., and Leavenworth, C. S. (1946). Agr. Expt. Sta. Connecticut Bull. 496, 1. Viets, F. G., Moxon, A. L., and Whitehead, E. I. (1946). Plant Physiol. 21, 271. Virtanen, A. I. (1938). Lanlbruks-Hogskol. Ann. 6, 429. Virtanen, A. I. (1939). Trans. Intern. Congr. Soil Sci. 3rd Congr. Oxford A, 4. Virtanen, A. I., and Arhimo, A. A. (1939). Suomen Kemislilehli l2B, 24.
304
J. W. H. LUGQ
Virtanen, A. I., and De Ley, J. (1948). Arch. Biochem. 16, 169. Virtanen, A. I., and Laine, T. (1937). Suomen Kemistilehti lOB, 24. Virtanen, A. I., and Laine, T. (1938a). Nature 141, 748. Virtanen, A. I., and Laine, T. (1938b). Nature 141, 165. Virtanen, A. I., and Laine, T. (1941). Biochem. 2.308, 213. Virtanen, A. I., and Laine, T. (1945). Suomen Kemietilehti 18B, 38. Virtanen, A. I., Linkola, H., Hakala, M., and Rautanen, N. (1946). Suomen Kemistilehti lQB, 83. Vivino, A. E., and Palmer, L. S. (1944). Arch. Biochem. 4, 129. Vorob'eva, M. N. (1939). Bull. acad. sci. U.R.S.S. SLr. biol., 1103. Walkley, J., and Petrie, A. H. K. (1941). Ann. Botany N. S. 6, 660. Warburg, O., and Negelein, E. (1920). Biochem. 2. 110, 66. Wildman, 5. G., and Gordon, S. A. (1942). Proc. Nall. Acad. Sci. U. 5. 28, 218. Wildman, S., and Bonner, J. (1946). Am. J . Botany 33, Suppl. 22a. Wildman, S. G., and Bonner, J. (1947). Arch. Biochem. 14, 381. Wilkins, H. L. (1937). Science 86, 526. Williams, G. (1916-1917). J. Agr. Sci. 8, 182. Williams, R. F. (1938). Australian J . Ezptl. Bid. Med. Sci. 16, 65. Wilson, P. W., and Burria, R. H. (1947). Bact. Revs. 11, 41. Wolf, F. T. (1948). Arch. Biochem. 16, 143. Wood, H. G., Werkman, C. H., Hemingway, A., and Nier, A. 0. (i941). J . Biol. Chem. 139, 483. Wood, J. G. (1941). Auatralian J . Ezptl. Biol. Med. Sci. 19, 313. Wood, J. G. (1945). Ann. Rev. Biochem. 14, 665. Wood, J. G. (1947). Personal communication. Wood, J. G., and Cruickshank, D. H. (1944). Austrdian J . Ezptl. Biol. Med. Sci. 22, 111.
Wood, J. G., Cruickshank, D. H., and Kuchel, R. H. (1943). Australian J . EzpU. Biol. Med. Sci.21, 37. Wood, J. G., Mercer, F. V., and Pedlow, C. (1944). Australian J . Ezpll. Biol. Med. Sci. 22, 37. Wood, J. G., and Petrie, A. H. K. (1938). Ann. Botany N. S. 2, 729. Wood, J. G., and Petrie, A. H. K. (1942). Australian J. Ezptl. Bid. Med. Sci. 20, 249.
Woods, E., Beeson, W. M., and Bolin, D. W. (1943). J. Nutrition 26, 327. Woods, M. W., and Du Buy, H. G. (1941). Phylopathology 21,978. Wormell, R. L. (1946). U. S. Pat. 2,406,650. Wynd, F. L. (1943). Botan. Rev. 9, 395. Yemm, E. W. (1937). Proc. Roy. SOC.London Bl23, 243.
Synthetic Fibers Made from Proteins BY HAROLD P. LUNDGREN Western Regional Research Laboratory, U.S . Department of Apiculture, Albany, California
CONTENTS Page
I. Introduction, . . . ......................... 11. General Consider vior.. . . . . . . . . . . . . . . . . . . . . 111. Preparation of Fibers from Proteins.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. Outlineof Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Chain Length and Fiber Strength.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Alkaline Agents in Protein Fiber Preparation ..................... 4. Factors Relating to Solubility, Denaturation, and Unfolding of Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Application of Detergents to Protein Fiber Preparation.. . . . . . . . . . . 6. Complex Formation of Silk Fibroin with Cupriethylenediamine. 7. Synthetic Fibers Made from Collagen.. . . . . . . . . . . . . . . . . . . . . . . 8. Extrusion of Powdered Proteins Moistened with Plasticizing Agents. 9. Synthetic Fibers of Proteins Mixed with Other Linear Polymers.. ... 10. Summary .......................................... IV. Molecular Basis for Mechanical Properties of 1. Introduction. ......................... 2. Orientation and Mechanical Properties. . 3. Crystallization in Fibers. 4. Folding and Unfolding of 5. Mechanical Properties Related to Chain Folding and Chain Interaction 6. Experimental Methods in the Study of Molecular Basis of Fiber Behavior. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7. Theoretical Interpretations of Stress-Strain 8. Thermoelastic Behavior as Related to Chain Interaction.. . . . . . . . . . . 9. Chain Interaction in Synthetic Fibers Made from Feather Keratin.. . . V. Summary and Conclusion. .. . ..... ........... References ..................... ......... .................
305 307 311 311 312 313 315 317
325 326
329
334 336 345
I. INTRODUCTION Notable advance in our knowledge of proteins came with the recognition of their behavior as typical linear polymer systems. As linear macromolecules we recognize proteins to be fiber-forming materials, yet to differ among themselves in this property partly because of differences in chain length and partly because of differing size, shape, chemical 305
306
HAROLD P. LUNDOREN
nature, and disposition of amino acid residues along the chains. We have come to recognize two principal types of molecular configuration in proteins-the fibrous and corpuscular. I n the one case, the chains are believed to be stabilized as more or less extended configurations and in the other as uniquely folded patterns. Present-day work on the fibrous proteins covers a wide range of objectives. Among these are (a) fundamental studies designed to further our understanding of the stabilizing forces between chains, ( b ) studies on the influence of composition on the ability of the chains to rearrange t o new configurations, and (c) investigations on the capacity of chains to interact with other molecular species such as water, electrolytes, and other proteins. Present study is also aimed toward elucidation of the supermolecular structure and mechanical behavior of muscle fibers (15,16,144, 145,159), the collagens (12,20,84,133,144), and the keratins (11,19,102, 103), and toward an understanding of protoplasm (142,148,149), its fiber-forming tendencies, and its mechanical properties. Examples of more immediate practical interest include studies on the stability of fibrous proteins in food products (99), including the baking qualities of wheat gluten (146,147) and the tenderness of meat products (90). Another phase of practical interest that has received much attention lately is the industrial utilization of the many cheap and available proteins. Since natural proteins are macromolecules, it would appear that only relatively simple manipulation might be necessary to convert them to fibers of commercial interest, in contrast to the complex operations generally required t o produce the purely synthetic polymers. Similar considerations hold for various other products possible from proteins, including films, plastics, coatings, and adhesives (42,101,139). I n order to realize these possibilities, certain peculiarities must be overcome or turned to advantage. Where proteins have already been used for such purposes, it is possible that new knowledge of the interacting polypeptide chains will provide the basis for further improvement in order to increase their range of usefulness. I n the present review, which is necessarily brief and limited, the practical point of view is maintained; the considerations to be discussed are nonetheless fundamental to the understanding of fibrous proteins in general. Aspects of solubilization, unfolding, and formation of fibers will be covered; orientation and mechanical properties of the fibers will be examined. Throughout the discussion emphasis will be placed on the contemporary approach to the understanding of the molecular basis for the stability and mechanical behavior of protein fibers. The present review will not go into the extensive subject of hardening and tanning of proteins nor deal with the allied subject of water absorption. Several
SYNTHETIC FIBERS MADE FROM PROTEINS
307
reviews and papers of significance dealing with the interactions of proteins with organic reagents including formaldehyde have appeared recently (22,49-52,54,73,134). Fibers made from proteins should be distinguished from the recently reported fibrous proteinlike materials made through modification of Leuchs’ synthesis (53,87,88,120,174). These new and purely synthetic materials exhibit interesting theoretical and practical possibilities; they promise further opportunity t o study the influence of structure on the physical properties of the polypeptide chain. Furthermore these materials appear to provide the basis for development of fibers having a wider range of utility than has been possible for fibers made by transformation of the natural proteins. Because the present discussion is concerned only with the fibers made from natural proteins, the basic concepts to be considered will apply also to future fibers made from purely synthetic polypeptides. In fibers made from polypeptides we may expect to find-and do find, as we shall observe presently-similarity of mechanical properties and structure to these properties and structures in natural protein fibers. In the cases studied the fibers made by man from proteins are true fibers according to mechanical and structural criteria. Some confusion exists in the terminology of fibers made from proteins. Various uses of the terms synthetic, artificial, and regenerated as applied to these fibers reflect indecision. It is true that the fibers made by transformation of natural proteins are not completely synthetic products, but as long as we consider the long chains as our building blocks it is correct to state that these are transformed synthetically into fibers. The term artificial implies that the fiber is “made or contrived by a r t ” and “made t o resemble a natural fiber” but it also suggests that the resemblance is only superficial. We avoid the use of artiJiciaZ because we find the similarities to be more characteristic than the differences. The term regeneratedjbers is selected to denote fibers made from proteins naturally occurring in the fibrous form, such as “regenerated collagen fibers.” Another term is needed to distinguish the fundamentally similar fibers made from corpuscular proteins. The committee D-13 of the American Society of Testing Materials on fibers has proposed (1946) the generic name Azlon t o denote fibers made from proteins in ,general.
11. GENERALCONSIDERATIONS OF PROTEIN CHAINBEHAVIOR So far as we know the polypeptide chain (Fig. 1) postulated by Emil Fischer is common to all proteins. The first demonstration that the extended polypeptide chain is the basis for protein fiber structure was the classical interpretation made by Meyer and Mark (1 17,118,123) of the X-ray diffraction pattern of silk fibroin (Brombyz mori). All
308
HAROLD P. LUNDGREN
evidence obtained from the fibrous proteins that have been studied is in accord with the chain theory for proteins (76,123,124). From the mass of accumulated data on proteins, as from other linear polymeric systems, there has emerged a picture of the typical fiber that is consistent with the many known properties of fibers. We recognize the basic fiber pattern of linear high polymers as a system of overlapping chain molecules joined laterally in random fashion in the longitudinal direction (55,57,69, 104,105). This structure, illustrated in Fig. 2, explains the characteristic fiber properties of high strength, chemical resistivity, and high modulus
POLYPEPTIDE CHAIN FIQ. 1. The flexible, yet aomewhat bulky, polypeptide chain. On the amino acid residues and peptide groups are attractive centers responsible for interactions which stabilize the corpuscular and fibrous proteins.
of elasticity ; it also explains the rubberlike properties of long-range elasticity and resilience. The properties of fibers in general lie between these extremes-the balance being determined by the degree to which the molecules interact subject to limitations imposed by charge effects and steric hindrance, as well as by the influence of environment. The protein fiber is similar in stability to the native corpuscular protein. The stability of both is determined by interchain attractive forces which have their origin in the various polar and polarizable groups along the principal chain (9,12,38,63,126,156). These attractive forces tend to preserve the protein structure against the dispersive forces of heat, solventa, and applied mechanical force. Where the native protein is characteristically susceptible to profound alteration through denaturation, dissociation, or aggregation-the result of changes in environmentthe fiber in similar manner is susceptible to analogous changes in mechanical properties with environment. Thus, for example, when a wet
309
BYNTKETIC FIBERS MADE FROM PROTEINS
collagen fiber is heated it undergoes contraction in length within a characteristic temperature range like that of denaturation (10,63,124). When silk fibroin is placed in saturated lithium iodide it first acquires rubberlike behavior; then finally all of the interchain attractions are severed and the silk fiber dissolves (170). The dissolved protein, considered in the molecular-kinetic sense, now behaves in the same manner NATIVE CORpuscULAR PROTEIN
FIBROUS PROTEIN
. . . . , _ ‘ I .
HEATJ
SOCVEN
.
,
....
DISSOGIATION
I FORMATION
HEAT
SOLVENTS
DISPERSED FLEXIBLE CHAINS
FIG. 2. Several states of organization among protein chains. Attractions that stabilize the corpuscular and fibrous proteins are dissociated by heat and solvents. When freed from their confinements, chains from either type behave, molecularkinetically, in similar manner. Fiber formation consists in the interaction of aligned, overlapping chains.
as any dissolved denatured protein. Such material has fiber-forming tendencies. The coagulation of proteins by heat, alcohols, acids, etc. has long been recognized as involving the two steps: the denaturation process followed by aggregation of the denatured protein (23,130). Astbury (7,13,14) was first to interpret the change in the X-ray diffraction pattern attendant on denaturation as “due to the liberation of peptide chains which aggregate en coagulation into parallel bundles like those found in keratin.” The formation of organized areas, the result of aggregation, is indicated by the sharpening of the X-ray diffraction pattern. Astbury showed how the denatured and partially reorganized
310
HAROLD P. LUNDGREN
system could be further organized through mechanical manipulation to give material showing the typical /%keratin pattern normally found with silk fibroin and stretched wool. Be and his associates (14,30) prepared fibers from native corpuscular proteins which had been dispersed in concentrated aqueous urea solutions. These considerations have provided direction for the more recent research on fibers made from proteins. Interestingly, the fiber-forming property of certain biologicals was observed long before the molecular basis for this behavior was recognized. Record of this property was made as far back as 16G4 by Robert Hooke, curator of the Royal Society in London and discoverer of the fundamental relation between the stress and strain properties of elastic materials. From observations made with a simple microscopc he drew the following remarkable conclusions, which he described in his Micrographia (74) :* “Of fine waled Silk, or Taffety” “ A pretty kinde of artificial stuff I have seen, looking almost like transparent Parchment, Horn, or Ising-ghss, and perhaps some such thing it may be made of, which being transparent, and of a glutinous nature, and easily mollified by keeping in water, as I found upon trial, had imbib’d, and did remain ting’d with a great variety of vivid colours, and to the naked eye, it look’d very like the substance of the Silk. And I have often thought that probably there might be a way out, to make an artificial glutinous composition, much resembling, if not full as good, nay bettcr then that Excrement, or whatever othcr substance it be out of which, the Silk-worm wire-draws his clew. If such a composition were found it were certainly an easier matter to find very quick ways of drawing it out into small wircs for use. I need not mention the use of such an Invention, nor the benefit that is likely to accrue to the finder, they being sufficiently obvious. This hint therefore, may, I hope, give some Ingenious inquisitive I’rrson an occasion of making somc trials, which if successful, I have my aim, and I suppose he will have no occasion to be displeascd.”
Hooke obviously was ahead of his time. It was not until a couple of centuries later a t the Paris Exhibition in 1889 that the first “artificial silk,” made from cellulose was demonstrated by Count de Chardonnet (48). The sensation it created stimulated further development and interest in other fiber-forming materials. Soon afterwards Millar (125) in Scotland obtained fibers from gelatin, egg albumin, and casein. All of these early “artificial” fibers were very poor in quality. Their subsequent improvement, chiefly during the past decade, has been supported by the developing concepts of the basic structural pattern of fibers, and provides material evidence that our present ideas of fiber construction are useful if not substantially correct. T h e development by C‘arothers (29) of polyamides (nylon-type polymers) provided recent stimulus to the further study of linear polymers of various sorts.
* A copy of this book is in the University of California (Berkeley) Library’s rare bookroom.
SYNTHETIC FIBERS MADE FROM PROTEINS
31 1
The expanding literature on proteins, and particularly the patent literature, contains abundant reference to their fiber-forming properties. This property of proteins, held in common with other natural and synthetic linear polymers, including cellulose (135,136) and the polyamides, is generally evident when these materials are separated from solutions under conditions which favor the overlapping of the elongated dissolved units resulting in the formation of fibrillar or stringy precipitates. Of physiological interest are the fiber-forming properties of fish proteins (43,112,113) and muscle proteins (36,159,169) and of nucleoproteins from cell nuclei (127) and sperm (140); of therapeutic significance are the fibrous products derived from fibrinogen (42) ; and of practical interest are the fibers made from various industrial proteins including silk fibroin (32,33,47,122,141,170),collagens (28), keratins (96,100,166), zein (35,37,158), proteins of peanuts (14,26,116,160,162-164), cottonseed (26), and soy beans (21,60,107,153,160), casein (107,109,110,138,151) and technical egg white (95,96,98). Although many of these proteins by themselves are not fiber-forming, they become so when treated in a special manner. We shall consider these conditions and limitations in the discussion which follows. 111. PREPARATION OF FIBERS FROM PROTEINS 1. Outline of Methods
Fibers have been made from proteins in several ways: by direct precipitation, by spreading on surfaces (8,34,61,77,130),and by extrusion of solid proteins moistened with water or other plasticizers (109,110, 131, 132,152). The most practical method for the large-scale production of fine-diameter filaments is a duplication of the natural process by which the silkworm produces its filaments. The fibroin filaments are made from a viscous water-soluble protein which coagulates as a result of the mechanical conditions of extrusion. The passage of the solution through the channels in the worm’s head, followed by drying of the stretched extruded material, would favor the alignment and interaction of the chains. Experience has shown that i t is practical to prepare synthetic fibers from concentrated viscous solutions (or melts in the case of polyamides) of long-chain materials by similarly extruding, precipitating (or cooling), and finally by stretching the resulting filaments as illustrated diagrammatically in Fig. 3. This procedure was adopted by Chardonnet and Millar and has since become the basis for the presentday large-scale processing of synthetic fibers. Numerous patents (153) have been issued on protein-spinning solutions, on modified manipulations, and on the after-treatment of the
312
HAROLD P. LUNDOREN
fibers. In general, concentrated viscous solutions, containing from 15 to 30 % of dissolved “spinnable ” protein, are forced through spinnerettes into a precipitating bath consisting usually of aqueous solutions of acids, inorganic salts, and, frequently, a heavy metal ion. The filaments are stretched and then chemically “ cured’’ by agents such as formaldehyde and acetic anhydride. Until recently these processes have been developed more by craftmanship than by science, but gradually the empirically
FIG. 3. Mechanics of fiber spinning. Viscous spinning solution is extruded through small spinnerette openings into precipitating bath. Filaments formed are washed, stretched, reeled up, and subsequently treated with chemical curing agents to improve resistance to water.
developed methods are being extended in accord with better understood theoretical foundations. 2. Chain Length and Fiber Strength
A first theoretical consideration in the making of fibers from linear macromolecules is whether the material is suitable with respect to chain length. Although little work has been done on proteins, the study of this requirement in the other systems, illustrated in Fig. 4, has provided the specifications of chains if desirable tensile strength is sought (44,106,155). It is found that chains must possess at least a certain minimum stretchedout length, commonly referred to as degree of polymerization, or simply DP. This lower limit is of the order of 40 to 80 DP, which for the average polypeptide chain corresponds to a molecular weight of not over 10,000. Below this value the prepared fibers or films cannot be expected to exhibit any significant strength and above this minimum the tenacity increases in proportion to chain length up to DP about 200 (protein
SYNTHETIC FIBERS MADE FROM PROTEINS
313
molecular weight about 24,000). Above this value the strength levels off and approaches a limiting value between 300 and 600 DP (protein molecular weight between 36,000 and 72,000). The increase in tenacity with chain length within the specified limits is believed to depend on the greater area of contact that is possible between the overlapping chains as well as the greater chance th at interacting groups are in juxtaposition. With longer chains it is believed that the looping back and forth prevents further increase in strength.
FIG.4. Relationship of strength of fibrous materials and the degree of polymerization of the chains from which they have been madc (after Mark, 106). Information presently available indicates that a corresponding relationship is valid for proteins. (1 g. per denier for the case of protein fibers is equivalent to 16,600 lb. per square inch, equal approxirnately to dry strmgth of wool; the strength of silk is about three times this value.) I>P is a direct function of chain length. Each DP unit for protein chain corresponds to cn. 120 units in niolecular weight.
From these considerations we understand why short-chain poly,peptides, for example, salmine of molecular weight 5,000, do not form fibers, whereas proteins of higher molecular weights can be transformed to fibers having desirable strength. For instance, egg albumin, of molecular weight 45,000, can be made into fibers whose dry strength values lie between those of wool and silk (96,151). For these reasons we should obtain knowledge of the potential chain length of a protein before we spend time trying to convert it t o fibers of a quality th a t theoretically may not be possible. Unfortunately many of the industrially interesting proteins have not been characterized with respect to molecular size.
3. Alkaline Agents in Protein Fiber Preparation
Also based on the foregoing considerations is the choice of solvent t o be used for the protein: the solvent obviously should not degrade the
314
HAROLD P. LUNDGREN
chains. Because proteins are particularly subject to hydrolysis, this matter becomes important when alkaline dispersing agents are used. Alkaline agents, including sodium hydroxide, have been most frequently employed, not only because they readily solubilize proteins, but also because they render them “ spinnable.” The latter property, as we shall discuss presently, is associated with the unfolding of the protein. A few proteins, myosin and various nucleoproteins for example, exist in their native condition as discrete and highly asymmetric particles so that their conversion to fibers is relatively simple. Thus fibers can be made from myosin simply by extrusion of a salt solution of the native protein into water (169). Waugh (167,168) has reported the conversion of insulin to fibrillar form by heat treatment a t 100°C. of solutions a t pH 2 to 2.5. Electron micrographs of this material reveal the presence of uniform fibrils measuring several microns in length and approximately 200 d in width. Such fibrils may be regenerated to yield products resembling insulin in crystallization, biological activity, and fibril formation. It appears that the insulin molecule has not undergone an unfolding by this treatment, but that the conversion to fibrils and regeneration involves the reversible linear aggregation of the regularly folded insulin molecules. But in general, corpuscular proteins require denaturation and unfolding, and partial alignment of the flexible chains to permit interaction in the overlapping fashion which characterizes a fiber. (By denaturation of the protein we refer to the severance of the stabilizing bonds as preliminary and distinct from the unfolding process.) Casein is unusual among the industrial proteins requiring denaturation, being among the easiest to manipulate because it responds in solutions that are only mildly alkaline. The recent researches by Peterson and associates (138) a t the Eastern Regional Laboratory, by the Bureau of Dairy Industry (171), both of the U. S. Department of Agriculture, and by the Aralac Company (107,153) have resulted in marked improvement in the spinning manipulations as well as in the quality of the fibers made from casein. Casein fibers a t their best resemble wool in textile qualities although they are generally lower in tensile strength and much lower in wet strength. Improved fibers have been made from alkaline dispersions of zein. This development a t the Northern Regional Laboratory by Croston, Eians, and associates (35,37) has resulted in fibers with good elastic elongation and whose dry tenacities lie between those of wool and silk. In these respects zein stands out among the various proteins studied as having superior fiber-forming properties. These investigators have found that a “precure” treatment of the fibers with formaldehyde
SYNTHETIC FIBERS MADE FROM PROTEINS
315
reduces the tendency to plastic flow during the stretching process. They have made significant improvement in the “boil resistance” of the fibers, zein fibers being normally especially susceptible to shrinkage upon exposure to hot water. After acetylation treatment the shrinkage characteristic of the fiber is largely eliminated, yet the wet strength, like that of the casein fibers, is only about half the dry value. However the wet strength is high enough for the fiber to merit serious consideration for commercial production. Similar progress has been made in the fibers made from alkaline dispersions of peanut protein in studies a t the Southern Regional Laboratory (116) and in England (160,162-164). Several other proteins, especially those of soybeans (21,60) and cottonseed (26) are being investigated by similar methods in this country and abroad. Whereas many of the proteins can be handled by mild to moderately alkaline procedures there are some proteins which cannot; egg albumin, for example, forms stiff gels in the protein concentrations customarily suitable for extrusion. Such gels will dissolve with higher concentrations of alkali, but not without serious degradation of the chains. Feathers, as another example, are practically pure protein and appear naturally to be fiber-forming and will dissolve in strongly alkaline solutions. Such dispersions, however, are similarly useless for fibers because of the extensive hydrolytic degradation. Considerations of this sort have stimulated research on the development of new and milder means for the manipulation of the various agricultural proteins available as surplus or waste materials (97,101).
4. Factors Relating to Solubility, Denaturation, and Unfolding of Proteins In selecting the appropriate alternate solvent medium for a protein, we should, therefore, look for a medium which, like alkali, solubilizes the protein and also unfolds it to permit the dispersed chains to form fibers. It is to be expected that this behavior will depend on such factors as the average chain length, the nature and distribution of polar groups, the flexibility of the chains, as well as the tightness of packing and the type and number of cross links. Both theory and experiment confirm the hypothesis that dissolved flexible chains such as are derived either from fibrous or corpuscular proteins prefer the thermodynamically more probable coiled configuration. To uncurl them requires the application of a restraining force. Such unfolding is favored, for example, by the presence of adsorbing surfaces, shear gradients, or by specific chemical agents whieh reduce the tendency to interaction of different portions of the molecule (130).
316
HAROLD P. LUNDGREN
Several properties of solutions of linear polymers indicate unfolding in solution. Unfolding is evident through increase in viscosity (3,4,79, 83). Along with the change in viscosity there is a corresponding improvement in the fiber-forming property. It has been customary, for this reason, to permit aging of the spinning solution, the so-called “ripening” having been found t o improve the spinning quality. The unfolding of protein chains is also recognized through change in the diffusion and the sedimentation behavior of the proteins (83,91,92,128). Both reveal change in the frictional resistance of the mobile protein as a result of TYPES OF CROSS LINKAGES
It
I:
FIQ. 5. Several possible cross linkages between protein chains. Commonly recognized are covalent (I) disulfide bond of cystine and three secondary (11) bonds: salt linkages, van dcr Waals’, and hydrogen bonds. Other possible interchain bonds between protein chains include peptide, thioether, ether, and ester.
change in shape. Unfolding, when marked, is also evident from flow birefringence (36). A considerable number of agents other than alkali have been found that favor denaturation and unfplding as well as the solubilieation of proteins (5,100,130,170). Among them are urea, guanidine salts, formamide, sodium salicylate, various detergents, and solutions of certain inorganic salts such as lithium iodide. The patent literature claims additional agents, which “improve the spinning ” of proteins; doubtless some of these would be of interest to investigate further. The nature of the agents required to solubilize and denature proteins provides indication of the types of bonds which stabilize them. All foregoing agents possess the power to sever hydrogen-bonded structures. Some proteins, notably certain keratins, also require cleavage of disulfide bonds before they are solubilized and unfolded by these agents. Still other proteins, for example, egg shell membrane protein (also keratin) , are insoluble in non-chain-degrading conditions. In general, however, proteins appear t o be stabilized through dissociable salt linkages, hydrogen bonds, and disulfide bonds (Fig. 5). The ease with which they are manipulated will depend on the nature and distribution of these bonds.
SYNTHETIC FIBERS MADE FROM PROTEINS
317
Moreover, the total attraction will increase cumulatively with the number of bonds per unit area of contact. It is not unlikely, however, that those proteins which tend most strongly to resist solubilization and unfolding are held together through other possible bond types; for instance (illustrated in Fig. 5) interchain peptide bonds, ether, thioether, and ester linkages. Undoubtedly the further study of the stability of natural proteins in comparison with the purely synthetic polypeptides will play a significant role in the achievement of better understanding of the molecular basis for protein stability behavior and the nature of their interchain attractions. A striking example of the influence of a specific agent on the solubilization and unfolding of proteins is seen in the action of certain detergents. The work on protein-detergents interactions has been reviewed recently by Putnam in Volume I V of this series. We shall concern ourselves here with those reactions which are related specifically to fiber formation (93-96,166).
5 . Application of Detergents lo Protein Fiber Preparation The fiber-forming property that appears when many proteins are treated with certain detergents is clearly evident when a precipitating salt or excess of a precipitating solvent is added to the solutions. For example, when equal volumes of 1% salt-free solutions of egg albumin and sodium alkylbenzene sulfonate (e.g., a commercial mixture of average composition, dodecylbenzene sulfonate) are mixed a t neutral p H and then treated with magnesium sulfate, a stringy precipitate is formed. As illustrated in Fig. G this precipitate may be drawn by hand into fibers. If, instead of salt, acetone is used, the precipitate which forms, as shown in Fig. 7, is fibrillar. When equal volumes of 20% solutions of egg albumin and sodium alkylbenxene sulfonate are mixed, the viscosity of the mixture increases and with time the solution acquires pronounced flow birefringence. As illustrated in Fig. 8, such solutions exhibit spinning properties desirable for extrusion in the conventional manner. The filaments that form when precipitation occurs in baths of acidified saturated magnesium sulfate solution may be reeled up in continuous fashion. The fiber-forming property conferred on proteins by the ionic detergents is found in mixtures made with more than a minimum limiting proportion of protein and detergent. The relationships between fiberforming behavior and proportions of protein and detergent, as determined by viscosity, electrophoresis, diffusion, and sedimentation analysis, are in accord with the results of chemical analysis of the precipitated materials. These results show that complexes are formed when proteins and
318
HAROLD P. LUNDGREN
Fra. 6. l~il~er-formiiigproperty of protciri-dctcrgent roiriplexes. Depending upon proportion of protck and detergent, salt-precipitated c,omplcxes vary from noncohering floccules to the stringy condition shown, which permits ready mnniprilation to flbers.
FIG.7. Fibrils from rgg alburriin by precipitation of protciii-detergelit (dod(~cy1benzene sulfonate) complex with excess acetone (magnification 3.5 X).
SYNTHETIC F I B E R S MADE FROM P R O T E I N S
319
the salts of ionic detergents are mixed in neutral solution. The combination involves two types of binding forces; the first-the minimum limiting interaction-is chiefly electrostatic in nature, irltrolving the salt linkages (or acid and basic groups) of the protein, and the second is nonelectro-
FIG.8. Extrusion of protein-detergent complcxrs :LS fililmcmts according to conventio~ialspinning procedure. Prccipitation occurs :LH viscwus solution conies into contact with bath of saturated magncsium sulfate.
static, and presumably involvcs forces that, normally bind detergent ions into micelles. Undoubtedly the electrostatic interactions are aided and stabilized by nonpolar van der Wad s forces between the protein and the detergent ions. The first re:tc.t,ion illustrated for an anionic detergent may be represented as follows :
coo-
I
YO-
In reacting with the salt linkages the detergent helps to wedge apart the chains and chain segments. Analyses confirm that the primary reaction of the detergent anions, as shown, is on the basic groups of the protein. The second reaction confers the characteristic stringiness to the precipi-
320
HAROLD P. LUNDOREN
tated complex as well as the desirable flow property to the concentrated solutions and may be represented as follows:
The unfolded protein chain. with the laterally attached detergent is shown diagrammatically in Fig. 9. When water is present the highly charged groups contributed by the second layer of detergent tends to
FIQ. 9. Proposed structure of protein-detergent complexes which accounts for their flow behavior in water. Two layers of anionic detergent are shown attached to unfolded chains, first layer bound through ionic forces to basic groups of the protein ;second layer bound to the residues of the ionically bound detergent through forces presumed similar to those which bind detergent ions as micelles. Highly charged groups on second layer serve t o inhibit interaction of neighboring protein chains. Thus detergent scrves t o plasticize chains to favor their flow in fiber formation.
repel similarity charged groups on adjacent chains and in this way tends to favor the flow of the chains past one another. Similar considerations hold for the interaction of proteins with cationic detergents; in this case, the primary interaction of the protein and detergent involves the acidic groups on the protein. I n accordance with experimental evidence we shall attempt to picture the transitions in chain organization involved in fiber formation, as illustrated in Fig. 10, starting with dispersed chains in the spinning solution and proceeding to the true fiber in which the chains have interacted over relatively large areas. Similar transitions will occur in the formation of fibers from the various linear polymers. The dispersed
SYNTHETIC FIBERS MADE FROM PROTEINS
321
chains in their viscous spinning solutions have the latitude to diffuse
as independent molecular-kinetic units, so-called macro-Brownian motion (85). Frequently such solutions show evidence of a damping-out of chain freedom as recognized by their tendency to form thixotropic gels. This structure, however, is generally broken down upon extrusion. The streamline flow through the spinneret favors the parallel alignment of the chains, indicated by flow birefringence. As the more or less aligned
FIG. 10. Transitions in chain organization in fiber spinning. Freshly precipitated fiber is weak and rubberlike in behavior. Interactions at isolated points (rnesomorphic order) increases with application of stretch. As chains are further aligned conditions favor three-dimensional lattice (crystallite) formation. Further stabilization of fiber network is brought about by treatment with chemical agents intended to cross link the chains as shown. Network stabilization provides for strength; chain freedom provides long-range elasticity; a balance of these properties is generally desired.
chains strike the precipitating bath they are exposed to conditions that favor their interaction. If the chains were not unfolded at this stage, their poor coherence as the result of interaction would yield a flocculent instead of fibrillar precipitate. The freshly formed fiber is generally weak in strength and rubberlike in elasticity. I n such a fiber the unreacted chain portions between the rather infrequent network juncture will still have freedom to undergo extensive coiling and uncoiling. This freedom, called micro-Brownian motion (85), accounts for the long range elastic behavior. As the chains become increasingly interacted, through stretching of the fiber, the micro-Brownian motion is progressively damped out and the material acquires improved strength a t the expense of the long range elasticity. When detergents are used the protein chains are prevented from extensive interaction; attempts to stretch the fiber result chiefly in chain slippage; it becomes necessary, therefore, to remove the detergent before
322
HAROLD P. LUNDGREN
this action materially improves the strength and reduces the brittleness of the dry fiber. The process of recovery of the detergent from the fiber consists in reversal of the reactions involved in the complex formation. It involves the extraction of the detergent with a solvent such as 60 to 70% aqueous acetone containing salt. The extracted fiber remains as substantially pure protein. The regeneration process provides further evidence on the nature of the detergent binding. When no salt is present during extraction, the
FIQ.11. Recovery of dodecylbenzeiie sulfonate detergent from egg albumin fiher. Detergent is extracted with 60 to 70% aqueous acctonc. When no inorganic salt is present, solvent recovers only detergent bound through micellar forces indicated by change in composition from the 50% employed for spinning to the 75% compoaition.
solvent recovers only the “extra” detergent held nonstoichiometrically by the nonelectrostatic forces. In this case the composition changes from t ha t employed in the spinning process to the minimum limiting composition corresponding to one detergent ion for each of the basic (or acidic) groups of the protein molecule. When salt is present the solvent removes essentially all the detergent. The course of extraction of detergent from egg albumin fibers complexed with dodecylbenzene sulfonate is shown in Fig. 11. Aqueous acetone in the absence of salt brings the composition from the 50 to the 75% protein, the latter approximating the limiting proportion of one detergent ion for each of the 43 (equivalents) basic groups present in the egg albumin molecule. I n the presence of salt the solvent also removes ionically bound detergent
SYNTHETIC F I B ER S MADE FROM PROTEINS
323
following the severance of its bonding to the protein by the following ion exchange reaction:
Extraction a t elevated temperature favors this reaction. After the detergent is removed, the fiber can be stretched in hot water or steam with a minimum of chain slippage t o produce a moderately high degree of crystalline orientation. Model fibers made in this manner from native crystalline egg albumin possess mechanical properties and X-ray diffraction patterns (Fig. 12) similar to those of the natural protein fibers.
FIG.12. X-ray diffraction photograph (by K. J. Palmer) of an oriented crystalline egg albumin prepared by the detergent method. Spacings correspond to ,%keratin patterns obtained from natural protein fibers, silk and stretched wool.
These fibers show evidence of the three degrees of chain order that characterize the typical fiber: the oriented crystalline areas; the less well organized, so-called mesomorphic, areas; and the unreacted fibrillar portions. The dry fiber strength is high, and for the most highly oriented fibers comparable with silk, but the wet strength is low. Similarly to other fibers made synthetically from proteins, the degree of chain stabilisation is insufficient to resist the influence of water. Treatment with curing agents only partially offsets this weakness. The detergent method has been applied t o feather keratin (96). As we have already pointed out in this discussion, it is first necessary with this protein, t o sever the disulfide cross linkages before the protein is solubilized. The soluble protein-detergent complexes which form with alkylbenzene sulfonate exhibit fiber-forming properties and molecular-
324
HAROLD P. LUNDGREN
kinetic behavior similar to the complexes made from egg albumin. Like the solutions of egg albumin complexes, the feather keratin solutions possess relatively high viscosity, exhibit marked flow birefringence, and precipitate with salt or acetone in fibrous form. Upon extraction of the dried fibers with 70% acetone containing salt, the fibers may similarly be stretched sufficiently in steam to exhibit crystalline orientation, although the degree of orientation attained with feather keratin is somewhat less than that with egg albumin. Moreover, and typical of other synthetic fibers made from proteins, the synthetic keratin fibers lose 50% or more in strength when wet even after curing with chemical agents according t o present methods. The manipulation of proteins by detergents has a counterpart in the conversion of cellulose to rayon (136). Cellulose is solubilizcd and spun as xanthates, or cuprammonium complexes and these are decomposed to regenerate the cellulose in the desired fiber form. Although similar manipulation of proteins with detergent is more involved than the more simple and commonly employed alkaline dispersion methods, the method, nevertheless, is feasible for the manipulation of the more difficultly soluble proteins such a s feather keratin. Furthermore the study of protein-detergent, complexes provides a pattern for the development of new and mild methods for the manipulation of polypeptide chains.
6. Complex Formation of Silk Fibroin with Cupriethylenediamine Another example of complex formation in the solubilization of a nrotein, also a fibrous protein, is that of silk fibroin dispersed in cupriethylenediamine (32,33,75). Coleman and Howitt have shown that the fibroin dissolves in a “6/8” solution of this agent-prepared by dissolving 6 g. of Cu(OI1)2 and 8 g. of ethylenediamine in 100 ml. of aqueous solution-to form complexes which precipitate upon addition of alcohol. The composition of these complexes indicates that the copper atoms are linked to pairs of neighboring peptide nitrogen atoms along the length of the chain. When the solutions are neutralized to pH 8 with acetic acid and then dialyzed, a portion of the protein precipitates and the remainder stays in water solution; the distribution depends on the time elapsed between solubilization and neutralization. The water-soluble protein possesses a molecular weight of about 30,000 and exhibits certain characteristics of corpuscular proteins. Upon precipitation, the protein becomes insoluble but when the solutions are layered on a clean surface of mercury and the solvent is evaporated in a vacuum over concentrated sulfuric acid or phosphorus pentoxide, a water-soluble film is produced. Such films when moistened show poor tensile strength. However, if
SYNTHETIC FIBERS MADE FROM PROTEINS
325
they are stretched they become insoluble, acquire increase in tensile strength, and now exhibit the @-keratinX-ray diffraction pattern. Coleman and Howitt emphasize the coiling tendency of the dissolved protein as a unique property. However, unless i t can be shown th a t the chains undergo a regular type of folding as in the native proteins, the coiling itself, a s we have already mentioned, is not unique. This behavior is to be expected of linear flexible chains when separated from restraining influences of their natural state. Such behavior is observed, for example, in dilute solutions of feather keratin.
7. Synthetic Fibers Made From Collagen Collagen is another protein of natural fibrous origin which has been studied for its fiber-forming properties. von Buzagh (28) reported that the collagen of calf’s tendons could be solubilized by dilute acetic acid and precipitated with salt as fibers. Recent study of these fibers has been carried out a t the Massachusetts Institute of Technology by F. 0. Schmitt and associates (personal communication). Using standard extrusion procedure they have prepared such fibers having outstanding dry strength, comparable in this property to zein and egg albumin. However, the synthetic collagen fibers, like the fibers made from the other proteins, lose considerable strength when wet, even after treatment using present curing agents. 8. Extrusion of Powdered Protein Moistened with Plasticizing Agents
Recent studies have been carried out in the preparation of bristles by the extrusion of powdered proteins moistened with plasticizing solvents and extruded with elevated pressure and temperature. By essentially this technique, Senti and associates (131,152) have demonstrated the conversion of a number of corpuscular proteins to the oriented crystalline state. I n their study the proteins were kneaded into plastic masses with water and then extruded. Denaturation by the mechanical shearing forces of extrusion was incomplete, but the reaction could be completed by heating the fibers in water or soaking them in various organic solvents, whereupon they became insoluble and elastic and could then be stretched to several times their original length. Oriented fibers with @-keratin structure were made in this manner from casein, zein, edestin, egg albumin, hemoglobin, and tobacco seed globulin. McMeekin (109,111) has studied the preparation and properties of bristles prepared by the extrusion of casein powder moistened with water. After curing treatment with formaldehyde and quinone the bristles exhibit mechanical properties that are of commercial interest.
326
HAROLD P. LUNDGREN
9. Synthetic Fibers o j Proteins Mixed with Other Linear Polymers The possibility of fibers from proteins mixed with other proteins, with cellulose, and other linear polymers has been claimed from time to time in the patent literature. One example of interest is cited in a patent issued to Graves (62) in which mixed fibers of protein with polyamides are claimed. The following proteins were found to mix with polyamides to make fibers or films of “increased strength and water resistance”: zein, casein, keratin, gelatin, silk fibroin, and soybean protein. For example, a zein-polyamide spinning solution is made up of 13 parts of zein, 18 parts of hexamethyleneadipamide, and 60 parts of 98% formic acid. The mixture is spun into heated air a t 110°C. The filaments formed upon evaporation of the solvent are stretched to give fibers whose dry strength is comparable t o that of silk. From the description of these and of other mixed fibers whiih have been reported it appears that their mechanical properties are not much better, if any, than the average of the components separately. This is to be expected unless the component chains of the mixed fiber can be made to fit into common lattice structures or unless they can be crosslinked through covalent bonds. I n this connection there is the interesting suggestion of Astbury (9) that the fiber structural period of nucleic acids is similar to th at of proteins. This condition has been suggested as having possible bearing in connection with the formation of the fibrous nucleoproteins.
10. Summary I n summary, we have seen that proteins vary in the ease with which they respond to solubilization und transformation to fibers, but th a t it is possible by means of ionic detergents to manipulate the more difficultly soluble proteins. With detergents the proteins are converted to soluble, spinable complexes which, after spinning, are decomposed, and the detergent is recovered, leaving the protein in the fibrous form. Of the various proteins investigated, certain ones, particularly egg albumin, zein, and collagen, appear better suited than others for the preparation of fibers having high strength. In this connection we have discussed the general relationship of polymer chain length to fiber strength but, since the molecular characterization has not been conducted on most of the proteins studied for fibers, it is difficult a t the present time to draw further conclusions. We shall turn our attention next to other factors that determine the strength, elastic elongation, and water resistance of the fibers. We shall deal specifically with the limitations imposed by the nature and propor-
327
SYNTHETIC FIBERS MADE FROM PROTEINS
tion of amino acid residues on the chains, the degree of orientation and crystallization of the chains, and the natbre of the forces responsible for chain interaction.
IV. MOLECULAR BASISFOR MECHANICAL PROPERTIES OF FIBERSMADE FROM PROTEINS 1 . Introduction
Progress has been made in recent years in the understanding of the molecular basis for the mechanical properties of fibers, films, plastics, and rubbers. Correlations have been made between the mechanical behavior of many of these products and the nature, size, and flexibility of the particular chains used in their preparation. Studies have sought to explain the behavior of these materials in creep and stress relaxation and the effect of temperature and solvents. The practical goal of these investigations is the establishment of broader bases for the development, where possible, of specific properties in each material for specific commercial uses. An excellent review of recent work on the mechanical properties of high polymers is presented by Alfrey (1). The description of nonprotein polymer behavior provides a model for the resolution of the more complex behavior of protein fibers. 2 . Orientation and Mechanical Properties
The marked relationship between the mechanical properties of fibers made from egg albumin and the degree of stretch applied during fiber formation is illustrated in Table I, taken from the study Nutting, Halwer, Copley, and Senti (132). Increase in stretch, as seen, leads to improved tenacity a t the expense of elastic elongation. Furthermore the wet-todry strength ratio is significantly increased. TABLE I Effect of Applied Stretch to Mechanical Properties of Fibers from Egg Albumin (132) Relative length
Tenacity, g./denier
Wet/dry tenacity ratio
Flexibility
1 .o 2.4 4.2 5.8
0.60 0.89 1.14 1.44
0.28
0.77
-
-
__
0.44
0.81
0.54 0.52
0.37 0.23 -
Applied stretch also leads to orientation in the fiber. Orientation refers to the uniformity of alignment of the chains and organized regions with respect to a specific direction in the fiber, usually the fiber axis.
328
HAROLD P. LUNDGREN
Orientation is recognized by changes in optical properties (17,18,56,151), in the X-ray diffraction pattern (17,18,46,137,149),and in characterizing changes in mechanical properties (106,137,150,151). Thus when fibers are oriented they become birefringent as a result of the difference in the refractive index in different directions in the fiber. Increase in orientation is recognized (see Fig. 12) by the breaking up of the X-ray diffraction rings into arcs and by transition of the arc patterns into well-defined spot (106,137,151). Furthermore the proportion of amorphous to well-defined pattern provides a measure of the degree of order in the fiber. Orientation is generally produced by stretching but it may also be obtained through pressing and rolling of the specimen. In this fashion it is possible to obtain more than one of the crystallite axes aligned with respect to preferred directions. Thus Senti et al. (151) have produced the biaxial orientation of crystallites in fibers made from egg albumin. In leading to the structural alignment, orientation permits the equalization of applied longitudinal stress over the cross section of the fiber, which tends to give the material higher tenacity. Also contributing to increased fiber strength is the crystallization which occurs as the result of orientation.
3. Crystallization in Fibers Crystallization occurs as the chains are deformed from their most probable coiled configuration toward the stretched-out condition where new attractive centers are brought face to face (45,56,106). The first stage of interaction, as we have mentioned, is the formation of regions of local order-the so-called mesomorphic areas; these in turn may grow into small crystallites and the smaller crystallites into larger ones. The crystallites are believed to resemble true crystals in consisting of units arranged in definite geometrical array, but to differ in being fringed at the ends where the interconnecting chains enter and leave (57,106). Orientation cannot lead to crystallization unless the chains can fit easily into lattice structure. The presence of bulky side groups can interfere; in this connection the presence of proline and hydroxyproline can interfere not only through their inherent bulkiness, but also through restriction of rotation necessarily imposed by these residues upon the main valence chain (10,76,106).
4. Foldzng and Unfolding of Chains in Fibers Harris has suggested wool and silk as fiber models for the preparation of fibers from proteins (69,70). Fibers of the “wool type” would be built of chains which do not crystallize well, but which like wool (poten-
SYNTHETIC FIBERS MADE FROM PROTEINS
329
tially at least) can be stabilized through occasional cross-linkages in order to supply strength and t o prevent the plastic flow resulting from slippage of the chains. Fibers of the “silk type,” on the other hand, would require chains which fit together well over relatively large areas to give sufficient cohesion t o resist plastic flow. Silk fibroin is characterized by its relatively high content of glycine, alanine, and serine; these make up over 80% of the total amino acid residues. Their presence contributes to making the chain relatively smooth. The close approach of neighboring chains permits a more firm hydrogen bonding through the peptide bonds of the chain backbone. Even though hydrogen bonds at their best are relatively weak, the total attraction is increased through the cumulative effect of a large number of bonds. Wool is much less crystalline than silk presumably, as Harris points out, because of the greater bulkiness of its side chains. About 50% of the weight of wool is in the side chains. Considering its lower crystallinity together with its greater proportion of hydrophilic groups, wool has a remarkably good wet strength compared with the fibers made synthetically from proteins. This behavior has been attributed t o the presence of the stabilizing disulfide cross-links in wool. When these are broken by reduction or by hydrolysis a t elevated temperatures, the wet wool fiber loses its unique tenacity and now behaves like the synthetic protein fibers. Because the compositions of natural proteins in general resemble wool more than silk, it follows that we should intensify our search for more effective means of introducing into the synthetic fibers covalent cross-linkages to resemble those in wool as one means of improving the present fibers made from proteins. As already indicated, the X-ray spacings characterizing the oriented fibers made from proteins correspond to those obtained with silk and stretched wool, that is, the chains appear t o be fully extended. The regularly folded, so-called a-pattern observed in unstretched wool has so far not been observed in synthetic fibers, even those made from dispersed keratins. 5. Mechanical Properties Related to Chain Folding and Chain Interaction
The folding and unfolding of protein chains undoubtedly plays a role in the mechanical behavior of protein fibers. The earlier contention (6,72) that the highly regular type of folding observed in the X-ray pattern of unstretched wool is the basis for its long range elongation has in recent years been considered doubtful. This folding is observed only in the crystallite fraction, the proportion of which, as we have already pointed out, is relatively low. It appears that we must look for the explanation of the long-range elasticity in the amorphous components of
330
HAROLD P. LUNDGREN
wool. A fresh approach to the understanding of the long-range elastic behavior of keratins has been made by Bull (24,25), based on the mechanical properties of hair, and by Mercer and Rees (114), based on electron micrographs of wool and its separated components. These investigators consider that reversible “gel-sol” transformations within the amorphous areas of the fibers are the limiting factors. We might consider transition of gel to sol under the influence of an applied force as analogous to the melting of ice under an applied pressure. When the stabilizing bonds in the fiber are broken, the chains are free to respond by unfolding or by slipping in order to relieve the tension. The bonds would reform in the stress-free position, the rearrangement being comparable in behavior to the “flow” of glacial ice. The folding and unfolding of proteins will necessarily depend on the size, shape, chemical nature, and disposition of amino residues along the chains. In a discussion of the space requirements for certain proposed structures for corpuscular proteins, Neurath (129) emphasizes the dominant role which glycine must play in the folding of the polypeptide chains; moreover he suggests that the more reactive residues determine the mode of folding of the chains through their mutual interaction. Mark (105,106) suggests that the tendency for certain regularity in the curled-up state seems to be distinctly more pronounced if easily polarizable bonds or hydrogen-bonding groups are regularly distributed along the chain as in proteins and polyamides. Certain mechanical properties of polyamides can be explained by the assumption of a regular folding of the chain in the quenched state. Furthermore, the cold drawing of nylon has been considered to be like the denaturation of a protein, as the conversion of chains from a regularly coiled “internal state of crystallization” to an extended state with “external crystallization”; the breaking of hydrogen bonds in both cases is believed necessary as a preliminary step to the unfolding. From the practical point of view, the response which the fiber exhibits to an applied stress largely determines its utility. We inquire, for example, how strong the fiber is, or if it is not broken, whether it remains permanently deformed when applied stress is removed; we are also interested in how quickly it responds, that is, whether or not it is highly resilient. All these characteristics will depend ultimately on molecular processes including the stretching of bonds, the folding and unfolding of flexible chains, and particularly in wet fibers on the slippage of the chains past one another. These responses will be limited by the nature and extent of the stabilizing bonds in the fiber network. The characteristic low wet strength and plastic flow exhibited by the fibers made from proteins is manifestation of the greater attraction which the many polar
SYNTHETIC FIBERS MADE FROM PROTEIN8
33 1
groups have for water over their normal tendency by mutual interaction to stabilize the fiber network. 6. Experimental Methods in the Study of Molecular Basis of Fiber Behavior The experimental approach to the understanding of fiber behavior (86,115,154) is obtained through measurement of such properties as ( 1 ) the load required to break the fiber, ( 2 )the elongation with time produced by a constant load (creep), (3) the change in force (stress) required to maintain constant elongation in the fiber which has been previously elongated (stress relaxation), (4) the variation in stress w4th temperature (thermoelastic behavior), ( 5 ) the change in length with constant rate of loading, and (6) the variation of stress with constant rate of elongation. Another criterion, the so-called 30 % index, which was introduced originally by Speakman (156), and has been used extensively in recent work on wool by Harris (69-71), is the ratio of the work required to elongate the fiber under selected test conditions to the value measured for the same fiber under normal conditions. I n the wet condition, chemically reduced wool is similar by this criterion to the artificial fibers made from proteins, confirming that the disulfide cross-links in wool are fundamentally responsible for its unique stability. The data obtained from these methods, treated in the light of recent theory, provide new insight into the mechanism of fiber behavior. Such data are especially valuable when studies are made on model fibers of known composition, with chemically altered fibers, and with measurements made in selected solvent environments and covering a range of temperature. In this connection the study of fibers of known and controlled composition made by the recently revived Leuchs’ synthesis would appear to offer interesting possibilities.
7. Theoretical Interpretations of Stress-Strain Behavior of Fibers Recent theoretical treatments of the mechanical basis of fiber behavior are based on the consideration of the two simultaneous processes of elastic and plastic deformation. The elastic processes are regarded as due to the uncoiling of chains and, to a lesser degree, to the stretching of bonds. The plastic processes are considered as the flow (or slippage) of chains past one another. Flow is irreversible and results when chains are not sufficiently strongly anchored to resist the applied mechanical stress. The elastic processes, on the other hand, are reversible, and behave like true springs according to the kinetic theory of elasticity. For the case of bond stretching, the elastic process behaves according to the known elastic characteristics of chemical bonds. It has become fashionable t o represent the behavior of fibers diagram-
332
HAROLD P. LUNDOREN
matically by mechanical models consisting of combinations of springs and dashpots connected in various combinations (1,2,3942,86,106,161). A general model is illustrated in Fig. 13. This model consists of two springs, which represent the elastic processes of flexible chains and bond stretching, and a single dashpot, which represents the chain slippage a t MODEL CHAIN
MECHANICAL MOOEL STRESS - STWN BEHAVIOR
Af MATHEMATICAL REPRESENTATION
FIG. 13. Deformation behavior of fibers. Deformation assumed to consist of elastic (reversible) processes of chain coiling and bond stretching and plastic (irreversible) processes of chain slipping such as occur a t points of weaker attachment. Elastic processes represented by springs, each characterized by a Hooke’s law constant (0, and gl). Plastic processes represented by a dashpot characterized by a viscosity ( q ) . Total force, j , on model is divided into the force, f l , on the spring on the left representing chain segment firmly attached, and the force, fi, on the Maxwell unit on the right representing chain segments subject to slippage as well as elastic behavior. Composite model exhibits stress-strain behavior similar to that of real fibers.
the weaker bonds. The spring in series combination with the dashpot is termed the Maxwell viscoelastic unit (108). The composite unit of the ‘spring in parallel with the Maxwell unit is intended to represent the deformational behavior of the fiber model. An applied force, f, on this model would be the sum of the partial forcesf1 on the spring a t the left and fa on the Maxwell unit. The resulting strain would depend on the two spring constants, g1 and gz, and the viscosity, q , of the dashpot. This simple model will exhibit deformational behavior that is characteristic of real fibers, yet in most cases the response of real fibers, especially the protein fibers, is considerably more complex. Attempts to
SYNTHETIC FIBERS MADE FROM PROTEINS
333
give a better account for the real behavior of fibers have been made by using additional mechanical elements in series and parallel combinations to represent distributions of molecular elements in the fiber network. In general such treatments have at best been only partially successful. In most of the applications, the assumption is made that the dashpot behavior is Newtonian, that is, the flow of viscous elements in the system is directly proportional to the applied stress on the element. In a modified approach to the molecular-mechanical treatment of fiber behavior, Eyring (3941,59,67,68,157) has introduced a new concept of nonlinear flow, in which he considers the process like any chemical reaction to involve the “jumping” of the reacting units over potential energy barriers. In the elementary flow process the moving units are considered to be (not necessarily rigidly determined) groups of atoms in the protein molecule acting as units. The potential barrier consists of the energy required to sever the stabilizing bonds, as well as that necessary to provide a hole into which the flowing unit can move. According to this treatment the movement of the dashpot is expressed mathematically by a hyperbolic sine relation between the rate of flow and the applied stress. The simple Newtonian behavior is the limiting condition toward which the hyperbolic sine behavior approaches as the applied stress is reduced in magnitude. Integration of the differential equation for the Maxwell-type unit with an Eyring dashpot leads to relations between stress, strain, and time, for the general model, which are in better agreement with the observed behavior of several fibers. The analysis of experimental curves leads to the constants which characterize the springs and dashpot; it also yields the free energy of activation for the flow process. These constants, as determined from one type of experiment, for example, strain with constant rate of loading, may then be used to determine the behavior under other conditions, such as creep or stress relaxation. Eyring considers that the elements of the simple model by themselves may represent distributions of molecular elements, so that the characteristic constants represent average values. While the agreement between real behavior of a fiber and that based on the simple Eyring model is good for several fibers studied, including silk and several rayons, marked deviations are observed in the case of wool, and also in the case of a fiber made from feather keratin. Eyring and Halsey (41,65,67) have suggested several possible reasons for these deviations; for example, better agreement is obtained for wool when unsymmetrical rather than symmetrical energy barriers are considered for the dashpots. Moreover deviations from the simple behavior may arise because of departure of the springs from simple Hooke behavior. Furthermore the springs may be converted to dashpots at higher elonga-
334
HAROLD P. LUNDGREN
tions. I n the case of the feather keratin fiber thixotropic effects were noted. Undoubtedly the various types and degrees of chain interaction that a r e possible in the protein fibers give rise to wide distributions of interaction ranging from weak van der Waals' attraction to the covalent disulfide bonds, all of which contribute to their highly complex mechanical response. It would appear that further knowledge of protein network interactions would help to give a better understanding of their mechanical behavior. 8. l'hermoelastic Behavior as Related to Chain Interaction
Another approach to the understanding of network interactions in protein fibers is gained through investigation of their thermoelastic behavior. The influence of of temperature on the elastic properties of a fiber may be compared, as illustrated in Fig. 14, to the influence of temG A S z LIQUID SOLID FREE CHAINS
s
Edlb+d
MEsg&[[HIc MEs~#/~[HIc
REAL GAS
IDEAL GAS
p s - RT
V I
__z
P = - A + - RT V
CRYSTALLITES
"FREE CHAINS
f=BT I
ASSSOCIATED CHAINS
f = -A
+ BT
-T FIG.14. Transition of free chains to fiber analogous to transition of gas to solid. Equilibrium force on stretched rubberlike fiber is determined by thermal kinetic motions of frec chain segments betwcen network junctures. Tension in stretched fiber corresponds to the pressure of a gas, being a direct function of temperature and related to kinetic motion. Mutual interaction in both systems similarly modify their ideal behavior. Transition of rubberlike behavior t o normal solid behavior is identified by change in slope of force-temperature curve.
perature on a gas. I n both cases we are dealing with thermal motion of molecules. Depending upon the freedom of the molecules to respond to the applied heat, the elastic tension of the stretched fiber will vary in a manner analogous to the pressure of the confined gas. This interpretation of rubberlike behavior of linear chain systems was introduced and
SYNTHETIC FIBERS MADE FROM PROTEINS
335
developed principally by Meyer (1 19), Busse (27), Karrer (80), Guth (64), Wall (165), Flory (46), and Eyring (66,161). The theory has been applied t o the determination of the internal energy and entropy contributions of the equilibrium force on various protein fibers including muscle (121a,124,172),collagen (89,121,124), elastin (89), silk (89), hair (24,25), and wool (89,173). The application of the kinetic theory provides a thermodynamic basis for the understanding of the physical properties of fibers. This approach is in accord with the suggestion of Gibbs (58) that I( the comprehension of the laws which goven any material system is greatly facilitated by considering the energy and entropy of the system in the various states of which it is capable.” With fibers we are dealing with chain networks, the stability of which is determined by the state of lateral cohesion between overlapping chains. The situation is fundamentally no different from that observed with other materials. The physical properties of all materials are related t o the degree of interaction among the molecules. Thus the solubility of a solid, the vapor pressure of a liquid, and the osmotic pressure of a solution are all determined by the escaping tendency of the molecules as limited by the degree of their mutual interaction. I n all cases the molecular interaction favors order among the molecules and tends to oppose the influence of thermal motion, which favors disorder. The transition of a gas t o a liquid and of a liquid to a solid are manifestations of increasing order similar to the transitions, which we have discussed, of free chains in spinning solutions to rubberlike fibers containing higher degree of order. I n the fiber, kinetic motion is limited to the free segments of chains between the network junctures. Thermal motion in the fibers, accordingly, corresponds to the whipping about of these free portions. I n whipping about, the chain ‘(prefers” the randomly coiled condition in which its components have greater elbow room, i.e., maximum configuration entropy. The force required t o displace the chains from this most probable configuration is analogous t o the pressure required t o compress a gas. When the temperature of the rubberlike fiber is increased the thermal motion, like the pressure of a heated gas, is increased. The result is that greater force is required to stretch the fiber (or if it is already stretched, to maintain constant elongation). I n highly organized fibers the first effect of temperature is to “melt” the organized areas. This results in a contraction in length when no force is applied to maintain constant elongation; otherwise the force increases. I n either case the transition which takes place results in a more rubberlike behavior of the fiber. The presence of solvents will lower the characteristic temperature region of thermal contraction. It follows from these considerations that an equation of state can be
336
HAROLD P. LUNDGREN
derived for the fiber which represents the force on the fiber at constant elongation as a function of temperature; this representation is analogous to an equation of state which defines the pressure of a gas a t constant volume as a function of temperature. We recall that the behavior of an ideal gas is represented by the familiar van’t Hoff equation in which the pressure is the simple direct function of temperature, namely P = R T / v = bT. According to the kinetic theory of flexible chain networks there exists a similar relation between the force and temperature at constant strain of the network. But just as the real gases are subject to molecular interaction which modifies their ideal behavior and which is represented by the introduction of an additional term to the equation of state, so it is with the flexible chain system; the force in the “real” fibers is more nearly represented by an equation of state of similar form, namely:
f=&A+BT This equation of state is a straight line function when the force is plotted against the temperature. The slope of the line, B, represents the contribution of the kinetic motion to the total force, whereas the intercept, A , represents the contribution of the attractive forces. It is apparent, therefore, that, from measurements of the reversible force on the fiber network determined as a function of the temperature in various solvent media, it is possible to gain insight into the nature and extent of chain interaction. Experiments confirm theories in showing that those environments which sever the interconnecting attractions will lower the force required to maintain constant elongation in the fiber. Through selection of appropriate solvent mixtures it becomes possible in this manner to isolate and study specific bond types. We shall review this behavior specifically for uncured synthetic fibers made from feather keratin by the detergent method discussed earlier (100). 9. Chain Interaction i n Synthetic Fibers Made jrom Feather Keratin
The transition of the oriented crystalline (keratin) fiber to the rubberlike state is analogous to the denaturation of a corpuscular protein. Both transitions occur when the systems are brought above a critical temperature region. When the isoelectric keratin fiber is heated in water, this transition is manifested abruptly at 4547°C. by a contraction in length. If a force is applied to counterbalance this tendency to thermal contraction, the transition is marked by an increase in the required force. The variation in force can be measured in several ways; for example, in the apparatus shown in Fig. 15, the tension in the fiber is communicated to a torsion wire on which a mirror is mounted. By means of an optical lever system the change in tension in the fiber is magnified and the dis-
SYNTHETIC FIBERS MADE FROM PROTEINS
337
placement is registered on an appropriate scale. The displacement is either measured directly or else photographed and then measured on the photographic record. Although slight changes in length of the fiber will necessarily occur, these can be made so small compared with the over-all length of the fiber sample that the fiber can be considered as behaving essentially a t constant length. The transition in state in the keratin fiber is illustrated in Fig. 16. The fiber is mounted a t zero tension a t a low temperature. As the
FIG.15. Apparatus for measurement of equilibrium force-temperature behavior of fibers. Tension in fiber, thermostated in chamber A, is communicated by stiff wire to torsion wire B; displacement of torsion wire as enlarged by optical lever system is recorded by the camera.
temperature is raised, the transition is seen t o occur at the critical region. With higher temperature, the fiber reaches a condition in which the force behaves as a straight line function of temperature with a positive slopeconditions which characterize rubberlike behavior. When the system is cooled below the original critical region, a slight deviation in the force from the linear behavior is observed. This deviation is attributed t o the initial stages of recrystallization of the still more or less aligned chains. If the fiber is maintained in this condition the crystallization process continues, but if the temperature is raised and lowered uninterruptedly the force varies reversibly as indicated over the relatively wide temperature range. The equilibrium of forces which defines the state of the rubberlike fiber is represented in Fig. 17. Tending to preserve the network structure are the interchain attractive forces. In addition there is a small and, for practical purposes, negligible contribution of the hydrostatic pressure
338
HAROLD P. LUNDOREN
TRANSITION OF STATE IN SYNTHETIC FEATHER KERATIN FIBER
ORIENTED CRYSTALLINE STATE
1
-
RUBBERLIKE
STATE
HEAT H,O
i
Fro. 16. Transition of state in synthetic feather keratin fiber.
of the surrounding medium. Opposing these forces and tending t o disperse the network chains are ( I ) the normal kinetic force of the chains, ( 2 ) the osmotic force involving the solvent medium, and (3) the applied mechanical force. The state of the fiber system is determined by those forces which predominate.
I'
+THERMAL FORCE
OsMonC FORCE
MECHANICAL W I N O FORCE
FIG.17. Balance of dispersive and cohesive forces in the fiber network-solvent system.
According to the first and second laws of thermodynamics a reversible change in the state of interaction among the chains, as designated by change in the internal energy, AE, minus the heat absorbed by the system, TAS, is equivalent to the total work done, that is: AE
- TAS=jU-PAV+AF,
SYNTHETIC FIBERS MADE FROM PROTEINS
339
where fA.L is the mechanical work of elongation, P A V is the hydrostatic work, and AFa is the osmotic work. After rearrangement and division by AL,we obtain:
f
=
( A H / U ) T - T(AS/AL)T - ( ~ , / U ) T
+
where AH = A E PAV. Now, since AFa = A H . - TAB. we may combine the energy and entropy changes t o yield:
f
=
(AH - A H , ) / A L v
-
T(AS
-
AS.)/ALr
but since i t is readily shown that the expressions: -T(AS -
AS,)/ALT and
+
+T ( A f / A T ) L
are equivalent, we may write f = f A B T , which is the equation of state that we have discussed previously. The force-temperature behavior of the keratin fiber in water is reversibly altered by the addition of salt and alcohols to the system. For the present we shall review the influence these agents have on the fiber above the critical temperature region (the critical region is shifted somewhat with change in solvent environment). Small increments of salt such as potassium or lithium chloride lower the equilibrium force; excess of added salt increases the force. I n these respects the fiber behavior resembles the salting-in and salting-out of soluble proteins. A shift in pH from the isoelectric region also causes a fall in the equilibrium force as do also increments of added ionic detergents. I n all these cases the principal influence of these added agents is on the intercept of the force-temperature curve, the slopes remaining not much changed from the value in water. Added alcohols similarly lower the equilibrium force; in this case, also, a minimum value is reached. For aqueous solutions of methyl, ethyl, or n-propyl alcohols this minimum occurs in solutions containing approximately equal amounts of alcohol and water. The minimum value is progressively lowered with change from methyl to ethyl to n-propyl alcohol. Furthermore, when the temperature is changed the effect of the alcohols and the salts is chiefly on the intercept value of the forcetemperature curves, the slopes being essentially independent of these solvents above the critical temperature regions. The latter behavior suggests that these agents likewise affect dominantly the internal energy of the system (ie., the network attractive forces) rather than the entropy (kinetic forces). The characteristic influence of the added salts on the system (illustrated also by shift in pH of the solvent from the isoelectric region, and by the addition of ionic detergents) suggests th a t the attractive forces
340
HAROLD P. LUNDGREN
involved are salt linkages of the network; that is, the reaction is considered comparable to the reaction of soluble dipolar ions with ions. For example, this reaction of the fiber network with inorganic salt may be described as follows:
L wa-/\coo-
Li+
The severance of salt linkages would account for the observed decreased force required to maintain constant elongation in the fiber when salt and alcohols are added above the critical temperature region. The observed behavior of the equilibrium force for a range of alcohol and salt concentrations is illustrated in Fig. 18. This figure represents a /
1.0
/ , 0.5
% 0
v 2 FIG. 18. Equilibrium force on the keratin fiber in alcohol-water-lithium chloride mixtures.
SYNTHETIC FIBERS MADE FROM PROTEINS
341
three-dimensional plot of the variation in the equilibrium force ratio of the fiber as a function of the ionic strength of the medium and the inverse square of the dielectric constant. The equilibrium force ratio is the ratio of the force in the solvent under consideration to the force in pure water selected as a standard for comparison. Do is the dielectric constant of pure water at the temperature of the experiment, and D is the dielectric constant of the solution.* Starting a t the upper left in the figure-the condition representing the isoelectric fiber in pure water-the equilibrium force ratio is seen to fall somewhat with increase in ionic strength. This is the expected behavior of dipolar ions interacting with ions. On the other hand the equilibrium force also falls when alcohols are added to the system. Furthermore the influence of the alcohols is nonspecific in solutions of low ionic strength, but is specific for each alcohol in solutions of higher ionic strength. In both cases an apparent limiting behavior of the force with the inverse square of the dielectric constant is observed, but this correspondence cannot be taken as evidence for dipolar ion-ion interaction as was done in the previous interpretation because of the apparently greater change in free energy in transfer of the keratin from water to the alcohol solution. The influence of the alcohol in salt solutions as well as in the salt-poor solutions must be attributed to other than simple electrostatic interactions, presumably hydrogen bond interactions. Hydrogen bond interaction of the alcohols and the peptide hydrogen can be formulated as follows: -N-
T h e inverse square function of the dielectric constant and direct function of the ionic strength provide convenient scales for graphically summarizing the influence of alcohols and ionic strength on the force-temperature behavior of the fibers. The significance of these particular functions rests on their relation with the change in free energy involved in simple electrostatic interaction of dipolar ions with ions. The interactions of amino acids, peptides, and proteins, for example, are expressed in the limiting condition of low ionic strength by the following general equation (31,82,141): AF = -RT log
y =
R T log S/Sa
=
K/D'r/2
This equation, a s derived, is valid only when referred to the specific solvent employed. Thus when the alcohol concentration of the system is changed it is also necessary to consider the change in free energy of transfer of the dipolar ion from water t o the alcohol solution.
342
HAROLD P. LUNDaREN
Similar reactions of the alcohols with hydrogen-bonded serine residues may also occur. It might be supposed that since the water molecule has two hydrogen atoms capable of forming hydrogen bonds it should be more effective in this respect than the alcohols. The preferential interaction of the fiber with alcohol in competition with alcohol would indicate that the R group of the alcohol must play a role in the interaction, possibly through interaction with nonpolar residues on the protein to hydrogen-bonding sites. Moreover the observed specific effects of the separate alcohols in salt solution may be ascribed to the well-known influence of salts in raising the activity of alcohols in water solution. The salting-out of the alcohols necessarily increases with increase in size of the hydrocarbon residue. However, when the oven-dried fibers are treated with the three absolute alcohols, combination with the protein takes place but in nonspecific fashion and with sufficient attraction that it is difficult t o remove the alcohols on heating a t elevated temperatures. On the other hand, if such fibers containing alcohol are placed in water the alcohols are spontaneously replaced. The similar behavior of alcohols on other materials, particularly the hydroxy compounds cellulose and pectin (78) , provides supporting evidence that alcohols combine through hydrogen bonding with these materials. Further supporting this view are the corresponding effects in the lowering of the force on the fiber network found when urea is used in place of the alcohols. The conclusion is drawn, therefore, that hydrogen bonds are the second type of network attractions affected by the alcohols and that in water solution the interaction is influenced by the activity of the alcohols as increased by added salt. Fig. 19 illustrates the reversible force-temperature curves for the isoelectric keratin fiber in water and in 30% (by weight) ethyl alcohol and water containing 0.01 M potassium chloride. The behavior in the alcohol-water-salt mixtures is characterized by a well-defined critical temperature region, above which the fiber is essentially rubberlikc, and below which the force exhibits a marked negative temperature coefficient that typifies normal solid behavior. This behavior at low temperatures, as we shall review presently, is ascribed to a gel formation within the fiber. The force-temperature behavior is reversible over many cycles of heating and cooling but, when a reducing agent (for example monothioglycol in 0.1% concentration) is added a t a low temperature as indicated in the figure, and the temperature is raised above the critical region, the force is no longer supported by the system. As the force falls t o zero, the fiber dissolves. It is apparent that the reducing agent severs a third type of bond that stabilizes the network in these fibers.
SYNTHETIC FIBERS MADE FROM PROTEINS
343
FIQ. 19. Force-temperature curves characterizing isoelcctric synthetic keratin fiber in water, in 30% (by weight) ethyl alcohol in water containing 0.01 Molar KCl, and when a reducing agent (monothioglycol) is added to the alcohol-water-salt system.
Similar response is observed when cyanide or sulfite is used a s the reducing agent. Tests with nitroprusside indicates t h a t the reduction involves disulfide bonds. The reaction with monothioglycol may be considered as follows: HOCIIrCHk3H
A
I
A
S-sCH2CH&H
.s"
When the warm solution of dissolved keratin is cooled the protein precipitates as clumps of gel (or the whole solution sets to a gel with higher protein concentration). Analogous to the transition in the fibers, the gelation occurs a t a well-defined temperature region; moreover the process is reversible with change in temperature. This similarity in behavior of the fiber and the solution suggests that a comparable gel formation involving the solvent occurs within the fiber below the critical temperature region. The latter is viewed as a process distinct from the reversal of chain crystallization discussed previously. Supporting this contention is parallel influence of urea on the fiber and on the solution. An excess of urea prevents gel formation of the solution and also inhibits the transition t o normal solid behavior in the fiber as indicated by the force-temperature curve below the critical temperature region. The
344
HAROLD P. LUNDQREN
behavior with urea also indicates that the gelation process involves hydrogen bonds with the solvent. These bonds are apparently stable in alcohol-water as long as the system is kept below the specific critical temperature. I n concluding this section on the force-temperature behavior of fibers made from feather keratin we have reviewed how the presence in the fiber of specific bond types may be demonstrated through the study of the influence of appropriate solvent mixtures. I n this manner the evidence was presented for the three types of bonds, illustrated in Fig. 20,
l1
+THERMAL FORCE
oswnc
FCRCE
MECHANICAL DEFORMING FORCE
FIQ.20. Three typcs of network stabilizing bonds demonstrated by the thermoelastic analysis of synthetic fibers made from feather keratin.
namely: (1) salt linkages which are severed through interaction with inorganic ions, by shift of pH from the isoelectric region and by the addition of ionic detergents; ( 2 ) hydrogen bonds, which above a critical temperature region, are severed through interaction with the alcohols, especially when the activity of the alcohol is raised by added salt; below the critical temperature the solvent forms gels with the keratin; and (3) disul$de bonds which are severed by reduction. Even though covalent disulfide bonds are present the low wet strength shows they are not present in sufficient number to provide the degree of network stabilization found in the natural keratins. It is entirely possible that the further study of these interactions in the protein fibers may lead to means of increasing the proportion of effective disulfide bonds. A practical application of the use of the combined solvent system alcohol-water-reducing agent-salt is seen in the solubilization of keratin directly from feathers (Fig. 21). Employing temperatures above the critical region (e.g., 70-80°C. for ethyl alcohol) the solvent extracts about 80% of the solid material as a maximum (the same as obtained by the use of detergents with reducing agents). When the solution is cooled below the critical region, or diluted with water, the keratin is
SYNTHETIC F I B E R S MADE FROM P R O T E I N S
345
either precipitated or forms gels with the solvent. The soluble protein is recovered by filtration or by pressing. I n this manner the alcoholwater systems, like the detcrgent systems discussed earlier, provide another procedure for the solubilization of feather keratin (100,101). A
faNETWORK ENERGY OF DRY FIBER
INCREASlNG SOLVATI
FIG. 21. Qualitative description of influence of solvent system alcohol-waterrequired to overcome network energy salt-reducing agent on mechanical work, fa, of fiber in dry condition, AH - T A S , as successively decreased by interaction with water, by interaction with ions, by influence of alcohols, and by reaction with reducing agents.
V. SUMMARY AND CONCLUSION Throughout this discussion, we have emphasized the practical aspects of synthetic fibers made from proteins. We have pointed out limitations in proteins, that some are more easily manipulated than others and that some give better fibers than others. I n all cases, however, the fibers lack sufficient network-stabilizing bonds, even after present curing treatments, to resist satisfactorily the effects of water. We have reviewed methods by which the stabilizing bonds can be studied and shown how it is possible through thermoelastic measurements on fibers made from feather keratin to identify several types of bonds that stabilize the fiber network. When considered for textile purposes, the fibers made from proteins are in demand in spite of present mechanical weaknesses. These fibers exhibit certain characteristics of specific interest. For example, they have pleasing appearance, desirable resilience, warmth, good affinity for dyes, and a favorable “hand.” Moreover, fabrics made of these fibers
346
HAROLD P. LUNDQREN
and also blends of these with other textile fibers have frequently a better eye appeal and draping quality than is present in other fabrics. Principally for these reasons, the fibers made from proteins have found a place in industry. Their range of utility is expected to broaden significantly when means for further stabilizing the fiber networks are available. When considered from the more general point of view, the study of protein fiber stability provides a means of gaining insight into the complex behavior of proteins under the many conditions to which they are exposed. Specifically, the proteins of foods respond favorably and unfavorably to heat and cold, to mechanical and osmotic forces, to desiccation and chemical influences; the proteins of muscles are subject to stimuli that determine their behavior in contraction, and the proteins of protoplasm are affected by environment, which determines cell behavior. I t is obvious that further knowledge of protein chain behavior under the various physical and chemical conditions will materially advance our understanding of these systems. For these investigations the study of the interactions and the folding and unfolding of protein chains as determined in the solid state will serve to supplement information gained from studies of proteins in solution. REFERENCES 1. Alfrey, T. (1948). Mechanical Behavior of High Polymers, in High Polymers. Vol. VI, Intersci. Pub., N. Y. 2. Alfrey, T.,and Doty, P. M. (1945). J . Applied. Phys. 16, 700. 3. Alfrey, T. (1947). J. Colloid Sci. 2, 99 4. Anson, M.L. (1939). J . Gen. Physiol. 23, 239. 5 . Anson, M.L., and Mirsky, A. E. (1932). Ibid. 16, 341. 6. Astbury, W.T.,and Woods, H. J. (1930). Nature 126, 913; (1933). Trans. Roy. SOC.London A232, 333. 7. Astbury, W.T.,Dickinson, S., and Bailey, K. (1935). Biochem. J . 29, 2351. 8. Astbury, W. T.,Bell, F., Gorter, E., and van Ormondt, J. (1938). Nature 142, 33. 9. Astbury, W.T. (1939). Ann. Rev. Biochem. 8, 113. 10. Astbury, W. T. (1940). J . Intern. SOC.Leather Trades’ Chemists 24, 69. 11. Astbury, W.T. (1942). J . Chem. Soc. 337. 12. Astbury, W.T. (1943). Advances in Enzyrnol. 3, 63. 13. Astbury, W. T. (1945). J . Teztile Inst. 36, P154. 14. Astbury, W.T. (1945). Nature 166, 501. 15. Astbury, W. T. (1947). Proc. Roy. SOC.London B134,303. 16. Bailey, K. (1944). Advances i n Protein Chem. 1, 289. 17. Baker, W.O.,and Fuller, C. 5. (1943). Ann. N . Y . Acad. Sci. 44,329. 18. Baker, W.O.,Fuller, C. S., and Pape, N. R. (1942). J . Am. Chem. SOC.64, 776. 19. Bear, R. 5. (1943). J . Am. Chem. SOC.66, 1784. 20. Bear, R. S. (1944). Ibid. 66, 1297.
SYNTHETIC FIBERS MADE FROM PROTEINS
347
21. Boyer, R. A. (1942). Modern Plastics lQ, (6) 48; (1940). Z n d . Eng. Chem. 54, 1549. 22. Brown, A. E., Gordon, W. G., Gall, E. C., and Jackson, R. W. (1944). Zbid. 36, 1171. 23. Bull, H. B. (1938). Cold Spring Harbor Symposia Quant. Biol. 6 , 140. 24. Bull, H. B. (1945). J . Am. Chem. SOC.67, 533. 25. Bull, H. B., and Gutmann, M. (1944). Zbid. 66, 1253. 26. Burnett, R. S., Roberts, E. J., and Parker, E. D. (1945). Znd. Eng. Chem. 37, 276. 27. Busse, W. T. (1932). J . Phys. Chem. 36, 2862. 28. Buzagh, A. von (1942). Kolloid-2. 101, 149. 29. Carothers, W. H. (1940). Collected Papers. Interscience, N. Y., 30. Chibnall, A. C., Bailey, K., and Astbury, W. T. (1937). British Pats. 467,704; 467,812. 31. Cohn, E. J., and Edsall, J. T. (1943). Proteins, Amino Acids and Peptides
Reinhold Pub. Corp., N. Y. 32. Coleman, D., and Howitt, F. 0. (1946). Symposium on Fibrous Proteins, p. 144. The Society of Dyers and Colourists; Chorley and Pickersgill Ltd., Leeds, England. (1945). Nature 166, 78. 33. Coleman, D., and Howitt, F. 0. (1947). Proc. Roy. SOC.London AlQO, 145. 34. Crisp, D. J. (1946). Symposium on Fibrous Proteins, p. 100. The Society of Dyers and Colourists; Chorley and Pickersgill Ltd. Leeds, England. 35. Croston, C. B., Evans, C. D., and Smith, A. K. (1945). Znd. Eng. Chem. 37, 1194. 36. Edsall, J. T. (1938). In the Chemistry of the Amino Acids and Proteins, p. 531. C. L. A. Schmidt, ed., C. C. Thomas, Springfield, Ill. 37. Evans, C. D., Croston, C. B., and Van Etten, C. (1947). Teztile Research J . 17, 562. 38. Eyring, H., and Stearn, A. E. (1939). Chem. Revs. 24, 253. 39. Eyring, H., and Halsey, G. (1946). Textile Research J . 16, 13. 40. Eyring, H., and Halsey, G. (1946). Ibid. 16, 124. 41. Eyring, H., and Halsey, G. (1947). J . CoZloid Sci. 2, 17. 42. Ferry, J. D. (1948). Advances in Prolein Chem. 4, 1. 43. Ferry, J. D. (1941). J . Biol. Chem. 138, 263. 44. Flory, P. J. (1945). J . Am. Chem. SOC.67, 2048. 45. Flory, P. J. (1947). J . Chem. Phys. 16, 397. 46. Flory, P. J., and Rehner, J., Jr. (1943). Zbid. 11, 512; (1943). Ann. N . Y . Acad. Sci. 44, 419. 47. Fo&, C. (1912). 2. Chem. u. Znd. Kolloide 10, 7. 48. Foltzer, J. (1924). Artificial Silk and Its Manufacture, p. 22. Pitmen and
Sons, London. 49. Fraenkel-Conrat, H., Brandon, B. A., and Olcott, H. S. (1947). J . Biol. Chem. 168, 99. 50. Fraenkel-Conrat, H., Cooper, M., and Olcott, H. S. (1945). J . Am. Chem. Soc. 67, 950. 51. Fraenkel-Conrat, H., and Olcott, H. S. (1946). Zbid. 68, 34. 52. Fraenkel-Conrat, H., and Olcott, H. S. (1948). J . Am. Chem. SOC.70, 2673. 53. Frankel, M., and Katchalski, E. (1943). Ibid. 66, 1670. 54. French, D., and Edsall, J. T. (1945). Advances in Protein Chem. 2 , 277.
348
HAROLD P. LUNDGREN
55. Freund, E. H., and Mark, €I. (1942). Rayon Textile Monthly 23, 515, 605. 56. Fuller, C. S., Baker, W. O., and Pape, N. R. (1940). J . Am. Chem. SOC.62, 3275. 57. Gerngross, O., Hermann, K., and Abits, W. (1930). Biochem. Z . 928, 409. 58. Gibbs, W. (1875). Trans. Conn. Acad. 3, 108; quoted in The Scientific Papers of J. Willard Gibbs Vol. I (1906), p. 53. Longmans, Green, and Co., Ifindon, p. 53. 59. Glasstone, S., Laidler, K. J., and Eyring, H. (1941). The Theory of Rate Processes. McGraw-Hill, New York. 60. Go, Y., and Noguti, Z. (1939). J . Chem. SOC.Japan 60, 427. 61. Gorter, E. (1937). Trans. Faraday SOC.33, 1125. 62. Graves, G. D. (1942). U. S. Pat. 2,289,775, July 14 to du Pont. 63. Gustavson, K. H. (1946). J . A m . Leather Chemists' Assoc. 41, 47. 64. Guth, E., and Mark, H. (1934). Monatsh. 66, 93. 65. Halsey, G. J. (1947). J . Applied Phys. 18, 1072. 66. Halsey, G., and Eyring, H. (1945). Textile Research J . 16, 451. 67. Halsey, G., and Eyring, H. (1946). Zbid. 16,329. 68. Halsey, G., White, H. J., Jr., and Eyring, 13. (1945). Zbid. 16,295. 69. Harris, M. (1945). Advancing Fronts in Chemistry, Vol. I, p. 165. Reinhold Pub. Corp., N. Y. 70. Harris, M., and Brown, A. E. (1946). Symposium on Fibrous Proteins, p. 203. The Society of Dyers and Colourists; Chorley and Pickersgill Ltd. Leeds, England. (Same article in Teztile Research J . 17,323.) 71. Harris, M., Misell, L. R., and Fourt, L. (1942). Znd. Eng. Chem. 34, 833. 72. Harrison, W. (1938). Am. Dyestufl ReptT. 27, 393. 73. Herriott, R. M. (1947). Advances i n Protein Chem. 3, 169. 74. Hooke, Robert (1665). Microgrnphia: or, Some Physiological Descriptions of Minute Bodies Made by Magnifying Glasses. Printed by Martyn and Allestry, printers t o the Royal Society, London. 75. Howitt, F. 0. (1946). Bibliography of the Technical Literature on Silk, p. 31. Hutchinson, London. 76. Huggins, M. L. (1943). Chem. Revs. 32, 195. 77. Hughes, A. H., and Rideal, E. K. (1932). Proc. Roy. Soc. London A137, 62. 78. Jansen, E. F., Waisbrot, S. W., and Rietz, E. (1944). Znd. Eng. Chem., Anal. Ed. 16, 523. 79. Jirgensons, B. (1942). J. prakt. Chem. 160, 120; (1946). J. Colloid Sci.1,539. 80. Karrer, E. (1932). Phys. Revs. 39, 857. 81. Katr, S., Halsey, G., and Eyring, H. (1946). Trztile Research J . 16, 378. 82. Kirkwood, J. G. (1934). J . Chem. Phys. 2 , 351. 83. Kraemer, E. 0. (1943). The Chemistry of Large Molecules, in Frontiers in Chemistry, Vol. I. p. 73. Intersci. Pub. N. Y. 84. Kratky, O., and Sekora, A. (1943). J . makromol. Cheni. 1, 113. 85. Kuhn, W. (1934). Kolloid-Z. 68, 2; (1939). Kolloid-2. 87, 3. 86. Leaderman, H. (1943). Elastic and Creep Properties of Filamentous Materials. Textile Foundation, Washington, D. C. 87. Leuchs, H., and Geiger, W. (1908). BeT. 41, 1721. 88. Leuchs, H., and Manasse, W. (1907). Zbid. 40, 3235. 89. Lloyd, D. J., and Garrod, M. (1946). Symposium on Fibrous Proteins, p. 24. The Society of Dyers and Colourists; Chorley and Pickersgill Ltd. Leeds, England. 90. Lowe, B., and Stewart, G . F. (1947). Food Technol. 1, 30.
349
SYNTHETIC FIBERS MADE FROM PROTEINS
91. Lundgren, €1. P. (1936). Nature 138, 122. 92. Lundgren, H.P., and Williams, J. W. (1939). J . Phys. Chem. 43, 989. 93. Lundgren, H. P. (1941). J . Am. Chem. Sac. 63, 2854. 94. Lundgren, H. P., Elam, D. W., and O’Connell, R. A. (1943). J . B i d . Chem. 149, 183. 95. Lundgren, H.P.,and O’Connell, R. A. (1944). Ind. Eng. Chem. 36, 370. 96. Tmdgren, IT. P. (1945). Teztile Research J . 16, 335. 97. Lundgren, H.P. (1946). Chemurgic Digest 6, 113. 98. Lundgren, H.P. (1947). U. 8. Pat. 2,425,550;2,459,708. 99. Lundgren, H.P. (1947). Food Technol. 1, 70. 100. Lundgren, H.P., Stein, A. M., Koorn, V. M., and O’Connell, R. A. (1948). J . Phys. & Colloid Chem. 62, 180. 101. Lundgren, €I. P. (1948). U . S. Egg Poultry Mag. 64, 8. 102. MacArthur, I. (1940). Wool Rev. Jan. page 9. (1943). Nature 162, 38. 103. MacArthur, I. (1946). Symposium on Fibrous Proteins. The Society of Dyers and Colourists; Chorley and Pickersgill Ltd. Leeds, England. 104. Mark, H. (1940). J . Phys. Chem. 44, 764. 105. Mark, H. (1942). Science i n Progress Third Series, 246. 106. M&k, H. (1942). Znd. Eng. Chem. 34,449, 1343. 107. Mauersberger, H. R.,ed. (1947). Matthews’ Textile Fibers. J. Wiley and Sons, N. Y. 108. Maxwell, J. C. (1868). Phil. Mag. 36, 134. 109. McMeekin, T.L. (1943). ASTM Bull. No. 126, 19. 110. McMeekin, ‘l’. I,., Reid, T. S., Warner, R. c., and Jackson, H. W. (1945). Ind. Eng. Chem. 37, 685. 111. McMeekin, T.L., and Warner, R. c. (1946). Ann. Rev. Biochem 16, 119. 112. Mecheels, 0. (1938). Melliand Textilber. 19, 579. 113. Mecheels, O.,and Essig, K.A. (1942). Zellwolle, Kunstseide, Seide 46, 167. 114. Mercer, E. H.,and Rees, A. L. G. (1946). Australian J . Ezptl. B i d . Med. Sci. 24, 175. 115. Meridith, R.J. (1945). J . Teztile Inst. 36, T147. 116. Merrifield, A. L., and Pomes, A. F. (1946). Teztile Research J . 16, 369. 117. Meyer, K. H., and Mark, H. (1928). Ber. 61, 1932. 118. Meyer, K. H., and Mark, H. (1930). Der Aufbau der hochpolymeren organishen Naturstoffe. Akademische Verlagsgesellschaft, Leipzig. 119. Meyer, K. H.,Von Susich, G., and Valkb, E. (1932). Kolloid-2. 69, 208. 120. Meyer, K. H.,and Go, Y. (1934). Helv. Chim. Acta 17, 1488. 121. Meyer, K. H.,and Ferri, C. (1936). Arch. ges. Physiol. (PjEzZgers)238, 78. 121a. Meyer, K. H.,and Picken, L. (1937). Proc. Roy. SOC.London Bl24, 29. 122. Meyer, K. H.,and Jeannerat, J. (1939). Helv. Chim. Acta 22, 22. 123. Meyer, K. H.,Fuld, M., and Klemm, 0. (1940). Ibid. 23, 1441. 124. Meyer, K. H. (1942). Natural and Synthetic High Polymers, Vol. IV. Intersci. Pub., N. Y. 125. Millar, A. (1899). J . Sac. Chem. Ind. 18, 16. 126. Mirsky, A. E.,and Pauling, L. (1936). Proc. Natl. Acad. Sci. U . S.22, 439. 127. Mirsky, A. E.,and Pollister, A. W. (19413). J. Gen. Physiol. 30, 117. 128. Neurath, H. (1942). Chem. Revs. 30, 357. 129. Neurath, I€. (1943). J . A m . Chem. Sac. 66, 2039. 130. Neurath, H.,Greenstein, J. P., Putnam, F. W., and Erickson, J. 0. (1944). Chem. Revs. 34, 157. 131. Nutting, G.C., Senti, F. R., and Copley, M. J. (1944). Science 99, (2573)328. d
350
HAROLD P. LUNDQREN
132. Nutting, G.C., Halwer, M., Copley, M. J., and Senti, F. R. (1946). Teztile Research J . 16, 599. 133. Nutting, G. C., and Borasky, R. (1948). J . Am. Leather Chemists' Assoc. 43, 96. 134. Olcott, H. S.,and Fraenkel-Conrat, H. (1947). Chem. Revs. 41,151. 135. Ott, E. (1943). Frontiers in Chemistry, Vol. I, p. 73. Intersci. Pub., N. Y. 136. Ott, E. (1943). High Polymers: Cellulose and Cellulose Derivatives, Vol V. Intersci. Pub., N. Y. 137. Palmer, K. J., and Calvin, J. A. (1943). J . Am. Chem. SOC.66, 2187. 138. Peterson, It. F., Caldwell, T. D., Hipp, N. J., Hellbach, R., and Jackson, R. W. (1945). Ind. Eng. Chem. 37, 492. 139. Pinner, S. H. (1946). Brit. Plastics 18, 313, 353. 140. Pollister, A. W.,and Mirsky, A. E. (1946). J . Gen. Physiol. 30, 101. 141. Ramsden, W. (1938). Nature 142, 1120. 142. Scarth, G.W. (1927). Protoplasma 2, 189. 143. Scatchard, C.,and Kirkwood, J. G. (1932). Physik. 2.33, 297. 144. Schmitt, F. 0. (1944). Advances i n Protein Chem. 1, 25. 145. Schmitt, F. O.,Bear, R. S., Hall, C. E., and Jakus, M. A. (1947). Ann. N . Y . Acad. Sci. 47,799. 146. Schofield, R. K.,and Scott Blair, G. W. (1937). Proc. Roy. SOC.London A160, 87. 147. Scott Blair, G . W. (1939). Cereal Chem. 16, 707. 148. Seifriz, W., and Plowe, J. Q. (1931). J . Rheol. 2, 263. 149. Seifriz, W.,ed. (1942). A Symposium on the Structure of Protoplasm. Iowa State Coll. Press, Ames, Iowa. 150. Senti, F. R. (1947). Am. DyestuiJ Reptr. 36, P230. 151. Senti, F. It., Copley, M. J., and Nutting, C. C. (1945). J . I'hys. Chem. 49,192. 152. Senti, F. R.,Eddy, C. R., and Nutting, G. C. (1943). J. Am. Chem. Soc. 66, 2473. 153. Sherman, J. V., and Sherman, S. L. (1946). The New Fibers. Van Nostrand Co., N. Y. 154. Smith, H. Dew. (1944). Teztile Fibers, an Engineering Approach to Their Properties and Utilization, Edgar Marburg Lecture, Am. SOC.Testing Materials, Proc. 44, 1. 155. Sookne, A. M., and Harris, M. (1945). Ind. Eng. Chem. 37, 478. 156. Speakman, J. B. (1928). Proc. Roy. SOC.London B103, 377. 157. Stein, R.,Halsey, G., and Eyring, H. (1946). 'I'eztile Research J . 16, 53. 158. Swallen, L. C. (1941). Ind. Eng. Chem. 33, 394. 159. Szent-GyGrgyi, A. (1947). Chemistry of Muscular Contraction. Academic Press, N. Y. 160. Thomson, R. H. K. (1946). Symposium on Fibrous Proteins, p. 173. The Society of Dyers and Colourists: Chorley and Pickersgill Ltd. Leeds, England. 161. Tobolsky, A., Powell, R. E., and Eyring, H. (1943). I n Frontiers in Chemistry, Vol. I, p. 125. Intersci. Pub.,". Y. 162. Traill, D. (1945). Chemietry & Zndudry No. 8,58. 163. Traill, D. (1945). J. Soe. Dyers Colourists 61, 150. 184. Traill, D. (1946). J. Teztile Inst. 37, P295. 165. Wall, F. T. (1942). J. Chem. Phys. 10, 485. 166. Ward, W. H.,High, L. M., and Lundgren, H. P. (1946). J . Polvmer Research 1, 22.
SYNTHETIC FIBERS MADE FROM PROTEINS
351
167. Waugh, I). F. (1944). J. A m . Chem. SOC.66, 663. 168. Waugh, D. F., Smith, J., and Fearing, D. F. (1948). Proc. SOC.Ezptl. B i d . M r d . 7, 131. 169. Weber, H. H . (1934). Arch. ges. Physiol. (Pflilger’s)236, 205. 170. Weimarn, P. P. von. (1926). Can. Chem. Met. 10, 227. (1927). Z n d . Eng. Chem. 19, 109. (1932). In Colloid Chemistry, Vol. IV, p. 363. Chemical Catalogue Co., Reinhold Pub. Corp., N. Y. 171. Whittier, E. O., and Gould, S. P. U. S. Pat. 2,140,274 (1938); 2,204,535 (1940). 172. Wohlisch, E. (1927). 2.B i d . 86, 325. 173. Woods, H. J. (1946). J. Colloid Sci. 1, 407. 174. Woodward, R. B., and Schramm, C. 13. (1947). J . A m . Chem. Soc. 69, 1551.
This Page Intentionally Left Blank
Some Protein-Chemical Aspects of Tanning Processes
BY K. H. GUSTAVSON Swedieh Tanning Research Inutituts, Stockholm, Sweden
CONTENTS I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. The Skin . . . . . . . . . . . . .
paqc 354
I I. Chemistry of Collagen. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6. Coordinative Reactions..
........
. . . . . . . . . . . . . . . 368 . . . . . . . . . . . . . . .369 . . . . . . . . . . . . . . . . . . 378
1. Structure of Basic Chromic Salts of Importance in Tanning.. 2. Factors Governing the Tanning Effect.. . . . . . . .
. . . . . . 379
a. Nature of the Anion... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
381
b. Basicity of Chromic Salt.. .................... c. Concentration of Chromium Salt.. . . . . . . . d. Neutral Salt Effect.. . . . . . . . . . . . . . . . . . . . . . . . . e. Hydrogen Ion Concentration.. f. Influence of Previous History of 3. Nature of Chrome-Collagen Compound.. ........................ 388 4. Some Important Properties of Chrome Leather of Theoretical Interest 392
.
.......................
396
c. Degree of Stabilization. ....
VIII. Tanning Power of Aldehydes.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Tanning with Formaldehyde. . . . . . . . . . . . . . . . . . . . . . . . . . . . . a. General Aspects of the Reaction.. ..........................
353
406
354
K. H. QUBTAVSON b. Participation of Lysine Groups. . . . . . . . . . .
c. Reaction of Arginine Gtoups.. . . . . . . . . . . . . . . . d. Participation of Peptide and Other Groups. . . . . . e. Influence of Solvent. . . . . . . . . . . . . . . . . f. Ewald Reaction., . . . . . . . . . . . . . . . . . . . . . . . . . IX. Quinone Tannage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . X. General Comments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Page 406 . 407
.
402)
411 41 1 . 411 413 415
I. INTRODUCTION At an early stage of civilization, the art of tanning was discovered. By tanning skin is converted into a material possessing the valuable properties of resisting water and putrefaction and of remaining soft and pliable. By incorporation of tanning agents within the hide structure, suitably pretreated and conditioned, leather is formed. A great many substances of different types and chemical compositions possess tanning potency, e.g., vegetable tannins, unsaturated oils, aldehydes, and, preeminently, basic chromium salts. Theoretically, tanning connotes stabilization of proteins by their irreversible combination with tanning agents. Practically, tanning may be defined as a process making the putrescible proteins of hide microbiologically and hydrothermally resistant, preserving certain properties of the original fiber structure, and, further, imparting certain desirable mechanical properties to the hide. The main operations in the preparation of hide for tanning are briefly as follows. The first step is the removal of salt and other extraneous matter from the salted hide by soaking it in water, restoring the natural water content of the fresh hide. In the subsequent liming process, the epidermal and subcutaneous layers are removed and certain accessory proteins dissolved or modified. The final result is the leather-forming pelt, mainly containing collagen, which after further conditioning is made neutral and simultaneously brought to the state of minimum swelling. The pelt is then ready for the tanning proper. 1. The Skin Since animal skin is the basis of leather, some knowledge of its microstructure is essential for an appreciation of the complicated reactions involved in tanning. The skin protein chiefly concerned in ordinary tanning is collagen, which forms the main part of the skin undergoing tanning, the corium or dermis. The major part of the raw material is of bovine origin. The skin of mammals (211) consists of three distinctive layers: ( I ) epidermis, ( 2 ) corium or skin substance, and ( 3 ) subcutaneous tissue. The epidermis and subcutaneous tissue are removed in the preparatory processes. The remaining corium consists of bundles of collagen fibers interwoven in all directions. The collagen fibers of mammalian hide are arranged in bundles. Fibrils make up the fiber, which appears to be interwoven in all directions. In bovine skin the fiber
PROTEIN-CHEMICAL ASPECTS OF TANNINQ
355
bundles generally are 50-100 p thick, the fibers about 25 p, and the fibrils 1 p or less in diameter. The channels between the bundles are fairly wide, averaging a few microns. In the pretreatmerfts for conditioning the hide for tannage, the space may be narrowed or widened according to the properties desired. The removal of interfibrillar proteins, albumins, globulins, and mucoids in the pretreatment8 tends to increase the permeability of the skin. By swelling and by introduction of large molecules into thc structure in the form of tanning agents the size of the channels is also altered. The influence of the internal structure of skin is important in tanning processes, topochemical factors being prominent. In practical tanning, the additional complication of uneven distribution of the tanning agent throughout the interior of the hide is caused by the polydispersity of the solutions of many important tanning agents, e.g., the vegetable tannins. This complication may be largely eliminated in laboratory studies by increased subdivision of the substrate, the collagen being used in the form of hide powder or single fibrr bundles. Single bundles are preferably applied in studies of theoretical problems by means of physicochemical and mechanical methods, since the macroweave s t h c t u r e is thus eliminated. The implication of the twophase systems with accompanying difficulty of attaining true equilibria must be borne in mind.
2. General Structure of Fibrous Proteins
The classification of the proteins into two maiu groups, the globular and the fibrous, is made on the basis of the arrangement of the protein chains, the peptide chains of fibrous proteins being more or less extended and oriented in parallel, whereas the units of the globular proteins are folded, no preferred direction of orientation being evident. The properties and reactivity of fibrous proteins are primarily governed by their chemical composition, with due regard given t o the organization of the protein structures. This fact is clearly brought out by comparing collagen and its secondary product, gelatin, which differ markedly with respect to their physical properties although their compositions are identical. The different degrees of organization of the peptide chains and the higher units account for the differences between collagen and gelatin. The collagen units are arranged in parallel and stabilized by valency forces between adjacent peptide chains and micelles (cross links). In the conversion of collagen into gelatin, some of these cross links are broken, resulting in shortening and disorganization of the protein chains. The aspect of organization of proteins was early recognized by Jordan Lloyd (112). Attention was called t o the interrelationship of chemical composition, degree of stabilization of the units, and the physicochemical reactivity of proteins, particularly the degree of swelling (water imbibition) a t the p H of maximum swelling and at the p H of minimum swelling (isoclectric range). The data of Table I (112) illustrate this point for some typical fibrous proteins (silk fibroin, keratin, collagen, gelatin, and
356
K . H. QUSTAVSON
muscle protein). The acid-binding capacities of the proteins mentioned are included in order t o show that the degree of swelling by acids is not a simple function of the amount of acid-binding protein groups. It is TABLE I Organization of Proteins and Degree of Water Imbibition (Swelling) -.
Protein
Silk fibroin Keratin Collagen Gelatin Muscle protein
-
Total Total bases, dicarboxylic acids, % 1 11 15 15 17
15 17 17
20
bfilliequivalents HCI fixed per g. protein 0.1 0.8 0.9 0.9 1 .o
-
Degree of swelling, % water taken up PH 5 20
22
30
35 490
260
-
1300 300
6500 1400
evident that fibrous proteins, mainly built up from amino acids with nonpolar groups, possess low affinity for water, whereas proteins containing numerous polar groups take up water readily. However, collagen and keratin, with practically the same degree of polarity and acid-binding capacity, show great differences in water uptake and degree of swelling. The different internal organizations of these proteins explains their behavior. Keratin contains the covalent cystine bridge, which resists the action of acids, maintaining the rigid structure in spite of the electrostatic repulsions of those protein groups which are positively charged in acid solutions. In collagen, on the other hand, the negatively charged R groups, t o a large extent internally compensated by electropositive k groups, forming saltlike cross links, are discharged and the cross links broken. By this disorganization of the chains, some of the second type of cohesive forces of the collagen structure, the coordinate cross links on the peptide groups (hydrogen bonds), are probably ruptured. Water is added to the protein by polar forces and by association on the coordinate bci of the peptide groups set free by the contraction of the main chains. The effective stabilization of keratin is due t o the disulfide bridges. This is evident from the fact that, on breaking the covalent bridge, the resulting products, the keratoses, swell in acid solution t o a degree comparable to that of collagen.
11. CHEMISTRY OF COLLAQEN 1. General Aspects
The amino acid composition of collagen (and gelatin) is more completely known than the compositions of most proteins. The early work
357
PROTEIN-CHEMICAL ASPECTS OF TANNING
was chiefly carried out on gelatin (10,28). In recent years determinations of the amino acid composition of collagen of known previous history and high degree of purity have been made by specific methods (24). The present discussion of the behavior of collagen as a dipolar structure was originally based upon the data of the amino acid composition of native collagen given by Chibnall in 1946 (24). Since this chapter was written, revised values obtained by the Cambridge School of Biochemistry in collaboration with the British Leather Manufacturers' Research Association have been published. Bowes and Kenten (14a) have tabulated the most probable values of the amino acid composition of native collagen of ox hide which had received no alkaline or enzymic treatment (Table 11). The total nitrogen content was 18.6%. The total nitrogen and nmide contents of this pure preparation of TABLE I1 Amino Acid Composition of Native Unlimed Collagen
Amino acid Amino nitrogen Glycirie Alanine Leucine Isoleucine Valine Phenylalanine Tyrosine Tryptophan Serine Threonine Cystine Methionine Proline Hydroxyproline Lysine Hydroxylysine Arginine Histidine Aspartic acid Glutamic acid Amide nitrogen Total found
1
0
b
N as % G. residues protein N per 100 g.
Mmol. per g.
Assumed Apparent number of minimum residues
mol. wt.
2.5 26.3 8.0
0.33 3.50 1.06
-
-
19.9 7.6
136 41
38,880 38,580
3.2
4.8
0.42
17
39,950
2.2 1.9 0.6 0.0 2.5 1.5 0.0 0.4 9.9 8.Oa 4.7 1.2 15.3 1.2 3.6 5.8 3.5 99.8
2.9 3.7 1.3 0.0 2.7 2.0 0.0 0.7 12.7 12.1 4 .O 1.1 7.9 0.7 5.5 10.0
0.29 0.25 0.08 0.00 0.33 0.20 0.00 0.05 1.32 1.07 0.31 0.08 0.51 0.05 0.47 0.77 0.47 10.76'
11 10 3
37,620 39,600 37,620
13
-
Determined on gelatin. Excluding amide nitrogen.
-
-
99.6*
8
39,130 40,160
-
-
2 51 41 12 3 20 2 19 30 18 419
37,640 38,760 38,580 38,400 37,680 39,380 37,640 39,750 38,910 38,740 38,730 (mean)
358
K. H. GUSTAVSON
native collagen are higher than those previously reported. Since these figures account for over 99% of the total nitrogen of collagen, it appears as the analysis of collagen is virtually complete. The amino acid composition of gelatin is nearly the same as that of collagen; the main difference is the low content of amide nitrogen of gelatin (0.5%). The new values of dicarboxylic acids were obtained by microbiological assays and the remainder by the chromatographic technique. In comparing these data t o the older values, the main differences are found in the contents of glutamic and aspartic acids, which are nearly twice as large as the earlier values. This will radically affect the proportion of basic (cationic) t o acidic (anionic) groups changing the ratio from 1.5:1 to 1 : 1.3. The adjusted figures of the dicarboxylic acids, and the resulting ratio of basic to acidic groups with due consideration of the deamidation of collagen taking place in the alkali treatment, remove the paradox that gelatin and limed collagen are isoelectric in the acid p H range, although according t o the old values they should contain a large excess of cationic groups. On the other hand the new data make the discrepancy between the figures obtained for basebinding capacity of collagen and its content of free carboxyl groups still wider. The minimum molecular weight calculated from the figures of Table I1 is about 39,000 and the average residue weight as 92.6, which figure exactly agrees with thc value obtained by Chibnall from the nitrogen distribution (23). Collagen contains a well-balanccd proportion of posit ivcly and negatively charged polar groups, securing a fair degree of ionic reactivity. This function is particularly important for its behavior in the preliminary processes of tanning and in the tanning processes proper. Characteristic for the collagen group of proteins are the large content of nonpolar amino acids, glycine and alanine, composing about a third of the total and, above all, the prominence of prolinc and hydroxyproline and the paucity of aromatic residues. The large number of prolines, built into the chains with the formation of -("()-If= links, introduces a rcguliLr intcrruplinks formed hy the remainder of the tion of the regular -CO-NfIamino acid residues. I t is also interesting to note that hydroxyl groups, mainly contributed by hydroxyproline and serine, form nearly one-half the total number of polar groups. The function of these groups, containing residual negative charge on oxygen (183), in the reactions of tanning agents with collagen has not yet been considered. Probably they are of importance as electrostatic centers and as hydrogen-bonding loci in coordination of vegetable tannins and chromic salts in these tannages. From investigations of collagen by X ray and electron microscope the
PROTEIN-CHEMICAL ASPECTS OF TANNING
359
presence of protein chains aligned in parallel has been experimentally proved; they are nearly completely stretched. The X-ray analysis shows a meridional spacing of about 2.9 b, corresponding to the average length of an amino acid unit in the length of the chain (3). Since the length of one residue of a fully stretched protein chain (silk fibroin) is 3.5 b (156), it is evident that the main chain of collagen is slightly contracted. This curling of the chains is probably caused by the presence of large proline and hydroxyproline residues and is thus of steric nature (3). The long spacing of about 640 b reported by Rear (6) may be taken as a measure of the over-all length of the molecule. Two equatorial spacings are of special significance: a regular lateral spacing of about 11-15 b, according to the degree of hydration of the fiber, and a less regular one of about 4.5 A, corresponding to the distance between side chains and between backbones, respectively (2). In the electron microscope collagen shows a banded structure (98,162, 179,180). Two distinct phases of collagen, one fully stretched and one with contracted chains, are indicated. The average spacing from one dark band to the next is 645 A, closely corresponding to the long meridional spacing found by X-ray analysis (6). The unexpected discovery of the large extensibility of collagen fibrils (179), not shown by bundles of fibers, is in line with the presrnce of folded sections in the chains. The possibility of the presence of different chains, each with a certain amino acid pattern associated by means of cross links-the molecular grid type of structure, a s indicated in insulin and other proteins according to Chibnall and coworkers (23)-is also of interest in this connection. The following presentation will be based upon the established concept that collagen contains parallel protein chains of the general form: -C €1--N H-CO-C H-
I
Ri
I
R*
The cohesion of the higher units is due to the attraction of oppositely charged side chains and to hydrogen bonds acting between prptide groups of adjacent chains. I n Astbury’s (3) model of collagen, obtained by interpretation of X-ray data and from figures of amino acid composition, the molecule is considered to be composed of parallel chains, practically fully extended, in which glycine rrsidues represent every third residue, proline and hydroxyproline anothcr third, the rest of the residues accounting for the remaining third. The presence of imino links in every third residue (proline and hydroxyproline being built into chain forming =N-COlinks) is considered to limit the rotational possibility about these atoms, leading to a spiral or slightly folded configuration of tlie chains a t these points. The bulky proline groups arc considered to be
360
K. H. OUSTAVSON
held on one side of the chain and the remaining R groups on the opposite side (3). A general concept may also be based upon Huggins’ model of collagen (109,168) with a spiral configuration of the main chain, the side groups projecting alternately above and below the main chain. The chains are considered to be grouped into layers, the individual chains held together mainly by means of hydrogen bonds, supplemented by electrostatic attraction of directed and undirected valency forces. Astbury’s (2) deductions led to the concept of periodicity of the protein chains. This hypothesis was further advanced by Bergmann and Niemann (11). The isolation of a tripeptide lysylprolylglycine from gelatin by Grassmann and Riederle ( 5 2 ) strengthened the periodicity claim. However, Chibnall and his school (23) have proved the limitations of the simple whole number rule applied to analytical data of proteins of known composition. The weakness of the periodic concept from the statistical and probability point of view has also been demonstrated (163) and further data obtained by the chromographic method of Martin and Synge (144,145) do not support the original postulate of periodicity. However, a certain sequence of residues seems probable. For a more precise theory information is needed regarding the general sequence of the amino acids in the subunits, the location of the reactive R groups and their spatial relation. However, these fine structural details are unknown. Additional evidence for the lattice structure of collagen is furnished by the birefringence of the fibers. The collagen fiber shows positive double refraction (119). By the incorporation of various tanning agents into the structure, the degree and sign of the birefringence are in many instances altered (121). Some tanning agents, e.g., the vegetable tannins may cause a shift from positive to negative values. Marriott (142) has found some relationship between the value of the birefringence of vegetabletanned collagen fibers (leather) and their abrasional resistance. Further, Kiintzel (121) asserts that it is possible from the study of the birefringence of tanned fibers to prove the occurrence of intra- or intermicellar types of combination of tanning agents with collagen. Thus, Kiintzel concludes, from the inversion in sign of the birefringence of collagen upon vegetable tannage, t h at the tannins are incorporated intramieellarly (121,126). The type of micellar reactions is important in tannage, particularly in chrome tanning. Electron microscope studies may give needed information (see, e.g., 122 and 162).
2. Stabilization of Collagen The cohesive forces of mammalian collagen, inter- and intramicellar, are generally considered to be of two main types, directed valency and undirected valency. The former consists of: (1) electrovalency, located
PROTEIN-CHEMICAL ASPECTS OF TANNINQ
361
a t polar R groups of the chains with opposite charge, forming a saltlike cross link, and ( 2 ) coordinate valency between the carbonyl groups of one chain and the imino group of an adjacent chain’s peptide group (hydrogen bond). The undirected valency is of the nature of electrostatic attraction or van der Waals forces. An evaluation of the relative importance of the two types of directed valency linkages may be obtained by measuring the hydrothermal stability of collagen (its shrinkage temperature) (68), after treatment with agents which are specific for the groups involved. Thus, P-naphthalenesulfonic acid will inactivate the basic groups completely, thereby eliminating the saltlike cross links, without swelling collagen and without interfering with the coordinate activity of the peptide groups (71). The change of shrinkage temperature caused by such inactivation may be applied for the evaluation of the degree of cohesion of the collagen lattice due to electrovalent interaction of polar R groups. The shrinkage temperature of mammalian collagen (65-68”C.) is decreased 10-12°C. by this treatment (71). Further, by pretreatment of bovine collagen in solutions of agents with a specific action on peptide groups and their coordination centra, as, for example, concentrated solutions of urea, shrinkage occurs a t room temperature. This means a lowering of the shrinkage temperature of some 40°C. (71,llG). Accordingly, it is evident that the main stabilization forces of mammalian collagen belong to the type of hydrogen bonds. The data indicate that they are responsible for at least three-fourths of the internal cohesion, the rest being supplied by the saltlike cross links. (See further Section 7.) The tensile strength of the fiber bundles apparently is not simply a function of the strength of the bonds of the molecular units (93), the cohesion of the fibers being an additional factor. Hence, a direct comparison of the shrinkage temperature and the mechanical strength of fibers appears not to be permissible. (Cf. 93 and 101.) In fibers in equilibrium with water, its dipole nature will probably influence the internal cohesion, especially the strength of attraction between oppositely charged R groups, affecting the strength of the wet and dry fibers. The distance between the ionic groups is also widened by hydration of collagen (1 15). The fiber strength generally decreases with increased moisture content. Exceedingly drastic drying may lead to markedly increased strength, probably caused by conversion of ionic cross links to covalent links. Some water of constitution (“bound water”), which amounts to about 20% of the weight of collagen (107), must be lost in drying collagen to 12% water content. It is not known if this water is withdrawn from the polar groups or from the peptide groups. It has been assumed by some workers (108) that the water molecule forms the link between
362
K.
n.
GUSTAVSON
internally compensated peptide groups as
\
/
N.H.*-O.H..OC the / H additional water taken up by collagen being coordinated on the molecule of water, functioning as a vehicle for cross linking. Data on the strength of skin pieces cannot be used for a n evaluation of the influence of various types of processing, e.g., tanning, on the strength of the fibers, since the tensile strength measured on a strip of skin really gives the strength of the fiber weave. This aspect of the problem, with due consideration of the effect of various tanning agents on ~ single fibers, was particularly stressed long ago by Jovanovits ( 1 1 3 14), Chernov (22), and Grassmann (50), and recently by Highberger (101). I t is noteworthy that the tensile strength on a collagen basis is greater for vegetable-tanned skin than for chrome-.tanned skin. However, the strength of chrome-tanned .fibers (about 30 kg./mm.*) is about thrice that of vegetable-tanned jibers on the same weight of original collagen (22). In the former instance the macro structure (weave) is the main factor, whereas the strength of the fibers represents the molecular structure of the tanned collagen. The greater strength of the chrome-tanned fiber is in line with the prevailing concept of the degrees of efficiency of the cross linking in the two tannages. \I
3. Binding of Hydrogen and Hydroxyl Ions The maximum acid-binding capacity of collagen (HCl) is reached a t equilibrium pII of about 1 and amounts to 0.9 milliequivalent hydrogen ion per gram collagen, in good agreement with the number of equivalents of basic groups (8). The base-binding capacity is less than half the value expected from considerations of the newest figures of free dicarboxylic acids of collagen. The values are, for collagen, 0.4-0.5 milliequivalent hydroxyl ion per gram protein (8,127,129). Earlier discussions of the acid- and base-binding capacity of collagen have been based on a proportion of 1.5 to 2 basic groups per acidic group (63). The reason for tfhis large difference in the base-binding values of collagen reported (8,127,129), one one hand, and the theoretical values calculated from the content of free dicarboxylic acid, on the other hand, is not clear. The detcrminat,ion of the maximum fixation of alkali by collagen from sodium hydroxide has been carried out in the p H interval 12-13. This determination is subject to large errors. Some errors are inherent in the method and are due to the presence of carbon dioxide and the difficulty of pH determinations a t such high alkalinity. Other errors are due to chemical changes in the collagen at prolonged interaction of alkali such as formation of soluble degradation products and alteration of
PROTEIN-CHEMICAL ASPECTS OF TANNING
363
collagen. Recent figures of the binding capacity of collagen (isoelectric calfskin) for barium hydroxide, a t equilibrium p H values of 12-13, obtained by means of gravimetric analysis of the substrate in equilibrium with the hydroxide solution, are in the range 0 . 5 4 . 6 milliequivalent Ba++ per gram collagen (94). These figures are the final ones, after due correction for sorbed solution in the pressed substrate and the water of hydration of collagen. Another possibility is that collagen requires still higher concentrations of hydroxyl ions than those heretofore employed for attainment of maximum base fixation. The base-binding capacity of gelatin is considerably greater, showing values of 0.7-0.9 milliequivalent hydroxyl ion per gram gelatin (25). The difference of 0 . 3 4 . 4 milliequivalent hydroxyl ion between gelatin and collagen can only in small part be accounted for by the deamidation taking place in the conversion of collagen into gelatin. The possibility that some part of the dicarboxylic groups of collagen is inactivated (for instance by interaction with arginine residues) seems worthy of consideration as a n eventuality if the present low values of the base-binding capacity of collagen prove to be correct. The alkali-binding power of collagen is greatly influenced by the degree of pretreatment of the hide with alkali in the liming process, which leads t o partial deamidation (9,100) and also to rupture of coordinated cross links (hydrogen bonds) and dislocation of part of the saltlike cross bonds, links. Also splitting of peptide links, including the -C!O-N= seems to occur (126)) resulting in a slight increase of hydrogen-ion- and hydroxyl-ion-binding side chains. I n systems for determination of the acid- and base-binding curves of collagen, it is of the utmost importance that the swelling of the protein be repressed by the presence of neutral salts such as NaCl and Na2S04. Otherwise, the different ionic distribution between the solid phase and the external solution will give rise to large errors, as will also the influence of the swelling of the substrate of the sorption of solvent and solute which may result in negative adsorption. This important aspect has been emphasized by Steinhardt and coworkers (189)) references being given t o their publications. The present discussion of the reaction of collagen with hydrogen ions was restricted t o the ideal type of a strong acid forming a practically completely ionized salt and thus avoiding the complications due to fixation of the anion by the protein. With more complicated acids the affinity of the anion must be included and considered, which will lead to greater fixation of hydrogen ions in the higher pH range up to the isoelectric point. This phase of the problem of acid fixation by proteins has been lucidly treated by Steinhardt and collaborators (189,190). I t has for a
364
E. H. GU8TAV80N
long while been realized by investigators of tanning processes that the anion affinity, or complex-forming tendency, is one of the governing reactions in certain tannages (vegetable tannage and fixation of synthetic tannins of sulfo acid type), a fact especially stressed by Otto (165) and by the present author (61). In many instances, e.g., in the irreversible fixation of high molecular lignosulfonic acid (Fig. l), the anion effect dominates over the factor of hydrogen ion concentration in regulating the course of the reaction, as is evident by fixation of anions in the pH range 5-8, i e . , on the alkaline side of the isoelectric point of pelt. 100
FIQ. 1.-Interaction of hydrochloric acid
90-
(I) and high molecular fraction of lignosul-
5a ao.u. - 70.2 E i6 0 -
fonic acid (11) with standard hide powder aa a function of hydrogen ion concentration of final solutions (78). (The irreversible fixation of lignosulfonic acid obtained indirectly by determination of the hydrochloric acid binding capacity of treated hide powder.)
. )
c
c 9
0 0
50-
-
40-
\
'c)
.-a v-
P
30-
2010
I
0
1
2
3
4
5
6
I
7
I
8
I
9
I
1
0
Final pH value
In the system collagen-aqueous solution of weak acids of high concentration, as e.g. acetic acid, the fixation of the acid in molecular form has been indicated (74). This additional uptake of acid molecules seems to be located in the peptide groups, leading to rupture of hydrogen bonds, evident in the increased reactivity of collagen thus treated toward coordination-active agents. Steinhardt and coworkers (191) have demonstrated the same effect on keratin by a different method. The titration curves of collagen may be analyzed by assigning the original pK values of the particular amino acids to the various amino acid residues (R groups) built into the protein chain. In such allocation of the original pK values to protein groups, certain difficulties have been experienced, for example, in the theory of formaldehyde tannage. It is at present generally recognized that the pK values of an amino acid do
PROTEIN-CHEMICAL ASPECTS OF TANNING
365
not necessarily apply to the amino acid residue built into a peptide chain, a fact especially stressed by Cohn and Edsall (25) and by Greenstein (53). The presence of charged groups on adjacent chains may bring about a large shift of the pK values and the influence of the most common link of protein chains, the peptide group, must also be recognized. Thus, the data of Stiasny and Scotti (195) show for polypeptides the marked influence of the accumulation of peptide groups of the same chain on the strength of the ionic end groups; Greenstein’s researches (53) point in the same direction. One must avoid drawing far-reaching conclusions by application of the method of assigning the pK values of the simple amino acids to the reactive groups of collagen, especially in irreversible systems. That collagen and gelatin, even in their simple reactions, are more complicated than implied by a direct evaluation of reacting groups by the pK values of the amino acids concerned is strikingly proved by comparing the hydrogen ion-binding curves of mammalian collagen, fish collagen, and gelatin in solutions of hydrochloric acid (82). These proteins show identical maximum fixation of hydrochloric acid. However, in the pH range 3-5, gelatin fixes about twice as many hydrogen ions as does native bovine collagen a t the same pH value; fish collagen is intermediate. The difference is probably the result of different degrees of internal inactivation of the polar groups, being a function of the degree of orientation of the chains. At a given pH value the discharge of those free or lightly compensated groups of gelatin would be expected to occur more easily and extensively than that of the strong pair of electrovalent links of native collagen. Since changes apparently take place in the internal compensation of polar groups in the pretreatment of skin for tanning, the degree of depolarization will be important for the behavior of collagen toward electrolytes, including some types of tanning agents. In the formation of soluble protein products by the interaction of acids and alkalis with collagen, neither the mode of attack of these agents nor the nature of the solubilized portion is known. A great deal of painstaking work remains to be done on these and related problems. 4. Swelling of Collagen and the Donnan E$ect
The swelling of hide and its effect upon the inter- and intramicellar space greatly influences the reactivity of hide for large molecules, e.g., vegetable tannins, since the degree of accessibility of the reacting protein groups is a function of the degree of swelling of the hide. The degree of swelling of collagen taking place in solutions of acids and alkalis nearly follows the curves of hydrogen ion and hydroxyl ion binding. By the brilliant application of the principles of the Donnan
366
K . H. GUSTAVSON
equilibrium to the swelling of gelatin and collagen in the classical work of Procter and Wilson (173,211), a basis was laid for quantitative work on protein swelling, even though it was later found that the phenomenon is not so simple as the ideal case considered by the pioneers, among whom the namc of Loeb (139) bclongs because of his outstanding experimental contributions to this problem. A number of problems such as the maximum points of swelling of collagen, the decreased degree of swelling incurred by further increase of the concentration of hydrogen and hydroxyl ions, and the repression of swelling by neutral salts are logically explained by the theory. I n the following sections attention will be called to a few problems to which this simple conception of ionic interaction is not applicable. It may be said that the occasional failure of the quantitative treatment of swelling of collagen according to the Procter-Wilson concept is due t o the complicated reactions taking place in the swelling, not amenable to mathematical treatment. First, the application of the Donnan equation requires complete ionization of the compound formed and a completely reversible system. Accordingly, the anion of the interacting acid should not possess affinity for the protein, and, further, molecular effects, so-called lyotropic effects, should be excluded. Another complication is presented by the changed degree of cohesion of the protein structure a t various degrees of swelling, especially rupture of hydrogen bonds a t high alkalinity, a complication also recognized by the pioneer investigators, which exerts a great effect not yet numerically calculable. The Donnan effect, which undoubtedly plays a prominent part in reactions of collagen, particularly in the preparatory processes, has become a less prominent factor because of the complications mentioned (see, e.g., 126). Furthermore, many problems can be explained equally well by other pliysicochemical treatments. The greatest importance of the Procter-Wilson theory has been its accentuation of physicochemical principles and technique in protein investigations. It constitutes the first attempt to describe the swelling phenomenon on a quantitative basis. It still remains a cornerstone of the theoretical foundations of tanning processes.
5 . Zsoelectric Point of Collagen I n determination of the isoelectric point of collagen, standard hide powder has usually been the substrate. Hide powder is practically pure collagen, obtained from thoroughly limed bovine hide which after deliming is made into a woolly powder. The isoelectric point of hide powder has by several independent methods been localized in the neighborhood of pH 5 , or a value practically coinciding with th a t of gelatin
PROTEIN-CHEMICAL ASPECTS OF TANNING
367
(139). Since the hide material undergoing tanning in practice is always limed, the importance of the isoelectric point of limed (alkali-treated) collagen is obvious. The correlation of the hydrogen ion concentration corresponding to the isoelectric point of this type of collagen with the figures of the contents of acidic and basic groups of collagen and their relative strength was earlier a puzzling problem, since collagen according to the previously accepted values of the amino acid distribution contained a large excess of basic groups. However, in view of the fact that not only the quantity of polar groups but also their strength enter into the equation of the location of the isoelectric point, the possibility that the pK values of the acidic groups are increased, compared to the value of the simple amino acids, by the ionic environment of the protein chains, could not be wholly discounted. It was reported by Briefer (17) and independently by Kraemer and Dexter (117) th at the isoelectric point of gelatin obtained from acid-treated pigskin, which had not previously been limed, corresponded t o p H values in the vicinity of 7. Limed pigskin gave a gelatin with a normal isoelectric point a t p H 4.8. A number of investigators later showed that the location of the isoelectric point of hide collagen is a function of its previous history, particularly the degree of alkali treatment (liming) (33). The final experimental proof of this assertion and an explanation of the displacement of the isoelectric point was given by Beek and Sookne (9) in cataphoretic investigations of collagen of different pretreatment. The isoelectric point of native (unlimed) collagen was found to be a t pH 7. Independently, Highberger (100) obtained figures in the p H range 7.5-8.0. The shift of the isoelectric point some 2 pH units toward the acid side, resulting from thorough liming of collagen, was ascribed to the deamidation taking place by the action of alkali on collagen, whereby the strong carboxyl group is set free (9,104). The new figures of the contents of dicarboxylic acids and amide groups of native collagen (Table 11) offer a rational explanation of the location of the isoelectric point of gelatin and limed collagen in the acid pH range, if due regard is paid to the hydrolysis of acid amide groups, with the formation of free carboxylic groups, taking place in the regular liming process (about half of the amide nitrogen being split by a few days’ treatment of hide in saturated lime solution (104)). Thus, limed collagen will contain about 1 .O mmol free carboxylic groups (1.24-0.24) and 0.87 mmol basic groups, or a surplus of acid groups (0.13 mmol/g. collagen). The new data also satisfactorily account for the localization of the isoelectric point of native (not alkali-treated) collagen in the p H range of 7-8, since the amount of fiee carboxylic groups is 0.77 mmol
368
K. H. QUSTAVBON
(1.24-0.47) and the content of basic groups 0.87 mmol or a surplus of 0.1 mmol basic groups. Highberger and Stecker (104) found that destruction of basic groups (arginine) is only of minor importance in shifting the isoelectric point. Furthermore Kuntzel (126) suggests that in the prolonged alkali treatment of collagen, which occurs in making gelatin, hydrolysis of peptide links may occur. The effect of the opening up of the -CO-N= group on the isoelectric reaction of collagen and gelatin may be an important factor. The location of the isoelectric point of collagen in combination with tanning agents is important for the theory of tanning and will be considered in that connection. 6 . Coordinative Reactions
In the complex reactions of tanning agents with collagen, coordination of the two components is in many instances of a n importance equal t o or greater than that of electrovalent reactions; the coordination is probably mainly localized in the peptide groups. Some of these groups are already internally compensated (hydrogen bonds). However, in view of the fact that limed collagen is generally concerned in fixation of tanning agents, and also since the alkaline pretreatment tends t o weaken or rupture the internal links of the peptide bonds by the swelling of the hide, coordination-active peptide groups should be available in the usual form linkage of limed collagen. Furthermore, the presence of the - CO-N= in every third residue, withdrawing loci for the compensation of adjacent -COgroups of peptide bonds, leaves coordinate loci available even in native collagen. It has been mentioned previously that disorganization of the lattice structure by contraction of the protein chains results in increasing the gap between some of the compensated peptide groups. Thus, the swelling of collagen by hydrotropic agents tends to increase the fixation of reactants of the coordinate type, e.g., vegetable tannins. Simultaneously, the internal cohesion of the structure is impaired, which may have undesirable practical consequences. The opening up of the structure will make the reactive groups more accessible to the tanning agent, and activation of peptide linkages will facilitate multipoint attachment of large molecules with numerous coordination-active groups on peptide groups (71). The coordinate type of reaction is sterically favored since -CO-NHgroups should be more easily accessible than the side chains of collagen to large molecules with regularly interspaced reactive groups. The coordinate type of reaction is indicated to be of governing importance for the fixation of vegetable tannins by collagen, as will be further shown in the section on vegetable tanning.
PROTEIN-CHEMICAL ASPECTS OF TANNING
369
7. Denaturation and Heat Shrinkage Collagen fibers, in the form of skin or tendon, in contact with water of 65-70°C. contract sharply at a given temperature (shrinkage temperature) to about one-third of their original length (34). The shrunken specimens feel gluelike and show rubberlike elasticity. The tensile strength of the fibers is immensely lowered. The original resistance of native collagen to trypsin is destroyed (35). Marked hydrolytic changes, in the form of splitting of peptide groups, do not seem to occur. The elementary composition, including the nitrogen content, is unchanged (50). The maximum acid-binding capacity is not affected (50). The water content remains the same (50). However, the X-ray diffraction pattern of the fiber becomes less well defined and the birefringence is decreased, indicating disorganization of the units constituting the fiber (3). By extension of the shrunken fibers the diffraction pattern is partially restored (3). By intramicellar introduction of certain tanning agents, e.g., aldehydes, almost complete reversal of the shrinkage is attained upon cooling the shrunken fiber structure (Ewald’s reaction, 34), probably due to the presence of aldehyde cross links (154). Similar to the action exerted by heat on moist fiber is the swelling by strong acids, which however only contract the fiber to two-thirds of its original length (121). The swelling incurred in brief treatment is reversed by a subsequent neutralization of the swelled collagen (141). However, upon prolonged interaction with strong acids, and particularly with alkali, the resulting swelling is only partly reversible (121). The contraction of the fiber by lyotropic agents is of an order comparable to the hydrothermal shrinkage. This change is irreversible (121). A close interrelationship of degree of swelling and hydrothermal stability is evident ( 16). The primary reaction of the shrinkage is in all probability a melting of the hydrated crystallites as assumed by Wohlisch (218) and by Meyer (21). Mirsky and Pauling (158) and Kuhn (118) have shown that the change from an oriented structure to a random state of configuration of the protein units is a natural consequence of the tendency of the structure to increase its entropy. The fibrous state is from the thermodynamic point of view artificial and labile, whereas the coiled globular state with a random distribution of the protein chains represents the more probable and stable state of protein configuration. Some investigators like Meyer and collaborators (21,154) consider the denaturation and shrinkage of collagen to be a reversible process. They consider that irreversible alterations occur only in fibers hydrolytically damaged during the denaturation. It seems, however, that the fiber wcave of the hide structure is irreversibly shrunk and permanently altered (120,123). This is indicated by its changed reactivity toward high-molecular compounds (71). However, in spite of the contraction of the chains and the disorganization of the structure, a certain orderly grouping of the units remains (120). Probably the saltlike crosd
370
K. H. GUSTAVSON
links still function as cohesive forces, although weakened by unfavorable steric conditions, maintaining the principal outline of the structural features. The hydrothermal denaturation also affects the hydrogen bonds, which are partly ruptured, resulting in noncompensated pcptide groups with coordinate loci available. This is indicated by the marked increase in the reactivity of collagen toward coordination-active compounds induced by the shrinkage (71,82). The denaturation does not affect the irreversible fixation of ionic agents, e.y., basic chromic salts of low to medium molecular size (two to six atoms of chromium in the molecule). However, the uptake of high molecular agents reacting mainly as coordination compounds, such as certain large chromium complexes and vegetable tannins, is drastically increased as shown by Table 111, in many instances a doubling of the ameunt of fixation being obtained (71).
The instantaneous shrinkage occurring at a certain temperature has its counterpart in similar although less drastic changes induced in the collagen structure by prolonged heating in water of temperature considerably lower than that causing rapid denaturation, for example, in water of about 50°C. This is evident from Grassmann’s data (50), and was more recently noted by Pankhurst (170) in prolonged treatment of pelt in water a t 45°C.(“incipient shrinkage”). Such changes will have practical importance in the pretreatment of hide in tanning. The incipient shrinkage leads to increased affinity of the treated collagen (4045°C.) for coordinate agents, although this is less pronounced than the effect of complete shrinkage (79). Similar changes in the reactivity of collagen to those induced by thermal shrinkage are obtained by the pretreatment of hide collagen with concentrated solutions of lyotropic agents, such as calcium thiocyanate, urea, and 2-3 M solutions of acetic acid (71). The maximum fixation of strong mineral acids such as HC1 is not changed by denaturation, as mentioned earlier. However, the fixation of hydrochloric acid in the p H range 2.5-6 is somewhat increased by denaturation. This is probably due to the easier discharge of the carboxyl groups, present as internally compensated ionic cross links, resulting from the widening of the gap between some of the polar groups, brought about by the steric changes caused by the thermal contraction (82). The fixation of hydroxyl ions is also slightly increased in the pH range 8-11 as a result of denaturation. Noteworthy is the finding of Wohlisch (218), later confirmed by Grassmann’s (50) investigation of single fibers, that the shrinkage temperature is increased when tension is applied to the fiber. This fact will explain why the shrinkage temperature from a compact part of a skin is higher than that of looser sections of the same skin. The fiber bundles of a close-textured weave are exposed to strain from adjacent fibers, whereas the effect of the neighboring fibers on fiber bundles in a loosely interwoven specimen is not apparent. The shrinkage tempera-
PROTEIN-CHEMICAL ASPECTS O F TANNING
371
tures of samples from various parts of the same skin generally do not differ more than 2-3°C. Fish skin collagen shows considerably lower hydrothermal stability than mammalian skin. The shrinkage temperature of fish skins generally is in the range 4@45"C., compared to temperatures of 6O-7O0C. for the mammalian type (71,72). It is noteworthy that gelatin films shrink a t 45°C. (170), which also is the shrinkage temperature of cod skin. It is probable t ha t in the teleost type of collagen saltlike cross links mainly hold the structure together; hydrogen bonding is of secondary importance. This is indicated by a comparison of the reactivity of bovine and fish collagens. Thus, e.g., high-molecular sulfonic acids such a s a-lignosulfonic acid, which is irreversibly fixed by both types of collagen in stoichiometric ratio, according to the available equivalents of basic protein groups, yield with bovine skin a leatherlike product with increased tensile strength compared to native skin, although the hydrothermal stability is not improved. By the corresponding interaction cod skin is converted into a gluelike mass (91). This finding may be interpreted to mean that the ionic pairs of cross links of collagen of fish skin are destroyed in the discharge of the carboxyl ions and by the irreversible attachment of lignosulfonate anions to the basic groups. The electrostatic attraction is eliminated and, since the major part of the peptide groups are not internally compensated, the natural cohesion of the structure is greatly impaired. The corresponding inactivation of the electrovalently compensated cross links of mammalian collagen has only a slight effect on this type of collagen because the main stabilizing links, the hydrogen bonds, are left intact, and also since collagen is not swelled by the sulfo acid. The complete dissolution of fish skin collagen, and also of ichtyocol and elastoidin, two special types of collagens investigated particularly by FaurC-Fremiet and his school (38), in dilute solutions of weak acids, e.g., acetic acid of 0.001 N strength (37), is also of interest in this connection. Mammalian skin collagen is not affected by such solutions, but collagen of tendon, as for example that of the rat's tail, dissolves (134, 159). It is probable that differences in histological structure (the intertwining of fiber bundles in the skin, as contrasted to parallel-grouped structures in the tendon) and in the ensheathing reticulin also play a role in the di&olution. The low degree of interlacing of the structure of fish skin collagen may also contribute to its lability toward acidic solutions.
Native fish skin, such as cod skin, which has been investigated in detail, is easily digested by trypsin (72,75) in contrast to native mammalian skin, which is practically inert toward proteinases (35). This finding indicates that the peptide groups of teleost collagen are largely uncompensated. Since swelling contracts the fibers in the direction of the long axis, it
372
K . H. GUSTAVSON
is evident that the shrinkage temperature (if there is any sense in measuring shrinkage temperatures in such cases) is lowered by acids and alkalis (16). By repressing the swelling of pickled pelt with salt, the dehydration of the structure results in a slightly improved hydrothermal stability (68,123). Complete inactivation of the ionic protein groups by means of a nonswelling acid, resulting in the complete destruction of saltlike cross links, should result in lowered shrinkage temperature (ca. 10-12°C.) (71). The increase obtained by acid saturated collagen from pickle with high salt concentrations evidently shows that some other changestending to increase the stability of the structure-more than outweigh the decrease due to the ionic discharge. Speculatively i t may be conceived that the dehydrating effect of the salt in conjunction with the action of the acid brings the backbone peptide groups closer to each other, increasing the strength of the hydrogen bonds (68). It is interesting that in pickle solutions with increasing concentrations of hydrochloric acid the shrinkage temperature remains constant or slightly increases, compared to neutral pelt, until a concentration of 1 N acid is reached. By further increase of the acid concentration the shrinkage temperature is greatly decreased. Thus, shrinkage takes place at room temperature in acid concentrations greater than 2 N (123). The latter change, which is independent of the neutral salt concentration, is evidently a specific molecular effect, probably due to association and coordination of hydrochloric acid molecules on the peptide links, leading to an irreversible rupture of a part of the stabilizing links at these points. This is an additional example of the dual nature of the action of electrolytes in concentrated solutions upon proteins. Similar effects, decreasing or increasing shrinkage temperatures of collagen according to the concentration of the solute, not involving complicating osmotic swelling phenomenon, are shown by ethanol (123). Mixtures of ethanol and water, containing less than 40 % ethanol decrease the shrinkage temperature of collagen, whereas higher concentrations increase it, the augmentation being about 10°C. in 80% ethanol. Noteworthy is thc behavior of fibers that are heat shrunk in the latter solution. By careful stretching of the fibers it is possible not only to restore them t o their original length but to obtain elongation up to double the original fiber length by applying further tension. The X-ray diagram of the extended denatured fibers shows chains aligned in parallel. This observation of the great extensibility of alcohol-treated collagen (123) is especially interesting in view of the extensibility of collagen fibrils (179). The shrinkage temperature of tanned skin is one of the fundamental criteria of tanning potency (148), although it is important not to evaluate tanning agents merely by this property (101).
8. Neutral Salt Action and Lyotropic Behavior
The effect of neutral salts and lyotropic agents on collagen in the isoelectric state is important practically as well as theoretically. From
PROTEIN-CHEMICAL ASPECTS O F TANNING
373
a practical standpoint the use of neutral salts, generally sodium chloride, for preservation of the raw hide may be mentioned, as well as the influence of salts on the neutral pelt previous to tanning. The interaction of neutral salts with collagen also involves problems of fundamental importance for the theory of protein reactivity. The specific action of neutral salt ions on soluble proteins, established by Hofmeister (106), applies also to the action of concentrated solutions of neutral salts on collagen. I n this instance the action is mainly a molecular effect. It was shown by Thomas and Foster (203) that, by treating hide powder in solutions of various neutral salts of about 1 M strength, a large percentage of collagen was solubilized by sodium thiocyanate, iodide, and bromide, whereas sodium sulfate acted as a preservative. Sodium chloride had only a slight effect. I n view of the wellknown fact that molecular compounds are formed between amino acids and neutral salts, isolated by Pfeiffer and his school (172), and since the degrees of solubilization of collagen by the various salts on the whole follows the order of the stability of the neutral salt compounds with amino acids, the effect has been considered to be an interaction of the neutral salt with the coordinate valency bonds on the peptide groups (hydrogen bonds) (55). Part of these stabilizing bonds are weakened and ruptured, resulting in creation of new coordinate loci. In reactions between amphoteric ionic structures such as collagen and simple electrolytes, ionic interaction of the type given by simple amino acids and neutral salts has also to be considered, according to the equation (172; cf. 1,135):
+
C O O - C H ~ . N H J + Me+X- e-Me+.COO-CHn.NH8+.X-
This type of reaction or electrostatic interaction probably dominates in dilute solutions. These have no dissolving effect upon collagen. In concentrated solutions of neutral salts, as for instance calcium chloride and thiocyanate, collagen is attacked, being largely dissolved and permanently changed in reactivity and properties. The destructive action of concentrated solutions of lyotropic neutral salts (1-2 M ) seems to be mainly a molecular effect. A proof of the changed reactivity of pretreated collagen, probably involving nonionic groups such as the peptide groups, is supplied by the marked increase of the fixation of high-molecular compounds, e.g., highly aggregated sulfito-sulfato chromiate and vegetable t.annins, compared to native collagen (55,71). The fixation of simple electrolytes, such as acids, alkalis, and chromium salts of low molecular weight, is not affected by the treatment. Table I11 contains some typical data, showing the degree of solubilization of hide powder upon 2 weeks’ treatment under
374
K. H . GUSTAVSON
sterile conditions in M solutions of various neutral salts. It also presents data on the fixation of ionic-reacting basic chromium sulfate, highmolecular chromiate, and wattle bark tannins (55). The table also includes results from series of pretreatment5 of neutral pelt (calfskin) with 8 M urea solution and with solutions of 3 M acetic acid (the stock subsequently neutralized and made isoelectric). The effect of hydrothermal denaturation of hide powder (2 minutes a t 70°C.) is also included, since this denaturation also modifies the reactivity of collagen (71). Chrome tanning was carried out in solutions containing 1 equivalent of chromium per liter a t 20°C. for 144 hours. TABLE I11 Influence of Pretreatment of Collagen on Its Reactivity --_ Fixation of tanning agents from Pretreatment
Hide powder Water (blank) M KCNS M CaCL Shrunk at 70°C. Pelt Water (blank) 8 M urea (37°C.) 3 M acetic acid
Collagen dissolved in pretreatment, %
3 24 32
-
0 54 21
66% acid Cr-sulfate, % Cr2W collagen
'Sulfitoehromiate, % CrzOa/ collagen
Wattle bark tannins, % tannin/ collagen
11.2 11.0 11.3 11.7
20.8 31.3 30.6 34.1
52 81
10.3 10.4 11.7
16.5
46
40.3 26.8
80 69
78 84
An excellent illustration of the ionic and molecular effects is presented by acetic acid. I n dilute solution, e.g., 0.1 N a t final p H values of about 3, its action upon collagen is nearly the same as that of a corresponding solution of hydrochloric acid of identical final pH; the effect is mainly due to the hydrogen ion concentration. With increasing hydrogen ion concentration of the solutions, up to pH 1, the swelling of collagen by hydrochloric acid increases until a maximum is reached at pH 2. In contrast to the behavior of strong acids, increased concentration of acetic acid leads to very much greater augmentation of the swelling (74). At the lowest pI-1 values the collagen loses its structural cohesion and goes into solution (74). This peptization is due to the molecular effect of the acetic acid. Since sodium acetate solutions of corresponding concentrations (> 3 M ) do not show any -solubilization effect on collagen and do not alter its coordinate reactivity, it is evident that neither hydrogen
375
PROTEIN-CHEMICAL ASPECTS O F TANNING
ions nor acetate anions are responsible for the drastic effect of acetic acid in concentrated, aqueous solutions. The effect is probably caused by the coordination of the monomeric acid molecules on the peptide groups which are thereby partly broken. The result will be diminished internal cohesion of the structure (74). It is noteworthy that strong solutions of mineral acids (above 2 M ) show this molecular type of swelling to a greater degree than the ionic type (123). Phenol reacts in a similar manner (123). The effect of urea on globular proteins, splitting the molecule into several fragments (19), has a counterpart in its behavior toward collagen. The fibers shrink ( 1 16) and the treated product shows markedly increased coordinate reactivity (71). Summing up, the effects of lyctropic neutral salts, organic substances of lyotropic function, weak organic acids in concentrated solution, and heat denaturation of collagen are of the same general type, resulting in a weakening and dislocation of the stabilizing cross links located on the peptide groups (hydrogen bonds). The application of collagen with such activated coordinate loci for the study of the mechanism of tanning processes has been a valuable adjunct as will be shown in the sections on tanning. 111. KERATOLYSIS AND
-4CTION OF
ALKALION
HIDE
The removal of hair and epidermal matter from the hide in order to obtain the leather-forming part, the corium, free from nonleather-forming constituents is accomplished by the liming process, which also brings the unhaired hide, the pelt, into a suitable state for tanning, activating certain protein groups, plumping the hide by water uptake, “opening u p ” fiber bundles, bringing about lateral splitting u p of the individual fibers, saponifying fats, and modifying or removing the accessory proteins, such as interfibrillar matter and reticulin. The hide is treated in a saturated solution of calcium hydroxide, containing a large excess of the alkali and a small amount of a specific depilatory, generally sodium sulfide or hydrosulfide. Within a few hours or days, according to the degree to which the action of the limes is sharpened with depilatory, the hair roots are loosened and unhairjng results. In the fundamental investigations of Merrill (149), the breaking of the disulfide bridge of keratin was established as the main reaction of depilation. This was further confirmed by Marriott (140) and by Windus and Turley (215). The depilatory action of a great number of organic sulfur compounds has been extensively investigated by Turley and Windus in a series of important papers (209,215).
376
K . H. QUSTAVSON
According to the modern view, the first reaction of unhairing is the breaking of the disulfide bridge by hydroxyl ions. The action of the depilatory is mainly secondary. Its main function is to prevent the formation of new links which otherwise would stabilize the keratin structure. The breaking of the disulfide cross link in alkaline solution leads to the formation of a thiol and a sulfenic acid group:
\
/
CH.CHz.S.S.CH2.C cystine link
-
&-OR
\
\
/
CH.CHz.SOH sulfenic acid group
+ HS.CHz.CII/
\
thiol group
Then the specific action of the depilatory enters. If only hydroxyl ions are present, the active groups formed by the splitting of the disulfide bridge react furthcr, with formation of new cross links between adjacent protein chains, which leads to increased stability. Among the many reaction mechanisms proposed, experimental evidence has been supplied for the formation of the -CHZSCHZ- link (lanthionine, 27,186). The specific unhairing agent apparently reacts with the sulfenic acid, group, inactivating it and preventing the reformation of cross links. A great many reducing agents possess this property, the most important technically being the alkali sulfides, primarily sodium sulfide (hydrosulfide). Other agents proposed are cyanides (140), thioglycolic acid (157), sulfites (140), and aliphatic amines (147). Among those mentioned the cyanidcs are particularly interesting from a theoretical point of view, since Cuthbcrtson and Phillips (27) recently reported that potassium cyanide in neutral solution converts all cystine of wool to lanthionine. The excellent unhairing action of secondary amines, e.g., dimethylamine, discovered by McLaughlin (147), has not yet been satisfactorily explained (cf. 143). It is noteworthy that the secretion of the cloth moth contains a reductase in alkaline medium (pH 10) which enables the larva to digest wool keratin, as the remarkable histochemical researches of Lindcrstr$mLang and Duspiva (138) show.
The presence of various types of keratins and related proteins in the epidermis of hide further complicates the problem of unhairing. The soft keratins, such as those of the mucous layers of the epidermis, show a lower degree of stability toward reducing agents and alkalis and are also more readily attacked by heat and enzymes than the hard keratins present in hair. This has been explained by the higher frequency of disulfide cross links in the latter. The lower stability of the epidermal keratins has been accounted for by the presence of a lavge part of the sulfur as sulfhydryl groups. However, the cystine content apparently is not the only factor governing the stability. If we take into account Block’s concept of keratin structure, postulating the molecular ratio of basic amino acids as the characteristic property of keratins, some clues to the varying stability of a-proteins may be obtained (177). This is shown by Rudall’s work on the hydrothermal behavior of various epidermal keratins and other a-proteins (177). With decreasing relative proportion of arginine, the stability of these proteins (keratins, myosin and fibrin) is decreased. The guanidyl group may then contribute to the stability of the keratin structure by intermolecular linkage. I n view of
PROTEIN-CHEMICAL ASPECTS OF TANNING
377
Rudall’s important finding (177) th at the mucous layers of the epidermis are readily dissolved in 50 % urea solution, yielding epidermin, the possibility of removal of epidermis by means of strong solutions of lyotropic agents is suggested as a method of unhairing.* The direct chemical action of alkalis on collagen was mentioned in connection with the discussion of the shift of the isoelectric point. It was pointed out that the main reaction apparently is hydrolysis of the amide groups of asparagine and glutamine. By treatment of collagen in lime for 9-10 days, ammonia in an amount equivalent to the hydrolyzed amide groups is formed (104). Partial destruction of the guanidine group of the arginine residue occurs in alkaline solutions, being considerable at high temperatures. However, this reaction does not seem to be appreciable in ordinary liming (104). The extent of the action of alkali on the peptide groups, leading to their cleavage, particularly the effect on the -CO-N= groups, is not known. The hydrolysis of the proline peptide groups should prove especially interesting. From the titration curves of collagen of different degrees of liming no indication of appreciable increase of ionic groups is obtained. However, it must be realized th a t even if the collagen molecule is halved the acid- and base-combining capacity will not show greater increase than 0.1 milliequivalent per gram collagen (1 4). I n conclusion, it may be pointed out that the alkaline swelling is not depressed by the addition of sodium chloride to solutions of sodium hydroxide, as would be expected from considerations of swelling as a Donnan effect. Such an explanation evidently fails in this instance, as aptly pointed out by Kuntzel (126). Sodium sulfate is the only common neutral salt functioning as a n active swelling depressant in alkaline solutions (129). One important function of the liming treatment of hide is the opening up of the fibers, making the structure accessible and permeable to large molecules. Another important aspect is the activation of ionic groups by the breaking up of the saltlike links through hydrolysis of the charged amino groups by the hydroxyl ions, which upon prolonged liming (swelling) involves partly irreversible changes. The increase of carboxylic groups by the deamidation process connotes increased ionic reactivity. Thus, the binding capacity of collagen for electrolytes and tanning agents, both chromium salts and vegetable tannins, is increased. The reactivity of collagens is further favored by the physical changes of the
* Kritainger (117a) reports that by immersion of fresh calfskin in a 10% solution of sodium chloride for 7 days under sterile conditions unhairing takes place. He attributes this to the removal of globulins, but probably the solvent action of the lyotropic agent, sodium chloride, on the epidermis is the main factor of depilation in this instance.
378
K. H. QUSTAVBON
hide in liming and also by the chemical changes in nonionic protein groups, since the swelling of the hide leads to rupture of some of the coordinate bonds between adjacent protein chains (66). The hydrogen bonds are not completely reformed in the subsequent deliming which reduces the swelling. Hence, the reactive protein groups are made more easily accessible.
IV. GENERALASPECTSOF TANNING In the definition of tanning in practical terms, the most striking physical, directly observable changes involved in the process form the criteria. By tanning the easily putrescible hide substance is made resistant to micro-organisms. Further, leather will resist water and moderate temperatures in the moist state and remain soft and flexible upon drying. Through the extended knowledge of the tanning processes acquired during the last decades, it is now possible to define tanning action in a more scientific manner. The first criterion of tanning potency of a substance is its capacity to form an irreversible combination with collagen, resistant to the action of water. Certain reactive protein groups are inactivated. However, the simple incorporation of an irreversibly fixed agent and the reduction of the water-binding capacity of collagen does not effect tanning. The second criterion is the stabilization of the collagen by the tanning agent, improving its resistance to heat and proteinases and preventing the “glueing together” of fibers upon drying (“leathery drying out ”) without detrimentally affecting the mechanical strength of the original hide structure. By the conditioning of the hide, its original resistance to the agents mentioned is lowered, chiefly due to diminished internal cohesion of the weave structure. The function of the tanning agent is to make up for this labilization, incurred in the prior conditioning of the hide, by making it into a more stable structure than it originally was, supplying stable cross links which rivet the chains together. The cross link concept was suggested by Meyer (153). From the practical point of view, it is not always necessary that all criteria of tanning should be realized. Thus, hide treated with agents which do not raise its shrinkage temperature, may nevertheless possess certain specific properties desired, e.g., great tensile strength or pliability. On the other hand, tanning agent making the hide resistant to high temperature may have other disadvantages, such as impairing the tensile strength. It is obvious that tanning agents do not form a distinct class of compounds chemically. On the contrary, substances of very dissimilar
PROTEIN-CHEMICAL ASPECTS OF TANNING
379
composition and nature possess tanning properties; one need mention only basic chromium salts, vegetable tannins, aldehydes, certain condensed phenols containing sulfonic acid groups, and unsaturated oils. The substances mentioned are the tanning agents of practice; chromium salts and natural vegetable tannins being the principal ones. Since the reaction of chromic salts with collagen (chrome tanning) is best known and offers a great many problems of general interest., the chemistry of chrome tanning will be comprehensively discussed.
V. REACTIONOF CHROMIUM COMPOUNDS WITH COLLAGEN (CHROMETANNING) Among inorganic substances the chromium compounds, notably the basic chromic salts, hold an unrivaled position as tanning agents. The strong complex-forming capacity of the chromium atom probably is not in itself the deciding factor for the excellent tanning potency. Thus trivalent cobalt is equally efficient as a complex former, but its compounds do not possess tanning potency. Chromium forms polynuclear compounds of an intermediate degree of stability, aqueous solutions forming compounds of the type -Cr-0-Crupon hydrolysis. However, such hydrolytic changes are not restricted to chromium, even the related metals (Fe and Al) show this tendency. Chromium differs insofar as i t forms polynuclear complexes which generally are in hydrolytic equilibrium and reactive. It seems that a combination of these properties is decisive for tanning action. The electronic structure of chromium, with three unpaired electrons entering into resonance with each other in the ionic and covalent chromic compounds, does not permit application of the magnetic method of differentiation of the type of bonds formed between chromic salt and collagen (171). 1. Structure of Basic Chromic Salts of Zmportance in Tanning In practice basic sulfates of chromium containing an equimolar amount of sodium sulfate are commonly employed in tanning. The basic sulfates differ from t h e basic chlorides insofar as the sulfate group possesses a marked tendency to be directly linked to the chromium atom, forming sulfato-chromic complexes (193). The 0 following is an example: (Cr
/
\
\
so,
/
Cr) SO,.
The basic sulfates exist in dilute
solutions, such as those generally employed in tanning, mainly as cationic complexes (77,80,81). Concentrated solutions (greater than 4 N) contain a large percentage of neutral molecules and even negatively charged chromium complexes (77,80,81). On the contrary, the electrochemical composition of basic chlorides is independent of the concentration of the solutions (77,80,81). Chromium salts possess, beside electro-
380
K. H. QUSTAVSON
valent function, the faculty of coordination. The number and nature of the groups in the internal sphere, the coordination sphere, determine the coordinate ability of the central atom. The effect of the nature of the complexly attached acid residues (acido groups) on the coordinate function of chromium is known in a general way (193). The aggregating tendency of the compound is also interrelated with the structure of the internal sphere of the complex. The sulfate group seems to exert optimal action on the coordinate activity of the chromium atom with regard to tanning. Chloride and nitrate groups are weak complex formers, generally being present in the electrovalent state (193). Some strong complex-forming organic acid residues, e.g., the oxalate and tartrate, apparently leave chromium only a slight chance of further coordinate function (193). The tanning potency of basic chromic salts is a function of the coordinate faculty, the degree of aggregation of the salt and its degree of ionization. The size of the complex is important for the polyfunctional activity in the multipoint fixation of the chromium compounds. The properties and behavior of chromium salts are advantageously considered from the point of view of the Werner coordination theory (193). Before proceeding any further with dctails, a general orientation in the probable mode of interaction between a solution of a basic chromic sulfate and collagen may be helpful. A 67% acid chromic sulfate is employed. It may be simply made from the hexaquosulfate by the addition of the required amount of sodium hydroxide to yield a basic sulfate with composition corresponding to the empirical formula Cr2(0H)2(S04)2-Na2S04. This solution is hydrolyzed, and, in the concentration employed for our experiment, 1 equivalent per liter chromium, gives a p H of about 3.0. The neutral collagen combines immediately with the free sulfuric acid. Simultaneously the positively charged sulfatochromium complexes, as e.g. (Cr20S04)++,are attracted to the negatively charged carboxyl groups of collagen, and react with them, the sulfate ions being compensated by the charged amino groups. The chromium complex is extremely firmly attached to collagen; indicating the conversion of the originally electrovalent bond into a type of greater stability (66). The depolarization of the chromium-collagen link is probably the net result of the penetration of the carboxyl group of collagen into the chromium complex, which also completes the reaction by coordination on adjacent coordination-active groups of the protein chains. The final result is the formation of a modified type of an internal complex salt with covalent as well as coordinate multipoint attachment of the chromium complex to groups of adjacent collagen chains, leading to a riveting together of the protein chains by means of the chromium complexes. The formation of chelate compounds explains the high degree of rigidity of the resulting structure and the stability of chrome leather (66,122).
2. Factors Governing the Tanning Efect These are mainly: the type of salt, its basicity (% acidity), type of compound, its concentration, the neutral salt content, the pH value of the system, the presence of complex-forming substances, the temperature of the tanning bath, and the time of interaction. The factors mentioned pertain to the chromium compounds. The nature and the state of collagen enter also heavily in the equation, particularly in regard to the practical issues and the qualities of the leather produced. Such factors are the availability of amphoionic groups, coordination potency of the
PROTEIN-CHEMICAL ASPECTS OF TANNING
381
pelt, and its degree of rigidity (if swelled or flaccid) in the initial stage of the tannage. Many of these factors are evidently interdependent and difficult to treat separately. a. Nature of the Anion. This factor is of primary importance for the tanning function of chromic salts. Nitrates are less satisfactory than chlorides, which in their turn are markedly inferior t o sulfates; the latter as previously mentioned are the most suitable tanning agents. Formato complexes are good tanning agents, especially in mixtures with sulfates (193). Some oxalate complexes possess tanning potency but most of them do not. Of the numerous organo complexes, the acetate and tartrate as well as the oxy acid compounds are devoid of tanning power. The influence of the anion on the tanning potency seems to be a function of its affinity for the central atom, indirectly affecting its coordinate function and secondary reactions (193). b. Basicity of Chromium Salt. The basicity or per cent acidity (expressed in per cent of the equivalents of basic or acid residues present in combination with chromium, calculated on the total amount of equivalents of chromium) (194) is in practice the primary factor. If the basicity is not suitably adjusted, the result of the tanning will be upsatisfactory. By adjusting the percentage of acidity of the chromic salt to the values of 50-66, an average molecular size with two to six atoms of chromium (111,175)is obtained, possessing good diffusibility through the fiber structure a n 4 a suitable degree of affinity for collagen. Sterically, the multipoint interaction is facilitated since further hydrolysis of the fixed chromium salts may occur in the secondary processes and in the subsequent treatments of the leather, yielding in situ still more aggregated complexes. Solutions of this type of salt generally have hydrogen ion concentrations in the range of pH 3-3.5. Hydrogen ion concentrations of this order are favorable for a suitable rate of tanning, since a great part of the carboxyl groups of the amphoionic collagen structure are present as ions. At pH values in the vicinity of 2 or below, only a small part of the carboxyl groups are present in the charged state, slowing up the reaction. On the other hand, a t higher p H values, e.g., 4.5-5,all the carboxyl groups are available in the ionic state. This would result in a very rapid combination. Furthermore, the basic chromic salts formed in this pH range will aggregate. By the rapid uptake of hydrolyzed acid by the neutral pelt, the hydrolysis proceeds. Highly aggregated chromium compounds may be precipitated on the outside of the hide, giving rise to a blocking of the penetration, leaving the interior of the hide in untanned or undertanned state (" case hardening"). c. Concentration of Chromium Salt. As is evident from Fig. 2 the chrome fixation by collagen from solution of basic chlorides increases
382
K . H . GUSTAVSON
steadily with increasing chrome content of the solution (97). This probably reflects the fact that the basic chromic chlorides do not form complexes of greatly loivered affinity for collagen (uncharged or anionic) in concentrated solutions. By interaction of solutions of basic chromic chlorides and sulfates with organic cation exchangers of the sulfonic acid type the same type of chrome fixation curves, as a function of chrome concentration, arc obtained as for corresponding collagen systems (81). The reaction of the basic chromic sulfates with collagen is complicated with increased chrome concentration on account of the formation of uncharged and negatively chargcd chromium complexes (see Fig. 2).
I 0
I
I
50
100
Cr203, q J L
FIG.2.-Fixation of chromium compounds by hide powd; as a function of the chrome content of solutions (66). Time of tanning, 48 hours; composition of chromic salts corresponding to Cr2(0H),X..2NaX. -x-x-, 78 % acid sulfate; ---o-o--, 66 % acid sulfate; -A-A-, 50 ’?& acid sulfate; -0-0-~ 00 % acid chloride.
These show markedly different reactivity toward collagen than the common type of cation chromium complexes present in dilute solutions (80,81). Thus, e.g., the 50% acid chromic sulfate of composition corresponding to the empirical formula Cr4(OH)e(S04)3-2Na2S041 contains only cationic chromium complexes in concentrations up to 1 equivalent per liter chromium. Their amount decreases steadily with increasing chrome concentration, uncharged chromium complexes mainly being formed, and in the most highly concentrated solutions anionic chromium. Thus, a t a concentration of 4 equivalents per liter chromium, the solution contained 60 % cations, 38 % uncharged complexes, and 2% anionic chromium complexes; in solutions containing 8 equivalents per liter chromium, the corresponding figures were 35, 60, and 5 (77,80,81). Since not only cationic chromium complexes are fixed by collagen from concentrated solutions of basic sulfates but also uncharged and negatively charged complexes, the composition of the resulting leather will be very much more complicated than that of collagen tanned in dilute
PROTEIN-CHEMICAL ASPECTS OF TANNINQ
383
solutions (80,81). Further complications arise from the unstable nature of the fixed noncationic complexes and their gradual conversion into cationic ones, for example, by washing the leather, which will result in redistribution of the forces between collagen and the attached chromium compound (68,94). The effect of the uptake of nonionic complexes by collagen is clearly evident in tanning with dilute solutions of basic chromium sulfate, used immediately upon dilution. Although the noncationic complexes are rapidly changed into cationic ones by the dilution (to 60-70% within 4-5 hours), the pelt will combine with the uncharged complexes during a brief duration of tannage, for example within 4 hours. The chrome fixation obtained will be markedly greater than that resulting from the aged solution, which contains only positively charged chromium complexes. Moreover, the composition of the fixed chromium complexes, measured by the ratio of equivalents of sulfate to chromium combined, will give higher % acidity of the stock tanned in freshly diluted solution, due to the fixation of the more acid uncharged complexes (94). Further, in the determination of the shrinkage temperature, on immersing the pressed and lightly rinsed leather in water at about 1°C. below the actual shrinkage temperature (this being localized by preliminary trials), i t will be found that the leather tanned in the aged solution, which contains only cationic chromium complexes, will shrink a t a temperature !5-6”C6 higher than that of the stock tanned in the freshly diluted solution, containing considerable amounts of noncationic complexes (94). A striking illustration of the importance of the mode of determining the shrinkage temperature is supplied by the hydrothermal behavior of the two types of leather. If the determination of the shrinkage temperature is carried out by gradual heating of the stock in water, commencing a t 40°C. and heating a t a rate of 2°C. per minute, the noncationic chromium complexes fixed by collagen have ample time for rearrangement and formation of new attachments to collagen. Both types of leather will show complete resistance toward boiling water whereas, in the instance of direct contact of the specimens with water of the actual shrinkage temperature, the values will be 93 and 99°C. for the leather tanned in the directly diluted and the aged solutions, respectively (94). In actual practice the final shrinkage temperature will be practically identical since in the subsequent washing and processing of the leather the complexes will reach the stable state by structural rearrangement (68). Kiintzel, Kinzer, and Stiasny (128) consider that the main cause of the diminished chrome fixation with increased concentration of chrome is the dehydrating effect of the sodium sulfate on collagen, accumulating in concentrated solutions of “chrome sulfate liquors.” The dehydrated hide should possess less affinity for chromium. They do not consider the constitutional factor of the chromium salt to be of any importance. However, it is difficult to conceive why the large amount of sodium chloride present in concentrated solutions of basic chromic chlorides should not have the same effect (66,97). However, collagen shows a very sharp rise in chrome fixation in tanning with very concentrated solutions of chromic chloride.
384
K. H . OUSTAVSON
The explanation on a constitutional basis receives experimental support by data from the action of neutral sulfates (Na2SO4)on basic sulfates in regard to the chrome fixation by collagen. The formation of noncationic complexes possessing low affinity for collagen runs parallel with the decreasing uptake of chromium from such solution by the cationic exchangers (81).
d. Neutral Salt Effect. The action of neutral salts in the processes of chrome tanning is an important but exceedingly complicated problem. As a rule, the salts mainly concerned are sodium sulfate, a regular constituent of chrome-tanning preparations, and sodium chloride, present in the pickled pelt or added to the tanning bath as a regulator of swelling. The primary role of the neutral salts as agents which depress fiwelling may be illustrated by tanning neutral pelt, or still better pickled pelt, in a solution of a basic chromic sulfate free from sodium sulfate (for example a salt prepared by reducing chromic acid in the presence of sulfuric acid by means of hydrogen peroxide). The pelt will swell and the penetration of the chromic salt through the hide structure will be greatly retarded and tanning practically prevented if highly basic chromic sulfate is used. The cross linking of collagen will be sterically hindered (65). Hence, the primary function of neutral salts in chrome tannin& is the osmotic balancing of the swelling of the hide which takes place in the initial stage of tanning by the preferential fixation of the free acid by collagen. With the concentrations of neutral salts generally used, of the order of a few tenths molar, present in an acid system together with the astringent chromic salt, any direct, specific action of the neutral salt on the protein is not detectable and hardly likely. The main secondary function of the neutral salt will be its effect on the physicochemical properties of the solution, such as alteration of the degree of hydrolysis and, above all, constitutional and electrochemical changes of the chromic salt. The important function of neutral salts in chrome tanning was discovered by Wilson (212-214) some 30 years ago. Fundamental studies of this effect have been carried out by Thomas and coworkers (4,199-201). For details regarding the effect of neutral salts on the hydrogen ion concentration of solutions of chromic salts these papers should be consulted. Only a brief summary of the main principles and results of the neutral salt effect in the various systems can be given. Sodium chloride markedly increases the reactivity of chromic chloride for collagen in dilute solution of chromic chloride, while in concentrated solution the effect is the reverse (59). This action has been explained by the influence of the salt on the composition of the coordinated sphere. By the excess of chloride ions, chlorine groups are being forced into the internal sphere (59). In solutions of very basic chromic chlorides this effect is further
PROTEIN-CHEMICAL ASPECTS O F TANNINQ
385
accentuated by aggregation of the chromic chloride (59). Since in both instances the equivalent weight of the chromium complex will be increased, greater chrome fixation will result. The positive constitutional effect more than counterbalances the negative effect of increased hydrogen ion concentration of the system. I n concentrated solutions of basic chromic chlorides the constitutional change of the complexes has already been attained. Accordingly, it is not further altered by the addition of sodium chloride. I n this instance the main effect of the salt will be to decrease the chrc me fixation by the increased hydrogen ion concentration which it brings about (59). The diminished affinity of basic chromic sulfate for collagen in the presence of sodium chloride seems primarily to be connected with the formation of mixed chloride-sulfate by double decomposition (66). By addition of sodium sulfate to solutions of chromic chlorides, chromic sulfates are formed because of the great complex affinity of sulfate groups. For particulars see Gustavson (56,66). The system basic chromic sulfate-sodium sulfate presents less complication, being a system of a common anion. The addition of small amounts of sodium sulfate to a salt-free solution of basic chromic sulfate tends to increase the chrome fixation by collagen (164). This effect has nothing to do with the neutral sulfate effect on the chromic sulfate. It is probably a purely osmotic effect of the neutral salt on the hide, suppressing its swelling, facilitating the uptake and diffusion of the chromic sulfate. By further addition of sodium sulfate the affinity of the chromic sulfate for collagen is decreased. An over-all decrease is also generally found by the addition of sodium sulfate to the systems of the commonly employed type of chromic sulfate containing sodium sulfate, such as Cr2(OH)2(SO4)2.Na2SO4(212). The retarding effect of sodium sulfate on the chrome fixation cannot be a pH effect, since the hydrogen ion concentration is decreased. The real cause is the change of the chromic sulfate in the direction of uncharged and negatively charged complexes, possessing less affinity for collagen than the cationic ones (66,81). Possibly the formation of addition compounds between the components of the system with lower degree of activity is an additional factor (201). The reaction of chromic sulfates with the cation exchanger as a function of added sodium sulfate (81) shows a similar trend to that of the chrome fixation by collagen. Analysis of solutions by means of the ionic exchange method shows, e.g., that a solution of 66% acid chromic sulfate (1 equivalent per liter chromium) contains practically all the chrome in cationic form. Upon the addition of sodium sulfate, making the solution 1 M in NapSO,, it contains 65% cationic, 6% anionic, and 29% uncharged complexes. Since concentrated solutions of basic chromic sulfate contain mainly noncatiouic complexes, the retarding effect of sodium sulfate upon the chrome fixation by collagen should in this case hardly be perceptible, since by the addition of neutral
386
K. H. QUSTAVSON
sulfate further shift of the equilibrium of the various forms should not be expected. Thia is also the case (201). An important property in the investigation of tanning processcs, easily measured and stated, is the hydrothermal stability of the leather as a function of the neutral salt content of the tanning bath, measured by the shrinkage temperature. Generally decreased chrome fixation and formation of more acid chromium compounds in the skin, associated with the neutral salt effect, will tend to decrease the shrinkage temperature. That is exemplified by the system chromic sulfate-neutral salts. Sodium sulfate affects the hydrothermal stability only slightly, although it lowers chrome fixation more than does sodium chloride, whereas sodium chloride greatly decreases the hydrothermal stability (66). This may be ascribed to the fact that chromium chlorides do not yield leather of such high shrinkage temperature as that tanned with chromic sulfates. The formation of chloro-chromium complcxes in leather tanned with solutions of basic chromic sulfates containing large amounts of sodium chloride explains the low degree of stability of this type of leather (66). Concerning the systems chromic chlorides-neutral salts, it may be said that addition of neutral sulfates as well as of neutral chlorides improves the hydrothermal stability of the treated collagen, the stabilization being particularly marked in the first instance. The explanation of the action of neutral sulfate is the formation of basic chromic sulfate, which cross links the protein chains more effectively than chromic chloride (66,130).
The neutral salt effect in chrome tanning, presenting intricate theoretical problems, is also of great importance to practical tanning. It forms a versatile tool to regulate the degree of swelling of pelt in the initial chrome fixation which largely determines the character of the final leather. e. Hydrogen Ion Concentration. The hydrogen ion concentration of the tanning system, employing neutral pelt, is a factor interrelated with the composition of the chromic salt, primarily its basicity. The extremes are: simple complexes and high hydrogen ion concentration of solutions of the normal salts and aggregated larger complexes with lower hydrogen ion concentration present in the basic salts. Thus, increased baficity of the chromic salt will make itself felt in two respects, both favorable for the tanning effect: ( I ) increased size of the complex, facilitating multipoint attachment to the protein, and ( 2 ) lowered hydrogen ion concentration of the tanning bath, which means a greater number of carboxyl groups of collagen in the ionic form. A competition for these ionic groups of collagen is set up between hydrogen ions and cationic chromium complexes. In the chrome tanning agents usually employed a certain upper limit to the pH values of the solutions is set by the fact that the incorporation of too many hydroxyl groups will lead to such a high degree of aggregation that the ensuing compound will be too bulky and insoluble. Undoubtedly the hydrogen ion concentration is a very important factor, in combination with the factors discussed earlier. However, a
PROTEIN-CHEMICAL ASPECTS OF TANNING
387
too one-sided accentuation of one single factor may be precarious as shown by recent tendencies (146). According t o the hydrolytic concept (32), as modified by McLaughlin (146), the acid-binding function of collagen is the governing factor in chrome tanning. This hypothesis is in conflict with scveral fundamental facts of physical chemistry and our knowledge of chrome tanning. Among the many experimental findings contradicting the hydrolytic hypothesis (88) the following may be mentioned: Since the acid fixation by collagen reaches equilibrium in 24-48 hours and the chrome and acid fixations are interrelated, the chrome fixation should come to a standstill within the time given (87,88). This is not the case. Furthermore, the chrome fixation by various proteins is not a direct function of their acid-binding capacity; the ratio of fixed chromium to fixed acid in equivalents shows values ranging from 1.0 for silk fibroin t o 10.3 for blood fibrin (88). If tanning occurs in a n unchangeable system as assumed by the hydrolytir concept, the chrome fixation should be independent of the temperature of tanning, since the fixation of strong acids is not affected by the temperature (87). In reality, the chrome fixation is greatly increased by augmented temperature (56,151,164). Collagen pickled with acid and in equilibrium a t a given p H value, e.g., 2.5, should not take up chrome from a solution of basic chromic sulfate of a higher p H value, e.g., p H 3.5, according to the hydrolytic concept. However, this is not true. The practice of chrome-tanning pickled stock of lower p H value than that of the tanning bath is in fact in direct contradiction t o the adsorption hypothesis. Finally, the excellent tanning artion of certain complex salts in solutions with p H values on the alkaline side of the isoelectric point of collagen invalidates this concept (57). Also the findings of the neutriil salt effect reported conflict with this hypothesis, since no direct relationship exists between the hydrogen ion concentration of the solution of chromic salt in the presence of neutral salts and the fixation of chromium by collagen (88). Still, the hydrolytic concept must be given the credit of having called needed attention to this important detail of the chrome-tanning mechanism, particularly Elod’s scholarly contributions (32) demonstrating the importance of changes in the chromic salt fixed by the protein during the tanning and in the subsequent processes.
The rate of chrome fixation is increased by increase of temperature (151). A number of factors are involved in the temperature function of the system. The equilihria in hydrolytic systems are complicated by constitutional changes of the solutes which are greatly temperaturedependent in regard to the degree of hydrolysis (pH), degree of aggregation of the basic chromium compounds, their constitution, and electrochemical behavior. Furthermore, the protein component, being an amphoionic structure, is highly temperature-dependent, the ratio of charged to uncharged protein groups decreasing with increasing temperature (87). f. Influence of Previous History of Collagen. The influence of the hydrogen ion concentration of the environment upon the degree of reactivity is a general phenomenon, discussed in connection with the various factors. The effect of various pretreatments of collagen, leading to certain irreversible changes of the protein, upon the chrome fixation
388
K. H. OUSTAVSON
will be briefly considered, since it supplies certain information regarding the mechanism of the reaction (66). The acid-binding capacity of collagen is not changed by its pretreatment in strong solutions of lyotropic neutral salts, as, e . g . , 1-2 M calcium chloride and thiocyanate, nor by previous prolonged immersion in solutions of hydrotropic agents, such as urea in concentrated solution, weak organic acids, or 2-3 M acetic acid; furthermore hydrothermal denaturation does not alter this acidbinding capacity. The fixation of chromium by collagen from solutions of simple cationic chromic sulfates and chlorides generally used in tanning is not influenced by these pretreatments, with the exception of the acetic acid pretreatment, which leads to increased chrome uptake, probably as a result of the liberation of acidic groups of collagen by means of the deamidation occurring in prolonged treatment with acid solutions. Accordingly, the conclusion seems justified that forces of coordinate nature do not govern the primary fixation of simple chromic salts. However, the fixation of highly aggregated salts, e . g . , those with acidities less than 50%, is a function of the pretreatment, considerably increased chrome fixation being noted by pretreated collagen. This increase may be of the order of 75-100% of the original fixation for the highly aggregated sulfito-sulfato chromiate, containing mainly uncharged and negatively charged chromium complexes (Table 111). The coordinate effect is also evident in the fixation of compounds of the type of tetraoxalatodiol chromiate (57), and generally in reactions involving coordination of the tanning agent on collagen (84). By prolonged pretreatment of native collagen in solutions of alkali in the range p H 12-13, the coordinate reactivity of collagen as well as its ionic reactivity by means of the carboxyl groups is increased (66,71). The uptake of the simpler types of chromium compounds which are not affected by changes in the coordinate property of collagen is increased by the alkali pretreatment, indicating ionization of carboxyl groups. 3. Nature of the Chrome-Collagen Compound
Even such small amounts of combined chromium as 0.5-1.5 g. per 100 g. collagen impart a high degree of stabilization. With chromium contents of or greater than 2-2.5 g. per 100 g. collagen, the leather will as a rule withstand the action of boiling water. I n earlier sections it was pointed out that a tentative explanation of the reaction mechanism of basic chromic salts with collagen is given by the concept of the formation of a modified type of internal complex salts (chelate compounds, involving groups of different protein chains). According to this concept, the initial reaction is an ionic interaction of cationic chromium complexes,
PROTEIN-CHEMICAL ASPECTS OF TANNING
389
such as (Cr20SOd),2n+, with the charged carboxyl groups of collagen, the sulfate ions being compensated by the NHs ions. The carboxyl groups, having a great tendency to complex formation and for a direct attachment to chromium, penetrate into the coordination sphere, forming a covalent-coordinate bond. Since several chromium atoms are present in large chromium complexes, and in view of the secondary aggregation of the fixed chrome complexes by further hydrolysis, possibilities may be a t hand for a multipoint interaction of one chainlike chromium complex with several carboxyl ions of the collagen lattice, resulting in the linking of adjacent protein chains by strong bonds by means of the chrome bridge. Shuttleworth's conductivity data for gelatin solutions containing chromic salt point to the inactivation of carboxyl ions a s the main reaction (182). This type of cross linking an olated chromium complex, containing several
/
oc \ HC /
NH
X--'C:..-OH,
/
\
'.
/
HO.
'b
O )
*\
HO, 'OH (CH,),COO-'CT-------NH,(CH,)~CH /
'.'OH
/
H2O-.C?-X
/
,% ,
\ HN
\
FIG. 3.
chromium atoms, by means of two or more carboxyl groups on adjacent protein chains, accounts sat,isfactorily for the chemical behavior of chromed collagen, particularly the increased reactivity of the basic protein groups resulting from the chrome tannage ((31). I n view of the new data (Table 11) of the free carboxyl groups, previous objections based on spatial and stoichiometric considerations are largely removed. The numerous hydroxyl groups may further act as supplementary links since the oxygen of this group possesses a residual negative charge (183). Moreover, the chromium atom possesses great coordination power. Coordinate centers are numerous in the protein chains in the form of basic groups and particularly as hydroxyl and peptide groups, containing both carbonyl and imino groups with coordination-active 0- and H- groups. A possible type of combination is illustrated by Fig. 3. The postulation of the formation of secondary stabilizing links between the chromium complex and these groups is chemically sound and may represent the most probable course of the reaction, although direct evidence for this concept has not yet been adduced. Indications th a t chrome fixation by collagen is intermicellar, only affecting the surface ionic groups of micelles,
390
K . H . GUSTAVSON
have been supplied by Kiintzel (122). The hardening of chrome leather upon drying and certain aspects of its reactivity are in accord with this view. The most direct indication is of a qualitative sort (130). The bluegreen solution of basic chromic sulfate, diffusing into a gelatin gel, imparts a strongly violet color to the layer penetrated. Chromium compounds containing carboxyl groups directly attached to chromium, e.g., chromium acetate, show the very same color. The direct interaction of carboxyl groups of gelatin with the chromium atom is thus indicated to take place. It was early recognized (54,62) that a prototype of the chrome-collagen compound was the structure of the internal complex salts which copper and chromium form with simple amino acids. These compounds have been extensively investigated by Ley (136), who together with Pfeiffer (172) first conceived them as internal complex salts. In such structures the central atom, chromium in this casc, completes its coordination number by primary (covalent) interaction with the carboxyl groups and coordination with the amino groups of the same molecule, forming a stable system. The diol-chromium glycinate (137) is the classical example of these interesting compounds and especially useful for the present problem:
In chrome tanning various groups of adjacent protein chains are probably incorporated into the coordination sphere of the aggregated chromium complex, resulting in the riveting together of different chains on the same complex. The nonionic nature and the high degree of stability of the formed chromium-collagen compound is thus explained. Kiintzei’s spectrophotometric investigations of chromium glycinates and the complexes formed between gelatin or degraded collagen and chromium nitrate show very similar type of curves (124,130). This similarity constitutes the best available indication of the presence of structures of the type of internal complex salt as the reaction product of collagen and gelatin with basic chromium salts. By inactivation of carboxyl groups of collagen by methylation, the chrome fixation decreases about 70%. I t is interesting to note that the chrome taken up by collagen lacking in ionic carboxyl groups does not sfabilize the structure, although the same amount of chromium in untreated collagen markedly increases its shrinkage temperature and
PROTEIN-CHEMICAL ASPECTS O F TANNING
39 1
results in tanning. Deamination of methylated collagen only causes a further slight decrease in chrome fixation (13a). These interesting findings are in harmony with results obtained in a study of the fixation of chromium by collagen, with its carboxyl groups completely discharged (pH 1.0) from solutions of basic chromic sulfates (68). Acid-saturated collagen fixes only small amounts of chromium, which do not exert tanning effects, as measured by shrinkage temperature and tryptic resistance of tlhe stock. These investigations also show that ionized carboxyl groups play a major part in the fixation of chromium and for the tanning effect. These findings further suggest that the chromium complexes do not combine with amino groups independently of carboxyl groups. The interaction of chromium complexes with electrovalent groups of collagen should be expected to shift the isoelectric point of the original protein. By removing the protein-bound acid of hide powder, tanned with basic chromic sulfate, by means of pyridine (60), and then carefully freeing it from the latter, the isoelectric point of the chromed collagen was located in the p H range 6-7 (212) by the dyestuff fixation technique (206). The isoelectric point of the original hide powder was 5.5. The dioxalatochromiate tannage shifted the isoelectric point toward the acid side, p H 4 . 0 4 . 5 being obtained (212). Thus, the fixation of cationic chromium complexes by collagen leads to inactivation of acid protein groups, whereas the anionic oxalato compound is indicated to combine with basic protein groups. In recent investigations by Theis (196), examining suspensions of tanned hide powder in a cataphoretic cell using buffer solutions, values in the opposite directions to those previously mentioned were obtained. According to Theis’ findings, the cationic chromium complexes should mainly inactivate basic protein groups. He found collagen tanned with chromic sulfate isoelectric a t p H 4.50. Determinations of the isoelectric point of chrome-sulfate-tanned hide powders, using borate buffers, which do not interefere with the composition of the fixed chromium complex, by cataphoresis, gave values of pH 6.5-7.0* The oxalato-chromed hide powder did not migrate a t pH 4.5.* By the use of complex-forming buffers, containing phosphate and acetate, the composition of the cationic chromium complex was radically changed, leading to dislocation of bonds. The sulfate-tanned hide powder then was isoelectric a t p H values below 5. The importance of employing buffers without complex affinity for chromium is strikingly demonstrated. By the electrokinetic technique of Neale (160), the isoelectric point of the pelt tanned with cationic chromium sulfate was * Determinations carried out at the Swedish Institute of Textile Research, Gothenburg (unpublished) by Mr. B. Olofsson whose cooperation is gratefully acknowledged.
392
K. H. GUBTAVBON
found to be in the vicinity of p H 7 ; that of oxalato stock a t p H 4.* These values were obtained in the absence of buffers in 0.02 M potassium chloride solutions. Theis’ findings are probably vitiated by the secondary action of the unsuitable buffers employed (phosphate and acetate). It is interesting to note that by light drying or acetone dehydration of chrome leather its isoelectric point is displaced toward the acid side (94), possibly due to gradual inactivation of basic protein groups.
4. Some Important Properties of Chrome Leather of Theoretical Interest The most significant property of chrome leather is its high degree of hydrothermal stability as mentioned previously. A well-tanned chrome leather will withstand a few minutes’ boiling in water. I n a recent method the determination of shrinkage temperature of leather is carried out in a glycerol-water mixture in order to assess shrinkage temperatures above 100°C. (148). Objections may be raised against the use of such mixtures since the environment is altered. The use of cod skin, immersed in water, for the determination of the degree of stabilization of the collagen lattice by chromium salts is free from this objection. Cod skin with a shrinkage temperature of 45°C. leaves a range of 55°C:. to the boiling point whereas bovine skin only gives a margin of about 30°C. A fully tanned chrome leather is not digested by trypsin, as first noted by Thomas and Seymour-Jones (208). At that time nothing definite was known about the relation of cross linking and chemical structure, on one hand, and the stability of proteins toward proteinases, on the other hand. Since trypsin generally is considered to hydrolyze peptide groups of proteins, it was natural to conclude that chrome fixation is localized to and inactivates peptide groups. However, i t was later found that native collagen with its internal cohesion intact is resistant toward trypsin and further that structures stabilized by means of covalent cross links, e.g., keratin, are not attacked. Accordingly, the inertness of chrome leather toward trypsin only shows that the protein structure is effectively stabilized by the new cross links (64). By chrome tanning of collagen, its reactivity for acid dyestuffs of the sulfonic acid type is greatly increased as is also the irreversible fixation of vegetable tannins (61). Since in the pH range (3-5) involved in the latter reaction collagen contains a large part of its ionic groups internally compensated and thus not available for interaction with the reactants mentioned, their fixation will be governed by the degree of ionization * Determinations carried out by Professor 5. M. Neale (College of Technology, Manchester, England), whose kind cooperation is gratefully acknowledged.
PROTEIN-CHEMICAL ASPECTS OF TANNING
393
of these groups according to the equilibrium pH of the system; and only that part of the total binding capacity of collagen can be utilized by the anionreacting agents. The chrome-tanned collagen, on the other hand, contains all or the greatest part of its basic groups in the free, reactive state, independent of the p H value of the system, as the result of the breaking up of the internal salt structure of collagen by the ionic interaction of the carboxyl ions with chromium complexes. I n order to obtain the same degree of fixation of vegetable tannins by collagen as by chromed collagen, extensive periods of tanning (several years) (169) and high hydrogen ion concentration of the tanning systems are required; whereas the maximum fixation of vegetable tannins by chromed collagen is reached within a few days’ time by applying vegetable tannins in solution of p H 4-5 (61). This activation of the vegetable tannin fixation is of practical importance and extensively applied in the manufacture of combination-tanned leather, although it was not until recently that this important function of the chrome-tanning process was fully realized. It is a n excellent example of the import.ance of the availability of reactive protein groups. The acido groups of the chromium complexes fixed by collagen may be completely or partly displaced by other groups, which may result in changed hydrothermal stability. In reactions in solutions certain anion series are shown, e.g., oxalate, acetate, formate, sulfate, chloride, and nitrate, with the oxalate possessing the greatest and the nitrate group the lowest degree of affinity for chromium (193). I n the interaction of complex-forming salts with chrome leather this order may be modified by the mass action effect. In the treatment of collagen tanned with basic sulfates, the protein-bound sulfuric acid being previously removed, only slight amounts of sulfate are present in ionic form in contrast to the high concentration of salt used for the treatment. The mass action effect of the solute will dominate over the complexforming tendency. This will explain why sulfato groups of chromium-sulfate-tanned collagen are nearly completely displaced by treating the stock for a few days in a 1-2 M solution of sodium chloride containing a group of low degree of complex formation (65,66). The original leather, standing the boiling test, will upon such a treatment show large shrinkage in boiling water. On the other hand, the treatment of a sulfate-tanned leather of a shrinkage temperature of, e.g., 93°C. for a few hours in a solution of 1 M sodium sulfate will mostly result in a leather resistant to boiling (66). In order to explain the stabilizing effect of sodium sulfate on the sulfato-chromecollagen eompound, it has been suggested that by the penetration of sulfate groups into the internal sphere several sulfate groups may become associated with one chromium atom in the polynuclear complex. This particular chromium atom with an excess of sulfate groups acquires negative charge (131). Accordingly the fixed complex will function as a n amphoionic structure; t h e negatively charged chromium atom serving as a locus for additional stabilization of the structure by its electrostatic interaction with positively charged protein groups. The concept of Latimer and Porter (133) regarding the residual charge of a particular atom, e.g., N in NHz or NHa+,applied by Smythe and Schmidt (183) to ferric proteins is also of interest in thia connection (68).
394
K.
€I QUBTAVSON .
These examples ehow the importance of the constitution of the chromium complex fixed by collagen for its hydrothermal stability and for other physical properties. Evidently the valency partition within the chromium complex, especially the nature of the acido groups, determines the efficiency of the chromium atom as a center of coordination, a function of fundamental importance for the stability of the chrome-collagen compound. The state of ionic protein groups and steric conditions of the protein chains are the main factors of the protein component.
VI. VEQETABLE-TANNINQ PROCESS 1. General Properties of Vegetable Tannins
The tannins, polyphenols of complicated structure, may conveniently be classified as: (1) Hydrolyzable tannins, which are esters of hexoses and phenolcarbonic acids, the simplest type being tannic acid. ( 2 ) Condensed tannins or the catechin type, the latter being the prototype for this important category. Hydrolyzable tannins containing ellagic acid are sometimes considered as a third group. Tannins produce on the tongue the sensation of puckering. Substances causing this effect are called astringent in proportion to the degree of the effect. A numerical classification of the various tannins and vegetable-tanning materials according to their degrees of astringency is not possible, since a number of factors are involved. The astringency appears to be a function of the molecular size of the tannins, the proportion of tannins to total solubles (degree of purity), the hydrogen ion concentration of the solution, the electrical charge of the particles, nature of anions and neutral salt present, temperature, and tannin content of the solution. Finally, the state of the protein substance is an important factor. It has been assumed that the affinity of the tannins for collagen, at comparable experimental conditions, is a measure of the astringency. I t has also been suggested that the type of protein groups involved in the fixation of the tannins by collagen may affect the astringency (70,83). Quebracho and tannic acid are classified as astringent tannins, whereas gambier and myrobalans, possessing lower degree of affinity for collagen, belong to the nonastringent group. Table I V contains data for some important tannins, regarding their source, type, degree of purity, pK value (5a), and molecular weight (30), measured by the depression of the freezing point of electrodialyzed solutions (1-2.5% total solubles). The molecular weight of the tannins purified by dialysis is in the vicinity of 1000-4000 (30,110) but also higher values are indicated; the
395
PROTEIN-CHEMICAL ASPECTS OF TANNING
molecular weight is influenced by the hydrogen ion concentration, high acidity tending to aggregation. The problem of the charge of the tannins has been variously interpreted. Some investigators have found them to be uncharged (18); others consider them to be charged (202). Since the polyphenols are very weak acids (pK > 6 ) , the ionization of the phenol groups should be marked only at rather high pH values and also their reactivity as electrolytes depressed a t low pH values. The tannins containing stronger acidic groups, e.g., carboxyl as in valonia (185), cerTABLE IV Some Properties of Vegetable Tanning Material8 Material
Source
Galls on leaves of species of Quercus and Rhu8 Hydrolyzable Fruits of Terminnlic Hydrolyzable Myrobalans chebula Wood of Caetanea Hydrolyzable Chestnut vesca Leaves of Uncaria Condensed Gambier species Condensed Mimosa (wattle Bark of Acacia species bark) Wood of Quebracho Condensed Quebracho colorado Wood of Quebracho Condensed .Sulfited colorado quebracho
Molecular weight
Purity
Tannic acid
89 62
5.c 4.5
3 100-3400 1900
76
5.c
1550
55
4.5
520
79
> 6.0
1600- 1700
89
> 6.0
2420
89
> 6.0
760
tainly react ionically with collagen in the pH range (3-5) usually encountered in vegetable tannage. Recent investigations of purified solutions of tannins by electrophoresis show a part of the wattle and quebracho tannins as well as the valonia tannins to be charged (29). However, the high-molecular fractions did not migrate (29). Reactions between vegetable tannins and collagen generally seem to involve both electrovalent and coordinate forces, the latter type being the dominating one in some tannins. Investigators of the mechanism of vegetable tannage have surprisingly enough emphasized either the one or the other function. However, the final answer will probably have to recognize both types of reaction, with preference for a particular reaction in certain instances. That will not necessarily mean that one type of tannin is attached by electrovalent reactions to the basic protein groups and another type of tannin coordinated on peptide groups. It rather
396
K. H. OUSTAVSON
seems that the very same large molecule may interact with collagen by means of both types of valencies (multipoint fixation). Mechanical deposition of the tanning material in hide probably accounts for a large part of leather formation. 2. Factors Governing the Reaction
a . Hydrogen Ion Concentration. The hydrogen ion concentration of the system is probably the foremost factor in vegetable tanning. Our present knowledge of this factor is due mainly to the pioneering investigations of Thomas and his school (205). This effect is very intricate. First, it involves the influence of the hydrogen ion concentration on the tannins, in regard to their charge, degree of aggregation, and secondary chemical reactions (at high pH values). Second, the effect of the pH of the system upon the collagen substrate is important, and no doubt is the deciding factor with respect to the degree of swelling, the accea-
0
2
4
6 8 Final pH
1
0
FIG.4.-Average curves of the irreversible fixation of vegetable tanning by hide powder as a function of the concentration of the tannin solutions at attained equi, 2 weeks’ tanning; - - -, 24 hours’ tanning. librium (212).
-
-
sibility of reactive groups, and the internal resistance of the hide structure to the diffusion of the tannins. In this tannage the state of the hide fiber weave enters heavily. Fig. 4 shows the average curves of the fixation of six commercial tannins by hide powder run in series for 24 hours and 2 weeks, as a function of the final hydrogen ion concentration of the systems (212). The sharp minimum of fixation in the pH range of the isoelectrio point of collagen is especially noteworthy, as is the greatly increased tannin fixation which occurs on lowering the pH value. On the alkaline side of the isoelectric point a marked fixation also takes place. However, a sharp drop is clearly evident a t pH values greater than 8. These fixation curves illustrate the irreversible fixation of tannins, i.e., the part
PROTEIN-CHEMICAL ASPECTS OF TANNING
397
which forms a compound with collagen resistant to water of p H values corresponding to the equilibrium p H values of the tanned stock. The total sorption of substances by collagen is far less pH-dependent than the irreversible tannin fixation. The curve of the irreversible tannin fixation after 2 weeks’ tannage is also more evened out in the whole pH range than the corresponding curve from the short period of tanning. The formation of coordinate compounds between tannins and collagen is indicated to be the main reaction in the pH-independent process. Simple prototyes of such molecular compounds between amino acids and anhydrides, on one hand, and phenols, on the other hand, have been isolated by Pfeiffer (172). The importance of such coordinated systems for the vegetable-tanning process has especially been emphasized by Freudenberg (44) and Stiasny (193). The coordinate reactivity of collagen should be pH-independent over the whole pH range if swelling and rupture of hydrogen bonds is excluded; this is shown in fixation of tannins in solutions of pH values lower than the pK values of the phenolic groups, since intact hydroxyl groups are necessary for coordination. The influence of the hydrogen ion concentration on the molecular weight of the tannins is a complicating factor, more or less prominent according to the type of tannins. b. Protein Groups Involved. The tannins contain numerous phenolic groups interspaced on the chainlike structure and also, more sparingly represented, ionic groups which interact with favorably located protein groups of adjacent chains, resulting in a multipoint attachment of the tannins. According to the nature of the tanning agent, especially the relative proportions of electrovalent and coordinate groups, and the experimental conditions, primarily hydrogen ion and tannin concentrations and time of interaction with a given collagen substrate, the preponderance of one of the two main binding types will be determined. Our present knowledge of the protein groups concerned in this tannage will be briefly reviewed. Regarding the basic groups, the first direct demonstration of their participation in the fixation was adduced by Thomas and collaborators (204), who showed that deamination of collagen (removal of c-amino groups of lysine) decreases its capacity for irreversible fixation of tannins. Further, by complete inactivation of the basic protein groups by means of irreversibly fixed sulfo acids, e.q., pentanaphthalenetetramethylenedisulfonic acid, a more or less marked decrease of the tannin fixative capacity of collagen will result; this constitutes further evidence for the importance of the basic groups (70). In a comparative study of tannin fixation by untreated hide powder and hide powder with inactivated basic groups, the following findings are of interest (70). The former will react with all available groups, ionic and nonionic, and the latter mainly with the nonionic groups (peptide
398
K. H. GUSTAVSON
groups). The difference of the fixation obtained in the two series represents the fraction of tannins fixed by the basic groups. With increasing hydrogen ion concentration the participation of the basic groups is more prominent. The hydrothermal stability, imparted to the fibers by the action of the tannins, is found to be associated mainly with the fraction of tannins attached to the basic groups. The equilibrium of the reaction with the basic groups is rapidly attained. It was further indicated that the uptake of fannins by collagen upon prolonged tanning is mainly accounted for by the fraction reacting with peptide groups. Further, a t increased concentrations of tannins as well as of hydrogen ions, the peptide-bound fraction of tannins is not affected; the increased reactivity is due mainly to the ionic protein groups (70). Further evidence for the participation of the basic groups in vegetable tannin fixation may be cited. Vegetable tannins displace rather completely the fixed highpolymeric phosphate in hexametaphosphate-treated hide (94). Displacement of tannins is further shown by certain sulfo acids, which have a selective affinity for the basic groups (85,92). Even 0.1 N solutions of mineral acids may displace some fixed tannins (92). The isoelectric point of hide powder tanned with wattle bark or quebracho tannins is displaced 1 pH unit toward the acid side, or from pH 5.5 to 4.0-4.5, indicating inactivation of basic protein groups (212).
Regarding the nonionic protein groups, primarily the peptide groups as loci for coordinate valency forces, a great array of facts may be cited as indication or evidence for their important function. First, the different behavior of vegetable tannins and simple sulfo acids toward water-soluble urea-formaldehyde condensation products, introduced by Grassmann and coworkers (51), merits attention. The Grassmann reagent, consisting of -NHCO-NHCH2*NH*CO*NH- units, forms heavy precipitates with weakly acid solutions of vegetable tannins, being the most sensitive tannin reagent known. On the contrary, condensed sulfo acids, such as the condensed naphthalenesulfonic compound earlier mentioned, which do not carry coordination-active groups, are not reacted upon or precipitated by the Grassmann reagent, which seems t o be specific for agents reacting with the main bonds of the urea-formaldehyde, the peptide groups. This fact together with certain considerations of the acidbinding capacity of vegetable-tanned collagen led Grassmann and coworkers (51) to the conclusion that the basic protein groups are not involved in the reaction of tannins with collagen. The data from the vegetable tanning of collagen with blocked basic groups also show the presence of nonionic protein groups with affinity for tannins (70). Further indications have been supplied by the study of the influence of certain pretreatments on the amount of irreversibly
PROTEIN-CHEMICAL A S P E C T S O F T A N N I N G
399
fixed tannins. Pretreatment of hide powder in solutions of lyotropic salts of 1-2 M strength does not alter the acid-binding capacity of collagen, but largely increases the fixation of tannins; the increase generally is of the order of 50-100% (55). The same effect, but t o a still larger degree, is obtained by pretreatments in 3 M acetic acid and in 6-8 M urea (71,79). Finally, by hydrothermal shrinkage of collagen the very same trend is shown. As mentioned in the section on chrome tanning, the fixation of simple ionogens is not affected by the pretreatments mentioned, which forms conclusive evidence for the participation of nonionic groups in the irreversible attachment of tannins to collagen (71). c. Degree of Stabilization. Further illustration of this point has been obtained by investigation of the stability of the irreversibly fixed vegetable tannins in leather toward repeated treatment in 6-8 M urea (83). The action of urea on soluble proteins is generally considered to be specifically directed toward the peptide groups, although in view of the dipolar nature of urea, involving resonance with a dipolar ionic form, reactions with ionic groups may also occur to some minor extent. I t is interesting to note that practically all or the greater part of the tannins of the nonastringent class are removed by one week’s treatment of the leather in 8 M urea. The detanned leather shows shrinkage temperatures below that of the original pelt, probably as a result of the denaturation of the detanned collagen by urea. Collagen in combination with astringent tannins loses about half the total amount of combined tannin. However, the extracted leather is not pelty and gluelike as in the abovementioned case but has retained its leathery properties; the shrinkage temperature is lowered only a few degrees. The order of the degree of removal of various tannins by urea solution is practically identical with the series arranged according to the ratio of tannins combined with peptide groups t o those attached to basic groups, commencing with myrobalan and ending up with wattle bark and quebracho tannins. The shrinkage temperature of collagen in combination with sorbed matter, including nontans and reversibly fixed tannins, is markedly increased by the removal of the sorbed matter; increases up to 6°C. have been recorded (168). This is not due to the p H effect on the leather, since even leather tanned a t pH 4-6 behaves similarly (85). Since i t is indicated t h at the initial stage of vegetable tanning involves primarily the basic protein groups, and the later stage coordination reactions (70), it is interesting t o note t h a t the optimum of hydrothermal stability is obtained at rather low percentages of combined tannin (short duration of tannage) (168,197). Further increase of the degree of tannage leads to a slight lowering of the shrinkage temperature. The great decrease of the stability induced by sorbed matter is probably due to a breaking up of hydrogen bonds through the association of sorbed matter on the peptide groups without cross-link formation. Vegetable tannage imparts to collagen
400
K. €I GIUSTAVSON .
increased resistancetoward trypsin (12,75). Of great interest is the drastic removal of fixed tannins by aqueous acetone, reported by Merrill and coworkers (150), and also the fact that tannins dissolved in ethanol are not fixed by collagen (20), which evidently shows that water must be involved in this tannage. By saturation of hide with tannins dissolved in water-miscible organic solvents, e.g., acetone, pressing out the excess of the solution, and subsequent immersion of the impregnated stock in water, tanning takes place, yielding vegetable-tanned leather (176).
3. General Comments on the Theory of Vegetable Tannage
I n their suggestive paper on “Collagen Structure and the Vegetable Tanning Process,” Braybrooks, McCandlish, and Atkin (16) justly remark that very few of the tanning theories take into consideration the very important factor of the molecular structure of the hide itself, and further the fact that the vegetable tanning agents are colloidal. The starting point in their discussion is the accessibility of the amino groups to the tannins. Modifying their concept by including the coordinate groups, which according to data to be discussed later are the groups chiefly affected by swelling, the accessibility factor means a n important advance in our conception of the vegetable tanning process. As pointed out by these authors, the vital factor is the swelling of the micellar structure of the hide in order to make it accessible to the tannins. The parallelism between the curves showing the effect of the p H value of the systems on the vegetable tannin fixation, on one hand, and on the degree of swelling of the hide, on the other hand, is striking. Braybrooks, McCandlish, and Atkin advance as a final proof of their assertion the fact that hydrogen ion concentration does not affect the tannin fixation after the hide structure has been “struck through ” (penetrated) by tannins and thus somewhat fixed by the light tannage. It is shown that the initial swelling is the governing factor in tannin fixation. The influence of the osmotic condition of the hide substrate on its affinity for vegetable tannins will explain the important practical finding of the English school ( 5 ) regarding the controlling importance of the content and type of neutral salts in the tannin solutions and the importance of the ratio of free and total acid of the solution for the vegetable tanning process. The primary role of the p H factor in vegetable tannage on the swelling of collagen was first recognized and experimentally indicated by Vogl (219). In order to demonstrate the fallacy of the Procter-Wilson extension of the Donnan effect (211) to vegetable tanning, Vogl carried out the following, very simple but convincing experiments. Portions of hide powder in equilibrium with aqueous solutions of p H 5 and 3, the latter accordingly considerably swelled, and the former not swelled a t all, were lightly treated with formaldehyde in order to fix the degree of
PROTEIN-CHEMICAL ASPECTS OF TANNING
40 1
swelling. The hide powder specimens were then tanned in solutions of various tannins at p H 3 and 5. The amounts of irreversibly fixed tannins obtained were the same in tanning a t p H 3 and 5 for the specimens of the same degree of initial swelling. This proves the contention of Vogl that the main effect of the hydrogen ion concentration in vegetable tanning is directed toward the protein, regulating its degree of swelling. The data may also be interpreted in terms of the activation of the protein groups involved in the tannage (ionic as well as coordinate valencies), as a function of the hydrogen ion concentration of the system. The importance of the degree of accessibility of the basic groups to tannins is also stressed by Page (167,168). It may be remarked that, from consideration of the amphoionic nature of collagen and the internal compensation of oppositely charged groups, it is self-evident th a t the activity of the basic groups is a function of the hydrogen ion concentration. Since the maximum degree of swelling as well as the corresponding point of tannin fixation coincide with the p H value a t which the complete discharge of carboxyl ions and maximum activation of charged basic groups is accomplished, the concept of accessibility does not necessarily need to be restricted to the osmotic swelling, although i t is likely th a t the micellar changes resulting from the disorganization, due to osmotic forces, in themselves are most important. Page cites some data of Beek (7), who in experiments on the acid equivalent of fully tanned vegetable leather found that only about half of the basic groups had combined with tannins. The maximum degree of inactivation of the basic groups was reached in the initial stage of tannage, a t rather low values of fixed tannin, further tannin combination not affecting the free basic groups. This finding is in complete agreement with the results of the tannin fixation by collagen with inactivated basic groups discussed earlier. Page mggests that the apparent inaccessibility of half of the basic groups to the tannins is due to the inaccessibility of certain basic groups to the high-molecular tannins, deduced from Huggins’ (109) model of collagen structure. Page’s hypothesis is in harmony with findings on the accessibility factor in the reaction of high-molecular sulfo acids to be discussed in the following section. The pH independence of vegetable tannin fixation by chrome-tanned hide (61) discussed earlier has also some bearing upon the problem of accessibility of reactive groups. Since optimal conditions for reactivity of collagen are created by the chrome tannage, the maximum fixation of tannins is easily obtained without the aid of hydrogen ions. Furthermore, topochemical complications due to diminished rate of diffusion of the tannins through the gel-like hide structure are eliminated since swelling is avoided.
402
K. H. GUSTAVSON
I n this connection, it is interesting to note that in tanning hide powder with wattle tannin, which is only slightly acid- and salt-sensitive, at pH 2 and 5 in the presence of 4 volume per cent salt, practically the same degree of vegetable tannin fixation is obtained (94). Since complete discharge of ionic carboxyl groups of collagen and consequently liberation of charged basic groups are obtained at pH 2, without altering the accessibility of coordinate loci on the peptidc groups, which requires disorganization of protein chains by swelling, the findings indicate the primary role of the coordinate type of reaction in vegetable tannage. Further, it proves the insufficiency of the concept of the activation of the basic groups and their accessibility as the regulating factor in the fixation of tannins by collagen.
VII. REACTION OF CONDENSED SULFOACIDS(SYNTANS) WITH COLLAGEN The synthetic tanning agents called syntans are generally condensation products of aromatic hydroxy compounds and formaldehyde, made soluble by introduction of the sulfonic acid group(s). The irreversible fixation of strong sulfo acids by hide protein is regulated by the stoichiometric acid-binding capacity of collagen, the degree of affinity of the sulfo acid anion for the basic protein groups, and the stability of the bonds formed. The reaction of the large sulfo acid molecule containing several reacting groups is primarily governed by the anion affinity, a problem discussed in connection with the reaction of acids with collagen. The effect of the molecular size of the sulfo acid on its reaction with collagen and the nature of the compound formed is illustrated by naphthalenesulfonic acid and its condensation products. @-Naphthalenesulfonic acid, which does not swell hide (pH 2), is partly irreversibly fixed by collagen. The breaking of ionic cross links of collagen by the fixation of the acid leads to decreased hydrothermal stability; the shrinkage temperature being lowered 10-12°C. By condensing three or four naphthalene units by means of formaldehyde and introducing two terminal sulfonic acid groups, the resulting compound will be irreversibly fixed to a large extent. These compounds will convert hide into a leatherlike product. The hydrothermal stability is not decreased. Compounds of still higher degree of condensation yield leather with shrinkage temperatures a few degrees higher than that of pelt. Both the sulfonic acid groups of the irreversibly fixed compounds interact with the protein. Since no coordination-active groups are present in this type of C O W pound, the behavior of these compounds toward collagen illustrates the importance of the molecular size of the tanning agent for stabilization of the proteins. More complicated types of compounds, indicated to react entirely by means of electrovalent forces, are present in the high-
PROTEIN-CHEMICAL ASPECTS O F TANNING
403
molecular fraction of lignosulfonic acid (76,78). These compounds, with molecular weights of 5,000-10,000 and with an average equivalent weight of about 500, are to a great extent irreversibly fixed by collagen. The equivalent weight of the fixed sulfo acid is also about 500 (96). I t has been proved that the lignosulfonic acid molecule, containing from ten to twenty sulfonic acid groups, is irreversibly attached to collagen by means of all available sulfonic acid groups (96). Further, the reversibly sorbed part of the lignosulfonic acid only reacts by means of two to four sulfonic acid groups of the ten to twenty groups present. By the multipoint attachment of the large molecule on collagen by means of numerous sulfo acid groups a highly resistant compound is formed. The previous discussion has concerned the mechanism of the reaction of collagen with such sulfo acids as predominantly react electrovalently. In the tanning agents of sulfo acid type, however, the presence of coordination-active substituents seems to be necessary (phenolic groups). In the remarkable work of Wolesensky (216)) reported in the early twenties, condensed polyphenols not containing sulfo groups were studied. Resorcinol and pyrogallol were condensed by means of formaldehyde to water-soluble products. Particularly the former compounds possessed excellent tanning properties, yielding in neutral solution a full and stable leather. The fixation was quite independent of the hydrogen ion concentration of the system within a wide pH range, which is typical for the coordinate type of protein reactions. However, the accumulation of phenolic groups on a n extended molecule does not lead to a sufficient degree of water solubility of the product, as shown by Wolesensky’s experiments. The correct balancing of the size of the molecules, or rather their size distribution, seems to be the fundamental problem, controlling water solubility and reactivity as well as the degree of irreversible fixation with multipoint linking of the various groups. Polydisperse systems are advantageous, and the proportion of ionic groups (generally sulfo) and coordinate groups (generally hydroxyl) is important. See also the papers of Croad (26) and Kuntzel and Schwank (132). The use of simple condensed sulfo acids for inactivation of the basic protein groups and the influence of this pretreatment of collagen on the subsequent retannage with vegetable tannins have been mentioned in the section on vegetable tannage. The behavior of this type of sulfo acids toward vegetable-tanned collagen throws further light on the mechanism of the two processes. In comparing the fixation of a strong mineral acid (HC1), a condensed naphthalenedisulfonic acid (two to three naphthalene units), and the high-molecular fraction of lignosulfonic acid by dried vegetable-tanned collagen subsequently hydrated a t optimum p H values of the fixation, the following interesting findings were obtained (96).
404
K . H. QUSTAVSON
The fixation of hydrochloric acid by the thoroughly hydrated, vegetabletanned hide powder (tanned by means of wattle, quebracho, and myrobalan tannins) was 85-95 % of the maximum binding capacity of untreated hide powder (collagen). The uptake of the naphthalenesulfonic acid by the vegetable-tanned collagen was 5 0 4 0 % of the figure for untreated collagen and finally the high-molecular lignosulfonic acid reacted with vegetable-tanned hide powder only to 3040 % of the maximum potency obtained by native collagen. Thus, e.g., the wattle-tanned hide powder with a combining capacity of 0.89 milliequivalent hydrochloric acid per gram collagen fixed only 0.31 milliequivalent lignosulfonic acid. The lignosulfonic-acid-treated, wattle-pretanned collagen fixed further 0.52 milliequivalent hydrochloric acid from 0.1 N solution. Hence, free acid-binding groups (carboxyl and amino) are present in vegetable-tanned hide powder in equilibrium with lignosulfonic acid of pH 1.5, the optimum for its fixation. I n the irreversible fixation of the high-molecular fraction of lignosulfonic acid by native collagen and by vegetable-tanned collagen, all sulfo groups react with the protein. Since i t has been demonstrated by various methods (76,78,95) that the reaction of lignosulfonic acid with collagen does not involve coordinate reactions, in the initial fixation a t least, the very drastic decrease of the irreversible fixation of lignosulfonic acid by vegetable-tanned collagen cannot be caused by the inactivation of coordinate loci of collagen by the vegetable tannins fixed. The probable explanation is to be found in the special nature of the lignosulfonic compound, its molecular size, and the presence of numerous strong acid groups in the molecule. Although the incorporation of large vegetable tannin molecules in the collagen lattice and the valency interaction of the same with the reactive protein groups do not materially hinder the reaction of small ions, such as hydrogen and chloride ions, with the ionic protein groups, some sort of steric hindrance may interfere with the reaction of collagen with the very large anion of lignosulfonic acid, which for its stabilization in the protein lattice requires a multipoint attachment. The blocking of the large anion will in its turn affect the fixation of the hydrogen ions, decreasing the same, since the reactions of hydrogen ions and anions of the sulfo acid are interdependent; the maintenance of electroneutrality of the system is essential. The sum total is lowered degree of fixation. The accessibility factor is also involved in the fixation of highly basic chromic salts by vegetable-pretanned hide. In retanning vegctable-tanned collagen by means of extremely basic chromic salts and high-molecular chromium compounds generally, the chrome fixation is only 10-20% of the corresponding fixation by native collagen. Dried vegetable-tanned hide powder has practically no affinity for chrome (94).
PROTEIN-CHEMICAL ASPECTS OF TANNING
405
VIII. TANNING POWER OF ALDEHYDES 1. Tanning Action of Various Aldehydes
Among the aldehydes, formaldehyde holds a unique position as a tanning agent. Since the chemistry of the reaction of proteins with formaldehyde (F.) has been comprehensively treated in Volume I1 of this series (43),only some special problems characteristic of the tanning process will be discussed. A survey of the general behavior of various types of aldehydes toward collagen will be included. Applying the usual criteria of tanning potency, i.e., the degree of hydrothermal stability, inertness to swelling agents and proteinases, and the “leathery drying” of the fibers of the treated skin, it will be found that, among the simple saturated aliphatic aldehydes, the tanning power of the first member of the series, formaldehyde; is outstanding, when the effects of the various aldehydes are compared in solutions of equal parts of water and acetone. Acetaldehyde has a feeble action and the higher homologs are devoid of tanning power. Among the more common unsaturated aldehydes, acrolein and crotonaldehyde show a fair degree of tanning potency (67). By introduction of ethyl and propyl groups in the 2 and 3 positions in acrolein, this property is lost (94). The dialdehydes are an interesting class from the point of view of tanning theory, since the presence of two aldehyde groups on a small molecule would be expected to facilitate the cross linking of adjacent protein chains, as first pointed out by Seligsberger (181). The simplest dialdehyde, glyoxal, is a fair tanning agent, particularly in certain organic solvents (89). Surprisingly enough, the methyl derivative, pyruvic aldehyde, shows excellent tanning properties; in some solvents it is fully comparable with that of formaldehyde (90). Both these aldehydes possess a very marked tendency to condensation and polymerization. However, the tanning effect is indicated to be mainly associated with the monomers. The aromatic aldehydes form a chapter in themselves (46,67). Benzaldehyde and related compounds with the CHO group built into the aromatic structure will combine with collagen, when used in organic solvents or water mixtures of the same (67). However, the effect of their combination with collagen is not a stabilization of the structure. On the contrary, they exert marked hydrotropic action or labilization of the protein lattice, similar to the action of many related organic compounds, e.g., simple phenols. The valency action of these aldehydes evidently is directed toward the hydrogen bond loci; rupturing part of these cross links. The tanning power of aldehydes of diphenols is inter-
406
U. A. GUSTAVSON
esting. However, since the methoxy derivatives behave similarly to the simple aldehydes, being devoid of tanning power, the tanning action shown by dihydroxy aldehydes is probably due to the presence of two phenolic groups, as indicated by Gerngross’ work (46). Aromatic aldehydes containing the CHO group in an aliphatic side chain usually possess weak tanning power (67). At the present state of our knowledge, i t is not possible to describe the tanning action of aldehydes on a common basis. Probably the activation of the aldehyde group by adjacent groups is the underlying cause of the apparently elusive behavior of the aldehydes toward proteins. 2. Tanning with Formaldehyde
a. General Aspects of the Reaction. By the fixation of F. by collagen the strength of the acidic groups is increased (13,99), probably as a result of the inactivation of the basic gr0up.s by F. and breaking up of the internal compensation of the electrovalent groups. Hence, the alkali binding is increased (48) and further, as a direct result of the inactivation of the. basic groups, the capacity of acid fixation decreased (45,192). The isoelectric point of gelatin and collagen is displaced about 1 pH unit to the acid side by F. tannage (47,212). b. Participation of Lysine Groups. Since the development of a quantitative method of determining fixed F. (15,102), an array of data obtained by numerous investigators indicate the main reaction of importance for the tanning function of F. to be located a t the c-amino group of the lysine residue. It is also generally considered that the tanning effect is a result of the formation of cross links by means of F. (46,153). The maximal fixation of F. a t the highest pH value in the range mentioned amounts to 0.4-0.5 millimole F. per gram collagen (15,102). Hence, it is nearly equivalent to the content of lysine residues although this may merely be a coincidence. The formation of -CH*cross links between two adjacent amino groups on different chains should only require half the amount of fixed F. found; this point has especially been stressed by Nitschmann and Hadorn (161) and by French and Edsall (43). It has also been pointed out (67) that the chance that two amino groups of different peptide chains should approach each other close enough for the formation of a methylene bridge should be rather slight (cf. 181). Hence, it was suggested that only a minute part of the bound F. is bridge-forming, the main part reacting without interlocking the collagen chains, with the formation of simple -NH*CHvOH structures. The drastic change of the internal cohesion of polystyrene polymers effected by the introduction of covalent bridges (187) (one bridge per 33,000 residues)
PROTEIN-CHEMICAL ASPECTS OF TANNING
407
in the form of divinylbenzene in the polystyrene linear polymers was mentioned as an analogous reaction. * A very important advance in our concept of F. tannage is marked by the researches of Nitschmann and Hadorn (161), who concluded from their comprehensive investigation of the reaction of F. with casein, that bridges the F. cross linking takes place by the formation of -CH2between an e-amino group and the imino group of the peptide link of adjacent chains. The concept of Nitschmann and Hadorn does away with the steric objection to the hypothesis of methylene bridge formation between two amino groups of differentchains, since the assumption of the proximity of amino and peptide groups on adjacent protein chains seems reasonable. Further, it also recognizes the bonding strength of the compounds formed, as remarked by French and Edsall (43), and agrees quantitatively with the ratio of one F. bound to one NHs group of F.-treated collagen of maximum F. content in its zone of maximum stability (pH 7-8). c. Reaction of Arginine Groups. The role of the guanidyl group of the arginine residue for the fixation of F. from solutions of final p H values greater than 8, indicated by the investigations of Highberger and Salcedo (103), is particularly interesting in view of the high pK value of this residue in the simple amino acid (pK about 13), which would require p H values about 3 or 4 units greater than the value given for reaction with the discharged guanidyl group, if the pK of the guanidyl group of arginine as amino acid applies to this group built into peptide chains. The very same discrepancy is noticeable between the p H of the reactivity of the lysine amino group (pH 5-8) and the p K of the simple amino acid (pK 9). It seems likely that the influence of the ionic environment of the protein groups as well as the effect,of F. may change the pK value of these amino acids considerably (69,86). The shift of the equilibrium NH3+-+ N H z groups, due t o inactivation of NHz groups by F. should not be ignored (73). The F. fixed by the arginine residue does not stabilize the protein structure, since it has been shown that deaminated collagen reacting with F. a t pH 11-13, although fixing 0.4t o 0.5 millimole F. per gram collagen, retains the shrinkage temperature of the original deaminated pelt and is more extensively swelled by strong acids and considerably less resistant to trypsin than the untreated deaminized stock (73). I n view of the possibility of cross linking by F. of the guanidine and lysine groups occurring in simple prototypes (42), the lack of stabilization strictly proves only the absence of cross links between arginine residues. Recent
* Cross links between pairs of amino groups by means of condensed F. are too labile to withstand washing (private communication from Dr. H. Fraenkel-Conrat).
408
K. H. GUSTAVSON
criticism of this view (101), based upon the difference of the shrinkage temperature of collagen treated a t pH 12 and F. collagen tanned at pH 12, is unwarranted because of lack of information, and further has no bearing upon the findings reported which are based on the behavior of deaminated collagen. Yet it must be pointed out that the interrelationship of shrinkage by heat and by swelling has been overlooked (101). In comparing pelt and F. pelt of such high alkalinity as pH 12 (the point of maximum alkaline swelling), the swelling of the blank (pelt) will yield values of shrinkage temperature far too low, whereas this effect is practically eliminated by F. tannage. The value of the shrinkage increase will be too large a t p H 12. If the experiments are carried out in 5 % solutions of swelling-depressant sodium sulfate, it will be found that the increased hydrothermal stability obtained by tanning a t p H 12 is practically equal t o that obtained in tanning a t p H 8, compared to the blanks (pelt) (93). This also rules out bridge formation between lysine and arginine groups by means of condensed F. (93). I n both instances of tanning the stabilizing reaction is apparently located in the e-amino groups involved in F. fixation a t the low pH range (73), which, of course, also is a part of the total F. fixation in the high pH range. Partial removal of guanidyl groups does not decrease the hydrothermal stability of collagen F. tanned a t optimal pH for these groups, pointing to the nonparticipation of the guanidyl group in the stabilization of collagen by F. bridge formation (69). I n practical tanning the pH range 5-8 is generally preferred (212). Precipitated chalk is an excellent pH regulator (pH 7.5-8). Since this range forms the zone of minimum swelling of collagen and the affinity of collagen for F. is sufficiently great a t these p H values, reactive protein groups being present, theory and practice agree. A marked lowering of the pH value would mean insufficient F. fixation and too low tanning action, while increasing the p H value of the system too far, would promote swelling, rupture of hydrogen bond cross links, lead t o disorganization of the structure, and accordingly create unfavorable steric conditions for cross linking. The interaction of F. with the basic groups is also shown in retanning chrome leather with F. It has been proved (73) that F. displaces the protein-bound acid of chrome-tanned collagen in equilibrium with a weakly acid solution of F. This is an additional proof of the discharge of NH8+ ions by F. in the acid range. If lightly chrome-tanned leather, preferably that tanned by means of basic chromic chlorides, is used, with shrinkage temperature in the range 85-90°C., the F. treatment results in a boiling-resistant leather (shrinkage temperature greater than 100°C.)
PROTEIN-CHEMICAL ASPECTS OF TANNING
409
in a p H milieu of 3 4 . A marked displacement of protein-bound mineral acid by F. is noted (73). Of practical importance also is the uie of F. in the retannage of vegetable-tanned leather, which also is hydrothermically stabilized, indicating the presence of free basic groups in vegetable-tanned leather (70,197). Some interaction of F. with the vegetable tannins fixed in the leather may also occur (70,73). d. Participation ojPeptide and Other Groups. The participation of the peptide groups in F. binding is an interesting part of the problem. A labile fixation of this type seems to occur over the whole p H range, being a pH-independent reaction. This type of reaction may be prominent in solutions of high F. content. However, it does not appear to contribute measurably to the stabilization of the lattice. This is shown in tanning pelt and deaminated pelt at low p H values, e.g., p H 2, adjusted by hydrochloric acid and eliminating the swelling by using 5 % sodium chloride solution containing high concentrations of F. (5-10%). The shrinkage temperature of the pelt will upon prolonged treatment (about 2 weeks) increase 5-10°C., whereas the shrinkage temperature of deaminated pelt remains unchanged (73). This also excludes cross linking of collagen through amide groups but leaves open a possible linking of amide and amino groups. However, in both instances the peptide groups are equally accessible to F. The deaminated pelt will bind F. probably on the peptide and amide groups. The fixation of F. by collagen in the pH range 1-3 is an exceedingly slow reaction, probably connected with the gradual discharge of NH*+ groups, according to: -NHs++ -NH2 H+, high F. concentration and prolonged interaction being required. The cross linking will gradually lead t o increased stability of the lattice (73). The kinetics of the F. binding does not fit a reaction with the amide group. This type of reaction (40,41,217)recently discovered in F. combination with amide-rich proteins (casein and vegetable proteins) cannot be prominent with limed collagen, which only contains small amounts of such groups. However, the greater F. fixation of native collagen (not alkali-treated) compared t o alkali-treated collagen (limed) from weakly acid solutions is, as Highberger (101) points out, logically explained by the F. fixation by means of the amide groups of the nat'ive collagen. The discharge of ammonium ions of collagen in tanning with F. in concentrated solutions a t low p H values is demonstrated by the diminished fixation of sulfuric acid from 0.1 N solutions containing 4 volume per cent sodium sulfate with increasing F. content (73). It is further indicated by the behavior of collagen with its basic
+
410
K . H . OUSTAVSON
groups completely inactivated by polymethylenenaphthalenedisulfonic acid that F. fixed by the nonbasic groups of collagen (amide and peptide) does not stabilize the protein lattice. Condensed naphthalenesulfonic acid is specific for the acid-binding groups and does not interfere with the affinity of the peptide linkages. The F. fixed by collagen with its basic groups completely blocked, probably attached mainly to the peptide groups, does not increase the shrinkage temperature (67). The cross linking of proteins by F. was suggested by Meyer (153). Experimental evidence for the condensation type of cross-linking protein chains by F. was first supplied by Nitschmann and coworkers (lGl), who showed that in the reaction of gaseous F. with casein water is liberated. Fraenkel-Conrat and collaborators (40,41) have furnished further proof. The recent researches of Fraenkel-Conrat and Olcott (42) have largely extended our knowledge of the mechanism of F. tannage. By model experiments, employing simple amides and amino compounds, these authors have demonstrated that a t room temperature and over the range of pH 3 to 7.5,-CH*bridges are formed in the interaction of F. with a primary amide and an amine or amino acid. Such F. condensation products could be isolated and characterized. The reaction between amino acids and amides is favored by alkaline medium, whereas the condensation of amines with amides is facilitated in weakly acid solution. Secondary amides (-COaNHR) do not react with F. to give cross linking. This finding evidently invalidates the hypothesis postulating cross linking of protein by F. through condensation of two adjacent imino groups of the peptide linkages. The improbability of such a linking has been indicated by other experiments, mentioned in the foregoing. The primary reaction of F. with proteins probably is the formation of methyl01 amines. Simple amines condense with guanidines and F., but amides and guanidine do not react. The results of experiments with proteins and macromolecular model compounds, such as polyglutamine, rich in amide groups, were in line with the findings of the simpler models. The authors consider that the F. tanning of proteins of average composition, i.e., containing equal numbers of amino and amide groups, is largely due to the secondary cross linking of amino and amide groups by means of condensed F. (methylene bridges). This reaction probably occurs in F. tanning of native collagen in slightly acid solution, as previously mentioned (101). However, in view of the low content of amide nitrogen of limed collagen, usually employed in practical tanning, this hypothesis does not appear to be applicable to the F. tannage of pelt and gelatin. Summing up, it may be said that the only firmly established fact is that the hydrothermal stabilization and tanning of collagen by F. is bound up with the c-amino groups of lysine residues.
PROTEIN-CHEMICAL ASPECTS OF TANNINQ
41 1
e. Znjluence of Solvent. Since F. in aqueous solution mainly exists as the monohydrate methylene glycol, CH,(OH)1 (210),whereas the carbonyl form predominates in organic solvents (210),it is noteworthy that F. dissolved in the common alcohols, acetone, dioxane, and benzene cxcrts as good a tanning action as it does in aqueous solution (67,178). Any influence of the dielectric constant of the solvent on F. fixation is not cvident (67). Hence, colloidal polymethylene glycols cannot be the effective tanning ngent (67). This speculation is also invalidated by the particular p H function of the F. tannage; the maximum F. fixation being located in the range of slight alkalinity under which conditions the polymerized glycol is rapidly converted into the monomer. In connection with the polymeric forms of F., attention will be called t o a detail of practice. In tanning with F. in neutral or slightly acid solutions great variations in the efficiency of the tannage are a t times experienced, probably because of the failure to take due precautions t o allow for sufficient ageing of the diluted formalin or adjusting its p H value for effective depolymerization. f. Ewald Reaction. Ewald (34)found that heat-shrunk F.-treated tendon spontaneously re-extends t o almost its original length upon cooling. This reaction is specific for F.-treated collagen fibers and has even been proposed as a tes: for both F. and collagen. Evidently, the re-elongation of the shrunk fiber on cooling is intimately connected with the changed micellar tension of the collagen lattice due t o its combination with F. (cross links) (123). It is interesting to note that native elastoidin fibers also show reversible hydrothermal contraction (36). Since this special collagen contains sulfur, the presence of sulfur bridges may explain the unique behavior of elastoidin. Kuntzel (123)applies the Meyer-Ferri (155)concept of the mechanism of the elastic behavior of tendon t o the reversed contraction of F. collagen. According to Meyer and Fcrri, two opposing tensional systems of opposite temperature coefficients arc a t work in elastic structures such a s tendons. In the melting and contraction of the fiber the lcngthwise tension is increased. The cross-sectional tension sets in upon cooling of the fiber, bringing about reorientation of the structural units and accordingly elongation of the fiber (123).
IX. QUINONETANNAGE The remarkable tanning power of p-benzoquinone was discovered 40 years ago by Meunier and Seywetz (152)in the course of their inveeti-
gation of the use of oxidized phenols as photographic developers and hardening agents for gelatin film. Quinone tannage was later comprehensively investigated by Thomas and Kelly (207). Its effect on subsequent vegetable tannage was also investigated. In the fundamental researches of Hilpert and Brauns (105)in 1925,the principles of the reaction of quinone with collagen were established by means of preparation of simple prototypes involving reaction be tween quinones and various amino compounds and by the study of the actual tanning process. The more recent investigation of Stecker and Highberger (188)is in excellent agreement with the older work (105). The established facts of quinone tannage are as follows. Benzoquinone tans well in alcoholic solution (212). In that, as in many other respects, it shows similarity t o F. The monomer is evidently the active tanning agent in this case. From aqueous solutions of pH values less
412
K. H. QUSTAVSON
than 7 collagen fixes the monomer, increasing acidity markedly lowering the degree of fixation. In slightly alkaline solutions polymerized guinones partake in the reactlion. Upon prolonged tanning, formation and fixation of such polymerization products also occur in solutions of pH less than 7. In the reaction of quinone with collagen in solutions of pH greater than 8 the polymerized products are mainly involved (188). The study of model compounds by Hilpert and Brauns (105) indicates that the old conception of the reaction (152), assuming the CO groups to be the active groups of tanning, is incorrect. Instead, the formation of compounds of the type O=
q;:
seems most likely. This con-
cept is also supported by the fact that tetrachloroquinone is devoid of tanning power (198). Quinone tannage can accordingly be formulated as: (1) A rapid reaction of the monomer with the amino groups of collagen, predominant in neutral and slightly acid media. (2) A relatively slow reaction of polymerized quinone in alkaline solutions, probably by attachment of the high molecular products to the peptide groups of collagen. This reaction also seems to occur as a secondary fixation to (1). I t is doubtful if this type of interaction results in tanning. A mechanical impregnation with preformed products and by means of substances polymerized in situ has also been made probable (188). The shrinkage temperature of collagen is raised about 25°C. by quinone tanning under optimal conditions (at pH 6). Hence, quinone is superior to F. in this respect. A recent investigation of Highberger (101) regarding the tensile strength of single fiber bundles of collagen (kangaroo tail tendon) is of interest in this connection. Tanning of the tendon by means of F. or quinone markedly decreases the tensile strength of dry fibers. These observations cause the author to question the general applicability of the cross-linking concept of the tanning mechanism to the mechanical stability of fibrous proteins, since the effect predicted on the basis of cross linking appears to be at variance with the findings. However, it seems possible that two different types of forces are involved in the hydrothermal and mechanical stabilization of collagen; the former being due to intra- and intermicellar cross linking and the latter also to interfibrillar forces (93). Some physiological aspects of quinone tanning may be mentioned. The possible role of quinonelike substances in the hardening and formation of larval cuticles has received a great deal of attention. Pryor (174) first recognized the hardening of insect cuticles &B a tanning reaction, probably due to the oxidation products of an o-dihydroxyphenol which combine with the water-soluble proteins to form a dark brown, insoluble, tanned protein. The recent researches of Fraenkel and Rudall (39) also s ~ p p l yexperimental indications of the formation of stiff nondeformable structures of insects by the stabilization of proteins b y means of quinonelike substances formed i n vivo by deanination of tyrosine. Glyoxal haa also been claimed as a hardening
PROTEIN-CHEMICAL ASPECTS OF TANNINa
413
agent for the nondarkening proteinous parts of insects (177). Finally, the increased solubility of certain naphthoquinone derivatives in solutions of serum albumin, mentioned by Edsall (31),is of interest in this connection. The tanning action of inorganic poly acids, e.g., tungstic and metaphosphoric, zirconium compounds, aluminum and ferric salts, as well as unsaturated oils, has not been included in this survey; the reason is not only consideration of space but primarily the lack of fundamental theoretical knowledge of these tanning systems, and secondly their limited application.
X. GENERALCOMMENTS I n the processes of tanning and leather formation complexities in addition to the intricacies of heterogeneous systems of high molecular proteins are introduced by the two phase nature of the systems. The insoluble lattice-structured protein differs in reactivity fundamentally from the soluble proteins, since factors of little or no importance for the latter are often of governing importance for the behavior of the fibrous protein. The inherent properties of a biological product, such as the weave structure of the fiber bundles, the organization and macro structure of the fibers and fibrils, and the micro and molecular structure of the micelles and protein chains, show variations according to the origin of the hide, the location of the sample tested, and the previous history of the hide prepared for tanning, including the nature of the pretreatments. Beside being a function of purely chemical factors, the interaction of tanning agents with hide protein is intimately bound up with the physical state of the substrate, particularly the rate of diffusion of the tanning agents, frequently in polydisperse solution, into the interior of the hide (macroscopic) and into the interior of the fibrils. Finally, the tanning agent may preferentially react on the surface groups of the micelles, intermicellarly, or penetrate into the micelles, reacting with the individual protein chains (intramicellarly). The attainment of equilibrium is a vexing problem in this type of reaction; and this difficulty is one of the greatest obstacles in the investigations of this field. Topochemical reactions are of major importance, and the conditions for reaction may accordingly be altered during the process. The main reaction and ultimate nature of tanning is incorporation of substances possessing affinity for various protein groups; the tanning agent being immobilized in the protein lattice. The protein enters into a more or less stable combination with the tanning agent which in its turn leads t o stabilization of the protein structure. The most effective way of obtaining this appears to be the function of the tanning agent as an artificial bridge between reactive protein groups of adjacent chains. The possibility of some kind of depolarization of the active valency centers of the protein by means of the tanning agent, the type of long range effect,
414
K. €GU8TAVSON I .
may also be involved, as is the simple inactivation of hydrophilic protein groups. It is likely that the strength of the valency forces between the fixed tanning agent and collagen does not in itself govern the efficiency of the stabilization. The spatial factor exerts a marked influence on the reactions, as, for instance, the size and shape of the introduced molecule and the distance between reactive protein groups and between those and the reactive loci of the tanning agent. Knowledge is lacking regarding the elementary background to the problem of the spatial relationship of the reactants, the sequence of amino acid residues, and their spatial environment. Furthermore, some tanning agents probably interact with collagen mainly intermicellarly, not affecting the interior, ultimate units, but only the surface-active groups of micelles and fibrils which offer quite different spatial environment and valency distribution than the more closely packed protein chains. The whole subject of fine structure of proteins of the collagen group, and the problem of the mechanism of tannage, definitely invites profound investigations by application of new tools and refined techniques. It is obvious that our conceptions of the nature of tanning processes are based upon and reflect the status of the fundamental chemistry of the interesting groups of substances concerned in these complicated reactions. The problem of tannage is not confined to industrial applications. It appears to be of general importance for the elucidation of the nature of metal-protein compounds, and other complexes of proteins and smaller molecules, with important biochemical functions. ADDENDUM ADDED I N PROOF
Bowes and Kenten (220) recently presented detailed data on the combination of modified collagens with various tanning agents; some results of this investigation were mentioned in the text (preliminary statement of €?owes (13a)). Of particular interest is their finding that no chromium is fixed by methylated collagen (with its carboxyl groups completely blocked) from dilute solutions of chromium sulfate of pH 24. The original shrinkage temperature of the methylated collagen is not changed. From these findings and the behavior of deaminated collagen, the authors concluded that the combination of chromium salts and collagen involves coordination of both the carboxyl and amino groups with the same chromium complex. Their data confirm earlier findings of the nonreactivity of collagen with its carboxyl groups in the nonionized state towards cationic chromium complexes (68). Since complete methylation of collagen requires repeated treatment by the methylating agent (methyl sulfate), the resulting preparation of collagen may be radically modified in many reHpects, as indicated by its behavior towards other tanning agents. Hence, precaution is needed in drawing conclusions. In two papers on the mechanism of the hydrothermal shrinkage of collagen (tendon) (221,222), Weir opened a new approach to the investigation of the nature of tanning processes. The thermodynamic treatment of tanning processes appears to
PROTEIN-CHEMICAL ASPECTS OF TA N N I N a
415
be (very) promising. He (221) measured the coefficient of the cubical expansion of hative tendon and tanned tendon in water of increasing temperature up to the point of complete shrinkage. The data are interpreted a8 indicating that shrinkage does not occur a t a characteristic temperature but is a rate process, involving a reaction of the first order. This is in agreement with earlier findings on skin collagen (50,79,170) and also with Rudall's observation of shrinkage curves of sigmoid form in the contraction of epidermin within a wide temperature range (177). Weir reported the heat of shrinkage of untreated tendon collagen to be 141 kcal./ mol., the entropy 349 cal./mol. deg., and the free energy 24.7 kcal./mol. at 60"C., with standard deviations of 15, 43 and 0.6 units, respectively. Weir (222) further attempted to divulge the type of binding between various tanning agents and collagen through determination of heat, entropy, and free energy changes in the shrinkage of tendon collagen, pretreated with various substances, including tanning agents. By applying the theory of absolute reaction rates to the shrinkage of tanned and otherwise pretreated collagen, he concluded that tannages with the salts of aluminum, iron and zirconium as well as with F., seem to reduce entropy more than heat, thereby increasing the free energy. The tannage with chromium salts is unique inasmuch as it increases not only the free energy but also the entropy of activation. The heat of activation, identified with the degree of disorganization occurring in the activating process, is drastically increased, and this trend is in agreement with the concept of chrome tanning as an internal stabilization of the collagen lattice by crosslinking the protein chains through chromium complexes. One per cent CraO, fixed by collagen is sufficient to produce the maximum degree of stabilization. Weir concluded that only a fraction of the acidic and basic protein groups possess the required spatial orientation to react with the chromium complex to form crosslinks. The amount of chromium combined with collagen in excess of 1 % probably combines with collagen in the same manner as the other tanning agents mentioned, decreasing the entropy of activation. Hence, this additional chrome will add nothing to the orientation of the protein chains and their stabilization. Incorporation of large amounts of chromium will mean some loss of orientation of the protein lattice. Weir's concept of the mechanism of denaturation of collagen, the type of stabilizing bonds, and the nature of the chrome tanning process are on the whole in harmony with the views on these problems outlined in the present chapter.
REFERENCES 1. Anslow, W. K., and King, H. (1927). Biochem. J . 21, 1168. 2. Astbury, W. T. (1933). Trans. Faraday SOC.29, 193. 3. Astbury, W. T. (1940). J . Intern. SOC.Leather Trades' Chemists 24, 69. 4. Baldwin, M. E. (1919). J . A m . Leather Chemists' Assoc. 14, 10. 5. Balfe, M. P. (1948). J . Intern. SOC.Leather Trades' Chemists 82, 39. 5s. Balfe. M. P. (1948). Progress in Leather Science. 111. 610. London. 9. (1944). . J . Amy Chem. SOC.66, 1927; (1942) 64,'729. 6. Bear, 7. Beek, J., Jr. (1930). I d . Eng. Chem. 22, 1373. 8. Beek, J., Jr. (1938). J . Research Natl. Bur. Standards 21, 117; J . A m . Leather Chemists' Assoc. 83, 621. 9. Beek, J., Jr., and Sookne, A. M. (1939). J . Research Natl. Bur. Standards 23, 271. 10. Bergmann, M. (1935). J . Biol. Chem. 110, 471. 11. Bergmann, M., and Niemann, C. (1936). J . Biol. Chem. 116, 77; (1937). 118, 301.
k.
416
K. H. OUSTAVSON
12. Bergmann, M., and Thiele, H. (1933). Collegium, 582. 13. Birch, T. W., and Harris, L. J. (1930). Biochem. J. 24, 1080. 13a. Bowes, J. H. (1948). Progress in Leather Science, 111, 535. London. 14. Bowes, J. H. (1946). Progress in Leather Science, I, 158. London. 14a. Bowes, J. H., and Kenten, R. H. (1948). Biochem. J . 43,363. 15. Bowes, J. H., and Pleass, W. B. (1939). J . Intern. SOC.Leather Trades Chemists 23, 365, 451, 499. 16. Braybrooks, W. E., McCandlish, D., and Atkin, W. R. (1939). J . Intern. SOC. Leather Trades' Chemists 23, 111, 135. 17. Briefer, M. (1928). Ind. Eng. Chem. 21, 266. 18. Bungenberg de Jong, H. G. (1923). Rec. trav. chim. 42, 437. 19. Burk, N. F., and Greenberg, D. M. (1930). J . Biol. Chem. 87, 197. 20. Chambard, P., and Mezey, E. (1925). J . Intern. Soc. Leather Trades' Chemists 9,57. 21. Cherbuliez, E., Jeannerat, J., and Meyer, K. H. (1938). 2. physiol. Chem. 266,241. 22. Chernov, N. W. (1936). J . Intern. soc. Leather Trades' Chemists 20, 121. 23. Chibnall, A. C. (1942). Proc. Roy. SOC.London B 131, 136. 24. Chibnall, A. c. (1946). J . Intern. SOC.Leather Trades' Chemists 30, 1. 25. Cohn, E. J., and Edsall, J. T. (1943). Proteins, Amino Acids and Peptidcs. Reinhold, New York. 26. Croad, R. B. (1923). J. Soc. Chem. Znd. London 42, 203. 27. Cuthbertson, W. R., and Phillips, H. (1945). Biochem. J . 39, 7. 28. Dakin, H. D. (1920). J. Biol. Chem. 44, 499. 29. Danielsson, C. E. Unpublished work, Institute of Physical Chemistry, Upsala. 30. Douglas, G. W., and Humphreys, F. E. (1937). J . Intern. Soc. Leather Trades' Chemists 21, 378. 31. Edsall, J. T. (1947). Advances in Protein Chem. 3, 472. 32. Elod, B., and Schachowskoy, T. (1939). Kolloid Beihejte 61, 1. 33. Elod, E., and Siegmund, W. (1934). Collegium, 281. 34. Ewald, A. (1919). 2. physiol. Chem. 106, 115, 135. 35. Ewald, A,, and Ktihne, E. (1876). Die Verdauung ale histologische Mcthodr, Heidelberg; quoted by Grassrnann (50). 36. Faurb-Frcmict, E. (1937). J. chim. phys. 34, 801. 37. FaurC-Fremiet, E., and Baudouy, C. (1937). Bull, SOC. chim. biol. 19, 1134. 38. FaurC-Fremiet, E., and Cougny, A . (1937). Bull. museum hist. nat. Paris 9, 188. 39. Fraenkel, G., and Rudall, K. M. (1947). Proc. Roy. SOC.London B134, 111. 40. Fraenkel-Conrat, H., Cooper, M., and Olcott, H. 8. (1945). J. Am. Chem. SOC.67, 950. 41. Fraenkel-Conrat, H., and Olcott, H. 8. (1946). J . A m . Chem. SOC.68, 34. 42. Fraenkel-Conrat, H., and Olcott, H. S. (1948). J . Am. Chem. SOC.70, 2673. 43. French, D., and Edsall, J. T. (1945). Advances in Protein Chem. 2 , 277. 44. Freudenberg, K. (1921). CoUegium, 353. 45. Gerngross, 0. (1920). Collegium, 2; (1921), 489. 46. Gerngross, 0. (1939). In Grassmann, W., Handbuch der Gerbereichcmie, II/2. Springer, Vienna. 47. Gerngross, O., and Bach, 8. (1923). Collegium, 377. 48. Gerngross, O., and Lowe, H. (1926). Collegium, 229. 49. Gerngross, O., and Ritter, A. (1920). Biochem. 2. 108,82.
PROTEIN-CHEMICAL ASPECTS OF TANNING
417
Grassmann, W. (1936). Kolioid-Z. 77, 205. Grassmann, W., Chuan Chii, P., and Schele, H. (1937). Collegium, 530. Grasemann, W., and Riederle, K. (1936). Biochem. Z. 284, 177. Greenstein, J. P. (1931). J. Biol. Chem. 93,479; (1932). 96, 465. Gustavson, K. H. (1923). J. Am. Leather Chemists’ Assoc. 18, 568; (1926). 21, 22. 55. Gustavson, K. H. (1926). Colloid Symposium Monograph 4, 79; J . Am. Leather Chemists’ Assoc. 21, 206. 56. Gustavson, K. H. (1926). Collegzum, 97. 57. Gustavson, K. H. (1926). J . Am. Chem. SOC.48, 2963. 58. Gustavson, K. H. (1927). J . Am. Leather Chemists’ Assoc. 22, 125; (1939). “ Allgemeine Theorie des Gerbvorganges” in Grassmann, W., Handbuch der Gerbereichemie, II/2. Springer, Vienna. 59. Gustavson, K. H. (1927). Ind. Eng. Chem. 19. 1015. 60. Gustsvson, K. H. (1927). J . Am. Leather Chemists’ Assoc. 22, 60. 61. Gustavson, K. H. (1927). J. Am. Leather Chemists Assoc. 22, 125; (1939). “Die Kombinationsgerbung ” in Grassmann, W., Handbuch der Gerbereichemie, II/2. Springer, Vienna. 62. Gustavson, K. H. (1931). J . Am. Leather Chemists’ Assoc. 26, 635. 63. Gustavson, K. H. (1932). Collegium, 775. 64. Gustavson, K. H. (1937). J . Intern. SOC.Leather Trades’ Chemists 21, 4. 65. Gustavson, K. H. (1937). In Stiasny Festschrift, p. 99. Roether, Darmstadt. 66. Gustavson, K. H. (1939). Theorie und Praxis der Chromgerbung in Grassmann, W., Handbuch der Gerbereichemie, II/2. Springer, Vienna. 67. Gustavson, K. H. (1940). J. Intern. Soc. Leather Trades’ Chemists 24, 377. 68. Gustavson, K. H. (1940). Suensk Kem. Tid. 62, 75. 69. Gustavson, K. H. (1940). Suensk Rem. Tid. 62, 261. 70. Gustavson, K. H. (1941). S u a s k Kem. Tid. 63, 324. 71. Gustavson, K. H. (1942). Biochem. Z. 311, 347. 72. Gustavson, K. H. (1942). Suensk Kem. Tid. 64, 74. 73. Gustavson, K. H. (1943). Kolloid 2.103, 43. 74. Gustavson, K. H. (1943). Suensk Kem. Tid. 66, 191. 75. Gustavson, K. H. (1943). Suensk Kem. Tid. 66, 249. 76. Qustavson, K. H. (1944). Ing. Vetenskaps Akad. Handlr. No. 177. 77. Gustavson, K. H. (1944). Suensk Kem. Tid. 66, 14. 78. Gustavson, K. H. (1945). Suensk Kem. Tid. 67, 35. 79. Gustavson, K. H. (1946). J. Am. Leather Chemists Assoc. 41, 47. 80. Gustavson, K. H. (1946). J . Colloid Sci. 1,397; J . Intern. SOC.Leather Trade2 Chemists 30, 264. 81. Gustavson, K. H. (1946). Suensk Kem. Tid. 68, 2, 274. 82. Gustavson, K. H. (1947). Acta Chem. Scand. 1, 581. 83. Gustavson, K. H. (1947). J. Am. Leather Chemists Assoc. 42, 13. 84. Gustavson, K. H. (1947). J. Am. Leather Chemists Assoc. 44, 201. 85. Gustavson, K. H. (1947). J. Am. Leather Chemists Assoc. 42, 313. 86. Gustavson, K. H. (1947). J. Biol. Chem. 169, 531. 87. Gustavson, K. H. (1947). J . Intern. SOC.Leather Trades’ Chemists 31, 181. 88. Gustavson, K. H. (1947). J . Phys. & Colloid Chem. 61, 1181. 89. Gustavson, K. H. (1947). Suensk Kem. Tid. 69, 98. 90. Gustavson, K. H. (1947). Saensk Kem. Tzd. 69, 159. 91. Gustavson, K. H. (1949). J . SOC.Leather Trades’ Chemists 33, 332. 50. 51. 52. 53. 54.
418 92. 93. 94. 95. 96.
K. H. OUSTAVSON
Gustaveon, K. H. (1948). J. Am. Leafher Chemists’ Aesoc. 48, 51. Gustavson, K.H. (1948). J . Am. Leather Chemists’ Assoc. u),741,744. Gustaveon, K. H. Unpubliihed work. Gustavson, K. H., and Larsson, A. (1947). Suensk Papperttidn. 60, 101. Gustavson, K. H., and Tomlinson, J. (1948). J . Soc. Leather Trades Chemist8
aa, 1136. 97. Gustavson, K. H., and Widen, P. J. (1926). Collegium, 153. 98. Hall, C. E., Jakus, M. A., and Schmitt, F. 0. ’(1942). J . A m . Chem. SOC.64, 1234. 89. Harris, L. J. (1923). Proc. Roy. SOC.London BB6, 442, 500. 100. Highberger, J. H. (1939). J . Am. Chem. SOC.61, 2302. 101. Highberger, J. H. (1947). J . Am. Leather Chemzsts’ Assoc. 41, 493. 102. Highbcrgcr, J. H., and Retrsch, C. E. (1939). J . Am. Leather Chemzstq’ Aesoc. 34,131. 103. Highberger, J. H., and Salcedo, I. 8. (1940). J . Am. Leather Chemists’ Asaoc. as, 11; (1941). 88,271. 104. Highberger, J. H., and Stecker, H. C. (1941). J . Am. Leather Chemzsts’ Assoc. 86,368. 105. Hilpert, S., and Brauns, F. (1925). Collegium, 64. 106. Hofmeister, F. (1888). Arch. ezptl. Path. Phurmkol. 14, 247. 107. Holland, H. C. (1943). J. Intern. Soc. Leather Trades’ Chemists 17, 207. 108. Huggina, M. L. (1942). Ann. Rev. Biochem. 11, 27. 109. Huggins, M. L. (1943). Chem. Reus. 81, 195. 110. Humphreys, F. E. (1934). J . Intern. SOC.Leather Trades’ Chemisfs 18, 178. 111. Jander, G., and Scheele, W. (1932). Z . anorg. allgent. Chem. 106, 241. 112. Jordan Lloyd, D. (1933). J . Intern. Soc. Leather Trades’ Chemists 17,208. Jordan Lloyd, D., and Marriott, R. H. (1936). Trans. Faraday SOC.31, 1. 113. Jovanovits, J. A. (1932). Collegium, 215. 114. Jovanovits, J. A., and Alge, A. (1932). Collegium, 222. 115. Katr, J. R., and Derksen, J. C. (1932). Rec. trau. chim. 61, 513. 116. Katr, J. R., and Weidinger, A. (1933). Collegium, 85, 290. 117. Kraemer, E. O., and Dexter, S. T. (1927). J . Phys. Chem. 81, 764. 117a. Kritringer, C. C. (1948). J . Am. Leather Chemists’ Assoc. 48, 675. 118. Kuhn, W. (1937). In Zur Entwicklung der Chemie der Hochpolymeren. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131.
Verlag Chemie, Berlin. Ktinteel, A. (1925). Collegium, 623. Ktinteel, A. (1929). Biochem. Z. IOB, 326. Kiinteel, A. (1929). Collegium, 207. Ktintzel, A. (1936). Collegium, 625. Ktintzel, A. (1937). In Stimny Festschrift, p. 191. Roether, Darmstadt. Ktintrel, A. (1940). Kolloid-Z. 01, 152. Kiintzel, A. (1941). Kolloid-Z. 06, 273. Ktintrel, A. (1944). In Graeemann, W. Handbuch der Gerbersichemie, I/1. Springer, Vienna. Ktintrel, A., and Buchheimer, K. (1930). Collegium, 205. Ktintrel, A., Kinrer, R., and Stiasny, E. (1934). Collegium, 213. Kthtzel, A., and Phillips, J. (1932). Collegium, 267. K h t s e l , A., and Riess, C. (1936). Collegium, 138. Kthtrel, A., and Rim, C. (1936). Collegium, 635.
PROTEIN-CHEMICAL ASPECTS OF TANNINQ
419
Kllntzel, A., and Schwank, M. (1940). Collegium, 441, 455, 489, 500. Latimer, W. M., and Porter, C. W. (1930). J . Am. Chem. Soc. 62,206. Leplat, G. (1933). Compt. rend. aoc. biol. 112, 1256. Leuthardt, F. (1932). Helv. Chim. A d a 16, 540. Ley, H. (1940). 2.Elcktrochem. 10, 954. Ley, H., and Ficken, K. (1912). Ber. 46, 377. LinderstrBm-Lang, K., and Duspiva, F. (1935). 2. phyaid. Chem. 257, 131 Loeb, J. (1922). Proteins and Theory of Colloidal Behavior. McGraw-Hill, New York. 140. Marriott, R. H. (1928). J. Intern. Soe. Leather Trades’ Chemists 12, 216, 281, 132. 133. 134. 135. 136. 137. 138. 139.
342. 141. 142. 143. 144.
Marriott, R. H. (1932). Biochem. J. 26,46. Marriott, R. H. (1935). J . Intern. Soc. Leaher Traded Chemiata 18, 246. Marriott, R. H. (1937). I n Stiasny Festschrift, p. 245. Roether, Darmstadt. Martin, A. J. P. (1946). Symposium on Fibrous Proteins. J . Soc. D y ~ 8
c01ourists. 145. Martin, A. J. P., and Synge, R. L. M. (1941). Biochem. J . S6,91, 13%. 146. McLaughlin, G. D., and Cameron, D. H. (1937). J . Phya. Chem. 41, 961. 147. McLaughlin, G. D., Highberger, J. H., and Moore, E. K. (1927). J . Am. Leather Chemiata’ Asaoc. 22, 345. 148. McLaughlin, G. D., and Theis, E. R. (1946). The Chemistry of Leather
Manufacture.
Reinhold, New York.
149. Merrill, H. B. (1927). J . Am. Leather Chemiata’ Aascrc. 22, 230; (1925). I d . Eng. Chem. 17,36. 150. Merrill, H. B., Cameron, D. H., Elliin, H. L., and Hall, C. P. (1947). J . Am. Leather Chemists’ Assoc. 42,536. 151. Merrill, H. B., and Schroeder, H. (1929). I d . Eng. C h m . 21, 1225. 152. Meunier, L.,and Seyewets, A. (1908). Compt. rend. 146, 987. 153. Meyer, K. H. (1929). Bimhem. 2. 208, 23. 154. Meyer, K. H. (1940). Die Hochpolymeren Verbindungen. Akademkhe 155. 156. 157. 158. 159. 180. 161. 162.
Verlagsgesellschaft, Leipsig. Meyer, K. H., and Ferri, C. (1935). Helv. Chim. A d a 18, 570. Meyer, K. H., and Mark, H. (1928). Ber. 61, 19. Michaelis, L. (1935). J . Am. Leather Chemiala’ Aaaoc. SO, 557. Mirsky, A. E., and Pauling, L. (1936). Proc. NaU. A d . Sci. U.S. 41,439. Nageotte, J. (1928). Compl. rend. aoc. bid. W, 15. Neale, S. M., and Petera, R. H. (1946). Trans. Faraduy SOC.11,478. Nitschmann, H., and Hadorn, H. (1944). Helv. Chim. A& 27, 277, 299. Nutting, G. C., and Boraaky, R. (1948). J . Am. Leather Chemists’ Aasoc.
4q96. 163. Ogston, A. G. (1943). Tram. Faraduy Soc. M,151; (1945). 41, 670. 164. Otto, G. (1928). Dissertation, Technische Hochschule, Karhruhe, pp. 23-25. 165. Otto, G. (1933). Collegium, 586; (1935). 371; (1938). 170. 1%. Page, R. 0. (1928). J . Am. Leather Chemists’ Assoc. 23,495; (1933). 28, 93. 167. Page, R. 0. (1933). J . Am. Leulbv Chemists’ Aasoc. 28, 93. Page, R. 0. (1944). J . Intern. Soc. Leather Traded Chemists 10,156. 168. Page, R. 0. (1944). J . Intern. soc. Leather Trades’ Cherniats 88, 168. 169. Page, R. O., and Holland, H. C . (1932). J . Am. Leather Chemista’ Asaoc. 27, 432.
420
K . H. QUSTAVSON
170. Pankhurst, K. G. A. (1947). Nature 168, 538. 171. Pauling, L. (1940). The Nature of the Chemical Bond. 2nd ed. Cornell
Univ. Press, Ithaca, N. Y. (1927). Organische Molekiilverbindungen, 2nd Ed. Enke, Stuttgart. 173. Procter, H. R., and Wilson, J. A. (1916). J . Chem. SOC.109, 307. 174. Pryor. M . G. M. (1940). Proc. Roy. SOC.London Bl28, 378, 393. 175. Itiess, C., and Barth, K. (1935). Collegium, 62. 176. Roddy, W. T. (1943). J . Am. Leather Chemists’ Assoc. 38, 184. 177. Rudall, K. M. (1946). Symposium on Fibrous Proteins, p. 21. J. SOC.Dyers and Colourisls. 178. Salcedo, I . S., mid Highberger, J . H. (1941). J . Am. Leather Chemists’ Assoc. 172. l’feiffer, P.
36, 271. 179. Schrnitt, F. 0 .
(1944). Advances in Protein Chem. 1, 26; J . Am. Leather Chemists’ Assoc. 39,340; (1945). Harvey Lectures 40, 249. 180. Schrnitt, F. O., Hall, C. E. and Jakus, M. A. (1942). J . Cellular Comp. Physiol. 20, 11. 181. Scligsberger, L. (1941). J . Am. Leather Chemists’ Assoc. 36, 669. 182. Shuttleworth, S. G. (1948). J . SOC.Leather Trades’ Chemists 32, 116. 183. Smythe, C. V., and Srhmidt, C. L. A. (1930). J. Biol. Chem. 88, 241. 184. Sokolov, S. J., and Kolyakova, G. E. (1935). Kolloid Z. 72, 74. 185. Sourlangas, S. D., and Atkin, W. R. (1943). J. Intern. SOC.Leather Trades’ Chemists 27, 183. 186. Speakman, J. B., and Whewell, C. S. (1936). J. SOC.Dyers Colourisls 62, 380. 187. Staudinger, H. (1937). Zur Entwicklung der Chemie der Hochpolymeren, pp. 151-153. Vrrlag Chemie, Berlin. 188. Stecker, H. C., and Highberger, J. H. (1942). J. Am. Leather Chemists’ Assoc.
37, 226. 180. Steinhardt, J.
(1944). Ann. N . Y. Acad. Sci. 61,287. 100. Steinhardt, J., Fugitt, C. H., and Harris, M. (1941). J . Research Natl. Bur. Standards 26, 293; (1942). 28, 201. 191. Steinhardt, J., Fugitt, C. H., and Harris, M. (1943). Zbid. 30, 123. 192. Stiasny, E. (1908). Collegium, 132. 193. Stiasny, E. (1931). Gerbereichemie. Steinkopff, Dresden. 194. Stiasuy, E., and Gustavson, K. H. (1946). J . Am. Leather Chemists’ Assoc. 41,516. 195. Stiasny, E., and Scotti, H. C. (1930). Ber. 63, 2977. 196. Theis, E. R. (1941). J . Am. Leather Chemists’ Assoc. 36, 449. 197. Theis, E. It., and Blum, W. A. (1942). J . Am. Leather Chemists’ Assoc. 37, 553. 198. Theis, E. It., and Lams, M. (1946). Unpublished work noted in (148) p. 409. 199. Thornas, A. W., and Baldwin, M. E. (1918). J . Am. Leather Chemists’ Assoc. 13, 192. 200. Thomas, A. W., and Baldwin, M. E. (1919). J . Am. Chem. SOC.41, 1981. 201. Thomas, A. W., and Foster, S. B. (1922). Znd. Eng. Chem. 14, 132. 202. Thomas, A. W., and Foster, 9. B. (1922). Znd. Eng. Chem. 14, 191; (1923). 16,707. 203. Thomas, A. W., and Foster, S. B. (1925). Znd. Eng. Chem. 17, 1162, 204. Thomas, A. W., and Foster, S. I3. (1926). J . Am. Chem. SOC.48,489.
PROTEIN-CHEMICAL ASPECTS O F TANNING
421
205. Thomas, A. W., and Kelly, M.W. (1923). Ind. Eng. Chem. 16, 1148; (1924). 16, 800; (1925). 17, 41; (1926). 18, 625. 206. Thomas, A. W., and Kelly, M.W. (1924). J . A m . Chem. SOC.44, 195. 207. Thomas, A. W.,and Kelly, M. W. (1924). Ind. Eng. Chem. 16, 925; (1926). 18, 383. 208. Thomas, A. W., and Seymour-Jones, F. L. (1924),Ind. Eng. Chem. 16, 157. 209. Turley, H. G.,and Windus, W. (1937). In Stiasny Festschrift, p. 396. Roether, Darmstadt. (1938). J . A m . Leather Chemists’ Assoc. 33, 246; (1941). 36, 603. 210. Walker, J. F. (1944). Formaldehyde. Reinhold, New York. 211. Wilson, J. A. (1928). The Chemistry of Leather Manufacture, Vol. I, 2nd Ed. Chemical Catalog Co., New York. 212. Wilson, J. A. (1929). The Chemistry of Leather Manufacture, Vol. 11. 213. Wilson, J. A.,and Gallun, E. A. (1920). J . A m . Leather Chemists’ Assoc. 16, 273. 214. Wilson, J. A,, and Kern, E. J. (1917). J . A m . Leather Chemists’ Assoc. 12,445. 215. Windus, W.,and Turley, H. G. (1938). J . Am. Leather Chemists’ Assoc. 33, 246; (1941). 36, 603. 216. Wolesensky, E. (1925). Natl. Bur. Standards U . S. Technol. Papers, No.302; (1926). No.20, No. 309. 217. Worniell, It. L., and Kaye, M. A.G. (1944). Nature 163,520;(1945). J . SOC. Chem. Ind. London 64, 75. 218. Wiihliuch, E. (1932). Bioche,n. Z.247, 329. 219. Vogl, I,. (1!)26). I>iss:rtation, Technische Ilochschule, Darmstadt. 220. Bowes, J. €I., and Kenten, R. I f . (1949). Riochem. J . 44, 142. 221. Weir, C. E. (1949). J . Am. Leather Chemists’ Assoc. 44, 108. 222, Weir, C.E. (1949). J . A m . Leather Chemists’ Assoc. 44. 142.
This Page Intentionally Left Blank
Proteins, Lipids, and Nucleic Acids in,Cell Structures and Functions BY ALBERT CLAUDE The Rockefeller Institute for Medical Research, New York, New York
CONTENTS Page
I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. The Cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111. The Nucleus.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. Chromosomes and Chromatin. . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Nucleolus. ...................................... IV. The Cytoplasm. ...................................... 1. Mitochondria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Microsomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Fibrous Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V. Constitution and Duplication of Living Matter.. . . . . . . . . . . . . . . . . . . . . . . VI. Phospholipids in Cell Structures and Functions.. . . . . . . . . . . . . . . . . . . . . . . VII. Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VIII. Nucleic Acids .............................................. References . . . . . . . . . . . . .. .. . . . . . . . .
428 428 430 432 432 433 434 436 437 . 439
I. INTRODUCTION Proteins are products and constituents of cells. I n the past, chemists have been increasingly successful in isolating proteins from their cellular environment, and in determining some of the specific properties that these compounds exhibit in vitro. There is no doubt that these properties are utilized by the cell, but it is not known to what extent such properties are modified by virtue of chemical association with other cellular constituents, and by the influence of other associated or competing biochemical systems. It is no more plausible to assume that the properties exerted by a protein in the cell are just those exhibited by the isolated protein in vitro than to consider th at the properties of amino acids separated from a protein hydrolyzate can entirely explain the properties of the original protein. In the cell, proteins are parts of structures of considerable complexity, of ten in association with phospholipids or nucleic acids. I n recent years, methods have become available which permit the separation of certain morphological constituents of the cell, in quantities sufficient for biochemical analysis. The present paper deals 423
424
ALBERT CLAUDE
with a n approach to the direct study of the nature of a number of cellular complexes, and the integration of biochemical functions. The reactivity of chemical compounds is determined not only by their elementary composition but, t o a much greater extent, by the spatial disposition of the individual atoms within the molecule, and by the configuration of the molecule as a whole. This property, which allows for considerable variations, is fully utilized by biological systems and is the basis of the specificity of biochemical reactions. The sensitivity of biological systems to molecular configuration was vividly demonstrated by Pasteur when he showed that Yenicillium glaucum could utilize the d but not the I form of tartaric acid (42). In recent years observations have been rapidly accumulating which show the importance of structural variations in the control and conduct of essential biological processes; minor changes, as slight as the rotation of a single carbon atom in the prosthetic group of an antigen, affect the reactivity of the entire protein complex, and the nature of the immunological response (2,31); analogs and isomers are found to play decisive roles in problems of nutrition and growth, and in the limitation and invasiveness of infectious diseases (27,62). Isomeric variations may prove to be sufficient to determine the genetic specificity of certain nuclear constituents, for example, in the case of the nucleic acids isolated by Avery and coworkers (3,34). At a higher level of organization, genetic studies have repeatedly shown the dependence of gene effect on the position of the gene along thc chromosome (26,55). It is clear that the larger the molecule or the molecular aggregate, the greater will be the opportunity for structural variations and for reciprocal influences between the various parts of the complex, and for ultimate effect on the properties of the complex as a whole. Active cellular components such as enzyme proteins, coenzymes, hormones, and vitamins are being isolated in increasing number, and the kinetics of their specific properties determined in vitro. In vitro, however, isolated chemical reactions follow a uniform course and eventually reach an equilibrium which, for the cell, means death. In living protoplasm active chemical constituents are linked to form three-dimensional systems of varying, and probably considerable, complexity ; substances known to be readily diffusible in water are held in place and consequently are not permitted to react freely. I n this manner, chemical processes can be channelled through preferential paths, slowed down, stopped, or speeded up, according to needs or to conditions established possibly at a distance rcmote from the locus where the particular reaction is taking place. It is in this structural arrangement, which permits a particular system t o determine and control the timing, order, and sequence of a
PROTEINB, LIPIDS, NUCLEIC ACIDS IN CELLS
425
great variety of individual chemical reactions that probably resides, for the biologist, the ultimate challenge of living matter. Morphological and cytochemical studies, especially in recent years, have indicated that important cellular functions are localized in definite areas of the cell. This is the case for the aerobic respiration (14,28,47) and the fatty-acid oxidation systems (30) which involve, among others, the cytochrome-linked enzyme structures. By means of the method of differential centrifugation (15,29), now finding wide application in cytological and histological studies, it is likely that our knowledge of the topography of intracellular functions will be considerably extended and that i t will be possible to map the distribution in the cell of the major biochemical processes. It will be the further task of modern cytology to investigate the constitution and architecture of the various integrated cell structures, a t the molecular level, and to discover in what way structure conditions biological processes and insures the continuance of biochemical cycles.
11. THE CELL Differentiated areas first recognized in cells were the nucleus and the region surrounding it, the cytoplasm. Late'r, delicate structures were found in both: in the nucleus, the nucleolus and the chromosomes; in the cytoplasm, mitochondria, centrospheres and centrioles, specialized vacuoles, and fat bodies. More recently submicroscopic constituents were discovered or postulated, namely, the microsomes (10,13) and the fibrous protein system which, through viscosity changes, is presumably responsible for certain movements of cytoplasm (51). The elements just mentioned can be detected either under the light microscope, directly, or indirectly with the help of special techniques (13,19); the electron microscope so far has added little, except for morphological details, to this usual complement of protoplasm. The fact t ha t many of the cell constituents could be stained differentially for examination under the microscope led to numerous attempts to adapt the sensitivity of chemical and biochemical tests to the dimensions of the cell, and t o conduct the analysis of the various cell structures a t the microscopic level. The methods that were tried suffered considerable limitations and uncertainties: the scope and application of the tests were generally restricted to those which would yield colored products of a density sufficient for detection under the microscope; furthermore, the reactions had t o be carried out in a cellular environment of unknown complexity, already profoundly modified by the preceding fixation. The validity of most histochemical tests so far proposed has repeatedly been
426
ALBERT CLAUDE
challenged and to date only a few, such as the Feulgen reaction for desoxyribonucleic acid, continue to be considered specific. Thanks to the method of differential centrifugation, perfected And systematically applied to cell studies during the past few years, it is now possible to separate a number of the morphological cell structures, namely, those which originally are of different sizes or densities, and to obtain distinct cell fractions in practically unlimited amounts (15). This fact a t once liberates the cytochemist from the limitations of the microscopic techniques and offers the possibility to subject the formed cellular constituents to wide and exhaustive biochemical analysis. By adaptations of this method, 2 to 3 g. of purified chromatin strands can be readily prepared in the course of 2 to 3 hours (21), 5 to 6 g. of purified mitochondria or microsomes in 4 to 6 hours (15). 111. THENUCLEUS 1. Chromosomes and Chromatin
Chromatin threads from the resting nucleus were first isolated by Claude and Potter (12,21)) and their method was later adopted, with slight modifications, by Mirsky and associates for the study of nucleoprotein complexes (37-39). ' The chromatin strands separated by differential centrifugation were found to contain, in per cent of dry weight: nitrogen, 15.57; phosphorus, 3.72; carbon, 45.60; and sulfur, 1.67. The amount of phosphorus, and chemical tests, indicated that as much as 40% of the chromatin complex was represented by nucleic acids of the desoxyribose type. Histological studies and chemical analysis have shown that cells are rich in lipids, principally phospholipids, which are found mostly as constituents of cytoplasmic structures (mitochondria and microsomes) and of the cellular and nuclear membranes. Lipids appear to be rare or absent in chromosomes and the only nuclear structure in which it has been detected, besides the nuclear membrane, is the nucleolus. The small amount of lipids recovered from the chromatin threads, 2.3% or less, may have been derived from cytoplasmic contaminants or from a certain proportion of the nucleoli which have been found to remain attached to some of the chromatin strands.* From cytological tests, and the direct chemical analysis of chromatin threads, it would appear that chromosomes are predominantly composed of nucleoproteins
* Isolated chromatin strands in dilute methyl green or methyl green-pyronine aolutions appear stained a bright green; frequently a relatively large, bean-shaped body staining pink or red, presumably the nucleolus,~isfound attached at one point on B ohrometin strand.
PROTEINS, LIPIDS, NUCLEIC ACIDS IN CELLS
427
extremely rich in desoxyribonucleic acids. There is evidence, especially based on the study of salivary gland chromosomes, that chromosomes are not structurally homogeneous, and it must be assumed that other substances are also present, possibly proteins of a different nature, such as the chromosomin of Stedman and Stedman (54),a finding confirmed later by Mirsky. Because of their abundance in the nucleus, their particular arrangement in chromosome structures, and their participation in the mitotic cycle, desoxyribose nucleoproteins have been postulated to play a leading role in the economy of the cell. This view has been consistently sustained by the continuous accumulation of indirect evidence which indicates that chromosomes are the repository of genetic characters, i.e., exert a controlling influence on the morphology and the functions of the cell. The way by which this influence is mediated to other parts of the cell is not known. It has been suggested that chromosomes are represented by systematic arrays of enzymes or proenzymes, representatives of which migrate to the cytoplasm where they take part in, or direct, metabolic processes (52,53,63,64). A definite example of this kind of action may be found in the type transformation of pneumococci in which a desoxyribonucleic acid has been shown to be the determinant and specific factor (3,34). The most concrete evidence of a functional relation between nucleus and cytoplasm appears in the striking observation of McClintock where a terminal segment of a chromosome in the somatic cells of maize (twothirds of the short arm of chromosome 9) becomes detached during mitosis and lags behind in the cytoplasm, becomes pycnotic, and is finally eliminated. This event is accompanied, in the direct descendants of this cell, by the loss of the tendency, or the ability, to synthesize chlorophyll (35). This case seems to provide visual demonstration of the dependence of plastids, elements which are essentially cytoplasmic, on chromatin. If chromatin normally penetrates the cytoplasm, it must be in a subtle manner, and in minute amounts, since histochemical tests have consistently failed to demonstrate desoxyribonucleic acid in the cytoplasm. These observations were confirmed in experiments in which the main cell constituents were separated by differential centrifugation, and analysis carried out in vitro. The results indicated that practically the entire desoxyribonucleic acid of the cell could be recovered with the nuclear fraction, while tests with the cytoplasmic fractions remained negative (15,46,47). Chromosomes, and their chief chemical constituents, nucleoproteins and nucleic acids, have received continued attention since they were discovered, 60 to 70 years ago. It is noteworthy that, in spite of this sustained interest, the biochemical function of nucleoproteins and
428
ALBERT CLAUDE
nucleic acids is still unknown, and that no tests exist whereby their action could be measured in vitro. The fact that “native” chromatin can be obtained readily in large quantities should encourage research in this field. 2. Nucleolus
The nucleolus, another distinctive constituent of the nucleus, contrasts in morphology and properties with the chromosomes and in many respects resembles cytoplasmic components, namely, mitochondria. This is demonstrated by similar staining properties and by the fact that both nucleoli and mitochondria contain appreciable amounts of lipids and nucleic acid of the ribose type. The size of the nucleolus may vary greatly from cell to cell, or in the course of the mitotic cycle, and morphologically, is not an homogeneous structure; light-microscope pictures frequently show the nucleolus to be composed of granules or globules of uniform size; this appearance is confirmed in electron micrographs of sections of guinea pig liver (unpublished observations) and of cells from tissue culture (16). Histochemical tests indicate that the nucleolus is probably rich in alkaline phosphatase (58). As in the case of other cell constituents, it is probable that the function of the nucleolus will not be fully understood until a way is found whereby nucleoli can be isolated and prepared in bulk, thus making possible a systematic study of their constitution and biochemical properties.
IV. THE CYTOPLASM Since the discovery of chromosomes and the initial work of Strasburger and Flemming (60), the attention and ingenuity of cytologists has been directed in large part toward the study of the phenomenon of mitosis, and only sporadically has it been devoted to the problems concerning the morphology and functions of cytoplasm. No doubt this interest first arose from the attractive orderliness of the mitotic process, and later received further encouragement and impetus from the science of genetics; these developments must have detracted research workers from the less rewarding investigations on cytoplasm. The countless fixatives empirically compounded year after year aimed at the fixation or, at least, the precipitation of chromatin; most of them featured acids or alcohols, two types of reagents which have destructive and dissolving effects on cytoplasm and their almost universal use led to misconceptions regarding the composition and structure of the cell. Experience has now shown that no single fixative can a t the same time preserve both nuclear and cytoplasmic structures, and fixatives found to be satisfactory either far the nucleus or for the cytoplasm are generally exclusive.
PROTEINS, LIPIDS, NUCLEIC ACIDS I N CELLS
429
Since the studies by Altmann (l), it was realized that acids and alcohols should be avoided, and that reagents with strong oxidative or reducing properties were good fixatives of cytoplasm. The best fixatives known for the preservation of cytoplasmic inclusions are neutral or only slightly acid and contain, alone or in combination, dichromates, osmium tetroxide, and formalin. This particular behavior toward fixatives can be understood from the knowledge of the chemical constitution of those formed elements of the cytoplasm which, by their abundance, contribute most to the microscopic picture, namely, mitochondria and microsomes. Each of these components represent 15 to 20% of the cell mass; they have been separated by differential centrifugation under various conditions and, in the isolated state, subjected t o chemical and biochemical analysis. Both mitochondria and microsomes have been found to be complex structures composed in large part of phospholipids, proteins, and ribose nucleotides especially in the form of nucleic acid (10,11,13,15). Mitochondria and microsomes react similarly toward acidification of the medium : increasing acidification produces slight agglutination to massive clumping; a t pH 3.5 both elements disintegrate with concomitant denaturation of proteins, separation of phospholipids, and solubilization of ribose nucleotides and nucleic acid (15). Similar events take place on the alkaline side in the neighborhood of p H 12. On the other hand, strong alcohol, especially if acid is present, will denature the proteins, dissolve the phospholipids, and reprecipitate nucleic acid. It is clear that the same destructive effects must obtain in the cell, and that the choice of chemically suited fixatives is of paramount importance in studies dealing with cell structures and the distribution of substances within the cytoplasm and the nucleus. In recent years extensive studies concerned with the microscopic detection of nucleotides by means of characteristic light absorption in the ultraviolet have given rise to far-reaching conclusions regarding the distribution and possible interchange of nucleic acids among the various cell structures (4-6,8,50). It must be noted that measurements by Caspersson and his followers have frequently been conducted on tissues fixed in mixtures containing large proportions of both acid and alcohol. Although such treatment produces preparations which are conveniently transparent, i t would be gratifying if the observations could be confirmed on material subjected to fixatives known to preserve best the morphology of the cell structures involved, namely, the nucleolus, microsomes, and mitochondria. It has been suggested that the pattern of ultraviolet absorption ascribed by these authors to nucleic acids, especially that in the nucleiis, is apparent only after cell injury (32).
430
ALBERT CLAUDE
As already noted, analysis reveals similar chemical constituents in mitochondria and microsomes, especially ribonucleic acid and phospholipids; lipositol has been found t o occur in equal proportions in both elements, i.e., 12% of the total lipids (13). It is not known whether the presence of similar substances in both elements indicates the existence of common properties. On the other hand, differences between mitochondria and microsomes are varied and numerous; in size, organization, functions, and degree of complexity they undoubtedly constitute two different classes of cytoplasmic entities. 1. Mitochondria
Mitochondria are relatively large elements, about 0.5 to 1 p in diameter, and therefore are detectable under the light microscope. They may vary considerably in shape, depending on the cell type, from short rods in liver cells to slender filaments several times the length of the cell in cxtcnded fibroblasts; the width, however, although variable aecording t o organs or cell species, is remarkably uniform and constant for a given cell type. In contrast to chromosomes which duplicate according to their length, mitochondria grow by elongation; in certain cells, for example, in the germ cells of Scorpio (59,61) and of the grasshopper,* they can be seen to be passively cut in apparently equal portions by the constricting furrow at the time of cell division. The major chemical constituents so far detected in mitochondria are lipids (25 to 30%)-two-thirds being represented by phospholipids-nucleotides and flavins, and ribonucleic acid ( 1 3 ~ 5 ) . A characteristic feature of mitochondria is the presence of a limiting membrane possessing semipermeable properties. Mitochondria respond osmotically to changes in the salt concentration of the medium; in hypotonic solution they round up and swell, depending on the degree of dilution; in water they may reach the size of a mammalian red cell, and finally disintegrate (I 3,15). Under physiological conditions, when osmotic variations are probably slight and localized, this selective membrane must play an important role in the exchange of fluid and metabolites between the functioning mitochondrion and the surrounding cytoplasm. The most significant development in recent years has been the demonstration that important cell functions are segregated in mitochondria and homologous “large granules” (14). It has been shown by quantitative measurements that the respiratory system which utilizes
* In the germ cells of the grasshopper, filamentous mitochrondria become arranged in bundles alongside the spindle at mctaphase, instead of in the ring formation characteristic of Scorpio, and are divided during cell division so that opposite halves arc retained by the two daughter cells.
PROTEINS, LIPIDS, NUCLEIC ACIDS I N CELLS
431
molecular oxygen is localized in these elements, probably exclusively. This is the case for important members of the system, namely, cytochrome oxidase, succinoxidase (28,47), and cytochrome c (49). Likewise, D-amino acid oxidase activity has been found exclusively in the mitochondria, or large granule fraction (13). Recently Kennedy and Lehninger (30) have shown that other important groups of enzymes, namely, the fatty-acid oxidase system and members of the Krebs cycle, are also segregated in mitochondria. Thus, these distinct cytoplasmic elements appear to constitute the power plants of the cell, where the energy of molecular oxygen is transferred and utilized and, in addition, are probably the site of active metabolic and synthetic processes. A mitochondrion, rod-shaped as in mammalian liver, 2 p long and 0.5 p in diameter, would have a volume of 0.4 p 3 and, for a density of 1.2, a corresponding net weight of 4.8 X lo-' y ; if the dry weight of mitochondria is not greater than th at found for the whole liver, i.e., 30% of the wet weight, the solid matter present in a single mitochondrion would amount to 1.4 X y. It has been shown that lipids account for 25 to 30 % of the mitochondrial substance (11,13); if we assume that another 20 t o 25% of the dry weight is represented by inorganic matter and compounds of low molecular weight, it appears that proteins may contribute approximately one-half of the mitochondrial body, or 7 x 10-8 y of the bulk. It follows that a mitochondrion of the size considered could accommodate a t least one million protein molecules, a figure calculated on the basis of proteins of average molecular weight of 35,000 (absolute weight, 6 x lO-'4 y). If we venture the further assumption th a t one mitochondrion has a complement of, let us say, 25 different enzymatic systems, such as those already identified (the cytochrome-linked system, the fatty-acid oxidase system, the so-called Krebs cycle) each system being composed of 20 different protein molecules, it is apparent that there could exist, simultaneously in the same mitochondrial unit, as many as 2000 duplicates of each of the 25 enzyme systems postulated. The possibility has already been discussed (17,lS) that special activities may be segregated in different cytoplasmic granules, thus allowing for an even greater variety of functions. The manner in which the various enzyme systems are integrated in the mitochondrial structure is not known. The type of growth by elongation and transverse subdivision of mitochondria suggests th a t the different enzyme systems present must be uniformly distributed, since uneven arrangement along the rod or filament would eventually produce a segregation of the various systems in divergent cell lines, in the course of successive cell division, and result in the progressive loss of functiohs. In the investigation and further understanding of the fine structure of
432
ALBERT CLAUDE
mitochondria it is probable that the biochemical methods, in conjunction with cell fractionation, will play a role similar to that being played by the science of genetics in the understanding of the organization and functions of chromosomes. 2. Microsomes The microsomes are small elements ranging in size approximately from 60 to 250 mp, and therefore not detectable under the light microscope. They were discovered through their concentration and isolation from tissue extracts, by means of centrifugation a t high speed (10); later evidence established their derivation from the cytoplasm and demonstrated that they constituted the basophilic component of the ground substance (13). In contrast to mitochondria no limiting membrane has been demonstrated around these elements although they are affected by changes in the salt concentration of the medium, and become more highly hydrated when in hypotonic solutions (15) ; in electron micrographs of osmic-acid-fixed preparations, the microsomes appear as discrete vesicles generally swollen to various degrees as a result of their passage through distilled water, during the washing to which they are subjected in the course of their preparation for electron microscopy (20,44). Present evidence does not permit one to decide whether this vesiculation of the microsomes is conditioned by a pre-existing semipermeable membrane or is an artifact of osmic acid fixation. Chemical analysis has shown that the microsomes are complex structures composed in large part of lipids and nucleoproteins; as much as 40% of the mass is represented by lipids, two-thirds of which are phospholipids, and 10 to 12% of which is lipositol; ribose nucleotides of low molecular weight have not been detected in appreciable amount but the microsomes have been found to be especially rich in ribonucleic acid. At least 60% of the ribonucleic acid of the cell has been recovered with the microsome fraction (15,29,47,48). The function of the microsomes in the cell is still obscure; biochemical tests so far have failed to detect in the microsome fraction characteristic and exclusive biochemical activities (14,28,47). Recent reports suggest that the microsome fraction may be rich in esterase (41). The thromboplastic activity of tissues appears t o reside in the microsomal fraction (9,15). Their relative abundance-they constitute 15 to 20% of the cell mass-and their high content of ribonucleic acid indicate that the microsomes must have an important share in the normal physiology of the cell. 3. Fibrous Matrix
Another morphological component of importame, long recognized in cells, is a fibrous framework associated with the motility and plasticity
PROTEINS, LIPIDS, NUCLEIC ACIDS I N CELLS
433
of protoplasm, and intracellular movements. In order to account for the peculiarities of the plasma gel layer, changes in cell shape, the activity and direction of cytoplasmic currents, and the formation of asters and spindles, a fibrous protein was postulated, and variations in the degree and type of aggregation of elongated molecules were assumed to he responsible for the properties of reversible gelation displayed by living cytoplasm (51). In some of the events mentioned, especially the production of asters and spindles, it is apparent that the centrosomes have an important and directing influence. The existence of a fibrous protein system is confirmed by the precipitation in the cytoplasm of an abundant fibrous network under the action of a variety of fixatives. The fibrous nature of the network is even more apparent in electron micrographs of cells fixed by acids, and by osmic acid and alcohol (19,44).
V. CONSTITUTION AND DUPLICATION OF LIVINQMATTER This brief review indicates that a cell is composed essentially of six different systems of structures, namely, chromosomes, nucleoli, mitochrondria, microsomes, a fibrous framework, and the centrioles. It must be noted that the chief cell components mentioned are fundamentally distinct in their chemical constitution, their morphology, and their biochemical functions. A living cell, therefore, is not a biochemical continuum but a composite entity, the sum of the interactions of associated elements which appear to be heterogeneous, if not autonomous. In this respect it is probable t b t , as the mechanisms of cellular synthesis are better understood, the concept of “self-duplication ” will be enlarged and extended to other components of the cell (45). In recent years, attention has been centered almost exclusively on the problem of reduplication of gene substance, although it is obvious that all the other essential cell structures are likewise reduplicated during cell growth, or a t the time of cell division. The template theory, elaborated to account for the supposedly unique process of gene reproduction assumes that the gene or the chromosome serves as a mold for the systematic apposition of new substance. This operation, if possible, would result in the production of a negative image, which would be unusable, and could even constitute a danger for the survival of the cell in view of the competing and interfering power that analogs and isomers are known to possess. For exact reproduction by the template process we would have to postulate the preliminary construction of an intermediate replica, to be discarded later. The template theory appears even less workable when viewed a t the molecular level, taking into account the numerous and diverse steps undoubtedly involved even in relatively simple reactions, such as the assembly of a variety of
434
ALBERT CLAUDE
amino acids into a specific protein. In discussing this type of reproduction it is usually forgotten that chromosomes are of considerable thickness and complexity, made up of numerous molecular layers, and it is difficult to visualize how template reduplication could thus operate in depth. Finally, the mode of growth by accretion would presuppose the production of the various building stones in excess so that enough of them may find their proper place, in time. This aspect of synthesis by statistical method also appears impractical and unsafe. I n the light of biochemical processes already known, it is conceivable that the duplication of essential and characteristic cell substances is the end result of a series of rigidly ordered chains of reactions, the final product in turn taking part a t some point, and thereby directing the specificity, of the same or other, interlocked, biochemical cycles. The specificity of a gene is not more striking, and probably not more difficult to achieve, than the structural and functional specificity of a proteolytic enzyme, or that of a polysaccharide. Thus the term “self-duplication” appears meaningless when applied t o those complex but highly organized cyclic biochemical processes leading to the production of new cell substances, and t o the reciprocal action that these may exert on the system that produced them. From chemical analysis and histochemical tests three groups of substances, namely, proteins, lipids, and nucleic acids, appear as the most conspicuous and most abundant constituents of cells. Among these, proteins and nucleic acids have already beqn assigned important functions in cellular physiology; although equally abundant, lipids have only occasionally attracted the attention of cytologists and the consensus seems to be that they play a role less essential than proteins or the nucleic acids in the economy of the cell.
VI. PHOSPHOLIPIDS IN CELLSTRUCTURES AND FUNCTIONS Phospholipids are known to take an important part in the formation of cellular membranes, and have been uniformly detected in the limiting niembrttnes of most morphological cell structures, i.e., the nucleus, mitochondria, vacuoles, secretory granules, and the cell itself. This property of phospholipids to constitute membranes has been beautifully investigated by Nageotte (40),and more recently by Dervichian (25). Even in their simple form, phosphatide films represent filters of high sensitivity, capable of concentrating selectively certain ions etc.; in vivo, they may be reinforced by oriented protein films or, as has been suggested, certain types of membranes may be composed of alternating patterns of phosphatides and proteins. Through their ability to form semipermeable membranes phospholipids have a selective role in the interchange of
PROTEINS, LIPIDS, NUCLEIC! A C I D S IN CELLS
435
water and solutes between nucleus and mitochondria, and the cytoplasmic fluid. About one-third of the total cell mass is made u p of lipids (36); a s already mentioned, the lipid content of mitochondria and microsomes is about 25 and 40%, respectively (13,15). I t has been found that the phospholipids present in these cell inclusions occur in firm chemical association with the rest of the structure since they cannot be removed even by prolonged extraction with organic solvents, such a s ether, benzene, or pentane, unless the complex is severely disrupted (unpublished obaervations). Release of phospholipids, as a rule concomitant with denaturation of the associated proteins and solubilization of ribonucleic acid and ribose nucleotides, is accomplished by a variety of conditions such as heating a t 55°C. for 30 minutes,* hydrogen ion concentrations of pH 3.0 and 12.0, treatment with chloroform, alcohol, or acetone, drying, and repeated freezing and thawing (15,22). The high lipid content of certain structures, for example, mitochondria, cannot be accounted for in the constitution of the membranes alone since upon lysis of the large granules in distilled water a residue is left which is richer in phospholipids than the intact elements; it may be significant t hat the original ribonucleic acid complement of mitochondria is found in this residue in association with the phospholipids, and in proportion close to th at existing in the microsomes (14). Lipids are commonly thought of as substances particularly adapted for supply and storage of energy, or as relatively static constituents of structures such a s membranes, the covering of nerve fibers, etc. This position of lipids is not unique; proteins, essential structural and specifically active constituents of cells, can be utilized as a source of energy if supplied in excess, and constitute many comparatively inert body structures. The abundance of phospholipids in the cell, their occurrence in constant proportion in highly active elements such as mitochondria, their presence in nucleoli and microsomes, in all cases in apparent chemical combination with proteins, ribonucleic acids, and ribose nucleotides (6,8,13,15), suggest that they are integral parts of metabolically active structures. Besides their potential value in the supply and transfer of energy, i t is probable that phospholipids have a definite role in the conduct of biochemical reactions; thus it is possible that, thanks to their
* It is interesting t h a t the lethal points for animal cells are found in the range 37-42°C. in vim, and 47-50°C. in uilro, when the phospholipid-nucleoprotein complexes are known to disintegrate (15). T h e first injury noted microscopically is in the mitochondria (33,43), but it is possible that other phospholipid-nucleoprotein structures may also be affected.
436
ALBERT CLAUDE
dual hydrophilic and hydrophobic (oleophilic) property, phospholipids may condition the physical state of certain biochemical systems. Under ordinary conditions, cytochrome oxidase and succinoxidase are not soluble in salt solutions or water, and i t has been suggested that cytochrome oxidase may exist as a lipoprotein complex (57). On the other hand, there is evidence that certain components of the cytochrome system which are readily soluble in water, such as cytochrome c, are not freely diffusible in the living cell, nor in isolated mitochondria (49). It is apparent that phospholipids, by their chemical association with native protein aggregates, may restrain the dispersion of the latter into the aqueous medium. On the other hand, phospholipids, partly through their hydrophilic properties, may help t o regulate the transfer and exchange of water not only through membranes, but also through the mesh of biochemical structures. VII. PROTEINS
It is probable that the high and often rigid specificity of biological reactions has its physical counterpart in equally rigid spatial configuration of the corresponding biochemical systems. Evidence for this is most abundant in the field of amino acid and protein chemistry where a narrow correspondence is often required between a specific enzyme and the configuration of the substrate. du Vigneaud and coworkers have shown that the rat could not utilize acetyl-D-tryptophan for growth purposes, whereas excellent growth was achieved with acetyl-L-tryptophan. The apparent reason is that the cells of the rat are lacking in the specific enzyme capable of hydrolyzing the acetyl group when the latter is attached to the unnatural isomer. When free D-tryptophan was provided in place of ctryptophan, excellent growth resulted, indicating that the body was equipped for the conversion of the compound into its Zevo-isomer (56). Obviously, high specificity limits the reactivity of the system and the freedom of reaction must decrease correspondingly with an increase in the complexity of the structure. Thus it is understandable that, in the construction of biological systems, an analog or isomer cannot substitute for another, since it would result in spatial distortion and therefore in a profound vitiation of the normal biochemical processes. It is clear that, at an early stage of evolution, a choice had to be made between the various chemical configurations available. It would be of interest to know if option for the Zevo-eniantiomorph entailed some definite advantages for the biological systems and for the organisms evolved on this selected pattern.
PROTEINS, LIPIDS, NUCLEIC ACIDS I N CELLS
437
VIII. NUCLEIC ACIDS The nucleic acids, so far, have not disclosed the variety of configurations characteristic of amino acids and proteins, although the possibility for variation exists and structural differentiation must be postulated, a t least in the case of desoxyribonucleic acid, if the latter has to account for the wide range of genetic potentialities generally ascribed to it. The hypothesis that ribo- and desoxyribonucleic acids can be converted one into the other during growth (4,6,7,50) is based on the simple assumption that the molecular differences between these compounds is slight, and perhaps not greater than the chemical difference observed between their sugars, and between uracil and thymine. Experience has shown that, given the proper tests, substances which a t first appeared elementarily alike, for example, polysaccharides or proteins from the same or from different cell species, could be differentiated into numerous isomers or homologs, each distinguished by physical or by highly specific biological properties. Likewise, it would be reasonable to suspect th a t structural differences will be found not only between ribo- and desoxyribonucleic acid molecules, but also among molecules of the same type of nucleic acid. If conversion or interchange is taking place in the cell between the two main types of nucleic acids, it would be more in keeping with the known specificity of physiological processes to expect it to occur through a complete breakdown of one, followed by synthesis of the other. Recent experiments of Cohen seem to clarify this point, so far as the growth of certain bacterial viruses is concerned (23,24). The observations dealing with the multiplication of TB bacteriophage in Escherichia coli B seem to show conclusively that the desoxyribonucleic acid of the newly formed bacteriophage particles was essentially built from the inorganic phosphorus of the medium. I n addition, it was found that the ribonucleic acid of the host cells remained inert, showing a very low, if any, turnover during the growth of the virus. Thus it would appear that the desoxyribonucleic acid of the bacteriophage was the product of an independent synthetic process and that the nucleic acids of the infected cells did not play a n active part in the reaction. The two types of nucleic acids have a different distribution in cells and i t can be inferred that, in these different locations, they perform different functions. I n animal cells, ribonucleic acid is found in formed elements, i e . , in nucleoli, in mitochondria, and in the microsomes (6,13,15,50); i t may be physiologically significant th a t in each case it occurs in association with appreciable amounts of phospholipids. On the basis of ultraviolet absorption studies and characteristic staining reactions, Caspersson and Brachet have endeavored to correlate
438
ALBERT CLAUDE
the presence of ribonucleic acid in the cell with protein synthesis (6-8,50). It has been pointed out in preceding papers (17,18)that the role attributed t o ribonucleic acid as a factor in protein synthesis is not the only hypothesis t ha t could be held to account for the abundance of this substance in certain cells. So far, the evidence produced has been by association, the abundance of ribonucleic acid being correlated with a presumed high protein requirement by the cell. Proteins and, more recently, nucleic acids remain the most popular structural constituents of the cell. It should not be forgotten, however, that lipids may account for as much as one-third of the cell mass, and th at growing and metabolically active cells have a high phospholipid turnover. Moreover, phospholipids and ribonucleic acid are frequently found together in the cell, being espxially concentrated in microsomes, mitochondria, and nucleoli. If topographical association is significant, it would be reasonable to expect that phospholipids and ribonucleic acid are also related metabolically. It should be pointed out that large quantities of ribonucleic acid in the cytoplavm coincide with another outstanding property of these cells. In general, the cells that have been found to have a high ribonuc!eic acid content, for example, embryonic and tumor cells, have been shown by the work of Warburg and followers to possess t o a high degree the power of anaerobic glycolysis. Thus, the abundance of ribonucleic acid in these cells may prove to be related in some way to their captLcity for anaerobic respiration. Mono- and dinucleotides of various types are known to be involved in energy transfer during metabolic processes and to take part in a variety of enzymatic reactions. It is conceivable th a t ribonucleic acid may play a comparable role in energetic reactions taking place in the course of anaerobic respiration. This view would seem to be supported by the concurrence of large amounts of ribonucleic acid and of active fermentative processes, in cells such as yeast and certain bacteria. It has been demonstrated that the power of aerobic respiration, ie., the transfer and utilization of molecular oxygen through the cytochromelinked system, is localized exclusively in the cytoplasm, precisely, in the large granules and mitochondria (14,28,47). From these findings, and from considerations such as those just presented, it appears possible that the two complementary respiratory mechanisms, a t least in animal cells, are segregated in separate cellular regions, the aerobic respiration being restricted, as already shown, t o the mitochondria, and the energy derived from anaerobic glycolysis being mediated through ribonucleicacid-containing elements. I n the cytoplasm, this function might be assumed, in the main, by the microsome system, a situation which would explain the highly basophilic character of the ground substance of cells placed in an environment where the supply of oxygen is failing, as in
PROTEINS, LIPIDS, NUCLEIC ACIDS IN CELLS
439
certain tumors, or naturally inadequate, as in young embryos. The nucleus, lacking the cytochrome-linked mechanism necessary for aerobic respiration, might be expected to obtain at least part of the energy for its growth from an anaerobic mechanism, possibly through the agency of the nucleolus.
REFERENCES 1. Altmann, R.
(1894). Die Elementarorganismen und ihre Beziehungen zu den Zellen. Von Veit, Leipzig. 2. Avery, 0. T., and Goebel, W. F. (1929). J . Exptl. Med. 60, 521. 3. Avery, 0. T., MacLeod, C. M., and McCarty, M. (1944). J . ExpU. Med. 79, 137. 4. Brachet, J.
(1933). Arch. Biol. Liege 44, 519. 5 . Brachet, J. (1940). Compl. rend. soc. biol. 133, 88. 6. Brachet, J. (1945). Embryologie Chimique. Masson, Paris. 7. Brachet, J. (1947). Cold Spring Harbor Symposia Quant. B i d . 12, 18. 8. Caspersson, T., and Schultz, J. (1940). Proc. Natl. Acud. Sci. U. S . 26, 507. 9. Chargaff, E. (1944). Advances i n Protein Chem. 1, 1. 10. Claude, A. (1938). Proc. SOC.Exptl. Biol. Med. 39, 398. 11. Claude, A. (1941). Cold Spring Harbor Symposia Quant. Biol. 9, 263. 12. Claude, A. (1942). Trans. N. Y. Acad. Sci., [2] 4 , 79. 13. Claude, A. (1943). Biol. Symposia 10, 111. 14. Claude, A. (1945). A.A.A.S. Research Conference on Cancer, July 31August 4, 1944. American Association for the Advancement of Science, p. 223. 15. Claude, A. (1946). J. Exptl. Med. 84, 51, 61. Public Health Labs. 26, 12. 16. Claude, A. (1946). Proc. N . Y . State ASSOC. 17. Claude, A. (1947). Conference on Structure in Relation to Cellular Function. New York Academy of Sciences, New York (in press). 18. Claude, A. (1948). Harvey Lectures Ser. 43, Jan. 15, 1948. 19. Claude, A,, and Fullam, E. F. (1946). J . Exptl. Med. 83, 499. 20. Claude, A., Porter, K. R., and Pickels, E. G. (1947). Cancer Research 7,421. 21. Claude, A., and Potter, J. S. (1943). J . Exptl. Med. 77, 345. 22. Claude, A., and Rothen, A. (1940). J . Ezptl. Med. 71, 619. 23. Cohen, S. S. (1948). J . Biol. Chem. 174, 281. 24. Cohen, S. S. (1948). J . Biol. Chem. 174, 295. 25. Dervichian, D., and Magnant, C. (1946). Bull. soc. chim. biol. 28, 419. 26. Dobzhansky, R. (1936). Biol. Revs. 11, 364. 27. Fox, S. W., Fling, M., and Bollenbsch, G. N. (1944). J . Biol. Chem. 166, 465. Fling, M., Minard, F. N., and Fox, S. W. (1947). J . Am. Chem. SOC.69, 2466. 28. Hogeboom, G. H., Claude, A., and Hotchkiss, R. D. (1946). J . Biol. Chem.
166, 615. 29. Hogeboom, G. H., Schneider, W. C., and Pallade, G. E. (1948). J . Biol. Chem. 172, 619. 30. Kennedy, E. P., and Lehninger, A. L. (1948). J . Biol. Chem. 172, 847. 31. Landsteiner, K. (1945). The Specificity of Serological-Rcactions. Harvard
Univ. Press, Cambridge, Mass. 32. Larionov, L. P., and Brumberg, E. M. (1948). A m . Rev. Soviet Med. 6, 111. 33. MacCardle, R. C. (1937). J . Morphol. 61, 613. 34. McCarty, M. (1946). B a d . Revs. 10, 63.
440
ALBERT CLAUDE
85. McClintock, B. (1946-47). Carnegie Inst. Wash. Year Book 48, 146. 36. Macheboeuf, M. (1936). Elals des Lipides dans la Matibe Viuande. Actualit&
Scientifiques et Industrielles. Hermann et Cie, Paris. Mirsky, A. E. (1947). Cold Spring Harbor Symposia Quant. Biol. 12, 143. Mirsky, A. E., and Pollieter, A. W. (1943). Biol. Symposia 10, 247. Mirsky, A. E., and Ria, H. (1947). J . Gen. Physiol. 31, 1, 7. Nageotte, J. (1937). Morphologie des Gels Lipoides. Actualit& Scientifiques et Industrielles. Hermann, Park 41. Omachi, A., Barnum, C. P., and Click, D. (1948). Proc. Soe. Ezptl. Biol. Med.
37. 38. 39. 40.
87, 133. 42. 43. 44. 45. 46. 47. 48. 49.
Pasteur, L. (1860). Compt. rend. 11, 298. Policard, A. (1912). Compt rend. soc. biol. 72, 228. Porter, K. R., Claude, A., and Fullam, E. F. (1945). J . EzpU. Med. 81, 233. Rhoades, M. M. (1946). Cold Spring Harbor Symposia Quant. Biol. 11, 202. Schneider, W. C. (1945). J. Biol. Chem. 181, 293. Schneider, W. C. (1946). J. Biol. Chem. 186, 585. Schneider, W.C. (1947). Cold Spring Harbor Symposia Quant. Biol. 12, 169. Schneider, W. C., Claude, A., and Hogeboom, G. H. (1948). J . Biol. Chem.
172, 451. 50. Schultz, J. (1941). Cold Spring Harbor Symposia Quant. B i d . 9, 55. 51. Seifriz, W., ed. (1942). The Structure of Protoplasm. Iowa State Coll. Fkess,
Amen, Iowa. 52. Spiegelman, 8. (1946). Cold Spring Harbor Symposia Quant. Biol. 11, 256. 53. Spiegelman, S., and Kamen, M. D. (1947). Cold Spring Harbor Symposia Quant. Biol. 12, 211. 54. Stedman, E., and Stedman, E. (1947). Cold Spring Harbor Symposia Quant. Biol. 12, 224. 55. Sturtevant, A. H., and Beadle, G. W. (1939). An Introduction to Genetics.
Saunders, Philadelphia. 56. du Vigneaud, V., Sealock, R. R., and Van Etten, C. (1932). J . B i d . Chem. 98, 565. 57. Wainio, W. W., Cooperstein, 9. J., Kollen, S., and Eichel, B. (1948). J . Biol. Chem. 173, 145. 58. Wicklund, E. (1948). Nalure 181, 556. 59. Wilson, E. B. (1916). Proc. Natl. Acad. Sci. U.S.2, 321. 60. Wilson, E. B. (1925). The Cell. 3rd ed., Macmillan, New York. 61. Wilson, E. B. (1931). J. Morphol. andphysiol. 62,429. 62. Woolley, D. W. (1947). Physiol. Reus. 27, 308; (1945-46). Haruey Lectures 41, 189. 63. Wright, S. (1941). Phyuiol. Reus. 21, 487. 64. Wright, 9. (1945). Am. Naturalist 79, 289.
Author Index Numbers in parentheses are reference numbers. They are included to assist in locating references in which the authors’ names are not mentioned in the text. Names in parentheses indicate coauthors. Numbers in italics indicate the page on which the reference is listed. Ezample: Akkerman, A. M., 44 (ref. 224),53, 80 meana that this author’s article is reference 224 on p. 44, that it is mentioned on p. 53, and is lkted on p. 80 at the end of the article. Allison, J. B., 158, 160, 161, 162, 163, 164, 166, 167, 168, 169, 170, 171, Abderhalden, E.,11, 13, 14, 15, 17 (ref. 177, 181, 182, 185, 187, 188, 189, 196,197 7), 36 (ref. 6,24), 37 (ref. 8, 16), 38 (ref. 8, 9, 12, 15, 16, 17, 24, 28, 30, Almquist, H. J., 244, ,996 32, 34,36), 39 (ref. 6,8,9, 10,12, 15, Altman, R. F. A., 241, 296 16,28,30, 33,34,37), 40 (ref. 14,26, Altmann, R.,429, 489 27, 36), 46, 47 (ref. 3), 49 (ref. 13), Alvaree-Tostado, C.,212 (ref. 60),226 50 (ref. 2, 3),51 (ref. 2,3, 13, 17),53 Anderson, E.G., 158 (see Borman), 197 (ref. 35), 54 (ref. 21, 35), 55, 57 Anderson, J. A., 158, 160, 161, 163, 164, 167, 168, 170 (see Allison), 171, 177, (ref. 7), 58, 61 (ref. 22),65,66 (ref. 181, 185, 187, 188, 196, 197 22), 67, 68, 69 (ref. 20, 25), 70 (ref. Anslow, W.K., 373 (ref. l),416 23, 25), 71 (ref. 18), 72, 76,76,78 Abderhalden, R., 46 (ref. 2), 50 (ref. 2), Anson, M.L., 139, 141, 148, 161,346 Archibald, R. M.,85, 148 51 (ref. 2), 76 Arenz, B., 279, 296 Abeniue, P. W., 5, 76 Abitz, W.,308 (ref. 57),328 (ref. 57),948 Arhimo, A. A., 271, 303 Armstrong, W.D., 182, 197 Ackerman, H.,268, 303 Arnold, C.,219, 222 (ref. 2),296 Adair, G.S., 128, 134, 148, 163 Adair, M. E., 128, 134,2 4 8 , 163 Artom, C.,57, 79 Asenjo, C. F., 198 Addis, T., 182, 187, 196 Adler, E.,271, 297 Astbury, W.T., 141, 143, 148, 306 (ref. Agatov, P.,240,294, ,996,299 11, 12, 15),308 (ref. 9, 12), 309, 310 Agner, K., 222, 226 (ref. 14, 30), 311 (ref. 8, 14), 316 h e s o n , A., 89, 163, 219 (ref. 135), 222 (ref. 4, 5), 326, 328 (ref. lo), 329 (ref. 135), (ref. 6), 346, $47, 359, 360, 369 Akkerman, A. M., 44 (ref. 224),53,80 (ref. 3), 395 (ref. 3), 416 Albanese, A. A., 89, 118, 119, 120, 148, Atkin, W.R., 369 (ref. 16),372 (ref. 16), 395 (ref. 185), 400, 426,4.90 158, 171 (see Cox), 176, 196, 198, Audrieth, L. F., 52 (ref. 286),82 234, 243, 296 Albaum, H. G., 276, ,996 Auerbach, G.,67 (ref. 194), 68 (ref. 194), 79 Alcock, R. S., 276, ,996 Avery, 0. T., 424 (ref. 2, 3), 427 (ref. 3), Alekseeva, T. S., 255,303 Alfrey, T., 316 (ref. 3), 327, 332 (ref. 439 Ayres, M. M.,180 (see Bosshardt), 197 1, 2), 3qs Alge, A., 362 (ref. 114),428 B AUing, E. L., 181 (see Zeldis), 184, 198, Bach, S., 406 (ref. 47),426 199, 2m Bacon, J. S. D., 240, $99 Allison, F. E., 274, 296, ,998 441 A
442
AUTHOR INDEX
Bawden, F. C., 254,255, 293,296 Baer, E., 25, 79 Beach, E.F., 105, 148, 202 (ref. 8), 203 Baernstein, H.D., 96, 103: 105, 148 (ref. 8),218 (ref. 8),226 Baertich, E.,46 (ref. 2), 50 (ref. 2), 51 Beadle, G. W., 424 (ref. 55),4.40 (ref. 2),76 Bahn, A,, 47 (ref. 3), 60 (ref. 3), 51 (ref. Beadles, J. R., 176, 197, 199 Bear, R. S., 306 (ref. 19, 20), 346, 360, 3), 65,76 359,416 Bailey, C. H., 262,303 Beatty, W.A., 67 (ref. 232),80 Bailey, J. L., 22 Bailey, K.,96,97,103,105, 106,108,112, Beek, J., Jr., 362 (ref. 8), 363 (ref. 9), 367,401,416 115, 133, 134, 136, 140, 141, 142, 147, 148, 148,218 (ref. 3), 826, 233, Beeson, W.M., 244, 304 247, 257, 296, 896, 306 (ref. 16), 309 Behrens, M.,265,298 Behrens, 0.K., 23, 37 (ref. 43), 43 (ref. (ref. 7),310 (ref. 30), 346, 347 248), 44 (ref. 248), 49 (ref. 44), 60 Bain, J. A., 215 (ref. 4), 826 (ref. 43), 63 (ref. 248), 62 (ref. 45), Baker, W.O., 328 (ref. 17, 18, 56), 348, 65 (ref. 45), 66, 76, 81, 82 343 Bell, F., 311 (ref. 8),346 Balbiano, L., 2,76 Belozersky, A. N., 240,265,268,296, %09 Baldwin, E.,285,295, 296 Baldwin, M. E., 384 (ref. 4, 199, 200), Belt, A. E., 182, 199 Benditt, E.P., 190, 197 395 (ref. 4),416, 420 Balfe, M. P., 394 (ref. 5a), 395 (ref. 5), Benedict, F. G., 264, 296 Berg, J. L., 182,197 400 (ref. 5),416 Ball, E. G.,219 (ref. 5), 220, 221 (ref. Bergell, P.,9 (ref. 144), 10 (ref. 144), 78 Berger, L.,264,206 log),2%6,2.27 Berggren, R.E. L., 204 (ref. 22),207,226 Ballou, G.A., 217 (ref. 6),226 Balls, A. K., 33 (ref. 40), 73, 76, 222 Bcrgmann, M.,2 (ref. 46), 3,6,6, 7 (ref. 78), 10 (ref. 145),11 (ref. 50), 15, 16 (ref. 7),226, 292, 293, 300 (ref. 51, 60,67,90, 115), 17 (ref. 64, Baly, E.C. C., 270,296 65, 189),18, 19, 20 (ref. 57, 189),21 Barkdoll, A. E., 31 (ref. 41),47, 60 (ref. (ref. 74), 23, 25, 26 (ref. 74, 77, 80, 41),76 85), 30 (ref. 187, 188), 31 (ref. 49, Barmore, M. A., 262, 296 86,276),33 (ref. 47, 62,64),34 (ref. Barnes, R. H., 162, 173, 174, 175, 180, 190), 36 (ref. 49, 56, 74, 77, 80, 89, 197 188, 189, 276, 278), 37, 38 (ref. 49, Barnum, C. P., 432 (ref. 41),440 56, 78, 80,278), 39 (ref. 49, 78, 80), Barth, K.,381 (ref. 175),420 40 (ref. 53, 80, 276, 277), 41 (ref. Bartholomew, E.T., 241,302 67, 74), 42 (ref. 74, 87,88), 43 (ref. Bartner, E.,1 1 1 (see Smith, E.L.), 112 60, 67, 74, 76, 78, 79, 87, 88, 187, (see Smith, E. L.), 113 (see Smith, 189, 216), 44 (ref. 74, 77, 78, 88, E. L.), 117 (see Smith, E.L.), 162, 187),45 (ref. 74,87,88), 46, 47 (ref. 212 (ref. 123), 213 (ref. 123), 218 50, 85,88, 187),48 (ref. 50, 188), 49 (ref. 123),2%8,263, SO2 (ref. 44,50, 53,56,60,62,66,67,74, Barton, R. W., 171 (see Cox),108 85, 115, 187, 188, 215), 50 (ref. 43, Bass, L. A., 38 (ref. 231), 80 60, 51, 53, 66, 74, 88, 188),51 (ref. Bates, M.J., 173, 174 (see Barnes), 175 65, 67, 88, 90), 63 (ref. 74), 55, 57 (see Barnes), 197 (ref. 81,65, 74, 279), 59, 60,61,62, Baudisch, O., 270,296 63, 64,66,68,69, 70, 71,74, 76, 77, Baudouy, C.,371 (ref. 37), 416 78, 79, 80, 88, 148, 277, 896, 357 Baumann, E.,64 (ref. 42),76 (ref. lo),360,400 (ref. 12),416, 416 Baumann, L.,72 (ref. 5), 76 Bernal, J. D., 133, 149, 254, 896 Baur, H., 40 (ref. 117), 78
AUTHOR INDEX
Bernhart, F. W., 191, 200 Bernstein, S. S., 105 (see Bench), 148,202 (ref. 8), 203 (ref. 8), 218 (ref. 8), 226 Berridge, N. J., 209, 226 Bertho, A., 24, 77 Best, R. J., 237, 254, 255, 256, 293, 294, 296, 300
Bezer, A. E., 263, 299 Billimoria, M. C . , 288, 301 Birch, T.W., 406 (ref. 13), 416 Birkhofer, L., 96, 103 (see Kuhn), 105 (see Kuhn), 139 (see Kuhn), 151 Bivshich, N., 240, 299 Bjorksth, J., 272, 296 Black, A., 117, 160, 244, 298 Black, H. C., 158 (see Borman), 197 Blake, M. A., 271, 303 Blank, P., 15 (ref. 146), Y8 Blaxter, 162, 197 Blish, M. J., 262, 296, 301 Block, H., 93, 103 (see Dunn), 110 (see Dunn), 160 Block, R. J., 97, 101, 102, 103, 105, 109, 110, 111, 112, 113, 114, 116, 117, 118, 119, 120, 129, 130, 14.9, 169, 192, 193, 194, 197, 199, 236, 237, 243, 244, 245, 246, 247, 250, 263, 296, 300 Blom, J., 274, 296 Blotter, L., 108 (see Lyman), 112 (see Kuiken), 113 (see Kuiken), 119 (see Lyman), 138 (see Lyman), 161, 235, 244, ,999 Blum, A. E., 110 (see Horn), 120 (see Horn), 161, 263, 298 Blum, W. A,, 399 (ref. 197), 409 (ref. 1971, &O Blumenthal, D., 96, 103, 105, 149, 247, ,996 Boivin, A., 268, 296 Bolin, D. W., 244, 301, 304 Bollenbach, G . N., 424 (ref. 27), 439 Bolling, D., 97, 102, 103, 105, 109, 110, 111, 112, 113, 114, 116, 117, 118, 119, 120, 129, 130, 149, 237, 243, 244, 245, 246, 250, 263, 296 Bondy, C., 241, 296 Bonem, P., 63 (ref. 118), 78 Bonner, D. M., 273, 303 Bonner, J., 260, 267, $04
443
Boothe, J. H., 45 (ref. 91a), 77 Borasky, R., 306 (ref. 133), 360,359 (ref. 162), 360 (ref. 162), 419 Borman, A., 158, 197 Borodin, J., 278, 283, 296 Borsook, H., 33 (ref. 92), 77 Bosshardt, D. K., 162, 173, 174, 175, 180, 197, 216, 226
Bot, G. M., 267, 296 Bourque, J. E., 184, 199 Boussingault, J. B., 277, 279, 296 Bower, F. O., 257, 296' Bowes, J. H., 143, 149, 357, 377 (ref. 14), 391 (ref. 13a), 406 (ref. 15), 414, 416, 421
Bowles, L. L., 182 (see Berg; Hall, W. R.), 197, 198 Bowman, D. E., 263, 296 Boyd, G. L., 211, 226 Boyd, M. J., 109, 149 Boyer, P. D., 217 (ref. 6), 226 Boyer, R. A., 311 (ref. 21), 315 (ref. 21), 347
Boyland, E., 253, 297 Brachet, J., 429 (ref. 4-6) 435 (ref. 6), 437, 438 (ref. 6, 7), 439 Brand, E., 85, 87, 88, 93, 95, 96, 100, 101, 102, 103, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 128, 130, 132, 133, 135, 136, 139, 142, 145, 149, 161, 162, 216, 218 (ref. 13, 61), 223, 226, 226 Brandon, B. A., 307 (ref. 49), 34'7 Branscombe, D. J., 235, 902 Brauns, F., 411, 412, 418 Braunstein, A. E., 20, 77, 271, 273, 296 Braybrooks, W. E., 369 (ref. 16), 372 (ref. 16), 400, 416 Brazier, M. A. B., 248, 296 Breusch, F. L., 285, 296 Bricker, M. L., 163, 197 Briefer, M., 367, 416 Brierley, P., 293, 296 Briggs, D. R., 217, 226 Brockmann, H., 36 (ref. 6), 39 (ref. 6), 76 Brown, A. E., 38 (ref. 275), 51 (ref. 275), 81, 307 (ref. 22), 328 (ref. 70), 331 (ref. 70), 347, 348 Brown, E. B., 250, 299
444
AUTHOR INDEX
Brown, J. H., 162, 163, 167, 168, 169, 170 (see Allison), 171, 182, 185, 189, 197 Brown, R., 287,296 Briickner, V., 45 (ref. 222), 80 Bruhre, P., 263, 301 Brumberg, E. M., 429 (ref. 32), 4.39 Bruah, M. K., 171, 172, 197, 200 Buadze, S., 17 (ref. 7), 57 (ref. 7), 76 Buchheimer, K., 362 (ref. 127), 418 Buehler, H. J., 124, 125, 143, 149 Bull, H. B., 128, 149, 215 (ref. 15), 226, 309 (ref. 23), 330, 335 (ref. 24, 25), 547
Bungenberg de Jong, H. G., 395 (ref. 18), 416
Burack, E., 185, 198 Burk, D., 274, 998 Burk, N. F., 128, 134, 1.69, 207, 226, 416 Burnett, R. S., 311 (ref. 26), 315 (ref. 261,347
Burris, R. H., 250, 251, 254, 274, 275, 296, 997,304
Burstram, H., 270, 29'7 Burton, I. F., 25 (ref. 94), 77 Busse, W. T., 335, 547 Butler, B., 103 (see Lyman), 161 von Buzagh, A., 311 (ref. 28), 325, 347 C
Cable, R. S., 223 (ref. 147), 226, 228 Cahill, W. M., 25 (ref. 94), 77 Caldwell, T. D., 311 (ref. 138), 314 (ref. 138), 560
Cameron, D. H., 387 (ref. 146), 400 (ref. 150), 419
Camien, M. N., 93, 101 (see Shankman), 103 (see Dunn), 108 (see Dunn), 110 (see Dunn), 111 (see Dunn), 119 (see Dunn), 120 (see Dunn), 1.69, 160, 162, 218 (ref. 115), 228, 250, 251, 252, 253, 254, 297
Campbell, R. M., 182, 187, 197, 198 Cannan, R. K., 85, 88, 93 (see Keston), 94, 100, 101 (see Keston), 106, 108, 114 (see Keston), 126 (see Longsworth), 135, 136, 138, 1.69, 161, 168, 214 (ref. 17), 216 (ref. 17), 218 (ref. 66), 826, 226, 263
Cannon, P. R., 158 (see Frazier), 189, 190, 191, 192, 194 (see Block), 197, 198, 200
Carman, G. G., 165, 174, 199 Carothers, W. H., 310, 347 Carpenter, C. M., 268, 303 Carpenter, D. C., 207, 228 Carpenter, L. M., 207, 228 Carter, H. E., 18, 77, 89, 149, 249, 297 Carter, J. R., 183 (see Madden), 198 Caspersmn, T., 265, 277, 297, 429, 435 (ref. 8), 437, 438 (ref. 8), 439 Caswell, M. C., 103 (see Stokes), 110 (see Stokes), 111 (see Stokes), 112 (see Stokes), 113 (see Stokes), 117 (see Stokes), 120 (see Stokes), 165, 256, 303 Cederquist, D. C., 245, 297 Chambard, P., 400 (ref. 20), 416 Chanutin, A., 207 (ref. 18), 226 de Chardonnet, Count, 310, 311 Chargaff, E., 266, 268, 297, 432 (ref. 9), 439
Cherbuliez, E., 205, 226, 369 (ref. 21), 416
Chernov, N. W., 362, 416 Chevalier, R., 221 (ref. 43), 226 Chibnall, A. C., 2 (ref. 99), 77, 85, 87, 93, 97, 101, 104, 105, 106, 107, 108, 113, 114, 115 (see Bailey), 116, 118, 119, 120, 130, 131, 132, 133, 134, 135, 136, 138, 139, 143, 147, 148, 149, 218 (ref. 3), 226, 233, 234, 235, 237, 238, 239, 241, 253, 257, 264, 267, 270, 271, 272, 277, 279, 281, 283, 284, 285, 297, 300, 310, 347, 357, 358, 359, 360, 416 Chick, H., 179, 197, 240, 297 Chou, C., 57 (ref. 279), 82 Chow, B. F., 125, 149, 181, 187, 197 Chuan Chti, P., 398 (ref. 51), 427 Christensen, L. K., 217 (ref. 59), 218, 226 Christian, W., 264 Chrzaszcz, T., 219 (ref. 21), 226 Circle, S. J., 262, 502 Clandinin, D. R., 193, 194, 197, 199 Clapp, 5. H., 3, 67 (ref. 253), 68, 70, 81 Clark, H., 200 Clark, L. C., 170 (see Murlin), 171 (see Murlin), 198, 199
445
AUTHOR INDEX
Clarke, H. T., 18 (ref. lOOa), 64 (ref. loo), 75 (ref. l a , 257), 77, 81, 96, 103, 105, 1.69, 233, 246, 247, 248, 249, 257, 296,300,303 Claude, A., 425 (ref. 10, 13-15, 28), 426, 427 (ref. 15), 428 (ref. 16), 429 (ref. 10, 11, 13, 15),430 (ref. 13-15), 431 (ref. 11, 13, 17, 18, 28, 49), 432 (ref. 10, 13-15, 20, 28, 44), 433 (ref. 19, 44), 435 (ref. 13-15, 22), 436 (ref. 49), 437 (ref. 13, 15), 438 (ref. 14, 17, 18,281, 439,440 Coghill, R. D., 250,297 Cohen, P.P.,276, 286,2996,297 Cohen, S. S., 437,4.99 Cohn, E. J., 3 (ref. 101), 33 (ref. 101), 77, 133, 144, 149, 204 (ref. 22), 207, 215 (ref, 33), ,926, 263, 297, 341 (ref. 31),347, 363 (ref. 25), 365, 416 Cole, W. H., 181, 187 (see Chow), 197 Coleman, D., 109, 110, 114, 116, 119, 149, 311 (ref. 32, 33), 324, 325,347 Colowick, S. P.,264, 296 Cone, L. H., 62 (ref. 147), 65,78 Consden, R.,3 (ref. 103),77,97, 136,1.49 Cook, A. H., 18 (ref. 103a),77 Cooper, M., 347, 409 (ref. 40), 410 (ref. 401,416 Cooperstein, S. J., 436 (ref. 57), 4-40 Copley, M. J., 311 (ref. 131, 132), 313 (ref. 151), 325 (ref. 131), 327, 328 (ref. 151),349,360 Copping, A. M., 179,197 Corbet, A. S., 270,897 Cori, C. F., 264, 296 Cornforth, J. W., 18 Corran, H.S.,220, 221 (ref. 23), 226 Cougny, A., 371 (ref. 38),41.6 Cowgill, G. R., 158, 159, 174, 185, 198 Cox, W.M., Jr., 159, 171,198 Coy, N. H., 212 (ref. 120),228 von Cramm, E., 43 (ref. 283), 44 (ref. 283),82 Cravens, W. W., 193, 194,197, 199 Crisp, D. J., 311 (ref. 34),347 Croad, R. B., 403, 416 Crook, E.M., 234, 255, ,997 Croston, C. B., 311 (ref. 35, 37), 314, 347 Crowfoot, D., 128, 133, 1.69, 215, 2.26
Crowther, C., 202,210, 212, 223 (ref. 25), 996 Cruickshank, D. H.,281, 282, 286, 287, 289,304 Csonka, F. A., 103, 149, 243, 249, 262, 297 Currie, B.T.,215 (ref. 15),226 Curtius, H., 9 (ref. 108), 77 Curtius, T.,6,7,9, 10, 15,21, 28, 77 Custer, J. H., 126 (see McMeekin), 127 (see McMeekin), 135 (see McMeekin), 161, 212 (ref. 80), 217 (ref. 79,81, 82), 227 Cuthbertson, W. R., 376, 416 Cutting, M. E. M., 240, 2997
D Dahlberg, A. C., 222,228 Dakin, H.D., 357 (ref. 28),416 Damodaran, M., 208, 226, 280, 297 Danielason, C. E., 395 (ref. 29),416 Darmon, 5. E., 112, 117, 1.69 Das, N. B., 271, 297 Davidson, J. N., 266, 277,297 Davidssohn, B.,2,78 Davies, W.L., 240, 297 Davis, B. D.,217,226 De, P. K., 274, 698 Dekker, C. A., 34 (ref. 112a), 48 (ref. 112b), 71 (ref. 112a, 112b), 72 (ref. 112b),73 (ref. 112b), 77 Delaporte, B., 265, ,997 De Ley, J., 259,304 Della Monica, E. S., 126 (see McMeekin), 127 (see McMeekin), 135 (see McMeekin), 161, 212 (ref. 80), 217 (ref. 79, 81, 82),227 Denes, K., 205 (ref. 42),226 Dent, C.E., 57, 77, 271, 252, 288, 297 Denton, C. A., 103,149 Derksen, J. C., 361 (ref. 115),418 Dervichian, D., 434,439 Desveaux, R.,270, 299 Deuel, H.J., Jr., 171,198 Deutsch, H. F., 202, 215 (ref. 4), 223 (ref. 28), 226 Dewan, J. G., 220 (ref. 23), 221 (ref. 23), 2.96 Dexter, S. T.,367, 418
446
AUTHOR INDEX
Dickineon, S., 309 (ref. 7), 3.bs Dirr, K., 61 (ref. 114), 64 (ref. 114, 123), 77, 78 Diskant, E. M., 100 (see Brand), 112 (see Brand), 149 Dittrich, W., 271, 297 Dixon, M., 220,S'86 Dobzhansky, R.,424 (ref. 26), 439 Doherty, D. G . , 16, 18, 37 (ref. 115), 49 (ref. 44, 115),76, 77, 886 Doty, D. M., 243,897 Doty, P.M., 332 (ref. 2),546 Douglas, G. W.,394 (ref. 30),416 Driscoll, P. E., 36 (ref. 19Oa), 38 (ref. Nos), 79 Dubs, R. J., 217, 886, 268 (see Robinow), 302 Dubuiseon, M., 140, 146,f@ Du Buy, H. G., 294,304 D b r , W.,79 DufrBnoy, J . , 294, 297 Dunn, M. S., 15 (ref. 267), 39 (ref. 267), 81, 85, 91, 92, 93, 100, 101, 103, 108, 110, 111, 119, 120, 149, 160, 168, 218 (ref. 115), 888, 250, 251, 252, 253, 254, 297 Duspiva, F., 376, 429 Dwyer, I. M., 103 (see Stokes), 110 (see Stokes), 111 (see Stokes), 112 (see Stokes), 113 (see Stokes), 116 (see Gunness), 117 (see Stokes), 120 (see Stokes), 160, 163, 256, 303 Dyer, E., 49 (ref. 116), 78
Edwards, L. E., 170 (see Murlin), 171 (see Murlin), 198, 199 Ehrensviird, G., 2, 23, 78 Eichel, B., 436 (ref. 57), 4-40 Eichel, H., 24, 79 Elam, D. W., 317 (ref. 94), 349 Eldred, N. R., 194,198 Ellenbogen, 128, 132 Ellison, H.L.,400 (ref. 150),419 Elman, R.,171 (see Cox), 187, 198 Elad, E., 367 (ref. 33),387, 416 Elsden, 5. R., 95, 160 Elvehjem, C. A., 103 (see Riesen), 162, 193, 194,197, 199 Engel, H., 211 (ref. 32), 826 Engel, L. L., 20,80 Enger, R.,58, 78 Engler, 231 Eppling, F. J., 274,296 Erickson, J. O . , 97 (see Neurath), 161, 309 (ref. 130), 311 (ref. 130), 315 (ref. 130),316 (ref. 130),349 Erlenmeyer, E., 16, 49 (ref. 121), 78 Erxleben, H.,253, 999 Essig, K.A., 349 von Euler, H., 271, 997 Evans, C.D., 311 (ref. 35, 37), 314, 347 Evans, R.J . , 193, 198 Everson, G. J . , 171 (see Swanson, P.), 199, 245, 897 Ewald, A.,369 (ref. 34, 35), 371 (ref. 35), 411,416 Eyring, H.,308 (ref. 38),332 (ref. 3941, 161), 333,335,347, 5.48,560
E F Eckerson, S. H., 270, 997 Fankuchen, I., 215 (ref. 33), 826, 254, Eckstein, H.C . , 250, 897 896 Eddy, C. R., 311 (ref. 152), 325 (ref. FaurB-Fremiet, E., 371,411 (ref. 36), 4f6 152),360 Edlbacher, S.,40 (ref. 117),63 (ref. 118), Fearing, D. F., 314 (ref. 168),361 Felix, K.,64 (ref. 123),78 78 Edaall, J. T., 3 (ref. 101), 33 (ref. 101), Ferguson, F. P., 162, 166, 182 (see Allison), 197 77, 85, 93, 95, 105, 107, 128, 133, 142, 144,14.9, 160, 207, 215 (ref. 33), Ferri, C.,335 (ref. 121), 34.9, 411, 419 216, ,926, 263, 897, 307 (ref. 54), 311 Ferry, E. L.,175,199 (ref. 36), 316 (ref. 36),341 (ref. 31), Ferry, J. D., 217 (ref. 34), 826, 306 (ref. 42), 311 (ref. 42, 43), 332 (ref. 42), 347, 363 (ref. 25), 365,405 (ref. 43), 406,407,413,416 347
AUTHOR INDEX
Feulgen, R., 265,,998 Fevold, H. L., 249,,999 Ficken, K.,390 (ref. 137), 429 Fiedler, A., 43 (ref. 148), 44 (ref. 148),78 Fiere, H. E., 24, 79 Fink, H., 250, ,998 Fischer, E.,1, 2,3,5,6,7,8,9,10,11, 12, 13, 14,15, 17,21,23,25,27,36 (ref. 132,134, 136,149, 158,160, 167),37, 38 (ref. 128, 129, 133, 134, 136, 140, 149, 161, 164, 167, 1681,39 (ref. 136, 138, 154, 164, 168), 40, 41, 43 (ref. 148,155,157),44 (ref. 139,148,155157), 46, 49, 50 (ref. 129, 150, 158, 166. 171),51 (ref. 129,139,140,165), 53 (ref. 150, 171), 54 (ref. 150, 171), 55,57,58,61 (ref. 172),62 (ref. 147), 63, 65, 66, 67, 69 (ref. 162), 70, 75 (ref. 133), 78,79,307 Fischer, H. 0. L., 25,79 Fisher, A. M., 132,262 Fishman, W.H., 57, 79 Fishmann, M. M., 267, 298,300 Fleiachmann, R., 37 (ref. 8), 38 (ref. 8), 39 (ref. 8),76 Flemming, 428 Fling, M., 424 (ref. 27), 4.39 Flory, P. J., 312 (ref. 44), 328 (ref. 45, 46),335,S47 Fo6, C., 311 (ref. 47),347 Fodor, A., 13, 15, 38 (ref. 9), 39 (ref. 9, lo), 76 Fogg, G. E.,274, ,998 Folin, 98,247 Foltzer, J., 310 (ref. 48),347 Fontaine, T.D., 262, 298,299 Foreman, F. W., 97,260 Forster, M.O., 24, 79 Foster, G. L., 85, 90, 93, 94, 101, 106, 108, 111, 112, 114, 116, 120, 122, 260,262 Foster, S. B., 373, 384 (ref. 201), 385 (ref. 201), 386 (ref. 201), 395 (ref. 202),397 (ref. 204),420 Fourcroy, A. F., 233, 698 Fourneau, E., 1, 5, 7 (ref. 149), 9, 27 (ref. 149),36 (ref. 149),38 (ref. 149), 78 Fourt, L., 331 (ref. 71),348 Fox, 8.W., 424 (ref. 27), 439
447
Fraenkel, G., 412, 426 Fraenkel-Conrat, A , 96, 119, 121, 129, 133, 135, 145,262 Fraenkel-Conrat, H., 76, 307 (ref. 49-52, 134),347,360,407, 409 (ref. 40,41), 410,426 Fraenkl, W., 120 (see Dunn), 160 Frampton, V. L., 235, 260, 293, 294, ,998 Frank, V. S., 32,54 (ref. 270a),81 Frankel, M.,7, 22, 67 (ref. 177), 79,80, 307 (ref. 53),347 Frazier, L. E., 158, 189, 191, 194 (see Block), 197,298 Freeland, J. C., 250, 251, 252, 254, 276, ,998 Freeman, S., 182, 298 French, D., 307 (ref. 54), 347, 405 (ref. 43),406, 407,416 Freudenberg, K.,13 (ref. 180), 24, 25, 79,397,426 Freund, E. H., 308 (ref. 55),348 Freundlich, H.,241, ,996 Fricke, R.,11, 50 (ref. 206), 80 Fried, S.,299 Friedes, R., 108 (see Hier), 110 (see Hier), 111 (see Hier), 112 (see Hier), 113 (see Hier), 116 (see Hier), 118 (see Hier), 119 (see Hier), 120 (see Hier), 260 Fritsch, F. E., 274,298 Frost, D. V., 159,298,199 Frllstllck, E., 16,49 (ref. 121), 78 Fruton, J. S.,3 (ref. 191,250),7 (ref. 78), 11 (ref. 50), 16 (ref. 51), 17 (ref. 186, 189, 304), 19 (ref. 54), 20 (ref. 189), 26 (ref. 80, 181), 30 (ref. 187, 188),31 (ref. 49),33 (ref. 52,54, 182, 186),34 (ref. 112a,190),35 (ref. 185), 36 (ref. 49, 80, 188, 189, 19Oa), 37 (ref. 78-80, 277),38 (ref. 49, 78, 80, 19Oa), 39 (ref. 49, 78, 80), 40 (ref. 53, 80, 184, 277), 41 (ref. 186), 43 (ref. 50, 78, 79, 187, 189), 44 (ref. 78, 187, 274), 45 (ref. 183),47 (ref. 50, 187), 48 (ref. 50, 112b, 188), 49 (ref. 50,53, 187, 188),50 (ref. 50,51, 53, 188), 51 (ref. 274), 56 (ref. 181, 186),57 (ref. 181,186),58 (ref. 181), 61 (ref. 55), 64 (ref. 55, 186), 68 (ref. 273a), 69 (ref. 273a), 70 (ref.
448
AUTHOR INDEX
273a), 71 (ref. 112a, 112b, 18h, 188), 72 (ref. 112b, 18h), 73 (ref. 112b),76, 77, 79, 82, 82, 277, f?96 Fuchs, F., 21,79 Fugitt, C. H., 363 (ref. l90), 364 (ref. 1911,MO Fuld, M.,307 (ref. 123), 308 (ref. 123), wQ
Fullam, E. F., 425 (ref. 19), 432 (ref. 44), 433 (ref. 19, 44),@9, 440 Fuller, C. S., 328 (ref. 17, 18, 56), 346, 34R G
Gale, E. F., 90, 108, 116, 118, 119, 120, 122, 130,260,218 (ref. 35),$26, 250, 251,252, 253,254, 276, $98 Gall, E. C.,307 (ref. 22),347 Gallacher, A. M., 295 Gallun, E.A., 384 (ref. 213),482 Calvin, J. A., 328 (ref. 137),360 Garrod, M.,335 (ref. 89),348 Caw, H. Z.,256, $98 Geidel, W.,65,76 Geiger, W.,21 (ref. 229),37 (ref. 229),49 (ref. 229), 80, 307 (ref. 87),348 Gerngross, O., 50 (ref. 150),53 (ref. 150), 54 (ref. 150), 78, 308 (ref. 57), 328 (ref. 57),348, 405 (ref. 46), 406,426 Gibbs, J. W., 335 (ref. 58), 348 Gibbs, W.,335,34.8 Giffhorn, A,, 220, 226 Gillespie, H. B., 64 (ref. loo),77 Glasstone, S.,333 (ref. 59),34.8 Gliek, D.,432 (ref. 41),440 Gluud, W.,14 (ref. 151), 78 Go, Y.,307 (ref. 120), 311 (ref. 60),315 (ref. 60),348, 349 Goebel, W.F., 424 (ref. 2), 439 Goettsch, E.,186 (see Weeeh), 200 Gohdes, W., 38 (ref. 12), 39 (ref. 12), 76
Goldberg, 8.C., 108 (see Dunn), 149 Goldwater, W. H., 116 (see Brand), 120 (see Brand), 149, 216 (ref. 13), 218 (ref. 13), 286 Goodloe, M. B., 202 (ref. 28), 223 (ref. 28), 286 Goralowna, C.,219 (ref. 21),$26
Gordon, A. H., 3 (ref. 103), 77, 88, 97, 109, 110, 136, 249, 160, 220 (ref. 23), 221 (ref. 23), 226 Gordon, 8.A., 267, 268, 298, SO4 Gordon, W. G., 223, $26, 228, 307 (ref. 2% 347 Gorter, E., 311 (ref. 8, 61), 346, 4148 Gortner, R. A., 257, 262, 298 von Gorup-Besanez, F., 278, $98 Gould, B. S., 220 (ref. 38),226 Gould, S. P., 314 (ref. 171),4162 Goyco, J. A., 298 Grbacher, C., 48 (ref. 192a),79 Grafe, K.,19,20 (ref. 57),36 (ref. 56),38 (ref. 56), 49 (ref. 56), 76 Graham, C. E., 108 (see Hier), 110 (see Hier), 111 (see Hier), 112 (see Hier), 113 (see Hier), 116 (see Hier), 118 (see Hier), 119 (see Hier), 120 (see Hier), 121,260 Graham, W. R., 221 (ref. 39),226 Graniek, S.,234, 267, 298 Grantham, J., 139, 2 4 9 Grassmann, W.,41 (ref. 193), 42 (ref. 193),43 (ref. 193),44 (ref. 193),67 (ref. 194), 68, 79, 360, 362, 363 (ref. 126), 364 (ref. 61), 366 (ref. 126),368 (ref. 126),369 (ref. 35, 50), 370,371 (ref. 35), 378 (ref. 66), 383 (ref. 66),385 (ref. 66), 386 (ref. 66), 388 (ref. 66), 389 (ref. el), 392 (ref. 61), 393 (ref. 61, 66), 398, 405 (ref. 46), 406 (ref. 46),415 (ref. 50), 416, 417, 418
Grau, C. R., 244, 296 Graves, G. D., 326,348 Green, D.E., 33 (ref. 182), 79, 220 (ref. 23), 221 (ref. 23),226, 273, 298 Greenberg, D. M., 128, 134, 249, 207, 286, 426
Greene, R. A., 250, 298 Greene, R. D., 92,95, 99, 102, 103, 105, 110, 111 (see Smith, E.L.), 112, 113 (see Smith, E. L.), 117, 118, 119, 120, 138, 260, 262, 212 (ref. 121123), 213, 214 (ref. 121), 218 (ref. 121-123), 828, 244, 263, 264, $98, 302 Greenstein, J. P., 6, 19, 38 (ref. 255b), 42 (ref. 197), 43 (ref. 197), 44 (ref.
449
AUTHOR INDEX
197), 51 (ref. 198a), 53 (ref. 196), 54, 59 (ref. 81), 61 (ref. 81), 62 (ref. Sl), 64 (ref. 196a), 65, 76, 79, 81,97, 105, 134, 139, 141, 160, 161, 255, 257, 266, 298, 309 (ref. 130), 311 (ref. 130), 315 (ref. 130), 316 (ref. 130), 3.49, 365, 417 Gregory, F. G., 283, 287, 289, 298 Grimaux, E., 2, 79 Grimmer, W., 222, 826 Grindley, H. S., 234, 242, 898 Granwall, A., 126, 135, 160,216, 217, 886 Groh, J., 205, 886 Gromyko, E., 294, 301 Groot, E. H., 240, 298 Grossfeld, I., 22 (ref. 223a), 80 Groves, M. L., 223 (ref. 148), 828 Griinert, H., 10 (ref. 207), 80 Gilnther, G., 271, 297 Guggenheim, M., 49 (ref. 14, 51 (ref. 13),76 Guggisberg, H., 207, 227 Guillermond, A., 266, 298 Guiltonneau, G., 221, 2.26 Guirard, B. M., 91, 108, 113, 118, 119, 120, 160,162 Gulick, A., 277, 298 Gumlich, O., 9 (ref. log), 77 Gunnese, M., 91, 103 (see Stokes), 106, 109, 110 (see Stokes), 111 (see Stokes), 112 (see Stokes), 113 (see Stokes), 116, 117 (see Stokes), 118 (see Stokes), 119 (see Stokes), 120 (see Stokes), 122, 160,162,163,248, 249, 250, 251, 252, 253, 254, 256, 258, 259, 303 Gunsalus, I. C., 273, 300 Curd, F. N., 189, 198,200 Gustavson, I(. H., 308 (ref. 63), 309 (ref. 63), 948 Gustavson, K. H., 361 (ref. 68, 71, 93), 362 (ref. 63), 363 (ref. 94), 364 (ref. 61, 74), 365 (ref. 82), 368 (ref. 71), 370 (ref. 71, 79, 82), 371 (ref. 71, 72, 75, 91), 372 (ref. 68, 71), 373 (ref. 55, 71), 374 (ref. 55, 71, 74), 375 (ref. 71, 74), 378 (ref. 66), 379 (ref. 77, 80, 81), 380 (ref. 66), 381 (ref. 194), 382 (ref. 77, 80, 81, 97), 383 (ref. 66, 68, @4,97), 384 (ref. 59, 65,
81), 385, 386 (ref. 66), 387 (ref. 56, 57, 87, 88), 388 (ref. 57, 66, 71, 84), 389 (ref. 61), 390 (ref. 54, 62), 391 (ref. 60, 68), 392 (ref. 61, 64,94), 393 (ref. 61, 65, 66, 68), 394 (ref. 70, 83), 397 (ref. 70), 398 (ref. 70, 85, 92, 94), 399 (ref. 55, 70, 71, 79, 83, 85), 400 (ref. 75), 401 (ref. 61), 402 (ref. 94), 403 (ref. 76, 78, 96), 404 (ref. 76, 78, 94, 95), 405 (ref. 67, 89, 90, 94), 406 (ref. 67), 407 (ref. 69, 73, 86), 408 (ref. 69, 73, 93), 409 (ref. 70, 73), 410 (ref. 67), 411 (ref. 67), 412 (ref. 93), 414 (ref. 68), 415 (ref. 79), 417,418,480 Gutfreund, H., 128, 132, 160 Guth, E., 335, 3.48 Gutmann, M., 330 (ref. 25), 335 (ref. 25), 347
H Haag, J. R., 236, 298 Haas, P., 45 (ref. 200), 79, 220 (ref. 44), 2.26 Haase, E., 40 (ref. 14), 76 Hac, L. R.,106, 108, 138, 160, 218 (ref. 45), 226,264, 298 Hadley, P., 92, I60 Hadorn, H., 406, 407, 410 (ref. 161), 419 Hafner, F. H., 244, 298 Hakala, M., 275, 304 Hale, F., 103 (see Lyman), 108 (see Lyman), 112 (see Kuiken), 113 (see Kuiken), 119 (see Lyman), 138 (see Lyman), 161,235, 244, 299 Hale, W. S., 222 (ref. 7),226 Hall, C. E.,960, 359 (ref. 98, 180), 418, 4.90 Hall, C.P., 400 (ref. 150), 419 Hall, D. H.,233, 897 Hall, L., 175 (see Zucker, L.), 179 (see Zucker, L.), 180 (see Zucker, L.), 800 Hall, W. K., 182, 197,198,199 Halpin, J. G.,193, 197 Halsey, G., 332 (ref. 39-41), 333, 335 (ref. 66), 947,344 360 Halsey, J. T., 16, 78 Halwer, M., 311 (ref. 132), 327, 360 Hamilton, T. S., 199,234, 242, 263, 298, 300
450
AUTHOR INDEX
Hammarsten, O., 202 (ref. 46),886 Hamoir, G.,140, 146,149 Hanby, W.E.,45 (ref. 201), 79,265,298 Handler, P.,18 (ref. 96),77 Hanke, M. T., 119,160 Hansen, R. G., 102, 103, 106, 108, 110, 111, 112, 113, 117, 118, 119, 120, 160, 213, 218 (ref. 48), 223 (ref. 47, 48), 826 Haneon, E. A., 234,267,898 Haneon, H. T., 36 (ref. 201a), 39 (ref. 201a),70 (ref. 201a),80 Harden, A., 222, 226 Harington, C. R., 3 (ref. 202), 31, 36 (ref. 203), 41 (ref. 203a), 42 (ref. 203),43 (ref. 202, 203), 44 (ref. 202, 203), 45 (ref. 203),47, 50 (ref. 204), 61, 52, 53 (ref. 203, 204), 54 (ref. 203,204),80, 100, 132, 133, I60 Harmon, K. M., 50 (ref. 240a),81 Harris, G., 18 (ref. 103a),77 Harris, J. I., 25, 49 (ref. 204a),80 Harris, L. J., 406 (ref. 13,99),416,418 Harris, M., 308 (ref. 69), 312 (ref. 155), 328, 329, 348, 360, 363 (ref. 190), 364 (ref. 191),420 Harris, R. H., 262,298 Harrison, H. C.,182, 188,198 Harrison, W.,329 (ref. 721,348 Harte, R. A., 176, 192, 198 Hartig, T.,260, 298 Hartree, E.F., 220,226 Hartung, W.H., 20 (ref. 205),80 Ham, E.,264 Havestadt, L., 11, 50 (ref. 206),80 Hawley, E. E., 170 (see Murlin), 171 (see Murlin), 198,199 Haydee, E., 182,197 Hayward, J. W., 244, 298 Hegediie, B.,43 (ref. 206a),45 (ref. 206a), 53 (ref. 206a), 80 Hegsted, D. M., 99, 111, 113, 160, 179, 198,246,9102 Heidelberger, M., 263, 299 Heiduschka, A., 220 (ref. 501, 226 Heilbron, I. M., 18 (ref. 103a), 77,270, d96 Heinsen, J., 198 Helferich, B., 10 (ref. 2071,80
Hellbach, R., 311 (ref. 138), 314 (ref. 138), 360 Hemingway, A., 285,304 Henderson, L.M.,102,103,106,108,110, 111, 112, 113, 116, 117, 118, 119, 120, 160, 218 (ref. 51), 826 Hendry, J. L., 149 Henze, E., 36 (ref. 260),81 Herbst, R. M., 20,80,81,276, 298 Hermann, K.,308 (ref. 57), 328 (ref. 57), 348 Herriott, R. M., 3 (ref. 212),80,125,126, 145, 160, 307 (ref. 73), 548 Hem, W.C.,72 (ref. 213),73,80,96,103, 105, 111, 117, 160, 163, 218 (ref. 129), 828,256,262, 298,300 Hetler, D. M., 250,298 Hevesy, G.,283, 298 Hewitt, F. O., 109, 110, 114, 116, 119, 149 Hier, 9. W., 108, 110, 111, 112, 113, 116, 118, 119,120, 121 (see Graham), 160 High, L. M., 311 (ref. 166),317 (ref. 166), 360 Highberger, J. H., 361 (ref. 101), 362, 363 (ref. loo), 367, 368, 372 (ref. 101), 376 (ref. 147), 377 (ref. 104), 406 (ref. 102), 407, 408 (ref. 101), 409, 410 (ref. 101), 411, 412, 418, 419,P O Hill, D. W., 208, 286 Hill, T.G., 45 (ref. 200),79,220 (ref. 44), 826
Hilpert, S., 411, 412, 418 Hipp, N.J., 223 (ref. 148),228,311 (ref. 138),314 (ref. 138),360 Hippius, A., 219 (ref. 52), 221, 226' Hird, F. J. R., 295 Hirsch, P.,37 (ref. 16), 38 (ref. 15, 16), 39 (ref. 15, 16), 76 Hirsch, R. R., 193 (see Russell), 199 Hirszowski, A., 38 (ref. 17), 51 (ref. 17), 76 von Hochstctter, H., 79 Hock, C. W., 182 (see Berg; Hall, W. R.), 297, 198 Hoffman, 0.D.,202 (ref. 8), 203 (ref. 8), 218 (ref. 8),226 Hoffman, W.F., 262, 298
451
AUTHOR INDEX
Hofmann, K., 43 (ref. 21b), 49 (ref. 215), 60 (ref. 214), 61 (ref. 214, 216), 64 (ref. 216), 80 Hofmeister, F., 2, 80, 373, 418 Hogeboom, G. H., 425 (ref. 28, 29), 431 (ref. 28, 49), 432 (ref. 28, 29), 436 (ref. 49), 438 (ref. 28), 4.99, GO Holland, H. C., 361 (ref. 107), 393 (ref. 169), 418, 419 Holman, R. L., 182, 185, 186, 198 Holt, L. E., Jr., 171 (see Cox), 198 Holter, H., 209, 226 Hooke, R., 310, 348 Hoover, S. R., 218 (ref. 88), 227, 274, 296
Hopkins, F. G., 51 (ref. 218), 80, 220, 827
Horn, M. J., 110, 120, 151, 263, 298 Homer, C. K., 274, 298 Horowite, N. I€., 273, 302 Hotchkiss, R. D., 3 (ref. 219), 80, 217 (ref. 75), 226, 280, 299, 425 (ref. 28), 431 (ref. 28), 432 (ref. 28), 438 (ref. 28), 4.39 Houston, J., 159 (see Kade), 198 Howe, E. E., 166, 199 Howe, P. E., 202, 211, 226, 227 Howitt, F. O., 311 (ref. 32, 33), 324, 325, 3447,348
Huddleson, I. F., 223 (ref. 110), 227 Hudson, D. P., 270, 296 Huffman, H. M., 33 (ref. 92), 77 Huggins, M. L., 308 (ref. 76), 328 (ref. 76), 348, [360, 361 (ref. log), 401, 418 Hughes, A. H., 311 (ref. 77), 348 Hull, R., 217, 225 Hulme, A. C . , 241, 286, 299 Hummel, F. C . , 105 (see Bectch), 148 Humphreys, E. M., 190, 197 Humphreys, F. E., 394 (ref. 30, 110), 416, 418
Hunt, M., 11, 22 (ref. 220), 36 (ref. 220), 40 (ref. 220), 42 (ref. 288), 44 (ref. 288), 55 (ref. 292), 61 (ref. 220, 221), 62 (ref. 220, 221), 65 (ref. 220, 221, 288), 80, 82 Hurwits, S. II., 182, 198 Hutchinson, J. C . D., 240, 299
I Ing, H. R., 32, 80 Irving, G. W., Jr., 34 (ref. 190), 7'9,262, 298, 899
Ivanovics, G., 45 (ref. 222), 80
J Jackson, R. W.,307 (ref. 22), 311 (ref. 110, 138), 314 (ref. 138), 347, 34.9, 360 Jacobs, W. A., 57, 78 Jacobsen, C. F., 2, 53 (ref. 222a), 80, 81, 100, 126, 132, 133, 135, 161, 217 (ref. 59), 218, 226 Jakus, M. A., 360, 359 (ref. 98, 180), 418, 420
Jameson, E., 212, 226 Jander, G., 381 (ref. l l l ) , 418 Janisch, R., 260, 299 Jansen, E. F., 342 (ref. 78),348 Jarrouse, H., 221 (ref. 43), 826 Jeannerat, J., 311 (ref. 122), 34.9, 369 (ref. 21), 416 Jensen, H., 105 (see du Vigneaud), 163 Jirgensons, B., 316 (ref. 79), 348 Johansen, D., 280, 299 Johansen, G., 217 (ref. 75), 226 Johanson, R., 244, 252, 299 Johns, C. O., 262, 299 Johnson, J. R., 18 (ref. lOOa), 75 (ref. lOOa), 77 Johnson, M., 262, 298 Johnson, R. M., 171, 198 Johnson, T. B., 250, 299 Johnson, W. A., 285, 299 Jones, D. B., 110 (see Horn), 120 (see Horn), 161, 244, 247, 262, 263, 297, 298, 299
Jones, M. J., 273, 299 Jordan Lloyd, D., 355, 418 Jovanovits, J. A., 362, 418 Just, F., 250, 298
K Kabat, E. A., 263, 299 Kade, C. F., Jr., 159, 198 Kamen, M. D., 277, 302, 427 (ref. 53), 440
Kann, E., 18 (ref. 58), 76
452
AUTHOR INDEX
Kapfhammer, J., 7 (ref. 223),80 Kardos, E.,205 (ref. 42), 926 Kariher, D.H., 183,199 Karrer, E.,335,348 Kassell, B.,87, 96, 103, 105, 109, 110, 116, 117, 118, 120 (see Brand), 135, 149, 161, 216 (ref. 13), 218 (ref. 13, 61),226, 226 Katchalski, E., 7, 22, 79, 80, 307 (ref. 531,347 Kattus, A. A., Jr., 183 (see Madden), 198 Katz, J. R., 361 (ref. 115, 116),418 Katz, S.,348 Kautzsch, K., 5 (ref. 153), 78 Kay, H. D., 208, 221 (ref. 39, 62), 226, 227 Kaye, M. A. G., 409 (ref. 217),421 Keilin, D.,220, 226, 275, 299 Kekwick, R. A., 135, 138,161, 214 Kelly, M. W.,391 (ref. 206), 396 (ref. 205),411, 42f Kelly, P. L., 221 (ref. 64, 65),226 Kemm, E.,8 (ref. 296), 48 (ref. 296), 88 Kemmerer, K. S., 171 (see Cox), 198 Kemp, A. R., 241,299 Kempe, M.,11, 71 (ref. 18), 72,76 Kennaway, E.I,., 63 (ref. 225),80 Kennedy, E. P., 425 (ref. 30),431, 439 Kenten, R. H., 143, 149, 357, 414, 416, 481
Kenyon, A. E., 270, 303 Kern, E.J., 384 (ref. 214), 491 Kerr, W.J., 182, 198 Keston, A. S., 93,94, 101, 114, 161, 218 (ref. 66),996,283, 298 Kibrick, A. C., 106, 108, 135 (see Cannan), 136 (see Cannan), 138, 149, f61,214 (ref. 17),216 (ref. 17),226 Kidd, D. E., 32, 45 (ref. 223b), 80 Kiesel, A., 240,299 King, F.E., 32,45 (ref. 223b), 80 King, H., 373 (ref. l), 416 Kinsman, G. M., 163,197 Kinzer, R.,383, 418 Kirkwood, J. G., 341 (ref. 82),348, 360 Klein, D., 108 (see Hier), 110 (see Hier), 111 (see Hier), 112 (see Hier), 118 (see Hier), 116 (see Hier), 118 (see Hier), 119 (see Hier), 120 (see Hier), 121 (see Graham), 160
Klein, G., 280, 299 Klemm, O., 307 (ref. 123), 308 (ref. 123), 349
Klemperer, F. W.,42 (ref. 197), 43 (ref. 197), 44 (ref. 197),65, 79 Klose, A. A., 249, 299 Klotz, I. M., 207, 226 Knight, C. A., 101, 102, 103, 105, 106, 108, 109, 110, 111, 112, 113, 114, 116, 117, 118, 119, 120, f61, 255, 256, 259,299 Kodama, S., 204, 227 Kogl, F., 441 (ref. 224),53, 80, 253, 999 KBhler, E.,260, 299 Kohler, F.,33 (ref. 40),73, 76 Koelker, A. H., 39 (ref. 154), 78 Koenigs, E.,41, 43 (ref. 155), 44 (ref. 155, 166),78 Koppel, W., 54 (ref. 21), 69 (ref. 20), 76 Koessler, K. K., 119, 160 Koster, H., 5 (ref. 82), 16 (ref. 60), 49 (ref. 60),61, 63, 76' Kollen, S.,436 (ref. 57),440 Kolyakova, G.E., 420 Komm, E., 220 (ref. 50),226 Koning, C. J., 219, 226 Kooper, W.D., 222 (ref. 69),226 Koorn, V. M., 311 (ref. loo), 316 (ref. loo), 336 (ref. loo), 345 (ref. loo),
349
Korn, A. H., 218 (ref. 88),227 Kossel, A,, 63 (ref. 225, 226),80 Kosterlitz, H. W., 182, 187, 197, 198 Kraemer, E. O., 316 (ref. 83), 348, 367, 418
Kraft, W. M., 80 Kratky, O., 306 (ref. 84),348 Kratzer, F. H., 244,296 Krauel, K., 159 (see Kade), 198 Krauss, B. H., 271,302 Krebs, H. A., 285, 299 Kritzinger, C.C., 377, 418 Kritzman, 20 Kritzmann, M. G.,271, 273, 296 Kropp, W.,43 (ref. 157),44 (ref. 157),78 Kuchel, R. H., 281,304 Klihne, E.,369 (ref. 35), 371 (ref. 35), 416
KUntzel, A., 360, 362 (ref. 127, 129),363 (ref. 126), 366 (ref. 126), 368, 369
AUTHOR INDEX
(ref. 120, 121, 123), 372 (ref. 123), 375 (ref. 123), 377, 380 (ref. 122), 383, 386 (ref. 130), 390, 393 (ref. 131), 403, 411, 418, 419 Kuhlmann, A. G., 262, 299 Kuhn, R., 96, 103, 105, 139, 161 Kuhn, W., 321 (ref. 85), 348, 369, 418 Kuiken, K. A., 93, 108 (aee Lyman), 112, 113, 115, 119 (see Lyman), 138 (see Lyman), 161, 235, 244, 299 Kuk, S.,67 (ref. 177), 79 Kulka, J. P., 181 (see Zeldis), 200 Kunitz, M., 126, 145, 161, 162, 263, 264, 299 Kuttner, A., 211, 226
L Laidler, K. J., 333 (ref. 59), 348 Laine, T., 273, 275, 304 Lamanna, C., 124 (see Buehler), 125 (see Buehler), 143 (see Buehler),
1.69 Lambotte, E., 9 (ref. 110), 77 Lampen, J. O., 273, 299 Lampman, C. E., 244, 301 Lams, M., 412 (ref. 198), 420 Landolt, H., 48 (ref. 192a), 79 Landsteiner, K., 424 (ref. 31), 439 Lane-Claypon, J. E., 222, 226 Larionov, L. P., 429 (ref. 32), 439 Larkin, J. B., 15 (ref. 267), 39 (ref. 267), 81
Larsmn, A., 404 (ref. 95), 418 Latimer, W. M., 393, 419 Laufer, M. A., 162 Lavine, T. F., 103, 161 Leach, M. F., 250, 299 Leaderman, H., 331 (ref. 86), 332 (ref. 861, 348
Leavenworth, C. S., 233, 235, 273, 281, 284, 288, Sol,303 Leeper, G. W., 270, 299 Lehmann, W., 209, 227 Lehninger, A. L., 425 (ref. 30), 431, @9 Lehoult, Y., 268, 303 Leinert, F., 36 (ref. 89), 61 (ref. 22), 66 (ref. 22), 68 (ref. 89), 69 (ref. 89), 70 (ref. 89), 76, 77 Leloir, L. F., 273, 298
453
Lemoigne, M., 270, 299 Leplat, G., 371 (ref. 134), 419 Le Quesne, W. J., 43 (ref. 227a), 44 (ref. 227a), 45 (ref. 227a), 80 Leuchs, H., 21, 36 (ref. 228), 37 (ref. 229), 49 (ref. 229), 80, 307 (ref. 87, 881, 348 Leutert, F., 24, 79 Leuthardt, F. M., 19 (ref. 198), 51 (ref. 198a), 79, 257, 298, 373 (ref. 135), 419
Levene, P. A., 5, 18 (ref. 236), 36 (ref. 235), 38 (ref. 231, 233, 234, 237), 39 (ref. 234, 237), 67 (ref. 232), 80, 208, 226
Levitt, J., 241, 299 LCVY, L., 9, 77 LCW,W., 182, 187, 196 Lewis, J. C., 108, 138, 161 Lewis, J. H., 211, 226 Ley, H., 390, 419 Li, C. H., 126, 127, 135, 161, 216, 226 Li, Tsan-Wen, 182, 198 Lilienfeld, L., 2, 80 Linderstrgim-Lang, K., 2, 81, 89, 100, 132, 133, 161, 204, 205, 209, 217 (ref. 75), 226, 227, 276, 280, 283, 898, 299, 376, 419 Linkola, H., 275, 304 Lintcel, W., 240, 299 Lipmann, F., 208, 227,'286, 299 Lipschitz, E., 36 (ref. 158), 49 (ref. 158), 50 (ref. 158), 78 Little, R. B., 211, 228 Lloyd, D. J., 335 (ref. 89), 348 Loeb, J., 366, 367 (ref. 139), 419 Lowe, H., 406 (ref. 48), 416 Logan, M. A., 109, 149 Long, C. N. H., 182, 188, 198 Longsworth, L. G., 126, 138, 161 Loo, Y. H., 89, 1.69 Loring, H. S., 44 (ref. 289), 52 (ref. 240, 286), 53 (ref. 240), 54, 62 (ref. 240), 81, 82 Lowe, B., 306 (ref. go), 3.48 Lowndes, J., 218 (ref. 102), 227 Luck, J. M., 217 (ref. 6), 226 Ludewig, S., 207 (ref. 18), 226 Lugg, J. W. H., 98, 103, 105, 116, 117, 119, 121, 161, 218 (ref. 78), 227, 233,
454
AUTHOR INDEX
234, 236, 236, 237, 238, 239, 240, 243, 244, 245, 240, 247, 248, 265, 268, 267, 258, 279, 293, 294, 298, 299,900 Lum, F. G., 217 (ref. 6),926 Lundgren, H. P., 306 (ref. 99, 101), 311 (ref. 96, 96, 98, 100, 166), 313 (ref. 96), 315 (ref. 97, lOl), 316 (ref. 91, 92, loo), 317 (ref. 93-96, 166), 323 (ref. 96), 336 (ref. 100), 346 (ref. 100, 101),349,360 Luniak, A., 3, 6, 11, 68,70, 78 Lyman, C. M., 93, 103, 108, 112 (see Kuiken), 113 (see Kuiken), 115, 119, 138, 161, 236, 244, 299
M Maack, J. E., 173, 174 (see Barnes), 176 (see Barnes), 197 MacAllister, R. V., 50 (ref. 240a), 81 MacArthur, I., 306 (ref. 102, 103),349 McCalla, A. G., 262,302 McCandlish, D.,369 (ref. 16), 372 (ref. 16), 400,416 MacCardle, R. C., 435 (ref. 33),439 McCarty, M., 424 (ref. 3, 34), 427 (ref. 3,341, 439 McClintock, B., 427, 440 McCoord, A. B., 181 (see Zeldis), 600 McDonald, M. R., 264,299 MacFarlane, A. S . , 135,161 McGavack, T.H., 171,199 McGinnis, J., 193, 198 McGowan, J. C., 235,30% Macheboeuf, M., 435 (ref. 36), 440 MacInnes, D.A,, 126 (see Longsworth), 138, 161 McIntyre, J. M., 168 McKinney, H. H., 292,293,SO0 McLaughlin, G. D., 372 (ref. 148), 376, 387, 392 (ref. 148),419 MeLennan, E.,295 MacLeod, C. M., 424 (ref. 3), 427 (ref. 3), 439 McMahan, J. R., 113, 118,161 McMeekin, T. L., 126, 127, 135, 161, 212 (ref. 80), 215, 217 (ref. 79, 81, 82), 223 (ref. 148), 867, 698, 311 (ref. 109, 110),325,349
McNaught, J. B., 183, 198 MacPherson, H. T., 88, 89, 90, 95, 98, 104, 110, 118, 119, 120, 122, 130, 161, 218 (ref. 84),227 Macrae, T. F., 240, 699 Macy, I. G., 105 (see Beach), 148, 202 (ref. 8),203 (ref. 81,218 (ref. 8), 926 Madden, 8. C., 183, 187,198 Miirkert, L., 13 (ref. 180), 79 Magnant, C.,434 (ref. 25), 439 Mahadevan, S., 280, 297 Mahdihassan, S., 265, d98 Mahoney, E. B., 182, 185, 187 (see Holman), 198 Maier, J., 24, 77 Manasse, W.,21 (ref. 230), 80,307 (ref. 881, w Manske, R. H. F., 32, 80 Marais, J. S . C., 236,SO0 March, M. E., 172 (see Murlin), 174 (see Murlin), 199 Marenzi, 247 Mark, H., 307, 308 (ref. 55, 104, 105), 312 (ref. 106), 313, 328 (ref. 106), 330, 332 (ref. 106), 335 (ref. 64), 348, 349, 359 (ref. 156))419 Marker, R. E., 5 (ref. 237),38 (ref. 237), 39 (ref. 237),80 Markley, A. L., 221,667 Marriott, It. H., 355 (ref. 112),360, 369 (ref. 141), 375, 376 (ref. 140, 143), 418, 419 Martell, A. E., 81 Martin, A. J. P., 3 (ref. 103, 242), 77, 81, 84,88,94,97,110, 123, 136,149, 160, 161, 237, 300,360,419 Martin, L. F., 292, 293,SO0 Maschke, O.,260,300 Maschrnann, E.,40 (ref. 243),81 Masket, A. V., 207 (ref. 18),866 Mason, T.C., 284,300,SO1 Massart, L., 221, 267 Matthes, K.,7 (ref. 223),80 Mauersberger, H.R., 311 (ref. 107),314 (ref. 107),349 Max, J., 11 (ref. 244), 36 (ref. 244), 37 (ref. 244),81 Maxwell, J. C., 332,349 Maycr, J., 8 (ref. 296), 48 (ref. 296),8.2 Mazur, A., 246, 247, 248, 249, 257, 300
465
AUTHOR INDEX
Mead, T. H., 3 (ref. 202), 31, 36 (ref. 203), 42 (ref. 203), 43 (ref. 202, 203), 44 (ref. 202, 203), 45 (ref. 203),
Miller, G. L., 43 (ref. 248), 44 (ref. 248, 289, 291), 53, 54 (ref. 289), 81, 89,
51, 80, 100, 132, 160 Mecchi, E., 244, 296 Mecheels, O., 311 (ref. 112, 113), 949 Mehl, J. W., 171, 198 Mehrhof, T. G., 193 (see Russell), 199 Melchers, G., 255, 300 Mellander, O., 204, 205, 206, 208, 218 (ref. 87), 219 (ref. 87), 227 Mellon, E. F., 218 (ref. 88), 227 Melnick, D., 158, 159, 174, 185, 186, 193, 198 Melville, D. B., 18 (ref. 96), 77 Melville, J., 44 (ref. 245), 45, 81 Mendel, L. B., 175, 193, 199 Menefee, S. G., 202 (ref. 89), 203 (ref. 89), 227 Menke, W., 234, 239, 267, 900 Mercadante, M., 278, 300 Mercer, E. H., 330, 349 Mercer, F. V., 281, 304 Meridith, R. J., 331 (ref. 115), 349 Merkel, R., 67, 70 (ref. 231, 76 Merrifield, A. L., 311 (ref. 116), 315 (ref. 1161, 349 Merrill, H. B., 375, 387 (ref. 151), 400, 419 Meunier, L., 411, 412 (ref. 152), 419 Meyer, C. E., 25 (ref. 2901, 82 Meyer, F., 205, 226 Meyer, H., 10, 81 Meyer, K. H., 307, 308 (ref. 123, 124), 309 (ref. 124), 311 (ref. 122), 335, 349, 359 (ref. 156), 369, 378, 406 (ref. 153), 410, 411, 416, 419 Mez, 257 Mezey, E., 400 (ref. 20), 416' Michael, G., 281, 300 Michaelis, L., 376 (ref. 157), 419 Miekeley, A., 17 (ref. 65), 18 (ref. 58), 49 (ref. 62), 51 (ref. 65), 55, 57 (ref. 61, 65), 76 Miescher, K., 75 (ref. 252, 294), 81, 82 Migliardi, C., 68 (ref. 247), 70 (ref. 247), 81 Millar, A., 310, 311, 349 Miller, E. J., 233, 234, 238, 241, 29Y, 300 Miller, G., 265, 300
Miller, L. L., 166 (see Whipple, G. H.), 171, 183, 184, 185, 189, 194, 198, 199, 200 Milne, 25 Minard, F. N., 424 (ref. 27), 439 Mirsky, A. E., 139, 141, 148, 161, 308 (ref. 126), 311 (ref. 127, 140), 316 (ref. 5), 346, 349, 360, 369, 419, 426,
105, 161
427,
.bdo
Mitchell, H. H., 156, 158, 162, 163, 164, 165, 169, 171, 173, 174, 176, 192, 193, 194, 197, 199, 236, 263,-300 Miaell, L. R., 331 (ref. 71), 348 Moeller, O., 247, ,999 Mohr, E., 15, 18, 33 (ref. 249), 81 Moir, F., 252, 299 Moir, R. J., 244, SO3 Monguillon, P., 270, 999 Moore, D. H., 216, 226 Moore, E. K., 376 (ref. 147), 419 Moore, S., 3, 36 (ref. 278), 38 (ref. 278), 67 (ref. 279), 81, 88, 95, 101, 102, 111, 112, 113, 116, 122, 161 Morehouse, M. G., 171, 198 Morel, M., 290, 900 Morgan, E. J., 220, 227 Morneweg, W . , 46 (ref. 2), 50 (ref. 2), 61 (ref. 2), 76 Moro, E., 219 (ref. 91), 221, 227 Morris, H. J., 274, 296 Morris, M., 223 (ref. 147), 228 Moseley, O., 103 (see Lyman), 161 Mothes, K., 272, 283, 284, 300 Mourgue, M., 264,901 Mowat, J. H., 45 (ref. 250a), 81 Moxon, A. L., 272, 303 Moyer, L. S., 267, 998, 300 Mueller, A. J., 159, 171 (see Cox), 198 Murlin, J. R., 170,171,172,174,198,199
rl Nageotte, J., 371 (ref. 159), 419, 434, 4-40
Nakamura, F. L., 176 (see Beadles), 197 Naeset, E. S., 172 (see Murlin), 174 (see Murlin), 199
456
AUTHOR INDEX
Olcott, H. S., 96, 108, 119, 121, 129, 133, 135, 138, 145, 161, 162, 307 (ref. 49-52, 134), 347, 360, 407 (ref. 42), 409 (ref. 40, 41), 410, 416 Olofsson, B., 391 Olsen, C., 283, 298 Olsen, R. T . , 198 Omachi, A., 432 (ref. 41), 440 Oncley, J. I,., 128, 132, 144, 217 (ref. 34), 226 Onslow, M. W., 263, 300 Oparin, A. I., 289, SO0 Orcutt, M. L., 211, 887 van Ormondt, J., 311 (ref. 8), 346 Orten, A. U., 184, 199 Orten, J. M., 184, 199 Osborne, T. B., 3, 67 (ref. 253), 68, 70, 81, 134, 162, 175, 193, 199, 202, 205, 206 (ref. 97), 223, 897, 298, 233, 235, 260, 261, 263, 264, 296, 300, 301 Oser, B. L., 193, 198 416 Nienburg, H., 68, 69 (ref. 25), 70 (ref. Ott, E., 311 (ref. 135, 136), 324 (ref. 136), 360 25), 76 Otto, E., 11, 21, 36 (ref. 160), ‘78 Nier, A. O., 285, 304 Nitschmann, H., 207, 209, 867,406, 407, Otto, G., 364, 385 (ref. 164), 387 (ref. 164), 419 410, 419 Overhoff, J., 41 (ref. 203a), 80 Niven, C. F., 182, 199 Overman, 0. R., 202 (ref. 89), 203 (ref. Noack, K., 239, 300 89), 287 Nocito, V., 273, 298 Noguti, Z., 311 (ref. 60), 315 (ref. 60), Owen, E. C., 236, 301 348 P Norman, W. H., 112 (see Kuiken), 113 (see Kuiken), 161, 235, 244, g99 Pacsu, E., 7, 26 (ref. 254), 81 Northrop, J. H., 126, 145, 161, 168 Nutting, G. C., 306 (ref. 133), 311 (ref. Paech, K., 285, 289, 301 131, 132, 150), 313 (ref. 151), 325 Page, It. O., 360 (ref. 1681, 393 (ref. 169), 399 (ref. 168), 401, 419 (ref. 131, 150), 327, 328 (ref. 151), 349, 360, 359 (ref. 162), 360 (ref. Pallade, G. E., 425 (rcf. 29), 432 (ref. 29), 162), 419 439 Palmer, A. H., 127, 135, 136 (see Can0 nan), 138 (see Cannan), 149, 162, 214, 215, 216, 217, 826, 28’7, 298 O’Connell, R. A., 311 (ref. 95, loo), 316 Palmer, K. J., 323, 328 (ref. 137), 360 (ref. loo), 317 (ref. 95, 96), 336 (ref. Palmer, L. S., 241, 304 Palmes, E. D., 256, 298 loo), 345 (ref. loo), 349 O’Doherty, K., 197 Pankhurst, K. G . A., 370, 371 (ref. 170), 415 (ref. 170), 490 Oesterling, M. J., 158 (see Borman), 197 Ogston, A. H., 128, 160, 168, 360 (ref. Pape, N. R., 328 (ref. 18, 56), 346, $48 Parker, E. D., 311 (ref. 26), 315 (ref. 26), 163), 419 O’Kane, D. E., 273, 300 347
Neale, S. M., 391, 392, 419 Nedvidek, R. D., 241, 30.2 Needham, J., 266, 300 Negelein, E., 264, 271, 304 Neglia, F. J . , 262, 300 Neher, R., 75 (ref. 251, 252), 81 Nelson, R. E., 120, 163 Neuberger, A., 13 (ref. 252a), 81, 89, 133, 139, 160, 161 Neumann, A., 36 (ref. 24), 38 (ref. 24), 76 Neurath, H., 97, 144, 161, 309 (ref. 130), 311 (ref. 130), 315 (ref. 130), 316 (ref. 128, 130), 330, 349 Nevens, W. B., 234, 242, 298 Neville, H. E., 89, 149 Newcomer, E. H., 266, 900 Newell, G. W., 199 Nicolet, B. H., 88, 89, 109, 110, 115, 161, 168, 208, 227 Niemann, C., 50 (ref. 240a), 81, 148, 360,
AUTHOR INDEX
Parsons, H. T., 245,297 Pssteur, L.,424,440 Patterson, W.I., 55 (ref. 292),82 Paul, W.,197 Pauling, L., 308 (ref. 126),349,369,379, (ref. 171),419, 420 Pavlova, M., 240,299 Payne, D.S., 244,301 Pearsall, W.H., 288,301 Pedersen, K. O.,134, 135,162,163,202, 207, 214, 215, 216, 222, 2.27, 238, 263,303 Pedlow, C., 281,SO4 Peakett, G. R., 223,227 Peters, R. H., 391 (ref. 160),419 Peterson, C. F., 244, 301 Peterson, R. F., 311 (ref. 138), 314,360 Petrie, A. H. K., 271, 276,277, 284,285, 286,289,301,304 Petrik, F. G.,268,301 Pfaltz, M. H., 36 (ref. 235),38 (ref. 233, 234), 39 (ref. 234),80 Pfeffer, W.,278, 283, 285,301 Pfeiffer, P.,373,390, 397,420 Pfeiffer, S. E., 249,303 Phillips, G.E., 249,297 Phillips, H., 376, 416 Phillips, J., 362 (ref. 129),377 (ref. 129), 418 Phillips, P. H., 102 (see Hansen), 103 (see Hansen), 106 (see Hansen), 108 (see Hansen), 110 (see Hansen), 111 (see Hansen), 112 (see Hansen), 113 (see Hansen), 117 (see Hansen), 118 (see Hansen), 119 (see Hansen), 120 (see Hansen), 160, 213 (ref. 48), 218 (ref. 48), 223 (ref. 47,48),826 Phillis, E.,284, 300,301 Pickels, E. G.,267, 302, 432 (ref. 20), 439
Picken, L., 335 (ref. 121a),349 Pinner, 9. H., 306 (ref. 139), 360 Piria, R.,277,301 Pirie, N. W., 50, 53 (ref. 255), 54 (ref. 255),81,125 127,166,235,254,255, 293,296,301 Pitt Rivers, R. V., 47, 50 (ref. 204), 51 (ref. 204), 53 (ref. 204), 54 (ref. 204),80 Pleass, W. B., 406 (ref. 15), 416
457
Plimmer, R. H. A., 218 (ref. 102), 227 Plowe, J. Q., 560 Policard, A., 435 (ref. 43), 4 0 Polis, B. D., 126 (see McMeekin), 127 (see McMeekin), 136 (see McMeekm), 161, 217 (ref. 81, 82), 227 Polis, E., 204 (ref. 141), 208, 222, 2d8 Pollister, A. W.,311 (ref. 127, 140), 349, 360, 426 (ref. 38), 440 Pollok, H.,61 (ref. 55),64 (ref. 55), 76 Pomes, A. F., 311 (ref. 116), 315 (ref. 1161,549 Pommerenke, W.T., 183, 199 Poo, L. J., 182, 187, 196 Porter, C. W., 393,419 Porter, K. R., 432 (ref. 20, 44), 433 (ref. 441,439,440 Porter, R. R., 98,99, 113, 129, 134, 135, 136, 138, 139, 141, 147, 148,162 Posternak, S.,208,267 Potter, J. S., 426,4.99 Potter, R. L., 102 Tsee Hansen), 103 (see Hanaen), 106 (see Hansen), 108 (see Hansen), 110 (see Hansen), 111 (see Hansen), 112 (see Hansen), 113 (see Hansen), 117 (see Hansen), 118 (see Hansen), 119 (see Hansen), 120 (see Hansen), 160, 213 (ref. 48), 218 (ref. 48), 223 (ref. 48), 226' Powell, R. E., 332 (ref. 161), 335 (ref. 161), 560 Prantl, 231 Prelog, V., 59,62 (ref. 255a),81 Preston, C.,287, 292,302,303 Prianishnikov, D., 278, 279, 280, 289, 301
Prianishnikov, D. N., 281,301 Price, V. E., 19, 38 (ref. 255b), 51 (ref. 198a),79, 81 Price, W. C., 255,300, 301 Pringsheim, E. G.,268,301 F'rocter, H. R.,366,420 Pryor, M.G. M., 412, 0 0 Pucher, G.W., 41,82,273,279,281,282, 284,288,303 Pund, E. R., 182 (see Berg), 197 Putnam, F. W., 97 (see Neurath), 161, 309 (ref. 130), 311 (ref. la), 315 (ref. 130), 316 (ref. 130), 317, 349
458
AUTHOR INDEK
Rieaen, W. H., 103, 166, 194, ls’g Riess, c., 381 (ref. 175), 386 (ref. la), 390 (ref. 130), 393 (ref. 131), 418, Quackenbush, F. W., 96, 103 (see Kuhn), 105 (see Kuhn), 139 (see 40 Rietz, E., 342 (ref. 78),348 Kuhn), 161 Quisenberry, J. H., 176 (see Beadles), 197 Riley, D., 128,149 Rimington, C., 208, 227 R Rinke, H., 26 (ref. 85),36 (ref. 258), 47 (ref. 85),49 (ref. 85),59 (ref. a), 61 (ref. 83, 84), 62 (ref. 83, 84, 258), Raistrick, H., 202, 210, 212, 223 (ref. 63 (ref. 83), 74 (ref. 85), 78, 77, 81 25), 926 Ris, H.,426 (ref. 39), 440 Ramachandran, B. V., 208, 226 Rischkov, V.L., 294,301 Ramamurti, T.K., 287,SO2 Riaser, W.C.,18 (ref. 97),77, 159, 199 Ramaawamy, R., 280, 297 Rittenberg, D.,90, 101, 162, 156, 199, Ramdas, K., 280,297 282,239,301,303 Ramsay, H., 64 (ref. 25ci),81 Ramaden, W.,311 (ref. 141), 341 (ref. Ritter, A., 416 Rivers, T. M., 255 (see Stanley, W. M.), 141),360 256 (see Stanley, W.M.), 302 Ranefeld, A. N., 91,162 Roberta, E. J., 311 (ref. 26), 315 (ref. 26), Raske, K., 38 (ref. 161),78 Ratner, B., 211, 226 347 Robinow, C. F., 268, SO1 Ratner, S.,75 (ref. 257),81 Robinson, R., 18 (ref. lOOa), 75 (ref. Ratti, R., 36 (ref. 260), 81 100a), 77 Rautanen, N., 273,275,301,SO4 Robscheit-Robbins, F. S., 156 (see Ravdin, I. S., 189, 198 Whipple, G.H.), 183, 184, 185, 189, Rees, A. L. G.,330,349 194, 198,199,800 Reea, M. W.,87 (see Chibnall), 88, 89, 95,96,97, 100, 104, 106 (see Bailey; Roche, J., 264,SO1 Chibnall), 108 (see Bailey; Chib- Rockland, L. B., 15 (ref. 267), 39 (ref. 267), 81, 85, 91, 92, 100, 101, 103, nall), 109, 110, 114 (see Chibnall), 108 (see Dunn), 111 (see Dunn) , 115, 118 (see Chibnall), 119 (see 119, 120 (see Dunn), f@, 160 Chibnall), 120 (see Chibnall), 121, 122, 130, 132, 136, 138 (see Chib- Roddy, W. T., 400 (ref. 176),&!O nall), 139, 140, 148, l@, 168, 218 Rodkey, F.L., 221 (ref. 108), 927 Rodney, G.,194, 198 (ref. 3, 104),,826,227,253, 897 Roepke, R. R., 271,273, 899 Reeves, E. B., 186 (see Weech), 900 Rehner, J., Jr., 328 (ref. 461, 338 (ref. Roesner, H.,55, 57 (ref. 163),78 Ronzoni, E.,101, 102, 103, 105,106, 108, 461,347 109, 110, 111, 112, 113, 116, 117, Reid, K., 222 (ref. 105),287 118, 119, 120, 130, 140, 147, 163 Reid, T.S., 311 (ref. 1101,349 Reif, G.,14 (ref. l62), 67 (ref. lSZ), 68 Rose, W. C., 3 (ref. 299), 88, 85, 156, 158,185,194,197,199,$00 (ref. 162),69 (ref. 162), 78 Ross, A. F.,102, 109, 110, 111, 116, 117, Retzch, C. E., 406 (ref. 102),418 118, 169,256, 301 Rhoades, M.M., 433 (ref. 45), 440 Ross, W. F., 31 (ref. 41,86), 47, 50 (ref. Rice, E.E., 158, 194,199 41), 59 (ref. 86), 60 (ref. 63,86), 61 Rice, F. E., 221, 927 (ref. 86), 62 (ref. 86), 76, 77 Rich, C. E., 263,301 Rowenbeck, H., 265,998 Richards, F.J., 286,287, SO1 Rossner, E.,40 (ref. 26), 76 Rideal, E. K., 311 (ref. 77), 348 RothBs, F.,263,301 Riederle, K.,360,417
Q
AUTHOR INDm
Rothemund, K. W.,25, 81 Rothen, A., 277, 290, 301, 435 (ref. 22), 439
Rouelle, M., 233,234,301 Rowland, S.J., 202 (ref. log), 227 Rudall, K. M., 376, 377, 412, 413 (ref. 177), 415, 416, 420 Ruggli, P., 36 (ref. 260), 81 Ruhland, W.,279,901 Russell, W. C., 193,199 Ryan, F. J., 100 (see Brand), 102, 110, 112, 116 (see Brand), 117, 120 (see Brand), 149, 162, 216 (ref. 13), 218 (ref. 13),226 Rydon, H. N., 45 (ref. 201),79, 265, 698
S Sahyun, M., 159 (see Kade), 198 Saidel, L. J., 101, 102, 109, 112, 116 (see 216 Brand), 117, 119, 120, f@, (ref. 13),218 (ref. 13),226 St. John, J. L., 193,198 Saito, M.,40 (ref. 27), 76 Salcedo, I. S., 407, 411 (ref. 178), 418, @O Salle, A. J., 250,251, 252,253,254, 297 Salsmann, L., 42 (ref. 87, 88), 43 (ref. 87,88),44 (ref. 88), 45 (ref. 87,88), 47 (ref. 88),50 (ref. 88), 51 (ref. 88), 77
San Clemente, C. L., 223 (ref. 110),6.97 Sandstedt, R. M., 262, 296, 901 Sanger, F., 98, 99, 101, 113, 129, 133, 134, 135, 136, 139, 141, 146, 147, 148,16.2 Sanni6, C., 5 (ref. 261), 36 (ref. 261), 81 Sarich, P.,176, 198 Scarth, G. W., 306 (ref. 142),960 Scatchard, G.,360 Schaal, E., 2, 81 Schachowskoy, T.,387 (ref. 32),416 Schanderl, H.,274, 275, 302 Schardinger, F.,220, 927 Schauts, E. J., 124 (see Buehler), 125 (see Buehler), 143 (see Buehler),
149 Scheele, W., 381 (ref. lll),418 Scheibler, H.,7 (ref. 164),37 (ref. 164),38 (ref. 164),39 (ref. 164), 79 Sohein, A. H., 264, 306
469
Schels, H., 398 (ref. 51), 417 Schenck, J. R., 199 Schenk, R., 219 (ref. 112),228 Schiff, H.,2, 81 Schlag, H.,211 (ref. 32), 626 Schleich, H.,17 (ref. 64), 19 (ref. 64),26 (ref. 80, 85), 36 (ref. 80, 89), 37 (ref. SO), 38 (ref. SO), 39 (ref. 80), 40 (ref. SO), 42 (ref. 88), 43 (ref. 88). 44 (ref. 88), 45 (ref. 88), 47 (ref. 85, 88),49 (ref. 85), 50 (ref. a),51 (ref. 88),59 (ref. 84),61 (ref. 84),62 (ref. 84), 68 (ref. 89), 69 (ref. 89), 70 (ref. 89), 74 (ref. 85), 76, 77 Schmid, 121 Schmidt, C. L. A., 33 (ref. 92), 77, 133, 162, 263,902,311 (ref. 36),316 (ref. 36), 347, 358 (ref. 183), 389 (ref. la),393,@O Schmidt, H. L., Jr., 182 (see Sydenstricker), 199 Schmitt, F. O., 266, 302, 306 (ref. 144, 145), 325, 360, 359 (ref. 98, 179, 180),372 (ref. 179),418, 4.90 Schmitt, V., 17 (ref. 65), 51 (ref. 65),57 (ref. 65), 76 Schneider, F.,26 (ref. 80), 36 (ref. SO), 37 (ref. SO), 38 (ref. 80, 264), 39 (ref. SO), 40 (ref. SO),41 (ref. 193), 42 (ref. 193), 43 (ref. 193), 44 (ref. 193,264), 45,62,76. 79, 81 Schneider, M. L., 205,226 Schneider, W. C., 425 (ref. 29, 47), 427 (ref. 46, 47), 431 (ref. 47, 49), 432 (ref. 29, 47, 48), 436 (ref. 49), 438 (ref. 47), 439, 4.60 Schoenheimer, R., 23, 36 (ref. 266), 37 (ref. 266),39 (ref. 266), 81, 156,199, 282,303 Schoeller, W., 49 (ref. 165),511 (ref. 165), 79
Schoenebeck, O., 67 (ref. 194), 68 (ref. 194),79 Schofield, R. K., 306 (ref. 146), 960 Schott, H.F., 15,39 (ref. 267),81 Schramm, C. H., 21,22,37(ref. 301),82, 307 (ref. 174),961 Schrauth, W., 5 (ref. 166), 6 (ref. 166), 50 (ref. 166), 79 Schroeder, H., 387 (ref. 151), 419
460
AUTHOR INDEX
Schroehr, G., 57, 81 Schryver, 9. B., 233, 297 Rchutte, E., 64 (ref. 269), 81 Schuteenberger, P., 2 Schuler, J., 37 (ref. l6), 38 (ref. 16, 28), 39 (ref. 16, 28), 76 Schultr, J., 285, 697, 429 (ref. 8, 50),435 (ref. 8), 437 (ref. 50), 438 (ref. 8,
W , 439, 440 Schulee, A., 36 (ref. 167), 38 (ref. 167), 79 Schulze, E., 278, 279, 283, 308 Schwab, G., 272, 279, 902 Schwank, M., 403, 419 Schweigert, B. S., 92, 103 (see Riesen), 166
Schwimmer, D., 171, 199 Scott, D. A., 132, 166 Bcott, V. C., 183, 198 Scott Blair, G. W., 306 (ref. 147), 960 Scotti, H. C., 365, 480 Sealock, R. R., 436 (ref. 56), 440 Sebelien, J., 210, 268 Seeler, A. O., 166, 199 Seeley, R. D., 160, 161, 162, 164, 166, 167, 168, 170 (see Allison), 171, 181, 182, l 8 5 , i 8 6 , i 8 7 , i 9 r l 199 Seibert, F. B., 265, 902 Seifrie, W., 306 (ref. 148, 1491, 960,425 (ref. 51), 433 (ref. 51), 440 Sekora, A., 306 (ref. 841, 948 Seligsberger,Id., 405, 406 (ref. 181), 420 Semmett, W. F., 223 (ref. 147), 226, 228 Sen, P. K., 283, 287, 289, 298 Senti, F. R., 228, 311 (ref. 131, 132, 152), 313 (ref. 151), 325,327,328, 3@,360 Serenyi, V., 205 (ref. 421, 226 Seycwetz, A., 411, 412 (ref. 152), 419 Seymour-Jones, F. L., 392, 421 Shankman, S., 91, 93, 101, 103 (see Dunn), 108 (see Dunn), 110 (see Dunn), 111 (see Dunn), 117 (see Dunn), 1.69, 160, 166, 218 (ref. 115), 228
Shaw, G., 18 (ref. 103a), 77 Shedlovsky, T., 126, 162 Sheehan, J. C., 32, 54 (ref. 270a), 81 Sheffield, F. M. L., 255, 697 Shemin, D., 20, 80, 82, 85, 90, 93, 94, 101, 106, 108, 114, 116, 120, 122, 162, 276, 289, 298, 301
Sherman, J. V., 311 (ref. 163), 314 (ref. 153), 360 Sherman, M. S., 274, 698 Sherman, 9. L., 311 (ref. 153), 314 (ref. 153), 360 Sherman, W. C., 245, 306 Shinn, L. A., 88, 109, 110, 115, 161, 162, 208, $27 Shore, A,, 88, 100, 162 Shuttleworth, 5. G., 389 (ref. 182), 420 Sickel, H., 38 (ref. 30), 39 (ref. 30), 58, 72 (ref. 29), '76 Sideris, C. P., 271, 90% Sieber, W., 21 (ref. 112), 77 Siegmund, W., 367 (ref. 33), 416 Sifferd, R. H., 3 (ref. 2721, 31, 36 (ref. 272), 52, 61 (ref. 272), 62 (ref. 272), 65 (ref. 272), 81 Sigmund, F., 22, 36 (ref. 273), 49 (ref. 273), 81, 82 Sigurgiersson, T., 255, 902 Silber, R. H., 166, 199 Simmonds, S., 34, 44 (ref. 274), 51 (ref. 274), 68 (ref. 273a), 69 (ref. 273a), 70 (ref. 273a), 81 Simms, H. S., 36 (ref. 235), 80 Simpson, F., 199 Sinclair, W. B., 241, 262, 298, 902 Singer, W., 38 (ref. 32), 76 Sjogren, B., 214, 228 Skoog, F., 284, 502 Slack, E. B., 179, 197 Slade, R. E., 235, 302 Slavin, H. B., 183, 199 Slein, M. W., 264, 696 Slonim, N. B., 37 (ref. 276a), 43 (ref. 276a), 44 (ref. 276a), 88 Smith, A. H., 241, 302 Smith, A. K., 262, 302,311 (ref. 35), 314 (ref. 35), 347 Smith, A. M., 238, 302 Smith, C. S., 38 (ref. 275), 51 (ref. 275) 81
Smith, E. L., 31 (ref. 2761, 36 (ref. 201a, 276), 37 (ref. 276a), 38 (ref. 275b), 39 (ref. 201a), 40 (ref. 276), 43 (ref. 276a), 44 (ref. 276a), 68, 69, 70 (ref. 201a, 276), 71 (ref. 275a, 276), 72, 80, 81, 82, 92, 95, 99, 102, 103, 105, 110, 111, 112, 113, 117, 118,
AUTHOR INDEX
119, 120, 138, 162, 202, 210, 212, 213, 214, 218 (ref. 117, 118, 121123), 223, 228, 263. 264, 267, 308 Smith, E. P., 121 (see Graham), 160 Smith, H. Dew., 331 (ref. 154), 360 Smith, H. P., 182, 199 Smith, J., 314 (ref. 168), 361 Smith, J. D., 275, 899 Smith, T., 211, 828 Smith, V. A., 36 (ref. 19Oa), 38 (ref. 190a), 79 Smith, W. H., 286, 299 Smuts, D. B., 236, 300 Smythe, C. V., 358 (ref. 183), 389 (ref. 183),393, 420 Snell, E. E., 85, 90, 91, 92, 101, 102, 103, 104, 106, 108, 110, 111, 112, 113, 116, 117, 118, 119, 120, 138, 160, 161, 168, 218 (ref. 45, 51), 226, 264, 276, 298, 302 Sgirensen, M., 135, 162, 214, 215, 223 (ref. 127), 228 S@rensen, S. P. L., 214, 215, 223 (ref. 127), 228 Sokolov, S. J., 4-20 Somers, G. F., 263, SO3 Sookne, A. M., 312 (ref. 155), 360, 363 (ref. 9), 367, 416 Soule, M. H., 250, 297 Sourlangas, S. D., 395 (ref. 185), 4.80 Soutoulov, A. N., 280, 302 Spath, H., 61 (ref. 114), 64 (ref. 114), 77 Speakman, J. B., 308 (ref. 156), 331, 360, 376 (ref. 186), 420 Spiegelman, S., 277, 302, 427 (ref. 52, 53), 440
Spielman, M., 75 (ref. 251), 81 Srb, A. M., 273, 502 Stahlschmidt, A,, 43 (ref. 157), 44 (ref. 157), 78 Stahmann, M. A., 37 (ref. 277), 40 (ref. 277), 82 Stamberg, 0. E., 244, 301 Stamm, G., 57 (ref. 279), 82 Stanley, P., 238, 300 Stanley, W. M., 162, 254, 255, 256, 277, 292, 298, 299, 302 Stare, F. J., 246, 302 Staudinger, H., 406 (ref. 187), 420 Staudt, W., 63 (ref. 226), 80
46 1
Stearn, A. E., 308 (ref. 38), 347 Stecker, H. C., 367 (ref. 104), 368, 377 (ref. 104), 411, 412 (ref. 188), 418, 420
Stedman, E., 266, 302, 427, 440 Stedman, E., 266, 302, 427, 4.40 Steenbock, H., 245, 897 Steffee, C. H., Jr., 158 (see Frszier), 189, 190, 191, 194 (see Block), 197, 198, 8rn Steiger, R. E., 5 (ref. 237), 18 (ref. 236), 38 (ref. 231, 237), 39 (ref. 237), 80 Stein, A. M., 311 (ref. loo), 316 (ref. loo), 336 (ref. loo), 345 (ref. loo), $49 Stein, R., 333 (ref. 157), 360 Stein, W. H., 3, 36 (ref. 278), 38 (ref. 278), 57 (ref. 279), 79, 81, 82, 85, 95, 101, 102, 103, 104, 111, 112, 113, 116, 122, 161, 162 Steingroever, J., 37 (ref. 168), 38 (ref. 168), 39 (ref. 168), 79 Steinhardt, J., 363, 364, 420 Stekol, J. A., 73 (ref. 279a), 82 Stepka, W., 297 Stepto, R. L., 197 Stern, F., 16 (ref. 67), 18 (ref. 66), 19 (ref. 66), 23 (ref. 67), 41 (ref. 67), 43 (ref. 67), 46 (ref. 67), 49 (ref. 66, 67), 50 (ref. 66), 51 (ref. 67), 76 Stern, K. G., 264, $02 Stevens, C. M., 18 (ref. 98), 25, 77 Steward, F. C., 270, 271, 273, 276, 287, 288, 289, 292, 297, 302, 303 Stewart, A. M., 244, 303 Stewart, C P., 220, 667 Stewart, G. F., 171 (see Swanson, P.), 199, 306 (ref. go), 348 Stiasny, E., 365, 380 (ref. 193), 381 (ref. 193, 194), 383, 393 (ref. 193), 397, 406 (ref. 192), 418, 420 Stockelback, L. S., 262, 303 Stockinger, H. E., 268, 303 Stokes, J. L., 91, 103, 106, 109, 110, 111, 112, 113, 116 (see Gunness), 117, 118, 119, 120, 122, 160, 166, 163, 248,249, 250,251,252, 253, 254, 256, 258, 259, SO3 Stoklasa, J., 220, 888 Stout, P. R., 287, 303 Straitiff, W. G., 241, 299
462
AUTHOR INDEX
Strasburger, 428 Straube, R. L., 190, 194 (see Block), 197 Street, H. E., 270, 273, 276, 287, 289, 303 Strohschein, F., 15, 18, 33 (ref. 249), 81 Strong, F. M., 162 Stuart, L. S., 244, 301 Stuart, N., 293, 296 Stueck, G. J., 88 (see Shore), 100 (see Shore), 162 Sturtevant, A. H., 424 (ref. 55), 440 SubbaRow, Y., 45 (ref. 280), 82 Sukhov, K. S., 255,303 Sullivan, J. T., 235, 303 Sullivan, M. X., 72 (ref. 213), 73, 80, 96, 103, 105, 111, 117, 160, 163, 218 (ref. 129), 828,256, 262, 298,300 Sumner, J. B., 263, 303 Sutherland, G. B. B. M., 112 (see Darmon), 117, 149 Suzuki, U., 7 (ref. 172), 10 (ref. 170), 12 (ref. 169), 27 (ref. 172), 50 (ref. 171), 53 (ref. 171), 54 (ref. 171), 58, 61 (ref. 172), 63, 66 (ref. 172), 67 (ref. 169), 79 Svanberg, O., 220, 228 Svedberg, T., 134, 163, 207, 214, 228, 263, 303 Swallen, L. C., 311 (ref. 168), 360 Swaneon, P. P., 171, 172, 197,199,200 Sydenstricker, V. P., 182, 197,198,199 Synge, R. L. M., 3 (ref. 242, 281, 282), 37 (ref. ZSZa), 40, 51 (ref. 282a), 60, 61 (ref. 282a), 62 (ref. 282a), 68 (ref. 282a), 69 (ref. 282a), 70 (ref. 282a), 81,82,84, 88,94, 95, 110, 123, 160, 161,237, 259, 300, 303,360, 419 Szent-Gyorgyi, A., 306 (ref. 159), 311 (ref. 159), 360 Seymamki, T. A., 199
T Takahashi, W. N., 235,260,293,294,298 Tamura, S., 250, 303 Tanner, F. 249, 303 Tatum, E. L., 44 (ref. 274), 51 (ref. 274), 81,273, 303 TaubBck, K., 280, 299 Taylor, E. S., 276, 303 Taylor, G. L., 128, 163
w.,
Taylor, M. W., 193 (see Rumell), 199 Taylor, 8. P., 48 (ref. 112b), 71 (ref. 112b), 72 (ref. 112b), 73 (ref. 112b), 77 Teague, D. M., 202 (ref. a), 203 (ref. 8), 218 (ref. 8), 226 Templeman, W. G., 287, 301 Thatcher, R. W., 222, 228 Theis, E. R., 372 (ref. 148), 391, 392 (ref. 148), 399 (ref. 197), 409 (ref. 197), 412 (ref. 198), 419,420 Theorell, H., 89, 163,219 (ref. 135), 222, 228 Thiele, H., 400 (ref. 12), 416 Thierfelder, €I., 43 (ref. 283), 44 (ref. 283), 82 Thomas, A. W., 373, 384, 385 (ref. 201), 386 (ref. 201), 391 (ref. 206), 392, 395 (ref. 202), 396, 397, 411, 420,4x1 Thomas, K., 156, 164, 169, 200 Thomas, L. E., 116, 163 Thomson, R. H. K., 311 (ref. 160), 315 (ref. 160), 360 Thurlow, S., 220, 826 Tiedjens, V. A., 271, 303 Tieteman, J. E., 6, 16 (ref. 116), 18 (ref. 116), 37 (ref. 115), 49 (ref. 115), 69, 70 (ref. 68), 71, 76,77 Timm, E., 239, 300,303 Timmerman, W. A., 211, 228 Tincker, M. A. H., 284,303 Tiselius, A., 94, 163 Toboleky, A., 332 (ref. 161), 335 (ref. lei), 360 Tomarelli, R. M., 191, 800 Tomlineon, J., 403 (ref. 96), 418 Tracy, P. H., 202 (ref. 89), 203 (ref. 89), 227 Traill, D., 311 (ref. 162-164), 315 (ref. 162-164), 360 Trasciatti, 2 Travers, J. J., 176, 192, 198 Trikojus, V. M., 295 Tristram, G. R., 85, 94, 95, 101, 102, 104, 108, 109, 110, 111, 112, 113, 114, 116, 117, 118, 122, 124, 129, 130, 143, 149, 163, 218 (ref. 138), 228,234, 235, 237, 241, 247, 257,303 Turley, H. G., 375, 421 Turner, J. S., 295
AUTHOR INDEX
U Udenfriend, S., 93 (see Keston), 94, 101 (see Keston), 114 (see Keston), 161, 218 (ref. 66), 886 Umlauft, W., 278, 308 Underhill, 184 Underwood, E. J,, 252, 299
V
463
Waisbrot, S. W., 342 (ref. 78), 348 Wakeman, A. J., 205, 206 (ref. 97), 227, 233,235,273,281,284,288,301, 303 Walden, P., 13,82 Waldschmidt-Leitz, E.,17 (ref. 298), 29 (ref. 298),82 Walker, J. F., 411 (ref. 210),421 Walkley, J., 284, 304 Wall, F.T., 335,360 Wallerstein, J. S., 264, 302 Wang, T.,238, 306 Warburg, O.,12,264,271,304, 438 Ward, W. H.,311 (ref. 166), 317 (ref. 166),560 Warner, R. C., 88, 100, 163, 204, 205, 206, 208, 209, 215, 222, 887, 928, 262, 298, 999, 311 (ref. llO), 325 (ref. lll),349 Washburn, M. R., 182 (see Niven), 199 Watson, G. M., 270,303 Waugh, D.F., 314,361 Weber, H. H., 311 (ref. 169), 314 (ref. 169),361 Weber, L. E., 38 (ref. 34),39 (ref. 34), 76 Weech, A. A., 186,2W Weidinger, A., 361 (ref. 116),418 Weidle, H.,46 (ref. 2), 50 (ref. 2), 51 (ref. 2), 76 von Weimarn, P. P., 309 (ref. 170),311 (ref. 170), 316 (ref. 170), 361 Weir, C. E., 414,415, 481 Weiss, S., 193, 198 Weizmann, C.,22 (ref. 176a),79 Weller, R. A., 234, 236, 237, 238, 239, 243, 245, 279,300 Wells, H.G., 202, 211, 223, 226' Werkman, C.H., 285,304 Werner, L.H., 75 (ref. 251, 294),81, 88 Wessely, F., 8, 21, 22, 36 (ref. 273), 48 (ref. 296),49 (ref. 273), 81, 8.2 Westall, R. G., 233, 697 Wettstein, A., 75 (ref. 251, 252, 294), 81,
Valk6, E., 335 (ref. 119),349 Vandendriessche, L., 221, 227 Vandevelde, A. J. J., 220,228 Van Etten, C., 311 (ref. 37),314 (ref. 37), 347, 436 (ref. 561,440 Van Lanen, J. M., 249, 303 Vam, H.M., 189, 198, 200 Vaseel, B., 105, 163 Velick, 5.F., 101,102,103, 105, 106,108, 109, 110, 111, 112, 113, 116, 117, 118, 119, 120, 130, 140, 147, 163 Vendrely, R., 268,296, 303 Venkatesan, T. R., 280, 297 Vickery, H.B., 4 (ref. 284), 41, 82, 84, 105, 118, 119, 141, 163, 233, 234, 242, 262, 267, 271, 273, 279, 281, 282, 284, 288, 303 Viets, F. G., 272,303 du Vigneaud, V., 3 (ref. 272), 11, 16 (ref. QO), 22 (ref. no), 23, 25 (ref. 290), 31, 36 (ref. 220, 272), 40 (ref. 220), 42 (ref. 288), 43 (ref. 248), 44 (ref. 248, 288, 289, 291), 46 (ref. go), 51 (ref. QO), 52,53, 54, 55, 61 (ref. 220, 221, 272), 62 (ref. 45, 220, 221, 240, 272),63 (ref. go),65,66,71 (ref. 69), 76, 77, 80, 81, 88, 105, 132, 161, 163, 436,G O Virtanen, A. I., 259, 271, 273, 274, 275, 303,so4 Vischer, 266 Vivino, A. E., 241, 304 Vlassopoulos, V.,39 (ref. 33),76 88 Vogl, L.,400,401, 421 Wetzel, K., 279, 301 Von Susich, G., 335 (ref. 119),349 Whewell, C.S., 376 (ref. 186),420 Vorob'eva, M. N., 292,SO4 Whipple, G. H., 156, 180, 182, 183, 184, Vovk, A. M., 255, 303 185, 187, 189, 194, 198, 199, 9LW SV White, A., 105,163 Wahling, H. B., 274, 696 White, H.J., Jr., 333 (ref. 68),348 Wainio, W. W., 436 (ref. 57), 4 0 White, J., 55, 89
464
AUTHOR INDEX
White, J. I., 188, 196 Wintersteiner, O., 105 (see du Vigneaud), Whitehead, E. I., 272, 303 163 Whittier, E. O., 314 (ref. 171), 361 Wissler, R. W., 158 (see Frazier), 189, Wichmann, A., 214, 228 190, 191, 194 (see Block), 19Y, 198, Wicklund, E., 428 (ref. 58), 440 do0 Widen, P. J., 381 (ref. 97), 383 (ref. 97). Witte, C., 16 (ref. 67), 23 (ref. 67), 41 (ref. 67), 43 (ref. 67), 46 (ref. 67), 49 418 Widmark, O., 5, 76 (ref. 67), 51 (ref. 67), 76 Wielend, P., 59, 62 (ref. 255a), 81 WBhlisch, E., 335 (ref. 172), 361, 369, Wildman, S. G., 260, 207, SO4 370, 421 Wilkins, €1. L., 234, 304 Wolesensky, E., 403, 421 Williams, E. F., 87 (see Chibnall),' 97 Wolf, F. T., 249, 304 (see Bailey), 106 (see Bailey; Chib- Womack, M., 3 (ref. 299), 82, 158 (see nall), 108 (see Bailey; Chibnall), 114 Borman), 185, 194, 197, 200 (see Chibnall), 115 (see Bailey), 118 Wood, H. G., 285, 304 (see Chibnall), 119 (see Chibnall), Wood, J. G., 270, 271, 275, 276, 281, 282, 120 (see Chibnall), 136 (see Bailey), 283, 284, 285, 286, 287, 289, 301, 138 (see Chibnall), 148, 149, 218 304 (ref. 3), 826,253, 697 Wood, J. L., 52 (ref. 300), 53 (ref. 300), 82 Williams, G., 240, 304 Wood, S., 103 (see Lyman), 161 Williams, H. H., 105 (see Beach), 148 Wood, T. R., 158 (see Borman), 197 Williams, J. W., 316 (ref. 92), 349 Woodman, H. E., 212, 228 Williams, R. C., 255, 301 Woods, E., 244, SO4 Williams, R. F., 286, 287, 301, 30.4 Woods, F. M., 183, 198 Williams, R. J., 108 (see Hac), 113 (see Woods, H. J., 329 (ref. 6), 335 (ref. 173), 346, 361 Guirard), 118 (see Guirard), 160, 264, S98 Woods, M. W., 294, 304 Williamson, M. B., 218 (ref. 144), 219 Woodward, R. B., 21, 22, 37 (ref. 301), 82, 307 (ref. 154), 361 (ref. 144), 838 Willman, W., 171 (see Brush), 172, 197, Woolley, D. W., 3 (ref. 302), 24, 57, 82 194, 200,424 (rcf. 62), 440 $00 Willstlitter, R., 17 (ref. 298), 29 (ref. Woolridge, R. L., 158 (see Frazier), l%J, 190, 191, 194 (see Block), 197, 198, 298), 8,9 200 Wilson, E. B., 428 (ref. 60), 430 (ref. 59, Worcester, J., 179, 198 6l), 440 Worden, A. N., 240, 299 Wilson, E. J., 7, 26 (ref. 254), 81 Work, T. S., 25, 49 (rcf. 204a), 80, 290 Wilson, H., 88 (see Shore), 100 (see Wormell, R. L., 242, Y04, 409 (ref. 217), Shore), 168 431 Wilson, J. A., 354 (ref. 211), 366, 384, Wright, S., 427 (ref. 63, 64), 440 385 (ref. 212), 391 (ref. 212), 396 Wrinch, 132 (ref. 212), 398 (ref. 212), 400 (ref. Wulff, H. J., 264 211), 406 (ref. 212), 408 (ref. 212), Wybert, E., 54 (ref. 35), Y6 411 (ref. 212), 420, 421 Wyckoff, R. W. G., 255, SO1 Wilson, P. W., 274, 275, 296,297, 304 Wyrnan, J., 141, 163 Wiltshire, G., 97, 106, 108, 115, 122, 132, Wynd, F. L., 292, 304 130, 138 Y Windus, W., 375, 421 Winnick, T., 110,163 Ydse, L. C., 180 (see Bosshirrdt), 197 Winternitz, O., 119, 163 Yemm, E. W., 281,304
AUTHOR INDEX
Young, G. T., 43 (ref. 227a), 44 (ref. 227a), 45 (ref. 227a), 53 (ref. 35), 80 Young, H. Y., 271, 302 Young, M., 175 (see Zucker, L.), 179 (see Zucker, L.), 180 (see Zucker, L.), 800
Yudkin, W. H., 17 (ref. 304), 82 Z
Zaitschek, A., 219 (ref. 146), %28 Zamenhof, S., 268, 297 Zeisset, W., 38 (ref. 36), 39 (ref. 37), 40 (ref. 36), 76, 76 Zeldis, L. J., 181, 182, 200 Zervas, L., 5 (ref. 82), 7 (ref. 78), 16 (ref. go), 21 (ref. 74), 25, 26 (ref. 74, 77, 80, 85), 31 (ref. 86), 36 (ref. 74,
465
77, 80, 89), 37, 38 (ref. 78, 80), 39 (ref. 78, 80), 40 (ref. 80), 41 (ref. 74), 42 (ref. 74, 87, 88), 43 (ref. 74, 76, 78, 79, 87, 88), 44 (ref. 74, 77, 78, 88),45 (ref. 74, 87, 88), 46 (ref. go), 47 (ref. 85, 88), 49 (ref. 74, 85), 50 (ref. 74, 88), 51 (ref. 88, go), 53 (ref. 74), 57 (ref. 74), 59 (ref. 81, 84, 86), 60 (ref. 86), 61 (ref. 74, 81, 83, 84, 86), 62 (ref. 81, 83, 84, 86), 63, 66, 68 (ref. 89), 69 (ref. 89), 70 (ref. 89), 71 (ref. 69), 74, 76, 77, 8% Zetzsche, F., 25, 81 Zittle, C. A., 120, 163 Zucker, L., 175, 179, 180, 200 Zucker, T. F., 175 (see Zucker, L.), 179 (see Zucker, L.), 180, 800
Subject Index A
Aldehydes (see also Formaldehyde), as tanning agents, 405-41 1 Aldehydrase, see Xanthine oxidase Aldolase, see Myogen A
Acetyldehydroalanine, synthesis of, 19 Acetyldehydrophenylalanine, synthesis of, 18 AhW, Acetyldehydrotyrosine, synthesis of, 18 amino acid composition of proteins of, N-Acetyldiketopiperazine, 71 247-249, 258 Actinom ycetes, Alkalis, amino acid composition of, 252, 258 effect on collagen, 368,377 Acylamino acids, synthesis of peptide Amandin, derivatives from chlorides of, 9 molecular weight of, 263 Amino acids, see also under names of Acyldehydroamino acids, preparation of, 19 individual members, analysis in protein hydrolyzates, 83-124 synthesis of peptides from azlactones of, 18 accuracy of methods, 99-106 Acyl halides, a-halogeno, by colorimetric methods, 98 preparation of optically active, 12, 13 by electrodialysis, 89 synthesis of peptides from, 11-15 by enzymatic methods, 90 Aerobacter aerogenes, by isotope dilution, 93-94 amino acid composition of protein by microbiological assay, 90-93 from, 252 by partition chromatography, 94ff. Alanine, 9 availability of, 193 derivatives of, 36 chlorides of, estimation in protein hydrolyeates, 115 preparation of, 11 comparison of methods for, 96, 114 synthesis of peptides from, 1 1 peptides of, 35-40 in collagen, 357ff. optical rotation of, 38, 39 destruction during protein hydrolysis, synthesis of, 3, 5, 13, 19,24,35ff. 123ff. distribution in muscle proteins, 140 sources of,114,129,130, 131,137, 140, in plant proteins, 235-241, 243-246, 142, 143, 145 DxAlanine anhydride, 5 247,252 stability of, 6 in purified proteins, 83-148 8-Alanyl-chistidine, see carnosine effect on body weight, 185 8-Alanyl-l-methyl-bhistidine, see Anessential, determination of, 158 serine diet and, 158, 195, 196 Albumin, 181 egg, Bee Ovalbumin effect on appetite, 190 on body weight, 190 plasma, amino acid composition of, 142 on nitrogen equilibrium, 158, 159 erythrocytes and, 190 serum, amino acid composition of, 96, 101, nitrogen-sparing action of, 172 102, 103, 105, 106, 109, 110, 111, serum proteins and, 190 112, 113, 116, 117, 118 utilization of, 192 466
SUBJECT INDEX
estimation of dicarboxylic, 97-98, 114 of sulfur-containing, 95-97 fatigue and, 158 nervousnem and, 158 nitrogen balance and, 157-159 non-essential, 171 from proteins, 180 regeneration of plasma proteins and, 189 relation between protein levels and, in plants, 284 requirements, 176 resolution of optically inactive, by proteolytic enzymes, 34 specificity of methods for determination of, in protein hydrolyzates, 99-106 stores in animals, 180 synthesis in liver, 180 of peptides from N-carbonic acid anhydrides of, 21-23 in plant t k u e s , 271, 274 unnatural isomers of, 170 pAminobenzoic acid, as component of folic acid, 73 peptidee of, 73 Aminobutyric acid, 9 4-Amino-n-butyryl peptide, synthesis of, 40 a-Aminobbutyryl peptide, synthesis of, 40 Ammonia, protein synthesis from, in plants, 271274 role in nitrogen metabolism. 281 Amylase, in milk, 219 activity of, as criterion for quality of, 220 Angiosperms, isolation of proteins from leaves of, 233240 Anorexia, amino acids and, 158 Anserine, structure of, 65 Antibodies, transfer from mother to infant, 212 Arachm, 262
467
Arginine, derivatives of, 61 effect on body weight, 190 estimation in protein hydrolyzates, 90 comparison of methods for, 96, 118 hemoglobin output and, 185 isolation from soybean oil meal, 194 nitrogen metabolism and, 169, 185 peptides of, 46, 62-64 optical rotation of, 62 synthesis of, 63, 64 sources of, 118, 129, 130, 131, 132, 136, 140, 142, 143, 145 Ascomycetes, proteins of, 250-254 Asparagine, role in protein metabolism of plants, 278 synthesis of L-, 42 Aspartic acid, 9 derivatives of, 43 estimation in protein hydrolyzates, 114 comparison of methods for, 96, 106 peptides of, 41ff. optical rotation of, 44 synthesis of, 41, 42 sourcesof, 106, 124, 131, 132, 136, 137, 138, 140, 142, 143, 145 @-LAspartyl-Lcysteinylglycine (Asparthione), synthesis of, 53 Aspergillus niger, amino acid composition of protein from, 251 Aucuba virus, amino acid composition of green, 256 of yellow, 256 effect on protein metabolism of host, 292 Auxin, in chloroplasts, 267 Azides, synthesis of peptide derivatives from, 9 a-Azidoacyl halides, preparation of, 24 synthesis of peptides from, 24 Azlactones, 18 preparation of, 16, 17, 18 synthesis of peptides from, 15-19, 20, 46 Azotobacter, nitrogen assimilation, 274
468
SUBJECT INDEX
Azotobacter vinelandii, amino acid composition of protein from, 254
B Bacillus anthracia, capsular substance of, 251, 253, 259, 265,269 chemical nature of, 251, 265 function of, 253 as source of polyglutamic acid peptides, 45 Bacillua brevis, amino acid cornposition of protein from, 252 Bacillus subtilis, amino acid composition of protein from, 252 Bacteria, proteins of, 251-254, 268 amino acid compoRition of, 258 p-Benzoquinone, as tanning agent, 411-413 Benzoyl-Lleucylglycinanilidc, enzymatic synthesis of, 33 Benzoyl-cleucyl-bleucinanilide, enzymatic synthesis of, 33 Benzoyl-ctyrosy lglycinanilide, enzymatic synthesis of, 33 “Biuret base,” 7 Body weight, effect of amino acids on, 189 Bristles, made from casein powder, 325 a-d-Bromoisocaprony Iglycylgl ycylglycine, synthesis of, 14 Bryophytes, proteins of, 231 amino acid composition of, 258
C Cancer, and plasma proteins, 181 Carbethoxyglycylglycylleucine ethyl ester, synthesis of, 7, 8 Carbobenzoxy-balanyl azide, 40 Carbobenzoxyamino acids, synthesis of peptides from chlorides of, 25-32
4Carbobenroxyaminooxazolidone-2, 57 Carbobenzoxy-~glutamyl-~-tyrosine, hydrolysis by pepsin, 47 synthesis of ethyl ester of, 30 Carbohydrases, in milk, 219 Carbohydrates, effect on protein synthesis in plants, 285 Carbonyl-bis-gl ycine, synthesis of, 8 Carbonyl-bis-glycylglycine, synthesis of, 8 Carnosine, 3 structure of, 65 synthesis of, 65 Casein, 204-209 amino acid composition of, 101, 103, 105, 106, 109, 110, 111, 112, 113, 114, 116-120, 206, 218, 224 chemical difference between human and cow’s, 204, 206-208 heterogeneity of, 204 hydro1yzates, biological evaluation of, 159 plasma proteins and, 187 fibers made from, 310, 311, 314, 325 from mixtures of, and polyamides, 326 in milk, 202 molecular weight of, 207 phosphopeptones from, 208 composition of, 208 plasma proteins and, 186, 187 preparation of a-,205-206 of 8-, 206 regeneration of liver proteins and, 188 rennet, 209 separation of a- and 8-, 205 Catalase, in milk, 222 Caulerpa racemoaa, 248 Cells, constituents of, 423-439 distribution of proteins in, 265-267, 423-439 duplication of living, 433 localization of specific functions in, 425 role of nucleic acids in, 434 of phospholipids in, 434 of proteins in, 434
SUBJECT INDEX
Chenopodiacae, amino acid composition of proteins in leaves of, 237 Chitin, 251 Chloroacetyldehydrophenylalanine,17 Chloroplasts, auxin in, 267 proteins of, 267 Chondrus erispans, amino acid composition of protein of, 248 Chromatin, 428 composition of, 426 isolation from cell nucleus, 426 Chromium, complex formation with collagen, 388392 salts of, reaction with collagen, 382 structure of basic, 379-380 and tanning potency, 380 as tanning agents, 354, 379-392 factors governing effect of, 379 Chromosomes, 425, 433 composition of, 426, 427 structure of, 427 Chromosomin, 266, 427 Choline, nitrogen-sparing action of, 172 Chymotrypsin, hydrolysis of peptides by, 47 peptides as substrates for, 47 synthesis of peptide derivatives by,
33 Chymotrypsinogen, amino acid composition of, 145 Clostridium botulinum type A toxin, amino acid composition of crystalline, 124, 143 Codium fragile, 248 Collagen, 306, 355 acid- and base-binding capacity of, 362-365 amino acid composition of, 143, 357, 358 chemistry of, 356-360 complex formation with chromium, 388-392 denaturation of, 369
469
effect of alkali on, 368, 377 of lyotropic agents on, 372-375 of neutral salts on, 372ff. of pretreatment on reactivity, 374 fibers made from, 311, 325, 326, 335 heat shrinkage of, 369 isoelectric point of, 3666. modified, reaction with tanning agents, 414 molecular weight of, 358 reaction with aldehydes, 405 with chromium compounds, 379,328 with condensed sulfo acids (syntans), 402 with vegetable tannins, 3 9 5 4 0 0 factors governing, 396-400 Stabilization of, 360-362 structure of, 355, 359, 360 swelling of, 356, 365, 368 Colostrum, 210-214 chemical composition of, 212 immune properties of, 211-213 protein content of, 211 pseudoglobulin, amino acid composition of, 102, 103, 105, 106, 109, 110, 111, 112, 113, 117, 118-120 Conarachin, 262 Conglutin, heat of combustion, 264 Cottonseed, fibers made from proteins of, 311, 315 Cruciferae, amino acid composition of proteins in leaves of, 237 Cucumber 3(4) virus, amino acid composition of, 256 Cupriethylenediamine, complex formation with silk fibroin, 324 Cysteine, derivatives of, 53 determination in intact proteins, 96 in protein hydrolyzates, 96, 111-1 13 comparison of methods for, 105 peptides of, 6, 49-55 hydrolysis by pepsin, 47 optical rotation of, 54 synthesis of, 32, 49-54 sourcesof, 105, 124, 131, 132, 135, 137, 140, 142, 143, 145
470
SUBJECT INDEX
LCysteine anhydride, hydrolysis of, 6 Cystine, 147 derivatives of, 54 estimation in protein hydrolyzates, 111-113
comparison of methods for, 105 hematopoiesis and, 184 nitrogen-sparing action of, 172 peptides of, 496. optical rotation of, 54 synthesis of, 49, 54, 55 regeneration of liver proteins and, 189 sources of, 105, 124, 131, 132, 137, 140, 142, 143, 146
stability of keratin and, 376 Cysloseira osmundaceae, 248 Cytochrome c, 431, 436 role in aerobic cell respiration, 438,439 Cytochrome oxidasc, 431 nature of, 436 Cytoplasm, 428-433 aerobic cell respiration and, 438 constituents of, 425,426,427,429, 433, 438
functional relation between cell nucleus and, 427 labile liver, 187 ribonucleic acid content of, 438
D Decarboxylaaes, amino acid estimation in protein hydrolyaates with, 90 sources of, 90 Dehydrogenase, in milk, 220 Dehydropeptidasc, 17 action on peptides, 17, 19 Depilation, of animal hide, 375-378 Desoxyribonucleic acids, 265, 268, 427 biological conversion to ribonucleic acid, 43 distribution in cells, 437 location in cells, 437 Detergents, application to protein fiber preparation, 317-324 complex formation with proteins, 318320
Diet, amino acids and, 158 effect on plasma proteins, 182 Digestibility, of dietary nitrogen, 162, 165 3,5-Diiodo-btyrosine, synthesis of peptides of, 49 Diketopiperazines, racemization of, 5, 6 synthesis of, 7 of dipeptides from, 5-6, 9 &fl-Dirnethyl-D-cysteine, as component of penicillin, 74
5,5-Dimethylthiaaolidine4 carboxylic acid methyl ester, 75 Donnan effect, 365, 366
E Edestan, conversion of edeatin to, 1 3 4 Edestin, amino acid composition of, 98, 101, 102, 103,105,106,109, 110,111, 112, 113, 114, 116-120, 131, 134, 264 conversion to edestan, 134 fibers made from, 325 heat of combustion, 264 molecular weight of, 98, 128, 134, 263 structure of, 134-135, 148
Egg protein, see also Ovalbumin, effect on growth, 177 nitrogen equilibrium and, 164 Egregria wnziesii, 248 Elastin, fibers made from, 335 Enzymes, crystalline, 263, 264 amino acid composition of, 145 in milk, 31S223 in mitochondria, 431 protein metabolism in plants and, 288290
proteolytic, in milk, 222 peptides and, 11, 14, 33, 34, 40 resolution of Dbamino acids by, 34, 35
specificity of, 11, 33
47 1
SUBJECT INDEX
DrEpiglucoaaminic acid, configuration of, 74 synthesis of peptide of, 74 Escherichiu coli, amino acid composition of protein from, 252 Esterases, in milk, 221 Estrus cycle, effect on milk lipase activity, 221 Euglobulin, 210, 213 kr whey, 203 Euphorbiacae, amino acid composition of proteins in leaves of, 237 Excelsin, molecular weight of, 263
F Fatigue, and amino acids, 158 Feathers (see also Keratin), fiber-forming properties of, 315 protein nature of, 315 Fibers, synthetic, from proteins, 305-346 alkaline agents in preparation of, 313-315
crystallization in, 328 deformation behavior of, 332 effect of water on, 345 interpretation of stress-strain behavior of, 331-334 molecular basis for mechanical properties of, 327-336 experimental methods in the study of, 331 muscle protein as source of, 306, 311, 335
polypeptide chains of, 308 relation between, and mechanical properties of, 312, 327-336 properties of, 345 stabilizing bonds in, 308, 309, 316, 339ff., 344, 345 stability of, 308, 346 structure of, 308 Fibrin, amino acid composition of, 101, 102, 103, 106, 113, 116, 117, 118, 120
evaluation of hydrolyzates of, 159 stability of, 376 Fibrinogen, amino acid composition of, 142 fiber-forming properties of, 311 Fishes, fiber-forming properties of protein8 from, 311 Flavin adenine nucleotide, flavin moiety of milk xanthin oxidase and, 220 Flavoproteins, isolation from milk, 220 Folic acid, structure of, 45, 73 Formaldehyde, reaction with proteins, 405, 410 as tanning agent, 405-411 Fucus furcatus, 248 Fungi, proteins of, 249-250 amino acid composition of, 250, 251, 258
nutritive value of, 249 G
Gelstin, acid-binding capacity of, 365 amino acid composition of, 143 base-binding capacity of, 365 fibers made from, 310 from mixtures of polyamides and, 326
isoelectric point of, 367 proline peptides from, 67 swelling of, 356, 366 Gliadin, 262, 263 amino acid composition of, 143, 264 hydrolyzation products of, 3 proline peptides from, 67, 68, 70 regeneration of liver proteins and, 188 solubility of, 147 Globulin, colostrum, 210, 211 amino acid composition of, 210, 211, 218
fibers made from, of tobacco seeds, 325 milk, composition of, 210 preparation from whey, 203, 210 in plasma proteina, 181, 182
472
SUBJECT INDEX
u-Globulin, amino acid composition of, 112 7-Globdin, amino acid composition of, 101, 102, 103, 105, 106, 109, 110, 111, 113, 116-120, 142 immune properties of plaema, 213 Gloedrichia echinulda, 248
rr
peptides of, 3, 35-40 optical rotation of, 38 synthesis of, 5, 6, 7, 11-12, 32, 40 preparation of polymers of, 22 role in tho folding of proteins, 330 source8 of, 129, 130, 131, 137, 140, 142, 143, 145
Glycine anhydride, Glucosaminic acid, hydrolyzation products of, 5 melting point, 36 peptides of, 74 Glutamic acid, synthesis of, 5 Glycinin, 262 derivatives of, 43 estimation in protein hydrolyzates, 80, Glycyl-Lalanyl-cleucine, structure of, 4 114, 115 comparison of methods for, 96, 108 Glycyldehydroalanine, synthesis of, 19 molecular weight of D( -)-, 265 Glycyldeh ydrophen ylalaninc, peptides of as substrate for dehydropeptidasr, 17 physiological role of, 45 synthesis of, 17 optical rotation of, 44 Glycyldehydrophenylalanyl-I-glutamic sources of, 45 acid, synthesis of, 17 synthesia of, 16, 17,32, 43,44, 45,46 Glycyltaurine, 55 polypeptide of D( -)-, in capsular sub- Gramicidin, 259 stance of Bm'llu8 anlhracis, 251, peptide nature of, 3 253, 265,268 Gramicidin S, sources of, 108, 124, 131, 132, 136, 137, peptide nature of, 3 138, 140, 142, 143, 145 Graminae, LGlutamine, amino acid composition of proteins in synthesia of, 45 leaves of, 237 a-LGlu tamyl-Lcysteinylglycine Growth, (isoglutathione), 54 effect of diet on, 179 7-L-Glutamyl-Lcysteinylglycine, see and nitrogen retention, 173 Glutathione, Gymnosperms, 7-D-Glutamyl-Lcysteinylglycine(epi-gproteins of, 231 glutathione), 53 H Glutathione, 3 structure of, 51 synthesis of, 51, 52, 53 Hematopoiesis, Glutelin, 263 and isoleucine, 184 Hemoglobin, Gluten, wheat effect on growth, 177 amino acid composition of, 98, 101, on nitrogen equilibrium, 164 102, 103, 105, 106, 109, 110, 111, fractionation products of, 262 112,113, 114, 116, 117-120, 137, 139 nutritive value and plasma proteins, fibers made from, 325 molecular weight of, 98, 128, 139 187 Glutenin, 262 in root nodules, 275 GIycine, structure of, 139, 147 Hexokinase, 264 derivatives of, 36 effect on nitrogen balance, 159 Hide, see also Collagen, Leather estimation in protein hydrolyzates, 110 depilation of, 375-378 comparieon of methods for, 96, 101 stability of vegetable tanned, 399
473
SUBJECT INDEX
Hippuric acid, 9 Histidine, derivatives of, 61 effect on body weight, 190 on hemoglobin output, 185 on nitrogen excretion, 185 estimation in protein hydrolyzates, 90 comparison of methods for, 96, 119 peptides of, optical rotation, 62 synthesis, 65-67 sourcesof, 119, 131, 137, 140,142, 143, 145 Histidine anhydride, stability of, 6 synthesis of, 7 Holmes’ masked virus, amino acid composition of, 256 Holmes’ rib grass virus, amino acid composition of, 256 Homocysteine, peptides of, 54, 73 Homocystine, peptides of, 54 Hordein, heat of combustion, 264 Hormones, protein metabolism in leaves and, 284 Hydantoin, conversion of peptides to derivatives of, 48 Hydrogen ion concentration, role in chrome tanning, 386-388 in formaldhyde tanning, 408 in vegetable tanning, 396, 400-402 Hydroxylysine, sources of, 143 Hydroxyproline, derivatives of, 69 estimation in protein hydrolyzates, 109 peptides of, optical rotation, 70 synthesis of, 71,72 sources of, 109, 143
I
fibers from, 314 molecular weight of, 128, 132 structure of, 132-133 Isoleucine, derivatives of, 37 synthesis of, 35,38, 39 effect on body weight, 190 on nitrogen balance, 158 estimation in protein hydrolyzates. 115, 117-119 comparison of methods for, 96, 112 hematopoiesis and, 184 nitrogen excretion and, 185 peptides of, 35 optical rotation, 39 sources of, 112,131,137,140,142, 143, 145
J J 14 D1 virus, amino acid composition of, 256
K Keratin, 306 amino acid composition of wool, 143 chain interaction in fibers made from feather, 336-346 chemical structure and stability of, 376 fibers made from, 311,323,324 from mixtures of, and polyamides, 326 solubility of, 316, 345 structure of, 356 water uptake by, 356 Keto acids, synthesis of peptides from, 19,20 Krebs cycle, 431 1
Lactalbumin, isolation of, 214 Infection, probable identity of crystalline, with proteins and, 180, 181 p-lactoglobulin, 214 Insulin , regeneration of liver proteins and, 188 amino acid composition of, 98, 101, source of, 203 102, 105, 106, 109, 110, 111, 112, Lactase, 113, 114, 116, 118-120, 130-132 in milk, 220
474
SUBJECT INDIJX
LaetobaciUus s p p . , amino acid composition of protein from, 252, 253 Lactoglobulin, composition of immune, 213, 218 immune properties of colostral, 213 isolation from colostrum, 212 #-Lactoglobulin, 144, 214 amino acid composition of, 96, 98, 101, 102, 103,105, 106,109, 110, 111, 112, 113, 114, 116, 117-120, 131, 135, 136, 138, 216, 218 denaturation of, 217-219 molecular weight of, 98, 128, 129, 215 preparation of, 214-215
probable identity with lactalbumin, 214
properties of, 215 solubility of, 216 structure of, 135, 136 whey ae source of, 203 Lactomucin, 214 Lactoperoxidase, isolation of crystalline from milk, 222 molecular weight of, 222 physical properties of, 223 Laminaria sp., amino acid composition of protein of, 248
Latex, protein from, 241 Leather (see also Collagen, Hide), chrome, 392-394 Leaves, proteins of, amino acid composition of, 236-240, 271
isolation of, 233-240 metabolism of, 281-283 hormonal control of, 284 nutritive value of, 236 relation between levels of, and amino acids, 284 and water, 284 Legurnelin, heat of combustion, 264 Leguminoaae, amino acid composition of proteins in leaves of, 237 Lcssoniopeie littoralis, 248 Leucine, derivatives of, 37
effect on body weight, 190 on nitrogen balance, 158 estimation in protein hydrolyeates, 90, 115
comparison of methods for, 96, 112 nitrogen excretion and, 186 peptides of, 6, 35ff. optical rotation, 38, 39 synthesis of, 12, 35-40 sourcesof, 112, 124, 130, 131, 137, 140, 142, 143, 145
synthesis of polymers of, 22 Leucine anhydride, stability of, 6 DGLeucyl-Dbalanine, isomers of, 4 Lignosulfonic acid, molecular weight of, 403 reaction with collagen, 403, 404 Lipase, in milk, 221 pitocin and, 221 relation between activity of, and estrus cycle, 221 Lipides, 434 as cell constituents, 426, 431, 435, 438
protein deficiency and, 182 role in cellular physiology, 435 Lipositol, 432 Liver, effect on growth, 180 protein metabolism and, 189 Lyotropic agents, effect on animal skin, 372-375, 377 Lysine, availability of, 194 derivatives of, 61 aa substrate for pancreatic trypsin, 60 effect on body weight, 190 on digestibility of wheat gluten, 165 on growth, 158 on nitrogen balance, 158 hemoglobin output and, 185 isolation from soybean oil meal, 194 methoda for estimation of, 96, 120 nitrogen excretion and, 185 peptides of, 58ff. optical rotation, 62 synthesis of, 58-60
475
GUBJECT INDEX
preparation of polymers, 22 sources of, 120,124,131, 137,140,142, 143, 145 Lysine anhydride, synthesis of, 7
M Macrocystis pyrifera, 248 Maize, amino acid composition of protein from, 244 Malnutrition, plasma proteins and, 181 Mesonin, 262 Methionine, availability of, 193 derivatives of, 71 effect on body weight, 190 on digestibility of casein, 165 of fibrin hydrolyzates, 165 on nitrogen balance, 158 estimation in protein hydrolyzates, 96, 111 comparison of methods for, 96, 103 hematopoiesis and, 184 nitrogen excretion and, 184 nitrogen-sparing action of, 171 peptides of, 72ff. optical rotation, 72 synthesis of, 73 protein metabolism and, 169, 189 regeneration of liver proteins and, 189 sources of, 103, 131,137, 140,142 143,
145 Microsomes, 425, 432, 433 composition of, 429,430,432,435,437 function in cell, 432,438 isolation of, 432 size of, 432 Milk, enzymes in, 219-223 isolation of flavoproteins from, 220 proteins of (see also under name of individual members), 201-225 amino acid composition of, 218, 219 distribution of, 202-203 properties of, 203-210 relationship to serum proteins, 223 separation of, 203-210 serum, see Whey
Mitochondria, 425,430-432,433 aerobic cell respiration and, 438 composition of, 429, 430, 431, 435, 437 enzymes in, 431 size of, 430 Molybdenum, effect on nitrogen assimilation, 274 Muscle, protein, fibers made from, 311,335 water uptake by, 356 Mustard gas, effect on proteins, 3 Myogen A, amino acid compoaition of, 140 source of, 147 Myoglobin amino acid composition of horse, 98, 137 molecular weight of horse, 98 structure of, 148 Myosin, amino acid composition of, 105, 140 sources of, 147 stability of, 376
N ,9-Naphthalenesulfonic acid, reaction with collagen, 402 Nervousness, amino acids and, 158 Nitrate, assimilation in plants, 270ff. Nitrogen, assimilation in plants, 274ff. effect of metals on, 274 role of ammonia in, 281 balance, 157, 160, 161, 163, 164, 167 essential amino acids and, 158, 159 repletion of plasma proteins and, 186 digestion, 162, 163 excretion, 157, 162, 163, 166, 168, 169, 170, 171, 172, 180, 183 diet and, 170 protein stores and, 161 fecal, 162, 163, 180 requirement in dietary proteins, 158 retention of, and growth, 173 and protein efficiency, 176
476
SUBJECT INDEX
Nucleic acids, 428,434, 437439 Pelvetia, in chromosomes, 427 as source of glutamic acid peptides, 45 in cytoplasm, 429 Penibillin, Nucleolus, 425,428,433 B,B-dimethyl-D-cysteine as component constituents of, 428,437 of, 74 Nucleoproteina, 428 Penicillium notalum, bacteria as source of, 268 amino acid composition of protein composition of, 427, from, 251 as constituents of chromosomes, 426, Pepsin, 427 amino acid composition of, 103, 145 function in cell, 427 hydrolysis of peptides by, 47 Nucleotides, 429 Pcptides (see also under names of indiNucleus, 425 vidual amino acids), functional relation between cytoplasm biological activity of, 3 and, 427 configuration of, 4 isolation of chromatin threads from, conversion to hydantoin derivatives, 420 48 derivatives of, synthesis, 9-11, 33-34 0 nomenclature, 4 Oranges, synthesis of, 1-75 amino acid composition of protein from amino acid chlorides, llff, from, 241 azlactone method for, 15-19 Ornithine, 60 by the carbobenzoxy method, 25-32 derivatives of, 61 from N-carbonic acid anhydrides of estimation in protein hydrolyzates, 90 amino acids, 21-23 peptides of, 618. by condensation of keto acids and optical rotation of, 62 amides, 19-20 synthesis of, 60-62 of peptide esters, 6-9 Osmunda claytonianu, from u-halogeno acyl halides, 11-15 isolation of protein of, 246 by partial hydrolysis of diketopiperaamino acid composition of, 247 zines, 5, 6 Ovalbumin, from phthalylamino acids, 32-34 amino acid composition of, 98, 101, from toluenesulfonylamino acids, 102, 103, 105, 106, 109, 110, 111, 23-24 112, 113, 114, 116, 117-120, 137, Phaseolin, heat of combustion, 264 138, 139 Phenylalanine, fibers from, 310,313,315,323,325,326 derivatives of, 49 relation between applied stretch and effect on body weight, 190 mechanical properties of, 327 on nitrogen metabolism, 158, 184 molecular weight of, 98, 128,313 on plasma protein output, 185 structure of, 139, 141 estimation in protein hydrolyzates, 115 Oxidases, in milk, 222 comparison of methods for, 96, 11 1 peptides of, 3, 6,11, 12, 32, 46ff. P optical rotation, 51 Papain, synthesis of, 46ff.,49 resolution of Dbglutamic acid by, 34 polymers of, synthesis, 22 synthesis of peptide derivatives by, 33 sourcesof, 111, 124,131, 137, 140, 142, Peanuts, 143, 145 protein of, 244,262 Phenylpyruvylamino acids, fibers from, 311, 315 synthesis of, 20
BUBJECT INDEX
Phormidium valderianum, amino acid composition of protein of, 248 Phosphatase, in milk, 221 properties of, 221 Phospholipides, as constituents of cytoplasm, 426,429, 430, 431, 432 role of, in the formation of cell membranes, 4346. Phosphopeptones, from casein, 208 composition of, 208 Phosphoserylglutamic acid, isolation from casein digests, 208 Phthalylamino acids, synthesis of peptides from, 32-34 Pitocin, activation of milk lipase by, 221 Placenta, transfer of antibodies through, 212 Plants (see also under names of the various divisions) assimilation of nitrate in, 270-271 proteins of, 229-295 homogeneity of, 261-263 metabolism of, 26S294 amino acids and, 284 carbohydrates and, 285 enzymes and, 288-290 respiration and, 286-288 water and, 284 synthesis of amino acids in, 271-277 utilization of ammonia by, 271-274 Plasmapheresis, 182, 185, 186 Pollen, proteins of, 241 Polyphenols, tanning properties of, 403 Potatoes, amino acid composition of protein from, 240 Prolamin, see Gliadin Proline, derivatives of, 69 estimation in protein hydrolyzates, 110-111 comparison of methods for, 96, 102 peptides of, 67ff. optical rotation, 70 from protein hydrolyzates, 67 synthesis of, 12, 14, 67-71
477
sources of, 102, 129, 130, 131, 137, 140, 142, 143, 145 Propionyldehydroaspartic acid, synthesis of, 18 Protease, in milk, 222 Proteins, acid-base relationships of, 3 action of mustard gas on, 3 amino acid composition of muscle, 140 and nutritive value of, 192 of plasma, 142 of purified, 129-148 analysis of hydrolyzates of, 88-125 biological evaluation of, 155-200 cellular, 423439 enzyme nature of, 259 determination of free amino groups in, 99 dietary, bioassay of, 188, 190 biological value of, 169 unidentified substances in, 185 effect on body weight, 176, 179, 180, 185 of food processing on nutritive value of, 173 of urea on soluble, 399 efficiency ratios, 175 fibrous, 305-346 chemical composition and properties of, 316, 355, 356 coagulation of, 309 complex formation with detergents, 318-320 conversion of corpuscular to crystalline, 325 denaturating agents for, 315, 316 peptide chains of, 315-317, 330 relation between molecular weight and properties of, 313 solubilizers for, 316 stabilizing bonds in, 316, 317, 373 effect of lyotropic agents:on, 373, 375 structure of, 355 and solubility, 316 suitability for fiber preparation, 326 infection and, 180, 181 iodination of, 3 liver, repletion of, 187
478
SUBJECT INDEX
milk, 201-225 amino acid composition of, 218, 219 of colostral, 21Ck214 distribution of, 202-203 properties of, 203-210 separation of, 203-210 molecular weights of, 128-129 nutritive value of, 156, 170, 175 of plants, 229-295 metabolism of, 269-294 asparagine and, 278 phylogenetic aspects of composition Of,
257-259
relation between type and function of, 265 synthesis of, 271-277 plasma, albumin content of, 181 cancer and, 181 malnutrition and, 181 nitrogen balance and, 186 regeneration of, 182, 185, 186 tuberculosis and, 181 potency ratios, 186 purity of, definition of, 125 determination of, 125-127 reaction with formaldehyde, 405 requirements of rats, 174 ribonucleic and cellular synthesis of 438
stability of a-,376 structure, 355 and solubility, 144, 147, 316 tryph-inhibiting, in soy beans, 263 unidentified factora in, 194 Proleus vulgari8, amino acid composition of protein from, 252 Protoplasm, 306 structure of, 266 Pseudoglobulin, 210 in colostrum, 211 in whey, 203 Pteridophytes, proteina of, 231, 246 amino acid composition of, 247, 258 Pteridium aquilinum, amino acid composition of protein from, 246, 247
R Respiration, aerobic, cytoplasm and, 438 role of cytochrome in, 438, 439 anaerobic, role of ribonucleic acid in, 438 relation between, and protein metabolism in plants 286-288 Rhizopus nigricans, amino acid composition of protein of, 25 1
Rhoddwula ruber, amino acid composition of protein of, 25
Ribonucleaae, amino acid composition of, 145 Ribonucleic acid, 265, 266, 430, 431 biological conversion to desoxyribonucleic acid, 437 cellular protein synthesis and, 438 distribution in cells, 438 function in cells, 437 location in cells, 437, 438 in microsomes, 432 RiCin, molecular weight, 263 source of, 263 toxicity of, 263 Roots, amino acid composition of proteina from, 240 hemoglobin in nodules of, 275
s Saccharomyces spp., amino acid composition of protein from, 251 Salmine, amino acid cornposition of, 98, 101, 102, 109, 112, 113, 114, 118, 129, 130 molecular weight of, 98, 128, 313 peptide nature of, 129 Salts, neutral, action on soluble proteins, 373 role in chrome tanning, 373, 384-386 Sarcosine, preparation of polymers of, 22
SUBJECT INDEX
Satgassum $uuilana, amino acid composition of protein from, 248 Satgassum Mtana, amino acid composition of protein from, 248 Schardinger enzyme, see Xanthine oxidase
seeds, proteins of, 242-246, 267, 268 amino acid composition'of, 243-246, 263-264 and nutritive value, 243-245, 246 distribution of, 245 heat of combustion of, 264 homogeneity of, 261 isolation of, 242-243 metabolism of, 277-281 molecular weights of, 263 physical properties of, 263 stability of, 261 toxicity of, 263 trypain-inhibiting activity of, 263 transaminaae activity in, 276 Selaginella unecinda, protein of, amino acid cornposition of, 247 isolation of, 246 Serine, derivatives of, 57 estimation in protein hydrolyzates, 115, 122 comparison of methods for, 96, 109 isolation of G,57 peptides of, 55ff. optical rotation of, 58 synthesis of, 55-58 aources of, 109, 130, 131, 132, 137, 140, 142, 143, 145 Serine anhydride, synthesis of, 7
-
sew,
relationship of milk proteins to proteins of, 223 Shrinkage, of animal skins, 369ff., 383 Silk fibroin, 309,359 amino acid composition of, 101, 102, 103, 106, 109, 110, 112, 113, 114, 116, 118-120, 143,329
479
complex formation with cupriethylenediamine, 324ff. fibers made from, 311, 335 from, and polyamides, 329 hydrolyzation producta of, 3, 57 molecular weight, 324 solubility of, 147 water uptake by, 356 X-ray diffraction of, 307 Skin, animal, see also Hide shrinkage of, 369ff. species differences in, 371 structure of, 354-355 Solanacae, amino acid composition of proteins in leaves of, 237 Soybeans, protein of, amino acid composition of, 244 fibers made from, 311, 315 of mixtures of, and polyamides, 326 molecular weight of, 263 trypsin-inhibiting activity of, 263 Soybean oil meal, effect of heat on nutritive value, 193 aa source of amino acids, 193ff. Spermatophytes, proteins of, 231-246 amino acid composition of, 258 SlaphyEoeoceus autew, amino acid composition of protein from, 252 Stseptoeoecwr fecalis, amino acid composition of protein from, 252 Streptogenin, effect on nutrition of mice, 194 Streptomyces griSeua, 251 amino acid composition of protein from, 252 Succinoxidaae, 431 Sulfo acids, condensed, see Sydtans Sunflower, amino acid composition of seed proteins of, 244 Swelling of fibrous proteins, 356, 365, 366 Syntans, nature of, 402 as tshning agents, 402405
480
BTJBJECT INDEX
T
Transamination, 20 in green plants, 271, 273 Tanning, Triose phosphate dehydrogenase, definition of, 354 amino acid composition of, 140 general aspecta of, 378-379 source of, 147 protein chemistry of, 353416 Tropomyosin, role of micellar reactions in, 360 amino acid composition of, 140 theory of vegetable tanning, 400-402 source of, 147 Tanning agents, 354,379ff. Trypsin, aldehydes as, 405-411 amino acids as substrate for, 60,64 chromium compounds aa, 379-392 inhibitory effect of soybean protein on, condensed sulfo acids (syntans) as, 263 402404 Tryptophan , effect on tensile strength, 362 biological isomerization of d-, 436 quinones 88,411-413 derivatives of, 71 reaction with hide proteins, 414 effect on body weight, 190 with modified collagens, 413 on nitrogen balance, 158, 159 vegetable, 394-396 nitrogen excretion, 184 molecular weights of, 395 estimation in protein hydrolyzates, properties of, 394-396 119, 121 reaction with collagen, 395 comparison of methods for, 117 factors governing, 396-340 fortification of casein hydrolyzate by, sources of, 395 159 Tannins, see Tanning agents, vegetable isolation from soybean oil meal, 194 Thallophytes, peptides of proteins of, 231, 247ff.,264-265 optical rotation, 71,72 amino acid composition of, 258 synthesis of, 11, 71, 72. 73 Thiazolidine4carboxylic acid methyl plasma protein output and, 185 ester, 75 sourcesof, 117, 124, 131, 137,140, 142, Thiazolones, 18 143, 145 Threonine, Tuberculin protein, 265 effect on body weight, 190 Tuberculosis, and plasma proteins, 181 estimation in protein hydrolyzates, Turnips, 115, 122 amino acid composition of protein comparison of methods for, 96, 110 from, 240 nitrogen metabolism and, 158, 184 Tyrocidine, 259 plasma proteins output and, 185 pcptidc nature of, 3 sources of, 110,115, 124, 131, 132,137, Tyrosine, 140, 142, 143, 145 derivatives of, 49,50 Tissues, estimation in protein hydrolyzntcs, 90, dehydropeptase in animal, 17 119 effect of protein depletion on, 182 Comparison of mctjhods for, 96, 116 Tobacco mosaic virus, peptides of, 46ff amino acid composition of, 101, 102, optical rotation of, 4G,51 103, 105, 106, 109, 110, 111, 112, synthesis of, 46,47 113, 114, 116, 117-120, 255, 256 sourcesof, 116,131, 137, 140,142, 143, effect on protein metabolism of host, 145 292-294 Tyrosine anhydride, Toluenesulfonylamino acids, stability of, 6 synthesie of peptides from 23-24, 57
48 1
SUBJECT INDEX
U Ulva ladwa, amino acid composition of protein of,
248 Urea, effect on soluble proteins, 399
V Valine, derivatives of, 37 effect on body weight, 190 on nitrogen equilibrium, 158 excretion, 185 estimation in protein hydrolyzates,
115ff. comparison of methods for, 96, 113 peptides of, 35ff. optical rotation, 38, 39 synthesis of, 40 sources of, 130, 131, 137, 140, 142, 143 Vanadium, effect on nitrogen assimilation, 274 ViNeeS, nucleoprotein nature of, 254 phytopathogenic, 254-256 amino acid composition of, 255,256-
258 and activity of, 259 crystalline, 255 effect on protein metabolism of host,
292-294
W Water, relation between protein levels and, in plants, 284 uptake of by fibrous proteins, 356 Wheat, proteins from, 244, 262 amino acid composition of, 244 Whey, composition of, 203 proteins of, 210-219 Wool, structure of, 329
X Xanthine oxidase, in cow’s milk, 220-221 molecular weight, 221 properties of, 221
Y Yeast, crystalline proteins from, 264 Z
Zein, amino acid composition of, 143, 264 fibers made from, 311, 314, 315, 325,
326 from mixtures of, and polyamides, 326 regeneration of liver proteins and, 881 solubility of, 147
This Page Intentionally Left Blank
