ADVANCES IN BIOCHEMICAL ENGINEERING Volume 6
Editors" T. K. Ghose, A. Fiechter, N. Blakebrough Managing Editor" A. Fiec...
143 downloads
1022 Views
9MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
ADVANCES IN BIOCHEMICAL ENGINEERING Volume 6
Editors" T. K. Ghose, A. Fiechter, N. Blakebrough Managing Editor" A. Fiechter
With 28 Figures
Springer-Verlag Berlin Heidelberg New York 1977
ISBN 3-540-08363-4 Springer-Verlag Berlin Heidelberg New York ISBN 0-387-08363-4 Springer-Verlag New York Heidelberg Berlin
This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically those of translation, reprinting, re-use of illustrations, broadcasting, reproduction by photocopying machine or similar means, and storage in data banks. Under § 54 of the German Copyright Law where copies are made for other than private use, a fee is payable to the publisher, the amount of the fee to be determined by agreement with the publisher. @ by Springer-Verlag Berlin • Heidelberg 1977 Library of Congress Catalog Card Number 72-152360 Printed in Germany The use of registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. Typesetting, printing, and bookbinding: Briihlsche Universitg.tsdruckerei GieBen. 2152/3140-543210
Editors Prof. Dr. T. K. Ghose Head, Biochemical Engineering Research Centre, Indian Institute of Technology Hauz Khas, New Delhi 110029/India Prof. Dr. A.Fiechter Eidgen. Techn. Hochschule, Mikrobiologisches Institut, Weinbergstrage 38, CH-8006 Ziirich Prof. Dr. N. Blakebrough University of Birmingham, Dept. Chemical Engineering, P.O.B. 363, Birmingham B15 2TT/England
Managing Editor Professor Dr. A.Fiechter Eidgen. Techn. Hochschule, Mikrobiologisches Institut, WeinbergstraBe 38, CH-8006 Ziirich
Editorial Board Prof. Dr. S. Aiba Biochemical Engineering Laboratory, Institute of Applied Microbiology, The University of Tokyo, Bunkyo-Ku, Tokyo, Japan
Prof. Dr. R. M. Lafferty Techn. Hochschule Graz, Institut fiir Biochem. Technol., Schl6gelgasse 9, A-8010 Graz Prof. Dr. M. Moo-Young
Prof. Dr. B. Atkinson University of Manchester, Dept. Chemical Engineering, Manchester / England
University of Waterloo, Faculty of Engineering, Dept. Chem. Eng., Waterloo, Ontario N21 3 G L / C a n a d a Dr. I. N~iesch
Dr, J.B6ing R6hm GmbH, Chem. Fabrik, Postf. 4166, D-6100 Darmstadt
Ciba-Geigy, K 4211 B 125, CH-4000 Basel
Prof. Dr. J.R.Bourne
Dr. L. K. Nyiri Dept. of Chem. Engineering, Lehigh University, Whitaker Lab., Bethlehem, PA 18015/USA
Eidgen. Techn. Hochschule, Techn. Chem. Lab., Universit~itsstraBe 6, CH-8006 Ziirich Dr. E. Bylinkina Head of Technology Dept., National Institute of Antibiotika, 3a Nagatinska Str., Moscow M-105 / USSR
Prof. Dr. H.J.Rehm Westf. Wilhelms Universitiit, Institut fiir Mikrobiologie, TibusstraBe 7--15, D-4400 M~nster Prof. Dr. P. L. Rogers
Prof. Dr. H.Dellweg Techn. Universit~it Berlin, Lehrstuhl fiir Biotechnologie, Seestral3e 13, D-1000 Berlin 65
School of Biological Technology, The University of New South Wales, PO Box 1, Kensington, New South Wales, Australia 2033
Dr. A.L.Demain
Prof. Dr. W. Schmidt-Lorenz
Massachusetts Institute of Technology, Dept. of Nutrition & Food Sc., Room 56-125, Cambridge, Mass. 02139/USA
Eidgen. Techn. Hochschule, Institut ftir Lebensmittelwissenschaft, Tannenstral3e 1, CH-8006 Ztirich
Prof. Dr. R.Finn School of Chemical Engineering, Olin Hall, Ithaca, NY 14853/USA
Prof. Dr. H. Suomalainen Director, The Finnish State Alcohol Monopoly, Alko, P.O.B. 350, 00101 Helsinki 10/Finland
Dr. K. Kieslich Schering AG, Werk Charlottenburg, Max-Dohrn-Strage, D-1000 Berlin 10
Prof. Dr. F. Wagner Ges. f. Molekularbiolog. Forschung, Mascheroder Weg 1, D-3301 St6ckheim
Contents
The Role of Thiobadllus ferrooxidans in HydrometaUurgicul Processes A. E, Torma, Socorro/New Mexico (USA)
Cellulase Biosynthesis and Hydrolysis of Cellulosic Substances T. K. Ghose, New Delhi (India)
39
Metabolism of Methanol by Yeasts H. Sahm, Braunschweig (Germany)
77
Control of Antibiotic Synthesis by Phosphate J. F. Martin, Salamanca (Spain)
105
The Role of ThiobaciUus ferrooxidans in Hydrometailurgical Processes Arpad E. Torma Department of Metallurgical and Materials Engineering, New Mexico Institute of Mining and Technology, Socorro, New Mexico 87801, USA
Contents 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Microbiological Background . . . . . . . . . . . . . . . . . . . . . . . . ~ ........... a) Morphology of T. f e r r o o x i d a n s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . b) Physiology o f T. f e r r o o x i d a n s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . c) Chemical Composition and Structure of T. f e r r o o x i d a n s . . . . . . . . . . . . . . . . . . d) Mechanisms o f Bacterial Action . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e) Kinetics o f Microbial Growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Factors Influencing Bacterial Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . a) Effect o f E n v i r o n m e n t . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . b) Effect o f pH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . c) Effect o f Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . d) Effect o f Nutrients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e) Effect o f Ferric Ion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . f) Effect o f Particle Size and Surface Area . . . . . . . . . . . . . . . . . . . . . . . . . . . g) Effect o f Regrinding o f the Leach Residue . . . . . . . . . . . . . . . . . . . . . . . . . . h) Effect o f Surface Active Agents and Organic Solvents . . . . . . . . . . . . . . . . . . . i) Effect o f Adaptation o f Bacteria to the Specific Substrate . . . . . . . . . . . . . . . . . 4. Microbiological Leach Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Microbiological Leaching o f Metal Sulfides . . . . . . . . . . . . . . . . . . . . . . . . . . . . a) Copper Sulfides Leaching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . b) Cobalt and Nickel Sulfides Leaching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . c) Zinc Sulfide Leaching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . d) Lead Sulfide Leaching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e) U r a n i u m Extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . f) Leaching o f Other Metal Sulfides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6. Biodegradation o f Non-Sulfide Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7. Different Aspects o f Bacterial Leaching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9. Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10. References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2 3 3 3 5
5 7 10 10 11 11 13 15 15 17 18 20 20 21 21 22 23 24 24 25 27 27 28 28 29
The present article illustrates the increased interest which is manifested in the microorganisms, Thiobacitlus f e r r o o x i d a n s , involved in the biohydrometatlurgical extraction processes. The wide varieties o f problems currently studied are very i m p o r t a n t in order to gain a better understanding about the factors which are governing the growth o f microorganisms, and as a consequence, the metal dissolution p h e n o m e n a . In several mining sites, the microbiological leaching techniques are currently practiced at industrial-scale, especially for recovery of copper and u r a n i u m from low-grade materials. However, an accurate assessment of further potential possibilities for the application o f
2
Arpad E. Torma
microorganisms in leaching metal sulfidesrequires a more fundamental knowledge about the interactions of the physical and chemical factors with the growth of T. ferrooxidans in pure and mixed cultures including heterotrophic and thermophilic cohabitants. Altogether, the future industrial exploitation of these microbiologicalleaching techniques are very attractive in many countries of the world.
1. I n t r o d u c t i o n In spite of the fact that bacterial oxidation of sulfide minerals has been occurring for centuries, microbiological leaching is only a recent development. The microorganisms, Thiobacillus ferrooxidans, responsible for this oxidation were first isolated in 1947 from the acid mine drainage of bituminous coal mines [ 1]. The presence of copper in mine drainage waters was observed by the Phoenicians, Romans, Arabs and Spaniards. The earliest leaching of copper from copper sulfide-bearing materials was recorded in 1970 at Rio Tinto in Spain [2]. However, the presence of bacteria in leach waters of the Rio Tinto mines was not confirmed until 1963 [3]. Dump leaching techniques were practised in the United States of America, Peru, Canada, Africa and in many other locations without knowing about the contribution of the microorganisms in these processes. The earliest report on microbiological leaching of metal sulfides was published in 1922, using some non identified autotrophic bacteria [4, 5] and suggesting that the biological treatment might be an economical way for the extraction of metals from low-grade sulfide-bearing ores. This idea was neglected for the next twenty five years until the discovery and characterization of the chemolithotrophic T. ferrooxidans [ 1 , 9 - I 1 ]. These bacteria can tolerate exceptionally high metal and hydrogen ion concentrations, for example, 120 g/1 of zinc [ 12], 72 g/1 of nickel [ 13], 30 g/1 of cobalt [ 13 ], 12 g/1 U3Oa [ 14], 55 g/1 of copper [15 ], 160 g/t of iron (II) [ 16] and an acid medium of pH 1.0 to 5.0 [ 17]. These facts are of considerable economic significance from metallurgical point of view and because, unlike many other fermentations, the bacterial leaching does not require an expensive sterilisation of the medium prior to inoculation. T. ferrooxidans are virtually ubiquitous. They can be found everywhere in nature, wherever an acidic environment is maintained in the presence of sulfide minerals [6-8]. The chemolithotrophic microorganisms [ 18] have the ability to utilize energy released from the metabolic oxidation of inorganic substrates [ 19, 20] such as reduced-valence inorganic sulfur compounds [21-23] and ferrous ion [24]. The chemical energy is converted by oxidative phosphorylation to ATP [25]. This is universally recognized to be the form of metabolic energy which can be utilized by the cell [26] for transportation work (substrate and nutrients into the cell and product out of the cell), mechanical work (muscle work for vibration and locomotion) and biosynthesis work (synthesis of cellular material). In this process the ATP is hydrolyzed to ADP and inorganic phosphate. These latter two species will be recombined into high energy carrier ATP in the followed-up respiration. The carbon metabolism by the chemolithotrophic bacteria may be either autotrophic, or facultative which represents a nutritional mode between the autotrophic and heterotrophic metabolism. The autotrophic capabilities of bacteria
The Rote of Thiobacillusferrooxidans in HydrometallurgicalProcesses
3
were established in 1887 [27] and can be defined as the ability to grow on strictly inorganic substrates providing energy for growth and carbon dioxide as the main source of carbon for the biosynthesis of cell materials [28, 30, 31 ]. The discovery of T. ferrooxidans, opened up an area of research which has had and will continue to have considerable economic significance. It represents a potential solution to the problem faced in many countries where continuing depletion of high-grade ore deposits has created a need to develop effective methods for recovering metals from low-grade sulfide ores. The microbiological leaching of metal sulfides can be defined as a biochemical oxidation process catalyzed by living organisms. However, only the insoluble sulfides are of commercial consequence. This process can be represented by the following simplified equation: MS + 2 02 microorganisms MSO4,
(1)
where M is a bivalent metal. When the oxidation product is insoluble, as it is the case, i.e., for the lead sulfide leaching, this fact can be used for selective leaching purpose [30] to separate the insoluble from the solubilized metals. The microbiological leaching processes involve complex interactions between the microorganisms, substrates and the nutrient concentrations, which are not yet completely understood. Altogether, a more economic use of these leaching processes require a better understanding of the various factors influencing bacterial growth and as a consequence, the microbiological metal dissolution processes.
2. Microbiological Background a) Morphology of T ferrooxidans
T. ferrooxidans possesses the following morphological characteristics: it is a motile, flagellated [32-34], non spore forming, Gram-negative, rod shaped (0.1 by 1.5/.tm) bacterium occurring single or occasionally in pairs [9, 10]. The growing cell goes through lag, log, stationary and death phases. When cell growth reaches about the double of a single cell size, it divides by binary fission. In the death cell, the mechanisms which regulate the permeability of the cell wall and cytoplasmic membrane do not function and the cell is plasmolyzed and broken down under the influence of the acid medium [35]. b) Physiology of T. ferrooxidans The microorganisms, T. ferrooxidans, derives the necessary energy for its life processes from oxidation of ferrous ion and of reduced-valence inorganic sulfur compounds and utilizes carbon dioxide for growth [33]. It is morphologically and, in some aspects, physiologically similar to T. thiooxidans, which is often present in acid mine drainage [36, 37]. The fundamental difference between the two species is generally recognized
4
Arpad E. Torma
to be the inability of T. thiooxidans to oxidize ferrous iron and insoluble metal sulfides [9, 10, 36]. Other bacteria have been identified from acid mine waters, oxidizing ferrous iron but not elemental sulfur or thiosulfate. It was considered to be a new genus and assigned the name of Ferrobaciltus ferrooxidans [38, 39]. Similarly, the name Ferrobacitlus sulfooxidans was assigned to a microorganism which utilized ferrous iron and elemental sulfur but not thiosulfate [40]. Subsequent investigations [41-45] indicated that the microorganisms (T. ferrooxidans, F. ferrooxidans and F. sulfooxidans) were identical and should be called T. ferrooxidans. All these organisms were capable of oxidizing elemental sulfur and thiosulfate in addition to ferrous ion [44]. The earlier, apparent fragmentation of the nomenclature and classification for this single species resulted from the use of different techniques in studying it. A new approach in the naming and classifying of bacteria is to reflect the manner in which present organisms are related by virtue of descent [53]. The increasing knowledge of comparative cytology and biochemistry led to the characterization of bacteria and blue-green algae as being procaryotic cells [54], possessing a simpler and an evolutionary more primitive structure than do all other cells (eucaryotic, i.e., possessing true mitotically dividing nuclei). The latest edition (8th) of the Bergey's Manual carries now the GC content of the DNA of each described nomenclatural type of organisms. The studies of DNA base composition appear to be the most beneficial when the cultures analysed are characterized by other biochemical means [55-57]. The DNA base composition of ferrous iron grown T. ferrooxidans, which belongs to the procaryote group of microorganisms, has been found to be in a narrow range of 56.0-57.0% GC [33]. However, recent studies [58, 59] on ferrous iron, chalcopyrite and lead sulfide grown T. ferrooxidans indicated 56.0, 60.1 and 54.4% GC respectively. The relatively important variations obtained for T. ferrooxidans grown on different substrate seemed to be not attributable to the analytical methods (melting temperature, cesium chloride density gradient centrifugation and ultraviolet absorbancy ratios) because of the good reproducibility of the results (5 t.0, 51.5, and 51.8% GC respectively) obtained for E. coli reference DNA. These variations cannot be explained by the adaption of T. ferrooxidans to the specific substrate. This problem is very complex and considerable caution must be exercised in extrapolating laboratory observations to microorganisms in their natural habitat [60]. Under natural conditions pure culture do not occur, growth rates are very slow by laboratory standards. In the development of heavy metal resistance in various organisms, it is difficult to decide whether adaptation, mutation, cohabitation or a combination of these is involved. However, the possibility exists that in these studies a kind of selection of microorganisms took place as suggested by other investigators [61-63] who isolated new species of bacteria from cultures of 7".ferrooxidans by changing environmental conditions and substrates. Data are available indicating that in the relatively strong sulfuric acid media [64] other (mixotrophic) microorganisms (algae, molds, protozoa and bacteria) than T. ferrooxidans can develop simultaneously. This view is supported by recent studies on microbial mutualism in ore leaching [65, 66], i.e., use of an aerobic, nitrogen fixing Beijerinckia lacticogenes in presence of T. ferrooxidans. Further, it has been pointed out [67], that considerable caution must be exercised when GC-data are compared from different laboratories.
The Role of Thiobacillusferrooxidans in HydrometallurgicalProcesses
5
c) Chemical Composition and Structure of T. ferrooxidans The first report [46] on cell composition of T. ferrooxidans indicated that it contained approximately 20% protein which consisted of 13 amino acids, and two B-vitamins: riboflavin and thiamine. These results have been refined [47] and the following cell composition has been obtained: 44% protein, 26% lipid, 15% carbohydrate, 10% ash, and at least, the two B-vitamins mentioned before. The cell structure of T. ferrooxidans has been found to be similar to that of other Gram-negative bacteria [47, 48]. The cell envelope, which is semi-permeable to nutrient [49, 50], appears to be composed of three osmophilic and three osmophobic layers, the total thickness measures 125 to 215 A [51, 52]. These layers are composed of lipoprotein, lipopolysaccharide, globular protein and peptidoglycan passing from outer to inner layers [51, 68]. The lipopolysaccharide layer consists of heptose, glucose, galactose, mannose and 2-keto-3-deoxyoctutosonate. Iron, mostly in the ferric form is associated with the lipopolysaccharide, suggesting it might serve as the initial binding site for the substrate. The peptidoglycan layer consists of glutaminic acid, a-e-diaminopimelic acid, alamine, glucosamine and muramic acid [69]. These two layers are of similar composition as those ofE. coli strains [70-72]. Phospholipids and neutral lipids are also related to the cell envelope structure [73]. The cytoplasma of T. ferrooxidans contains ribosomes, nuclear materials and cell inclusions [52, 68, 69]. At the present time, the relation between the structure and the function of cell envelope of T. ferrooxidans remains to be obscure. However, a better understanding of these problems would be very important from the point of view of evolution of chemolithotrophs and elucidation of oxidation pathways of insoluble inorganic sulfur substrates. d) Mechanisms of Bacterial Action The fact that T. ferrooxidans is capable of oxidizing ferrous ion and the reduced-valence forms of inorganic sulfur compounds, is an indication that it should have an enzymatic system similar to those of the iron and sulfur oxidizing bacteria. With regard to the mechanisms of metabolism of these substrates, there exist still different opinions over the basic concepts of oxidation pathways. There is agreement however, that the solid substrates must be rendered soluble before the bacterial oxidation could take place [ 17, 74] and the same time, nutrients have to be available in the environment of the contacted mineral surface. d. I) Oxidation of Inorganic Sulfur Compounds. The metabolism of inorganic sulfur compounds has been studied extensively [20, 21, 75, 76, 77, 80, 81]. The inorganic sulfide to sulfate oxidation can be represented by a simplified schema as follows: S -2
~
S +6 + 8 e - .
(2)
In this reaction 8 electrons are removed from the substrate and will be carried out through a series of intermediate products: Sx 2, SO -2, SO2-2, $20~-2, $2042, SxO~-2, SO~ 2, and so on. However, many of these products are highly unstable [78, 92] and probably
6
Arpad E. Torma
could not exist under physiological conditions. It has been proposed [79] that the first intermediate of sulfur oxidation is sulfite: SO+ 02 + H20 bacteria H~SO3
(3)
which is probably the key reaction in the pathways of the oxidation of sulfides, polysulfides or polythionates. On the basis of the available data [22, 82-85] the following scheme can be suggested for sulfides oxidation by T. ferrooxidans:
t6 S-2
t7
~ S0
1
~ 52032
2
1~ SO~"2
i, $4062
3
4
~, S042.
(4)
5
~8 9
$3062
]0
Reactions 1-4 are catalyzed by the sulfur-oxidizing enzyme [89, 83, 90, 91] where sulfite is the product of the reaction. Sulfite is oxidized to sulfate [85] by sulfite-oxidase (reaction 5) with the formation of ATP [88]. Thiosulfate is formed [84] by oxidation of elemental sulfur (reaction 7). It can be cleaved [82] by rhodanese (reaction 8) to sulfite and elemental sulfur. The thiosulfate is oxidized to tetrathionate (reaction 3) by the thiosutfate-oxidizing enzyme [84]. The reactions 8-10 may exist if T. ferrooxidans species behave like other Thiobacilli [86, 87]. d. 2) Oxidation of Ferrous Ion The oxidation of ferrous ion has been studied by many investigators [34, 93-99] using
T. ferrooxidans: Fe +2 ~ Fe +a + e-.
(S)
A mechanism of iron oxidation is proposed [24] in which the complexing of ferrous ion with molecular oxygen precedes the electron transport involving sulfate [ 100-102] and phosphatidylserine [73, 103] in the oxidation. The isolation of cytochrome c and cytochrome a from T. ferrooxidans [ 104] allowed the postulation of iron oxidation through the cytochrome system with oxygen as the final electron acceptor [96, 104-106]:
Fe +3
)
reduced cytochrome c
)
\ oxidized cytochrome a
)
(6) H20
It has been reported that the cytochrome c reductase is associated with DNA [107] and the nucleic acid involved was RNA [ 108]. Further, the iron-cytochrome c reductase was
The Role of Thiobacillusferrooxidans in HydrometallurgicalProcesses
7
inhibited by copper, nickel, chromium and mercury ions [ 107, 109]. The resting cell suspensions consume oxygen at high rate [93, 94, 96], however the highest rate is reported to be 21000 ~10~Jmg cell N/h [95]. The mean generation time on ferrous ion was reported to be 7.3 h [41].
e) Kinetics of Microbial Growth Studies of microbial growth involve periodic observations of bacterial growth, substrate utilization and product formation. Kinetic analysis makes the connection between metabolism and growth of the living organisms [26]. This fundamental information allows one to make maximum advantage of bacterial conversion, and also, suggests techniques which could lead to an increase in the yield of the desired product [123]. However, the biokinetic modeling is especially difficult because of the many different metabolic pathways and side reactions involved which are important to the tile of bacterial cell. Major complications in modeling come from the fact that many of the reaction mechanisms of the cell's metabolisms are not completely understood. Factors influencing bacterial growth are numerous, and the biological knowledge and mathematical tools necessary for the formulation and study of a completely general model do not exist [110]. The kinetic character of individual growth processes differs widely. However, certain characteristics typical for a group permit their classification in three different ways: phenomenological [ 11, 112], thermodynamic [ 113, 114] and kinetic [ 115]. The hypothesis that the enzyme (E) and the substrate (S) form a complex (ES) in enzyme catalyzed product (P) formation was originally derived in 1913 [ 116, 121]: k~ , , k3 E + S ~, /ES)-> E + P .
(7>
In batch systems where the substrate concentration is a limiting factor the growth rate (V) may be expressed [117] by:
V = Vm S/(K + S),
(8)
where Vm is the maximum growth rate and K constant. A plot of 1/Vversus 1/S gives a linear relationship [118] where the intercept is 1/Vm and the slope K/Vm. There exist many alternative forms [ 119, 120, 122] of the linearized Eq. (8). When sulfide minerals are used as substrate, S can be represented by the specific or total surface area [ 124]. With increasing fineness of the particles, and with all other nutrients in excess, the metal extraction rate, which is proportional to the growth rate [ 125], increases towards a limit. A more general form for bacterial growth limited by a single nutrient is given by the next equation [126]: V= Vm (I - exp (1 - S/K)).
(9)
8
Arpad E. Torma
When S/K is small the Eq. (9) becomes approximately equal to Eq. (8), which is known to be the hyperbolic rate equation relating the effect of a single limiting nutrient on the specific growth rate. However, in the most important processes this condition is not maintained and more than one substrate is used. For those cases Eq. (8) has to be modified. For example for the two-substrate reaction [127] it can be written: V = VmS~S2/(1 + K1S1) (1 + K2S2).
(10)
Similarly an equation has been developed for ternary complex formation [ 128] in which, ordered sequence of substrate addition is supposed. Despite the fact that most workers [117, 126, 129-132, 162] have regarded the specific growth rate of a population as a single function of the concentration of the limiting substrate, others were able to show that it is also a function of the population density [133,134] and the mass transfer or the assimilation processes [135]. These authors [133-135] also claim general applicability of their models to both batch and continuous systems. For a single stage continuous culture [ 182] the rate of bacterial growth, dX/dt, and substrate utilization, dS/dt, can be defined as follows: d X _ (la - D ) X dt
(11)
~
(12)
and = (So -- S) - D ~X y"
Under steady state conditions the specific growth rate,/~, is equal to the dilution rate, D, which is the reciprocal of the mean holding time. Hence the bacterial output and the substrate utilization can be given by: D X = Y(So - S) = Y ( S o - - K bLm C---D)
(13)
S= K
(14)
since D
/2m - D"
Further, the efficiency (r?) of substrate utilization can be expressed by: So-S
So_K _
~7- So
D
/1 m -- O
(15)
So
Equation (15) shows that the maximum substrate utilization could be obtained at low dilution rates, whereas the maximum bacterial growth respectively substrate consumption, will be achieved at high dilution rates. It has been shown [183] that for the oxidation of one mole of the limiting substrate,
The Role of Thiobacillus ferrooxidans in HydrometallurgicatProcesses
9
B moles of oxygen are required. The rate of oxygen uptake, dC/dt, may be expressed by: (I 6)
dC _ - B dS dt dr
It follows, therefore, that the rate of oxygen uptake is directly proportional to the growth rate. dC _ B d X _ B dt Y dt I£X.
(17)
For steady-state condition ~t = D and Y = x/(so - s).
(18)
Incorporating these terms we get ~-
BU(So
S) ~- BDSo.
(19)
S may be neglected since it is usually small as compared to So. The oxygen demand of a bacterial culture in a continuous system is directly proportional to the dilution rate and the concentration of substrate in the inlet medium. Some kind of inhibitory products of metabolism are always formed and accumulating during the growth processes. These will compete with the substrate for the active sites of the enzyme and could result in a diminution of product formation and the number of viable microorganisms. Several models are proposed [136, 137] to describe the relationship between the growth rate and the concentration of the toxic substances. Also, equations have been derived to fit the sigmoid type growth curves by exponential expressions [ 138-140] or by logistic exponential forms [ 141-154]. The product formation in product limited cultures [ 155] can be described by the following equation:
1 dP__l X di
dX+~
" X di-
u,
(20)
where X is cell number and a and b are constants, The first term on the right hand side is an expression for growth associated product formation and the second term for nongrowth associated product formation. Based on Eq. (20), a plot of the specific rate of product formation, 1 dP, versus the specific growth rate, ~ ~-'dXshould give a straight line, where b would be the intercept and a the slope of the regression line. Further models have been developed describing simultaneous effect of product and substrate [ 156, 157], cell age distribution [ 158-t61 ] and statistics of cell division [ 163171] on the growth or product formation rate. Growth curves have been approximated and described by mathematical expressions involving the cell age and size distribution [ 172-174] the cell composition (ribosome, protein and nucleic acid) [ 175-178] and allosteric enzyme actions [ 179--181 ].
10
Arpad E. Torma
As shown, the current literature contains many examples of mathematical models. These can be applied in description of many of the microbiological phenomena or they may be used to derive new expressions. However, in the choice of a model which quantitatively describes the biological phenomena, one should be certain that it has generality and predictive ability [184]. This mathematical model should further depend on what is already known about the system and on what type of results are expected.
3. F a c t o r s Influencing Bacterial Activity The metabolic activity of T. ferrooxidans growing on mineral sulfide is largely influenced by environmental factors. In this inorganic system, the available energy from the substrate is only in the form of electrons. A maximum rate of metal extraction can be achieved when the environment is maintained at the optimum leaching conditions. The problem is in turn an engineering one. It consists of transporting the limiting nutrient and substrate materials into the medium with a rate required for maximum bacterial growth. a) Effect of Environment The limits of bacterial activity of the natural environment have been studied [ 182, 185, 186, 188, 189] in terms of the pH and the oxidation reduction potential (Eh). This latter is defined by:
Eh = E ° + R nFT logK,
(21)
where K is the reaction equilibrium constant. The Eh-scale extends from +850 to - 4 5 0 mV, while the pH ranges from 1.0 to 10.2. However, these values do not represent the extreme limits [187], life still exists outside of them. A more sensitive indication of the effect of the environment on the activities of microorganisms is given by: rH2 = --log[all2]
(22)
where aH 2 is the activity of molecular hydrogen, which can be defined from the next equation: H2 ~ 2 H + + 2 e - .
(23)
Through application of Eq. (21) on Eq. (23) a relation between Eh and rH2 can be derived: rH2 _- 0 E. ~h
+ 2 pH at 25 °C.
(24)
The Role of Thiobacillusferrooxidans in HydrometallurgicalProcesses
11
The rH2 values may range from 0 to 41.7 and are directly related to Gibb's free energy change: AG = - n F AEh.
(25)
Further, AG is related to the equilibrium constant representing the metabolic conversion of substrate: AG = --RT log K.
(26)
The AG-values correspond to the maximum amount of energy which is available for the microorganisms in form of ATP, to do work. However, no fondamental studies exist concerning the relation between Eh, rH2, K, and AG for T. ferrooxidans. Such studies should be undertaken to gain a better understanding on the energetic phenomena associated with the microbiological leaching of metal sulfides.
b) Effect of pH The biological oxidation of ferrous ion and metal sulfides involves movements of hydrogen ions as well as of electrons. Therefore, the pH has a definite effect on their metabolism. The influence of pH on the activity of T. ferrooxidans has been studied by the majority of investigators [ 17, 190] in the range of 1 to 5. Optimum pH-values were reported to lie between 2.3 and 2.5 for chalcopyrite [3, 15, 191], zinc sulfide [192], chalcocite [193, 195], covetlite [193] and ferrous ion [194]. These values are derived in terms of shortest lag time, fastest rates of substrat oxidation or highest yield of metal extraction. The acidophilic character of T. ferrooxidans is shared by a number of microorganisms: T. thiooxidans [37], algae, molds, protozoa [64], yeast [196] and so on. However, no data are available to indicate the interactions which may exist between these microorganisms and T. ferrooxidans during the leaching of the sulfide-bearing ores.
c) Effect of Temperature The temperature range in which microbiological leaching of metal sulfides by T. ferrooxidans functions best, is limited. There are two competing factors to be considered in this process: the usual rise in the reaction rate with an increase in temperature (activation) and at the same time, an increase in the rate of thermal death of the microorganisms (inactivation) due to denaturation of the proteins [ 197, 198]. This denaturation reaction leads to a loss in the biological activity and as the temperature increases, the denaturation becomes much faster than the sulfide oxidation until the thermal death of the microorganisms occurs. The optimum temperature has been found to be in the range of 25 to 45 °C [199] for the different strains of T. ferrooxidans. For ferrous iron oxidation optimum values reported are 28 °C [200], 32 °C [40], and 35 °C [ 194, 201]. For metal sulfide leaching the optimum temperature is 35 °C [3,192, 193, 194, 202]. The biological oxidation ceases at around 55 °C and at higher temperatures only chemical oxidation occurs
12
Arpad E. Torma
[202]. No minimum temperature has been established for growth of T. ferrooxidans, but is is generally accepted that bacterial activity stops at the freezing point of the culture media. On the basis of its temperature optimum, 35 °C, this bacterium can be classified as a mesophilic organism, despite its soil origin. Soil organisms are generally psychrophilic. However, the temperature range for growth and reproduction of different microorganisms extends from - 1 8 ° to 104 °C [187]. These limits exceed those defining the stability field of pure water under one atmosphere, but they do not exceed the stability field of water in the liquid state when it is impure and under variable pressure. According to relatively recent studies [284-286, 290], the microbiological leaching of metal sulfides can be realized in the temperature range of 50 ° to 80 °C by thermophilic acidophilic thiobacteria [287-289]. These microorganisms are present in nature and are associated with the oxidation of reduced valence inorganic sulfur compounds. These bacteria present new possibilities for leaching, since the increase in the temperature may result in an increase in the rate of metal extraction. Further studies are needed to understand this phenomenon. A quantitative treatment of the effect of temperature can be given by the temperature coefficient [203]: 1o kl T~ e,o:
-
(27)
The Q~o-value for many chemical reaction is about 2; that is, the reaction rate doubles approximately for each 10 °C rise in the temperature. The respective Q~o-values for growth of T ferrooxidans on different substrates are given in Table 1. The dependence Table 1. Comparisonof Qlo-and Ea,i-values for oxidation of different substrates by Thiobaeillus
ferrooxidans Substrate
Temperature range
Qlo
ZnS Cu~S CuS FeSO4 FeSO4 CuS Cu~S
Ea, i
References
kcal/mole
°C
25-30 25-30 25-30 23-32 40-45 40-45 40-45
2.0 2.4 t.9 1.8 -
-12.0 -16.3 -11.7 -13.9 53.3 55.5 61.5
192 193 193 99 99 193 193
of reaction rate constant on temperature is merely an extension of the kinetic theory of matter. As the thermal energy increases, thereby increasing the electron spin, the concentration of molecules with this higher level of energy rises. That means, there are more molecules with an energy of activation [204], which can be expressed mathematically as follows: ga = - R din All
(k)
/Th
(28) "
The Role of Thiobacillus ferrooxidans in HydrometallurgicalProcesses
13
The temperature-dependence of Ea in enzymatic reactions has recently been discussed in detail [205]. The available literature data on activation energies are shown in Table 1.
d) Effect of Nutrients The nutrients requirements of T. ferrooxidans are normal for a chemosynthetic autotroph and a number of growth media are described in the literature [200, 206, 207]. The liquid media most frequently used, are compared in Table 2. This organism synthetizes its cell materials from inorganic sources [208] which are: carbon dioxide (as carbon source for cell growth), ammonium sulfate and dipotassium hydrogen phosphate (as nitrogen and phosphate sources) and potassium chloride, magnesium sulfate and calcium nitrate (as growth factors). Table 2. Liquid media for Thiobacillus ferrooxidans Components
Reference 206 in g
Reference 200 in g
Basal salts (NH4)2SO4 KC1 K2HPO, MgSO4 • 7 H~O Ca(NO3)~ Dist. H20 10 N H2SO4
0.05 0.05 0.05 0.50 0.01 1000 ml pH = 3.5
3.00 0.10 0.50 0.50 0.01 to 700 mt 1.00 ml
10 ml of a 10% W/V solution
300 ml of a 14.74% W/V solution
Energy source FeSO, • 7 H~O
Where IV/V = weight per volume. Studies on the effect of nutrient concentrations on the microbiological leaching of a zinc sulfide concentrate [ 192] indicated that ammonium concentration controlled the yield, while the phosphate concentration affected the rate of zinc extraction. The concentration of the growth factors had no detectable effect on substrate oxidation [192, 209]. These minor nutrients were required by the organisms in such small quantities, that any requirement beyond the amounts contained as impurities in the ammonium and phosphate salts or in the metal sulfide concentrate could not be demonstrated. In the acid medium of sulfide leaching, the solubility of oxygen and carbon dioxide is low, and a maximum rate of mass transfer of these gases is required to maintain a limitation free bacterial growth. According to Eq. (1) the transformation of one kg of sulfur of the sulfide to sulfate requires two kg of oxygen. When a flooding type of leaching would be practized with a 100% efficiency, this quantity of oxygen could be supplied in a volume of about 2800001. The solubility of oxygen is about 7 ppm at 35 °C. From this demonstration it is evident that cyclic leaching systems provided with agitation are preferred at the industrial scale. Therefore to take maximum advantage of the bacterial leaching process,
14
Arpad E. Torma
information is required about the oxygen mass transfer into the leach solution. It is known [210] that the oxygen mass transfer can be described by the Fick's laws for steady state diffusion through a static film [211 ]: (29)
kl = Dg/x I
for penetration at a laminar flow [212]: ! kl =2...(Dg/nO) 2
(30)
and for the surface renewal [213]: 1
kt = (Dgs) 2.
(31)
In most cases, oxygen is supplied by bubbling air into the nutrient media. A measure of the oxygen mass transfer performance of a given bioreactor may be expressed by the volumetric oxygen mass transfer coefficient, which can be evaluated experimentally [214, 216] using sodium sulfite solutions. This measure is based on the next equation: dC1 _ kla (C* - C~). dt
(32)
In order to achieve the highest oxygen transfer and thus the highest value of kla in a growing culture of microorganisms, the rate of agitation must be optimized [216]. For copper sulfide leaching, the rate of metal extraction can be related to dissolved oxygen concentration [217]: ~[M+21 = k [o2]n dT exp(zkE/R T)"
(33)
The fixation of carbon dioxide by the lithotrophic bacteria is a reduction process which requires energy. Therefore, this process would thermodynamically be impossible without the simultaneous assimilation of metabolic energy [218]. The Calvin cycle has been suggested [98,219,220] as a possible way for the carbon dioxide fixation by T. ferrooxidans. By studying the effect of carbon dioxide on bacterial activity, it has been observed [221 ] that the rate of pyrite oxidation by T. ferrooxidans gradually decreased upon removing the carbon dioxide from air used for aeration. Increasing the carbon dioxide content of air, stimulated the growth of T. ferrooxidans [222] using ferrous ion as the substrate. The best results were obtained for this oxidation process [43, 182] when the carbon dioxide content was increased to 2%. Other investigators [223] utilized an atmosphere containing approximately 0.1% carbon dioxide during bacterial leaching of metal sulfide concentrates. Studying the effect of solid substrate concentration (ZnS) at different carbon dioxide concentrations varying from 0.03 to 7.92%, the optimum C02 concentration has been found to be 0.2% [ 12] in terms of highest rate of zinc extraction. However, all these foregoing data of the effect of carbon dioxide on growth of T. ferrooxidans are preliminary in nature. A more appropriate approach to the
The Role of Thiobacillus ferrooxidans in Hydrometallurgical Processes
15
elucidation of this effect and determination of optimum CO2 concentration must be done similarly as it is outlined for the effect of oxygen concentration. Therefore, quantitative studies are sought relating the carbon dioxide mass transfer coefficient to the growth rate of T. ferrooxidans or to the rate of metal extraction. The ferrous ion oxidation by T. ferrooxidans requires the presence of sulfate ion [ 100, 102], probably as a complexing agent [24]. When it is cultured on elemental sulfur or metal sulfides, no addition of external sulfate is required, since the oxidation product is sulfate. e) Effect of Ferric Ion The microbiological leaching of metal sulfides is accelerated in presence of ferric ion, which is known to be an oxidizing agent [224]. This catalytic effect of iron can be expressed as follows: MS + Fe2(SO4)3 ~ MSO4 + 2 FeSO4 + S°.
(34)
Where M is a bivalent metal. The elemental sulfur which has been set free in Eq. (34) will be oxidized to sulfuric acid by T. ferrooxidans [44, 94]: SO+ 1½ 02 + H20 bacteria H2SO4.
(35)
Similarly, the ferrous iron is reoxidized by the microorganisms [200]: 2 FeSO4 + ½ 02 + H2SO4 bacteria Fe2(SO4)3 + H20
(36)
and then the iron redox cycle is repeated. Several investigators [64, 182, 191,226-228] discussed the effect of iron, although, quantitative determinations were reported only recently [ 13,193] dealing with the oxidation of analytic pure metal sulfides. These studies indicated that the microorganisms were able to oxidize the iron free sulfide substrates (NiS, CoS, Cu2S, and CuS) and in the presence of ferric iron concentrations, 10 -4 to 10 -2 mole/l, the rate of metal extractions more than doubled. Higher iron concentrations were not effective in this process. The redox potential of the ferrous/ferric couple at 25 °C [225] is
Eh = 0.771 + 0.0591 log aFe+3 aFe+2
(37)
Equation (37) indicates that solutions containing even one part of ferric ion per million parts of ferrous ion, have an oxidizing potential higher than +0.4 V, and consequently, these solutions could attack most metal sulfides. f) Effect of Particle Size and Surface Area Although metal sulfides are widely oxidized by the microorganisms [ 17, 76, 98, 2 2 8 233], the surface chemistry involved is known far from well. In aqueous suspensions of solid particles, the ions to be oxidized are supplied by the mineral surface. The concen-
16
Arpad E. Torma
tration of these ions is controlled by many phenomena, i.e., by the solubility product, the hydrolysis reactions, the presence of inert electrolytes, the surface tension, the particle size or surface area and the redox-potential. Thus, a better understanding of these phenomena and the interaction between them and the microorganisms is necessary to achieve a maximum advantage of the metabolic conversion of metal sulfides. One of the major requirement for the oxidation of metal sulfides by the bacteria is the availability of the substrate. The ideal condition exists when the substrate is soluble such as with ferrous sulfate. For insoluble substrates, another requirement is an adequate exposure of the sulfide minerals. It has been found [36] that ball milled samples of pyrite and marcasite were more susceptible to action of T. ferrooxidans than the coarser ones. By reducing the particle size of molybdenite from 280 to 230 gtm, an increase in the oxidation of the smaller particle size fraction has been observed [207]. Using different size fractions of a pyrite sample [ 191 ], the best results were obtained with the smallest particle size fraction of 42 gtm. These studies [ 191] emphasized the importance of the solid surface area upon the bacterial performance. Similar studies [234] reported an increase in the rate of oxidation of pyritic samples, upon decreasing the particle size from 220 to 42 gm. Using chalcopyrite concentrates, it was reported [3,235] that copper is more rapidly extracted from fine particles than from coarser ones and the best results were obtained with 42/lm size fraction. This finding was confirmed by other investigators [236, 237]. Using synthetic copper sulfide [64], it was reported that the rate of bacterial oxidation is almost doubled when the particle size is reduced from 104 to 64/~m size. The importance of particle size on bacterial leaching of chalcopyrite substrate has been discussed [ t 25, 182] and its effect demonstrated using subsieve fractions [238]. The biodegradation of a subsieve sphalerite concentrate has been studied [192] and the maximum rate of zinc extraction determined [239] to be 574.0, 565.7,569.8, and 569.2 mg • 1-1. h - i in terms of substrate concentration, specific surface area, total surface area and particle size, respectively. These values were in good agreement with 588.2 mg- 1- l • h - I , which was obtained for the oxidation of an analytically pure zinc sulfide [240]. These foregoing kinetic data were derived at normal air used for aeration. Increasing the carbon dioxide content, resulted zinc extraction rates as high as 1200.0 mg • 1-1 . h-1 and final zinc concentrations in solution of about 120.0 g. 1- l [ 12]. These kinetic values were confirmed by continuous culture of T. ferrooxidans on a zinc sulfide concentrate [241]. In the case of solid substrates, the microorganisms cannot attack the substrate in the interior until the outer material is dissolved. Therefore, the substrate concentration is synonymous with the particle size, specific surface area and total surface area expressions. Decreasing the size of the particles means increasing the particle specific surface area and the total surface area per unit volume of leach medium, without increasing the particle mass. On the other hand, increasing the pulp density (substrate concentration) means again an increase in the total surface area per unit volume but this time by increasing the particle mass. When low-grade ore is leached, the above considerations cannot always be applied [64]. Since the decrease in the particle size not only produces more sulfide surface but also a relatively larger surface of the accompanying host rock. In this case, the decrease in the particle size is synonymous with the dilution of the
The Role of Thiobacillus ferrooxidans in HydrometallurgicalProcesses
17
sulfide ore. In practice the optimum particle size has to be determined for each kind of ore to be leached. This size will be dictated by the relative economics which should be based on the two alternatives [30]: the gains resulted in the increase of the rate of metal extraction by decreasing the particle size and the grinding costs. For large scale in-situ operations, the exposure of the sulfide minerals can be accomplished by blasting of the ore body. The studies of nuclear underground explosion [242-248] indicated that fragmentation of large volumes of massive ore bodies could be realized cheaply and safely by this mean.
g) Effect of Regrinding of the Leach Residue During the microbiological leaching of naturally occurring metal sulfides, large quantities of iron are solublized which may partially precipitate and cover the surface of the solid particles. This precipitate layer then inhibits the further bacterial action [99, 199, 249,250]. The reaction rate of metal sulfides has its maximum value at the start when the maximum surface area is available for reaction. From there on, it decreases due to resistance which results from diffusion through the basic ferric hydroxide deposit and the reacted shell. According to the next hydrolysis reactions [251,252] the iron precipitate has the following composition: Fe2(SO4)a + 2 H20 ~ {2 Fe(0H)S04} + H2S04,
(38)
11 Fe2(SO4)s + 6 H20 ~ {H[Fe(SO4)2 • 2 Fe(OH)s]} + 2½ H2SO4,
(39)
Fe2(SO4)s + 6 H20 ~ {2 Fe(OH)s} + 3 H2SO4.
(40)
The above reactions remove iron from the solution by yielding sulfuric acid. The remaining total ferric iron concentration in solution is a function of the solubility constant of hydroxides and the hydrogen ion activity [253] and may be expressed by the following form [254]: [Fe +3 ] = [Fe +3 ] + [Fe(OH) +2] + [Fe(OH)~ ] + [Fe2(OH)~ 4] + [Fe(S04) +] + [Fe(S04)~-].
(41)
The iron chemistry involved in the microbiological leaching of metal sulfides has been very little studied and further investigations are needed to gain a better understanding about the role of these processes. The inhibitory effect of the iron precipitate resides in the fact that it inhibits the intimate contact between the microorganisms and the surface of the sulfide mineral (substrate) and the reaction will be controlled by the diffusion phenomena [250]. It has been suggested [250, 252] that these difficulties can be overcome and an almost complete extraction of metals can be achieved, when the leach residue is reground to expose new surfaces for releaching. Depending on the nature of substrate and the final yield of extraction projected, there are methods using one [ 13] or two [ 15] regrinding steps.
18
Arpad E. Torma
h) Effects of Surface Active Agents and Organic Solvents Surface active agents have a wide range of utility in their industrial applications as wetting agents, detergents and emulsifying agents. With respect to the opportunities for application to systems of biological interest, it was reported [255] that these compounds exhibit marked effectiveness in phenomena such as precipitation, complex formation and denaturation o f proteins, cytolysis o f cells, destruction o f microorganisms, and inactivation o f viruses. Despite o f these facts, a number o f studies were published [ 2 2 3 , 2 5 6 - 2 5 8 ] stating that the microbiological metal extraction was accelerated in presence o f the wetting agents, although, further investigations [250, 2 5 9 - 2 6 1 ] contradicted this effect o f the surfactants. The surface active agents reduce the surface tension o f the leach solution [ 2 6 0 - 2 6 5 ] and the oxygen mass transfer [210]. In normal culture media the surface tension is in the order o f 45 to 54 dyne/cm as compared with 72 for water [266]. When the surface tension is progressively reduced below 40 dyne/cm, bacterial growth is increasingly inhibited and below 30 dyne/cm only limited or no growth will take place [261,264]. F o r example, the decrease in the chalcopyrite oxidation ability of 7'. ferrooxidans is shown in Table 3 as a function o f the wetting agent
Table 3. Effect of wetting agents on chalcopyrite oxidation ability of Thiobacillusferrooxidans (260). Experimental conditions in Warburg apparatus: 200 mg of CuFeS~, 2 ml of basal salts nutrient medium (200) containing various concentrations of wetting agents and 0.3 ml of bacterial suspension containing 5,8 mg of protein (268) in a total volume of 2.3 ml Wetting agent concentration (ppm)
Specific rate of oxygen uptake (~10~. h-1 . mg-1 protein)a Tween 20
Tween 40
Tween 60
Tween 80
0 10 100 1000 10000 Steril controls (no Tween)
71.7 35.6 30.1 18.2 18.1
71.7 47.0 35.0 13.5 14.47
71.7 39,5 36.9 15.3 14.9
71.7 45.1 36.9 18.4 17.4
± 0.62 ± 0.27 ± 0,17 ± 0.12 -+0.12
5.6 ± 0.03
-+ 0.62 -+0.23 ± 0.23 ± 0.06 ± 0.10
5.66 ± 0.03
± 0.62 ± 0.23 ± 0.23 ± 0,06 ± 0.06
5.6 ± 0.03
-+ 0.62 +- 0.23 ± 0.17 ± 0.12 ± 0.15
5.6 ± 0.03
a Each figure is the mean value of three individual experiments ± SEM (standard error of the mean). concentration [260]. Similarly it was reported that the ferrous iron oxidation by T. ferrooxidans was depressed by sodium dodecylsulfate surfactant [267] and the covellite oxidation by Tergitot, Tween 20 and Triton X-100 surfactants [64]. An inhibitory effect has been observed for other surface active substances [ 2 7 2 - 2 7 4 ] , and flotation agents [275] on ferrous ion and copper sulfides oxidation. It is known [ 2 2 2 , 2 6 9 - 2 7 1 ] that the lithotrophic bacteria (T. thiooxidans and T. ferrooxidans) secrete certain chelating and wetting agents into the leach solutions. The microorganisms can probably cover their need for the surfactants and do not require any external addition o f these agents. All the before mentioned surface active substances (inhibitory organic compounds) may adsorb at the solid-liquid interface and affect the iron and sulfide oxidizing enzyme
The Role of Thiobacillus ferrooxidans in HydrometallurgiealProcesses
19
systems of bacteria, or they may react with (or disrupt) the bacterial envelope, thus influencing certain nutritive or growth processes. Further, the fact that the surface active agents may be adsorbed at the solid-liquid interface could be used to develop new techniques for separation of microorganisms [280-282] from the solid substrate in the aerated liquid media rather than to enhance bacterial activity. The concept of separation of particles in aqueous suspension by foaming, using the surface active agents, has long been known and applied in the field of mineral engineering, and is the basis for flotation processes [283]. The separation of microorganisms by flotation would potentially be important since this method may be relatively rapid and easy to apply. In addition, the fundamental investigations on this separation process may provide basic information concerning the structures and surface properties of the microorganisms and contribute to a better understanding of substrate assimilation processes. The microbiological leaching of ores and minerals can be represented schematically as shown in Fig. 1. In the solvent extraction, the dissolved metals are transferred during loading from the aqueous phase into the organic phase. After separation of phases, the
ORE LEACHING --i
RESIDUE
S/L .... I
2
SOLVENT 1
EXTRiCTtON
STRIPPING Fig. 1. Schematicalrepresentation of a microbiological cyclic leachingprocess
RECOVERYOFMETALS
aqueous phase is recycled to the leaching, the metals released from the organic phase by stripping and recovered, i.e., by electrolysis [276, 277]. The recycling solution will contain different amounts of organic matter depending upon the solubility of the solvents used and the mechanical losses due to entrainments, which may affect bacterial activity. Recent studies [278, 279] investigated this phenomenon and found that all commer-
20
Arpad E. Torma
cially available solvents used primarily for extraction of metals from aqueous leach solutions decrease the surface tension of the aqueous phase, and the chalcopyrite oxidation ability of Z ferrooxidans. The organic solvents aresimilar to, and probably act in the same way as, surfactants. Therefore, the recovery of metals by solvent extraction technique from the bacterial leach solutions must take into account the fact that the dissolved organic matter adversely affects bacterial activity and as a consequence, results in a diminution of solubilisation of metals from ore-leaching materials. These dissolved organics represent an economic loss and may produce pollution if these organic compound-containing solutions are released into the environment. To reduce these effects, it is recommended [279] that where solvent extraction is to be used to recover metals from the bacterial leach solutions, the aqueous recycling liquor be treated with activated charcoal or other inert absorbents to remove all organic matter (dissolved or entrained) before being returned to leaching or released in the environment.
i) Effect of Adaptation of Bacteria to the Specific Substrate The reproducibility of the results of microbiological leaching of metal sulfides requires adaptation of the microorganisms to the specific substrate prior to experimentation [291]. Generally, three to four transfers are necessary to achieve adaptation [292]. For subculturing and experiments an aliquot of a latedog-phase culture has to be used.
4. Microbiological L e a c h T e c h n i q u e s The earlier laboratory studies on the microbiological leaching of sulfide ores were carried out with air lift percolators [ 191, 221], Warburg respirometers [94] and with stationary leach bottles [235]. The oxygen supply is poor in both, the percolator and the stationary leach bottle techniques, whereas the size and the principle of the Warburg apparatus render it unsuitable for practical leaching. A method, which produces rapid aeration and an accelerated rate of leaching, is gyrotory shaking [236], using Erlenmeyer flasks. For example, using percolators for the bacterial leaching of chalcopyrite, 2.7% of copper is extracted from one sample and 6.1% from an other in 70 days [221]; and using the shake flask technique, 72% of copper was extracted in 12 days from the same type of ore [265]. This comparison shows the superiority of the shaking technique. However, besides gyrotory shaking, there are many other types of mixers available for use in laboratory leaching experiments, e.g., air spargers, magnetic stirrers and reciprocating shakers. The effect of these techniques on microbiological copper extraction was compared [293]. It was reported that magnetic stirring and reciprocating shaking gave results comparable with those for gyrotory shaking. Laboratory column leaching techniques [294] may simulate the commercial procedure for heap or dump leaching. A sample of ore is placed in a column and the liquid medium is circulated through it by air lift similarly as in percolators. One concurrent effect of this technique is to provide the column with oxygen and carbon dioxide saturated
The Role of Thiobacillus ferrooxidans in Hydrometallurgical Processes
21
medium. Although, because of the high oxygen requirement of the process, oxygen still may be a limiting factor. Another leach technique which may be used for microbiological metal extraction is the tank leaching technique [238,250]. This method is particularly useful for evaluating high-grade materials and provides for easy control of all the important parameters influencing this type of leaching. The interpretation of the results of tank leaching experiments also may contribute to a future acceptance of the bioleaching technique for recovery of metals from certain sulfide-bearing concentrates, which are presently recovered by conventional hydro- or pyrometallurgy. So far, only three different types of commercial scale microbiological leach techniques are practised for recovery of copper from low-grade ore. They are by dump leaching, heap leaching [189,295,296] and uranium by in-situ leaching [297-299]. In dump leaching the low-grade copper sulfide-bearing ore containing 0.1 to 0.4% copper, in most open mining sites is deposited by trucks on impermeable ground and shaped by bulldozers into a truncated cone. Larger dumps may be as high as 200 m, the diameter at the top 80 m and at the bottom 200 m. The leach solution is applied at the top, and after percolating through the ore, it is collected at the bottom. A typical pregnant solution may contain 1 to 3 g • 1-1 of copper, which is generally recovered by concentration on shredded scrap iron cans [300], CuSO4 + Fe ° ~ Cu ° + FeSO4
(42)
or by solvent extraction technique [276]. Heap leaching differs mainly from dump leaching in that it generally leaches finely dissiminated oxide-sulfide-bearing ores, which are unsuitable for concentration by flotation [296]. The leaching is carried out in beton tanks or vats equipped with a false bottom to allow the solution to percolate through the ore. These are usually arranged in countercurrent systems. Tanks having a capacity of 12 000 tons of ore are in common use.
5. Microbiological Leaching o f Metal Sulfides a) Copper Sulfides Leaching Chalcopyrite is a widespread mineral and one of the main source of copper in the world. Many investigators [ 1 5 , 1 9 1 , 2 0 7 , 2 2 6 , 2 3 8 , 2 9 5 , 3 0 2 , 3 0 7 - 3 1 0 ] studied the applicability of the microbiological leaching technique in the solubilisation and recovery of copper from chalcopyrite. The reactions involved in the biodegradation of copper sulfides are complex and many environmental conditions may effect bacterial activity. The microbiological leaching of chalcopyrite can be represented by the following electro-chemical reactions [301]: anodic reactions: 2 CuFeS2 + 16 H20 + H2SO4 ~ 2 Cu +2 + 2 Fe +3 + 5 SO~2 + 34 H + + 34 e - ;
(43)
22
Arpad E. Torma
cathodic reaction: 34 H + + 3 4 e - + 8½ 02 ~ 17 H20.
(44)
T. ferrooxidans catalyzes the above reactions (43) and (44), which are summarized in the next over-all equation: 2 CuFeS2 + 8½ 02 + H2SO, bacteria 2 CuSO4 + Fe2(SO4)3 + H20.
(45)
Pyrite and pyrrhotite associated with the chalcopyrite mineral are also oxidized by the microorganisms to ferric sulfate and sulfuric acid. The ferric sulfate contributes to the oxidization of chalcopyrite [303]: CuFeS2 + 2 Fe2(SO4)3 ~ CuSO4 + 5 FeSO4 + 2 S°.
(46)
The elemental sulfur and ferrous iron produced in reaction 46 are oxidized to sulfuric acid and ferric sulfate by the microorganisms. Recent studies [ 15,238,250, 301 ] on chalcopyrite concentrate leaching suggested the feasibility of this method for recovery of copper from chalcopyrite bearing concentrates on an industrial scale. The rates of copper release as high as 725 mg. 1-1 • h -1 were realized [238] and yields almost 100% have been obtained [15,250] when the leach residue was regrinded and releached. The oxidation mechanism for chalcopyrite [ 193,195], which is one of the most important copper mineral being leached in the southwestern United States, can be expressed by the following equation: Cu2S + ½ 02 + H2SO 4 bacteria CuS + CuSO4 + H20.
(47)
The CuS corresponding to the naturally occuring covellite, will be further oxidized [221,304, 305] by the microorganisms: CuS + 2 02 bacteria CuSO4.
(48)
Investigations [ 193,221,304, 306] on oxidation of chalcocite and covellite indicated that these substrates can be oxidized by T. ferrooxidans in complete absence of iron. However, addition of small amounts of ferric sulfate improved the rate of copper dissolution.
b) Cobalt and Nickel Sulfides Leaching The presence of cobalt in biological systems is well known [311 ], this metal being a key element of vitamin B12. The earlier literature [312] indicated that T. ferrooxidans is able to solubilize cobalt from sulfide-bearing minerals. Laboratory studies [256, 259] on the bacterial leaching of nickel sulfides resulted relatively high yield. This was attributed to the beneficial effect of surfactants. Further, it was found [313] that bacterial
The Role of Thiobacillusferrooxidans in HydrometallurgicalProcesses
23
activity is increasingly inhibited when nitrogen and phosphorous salts deficient leach media were used for nickel sulfide oxidation. The principal nickel-bearing sulfide mineral in many countries is pentlandite, in which the nickel is replaced by cobalt in the ratios of Ni:Co = 50:1 to 50:2. The bacterial oxidation of this mineral can be given [252] as follows: (Ni, Fe)gSa + 175 02 + 3¼ H2SO4 bacteria 4½ NiSO4
(49)
+ 2~- Fe2(S04)~ + 3¼ H20. Based on semi industrial-scale experiments using a pentlandite ore and concentrate a cyclic batch microbiological extraction process has been suggested for recovery of nickel and cobalt [314]. This method includes regrinding of the leach residue, after the nickel extraction reaches about 50 to 60%, to produce new substrate surface for bacterial action and the extraction of nickel is completed in a second leaching step. The rate of nickel extraction as high as 222 mg - 1- l • h -1 has been achieved. The microbiological leaching of a pentlandite concentrate has been investigated using percolators and T. ferrooxidans together with a nitrogen-fixing organism, Bei]erinekia laeticogenes [66]. It is claimed that during the leaching, T. ferrooxidans fixed atmospheric carbon dioxide for themselves as well as for the nitrogen fixing organisms and the nitrogen f'ixing bacteria fixed nitrogen to meet the requirements of both organisms in nutrient media devoid of any added carbon or nitrogen sources. This symbiotic association was found to be beneficial with respect to the metal extraction rate and yield. c) Zinc Sulfide Leaching For the oxidation of zinc sulfide the following overall equation can be written [240]: ZnS + 2 02
bacteria
• ZnSO4.
(50)
The capability of T. ferrooxidans to oxidize ZnS has been well demonstrated using naturally occuring sulfide-bearing minerals [ 191, 192,227,239,303, 316], or chemically prepared synthetic zinc sulfides [124,240, 301,317]. The semi-industrial studies resulted very fast zinc extractions of about 1300 g - 1-1 . h-1 [241] and dissolved zinc concentrations as high as about 120 g • 1-1 [12]. This latter leach solution was subjected to zinc electrowinning studies. Current efficiency was reported to be approximately 78%, which is somewhat lower than acceptable commercial levels (90%). It was suggested that minor modification of the purification procedure prior to electrowinning may overcome this deficiency. However, the quality of the deposited zinc plate was considered as satisfactory. The operating costs [318] for microbiological leaching of zinc sulfide concentrates including recovery of zinc by electrolysis were estimated to be comparable to an acid pressure leaching system.
24
Arpad E. Torma
d) Lead Sulfide Leaching During microbiological leaching, metal sulfides are oxidized to sulfates [291 ], which are, with the exception of lead sulfate, PbS + 2 02 .bacteria PbSO4
(51)
soluble in the aqueous nutrient media. This fact can be used to separate lead from solubilized metals such as cadmium, copper and zinc which always accompany an offgrade lead sulfide concentrate [30], which is not utilizable for lead recovery in a conventional smelting process. This material arises wherever lead, zinc, copper, iron, and cadmium occur together when the mineralisation consists of intergrowths of their sulfides and occasionally some graphite. Quantitative recovery of individual minerals is often rendered impossible because of fine mineralisation and differences in grindability, thus the grind for optimum liberation of chalcopyrite may result in over-grinding of galena. This will result substantial losses in the recovery of lead sulfide in flotation procedures. The economical viability of bacterial leaching of an off-grade lead sulfide concentrate will depend to a large extent upon the additional recovery of zinc, copper and cadmium as opposed to traditional smelting of the unleached concentrate. The only study available [30] indicated that during the oxidation about 50% of PbS is converted to PbSO4 while the zinc, copper, and cadmium contained in the concentrate were solubilized. Further, a process scheme is suggested for leaching and recovery of dissolved metals from a lead sulfide concentrate. The residue of this leaching is a highgrade lead concentrate, which can be used in a conventional smelting process for lead recovery.
e) Uranium Extraction The role of 7".ferrooxidans in the solubilization of uranium from its minerals is in oxidizing ferrous sulfate to ferric sulfate [as shown in Eq. (36)] which reacts with the uranium minerals [299,320-327] according to the following equation: UO2 + Fe2(SO4)3 ~ UO2SO4 + 2 FeSO4.
(52)
The tetravalent form of uranium present in the ore is insoluble in the leach solution while the oxidized hexavalent form is soluble [328-330]. Ferric sulfate and sulfuric acid can be produced by bacterial oxidation of iron sulfides such as pyrite [331,332] which is generally present in uranium ores, as expressed in the next reaction: 2 FeS2 + H20 + 7½ 02 bacteria Fe2(SO4)3 + H2SO4.
(53)
In the uranium leaching process, the ferric ion serves as the oxidizing agent [224] and is constantly regenerated by the microorganisms [99]. When all growth factors (temperature, pH, nutrients, and substrate) are optimized, the bacterial reproduction still can be limited by the availability of oxygen [216] as suggested by Eq. (53). Therefore, to
The Role of Thiobacillusferrooxidans in HydrometallurgicalProcesses
25
take maximum advantage of this uranium extraction process, information is required about the oxygen mass transfer into the leach solution. The applicability of bacterial leaching technique to the recovery of uranium from lowgrade ores and the influence of different parameters affecting these reactions have been investigated by numerous workers. 85% uranium extraction was realized [320] from an ore (particle size of about 6.3 mm) by T. ferrooxidans using percolation technique during a period of 20 weeks. Using similar leaching conditions it was observed [333] that the leaching time was directly proportional to the inital iron concentration. A 90% extraction of uranium was realized from an ore (particle size of about 4.0 mm) in the presence of iron concentrations varying from 0 to 1.0 g. 1-1 during 20 weeks. Other investigators [334], using different size of column type percolators, obtained 80 to 90% uranium recoveries from 0.5 to 79 kg of ore ground to a particle size of about 12.5 mm during 2 to 3 weeks. It was found [335,336] that T. thiooxidans is also effective in uranium extraction. These organisms were able to tolerate as high as 12.5 g - 1-1 uranium, while T. ferrooxidans endured up t o 12.0 g • 1-1 uranium oxide (U308) [238]. Studies on ferrous iron oxidation and uranium extraction from low-grade ores in continuous systems indicated that 80 to 90% uranium extractions can be achieved. f) Leaching of Other Metal Sulfides The microorganisms, T. ferrooxidans, were found to be able to oxidize orpiment, As2S3 [338], arsenopyrite, AsFeS [339], and enargite, Cu3(As, Sb)S4 [339]. The arsenopyrite oxidation is accomplished as a result of electrochemical reactions [342] and the overall reaction can be given by AsFeS + 3½ 02 + H20 bacteria FeAsO4 + H2SO4.
(54)
Selective leaching of arsenic, copper, tin and gold from complex sulfide-bearing concentrates was reported [340-342] and a flow-scheet suggested for treatment at an industrial scale. Bismuth extraction by microorganisms is accomplished by an indirect way [343] using ferric ion as an oxidizing agent: Bi2Sa + 6 Fe +3 + 6 02 ~ Bi2(SO4)3+ 6 Fe +2.
(55)
Then ferrous iron is reoxidized by T. ferrooxidans according to Eq. (5). The microbiological leaching of molybdenite can be obtained as follows [207]: 2 MoS2 + 9 02 + 6 H20 bacteria 2 H2MoO4 + 4 H2SO4.
(56)
This reaction was reported to be sevenfold faster in the presence of T. ferrooxidans than in sterile controls. The ability of T. ferrooxidans to oxidize antimony-bearing complex sulfide minerals [344, 345], low-grade stibnite [309] and synthetic antimony sulfides [291 ] has been reported. Evidence has been presented that the free energy released by the oxidation of
26
Arpad E. Torma
trivalent antimony to higher oxides suffices to support autotrophic growth [346]. This finding led to the discovery of a new antimony-oxidizing microorganism, Stibiobacter senarmontii [345]. It has been found that during the oxidation of stibnite by T. ferrooxidans the following reactions take place [292, 347]: Sb2S3 + 6 O2 bacteria Sb2(SO4)3.
(57)
The antimony(Ill) sulfate is partially hydrolyzed to produce an insoluble antimony(III) oxide sulfate: Sb2(504)3 + 2 H20 ~ (Sb0)2S04 + 2 H2SO 4.
(58)
Further, antimony(III) sulfate is partially oxidized to antimony(V) sulfate: Sb2(SO4)3 + 02 + 2 H2SO4 bacteria] (Sb02)2SO4 + 4 H2SO4.
(59)
The direct involvement of T. ferrooxidans in Eq. (59) could not be established. Further, the antimony(V) sulfate hydrolyzed to insoluble antimony(V) bioxide sulfate: 5b2(S04) s + 4 H20 ~ (8b02)2S04 + 4 H2SO4.
(60)
The total dissolved-antimony concentration was expressed by the following equation [292]: [Sb]tot = [Sb +3 ] + [Sb +5 ] + [SbO+ ] + [SbO~].
(61)
Vanadium was found to be oxidized by the bacteria [348] to its pentavalent state and precipitated in form of iron vanadate. The leaching is carried out in a series of vats and the precipitated iron vanadate is continuously removed from the last vessel. Microbiological leaching of low-grade manganese ores [349,350] were carried out using T. thiooxidans. The yield of manganese extractions varied between 70 and 99%. In spite of high toxicity of selenium [351 ], T. ferrooxidans were reported to be able to oxidize copper selenide [352]: CuSe + H2SO4 + 1 02 bacteria CuSO4 + H20 + Se°.
(62)
According to Eq. (62) the energy available corresponds to the removal of two electrons from the selenide. This represents 25% of the energy liberated in the oxidation of sulfides. Despite the similarity in chemical behavior of sulfur and selenium, their metabolic oxidation by T. ferrooxidans seems to be different. Further, unidentified autotrophic bacteria have been demonstrated to oxidize elemental selenium to selenic acid [353]. Fungi and species of Scopulariopsis, Penicillium and Aspergillus were also indicated to produce dimethylselenide from inorganic selenium compounds [354, 355].
The Role of Thiobacillusferrooxidans in HydrometallurgicalProcesses
27
6. Biodegradation of Non-Sulfide Materials The microbiological leaching of minerals is not limited to the action of Thiobacteria [356]. Different type of mineral degradation process may use intermediate metabolites, as, i.e., complexing agents which may bring metal into solution. For example, Agrobacterium tumefaciens [357,358] was reported to solubilize metalic gold up to about 1.5 ppm. It is suggested [342] that after adaptation, certain bacteria may dissolve gold from sands and ores up to about 15 g. 1-1. Similarly, decomposition of natural silicates such as wollastonite (CaSiO3), chabazite (CaA12Si4012 - 6 H20) and mineral phosphates has been realized [359] with Pseudomonas sp. This latter releases into solution ketogluconic acid which forms chelates with calcium, potassium, magnesium and other metals. Hence is the decomposition process. Fungal attack [360] on rocks may alter the rock structure and result their partial decomposition. Penicillium simplicissimum reported to solubilize titanium from granitic rocks [361] up to 80%. Reports are available [362] indicating that various fungi may reduce iron oxides. At the present time, the heterophic microorganisms appear to be less appropriate for commercial scale recovery of metals from minerals, than the Thiobacilli. However, further studies are needed to elucidate the possible interactions between the heterophic and the autotrophic microorganisms in symbiotic cultures. It is possible that the acidophilic sulfure-oxidizing bacteria [285,287] which are able to grow both heterotrophically and autotrophically and at increased temperatures, may be an asset in solving certain leaching problems.
7. D i f f e r e n t Aspects of Bacterial Leaching The current tendency in the mineral industry is still to exploit large and high-grade ore deposits, and to leave aside the low- or medium-grade ores. The reason for this is the relative lack in the versatility of the classical pyrometallurgical methods to recover economically the metals from lower grade sources [363]. These difficulties may be overcome by application of hydro- and/or biohydro-metallurgical methods. This latter is characterized by certain advantages over nonbiological conventional processes that it requires low capital investment and low operating costs [238]. It can be built and practised near the mining sites and used as a complementary treatment to the smelting process. The microbiological leaching technique does not require high temperature or pressure. It is easy to operate and control. It produces waste acid which can be used for treatment of oxide materials. The process does not contribute to air pollution. The microbiological leaching technique has great potential possibilities. However, there is no immediate prospect that it will compete with the smelting processes for conventional concentrates. Under certain conditions concentrate leaching could be practised economically in the following fields: - Off-grade complex sulfide-bearing concentrates which are not amenable to conventional metallurgical processes, such as finely dispersed lead and zinc ores, copper, arsenic, tin, silver and gold concentrates. - Concentrates produced in remote locations which are subject to high transportation
28
Arpad E. Torma
costs or, where the mineral deposit is small and consequently, no expensive installation is justified for conventional metal recovery process. After recovery of metals by cementation, the recycling solution contains considerable concentrations of iron and sulfuric acid. When solvent extraction is practised in connection with metal recovery, then in addition to iron and sulfuric acid this solution contains different concentrations of dissolved organic matter. In order to avoid aquatic or environmental pollutions, these solutions have to be treated if they are to be released into the environment. For example, the solutions have to be treated with calcium hydroxide to neutralize excess acidity and to precipitate iron in form ofjarosite type basic ferric hydroxide. The organic matter has to be removed by absorption on active carbon from the recycling solution or from the solution to be released into the streams in the environment.
8. C o n c l u s i o n The present study reports on the problems which are associated with the role of T. ferrooxidans involved in the biohydrometallurgical operations of sulfide-bearing materials. This method is presently practised commercially for recovery of copper and uranium from low-grade materials. However, possibilities are indicated where this method could be applied on an industrial scale for extraction of metals from sulfide-bearing concentrates. Microbiological leaching is influenced by a number of parameters and it functions best if carried out at optimum leaching conditions which can be summarized as follows: temperature 35 °C; pH 2.3; nutrients: ammonium sulfate 3 g - 1-1 and dipotassium hydrogen phosphate 0.5 g -1-1 carbon dioxide: 0.2% in air; oxygen: an intensive aeration is required to assure maximum oxygen mass transfer into the leach solution; mineral exposure: the smallest particle size of the solid sulfide substrate will assure the highest rate and yield of metal extraction; Eh must be kept below 500 mV in order to avoid jarosite type, basic ferric hydroxide precipitation on the surface of the soil substrate; and leach technique: wherever possible, a cyclic leaching process should be applied.
9. N o m e n c l a t u r e a aH 2
b B
cl C*
dC/dt dIM+2]/dt dP/dt dS/dt dX/dt
constant activity of molecular hydrogen, m g . 1- l constant number of mole of 02, mg. 1-1 actual concentration of dissolved liquid, mg - 1-1 concentration of dissolved oxygen at saturation, mg. 1-1 rate of oxygen uptake, mg. 1-1 . h - 1 rate of metal extraction, mg - 1-1 . h rate of product formation, mg • 1-1 - h -1 rate of substrate utilisation, mg • 1-1 -h -1 bacterial growth rate, mg • 1-1 . h - 1
The Role of Thiobaciltus ferrooxidans in Hydrometallurgical Processes
D Dg E Eo
E~ Eh F AG k kl, k2, k3 kl kLa K, KI, K2 n
P Qlo R $
so, s, s~, s2 T, TI, T2 V
Vm x1
X Y At Atm
~7 0
29
dilution rate, h -1 diffusivity of dissolved gas in water, cm 2. s- 1 or m 2. h - 1 enzyme, mg • 1-1 normal electrode potential, volt energy of activation, kcal • m o l e - 1 redox potential, volt Faraday number, coulomb Gibb's free energy change, kcal • mole-1 proportionality constant reaction rate constants, mg • 1- ! • h -1 oxygen mass transfer in Eqs. (29)-(31), m g . 1-1 • h - l volumetric oxygen mass transfer coefficient, h - 1 ½ V m , m g - 1 - I - h -1 number of electrons [Eq. (21)] or proportionality constant [Eq. (33)] product, mg • 1-1 temperature coefficient gas constant, cal - deg- 1 . m o l e - 1 surface renewal rate, h - 1 substrates, mg• 1-1 absolute temperatures, °K rate of growth or rate of metal extraction, nag • 1-~ • h - i maximum value of V, mg• 1-1 . h - x effective thickness o f surface film, cm cell number or concentration, mg yield, g bacteria/g substrate specific growth rate, h - 1 maximum value o f At, h-1 efficiency of substrate utilization, % time, h
10. R e f e r e n c e s 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.
Colmer, A. R., Hinkle, M. E.: Science 106, 253 (1947). Taylor, J. H., Whelan, P. F.: Trans. Inst. Mining Met. 52, 36 (1943). Razzell, W. E., Trussell, P. C.: J. Bacteriol. 85, 595 (1963). Rudolfs, W.: Soil. Sci. 14, 135 (1922). Rudolfs, W., Helbronner, A.: Soil. Sci. 14, 459 (1922). Joseph, J. M.: Ohio J. Sci. 53,123 (1953). Lorenz, W. C.: U. S. Bur. Min. Rept. Inv. 8080 (1961). Ito, I. Suiyokwai-Shi (Japan) 15,291 (1964). Colmer, A. R., Temple, K. L., Hinkle, M. E.: J. Bacteriol. 59, 317 (1950). Temple, K. L., Colmer, A. R.: J. Bacteriol. 62, 605 (1951). Colmer, A. R.: J. Bacteriol. 83, 761 (1962). Torma, A. E., Walden, C. C., Duncan, D. W., Branion, R. M. R.: Biotechnol. Bioeng. 14, 777 (1972). Torma, A. E.: Canadian Patent no. 960463, 1975 and English Patent No. 1382357, 1975.
30 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61.
Arpad E. Torma Duncan, D. W., Bruynesteyn, A.: Trans. Can. Min. Met. Bull. 74, 32 (1971). Sakaguchi, H., Silver, M., Torma, A. E.: Biotechnol. Bioeng. 18, 1091 (1976). Chernyak, A. S., Mineev, G. G.: Sovjet J. Non-Ferrous Metals 7, 28 (1966). Silverman, M. P., Ehrlich, H. L.: Adv. Appl. Microbiol. 6, 153 (1964). Zavarzin, G. A.: Lithotrophic Bacteria. Moscow: Nauka 1972. Kelly, D. P.: Ann. Rev. Microbiol. 25, 177 (1971). Peck, Jr., H~ D.: Ann. Rev. Microbiol. 22, 489 (1968). Peck, Jr., H. D.: Bacteriol. Rev. 62, 67 (1962). Sinka, D. B., Walden, C. C.: Can. J. Microbiol. 12, 1041 (1966). Roy, A. B., Trudinger, P. A.: The Biochemistry of Inorganic Compounds of Sulphur. Cambridge: University Press 1970. Dugan, P. R., Lundgren, D. G.: J. Bacteriol. 89, 825 (1965). Burton, K.: Nature 181, 1594 (1958). Torma, A. E., Panneton, J. J.: Introduction to Bacterial Kinetics and Modeling. No. S-129, Quebec: Quebec Official Publisher 1973. Winogradsky, S.: Botan. Ztg. 45,489 (1887). Starkey, R. L.: Bacteriol. Rev. 62, 142 (1962). Umbreit, W. W.: Bacteriot. Rev. 62, 145 (1962). Torma, A. E., Subramanian, K. N.: lnternat. J. Miner. Process. 1, 125 (1974). Lamanna, C., Mallette, F.: Basic Bacteriology. Baltimore: The Williams & Wilkins Co. 1965, Breed, R. S., Murray, E. G. D., Smith, N. R.: In Bergey's Manual of Determinative Bacteriology. 7th Ed., p. 78. Baltimore: The Williams & Wilkins Co. 1957. Vishniac, W. V.: In Bergey's Manual of Determinative Bacteriology. 8th Ed., p. 456. The Williams & Wilkins Co. 1974. Beck, J. V., Elsden, S. R.: J. Gen. Microbiol. 19, 1 (1958). Karavaiko, G. I., Avakyan, A. A.: Microbiologiya 39, 950 (1970). Temple, K. L., Delchamps, E, W.: Appl. Microbiol. 1,255 (1953). Tuttle, J. H., Randles, C. I., Dugan, P. R.: J. Bacteriol. 95, 1495 (1968). Leathen, W. W., Braley, S. A.: Bacteriol. Pract. p. 44 (1954). Leathen, W. W., Kinsel, N. A., Braley, S. A.: J. Bacteriol. 72, 700 (1956). Kinsel, N. A.: J. Bacteriol. 80, 628 (1960). Unz, R. F., Lundgren, D. G.: Soil. Sci. 92, 302 (1961). Ivanov, V. I., Lyalikova, N. N.: Mikrobiologiya 31,382 (1962). Beck, J. V., Shafia, F. M.: J. Bacteriol. 88, 850 (1964). Hutchinson, M. Johnstone, K. I., White, D.: J. Gen. Microbiol. 44, 373 (1966). Kelly, D. P., Tuovinen, O. H.: Int. J. Syst. Bacteriol. 22, 170 (1972). Lundgren, D. G., Andersen, K. J., Remsen, C. C.: Bacteriol. Proc. p. 25 (1962). Lundgren, D. G., Andersen, K. J., Remsen, C. C., Mahoney, R. P.: Dev. Ind. Microbiol. 6, 250 (1964). Holt, S. C., Shiveley, J. M., Greenawalt, J. W.: Can. J. Microbiol. 20, 1347 (1974). Mitchell, P.: Nature 180, 134 (1957). Csaky, T. Z.: Ann. Rev. Physiol. 27,415 (1965). Remsen, C. C., Lundgren, D. G.: J. Bacteriol. 92, 1765 (1966). Avakyan, A. A., Karavaiko, G. t.: Mikrobiologiya 39, 855 (1970). Mandel, M.: Ann. Rev. Microbiol. 23,239 (1969). Stanier, R. Y., van Niel, C.: Arch. Mikrobiol. 42, 17 (1962). Mandet, M.: J. Gen. Microbiol. 43,273 (1966). Stanier, R. Y., PaUeroni, N. J., Doudoroff, M.: J. Gen. Microbiol. 43, 159 (1966). Redfearn, M. S. Palleroni, N. J., Stainer, R. Y.: J. Gen. Microbiol. 43,293 (1966). Guay, R., Silver, M., Torma, A. E.: IRCS Med. Sci. Libr. Compend. 3, 417 (1975). Guay, R., Silver, M., Torma, A. E.: Rev. Can. Biol. 35, 61 (1976). Sadler, W. R., Trudinger, P. A.: Miner. Depos. 2 , 1 5 8 (1967). Balashova, V. V., Vedenina, I. Y., Markosyan, G. E., Zavarzin, G. A.: Mikrobiologiya 43, 581 (1974).
The Role of Thiobacillusferrooxidans in Hydrometallurgieat Processes 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73.
Lyalikova, N. N.: Mikrobiologiya 42, 941 (1974). Guay, R., Silver, M.: Can. J. Microbiol. 21,281 (1975). Ehrlich, H. L., Fox, S. I.: Biotechnol. Bioeng. 9, 471 (1967). Tsuchiya, H. M., Trivedi, N. C., Shuler, M. L.: Biotechnol. Bioeng. 16, 991 (1974). Trivedi, N. C., Tsuchiya, H. M.: Int. J. Min. Process. 2, t (1975). Gasser, F., Mandel, M.: J. Bacteriol. 96, 58 (1968). Wang, W. S., Korczynski, M. S., Lundgren, D. G.: J. Bacteriol. 104, 556 (1970). Wang, W. S., Lundgren, D. G.: J. Bacteriol. 95, 1851 (i968). Primosigh, J., Pelzer, H., Maass, D., Weidel, W.: Biochem. Biophys. Aeta. 46, 68 (1960). Mandelstam, J.: Biochem. J. 84, 294 (1962). Osborn, M. J.: Ann. Rev. Biochem. 38, 501 (1969). Korczynski, M. S., Agate, D. A., Lundgren, D. G.: Biochem. Biophys. Res. Commun. 29, 457 (1967). 74. Vogler, K. G., Umbreit, W. W.: Soil Sci. 51,331 (1941). 75. Vishniac, W., Santer, M.: Bacteriol. Rev. 21,159 (1957). 76. Trudinger, P. A.: Rev. Pure Appl. Chem. 17, 1 (1967). 77. Kelly, D. P.: Aust. J. Sci. 31,165 (1968). 78. Szekeres, L.: Talanta 21, 1 (1974). 79. Suzuki, I., Silver, M.: Biochim. Biophys. Acta 122, 22 (1966). 80. Lees, H.: Ann. Rev. Microbiol. 14, 83 (1960). 81. Suzuki, I.: Ann. Rev. Microbiol. 28, 85 (1974). 82. Tabita, R., Silver, M., Lundgren, D. G.: Can. J. Biochem. 47, 1141 (1969). 83. Silver, M., Lundgren, D. G.: Can. J. Biochem. 46, 457 (1968). 84. Silver, M., Lundgren, D. G.: Can. J. Biochem. 46, 1215 (1968). 85. Vestal, J. R., Lundgren, D. G.: Can. J. Biochem. 49, 1125 (1971). 86. Vishniac, W.: J. Bacteriol. 64, 363 (1952). 87. Trudinger, P. A.: Biochem. J. 90, 640 (1964). 88. Adapoe, C., Silver, M.: Can. J. Microbiol. 21, 1 (1975). 89. Margalith, P., Silver, M., Lundgren, D. G.: J. Bacteriol. 92, 1706 (1966). 90. Gleen, H., Quastel, J. H.: Appl. Microbiol. 1, 70 (1953). 91. Pankhurst, E. S.: J. Gen. Microbiol. 34, 427 (1964). 92. Pollard, F. H., Jones, D. J., Nickless, G.: J. Chromatog. 15, 393 (1964). 93. Silverman, M. P., Lundgren, D. G.: J. Bacteriol. 78, 326 (1959). 94. Beck, J. V.: J. Bacteriol. 79, 502 (1960). 95. Landesman, J. E., Duncan, D. W., Walden, C. C.: Can. J. Microbiol. 12, 25 (1966). 96. Tikhonova, G. W., Lisenkova, L. L., Doman, N. G., Skulachev, V. P.: Biokimiya 32, 725 (1967). 97. MacDonald, D. G., Clark, R. H.: Can. J. Chem. Eng. 48, 669 (1970). 98. Tuovinen, O. H., Kelly, D. P.: Z. Allg. Mikrobiol. 12, 311 (1972). 99. Guay, R., Torma, A. E., Silver, M.: Ann. Microbiol. (Inst. Pasteur) 126B, 209 (1975). 100. Lazaroff, N.: J. Bacteriol. 85, 78 (1963). 101. Dugan, P. R., Lundgren, D. G.: Bacteriol. Proc. p. 92 (1964). 102. Tuovinen, O. H., Kelley, B. C., Nicholas, D. J. D.: Arch. Microbiol. 105, 123 (1975). 103. Agate, A. D., Vishniac, W.: Bacteriol. Proc. p. 50 (1970). 104. Vernon, L. P., Mangum, J. H., Beck, J. V., Shafia, F. M.: Arch. Biochem. Biophys. 88, 227 (1960). 105. Blaytock, B. A., Nason, A.: J. Biol. Chem. 238, 3453 (1963). 106. Yates, M. G., Nason, A.: J. Biol. Chem. 241, 4872 (1966). 107. Yates, M. G., Nason, A.: J. Biol. Chem. 241, 4861 (1966). 108. Din, G. A., Suzuki, I.: Can. J. Biochem. 45, 1547 (1967). 109. Din, G. A., Suzuki, I., Lees, H.: Can. J. Biochem. 45, 1523 (1967). 110. Tsuchiya, H. M., Fredrickson, A. G., Aris, R.: Adv. Chem. Eng. 6, 125 (1966). 11 t. Gaden, Jr., E. L.: Chem. Ind. (London) p. 145 (1955). 112. Maxon, W. D.: Appl. Microbiol. 3, 110 (1955).
31
32 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. t36. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 165.
Arpad E. Torma Deindoerfer, F. H.: Adv. Appl. Microbiol. 2, 321 (1960). Calam, C. T., Driver, N., Bowers, R. H.: J. Appl. Chem. 1,209 (1951). Deindoerfer, F. H., Humphrey, A. E.: Ind. Eng. Chem. 51,809 (1959). Michaelis, L., Menten, M. L.: Biochem. Z. 49, 333 (1913). Monod, J.: Ann. Rev. Microbiol. 3, 371 (1949). Lineweaver, H., Burk, D.: J. Ann. Chem. Soc. 56,658 (1934). Hofstee, B. H. J.: Science 116, 329 (1952). Hofstee, B. H. J.: Enzymol. 17, 273 (1956). Konak, A. R.: Biotechnol. Bioeng. 17, 1551 (1975). Williamson, K. J., McCarty, P. L.: Biotechnol. Bioeng. 17, 915 (1975). Topiwala, H. H.: Biotechnol. Bioeng. Symp. No. 4,681 (1974). Torma, A. E., Legault, G., Kougiomoutzakis, D., Ouellet, R.: Can. J. Chem. Eng. 52, 515 (1974). MeGoran, C. J. M., Duncan, D. W., Walden, C. C.: Can. J. Mierobiol. 15, 135 (1969). Spicer, C. C.: Biometrics 11,225 (1955). Laidler, K. J., Socquet, I. M.: J. Phys. Colloid Chem. 54, 530 (1950). Segal, H. L., Kachamar, J. F., Boyer, P. H.: Enzymol. 15, 187 (1952). Novick, A., Szilard, L.: Science 112, 715 (1950). Hebert, D., Elsworth, R., Telling, R. C.: J. Gen. Microbiol. 14, 601 (1956). Novick, A., Szilard, L.: Proc, Nat. Acad. Sci. 36, 708 (1950). Hebert, D.: Soc. Chem. Ind. 12, 21 (1961). Contois, D. E.: J. Gen. Microbiol. 21, 40 (1959). Fujimoto, Y.: J. Theoret. Biol. 5,171(1963). Verhoff, F. H., Sundaresan, K. R., Tenney, M. W.: Biotechnol. Bioeng. 14, 411 (1972). Reid, A. E.: Bull. Math. Biophys. 14, 313 (1952). Finn, R. K.: J. Ferm. Technol. 44, 305 (1966). Schumacker, F. X.: J. Forest. 37, 819 (1939). Winsor, C. P.: Proc. Nat. Acad. Sci. 18, 1 (1932). yon Bertalanffy, L.: Hum. Biol. 10, 181 (1938). yon Bertalanffy, L.: Biol. Zentralbl. 61,510 (1941). Dewitt, C. C.: Ind. Eng. Chem. 35,695 (1943). Stevens, W. L.: Biometrics 7,247 (1951). Hartley, H. O.: Biometrika 35, 32 (1948). Pimentel-Gomes, F.: Biometrics 9, 498 (1953). Patterson, H. D.: Biometrics 12, 323 (1956). yon Bertalanffy, L.: Quart. Rev. Biol. 32, 218 (1957). Finney, D. J.: Biometrika 45,370 (1958). Patterson, H. D.: Biometrika 45,389 (1958). Patterson, H. D., Lipton, S.: Biometrika 46, 281 (1959). Hartley, H. O.: Biometrika 46, 293 (1959). Richard, F. J.: J. Exper. Bot. 10, 290 (1959). Grosenbaugh, L. R.: Biometrics 21,708 (1965). Edwards, V. H., Wilke, C. R.: Biotechnol. Bioeng. 10, 205 (1968). Luedeking, R., Piret, E. L.: J. Biochem. Microbiol. Techn. Eng. 1,393 (1959). Chen, J. W., Koepsell, H. J. Maxon, W. D.: Biotechnol. Bioeng. 4, 65 (1962). Maxon, W. D., Chen, J. W.: J. Ferm. Technol. 44,255 (1966). Powell, E. O.: Biometrika 42, 16 (1955). Shu, P.: J. Biochem. Microbiol. Technol. Eng. 3, 95 (1961). Maruyama, Y., Hayashi, K.: J. Ferm. Technol. 44, 227 (1966). Kobayashi, J.: J. Ferm. Technol. 44, 233 (1966). Dabes, J. N., Finn, R. K., Wilke, C. R.: Biotechnol. Bioeng. 15, 1159 (1973). Kelly, C. D., Rahn, O.: J. Bacteriol. 23, 147 (1932). Rahn, O.: J. Gen. Physiol. 15, 257 (1931-32). Finney, D. J., Martin, L.: Biometrics 7, 133 (1951).
The Role of Thiobacillus ferrooxidans in Hydrometallurgical Processes 166. 167. 168. 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180. 181. 182. 183. 184. 185. 186. 187. 188. 189. 190. 191. 192. 193. 194. 195. 196. 197. 198. 199. 200. 201. 202. 203. 204. 205. 206. 207. 208. 209. 210. 211. 212. 213. 214.
33
Kendall, D. G.: Biometrika 35, 316 (1948). Powell, E. O.: J. Gen. Microbiol. 15,492 (1956). PoweU, E. O.: J. Gen. Microbiol. 18, 382 (1958). Campbell, A.: Bact. Rev. 21,263 (1957). Koch, A. L., Schaecter, M.: J. Gen. Microbiol. 29, 435 (1962). Schaecter, M., Williamson, J. P., Hood, Jr., J. R., Koch, A. L.: J. Germ. Microbiol. 29, 421 (1962). Fredrickson, A. G., Tsuchiya, H. M.: Am. Inst. Chem. Eng. J. 9, 459 (1963). Benzer, A.: Biochem. Biophys. Acta 11,383 (1953). Terui, G.: J. Ferm. Technol. 44, 242 (1966). Yeisley, W. G., Pollard, E. C.: J. Theoret. Biol. 7,485 (1964). Swanson, C. H., Aris, R., Fredrickson, A. G., Tsuchiya, H. M.: J. Theoret. Biol. 12, 228 (1966). Ramkrishna, D , Fredrickson, A. G., Tsuchiya, H. M.: J. Ferm. Technol. 44, 203 (1966). Ramkrishna, D., Fredrickson, A. G., Tsuchiya, H. M.: Biotechnol. Bioeng. 9, 129 (1967). Monod, J., Changeux, J. P., Jacob, F.: J. Mol. Biol. 6, 306 (1963). Monod, J., Wyman, J., Changeux, J. P.: J. Mol. Biol. 8, 88 (1965). Changeux, J. P., Gerhart, J. C., Schachman, H. K.: Biochem. 7, 531 (1968). Moss, J. F., Andersen, J. E.: Austral. Inst. Min. Met. Proc. 225, 15 (1968). Pirt, S. J.: J. Gen. Microbiol. 16, 59 (1957). Nooney, G. C.: J. Theroet. Biol. 9, 239 (1965). Baas Becking, L. G. M., Kaplan, I. R., Moore, D.: J. Geol. 68, 243 (1960). Baas Becking, L. G. M., Moore, D.: Econ. Geol. 56, 259 (1961). Vallentyne, J. R.: Ann. New York Acad. Sci. 108, 342 (1967). Zobell, C. E.: Ann. Assoc. Petrol Geol. 30, 477 (1946). Andersen, J. E., Allman, M. B.: Austral. Inst. Min. Met. Proc. 225, 27 (1968). Razzell, W. E.: Trans. Can. Inst. Min. Met. 65, 135 (1962). Malouf, E. E., Prater, J. D.: J. Met. 13, 353 (1961). Torma, A. E., Walden, C. C., Branion, R. M. R.: Biotechnol. Bioeng. 12, 501 (1970). Sakaguchi, H., Torma, A. E., Silver, M.: Appl. Environm. Microbiol. 31, 7 (1976). Guay, R.: Ph. D.-thesis. Quebec: Laval University 1977. Imai, K., Sakaguchi, H., Sugio, T., Tano, T.: J. Ferment. Technol. 51,865 (1973). Ehrlich, H. L.: J. Bacteriol. 86, 350 (1963). Hulett, J. R.: Quart. Rev. 18, 227 (1964). Han, M. H.: J. Theor. Biol. 35,543 (1972). Marchlenwitz, B., Schwartz, W.: Z. Allg. Mikrobiol. 1,100 (1961). Silverman, M. P., Lundgren, D. G.: J. Bacteriol. 77, 642 (1959). Pluskota, B., Zmudzinski, K.: Rudy i Metale Niezelazne (Poland) 14, 34 (1969). Bryner, L. C., Walker, R. B., Palmer, R.: Trans. Soc. Min. Eng. AIME 238, 56 (1967). Trumbore, R. H.: In: The Cell Chemistry and Function, p. 102. Saint Louis: The C. V. Mosby Company 1966. Arrhenius, S.: Z. Phys. Chem. 4, 226 (1889). Perlmutter-Hayman, B.: In: Progress in Inorganic Chemistry. S. J. Lippard, p. 229. New York: Interscience Publication by John Wiley & Sons, Inc. 1976. Leathen, W. W., Braley, S. A., McIntyre, L. D.: Appl. Microbiol. 1, 61 (1953). Bryner, L. S., Anderson, R.: Ind. Eng. Chem. 49, 1721 (1957). Corrick, J. D., Sutton, J. A.: U. S. Bur. Min. Rept. Inv. 6714 (1963). Tuovinen, O. H., Niemela, S. I., Gyllenberg, H. G.: Biotechnol. Bioeng. 13, 517 (1971). Aiba, S., Humphrey, A. E., Millis, N. F.: Biochemical Engineering. 1st ed. New York: Academic Press: 1965. Withman, W. G.: Chem. Met. Eng. 29, 146 (1923). Higbie, R.: Trans. Am. Inst. Chem. Eng. 31,365 (1935). Danckwerts, P. V.: Ind. Chem. Eng. 43, 1460 (1951). Cooper, C. M., Fernstrom, G. A., Miller, S. A.: Ind. Eng. Chem. 36, 504 (1944).
34 215. 216. 217. 218. 219. 220. 221. 222. 223. 224. 225. 226. 227. 228. 229. 230. 231. 232. 233. 234. 235. 236. 237. 238. 239. 240. 241. 242. 243. 244. 245. 246. 247. 248. 249. 250. 251. 252. 253. 254. 255. 256. 257. 258. 259. 260. 261. 262.
Arpad E. Torma Bandyopadhyay, B., Humphrey, A. E., Taguchi, H.: Biotechnol. Bioeng. 9, 533 (1967). Guay, R., Silver, M., Torma, A. E.: Biotechnol. Bioeng. 19,727 (1977). Harris, J. A.: Proc. Aust. Inst. Min. Met. 230, 81 (1969). Kiesow, L.: Proc. Nat. Acad. Sci. 52, 980 (1964). Gale, N, L., Beck, J. V.: J. Bacteriol. 94, 1052 (1967). Din, G. A., Suzuki, I., Lees, H.: Can. J. Microbiol. 13, 1413 (1967). Bryner, L. C., Beck, J. V., Davis, D. B., Wilson, D. G.: Ind. Eng. Chem. 46, 2587 (1954). Schnaitman, C., Lundgren, D. G.: Can. J. Microbiol. 11, 23 (1965). Bruynesteyn, A., Duncan, D. W.: Can. Met. Quart. 10, 57 (t971). Dutrizac, J. E., Macdonald, R. J. C.: Miner. Sci. Eng. 6, 59 (1974). Peters, E.: In: AIME Short Cours in Bio-extractive Mining. p. 46. Denver: Am. Inst. Min. Eng. 1970. Ivanov, V. L, Nagirnyaik F. I., Stepanov, B. A.: Mikrobiologiya 30, 688 (1961). Szolnoki, J., Bognar, L : Acta Geol. Acad. Sci. Hung. 8, 179 (1964). Beck, J. V.: Biotechnol. Bioeng. 9, 487 (1967). Tuovinen, O. H., Kelly, D. P.: Internat. Met. Rev. 19, 21 (1974). Karavaiko, G. I., Kuznesov, S. I., Golomzik, A. E.: Role of Microorganisms in Leaching of Metal from Ores. Moscow: Nauka 1972. Le Roux, N. W.: In: Microbial Aspects of Metallurgy. J. D. A. Miller p. 195. Aylesburg: Medical and Technical Publishing Company Ltd. 1970, Schwartz, W.: Bild Wiss. 13, 60 (1976). Torma, A. E., Tabi, M.: Ingenieur 59, (294) 2 (1973). Silverman, M. P., Rogoff, M. H., Wender, I.: Appl. Microbiol. 9, 491 (1961). Razzell, W. E., Trussell, P. C.: Appl. Microbiol. 11,105 (1963). Duncan, D. W., Trussell, P. C., Walden, C. C.: Appt. Microbiol. 12, 122 (1964). Duncan, D. W., Walden, C. C., Trussell, P. C.: Can. Min. Met. Bull. 59, 1075 (1966). Bruynesteyn, A., Duncan, D. W.: Can. Met. Quart. 10, 57 (1971). Torma, A. E., Guay, R.: Naturaliste Can. 103, 133 (1976). Torma, A. E.: Rev. Can. Biol. 30, 209 (1971). Gormely, L. S., Duncan, D. W., Branion, R. M. R., Pinder, K. L.: Biotechnol. Bioeng. 17, 31 (1975). Arnold, W. D., Crouse, D. T.: Trans. Soc. Min. Eng. AtME 258, 311 (1975). Rabb, D. D.: Mining Eng. 16, 48 (1964). Braun, R. L., Mallon, R. G.: Trans. Soc. Min. Eng. AIME 258, 103 (1975). Hansen, S. M.: World Mining 18, 56 (1965). Coffer, H. F.: World Mining, 19, 34 (1966). Hardwick, W. R.: U. S. Bur. Min. Rept. Invest. 6996 (1967). Hamburger, R.: Min. Congr. J. 56, 43 (1970). Sheffer, H. W., Evans, L. G.: U. S. Bur. Min. Inf. Circ. 8341 (1968). Torma, A. E.: In: Proceedings of the Third International Biodegradation Symposium. J. M. Sharpley and A. M. Kaplan, p. 937. London: Applied Science Publishers Ltd. 1976. Ivarson, K, C.: Can. J. Soil Sci. 53,315 (1973). Torma, A. E.: TMS, Paper Selection No. A72;7. New York: Met. Soc. AIME 1972. Stumm, W., Lee, G. F.: Schweiz. Z. Hydrologie 22, 295 (1960). Bhappu, R. B., Johnson, P. H., Brierley, J. A., Reynolds, D. H.: Trans. Soc. Min. Eng. AIME 244, 307 (1969). Glassman, H. N.: Bacteriot. Rev. 12, 105 (19¢8). Duncan, D. W., Trussell, P. C.: Can. Met. Quart. 3, 43 (1964). Duncan, D. W., Teather, C. J.: U. S. Patent No. 3266889 (1966). Duncan, D. W. Teather, C. J.: Can. Patent No. 780405 (1968). Duncan, D. W.: Austral. Min. Nov. 1 (1967). Torma, A. E., Gabra, G. G.: IRCS Med. Sci. 3, 228 (1975). Torma, A. E., Gabra, G. G., Guay, R., Silver, M.: Hydromet. 1,301 (1976). Gibbs, W. M., Batchelor, H. W., Sickels, T. N.: J. Bacteriol. 11,393 (1926).
The Role of Thiobacillus ferrooxidans in HydrometaUurgical Processes 263. 264. 265. 266. 267. 268. 269. 270. 271. 272. 273. 274. 275. 276. 277. 278. 279. 280. 281. 282. 283. 284. 285. 286. 287. 288. 289. 290. 291. 292. 293. 294. 295. 296. 297. 298. 299. 300. 301. 302. 303. 304. 305. 306. 307. 308. 309. 310. 311. 312.
35
Pizarro, O. R.: J. Bacteriol. 13, 387 (1927). Alexander, A. E., Soltys, M. A.: J. Path. Bacteriol. 58, 37 (1946). Von Riesen, V. L.: Trans. Kansas Acad. Sci. 58, 337 (1955). Marshall, M. S.: J. Infect. Diseases 35,526 (1924). Imai, K., Sugio, T., Yasuhara, T., Tano, T.: Paper presented at the joint meeting of the Min. Met. Inst. Japan arid the AIME, Tokyo, May 24-27, 1972. Lowry, O. H., Rosebrough, N. J., Farr, A. L., Randall, R. J.: J. Biol. Chem. 193, 265 (1951). Jones, G. E., Starkey, R. L.: J. Bacteriol. 82, 788 (1961). Borichewski, R. M.: J. Bacteriol. 93, 597 (1967). Shaeffer, W. I., Umbreit, W. W.: J. Bacteriol. 85,492 (1963). Usami, S., Sugitani, T.: J. Ferm. Technol. 49, 587 (1971). Tuttle, J. H., Dugan, P. R.: Can. J. Microbiol. 22, 719 (1976). Mikulski, T.: Rudy i Metale Niezetazne (Poland) 20, 576 (1975). Kulebakin, V. G.: Tr. Biol. Inst. Akad Nauk SSSR, Sib. Otd. 25, 133 (1975). Flett, D. S.: Miner. Sci. Eng. 2, 17 (1970). O'Neill, C. E., Ettel, V. A., Oliver, A. J., Itzkovitch, I. J.: Can. Min. Met. Bull, 69, (774), 86 (1976). Itzkovitch, I. J., Torma, A. E.: IRCS Med. Sci. 4, 155 (1976). Torma, A. E., Itzkovitch, I. J.: Appl. Env. Microbiol. 32, 102 (1976). Boyles, W. A., Lincoln, R. E.: Appl. Microbiol. 6, 327 (1958). Gaudin, A. M., Mular, A. L., O'Connor, R. F.: Appl. Microbiol. 8, 84 (1960). Rubin, A. J.: Biotechnol. Bioeng. 10, 89 (1968). Gaudin, A. M.: Flotation. New York: McGraw-Hill Book Co. Inc. 1957. Fliermans, C. B., Brock, T. D.: J. Bacteriol. 111,343 (1972). Brock, T. D., Brock, K. M., Belly, R. T., Weiss, R. L.: Arch. Mikrobiol. 84, 54 (1972). Williams, R. A. D., Hoare, D. S.: J. Gen. Microbiol. 70, 555 (1972). Brierley, C. L., Murr, L. E.: Science 179, 448 (1973). Brierley, C. L.: J. Less-Common Met. 36, 237 (1974). Murr, L. E., Berry, V. K.: Hydromet. 2, I1 (1976). De Rosa, M., Gambacorta, A., Bu'Lock, J. D.: J. Gen. Microbiol. 86, 156 (1975). Silver, M., Torma, A. E.: Can. J. Microbiol. 20, 141 (1974). Torma, A. E., Gabra, G. G.: Antonie Van Leeuvenhoek 43, 1 (1977). Duncan, D. W., Walden, C. C., Trussell, P. C., Lowe, E. A.: Trans, Soc. Min. Eng. AIME 238, 1 (1967). Brimhall, D. B., Wadsworth, M. E.: Trans. Soc. Min. Eng. AIME 254, 68 (1973). Malouf, E. E.: Mining Eng. Nov., 43 (1971). Evans, L. G., Sheffer, H. W.: Min. Congr. J. 54, 96 (1968). Fisher, J. R., Lendrum, F. C., MacDermid, B. C.: Can. Min. Met. Bull. 59, 588 (1966). MacGregor, R. A.: Can. Min. Met. Bull. 59, 583 (1966). Matic, M., Mrost, M.: South Afr. Ind. Chem. 18, 127 (1964). Spedden, H. R., Malouf, E. E., Prater, J. D.: J. Met. 18, 1137 (1966). Torma, A. E., Legault, G.: Ann. Microbiol. (Inst. Pasteur) 124A, 111 (1973). Sutton, J. A., Corrick, J. D.: U. S. Bur. Min. Rept. Inv. 6714 (1965). Zimmerley, S. R., Wilson, D. G., Prater, J. D.: U. S. Patent No. 2829964 (1958). Nielsen, A. M., Beck, J. V.: Science 175, 1124 (1972). Duncan, D. W., Walden, C. C.: Dev. Ind. Microbiol. 13, 66 (1972). Corrans, I. J., Harris, B., Ralph, B. J.: J. South Afr. Inst. Min. Met. 72, 221 (1972). Gupta, A., Saroj, K. K., Thakur, D. N.: Chemical Era 11, 6 (1975). Dinic, M., Pustric, S., Lazic, L.: Rudarski Glasnik (Belgrad) No. 2, 47 (1970). Rossi, G.: Resoconti Dell'Associazione Mineraria Sarda 76(4), 1 (1971). Ito, I.: Kogys Kagakuzasshi (Japan) 72, 437 (1969). Young, R. S.: Sci. Progr. 44, 16 (1956). De Cuyper, J. A.: In: Unit Process in Hydrometallurgy. M. E. Wadsworth, p. 126. New York: Gordon and Breach Science Publ. 1964.
36 313. 314. 315. 316. 317. 318. 319.
320. 321. 322. 323. 324. 325. 326. 327. 328. 329. 330. 331. 332. 333. 334. 335. 336. 337. 338. 339. 340. 341. 342. 343. 344. 345. 346. 347. 348. 349. 350. 351. 352. 353. 354.
Arpad E. Torma Karavaiko, G. I., Moshnyakova, S. A.: Microbiologiya 40, 551 (1971). Torma, A. E.: Australian Patent No. 474361 (1976). Dutrizac, J. E., MacDonald, R. J. C.: Can. Min. Met. Bull 67(743), 169 (1974). Marchlewitz, B., Hasche, D., Schwartz, W.: Z. Allg. Mikrobiol. 1,179 (1961). Lyalikova, N. N.: Tr. Mosk. Abshchest. lspyt. Prir. Otdel. Biol. 24, 211 (1966). McElroy, R. O., Duncan, D. W.: Paper presented at the Conference of Can. Inst. Met., in Quebec City, August 26-29, 1973. Karavaiko, G. I., Lyalikova, N. N., Pivovarova, T. A.: In: Ecological and Geochemical Activity of Microorganisms. (Russ.) M. V. Ivanov, p. 25. Pushino, Academy of Sciences of USSR, Scientific Centre of Biological Investigations, 1976. Miller, R. P., Napier, E., Wells, R. A., Andsley, A., Daborn, G. R.: Inst. Min. Met. Bull. 674, 217 (1963). Marjanovic, D.: Mikrobiologija 2, 243 (1965). Mouret, M.: REv. Ind. Minerale 50, 415 (1968). MacGregor, R. A.: Nuclear Applications 6, 68 (1969). McCreedy, H. H., Harrison, V. F., Gow, W. A.: Can. Inst. Min. Met. Bull. 62(682), 135 (1969). Zajic, J. E., Ng. K. S.: Dev. Ind. Microbioh 11,413 (1970). Viragh, K., Szolnoki, J.: Fold. Kozl. Bull. Hung. Geol. Soc. 100, 43 (1970). Lloyd, P. J., Mrost, M.: In: Recovery Uranium Proc. Symp. 1970, p. 223, Vienna, IAEA, 1971. Barbic, F. F., Krajincanic, B. V.: Mikrobiologiya 41,346 (1972). Kulshrestha, S. C., Jayaram, K. M. V., Dar, K. K.: NMLTechn. J. (India) 15, 42 (1973). Tuovinen, O. H.: Atomic Ener. Rev. 10, 251 (1972). Silverman, M. P.: J. Bacteriol. 94, 1046 (1967). Guay, R., Silver, M., Torma, A. E.: European J. Appl. Microbiol. 3, 157 (1976). Harrison, V. F., Gow, W. A., Hughson, M. R.: J. Met. 18, 1189 (1966). Tomizuka, N., Takahara, Y.: In: Proc. IV IFS. Fermentation Technology Today. G. Terni, p. 513. Osaka: Soc. Ferment. Technol. 1972. Ebner, H. G., Schwartz, W.: Erzmetall 26, 484 (1973). Ebner, H. G., Schwartz, W.: Z. Al|g. Mikrobiol. 14, 93 (1974). Tomizuka, N., Yagisawa, M., Someya, J., Takahara, Y.: Agr. Biol. Chem. 40, 1019 (1976). Ehrlich, H. L.: Econ. Geol. 58, 99t (1963). Ehrlich, H. L.: Econ. Geol. 59, 1306 (1964). Polkin, S. I., Adamov, E. V., Panin, V. V., Kim, D. K., Lavrinenko, A. A.: Izv. Vyssh. Ucheb. Zaved. Tsvet. Met. 15, 27 (1972). Kamalov, M. R., Karavaiko, G. I., Ilyaletdinov, A. N., Abdrashitova, S. A.: Izv. Kaz. SSR. Set. Biol. 1, 37 (1973). Polkin, S. I., Panin, V. V., Adamov, E. V., Karavaiko, G. I., Chernyak, A. S.: Paper No. 33, Presented at the Xlth Internat. Miner. Process. Congress in Cagliari, 1975. Mizoguchi, T., Izaki, K., Takahashi, H., Okabe, T.: Kogyo Kagaku Zasshi (Japan) 73, 1811 (1970). Karavaiko, G.I. Uspekhi Mikrobiologii (Russ.) 6, 174 (1970). Lyalikova, N. N., Shlayn, L. B., Unanova, O. G., Anisimova, L. S.: Izv. Akad. Nauk SSSR. Ser. Biol. 4, 564 (1972). Lyalikova, N. N.: Docl. Akad. Nauk SSSR. 205, 1228 (1971). Torma, A. E.: In: Abstracts of Papers. 5th lnternat. Ferment. Syrup. H. Dellweg, p. 455. Berlin: Westkreuz-Druckerei und Verlag 1976. Goren, M. B.: Australian Patent No. 274690 (1967). Perkins, E. C., Novielli, F.: U. S. Bur. Min. Rept. Inv. 6102 (1962). Imai, K., Tano. T.: Hakko Kyokaishi (Japan) 25, 166 (1967). Cerwenka, Jr., A. E., Cooper, W. C.: Arch. Environ. Health 3, 189 (1961). Torma, A. E., Habashi, F.: Can. J. Microbiol. 18, 1780 (1972). Lipman, J. G., Waksman, S. A.: Science 57, 1463 (1923). Fleming, R. W., Alexander, M.: Appl. Microbiol. 24,424 (1972).
The Role of Thiobacillus ferrooxidans in HydrometaUurgical Processes 355. 356. 357. 358. 359. 360. 361. 362. 363.
Francis, A. J., Duxbury, J. M., Alexander, M.: Appl. Microbiol. 28, 248 (1974). Zajic, J. E.: Microbial Biogeochemistry. New York: Academic Press 1969. Pares, Y.: Ann Microbiol. (Inst. Pasteur) 107, 141 (1964). Pares, Y.: Rev. Ind. Minerale 50, 408 (1968). Duff, R. B., Webley, D. M. Scott, R. O.: Soil Sci. 95, 105 (1963). Silverman, M. P., Munoz, E. F.: Science 169, 985 (1970). Silverman, M. P., Munoz, E. F.: Appl. Microbiol. 22, 923 (1971). Ottow. J. C. G., yon Klopotek, A.: Appl. Microbiot. 18, 41 (1969). Rangachari, P. N.: Chem. Eng. World 10, 77 (1975).
37
Cellulase Biosynthesis and Hydrolysis of Cellulosic Substances T. K. Ghose Biochemical Engineering Research Centre, Indian Institute of Technology, New Delhi 110029, India
Contents 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Cellulose as a Substrate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Productivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Availability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Chemical and Energy Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4 Pretreatment of Cellulosics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Biochemical Nature of Cellulose-CellulaseSystem . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Lignin and Its Degradation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Cellulolytic Enzymes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Production of Cellulases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 Role of #-l,4-Glucosidase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Role of Cellobiose, Surfactants, and pH . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Kinetics of Cellulase Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Hydrolysis of Cellulosic Substances and the Cellulase Adsorption-Desorption Mechanism . 5.1 Kinetics of Cellulose Hydrolysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8. Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9. References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
39 40 40 40 42 43 47 47 49 53 54 55 63 65 68 71 72 73 74
1. I n t r o d u c t i o n The importance of cellulose as the earth's most abundant renewable resource has been realized, though not yet totally accepted. This is the consequence of two global phen o m e n a - e n e r g y crisis and pollution. While both these problems are on the ascending scale the industrially developed nations have begun to search for solutions which are likely to combine both the problems into one, as far as utilization or recycle of cellulosic substances are concerned. In the USA, the city-refuse produced daily alone amounts to more than 55 million tons, about half of which is cellulosic. It is reported [38] that USA spends nearly 1.5 billion dollars every year just for collection and disposal of this refuse. However, solution of pollution problems is p e r s e a different matter altogether. One estimate [31] suggests that in 1970 the annual h u m a n use of commercial energy was 52.9 x 10 Is kcal, while the annual solar energy incident on the earth's surface
40
T.K. Ghose
was 876960 x 10 ~s kcal. Global photosynthesis was able to convert only 605 x 10 ~s kcal of this energy, which is even less than 0.07% of the costless energy which the earth receives. The annual net yield of photosynthesis is 1.8 x 1012 tons of biodegradable substances, 40% of which is cellulose [92]. l f o n l y the mechanism o f photosynthesis were used to capture the entire incident solar energy for human use, 535. 7 × 1015 kcal might have been available annually-this is about ten times the global consumption of energy in 1970. Per capita availability of this energy to the 8 billion world population around the year 2000 would be 67 x 106 kcal annually. This is even less than the 1970 US per capita consumption, but more than four times the 1970 global average per capita energy consumption. Since the world's fossil fuel resources are depleting and the nuclear energy production is facing many significant questions which still have not been clarified, the availability of solar energy needs to be linked up with cellulose productivity and availability.
2.
Cellulose as a S u b s t r a t e
2.1 Productivity It is reported [4] that theoretically about 12.9 mg of pure cellulose can be produced per kcal of total incident radiation on the earth. According to this estimate, the productivity of cellulose on a global photosynthesis basis can be as high as 11.4 x 10 is kg every year. It is, however, not to be presumed that all the energy available in crops can be obtained merely by utilizing their cellulose content. For example, Heichel [30] estimated that if the 40% cellulose content of sugar cane crop could be made available as fermentation substrate at a 25% increase in energy input, the ratio of output to input energy for the crop would decline from 22 to 7 cal. ca1-1. If another 25% increase in energy input is needed to provide the facilities for fermentation, which would extract about 50% of the energy in the substrate in the form of ethanol, the ratio of output to input energy for sugar cane would further decline to 2.8 cal. ca1-1. From the point of view of energy returns on investment, sugar cane appears to be the most efficient crop and produces about 22 calories of total energy per calorie of cultural energy (energy needed to produce the crop in million cal per ha and year). If agricultural productivity exceeds the food requirements of the world, the residual agricultural waste substances, mostly cellulosics, may become an excellent substitute for petroleum as energy source, provided that the presently discarded crop wastes are properly recycled for the retrieval of bioenergy contained in them. While it is generally true that the modem agricultural production technology is extravagant in its consumption of cultural energy, yet more than 50% of U. K.'s food needs are met from home production employing only 3% of its fossil fuel bill [80]. 2.2 Availability The estimated [72] volume of forest resource available (coniferous and broad leaved) is 324 × 109 m a, excluding fuel wood. The total removal of wood solids during the period from 1970 to 72 was 1278 x 106 m 3 out of an estimated growth of nearly 3 200 × 106 m ~
CeUulase Biosynthesisand Hydrolysisof Cellulosic Substances
41
per annum (~ 40%). However, the removal based on the growing stock (306 x 1 0 9 m s) was only 0.5%. The highest rate of removal [87] takes place in Western Europe (2.5%), and the lowest in Latin America (0.05%). Non-wood plant fibres represent a major resource of fibrous raw materials in many countries, particularly in the developing ones. Because of widespread deforestation and lack of fossil fuels in certain areas, importance of non-wood plants mostly of agricultural origin are expected to increase in many developing countries. Several reports suggest that enzymatic hydrolysis of non-wood plant materials is significantly important, as these substances are also easily ameanable to chemical modification to make them more susceptible to enzyme attack. Agrictfltural residues, such as sugar cane bagasse, cereal, and rice straw, naturally growing plants, such as bamboo, papyrus, esparto, and other grasses, and fibre-producing plants, such as hemp, jute, kenaf, ramie flax, okra, etc. and pure fiber-producing plants, such as cotton linters, sisal, etc. are of major importance in our consideration for bioconversion. Estimated availability and utilization of plant-fibre raw materials in 1972 has been reported [87]. It is evident that the availability of the total amount of non-woody plant cellulosics on a global basis is extremely high and is more than a billion metric ton, of which only a little more than six million tons of non-wood agricultural plant materials are presently utilized. Sugar cane bagasse is the most important amongst these materials; not only because its energy (solar) utilization capacitiy is the highest amongst the growing crops, but its yield is about 80 tons. ha -1 (wheat = 1, grass = 2, tree = 20). Moreover, bagasse has a high a-cellulose content, ( ~ 33%) with substantial pentosans (30%) and low ash (2.4%); (rice straw = 17.5%, wheat straw = 11.0% ash). Domestic solid wastes and effluents from paper and pulp industries constitute an important cellulosic resource. The most important feature is that all these materials have zero to negative cost at the site of origin, but are expensive to collect and too bulky to store. Satisfactory separation and pretreatment processes being absolutely essential for any recycle of cellulosic substances, it is largely a question whether a delinking process would enable one to reuse the waste solids as a partial feed stock to the parent process or to employ them as fermentation substrate. As there is a lack of data, it does not seem that a ready answer is available to either of these problems. However, one of the chief advantages of employing these solids in bioconversion processes is their ability to be almost totally converted into microbial biomass or small sugars, requiring practically no pretreatment other than size reduction. With the improved pretreatment technology in the manufacture of paper and pulp, the recycling of waste cellulosic solids delinked from the process stream may appear to be equally or more expensive than the equivalent quantities of fresh materials entering the process stream. In developed countries, starchy grain is presently used as a chief source of livestock feed and is also employed in large quantities as pig and poultry feed. It is thus obvious that the more we are able to use the tremendous quantities of lignocellulosic waste materials, such as bagasse, cereal straws, oil seed stalks, and urban paper wastes after partial conversion or total hydrolysis by enzymes or even by whole microbial cells, the more grain can be released for consumption where it is most needed by the human population.
42
T.K. Ghose
It should be remembered that municipal wastes constitute a large proportion of cellulosic substances. In the United Kingdom alone, this is equivalent to more than 53% of all cellulosic substrates (33.6 million per annum) produced. This is generally true in countries whose food needs are supplied largely from imported processed stuffs. Available figures indicate that nearly 12% of all cellulosic wastes available in the U. K. is recycled every year and a substantial quantum of crop wastes is utilized for silage [73]. Also, the characteristics of domestic refuse are changing rapidly, with more paper, glass, and putrescents and less cinders and ash. This trend is most dominant in the case of apartment-house living. It is most likely that cellulosic constituents of domestic refuse could be an important candidate for hydrolysis in the not distant future. So far, hydrolysis of city refuse as a disposal process has not been applied, although wood hydrolysis was a war-time industry on an extensive scale in the USA. However, the only plant surviving until 1947 based on Madison Process of H2SO4 hydrolysis produced nearly 50 U. S. gallons of ethanol per ton of wood. Two most striking disadvantages of the acid hydrolysis process are the decomposition of sugars once produced by hydrolysis to the extent of nearly 30% of initial cellulose content and the need of nearly 44% additional heat to maintain the reactor conditions (plug flow, 0.4% acid, 230 °C and 29.5 kg • cm -2) through the process. Yet acid hydrolysis of cellulosics is still used in the Soviet Union [44]. In a recent report on the evaluation of substrates for the production of sugars from waste cellulose by enzymatic hydrolysis [3], availability and susceptibility of thirtyfive representative cellulosic wastes have been described. Assuming that we have a good cellulase enzyme and a saccharification technology, it is essential that the following aspects for a proper analysis of saccharification process be taken into consideration: (a) a positive margin between the total recovery of cellulosic wastes and the total demand for recycled pulp, and (b) administrative measures to ensure a stability in the price of waste pulp recovered for reuse.
2.3 Chemical and Energy Aspects Because of its unlimited availability, it is very likely that this natural polymer, being the most important renewable reservoir of solar energy, could be used in a variety of applications within the next few decades. The very chemical characteristics which make cellulose a somewhat difficult degradable substrate are those which make it nature's most important polymer. It is a hydrophilic linear glucose polymer, a polyalcohol more or less uniformly composed of anhydroglucose units coupled to each other by 1,4-glucosidic bonds with the hydroxyl group in the equilateral and H-atoms in the axial positions. Every other chain unit is rotated 180° around the main axis, which results in a strain-free linear configuration with practically no steric hindrance. The glucosidic linkages acting as functional groups and the three OH groups together determine the chemical properties of cellulose. All significant chemical and enzymatic reactions occur at these locations. The native cellulose structure has more than 10000/3-anhydroglucose residues forming a long chain molecule. It means that the mol. wt. of native cellulose is
Cellulase Biosynthesisand Hydrolysisof CellulosicSubstances
43
about 1.5 x 10 6. The total length of this molecule is about 5/am because of the length of each anhydroglucose unit being 5.15 A, while the t~-cellulose contains as many as 10 to 15 X 10 3 glucose units per molecule; 7-cellulose contains only < 15 units of glucose. The smallest structural aggregate of cellulose chains in a fibre of cellulose is the elementary fibril. There is a wide difference of opinion that prevails in regard to the number of cellulose chains in elementary fibrils and microfibrils. In elementary fibrils, areas of complete order, the crystallites and alternates with less ordered and amorphous areas are found. On an average the length of crystallites in native cellulose is 1000 -+ 200 A and that of amorphous areas 300 to 400 A. All our present knowledge, or the lack of it, concerning the biodegradability of cellulose molecule is the distribution and the configuration of these two areas in the natural substance. Wood cellulose occurs in the presence of hemicellulose and liguin-a nonpolysaccharide. Lignin being the cementing material holding the structural support of wood and constituting a large number of plant materials is extremely difficult. The white rot fungus Sporotricumpulverulentumand a cellulase negative mutant strain of it (cell 44)have been reported [2] to have varying degrees of ability to degrade liguin of kraft paper and native wood. The wild type of this fungus, which has both cellulase and lignase activities, degraded kraft lignin better than the mutant. Dimensionally there are three kinds of energy bonds associated with the stability of native cellulose molecule. Energy content is about 73 kcal. mole -1, of which about 68% accounts for covalent bonding, nearly 20% for the H-bonding, and the rest are due to van der Waals forces. Because the crystallinity of various cellulosic materials is widely different (73% in cotton linters, 29% in corn cobs) and because of the fact that the activation energy of the cellulase components vary markedly in respect to substrates they attack (C1 on CMC = 52 kJ. mole-l; C1 on cotton = 14.7 kJ. mole -1 and C1 + Cx on CMC = 26.9 kJ. mole-land C1 + Cx on cotton --- 18.9 kJ. mole -1) one would expect a wide range of conversion productivities achievable with different cellulosic substances. Current biochemical knowledge of C1 and Cx is discussed in Section 3.2.
2.4 Pretreatment of Cellulosics
Cellulosic materials of different origin, virgin or processed, pose a variety of problems in relation to their acceptability for enzymatic hydrolysis, besides economic questions. It is difficult, if not impossible, to establish a theory explaining penetration of the cellulase enzyme into the fibre, particularly when the cellulose is in the native form. It is therefore necessary to obtain data on the composition and characteristics of cellulosic materials which are to be considered for hydrolysis by enzyme action. Even when the ultimate chemical composition of two cellulosic materials is approximately the same their response to cellulase attack can be surprisingly different. The structural features of cellulosic materials that determine their susceptibility to enzymatic degradation include: - the size and diffusivity of the enzyme molecules in relation to the size and surface properties of the gross capillaries and the space between microfibrlls and the cellulose molecules in the amorphous regions,
44
T.K. Ghose
- the degree of crystaIlinity and degree of polymerization (DP) of the celluloses, the nature of substances with which the cellulose is associated. The capillary voids in wood and cotton fibres fall into two categories: (a) gross capillaries such as cell lumina, pit aperture, and pit membrane pores that are visible in the light microscope and (b) cell wall capillaries, such as the spaces between microfibrils and the cellulose molecules in the amorphous regions. The influence of the degree of crystallinity on the susceptibility of cellulose to enzymatic hydrolysis has been studied [66, 76, 88]. Using X-ray-defraction data Norkrans [66] conclusively demonstrated that the cellulase enzymes readily degrade more easily accessible amorphous portions of regenerated cellulose but are unable to attack the less accessible crystalline portion. Similar to the case of acid hydrolysis, a significant increase in crystallinity is observed during enzymatic hydrolysis of cellulose [ 19]. It has been demonstrated that the crystalline residue left after enzymatic hydrolysis can be rendered again into active form by physical treatments such as milling, heating, or both, and their original rate of hydrolysis could be regained. Of the four different forms (I-IV) in which cellulose occurs [32], only the cellulose I is found in native cellulose and the other three forms cannot be produced without altering the degree of crystallinity of the material. Therefore, any study of hydrolysis of cellulose II (viscose, cellophane, and mercerized cotton), cellulose III (cellulose treated with anhydrous ethylamine), and cellulose IV (cellulose treated at some high temperature) toward enzymatic hydrolysis cannot produce any effective data on the extent of accessibility and susceptibility of Cellulose-I. It has been reported [74] that the activation energies of cellulase biosynthesis processes induced by the four different forms of celluloses are markedly different. This suggests that the enzyme synthesizing system of the fungus can locate the most active site of the enzyme so as to accommodate a given structural form of the crystalline cellulose. The distinct presence of hydrogen peroxide produced by Poria monticola and a number of brown rot fungi in association with traces of iron has been demonstrated [29, 48] to catalyse depolymerization of wood cellulose. This suggests that microbial degradation of cellulose fibres may be accomplished at least partly by inorganic, non-protein metabolites which are substantially smaller in molecular weight. In association with Fe ÷+, hydrogen peroxide is said to display [50] three broad features which deserve consideration in the pretreatment of cellulosic materials: - Extensive depolymerization and a rapid increase in alkali solubility occuring at low weight loss. Lignin is altered and part of it is removed; but whether this improves the digestibility of residual wood by the ruminants is not known. It also increases swelling of the polymer. - At high ratios of H202 to Fe ++, the C6 of the gtucosyl unit in cellulose is oxidized to an uronic acid; at lower ratios the C2 and Ca carbons are oxidized to carbonyl groups and finally the pyranose ring is split. These reactions are regulated by pH. Because the mild conditions of pH and temperature are involved and the final products are non-toxic, the H202 + Fe++ systems appear to be significantly important as a potential pretreatment method of cellulosic substances for subsequent enzymatic hydrolysis from both engineering and ecological points of views. Koenigs [49] reported that nine brown rot fungi generated 63 units of H202 in 2 h, -
-
Cellulase Biosynthesisand Hydrolysisof CellulosicSubstances
45
while 12 white rot variety produced none. Production of H202 by the brown rot fungi was quantified by a comparison of H202 generated through glucose oxidase. It is reported further that Lenzites trabea, a brown rot fungus excreted 250 ~g. H202 per g wood in 24 h, and after 72 h the amount of H202 produced would be sufficient to depolymerize the residual cellulose to 25% of the original value. It is observed that although the ability to produce H202 by the brown rot fungi varies, it is the immobilization of Fe 2÷ and the capacity of the system to produce acid which lowers the pH, which are the critical aspects of the H:O2 + Fe 2+ system. The large variation of DP (10000-14000 glucose units per molecule of a-cellulose) are expected to affect the rate of hydrolysis substantially in particular by the enzyme system which chops the cellulose molecules by an endwise mechanism [39]. More recent efforts indicate [71 ] that exoglucanases are the main cellulase components involved in the degradation of native cellulose. With the degree of depolymerization decreasing during acid hydrolysis process, the chopped ends of cellulose chains of the amorphous regions develop a tendency to recombine (recrystallize). This makes the residues more resistant to hydrolysis. When the DP of cellulose is reduced to a point that the molecules become soluble and they no longer maintain their structural relationships one with the other, there is a great increase in the susceptibility to enzyme hydrolysis. Whether such a change is due to chain shortening or increased solubilization enhancing the hydrolysis is a matter not yet fully understood. Any treatment which will alter the proportion of crystalline material or the degree of perfection (parallelism) of the crystallites present may affect the susceptibility of the material to enzymatic attack. The accessibility of cellulose to the extracellular cellulolyric enzymes is determined in part by its distribution within the cell wall and by the nature of the structural relationship among the various cell wall constituents. The extraneous materials in cotton and wood (waxes, oils, fats, proteins, resins which are soluble in solvents such as alcohols, benzene, acetone, etc.) seem to affect the susceptibility of cellulose fibers to enzymatic degradation in two ways: - various substances deposited within the fine capillary structure of the cell wall reduce the accessibility of the cellulose to extracellular enzymes, and - certain specific enzyme inhibitors act directly to reduce the rate of enzymatic hydrolysis. A chemical bond between lignin and cellulose has been postulated [ 17]. Present evidence, however, suggests that the association is largely physical in nature, the lignin and amorphous cellulose forming a mutually interpenetrating system of high polymers. Lignin decreases the accessibility of wood fibers. Interestingly, this decreased accessibility to enzymes occurs without a corresponding decrease in moisture.retaining capacity compared to cotton, possibly because of the ability of small water molecules to preferentially permeate. Pretreatment of cellulosic materials that would increase susceptibility and surface to enzymatic hydrolysis include (a) milling and heat treatment (b) swelling, and (c) dissolving and regeneration. Several methods have been suggested [ 19, 61 ] for rendering various cellulosic substances more susceptible to enzymatic hydrolysis. The idea of heat and milling treatments of cellulosic raw materials to reduce crystallinity and to render them more susceptible to acid hydrolysis was introduced by Krupnova
46
T.K. Ghose
etal. [51]. It has been observed by Ghose and Kostick [19] and Mandels et at. [58] that hammer milling gave good size reduction and increased bulk density, but with no gain in susceptibility. Fluid energy milling and wet colloid milling reduce particle size and somewhat increased susceptibility, but only after extensive treatment at high cost. Heat-treated cellulose has been found to give rise to much lower yields of cellulase (Fig. 1). The inhibitory effects may be due to one or more of the following factors: (a) excessive drying of the fibers, (b) oxidation of a part of the cellulose, (c) reorientation of the cellulose molecule and decrease in cell growth of the cellulose surface, and (d) appearance of toxic substances being produced during heat treatment which are inhibitory to cellulase synthesis. The experimental data illustrated in Fig. 1 describing the effects of heated cellulose on the inhibition of cellulase synthesis and enhancement of cellulose saccharification are important observations [ 19, 22]. On the basis of what is already known, it can be considered that chemical pretreatment, such as alkali swelling or dissolving or regenerating cellulose, increases the availability of active substrate, but the products are of very low density so that suspensions of 4 to 5%
35 No
H~at
Treatment
15° -~4o
25
$ 8 lo
Zo
40
80
ZOO
Incubotion T i m ¢ , H r s . ~
1
Cf:- Ghos¢ & Kostick, Adv.Chcm.Scr., ACS,9._55 415, (1969)
CL u_
i
1OO
t
i
i
125 150 175 Tempcrotur¢ ('C }
i
200 =
Fig. 1. Effect of heat treatment of cellulose on cellulase biosynthesis, and effect of heating and milling of cellulose on hydrolysis. Culture conditions: volume 100 ml in 500 ml shake flask; inoculum 5% (v/v) of a 3-day-old T. viride mycelium suspension; incubation at 29 °C. Cellutase activity measured after 14 days. Substrate is cotton cellulose heated for 30 rain at various temperatures; 0.2% Tween 80 was added after 24 h of growth. S l = Solka floc, a spruce pulp $2 = Solka floc pot-milled $3 = Solka floc heated (200 °C, 25 min) and pot-milled $4 = Solka floc pot-milled and heated (200 °C, 25 min) S~ = Solka floc pot-milled, heated (200 °C, 25 min), enzyme digested, washed and dried (60 °C) S 6 = Solka floc pot-milled, heated (200 °C, 25 min), enzyme digested, washed and dried (60 °C), heated (as before) and pot-milled again
Cellulase Biosynthesisand Hydrolysisof Cellulosic Substances
47
are too thick to agitate or transport. More concentrated suspensions are required so that sugar solutions of reasonable concentration can be attained. Because of the generation of process wastes and the necessity, in many cases, of keeping the products wet for maximum susceptibility, the chemical approach may not be practical. Opinions may, however, differ. Although cellulose is difficultt to mill efficiently because of its fibrous and resilient nature, ball milling has been found to be the most successfull treatment. It reduces crystallinity and increases contact surface, yielding products with maximum availability and high bulk density so that highly concentrated (10-30% cellulose) suspensions can be handled easily in continuous stirred reactors to obtain higher concentrations of glucose syrup [18, 20]. This is a key consideration in the design of a hydrolyzer reactor for a fast rate enzymatic saccharification of cellulose. It should be possible to make a compromise between the milling cost and the concentration of cellulose suspension. 3. B i o c h e m i c a l N a t u r e o f Cellulose-Cellulase S y s t e m The mode of action of the cellulase system on pure and native celluloses has been the subject of study by a large number of workers beginning with the original hypothesis suggested by Reese etal. [75]. Microorganism displaying characteristic biochemical behaviours in the breakdown of cellulosic substances include Trichoderma viride,
Trichoderma koningii, Coniphora cerebella, Sporotrichum pulverulentum, Polyporus adustus, Myrothesium verrucaria, Penicillium funiculosum, Fusarium solanL Aspergillus awamori, Aspergillus wentii, and Coniphora thermophila. The majority of the works published have been based on the biochemical sequences of cellulose breakdown in the extracellular enzyme produced by the T. viride. The parent native strain isolated at the US Army Natick Laboratory, Mass., USA, designated as Tt, QM 6A, had undergone mutation [57], and an improved strain Tv QM 9414 (with three fold enhanced cellulase activity than T~ QM6A) is available for experimental studies from the ATCC, USA (ATCC - 26921). The other fairly well studied organisms are T. koningii [99] and S. pulverulentum [16]. The simplest definition of cellulases can be expressed as a system of enzymes which must be capable of breaking down the highly ordered cellulose polymer into sufficiently small sugars which are able to pass through the microbial cell walls. Such a definition is, however, not applicable to most specific enzymes. Cellulose partly degraded or rendered into active form or soluble derivatives can be hydrolyzed by enzyme components isolated from crude culture filtrates. But these components individually or a mixture of them cannot be considered as "cellulase system" as they are unable to breakdown the highly ordered cellulose even in the absence of lignin (Table 4). According to definition, a complete cetlulase system is found only in relatively few microorganisms, namely T. viride [6, 7, 53, 64, 65, 68-70, 84] and T. koningii [27, 28, 37, 95, 98].
3.1 Lignin and its Degradation Like cellulose, lignin is degraded by microorganisms effectively, continuously, and on a large scale. However, in vitro degradation oflignin by enzymes has yet to be demon-
48
T.K. Ghose
strated. In any bioconversion process of practical significance it is important to recognize the need for use of cellulases in association with lignin degradation enzyme systems, or to convert the lignocellulosics into microbial biomass directly. It is almost impossible to degrade totally significant substrates such as bagasse, most other agricultural residues, mechanical pulp, and cattle manure solids without the participation of lignin degrading organisms or enzymes. Because of the excellent ability to degrade lignins in wood by the white rot type of fungi, some of them may ultimately be used in the bioconversion of lignocellulosics into fungal proteins. Pioneering studies have been carried out by the US Forest Laboratories, Madison, Wisconsin [10, 4 0 - 4 7 , 93] and by the Swedish Forest Products Research Laboratory, Stockholm [ 1, 13-15, 81, 89-91]. It has not yet been possible to characterize and describe any one of the enzyme systems involved in the microbial degradation of lignins. It has been suggested [44, 45 ] that, - a cell-free lignin-degrading enzyme system is likely to be a complex mixture of enzymes and coenzymes; it is an extraceUular multienzyme system, and - for biological conversion of cellulose whole organism (whole rot fungi type) rather than isolated enzymes may be more useful. The poor state of our present knowledge of lignin degradation and the involved enzymecoenzyme system(s) is due to the lack of a good, reliable, and sensitive method for measurement of lignin degradation. However, some interesting data on 1+C-labeled synthetic lignin on bioconversion by a wood destroying white rot fungus, Polyporus resinosus alone and by a mixed microflora from soil [47] demonstrated that ~aCO2 release from the system is far more (nearly double) in the case of mixed microflora than in the case ofPolyporus resinosus (Fig. 2). Oxidation of a-carbinol groups in lignin during degradation by white rot fungi can follow indirectly from phenol oxidase oxidation through laccase or peroxidase, although a key role for such non-specific catalysis appears unlikely. The progress in the studies of enzymatic degradation of cellulose into sugar has gone far beyond the fundamental biochemical knowledge on the mechanism of such a process. This is particularly true in the case oflignin degradation. Techniques of delignifi-
4~ o
I
I
MIXED SOIL
5g,,,o z- o,
/I /
/p
°
+
o l -,.r
l 5
t l I0 15 TIME (DAYS)
2O
Fig. 2. Release of 14CO2 from 14labeledsynthetic lignin on incubation with a wood-destroying white-rot fungus, l"olyporus resinosus, and with mixed microflora from soil [471
Cellulase Biosynthesis and Hydrolysis of Cellulosic Substances
49
cation and saccharification of rice straw, bagasse, and sawdust with T. viride cellulase have been carefully worked out and reported [85]. Some saccharification data of the deliguified bagasse and rice straw with cellulase Onozuka Tv CL 7 in shake cultures are shown in Table 1.
Table 1. Saecharification of delignified bagasse and rice straw with cellulase Onozuka TvCL7 [851 Substrate
Bagasse Rice straw
Substrate
After 24 h
incubation
After 48 h
incubation
Cone. (%)
sugar (%)
decom. (%)
sugar (%)
decom. (%)
10 15 10 15
9.69 12.57 8.58 11.15
87.2 71.2 77.2 63.2
9.79 15.07 9.38 13.59
88.1 85.4 84.4 77.0
Substrates were delignified by boiling with a 1% NaOH solution for 3 h in a 100 ml Erlenmeyer flask; 10 g of delignified bagasseor rice straw and 90 ml of a 3% cellulase Onozuka TvCL7 solution or 15 g of the same substrate and 85 ml of the same enzyme solution were incubated at pH 5.0, 45 °C for 24 to 48 h. A substantial increase in the decomposition of bagasse and rice straw, as well as in the yield of sugar, is evidenced in the delignified samples. More than 90% of decomposition of treated sawdust of broad-leaved trees could be achieved by pretreatment of the material with peracetic acid. From the point of view of cost and scale of operation, it is worthwhile to examine the results with regard to any large scale exploitation of these materials. It seems quite important that the delignification process, either chemical or microbial, is an essential step for any large-scale utilization oflignocellulosic substances. It is interesting to note that while CO2 and H20 are the two end products of lignin decom. position by white rot fungus, when the process is interrupted at any point, the remaining lignin (most of it) remains unaltered, and the portion interfaced with the fungus and thus accessible to the enzyme is greatly modified [46]. It has been possible to separate and purify (free of non-lignin components) this modified lignin [44]. A cellulase negative mutant of white rot fungus, S. pulverulentum, has been used [2] to delignify birch chips to produce what is called biological pulp of longer fibers compared to those not subjected to the microbial treatment. Because of the reduction of the original lignin up to 17% after six weeks of treatment, it is felt that the cost of mechanical pulp production is likely to become substantially lower on account of lower energy consumption for the treated pulp. 3.2 Cellulolytic Enzymes During the last five years, major contributions made toward our understanding of the ceUulase complex and its mode of action have principally been confined to organisms such as Trichoderma viride, Trichoderma koningii, and Sporotrichurn pulverulentum
50
T.K. Ghose
(Chrysosporium lignorum), although a number of other fungi, bacterial species and of late Thermoactinomyces sp have been studied [36]. But they are more concerned with application aspects than with clarifying the biochemical mechanism of cellulose breakdown. In the text, the activities of different cellulase components are expressed as (i) Cractivity measured against highly ordered cellulosic substances (cotton), (ii) Cx-activity toward soluble cellulosic compounds (carboxymethyl cellulose), and (iii) FP-activity measured against Whatman No. 1 filter paper. The isolated and purified exo- and endo-glucanases from culture filtrates of cellulase-producing organisms do not fully identify the characteristics of C1 and Cx, respectively. One view [71 ] is that the presence of a hitherto unknown initiating factor (enzyme) might be responsible for converting the native cellulose to an activated state which is attacked by exo- and endo-glucanases to produce sugars. The basic and the most important question relating to the synergistic action between Ca and Cx components at the site of attack has not been answered. The other perplexing issue remaining unresolved is how and why the C1 and Cx enzymes are seperately capable of hydrolyzing swollen cellulose, but not highly ordered cellulose; yet, when both of them are in solution they are capable of hydrolyzing highly ordered cellulose with comparative ease [97]. Cellulase preparations produced commercially from T. viride have been reported [6, 84] to have a Cx component not identical with C1 components isolated from T. koningii or from Sporotrichum pulverulentum [ 12]. While the literature on cellulase is abundant in its descriptions of behavior and characteristics of C1 components isolated from various fungi, it has always been a problem to precisely compare the exact hydrolytic activities of the various C1 components because of the variety of assay procedures used. These various procedures are described by many authors [52, 60, 85, 96]. Purification procedures such as gel filtration, ion-exchange chromatography for the fractionation of C 1, Cx, and/3-glucosidase components of various extracellular cellulase culture filtrates have been reported [28, 98]. Trichoderma koningii is said to contain one C1, five Cx, and two t3,1-4 glucosidases [99]. Halliwell [29], however, reported that the multi-enzyme cellulase systems of T. koningii and M. verrucaria consist of two cetlulases-exo- and endo-glucanases, and two dextrinases-CM cellulase and cellobiase, which together saccharily native cellulose, its primary degraded product (short fibers) and degraded forms of polysaccharides. Both the cellulase systems produce cellobiase. The endo-cellulase also yields glucose and cellodextrins. Cellodextrins with cellobiose inhibit both the cellulases (endo- and exo-acting forms and dextrinases). Glucose inhibits endo-cellulases. From these observations he concluded that the four component multienzyme system with no reference to C1 can effectively carry out cellulolytic activity (short fiber formation and saccharification) if the substrate is susceptible and no inhibition on the system is exercised. Such an observation can neither establish the exact picture of C1, if it existed, in relation to exo-enzymes, nor can it disprove the role of C, in the activation of ordered cellulose into the susceptible form. Thus, the question whether regions of less ordered cellulose (amorphous, susceptible?) chains exist in the microfibril, which is of fundamental importance in our understanding of the mechanism of cellulolysis, has not yet been answered.
CeUulase Biosynthesis and Hydrolysis of Cellulosic Substances
51
However, recent studies [100] suggest that C1 enzyme cannot be considered to have a C~ function in the sense that it initiates attack on highly ordered cellulosic substrates. This would seem to be a characteristics attribute of Cx rather than C~. It is now clear that both these enzyme components in association with/3-glucosidase must be present in the solution for the manifestation of maximum, if not the entire, cellulase activity (Table 4). It is, therefore, reasonable to assume that C~ enzyme will be participating in the hydrolysis of the new chain ends that are created by the randomly acting Cx. Pettersson [71] isolated and purified four cellulolytic components of T. viride (Table 2) to the stage of physico-chemical homogeneity, one of which is characterized as an exo/3-1,4 glucanase [6]. It saccharifies microcrystalline cellulose (Avicel) up to 80% (w/w) and native cotton cellulose (DP 3000) up to 40% (w/w). The isolated exo-enzyme is most likely identical with the C1 purified from S. pulverulentum [12], T. koningii [98], and T. viride [55, 84]. Table 2. Some properties of cellulolytic enzymes isolated from Trichoderma viride [71] Activity toward different substrates Type of enzyme
Molecular weight
Isoelectric point
Carbohydrate content %
CMC
MicroReprecrystalline cipitated cellulose cellulose
Cellotetraose
Exo-/3-1,4-glueanase Endo-#-l,4-glueanase I Endo-t3-1,4-glucanase II #-glucosidase
42000 12500 50000 47000
3,79 4.60 3.39 5.74
9 21 12 0
+ + -
+ -
+ + + +
+ + + -
The enzyme active against Avicel removed cellobiose units from the non-reducing ends of/3-1,4 glucanase, a fact which has been observed by most workers in the field. The exo-glucanase was also found to be inhibited by the reaction product cellobiose. The component active against cellobiose was also found to be able to attack oligosaccharides at the/3-1,4 linkage, including the analogue p-nitrophenyl-/3 glucoside. When cellobiase is associated with exo-glucanase, the degradation of microcrystaUine cellulose is enhanced. Internal linkages of CMC or swollen cellulose at the/3-1,4 positions are attacked by the endoglucanases; they also cause a significant reduction in DP when incubated with cotton. The scheme of purification of various cellulolytic components of T. viride cellulase is shown in Fig. 3. [71]. Based on the studies carried out with T. viride ceUulases, a modified scheme (Fig. 4) of the sequential degradation of native cellulose into glucose has been proposed [71 ]. Although this scheme is yet to be confirmed, it gets partial support from the results of Halliwell's work [29] discussed earlier. Basic assumptions for the Pettersson model are: - Regions of low crystaUinity in the cellulose fiber are attacked by endo-glucanases and free chain ends are created. Exo-glucanases start the degradation from chain ends by hydrolyticaUy removing ceUobiose units. -
52
T.K. Ghose
Purification Scheme for Isolation of Four Different Cellulolytic Enzymesfrom Trichoderma viride
Crude Cellulase
I
Molecular-sieve Chromatography (Bio-Gel P-10)
1
Ion-Exchange Chromatography (DEAE-Sephadex A-50) 1
t
Chromatography (Arginine-sepharose 6 B)
Ion-Exchange Chromatography (SE-SephadexC-50) t
I
I
Bio-Specific Chromatography (Avicel)
Isoelectric Focusing (pH gradient 5--7)
lsoelectric Focusing (pH gradient 2,3--3,3
Molecular-Sieve Chromatography (Bio-Gel P-60)
i
Ion-Exchange
IsoelectricFocusing (pH gradient 3,0--4,2)
t
Isoetectric Focusing (pH gradient 4,1--5,2)
I
J
Molecular-Sieve Chromatography (Bio-Gel P-60)
Molecular-sieve Chromatography (Bio-Gel P-30)
Molecular-Sieve Chromatography (Bio-Gel P-60)
I
'-[Endo #-1,4-GlucanaseII !
[ #-Glucosidase ]
Fig. 3. Fractionation scheme for the purification of four different ceUulolytic enzymes from Tricho-
derma viride 1711
Endo-glucanase ~ Cellulosea Exo-glueanase Cellulosea • Cellobiose ~-Glucosidase Cellobiose , 2 Glucose
1. Native cellulose 2. 3.
a formed from native cellulose by the action of the endo-glucanase on non-crystalline regions of the cellulose fiber. Free chain ends are created. Fig. 4. A modified mechanism for enzymatic cellulose degradation [711
- Cellobiose is hydrolyzed to glucose through the action of a ~-glucosidase. Purification of endo- and exo-glucanases from the culture filtrates of S. pulverulentum [14, 15] made it possible to reconstruct the culture filtrate using the purified endoand exo-glucanases in the same proportion as in the original filtrate. The results of degradation of cellulose (dewaxed cotton) are presented in Table 3 [16]; it can be seen that the original filtrate shows about two and half times more cellulose degradation than the "reconstructed" culture filtrate having only exo- and endo-glucanases. Moreover, the effect of a non-oxidative atmosphere on the culture filtrate toward the degradation of cellulose was found similar to that of the "reconstructed" mixture of endo-
CeUulase Biosynthesis and Hydrolysis of Cellulosic Substances
53
and exo-glucanases. It might suggest that the missing enzyme component in the "reconstructed" filtrate is of oxidative nature. Comparative studies of the enzymes of four different organisms in respect to cellulose degradation by concentrated enzymes in 02 and N2 atmosphere [16] also appear to support that cellulose degradation is somehow associated with an oxidative process. A new wood-degrading enzyme, cellobiosequinone oxidoreductase-detected [89, 90] in the culture filtrate of white rot fungi and participating in the degradation of lignin-is most likely to be active also in the degradation of cellulose. Eriksson [16] reported a cellulase negative mutant (cell 44) ofS. pulverulentum which was able to degrade Kraft lignin as well as wood lignin without having either cellulases or cellobiose-quinone oxidoreductase activity. However, according to the same authors, in the lignin degradation by Pleurotus ostreatus, the presence of cellulose was found to favor the degradation. It has not, however, been possible until now to establish a confirmed scheme of biochemical breakdown of either cell,lose or lignin, nor to determine what, if any, missing factor(s) of enzyme component(s) might be linking the two very complicated processes of nature. That we have not yet been able to reconstitute the cellulase system by isolating and remixing the components obtained from the original culture fdtrate is also evidenced from the data (Table 4) of Wood [97]. Table 3. Degradation of cotton cellulose by enzyme from S. pulverulentum | 161 Enzyme preparation
Cellulose degradation weight loss (%)
Concentrated culture solution Concentrated culture solution Mixture of endo- and exo-glucanases Endo-#-l,4-glucanases Exo-~-1,4-glucanase
52.1 (oxygen atmosphere) 21.5 (nitrogen atmosphere) 20.0 0.0 0.0
Table 4. Relative eeltutase activities of the components of Z koningii cellulase separated on DEAE-sephadex 1971 Enzyme
Recovery of cellulase activity (%)
Cx + #-glucosidase C~ C~ + Cx + #-glucosidase Original culture filtrate
3 4 96 100
4. P r o d u c t i o n o f Cellulases During the last few years several contributions have appeared on the production of cellulases mostly by the fungus Trichoderma viride [8, 22, 59, 67]. Another author [83] reported on the constitutive nature of a system of cellulase enzymes produced by a bacterium P. fluorescens var cellulosa and the transition between repression and derepression of cellulase formation by this organism occurring in a rather narrow range of sugar
54
T.K. Ghose
concentrations (4 and 6 mM of glucose). Moreover, extracellular cellulases (avicelase and CMCase) and a-amylase were found predominantly in the membrane-bound polyribosome fractions, whereas intracellular enzymes (cellobiase and ~-glucosidase) were found almost equally in both free- and membrane-bound polyribosome fractions. Enari et aL [ 11 ] employed 1% glucose, 2% Solka Floc and 0.5% yeast extract and mineral nutrients in the medium for growth and production o f ~ , l - 4 glucanases byA. awamori. Catabolic repression of glucose and its concentration limits have not been discussed. For T. viride conditions for maximum cellulase production are not the same as the conditions for optimum growth. During metabolism of soluble carbohydrates, a rapid fall in pH and an uptake of sugars take place. Metabolism of cellulose as an insoluble substrate could be as fast as soluble carbohydrates. But because the ordered regions differ from the susceptible portions of cellulosic substrate, maintenance of a continuous culture on cellulose is apparently very difficult. Rapid disappearance of sugars (cellulose) causes a fall in the pH of the system. This in turn causes the disappearance of cellulases. It seems the sensitivity of/3-1,4 glucosidase is more marked toward low pH than the saccharifying cellulases (Table 5). The faster the rate of utilization of any carTable 5. Effect of addition of glucose to a cellulase-induced culture with and without pH control [591 Hours after glucose addition pH uncontrolled
pH controlled at 5.0
0
22
0
22
Glucosidase
0.146
0
0.183
0.159 units, m1-1
Saccharifying cellulase
0.48
0.32
0.77
0.75 units, m1-1
QM 9414 was grown in a 10-1 fermenter on 0.5% BW 200 with pH not allowed to drop below 4.0. After pH rose (cellulose consumed), 0.5% glucose was added and pH-controlled not to fall below 5.0. Glucose was rapidly consumed. After pH rose again, 0.5% glucose was added, pH was not controlled and fell to 2.4 as glucose was more slowly consumed bon source (including cellulose), the higher the production of acids, which results in the loss of cellulase activities [59]. Acid production during growth on cellulose may have a regulatory function. The activity of/3-glucosidase may control the flow of glucose derived from cellulose into the cell. This process is thus likely to affect considerably the activities of both saccharifying cellulases and t3-glucosidase. It has been demonstrated [59] that the addition of about 0.5% glucose to a culture growing on cellulose brings about a rapid fall in pH to 2.5, as the glucose is consumed with a simultaneous loss in the activities of j3-glucosidase and a portion of saccharifying celtulase. In the case of the continued presence of cellulose in the system, cellulase activity is likely to reappear after all the glucose is consumed.
4.1 Role of ~-l,4-Glucosidase When the T. viride medium is buffered at pH 5.0, the extracellular/3-glucosidase activity increases. When a low concentration of cellulose (< 0.2%) is hydrolyzed with T. viride
Cellulase Biosynthesisand Hydrolysisof CellulosicSubstances
55
cellulase, most of the soluble product is glucose. In the presence of a/3-glucosidase inhibitor, like nojirimycin [82], most of the soluble product appears as cellobiose; nearly 90% of the glucose is produced from the cleavage of cellobiose. A similar value is obtained when/3-glucosidase is inactivated by acid or inhibited by nojirimicin. It therefore supports the endoglucanase-cellobiohydrolase-t3-glucosidase hypothesis for cellulolysis at low concentrations of milled cellulose. However, when higher concentrations of milled cellulose (10%) are used, a different mechanism, other than the hydrolytic cleavage of cellobiose, causes appearance of glucose [82]. The amount of glucose produced from the low/3-glucosidase-cellulase preparation is almost the same as with high/3-glucosidase activity. Also, in the presence of low/3-glucosidase activity more glucose is produced from milled cellulose than from cellobiose. This suggests a second route of glucose production in addition to the hydrolytic cleavage of cellobiose by/3-glucosidase. This route does not appear to be operative when crystalline cellulose is hydrolyzed (cotton, avicel) and significantly more glucose is produced when the available/3-glucosidase activity is higher. It has also been reported [35] that the celhilase synthesis mechanism is derepressed during slow growth on other sugars. When T. viride is grown on cellulose, much higher levels of extracellular proteins are produced than when it is grown on glucose. Thus, the rate of fall of pH due to increased consumption of NH~may be greater. During the phase of rapid fall in pH exo- (saccharifying) and endo-cellulases are induced [59], which suggests that cellulase synthesis is not brought about by a slow feed of carbohydrates but that the induction occurs during the high metabolic activity. However, in complete absence of cellulose, step feeding of cellobiose in low concentrations enhances cellulase production (discussed later) [22]. Because of the multisubstrate nature of cellulosic materials, their concentration during the induction process is important. Different cellulosic substrates have been studied [22], and the optimum concentrations maximizing cellulase activity demonstrated (Fig. 5). It seems that the cellulase synthesis is related to the extent of microcrystalline cellulose present in the inducer.
4.2 Role ofCellobiose, Surfactants, and pH Extensive studies on the effect of cellobiose used as an inducer in the presence and absence of cellulose, as well as the manner in which cellobiose is made available to the T. viride culture, have been carried out [22]. The results indicate (Figs. 6-8) that cellobiose repression per se (?) or through glucose can be externally controlled with or without cellulose being available to the system, provided it is possible to make good the rapid uptake of NH~ by pH control. Effects of surfactants as agents to accelerate release of cellulases across the membrane may be attributed to a mechanism of modification of the membrane increasing its permeability [77]; this change is reflected in the metabolism of the organism. Effects of the presence of surface active agents on the release of celhilase are demonstrated [22]. A single feeding of cellobiose at 0.01% level slightly enhanced the cellulase yields compared to no feeding of cellobiose, but the yields were lower than those obtained using step feeding at 0.005% per day. These results appear to confirm that not only does an inducer-repressor system exist controlled by a feedback mechanism, but some catabolic
56
T.K. Ghose []
VP Cellulose
28
[ ] Sotka Floc 24
Fitter Paper Pulp
2° i --16
=~ 12 I-L
B
t, 4
0.25
0.5 0"70 1.O CeLlulose Concentration (%)
1,5
2.0
i.
Fig. 5. Effect of concentration of cellulose as inducer [22]. Culture volume was 100 ml shake flask. Inoculum 5% (v/v) of 3-day-old mycelial suspension of T. viride QM 9123. Incubation temperature 29 °C. Cellulase activity measured after 14 days. VP cellulose is purified powdered cellulose of native cotton obtained from VP Chest Institute, Delhi University, Delhi. Solka Floe 40 obtained from Dr. Mary Mandels, US Army Laboratories, Natick, Mass., USA
1.6 FPA
~I,2
g
~
j4
~o.,
3(
6
pH
TIME
(DAYS)
~--~
Fig. 6. CeUulasebiosynthesis on I% ceI1obioseas carbon source [22]. Culture conditions: Vol. 100 m]
shake culture (T. viride QM 9123). Inoculum: 5% v/v of a 3-day-old mycelial suspension. Incubation temperature: 29 °C. Activity units are defined in the cited literature
8
1-6
cx !
~
80
o
a
~
_
£ c
o.,
yy
fl
#,
2
4
i
i
6 8 10 t2 Incubation PGriod cloys
14
m
i
16
18
Fig. 7. Induced biosynthesis o f cellu]ase by step feeding o f eellobiose at 0.005%, added daily to the
Tv culture containing 0.7% VP cellulose at controlled pH 5.0 [22]. Same culture conditions as in Fig. 6. Cellobiose feeding started after 48 h of growth; the medium contained 0.5% Tween 80 added after 24 h of growth
80
y Feed}ng
~
~ O n c , OnceFeeding Doily . No
",.4
Cx
Once
,180 ,420
Feeding
No NO
" "" ""
3OO
1
FPU -FPU
4O
1
Doil y Feeding Once 180 _
._,-_ c
o
?_
0,.
"20 ,,6
J
50
2
4
6 8 10 Tim,¢ (Doys) ...........
12
'~4
16
4--
Fig. 8. Effect of feeding cellobiose (0.01%) once into a Tv culture containing 0.7% VP cellulose on cellulase biosynthesis. Upper, middle, and lower shaded areas represent Cx, FPU, and C l units m l - 1 , respectively, when ceUobiose was added once during the incubation period starting from 2rid to 12th day at 0.01% level [221. Same culture conditions as in Fig. 7. Cellobiose feeding started after 48h e - e - o Cx, o - o - o FPU, o - o - o C1 (daily addition of cellobiose at 0.005%) A--A_A Cx ' z~-zx-zxFPU (no cellobiose addition)
58
T.K. Ghose 100
70 80 60
60 >"
(
UA
50 4o z,C
20 3G 0
i ~r 20 40
"~-~-~__~.___b--i I i I o 60 80 lOO 12o 14o 16o 18o 200 220 hours
Fig. 9. Growth and enzyme production of Trichoderma on cellulose. QM 9414 on 0.75% BW 200 (baH-milled pulp) + 0.075% proteose peptone, 0.2% Tween 80, spore inocutum 159] o - o - o pH z~_~_A Saccharifying ceUulase (FP) 100 = 1 unit • m1-1 •, - m - . Endo-cellulase (Cx) 100 = 55 units • m1-1 , - , - # 13-glucosidase (13) 100 = 0.10 unit • m1-1 . - o - o Extracellular protein (P) 100 = 2 mg. m1-1
repression is likely to be involved in tile regulation of cellulase synthesis. It is also to be seen that the induction mechanism at a much higher concentration (0.5%) o f cellobiose is markedly suppressed. This effect can be attributed to be a repression caused by glucose appearing through the action 13-glucosidase on cellobiose. These findings are also confirmed by the data of Mandels et al. [59] shown in Table 6. Changes in pH o f an extracellular cellulase-producing system (like T. viride) can be used as a measure of the rate of cellulose consumption (Table 7). When conditions o f cellulase synthesis are not very favorable T. viride culture grows slowly and little cetlulase is produced, and the drop in the medium pH is also small. Under favorable conditions large quantities of cellulases are produced, and the rate of pH fall is higher than with glucose (Fig. 9; Table 7), suggesting that cellulose can be consumed as rapidly as glucose during the early phase of growth. However, there is no doubt that cellulose-grown cultures of T. viride produce much higher cellulases as extra cellular protein than when grown on a soluble carbohydrate (glucose or cellobiose). Both soluble protein and FP activity are markedly enhanced in the case o f T. viride culture when pH is controlled at 5.0 (Fig. 10). Mostly in industrial microbiological work starting pH values are reported in the case of batch fermentations; media for such cases are designed in such a way that a narrow limit of pH range can be maintained. An examination o f the effects o f pH on
Cellulase Biosynthesis and Hydrolysis of Cellulosic Substances
59
Table 6. Effect of addition of glucose to cellulose cultures on filter paper activity, soluble and total proteins Day
1% glucose
1% cellulose
1% cellulose + 1% glucose added at 0-day
1-day
2-day
5-day
Filter paper activity 0 0.12 NT 0.03 NT 0.14 0.09
0 1.73 NT 1.43 NT 1.85 1.76
0 0 NT 0 NT 0.64 1.26
0 0 NT 0.38 NT 0.99 1.19
0 0 NT 0.07 0.02 0.57 0.91
0 1.73 0.56 0.50 0.49 0.52 0.76
Soluble protein (mg- m1-1 ) 2 0.09 5 0.04 6 NT 7 0.02 8 NT 9 0.06 14 0.12
0.02 0.34 NT 0.37 NT 0.38 0.61
0.09 0.05 NT 0.02 NT 0.06 0.27
0.09 0.05 NT 0.04 NT 0.13 0.3t
0.02 0.05 NT 0.01 0.03 0.05 0.13
0.02 0.34 0.05 0.04 0.06 0.07 0.16
Total protein (mg- mt- 1) 2 0.54 5 0.58 6 NT 7 0.37 8 NT 9 0.24 14 0.29
0.13 0.62 NT 0.76 NT 0.55 0.89
0.60 0.73 NT 0.62 NT 0.36 0.80
0.53 0.58 NT 0.41 NT 0.43 0.80
0.13 0.39 NT 0.46 0.25 0.31 0.98
0.13 0.62 0.73 0.94 0.54 0.60 0.74
2 5 6 7 8
9 14
T. viride QM6a grown on 1% glucose with no additives or 1% cellulose plus 0.1% peptone and 0.2% Tween 80. 1% glucose added to certain cultures as indicated [591.
cell growth, glucose uptake rates, and dissolved 02 profdes in T. viride cultures (Fig. 11), vis-a-vis those under uncontrolled pH conditions (Fig. 12) at 28 °C, reveal that the organism grows in an exponential manner. The yield coefficient Y, (kg. cells). (kg. glucose) -1 was found to remain substantially constant at 0.4, and the maximum specific growth rate was linearly related to the hydrogen ion concentration by (/gm)pH = /'1° (1 -- 309 [Hq)
(1)
giving values of 0a°)28 = 0.104 h -1 and tt2s = 0 at pH = 2.48. An uncontrolled pH run showed a lag in the response to the drop in pH which could be a delay in the membrane resistance response [9]. The authors have summarized the results of controlled pH (Table 8). They indicate that T. viride cells grown on glucose give a value o f yield constant of 0.5 at pH between 3.0 and 4.0.
60
T. K. Ghose 3+6
2-B
2.0
t-2 ot~in
<( ~L LL
0-4
0'5
1"0 1-5 2.0 2.5 Aerotion Rote (vvm)
3"0
Fig. 10. Effect of pH control at 5.0 on cellulase biosynthesis in 5-1 stirred tank reactor (STR) I221. T. viride QM 9123 culture medium containing 0.2% Tween 80 after 48 h of growth. Agitation i00 rpm. Cellulase activity and protein were measured after 14 days. Temperature 29 °C and inoculum 5% v]v of a 3-day-old mycelial suspension o - o controlled pH (5.0) o - o no pH control
Table 7. Relation of substrate consumption and pH fall I59] Carbon source
Slope A sugar
Slope A pH
Xyiose Glucose Cellobiose Lactose L-Arabinose D-Arabinose Cellulose
0.357 0.253 0.225 0.139 0.100 0.060 -
0.180 0.156 0.175 0.076 0.063 0.055 0.246
za sugar = sugar consumption (mg- m1-1 • h - l ) , A pH = pH change (units. h - 1 )
Table 8. Yield coefficient of T. viride grown in glucose as a function of pH [9] Run
pH
[ H+]
~m (h-1 )
1 2 3 4
2.5 2.7 3.0 4.0
0.00316 0.00200 0.00100 0.00010
0.001 0.039 0.077 0.097
Y
0.39
0.40 0.40
2N KOH used for control (cm ~) 0 137 197 281
Cellulase Biosynthesis and Hydrolysis of Cellulosic Substances
61
J/3W27
*6
o
2
o
~
0
l 10
I20
0
I 10
I 20
100
-I 30
I 30
25
-' 40
,50
l60
, 70
t 80
t0 9
z,0
50
I 60
I 70
I "~ I 80 90
10 0
i00
o 2.7
so
~5
0
10
20
30
z,0 50 60 T i m e (h)
Fig. I t . Effect o f controlled pH on cell growth, glucose uptake, and QM 9123 culture [8]. Reactor dimensions and operating conditions Diameter 228 mm Impeller blade height Height 456 mm No. of baffles Liquid submerged 270 mm Type of sparger No. of impellers 1 Impeller speed Type of impeller 6 bladed disc turbine Air flow rate Impeller diameter 114 mm Operating Vol. Impeller height 89 mm Inoculum Vol. Impeller blade width 19 mm
70
80
90
d. o. profile in T. viride 38 mm 4 X 23 mm Single orifice 400 rpm 1 wm 10 1 500 cm a (x 0 = 0.5 kg • m - a )
100
62
T.K. Ghose
3 X
o
5 4
o
1,5
•
81
3 1C
0
10
20
30
Time(h)
40
50
60
Fig. 12. Effect of uncontrolled pH on cell growth and glucose uptake by T. viride QM 9123 [8]. Reactor and operating conditions same as in Fig. 11
Experimental data on the effects of glucose concentration on the synthesis of cellulases expressed as FPU have been given [8]. They seem to reveal that an activity as high as 2.5 FPU can be produced in 70 h using glucose as the growth-and cellulase-inducing carbon and energy source. Oxygen limitation effects and aeration rates on the enzyme production were also studied by Brown et al. [8] and Ghose et al. [22]. Some relevant data on these investigations are presented in Table 9 and Fig. 9. It is reported [56] that the synthesis of cellulases in commercial glucose medium is markedly affected by the sophorose content of the sugar and the conditions of the batch culture. Sophorose present as an impurity in commercial glucose obtained from acid hydrolysis of cellulose acts as an inducer of cellulase synthesis by T. viride. However, it does not serve as an inducer of celtulase synthesis by P. fluorescens [83]. Unknown effects are to be noted, particularly the interactions between the starting pH, inoculum size, and aeration capacity. Considerable information clarifying these effects for both continuous and batch production of cellulases are necessary. Mendels et al. [59] are of the view that continous culture of T. viricle does not favor rapid growth and enzyme production on cellulose. The fact that cellulose is the only carbon source responsible for the biosynthesis of the cellulase components of the T. viride enzyme system calls for indepth investigations on the production of cellulases with cellulose in continuous culture, even though it is not an easy technique. Rate of cellulase enzyme synthesis by T. viride vis-a-vis inoculum size and quality has been investigated [671. It is suggested that the initial slow growth rate of T. viride culture is probably due to two factors: (1) the peptone is the most readily available substrate and it is consumed by deamination causing a fall in pH of the system, and (2) the
63
Cellulase Biosynthesis and Hydrolysis of Cellulosic Substances Table 9. Oxygen limitation effects on the growth and cellulase synthesis by T. viride [8] Run
M a x i m u m Lowest cells pH (kg. m-3)
Time at lowest pH (h)
Oxygen rise fh)
Glucose Cellulase Maximum zero (h) start (h) cellulase (FPU)
30/2X 30/3.5X 30/5X 30/7X
4.8 (22)a 6.0 (32) 9.2 (42) 12.2 (44)
1 1 1 10
21 25 32 33
21 30 35 41
2.8 (22) 2.6 (38) 2.6 (42) 2.9 (38)
18 30 45 41
2.6 1.3 0.9 1.8
aNumber in brackets in time (h) at which the particular value occurred inocula are high in cellulase enzymes and this causes cellulosic substrates to produce small sugars. As the metabolic shift from protein consumption to carbohydrate begins, growth becomes fairly rapid and a shift in the respiratory quotient {RQ) value takes place. Depending on the size and age of the inoculum, RQ values change substantially as long as the system is not protein-limited. This process is associated with a rapid rate of pH decline. The cycle of increase and decrease in the RQ values seems to be associated with rapid and sluggish growth rates; subsequently, a steady state is reached when the hydrolysis of cellulose-producing sugars and the consumption of sugars by the cells become equal. The authors concluded that rapid growth of T. viride on cellulose is not necessarily associated with enzyme production and substantial FP assays are achievable in less than 24 h; also, that cultures with cellulase activities as low as 1.4 FPU may make excellent inocula and this level of enzyme activity may be achieved in about 20 h. 4.3 Kinetics of Cellulase Production Because of the insoluble nature of the multisubstrate cellulose and lack of complete knowledge of the biochemical mechanism of cellulase synthesis, very few data are available on the kinetic analysis of the system. In the early part of this review it was discussed that the supply of cellobiose to the culture at a rate the cells are able to consume (step feeding), with the result that no glucose is left to exercise any catabolic repression, produced appreciably higher soluble proteins and increased cellulase yields. Also, addition of Tween 80 (surface active agent) at 0.2% level increased the Cx activity by a factor of two. The time course of T. viride cellulase culture indicates that enzyme biosynthesis begins near or at the end of the phase of cell growth (Fig. 6). It then suggests that the specific enzyme synthesis rate (e) has a negative correlation with the specific growth rate (/a) (Fig. 13). On the basis of analysis of the kinetic data, the following postulates were suggested [22] - the biosynthesis of T. viride cellulases is a non-growth associated system, - the measured increase in the cellulase activity (FPA) is likely to be a product of the de novo synthesis of the peptide chain of the enzyme, - the mRNA for cellulases appears to be very stable compared to those responsible for the synthesis of a great majority of enzymes. An analysis has been made of a two-stage CSTR operation using glucose in the first stage for the purpose of cell mass production and cellulose for the induction of cellulase
64
T.K. Ghose 1-8
.OO9
30v
I
.025
O05~
~X
D1S
903~ hi
.2
8 .u
g Q.
CI.
,005 m 301
2
4
6
B 10 TIME (DAYS)
12
14
16
1B
'~
Fig. 13. Kinetics of ceUulase biosynthesis by T. viride QM 9123 on cellobiose medium with Tween 80 [221. Culture conditions same as in Fig. 6
in the second [62]. The concentration of biomass obtained in the first stage at a dilution rate of 0.21 h - l was 2.2 rag. m1-1, which corresponded to an insoluble protein content of 0.72 mg- m l - [ In the second stage, the insoluble protein content decreased steadily from 0.72 mg. ml - l at a dilution rate of 0.02 h -1 to 0.45 mg • m1-1 at a dilution rate of 0.0125 h -1. In the absence of any other carbon source available in the induction stage, a necessary increase in residence time beyond a certain limit produced a diminishing effect on the usual turn-over rate of proteins of normal healthy fungal biomass. This resulted in a decrease of the metabolic coefficient of the system, and thus the release of active enzyme as a product of the stage occurred. However, the authors reported a 4.13-fold increase in productivity of cellulases achieved in a two-stage system as compared to a semi-continuous operation of a single stage CSTR employing 1% Solka Floc, which produced an average enzyme productivity of 5.39 x 10-a FPA m1-1 h -1 [57]. Although it has been reported [62] that celtulase biosynthesis is growth associated, it is not quite clear why a two-stage (growth on glucose in the first and induction on cellulose in the second) set-up is warranted. In a more recent report [63] a kinetic model has been proposed for reassessment of cellul~ and amylase activities of a newly isolated strain of Trichoderma as a function of specific uptake rate of carbon source based on operon theory: logE = - n p log v + log /1
otb KZmax
k"
'
(2)
Cellulase Biosynthesis and Hydrolysis of Cellulosic Substances
65
where E = Specific enzyme concentration, units -1-1, = Cell mass concentration, g. 1-1, X Maximum value of A E / A X commensurate with total operator, Zmax = n = Number of co-repressor molecules bound with one repressor molecule, = Empirical exponent, P = Specific uptake rate of carbon source, m. mole sugar g- cell-1 - h -1, P Ol = Ratio of free operator to the total operator, = avalue when corepressor concentration is maximum, o~b k = Proportionality constant, K = Equilibrium constant. The value of E / X are significantly affected by the specific uptake rates of carbon source. The specific rate of extracellular hydrolase enzyme synthesis has been expressed in the usual manner, namely,
e - xdtl dE _ alav-~
(3)
The value of b for cellulase was found to be 4.4 and e is remarkably retarded as the value of v increases. This suggests that a controlled addition of carbon source (glucose, cellobiose) to the culture medium, so as to restrain the specific rate of carbon source utilization by Tn'choderma sp., might enhance the production of saccharifying enzymes. An optimal design of saccharifying enzyme production system is likely to be possible through a coupling of e with other rate equations such as growth, carbon source uptake, and respiration. The report does not deal with any data on the exo- and endo-glucanases or/3,1-4 glucosidase components of the strain, nor to what magnitude the v values affect the cellulase components when tested in isolation. It is certainly necessary to have more work done on the kinetics of cellulase synthesis to confirm or prove otherwise the various postulates proposed so far.
5. H y d r o l y s i s o f Cellulosic S u b s t a n c e s a n d the Cellulase A d s o r p t i o n Desorption Mechanism One of the most significant and important aspects of cellulose hydrolysis is the process of adsorption of the active components of cellulases and their recovery from the adsorbed phase, in a manner that the initial ratios of the active components are maintained as far as possible. This is important because of the high cost of producing cellulase enzyme. It has been conclusively established by the enzymologists working with the cellulase enzyme that most cellulosic substrates such as paper, agricultural residues, like bagasse and rice husk, etc. contain some loose or free ends of cellulose chains which are reactive, some highly ordered crystalline cellulose, and a range of partially crystalline to partially amorphous cellulose, which exhibit restricted activity to hydrolysis by Cx, and thus require both C1 and Cx components for total hydrolysis. However, complete hydrolysis may not be possible even if the enzyme employed is complete in its active components, because of accessibility problems such as low ratios of amorphous to ordered structure
66
T.K. Ghose
or large impurities present in the substrate. Given the most favorable conditions with respect to substrate and cellulase, it is expected that the specific cellulolytic components will take part in the degradation of the substrate in proper sequence. The question is how one then can recover and reuse the enzyme if need be. The problem is complicated. The most active component of the enzyme system is supposed to be adsorbed first, followed by other components. The desorption process is likely to follow in the reverse order, i. e. the least active components are released first followed by the more active ones. Because of the appearance of large quantities of sugar because of almost instantaneous hydrolysis of the accessible regions of cellulose immediately after adsorption of the active components unter favorable conditions of pH and temperature, it is very difficult to measure the activity distribution between the solid cellulosic phase and the homogeneous liquid phase. Adsorption-desorption equilibrium data with respect to realistic substrates are meagre. It is already known [25, 94] that when cellulosic substrates, such as newsprints and bagasse, are exposed to cellulases of increased activities, the exponential rates of initial hydrolysis also increase proportionally. Samples of 5% suspensions of newsprints were subjected to T. viride cellulase (FPA - 3.5); the enzyme is very strongly and rapidly adsorbed initially and is gradually released later (Fig. 14) [94]. The percentage of original activities remaining in the solution was evaluated for C1, Cx, and FPA by converting the respective activities after hydrolysis to enzyme protein equivalent, employing enzyme after precipitation. It can be seen that while there is not much of a difference in the amount of Cx activity in the hydrolysate between 45 ° and 50 °C, the release of FPA and C1 activities in different amounts, due to denaturation of the active protein at 50 °C as suggested by the authors, needs to be verified. Further studies on cellulase adsorption equilibria on newsprint [94] reveal that Cx and C1 are adsorbed to about the same extent. Based on these values the authors have assumed a distribution coefficient of 0.04 FP un!ts!ml . . as equivalent protein for the purpose t~r umts/g sonos of designing the enzyme recovery system. However, when the hydrolyzed solid residues were centrifuged from the original sugar solution and resuspended in buffer solution to determine the amount of enzyme released by desorption, no activity could be detected in the supernatant. Based on these results, a modified process design has been proposed for a combined process of adsorption, hydrolysis, cell recycle, and enzyme reuse [94]. A somewhat different approach [23] has been made through extensive studies in the adsorption-desorption of single and mixed enzymes on sugar-cane bagasse, in an effort to clarify the observed increase in the rate of bagasse hydrolysis by the mixed enzyme. There was also the desire to determine the equilibrium characteristics of cellulase components for the purpose of their recovery and reuse. Because hemicellulose was one of the substrates, xylanase activity was assayed in addition to C~ and Cx. The culture filtrates of T. viride QM 9414 andA. wenti Pt 2804 were precipitated by acetone (1 : 3 v/v) and dissolved separately in 0.05M citrate buffer (pH 4.8) to get the enzyme solutions of the same protein level. These preparations were used to obtain mixed enzymes of different proportion. The effectiveness of the mixed enzyme (TV + A w ) on bagasse substrate (t~-cellulose-46%, hemicellulose-25%, lignin-20.5%; size-150 + 160 mesh) with a solid:
Cellulase Biosynthesis and Hydrolysis of Cellulosic Substances
67
IOO 8O
I i i i
6O 40
~
5
"
C • 50*C
20 o
IO0
~-
80
1
C
=
~.
l
I
A -----.._.__._~_
6o -6
t
Zx45"C
40
A 50"C
20 I,
iO0
I
o 45"C
8o
• 50"C
..._n Fig. 14. Release of adsorbed enzyme (Tv QM 9414) at pH 4.8 during the hydrolysis of ground newsprint for initial FPA of 3.5 and 5% solids after 40 min of contact [941
4O
]r-"--------.............i -
~ zo 0
I 0
I ~0
I 20 Time (hr)
I 30
I 40
50
liquid ratio of 1 : 20 (w/v) are shown in Table 10. The increased rate of hydrolysis of bagasse in a modified enzyme solution (Tv : A w = 1 : 1) compared to T. viride enzyme alone is due to a higher capacity of adsorption of the xylanase component on bagasse (Table 11). None of the enzymes having low Xy activity in T. viride enzyme or low C1 and Cx activities in A. wentii enzyme could hydrolyze bagasse at the rate the mixed enzyme did. The adsorbed values of different enzyme components, however, do not represent the equilibrium values between the solid and the homogeneous phases under the conditions of rapid hydrolysis (50 °C, pH 4.8). From the data on the effects of temperature on the adsorption of mixed enzyme components on bagasse, it has been found that during batch contact, the adsorption of C1, Cx, and Xy components decreases with an increase in temperature from 5 to 50 °C, while the specific adsorption progressively increases with the highest value at 50 °C. The pH effects showed that the adsorption of enzyme components increases on either side of the scale with minimum values at 4.8, whereas the specific adsorption shows the maximum values at 4.8. Thus, specific adsorption of active cellulase components (activity units per mg protein) was found to be maximum at 50 °C and pH 4.8, which supports the observed increased rate of hydrolysis of bagasse under these conditions. Distribution coefficients (DC = ratio of activity units per ml in the liquid phase to that remained adsorbed per gm on solid phase under equilibrium conditions) of C1, Cx, and Xy components were determined every 15 min of contact at 5 °C between bagasse (500 mg) and the mixed enzyme in a 1 : 20 (w/v) proportion. This was followed by
68
T.K. Ghose
Table 10. Rate of hydrolysis of cellulose and bagasse in single and mixed enzyme at constant protein level at 50 °C and pH 4.8 [231 ~
~
e
Hydrolysis in first h mg Rs produced-m1-1 Tv
Aw
Tv : Aw 2:1
Cellulose (5%)
3.2
0.9
Bagasse (5%)
1.8
1.6
Substrate
Tv : Aw 1:2
Tv : Aw 1:1
2.7
1.9
2.5
2.8
2.6
3.4
x.~
Enzyme-culture f'fltrate of (i) T. viride QM 9414 (protein-1.8 mg • m1-1 , CI-8.7 × 10 -3 IU • m1-1, Cx-36.6 IU. m1-1, FP activity-l.36 IU - m1-1 , Xy-7.8 IU. m1-1 , (ii) A. wentii, Pt 2804 (iarotein1.8 mg- m1-1, C~-3.86 × 10- 3 IU • m1-1 , Cx-7.1 IU • m1-1 , FP activity-0.6 IU • m1-1, Xy-16.6 IU. m1-1
Table 11. Adsorption of different enzyme components on bagasse at 50 °C, pH 4.8 after 15 min of contact time [231, figures in parentheses refer to initial values Enzyme solution
Adsorbed values of Protein (mg. m1-1 )
Culture fdtrate of T. viride Culture filtrate ofA. wenti Culture filtrate ofTv +Awin 1 : 1 ratio
0.305 (1.8) 0.292 (1.8) 0.28 (1.8)
C1 × 103 (IU- m1-1 ) 2.62 (8.7) 1.27 (3.86) 1.87 (6.4)
Cx (IU • m1-1 ) 9.5 (36.6) 1.65 (7.1) 6.47 (26)
Xy (IU. m1-1) 2.12 (7.8) 5.0 (16.6) 3.5 (12)
three washings with 10 ml o f 0.05 M citrate buffer (pH 4.8) each time and separating the supernatant by centrifugation. Based on these values o f DC ( C t - 0 . 0 2 , C x - 0 . 0 1 7 and X y - 0 . 0 1 2 ) , an effective design for the recovery o f the enzyme components has become possible. It is important to be careful about what one does in the assay o f enzymes in the hydrolysate under conditions o f rapid saccharification (50 °C, p H - 4 . 8 and highly active ceUulase), as the amount o f active proteins capable o f displaying various component activities of celtulase may be under certain conditions as low as 30% o f the total protein o f the liquid homogenous phase. Before we a t t e m p t to determine the cellulolytic activities in hydrolysate having appreciable amounts o f sugars at 50 °C, we should have a dependable method of identifying active cellulolytic components and the non-active protein fractions.
5.1 Kinetics of Cellulose Hydrolysis Not much attention has been given to the study of cellulose-cellulase kinetics. Only a few contributions [21, 33, 34, 78] which appeared during the last five years dealt with
Cellulase Biosynthesis and Hydrolysis of Cellulosic Substances
69
the kinetic modelling on enzymatic hydrolysis of cellulosic substances. Lack of any valid clarification of the biochemical mechanism of the hydrolysis process and a marginal economic feasibility of enzymatic hydrolysis of cellulose could not possibly draw sufficient interest toward engineering analysis of the hydrolysis kinetics. However, it may be useful to review the current status of our knowledge in this field. Ghose and Das [21] reported that enzymatic hydrolysis of cellulosic materials is a noncompetitive product-inhibited system. They based their analysis on the two products of pure cellulose hydrolysis, cellobiose, and glucose as seen in the T. viride enzyme system. From the magnitude of the value of activation energy (14870 cal • mole - l ) for the rice huU-celtulase system they suggested that the overall process of cellulose depotymerization and saccharification could be either diffusion- or biochemical reaction rate-controlled. The Lineweaver Burk model for the system is v
X
Kp
+ ~,
(4)
where Rate of product formation, time -1, V = Maximum rate of product formation, time - l , gm = Michaelis constant, mg. m1-1, p = Product concentration, mg • mlKp = Equilibrium constant for cellobiose and glucose, S = Substrate concentration, mg • m1-1 . The quantitative nature of the inhibition process by both glucose and cellobiose has been described [21]. Later Ross and Updegraff [78] developed a method of correlation that could isolate the two effects (diffusion and reaction) which permits identification of zones where one or the other process is in command. They worked with direct bioconversion of cellulosics (newsprint) into M y r o t h e c i u m verrucaria cells. The assumption of Fickian diffusion process coupled with Michaelis-Menten kinetics resulted in suggestions which appear reasonable. Two zones of influence are evident from the data presented by them: (1) a zone of biochemical-reaction control which is reached ever a prolonged time, and (2) a diffusion-controlled region which is encountered over short periods only. It is not quite clear whether fresh substrate coming in contact with the new cells producing a high reaction rate or the dispersion of the cellulosic substances is retarded by clump formation or whether a combination of both these processes are in operation. The model proposed by them represents both of the regions: M =-
where M = t = D = C =
D
~-~ Ix = o d t = k t - K m
~X/r~-~,
Mass uptake at boundary, Time, Diffusion coeff, according to Fick's Second Law, Concentration of product,
(5)
70
T.K. Ghose
x = Distance normal to reaction surface, k = Reaction rate constant, K m = Michaelis-Menten constant, A value of Kx/-n-D >>k indicates a reaction-controlling regime. Because a clear mechanism of enzymatic depolymerization and saccharification of cellulose is still lacking, it is not quite clear exactly what is diffusing, i. e., what kind of enzyme component and what product? A non-competitive-product-inhibited model based on the assumption that all soluble carbohydrate is cellobiose (3-1,4-glucosidase absent?) predicts the experimental data [33] very well, as tong as substrate concentration is more than 0.3%. As a first approximation, the equation for the non-competitive-product-inhibited cellulose hydrolysis model has been suggested by the authors as
t = 3- 27 (1 +
\
!G_s] ~ In [G2] 4.68 x 10 - a ] [G2]o- G2
6.58 x 10-3G~ +
[G2] 2 9.56 x 10 -3'
(6) where t = Reaction time, [G2] = CeUiobiose in polymefized form (insoluble), G2 = Cellobiose in soluble form. The authors also concluded that the simplified product inhibited model is an approximate kinetic description of overall cellulose kinetics. From these three studies it can be assumed that a multisubstrate model with substrate- [86] and product-inhibition [21,33] functions should be the most updated description of cellulose-cellulase kinetics. The kinetic constants would be determined as a function of the manner and extent of pretreatment. Huang [34] also studied enzymatic hydrolysis of cellulose by T. viride cellulase in a batch reactor at several substrate levels and enzyme concentrations at 50 °C and pH 4.8. A kinetic model has been proposed, based on adsorption of enzyme on the solid substrate surface followed by sluggish reaction and subsequent product inhibition. It correlates data for up to 70% conversion of the insoluble substrate into product. The model is a first order rate equation in the linear form: 1 _
vo
I+K,(E)o
k z X l m K l (E)o
. ~_1 + 1
(S)0
.
(7)
k2(E)o
Vois obtained by fitting the product concentration-time data to a polynomial form. This eliminates the initial rate measurement, reproducibility scatter, and a more reliable value is obtained. The final form of the equation (5) in terms of time of hydrolysis is t=
Xlm K1 (S)o + 1 + K i (E)o (p) + 1 + KI (E)o + K3 (S)o (e)Z k2 X~m K I (E)o (S)o 2 k 2 Xl m K1 (E)o (8)2 1 + K1 (E)o + K3(S)o (p)3. + 3 k 2 X t r n K l (E)0 (S)~
(8)
Cellulase Biosynthesis and Hydrolysis of Cellulosic Substances
71
The coefficient of the (P) term directly gives the value of 1/Vo as long as all the polynomial coefficients are positive. The model is based on the accepted reaction mechanism kl E + S ~ X1, k_l
(9)
X 1 k2 E + P
(10)
E+p5 3 k_ 3 X2
(11)
and the assumption that the enzyme must first be adsorbed on the cellulose surface, according to Langmuir adsorption isotherm Xl _
Xlm
KI(E)
1 + K , (E)'
(12)
where X 1 = Amount of enzyme adsorbed per unit mass of substrate, x2 = Amount of product formed per unit mass of substrate, X l m = Saturation amount of enzyme adsorbed per unit mass of substrate, (E) = Enzyme concentration, kl K1 k_ l " Since (S) >> (E), a steady state consumption is valid and dX2 _ O. dt
Thus,
(E)o = (E) + (X1) + (X2)
and
v
_ d(P)_ k2 (X1). dt
(13)
One of the basic shortcomings in all models so far proposed and discussed here is the lack of information on the amount of active cellulase that is kept adsorbed by the lignin in the case of native substrates, even though lignin is not said to be participating in the hydrolysis of cellulosics.
6. Conclusion
In the present review, all aspects of the cellulose-cellulase system have been covered, as far as information is available. Lack of our understanding of the mechanism of cellulose degradation in the terms of biochemical steps has been discussed, It is obvious that the structural and chemical properties of cellulosic substances are intimately linked with the process of degradation of the substrate by the enzyme system. The work of Pettersson [71] on T. viride enzymes, of Wood and McCrae [99] and of Halliwell [29] on T. koningii
72
T.K. Ghose
enzymes, and of Eriksson and Pettersson [14, 15] on S. pulverulentum enzymes is of significance. So far no one has been able to establish that the "reconstructed" cellulase consisting of the separated and identified fractions is identical with the culture filtrates of T. viride, T. koningii, and S. pulverulentum. The findings of Suzuki [83] on the constitutive nature of a system of membrane-bound cellulase enzyme produced by the bacterium P. fluorescens vat. cellulosa and on the transition between repression and de-repression of cellulase formation occurring in a narrow range of sugar concentration appear to be a very interesting contribution to the effort to clarify the nature of the enzyme system. A metabolic regulation of the fungal activities, particularly of T. viride, may lead to an adapted or even a mutant strain that is ~-1,4 glucosidase negative. If wide use for cellobiose is found, end-products of cellulose degradation by such a strain could be of important economic significance, even though some amount of glucose would still appear in the hydrolysate because of endo-glucanase activity [82]. Results of the initial studies of Ghose et al. [24] show that under a properly controlled batch culture cellobiose sugar can be quantitatively rejected by yeast cells grown on a mixture of glucose and cellobiose. The work of Wilke and Yang [94] and Ghose et at [23, 24] on the adsorptiondesorption of cellulase on cellulose and bagasse are promising for both the recovery and reuse of cellulase. It seems that kinetic models of hydrolysis of cellulosic substrates for the design of optimum hydrolysis reactors can only be done empirically [21, 33, 34, 78] and it has not been possible to propose any realistic models on account of poor knowledge of the basic enzymatic reaction mechanism. A few significant areas of in-depth investigation in the cellulose-cellulase field include: - development of thermophilic strains of fungus that are capable of degrading native cellulose, - mixed enzymesand mixed cultures of two strains on a given substrate. This is a very promising area for future endeavors, - dependable methods of recovery and reuse of ceUulasefrom hydrolystate, - development of a reliable method of continuous enzymatic hydrolysisof cellulosic substances, - development of on-line monitoring and analysis of cellulolytic activities during biosynthesis of the enzyme. In this rather extensive review, materials pertaining to direct conversion of cellulosic substances into microbial biomass have not been covered, although some general discussions have been brought in. This is a new and fast-developing area and any review, however promptly compiled, is likely to become old very soon. In the Biotechnology Report [5] it has been concluded that cellulosic wastes are a large untapped source of SCP. The high prices for protein may lead to several processes that are of economic interest. It is felt that direct fermentation of cellulosic substrate by thermophilic actinomycetes offers most promise of becoming a commercially viable process. Perhaps a separate review of the entire area would be of significant interest in about a year.
7.
Acknowledgements
This review was compiled during the author's stay at the Mikrobiologisches Institut, EidgenOssische Technische Hochschule, Ziirich, as a visiting professor in the summer of 1976. The author sincerely thanks Professor A. Fiechter, Head of the Institute, for providing necessary facilities for the work. The financial assistance provided by the
Cellulase Biosynthesis and Hydrolysis of Cellulosic Substances
73
Swiss Technical Cooperation Scheme, Berne, and the deputation leave granted by the Indian Institute of Technology, Delhi are greatly appreciated. Some useful suggestions made by Dr. Mary Mandels of US Army Development Center, Natick, Mass., USA, and Mr. 1(. S. Bisaria of lIT Delhi are gratefully acknowledged.
8. Nomenclature t/
b C D E E (E)o (G2) G2 k k, k~, k 3
ab K Zmax/kn [Eq. (3)] np [Eq. (3)1, concentration of product, ML- 3 Diffusion coefficient according to Fick's second law specific enzyme concentration [Eq. (2)], L- 3 enzyme concentration [Eq. (12)1, ML- 3 initial enzyme concentration, L - 3 cellobiose in polymerized form (insoluble) cellobiose in soluble form proportionality constant, [Eq. (2)1 dimensionless reaction rate constant, T - 1
k_ 3
K, K~, K 3
Km Kp M n
P P
(P) S (S)o t P
V ~'o
X X
X1 X1 m
x~ Y Zmax c~b 6 V
/z /Jan (Um)pH
equilibrium constant, dimensionless Michaelis constant, ML - 3 equilibrium constant for cellobiose and glucose, dimensionless mass uptake at boundary, ML- 2 number of corepressor molecule bound with one repressor molecule, dimensionsless empirical exponent IEq. (2)1, dimensionless product concentration [Eq. (4)], ML- 3 product concentration, ML- 3 substrate concentration, ML- 3 initial substrate concentration, ML-3 time, T - 1 rate of reaction or product formation, T - 1 maximum rate of product formation, T -1 initial rate of reaction, T - 1 cell mass concentration, ML- 3 distance normal to reaction surface, L amount of enzyme adsorbed per unit mass of substrate, dimensionless saturation amount of enzyme adsorbed per unit mass of substrate, dimensionless amount of product formed per unit mass of substrate, dimensionless yield coefficient, dimensionless maximum value of AE/AX, dimensionless ratio of free operator to total operator c~value when corepressor concentration is maximum specific enzyme synthesis rate, T -1 specific uptake rate of carbon source, T -1 specific growth rate, T -1 maximum specific growth rate at given pH and temperature, T -1 Um as function of pH at given temperature, T -1 maximum (/~m)pH at given temperature, T -1
74
T.K. Ghose
9. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38.
Almin, K. E., Eriksson, K. E., Pettersson, B.: Europ. J. Biochem. 51,207 (1975). Ander, P., Eriksson, K. E.: Svensk Papperstidming 18,641 (1975). Andren, R. K., Mandels, M., Medeiros, J. E.: Appl. Polym. Syrup. 28,205 (1975). Bassham, J. A.: Biotechnol. Bioeng. Symp. 5, 9 (1975). Bellamy, W. D.: Biotechnol. Bioeng. 16, 869 (1974). Berghem, L. E. R., Pettersson, L. G.: Europ. J. Biochem. 37, 2 (1973). Berghem, L. E. R., Pettersson, L. G.: Europ. J. Biochem. 46, 295 (1974). Brown, D. E., Halsted, D. J., Howard, P.: Proc. Symp. Enz. Hydrol. Cellulose, Aulanko, Finland, Eds.: M. Bailey, T.-M. Enari, and M. Linko, March 1975, p. 137. Brown, D. E., Halsted, D. J.: Biotechnol. Bioeng. 17, 1199, (1975). Crawford, R. L., Kirk, T. K., Harkin, J. M., McCoy, E.: Appl. Microbiol. 25,322 (1973). Enari, T.-M., Markkanen, P., Korhonen, E.: Proc. Symp. Enz. Hydrol. Cellulose, Aulanko, Finland, Eds.: M. Bailey, T.-M. Enari, and M. Linki, March, 1975, p. 137. Eriksson, K.-E., Pettersson, B.: In: Biodegradation of Materials, Vol. 2. Appl. Sc. Pub. Ltd., London, 1972, p. 116. Eriksson, K.-E., Pettersson, B., Westermark, V.: FEBS Letters 49, 282 (1975). Eriksson, K.-E., Pettersson, B.: Europ. J. Biochem. 51,193 (1975). Eriksson, K.-E.: Europ. J. Biochem. 51,213 (1975). Eriksson, K.-E.: Proc. Symp. Enz. Hydrol. Cellulose, Aulanko, Finland, Eds.: M. Bailey, T.-M. Enari, and Linko, March, 1975, p. 263. Forss, K.: Proc. Symp. Enz. Hydrol. Cellulose, Aulanko, Finland, Eds.: M. Bailey, T.-M. Enari, and M. Linko, March, 1975, p. 26 Ghose, T. K.: Biotechnol. Bioeng. 11,239 (1969). Ghose, T. K., Kostick, J. A.: Symp. Adv. Chem. Set. (ACS) 95,415, (1969). Ghose, T. K., Kostick, J. A.: Biotechnol. Bioeng. 12, 921 (1970). Ghose, T. K., Das, K.: Adv. Biochem. Engng., Vol. I. Eds.: T. K. Ghose and A Fiechter. Springer-Verlag, Berlin- Heidelberg-New York, 1971, p. 55. Ghose T. K., Pathak, A. N., Bisaria, V. S.: Proc. Symp. Enz. Hydrol. Cellulose, Aulanko, Finland, Eds.: M. Bailey, T.-M. Enari, and M. Linko, March 1975, p. 111. Ghose, T. K., Bisaria, V. S., Dwivedi, C. P.: Abs. Papers V Int. Ferment, Symp., Berlin, Ed. H. Dellweg, 1976, p. 439. Ghose, T. K., Bhargava, R., Bajpai, R. K.: Biotechnol. Bioeng. 19, 605 (1977). Ghose, T. K., Dwivedi, C. P.: personal Communication, 1976. Halliwell, G.: Biochem. J. 95, 35 (1964). Halliwell, G.: Biochem. J. 95,270 (1965). HalliweU, G., Riaz, M.: Biochem. J. 135,587 (1973). Halliwell, G.: Proc. Symp. Enz. Hydrol. Cellulose, Aulanko, Finland, Eds.: M. Bailey, T.-M. Enari, and M. Linko, March, 1975, p. 319. Heichel, G. H.: Biotechnol. Bioeng. Symp. 5, 43 (1975). Holdren, J. P., Ehrlich, P. R.: Am. Sc. 62,282 (1974). Honeyman, J.: In: Recent Advances in the Chemistry of Cellulose and Starch, Interscience N. Y., 1959. Howell, J. A., Stuck, J. D.: Biotechnol. Bioeng. 17,873 (1975). Huang, A. A.: Biotechnol. Bioeng. 17, 1421 (1975). Hulme, M. A., Stranks, D. W.: Nature [Londonl 226,469 (1970). Humphrey, A. E., Armiger, W. A., Lee, E., Moreira, A.: Abs. Papers V Int. Ferment. Syrup. Berlin, Ed. H. Dellweg, 1976, p. 431. Iwasaki, T., Hayashi, K., Fubatsu, M.: J. Biochem. (Tokyo) 57,467 (1965). Keagy, D. M.: In: Testimony of Public Hearing on National Resources, Planning and Public Works, Terrace, Calif., US Dept. of HEW, National Center for Urban and Industrial Health, March 1, 1968, p. 20.
CeUulase Biosynthesis and Hydrolysis of Cellulosic Substances 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82.
75
King, K. W.: In: Adv, Enz, Hydrol. of Cellulose and Related Materials, Ed. E. T. Reese, Pergamon Press, London, 1963, p. 159. Kirk T. K., Harkin, J. M., Cowling, E. B.: Biochem. Biophys. Acta 165,145 (1968). Kirk T. K., Larsson, S., Miksche, G. E.: Acta. Chem. Scand. 24, 1470 (1970). Kirk T. K." Ann. Rev. Phytopathol. 9, 185 (1971). Kirk T. K., Lorenz, L. F.: Appl. Microbiol. 26, 173 (1973). Kirk T. K., Chang, H.: Holzforschung 28,217 (1974). Kirk T. K., Chang, H.: Holzforschung 29, 56 (1975). Kirk T. K., Chang, H., Lorenz, L. F.: Wood So. Technol. 9, 81 (1974). Kirk T. K.: Biotechnol. Bioeng. Syrup. 5,139 (1975). Koenigs, J. W.: Mater. Organ. 7, 133 (1972). Koenigs, J. W.: Arch. Mikrobiol. 99, 129 (1974). Koenigs, J. W." Biotechnol. Bioeng. Symp. 5,151 (1975). Krupnova, A. V.: Gidrolizn. Lesokhim. Prom. (USSR) 16 (3), 8 (1963). Leisola, M.: Proc. Samp. Enz. Hydrol. Cellulose, Aulanko, Finland, Eds.: M. Bailey, T.-M. Enari, and M. Linko, March 1975, p. 297. Li, E. H., Flora, R. M., King, K. W.: Arch. Biochem. Biophys. 111,439 (1965). Linko, M.: Kemia-Kemi2,602 (1975). Liu, T. H.: Ph.D. Thesis, Virginia Polytechnic Inst. Blakesburg, Va, USA, 1970. Mandels, M., Parrish, F., Reese, E. 7".:J. Bact. 83,400 (1962). Mandels, M., Weber, J., Parizek, R.: Appl. Microbiol. 21,152 (1971). Mandels, M., Hontz, L., Nystrom, J.: Biotechnol. Bioeng. 16, 1471 (1974). Mandels, M., Sternberg, D., Andreotti, R. E.: Proc. Symp. Enz. Hydrol. Cellulose, Aulanko, Finland, Eds.: M. Bailey, T.-M. Enari, and M. Linko, March 1975, p. 81. Mandels, M., Sternberg, D.: J. Ferment. Tech. (Jap.) 54 (4), 267 (1976). Millet, M. A., Baker, A. J., Sattern, L. D.: Biotechnol. Bioeng. Syrup. 5,193 (1975). Mitra, G., Wilke, C. R.: Enzyme Utilization of Waste Cellulosics Rept. No. LBL 2334 Rev. Univ. of Calif., Berkeley, 1975, p. 92. Nagai, S., Onodera, M., Aiba, S.: Europ. J. Appl. Microbiot. 3, 9, (1976). Nisizawa, K.: J. Ferment. Technol. (Japan) 51,267 (1973). Niwa, T., Kawamura, K., Nisizawa, K.: Proc. V Syrup. Cellulases and Related Enzymes. Cellulase Assoc., Osaka Univ. Japan, 1965, p. 44. Norkrans, B.: Physiol. Plant. 3, 75 (1950). Nystrom, J. M., Kornuta, K. A.: Proc. Symp. Enz. Hydrol. Cellulose, Aulanko, Finland, Eds.: M. Bailey, T.-M. Enari and, M. Linko, March 1975, p. 181. Ogawa, K., Toyama, N.: Proc. V Symp. CeUulasesand Related Enzymes. Cellulase Assoc., Osaka University, Japan, 1965, p. 85. Ogawa, K., Toyama, N.: J. Ferment. Technol. (Jap.) 45,671 (1967). Ogawa, K., Toyama, N.: J. Ferment. Technol. (Jap.) 46, 367(1968). Pettersson, L. G.: Proc. Symp. Enz. Hydrol. Cellulose, Aulanko, Finland, March 1975, p. 255. Pringle, S. L.: EUCEPA Proc. Int. Symp. (Report No. 3) Madrid, Mat, 1974. Poole, N. J., Smith, L: In: Octagon Papers 3, Dept. of Extra-Mural Studies, University of Manchester, 1975, p. 85. Rautala, G. S.: Ph.D. Thesis, Virginia Polytech. Inst., Blakesburg, Va. USA, 1967. Reese, E. T., Siu, R. G. H., Levinson, H. S.: J. Bact. 59,485 (1950). Reese, E. T., Segal, L., Tripp, V. M.: Text. Res. J. 27, 626 (1957). Reese, E. T., Maguire, A.: Dev. Ind. Microbiol. 12,212 (1971). Ross, L. W., Updegraff, D. M.: Biotechnol. Bioeng. 13, 99 (1971). Selby, K., Maitland, C. C.: Biochem. J. 104, 716 (1967). Smith, D. L. O.: In" Octagon Papers 3, Pub. Dept. of Extra-Mural Studies. University of Manchester, 1975, p. 100. Streamer, M., Eriksson, K.-E., Pettersson, B.: Europ. J. Biochem. 59, 607 (1975). Sternberg, D.: Appl. Environ. Microbiol. 31 (5), 648 (1976).
76 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100.
T.K. Ghose Suzuki, H.: Proc. Symp. Enz. Hydrol. Cellulose, Aulanko, Finland, Eds.: M. Barley, T.-M. Enari, and M. Linko, March 1975, p. 155. Tomita, Y., Suzuki, H., Nisizawa, K.: J. Ferm. Technol. 52,233 (1974). Toyama, N., Ogawa, K.: Biotechnol. Bioeng. Symp. 5,255 (1975). Van Dyke, B.: Ph.D. dissertation, MIT, Mass., USA, Sept., 1972. Virkola, N.-E.: Proc. Symp. Enz. Hydrot. Cellulose. Aulanko, Finland, Eds.: M. Barley, T.-M. Enari, and M. Linko, March, 1975, p. 26. Walseth,C. S.: Tappi 35,228 (1952). Westermark, U., Etiksson, K. E.: Acta. Chem. Scand. B 28,204 (1974). Westerm~k, U., Eriksson, K. E.: Acta. Chem. Scand. B 28, 209 (1974). Westermark, U., Eriksson, K. E.: Acta. Chem. Scand. B 29,419 (1975). Whittaker, R. H.: In: Communities and Ecosystem, Macmillan, New York, 1970. Wilcox, W. W.: In: US Forest Ser. Res. Paper FPL 70, Forest Products Lab., Madison, Wisc., 1968. Wilke,C. R., Yang, R. D.: Proc. Symp. Enz. Hydrol. Cellulose. Aulanko, Finland, Eds.: M. Barley, T.-M. Enari, and M. Linko, March, 1975, p. 485. Wood,T. M.: Biochem. J., 109, 217 (1968). Wood, T. M.: Biochem. J , 115,457 (1969). Wood, T. M.: Proc. 4th Int. Ferment. Symp., Jap. Ferment. Soc., Tokyo, 1972, p. 717. Wood, T. M., McCrae, S. I.: Biochem. J. 128, 1183 (1972). Wood,T. M., McCrae, S. I.: Proc. Symp. Enz. Hydrol. Cellulose, Aulanko, Finland, Eds.: M. Barley, T.-M. Enari, and M. Linko, March 1975, p. 231. Wood,T. M.: Biotechnol. Bioeng. Symp. 5, p. 111 (1975).
Metabolism of Methanol by Yeasts H. Sahm* Lehrstuhl fiir Biochemie und Biotechnologie der Technischen Universit/it Braunschweig und Gesellschaft for Biotechnologische Forschung, D-3300 Braunschweig-St6ckheim, West Germany. Present address: Institut ftir Biotechnologie der KFA Jiilich, D-5170 Jiilich, West Germany
Contents 1. 2. 2 a) 2 b) 2 c) 2 d) 3. 3 a) 3 b) 4. 4 a) 4 b) 5, 5 a) 5 b) 6.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dissimilation of Methanol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Oxidation of Methanol to Formaldehyde . . . . . . . . . . . . . . . . . . . . . . . . . . . . Oxidation of Formaldehyde to Formate . . . . . . . . . . . . . . . . . . . . . . . . . . . . Oxidation of Formate to Carbon Dioxide . . . . . . . . . . . . . . . . . . . . . . . . . . . Intracellular Localization of the Enzymes of the Dissimilatory Pathway of Methanol Assimilation of Methanol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Isotope Studies with Whole Yeast Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Enzyme Studies with Cell-Free Extracts . . . . . . . . . . . . . . . . . . . . . . . . . . . . Regulation of Enzymes of the Methanol Metabolism . . . . . . . . . . . . . . . . . . . . . Enzymes for Oxidation of Methanol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Regulation of the Flavin Synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Single-Cell-Protein(SCP) Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Production of Metabolites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
77 79 79 82 84 . . 86 88 91 92 94 94 96 98 98 99 100
1. I n t r o d u c t i o n In recent years, microbial utilization of noncarbohydrate carbon and energy sources, such as liquid n-paraffins, methane, or methanol for the production of single-cell proteins and biochemicals has been of growing interest all over the world. As an industrial fermentation substrate, methanol has several advantages over liquid n-paraffins. Methanol is available at a low cost with high purity, it is completely miscible with water in all proportions, and it can be easily and cheaply produced. As a consequence of the interest of industry in methanol-utilizing microorganisms, many studies on the isolation, identification, and physiology of those organisms were made during the last five to ten years. At present more than one hundred different methanol-assimilating bacteria are known that belong to several genera (Table t). These bacteria can be subdivided into two classes based on their growth substrate: 1. Obligate methylotrophs which grow only on carbon compounds containing no carboncarbon bonds (Crcompounds), e.g. methane, methanol, or methylamine; 2. Facultative methylotrophs which grow not only on C r c o m p o u n d s but also on substances such as glucose or succinate.
78
H.
Sahm
Table 1. Methanol-utilizing groups of bacteria 1. Obligate methylotrophs Methylobacter [11 Methylococcus [1, 21 Methylomonas I1, 3, 4] 2. Facultative methylotrophs Arthrobacter 15] Bacillus [5, 6] Hyphomicrobium [7] Klebsiella [ 81 Micrococcus 191
Methytocystis [1 l Methylosinus [11
Protarninobacter [ 101 Pseudomonas I 11, 121 Rhodopseudomonas [ 13, 141 Streptomyces [15 ] Vibrio I 161
The existence of yeasts capable of growing on methanol as the sole source of carbon and energy was discovered very recently. Ogata and coworkers published for the first time in 1969 a report on a yeast-utilizing methanol [ 17]. For the enrichment and isolation of methanol-assimilating yeasts it seems essential to suppress bacterial growth by the addition o f antibiotics to the enrichtment medium [ 18]. Van Dijken and Harder worked out optimal conditions for the isolation of methanol-utilizing yeasts; these are summarized in Table 2 [ 19]. Soil samples containing wood or flower residues are wellsuited for isolation of methanol-utilizing yeasts-probably since methanol is formed from the methoxy groups o f lignin or pectin.
Table 2. Optimal conditions for the isolation of methanol-assimilating yeasts [ 191 Factors influencing enrich ment
Optimum conditions
pH of the medium
4.5 (optimum pH for growth of yeasts on methanol is 5.0-6.0)
Antibiotics
Required in the physiological range of pH values for growth
Vitamin mixture
Necessary, since most methanol-assimilating yeasts require biotin and/or thiamin 0.1 to 0.5% (v/v). At methanol concentrations above 0.5% inhibition of growth occurred in some species
Methanol concentration Nature of inoculum
Soil samples rich in organic matter were good sources
As shown in Table 3, at present only a limited number of yeast species of a few genera (Candida, Hansenula, Pichia, Torulopsis) and only three fungi are known to grow on methanol. Therefore, the capacity of utilizing methanol as sole source of carbon and energy seems to be less widespread among yeasts and fungi than among bacteria. All methanol-assimilating yeasts tested so far are facultatively methylotrophic and unable to grow on the following Cl-compounds: methane, methylamine, formaldehyde, and formate as sole carbon- and energy source. However, the three methanol-utilizing fungi
Metabolism of Methanol by Yeasts
79
Table 3. Yeasts and fungi capable of growth on methanol 1. Yeasts Candida boidinii [18, 19, 201 rnethanolica 1211
N-16 1221 parapsilosis [23 ] Hansenula capsulata I201 glycozyma I201 henricii minuta 1201 non[ermentans [ 201 philodendra [201 polymorpha I20, 24] wickerhamff [20, 231
Kloeckera sp. No. 2201 a [17] Pichia haplophila 1231 lindnerii [ 251 pastoris 120, 231 pinus [201 trehalophila 1201 Torutopsis glabrata 126l methanodomerquii [271 methanoIovescens 1211 methanosorbosa [271 molischiana [ 201 nemodendra [ 201 nitratophila I201 pinus [20, 191
2. Fungi Gliocladhtm deliquescens 128] Paecilomyces varioti 1281 Trichoderma lignorum 1291
a Some microbiologists assert that this yeast should be defined as a strain of Candida boidinii
can also grow on formaldehyde, formate, and/or methylamine [28, 29]. Since some comprehensive reviews on Cl-metabolism in bacteria have been publsihed during the last few years [ 3 0 - 3 2 ] , this review concentrates in particular on the metabolism of methanol by yeast. An insight into possible applications of methanol-utilizing microorganisms in biotechnological processes is included.
2.
Dissimilation of Methanol
2 a) Oxidation of Methanol to Formaldehyde In methane- and methanol-utilizing bacteria, an enzyme catalyzing the oxidation of methanol to formaldehyde is present which has a gratifying uniformity among different bacterial strains. In vitro this enzyme is independent of NAD or NADP but absolutely dependent on the artificial hydrogen acceptor phenanzine methosulfate and on ammonium ions or methylamine as activators [33-38]. The methanol dehydrogenase is not specific for methanol as substrate; it also catalyzes the oxidation of other primary aliphatic alcohols up to a chain length of C16, as well as formaldehyde and acetaldehyde [37]. It has been suggested that the enzyme contains a pteridine as a prosthetic group, but the mechanism of this enzyme is still unknown [39]. The pteridine may act as a hydrogen acceptor; there is some evidence that in vivo the electrons obtained from the oxidation of methanol to formaldehyde are transferred to a cytochrome c [40, 41 ].
80
H. Sahm
Furthermore, it seems that methanol dehydrogenase is membrane-bound [42]. Studies of the oxidation of methanol by several methanol-utilizing yeasts have shown that formaldehyde is also the first intermediate in the sequential methanol oxidation. The enzyme which catalyzes this step has been purified and characterized from the yeasts, Kloeckera sp. [43, 44], Candida boidinii [45], Candida N-16 [22], and Hansenula polymorpha [46]. All these enzymes isolated from different yeast species are very similar. The visible absorption spectrum of the native enzyme with two peaks at 460 mm and 375 mm suggests the possible contribution of a flavin derivative to the enzyme as prosthetic group. This flavin was identified as FAD which is noncovalently bound to the enzyme protein [44, 45, 47]. The methanol-oxidizing enzyme of yeasts is thus a flavoprotein containing FAD as prosthetic group. The molecular weight of the enzyme was calculated to be 600000 to 670000. The enzyme is composed of eight identical subunits (mol. wt. 74 000 to 83 000) and contains 8 moles of FAD per mole of enzyme [45, 47]. Electronmicroscopic analysis of the enzyme of Kloeckera sp. and Hansenula polymorpha showed that each enzyme molecule is an octad aggregate, composed of two tetragons face to face [47]. The enzyme catalyzing the oxidation of methanol to formaldehyde uses oxygen as electron acceptor; therefore the stoichiometry of the reaction can be described by the following equation: CH3OH + 02 ~ HCHO + H20~. The enzyme is thus an alcohol oxidase. Studies for substrate specificity have shown that alcohol oxidase is not specific for methanol. It is also active towards ethanol, 1-propanol, 1-butanol, 2-propen-l-ol, 2-buten-l-ol, 2-chloroethanol, 2-bromoethanol, and formaldehyde; but it is inactive toward alcohols of longer alkyl-chain lengths than Cs, secondary or tertiary alcohols, and aromatic alcohols [44-47]. However, the Kinvalues indicate that methanol is the best substrate for this enzyme; the affinity of alcohol oxidase for alcohols decreased with the increasing length of the alkyl chain. There can be little doubt that this alcohol oxidase is a key enzyme in the metabolism of methanol by methanol-utilizing yeasts. It has also been found in some methanol-assimilating Pichia and Torulopsis strains by Kato et al. [48]. The necessary involvement of the enzyme in methanol oxidation by Candida boidinii is evident from the isolation of the mutant 4s, which lacks alcohol oxidase. While this mutant is unable either to oxidize or to grow on methanol, growth on glucose or ethanol is unimpaired in comparison with the wild type [45]. Furthermore, alcohol oxidase activity is very low or even nondetectable in yeasts grown on glucose, ethanol, or glycerol; during adaptation from glucose to methanol enzyme activity increases before growth on methanol begins [45, 48]. Recently Metha described a NAD-linked and glutathione-requiring methanol-oxidizing enzyme in CandMa boidinii and Pichia pinus [49]. Since in several alcohol-oxidase-negative mutants of Candida boidinii this NAD.linked oxidation of methanol was undetectable [50], it seems that this NAD-dependent methanol-oxidizing reaction is catalyzed by the following two enzymes: the alcohol oxidase and the NAD-linked glutathionerequiring formaldehyde dehydrogenase found in methanol-utilizing yeasts. NAD-linked alcohol dehydrogenase activity was found in all the methanol-utilizing yeasts tested.
Metabolism of Methanol by Yeasts
81
However, this enzyme from Candida boidinii does not catalyze the oxidation o f methanol and is constitutively formed [45, 48]. Therefore, at the present stage o f our knowledge two quite different groups o f methanoloxidizing enzymes exist in methanol-assimilating microorganisms, the one in bacteria and the other in yeasts. The main differences between these two enzyme groups are summarized in Table 4. Table 4. Comparison of properties of methanol dehydrogenase from methane- and methanolutilizing bacteria and of alcohol oxidase from methanol-assimilating yeasts
Prosthetic group Activation Molecular weight hmax KmCH3OH Substrates
Methanol dehydrogenase from bacteria
Alcohol oxidase from yeasts
Pteridine NH+, methylamine 120 000-140 000 (2 subunits) 280, 350 nm 0.02 mM Primary aliphatic alcohols up to a chain length of C16, formaldehyde and acetaldehyde
FAD 6000 000-670 000 (8 subunits) 280, 375,460 nm 0.2-2 mM Primary aliphatic alcohols up to a chain length of C5, and formaldehyde
In view of the H202 generation o f the alcohol oxidase reaction in methanol-utilizing yeasts catalase is indispensable for the degradation o f this toxic compound. As shown in Table 5, there are significant differences in catalase activity o f ceU-free extracts from
Table 5. Catalase activity of Candida boidinii grown on different carbon sources [52] Specific activity of catalase (U/mg protein) in cells of Candida boidinii harvested Carbon source
in the exponential growth phase
in the stationary growth phase
Ethanol Glucose Methanol
18 70 1450
320 310 1430
Candida boidinii grown on different carbon sources. When cells were harvested in the stationary growth phase, the catalase activity in methanol-grown cells is increased about fivefbld in comparison with cells cultivated on glucose or ethanol. The difference was much greater when cells were harvested at the exponential growth phase, since catalase is controlled by catabolite repression of glucose or ethanol [51, 52]. After transfer of glucose-grown cells o f Candida boidinii to a methanol medium, catalase activity increases very rapidly within 5 h [51 ].
82
H. Sahm
It is well known that catalase cannot only catalyze the cleavage of H202 to H20 and oxygen, but it can also function peroxidatively, oxidizing metabolites with H202 generated by oxidases. Methanol has long been known to serve as an electron donor for the peroxidative activity of catalase [53]; therefore several workers have suggested that catalase may also oxidize methanol during growth of yeasts on methanol [22, 51 ]. Van Dijken et at. partially purified catalase from the methanol-grown yeast Hansenula polymorpha and showed that if the ratio of the rate of hydrogen peroxide production to the catalase haem concentration is low (50/min) catalase is capable of oxidizing methanol, formaldehyde, and formate in the presence of H202 in vitro. However, if the ratio of the rate of hydrogen peroxide production to the catalase haem concentration is high (1000/min) peroxide is decomposed to oxygen and water [54]. Since the ratio of the intracellular concentration of catalase to alcohol oxidase and the intracellular concentration of hydrogen donors are not known, it is uncertain whether or not the hydrogen peroxide generated by alcohol oxidase is employed for the oxidation of methanol, formaldehyde, and formate in vivo. As shown previously in methanol-utilizing yeasts, alcohol oxidase and catalase are localized in microbodies [55-57], Roggenkamp demonstrated the peroxisomal nature of the isolated microbodies [52]. Therefore, we may be sure that methanol is oxidized to formaldehyde in methanol-assimilating yeasts by two reactions: The first is via an alcohol oxidase which requires oxygen and produces hydrogen peroxide besides formaldehyde. The second is the peroxidative action of catalase using the hydrogen peroxide which is generated by alcohol oxidase. 2 b) Oxidation of Formaldehyde to Formate Two enzymes which catalyze the dehydrogenation of formaldehyde have been characterized in bacteria grown on methanol, namely an NAD-linked formaldehyde dehydrogenase which requires reduced glutathione (GSH): (GSH) HCHO + NAD + H20 . . . . HCOOH + NADH2 [34,58] and an NAD-independent aldehyde dehydrogenase which may be a flavoprotein [34, 59]. This enzyme is assayed with the artificial hydrogen acceptor dichlorophenol-indophenol (DCPIP) and a wide variety of aldehydes can act as substrate: HCHO + DCPIP + H20 -+ HCOOH + DCPIPH2. The dual specificity of methanol dehydrogenase toward methanol and formaldehyde raises another possibility for formaldehyde oxidation. For example, Methylococcus capsulatus has none of the two aldehyde dehydrogenases described above, and therefore it has been suggested that in this organism formaldehyde is oxidized to formate by the methanol dehydrogenase [60]. In all methanol-utilizing yeasts studied so far an NAD-linked and glutathione-dependent formaldehyde dehydrogenase has been found [22, 48, 61, 62]. Formaldehyde dehydrogenase was purified 130-fold from Candida boidinii and was then homogeneous as
Metabolism of Methanolby Yeasts
83
judged by acrylamide gel electrophoresis [63]. The molecular weight of this enzyme was calculated to be 80000; for the subunits a molecular weight of 40000 was estimated. Therefore, it can be concluded that the formaldehyde dehydrogenase from Candida boidinii is a dimer composed of two probably identical subunits. The enzyme specifically requires gtutathione for activity, other thiol compounds such as cysteine, ~-mercaptoethanol, or thioglycolate were not able to replace glutathione. Besides formaldehyde, methylglyoxal also served as substrate for the enzyme, but no activity was found toward acetaldehyde, propionaldehyde, glycolaldehyde, or glyoxal [63]. A very similar enzyme has been isolated from Kloeckera sp. [62] and Hansenula polymorpha [46]. It has been assumed that the hemimercaptal spontaneously formed between formaldehyde and glutathione is the true substrate of formaldehyde dehydrogenases from human and beef liver and baker's yeast [64-67]. Using highly purified formaldehyde dehydrogenases from Candida boidinii and Hansenula polymorpha it was demonstrated that the product of the oxidation of formaldehyde is S-formylglutathione [46, 63]. Furthermore, only S.formylglutathione, but not formate, was able to support the reverse reaction of these enzymes. S-formylglutathione can hydrolyze nonenzymically to yield formate and glutathione but in human liver, as in Candida boidinii, a hydrolase seems present which catalyzes this reaction [63, 67]. Van Dijken reported that formate dehydrogenase from Hansenula polymorpha itself might have hydrolase activity [46]. It can be concluded that in methanol-utilizing yeasts formaldehyde is oxidized to formate by the following steps:
1)
2)
3)
H I H-C=O
H I H-C-OH I GS H-C=O I GS
H spontaneous l + GSH , H-C-OH I GS + NAD
+ H20
Formaldehyde , H-C=O dehydrogenase I GS
Hydrolysis , H-C=O I HO
+ NADH2
+ GSH
Recently it was found that formaldehyde can also be oxidized by alcohol oxidase of methanol-assimilating yeasts according to the following reaction: HCHO + 02 + H20 ~ HCOOH + H202 [46,68,69]. Sped et al. showed that the ability of formaldehyde to serve as a substrate for methanol dehydrogenase from bacteria is based on the fact that in aqueous solution formaldehyde is more than 99.9% hydrated [37]. This fact might also explain why formaldehyde is oxidized by alcohol oxidase from methanol-utilizing yeasts. At present it is not possible
84
H. Sahm
to quantify the role of alcohol oxidase in the oxidation of formaldehyde during growth of yeasts on methanol. The experiments on the oxidation of formaldehyde by cell suspensions of the wild type and a mutant 4s, lacking alcohol oxidase activity, of Candida boidinii suggest that alcohol oxidase is not implicated in formaldehyde oxidation in vivo [68]. However, these results may be of little relevance, because when cells grow on methanol the conditions in vivo may be different. A possible oxidation of formaldehyde to formate by alcohol oxidase and/or also by the peroxidative action of catalase [54] is of considerable importance for the overall energetics of the dissimilation of methanol in yeasts and therefore for the yield of biomass on methanol.
2 c) Oxidation of Formate to Carbon Dioxide In methanol.utilizing microorganisms studied so far an NAD-dependent formate dehydrogenase has been found to catalyze the last step of the dissimilation of methanol, the oxidation of formate to CO2. This enzymatic reaction can be described by the following equation: HCOOH + NAD -+ C02 + NADH2 [34,48,70]. Formate dehydrogenase was purified 19-fold from cell-flee extracts of methanol-grown Candida boidinii [63]. The molecular weight of the enzyme was determined to be 74000. After dissociation into subunits, a molecular weight of the polypeptide chains was estimated to be 36000. Since this is approximately half of the value found under native conditions, it seems that formate dehydrogenase from Candida boidinii is a dimer, composed of two probably identical subunits. The enzyme is specific for formate and NAD; NADP could not be used as electron acceptor. The Michaelis constants Km were found to be 13 mM for formate and 0.09 mM for NAD [63]. The reverse reaction, namely reduction of sodium bicarbonate by NADH2; could be detected but the reaction rate was found to be about 1/5000 of the dehydrogenase activity. This demonstrates that the equilibrium of the reaction strongly favors the dehydrogenation of formate. Very similar enzymes has been isolated from Kloeckera sp. [71 ] and Hansenula polyrnorpha [46]. While rather high Km-values of formate have also been reported for the formate dehydrogenases from Kloeckera sp. (22 mM) and Hansenula polyrnorpha (40 mM), recently van Dijken showed that formate dehydrogenase from Hansenula polymorpha has a 40-fold higher affinity for S-formytglutathione than for formate [46]. Furthermore, he could demonstrate that this enzyme hydrolyzes S-formylglutathione only in the presence of NAD. It is therefore suggested that the hydrolysis of S-formylglutathione by the formate dehydrogenase may lead to the formation of enzyme-bound formate, which is subsequently oxidized, Since similar results were obtained for the formate dehydrogenases of Candida boidinii and Pichia p&us tested in cell-flee extracts [46], it seems that the postulated hydrolysis of S-formylglutathione followed by oxidation of enzyme-bound formate is a general property of formate dehydrogenases of methanolutilizing yeasts.
Metabolism of Methanol by Yeasts
85
Since in all methanol-assimilating yeasts studied so far, the same or very similar enzymes responsible for the oxidation of methanol were found [48], methanol is most likely oxidized by yeasts via a pathway shown in Fig. 1.
GSH= H20
S CH3OH ~ ~ ,~HCHO FAD FADH2 NAD NADH2 GS NAD NADH2 Fig. 1. Pathway of methanol oxidation in methanol-utilizing yeasts. I: Alcohol oxidase, II: catatase; III: formaldehyde dehydrogenase; IV: formate dehydrogenase
CH30~a 02 02 HCHO"~ 2H20
Obligate methylotrophic bacteria assimilating Cl-compounds by the ribulose monophosphate cycle contain highly specific activities of glucose-6-phosphate dehydrogenase and 6-phosphogluconate dehydrogenase [72]. These two enzymes, together with hexulose phosphate synthase, hexulose phosphate isomerase, and glucose-6-phosphate isomerase, constitute a cyclic mechanism for the complete oxidation of formaldehyde to CO2, as shown in Fig. 2 [73]. This dissimilatory ribulose monophosphate cycle seems to be responsible for the oxidation of formaldehyde to CO: in some trimethylamineutilizing bacteria which apparently have no formaldehyde and formate dehydrogenases [74]. The presence of a ribulose monophosphate cycle and of glucose-6-phosphate isomerase, glucose-6-phosphate dehydrogenase, and 6-phosphogluconate dehydrogenase in Candida boidinii (Table 9) indicates that this cycle may also contribute to the oxidation of formaldehyde in yeasts. Since in these organisms formaldehyde and formate dehydrogenases are present, it may be speculated that this cyclic oxidation of formaldehyde might serve as a source of NADPH2 for biosynthetic purposes, as in Methylococcus capsulatus and Pseudomonas methanica [73].
Ribulose -5~5P~-~ 3 -
l-lexulose-6-P
V
NADPH2 Fig. 2. Dissimilatory ributose monophosphate cycle for the cyclic oxidation of formaldehyde. I: Hexulose phosphate synthase; II: 3-hexulose phosphate isomerase; III: phosphoglucoisomerase; IV: glucose-6-phosphate dehydrogenase; V: phosphogluconate dehydrogenase
Fru( '~£~se-6-P---~ Asslmdohon
/
NADP- J ~
6-O-Gluconate
Glucose-6-P
NADPH2 NADP
86
H. Sahm
2 d) lntracellular Localization of the Enzymes of the Dissimilatory Pathway of Methanol In plant and animal cells, flavin oxidase and catalase have been found to be at least partly located in microbodies [75, 76]. Avers and Federman [77] and Perlman and Mahler [78] demonstrate that microbodies exist in the yeast Saccharomyces cerevisisae. They detected several enzymes, including catalase, in the microbodies. Recently, Osumi et al. reported an increase of microbodies per cell that paralleled the increase of catalase activity in Candida tropicalis, when cells were transferred from a malt-extract medium to a medium containing n-alkanes [79]. Therefore, several groups investigated whether alcohol oxidase and catalase in methanolutilizing yeasts are also located in microbodies. As shown in Fig. 3 a, cells of Candida boidinii grown on methanol contained a cluster of bodies varying in size from 0.4 to 1/~m in diameter, containing crystalloid inclusions. These cytoplasmic organelles are surrounded by a single membrane and they may therefore be described as microbodies [80]. The number of organelles per sectioned cell was usually five or six, although up to 10 to 12 were found in rare cases. Microbodies were also observed in young buds. However, when Candida boidinii was grown on glucose or ethanol, microbodies were hardly detectable and their dimensions were smaller (0.2-0.4/lm) (Fig. 3 b). Similar small microbodies were observed in cells of the yeast Saccharomyces cerevisiae [77]. When cells of Candida boidinff or Kloeckera sp. grown on glucose were transferred to methanol medium, the size, and perhaps also the number of microbodies per sectioned cell, increased during a 5 to 8-h incubation time [52, 81 ]. After this period of cultivation, the size and number of the microbodies remained constant. Fukui et at. and van Dijken et al. also examined some other methanol-utilizing yeast belonging to the genera Hansenula, Kloeckera, Pichia, and Torulopsis with respect to their ultrastructure; and the large microbodies were also observed only in the methanol-grown cells of these genera [82, 83]. These results indicate that these microbodies may be associated with methanol metabolism. In order to study the occurrence of alcohol oxidase and catalase in these microbodies of methanol-grown yeasts cells, these organelles were isolated from Candida boidinii and Kloeckera sp. using differential centrifugation techniques [56, 57]. Spheroplast lysates of Candida boidinii were separated on noncontinuous Ficoll density gradients, resulting in a mitochondrial fraction and a fraction containing alcohol oxidase and catalase. Electron-microscopical examinations of this fraction revealed particles of 0.4 to 0.6/~m in diameter, which were sheroid and contained crystalloid inclusions (Fig. 4). Therefore, these particles have the same features as the microbodies in situ. Gel-electrophoretic studies of the microbody fraction demonstrated that alcohol oxidase and catalase are the main enzymes present in microbodies of Candida boidinii grown on methanol [56]. Localization of catalase and alcohol oxidase in microbodies of methanol grown yeasts was also demonstrated by cytochemical staining techniques [46, 55, 83]. Since in the microbodies the total catalase activity was at least 103 times higher than the activity of alcohol oxidase, catalase may well function mainly peroxidatively. The crystalloid inclusions observed in these microbodies are associated with alcohol oxidase and catalase, as shown cytochemically by the use of cerium chloride and
Metabolism of Methanol by Yeasts
Fig. 3a and b. Candida boidinii grown on methanol (a), grown on glucose (b). M: Microbody, N: Nucleus, V: Vacuole [801
87
88
H. Sahm
Fig, 4. Isolated microbodiescontainingcrystalloidinclusions1561
3,3-diaminobenzidine [46, 83]. Sahm et al. also found a close relationship of alcohol oxidase to the crystalloid by using a mutant lacking alcohol oxidase [80]. Furthermore, electromicroscopical studies indicated a correlation between the increase of alcohol oxidase and catalase activities and the development of crystalloids in the microbodies after transferring glucose-grown cells of Candida boidinii to a methanol medium [52]. All these results demonstrate in the crystalloid inclusions an intimate relationship between the hydrogen-peroxide-generating enzyme alcohol oxidase and the hydrogenperoxide-decomposing enzyme catalase. Since the activity of alcohol oxidase purified from Candida N-16 is irreversibly inhibited by 2.5 mM of hydrogen peroxide for 15 min [69], it is important that hydrogen peroxide be immediately decomposed. Formaldehyde dehydrogenase and formate dehydrogenase were observed exclusively in the cytoplasm [80, 81].
3. Assimilation o f M e t h a n o l Quayle and coworkers have found that in bacteria three different pathways exist by which reduced Cl-compounds are assimilated into new cell material [31 ]. The pathways involve either the incorporation of the C1 in the oxidized step of CO2 into phosphorylated sugar-the ribulose diphosphate cycle (Calvin cycle)-or the condensation of formaldehyde, either with a phosphorylated sugar-the ribulose monophosphate cycle (Quayle cycle)-or with glycine-the serine pathway.
Metabolism of Methanol by Yeasts
89
For a long time only one organism, Pseudomonas oxalaticus, was known that uses the ribulose diphospate cycle for Cl-assimilation during growth on formate [84-88]. Very recently this pathway was also found in Micrococcus denitrificans grown on methanol [9] and the photosynthetic bacterium Rhodopseudomonas acidophila, which grows anaerobically in the light on methanol [89]. In contrast to the ribulose diphosphate cycle, formaldehyde instead of CO2 is fixed in the ribulose monophosphate cycle; therefore its occurrence has so far been described only in microorganisms growing on CFcompounds containing methyl groups. The key enzyme of this cycle, hexulose phosphate synthase, catalyzing the condensation of formaldehyde with ribulose-5-phosphate to form o-arabino-3-hexulose-6-phosphate, is particulate in methane-utilizing bacteria [90, 91 ], while in some bacteria which cannot grow on methane but can grow on methanol, this enzyme is soluble [92, 93]. The acceptor molecule, ribulose-5-phosphate, is regenerated either by a series of transaldolase- and transketolase-catalyzed reactions or by reactions analogous to those of the ribulose diphosphate cycle [73, 74]. As shown in Fig. 5, two alternative routes for cleavage of fructose-6-phosphate are found: one route involves fructose diphosphate aldolase and the other part of the Entner-Doudoroff pathway [73, 74].
\
~NADPH z
Fig. 5. Ribulose monophosphate cycle (Quayle cycle) of formaldehyde fixation
In the first step of the serine pathway the Cl-compund is incorporated at the oxidation level of formaldehyde by a condensation with glycine to form serine. In the facultative methylotrophic bacterium XX, two species of serine transhydroxy-methylase catalyzing this reaction were found and there is evidence that a similar situation may occur in Pseudomonas AM1 [94, 95]. One of these enzymes functions in the serine pathway and is induced during growth on methanol; the other isoenzyme provides for glycine synthesis form serine during growth on succinate etc. As shown in Fig. 6, serine is converted to malate, which is activated to malyl-CoA and then cleaved to glyoxylate and acetyl-CoA. Glyoxylate is converted to glycine and recycled into the pathway. The balance of the serine pathway is therefore the synthesis of acetyl-CoA from formaldehyde and C02. Acetyl-CoA is further oxidized to glyoxylate by some reactions of the glyoxylate cycle
90
H. Sahm
Hydr~x~Fyruvate fNADH2
te
ADP+Pi
I
~'CoA
~" N A D
ATP j
GIyce 'ate
t,t,al Ire
,--ATP
NAD-,,
NADH2~ Oxalocetate
~ADP
2-P-Gt!:erate
~ Phospho-/ enolpyruvate
Fig. 6. Serine pathway of formaldehyde fixation
and is then used for the synthesis of C3-compounds via the serine pathway [96, 97]. However, there is a group of bacteria examplified by Pseudomonas AM1 which grow on C1-compounds by using the serine pathway; but the key enzyme of the glyoxylate cycle isocitrate lyase is absent [98]. At present it is not known how acetyl-CoA is further metabolized in these bacteria. Among the three different assimilation pathways found in methanol-utilizing bacteria, the ribulose diphospahte cycle is energetically most unfavorable, while the ribulose monophosphate cycle is from the energetic point of view more favorable than the serine pathway (Table 6). It is evident that the different energy requirements of the different C~-assimilation pathways should affect the cell yield, at least when the energy
Table 6. Overall energy requirement for the synthesis of 1 molecule pyruvate from 3 molecules methanol by the different pathways of C~-assimilation. To enable a comparison between the different pathways, it was assumed that cells can produce: 2 molecules NADH~ by the oxidation of 1 molecule methanol to CO~, 3ATP by the oxidation of NADH~ to NAD and 2ATP by the oxidation of FADH 2 to FAD
Ribulose diphosphate cycle: 3CO~ + 4ATP + 5H20
~
pyruvate + 4ADP + 4P;
Ributose monophosphate cycle a) Fructose diphosphate aldolase 3HCHO + ADP + Pi + NAD ~ pyruvate + ATP + NADH 2 b) Part of Entner-Doudoroff pathway 3HCHO + NADP ~ pyruvate + NADPH 2
3. Serine pathway 2HCHO + CO s ~
pyruvate
Metabolism of Methanolby Yeasts
91
yield of the oxidation on the Crsubstrate is equal [99]. Therefore, it seems likely that a Crutilizing microorganism possessing the ribulose monophosphate cycle of Crfixation is more suitable for the production of single-cell protein than an organism using the serine pathway. The currently available data actually show that all bacteria selected for processes for producing SCP from methanol assimilate formaldehyde via the ribulose monophosphate pathway [6, 94, 100]. 3 a) Isotope Studies with Whole Yeast Ceils The first indications of the nature of the Crassimilation pathways in bacteria were obtained by Quayle and his colleagues from the excellent studies of short-term incubation experiments with 14C-species of Crcompounds into stable nonvolatile alcohol. soluble metabolites of growing cells. This method was used by Fujii et al. to investigate the assimilation pathway of methanol in the yeast Candida N-16 [101,102]. When methanol-grown yeast cells were incubated with ~4C-methanol, a number of radioactivelabelled metabolites were rapidly formed. The analysis of the early kinetics of the incorporation of ~4C-methanol showed that over 70% of the total radioactivity fixed for the first 10 s was found in phosphorylated derivatives of fructose (42%), glucose (18%), and mannose (10%). Since the percentage distribution of radioactivity in fructose phosphate decreased as time elapsed, fructose phosphate may be a primary stable intermediate in the assimilation of methanol by Candida N-16. When glucose grown cells were incubated with ~4C-methanol, the radioactivity fixed into metabolites was negligible. The percentage distribution of radioactivity incorporated from ~4C-formaldehyde into metabolites by Candida N-16 was very similar to that found from the incorporation of 14C-methanol. Over 50% of the total ~4C-formaldehyde fixed for the first 20 s was found in phosphorylated derivatives of fructose (22%), glucose (18%), and mannose (12%). This result suggests that fructose phosphate is also a primary stable compound in formaldehyde incorporation by methanol-grown cells. When glucose-grown cells of Candida N-16 were incubated with 14C-formaldehyde, negligible incorporation into metabolites was observed. The incorporation of ~4C-formate into metabolites of methanol-grown cells of Candida N-16 was much smaller than of ~4C-methanol or ~4C-formaldehyde. At the earliest period the highest percentage of the total radioactivity fixed from ~4C-formate was found in serine (38%). The percentage distribution decreased as time elapsed, suggesting that serine may be a primary compound in formate assimilation. Almost the same results were obtained with glucose-grown yeast cells. Since Canclida N-16 cannot synthesize cell compounds from CO2 in the absence of an energy source, the incorporation of ~4C-bicarbonate was examined in the presence of methanol. Most of the total radioactivity flexed for the first 15 s was found in aspartate (72%) and malate (18%); no radioactivity was found in sugar phosphates at this time. The percentage distribution of aspartate showed a negative slope against time. Therefore this metabolite might be a primary intermediate in bicarbonate assimilation. It was found that Candida N-16 contains high activity of pyruvate carboxylase. All these results with whole cells ofCandida N-16 together suggest that fructose phos-
92
H. Sahm
phate may be a primary intermediate in the assimilation of methanol in Candida N-16. The assimilatory pathway of methanol may be: Methanol ~ Formaldehyde ~ Hexulose phosphate -~ Cell constituents, identical or similar to the ribulose monophosphate cycle found in some groups of methane- and methanol-utilizing bacteria. 3 b) Enzyme Studies with Cell-Free Extracts When cell-free extracts from different methanol-assimilating yeasts were tested for their ability to incorporate 14C-formaldehyde into hexose phosphates, a very low specific activity was found compared to the results of bacteria [103, 104]. Recently it was demonstrated that ATP is required for the incorporation of formaldehyde into hexose phosphates by cell-free extracts of Candida N-16 and Candida boidinii dialyzed overnight [105,106]. Other nucleotides, such as GTP, ITP, or UTP, cannot be used instead of ATP. Furthermore, no hexulose phosphate synthase activity was observed when ribulose-1,5-diphosphate was used in the assay system instead of ribose-5-phosphate, which can be converted to ribulose-5-phosphate by the high-ribose-5-phosphate isomerase activity in the cell-free extract (Table 8) [ 107]. The role of ATP in this reaction is still unknown, however it may be that ATP acts as a positive effector on the activity of the first enzyme of the C~-assimilation pathway. Since the dissimilatory and assimilatory pathways of the methanol metabolism'branches off at the formadehyde, the key step of the assimilatory pathway may be regulated by energy charge. Studies with cell-free extracts of Candida boidinii showed that the hexulose phosphate synthase has a requirement for Mg2+ [108]. While the hexulose phosphate synthase from some C~-utilizing bacteria can be activated not only by Mg2+, but also by Mn 2+, and partially by Cd 2+, Co 2+, Zn 2+ [73, 92], these metal ions were inhibitory to the enzyme from Candida boidinii. As shown in Table 7 the Michaelis constants and the pH-optimum of the hexulose phosphate synthase from Candida boidinii were found to be similar to these of the enzyme from the obligate methylotrophic bacterium Methylo-
Table 7. Comparison of properties of hexulose phosphate synthase from different methane or methanol-utilizing microorganisms [91,921 Hexulose phosphate synthase from
Activation Inhibition Km-values: HCHO Ribulose-5-P pH optimum Molecular weight Localization
Methylococcus capsulatus [91 ]
Methylomonas M15 I92]
Candida boidinii
NM:: or Mn2+ or Cu2+
Mg2+ or Mn2+ Ni2+ or Cu2+
Mg2+ and ATP Mn2+ or Ni2+
0.49 mM 0.08 mM 7.0 310000 (6 subunits) membrane bound
1.1 mM 1.6 mM 7.5 -8.0 43000 (2 subunits) soluble
0.9 mM 1.5 mM 8.0 about 100000 soluble (?)
Metabolism of Methanol by Yeasts
93
monas M 15. So far, detailed information is not available for the hexulose phosphate synthase from methanol.grown yeasts since this enzyme is very unstable [107]. It will be of great interest to see whether or not the hexulose phosphate synthase is located in the microbodies of methanol-utilizing yeasts. The presence of the enzymes necessary for the operation of the complete ribulose monophosphate cycle were studies in cell-free extract of Candida boidinii [ 109]. As shown in Table 8, the four enzymes necessary for the rearrangement of two molecules of fructose 6-phosphate and one molecule of glyceraldehyde-3-phosphate to three molecules of ribulose-5-phosphate, transaldolase, transketolase, ribulose.5.phosphate-3epimerase, and phosphoriboisomerase are abundant in similar high activities in this methanol-grown yeast as in Cl-assimilating bacteria, which use the ribulose monophosphate Table 8. Specificactivities of ribulose monophosphate cycle enzymes in cell-freeextract of Candida boidinii
Enzyme
Specific activity (umol/min/mg of protein)
Hexulose phosphate synthase 3-Hexulose-phosphateisomerase Transaldolase Transketolase Ribulose-5-phosphate 3-epimerase Phosphoriboisomerase Phosphofructokinase Fructose-1,6-diphosphate aldolase Fructose-1,6-diphosphatase Sedoheptulose-1,7-diphosphatase 6-Phosphogluconatedehydrase/phospho-2-keto3-deoygluconatealdolase
0.2 not tested 0.74 0.25 2.5 1.3 0.06 0.94 0.00 0.00 0.00
cycle. Sedoheptulose-l,7-diphosphatase and fructose-1,6-diphosphatase could not be found in Candida boidinii. Since the specific enzymes of the Entner-Doudoroff pathway, 6-phosphogtuconate dehydrase and phospho-2-keto-3-deoxygluconate aldolase, were also not detectable in this yeast strain, fructose-1,6-diphosphate-aldolase may be responsible for the cleavage of fructose-1,6-phosphate to glyceraldehyde-3-phosphate and dihydroxyaceton-phosphate. Hydroxypyruvate reductase, a key enzyme of the serine pathway, and also ribulose diphosphate carboxylase, the key enzyme of the ribulose diphosphate cycle, were not detectable in CandMa boidinii [108]. Cells of Candida boidinii grown on glucose lack hexulose phosphate synthase activity the enzyme is induced during adaption from glucose to methanol medium (Fig. 7). The combined results of radioactive experiments with whole cells using ~4C-labeled Cl-compounds in conjunction with enzymic analysis of methanol grown cell-free extracts indicate that methanol is assimilated by yeasts in a sugar phosphate pathway similar in concept, if not in absolute molecular detail, to the ribulose monophosphate cycle in C~-utilizing bacteria. However, it is necessary to purify the hexulose phosphate synthase
94
H. Sahm
ol °
Spec. activity
Spec. activity
0.9~ 0.8
®
07 0,6 0.5 OZ'
®
0,1
® 10
2O
-,," t (h)
Fig. 7. Changes in specific activities of alcohol oxidase (1), catatase (2), formaldehyde dehydrogenase (3), formate dehydrogenase (4), and hexulose phosphate synthase (5) in Can. dida boidinii after transferring glucose-grown cells into basal medium containing 300 mM methanol
from yeasts to find out the substrate for this enzyme and also its reaction product. Tye and Willets reported that the ribulose monophosphate cycle of formaldehyde fixation may operate also in the fungus Trichoderma lignorum during growth on methanol [29]. However, the two fungi Gliocladium deliquescens and Paecilomyces varioti, both growing on methanol, formaldehyde, or formate, assimilate the Ct-compounds via the serine pathway as shown by short-term incubation experiments with t4C-labled Ctcompounds and by enzymic studies [28].
4.
Regulation of Enzymes of the Methanol Metabolism
4 a) Enzymes for Oxidation of Methanol In six different types of methanol-utilizing yeasts, Kato et al. found that specific activities of alcohol oxidase, catalase, formaldehyde dehydrogenase, and formate dehydrogenase were greatly increased in methanol-grown cells compared with glucose grown cells [48]. As shown in Fig. 7, the activities of these four enzymes and also of hexulose phosphate synthase increased significantly within 6 to 8 h, when glucose-grown cells of Candida boidinii were transferred to a medium containing 300 mM methanol as sole carbon and energy source [ 106]. While van Dijken has found in a methanol-limited chemostat culture of Hansenula polymorpha, that the level of alcohol oxidase increased from about 7% of the total soluble protein at a dilution rate of 0.16 h - t to 20% at a dilution rate of 0.03 h -1 [46], in batch culture of Candida boidinii a concentration of about 100 mM methanol at a cell density of 3 mg cell dry weight per ml is necessary for the maximal induction of these enzymes in glucose grown cells. Very similar results were obtained with some mutant strains of Candida boidinii which lack alcohol oxidase activity (Fig. 8 a).
Metabolism of Methanol by Yeasts
95 Spec activities of ®~ ,©,® 1.0
Spec.activity of catalose ® 2000
O9 08
1500
07 0.6
Fig. 8a-c. Changes in specific activities of alcohol oxidase (1), catalase (2), formaldehyde dehydrogenase (3), formate dehydrogenase (4) and hexulose phosphate synthase (5) in Candida boidinii after transferring glucose-grown cells into basal medium containing different concentrations of methanol (a), fon;,aldehyde (b) or sodium formate (c). Cells were harvested after an incubation time of 16 h Spec. (activities of O,C~,©,®
Spec. activity of cotolose
1.l0
0.9
0.5
1000
04
®
0,3 02 01 50
a
100 .----t CH30H (mM)
Spec activities of @.@.@,@ 10
Spec activity of catalase il2000
09 Q
O8
0.8 1500
@ 1500
0.7 O6
06 0.5
1000
@
O5
1000
O4
0.4
::tt/
__.
\o
500
02
/
01
o
50 1
2
3
/.
5
6 HCHO (mM)
500
"®
"~@-/* HCO0 NG 100 206 (mM}
c
Although Candida boidinii is not able to grow on formaldehyde or formate as sole source of carbon and energy, these two Cl-compounds still induce the enzymes of the dissimilatory pathway of methanol in this yeast. When glucose-grown cells were transferred to a medium containing 2 to 3 mM formaldehyde alcohol oxidase, catalase, formaldehyde dehydrogenase, and formate dehydrogenase were induced, besides catatase, to half of their maximal specific activities. Since formaldehyde is toxic in concentrations above 6 mM for this yeast strain, the addition of 6 mM formaldehyde to the culture medium gives only a quarter of the maximal activities of these enzymes. It is interesting that hexulose phosphate synthase is not induced under these conditions, this may explain why Candida boidinii cannot grow on formaldehyde as sole source of
96
H. Sahm
carbon and energy [18]. The enzymes of the dissimilatory pathway of methanol were maximally induced by 10 to 20 mM sodium formate in the medium. Under these conditions hexulose phosphate synthase activity increased only a little. Induction is not the only form of control over the synthesis of these enzymes of the methanol metabolism; the presence of glucose, ethanol, fructose, or acetate in the culture medium has a striking effect on the synthesis of these enzymes in Candida boidinii. If this yeast is inoculated into medium containing glucose and methanol, the former so severely represses synthesis of alcohol oxidase, formaldehyde dehydrogenase, formate dehydrogenase, and hexulose phosphate synthase that no methanol is utilized until glucose is exhausted; a biphasic or diauxic growth curve is thereby obtained [61 ]. Sodium pyruvate, sodium succinate, or glycerol showed no catabolite repression of these enzymes of the methanol metabolism. It is uncertain whether the structural genes for alcohol oxidase, formaldehyde dehydrogenase, and formate dehydrogenase are clustered together to form an operon. Since the dissimilatory and assimilatory ribulose monophosphate cycles branch off at fructose-6-phosphate (Fig. 2), the regulation of the three enzymes involved in the oxidation of fructose-6-phosphate to ribulose-5-phosphate and CO2 was investigated in Candida boidinii. As shown in Table 9 the specific activites of phosphoglucoisomerase,
Table 9. Specific Activitiesof phosphoglucoisomerase,glucose-6-phosphatedehydrogenaseand 6-phosphogluconate dehydrogenasein cell-greeextracts of Candida boidinii grown on different carbon sources [ 109l Specific activities (vmol/min/mg protein) Carbon source
Phosphoglucoisomerase
Glucose-6-phospate- Phosphogluconate dehydrogenase dehydrogenase
Ethanol Glucose Methanol
0.64 1.20 0.38
0.25 0.60 0.25
0.10 0, 25 0.12
glucose-6-phosphate dehydrogenase and 6-phosphogluconate dehydrogenase are only half in methanol-grown cells in comparison with cells grown on glucose medium. Since in Candida boidinii only a very low specific activity of a-ketoglutarate dehydrogenase was found (0.002/lmol/min/Ing protein), glucose may be metabolized primarily via the oxidative pentose phosphate cycle. Furthermore, glucose-6-phosphate dehydrogenase and 6-phosphogluconate dehydrogenase of this yeast are inhibited in a non-competitive manner by 1 mM NADH2 to 70% and by 1 mM NADPH2 to 100%. This result suggests that the cyclic oxidation of formaldehyde may be regulated by the NAD/NADH2 and NADP/NADPH2 ratio in cells. 4 b) Regulation of the Flavin Synthesis Since the inducible FAD-dependent alcohol oxidase forms about 10% of the total soluble protein of Candida boidinii [45], the regulation of the FAD-biosynthesis in this
Metabolism of Methanol by Yeasts
97
yeast strain was studied. The specific activities of riboflavin synthase, riboflavin kinase, and FMN adenylyltransferase are considerably higher in ceils grown on methanol than in cells grown on ethanol, glucose, glycerol, or mannitol (Table 10). When Candida boidinii was cultured on glucose and then transferred into a methanol medium, FMNadenylyltransferase activity increased about 6 to 8-fold in comparison with that from glucose-grown cells (Fig. 9) [1 t0]. The full activity is reached after an incubation time of about 3 h, whereas the full activity of alcohol oxidase is reached after 10 h.
Table 10. Specific activities of riboflavin synthetase, riboflavin kinase, FMN adenyltransferase, and alcohol oxidase in cell-free extracts of Candida boidinii grown on different carbon and energy sources I 1101 Specific activities (~mol/min/mg protein) Carbon source
Riboflavin synthase
Riboflavin kinase
FMN adenylyltransferase
Alcohol oxidase
Ethanol Glucose Glycerin Mannitol Methanol
0.13 0.12 0.10 0.18 0.26
0.21 0.16 O. 19 0,17 0.32
0.062 0.046 0.052 0.056 0.382
0.00 0.00 0.00 0.00 0.17
018
044
>,
040
02
_,2 0 3 6
i u
~ 032
-E_-
-- 0 2 8 o 024. o
g
020
01 ~o "5
~ , 016-
Fig. 9. Induction of alcohol oxidase (1) and FMN adenyltransferase (2) in Candida boidinii after transferring glucose-grown cells into methanol medium
lll0l
<
012o z O08-
X u. Q04-
IL, I
2
3 4 time ( h )
5
6
?
In order to determine if there is a relationship between the synthesis of FAD-dependent alcohol oxidase and the regulation of FMN adenylyltransferase, some mutants of Candida boidinii were used which completely lack alcohol oxidase activity [ 110]. Two different types of mutants were used. One type of mutant contained no cross-reacting material of alcohol oxidase protein. When these mutants grown on glucose were transferred to a methanol medium, FMN adenylyltransferase was not induced, while the
98
H. Sahm
activities of catalase, formaldehyde dehydrogenase, and formate dehydrogenase increased to the same extent as in wild-type cells. However, when the second type of mutant containing inactive alcohol oxidase protein as indicated by immunoreaction were transferred into a methanol medium, FMN adenyly]transferase was also induced. These results indicate that in Candida boidinii the induction of FMN adenylyltransferase is evidently influenced by the synthesis of alcohol oxidase protein. However, further studies are necessary to determine the mechanism of the regulation of flavinsynthesis in this methanol-utilizing yeast.
5. A p p l i c a t i o n s
5 a) Single-Cell-Protein (SCP) Production The increasing demand for protein for human food and animal feed makes it necessary to develop new processes for producing low-cost protein with high nutritive value. Single-cell microorganisms, yeast in particular, have long been known to be sources of good quality protein. Much work has been done over the past decade toward the development of large-scale processes for the production of SCP by microorganisms capable of utilizing liquid hydrocarbons. However, these substrates have several disadvantages: low water solubility, high oxygen requirement, and high heat generation. Furthermore, it is necessary to refine the biomass grown on hydrocarbons by solvent treatment in the final step. In the last few years the possibility of producing SCP from methanol by microorganisms has been explored in a number of laboratories. Levine and Cooney isolated and studied a thermotolerant methanol-utilizing yeast identified as a strain of Hansenula polymorpha [24]. This yeast grows optimally with a maximum specific growth rate of 0.22 h-~ at pH 4.5 to 5.5 and at temperatures of 37 ° to 42 °C. When the methanol concentration in the culture medium was greater than 1% (v/v) growth rate decreased. The maximal growth yield in a methanol-limited chemostat culture has been found to be 0.36 g of dry cell weight per g of methanol. Reuss et al. [111] and Pilat and Prokop [112] observed with Candida boidinii a gradual decrease in the biomass yield and an increase in CO2production with increasing methanol concentration in the culture medium. It seems that the lower cell yields at higher concentrations of methanol are due to the uncoupling of methanol oxidation and production of ATP. The higher oxidation rate at higher methanol concentrations seems to be beneficial from the viewpoint of detoxification of the environment by the microorganism. Analysis of extended culture data of Candida boidinii indicates a Ks-constant (this value indicates the substrate concentration at which the specific growth rate is one-half of the maximum) of 550 mg methanol/liter and a maximal specific growth rate of 0.20 h -1 [113]. Because of this relatively high value of Ks, methanol concentrations greater than 0.6% (v/v) are necessary to reach maximal growth rate. At such levels, however, methanol inhibits growth and the cell yield does not attain the maximal value. Consequently, a balance must be struck in order to reach an acceptable compromise between optimal cell yield and optimal growth rate. Protein contents of methanol-utilizing yeasts have been reported to range from 35 to
Metabolism of Methanol by Yeasts
99
50% and total nucleic acid content varies from 5.0 to 7.0% [114]. The amino acid prof'fle of these yeast proteins indicate that the contents of essential amino acids are above the FAO standard, with the exception of methionine. This methionine deficiency can readily be overcome by the addition of synthetic methionine to the harvested cell material. Safety is one of the most important factors in determining the acceptability of SCP; therefore many different toxicity tests were carried out. So far, no evidence has been obtained indicating that these methanol-utilizing yeasts are more toxic than Baker's yeast and therefore can generally be regarded as nontoxic. It is suggested that protein from these yeasts will prove useful in the preparation of food for human consumption [115]. Yeasts have some advantages in comparison with bacteria: a) Recovery by either centrifugation or filtration is easy and cheap because of the large cells. b) Nucleic acid content is tow. c) Yeast, as SCP, is psychologically more palatable for human consumption. Yeasts, however, are not without their disadvantages as protein sources: a) They have a lower protein content than bacteria. b) Their protein has a low methionine level. c) They have a lower specific growth rate than some methanol-utilizing bacteria which grow on methanol with a maximal specific growth rate of 0.5 h - 1 [4, 116, 117].
5 b) Production of Metabolites While during the last few years production of single-cell protein using methanol as the sole source of carbon and energy has been studied with different microorganisms, only a few reports have been published concerning production of useful raetabolites from methanol. In a patent issued to Oknmura et al., examples are given of processes for the production by methanol-utilizing yeasts of amino acids: L-alanine (0.3 g/l); L-glutamic acid (1.5 g/I); L-lysine (0.5 g/l); L-threonine (0.5 g/I); L-valine (0.5 g/l); of some organic acids: citric acid (0.7 g/l); c~-ketoglutaric acid (0.5 g/l) and of hypoxanthine (0.3 g/1) [23]. Candida boidinii grown on methanol excreted about 1 g/1 citric acid if fluoroacetate was added to the culture medium [118]. Oki et Kitai isolated two obligate methanol-utilizing bacteria, named Methylomonas methylovora, which accumulated 6.8 g/1 L-glutamic acid after 48 h of fermentation with a yield of about 11% based on the methanol consumed [119]. Since it is known that methanol is assimilated by some groups of microorganisms via the serine pathway, it seemed attractive to investigate the production of L-serine by these organisms. Keune et al. isolated a facultative methylotrophic bacterium, Pseudomonas 3ab, which utilizes the serine pathway of formaldehyde fixation [ 120]. This organism produced only small amounts of L-serine (40 rag/l) when it was cultivated on mineral salt medium with methanol, pyruvate, lactate, or malate as carbon and energy sources. However, when glycine (20 g/l) and methanol (8 g/1) were added to methanolgrown cells at the end of the exponential growth phase and the pH was adjusted to 8.5, a remarkable increase in L-serine production was observed (4.7 g/1). Since there is some
100
H. Sahm
evidence that the carbon skeleton of L-serine is derived chiefly from glycine and methanol, L-serine is probably synthesized via the following steps: Methanol -~ Formaldehyde
Serinehydroxymethyltransferase ~r. , L-Serine Glyclne
Considering the involvement of vitamin B12 in methyl transfer reactions of many organisms, Fukui and coworkers and Kamikubo and coworkers investigated the production of vitamin B12 by methanol-assimilating bacteria [8, 121 ]. The vitamin B12 productivity of a pink-pigmented bacterial strain growing on methanol was dependent on Co 2+ and attained values of 2.6 mg/1 of medium [121 ]. Nishio et al. isolated a methanol-utilizing bacterium identified as Klebsiella sp., which produced vitamin B12 in detectable amounts only when methanol was used as the sole carbon source [8]. The maximal vitamin B12production by this strain was 140/2g/1 of medium. Since some enzymes o f flavin synthesis are induced in Candida boidinii grown on methanol, it might be possible to produce riboflavin or its phosphorylated and adenylated compounds, riboflavin-5-phosphate and flavinadenine dinucleotide, by methanol-utilizing yeasts. Recently a facultative methylotrophic Streptomyces strain was isolated, which produced a new antibiotic cryomycin at low growth temperature and a water-soluble basic antibiotic at moderate temperatures from hydrocarbons, methylacetate, sodium acetate or methanol [ 122]. Finally, it should be pointed out that the bacterial utilization of methanol is often accompanied by extensive production of water-soluble polysaccharides which are of commercial interest [ 123, 124]. Although the amount of all these metabolites produced from methanol is as yet far from acceptable industrial levels, these data nevertheless indicate the potential o f methanol, not only as a carbon source for SCP production but also as a starting point for many other fermentation processes.
Acknowledgement 1am grateful to Professor J. R. Quayle for his advice and correction of the English translation.
6. R e f e r e n c e s 1. 2. 3. 4. 5. 6. 7. 8. 9.
Whittenbury, R., Phillips, K. C., Wilkinson, J. F.: J. Gen. Microbiol. 61,205 (1970). Foster, J. W., Davis, R. H.: J. Bacteriol. 91, 1924 (1966). Dworkin, M., Foster, J. W.: J. Bacteriol. 72, 646 (1956). Sahm, H., Wagner, F.: Europ. J. Appl. Microbiol. 1,147 (1975). Akiba, T., Ueyama, H., Seki, M., Fukimbara, T.: J. Ferment. Technol. 48, 323 (1970). Myers,P. A., Zatman, L. J.: Biochem. J. 121, p. 10 (1971). Hirsch, P., Conti, S. F.: Arch. Mikrobiol. 48, 339 (1964). Nishio, N., Yano, T., Kamikubo, T.: Agric. Biol. Chem. 39, 21 (1975). Cox, R. B., Quayle, J. R.: Biochem. J. 150, 569 (1975).
Metabolism of Methanol by Yeasts 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61.
101
Den Dooren de Jong, L. E.: Zbl. Bakt. Parasit. (Abt. II) 71, 193 (1927). Peel, D., Quayle, J. R.: Biochem. J. 81,465 (1961). Anthony, C., Zatman, L. J.: Biochem. J. 92, 609 (1964). Quayle, J. R., Pfennig, N.: Arch. Microbiol. 102, 193 (1975). Douthit, H. A., Pfennig, N.: Arch. Microbiol. 107, 233 (1976). Kato, N., Tsuji, K., Tani, J., Ogata, K.: J. Ferment. Technol. 52, 917 (1974). Bassalik, K.: Jb. Wiss. Bot. 53, 255 (1913). Ogata, K., Nishikawa, tL, Ohsugi, M.: Agric. Biol. Chem. 33, 1519 (1969). Sahm, H., Wagner, F.: Arch. Mikrobiol. 84, 29 (1972). van Dijken, J. P., Harder, W.: J. Gen. Microbiol. 84, 409 (1974). Hazen, W., de Bruyn, J. C., Bos, P.: Arch. Mikrobiol. 87, 185 (1972). Oki, T., Kuono, K., Kitai, A., Ozaki, A.: J. Gen. Appl. Microbiol. 18, 295 (1972). Fujii, T., Tonomura, K.: Agric. Biol. Chem. 36, 2297 (1972). Okomura, S., Yamanoi, A., Tsugawa, R., Nakase, T.: Patent specification 1210770, London/ England (1970). Levine, D. W., Cooney, C. L.: Appl. Microbiol. 26, 982 (1973). Henninger, W., Windisch, S.: Arch. Microbiol. 105, 47 (1975). Asthana, H., Humphrey, A. E., Moritz, V.: Biotechnol. Bioeng. 13, 923 (1971). Yokote, J., Sugimoto, M., Abe, S.: J. Ferment. Technol. 52, 201 (1974). Sakaguchi, K., Kurane, R., Murata, M.: Agr. Biol. Chem. 39, 1695 (1975). Tye, R. J., Willetts, A. J.: J. Gen. Microbiol. 77, i(1973). Ribbons, D. W., Harrison, J. E., Wadzinski, A. M.: Ann. Rev. Microbiol. 24, 135 (1970). Quayle, J. R.: Adv. Microbial Physiol. 7, 1 t9 (1972). Anthony, C.: Sci. Progr. Oxford 62, 167 (1975). Anthony, C., Zatman, L. J.: Biochem. J. 92, 614 (1964). Johnson, P. A., Quayle, J. R.: Biochem. J. 93, 281 (1964). Patel, R. N., Bose, H. R., Mandy, W. J., Hoare, D. S.: J. Bacteriol. 110, 570 (1972). Mehta, R. J.: Antonie van Leeuwenhoek 39, 303 (1973). Sperl, G. T., Forrest, H. S., Gibson, D. T.: J. Bacteriol. 118, 541 (1974). Harder, W., Attwood, M. M.: Antonie van Leeuwenhoek 41,421 (1975). Anthony, C., Zatman, L. J.: Biochem. J. 104, 960 (1967). Anthony, C.: Biochem. J. 146, 289 (1975). Tonge, G. M., Harrison, D. E. F., Knowles, C. J., Higgins, I. J.: FEBS Letters 58, 239 (1975). Wadzinski, A. M., Ribbons, D. W.: J. Bacteriol. 122, 1364 (1975). Tani, Y., Miya, T., Nishikawa, H., Ogata, K.: Agric. Biol. Chem. 36, 68 (1972). Tani, Y., Miya, T., Ogata, K.: Agr. Biol. Chem. 36, 76 (1972). Sahm, H., Wagner, F.: Europ. J. Biochem. 36, 250 (1973). van Dijken, J. P.: Ph.D. thesis, University Groningen, Holland (1976). Kato, N., Omori, Y., Tani, A., Ogata, K.: Europ. J. Biochem. 64, 341 (1976). Kato, N., Tani, Y., Ogata, K.: Agric. Biol. Chem. 38, 675 (1974). Mehta, R. J.: J. Bacteriol. 124, 1165 (1975). Sahm, H.: Unpublished results. Roggenkamp, R., Sahm, H., Wagner, F.: FEBS Letters 41,283 (1974). Roggenkamp, R.: Ph.D. thesis Techn. Univ. Braunschweig, West Germany (1976). Keitin, D., Hartree, E. F.: Biochem. J. 39, 293 (1945). van Dijken, J. P., Otto, R., Harder, W.: Arch. Microbiol. 106, 221 (1975). van Dijken, J. P., Veenhuis, M., Vermeulen, C. A., Harder, W.: Arch. Microbiol. 105, 261 (1975). Roggenkamp, R., Sahm, H., Hinkelmann, W., Wagner, F.: Europ. J. Biochem. 59, 231 (1975). Fukui, S., Kawamoto, S., Yasuhara, S., Tanaka, A.: Europ. J. Biochem. 59, 561 (1975). Harrington, A. A., Kallio, R. E.: Can. J. Microbiol. 6, 1 (1960). Colby, J., Zatman, L. J.: Biochem. J. 132, 101 (1973). Patel, R. N., Hoare, D. S.: J. Bacteriol. 107, 187 (1971). Sahm, H., Wagner, F.: Arch. Mikrobiol. 90, 263 (1973).
102 62. 63. 64 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112.
H. Sahm Kato, N., Tamaoki, T., Tani, Y., Ogata, K.: Agr. Biol. Chem. 36, 2411 (1972). Schiitte, H., Flot~dorf, J., S',dam, H., Kula, R. M.: Euro. J. Biochem. 62, 151 (1976). Budenstein, A. Z.: Ph.D. thesis, Yale University, USA (1955). Strittmatter, P., Ball, E. G.: J. Biol. Chem. 213,445 (1955). Rose, Z. B., Racker, E.: J. Biol. Chem. 237, 3279 (1962). Uotita, L., Koivusalo, M.: J. Biol. Chem. 249, 7653 (t974). Sahm, H.: Arch. Microbiol. 105,179 (1975). Fujii, T., Tonomura, K.: Agric. Biol. Chem. 39, 2325 (1975). Kaneda, T., Roxburgh, J. M.: Can. J. Microbiol. 5, 187 (1959). Kato, N., Kano, M., Tani, Y., Ogata, K.: Agric. Biol. Chem. 38, 111 (1974). Davey, J. F., Whittenbury, R., Wilkinson, J. F.: Arch. Mikrobiol. 87, 359 (1972). StrUm, T., Ferenci, T., Quayte, J. R.: Biochem. J. 144, 465 (1974). Colby, J., Zatman, L. J.: Biochem. J. 148, 513 (1975). De Duve, C., Baudhuin, P.: Physiol. Rev. 46,323 (1966). TolberL N. E.: Ann. Rev. Plant Physiol. 22, 45 (1971). Avers, C. J., Federman, M.: J. Cell Biol. 37,555 (1968). Perlman, P. S., Mahler, H. R.: Arch. Biochem. Biophys. 136, 245 (1970). Osumi, M., Miwa, N., Teranishi, Y., Tanaka, A., Fukui, S.: Arch. Microbiol. 99, 181 (1974). Sahm, H., Roggenkamp, R., Wagner, F., Hinkelmann, W.: J. Gen. Microbiol. 88, 218 (1975). Tanaka, A., Yasuhara, S., Kawamoto, S., Fukui, S., Osumi, M.: J. Bacteriol. 126, 919 (1976). van Dijken, J. P., Veenhuis, M., Kreger-van Rij, N. J.-W., Harder, W.: Arch. Microbiol. 102, 41 (1975). Fukui, S., Tanaka, A., Kawamoto, S., Yasuhara, S., Teranishi, Y., Osumi, M.: J. Bacteriol 123, 317 (1975). Quayle, J. R.: Ann. Rev. Microbiol. 15,119 (1961). Quayle, J. R., Keech, D. B.: Biochem. J. 72, 623 (1959). Quayle, J. R., Keech, D. B.: Biochem. J. 72, 631 (1959). Quayle, J. R., Keech, D. B.: Biochem. J. 75,515 (1960). Blackmore, M. A., Quayle, J. R.: Biochem. J. 107, 705 (1968). Sahm, H., Cox, R. B., Quayle, J. R.: J. Gen. Microbiol. 94, 313 (1976). Lawrence, A. J., Quayle, J. R.: J. Gen. Microbiol. 63,371 (1970). Ferenci, T., Str6m, T., Quayle, J. R.: Biochem. J. 144, 477 (1974). Sahm, H., Schiitte, H., Kula, M. R.: Europ. J. Biochem. 66, 591 (1976). Stieglitz, B., Mateles, R. I.: J. Bacteriot. 114, 390 (1973). Harder, W., Quayle, J. R.: Biochem. J. 121,753 (1971). O'Connor, M. L., Hanson, R. S.: J. Bacteriol. 124, 985 (1975). Hersh, L. B., Bellion, E.: Biochem. Biophys. Res. Commun. 48, 712 (1972). Harder, W., Attwood, M. M., Quayle, J. R.: J. Gen. Microbiol. 78, 155 (1973). Quayte, J. R.: Proc. Internat. Syrup. Microbial growth on Ct-compounds, Tokyo, 59 (1975). van Dijken, J. P., Harder, W.: Biotechn. Bioeng. 17, 15 (1975). MacLennan, D. G., Gow, J. S., Stringer, D. A.: Process Biochem. 8, 22 (1973). Fujii, T., Tonomura, K.: Agric. Biol. Chem. 37,447 (1973). Fujii, T., Asada, Y., Tonomura, K.: Agric. Biol. Chem. 38, 1121 (1974). Diel, F., Held, W., Schlanderer, G., Dellweg, H.: FEBS Letters 38, 274 (1974). Trozenko, J. A., Bykonskaje, S. W., Kirikonze, N. N.: Biochimica 211, 1230 (1973). Fujii, T., Tonomura, K.: Agric. Biol. Chem. 38, 1763 (1974). Sahm, H., Dix, B., Eggeling, L., Roggenkamp, R:: Proc. V. lnternat. Fermentation Syrup. Berlin (1976). Sahm, H.: Habilitationsarbeit, Techn. Univ. Braunschweig, West Germany (1975). Sahm, H., Wagner, F.: Arch. Microbiol. 97, 163 (1974). Dix, B.: Diplomarbeit, Techn. Univ. Braunschweig, West Germany (1975). Eggeling, L.: Ph.D. thesis Techn. Univ. Braunschweig, West Germany (1976). Reuss, M., Sahm, H., Wagner, F.: Chem.-Ing.-Tech. 46,669 (1974). Pilat, P., Prokop, A.: Biotechnol. Bioeng. 17, 1717 (1975).
Metabolism of Methanol by Yeasts 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124.
103
Reuss, M., Gnieser, J., Reng, H. G., Wagner, F.: Europ. J. Appl. Microbiol. 1,295 (1975). Cooney, C. L., Levine, D. W.: In Single Cell Protein, II, p. 402, MIT Press (1975). Cardini, G., di Fiore, L., Zotti, A.: Proc. V. Internat. Fermentation Symp. Berlin (1976). Amano, Y., Sawada, H., Takada, N., Terni, G.: J. Ferment. Technol. 53, 315 (1975). Dostalek, M., Molin, N.: In Single Cell Protein, II, p. 385, MIT Press (1975). Sisting, P.: Ph.D. thesis, Tech. Univ. Braunschweig, West Germany (1977). Oki, T., Kitai, A.: Process Biochem. 11, 31 (1974). Keune, H., Sahm, H., Wagner, F.: Europ. J. Appl. Microbiol. 2, 175 (1976). Toraya, Yongsmith, B., Tanaka, A., Fukui, S.: Appl. Microbiol. 30, 477 (1975). Yoshida, N., Tani, Y, Ogata, K.: J. Antibiotics 25, 546 (1972). Tannahitl, A., Finn, R. K.: ACS Meeting, II1,, USA (1970). Chalfan, Y., Mateles, R. I.: AppL MicrobioL 23, 135 (1972).
Control of Antibiotic Synthesis by Phosphate Juan Francisco Martin Instituto de Microbiolog/a Bioquimica, CSIC, and Departamento de Microbiologia, Facultad de Ciencias, Universidad de Salamanca, Salamanca, Spain
Contents 1. Introduction. Historical Development of Phosphate-Limited Antibiotic Fermentations - • 105 2. Why Phosphate Regulation? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 3. Postulated Mechanisms of the Phosphate Effect . . . . . . . . . . . . . . . . . . . . . . . . . 108 3.1 Phosphate Favors Primary Metabolism; Shift Down in Primary Metabolism Derepresses Secondary Metabolism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 3.2 Phosphate Shifts Carbohydrate Catabolic Pathways . . . . . . . . . . . . . . . . . . . . 110 3.3 Phosphate Limits Synthesis of the Inducer of the Antibiotic Pathway . . . . . . . . . 111 3.4 Phosphate Inhibits the Formation of Antibiotic Precursors . . . . . . . . . . . . . . . 112 3.5 Phosphate lnhibits or Represses Phosphatases Necessary for Antibiotic Biosynthesis . 114 3.6 Other Possible Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116 4. Studies on the Phosphate Effect Using Resting Ceils . . . . . . . . . . . . . . . . . . . . . . 116 5. Possible Intracellular Effectors Which Might Regulate Gene Expression Related to Antibiotic Synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 5.1 Cyclic Nucleotides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 5.2 ATP and Adenylate Energy Charge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 5.3 Polyphosphates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 5.4 Highly Phosphorylated Nucleotides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 6. Alternative Levels of Control of Antibiotic Synthesis . . . . . . . . . . . . . . . . . . . . . . 122 7. Summary and Future Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
1. I n t r o d u c t i o n . H i s t o r i c a l D e v e l o p m e n t o f P h o s p h a t e - L i m i t e d Antibiotic
Fermentations The influence of phosphate in antibiotics production has been known for many years. Following the discovery of new antibiotics, nutritional studies designed to improve production revealed that high concentrations of inorganic phosphate inhibited the formation o f many antibiotics while stimulating vegetative growth. For example, studies on the production o f chlortetracycline revealed as early as 1954 that phosphate has a strong inhibitory effect [ 15, 17, 58, 110, 1 2 3 - 1 2 5 ]. Similarly, the effect o f phosphate in the biosynthesis of streptomycin and oxytetracycline was recognized many years ago [4, 19, 107, 155]. Phosphate inhibition of the biosynthesis of additional aminoglycoside antibiotics was soon recognized, e.g., neomycin [37, 83, 84], butirosin [29], and kana. mycin [10]. Inhibition was also seen with peptide antibiotics, such as bacitracin [50], polymyxin [ 128 ], gramicidin S [ 142 ], monamycin [ 51 ], viomycin [ 106 ], actinomycin [66], and cephamycin [3], and with other antibiotics such as novobiocin [57], vanco-
106
J.F. Martin
mycin [96], ristomycin [40, 139], fosfomycin [120], pyocyanine [43], and the adeninecontaining antibiotics A-9145 [ 16]. Also extremely sensitive to phosphate is the biosynthesis of antibiotics derived by the polyketide pathway such as cycloheximide [2] and the potyene macrolide antibiotics [86]. Early observations of the effect of phosphate on polyene biosynthesis were made by Popova et al. [109] in the case of nystatin and Donovick and Brown [37] with amphotericin B. Other more recent examples include ayfactin [1 ], candicidin [79, 88], candidin [Martin, unpublished], levorin [ 12], and mycoheptin [ 133]. Many antibiotics are only produced at concentrations of inorganic phosphate suboptimal for growth. Phosphate in the range of 0.3 to 500 mM permits excellent cell growth, whereas 10 mM phosphate often suppresses biosynthesis of antibiotics (Table 1). Permissive concentrations differ slightly between the different antibiotic processes. As shown in Table 1, the biosynthesis of polyene antibiotics seems especially sensitive to phosphate. Phosphate inhibition of antibiotic synthesis requires that industrial production be carried out under phosphate limitation. This is done routinely in the antibiotics industry. However, phosphate limitation can lead to low mycelial concentrations; which is undesirable since the final accumulation of antibiotic is the result of both the number of producing cells and the specific production per cell. The antibiotics industry usually avoids phosphate inhibition by empirically selecting media of low phosphate concentration or by addition of phosphate-complexing agents to the production broth. This latter practice may interfere with other aspects of the metabolism of the producer cell. Taking into account the general nature of the phosphate effect and its great industrial significance, the paucity of available information is somewhat surprising. Only in recent years have some publications appeared which deal with the possible mechanism(s) of the phosphate effect. Regulation by phosphate does not only apply to the formation of antibiotics. The synthesis of many secondary metabolites which lack antibiotic activity is also sensitive to phosphate, e.g., ergot alkaloids [ 119] and bikaverin [ 18].
2. Why Phosphate Regulation? From the viewpoint of the microbial cell, both the primary and the secondary metabolism of the cell must be integrated into an overall process which confers a selective advantage for the producing organism. Regulation is an essential condition of life for rapid adaptation to changing environments. Production of antibiotics may seem at first glance to be a wasteful process for the producing cell if secondary metabolites have no role in the producing cell [ 152]. However, antibiotic production would not have persisted in the evolutionary process if antibiotics play no role in the survival of the producing strain. Although antibiotic production seems to be dispensable under pure culture conditions in the laboratory (as judged from the viability of nonproducing strains), we have practically no information on the importance of antibiotics for survival in nature. Phosphate regulation of antibiotic production can be visualized as a means by which the primary and secondary metabolism respond in a manner appropriate to the needs
107
Control of Antibiotic Synthesis by Phosphate Table 1. Antibiotic processes which are inhibited by inorganic phosphate Antibiotic
Producer strain
Range of inorganic orthophosphate permitting antibiotic production (mM)
References
Streptomycin Novobiocin Monamycin Neomycin Chlortetracycline Oxytetracycline Vancomycin Viomycin Bacitraein Ristomycin Actinomycin Polymyxin Tetracycline Kanamycin Cycloheximide A-9145 Pyocyanine Prodigiosin Butirosin Coryneicins Oleandomycin Gramicidin S Gramicidin S Cephamycin
Streptomyces griseus Streptomyces niveus Streptomyces ]amaicensis Streptomyces fradiae Streptomyces aureofaciens Streptomyces rimosus Streptomyces orientalis Streptomyces sp. Bacillus licheniformis Proactynomyces fructiferi Streptomyces antibioticus Bacillus polymyxa Streptomyces aureofaciens Streptomyces kanamyceticus Streptomyces griseus Streptomyces griseolus Pseudornonas aeruginosa Serratia marcescens Bacillus circulans Corynebacterium sp. Streptomycesantibioticus Bacillus brevis Bacillus brevis Streptomyces clavuligerus
1.5-15 9-40 0.2-0.4
[1071 [571 [511 [371 I I I 0, 581 11551 [96 ] [ 1061 [50] 1401 1661 1281 [84a1 [10l [21 [161 [63 a, 431 [1531 [291 [131, 1351 [136, 1371 I69, 1401 11421 [31
Polyene antibiotics Amphotericin B Ayfactin Candicidin Candidin
Streptomyces nodusus Streptomyces aureofaciens Streptomyces griseus Streptornyces viridoflavus
1.5- 2.2 1- 17 0,5-5 0.5-5
Levorin Mycoheptin Nystatin Ristomycin
Streptomyces levoris Streptoverticillium mycoheptinicum Streptomyces noursei Proactinomycesfructiferi var. ristomycin <
0.3-4 3.5 a 1.6-2.2 2.8
1--5 2-10 1- 7 1- 8 0.1 - 1 0.2-5 1.4-17 0.14-0.2 2.2-5.7 0.05- 0.5 0.28- 2.24 O.1a 0.05-0.2 < 5.6 2.24a 0.5a 10-60 50a 25
[37, t 331 [1l [79, 881 Martin (unpublished) I121 I 1331 I37, 1331 [138, 1391
a Optimal concentration of phosphate for antibiotic production
of the organism under a variety of conditions. Present evidence suggests that many regulatory mechanisms, e.g., induction, carbon and nitrogen catabolite repression, and end-product regulation [32], involve a direct interaction between an allosteric enzyme and a metabolite of low molecular weight. Phosphate regulation of secondary metabolism may also qualify in this respect (see later).
108
J.F. Marffn
3. Postulated Mechanisms o f the Phosphate E f f e c t Several mechanisms have been proposed to explain the effect of phosphate on the production of secondary metabolites (idiolites). They are as follows. 3.1 Phosphate Favors Primary Metabolism; Shift Down in Primary Metabolism Derepresses Secondary Metabolism To understand the inhibitory effect of phosphate in secondary metabolism, it is first necessary to consider the role of phosphate as an effector of a large number of enzymic reactions of primary metabolism. Phosphate controls DNA, RNA, and protein synthesis, carbohydrate metabolism, cellular respiration, and ATP levels in antibiotic producers. Perlman and Wagman [ 136] showed that levels of phosphate inhibitory to streptomycin formation increased the rate of glucose utilization. Di Marco [36] reported that excess inorganic phosphate stimulated glucose utilization and increased the mycelial concentration of DNA, RNA, and protein in Streptomyces aureofaciens during the biosynthesis of chlortetracycline. Hostalek [58] demonstrated that by increasing the inorganic phosphate content of the medium, the rate of respiration was increased at the same time that the synthesis of chlortetracycline and pentose phosphate pathway activity were inhibited. Pass and Raczynska-Bojanowska [ 106] reported that phosphate controls both mycelial growth and the biosynthesis of viomycin. Phosphate addition results in a large increase of the stored carbohydrate in the cell [96]. Upon addition of 5 mM phosphate to cells already producing candicidin, the oxygen uptake of the culture increased sharply and antibiotic synthesis was soon inhibited [79, 91]. These changes were followed by an increased rate of glucose utilization and an increase in mycelial dry weight. This shift from secondary (antibiotic-producing) to primary (growth-related) metabolism can be initiated by phosphate addition at almost any time during the course of fermentation [79]. Phosphate addition appears to disrupt the typical diphasic behavior of antibiotic fermentations. In a normal phosphate-limited fermentation, rates of protein and RNA synthesis increase rapidly during the first 15 to 24 h and then decrease to minimum values at the time when antibiotic is first detected [96, 78 a]. This decrease of the rate of RNA synthesis (which might be involved in triggering the onset of secondary metabolism) is characteristic of the end of the trophophase. When a high concentration of phosphate is added, this decrease does not occur and RNA synthesis continues at a very high rate. When a late decrease in RNA synthesis occurs, onset of antibiotic synthesis takes place again, although with a considerable delay (50 h or more). The transition from trophophase to idiophase inPenicillium urticae, as measured by the appearance of 6-methyl salicyclic acid, is also delayed by addition of inorganic phosphate [21 ]. Phosphate addition to ristomycin-producing cultures ofProactinomyces vulgaris produced a shift of metabolism, demonstrated by a marked accumulation in the medium of volatile organic acids (formic and acetic acids) and a decrease in ketoacids (pyruvic and acetoacetic acids) [ 138, 139]. There is little doubt that in a large number of microbial processes phosphate exhaustion is related to a decrease in RNA synthesis and the initiation of antibiotic production.
Control of Antibiotic Synthesisby Phosphate
109
Labeled phosphate added at inoculation is rapidly exhausted in control candicidin fermentations, whereas high levels remain in phosphate-supplemented cultures [78 a]. Most of this excess phosphate is utilized in the biosynthesis of nucleic acids [Martin and Demain, submitted] while some seems to be accumulated as intraceUular polyphosphate (see below). The ability of phosphate to delay the onset of antibiotic synthesis is presumably due to the time taken for phosphate exhaustion. A shiftdown in primary metabolism due to exhaustion of phosphate may trigger the onset of secondary metabolism. This hypothesis is supported by the finding that phosphate is often the nutrient exhausted from the medium at the time of onset of antibiotic synthesis. This has been seen in the production of tetracycline [58], vancomycin [96], and levorin [12]. Lopatnev and co-workers [80] observed that the extracellular phosphate concentration in nystain fermentation decreased during growth and remained at a very low level during the production phase. Recent studies by Liras and Martin [78 a] using 32P-labeled phosphate confirmed these results, indicating that phosphate is exhausted from the medium before onset of candicidin synthesis; phosphate disappearance was due to intracellular accumulation and nucleic acid synthesis. Intemal conversion of phosphate to polyphosphate (volutin) granules in cultures producing antibiotics was suggested by Prokofieva-Belgovskaya and Popova [110] and was confirmed recently by Hostalek et al. [61 ]. In continuous culture studies of chlortetracycline production, it was shown that although the incoming medium contained 50 to 60/~g of phosphorus per mt, the phosphorus level in the chemostat was undetectable at dilution rates of O. 152, 0.138, and 0.092 h -1. Constant levels of antibiotics were obtained at these three dilution rates throughout the fermentations. At a very low dilution rate (0.049 h - ~), the chlortetracycline level decreased during the fermentation, suggesting that at this growth rate, the intracellular phosphate was sufficient to inhibit chlortetracycline biosynthesis [127]. According to Bu'Lock [20] and Bu'Lock et al. [23] exhaustion of any nutrient makes balanced growth impossible and brings on biochemical differentiation, autolysis following carbon exhaustion and secondary metabolism following nitrogen or phosphorus exhaustion [48]. Bu'Lock proposed that nutrient limitation is a continual process rather than a sudden event. Different degrees of nutrient limitation are, therefore, required to trigger the onset of different secondary metabolic processes. Thus, in glycine-limited batch fermentations of Gibberellafu/ikuroi, bikaverin synthesis began at about 30 h when the glycine level had fallen to about 13 mM, whereas the onset of gibberellin synthesis occurred 20 h later at a lower concentration of glycine. In chemostat runs, bikaverin synthesis occurred at specific growth rates (g) lower than 0.05 h - 1 while gibberellin synthesis took place at less than 0.01 h-1. It thus appears that antibiotic synthesis is inhibited during balanced growth at high growth rates and its expression requires a degree of growth limitation which may be different in different producer strains and for different antibiotics. The possible involvement of nucleoside polyphosphates as effectors mediating the growth-rate determined regulation of antibiotic synthesis will be discussed below.
110
J.F. Martfn
3.2 Phosphate Shifts Carbohydrate Catabolic Pathways This hypothesis [56, 58] suggests that an increased phosphate concentration in the medium decreases the activity of the pentose phosphate (hexose monophosphate) pathway favoring glycolysis. In support, inhibitors of glycolysis such as fluoride, iodoacetate, and benzylthiocyanate stimulate, under certain circumstances, the biosynthesis of chlortetracycline. Using C1- or C6-1abeled glucose, it could be demonstrated that mycelium growing in medium supplemented with phosphate showed lower values of the C1/C6 quotient (i.e., 1"CO2formed from 14Cl-glucose/14CO2evolved from 14C6glucose) than did control mycelium. A low C1/C6 quotient suggests a preferenctial metabolism through the glycolytic pathway. However, this method of calculation of the relative importance of alternative pathways of glucose catabolism [ 151 ] has been seriously criticized. A switch in the carbohydrate catabolic pathway has previously been described in the transition from trophophase to idiophase. Bu'Lock and co-workers [21] using this method concluded that ~4Cl-glucose metabolism, artd therefore the pentose phosphate route, clearly predominates in the rapid growth phase (trophophase), whereas glycolysis is more important in the production phase (idiophase) in P. urticae. Similarly, Heath and Koffler [55] noted in P. chrysogenum that the pentose phosphate pathway is preferentially used during the growth phase of Penicillium prior to penicillin production. The C 1/C6 quotient values show that in tetracycline biosynthesis, the pentose cycle participates in the catabolism of glucose between the 12th and the 18th h, that is, at the time of onset of tetracycline biosynthesis [65]. Phosphate control of the oleandomycin (a macrolide antibiotic) biosynthesis by S. antibioticus may occur by a similar mechanism. Orthophosphate inhibits the hexose monophosphate pathway, decreasing the rates of hexose utilization and pyruvate formation [136]. Since the major pathway of carbohydrate metabolism in S. antibioticus appears to be the hexose monophosphate pathway, its inhibition by phosphate could account for the inhibition of oleandomycin synthesis [ 137]. If phosphate switches the carbohydrate catabolic pathways from the hexose monophosphate shunt to glycolysis, then NADPH would become the limiting factor in antibiotic synthesis. This seems to be the case in the biosynthesis of a number of polyketidederived compounds, such as tetracycline [11 ]. Gatenbeck and Hermodsson [47] have established that the concentration of reduced pyridine nucleotides determines the in vitro incorporation of acetate into polyketide-derived compounds or fatty acids. Moreover, the switch of the carbohydrate catabolic pathways may explain the decreased synthesis of aromatic precursors of several antibiotics, since the initial reaction of the aromatic pathway is the condensation of phosphoenolpymvate and eythrose-4-phosphate, the latter being formed through the pentose phosphate cycle. This model of the phosphate effect has received little attention in recent years. Future studies will have to determine whether there are allosteric enzymes sensitive to phosphate, and ascertain the definitive reason why the utilization of one pathway or the other results in altered synthesis of antibiotic.
Control of Antibiotic Synthesis by Phosphate 3.3 Phosphate Limits Synthesis of the Inducer of the Antibiotic
111
Pathway
This hypothesis has been receiving increased attention recently, owing to the discovery of inducers of antibiotic synthesis. Such effectors of antibiotic synthases are recognized by their effect on product formation when added during the growth phase and their lack of effect when added during the idiophase, i.e., they do not behave as activators or precursors. Thus, methionine appears to be an inducer of cephalosporin and fosfomycin synthesis [38, 120], while staphylomycin formation is brought about by an inducer produced by the culture itself [ 154]. 4-Trans.n-propyl-L-proline is apparently an inducer of the emericins, peptide antibiotics produced by Emericellopsis microspora, even though none contains propylpoline [5]. Perhaps one of the more intriguing induction effects is that of factor A in streptomycin biosynthesis. Factor A, which has a structure totally unrelated to streptomycin, induces streptomycin synthesis in nonproducer mutants. Addition of 1/ag of factor A results in the production of 3 to 5 mg of streptomycin [67, 68]. The best known example of induction is that of tryptophan in the production of ergot alkaloids by Claviceps. Tryptophan is both a precursor and an inducer of alkaloid synthesis. This amino acid and some of its analogues (e.g., thiotryptophan) act by inducing dimethylallytryptophan synthase (DMAT synthase) [41, 70, 119] the first enzyme of alkaloid biosynthesis. Increased DMAT-synthase levels, which correlate with the degree of alkaloid formation, are due to higher intracellular pools of tryptophan [22]. Analyses of the endogenous mycelial tryptophan concentration in Claviceps have shown that there is a two. to threefold transient increase in the level of free tryptophan in the mycelium during the transition period from growth to alkaloid production [ 119]. In early studies it was noted that a high concentration of inorganic phosphate inhibits alkaloid synthesis [34, 93, 132]. It was postulated that excess phosphate stimulates nucleic acid and protein synthesis, thus extending the growth phase. The effect of phosphate was thought to be due to the prolongation of the growth phase. However, this hypothesis was disproved by Robbers et al. [119] who found that L-tryptophan overcomes the inhibitory effect of high phosphate concentration on alkaloid production, indicating that the phosphate effect was mediated through tryptophan. It was suggested that high concentrations of phosphate limit the formation of tryptophan. The results of Rao and Gupta [115] who described the inhibition of tryptophan synthase by phosphate in alkaloid-producing cultures of Aspergillus fumigatus support this hypothesis. The recent studies of Krupinski et al. [70] revealed that thiotryptophan (like tryptophan itself) can overcome the blocking of alkaloid synthesis by inorganic phosphate. The result~ with thiotryptophan show that the phosphate effect cannot be explained merely on the basis of inhibition of synthesis of tryptophan as a substrate. It is possible, though, that in high-phosphate cultures alkaloid-forming enzymes are not formed, apparently because an inducing level of tryptophan is not reached. An interpretation linking this explanation of the phosphate effect with the model of a shift in the carbohydrate catabolic pathways is as follows. Phosphate causes a shift from the pentose phosphate shunt to the glycolytic pathway with a resultant decrease in the activity of the tryptophan biosynthetic pathway. Thiotryptophan would still induce DAMT synthase, even in high phosphate concentration.
112
J.F. Martfn
3.4 Phosphate Inhibits the Formation of Antibiotic Precursors Phosphate affects the levels of precursors involved in antibiotic synthesis. Knowledge of precursors and metabolic intermediates is scanty, as is information concerning the phosphate effect at this level. There is, however, some information that deserves consideration. The production of tetracyclines, polyene and nonpolyene macrolides, and other polyketide-derived antibiotics synthesized by a condensation of acetyl-CoA and malonyl-CoA [30, 85,143] is extremely sensitive to phosphate. Formation of the elongation units (malonyl-CoA and/or methylmalonyl-CoA) that participate in the biosynthesis of these antibiotics proceeds via carboxylation of acetyl-CoA by a reaction catalyzed by acetyl-CoA carboxylase (acetyl-CoA ligase, EC 6.4.1.2), or by a transcarboxylation reaction in which oxalacetate is the carboxyl group donor, catalyzed by methylmalonyl-CoA pyruvate carboxyltransferase (EC 2.1.3.1) [87, 113]. Oxalacetate formation by the action of phosphoenolpyruvate carboxylase, an anaplerotic enzyme, has been described as the limiting step in the biosynthesis of polyene macrotides [ 114]. Inorganic phosphate, as well as ATP, inhibits the activity of phosphoenolpyruvate carboxylase in S. aureofaciens [ 144] and thus the formation of oxalacetate units needed for polyene biosynthesis. According to Hostalek and co-workers [60] the levels of individual tricarboxylic acid cycle enzymes in high-production (chlortetracycline) strains are substantially lower than in low-producing strains. Three carbon intermediates formed during carbohydrate metabolism do not pass through the tricarboxilix acid cycle, but rather are directly involved in the antibiotic biosynthetic process. The decrease in the levels of the tricarboxylic acid cycle enzymes correlates with a decreased capacity of the respiratory chain and a lower rate of ATP formation in high-producing strains (Fig. 1). Phosphate, which activates the respiration rate and macromolecule synthesis, may increase the activity of the Krebs cycle, thus decreasing the concentration of precursor available for antibiotic synthesis. Other examples are the inhibition of tryptophan biosynthesis in ergot alkaloid formation by Claviceps purpurea (see above) and of the synthesis of inositol, a precursor of the streptidine moiety of streptomycin. Although not studied in an antibiotic producer, it has been found in Neurospora crassa that phosphate addition leads to a high concentration of pyrophosphate which is a competitive inhibitor of D-glucose6-phosphate-cycloaldolase (NAD+-dependent) [ 108]. This enzyme catalyzes the conversion of glucose-6-phosphate to inositol-l.phosphate and seems to be the only enzyme responsible for inositol biosynthesis in all organisms so far examined. Since the reaction is presumably part of the streptidine biosynthetic sequence, this mechanism could contribute to phosphate inhibition of streptomycin biosynthesis. In the biosynthesis of peptide antibiotics, pyrophosphate may inhibit the amino acid activation reactions by a feedback mechanism. Biosynthesis of these antibiotics proceeds through activation reactions by ATP to form aminoacyl adenylates, releasing pyrophosphate [76]. Pyrophosphate as well as phosphate is in effect inhibitory for gramicidin S formation [142].
Control of Antibiotic Synthesis by Phosphate
113
>. I-.-
1!, ~
~
Z" IzlO
~o
~ o~
oF-_z,~
o0..~E ,~
~'
I..-
<[
~g ~.~ b_ o i
..J ~
~
.~1
..
.~
~
o__,
,, .._~ ~, ~
? .__,...,
'~
o
N
"
"'
~
0 t)
I--
0
~.
..1
n-
~
-r
_
,
0
---
I
?~oa
,~
b_
_J Y 0
tN "7--1-
÷ Zb-
~ ~
oN ~..~
~b.
•~ o ~
t
"-
7
~
114
J.F. Martin
3.5 Phosphate Inhibits or Represses Phosphatases Necessary for Antibiotic Biosynthesis In the biosynthesis of certain antibiotics, such as streptomycin [98], viomycin [ 106], and neomycin [82], some intermediates are phosphorylated, whereas the end products are not. Phosphatases which cleave phosphorylated intermediates could be regulated via feedback inhibition or repression by inorganic phosphate. The biosynthesis of streptomycin, which is markedly inhibited by phosphate, includes at least three phosphatecleaving steps in the formation of the streptidine moiety [33] (Table 2). Table 2. Phosphate-cleavagesteps in the biosynthesisof streptomycin 1) 2) 3) 4)
D-Myo-inositol-P--->Myo-inositol+ Pi O-phosphoryl-N-amidino-scyllo-inosamine ~ N-amidino-scyllo-inosamine + Pi
O-phosphoryl-streptidine~ Streptidine + Pi Streptomycin phosphate ~ Streptomycin + Pi
Miller and Walker [97] reported that a phosphorylated derivative of streptomycin accumulates in cultures of S t r e p t o m y c e s griseus growing in excess inorganic phosphate. The phosphorylated derivative, which contains a phosphate ester at position 6 of the streptidine moiety, is biologically inactive [98,102, 103]. Experiments of Walker and co-workers suggest that the phosphate group is introduced during biosynthesis of the streptidine moiety [ 146]. The dihydrostreptose moiety is transferred enzymatically from dTDP-dihydrostreptose to streptidine-6-phosphate to form O-a-L-dihydrostreptose (1 -+ 4)-streptidine-6-phosphate [67a]. Streptomycin-P could be therefore an obligatory intermediate in the biosynthesis of streptomycin (Fig. 2). NH
~H
C-NH 2 I~IH
C-NH2 NH
H
H
~. C-HH
CI'I2OH
I~IH
0
OH
0
OH
0 .
0
/
o
0 ~,,,,,J
o
OH
Glucose
myo-Inositol
o
OH
Streptidine-P Streptomycin-PStreptomycin
Fig. 2. Pathway of biosynthesis of streptomycin. Streptidinephosphate and streptomycinphosphate are phosphorylated intermediates involvedin streptomycin biosynthesis (see text)
Control of Antibiotic Synthesis by Phosphate
115
In support of this concept, streptidine-P (but not free streptidine) has been found in mycelium fed with myo-~4C-inositol. These findings were confirmed by the group of Nomi [ 101 ], who proved that streptidinephosphate is incorporated into streptomycinphosphate. The final intermediate in streptomycin biosynthesis is thus streptomycinphosphate, with the phosphate esterified to the C6 hydroxyl group of the streptidine moiety [103]. Free streptomycin may be phosphorylated at the C6 of the streptidine moiety of streptomycin [99, 100] by a phosphorylating enzyme in presence of ATP and Mg+÷. Recent work by Walker and co-workers [ 147, 149, 150] described three streptomycin kinases with different substrate specificities: streptomycin-6-kinase (stable, acting at C6 of the streptidine moiety), dihydrostreptomycin-3'-kinase (labile, acting at the 3'-carbon of the streptose moiety) and streptomycin-3"-kinase (labile, phosphorylating the 3"carbon of the N-methyl-L-glucosamine moiety. They are able to mono. or diphosphorylate (but not triphosphorylate) the streptomycin molecule. It is interesting that the streptomycin-3"-kinase is believed to phosphorylate at the same position as a streptomycin resistance enzyme in E. coli and Pseudomonas aeruginosa coded in an R factor [ 149], lending support to the hypothesis that certain R factors carrying enzymes for antibiotic inactivation may have originated in Streptomyces [ 13]. Streptomycin-producing strains may thus utilize the kinases to maintain streptomycin, a potential inhibitor of protein synthesis, in a nontoxic form. Streptomycin and streptomycin.P may represent active and inactive forms of a normal physiological regulator of protein synthesis in streptomycin-producing strains. Young cells of Streptomyces griseus are indeed sensitive to exogenous streptomycin [25, 26]. In this context, phosphate regulation of streptomycin synthesis may be understood as a protective mechanism of protein synthesis during the growth phase in streptomycin-producing strains. Streptomycin phosphatase, an idiophase enzyme which is present in streptomycin producers, but not in strains which cannot produce streptomycin, is inhibited, but not repressed by phosphate [ 148]. Similar mechanisms have been seen in the biosynthesis of neomycin B by Streptomyces fradiae. Neomycin phosphates have been described; these include neomycin B pyrophosphate and neomycin C pyrophosphate, in which the pyrophosphate group is attached to an amino group of the antibiotic [82]. The alkaline phosphatase of S. fradiae is inhibited and repressed by inorganic phosphate [84]. The alkaline phosphatase of S. fradiae is synthesized late in the fermentation and there is a direct relationship between enzyme activity and neomycin formation. The appearance of alkaline phosphatase is due to de novo protein synthesis, demonstrated by the inhibition of its synthesis with chloramphenicol [9]. Phosphate inhibition of alkaline phosphatase in S. orientalis may be responsible for the inhibition of vancomycin formation, although no phosphorylated intermediates have been reported so far [96]. A phosphorylated intermediate (or metabolite) of verdamicin, another aminoglycoside antibiotic, has been reported by Lee et aL [78], but the nature and the biosynthetic role of such phosphorylated intermediates have yet to be explained. The alkaline phosphatase of Proactinomyces fructiferi var. ristomicini, the producer of the antibiotic ristomycin, is completely repressed at concentrations of inorganic phos-
116
J.F. Martfn
phate that inhibit antibiotic production. No phosphorylated intermediates of ristomycin are known. It is of interest that the final product (ristomycin), when added to the fermentation, lowered by two to three times the activity of alkaline phosphatase [130, 139]. 3.6 Other Possible Mechanisms Weinberg [ 152] suggested that inorganic phosphate suppresses antibiotic production by depriving the cell of an essential metal. It is well known that phosphate readily forms precipitates with calcium, magnesium, iron, and other metals. Furthermore, nonspecific stimulation of antibiotic synthesis by several metal ions is known [ 152]. However, Liu et al. [79] failed to show a neutralization effect by zinc or magnesium ions on phosphate inhibition of candicidin biosynthesis. Haavik [50] claimed that phosphate inhibition of bacitracin synthesis is not due to metal deprivation, since bacitracin production is restored by further increasing the concentration of phosphate. He postulated instead that the effect of phosphate on bacitracin production is due to pH changes. However, experiments with controlled pH necessary to support his claim were not carried out. Studies of candicidin synthesis in pH-controlled fermentors suggest that the pH of the medium has nothing to do with the effect of phosphate on the biosynthesis of this antibiotic [79]. From the information cited in the above sections, it would appear that phosphate acts by different mechanisms in inhibiting the biosynthesis of various antibiotics. However, it is possible that these different mechanisms are mediated by a common effector, such as the intracellular ATP concentration or the adeffylate energy charge of the cell.
4. Studies o n the P h o s p h a t e E f f e c t Using Resting Cells Most studies concerned with the phosphate-induced shift of metabolism have been carried out in long-term batch experiments in which the effect could not be clearly attributed to an interaction with primary or secondary metabolism. To avoid this problem, a phosphate-limited resting cell system in chemically defined medium [89, 90] and continuous culture studies have been used [ 127]. Production of candicidin by S. griseus in such a resting cell system was linear for at least 36 h without evidence of growth. Short-term experiments in the resting cell system differentiated factors affecting mycetial growth from those acting on antibiotic production. Addition of 5-10 mM phosphate decreased the rate of antibiotic formation as well as the incorporation of labeled precursors into candicidin (Fig. 3). [88]. The inhibition was concentrationdependent and not due to an effect on the uptake of the precursors. Synthesis of protein and RNA were not affected [88]. Thus phosphate inhibition of candicidin synthesis in short-term experiments is independent of the reactivation of protein or RNA synthesis which one sees upon phosphate supplementation in long-term experiments.
Control of Antibiotic Synthesis by Phosphate
117 E
|
190
D CONTROL
"o 80
Z
e,3
CONTROL
~ 70 z <
P.OSPHATEO0 IO"M_
t) O I.z
i,o
PHOSPHATE O ,s
(IO x IO-3M )
o zI lid i.< n_ O a,. Q.
+~
1,
20
~
~ ~o
j°
9 2o 0
IO
! O
0
~ 40
,o
d 80
120
I 20
160 TIME
I 40
I 60
I 80
I I00
(re|n)
Fig. 3. Phosphate inhibition of candicidin synthesis by resting cells o f S. p h o s p h a t e or c-AMP effect on protein synthesis (right).
griseus (left).
Note tack of
5. Possible Intracellular Effectors Which Might Regulate Gene Expression Related to Antibiotic Synthesis It is clear from above that phosphate concentration controls the differential expression of growth vs. antibiotic synthesis. It is unclear, however, how the nutritional message is transmitted to those mechanisms which control expression of antibiotic production genes. Is phosphate the ultimate effector, or does it simply regulate the level of an intracellular effector, which in turn controls antibiotic synthesis? Such phosphatecontrolled intracellular effectors could include cyclic nucleotides, ATP, the adenylate energy charge, polyphosphates, and highly phosphorylated adenine or guanine nucleotides.
5.1 Cyclic Nucleotides Adenosine-3':5'-cyclic monophosphate (c-AMP) exerts its action on prokaryotes by stimulating the synthesis of inducible enzymes and reversing catabolite repression.
120
118
J.F. Martin
According to Rickenberg [ 118] synthesis of secondary metabolites might well be regulated by c-AMP since they are only formed after bacterial growth has slowed or stopped. Bu'Lock [20] also suggested that c-AMP might be an effector transmitting nutritional signals that control the production of secondary metabolites. However, there are no experimental facts supporting these hypotheses, c-AMP has not been reported in the Actinomycetales [ 118], nor in bacilli [52]. The actinomycetes and the bacilli are the two major antibiotic-producing taxonomic groups. The only reports possibly implicating c-AMP in secondary metabolism are preliminary ones: a) the stimulation of the synthesis of m-hydroxybenzyl alcohol dehydrogenase, and enzyme of patulin biosynthesis in Penicillim urticae, by dibutyryl c-AMP [42]; b) the reversal by c-AMP of the ethidium bromide and acridine orange inhibition of O-demathylpuromycin methyltransferase, a catabolite-sensitive enzyme involved in puromycin biosynthesis in S. albaniger [ 121 ]. Guanosine-3':5'-cyctic GMP (c-GMP), however, has been reported in Bacillus [ 14] but no firm evidence exists concerning its role. 5.2 ATP and Adenylate Energy Charge The involvement of ATP as an effector of antibiotic synthesis was suggested initially by Silaeva et at. [ 126] who reported lower concentrations of adenine nucteotides in strains producing gramicidin S than in nonproducing strains of Bacillus brevis. An inverse correlation appears to exist between the intracellular level of ATP and synthesis of chlortetracycline [31,64]. A recent report [45] refers to a possible regulatory effect of ATP in tyrothricine biosynthesis, although no definitive evidence was provided. Martin and Demain [92] found that exogenous nucleotides inhibited candicidin synthesis in a similar way to inorganic phosphate. All nucleotides tested, regardless of the esterified position, were inhibitors, whereas nucleosides and free bases were not. This suggests that the inhibitory effect is related to the phosphate groups contained in the nucleotides. Uptake studies showed that AMP was dephosphorylated before or during transport into the cells. The phosphate moiety was taken up much more rapidly by the cell than the adenine moiety. Thus, in the case of candicidin biosynthesis the effect of exogenous nucleotides is merely an inorganic phosphate effect. The intracellular concentration of ATP or the adenylate energy charge of the cell controls sporogenesis in Bacillus subtilis [63]. It is also possible that peptide antibiotic formation in B. subtilis is similarly regulated [122]. MartEn and co-workers [Martin et al., submitted] found, in the candicidin-producing strain of S. griseus a two- to threefold increase in intracellular ATP within 5 min after inorganic phosphate addition. This rapid increase in the ATP pool occurred before the observed inhibition of antibiotic synthesis, as determined by incorporation of labeled precursors into candicidin. The same was noted upon addition of GMP to the resting cells. If the increase in ATP is related to the inhibition of biosynthesis, it is important to determine whether ATP itself or the adenylate energy charge is the true regulatory effector. Energy charge (EC) is defined [6, 7] as the mole fraction of ATP plus half the mole fraction of ADP. Atkinson and co-workers propose that it is a regulatory parameter coordinating energyutilizing and energy-generating metabolic pathways of primary metabolism. (ATP) + ½(ADP) EC = (ATP) + (ADP) + (AMP) "
Control of Antibiotic Synthesisby Phosphate
119
EC is a linear measure of the amount of energy stored at any time in the adenylate system. Atkinson [7] pointed out that regulatory enzymes which generate ATP are highly active at low levels of EC and their activities decrease sharply as EC is raised above 0.75. Regulatory enzymes which consume ATP exhibit very little activity at low EC levels and their activities increase sharply above 0.75. Atkinson and co-workers [8, 27] maintain that EC is more important than the absolute concentration of ATP. This hypothesis has come under criticism and some reports indicate that large changes in the total concentration of adenine nucleotides which affect enzyme activity do not result in significant variations in the adenylate energy charge [112, 35]. The EC of resting cells of S. griseus, is 0.8 and changes only slightly following addition of 10 mM phosphate [Martfn et al., submitted]. Thus, concentrations of phosphate inhibitory for candicidin synthesis increase intracellular ATP to a much greater extent than EC. Additional evidence supporting a regulatory role for ATP in antibiotic production was obtained following the changes in the intracellular level of ATP as a function of time during batch cultures ofS. griseus. ATP was high during the rapid growth phase, but declined rapidly at the onset of antibiotic biosynthesis when phosphate was depleted; ATP remained at this low level during the candicidin production phase [78 a]. A similar pattern was observed in S. aureofaciens during the tetracycline fermentation [64, 72]. ATP concentration, and therefore its allosteric effects, might be controlled by phosphate through the action of adenosine triphosphatases (ATPases). Hostalek et at [61 ] have reported that the total ATPase activity of S. aureofaciens rises rapidly after 24 h of cultivation, when the growth phase of the culture terminates. Two enzyme activities were described: a Mg2÷ ATPase typically associated with bacterial membranes and a cytoplasmic ATPase. The synthesis of the latter was repressed by inorganic phosphate. It appears then that exhaustion of phosphate from the broth leads to a drop in the ATP content of the cell, accompanied by a derepression, and/or release from inhibition, of antibiotic.forming enzymes (see below). Such a rise in enzyme activity, together with accumulation of antibiotic precursors due to idling of macromolecular biosynthesis could result in the onset of antibiotic synthesis (Fig. 1). The shift of carbohydrate catabolic pathways upon phosphate addition proposed by Hostalek [58] (see Section 3.2) could be interpreted as an ATP-mediated regulatory effect. Since ATP (or EC) acts as a fine control of key branching enzymes such as phosphofructokinase, citrate synthase, and NAD(P)H-requiring isocitrate dehydrogenase [7], changes in the level of adenine nucleotides could result in a shift of the metabolic pattern of the cell [59, 60, 144]. 5.3 Polyphosphates Early results suggested that under conditions of excess phosphate, cells of S. aureofaciens accumulate polyphosphates [49, 110]. The role of polyphosphates, if any, in the regulation of primary and secondary metabolism is unclear. Their function as a phosphorus reserve and their role in the regulation of phosphorus economy of the cell have been reviewed by Harold [54]. He postulated that the physiological role of polyphosphates is that of regulating the levels of inorganic phosphates, ATP, and ADP, and
120
J.F. Martfn
that the polyphosphate cycle serves as a metabolic phosphate buffer. Polyphosphate accumulation is generally associated with nutritional conditions in which growth (and thus nucleic acid and phospholipid synthesis) is inhibited, but assimilation of inorganic phosphate from the medium continues, lntracellular release of phosphate (or nucleotides) by degradation of nucleic acids is balanced by polyphosphate accumulation [53]. Storage of polyphosphate is regarded then as a mechanism for the regulation of the intracellular inorganic phosphate level. Phosphate starvation leads to derepression of phosphatases, which makes available inorganic phosphate from organic phosphate esters, producing a phosphate excess; released phosphate is sequestered as polyphosphate. With this idea in mind, we may discuss a possible role of polyphosphate in antibiotic synthesis. Several studies have been published on the significance of polyphosphate in chlortetracycline production by S. aureofaciens. Hostalek and co-workers [ 61, 62] pointed out the possible significance of polyphosphate: D-glucose-6-phosphotransferase (polyphosphate glucokinase, EC 2.7.1.63) in the chlortetracycline biosynthesis. Activity of polyphosphate glucokinase in a high-producing S. aureofaciens strain appears after 16 to 20 h of growth, at the beginning of the phase of chlortetracycline production (Fig. 4), levelling off after 48 h of cultivation. It was suggested that this enzyme, which
O
O,
~o
11.5
O
~a
ao2
1.o
o.o,
!o.5
®
o
O, o~ 0
24
48
72
hours
Fig. 4. Activityof enzymesinvolvedin polyphosphatebiosynthesisduring tetracyclinefermentation. Redrawnfrom Hostateket aL [621. ATP-glucotinase(o), polyphosphateglucokinase(o), intracellular ATP (a), chlortetracycline(*), and dry weight (A)
Control of Antibiotic Synthesis by Phosphate
121
uses polyphosphate to phosphorylate glucose, participates in an alternative non-ATP route of glucose phosphorylation during the antibiotic production phase. The glucose6-phosphate formed in the reaction would be converted glycolytically to the antibiotic precursor malonyl-CoA. Since ATP levels and TCA cycle activities during the production phase are quite low (see above), this altemative form of glucose phosphorylation would represent a considerable saving of ATP in the antibiotic-producing cell. Polyphosphate would act as phosphorus and energy donor during the process of chlortetracycline biosynthesis. Polyphosphate glucokinase activity appears to be higher during the production phase in a high-producing strain than in a low-producing strain of S. aureofaciens [61]. Supporting this hypothesis is the fact that the alternate glycolytic pathway enzyme, ATP:glucose-6-phosphotransferase (EC 2.7.1.2), also known as ATP glucokinase or hexokinase, is very active during the rapid growth phase and declines thereafter, i.e. the maximum activity of ATP glucokinase occurs at the time of high intracellular ATP content. Further studies in resting cells or in vitro systems are required to confirm the role of pholyphosphate in antibiotic formation. If this hypothesis is confirmed, the intracellular content of polyphosphate would be an important factor in antibiotic synthesis. When and how is the high energy polyphosphate formed in S. aureofaciens? Two polyphosphate synthesizing enzymes were studied by Hostalek and co-workers [61] in high- and low-producing strains. Of these, the ATPpolyphosphate transferase does not appear to be involved in polyphosphate biosynthesis in these organisms. The second enzyme, 1,3-diphosphoglycerate: polyphosphate phosphotransferase (which catalyzes the transfer of high-energy phosphate from 1,3-diphosphoglyceric acid to polyphosphate), appears more important [72]. Its activity doubles at the beginning of chlortetracycline production. Some difficulty in the interpretation of this hypothesis arises from the heterogeneous composition and Origin of the cellular polyphosphates. The high-molecular-weight polyphosphate is localized in the peripheral regions of the cell close to the cytoplasmic membrane and might be involved in active transport of glucose [ 141]. Medium-weight polyphosphates are mainly present in the nucleus and may be associated with nucleic acid metabolism [74]. The acid-soluble lowmolecular.weight polyphosphate is found in the cytoplasm as volutin granules [71 ]. The latter polyphosphate is thought to be the donor of phosphate during chlortetracycline biosynthesis. The polyphosphate-synthesizing 1,3-diphosphoglycerate:polyphosphate phosphotransferase is localized in the volutin granules in Neurospora crassa [73]. Many questions remain concerning the control of polyphosphate accumulation, polyphosphate glucokinase activity, and the initiation of antibiotic synthesis. Hostalek and co-workers [62, 75] suggest that it is the level of ATP or EC which regulates the activities of ATP-glucokinase, polyphosphate glucokinase, and 1,3-diphosphoglycerate polyphosphate phosphotransferase, and chlortetracycline formation. 5.4 Higlfly Phosphorylated Nucleotides In the last few years, knowledge has been obtained on the mechanisms that control macromolecular synthesis during nutritional downshift. A particular aspect is the control of stable RNA synthesis following deprivation of a required amino acid, limitation of tRNA aminoacylation, or addition of inhibitors of amino acid synthesis. Wild-type
122
J.F. Martfn
E. coti shows a stringent control of RNA synthesis following amino acid starvation, whereas relaxed mutants ofE. coli abnormally accumulate RNA after amino acid starvation. This mutation has been mapped to the locus designated rel A [24]. Stringent
control is mediated by highly phosphorylated guanosine nucleotides. Both guanosine tetraphosphate (ppGpp) and guanosine pentaphosphate (pppGpp) are accumulated by stringent strains orE. coli during amino acid starvation but not by relaxed mutants. Moreover these highly phosphorylated nucleotides decrease after supplementing with amino acids. These highly phosphorylated nucleotides are effectors that function as inhibitors of RNA synthesis in response to a limitation in protein synthesis. The mechanisms involved in shiftsdown of nutrients other than amino acids, e.g., carbon step-downs [128], depletion of energy by interruption of aeration or cyanide treatment [39, 81 ] and nitrogen or sulfur starvation [39], are less clear; they involve ppGpp accumulation, but are independent of the tel A gene product, i.e., they occur identically in stringent or relaxed strains. According to Cashel [24], tel-A-independent accumulation of ppGpp suggests that ppGpp may participate in growth-rate regulation of RNA synthesis during the growth phase. Results of Lazzarini et aL [77] and Sokawa et al. [ 129] support this hypothesis. They found an inverse correlation between the basal levels of ppGpp (in both stringent and relaxed strains) and the RNA content per unit of DNA, over a range of growth rates corresponding to 40-205 min doubling time. As indicated previously, onset of antibiotic synthesis is a cellular response to a nutritional shiftdown. Derepression of antibiotic synthesis could be interpreted as the result of a stringent-type response of the cell to nutrient limitation. It is not known, however, whether this response is mediated by highly phosphorylated nucleotides. Adenosine-5'triphosphate-3'-diphosphate (ppApp)-synthesizing acitivity has recently been detected in five species of Streptomyces [104, 105]. It is also important to note that highly phosphorylated nucleotides of adenine (ppApp and pppApp) may be involved in the onset of sporulation in B. subtilis [116]. The stringent response of B. subtilis to amino acid deficiency, however, is mediated by guanosine tetraphosphate and guanosine pentaphosphate [ 117]. The postulated role of adenosine tetraphosphate and adenosine pentaphosphate in the regulation of spomlation may be relevant to antibiotic synthesis if we consider the close connection between sporulation and formation of peptide antibiotics by bacilli [122].
6. Alternative Levels of C o n t r o l of A n t i b i o t i c Synthesis The "late" expression of the synthesis of antibiotics and other secondary metabolites following the phase of intense growth is a well-known phenomenon. This sequential expression of genetic information coding for growth and antibiotic synthesis appears to be mediated by phosphate (see above). Generally speaking, the most common type of control mechanism involves repression of antibiotic synthases during the growth phase, their formation occurring only after the growth rate decreases (transcriptional control). The formation of many antibiotic-synthesizing enzymes follows this pattern, e.g., amidino-transferase in the biosynthesis of streptomycin [ 145], acyltransferase in
Control of Antibiotic Synthesisby Phosphate
123
the biosynthesis of penicillin [ 111 ], gramicidin S synthase [95, 134] and bacitracin synthase [44] in the biosynthesis of peptide antibiotics, phenoxazinone synthase in the formation of actinomycin [46], GTP-8-formylhydrolase in the biosynthesis of toyocamycin and sangivamycin [40a], and TDP-glucose oxidoreductase and TDP-4-keto-6deoxy-D-glucose transmethylase in the biosynthesis of tylosin [94]. Liras and Martfn [78 a], have recently observed derepression of candicidin-synthesizing enzymes after growth. These enzymes were not formed during the first 15 h of the batch, even though the highest rate of RNA synthesis occurred during this period. A transcriptional control by phosphate has therefore been suggested in the regulation of late expression of antibiotic synthesis. Repression of gene expression is not, however, the only mechanism involved in late expression of antibiotic synthesis. As mentioned above, addition of excess phosphate to candicidin-producing cultures immediately inhibits antibiotic synthesis [88]. This inhibition can be observed within 20 min following phosphate addition. When the effect of phosphate was studied under conditions of complete absence of RNA or protein synthesis, it was found that phosphate has the same inhibitory effect. Such an inhibitory effect independent of protein synthesis presumably involves inhibition of preformed antibiotic synthases [Martin and Demain, unpublished; 86]. It is thus probable that phosphate regulation acts, as do other regulatory mechanisms, at two levels: first, it rapidly inhibits the activity of antibiotic synthases; second, it represses formation of antibiotic synthases.
7. S u m m a r y and F u t u r e O u t l o o k There is a considerable body of evidence to show that phosphate has a strong inhibitory effect on the biosynthesis of many antibiotics. Some of the possible mechanisms of the phosphate inhibitory effect are also clear in the particular cases in which they have been studied. Thus, it is well known that phosphate inhibits or represses phosphatases involved in the biosynthesis of aminoglycoside antibiotics. Other mechanisms may be correct as well. What we do not know is if there is a common effector mediating these different reported mechanisms. It is generally accepted that onset of antibiotic synthesis is the cellular response to a nutritional downshift, usually phosphate depletion. It is unclear, however, how the nutritional message is transmitted to those mechanisms which control expression of antibiotic production genes. Present evidence suggests that ATP may be the effector controlling expression of antibiotic synthesis, although highly phosphorylated adenine or guanine nucleotides, or even cyclic nucleotides, cannot be excluded at this time. The final answer to these questions requires extensive research on the antibiotic-synthesizing enzymes and on the regulatory mechanisms controlling the formation and activity of these enzymes. Acknowledgements I thank Dr. Arnold L. Demain for editing the manuscript and Dr. Julio R. Villanueva for continuous support.
124
J . F . Martfn
References 1. 2. 3.
Abou-Zeid, A., Abou-el-Atta, A.: Zbl. Bakt. Parasitenk. Abt. II, 126, 371 (1971). Abou-Zeid, A.: Acta Microbiol. Pol. 4, 83 (1972). Aharanowitz, Y., Demain, A. L.: Proc. 16th ICAAC Conference, p. 49, American Society for Microbiology, Washington, D. C. (1976). 4. Alikhanian, S. I., Mindlin, S. Z., Orlova, N. V., Verkdaovtzeva, T. P.: Appl. Microbiol. 7, 141 (1959). 5. Argoudelis, A. D., Johnson, L. E.: J. Antibiot. 27, 274 (1974). 6. Atkinson, D. E., Walton, G. M.: J. Biol. Chem. 242, 3239 (1967). 7. Atkinson, D. E.: Ann. Rev. Microbiot. 23, 47 (1969). 8. Ball, W. J., Atkinson, D. E.: J. Bacteriol. 121,975 (1975). 9. Bandyopadhyay, S. K., Majumdar, S. K.: Antimicrob. Agents Chemother. 5,431 (1974). 10. Basak, J., Majumdar, S. K.: Antimicrob. Agents Chemother: 8, 391 (1975). I 1. Behal, V., Cudlin, J., Vanek, Z.: Folia Microbiol. 14, 117 (1969). 12. Belousova, I. I., Lishnevskaya, E. B., Elgart, R. E., Tereshin, I. M.: Antibiotiki 15, 224 (1970). 13. Benveniste, R., Davies, J.: Proc. Nat. Acad. Sci. USA 70, 2276 (1973). 14. Bernlohr, R. W., Haddox, M. K., Goldberg, N. D.: J. Biol. Chem. 249, 4329 (1976). 15. Biffi, G., Boretti, G., Di Marco, A., Pennella, P.: Appl. Microbiol. 2, 288 (1954). 16. Boeck, L. D., Clem, G. M., Wilson, M. M., Westhead, J. E.: Antimicrob. Agents Chemother. 3, 49 (1973). 17. Boretti, G., Di Marco, A., Julita, P., Raggi, F., Bardi, U.: Giorn. Microbiol. 1,406 (1956). 18. Brewer, D., Arsenault, G. P., Wright, J. L. C., Vining, L. C.: J. Antibiot. 26, 778 (1973). 19. Brinberg, S. L., Grabovskaya, O. B.: Mikrobiologiya 27,407 (1958). 20. Bu'Lock, J. D.: In: Industrial Aspects of Biochemistry. B. Spencer (ed.), Amsterdam: North Holland/Elsevier 1974, p. 335. 21. Bu'Lock, J. D., Hamilton, D., Hulme, M. A., Powell, A. J., Sepherd, D., Smalley, H. M., Smith, G. N.: Can. J. Microbiol. 11,765 (1965). 22. Bu'Lock, J. D., Barr, J. G.: Lloydia 31,342 (1968). 23. Bu'Lock, J. D., Detroy, R. W., Hostalek, Z., Munin-al-Shakarchi, A.: Trans. Br. Mycol. Soc. 62, 377 (1974). 24. Cashel, M.: Ann. Rev. Microbiol. 29, 301 (1975). 25. Celia, R., Vining, L. C.: Can. J. Microbiol. 20, 1591 (1974). 26. Celia, R., Vining, L. C.: Can. J. Microbiot. 21,463 (1975). 27. Chapman, A, G., Fall, L., Atkinson, D. E.: J. Bacteriol. 108, 1072 (1971). 28. Chigaleichick, A. G., Makhfud, M. K., Nefelova, M. V.: Mikrobiologiya 42, 76 (1973). 29. Claridge, C. A., Bush, J. A., Defuria, M. D., Price, K. E.: Develop. Ind, Microbiol. 15, 101 (1974). 30. Corcoran, J. W.: In: Genetics of Industrial Microorganisms, Z. Vanek, Z. Hostalek, and J. Cudlin (eds.), Prague: Academia 1973, Vol. II, p. 339. 31. Curdova, E., Kremen, A., Vanek, Z., Hostalek, Z.: Fotia Microbiol. 21,481 (1976). 32. Demain, A. L.: J. Appt. Chem. Biotechnot. 22, 345 (1972). 33. Demain, A. L., Inamine, E.: Bacteriol. Rev. 34, 1 (1970). 34. De Waart, C., Taber, W. A.: Can. J. Microbiol. 6, 675 (1960). 35. Dietzler, D. N., Lais, C. J., Magnani, J. L., Leckie, M. P.: Biochem. Biophys. Res. Commun. 60, 875 (1974). 36. Di Marco, A.: G. Microbiol. 2, 285 (1956). 37. Donovick, R., Brown, W. E.: In: Biogenesis of Antibiotic Substances, Z. Vanek and Z. Hostalek (eds.), Prague: Academia 1965, p. 281. 38. Drew, S. W., Demain, A. L.: Biotechnol. Bioeng. 15,743 (1973). 39. Edlin, G., Donini, P.: J. Biol. Chem. 246, 4371 (1971). 40. Egor0v, N. S., Toropova, E. G., Suchkova, L. A.: Mikorbiologiya 40, 475 (1971). 40.a Elstner, E. F., Suhaldonik, R. J.: J. Biol. Chem. 246, 6973 (1971).
Control of Antibiotic Synthesis by Phosphate
125
41. Floss, H. G., Robbers, J. E., Heinstein, P. F.: Recent Adv. Phytochem. 8, 141 (1974). 42. Forrester, P. I., Gaucher, G. M.: Biochemistry 11, 1108 (1972). 43. Frank, L. H., DeMoss, R. D.: J. Bacteriol. 77, 776 (1959). 44. Froyshov, O., Laland, S. G.: Eur. J. Biochem. 46, 235 (1974). 45. Fynn, G. H., Davidson, J. A.: J. Gen. Microbiol. 94, 68 (1976). 46. Gallo, M., Katz, E.: J. Bacteriol. 109. 659 (1972). 47. Gatenbeck, S., Hermodsson, S.: Acta Chem. Scand. 19, 65 (1965). 48. Grootwassink, J. W. D., Ehman, J. O., Gaucher, G. M.: In: Abstracts 5th Int. Ferm. Symp. H. Dellweg (ed.), p. 228, Berlin (1976). 49. Guberniev, M. A., Listvinova, S. N., Torbochkina, L. t., Katz, L. N.: Antibiotiki 6, 24 (1959). 50. Haavik, H. I.: J. Gem Mierobiol. 84, 226 (1974). 51. Hall, M. J., Hassall, C. H.: Appl. Microbiol. 19, 109 (1970). 52. Hanson, R. S.: In: Spores VI. P. Gerhardt, R. N. Costilow, and H. L. Sadoff (eds.), p. 318. Washington D. C., American Society for Microbiology 1975. 53. Harold, F. M.: Biochim. Biophys. Acta 45, 172 (1960). 54. Harold, F. M.: Bacteriol. Rev. 30, 772 (1966). 55. Heath, E. C., Koffler, H.: J. Bacteriol. 71,174 (1956). 56. Herold, M., Hostalek, Z.: In: Biogenesis of Antibiotic Substances. Z. Vanek, and Z. Hostalek (eds.), Prague: Academia 1965, p. 93. 57. Hoeksema, H., Smith, C. G.: Prog. Ind. Microbiol. 3, 91 (1961). 58. Hostalek, Z.: Folia Microbiol. 9, 78 (1964). 59. Hostalek, Z., Tinterova, M., Jechova, V., Blumauerova, M., Suchy, J., Vanek, Z.: Biotechnol. Bioeng. 11,539 (1969). 60. Hostalek, Z., Ryabushko, T. A., Cudlin, J., Vanek, Z. Folia Microbiol. 14, 121 (1969). 61. Hostalek, Z., Behal, V., Curdova, E., Vanek, Z.: Post. Hig. Med. Dosw. 28, 527 (1974). 62. Hostalek, Z., Tobek, I., Bobyk, M. A., Kulayev, I. S.: Folia Microbiol. 21,131 (1976). 63. Hutchinson, K. W., Hanson, R. S.: J. Bacteriol. 119, 70 (1974). 63.a lngledew, W. M., Campbell, J. J. R.: Can. J. Microbiol. 15,595 (1969). 64. Janglova, Z., Suchy, J., Vanek, Z.: Folia Microbiol. 14, 208 (1969). 65. Jechova, V., Hostalek, Z., Vanek, Z.: Folia Microbiol. 14, 128 (1969). 66. Katz, E., Pienta, P., Sivak, A.: Appl. Microbiol. 6, 236 (1958). 67. Khokhlov, A. S., Anisova, L. N., Tovarova, I. I.: Z. Allg. Mikrobiol. 13, 647 (1973). 67.a Kniep, B., Grisebach, H.: FEBS Letters 65, 44 (1976). 68. Kornitskaya, E. Y., Tovarova, I. I., Soifer, V. S., Khokhlov, A. S.: Antibiotika 21, 10 (1976). 69. Korshunov, V. V., Egorov, N. S.: Mikrobiologiya 31,515 (1962). 70. Krupinski, V. M., Robbers, J. E., Floss, H. G.: J. Bacteriol. 125, 158 (1976). 71. Kulayev, I. S., Belozerskii, A. N., Ostrovskii, D. N.: Biokhimiya 26, 188 (1961). 72. Kulayev, I. S., Szymona, O., Bobyk, M. A.: Biokhimiya 33, 419 (1968). 73. Kulayev, I. S., Konoshenko, G. I. Dokl. Akad. Nauk SSSR, 200, 477 (1971). 74. Kulayev, I. S., Vagabov, V. M., Tsiomenko, A. B.: Dokl. Akad. Nauk SSSR 204, 734 (1972). 75. Kulayev, I. S., Bobyk, M. A., Tobek, I., Hostalek, Z.: Biokhimiya 41, 343 (1976). 76. Laland, S. G., Zinmer, T. L.: Essays Biochem. 9, 31 (1973). 77. Lazzarini, R. A., Cashel, M., Gallant, J.: J. Biol. Chem. 246, 4381 (1971). 78. Lee, B. K., Condon, R. G., Wagman, G. H., Weinstein, M. J.: Antimicrob. Agents Chemother. 10, 363 (1976). 78.a Liras, P., Villanueva, J. R., Martfia, J. F.: J. Gen. Microbiol. In press (1977). 79. Liu, C. M., McDaniel, L. E., Schaffner, C. P.: Antimicrob. Agents Chemother. 7, 196 (1975). 80. Lopatnev, S. V., Kleiner, G. I., Bylinkina, E. S.: Antibiotiki 18, 608 (1973). 81. Lund, E., Kjeldgaard, N. O.: Euro. J. Biochem. 28, 316 (1972). 82. Majumdar, M. K., Majumdar, S. K.: Biochem. J. 120, 271 (1970). 83. Majumdar, M. K., Majumdar, S. K.: Folia Microbiol. 16, 285 (1971). 84. Majumdar, M. K., Majumdar, S. K.: Biochem. J. 122, 397 (1971). 84.a Makarevich, V. G., Upiter, G. D., Slugina, M. D., Tarasova, S. S., Gravit, N. F.: Antibiotiki 20, 295 (1975).
126 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131.
J . F . Martin Martfn, J. F.: Dev. Ind. Microbiol. 17, 223 (1976). Martin, J. F.: In: Microbiology 1976. D. Schlessinger (ed.), p. 548. Washington D. C., American Society for Microbiology 1976. Martin, J. F.: Ann. Rev. Microbiol. 31, 13 (1977). Martin, J. F., Demain, A. L.: Biochem. Biophys. Res. Commun. 71, 1103 (1976). Martin, J. F., Liras, P.: J. Antibiot. 29, 1306 (1976). Martin, J. F., McDaniel, L. E.: Eur. J. Appl. Microbiol. 3, 135 (1976). Martin, J. F., McDaniel, L. E.: Adv. Appl. Microbiol. 21, 1 (1977). Martin, J. F., Demain, A. L.: Can. J. Microbiol, in press (1977). Mary, N. Y., Kelleher, W. J., Schwarting, A. E.: Lloydia 28, 218 (1965). Mattern, H., Brillinger, G. U., Pape, H.: Arch. Mikrobiol. 88, 37 (1973). Matteo, C. C., Glade, M., Tanaka, A., Piret, J., Demain, A. L.: Biotechnol. Bioeng. 17, 129 (1975). Mertz, F. P., Doolin, L. E.: Can. J. Microbiot. 19, 263 (t973). Miller, A. L., Walker, J. B.: J. Bacteriol. 99, 401 (1969). Miller, A. L., Walker, J. B.: J. Bacteriol. 104, 8 (1970). Nimi, O., lto, G., Sueda, S., Nomi, R.: Agr. Biol. Chem. 35,848 (1971). Nimi, O., lto, G., Ohata, Y., Funayama, S., Nomi, R.: Agr. Biol. Chem. 35,856 (1971). Nimi, O., Norimoto, Y., Nomi, R.: Agr. Biol. Chem. 35, 1819 (1971). Nomi, R., Nimi, O., Kado, T.: Agr. Biol. Chem. 32, 1256 (1968). Nomi, R., Nimi, O.: Agr. Biol. Chem. 33, 1459 (1969). Oki, T., Yoshimoto, A., Sato, S., Takamatsu, A.: Biochim. Biophys. Acta, 410, 262 (1975). Oki, T., Yoshimoto, A., Ogasawara, T., Sato, S., Takamatsu, A.: Arch. Mikrobiol. 107, 183 (1976). Pass, L., Raczynska-Bojanowska, K.: Acta Biochim. Pot. 15, 355 (1968). Perlman, D., Wagman, G. H.: J. Bacteriol. 63, 253 (1951). Pifia, M. Z., Brunner, A., Chagoya, V., Pifia, E.: Euro. J. Biochem. 30, 463 (1972). Popova, L. A., Zavileyskaja, G. F., Dyghern, N. T., Pestereva, G. D.: Antibiotiki 6, 34 (1961). Prokofieva-Belgovskaya, A., Popova, L.: J. Gen. Microbiol. 20, 462 (1959). Pruess, D. L., Johnson, M. J.: J. Bacteriol. 94, 1502 (1967). Purich, A. L., From, H. J.: J. Biol. Chem. 247,249 (1972). Rafalski, A., Raczynska-Bojanowska, K.: Acta Biochim. Pol., 19, 71 (1972). Rafalski, A., Raczynska-Bojanowska, K.: Aeta Microbiol. Pol. 5, 87 (1973). Rao, K. K., Gupta, A. R.: Naturwissensehaften 62, 394 (1975). Rhaese, H. J., Groscourth, R.: FEBS Letters 44, 87 (1974). Rhaese, H. J., DichtelmuUer, H., Grade, R.: Eur. J. Biochem. 56,385 (1975). Rickemberg, H. ~L: Ann. Rev. Microbiol. 28, 353 (1974). Robers, J. E., Robertson, L. W., Hornemann, K. M., Jindra, A., Floss, H. G.: J. Bacteriol. 112, 791 (1972). Rogers, T. O., Birnbaum, J.: Antimicrob. Agents Chemother. 5, 121-132 (1974). Sankaran, L., Pogell, B. M.: Nature (New Biol) 245,257 (1973). Schaeffer, P.: Bacteriol. Rev. 33, 48 (1969). Shen, S. C., Soong, H. Y., Chen, J. P., Hung, M. M., Yin, H. C.: Acta Biol. Exp. Sinica 5, 249 (1956). Shen, S. C., Chen, J. P., Hung, M. M.: Acta Physiol. Sinica 21,302 (1957). Shen, S. C., Chen, J. P., Koo, T. A.: Scientia Sinica 8, 733 (1959). Silaeva, S. A., Glazer, V. M., Shestakov, S. V., Prokofiev, M.A.: Biokhimiya 30, 947 (1965). Sikyta, B., Slezak, J., Herold, M.: In: Continuous Cultivation of Microorganisms. I. Malek (ed.), Prague: Academia 1964, p. 173. Sokawa, Y., Sokawa, J., Kaziro, Y.: Nature (New Biol.) 234, 7 (1971). Sokawa, Y., Sokawa, J., Kaziro, Y.: Cell 5, 69 (1975). Suchkova, L. A., Toropova, E. G., Egorov, N. S., Tyrsin, Y. A.: Antibiotiki 17, 810 (1972). Suzuki, T., Honda, H., Katsumata, R.: Agr. Biol. Chem. 36, 2223 (1972).
Control of Antibiotic Synthesis by Phosphate 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155.
127
Taber, W. A.: Lloydia 30, 39 (1967). Tereshin, I. M.: Polyene Antibiotics. Present and Future. Tokyo: University of Tokyo Press 1976. Tomino, S., Yamada, M., Itom, H., Kurahashi, K.: Biochemistry 6, 2552 (1967). Tomita, F., Nakano, H., Suzuki, T.: Agr. Biol. Chem. 38, 1673 (1974). Torbochkina, L. I., Dormidochina, T. A., Zaitzeva, L. P.: Mikrobiologiya 33, 162 (1964). Torbochkina, L. I., Dormidochina, T. A.: Mikrobiologiya 33,325 (1964). Toropova, E. G., Egorov, N. S., Egorova, O. A., Suchkova, L. A.: Antibiotiki 18, 3 (1973). Toropova, E. G., Egorov, N. S., Suchkova, L. A.: Antibiotiki 18, 587 (1973). Uladova, T. P., Federova, R. I.: Mikrobiologiya 34, 631 (1965). Umnov, A. M., Steblyak, A. G., Umnova, N. S., Mansurova, S. E., Kulayev, I. S.: MikrobioIogiya 44, 113 (1975). Vandamme, E., Demain, A. L.: Dev. Ind. Microbiol. 17, 51 (1976). Vanek, Z., Hostalek, Z., Blumauerova, M., Mikulik, K., Podojil, M., Behal, V., Jechova, V.: Pure Appl. Chem. 34, 463 (1973). Vorisek, J., Powell, A. J., Vanek, Z.: Folia Microbiol. 14, 398 (1969). Walker, J. B., Hnilica, V. S.: Biochim. Biophys. Acta 89, 473 (1964). Walker, J. B., Walker, M. S.: Biochemistry 6, 3821 (1967). Walker, M. S., Walker, J. B.: J. Biol. Chem. 245, 6683 (1970). Walker, M. S., Walker, J. B.: J. Biol. Chem. 246, 7034 (1971). Walker, J. B., Skorvaga, M.: J. Biol. Chem. 248, 2435 (1973). Walker, J. B., Skorvaga, M.: J. Biol. Chem. 248, 2441 (1973). Wang, C. H., Stern, I., Gilmour, C. M., Klungsoyr, S., Reed, D. J., Biaty, J. J., Christensen, B. E., Cheldelin, V. H.: J. Bacteriol. 76, 208 (1958). Weinberg, E. D.: Adv. Microbial Physiol. 4, 1 (1970). Weinberg, E. D.: Dev. Ind. Microbiol. 15, 70 (1974). Yaginomoto, M., Terui, G.: J. Ferment. Technol. 49, 611 (1971). Zygmunt, W. A.: Appl. Microbiol. 12, 195 (1964).