International Review of Cytology: A Survey of Cell Biology, Volume 41

INTERNATIONAL REVIEW OF CYTOLOGY VOLUME41 ADVISORY EDITORS H. W. BEAMS HOWARD A. BERN W. BERNHARD GARY G. BORISY ROB...

Author: Geoffrey H Bourne | James Frederick Danielli

12 downloads 928 Views 17MB 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

DOWNLOAD PDF

INTERNATIONAL

REVIEW OF CYTOLOGY VOLUME41

ADVISORY EDITORS H. W. BEAMS HOWARD A. BERN W. BERNHARD GARY G. BORISY ROBERT W. BRIGGS R. COUTEAUX

B. DAVIS N. B. EVERETT DON FAWCETT MICHAEL FELDMAN WINFRID KRONE K. KUROSUMI MARIAN0 LA VIA

GIUSEPPE MILLONIG MONTROSE J. MOSES ANDREAS OKSCHE VLADIMIR R. PANTIC LIONEL I. REBHUN JEAN PAUL REVEL WILFRED STEIN ELTON STUBBLEFIELD H. SWIFT J. B. THOMAS TADASHI UTAKOJI ROY WIDDUS A. L. YUDIN

INTERNATIONAL

Review of Cytology EDITED BY

G. H. BOURNE

J. F. DANIELLI

Yerkes Regional Primate Research Center Emo y University Atlanta, Georgia

Center f o r Theoretical Biology State University of New York a t Buflalo Buffalo, New York

ASSISTANT EDITOR K. W. JEON Department of Zoology University of Tennessee Knoxville, Tennessee

VOLUME41

ACADEMIC PRESS New York San Francisco London 1975 A Subsidiary of Harcourt Brace ]ovanovich, Publishers

COPYRIGHT 0 1975. BY ACADEMIC PRESS,INC. ALL RIGHTS RESERVED. N O PART OF THIS PUBLICATION MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM OR BY ANY MEANS, ELECTRONIC OR MECHANICAL, INCLUDING PHOTOCOPY, RECORDING, OR ANY INFORMATION STORAGE AND RETRIEVAL SYSTEM, WITHOUT PERMISSION IN WRITING FROM THE PUBLISHER.

ACADEMIC PRESS, INC.

111 Fifth Avenue,New York, New York 10003

United Kingdom Edition published by ACADEMIC PRESS, INC. (LONDON) LTD. 24/28 Oval Road, London N W l

LIBRARY OF CONGRESS CATALOG CARD NUMBER: 52-5203 ISBN 0-12-364341 -4 PRINTED IN T H E UNITED STATES OF AMERICA

Contents LIST OF CONTRIBUTORS .

.

.

.

.

.

.

.

.

.

.

ix

The Attachment of the Bacterial Chromosome to the Cell Membrane PAUL

J . LEIROWITZAND MOSELIO SCHAECHTEH

. . . . . . . . . . . . I . Introduction . . . . . . . . . I1 . Morphological Considerations . . . . . 111. The Mode of Segregation of the Cell Membrane . . . . . IV . Polarization in the Segregation of Chromosomes . . . . . V . Methodological Considerations of Cell Fractionation VI . Is the Chromosome Attached to a Special Region of the Membrane? . . VII . Is the Origin of DNA Replication Attached to the Membrane? . . . . VIII Does DNA Replication Take Place on the Membrane? . . . . . . IX . Is the Chromosome Attached at Many Sites? . . X The Nature of the Binding of the Chromosome to the Membrane . XI The Role of the Membrane in Maintaining the Compactness of the . . . . . . . . . . . . Nucleoid . XI1 . Conclusions . . . . . . . . . . . . . References . . . . . . . . . . . . .

. . .

1 3 13 15 17 19 20 20 22 23

24 25 26

Regulation of the Lactose Operon in Escherichia coli by CAMP G . CARPENTERAND B . H . SELLS

. . . . . . I . Introduction. I1 CAMP Formation and Catabolite Repression 111. In Vioo Evidence of Site of cAMP Action . . . . IV cAMP Action in Vitro . V Additional Aspects of Lac Operon Regulation References . . . . . . .

. . .

. . . . . .

. . . . . .

. . . . . .

. . .

.

. .

. . .

. . .

.

.

.

.

. .

. .

29 36 43 46 54 55

Regulation of Microtubules in Tetrahymena

.

NORMAN E WILLIAMS

I . Introduction.

.

.

.

.

.

.

I1. Regulatory Patterns and the Cell Cycle .

.

. .

. 111. Microtubule Stability and Regression IV. The Dynamic Nature of Formed Microtubules . . V . Control of Microtubule Formation . . .

VI * Epilog: The Cell Cycle Revisited . . . . References . V

. .

. .

.

. .

.

. .

.

. .

. .

. .

.

.

.

.

.

.

.

.

.

.

. .

. .

. .

. .

. .

59 60 68 74 77 83 84

vi

CONTENTS

Cellular Receptors and Mechanisms of Action of Steroid Hormones SHUTSUNG LIAO

I . Introduction.

.

.

.

.

I1. Steroid-Binding Proteins in Blood

. V. VI .

.

. Steroid Receptors in Target Tissues .

.

. .

.

. .

.

. .

111 IV. Cytoplasmic-Nuclear Interaction of Steroid Receptors

Gene Expression and Steroid Receptor Concluding Remarks . . . . References . . . . . .

. . .

. . .

. . .

.

. . . . . .

.

. . . . . .

.

. . . . . .

.

87 90 92 127 139 151 157

173 177 205 234 235

. . . . . .

A Cell Culture Approach to the Study of Anterior Pituitary Cells A . TIXIEH.VIDAL. D . GOURDJI. AND c . TOUGAHD I . Introduction .

.

.

.

.

.

.

.

.

.

I1. Characteristics of Anterior Pituitary Cells Grown in Vitro

.

111

Reactivity to Specific Regulating Agents

IV. Conclusion . References

.

.

.

.

.

.

.

.

.

.

.

. . .

. . .

. . .

. . .

.

.

.

.

.

.

. . .

. . .

. . .

Immunohistochemical Demonstration of Neurophysin in the Hypothalamoneurohypophysial System

.

W B . WATKINS

. . . . . . . . . . . . I . Introduction . I1 . Relationship between Neurosecretory Material and Neurophysin . . . . . . . . . 111. Methods of Extraction of Neurophysin . . . . . . . IV. Purification of Neurophysin Antigens V. Production of Antibodies against Neurophysin . . . . . . VI . General Considerations of Antibody Production and Detection . . . . . . . . VII . Immunohistochemical Techniques . VIII . Demonstration of Neurophysin in the Hypothalamoneurohypophysial System Using Cross-Species Reactive Antineurophysin . . . . Ix. Use of Species-Specific Antisera for the Demonstration of Neurophysin . . . . . . . . . . . . . X . Conclusions . References . . . . . . . . . . . . .

241 243 244 246 250 255 256 260 279 280 28 1

vii

CONTENTS

The Visual System of the Horseshoe Crab

Limulus polyphemus WOLF H . FAHRENBACH

I. I1 111. IV V VI VII VIII . IX

. . . . . . X. XI .

Introduction. . . . . Dioptric Structures . . Pigment Cells . . . Neuroglial Cells . . Receptor Cells . . . Basal Lamina and Hemoc:eel . . Axons and Plexus Optic Nerves . . . Optic Centers . . . Miscellaneous Aspects . Vision and Behavior . References . . . .

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

. .

. .

. .

. .

. .

. .

.

.

.

.

.

.

.

.

SUBJECT INDEX . . . . . CONTENTS OF PREVIOUS VOLUMES .

. .

. .

. .

. .

. .

. .

. .

.

.

.

.

.

.

.

.

.

.

.

.

. . .

.

.

.

. . .

.

.

.

. . .

.

. . .

.

. . .

.

. . .

.

. . .

.

.

.

.

. . . .

.

.

.

.

. . . .

.

. .

.

. .

285 287 296 304 306 318 321 333 335 338 341 344

351 354

This Page Intentionally Left Blank

List of Contributors Numbers in parentheses indicate the pages on which the authors’ contributions begin.

G. CARPENTER* (29), Laboratories of Molecular Biology, Faculty of Medicine, Memorial University of Newfoundland, S t . John’s, Newfoundland, Canada WOLF H. FAHRENBACH (285), Laboratory of Electron Microscopy, Oregon Regional Primate Research Center, Beaverton, Oregon

D. GOURDJI(173), Groupe de Neuroendocrinologie Cellulaire, Laboratoire de Physiologie Cellulaire, Collbge d e France, Paris, France PAULJ. LEIBOWITZ (l),Department of Molecular Biology and Microbiology, Tufts University School of Medicine, Boston, Massachusetts SHUTSUNG LIAO (87), The Ben May Laboratory for Cancer Research and the Department of Biochemistry, T h e University of Chicago, Chicago, I11inois

MOSELIO SCHAECHTER(l),Department of Molecular Biology and Microbiology, Tufts University School of Medicine, Boston, Massachusetts B. H. SELLS(29), Laboratories of Molecular Biology, Faculty of Medicine, Memorial University of Newfoundland, S t . John’s, Newfoundland, Canada

C. TOUGARD (173), Groupe d e Neuroendocrinologie Cellulaire, Laboratoire de Physiologie Cellulaire, Collhge de France, Paris, France A. TIXIER-VIDAL (173), Groupe de Neuroendocrinologie Cellulaire,

Laboratoire de Physiologie Cellulaire, Collhge de France, Paris, France W. B. WATKINS (241), Postgraduate School of Obstetrics and Gynaecology, University of Auckland, Auckland, New Zealand NORMAN E. WILLIAMS (59), Department

of Zoology, University of Zowa,

Iowa City, Iowa

’ Present address: Department of Biochemistry, Vanderbilt University, Nashville, Tennessee. ix

This Page Intentionally Left Blank

The Attachment of the Bacterial Chromosome to the Cell Membrane PAUL J. LEIBOWITZAND MOSELIO SCHAECHTER Department of Molecular Biology and Microbiology, TUBSUniversity School of Medicine, Boston, Massachusetts

. . . . . . . The Nucleoid . . . . . The Membrane . . . .

I. Introduction

11. Morphological Considerations

111.

IV. V. VI. VII. VIII. IX. X. XI.

XII.

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

A. B. C. Morphological Evidence for the Association between Nucleoids and the Membrane . . . . . . . The Mode of Segregation of the Cell Membrane. . . . Polarization in the Segregation of Chromosomes . . , Methodological Considerations of Cell Fractionation . . Is the Chromosome Attached to a Special Region of the Membrane? . . . . . . . . . . . . Is the Origin of DNA Replication Attached to the Membrane? . . . . . . . . . . . . . Does DNA Replication Take Place on the Membrane?. . Is the Chromosome Attached at Many Sites? . . . . The Nature of the Binding of the Chromosome to the Membrane . . . . . . . . . . . . . The Role of the Membrane in Maintaining the Compactness of the Nucleoid . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . References . . . . . . . . . . . . .

5 13 15 17

19 20 20 22 23 24 25 26

I. Introduction

Prokaryotic cells are small, do not contain organelles limited by independent membrane systems and, instead of a complex nucleus, possess a pleomorphic central region of condensed DNA, the nucleoid. Nucleoids do not divide by mitosis and, at least in Escherichia coli, consist of single chromosomes. The DNA is in the form of a circular duplex molecule, and its replication takes place in both directions from a distinct starting point. Bacteria have the ability to respond to changes in environmental conditions by grossly altering their size and macromolecular composition. Thus, as a response to changes in the kind of nutrients provided, growing bacteria may vary in size or in RNA content by a factor of 10 or more. They change from one physiological state to another in a remarkably efficient and rapid manner. A review of 1

2

P. J. LEIBOWITZ AND M. SCHAECHTER

these attributes appears in a book by Maalfle and Kjeldgaard (1966). Bacteria are efficient and structurally simple; consequently they make multiple uses of cellular structures. In this vein, the synthesis and regulation of several macromolecules have been shown to be related to the behavior of the cell membrane. The subject of this article is one of these relationships, the connection of the bacterial chromosome with the cell membrane. Concern for this subject originated with a proposal by Jacob, Brenner, and Cuzin for the control of DNA replication in bacteria. They formally termed the unit of DNA replication the replicon, and proposed that initiation of replication is controlled by diffusible gene products (Jacob et al., 1963; Jacob and Brenner, 1963): A structural gene produces an initiator which acts upon a region of the chromosome at a specific site, the origin. Replication begins at the origin and proceeds linearly until the entire chromosome has been duplicated. Included in the model is the proposition that the chromosome is attached to the bacterial membrane. The model suggests that the DNA-synthesizing complex is fixed to the bacterial membrane and that the DNA moves through this complex. The membrane is thought to provide a mechanism for the segregation of the daughter replicons by growth of the cell surface between their sites of attachment. The bacterial membrane would thereby perform the function of the mitotic apparatus of higher organisms, as well as being the site of DNA synthesis. This model predicts the existence of membrane componer,ts that recognize specific sites on the chromosome. The replicon model stimulated a search for the association between the bacterial chromosome and the cell membrane. In this article we present morphological, genetic, and biochemical evidence for this association. We attempt to provide tentative state-of-the-art answers to the following questions: (1) Is DNA attached to the membrane? (2) If so, at how many sites? (3) If there are several sites, are they alike in function? (4) Is attachment possible along any region of the genome? ( 5 ) Is the membrane unique at the site of attachment? Answers to these questions are tentative, because the necessary methodology is in an early state of development. In very few cases have findings been confirmed by unrelated techniques. Even more difficult are the following questions: (6) Do all attachment points exist at the same time? (7)At what time in the cell cycle are attachment points born? (8) Do attachment points remain at their site of birth? Many researchers have interpreted their data in terms of a connection between the bacterial chromosome and the cytoplasmic mem-

BACTERIAL CHROMOSOME AND CELL MEMBRANE

3

brane. We feel that much of this work is only tangentially related to the subject of this article. For this reason we have selected evidence that comes closest to providing answers to the questions listed above, and which we feel deals directly with the issue of whether or not attachment exists. In addition, we have omitted work that is relevant to this field but which appears to lie beyond the conceptual framework of this article. The special topic of the attachment of bacteriophage DNA to the bacterial cell membrane has been reviewed recently (Siege1 and Schaechter, 1973). 11. Morphological Considerations

A. THE NUCLEOID

Perhaps the most convincing proof for the existence of nucleoids comes from observations of living E. coli with the phase-contrast microscope. Nucleoids can be seen when cells are grown in media of high refractive index. In time-lapse motion pictures nucleoids are first seen to change in conformation as they divide, and then to segregate into daughter cells prior to the completion of cell division (Adler et al., 1969). There has never been any demonstration of a membrane separating nucleoids from the cytoplasm, nor is there evidence of any of the elements of a mitotic apparatus. The DNA of E. coli consists of a single circular duplex molecule about 1100 pm in length (Cairns, 1963). Since the apparent volume of the nucleoid of this cell is about 0.1 pm3, it follows that the chromosome must be folded on itself and exist in a phase state quite unlike that of DNA in solution. Although this represents a very high concentration of DNA, the nucleoid is considerably less dense than its surrounding cytoplasm. The degree of condensation of the nucleoid observed in the electron microscope varies with the method of fixation. Freeze-etched preparations of unfixed bacteria reveal no clear distinction between the nucleoid and cytoplasm, but typical nucleoids are seen occasionally if cells are fixed with osmium tetroxide (Nanninga, 1968). The Ryter-Kellenberger (R-K) procedure is currently the most widely used fixation method for ultrathin sectioning of bacteria. It employs osmium tetroxide for both prefixation and fixation. With this fixation nucleoids are most frequently seen in the central portion of the cell, and they show bundles of fibers with dimensions similar to those of DNA (Kellenberger et al., 1958). Prefixation in glutaraldehyde followed by fixation with osmium tetroxide (G-0 fixa-

4

P. J. LEIBOWITZ AND M. SCHAECHTER

tion) causes the nucleoplasm to appear in a dispersed configuration rather than as a centrally located body (Margaretten et al., 1966; McCandless et al., 1968). Recent work by M. L. Higgins and L. Daneo-Moore (personal communication) suggests that degradation of RNA, which takes place during R-K but not G - 0 fixation, may lead to further condensation of the nucleoids. Despite the uncertainties introduced by these studies, R-K fixation results in sections which conform to what is expected from studies on living bacteria with the phase microscope. B. THE MEMBRANE The cytoplasm of bacteria is bounded by a trilaminar membrane, about 8-10 nm thick. Outside this membrane is the cell wall and outside it, in gram negative cells, is a membranelike outer layer. The inner membrane has intracytoplasmic involutions termed mesosomes, which vary in complexity among taxonomic groups (FitzJames, 1960; van Iterson, 1961; Glauert et al., 1961; Glauert, 1962).A comprehensive review of these structures has recently appeared (Reusch and Berger, 1973). The mesosomes of gram-positive bacteria appear as extensions of the cytoplasmic membrane forming saclike structures (outer mesosomal membranes) filled with vesicles, tubules, and/or lamellae (internal mesosomal membranes) (e.g., Bacilli. Ryter and Jacob, 1966, van Iterson, 1961, 1965, Fitz-James, 1960; Holt and Leadbetter, 1969; Listeria monocytogenes: Edwards and Stevens, 1963; Mycobacteria: Imaeda and Ogura, 1963; Streptomyces: Glauert, 1962). These internal structures are considered in turn to be invaginations of the sac, or the outer mesosomal membrane (Fitz-James, 1960; Ryter and Jacob, 1966). Mesosomes are also found in gram-negative bacteria (E. coli: Kaye and Chapman, 1963; Steed and Murray, 1966; Pseudomonas aemginosa: H o h a n n et al., 1973; Spirillum serpens: Steed and Murray, 1966; Caulobacter: Stove Poindexter and Cohen-Bazire, 1964). Ultrathin sections usually reveal that these mesosomes are uncomplicated structures, most often containing lamellae which apparently result from delicate foldings of the plasma membrane. Mesosomes in gram-negative bacteria are probably devoid of tubules. As with gram-positive bacteria, there is variation in structure among taxonomic groups. There is disagreement on the true morphology of mesosomes and on their number and location in the cell (Remsen, 1968; Nanninga, 1968; Highton, 1969, 1970a,b; Burdett and Rogers, 1970; Rogers, 1970). It must be emphasized that the morphology of the mesosome

BACTERIAL CHROMOSOME AND CELL MEMBRANE

5

is altered markedly by the conditions of fixation (e.g., Burdett and Rogers, 1970). In ultrathin sections mesosomes are seen to be touching the nucleoids of dividing cells, and to be continuous with division septa. There is evidence to implicate these structures in DNA replication (Higgins and Daneo-Moore, 1972), in the segregation of chromosomes (Ryter and Jacob, 1963), in the location of membrane and cross-wall synthesis and prespore septation (Ellar et al., 1967; Steed and Murray, 1966; Chapman and Hillier, 1953; Fitz-James, 1960, 1967; Freese, 1973), in subcellular degradative activities (lysosomal functions) (Reusch and Berger, 1972), and in oxidative function (van Iterson and Leene, 1964; Ferrandes et al., 1966). The morphological development of mesosomes was followed in synchromously dividing Bacillus megateriiurn by Ellar et al. (1967). Mesosomes develop by an initial concentric infolding of the cytoplasmic membrane and eventually assume a saclike shape. Cross wall formation begins at the base of these mesosomes which are located at the center of the cells. This implicates them in the initiation of cross wall synthesis. Later, the mesosome is seen on both sides of the developing cross wall, which suggests that it is also involved in the synthesis of cross walls. These central mesosomes are often associated with nucleoids, as are other mesosomes located at the poles. From these morphological considerations it seems likely that mesosomes are responsible for thickening of the cell wall prior to cell separation, and for initiation and synthesis of the cross wall. This has not yet been borne out by fractionation studies, since mesosomes have been found not to be particularly rich in enzymes and precursors involved in membrane or wall synthesis (Patch and Landman, 1971; Reusch and Berger, 1972). However, Nanninga (1968) showed differences in the freeze-etched surface structure of mesosomes and cytoplasmic membranes and concluded that mesosomes may in fact differ from the rest of the cytoplasmic membrane. This subject has been reviewed by Reusch and Berger (1973).

c.

MORPHOLOGICALEVIDENCEFOR THE ASSOCIATION BETWEEN NUCLEOIDSAND THE MEMBRANE In ultrathin sections the nucleoid is located in a central region of the cell and is not in obvious contact with the peripheral membrane. For this reason the morphological association between them escaped detection for many years. Upon closer examination the nuclear regions and mesosomes of both gram-positive and gram-negative

6

P. J. LEIBOWITZ AND M. SCHAECHTER

FIG. 1. Ultrathin sections of growing B . subtilis showing the association of mesosomes (M)with nucleoids (N). Fixation by the R-K method. (From Jacob et al., 1966, reproduced with permission from the publishers, The Royal Society, and the authors.)

BACTERIAL CHROMOSOME AND CELL MEMBRANE

7

FIG. 2. Ultrathin section of a mesosome in B. subtilis after prefixation with the G O method. x90,OOO. (From Ryter, 1968, reproduced with permission from the publishers, The American Society for Microbiology, and the author.)

FIG. 3. Mesosomes (M) of E . coli, which appear as delicate folds of membrane in contact with the bacterial nucleoid. X85,600. (From Ryter and Jacob, 1966, reproduced with permission from the publishers, Masson et Cie., Editeurs, and the authors.)

8

P. J. LEIBOWITZ AND M. SCHAECHTER

cells can nearly always be shown to be touching (see Figs. 1-3). They usually have considerable surface contact and often penetrate one another (van Iterson, 1961; Ryter and Jacob, 1964, 1966; Ellar et al., 1967; Pontefract et al., 1969; Remsen, 1968; Hoffmann et al., 1973). The contact is less dramatic in most gram-negative cells, because their mesosomes are smaller. In fact, in a mutant of E. coli which forms extensive intracytoplasmic membranes, the contact of the DNA with these membranes is readily evident (Altenburg and Suit, 1970; Altenberg et al., 1970). Offhand it should not be surprising to find that the pleomorphic nuclear region occasionally makes contact with the mesosome. If the association were fortuitous, however, one would expect great disparity among individual cells. Ryter and Jacob (1964) determined that the nucleoid and mesosomes of Bacillus subtilis were visibly linked in each of 20 serially sectioned cells which included all stages of the cell cycle. The nucleoid was associated with either one or two mesosomes, depending on the stage of growth (Ryter, 1968). Smaller nucleoids appeared to be attached to one mesosome, while larger ones were often attached to two. Consequently, it was possible to arrange the three-dimensional constructs in an order thought to reflect the cell cycle, Initially, the two nucleoids in each cell are seen attached to two separate mesosomes; as the chromosome replicates, mesosomes seem to split in two, each maintaining contact with one of the two newly formed nucleoids; segregation of nucleoids is accomplished by the growth of the membrane between them; after this segregation process begins, the cell septum starts to form. There is considerable disagreement with this model, at least in its simplest form. Several investigators have found that mesosomes do not arise by division but are formed de n o w at the site of septum formation. This was reported for B . rnegaterium by Ellar et al. (1967), Streptococcus faecalis by Higgins and Shockman (1970a, 1971), and E. coli by Pontefract et al. (1969). There is evidence that the nucleoid is always associated both with a polar mesosome and with the newly synthesized, septa1 mesosome (Ellar et al., 1967; Pontefract et al., 1969). Mesosomes that form at septa become polar mesosomes in daughter cells. Since two polar mesosomes within one cell arise during different cell division cycles, it should be possible in future work to distinguish between an old segregation apparatus and a new one. There are also indications that nucleoids may not always be associated with mesosomes. For instance, Highton (1970b)found that

BACTERIAL CHROMOSOME AND CELL MEMBRANE

9

multinucleated B . subtilis cells contain fewer mesosomes than nucleoids. There is evidence that in s.faecalis mesosomes may not participate in nucleoid segregation. This spherical bacterium has an equatorial band on the external surface of the wall which marks the site of new cross wall synthesis and the boundary between old and new wall. Upon initiation of wall synthesis these bands split, double in number, and move to a subequatorial position. Each daughter cell has an equatorial band from the preceding generation which marks the initiation site of wall growth for that generation. Mesosomes are usually seen just beneath an equatorial wall band on the cell surface, and are attached to the base of the septal membrane by a membranous stalk (Higgins and Shockman, 1970b, 1971; see Fig. 4). Mesosomes located near the septum are most often seen penetrating the nuclear mass. Mesosome formation precedes cross wall formation. The mesosome appears to maintain direct contact with the septum only during the early part of the cell cycle. The septal connection is lost prior to the completion of the cross wall at the time the nucleoid appears to have divided into two masses. Two new mesosomes are now found beneath wall bands in the developing daughter cell (Higgins and Shockman, 1971; Higgins and DaneoMoore, 1972). The effect of selective inhibition of DNA, RNA, and protein synthesis on the development of mesosomes was studied in S. faecalis by Higgins and Daneo-Moore (1972). It has been shown that the cross-sectional area of mesosomes increases rapidly during amino acid starvation (Higgins and Shockman, 1970b). These authors proposed that the increase in mesosome size might be related to continued DNA synthesis since, during amino acid starvation, RNA synthesis is shut off and the rate of protein synthesis decreases (Ziegler and Daneo-Moore, 1971).They suggest that the termination of DNA replication might result in activation of the regions of the envelope involved in segregation to form a site for the formation of a new mesosome. They suggest that mesosomes in this organism are necessary for the initiation of the cross wall and for DNA replication, but not for cross wall formation or nuclear segregation. Segregation would take place through direct attachment to the cytoplasmic membrane. It is not known if these discrepancies in the behavior of mesosomes are due to differences among various species of bacteria. It is likely that they are due to a combination of many factors, including fixation artifacts and differences in the physiological state of the

10

P. J. LEIBOWITZ AND M. SCHAECHTER

Wall Band

\

62 New Wall Synthesis

FIG.4. Diagrammatic representation of the cell division cycle for Streptococcus faecalis. The model proposes that linear wall elongation is a unitary process which results from wall synthetic activity at the leading edges of the nascent cross wall. The diplococcus in A is in the process of growing new wall at its cross wall and segregating its nuclear material to the two nascent daughter cocci. In rapidly growing exponential phase cultures before completion of the central cross wall, new sites of wall elongation are established at the equators of each of the daughter cells at the junction of old, polar wall (stippled) and new equatorial wall beneath a band of wall material that encircles the equator (B).Beneath each band a mesosome is formed while the nucleoids separate and the mesosome at the central site is lost. The mesosome appears to be attached to the plasma membrane by a thin membranous stalk (BI). Invagination of the septa1 membrane appears to be accompanied by centripetal cross wall penetration (BZ).A notch is then formed at the base of the nascent cross wall which creates two new wall bands (B3). Wall elongation at the base of the cross wall pushes newly made wall outward. At the base of the cross wall, the new wall peels apart into peripheral wall, pushing the wall bands apart (B4). When sufficient new wall is made so that the wall bands are pushed to a subequatorial position (e.g., from C to A to B) a new cross wall cycle is initiated. Meanwhile the initial cross wall centripetally penetrates into

BACTERIAL CHROMOSOME AND CELL MEMBRANE

11

cells. We feel that precise knowledge of the number and location of mesosomes awaits a detailed analysis of synchronously dividing cells growing at different rates. Linkage of the nucleoid to mesosomes has been reported to persist throughout the first stages of sporulation (Ryter and Jacob, 1964). In the first stage the contact appears to be mediated through one of the polar mesosomes which eventually participates in the development of the spore membrane. Later, the nucleoid migrates to a peripheral position in the spore cytoplasm, the mesosome disappears, and vesicles suggestive of mesosome tubules are found along the spore membrane. In the last stage of sporulation, the nucleoid is connected directly to the spore membrane. It is not clear if this represents the initial contact between the nucleoid and mesosome seen in the first stage of sporulation, or if a new contact point is formed. The sequence of events in nuclear division has also been studied in spore germination (Ryter, 1967). Early in germination of B. subtilis spores, the nucleoid assumes a central position, becomes an axial filament, and is connected to the spore membrane by a huge mesosome. Later, the number and size of mesosomes vary, and they are not always in contact with the nucleoid. Nonetheless, the nucleoid is linked to the membrane at two sites, either through mesosomes, by direct attachment to the membrane, or both. The distance between the attachment points increases as the cell elongates, but the distance from each attachment point to the pole of the cell seems to remain the same. This suggests that the membrane grows by the deposition of new material at the equator of the cell. Occasional sections reveal the presence of small nonmesosome structures in the membrane to which the nuclear fibrils are attached (Ryter, 1967). Mesosomes are evaginated when cells are plasmolyzed in hypertonic medium or when spheroplasts (wall-less or wall-deficient cells) are prepared. In such cases the nucleoid is found at the periphery of the cell, as if it had been dragged toward the cell surface by its attachment to the membrane (Ryter and Landman, 1964, 1967; Ryter and Jacob, 1964, 1966). One would expect that upon extrusion of the mesosome the chromosome would be linked to the portion of the membrane that was the cytoplasmic surface of the mesosome. In the cell, dividing it into two daughter cocci. At all times the body of the mesosome appears to be associated with the nucleoid. Doubling of the number of mesosomes seems to precede completion of the cross wall by a significant interval. Nucleoid shapes and the position of mesosomes are based on projections of reconstructions of serially sectioned cells. (From Higgins and Shockman, 1971, with permission of CRC Press, Inc., and by courtesy of the authors.)

12

P. J. LEIBOWITZ AND M. SCHAECHTER

BACTERIAL CHROMOSOME AND CELL MEMBRANE

13

some sections of spheroplasts, the cytoplasmic membrane is markedly invaginated toward the nucleoid. In B . subtilis spheroplasts remnants of mesosome tubules are seen attached to the outer surface of the membrane in contact with nuclear fibrils (see Fig. 5). This suggests that the association is not fortuitous. Spheroplasts of B . subtilis do not contain mesosomes and cannot divide unless transferred to special reversion media. Here the protoplasm apparently divides without the intervention of mesosomes (Ryter and Landman, 1964). Direct contact of the nucleoid with the membrane persists throughout reversion, but is not readily observed in the newly reverted bacillary form. During reversion mesosomes are morphologically atypical. In these cells contact with the membrane appears to be mediated not through mesosomes, but through a small vesicular attachment apparatus similar to that found in germinating spores by Ryter (1967), Ryter and Landman (1967), and Landman et al. (1968).These investigators proposed that mesosomes are not required for reversion of spheroplasts and may not be essential for any cellular function, since their absence has no noticeable effect on DNA replication, chromosome segregation, or cell division. They point out the need for a thorough investigation of the small attachment apparatus often seen connecting the nucleoid to the membrane in the absence of obvious mesosomes. 111. The Mode of Segregation of the Cell Membrane

We now consider how the membrane of bacteria “grows,” that is, where on its surface new material is deposited. It is important to consider the various models of membrane growth because each makes different predictions on the role of the membrane in chromosome segregation. If the cell membrane were synthesized solely at the equator of the cell, its lateral displacement would lead directly to segregation of chromosomes attached to it. Two newly replicated chromosomes would separate from each other by the intercalation of new membrane material between their points of attachment. If, however, the membrane is synthesized at many sites on its surface, its role in chromosome segregation would be either passive or more complicated. FIG.5. Protoplasts of B . subtilis showing remnants of mesosomal tubules (M) still attached to the outer surface of membrane. Nuclear fibrils are in contact with this portion of the membrane, believed to be the outer membrane of the mesosome prior to extrusion. Top, X80,OOO; bottom, X 120,000. (From Ryter and Jacob, 1966, reproduced with permission from the puhlishers,Masson et Cie., Editeurs, and the authors.)

14

P. J. LEIBOWITZ AND M. SCHAECHTER

Only a few experiments have been reported that deal with this point, and they are not conclusive. There is only one report of a direct attempt to determine the site of synthesis by pulse-labeling cells with specific membrane precursors and determining the location of newly synthesized material by radioautography. These investigators found that newly made lipids were not localized at specific membrane sites (Mindich and Dales, 1972). Most of the relevant studies dealt with this question by determining the pattern of segregation of the membrane. They are based on the following reasoning. If there is one site of membrane synthesis and it is at the equator of the cell, then new and old membrane may be distributed among daughter cells in a semiconseruative pattern. Conversely, if there are many sites of synthesis, old and new membrane will be dispersed among progeny cells. Several investigators have used the fact that bacterial flagella are attached to the membrane to follow the pattern of membrane segregation. Thus Ryter (1971) used a mutant of B. subtilis which synthesizes flagella at 37°C but not at 46°C. When this mutant was shifted from 37" to 46"C, the old flagella were not distributed randomly among the daughter cells but followed the distribution expected from an equatorial deposition of new membrane material. This result differs from an earlier one by Quadling and Stocker (1962) obtained with Salmonella typhimurium. However, the assumption that the distribution of flagella reflects that of the membrane site to which they are connected may not be equally valid for both gram-positive and gram-negative bacteria. Analogous experiments were carried out by following the distribution of membrane lipids when cells labeled with specific precursors were grown in unlabeled medium. The distribution of label in individual cells was followed by density gradient centrifugation of membrane fragments (Wilson and Fox, 1971), or by radioautography of whole cells (Lin et al., 1971; Green and Schaechter, 1972). It was found that the label became dispersed over the progeny population, suggesting that new membrane was deposited at many sites. This interpretation is clouded, however, by the possibility that the membrane is sufficiently fluid to permit the rapid lateral movement of newly made pieces. The pattern of segregation would then not correspond to the pattern of membrane synthesis. One kind of membrane protein, the permease involved in the transport of P-galactosides, has been reported to segregate semiconseruatively (Kepes and Autissier, 1972).When E. coli is grown under conditions that prevent the synthesis of new permease, old permease

BACTERIAL CHROMOSOME AND CELL MEMBRANE

15

molecules are not segregated randomly but are conserved in a few of the progeny cells. It is not clear at present whether the discrepancy between this result and those described above reflects differences in the behavior of various membrane constituents or in the methods employed.

IV.

Polarization in the Segregation of Chromosomes

If the progeny of a single cell could be lined up in a row that reflects their genealogy, would the original DNA strands always be found in certain cells, or would they segregate randomly? In other words, is the segregation process polarized? This question is relevant to our concerns because polarization in segregation suggests that DNA is attached to the membrane. Perhaps the simplest mechanism to explain polarization is that some time after replication the parental strands of each daughter chromosome become permanently attached to two different sites on the membrane. The relevant experiments have been done by growing cells whose DNA had been previously labeled in unlabeled medium under conditions in which the genealogical relationships are spatially maintained. This can be achieved by growth on agar in the presence of methylcellulose which restrains cell movement so that progeny cells are maintained as chains. The position of original parental DNA can then be determined by radioautography. Early work done by this or by analogous methods gave contradictory evidence. Thus Chai and Lark (1967, 1970) and Eberle and Lark (1966)found that the pattern of chromosome segregation was nonrandom in Lactobacillus acidophilus, B . subtilis, and E . coli. Ryter et al. (1968),Ryter (1968),and Lin et al. (1971)found the opposite, that the parental DNA was dispersed randomly over the progeny cells. To a large extent this discrepancy may be attributed to a physiological peculiarity of rodshaped bacteria, namely, that the number of nuclei per rod is dependent on the growth rate (e.g., Maalge and Kjelgaard, 1966). Moreover, the number of points on each chromosome where replication takes place also varies with the growth rate. Thus the number of “units of DNA conservation” in the original cell, that is, the number of labeled individual DNA strands, varies from a minimum of 2 at very slow growth rates, to 8 or 16 at fast growth rates. A high number of labeled strands in the original cell obscures segregation data, making the analysis of such experiments intricate and laborious. Perhaps the most thorough study of this point has been that of Pierucci and Zuchowski (1973),who reported that chromosome seg-

16

P. J. LEIBOWITZ AND M. SCHAECHTER

regation is indeed not random. They studied E. coli B/r grown at slow growth rates, Experimental and theoretical distribution patterns were determined for the location of labeled cells in chains. They found that all cells became unlabeled with unequal frequency, and concluded that segregation is polarized. They considered various models in detail, and the data appear to fit best the following. Upon replication the newly made strand does not immediately attach to the segregation structure, but becomes stably bound at the time it is first used as a template, namely, one round of replication later. Furthermore, only one strand of the duplex is capable of attachment. This conclusion differs from that of Lark and co-workers, who proposed that either strand of the duplex is capable of attachment to the membrane the first time it is used as a template. These data obtained by Pierucci and Zuchowski may fit other models of segregation not yet conceived, and it is not clear to us if the fit of their data is sufficiently good to be taken as proof of any one model. In a particularly interesting experiment, Chai and Lark (1967) determined that the segregation of the DNA of L. acidophilus is coupled to the segregation of the envelope of the cell. They did this by exposing the cells to tritiated thymidine and to fluorescent antibody made against the cell envelope. Upon subsequent growth in medium not containing either of these markers, they determined the proportion of cells containing both original DNA and original envelope material. The fraction of such cells was in fact much higher than expected from independent segregation of both components. It appears therefore from all considerations, that segregation of the bacterial chromosome is a polarized event, a fact that can best be explained by its attachment to the membrane. The proposition that the chromosome is attached to the membrane has been tested by determining whether different replicons contained within the same cell are distributed randomly or not. This has been done by comparing the segregation pattern of the chromosome and of an extrachromosomal element, the F' episome. This plasmid, in addition to carrying genes that impart fertility on E. coZi, also carries other markers that can be used to detect its presence or to turn on or off its replication selectively. This allows a large set of ingenious experiments to show whether or not the chromosome and the episome cosegregate into the progeny cells. It was shown that under a variety of circumstances these replicons do not segregate randomly and that a mechanism must exist to insure unilateral distribution (Cuzin and Jacob, 1965; Hohn and Korn, 1969). Since the two genomes are not known to be directly linked to one another, it

BACTERIAL CHROMOSOME AND CELL MEMBRANE

17

follows that they may both be linked to the cell membrane which becomes the agent of their joint segregation. This was among the earliest evidence for the replicon model and for the suggestion that bacterial DNA is attached to the membrane.

V. Methodological Considerations of Cell Fractionation At present it is not possible to describe the properties of the site of the membrane to which DNA is attached. It is not known how large this site is and whether it differs structurally from the rest of the membrane. Therefore operational criteria for the fractionation and isolation of a DNA-membrane complex do not exist. If one includes as many relevant constituents as possible and perhaps isolates the whole membrane and its attached DNA, the fraction may contain much extraneous material, some of which may be entrained artifactually. However, it may be thought desirable to isolate only the region of the membrane that is in direct apposition to the DNA. This is not easy to do, since it is impossible, a priori, to tell when a complex contains only and all the relevant constituents. The obvious problems include how to determine if a complex exists in the cells or is formed during breakage, and how to tell if some relevant components may have become detached by mechanical or enzymic action. In general, artifacts of aggregation that arise after cell breakage may be assessed by reconstruction experiments. Fractionation of labile cell components should only be carried out after breaking bacteria in a gentle manner. For this reason almost all the methods described below deal with spheroplasts, bacteria that lack some wall constituents and are sensitive to detergents, osmotic shock, or relatively weak mechanical forces. There are four classes of techniques in current use for the retrieval of DNA-membrane complexes: 1. Rapid sedimenting complexes (RSCs). Since DNA-membrane complexes are heavy relative to other cell constituents, they can be retrieved by centrifugation through sucrose gradients. Usually, such gradients are made over a cushion of cesium chloride or denser sucrose, and material that pellets to this interface may be called a RSC. Free DNA released from the membrane by shearing or nuclease action remains at the top of the gradient. In many of the reports that describe this technique (in several variations), the RSC contains much membrane material and relatively little of the total cell DNA. Under the conditions used membrane

18

P. J. LEIBOWITZ AND M. SCHAECHTER

fragments probably vesiculate, which may lead to the mechanical entrainment of DNA in membrane vesicles. While exogenously added purified DNA is not entrained in RSCs, it is not known if DNA as it exists in a cell with its normal ligands may in fact be entrained. 2. Complexes obtained by isopycnic centrifugation or electrophoresis. DNA-membrane complexes can be expected to possess buoyant densities different from those of other cell constituents. Ivarie and P h e (1970) separated newly synthesized from uniformly labeled DNA on linear renografin density gradients. An advantage of this technique is the short time required for the membrane-DNA complex to reach its equilibrium and for its separation from the other cellular components. The complex they isolated is enriched for markers near the origin and terminus. Daniels (1971)fractionated the cell membrane into various portions, one of which contains DNA and can be banded in equilibrium gradients. Since the DNA is found in only one or two of nine fractions, it appears unlikely that it is entrained nonspecifically in membrane vesicles. In an analogous fashion, DNA-membrane complexes can be expected to have a distinct mobility in an electric field and may be retrieved by electrophoresis. This has been done recently by Olsen et al. (1974). 3. M bands. When cell lysates are mixed with crystals obtained by adding magnesium salts to Sarkosyl (the detergent sodium lauroyl sarcosinate), a characteristic band is formed after centrifugation through sucrose gradients. This band (M band) contains crystals to which membrane-DNA complexes adhere. There are reasons to believe that this adhesion is due to an affinity between the hydrophobic surface of the crystals and the membrane. Membranes alone attach to the crystals, while native DNA does not. Therefore it is assumed that DNA is found in M bands because it is bound to the membrane (Tremblay et al., 1969; Earhart et al., 1968). It is possible to increase the proportion of membrane in the M band by letting the lysate-crystal mixture stand or to decrease it by using Triton X-100 as the lysing detergent. Nearly all the DNA of the cell is attached to as little as 4% of the total membrane (Ballesta et al., 1972). Hence different portions of the membrane appear to be heterogeneous in their affinities for the crystals, with the DNA-bearing portion having a very high affinity. 4. Zsolated nucleoids. When lysis of E. coli is carried out in the presence of 1 M sodium chloride, the DNA does not become extended but remains in a folded, compact state. These nuclear bodies can be isolated by centrifugation through sucrose gradients. Unlike DNA in solution, they do not contribute significantly to the viscosity

BACTERIAL CHROMOSOME AND CELL MEMBRANE

19

of the medium and have a high sedimentation coefficient. They contain a small amount of protein, nascent RNA, and some membrane constituents (Stonington and Pettijohn, 1971). They can be isolated as membrane-attached bodies containing either as much as 20% of the cell membrane or nearly free of membrane material. These compact nucleoids contain nearly all the RNA polymerase found in the cell. They sediment to positions in sucrose gradients, which apparently correspond to their stage of replication at the time of preparation (Worcel and Burgi, 1972).

VI. Is the Chromosome Attached to a Special Region of the Membrane?

This is a difficultquestion, since there are no satisfactory means for the chemical characterization of different portions of the bacterial membrane. Thus mesosomes, the likely candidates for the membrane region to which DNA is bound, are not decisively different from the rest of the membrane. One of the recognizable membrane components in “membraneattached” nuclear bodies (Stonington and Pettijohn, 1971)is a peptide which is found in the outer layer of E. coli (Worcel and Burgi, 1974). Other membrane components may be present, but are not sufficiently distinct for identification. This finding suggests that under certain conditions the DNA-membrane complex contains both inner and outer membrane components. These two layers are in close apposition at about 200 places over the cell surface (Bayer, 1968a,b), and it seems possible that these sites of contact may play a special role in DNA binding. The membrane component of M bands or of fractions isolated by an analogous method have a different phospholipid composition than the average membrane (Ballesta et al., 1972; Daniels, 1971). They are richer in phosphatidylethanolamine and contain less phosphatidylglycerol and cardiolipin, the other two major phospholipids of E. coli or B . megaterium. In rebanding experiments the portion of the membrane contained in M bands was shown to have a higher affinity for magnesium-Sarkosyl crystals than the rest of the membrane (Ballesta et al., 1972). Daniels (1971) did not find that the DNAbearing fraction had different proteins than the rest of the membrane when examined by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis. However, the large number of peptides seen in these preparations made it difficult to determine the presence of unique components.

20

P. J. LEIBOWITZ AND M. SCHAECHTER

VII. Is the Origin of DNA Replication Attached to the Membrane? The best evidence comes from an analysis of the genetic markers found in DNA-membrane complexes. Sueoka and Quinn (1968)used transformation in B . subtilis to show that RSCs containing 5-10% of the total DNA were substantially enriched in markers located at the origin of chromosome replication in this organism. As expected from the pattern of DNA replication, this enrichment was greater in cells growing at fast rates (O’Sullivan and Sueoka, 1972). These results were confirmed by Snyder and Young (1969), Ivarie and PBne (1970), and Yamaguchi et al. (1971), and extended to E. coli (Fielding and Fox, 1970) where the origin of replication was located by radioactive labeling. Several investigators also reported that the membrane-bound fraction is enriched in markers located at the terminus of DNA replication in B . subtilis (Sueoka and Quinn, 1968; Snyder and Young, 1969; Ivarie and PBne, 1970). A unique, but as yet uncharacterized region of the DNA of Mycoplasma gallisepticum is permanently attached to the membrane (Quinlan and Maniloff, 1973).

VIII. Does DNA Replication Take Place on the Membrane? While this question precedes many others historically, a definitive answer is not yet available. The strongest hint came from the finding that in bacterial conjugation genetic markers are transferred at the time of their synthesis (Jacob et al., 1963). This implies that in the donor cell DNA replicates at or near the site on the membrane involved in the formation of the conjugal bridge. Several investigators have reported the isolation of RSCs enriched in newly made DNA (Ganesan and Lederberg, 1965; Smith and Hanawalt, 1967; Ivarie and PBne, 1970; Fuchs and Hanawalt, 1970; Yamagcchi et al., 1971; Quinlan and Maniloff, 1972; Fujita et al., 1973). The extent of enrichment in newly made DNA in RSCs is not very great. In fact, in most experiments the enrichment factor: Newly made DNA in RSC/Total DNA in RSC Newly made DNA not in RSC/Total DNA not in RSC

vanes from unity (no enrichment) as reported by Yamaguchi et al. (1971), to a value of 2 to 4 in other reports. The value rarely approaches that expected if DNA were attached to the membrane uniquely at its point of replication. However, before concluding that

BACTERIAL CHROMOSOME AND CELL MEMBRANE

21

replication does not take place on the membrane, it should be pointed out that a number of experimental complications do not permit a rigorous analysis of this question. Thus Yamaguchi et al. (1971) found that the extent of enrichment in newly made DNA varied according to the method of preparation of RSCs. Among the difficulties encountered may be the possibility that the conditions of fractionation release the replicating complex from the membrane. It is possible in fact to isolate what appears to be a replicating complex in a fraction devoid of membrane material (Fuchs and Hanawalt, 1970). Likewise, the content of DNA polymerase I in RSCs depends on ionic conditions (Ivarie and PBne, 1970). Newly made DNA has unusual properties which may influence its behavior in cell fractionation. According to current views, DNA is replicated in short stretches (Okazaki fragments) using an RNA primer (Okazaki et al., 1973). It is rapidly detached from this RNA and is linked to the older portion of the chromosome. The physical properties of newly made DNA can be expected to differ from those of the rest of the chromosome. The nature of these differences is not known, but is relevant to the problem of membrane attachment. We have observed that, in E. coli infected with bacteriophage T4,DNA labeled for very short periods is not found in the M band, while DNA labeled for a longer time is (Leibowitz and Schaechter, unpublished data). At least some of this DNA is in the form of Okazaki fragments which ostensibly do not remain attached to their complementary strand under the conditions of fractionation. This result can give the false impression that DNA is not synthesized on the membrane. When an experiment is carried out with the M-band technique, the finding is open to this interpretation, since M bands contain virtually all the DNA. However, this is not obvious when an RSC method is used, since the membrane fraction usually contains only a fraction of the total DNA. These difficulties may not exist for all systems, since in Pneumococcus newly made DNA of the size of Okazaki fragments is found exclusively in the M band (Firshein, 1972). Conversely, instead of being excluded, newly made DNA may be artifactually entrained in RSCs. Newly made DNA can be easily denatured (Kidson, 1960), a fact that may be quite relevant because denatured DNA tends to stick to surfaces, including cell membranes (Dworsky and Schaechter, 1973). Another factor that would lead to an apparent decrease in the enrichment in newly made DNA in RSCs is that DNA may be attached at many sites, perhaps including those where replication takes place. If so, what remains bound to the

22

P. J. LEIBOWITZ AND M. SCHAECHTER

membrane after shearing is a large amount of DNA that is not the replicating region. This may influence the detection of enrichment in newly made DNA.

IX. Is the Chromosome Attached at Many Sites? The replicon model proposes that the chromosome is attached to the membrane at the site of DNA replication. There are several reports that this is not the only site of attachment, and that the chromosomes of E. coEi and €4. subtilis are attached at many places. These data are derived from experiments in which a given number of breaks are introduced in the DNA and the proportion of DNA released from the membrane is measured. The number of breaks, estimated from molecular weight measurements, required to release a given amount of DNA allows one to calculate the number of attachment sites. Breaks can be introduced by x-ray or y-ray irradiation or by shearing lysates. The number of attachment sites reported from experiments of this type is in the order of 10 to 30 for E. coli (Rosenberg and Cavalieri, 1968; Dworsky and Schaechter, 1973) and as many as 70 to 90 in B. subtilis (Ivarie and P h e , 1973). Different methods give different numbers (Ivarie and Phne, 1973; Dworsky and Schaechter, 1973). These estimates are computed on the assumption that the DNA loops between two adjacent attachment sites are similar in length. If a few loops are very long and many are very short, their total number would be seriously underestimated. At present, the available data do little more than suggest that the likely number of attachment sites is large but not as large as the number of folds required to pack the bacterial chromosome inside the bacteria. The E. coli chromosome is about 1100 pm in length (Cairns, 1963) and must be folded well over lo00 times to fit inside the cell. Likewise, the number of attachment sites is considerably smaller than the number of RNA polymerase molecules present in E . coli growing under conditions similar to those of these experiments (it has been estimated that the total number of RNA polymerase molecules is about 7000; Matzura et al., 1973). However, the number of attachment sites is larger than that of origins of replication or sites where replication takes place. The actual numbers would depend on the rate of growth (Helmstetter and Cooper, 1968), but for the cells in question would vary between about one and three and one and seven, respectively. Since the frequency of initiation of RNA synthesis (Bremer and Yuan, 1968) and DNA synthesis (Helmstetter and Cooper, 1968) is dependent on the growth rate, one would anticipate

BACTERIAL CHROMOSOME AND CELL MEMBRANE

23

that the absolute numbers of attachment points would be determined by the growth medium. For these reasons it is unlikely that the attachment of DNA to the membrane is related to a single function. There is in fact an indication that these attachment sites are of more than one kind. When cells are treated with the antibiotic rifampin, which inhibits transcription by binding to free RNA polymerase, the number of attachment sites decreases four- to fivefold (Dworsky and Schaechter, 1973).This suggests that there are functional or structural differences between the attachment sites that are destroyed by rifampin treatment and those that remain after such treatment. Since the estimated number of rifampin-resistant sites was two to five, it is tempting to think that they may represent the origins of DNA replication. The sites that are sensitive to rifampin (8 to 20 in number) depend on some property of RNA polymerase, since they are unaffected by treatment of a mutant whose RNA polymerase is insensitive to the drug. It is tempting to ascribe to these sites a role in transcription. This would have to be a special process, since it is likely that transcription takes place at many more sites than 8 to 20. Perhaps the repeated transcription of some cistrons (e.g., the rRNA cistrons) differs from the less frequent transcription of other cistrons, and may occur in contact with the membrane. These notions are totally unproved but can be subjected to experimental scrutiny.

X. The Nature of the Binding of the Chromosome to the Membrane In this area we are limited entirely to speculations. There are only hints and few facts, since there are no enzymes or chemicals known that remove the entire bacterial chromosome from the membrane in a manner that is nondestructive to either component. Membranes and DNA may be easily dissociated from one another, suggesting that they are not bound through covalent bonds. However, such bonds could easily escape detection. There is a suggestion that some of the attachment may be due to the adherence to the membrane of denatured regions of the chromosome. In reconstruction experiments it was shown that, while native DNA does not stick to membrane, heat-denatured DNA does (Dworsky and Schaechter, 1973).It is not known if extensive regions of denaturation exist in the chromosomes of growing bacteria. These may be found, for example, at the sites of replication, repair, recombination, or transcription. It has been shown that transcription may

24

P. J. LEIBOWITZ AND M. SCHAECHTER

not require denaturation over more than a few base pairs (Saucier and Wang, 1972),but special regions of frequent transcription may in fact result in the formation of larger “bubbles” of denatured DNA. Such regions may correspond to the attachment sites that are destroyed by rifampin treatment (see Section IX). RNA may play a role in the attachment of DNA to the membrane since Earhart et al. (1973) found that RNase liberates both E. coli and T4 phage DNA from the membrane. However, this effect may be indirect since it requires the presence of endonuclease I and is not seen in mutants that lack this enzyme. It is likely that at sites other than the origin of replication DNA is bound in a dynamic sense and, rather than being permanently attached, rapidly moves through the site. This is necessarily true if replication takes place on the membrane, and is likely to be the case if the attachment is at sites of special types of transcription. For this reason the involvement of single-stranded DNA is particularly alluring, since DNA denaturation brings about a readily reversible change which has great consequences for the physical properties of the molecule.

XI, The Role of the Membrane in Maintaining the Compactness of the Nucleoid Highly compact nucleoids can be isolated by lysing cells in the presence of 1M sodium chloride (Stonington and Pettijohn, 1971). It has been proposed by Worcel and Burgi (1972) that their compactness is the result of supercoiling of the DNA. These investigators found that the entire chromosome does not act as a single supercoiled circle but rather as if it were separated into 10 to 80 individual supercoils. These loops are thought to be held in their individual position by a “core” of RNA, because thev are unfolded upon treatment with RNase (Stonington and Pettijohn, 1971). It is tempting to relate this to the finding obtained by in vivo treatment with the drug rifampin, which induces both the unfolding of the nucleoids and the elimination of most of the attachment sites of the DNA to the membrane (Dworsky and Schaechter, 1973). It seems possible that the separation of the supercoiled loops is due to their attachment to the membrane, and that the RNA core is located on the membrane and may consist of a rifampin-sensitive structure. In favor of this notion is the coincidence of the estimated number of loops (10 to 80) and the rifampin-sensitive attachment sites (about 8 to 20),and the analogous action of rifampin in vivo and RNase in vitro. However, the amount of membrane found in isolated nucleoids prepared at room temperature is as little as 0.1% of the total. This amount, while very small,

BACTERLAL CHROMOSOME AND CELL MEMBRANE

25

may constitute an important portion of the membrane, since these purified nucleoids will form M bands (Wright and Michaelis, personal communication). The notion that the chromosome exists in the cell in a compact state is being challenged (see Section 11,B). The generally accepted picture of a central compact body may turn out to be inaccurate. Yet it seems unlikely that the bacterial chromosome exists in a random state within cells. Whatever its organization, it seems likely that the attachment to the membrane plays an important role.

XII. Conclusions It should be evident from what we have written above that we believe that work in this field is not at a conclusive stage. Therefore we believe that it is not yet possible to give an accurate picture of the role of membrane attachment in different functions of the bacterial chromosome. However, many of the available data are suggestive and promise further developments. We would like to summarize what is known to date, and what to us is physiologically plausible. The available evidence suggests that the replicon model is correct in postulating that nuclear segregation in bacteria takes place by virtue of the attachment of daughter chromosomes to different sites on the membrane. The relevant conclusions are the origin of DNA replication appears to be attached to the membrane, perhaps in a specific manner, and that daughter chromosomes probably segregate in a polarized fashion. Missing is the evidence that membrane growth takes place in a semiconservative manner which would immediately enable it to function as the segregation apparatus. However, the evidence to the contrary is not conclusive. The evidence on the behavior of mesosomes, the most likely sites of DNA attachment, is particularly confusing. It is not clear whether they split in two or arise de novo. As described in Section VIII, the evidence that DNA replication takes place at the membrane is not strong. The biochemistry of DNA replication is not fully elucidated and is certainly complex. It is not surprising that in vitro systems do not help determine if the membrane is involved in this process. There are better indications that transcription may be involved in membrane attachment, perhaps by creating special regions of DNA denaturation (Sections X and XI). In conclusion, there seem to be good reasons to believe that the bacterial chromosome is attached to the cell membrane. It is not quite as easy to assign to this attachment a role in chromosome segregation, replication, or transcription. The methodological dif-

26

P. J. LEIBOWITZ AND M. SCHAECHTER

ficulties appear great but not impossible, and it can be expected that further work will elucidate many of the current questions. It is tempting to think that the relative structural simplicity of bacteria necessitates that its structures be used in several ways, allowing multiple regulatory connections between various functions. If so, the bacterial cell membrane is a busy place. ACKNOWLEDGMENTS

The helpful comments of B. Beck, P. Dworsky, L. Daneo-Moore, C. Earhart, J. Fox, S. Guterman, M. Higgins, M. Inouye, L. A. McNicol, J. T. Park, A. Ryter, and G. Shockman are gratefully acknowledged. Work from the authors’ laboratory was supported by grants A1 05103 and A1 09465 from the United States Public Health Service. P.J. Leibowitz is a postdoctoral research fellow of the Public Health Service. Note added in pro05 Since the submission of this review at least two important papers have appeared. D. L. Parker and D. A. Glaser [(1974),J. Mol. B i d . 87,153-1681 found that both the origin and the site of chromosome replication in E. coli are membrane-bound. This study represents a thorough application of available techniques. C. E. Helmstetter [(1974),J. Mol. Biol. 84, 1-19,21-36] has proposed that the synthesis of a special region of the membrane serves as a signal for initiation of chromosome replication in E . coli. REFERENCES Adler, H. I.. Fisher, W. D., and Hardigree, A. A. (1969).Trans. N. Y. Acad. Sci. 31,

1050-1070.

Altenburg, B. C., and Suit, J. C. (1970).J. Bacteriol. 103,227-237. Altenburg, B. C., Suit, J. C., and Brinkley, B. R. (1970).J. Bacteriol. 104, 529-555. Bdesta, J. P.. CundlifFe, E., Daniels, M. J., Silverstein, J. L., Susskind, M.M., and Schaechter, M. (1972).J. Bacteriol. 112,195-199. Bayer, M. E. (1968a).J . Gen. Mlcrobiol. 53,395-404. Bayer, M. E. (1968b).J . Vtrol. 2,346-356. Bremer, H.,and Yuan, D. (1968).]. Mol. Blol. 38, 163-180. Burdett, I. D. J., and Rogers, H. J. (1970).J. Ultrastruct. Res. 30,354-367. Cairns, J. (1963).Cold Spring Harbor Symp. Quant. Bid. 28,43-46. Chai, N., and Lark, K. G. (1967).J . Bacteriol. 94,415-421. Chai, N., and Lark, K. G. (1970).J . Bacteriol. 104,401-409. Chapman, G. B., and Hillier, J. (1953).J. Bactertol. 66,362-373. Cuzin, F., and Jacob, F. (1965).C.R. Acad. Sci. 260,5411-5414. Daniels, M. J. (1871).Btochem. J. 122, 197-207. Dworsky. P., and Schaechter, M. (1973).J. Bacteriol. 116, 1364-1374. Earhart, C. F., Sauri, C. J., Fletcher, G., and Wolff, J. (1973).J. Vtrol. 11, 527-534. Earhart, C. F., Tremblay, G. Y.,Daniels, M. J., and Schaechter, M. (1968).Cold Spring Harbor Symp. Quant. B i d . 33,707-710. Eberle, H., and Lark, K. G. (1966).J . Mol. Btol. 22, 187-191. Edwards, M.R.,and Stevens, R. W. (1963).J . Bacteriol. 86,414-428. Ellar, D. J., Lundgren, D. G., and Slepecky, R.A. (1967).]. Bactertol. 94, 1189-1205.

BACTERIAL CHROMOSOME AND CELL MEMBRANE

27

Ferrandes, B., Chaix, P., and Ryter, A. (1966).C.R. Acad. Sci. 263,1632-1635. Fielding, P., and Fox, C. F. (1970).Biochem. Biophys. Res. Commun. 41, 157-162. Firshein, W. (1972).J . Mol. Biol. 70,383-397. Fitz-James, P.C. (1960).J . Biophys. Biochem. Cytol. 8,507-528. Fitz-James, P.C. (1967).In “Symposium on Microbial Protoplasts, Spheroplasts, and L-Forms” (L. B. Guze, ed.), pp. 124-143.Williams & Wilkins, Baltimore, Maryland. Freese, E. B. (1973).J . Gen. Microbiol. 75, 187-190. Fuchs, E.,and Hanawalt, P. (1970).J . Mol. B i d . 52,301-322. Fujita, H., Komano, T., and Maruyama, Y. (1973).Biochem. Biophys. Res. Commun. 52,1361-1367. Ganesan, A. T.,and Lederberg, J. (1965).Biochem. Biophys. Res. Commun. 18, 824-835. Glauert, A. M. (1962).Brit. Med. Bull. 18,245-250. Glauert, A. M., Brieger, E. M., and Allen, J. M. (1961).E r p . Cell Res. 22, 73-85. Green, E. W.,and Schaechter, M. (1972).Proc. Nat. Acad. Sci. US.69,2312-2316. Helmstetter, C. W., and Cooper, S. (1968).J . Mol. B i d . 31,507-518. Higgins, M.L., and Daneo-Moore, L. (1972).J . Bacteriol. 109, 1221-1231. Higgins, M. L., and Shockman, G. D. (1970a).J. Bacteriol. 101,643-648. Higgins, M. L., and Shockman, G. D. (1970b).J. Bacteriol. 103,244-254. Higgins, M. L., and Shockman, G. D. (1971).Chem. Rubber Co.Critical Reo. Microb i d . 1,29-72. Highton, P. J. (1969).J . Ultrastruct. Res. 26, 130-147. Highton, P. J. (1970a).J . Ultrastruct. Res. 31,247-259. Highton, P. J. (1970b).J . Ultrastruct. Res. 31,260-271. H o h a n n , H., Geftic, S. G., Heymann, H., and Adair, F. W. (1973).J.Bacteriol. 114, 434-438. Hohn, B., and Kom, D. (1969).J . Mol. Biol. 45,385-395. Holt, S. C., and Leadbetter, E. R. (1969).Bacteriol. Rev. 33,346-378. Imaeda, T.,and Ogura, M. (1963).J . Bacteriol. 85,151-163. Ivarie, R. D.,and PBne, J. J. (1970).J. Bacteriol. 104,839-850. Ivarie, R. D., and PBne, J. J. (1973).J . Bacteriol. 114,571-576. Jacob, F., and Brenner, S. (1963).C. R . Acad. Sci. 256,298-300. Jacob, F., Brenner, S., and Cuzin, F. (1963).Cold Spring Harbor Symp. Quant. B i d . 28,329-348. Jacob, F., Ryter, A., and Cuzin, F. (1966).Proc. Roy. Soc., Ser. B . 164, 267-278. Kaye, J. J., and Chapman, G. B. (1963).J. Bacteriol. 86,536-543. Kellenberger, E . , Ryter, A., and SBchaud, J. (1958).J . Biophys. Biochem. Cytol. 4, 671-678. Kepes, A., and Autissier, F. (1972).Biochim. Biophys. Acta 265,443-469. Kidson, C. (1966).J . Mol. Biol. 17, 1-9. Landman, 0.E.,Ryter, A., and FrBhel, C. (1968).J. Bacteriol. 96,2154-2170. Lin, E.C. C., Hirota, Y., and Jacob, F. (1971).J. Bacteriol. 108,375-385. Maal@e,O., and Kjeldgaard, N. 0. (1966).In “Control of Macromolecular Synthesis” (B. Davis, ed.). Benjamin, New York. Bacteriol. 96, McCandless, R. G., Cohen, M., Kalmanson, G., and Guze, L. B. (1968).J. 1400-1412. Margaretten, W.,Morgan, C., Rosenkranz, H. S., and Rose, H. M. (1966).J. Bacteriol. 91,823-833. Matzura, H., Hansen, B. S., and Zeuthen, J. (1973).J . Mol. B i d . 74,9-20. Mindich, L., and Dales, S. (1972).J . Cell B i d . 55,32-41. Nanninga, N. (1968).J. Cell B i d . 39,251-263.

28

P. J. LEIBOWITZ AND M. SCHAECHTER

Okazaki, R.,Sugino, A,, Hirose, S., Okazaki, T., Imae, Y., Kainuma-Kuroda, R.,Ogawa, T., Arisawa, M., and Kurosawa, Y. (1973).In “DNA Synthesis In Vitro” (R.D. Wells and R. B. Inman, eds.), pp. 83-106. University Park Press, Baltimore, Maryland, J . Bacteriol. Olsen, W. L., Heidrich, H.-G., Hofschneider, P. H., and Hannig, K. (1974). 118,646-653.

O’Sullivan, M. A., and Sueoka, N. i1972).J . Mol. Biol. 69,237-248. Patch, C. T.,and Landman, 0. E. (1971).J . Bacteriol. 107,345-357. Pierucci, 0.. and Zuchowski, C.(1973).J . Mol. B i d . 80,477-503. Pontefract, R. D.,Bergeron, G., and Thatcher, F. S. (1969).J. Bacteriol. 97, 367-375. Quadling, C.,and Stocker, B. A. D. (1962).J . Gen. Mfcrobiol. 28,257-270. Quinlan, D.C.,and Maniloff, J. (1972).J . Bacteriol. 112,1375-1379. Quidan, D. C., and Maniloff, J. (1973).J. Bacterlol. 115,117-120. Remsen, C.C.(1968).Arch. Mikrobfol. 61,4047. Reusch, V. M., and Berger, M. M. (1972).Fed. Proc., Fed. Amer. SOC. Erp. Biol. 31, 413. Reusch, V. M., and Berger, M.M. (1973).Biochim. Biophys. Acta 300,79-104. Rogers, H. J. (1970).Bactetiol. Reo. 34,194-214. Rosenberg, B. H., and Cavalieri, L. F. (1968).Cold Spring Harbor Symp. Quant. Biol. 33,65-72. Ryter, A. (1967).Folfa Mimbiol. (Prague) 12,283-290. Ryter, A. (1968).Bacteriol. Reu. 32,39-54. Ryter, A. (1971).Ann. Inst. Pasteur, Paris 121,271-288. Ryter, A,, and Jacob, F. (1963).C. R. Acad. Scf. 257,3060-3063. Ryter, A,, and Jacob, F. (1964).Ann. Inst. Pasteur, Paris 107,384400. Ryter, A,, and Jacob, F. (1966).Ann. Inst. Pasteur, Paris 110,801-812. Ryter, A,, and Landman, 0. E. (1964).J . Bacteriol. 88,457-467. Ryter, A., and Landman, 0. E. (1967).In “Symposium on Microbial Protoplasts, Spheroplasts, and L-Foms” (L. B. Guze, ed.), pp. 110-123. Williams & Wilkins, Baltimore, Maryland. Ryter, A., Hirota, Y.. and Jacob, F. (1968).Cold Spring Harbor Symp. Quant. Biol. 33, 669-676. Salton, M. R.J. (1971).Chem. Rubber Co.Critical Reo. Mfcrobiol. I, 29-72. Saucier, J. M., and Wang, J. C. (1972).Nature (London). New Biol. 239, 167-170. Siegel, P. J., and Schaechter. M. (1973).Annu. Reo. Microbiol. 27,261-282. Smith, D. W., and Hanawalt, P. C. (1967).Btochim. Biophys. Acta 149, 519-531. Snyder, R. W., and Young, F. E. (1969).Biochem. Biophys. Res. Commun. 35, 354-362. Steed, P., and Murray, R. G. E.(1966).Can. J . Microbiol. 12,263-270. Stonington, 0. G., and Pettijohn, D. E. (1971).Proc. Nut. Acad. Sci. U.S.68,6-9. Stove Poindexter, J. L., and Cohen-Bazire, G. (1964). J . Cell Biol. 23,587-607. Sueoka, N.. and Quinn, W. G. (1968).Cold Spring Harbor Symp. Quant. Biol. 33, 695-705. Tremblay, G. Y., Daniels, M. H., and Schaechter, M. (1969).J . Mol. Biol. 40,65-76. van Iterson, W. (1961).J . Biophys. Bfochem. Cytol. 9, 183-192. van Iterson, W. (1965).Bmteriol. Reo. 29,299-325. van Iterson, W., and Leene, W. (1964).J . Cell Biol. 20,361-375. Wilson, G., and Fox, C. F. (1971).Biochem. Biophys. Res. Commun. 44, 503-509. Worcel, A., and Burgi, E. (1972).J. Mol. Biol. 71,127-147. Worcel, A., and Burgi, E. (1974).J. Mol. Biol. 82,Ql-105. Yamaguchi, K., Murakawi, S., and Yoshikawa, H. (1971).Biochem. Biophys. Res. Commun. 44,1559-1565. Ziegler, R. J., and Daneo-Moore, L. (1971).J . Bacteriol. 105, 190-199.

Regulation of the Lactose Operon in Escherichia coli by cAMP G. CARPENTER'AND B. H. SELLS Laboratories of Molecular Biology, Faculty of Medicine, Memorial University of Newfoundland, St. John's, Newfoundland, Canada

. . . . . . . . . . . . . . . . . .

29

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

34 36 36

I. Introduction. A. Role of CAMP. . . . . . . . . . . . . . . . . B. Organization of the Lac Operon . . . . . . . . . . C. Nature of the Inducer 11. cAMP Formation and Catabolite Repression A. cAMP and the Catabolite Repression of &Calactosidase . . B. Cellular Concentrations of cAMP C. Enzymes Involved in cAMP Metabolism. . . . . . . D. Control of Catabolite Repression of fl-Galactosidase Synthesis. . . . . . . . . . . . . . . . . . . 111. In Vioo Evidence of Site of CAMP Action. . . . . . . . . A. Transcriptional or Translational Control. B. cAMP and Lac Promoter Mutants . . . . . . . . . . IV. cAMPActioninVitro. . . . . . . . . . . . . . . . A. In Vitro Synthesis of /ibGalactosidase. . . . . . . . . B. Mediation of cAMP Action C. Properties of CAMP-Receptor Protein . . . . . . . . D. Initiation of Lac mRNA Synthesis in Vitro . . . . . . V. Additional Aspects of Lac Operon Regulation . . . . . . . References

. . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

30 31

37 39 42 43 43 45 46 46 47 49 51

54 55

I. Introduction

Over the past two decades, significant progress has been made in explaining the general mechanism involved in the synthesis of protein from information encoded in the polydeoxynucleotide sequences in DNA. How synthesis of a particular protein is controlled has been resolved only in a few instances. The system that has received the greatest attention and which is best understood is the lactose (lac) operon in Escherichia coli. This article, which deals with a description of the lac operon, examines the various points of control in the synthesis of the enzyme pgalactosidase. Protein synthesis in its broadest sense consists of two processes -transcription and translation. The evidence currently avail-

' Present address: Department of Biochemistry, Vanderbilt University, Nashville, Tennessee. 29

30

G. CARPENTER AND B. H. SELLS

able concerning pgalactosidase formation suggests that most, if not all, the control of its expression occurs at the level of transcription. This control involves regulation by several components which interact with genetic material at distinct locations within the lac operon. This interaction is influenced by several low molecular weight compounds whose concentration modulates the expression of the operon. The model described for pgalactosidase synthesis may provide a prototype for understanding how production of other specific proteins is regulated. Following the pioneering work of Jacob and Monod on enzyme induction, a discovery which greatly aided in our understanding of the mechanisms that regulate pgalactosidase synthesis, was the recognition that the metabolite cyclic AMP (adenosine 3’,5’-monophosphate) (CAMP)is involved in the process. This cyclic nucleotide, which has an influence on a variety of biological systems, is discussed specifically in terms of its effect on the lac operon. A. ROLE OF cAMP Since 1957, when Sutherland and co-workers (Rall and Sutherland, 1958; Rall et al., 1957; Sutherland and Rall, 1957, 1958) identified CAMP as the biological factor that mediates the stimulatory effect of epinephrine and glucagon on glycogenolysis in rat liver, the nucleotide has been implicated in the regulation of many diverse physiological processes. The role of cAMP in mediating hormone action was explained by Sutherland et al. (1965) in their second messenger hypothesis. This hypothesis postulates that many polypeptide hormones interact with the cytoplasmic membranes of target cells and increase the activity of the membrane-bound enzyme adenyl cyclase which catalyzes the synthesis of cAMP from ATP. Increased intracellular levels of cAMP then bring about the appropriate biochemical response. For his discovery of cAMP and his identification of its physiological role, Sutherland was awarded the Nobel prize for medicine in 1971. Detailed review articles concerned with the role of cAMP in mammalian cells have been published (Jost and Rickenberg, 1971; Pastan and Perlman, 1971; Robinson et al., 1968). 1. Presence of CAMP in Bacteria The presence of this cyclic nucleotide in bacteria was first reported by Makman and Sutherland (1963), who identified cAMP in extracts of E. cold. The same year, Okabayashi et al. (1963) identified the nucleotide in extracts of Breuibactedum liquefuciens. Two years

REGULATION OF THE LACTOSE OPERON

31

later, Makman and Sutherland (1965) noted a correlation between the intracellular concentration of cAMP and the amount of glucose in the culture medium. As the medium became depleted of glucose, the intracellular level of cAMP increased sharply. When a suspension of E . coli was grown in excess glucose and transferred to a phosphate buffer without glucose, the concentration of cAMP increased from lo-' to M within 60 minutes. The addition of glucose to the phosphate buffer completely prevented the rise in cAMP concentration. Although it had been known for many years that glucose exerts a repressive effect on the synthesis of many enzymes in microorganisms, it was not until 1968 that Perlman and Pastan (1968a)and U11mann and Monod (1968) made the initial discovery that cAMP can overcome the glucose repression of pgalactosidase synthesis. This finding implicated this cyclic nucleotide in the regulation of enzyme synthesis in prokaryotic cells. During the past 5 years, many investigators have succeeded in elaborating the mechanism by which cAMP controls pgalactosidase synthesis. This system now represents one of the best understood in terms of regulation at the gene level. 2. Biological Events Efiected b y cAMP Recently, several reviews concerning the role of cAMP in enzyme synthesis in prokaryotic cells have been published (Pastan and Perlman, 1970; Perlman and Pastan, 1971, Reznikoff, 1972; Zubay and Chambers, 1971). As shown in Table I, cAMP is involved in the regulation of a large and diverse number of biological processes in bacteria. The list is undoubtedly incomplete, and the involvement of cAMP in processes other than enzyme synthesis may be indirect and reflect a requirement for the synthesis of a particular protein. With the exception of glutamate synthetase and glutaminase, cAMP regulates enzyme synthesis in a positive manner. Since the role of cAMP is best understood in the case of p g a l a c tosidase synthesis, we consider in this article the regulatory mechanisms by which cAMP controls the formation of this enzyme. B. ORGANIZATION OF THE LAC OPERON PGalactosidase is one of three enzymes coded for by a segment of DNA on the E. coli genome known as the lac operon. On the addition of lactose to the growth medium, the intracellular levels of these enzymes increased more than 1000-fold within a few minutes, enabling the bacteria to adapt to a new nutritional environment rapidly. PGalactosidase, which is coded for by the z cistron of the lac operon,

32

G. CARPENTER AND B. H. SELLS

TABLE I EFFECTSOF CAMP ON BACTERIAL METABOLISM Process Enzyme synthesis SGalactosidase Lac permease Galactokinase

Glycerol kinase a-Glycerolphosphate permease Arabinase operon regulator protein L-Arabinose permease LArabinose isomerase Fructose enzyme I1 (phosphotransferase) Tryptophanase DSerine deaminase Thymidine phosphorylase Threonine deaminase Pseudouridine kinase Pseudouridylate synthetase Chloramphenicol acetyl transferase GMP reductase Glutamine synthetase Glutamate dehydrogenase Glutaminase A Cytochromes Streptomycin adenylcyl transferase Oxidative phosphorylation Lysogeny Flagellar formation Bioluminescence Transformation Plasmid DNA replication Colicin DNA replication Protein kinase activity

Reference Perlman and Pastan (1Q68a); Ullmann and Monod (1963) de Crombrugghe et 01. (196Qb) Tao h Schweiger (1970); de Crombrugghe et al. (1Q69b); Nissley et al. (1971); Park et al. (1971); Wetekam et al. (1971) de Crombrugghe et al. (lQ6Qb) de Crombrugghe et al. (1969b) Yang and Zubay (1973) de Crombrugghe et al. (186913) Nakazawa and Yokota (1973) de Crombrugghe et al. (1969b) del Camp0 et al. (1970); Perlman and Pastan (1988a); Ramirez et al. (1072) de Crombrugghe et al. (1869b); McFall (1973) de Crombrugghe et al. (1QBQb) Shizuta and Hayaishi (1970) Perlman and Pastan (1971) Perlman and Pastan (1971) de Crombrugghe et al. (1972); Harwood and Smith (1971) Benson et al. (1971) Prusiner et al. (1972) Prusiner et al. (1972) Prusiner et al. (1972) Broman and Dobrogosz (1972) Harwood and Smith (1971) Hempfling and Breman (1971) Grodzicker et al. (1972) Yokota and Cots (1970) Nealson et 01. (1972) Wise et al. (1973) Katz et al. (1973) Nakazawa and Tamada (1972) Khandelwal et al. (1973); Kuo and Greengard (1869)

REGULATION OF THE LACTOSE OPERON

33

hydrolyzes lactose to glucose and galactose. The enzymology of p galactosidase has been discussed by Zabin and Fowler (1970). The molecular weight of the active enzyme is 540,000 daltons, and the enzyme is a tetramer composed of four identical polypeptide chains. The product of the z cistron therefore, is a polypeptide of 135,000 daltons. Although the synthesis of active enzyme is dependent on subunit assembly, Adamson et al. (1970)have reported that subunit assembly is not a rate-limiting factor during Pgalactosidase induction. The y cistron of the lac operon codes for a membrane-localized protein, referred to as the lac permease or M protein, which is responsible for the transport of pgalactosides such as lactose into intact cells. The lac permease is composed of subunits which have a molecular weight of 30,000 daltons; however, the active form of this enzyme is not known. Kennedy (1970) and Kepes (1971) have reviewed various aspects of the lac permease system. The third enzyme of the lac operon, thiogalactoside transacetylase, is a product of the a cistron of this operon. The gene product of this cistron is a 32,000-dalton polypeptide. Apparently, the active enzyme exists as a dimer of two identical subunits. Although thiogalactoside transacetylase activity can be detected in crude extracts by the transfer of a radioactive acetyl group from acetyl CoA to a thiogalactoside acceptor, the role of this enzyme in lactose metabolism remains obscure. Strains carrying a deletion of the a cistron grow as well as wild-type strains on lactose (Fox et al., 1966). The enzyme may be involved in a secondary pathway or side reaction during lactose metabolism, and has been discussed at length by Zabin and Fowler (1970)and Kennedy (1970). The arrangement of the z, y, and a cistrons in the lac operon, together with the regulatory sites that control the expression of these cistrons, is shown in Fig. 1. The i cistron, which is located just to the left of the lac operon,

FIG.1. Diagramatic arrangement of components of lac operon (not drawn to scale).

34

G. CARPENTER AND B. H. SELLS

codes for the lac repressor. This protein is thought to be constitutively synthesized, and has a low level of production as indicated by the fact that there are approximately 10 molecules of active repressor per haploid genome. The active repressor is a tetrameric molecule composed of four identical subunits. The product of the i cistrons is then a polypeptide with a molecular weight of 38,000 daltons. The nature of the repressor has been discussed in detail by Gilbert and Miiller-Hill (1970)and Bourgeois (1971). The repressor prevents the synthesis of pgalactosidase, lac permease, and transacetylase by binding to the operator ( 0 )region of the lac operon. The p region is the promoter of this operon, and is the site on the DNA to which RNA polymerase binds and at which the initiation of transcription takes place. The binding of repressor between the promoter and the z cistron blocks the transcription of the z, y, and a cistrons and constitutes a negative type of regulation. Synthesis of the enzymes of the lac operon occurs when an inducer is added to the system. The inducer combines with the repressor and decreases the affinity of the repressor for the operator site on the DNA. When the repressor is removed from the operator, transcription of the z, y, and a cistrons takes place. Although the effect of added inducer is commonly referred to as induction of the lac operon, the term derepression is more technically correct. We follow the current usage, however, and refer to induction of pgalactosidase by inducer.

C. NATURE OF THE INDUCER In the presence of lactose, the intracellular levels of the enzymes of the lac operon are increased over 1000-fold within a matter of minutes. Although valuable insights into the mechanism of the induction process have been obtained since the original discoveries of Jacob and Monod, the identity of the natural inducer of the lac operon was only recently established. The effectiveness of synthetic inducers such as isopropylthiogalactoside(IPTG) and thiomethylgalactoside (TMG) in early work obviated the necessity for identifying the natural inducer. Investigations by Burstein et al. (1965) suggested that the lactose molecule itself is not the natural inducer of the lac operon. This conclusion was based on the observation that strains lacking Pgalactosidase are unable to produce transacetylase in the presence of lactose. These workers demonstrated that the presence of some pgalactosidase is necessary to form inducer by a transgalactosidation reaction. Miiller-Hill et al. (1964)had previously observed that allolactose (1,6-O-P-D-galaCtOpyranOSyl-D-glUCOSe),

REGULATION OF THE LACTOSE OPERON

35

an isomer of lactose (1,4O-~-~galactopyranosyl-~-glucose) was an effective inducer of Pgalactosidase in uiuo. Recently, Jobe and Bourgeois (1972)presented strong evidence identifying allolactose as the natural inducer of the lac operon. These investigators showed that allolactose, when added to crude extracts, is the only sugar bound to subsequently purified repressor, and that allolactose acts as an efficient inducer in uitro by destabilizing the operator-repressor complex. They further extended the results of other workers by showing that allolactose is an effective inducer of the lac operon in uiuo. From their studies they also demonstrated that P-galactosidase catalyzes allolactose formation by a rearrangement reaction which converts the 1-4linkage of lactose to a 1-6linkage. The fact that active P-galactosidase is required to produce the inducer ensures that inactive protein forms of the lac operon enzymes are not made and energy wasted as a result of mutations in the z gene or cellular conditions unfavorable to P-galactosidase activity. A mechanism that prevents the expenditure of cellular energy on the synthesis of inactive proteins is illustrated by studies demonstrating that lactose acts as an antiinducer of the lac operon. Jobe and Bourgeois (1973)showed that high concentrations of lactose inhibit the induction of the lac operon in uiuo. That lactose acts as an antiinducer was shown by experiments in uitro,which demonstrated that lactose competitively inhibits the binding of inducers to purified repressor and that lactose stabilizes the repressor-operator complex. These results indicate that E. coli has evolved mechanisms to ensure: (1)that only enzymically active molecules of Pgalactosidase are induced, and (2)that in the absence of active P-galactosidase induction of the entire lac operon is prevented. A model of the effect of lactose and allolactose upon the repressor-operator binding constants is shown in Fig. 2.

FIG.2. Interaction of lactose and allolactose with the lac repressor. (From Jobe and Bourgeois, 1973.)

36

G. CARPENTER A N D B. H. SELLS

11. CAMP Formation and Catabolite Repression

Although the inhibitory effect of metabolites, particularly glucose, on the production of enzymes in bacteria has long been recognized, an understanding of this process, initially termed catabolite repression by Magasanik (1961), has only recently been uncovered. The reader is referred to the review by Paigen and Williams (1970) for extensive information concerning catabolite repression and various aspects of bacterial metabolism. A. CAMP AND OF

THE CATABOLITE REPRESSION

~GALACTOSIDASE

The studies of Makman and Sutherland (1965)provided the initial hint that eventually led to an understanding of the mechanism of glucose repression of enzyme synthesis. In 1965,these workers demonstrated a significant effect of glucose on intracellular concentrations of CAMP. The addition of glucose to a suspension of bacteria was shown to reduce markedly the intracellular levels of CAMP, although the physiological role of CAMP was not recognized at that time.

0

1;'be 0.03

0.04 I

0.105

0.06 I

Optical density 560 n m I FIG.3. EfFect of glucose and CAMPon the rate of synthesis of @gdactosidase. All samples contained IPTC. Glucose (closed circles) or glucose and CAMP (5 mM) (squares) was added at the arrow. Glycerol (open circles). (From Pastan and Perlman, 1970.) Copyright 1970 by the American Association of the Advancement of Science.

REGULATION OF THE LACTOSE OPERON

37

In 1968, the experiments of Perlman and Pastan ( 1 W a ) and Ullmann and Monod (1968)demonstrated that the addition of cAMP to E , coli growing in a glucose medium relieved the repressive effect of glucose on pgalactosidase synthesis (Fig. 3). The relationship of glucose, CAMP,and pgalactosidase synthesis raises two broad questions. How do metabolites such as glucose affect the intracellular levels of CAMP?What is the mechanism by which cAMP regulates the synthesis of pgalactosidase? Investigations pertinent to the former question are discussed in the remaining portion of this s e e tion, while studies concerning the latter question are considered in Section IV.

B. CELLULAR CONCENTRATIONS OF cAMP Buettner et al. (1973) have studied the intracellular concentrations of cAMP in strains of E. coli sensitive or resistant to catabolite repression. The catabolite repression-resistant (CR-) strains examined by these investigators either possess little cAMP phosphodiesterase activity (Crookes strain) or are phosphodiesterase-negative (strain AB257Pe') mutants. Phosphodiesterase is the enzyme that degrades cAMP to AMP. Presumably, these two strains are resistant to catabolite repression because their intracellular levels of cAMP are higher than those of strains that possess active phosphodiesterase and are sensitive to catabolite repression (CR+ strains). The data in Table I1 show the relationships among growth on different carbon sources, intracellular cAMP concentrations, and pgalactosidase levels. These results indicate that there is an inverse correlation between cell growth rate and intracellular cAMP levels. The CRstrain had slightly higher levels of cAMP than the CR+ strains at most growth rates. During growth on glucose the difference in cAMP concentration was approximately twofold. The results indicate, however, that catabolite repression of pgalactosidase formation is not mediated solely by CAMP. While pgalactosidase specific activity is approximately 300- to 400-fold higher in CR- strains compared to CR+strains during growth on a mixture of glucose and gluconate, the intracellular levels of cAMP are nearly equal in the different strains. Furthermore, the specific activity of pgalactosidase in the CR+ strain increases 20-fold when transferred from a medium containing a mixture of glucose and gluconate to a medium containing glucose. A possible explanation for these inconsistent relationships is that carbohydrates that repress pgalactosidase synthesis may interfere with the entry of inducers into intact cells (Kepes, 1960; Magasanik, 1970). Experiments using constitutive strains would obviate this difficulty

38

G. CARPENTER AND B. H. SELLS TABLE I1 CELLULAR CONCENTRATION OF CAMP AND ~GALACTOSIDASE ACTIVITY DURING EXPONENTIAL GROWTH’

Shin

AB257 (CR+)

AB257”-’ (CR-)

Crookes (CR-)

Source of carbonb Glucose plus gluconate Glucose Glycerol Succinate Acetate L-Proline Glucose plus gluconate Glucose Glycerol Succinate Glucose plus gluconate Glucose Glycerol Succinate

Generation time (minutes)

Intracellular cAMP (10-5 M )

/?-Galactosidase specific activity‘

57

1.2

10

56 96 170 270 218 58

1.2 4.3 5.0 5.5 24.0 0.8

200 12,500 14,500 19,800 43.800 4,000

72 66 157 48

2.3 10.2 9.3 0.8

11,000 13,000 13,400 3,500

53 59 86

2.2 4.5 6.2

11,000 13,000 15,400

a From Buettner et ol. (1973). with permission of the American Society for Microbiology. b Concentrations were: glucose and gluconate, 10 mM each; glycerol, 20 mM; succinate, 15 mM; acetate, 30 mM; L-Proline, 30 mM. Specific pgalactosidase activity is given as nanomoles of o-nitrophenol formed per minute per milligram of bacterial protein at 37°C.

and provide a better model to study the correlation between Pgalactosidase levels and intracellular CAMP. Pastan et al. (1969) showed that the addition of high concentrations of cAMP to a glucosegluconate medium reverses the catabolite repression encountered in this medium. Numerous workers (Buettner et al., 1973; Monod et aZ., 1970; Peterkofsky and Gazdar, 1971) repeated the initial observations of Makman and Sutherland (1965), which demonstrated that cellular concentrations of cAMP increase quickly when the E. coli medium is depleted of glucose. A similar observation has been made with yeast (Sy and Richter, 1972). This change in cAMP concentration in E. coli takes place rapidly when the glucose in the medium is exhausted. Within a 15-minute period the intracellular levels of cAMP rose 1000-fold. Since addition of chloramphenicol to the medium does not

REGULATION OF THE LACTOSE OPERON

-\

GIUWS~

ATp

adeny cyclasc

f .......

-

Intermediary metabolism

...............:.............

..",.................".............. '"'

e i E

39

Phospho--,/-

'\

' ' AMP

CATABOUTE REPRESSOR .,

./'

diesterase

-m

LAC OPERON

.".......................

---------

"

inhibition stimulation

FIG.4. Indirect control of the lac operon by a catabolite repressor.

prevent the increase in cAMP upon removal of glucose, it is unlikely that new protein synthesis is required for this change to occur. Several investigators (Makman and Sutherland, 1965; Peterkofsky and Gazdar, 1971) have suggested that excretion of CAMP into the medium is an important mechanism in the regulation of CAMP levels. The results of Buettner et al. (1973) indicate, however, that the intracellular concentration of cAMP bears little relationship to the extracellular concentration. They also demonstrated that the high intracellular levels of cAMP produced in intact cells by depleting the medium of glucose can be dramatically lowered by the addition of glucose to the cells. In this case, the decrease in intracellular cAMP is not reflected by the accumulation of cAMP in the medium. These observations suggest that the activities of adenyl cyclase and cAMP phosphodiesterase play a larger role than the export mechanism in controlling the intracellular concentration of CAMP.

C. ENZYMES INVOLVED IN cAMP METABOLISM 1. Aden yl Cyclase Adenyl cyclase catalyzes the conversion of ATP to cAMP and pyrophosphate in bacterial and eukaryotic cells. The importance of this enzyme to carbohydrate metabolism is demonstrated by the isolation of a mutant of E. coli deficient in adenyl cyclase activity (Perlman and Pastan, 1969). In the absence of exogenous CAMP,this mutant is unable to grow on lactose, maltose, arabinose, mannitol, or glycerol, and grows poorly on glucose, fructose, or galactose. Peterkofsky and Gazdar (1973)devised a procedure to measure the

40

G . CARPENTER AND B. H. SELLS

activity of adenyl cyclase in uiuo. This method involves labeling cAMP in uiuo from precursor adenosine, followed by quantitative analysis of the labeled nucleotide. When the labeling period is short and the specific activity of the ATP pool is measured, the rate of incorporation of radioactivity corresponds to a determination of adenyl cyclase activity in uioo. The results indicate that increases in cAMP levels that occur when glucose in a growth medium is exhausted are accompanied by an equivalent increase in the activity of adenyl cyclase. Measurements in uitro of adenyl cyclase activity in extracts prepared from yeast cells at various stages of growth in a glucose medium also showed that the enzyme activity increases when the medium is depleted of glucose (Sy and Richter, 1972). Abou-Sabe’ and Nardi (1973)recently isolated a catabolite repression-resistant mutant of E . cold which synthesizes cAMP during growth on glucose at a rate equivalent to that obtained during growth on glycerol. During growth on a glucose medium, synthesis of both Pgalactosidase and cAMP was stimulated. When the adenyl cyclase of this strain was isolated in a membrane fraction, the specific activity of the enzyme was increased by the addition of glucose to the reaction mixture. Solubilization of the adenyl cyclase resulted in an enzyme preparation which was insensitive to stimulation by glucose. These results suggest that glucose may interact directly or indirectly with the membrane-bound adenyl cyclase in E. colt and thereby regulate the intracellular levels of CAMP. This situation is strikingly similar to hormonal interactions with membrane-bound adenyl cyclase in eukaryotic cells. Purification of E . colt adenyl cyclase has proved difficult because of its association with particulate matter and instability after solubilization (Kepes, 1960).Although Ide (1969)was unable to extract the enzyme with detergents, Tao and Lipmann (1969)and Tao and Huberman (1970)found that it could be removed from the particulate fraction by washing with buffer. Both groups observed that the enzyme requires Mg+*and that its activity is inhibited by pyridoxal phosphate, oxaloacetate, pyruvate, malate, and ribose 5phosphate. The significance of these inhibitory compounds is questionable, as ATP was shown to be degraded by competing enzymes when pyruvate, ribose $phosphate, or oxaloacetate was added to the reaction mixture. The molecular weight of the adenyl cyclase was estimated at 110,OOO by sucrose density gradient centrifugation (Tao and Huberman, 1970). The adenyl cyclase of B. Ztquefaciens was studied by Hirata and Hayaishi (1965,1967) and Ide and co-workers (1967). In con-

REGULATION OF THE LACTOSE OPERON

41

bast to the E . coZi enzyme, they found that adenyl cyclase in B . Ziquefuciens is not associated with the particulate fraction but is found in the cell sap. Furthermore, the enzyme is dependent on pyruvate. Ide (1971) surveyed the adenyl cyclases of 21 strains of bacteria and reported that, in bacteria in which adenyl cyclase was found in the supernatant fraction, the enzyme was activated by pyruvate. In strains in which adenyl cyclase was located in the particulate fraction, the enzyme was not activated by pyruvate. Khandelwal and Hamilton (1971, 1972) purified 3200-fold the adenyl cyclase from Streptococcus suZiuurius and studied the effect of various metabolites on the purified enzyme. The enzyme was inhibited by ADP and various nucleoside triphosphates in a competitive manner. Inhibition by these compounds was also reported in E. coZi (Tao and Huberman, 1970) and B. Ziquefuciens (Hirata and Hayaishi, 1967). Inhibition of the S. suZiuurius enzyme was also observed when various diphosphate glucose nucleosides were added to the reaction mixture. An examination of the effect of individual glycolytic intermediates on adenyl cyclase activity showed that glucose 6-phosphate, glucose 1-phosphate, 2-phosphoglycerate, and pyruvate stimulated the enzyme up to 40%. Citrate and lactate inhibited the enzyme up to 50%. In the presence of fructose 6phosphate, fructose l,&diphosphate, glyceraldehyde 3-phosphate, and phosphoenolpyruvate, enzyme activity was either enhanced or inhibited, depending on the concentration of the compound present. Khandelwal and Hamilton (1972) have discussed the relationship between the adenyl cyclase system and the activity of various glycolytic enzymes. 2. Phosphodiesteruse Ide (1971)demonstrated the presence of CAMPphosphodiesterase activity in a variety of bacterial species. The enzyme from Serrutiu murcescans was purified over 1000-fold; however, the possible regulation of enzyme activity by metabolic effectors was not reported (Okabayashi and Ide, 1970). Aboud and Burger (1971b) have reported that the phosphodiesterase activity of E. cold is much lower in cells grown on glucose than in cells grown on glycerol. When cells were grown on glucose plus CAMP,the levels of phosphodiesterase increased to that obtained with cells grown in a glycerol-containing medium. Although the mechanisms by which phosphodiesterase activity was lowered during growth on glucose were not determined, these investigators suggested that the differences were due to the rate of enzymes synthesis. The hypothesis that synthesis of phospho-

42

G. CARPENTER AND B. H. SELLS

diesterase is controlled by the type of carbohydrate present and/or intracellular cAMP levels, however, can be inferred only from measurements of enzyme activity. Peterkofsky and Gazdar (1971)did not find a correlation between the amount of glucose in the medium and the activity of phosphodiesterase. Monod, Janecek and Rickenberg (1970) purified phosphodiesterase from E. coli. They found that the enzyme was composed of two protein components, separable by differential centrifugation, and a third dialyzable component. Since the activity of the dialyzed enzyme was increased by the addition of phosphorylated hexoses or pentoses, they suggested that the dialyzable factor is a metabolic intermediate. Subsequently, Monod and Rickenberg (1971) reported that one of the protein components (C-11) bound cAMP but did not hydrolyze it. The other protein component ((2-1)catalyzed the reduo tion of glutathione in the presence of NADPH or NADH. They reported that the cAMP hydrolysis took place when C-I and C-I1 were mixed with NADPH, NADH, or a thiol. These investigators proposed that the concentration of cellular reducing equivalents regulates cAMP phosphodiesterase activity. Monod and Rickenberg also reported that a mutant resistant to catabolite repression was defective in component C-I. D. CONTROLOF CATABOLITE REPRESSION OF ~GALACTOSIDASE SYNTHESIS Studies on the control of intracellular levels of cAMP by adenyl cyclase and phosphodiesterase activity suggest a possible mechanism by which a catabolite repressor molecule(s) indirectly regulates the synthesis of enzymes such as pgalactosidase. This model predicts that, during growth on glucose or other carbohydrates that give rise to catabolite repression, metabolites are generated which influence the intracellular concentration of cAMP by interacting with adenyl cyclase and/or phosphodiesterase (de Crombrugghe et al., 1969a; Goldenbaum et al., 1970; Monod et al., 1969). The ability of various metabolites to activate or inhibit these enzymes in oitro suggests that the formation of cAMP may be controlled by intermediary metabolism. This model is illustrated in Fig. 4. Buettner et al. (1973) showed that the transient catabolite repression of pgalao tosidase synthesis that occurs during the initial exposure of a suspension of glycerol-grown E. coli to glucose is not accompanied by a detectable change in the intracellular concentration of CAMP.The basis of transient catabolite repression is obscure, but these results indicate that transient repression may influence the expression of the

REGULATION OF THE LACTOSE OPERON

43

lac operon by means other than alteration of cAMP levels. Although interaction of the hypothetical catabolite repressor with the lac permease to prevent entry of inducer molecules into the system is a possible explanation for transient repression, this possibility appears unlikely since constitutive mutants also are sensitive to transient catabolite repression (Tyler and Magasanik, 1969). Magasanik (1970) has suggested that the catabolite repressor may act directly on the initiation of transcription of the lac operon. However, no evidence is available to support this proposal. 111. In V i m Evidence of Site of cAMP Action A. TRANSCRIPTIONAL OR TRANSLATIONAL CONTROL

Essential to the understanding of the mechanism by which cAMP regulates pgalactosidase synthesis are experiments demonstrating whether or not this cyclic nucleotide exerts its effect at the level of mRNA transcription or translation. Conclusive evidence of a direct effect of cAMP on one or the other depends on experiments in which the two processes can be separated experimentally. This is a difficult condition to achieve in uiuo, since transcription and translation may be coupled processes. These processes may be physically coupled either physiologically, as suggested by Stent (1964), or in the sense of being controlled by regulatory molecules common to both. Although the relationship of transcription to translation has been widely studied, the crucial question whether or not the two processes occur independently in d u o has not been resolved. Most of the evidence obtained with intact cells has supported the suggestion that cAMP acts at the transcriptional level. Nakada and Magasanik (1964), Sells (1965), and Kepes (1963) performed experiments in which transcription of the lac operon was dissociated from translation by exposure of cells to inducer for a very short period of time, followed by removal of the inducer by filtration or dilution. These experiments showed that the presence of glucose inhibited transcription of lac mRNA but did not affect its translation. These results suggest strongly that the site of catabolite repression is the transcription of mRNA. This observation was substantiated by Perlman and Pastan (1968b), who reported that cAMP increased the transcription of mRNA for pgalactosidase in the presence of glucose when protein synthesis was inhibited by chloramphenicol or amino acid starvation. The amount of mRNA accumulated in these studies was judged by the formation of active Pgalactosidase on the removal

44

C. CARPENTER AND B. H. SELLS

of inducer, glucose, CAMP, and protein synthesis inhibition. These experiments, however, do not measure mRNA molecules made and degraded during the inhibition of protein synthesis. Evidence has been presented implying that cAMP influences the translation of pgalactosidase (Paigen and Williams, 1970; Parks et al., 1971) and tryptophanase (Pastan and Perlman, 1968a). These conclusions were based on studies in which proflavin was used to inhibit transcription. Although this dye has no effect on protein synthesis, Conde et al. (1971) reported that cAMP interferes with its inhibition of RNA synthesis. Similar experiments using rifampicin to inhibit transcription have shown, however, that tryptophanase synthesis (del Camp0 et al., 1970; Ramirez et al., 1972) and Bgalao

"[

5O0r A

SECONDS

FIG. 5. Kinetics of accumulation of labeled lac mRNA. Escherkhia coll was

grown to log phase. Thirty milliliters of this culture were placed in each of three

M).and IPTC, glucose, and CAMP flasks. IPTG (5x lo-' M),IPTG and glucose (lo-*M )were added simultaneously to the cultures 5% minutes before labeling. Specimens were removed after 5 minutes for pgalactosidase assay, and 30 seconds later (time zero) 1.5 mCi of uridine3H was added simultaneously to each flask. Samples of 10 ml each were then removed at 30,60,and 180 seconds for RNA extraction. Following measurement of optical density at 260 nm, acid-precipitable counts in 0.5 pg of RNA were determined for each preparation; the specific activities were computed and plotted in (A). RNA species, 0.5 pg each, were then hybridized for 20 hours at 75% with filters containing 0.3 pg of Ah80 or AMOdlac W-labeled in the presence of 22 pg of unlabeled R N A from a lac deletion strain. Counts bound to the hh80 filters were subtracted from counts bound to the kh8Odlac filters, and the differences were assumed to measure lac specific counts in each preparation. These values are plotted in (B)as a function of time. Induced culture (triangles); glucose-repressed culture (circles); CAMP-treated culture (squares). (From Varmus et al., 1970.)

REGULATION OF THE LACTOSE OPERON

45

tosidase synthesis (Jacquet and Kepes, 1969)are controlled at the transcriptional level by glucose and CAMP. The most conclusive evidence that cAMP regulates Pgalactosidase synthesis at the transcriptional level has been provided by experiments that titrate the level of lac mRNA by the DNA-RNA hybridization (Varmus et al., 1970).The effect of glucose and CAMP on the formation of lac mRNA is shown in Fig. 5. From studies by Varmus et al. (1970),which demonstrated that glucose and cAMP do not alter the rate of degradation of lac mRNA, it was concluded that cAMP and glucose monitor the rate of synthesis of mRNA. Also, Miller et al. (1971)showed that the synthesis of mRNA from the galactose (gal) operon is regulated by cAMP and glucose in a manner similar to the control of lac operon mRNA.

B. CAMP AND LAC PROMOTER MUTANTS Following the discovery that cAMP stimulates pgalactosidase synthesis by increasing the level of transcription, attempts were made to define the precise site of action of this cyclic nucleotide. Ullmann and Monod (1968)reported that the effect of cAMP is distinct from the repressor-operator control of the lac operon, since cAMP is active in mutants of both repressor and operator loci. This observation prompted investigation of the promoter region of the lac operon as a possible site of action of CAMP. To initiate these studies, experiments were performed to investigate pgalactosidase synthesis in promoter mutants of the lac operon. Examination of these mutants, originally isolated by Scaife and Beckwith (1966),led to the conclusion that the promoter is distinct from the operator. In support of this belief is the demonstration that promoter mutants are insensitive to catabolite repression, whereas operator constitutive mutations have no effect on catabolite repression. Mapping experiments indicate that the promoter region is located between the lac repressor cistron (i locus) and the lac operator (Ippen et aZ., 1968; Miller et al., 1968). Several laboratories have now concluded that the promoter region is the site of action of catabolite repression and cAMP (Pastan and Perlman, 1968b; Perlman et al., 1969; Silverstone et al., 1969, 1970;Yudkin, 1970). Although data have been presented indicating that catabolite repressors and cAMP do not interfere with lac repressor-operator interaction (Miller et al., 1968;Tao and Huberman, 1970;Yudkin and Reddy, 1971),it is by no means clear from studies in uiuo that the promoter and operator regions are not overlapping areas (Beckwith et aZ., 1972; Smith and Sadler, 1971).Studies by Beckwith et al.

46

C. CARPENTER AND B. H. SELLS

(1972) indicate, however, that there are two distinct sites on the promoter. One of these is a site that normally promotes a low level of lac transcription, possibly by RNA polymerase holoenzyme alone. The second site is one through which cAMP and its binding protein stimulate lac transcription. Since much information pertinent to the function of cAMP and the promoter has been obtained from experiments with cell-free systems, the promoter function is discussed in more detail in Section IV.

IV. CAMPAction in Vitro A. In Vitro SYNTHESISOF ~GALACTOSIDASE Although studies of cAMP action in uiuo have provided considerable insight concerning the regulation of pgalactosidase activity, the validity of conclusions has been limited by the complexity of the intact cell. A detailed analysis of Pgalactosidase synthesis was made possible by the development by Lederman and Zubay (1968) of a DNA-directed system capable of synthesizing active Pgalactosidase molecules in uitro. Although synthesis of pgalactosidase in this system is very inefficient compared to synthesis in d u o , nearly all the effects, that is, repression, induction, observed in uiuo can also be observed in uitro. Whether the effects observed in uitro are produced by similar mechanisms cannot be affirmed beyond doubt, since the conditions present in uiuo may not be faithfully replicated in uitro. A cell-free system has been developed to provide optimal conditions specifically for the synthesis of pgalactosidase. Several modifications of the reaction system have increased the efficiency of /3galactosidase synthesis to a larger extent than the synthesis of total protein, as judged by the incorporation of radioactive amino acid into acid-insoluble material. The most significant advance in the development of this system was the use of bacteriophage DNA which by genetic recombination had E. coli lac DNA incorporated into its genome. Since the phage genome is considerably smaller than the E. coli genome, the use of phage DNA in uitro meant considerable enrichment (1Wfold) of the lac operon in the system. A crude soluble fraction (S-30) of E. coli was used in this system to provide the necessary RNA and protein components required for transcription and translation. The S-30 was prepared from a strain of E. coli carrying a deletion of the lac operon, and thus addition of pgalactosidase to the assay system was avoided.

REGULATION OF THE LACTOSE OPERON

47

Shortly after cAMP stimulation of Pgalactosidase synthesis in intact cells was first reported, Chambers and Zubay (1969; Zubay and Chambers, 1969) examined the effect of this nucleotide on enzyme synthesis in uitro. They showed that addition of 1 mM cAMP to a cell-free system increased the synthesis of Pgalactosidase 30-fold in uitro. At the same time, it also increased the efficiency of repression in oitro by the lac repressor from 50 to 95%. The effect on repression was explained by assuming that the presence of cAMP produces a higher percentage of transcriptional starts at the promoter site. The irrepressible synthesis observed in the absence of cAMP was thought to result from incorrect starts, probably near the beginning of the z cistron. The repression observed in this in uitro system was responsive to the inducer IPTG. In the presence of repressor, cAMP had little effect on Pgalactosidase synthesis. This fact, coupled with the increased efficiency of repression, suggested that cAMP increases the correct initiation of lac mRNA synthesis at the promoter site. That cAMP increases the amount of mRNA synthesized in uitro was demonstrated by de Crombrugghe et al. (1970). Using a DNARNA hybridization technique to detect lac mRNA, they showed that cAMP increased the rate of synthesis of lac mRNA 10-fold. Since cAMP did not increase the overall rate of RNA synthesis (the lac operon accounted for only 5% of the DNA), this constitutes evidence for a specific effect on lac mRNA synthesis. B. MEDIATION OF cAMP ACTION Although cAMP was observed to increase specifically the rate of synthesis of catabolite-repressible enzymes such as Pgalactosidase and galactokinase in uiuo and in uitro, it was difficult to imagine how this mononucleotide was able to recognize a specific DNA region. That another molecule might mediate the cAMP effect was indicated by an analysis of mutants unable to metabolize several carbohydrates, such as lactose, arabinose, and maltose (Perlman and Pastan, 1969). These mutants were divided into two groups based on their response to exogenous CAMP.In the presence of CAMP,one group of mutants was able to utilize the carbon compounds that were not metabolized in the absence of CAMP.The supression of the phenotype of these mutants by exogenous cAMP suggested that these strains are deficient in their ability to synthesize CAMP. Subsequent studies showed that many of these strains do not possess appreciable adenyl cyclase activity. A second class of mutants unable to synthesize catabolite-repressible enzymes did not respond to exogenous cAMP

48

G. CARPENTER AND B. H. SELLS

and appeared to have active adenyl cyclase systems. This observation suggested that the defect in this group of mutants might be in a molecule that mediates the action of cAMP on lac DNA. 1. Discoueqj of CAMP-Receptor Protein Further understanding of this system was provided by the discovery of a protein in the extract of wild-type cells that bound cAMP as reported by Zubay et al. (1970a) and Emmer et al. (1970). The mutants described above that were not responsive to added cAMP were tested for the ability of their cell-free extracts to promote pgalactosidase synthesis in oitro. The soluble fraction (S-30)prepared from these mutants was inactive in uitro, even in the presence of CAMP. However, when the purified CAMP binding protein, designated CAP or CPR, was added to these extracts together with CAMP,the synthesis of /3-galactosidase was stimulated fivefold. No stimulation was observed in the absence of CAMP.If extracts from wild-type cells were used, addition of the cAMP binding protein did not increase the amount of pgalactosidase synthesized in uitro. These results indicated that the mutants unable to synthesize pgalactosidase in the presence of cAMP were defective in this binding protein which was essential for cAMP action. 2. Eflect of the Receptor Protein on CAMPAction in Vitro Perlman et al. (1970) and Arditti and co-workers (1970) demonstrated that the addition of cAMP and its binding protein increased the transcription of lac mRNA in uttro. The cell-free lac system was refined by de Crombrugghe et al. (1971b,c) and Eron et al. (1971; Eron and Block, 1971). The crude extract previously used for pgalactosidase synthesis was eliminated and replaced by purified components which were required for the transcription of lac mRNA. The essential elements of this purified system were double-stranded lac DNA, RNA polymerase containing sigma factor, and CAMP. In this system, transcription of RNA from the DNA template occurred on both strands of the double-stranded DNA template. However, the correct information for the lac operon is contained on only one strand, referred to as the correct strand, and mRNA synthesized from the other strand does not direct the synthesis of active enzymes. Both groups of investigators demonstrated that the addition of cAMP and its binding protein to the reaction mixture greatly increased the percentage of lac mRNA transcribed from the correct strand of the DNA template. Eron et a2. (1971) demonstrated that the sigma factor of

REGULATION OF THE LACTOSE OPERON

49

RNA polymerase was also necessary for stimulation of lac transcription. Sigma factor (Losick, 1972)is a small protein which affects the initiation of transcription of mRNA from certain promoters. The requirement for sigma factor in the lac system suggests that cAMP and its binding protein are not transcriptional factors that replace sigma factor to alter the specificity of RNA polymerase for catabolitesensitive promoters. However, this observation does not preclude the possibility that cAMP and its binding protein may modify RNA polymerase in a way other than replacement of sigma factor. An obvious possibility is that the cAMP binding protein may have a CAMPdependent protein kinase activity. Although protein kinase activity has been demonstrated in E. coli (Kuo and Greengard, 1969),no evidence has been reported indicating that the cAMP binding protein has kinase activity. Nevertheless, these studies have demonstrated that the action of cAMP on the lac operon is mediated by a protein molecule, and that the transcription of lac mRNA is stimulated in uitro by cAMP and its binding protein. It should be realized, however, that these results do not eliminate the possibility that cAMP also influences the translation of lac mRNA, as only transcription occurs in this system. Parks et al. (1971)and Nissley et al. (1971)showed that synthesis of galactokinase in uitro and transcription of the gal operon in uitro are also dependent on cAMP and the same binding protein required for control of the lac operon. C. PROPERTIESOF CAMP-RECEPTORPROTEIN

1. Physical Properties Investigations have been reported by several workers on the characterization of the cAMP binding protein. Anderson et al. (1971)and Riggs et ul. (1971)purified the protein to apparent homogeneity, and from sedimentation equilibrium studies concluded that its molecular weight is 45,000daltons. The protein is composed of two identical subunits as judged by sedimentation equilibrium studies in 6 M guanidine, disc gel electrophoresis in sodium dodecyl sulfate and identical amino terminal amino acid sequences. The binding protein is a basic molecule with an isoelectric point of 9.12and binds cAMP liters mole-'. Although with an association constant of 1.1 X studies suggest that there is one binding site per dimer, they do not rule out the possibility of two sites and a cooperative binding mechanism in which binding at the first site decreases binding affinity at the second site. Emmer et al. (1970)and Zubay et ul. (1970a,b)have

50

G . CARPENTER AND B. H. SELLS

indicated that cAMP is reversibly bound and does not undergo chemical alteration during its attachment to the receptor protein. Based on the results of purification of the binding protein, Anderson et al. (1971) estimated that it constitutes 0.10% of the total protein in E. coli. This suggests that there are approximately 1300 molecules of binding protein per cell in E. coli. Recently, Saunders and McGeoch (1973) isolated a mutant of E. coli that appears to contain an altered cAMP binding protein. This altered binding protein stimulates pgalactosidase synthesis in vitro and in uiuo in the presence of cGMP. Although cGMP has been detected in E. coli (Hardman et al., 1971), no information has been presented defining its biological role. The antagonistic effect of cGMP on several CAMPdependent systems in eukaryotes has recently been discussed (Kolata, 1973). It remains to be determined whether cGMP is physiologically important in controlling CAMP-dependent systems, such as the lac operon, in bacteria. Nissley et al. (1972) investigated the effect of various cAMP analogs on the binding of cAMP to the receptor protein, and on the activity of receptor protein-dependent transcription of the gal operon in uitro. Their results iiidicated that the purine base, the ribose moiety, and the cyclic phosphate group are involved in the binding of the nucleotide to its receptor protein. Thus there is a marked specificity for cAMP to elicit the correct conformational change in the binding protein in order to stimulate transcription. 2. Binding to DNA Interaction of the receptor protein with other macromolecules in the presence and absence of cAMP has been studied. Binding of the receptor protein to RNA polymerase has not been detected (Nissley et al., 1972; Zubay et al., 1970b).The binding of the CAMP-receptor protein to DNA was studied by Riggs and co-workers (1971) and by Nissley et al. (1972). These workers demonstrated that CAMPreceptor protein binds to DNA and that the binding is dependent on the presence of CAMP.The binding is inhibited by cGMP, a competitive inhibitor of CAMP. It has not yet been possible to demonstrate specific binding to catabolite repression-sensitive genes or promoters. Binding has been observed with DNAs from salmon sperm, calf thymus, and chicken blood. The CAMP-receptor protein complex also binds to poly (dAT), denatured DNA, and single-stranded DNA. The binding is specific for DNA, however, as no binding was observed with rRNA or tRNA. Krakow and Pastan (1973) studied the effect of proteolysis on the binding of the receptor protein to DNA.

REGULATION OF THE LACTOSE OPERON

51

They found that receptor protein binds to DNA in a CAMP-dependent manner at pH 8, but in a CAMP-independent manner at pH 6. These investigators suggest that CAMP-independent binding at pH 6 is the result of an increase in the net positive charge of the receptor protein. As mentioned previously, the isoelectric point of the receptor protein is 9.12. If the receptor protein is treated with a proteolytic enzyme (subtilisin, trypsin, or chymotrypsin) in the absence of CAMP,DNA binding at pH 8 or 6 is not affected. However, if proteolysis occurs in the presence of CAMP,binding to DNA at pH 8 is abolished, but the binding at pH 6 is not affected. Gel electrophoresis in sodium dodecyl sulfate indicates that the 22,500dalton subunit of the binding protein is reduced to a 12,500-dalton fragment by proteolysis. Treatment with proteases in the absence of cAMP had no effect on the 22,500-dalton subunit. Krakow and Pastan suggest that cAMP induces a conformational change in the receptor protein which is necessary for DNA binding and which accounts for susceptibility to the action of proteolytic enzymes. Studies of the interaction of cAMP with its receptor protein and the interaction of this complex with DNA in uitro will undoubtedly provide considerable information on the mechanism of gene expression. D. INITIATIONOF

LAC

mRNA SYNTHESISin Vitro

1. Effect of Lac Promoter Mutations Eron and Block (1971) showed that when DNA prepared from bacteriophage DNA harboring promoter mutations of the lac operon is used as a template for lac transcription in vitro, lac mRNA is synthesized in a manner that reflects, qualitatively, the effect of these promoter mutations on Pgalactosidase synthesis in uiuo. When a preparation of DNA carrying a partial deletion of the lac promoter was used, 3-to 10-fold less lac mRNA was produced compared to wildtype lac DNA. The level of lac RNA was not increased by the addition of cAMP or cAMP binding protein to the reaction mixture. In viuo this mutation results in a 100-fold reduction in Pgalactosidase synthesis. These results suggest that the deletion in this mutant, L1, may have removed a portion of the lac promoter involved in the binding of CAMP-receptor protein. Eron and Block (1971) also examined the effect in oitro of promoter mutations that increase the synthesis of Pgalactosidase in viuo. When used to direct the transcription of lac RNA in vitro, these “superpromoter” mutations yielded increased levels of mRNA for the lac operon.

52

G. CARPENTER AND B. H. SELLS

These results indicate that the levels of mRNA from the lac operon are regulated by the nucleotide sequence of the promoter region and the CAMP-receptor protein complex in uitro as well as in uiuo. 2. Formation of Rifampicin-Resistant Znitiation Complexes

De Crombrugghe et al. (1971b)have provided evidence indicating that CAMP and the receptor protein are required for the binding of RNA polymerase to the lac promoter and the formation of a rifampicin-resistant transcription initiation complex. Rifampicin is known to bind to the beta subunit of E. coli RNA polymerase and thereby prevent the initiation of RNA synthesis. In the absence of rifampicin and nucleotides, however, RNA polymerase binds to DNA and forms an initiation complex capable of transcribing RNA when the nucleotides are added together with rifampicin. De Crombrugghe et al. (1971b) found that, when CAMPand the receptor protein were added together with RNA polymerase to DNA, transcription of RNA increased approximately 30-fold when the nucleotides and rifampicin were added to the system. This indicates that the cAMP-receptor protein complex increases the transcription of lac RNA by increasing the binding of RNA polymerase to DNA. It is possible that the CAMP-receptor protein complex may increase the binding of RNA polymerase to DNA by either of two mechanisms. The effector complex might interact with the polymerase in such a manner as to increase the affinity of the polymerase for proper DNA nucleotide sequences. This is unlikely, however, as no physical or enzymic interactions between the effector complex and RNA polymerase have been detected. As previously mentioned, the CAMP-receptor protein complex binds to DNA. This suggests that the effector complex may act by affecting the DNA binding sites for RNA polymerase in a positive manner. De Crombrugghe et al. (1971a) analyzed the binding of RNA polymerase and the effector complex to DNA by varying the concentrations of molecules in the formation of rifampicin-resistant initiation complexes. When the RNA polymerase concentration was held constant and the amount of DNA vaned, the concentration of CAMPreceptor protein required for maximal lac transcription was directly proportional to the amount of DNA. This suggests that the CAMPreceptor protein complex binds to DNA in a manner independent of RNA polymerase. De Crombrugghe et al. (1971a) also investigated the binding of RNA polymerase and CAMP-receptor protein to wild-type lac DNA and to DNA carrying a superpromoter mutation in the lac operon. In

REGULATION OF THE LACTOSE OPERON

53

the presence of high levels of CAMP-receptor protein and a fixed amount of DNA, the concentration of RNA polymerase needed for maximal transcription from each template was nearly identical, although twice as much RNA was transcribed from the template carrying the promoter mutation. However, when the levels of polymerase and DNA were held constant, higher levels of transcription of lac RNA from the DNA with the promoter mutation were achieved at a significantly lower concentration of CAMP-receptor protein than was required with the DNA template carrying a wild-type promoter. Based on these observations de Crombrugghe and co-workers (1971a) have proposed that the following sequence of steps occurs during the initiation of lac transcription:

+

1. CAMP CRP + CAMP-CRP 2. CAMP-CRP lac DNA + CAMP-CRP-lac DNA 3. CAMP-CRP-lac DNA RNA polymerase + CAMP-CRP-lac DNA-RNA polymerase

+

+

The third step is probably a complicated process involving several reactions, and additional steps may be required prior to the actual formation of nucleotide bonds. The molecular mechanisms by which the CAMP-receptor protein alters the conformation of DNA so as to increase the binding of polymerase are not known. Beckwith and co-workers (1972) have presented genetic evidence suggesting that the promoter region of the lac operon contains two sites. They suggest that an operator distal site in the promoter, defined by the promoter deletion L1, is involved in the binding of the CAMP-receptor protein complex to the DNA template. An operator proximal site in the promoter is proposed to be the binding site for RNA polymerase. This proposal predicts that promoter mutants exist which affect the binding of RNA polymerase to the promoter but which do not affect the binding of CAMP-receptor protein to the DNA. To date, no such mutants have been reported.

3. Effect of the Repressor on Lac Transcription An important consideration in the regulation of the lac operon is the possible overlap of the operator and promoter regions. As pointed out by Beckwith et al. (1972), promoter mutants analyzed to date are those that affect the CAMP-receptor protein binding site in the promoter, and therefore cannot be used to define the operator proximal portion of the promoter. Smith and Sadler (1971) isolated

54

G . CARPENTER AND B. H. SELLS

presumed operator mutants that affect the level of lac operon expression. This suggests a possible overlap between the operator and promoter regions of the E . coZi lac operon. Chen et d.(1971) studied the effect of the lac repressor on the binding of RNA polymerase to lac DNA in uitro. Their results indicate that the repressor and polymerase bind to DNA independently. However, competitive binding between repressor and polymerase occurred when DNA with a superpromoter mutation in the lac operon was used. Since a single mutation can bring about an overlap of repressor and polymerase binding sites, these two sites must be very close if not directly adjacent to each other.

V. Additional Aspects of Lac Operon Regulation Although the CAMP-controlled initiation of mRNA synthesis at the lac promoter undoubtedly plays a major role in regulating the expression of the lac operon, it should be pointed out that other mechanisms may be involved in this complex process of gene activity. To date these additional processes are not well understood, and their physiological significance remains to be established. Therefore these areas are not within the scope of this article and are mentioned only briefly. Sequence studies of the mRNAs from the trp (Bronson et al., 1973) and lac (Maizels, 1973) operon have been reported. These studies indicate that the translation initiation sites of both messengers are preceded by large nucleotide sequences. It is not known whether these sequences have a role in regulating the expression of these operons. Numerous investigators have examined the proposed coupling of the transcription and translation processes in E. co2i. These studies of metabolic integration have not yielded conclusive results as to the possible regulatory role of protein synthesis in the synthesis of mRNA. Obviously, if the two processes are coupled in a physical manner as proposed by Stent (1964), or coupled by common regulatory elements, any process that affects protein synthesis exerts an important influence on transcription. The mechanism by which the expression of the lac operon results in a natural polarity of enzyme synthesis is not known (Miller, 1970). Polarity refers to the fact that equal amounts of enzyme are not synthesized from the three structural genes of the lac operon. The amount of enzyme synthesized decreases as the distance from the

REGULATION OF THE LACTOSE OPERON

55

site of RNA initiation increases. Since the structural genes are transcribed as one unit of RNA, this polar effect must be due to a differential translation of different segments of the mRNA. This could result from the secondary structure of the RNA, increased degradation of the 3’ end of the message compared to the 5’ end, or termination of transcription within the operon. Evidence has been presented from various systems to substantiate each of these explanations. However, the evidence does not yet appear conclusive. REFERENCES Aboud, M., and Burger, M. (1970). Biochem. Biophys. Res. Commun. 38,1023. Aboud, M., and Burger, M. (1971a).Biochem. J. 122,219. Aboud, M., and Burger, M. (1971b). Biochem. Biophys. Res. Commun. 43,174. Abou-Sabe’, M. A. (1973). Nature (London),New Biol. 243, 182. Abou-Sabe’, M. A., and Nardi, R. (1973). Biochem. Biophys. Res. Commun. 51, 551. Adamson, L.,Gross, C., and Novick, A. (1970).In “The Lactose Operon” (J.R. Beckwith and D. Zipser, eds.), Dp. 27-47. Cold Spring Harbor Lab., Cold Spring Harbor, New York. Anderson, W. B., Schneider, A. B., Emmer, M., Perlman, R. L.,and Pastan, I. (1971).J. Biol. Chem. 246,5929. Arditti, R. R.,Eron, L.,Zubay, C., Tocchini-Valentini, G.,Conway, S., and Beckwith, J. (1970). Cold Spring Harbor Symp. Quant. Biol. 35,419. Beckwith, J . R. (1970). In “The Lactose Operon” (J. R. Beckwith and D. Zipser, eds.), pp. 5-26. Cold Spring Harbor Lab., Cold Spring Harbor, New York. Beckwith, J. R., Grodzickev, T., and Arditti, R. (1972).J . Mol. Biol. 69, 155. Benson, C. E., Brehmeyer, B. A,, and Gots, J. S . (1971). Biochem. Biophys. Res. Commun. 43,1089. Bourgeois, S. (1971). Cum Top. Cell. Regul. 4,39. Broman, R. B., and Dobrogosz, W. J. (1972). Bacteriol. Proc. p. 176. Bronson, M. J., Squires, C., and Yanofsky, C. (1973). Proc. Nut. Acad. Sci. U . S . 70, 2335.

Buettner, M. J., Spitz, E., and Rickenberg, H. V. (1973).J . Bacteriol. 114,1068. Burstein, C., Cohn, M., Kepes, A., and Monod, J. (1965). Biochim. Biophys. Acta 95, 634. Chambers, D. A., and Zubay, G. (1969).Proc. Nut. Acad. Sci. U . S . 63,118. Chen, B., de Crombrugghe, B., Anderson, W. B., Gottesman, M.E.,Perlman, R. L., and Pastan, I. (1971). Nature (London),New Biol. 233,67. Conde, F., del Campo, F. F., and Ramirez, J. M. (1971). FEBS (Fed. Eur. Biochem. SOC.), Lett. 16, 156. de Crombrugghe, B., Perlman, R. L.,and Pastan, I. (1969a).]. Biol. Chem. 244,5828. de Crombrugghe, B., Perlman, R. L., Vannus, H. E., and Pastan, I. (1969b).J. Biol. Chem. 244,5828. de Crombrugghe, B., Varmus, H.E., Perlman, R. L., and Pastan, I. (1970). Biochem. Biophys. Res. Commun. 38,894. de Crombrugghe, B.. Chen, B., Anderson, W. B., Gottesman, M. E., Perlrhan, R. L., and Pastan, I. (1971a).J . Biol. Chern. 246, 7343. de Crombrugghe, B., Chen, B., Anderson, W. B., Nissley, P., Gottesman, M. E., Pastan, I., and Perlman, R. L. (1971b). Nature (London),New Biol. 231,139.

56

G. CARPENTER AND B. H. SELLS

de Crombrugghe, B., Chen, B., Gottesman, M., Pastan, I., Varmus, H. E., Emmer, M., Nature (London),New Biol. 230,37. and Perlman, R. L. (1971~). de Crombrugghe, B., Shaw, B., Rosner. J., and Pastan, I. (1972).Nature (London), New Biol. 241,237. del Campo, F. F., Ramirez, J. M., and Canovas. J. L. (1970).Biochem. Biophys. Res. Commun. 40,77. Emmer, M., de Crombrugghe, B., Pastan. I., and Perlman, R. L. (1970).Proc. Nut. Acad. Sci. U.S. 66,480. Eron, L. and Block, R. (1971).Proc. Nut. Acad. Sci. U.S. 68,1828. Eron, L., Arditti, R.,Zubay, G., Connaway, S.,and Beckwith, J. R. (1971).Proc. Nut. Acad. Sci. U.S. 68,215. Fox, C. F., Beckwith, J. R., Epstein, W., and Signer, E. R. (1966).J. Mol. Biol. 19,576. Gilbert, W.,and Muller-Hill. B. (1970).In “The Lactose Operon” (J. R. Beckwith and D. Zipser, eds.), pp. 93-109. Cold Spring Harbor Lab., Cold Spring Harbor, New York. Goldenbaum, P. E., Broman, R. L., and Dobrogosz. W. J. (1970)./. Bacteriol. 103,663. Crodzicker, T.,Arditti, R. R.,and Eisen, H. (1972).Proc. Nut. Acad. Sci. U.S. 69,366. Hardman, J,, Robinson, G., and Sutherland, E. (1971).Annu. Reo. Physbl. 33, 311. Harwood, J., and Smith, D. H.(1971).Biochem. Bfophyr. Res. Commun. 42,57. Hempfling, W. P., and Breman, D. K. (1971).Biochem. Biophys. Res. Commun. 45, 924. H h t a . M., and Hayaishi, 0. (1965). Biochem. Biophys. Res. Commun. 21,361. Hirata, M., and Hayaishi, 0.(1967).Biochim. Biophys. Acta 149, 1. Ide, M. (1969).Biochem. Biophys. Res. Commun. 36,42. Ide, M. (1971).Arch. Biochem. Biophys. 144,262. Ide, M., Yoshimoto, A., and Dkabayashi, T.(1967).J.Bacteriol. 94,317. Ippen, K., Miller, J. H., Scaife, J., and Beckwith, J. (1968).Nature (London) 217, 825. Jacquet, M., and Kepes, A. (1969).Biochem. Biophys. Res. Commun. 36,84. Jobe, A,, and Bourgeois, S. (1972).J. Mol. Biol. 69,397. Jobe, A., and Bourgeois, S. (1973).J . Mol. Biol. 75,303. Jost, J., and Rickenberg, H. V. (1971).Annu. Reu. Biochem. 40,771. Kak, L., Kingsburry, D. T.,and Helinski, D. R. (1973).J . Bacterfol. 114,577. Kennedy, E. P. (1970).In “The Lactose Operon” (J.R. Beckwith and D. Zipser, eds.), pp. 49-92. Cold Spring Harbor Lab., Cold Spring Harbor, New York. Kepes. A. (1960). Biochim. Biophys. Acta 40,70. Kepes, A. (1963).Biochim. Biophys. Acta 76,293. Kepes, A. (1971).J . Membrane Biol. 4,87. Khandelwal, R. L., and Hamilton, I. R. (1971).J . Biol. Chem. 426,3297. Khandelwal, R. L., and Hamilton, I. R. (1972).Arch. Biochem. Biophys. 151,75. Khandelwal, R. L.. Spearman, T. N., and Hamilton, I. R. (1973).FEBS (Fed. Eur. Biochem. SOC.), Lett. 31,246. Kolata, G. B. (1973).Science 182,149. Krakow, J. S., and Pastan, 1. (1973).Proc. Nut. Acad. Sci. U.S. 70,2529. Kuo, J. F., and Greengard, P. (1969).J . Biol. Chem. 244,3417. Lederman, M., and Zubay. G. (1968).Biochem. Biophys. Res. Commun. 32,710. Losick, R. (1972).Annu. Reo. Biochem. 41,409. McFall. E. (1973).J . Bacteriol. 113,781. Magasanik, B. (1961).Cold Spring Harbor Symp. Quant. Biol. 26,249. Magasanik, B. (1970).In ‘The Lactose Operon” (J.R. Beckwith and D. Zipser, eds.), pp. 189-219.Cold Spring Harbor Lab., Cold Spring Harbor, New York.

REGULATION OF THE LACTOSE OPERON

57

Maizels, N. M. (1973). Proc. Nut. Acad. Scf. U.S . 70,3585. Makman, R. S., and Sutherland, E. W. (1963). Fed. Proc., Fed. Amer. Soc. E r p . Biol. 22, 470.

Makman, R. S., and Sutherland, E. W. (1965). J . Biol. Chem. 240,1309. Miller, J. H.(1970). In “The Lactose Operon” (J. R.Beckwith and D. Zipser, eds.), pp. 173-188. Cold Spring Harbor Lab., Cold Spring Harbor, New York. Miller, J. H., Ipper, K, Scaife, J. G., and Beckwith, J. R. (1968). J . Mol. B i d . 38,413. Miller, Z., Vannus, H. E., Parks, J. S., Perlman, R. L., and Pastan, I. (1971). J. Biol. Chem. 246,2898. Monod, D., and Rickenberg, H. V. (1971). Bacteriol. Proc., p. 155. Monod, D., Janecek, J., and Rickenberg, H. V. (1969). Biochem. Biophys. Res. Commun. 35,584. Monod, D., Janecek, J., and Rickenberg, H.V. (1970). In “The Lactose Operon” (J. R. Beckwith and D. Zipser, eds.), pp. 393-400. Cold Spring Harbor Lab., Cold Spring Harbor, New York. Miiller-Hill, B., Rickenberg, H. V., and Wallenfels, K. (1964). J. Mol. Biol. 10, 303. Nakada, D., and Magasanik, B. (1964). Biochtm. Biophys. Acta 61,835. Nakazawa, A., and Tamada, T. (1972). Biochem. Biophys. Res. Commun. 49, 977. Nakazawa, T., and Yokota, J. (1973). J . Bacteriol. 113, 1412. Nealson, K. H., Eberhard, A., and Hastings, J. W. (1972). Proc. Nut. Acad. Sci. U.S . 69, 1073.

Nissley, S. P., Anderson, W. B., Gottesman, M. E., Perlman, R. L., and Pastan, I. (1971). J . Biol. Chem. 246,2419. Nissley, P., Anderson, W.B., Gallo, M., Pastan, I., and Perlman, R. L. (1972). J . Biol. Chem. 247,4264. Okabayashi, T., and Ide, M. (1970). Biochim. Biophys. Acta 220,116. Okabayashi, T., Yoshimoto, A., and Ide, M. (1963). J . Bacteriol. 86,930. Paigen, K., and Williams, B. (1970). Adoan. Microbial Physiol. 4,252. Parks, J. S., Gottesman. M., Perlman, R. L., and Pastan, 1. (1971). J . Biol. Chem. 246, 2,419.

Pastan, I., and Perlman, R. L. (1968a). J . Biol. Chem. 244,2226. Pastan, I., and Perlman, R. L.(1968b). Proc. Nut. Acad. Sci. U.S . 61, 1336. Pasta, I., and Perlman, R. L. (1970). Science 169,339. Pastan, I., and Perlman, R. L. (1971). Nature (London),New Biol. 229,s. Perlman, R. L., and Pastan, I. (1968a). Biochem. Biophys. Res. Commun. 30, 656. Perlman, R. L., and Pastan, I. (1968b).J. Biol. Chem. 243,5420. Perlman, R. L., and Pastan, I. (1969). Biochem. Biophys. Res. Commun. 37,151. Perlman, R. L., and Pastan, I. (1971). Cum. Top. Cell. Regul. 3, 117. Perlman, R. L., de Crombrugghe, B., and Pastan, I. (1969). Nature (London) 223,810. Perlman, R. L., Chen, B., de Crombrugghe, B., Emmer, M., Gottesman, M., Varmus, H., and Pastan, I. (1970). Cold Spring Harbor Symp, Quant. Biol. 35,419. Peterkofsky, A,, and Gazdar, C. (1971). Proc. Nut. Acad. Sci. U.S . 68,2794. Peterkofsky, A,, and Gazdar, C. (1973). Proc. Nut. Acad. Sci. U.S . 70,2149. Prusiner, S., Miller, R. E., and Valentine, R. C. (1972). Proc. Nut. Acad. Sci. U.S . 69, 2922.

Rall, T. W., and Sutherland, E. W. (1958). J . Biol. Chem. 232, 1065. Rall, T. W., Sutherland, E. W., and Berthet, J. (1957). J . Biol. Chem. 224,463. Ramirez, J. M., Conde, F., and del Campo, F. F. (1972). Eur. J . Biochem. 25, 471. Reznikoff, W. S. (1972). Annu. Reu. Genet. 6, 133. Riggs, A. D., Reiness, G., and Zubay, G. (1971). Proc. Nut. Acad. Sci. U . S . 68, 1222.

58

G. CARPENTER AND B. H. SELLS

Robinson, G. A., Butcher, R. W., and Sutherland, E. W. (1968).Annu. Reu. Biochem.

37,149.

Saunders, R., and McGeoch, D. (1973).Proc. Nat. Acad. Sci. U.S . 70, 1017. Scaife, J., and Beckwith, J. R. (1966).Cold Spring Harbor Symp. Quant. Biol. 31,403. Sells, B. H. (1965).Science 148,371. Shizuta, Y., and Hayaishi, 0. (1970).J.Biol. Chem. 245,5416. Silverstone, A. E.. Magasanck, B., Reznikoff, W. S.,Miller, J. H., and Beckwith, J. R. (1969).Nature (London) 221,1012. Silverstone, A. E.. Arditti, R., and Magasanik, B. (1970).Proc. Nat. Acad. Sci. U.S . 66,

773.

Smith, T. F., and Sadler, J. R. (1971).J . Mol. Bfol. 59,273. Stent, G.S. (1964).Science 144,816. Sutherland, E. W., and Rall, T. W. (1957).J. Amer. Chem. Soc. 79,3608. Sutherland, E. W., and Rall, T. W. (1958).J . Biol. Chem. 232,1077. Sutherland, E. W., @ye, I., and Butcher, R. W.(1965).Recent Progr. Horn. Res. 21,

623.

Sy, J.. and Richter, D. (1972).Biochemistry 11,2788. Tao, M., and Huberman, A. 11970).Arch. Biochem. Biophys. 141,236. Tao, M., and Lipmann, F. (1969).Proc. Nat. Acad. Sci. U.S. 63,86. Tao, M., and Schweiger, M. (1970).J . Bacterfol. 102,138. Tyler, B., and Magasanik, B. (1969). 1. Bacteriol. 97,550. Ullmann, A,, and Monod, J. (1968).FEBS (Fed. Eur. Biochem. Soc.) Lett. 2,57. Varmus, H. E.,Perlman, R. L., and Pastan, I. (1970).J . Biol. Chem. 245,2259. Wetekam, W.. Staack, K., and Ehring, R. (1971).Mol. Gen. Genet. 112,14. Wise, E. M., Alexander, S. P., and Powers, M.(1973).Proc. Nut. Acad. Sci. U.S. 70,

471.

Yang, H. L., and Zubay, G. (1973).Mol. Gen. Genet. 122,131. Yokota, T., and Cots, J. S. (1970).J. Bacteriol. 103,513. Yudkin, M. D.(1970).Biochem. J. 118,741. Yudkin, M. D.,and Reddy, S. (1971).Mol. Gen. Genet. 113,297. Zabin, I., and Fowler, A. V. (1970).In “The Lactose Operon” (J. R. Beckwith and D. Zipser, eds.), pp. 27-47. Cold Spring Harbor Lab., Cold Spring Harbor, New York. Zubay, G., and Chambers, D. A. (1969).Cold Spring Harbor Symp. Quant. Bid. 34,

753.

Zubay, G., and Chambers, D. A. (1971).In “Metabolic Regulation” (H. J. Vogel, ed.) Metabolic Pathways, Vol. 5, pp. 297-347. Academic Press, New York. Zubay, G., Schwartz, D., and Beckwith, J. R. (1970a).Proc. Nat. Acad. Sci. U. S. 66,

104.

Zubay, G., Schwartz, D., and Beckwith, J. R. (1970b).Cold Spring Harbor Symp. Quant. Biol. 1,433.

Regulation of Microtubules in Tetrahymena' NORMAN E. WILLIAMS Department of Zoology, Unioersity of Iowa, Zowa City, Zowa

I. Introduction . . . . . . . . . . . . . 11. Regulatory Patterns and the Cell Cycle A. The Somatic Ciliature in the Cell Cycle . . . . . B. The Oral Apparatus in the Cell Cycle. C. Nuclear Microtubules . . . . . . . . D. Cycle-Independent Formation and Regression of Micro. . . . . . . . . . . . . tubules. 111. Microtubule Stability and Regression A. Somatic and Oral Microtubules . . . . . . . B. Nuclear Microtubules . . . . . . . . . . C. Conclusion. IV. The Dynamic Nature of Formed Microtubules . . . . A. Evidence in Tetrahymena . . . . . . . . . B. Turnover and Subunit Exchange . . . . . . . V. Control of Microtubule Formation. A. Tubulin Synthesis in Relation to Microtubule Formation B. Synthesis-Dependent Assembly of Preexisting Tubulin VI. Epilog: The Cell Cycle Revisited. References

. . . . . . . . . . .

. . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

59 60 60 63

65 66 68 69 72 73 74 74 75 77

77 79 83 84

I. Introduction The recent development of in uitro systems for the study of microtubule assembly (Weisenberg, 1972; Borisy and Olmsted, 1972; Shelanski et d., 1973) offers an exciting and promising approach to the study of problems having fundamental significance in cell biology. However, in uiuo studies of microtubules should continue to be of importance for the total solution of many of these problems. Living cells possess the ability to form microtubules in specific places at specific times, and also to resorb them in a similarly controlled fashion. A combined approach, using both in uiuo and in vitro studies, will probably be required to develop a satisfactory understanding of the mechanisms by which cells regulate microtubules and microtubule complexes. This discussion is a review of the information available concerning the regulation of microtubules in the ciliate Tetrahymena pyraformis. The formation and regression of micro-

' Supported by

grant number GB-41389 from the National Science Foundation. 59

60

NORMAN E. WILLIAMS

tubules in Tetruhymena is detailed first within the context of the cell cycle. Following this, general problems are discussed in relation to microtubule regulation in other systems. 11. Regulatory Patterns and the Cell Cycle

A. THE SOMATIC CILIATUREIN

THE

CELL CYCLE

The term somatic ciliature refers to the total body ciliature exclusive of the oral apparatus, or feeding structure. This system is extensive in Tetruhyrnenu, as is seen in Figs. 1 and 2. It is a major repository of microtubules. Each somatic cilium, containing the familiar 9 2 array of microtubules, extends outward from a basal body, or kinetosome, located beneath the surface within the cell cortex. Here

+

FIG.1. Scanning electron micrograph of T. pyrffonnfs showing the oral apparatus and somatic cilia, The bar indicates 10 q. Micrograph provided by J. J. Ruffolo, Jr.

REGULATION OF MICROTUBULES IN

Tetmhymena

61

FIG.2. Tetruhyrnenu late in the cell cycle, showing somatic cilia, the anterior oral apparatus, and a newly forming oral apparatus (lower left) with many short cilia. Micrograph provided by J. J. Ruffolo, Jr.

each basal body is associated in a regular way with other structural elements, including precisely oriented bundles of additional cortical microtubules. A basal body, together with its attached cilium and associated cortical elements, is called a ciliary unit. A study by Allen (1967) provides a useful three-dimensional reconstruction of the somatic cortex of Tetrahymena. 1. Number and Distribution of Ciliary Units There are typically 17 to 19 longitudinal rows of somatic cilia (meridians or kineties) in T. pyrifomis, as can best be seen in the denuded cell in Fig. 4. Two of these, the ones abutting on the oral apparatus, are shorter than the remaining ones. Cells within a given strain may have fewer or more than 17 to 19 ciliary rows. This type of variation within a population increases in response to certain (usually deleterious) environmental changes. However, most cells

62

NORMAN E. WILLIAMS

return to near 18 rows (the stability center) under optimal growing conditions. Although the stability center is likely under genic control in Tetrahymena, as it is in the ciliate Euplotes (Heckmann and Frankel, 1968; Frankel, 1973), a genetic analysis of deviations from the stability center suggests that mechanisms of cortical inheritance operate as well in the determination of ciliary row number (Nanney, 1966). The first estimate of the total number of somatic ciliary units in Tetrahymena, by Furgason (1940), was close to 800. The strain he used is no longer available. An estimate in the commonly used strain GL can be made from the counts of Williams and Scherbaum (1959); here the total appears to progress from about 600 to 1200 over the cell cycle. Data on a strain in syngen 1by Nanney (1971a)provide an estimate of from about 450 to 900 over the cell cycle. Counts of basal bodies in cells at a common stage in the cell cycle in 24 strains representing 12 syngens (Nanney and Chow, 1974) show that considerable variation in total basal body number exists among strains. A point of great interest is the finding that the total number of ciliary units is apparently regulated independently of the number of ciliary rows. Nanney (1971b) has shown that cells of a single strain with different row numbers (from 16 to 25) show nearly equivalent numbers of total ciliary units. 2. Pattern of Increase and Segregation New ciliary units are added adjacent to old units within the rows. Basal bodies appear first, immediately anterior to old basal bodies. Next, they move anteriorly to occupy a position midway between old ciliary units, and then sprout cilia (Williams and Scherbaum, 1959). The process has been described at the ultrastructural level by Allen (1969). Probasal bodies form in what is now considered classic fashion, elongate, and tilt to contact the cell surface with their distal ends. Other cortical elements which are part of the ciliary unit begin to form during tilting. Ciliary outgrowth is the final event. An early morphological study of the pattern of increase in somatic ciliary units in Tetrahgmena (Williams and Scherbaum, 1959) showed that developing units are found in all regions of the body and at all times in the cell cycle. Detailed quantitative analyses have been carried out subsequently by Nanney (1971a) and Perlman (1973). The conclusion that emerges from these studies is that, to a first approximation, the pattern of increase is uniform in time (throughout the cell cycle) and space (all ciliary rows and regions within rows). There appears to be, however, a rate change in two of

REGULATION OF MICROTUBULES IN Tetruhymenu

63

the rows near the new mouth during oral apparatus formation, and a slight overall reduction in the rate of increase within the cell at this time. Light microscopic observations show that equatorial breaks occur in the somatic ciliary rows beginning about three-quarters of the way through the cell cycle (Fig. 3); this has not yet been characterized at the ultrastructural level. The fission furrow forms subsequently, and the half-rows are segregated into the daughter cells.

B. THE ORAL APPARATUS I N THE CELL CYCLE Another 160 to 170 basal bodies, also associated with additional microtubules and fibrous material intracellularly, are found in the oral apparatus of T. pydfomis (Nilsson and Williams, 1966;Forer et ul., 1970; reviewed in Elliott and Kennedy, 1973). These basal bodies, most of them ciliated, occur in four groups, thus constituting the four compound ciliary structures of membranous nature from which Tetruhymenu derives its name. Three of the oral ciliary groups, called membranelles, are found along the (cell’s) left wall of

T.P

I V/BA 0

0.68

0.78

1.00

FIG. 3. Outline of major events in the cell cycle of T.p y z i f o n i s GL. Formation of the oral apparatus for the posterior division product is indicated in the cell outlines at the top. The three membranelles (M) and the single undulating membrane (UM) of the oral apparatus are designated on the left. Macronuclear development is outlined in the lower figures. The approximate times in the cell cycle at which the given configurations occur are indicated by the bar at the bottom. The transition point (T.P.), which occurs in strain GL at about 78% of the way through the cell cycle, is the time of an abrupt transition from high to low sensitivity to a wide range of chemical and physical agents (discussed further in the text). The shaded zone indicates the part of the cell cycle during which oral primordium resorption can readily be induced.

64

NORMAN E. WILLIAMS

the oral cavity (M in Fig. 3).Each consists of three tightly packed rows of ciliated basal bodies; the outside membranelle (Ml)has about 57 cilia, the middle one (M2)has about 42,and the inner one (M3)has about 15.Membranellar cilia can be seen in the center of the old oral apparatus in Fig. 1, and in the new oral apparatus in Fig. 2.It is the undulating membrane that has nonciliated basal bodies. This structure is found along the right margin of the oral cavity, as indicated in Fig. 3 (labeled UM). It consists of a single row of about 27 ciliated basal bodies (to the viewer’s left in Figs. 1 and 4)and an adjacent row of about 27 nonciliated basal bodies next to the row bearing the cilia. The extensive tubular and filamentous interconneo tions within the oral apparatus at the level of the basal bodies (see Elliott and Kennedy, 1973,Fig. 14)permit its isolation as a cell fraction.

1. Changes in the Old Oral Apparatus No growth or morphological change occurs in the old oral apparatus of T . pyrffomis during the first three-quarters of the cell cycle. Although there have been no detailed quantitative studies, casual observation suggests that the overall size of the oral apparatus does not change in the cell cycle, nor does it change even during starvation (cf. Frankel, 1970, Figs. 10 and 12). Constant size is also suggested by the fact that the basal body counts that have been made (see above), although not numerous, have been fairly uniform. If any change occurs, it is probably slight. Extensive alterations occur in the old mouth, however, during the last quarter of the cell cycle. The old oral apparatus comes to lie on the surface of the cell, loses some fibers and the undulating membrane cilia, and shows a shortening of the membranellar cilia (see Frankel and Williams, 1973;Buhse et al., 1973).These remarkable regressive changes bring the old mouth to the precise state of the newly forming mouth in the cell at this time. Subsequently, during cleavage, the old oral apparatus redevelops the missing components in synchrony with completion of development of the new oral apparatus. The adaptive significance of these changes in the old oral apparatus during division is a mystery; perhaps there is some undetected growth, or deterioration, during the cell cycle which must be compensated for.

2. Formation of the New Oral Apparatus The old oral apparatus is inherited by the anterior division product, whereas the new oral apparatus is formed for the posterior

REGULATION OF MICROTUBULES IN

Tetrahymena

65

division product in the midequatorial region of the parent cell cortex (Figs. 2 and 3). The first sign of oral apparatus formation is the appearance of a stomatogenic field of unciliated basal bodies. This occurs just prior to the half-way point in the cell cycle. The site of formation is nearly always adjacent to the right member of the pair of somatic ciliary rows which abut on the old mouth, in the manner shown in Fig. 3. This row is the so-called stomatogenic meridian. Ciliary growth begins early. Next, the nascent ciliary units move into double files to form three membranellar anlagen; the third row in each membranelle is added later. Undulating membrane formation starts later than membranelle formation, but the two processes overlap. The development described thus far occurs on the surface of the cell. The oral cavity is formed subsequently, along with various associated structures. Electron micrographs of the developing oral apparatus can be found in a recent article by Williams and Frankel (1973);the reader is referred to Frankel and Williams (1973) for further details of stomatogenesis and references to the literature. C. NUCLEARMICROTUBULES Tetrahymena typically have two different nuclei. One of the two types, the macronucleus, is hyperpolyploid, yet shows no visible chromosomes. This nucleus is found in all viable cells. The other, the micronucleus, is diploid, and shows chromosomes in both mitosis and meiosis, yet is absent in many strains. Unfortunately, most studies of microtubules and the cell cycle have been made with amicronucleate strains. Such a strain is GL, shown in Fig. 3. Most workers have failed to find microtubules in the interphase macronucleus (Elliott and Kennedy, 1973). Macronuclear microtubules appear in abundance, however, during macronuclear elongation. This process begins just prior to the onset of cytoplasmic cleavage, proceeds rapidly, and ends in the segregation of daughter nuclei. Macronuclear microtubules were first seen by Roth and Minick (1961) who described them as fibrillar elements in the dividing nucleus. Later workers showed them to be tubular fibers, and described their orientation. They have been described by Falk et al. (1968)and It0 et al. (1968)as occurring singly or in bundles, with a tendency to occur peripherally and show attachment to the inner nuclear membrane. Attachments to chromatin and nucleoli were also frequent, which led both groups of investigators to suggest that the microtubules may serve to anchor these structures to the nuclear membrane, Both groups also suggested that this might in some way assure the independence and/or segregation of diploid subnuclei.

66

NORMAN E. WILLIAMS

The membrane attachment of nuclear microtubules has been confirmed by Wunderlich and Speth (1970); Tamura et al. (1969) have noted that the tubules tend to be oriented along the long axis of the dividing macronuclei. The correlation between the appearance of microtubules and the elongation of the macronucleus suggests a possible role of microtubules in this process, and there is some experimental evidence for this (see Section 111).Other postulated functions remain, as do many things about the ciliate macronucleus, highly conjectural. The micronucleus, when present, divides mitotically within the persisting micronuclear membrane. Micronuclear and macronuclear division are well separated in time; the former begins earlier, about two-thirds of the way into the cell cycle, and is completed about the time macronuclear division begins. Elliott and Kennedy (1973) showed the presence of microtubules within the dividing micronucleus of Tetruahymena. There are few descriptive (and no experimental) data available on these tubules. AND REGRESSION OF D. CYCLE-INDEPENDENTFORMATION MICROTUBULES

Ciliary regeneration and oral replacement are two processes involving formation of microtubules which can occur in nongrowing cells, In addition, oral replacement always involves regression of microtubules, and ciliary regeneration may be preceded by ciliary resorption in some situations. Both processes, described in Section D, may be regarded as programs for microtubule regulation which are simplified by being uncoupled from the division cycle. As such, they offer certain advantages for experimental inquiry into the minimal mechanisms involved in formation and regression of microtubules. 1. Ciliary Regeneration

Tetrahgmenu can be divested of their cilia under conditions that leave the cells viable and capable of regeneration (Child, 1965; Rosenbaum and Carlson, 1969). The method involves concentrating the cells in a solution of appropriate ionic composition in the cold at pH 6.0, subjecting them to multiple shearings with a glass syringe and an 18-gauge needle, and then placing them in a recovery medium. Used first with strain W,the procedure works equally well, without modification, with strain GL (Nelsen, 1974; Rannestad, 1974). Partial deciliation can be accomplished by omitting shearing (Rannestad, 1974).

REGULATION OF MICROTUBULES IN

Tetmhyrnenu

67

Some cells may retain their oral cilia subsequent to the amputation procedure, but these drop off after about 10 minutes in the recovery medium. Regenerating cilia can first be detected by scanning electron microscopy (SEM)about 12 minutes after deciliation. The SEM observations of Rannestad (1974) and Nelsen (1974) have confirmed the general pattern of regeneration suggested by the original light microscope observations of Rosenbaum and Carlson (1969). The oral cilia appear and elongate prior to the majority of somatic cilia (Fig. 4). Ciliary growth within the mouth is synchronous, unlike the situation in oral apparatus development. The appearance and elongation of somatic cilia is asynchronous; also, early cilia tend to be distributed throughout all regions of the body (Fig. 4). The average cell regains motility about 45 minutes after amputation, but does not

FIG.4. Ciliary regeneration in Tetrahymena. This is a scanning electron micrograph of a cell 30 minutes after deciliation. Note that all the oral cilia have begun to grow out, but only a few scattered somatic cilia are visible. Micrograph provided by E. M. Nelsen.

68

NORMAN E. WILLIAMS

have the full complement of mature cilia until 100 minutes after deciliation. 2. Oral Replacement Members of the genus Tetrahymena can form an oral apparatus in the anterior region of the cell, as well as at the equator (reviewed by Frankel and Williams, 1973).In this case the new mouth replaces the old one which is resorbed. In T. pyriformis, oral replacement occurs in nondividing cells, and the new mouth is apparently identical to the one that is resorbed. The adaptive significance of this is not immediately apparent. Oral replacement occurs typically in T . pyrtfomnis after a nutritional shift-down. It has been found that the process can be synchronized in mass populations by administering the classic heat-shock program for induced division synchrony, after first starving the cells for amino acids (Frankel, 1970).As far as we know, the new oral apparatus forms in exactly the same way it does during division (see Section II,B), only the location is different. The replacement primordium forms immediately posterior to the old mouth, and then moves anteriorly to take the latter's place when resorption is complete. The simultaneous resorption of one mouth and formation of another, separated by only a few micrometers, is a dramatic demonstration of the precision of the spatial controls that operate in the ciliate cortex. Electon microscope study of oral replacement has shown that old oral cilia are resorbed either in situ, or after withdrawal into the cytoplasm; both methods operate simultaneously within each cell (Williams and Frankel, 1973).Cilia withdrawn into the cytoplasm show no associated membrane material, neither their own nor those of autophagic vacuoles. The absence of autophagic vacuoles suggests disassembly without degradation to amino acids. 111. Microtubule Stability and Regression

Microtubules are classified as stable or labile (Behnke, 1970)on the basis of their response to cold, colchicine, and high hydrostatic pressure. Labile tubules are directly disrupted by these agents, whereas stable tubules are not. These agents block the formation of both kinds of tubules, however. It will be seen in the present section that most Tetrahgmena microtubule systems, and perhaps all, either survive these treatments or are lost as an indirect consequence of them. The conditions under which these and other agents bring about microtubule regression in Tetrahymena are discussed,

REGULATION OF MICROTUBULES IN

Tetrahymena

69

together with the information available on the possible mechanisms involved. A. SOMATICAND ORALMICROTUBULES Cold, colchicine, and pressure all block the formation of microtubules in Tetrahymena, as is true in other systems. Regarding disruption of microtubules, the results are more complex. The oral apparatus is discussed first because there is more information about this system.

1. Oral Structures Prior to the Transition Point The responses of both the anterior oral apparatus and the developing oral primordium differ on either side of the transition point (T.P. in Fig. 3). The transition point, which occurs at approximately 78% of the way through the cell cycle, is the time of transition from sensitivity to relative insensitivity in the cell cycle. Treatments by a wide variety of agents prior to the transition point result in excess division delays, that is, delays in excess of the duration of the treatment. Treatments after the transition point do not produce significant excess division delays (see Zeuthen and Rasmussen, 1972). The anterior oral apparatus does not regress as a result of treatment with any of the three agents under consideration if exposed prior to the transition point. The developing primordium does not regress, at least to any significant degree, when treated prior to the onset of membranellar organization. This occurs at about 62% of the way through the cell cycle (Fig. 3). From this time until the transition point, however, the developing oral primordium responds to treatment with cold (Frankel, 1962),colchicine (Nelsen, 1970),and pressure (Simpson and Williams, 1970)with a dramatic regression of all structural elements, including microtubules. This period of sensitivity of the developing primordium is indicated by the shaded zone in Fig. 3. Inducible regression notwithstanding, the oral primordium microtubules during this sensitive period should probably not be regarded as labile in the sense that this term has been applied to mitotic or other cytoplasmic microtubules. The reason is that there is considerable evidence suggesting that oral primordium resorption is an indirect consequence of treatment with cold, colchicine, or pressure. For example, it is well known that the loss of labile microtubules under the influence of these agents is rapid, reversible, and energy-independent. Oral primordium resorption, however, is slower, of an all-or-none character, and energy-dependent (Frankel, 1967).Partic-

70

NORMAN E. WILLIAMS

ularly illustrative of the fundamental difference between the two types of regression is the response to cold. Labile microtubules disappear during the cold treatment, whereas the oral primordium disappears only after the cold treatment is over and the cells have been returned to normal temperatures. Ultrastructural analysis has shown that every identifiable microtubular component of the developing oral apparatus is intact after 60 minutes at 2°C (R. J. Williams, unpublished). Loss of tubules follows only upon rewarming the cells. It must therefore be concluded that the loss of primordium microtubules is not due to a simple equilibrium shift toward dissociation of subunits, but rather to a process of active cellular degradation, possibly of enzymic nature, as originally concluded by Frankel (1967).It is known, furthermore, that the resorption mechanism can be triggered by heat shocks and a wide variety of metabolic inhibitors (see Frankel and Williams, 1973), as well as agents that affect microtubules. It thus appears that the period just before the transition point is a period of general cellular sensitivity, such that any perturbation of critical cellular processes leads to the initiation of oral primordium resorption, as well as to very long excess division delays (see Section VI for further discussion). 2. Oral Structures Following the Transition Point The situation after the transition point is very different. First, the anterior oral apparatus is now resorption-susceptible, probably because it too is undergoing developmental changes at this time (see Section 11,B).Second, metabolic inhibitors do not cause excess division delays, nor do cold, colchicine, or pressure. Finally, metabolic inhibitors and cold shocks do not trigger oral resorption, but colchicine and pressure do (in both oral apparatuses). The application of the latter two agents after the transition point therefore has severe consequences for the next generation; it produces daughter cells with nonfunctional mouths. The cells recover, however. Resorption of the old structures occurs, and new mouths form to take their place. Progress toward division, and development of both mouths, are halted by cold treatments initiated after the transition point. Upon rewarming the cells all development resumes and normal daughter cells are produced with no significant excess division delays and no oral resorption (Frankel, 1962). This establishes that blocking formation of microtubules per se is insufficient to trigger resorption, at least after the transition point.

REGULATION OF MICROTUBULES IN Tetrahymena

71

If this is true, and oral microtubules are stable, what are the mechanisms by which colchicine and pressure initiate oral resorption after the transition point? Nelsen (1970) showed that colchicine stops development of both oral apparatuses while permitting cell division to continue. In this case, unlike the situation with cold shocks, there is a dif5erentiaZ effect on cell development. It is perhaps the resulting imbalance that may be viewed as the specific trigger for posttransition oral resorption. Pressure must operate differently, because there is no differential blocking effect; both cleavage and oral development stop during treatment. Upon release of pressure, however, cleavage goes forward and oral development goes “backward.” An electron microscope analysis carried out on posttransition cells fixed under pressure may provide the answer. Moore (1972) reported no loss of microtubules, but what may be described as microtubule displacements were abundant. For example, ciliated basal bodies were found in the endoplasm, and noncylindrical associations of basal body triplets were seen. Perhaps such disruptions of microtubule positioning represent the specific stimulus for pressure-induced resorption, rather than tubule breakdown, or the blocking (or differential blocking) of microtubule formation in posttransition development. As Moore points out, however, the alterations after 10 minutes under 7000 psi pressure could also represent the first stages in the resorption process rather than the direct influences of pressure. Ultrastructural descriptions of oral regression are given by Moore (1972)and Williams and Frankel (1973).Cilia are resorbed in situ or intracytoplasmically. The ciliary doublets first lose the outer wall of the B tubule, then the inner wall, and last the A tubule. Doublet regression is asynchronous within a cilium. Basal bodies seem to disconnect triplets and then disintegrate into piles of granular material. All intracytoplasmic breakdown stages occur free in the cytoplasm; no relation to autophagic vacuoles has been seen in Tetrahymena. 3. Somatic Microtubules Regression of microtubules of the somatic ciliary units in Tetrahymena has not been reported in connection with treatment by cold or colchicine. There is a report, however, that notes some loss of certain cortical microtubules following the application of 7500 psi of pressure for 10 minutes (Kennedy and Zimmerman, 1970). A similar study, using 7000 psi for 10 minutes, failed to note these effects (Moore, 1972). If confirmed, the somatic regressions may be due either to direct pressure effects or, as in the case of the oral appara-

72

NORMAN E. WILLIAMS

tus, to the initiation of active resorption mechanisms. Further work is required before it can be determined whether or not there are any labile microtubules in the cortex of Tetrahyrnena.

B. NUCLEARMICROTUBULES Macronuclear microtubules form during macronuclear elongation (Fig. 3), which occurs only after the transition point Kennedy (1969) and Wunderlich and Speth (1970) showed that the application of colchicine relatively early in the cell cycle prevents the formation of nuclear microtubules. This is undoubtedly correct. However, the usefulness of this demonstration is diminished considerably by the fact that the same result is almost certain to be obtained with the application of cycloheximide, or any of several agents, prior to the transition point. Colchicine, and all the rest, simply prevent the cell from entering the phase of the cell cycle to which nuclear microtubule formation is restricted. These studies therefore accomplish the dubious feat of nonspecifically preventing the formation of nuclear microtubules with an agent that specifically blocks microtubule formation. This criticism is circumvented in a precise study by Tamura et al. (1969),in which colchicine was added to synchronized cells after the transition point. They found that colchicine applied at this time had no blocking or delaying effect on either macronuclear or cell division. However, nuclear cleavage was typically unequal. Many daughter cells were produced with macronuclei that were either smaller or larger than normal, and some had no nuclei at all. On the basis of these results and their electron microscope observations, Tamura et al. (1969) suggested that colchicine (1) blocks the formation of microtubules required for macronuclear elongation, and (2) causes the loss of existing macronuclear microtubules. It is possible that a block in the formation of tubules alone is sufficient to cause unequal nuclear divisions, thus the evidence for loss of preexisting tubules is entirely ultrastructural. Their electron microscope observations were not extensive, however, and it is possible that the nuclei they observed were cleaved nuclei; microtubules disappear normally very soon after nuclear division, which is completed well before cytoplasmic cleavage is over. For the present, the suggestion that the microtubules of the nucleus are colchicine-labile should perhaps be regarded as tentative. What is more certain is that colchicine blocks the formation of macronuclear microtubules required for nuclear elongation, which in turn leads to unequal macronuclear divisions. Studies of the effects of cold and high hydrostatic pressure

REGULATION OF MICROTUBULES IN Tetrahymena

73

on macronuclear microtubules have not been reported. No information is available regarding micronuclear microtubules.

CONCLUSION Most, and perhaps all, of the microtubules in Tetrahymena are stable, that is, they are not directly dissociated by cold shocks, colchicine, or high hydrostatic pressure. The possible exceptions are nuclear microtubules and certain somatic tubules. All three agents, as well as others, can lead to the loss of microtubules in Tetrahymena, however. This effect is mediated by an active resorption mechanism which can be triggered within the cell. Whether resorption is initiated depends on the agent, the microtubule system in question, and the position in the cell cycle. Thus the resorption mechanism is highly discriminatory, both spatially and temporally. The control of resorption is one of the most important problems in the in uiuo regulation of microtubules. It is important in Tetrahymena, and probably other cell types as well. It is even conceivable that some of the “labile” microtubules in other systems may, upon closer examination, prove to be stable microtubules which are regulated by resorption mechanisms. The studies with Tetrahymena suggest several possible activating circumstances for the resorption of stable microtubules, all or only some of which may actually apply. C.

1. Subtle damage to microtubules may be caused by some agents, particularly high hydrostatic pressure. Tetrahymena microtubules tend to persist intact under pressure, but some sort of subtle intratubule damage cannot be ruled out as a trigger for the subsequent microtubule resorption process. 2. Microtubule displacements produced by some agents may represent a specific stimulus for resorption. This is suggested by the fact that microtubules in abnormal locations have been seen soon after the application of high hydrostatic pressure in some instances, whereas the tubules themselves appeared to be intact. 3. Blocking the development of microtubules per se may lead to the resorption of others already present. Although this may be true in some cases, it has been shown not to apply to cold-arrested oral primordia after the transition point. 4. An imbalance created by blocking microtubule formation without arresting other cell cycle developmental processes may be a specific trigger for resorption. This is suggested above in the case of colchicine-blocked oral primordium development after the transition

74

NORMAN E. WILLIAMS

point, and may apply as well to this system prior to the transition point. 5. An imbalance created by arresting progress in the cell cycle without blocking microtubule formation may also constitute a specific stimulus for microtubule resorption in some cases. Perhaps the resorption of the oral primordium that follows the application of diverse metabolic inhibitors prior to the transition point is brought about in this manner. 6. Abnormal patterns of microtubules, such as those formed after prolonged exposure to high temperature or colchicine, may trigger the resorption mechanism (see Frankel and Williams, 1973). 7. Specific microtubule resorptions may be included in the program for development in the cell cycle. Examples from Tetrahymena are the regression that occurs in the anterior oral apparatus after the transition point, and the disappearance of nuclear microtubules after division.

IV. The Dynamic Nature of Formed Microtubules A. EVIDENCEIN Tetrahymena Ciliary and oral apparatus proteins have been shown to be turning over in nondeveloping, or morphostatic, structures. This was first demonstrated in an autoradiographic study of the oral apparatus (Williams et al., 1969). By using synchronized cells it was possible to show that pulse-administered radioactive amino acids are incorporated into proteins which can be localized within the nondeveloping anterior oral apparatus of growing cells. Furthermore, the rate was fairly high; incorporation into the developing oral primordium was greater by less than a factor of 2. A relatively rapid decline in specific activity of prelabeled anterior oral apparatus protein was also demonstrated following a chase with unlabeled amino acids. These results establish the dynamic character of the nondeveloping oral apparatus; proteins are in a constant state of flux within the system. Moreover, the flux rate is high enough to create problems of interpretation in certain types of experiments dealing with biogenesis. The flux of microtubule proteins, specifically, has been demonstrated within the morphostatic oral apparatus. In one experiment (Nelsen and Williams, unpublished), prelabeled cells were washed and suspended in fresh medium containing no labeled amino acids. Microtubule proteins (tubulins) were prepared from oral apparatuses isolated near the beginning of the chase and again after 9-13 hours

REGULATION OF MICROTUBULES IN

Tetrahymena

75

(the development of new oral apparatuses was prevented during this period by intermittent heat shocks). Specific activity measurements from the tubulins of the two samples indicated a decline in specific activity of about 2.5% per hour (subtracting the heat periods, which also block turnover). These data, as well as previously published information (Williams and Nelsen, 1973), reflect the dynamic nature of nondeveloping oral apparatus microtubules. So far, the discussion pertains to the intracytoplasmic components of the oral apparatus only (of which basal bodies constitute the major fraction); this is because the oral cilia are removed during the oral apparatus isolation procedure. Nelsen (1974) investigated turnover of the microtubule proteins of morphostatic cilia in Tetrahymena, however, and presents evidence for its occurrence there as well. In these studies cells were pulse-labeled with radioactive amino acids and then washed into a chase medium incapable of supporting growth. It was found that the microtubule proteins of ciliary axonemes show a 2% per hour decline in specific activity which cannot be attributed to morphogenesis. Furthermore, it has been found that the microtubule proteins in nondeveloping cilia become labeled in pulse experiments at fully one-half the rate of labeling during ciliary regeneration (Nelsen, unpublished).

B. TURNOVER AND SUBUNIT EXCHANGE It appears from the above considerations that the protein subunits of nondeveloping microtubules in Tetrahymena are in a constant state of flux, entering and leaving the structures, both in growing and nongrowing cells. This process is referred to here as subunit exchange, or exchange. It has been referred to previously as turnover; however, this term is commonly used for the simultaneous synthesis and degradation of protein that occurs extensively in nongrowing cells. A certain amount of ambiguity is introduced therefore by referring to the flux of proteins within a structure as turnover, particularly when radioactive amino acids are used in nongrowing cells to demonstrate it. A terminology that distinguishes exchange from turnover and synthesis also focuses more clearly on questions of the interrelationships between these processes. For example, how do rate changes or blocks in synthesis or turnover affect subunit exchange, and vice versa? Answers to questions of this type would be greatly facilitated by the development of methods for isolating soluble subunits from the cytoplasm. This has not yet been reported for Tetrahymena tubulin. This view of subunit exchange has implications for the interpreta-

76

NORMAN E. WILLIAMS

tion of various types of labeling experiments. For example, it is recognized that drawing conclusions about the rate at which exchange occurs from the rate of incorporation of labeled amino acids into the macromolecular subunits isolated from structures is not justified unless it is known that the rate of synthesis, or turnover, is not limiting. In other types of experiments, independent information on subunit exchange may be of consequence for the interpretation of experiments dealing with such things as pool sizes or induced synthesis. For instance, a ratio of incorporation of labeled amino acids into the proteins of regenerating versus nonregenerating structures of, say, 20 :1, is suggestive of induced synthesis during regeneration. This result can also be obtained with no induced synthesis during regeneration, however, in a system that has a large precursor pool of macromolecular subunits which turns over rapidly but which exchanges with morphostatic structures at a low rate. These examples are not intended to represent actualities, but to illustrate the possibilities that must be considered until more is known about the relationships between synthesis, turnover and subunit exchange. The apparently high exchange rate of tubulin in Tetrahymena ciliary and oral microtubules may not be typical of all systems. In another ciliate, Eupbtes, Ruffolo (1970)showed that labeled proteins are differentially retained in the original oral apparatus, that is, the one that was formed during the pulse, for at least three generations of growth and division under chase conditions. G. W. Grimes (personal communication) had showed the same thing in Oxytricha oral cilia over eight generations. This type of relatively fixed localization of labeled proteins does not occur in Tetrahymena (Williams et al., 1969), and it suggests that there may be very different subunit exchange rates in the oral systems of these hypotrichous ciliates and Tetrahymena. The oral microtubules of ciliates in general are stable to cold, colchicine, and high hydrostatic pressure. It has been suggested that this type of stability may be due to low turnover (exchange) of microtubule subunits in these microtubules (Behnke, 1970; Raff et al., 1971; Olmsted and Borisey, 1973).The high exchange rates of stable microtubules in Tetrahymena, and the lack of correspondence of exchange rates with degrees of stability in the ciliates mentioned above, do not appear to support this idea. The site, mechanism, and adaptive significance of subunit exchange are all unclear at the present time. Subunits may in theory exchange within microtubules in restricted zones, or at sites distributed randomly throughout. The radioautographic data of Rosenbaum

REGULATION OF MICROTUBULES IN

Tetrahymena

77

and Child (1967) are more suggestive of the latter possibility, although this has not been established. The mechanism of exchange within microtubules is also unknown, and there is little to suggest at the present time what its adaptive significance might be. One possibility is that it may provide a mechanism for repairing microtubules without resorbing and re-forming them.

V. Control of Microtubule Formation A. TUBULINSYNTHESIS IN RELATION TO MICROTUBULEFORMATION

Studies with inhibitors have shown that protein synthesis is required for the formation of the microtubule systems of Tetrahymena in most instances, but not in all. Protein synthesis is required for ciliary regeneration (Rosenbaum and Carlson, 1969; Nelsen, 1974; Rannestad, 1974), oral replacement (Frankel, 1970), and for oral (and probably somatic) microtubule formation in the cell cycle prior to the transition point (see Frankel and Williams, 1973). Protein synthesis is not required for somatic, oral, or macronuclear microtubule formation in the cell cycle after the transition point. This information, together with the fact that microtubules are composed primarily, if not exclusively, of tubulin, readily suggests the hypothesis that tubulin synthesis may be limiting for microtubule formation, that is, the assembly of microtubules may be controlled by the supply of tubulin subunits. According to this hypothesis, microtubule formation during oral replacement, ciliary regeneration, and development in the cell cycle before the transition point would involve de nouo synthesis of tubulin; microtubule formation after the transition point would proceed in the absence of protein synthesis, presumably because of the attainment of a threshold store of tubulin sufficient to insure the completion of cell and nuclear division. This hypothesis has been tested by exploring the relationship between the synthesis of tubulin and the formation of microtubules during oral replacement (Williams and Nelsen, 1973) and ciliary regeneration (Nelsen, 1974). Both studies used nongrowing cells in which the absence of net synthesis of protein provides an optimal background for detecting specific synthesis associated with morphogenesis. The experimental design used in both studies can be formulated in general terms as in Fig. 5. The cells were prelabeled by growth in labeled amino acids (solid half-circles in Fig. 5). Radioactive tubulin was then synthesized (solid circles), and both ciliary

78

NORMAN E. WILLIAMS A

Smil assembly

T T

FIG.5. Experimental design used to test for induced synthesis of tubulin during ciliary regeneration and oral replacement in Tetahymena. Half-circles represent amino acids, complete circles represent tubulin, and aligned circles represent microtubules. Solid symbols denote labeled, and open symbols unlabeled, amino acids and tubulin. Result Cz indicates induced synthesis of tubulin during regeneration, whereas result C1 indicates the use of preexisting tubulin with little or no induced synthesis of tubulin during regeneration. Result C, was obtained in both oral replacement and ciliary regeneration. Further discussion is presented in the text.

and oral microtubules (aligned circles) were formed from labeled tubulin. The cells were next washed and transferred to a medium incapable of supporting growth, which also contained an excess of unlabeled amino acids (open half-circles). The old structures were then caused to regress or were physically removed (B); synchronous oral replacement was induced in one study, and the cells were deciliated in the other. The structures were then allowed to regenerate (C).If the new microtubules are formed from tubulin synthesized de nouo, they should contain predominantly unlabeled tubulin (C,). If, however, the new microtubules contain predominantly labeled tubulin (C,), their formation must be supported primarily by a pool of previously synthesized tubulin. This pool might be the old oral apparatus in the case of oral replacement (reutilization), but would necessarily be a soluble fraction in the case of ciliary regeneration because the old cilia were discarded. The critical information obtained was the relation between the specific activities of tubulin from the prelabeled (A) and regenerated (C, and C,) microtubules. The data in both studies conform to result C , in Fig. 5. The measurements indicate that about 6% of the tubulin of the oral basal bodies and associated microtubules, and less than 10% of the tubulin in ciliary microtubules, were synthesized during regeneration of the structures. It was concluded that the immediate requirement for protein synthesis during microtubule formation is not for bulk supply of tubulin subunits. Rather, the assembly of preexisting tubulin subunits must be controlled by the synthesis of some other factor (Fig. 6).

REGULATION OF MICROTUBULES IN Tetrahymena

/\Assembly

79

r=u n

t2 w e s

Factor

Tubulin

FIG.6. Scheme for the regulation of tubulin assembly in Tetruhyrnena.Ciliary and oral microtubules appear to be formed largely from preexisting stores of tubulin (see Fig. 5), yet de nouo synthesis of protein is required for assembly. The protein that must be synthesized is designated assembly factor in this figure. Assembly factor(s) may be a structural component of cilia, because ciliary regeneration can be obtained in the absence of protein synthesis if resorption of old cilia is induced. See text for further discussion.

Further studies suggest that this view of tubulin synthesis and assembly also applies to growing Tetrahymena. Bieber and Stone (1972)found that vinblastine (VLB) blocks the formation of microtubules in growing Tetrahymena but does not block synthesis of tubulin. By using appropriate pulse and pulse-chase experiments, they showed that ciliary microtubules formed during one doubling in cell number after release from a VLB block contained more tubulin synthesized during the block than after the block. Furthermore, the tubulin pool present at the end of the VLB block was unable to assemble into microtubules in the presence of cycloheximide. These data support the general conclusion that microtubule formation in Tetrahymena may involve extensive use of stored cytoplasmic tubulin, and that the assembly of this preexisting tubulin is synthesisdependent. The use of proteins from a previously synthesized pool was also shown in oral primordium development in growing cells by Williams et al. (1969). The chase experiments of Rannestad and Williams (1971)showing a decline in specific activity of oral apparatus tubulin of 50% per generation probably indicate a balanced increase in both pool and oral apparatus tubulin, rather than the lack of a cytoplasmic tubulin pool as originally suggested.

B.

SYNTHESIS-DEPENDENT

ASSEMBLYOF

PREEXISTING TUBULIN 1. Nature of the Synthesis Requirement What is the nature of the gene product presumed to regulate tubulin assembly (assembly factor in Fig. 6), and how does it operate? There may be more than one protein involved of course, and more than one mode of control. This appears to be probable, at

80

NORMAN E. WILLIAMS

least in ciliates which have many microtubule systems showing somewhat independent patterns of regulation. Given that there are pools of unassembled tubulin within the cell, and considering the thermodynamics of assembly (InouB, 1964), microtubule self-assembly is to be expected, unless there is either an assembly repressor (negative control) or some catalytic or structural component that must interact with tubulin before it can assemble (positive control). There is evidence for the latter type of control in other systems. For example, Mason and Schatz (1973) present evidence that mitochondrially synthesized cytochrome-c oxidase subunits regulate the assembly of the cytoplasmically synthesized subunits of this enzyme complex. A control mechanism for regulating the biogenesis of structures in which preexisting stores of proteins may be induced to assemble by the temporally regulated synthesis of one required component was predicted by King and Mykolajewcz (1973) on the basis of their studies of T4 assembly, although this does not occur in phage. The studies in Tetrahyrnena reviewed above suggest that it might be possible to identify tubulin assembly factors by their pattern of regulation. It has been found that exposure to a sublethal heat shock blocks the assembly of tubulin into microtubules (both net and via exchange), but does not block synthesis of total cell protein (Williams and Nelsen, 1973). Similarly, it has been shown that VLB blocks assembly of microtubules but does not block the synthesis of tubulin (Bieber and Stone, 1972). Both procedures may block the synthesis of tubulin assembly factors differentially, because the inability of tubules to form after these treatments in the absence of protein synthesis implies that there is no accumulation of these factors during the blocks. Alternatively, heat and VLB might impose an instability upon these proteins differentially. If the postulated tubulin assembly proteins are elements that are incorporated into structure, they should be easier to identify than agents acting catalytically or by means of negative control mechanisms. There is some indication, discussed in Section By2, that the proteins regulating ciliary regeneration are structural components of cilia, and that they might be identified by their pattern of de novo synthesis during regeneration.

2. ControZ of Microtubule Formation in Tetrahymena Evidence suggesting that the protein factor required for tubulin assembly during ciliogenesis may be a structural component of cilia comes from the work of Rannestad (1974), in which an exception to

REGULATION OF MICROTUBULES IN

Tetrahymena

81

the rule that ciliary regeneration requires de nouo protein synthesis is reported. The permissive condition is that the cells resorb some of their preexisting cilia. This was found to occur after partial deciliation of Tetrahymena cells, which can be accomplished by eliminating the shearing step in the deciliation procedure. Quantitative study of a timed series of scanning electron micrographs has established that the cilia remaining attached to the body are not cast off, but are instead withdrawn into the cytoplasm within the first 20 minutes after partial deciliation. Regeneration of new cilia begins before resorption of the old cilia is complete. This regeneration, unlike regeneration in fully deciliated cells, is not dependent on protein synthesis. Thus it appears that the protein factor controlling ciliary regeneration can either be synthesized de nouo, or obtained by resorbing old cilia (Fig. 6). The measurements show that complete regeneration of body cilia does not occur in cycloheximide in partially deciliated cells. However, significantly more ciliary material is formed than is resorbed. Rannestad has suggested that the difference comes from cytoplasmic stores, and that material from the resorbed cilia in some nonstoichiometric way potentiates utilization of pool material. If the synthesis of a regulatory protein does control the assembly of preexisting tubulin, the former should have a much higher specific activity than the latter in experiments in which labeled amino acids are provided during ciliary regeneration. If, additionally, the regulatory protein is a structural component of cilia, as suggested by Rannestad's experiments, it should be possible to isolate it from cilia and to identify it by its relatively high specific activity in this type of experiment. Following this approach, Nelsen (1974)has suggestive evidence that a high-molecular-weight component of cilia isolated on SDS gels may contain the regulatory protein(s). The method of fractionation makes it impossible to say whether this material is a component of microtubules or of some other part of the cilium. In theory it could be either; if tubulin does not assemble without it, a critical interaction with tubulin is implied whether or not it is assembled into microtubules. However, it may be significant that microtubules assembled in uitro from assembly-purified tubulin apparently also contain some high-molecular-weight material (Granett et al., 1973). Some peculiarities of the oral apparatus system in Tetrahymena suggest that it may be regulated differently. It has been found in Tetrahymena (Nelsen, 1974),and also in Stentor (Plapp and Burchill, 1972),that oral cilia can regenerate, at least to a considerable extent, in the absence of protein synthesis. This is something that somatic

82

NORMAN E. WILLIAMS

cilia cannot do. Either there are different assembly regulators for the two types of cilia, or at least oral cilia have first priority for limited supplies of a common factor. Another difference is that oral apparatus formation cannot be supported in the absence of protein synthesis by the resorption of old structures. The model in Fig. 6 can be modified to accommodate this difference by imagining that the assembly factor for oral microtubules is synthesized in conjunction with a protein that inhibits its function. Oral development would occur only when de nouo synthesis provides for neutralization of the inhibitor. Such a mechanism of negative control could account for the inability of the resorption of previous oral structures alone to promote oral development; resorption would make assembly factor available but would not inactivate the assembly inhibitor. These considerations perhaps do little more than emphasize the complexity that will be required to explain the total regulation of microtubules within a single cell. They lead additionally, however, to possible explanations for the change from synthesis-dependent to synthesis-independent microtubule formation, which occurs at the transition point in the cell cycle of Tetrahymena. According to the model in Fig, 6, transition may be correlated with the accumulation of enough assembly factor to ensure completion of the somatic, oral, and macronuclear microtubule formation required for cell and nuclear division. According to the negative control mechanism discussed above, transition could be achieved only if the assembly inhibitor were also inactivated simultaneously.

3. Assembly Factor Models Applied to Microtubules in Other Cell TY Pes There is evidence that mechanisms of the general sort discussed above may operate in several other cell types. Shubert et aZ. (1971) have established that neurotubule assembly during formation of cell processes in neuroblastoma cells in culture does not occur in the absence of protein synthesis. As in Tetrahymena, the required synthesis appears not to be of tubulin. Morgan and Seeds (1973) found no increase in the amount of tubulin during neurite formation, and Yamada and Wessels (1971) report the same for neutite production by cultured dorsal root ganglia under the influence of nerve growth hormone. Tubulin was measured by the colchicine-binding assay in both studies. Both mitotic spindle formation and ciliogenesis in sea urchin embryos require de nouo synthesis of protein, which in neither case appears to be tubulin. It was first shown by Hultin (1961), and later by

REGULATION OF MICROTUBULES IN

Tetrahymena

83

Wilt et al. (1967),that protein synthesis is required for formation of the cleavage spindle in sea urchin eggs. Borisy and Taylor (1967) then discovered that unfertilized eggs contain large quantities of tubulin, and Raff et al. (1971)subsequently demonstrated large cytoplasmic tubulin pools from fertilization through the gastrula state. It is generally agreed that the formation of the mitotic spindle and the appearance of cilia at the blastula stage must both be controlled in some manner other than by the induced synthesis of tubulin. Stephens (1972)studied the synthesis of ciliary proteins during ciliogenesis in sea urchin blastula and has reported that at least six minor components can be identified that are synthesized at relatively rapid rates during ciliogenesis. He concludes that bulk ciliary proteins, including tubulin, are made prior to ciliation in considerable excess, and suggests that the morphogenetic process is marked by a round of de nouo synthesis of minor, but critical, structural components. Stephens (1972)further suggests that the regeneration of cilia reported to occur in the absence of protein synthesis in older embryos (Auclair and Siegel, 1966)may be supported by the earlier synthesis of multiple rounds of limiting proteins in this case. Regeneration of flagella in Chlamydomonas may be controlled in a similar manner. As in Tetrahymena, protein synthesis is required for regeneration (Rosenbaum et al., 1969),yet there is evidence that the synthesis of tubulin is not limiting (Rosenbaum, personal communication). Resorption of old flagella in Chlamydomonas can also support regeneration of new flagella in the absence of protein synthesis (Rosenbaum et al., 1969;Coyne and Rosenbaum, 1970).

VI. Epilog: The Cell Cycle Revisited The relation between the regulation of microtubules and the mechanism of heat-induced division synchrony in Tetrahymena requires a brief comment. Zeuthen and co-workers have shown that heat shocks produce excess division delays (setbacks) prior to the transition point, which are greater in older cells and lesser in younger cells (see Zeuthen and Rasmussen, 1972). By this means cells are brought together in time, and then progress in synchrony toward division after the heat-shock treatment. A recent hypothesis of the mechanism by which setbacks are induced visualizes the progressive development of structural entities within the cell, which are destroyed by the heat shocks (Zeuthen and Williams, 1969; Zeuthen, 1971; Zeuthen and Rasmussen, 1972). The question then arises whether microtubule systems in Tetrahymena constitute such

84

NORMAN E. WILLIAMS

developing structural entities. Certainly the developing oral primordium is set back, that is, resorbed, if treated during the membranellar organizing period (see Section 111). However, it is unlikely that excess division delays, and therefore induced division synchrony, result from direct effects on microtubular systems in Tetruhymenu. The major reason for this conclusion is the lack of a strict correlation between the resorption of microtubules and the production of excess division delays. On the one hand, a variety of agents, including colchicine (Nelsen, 1970) and VLB (Williams and Stone, unpublished), produces excess division delays at all times prior to the transition point yet, as far as we know, leads to a loss of microtubules only in the oral primordium, and here only part of the time. On the other hand, the oral primordium can be set back (caused to resorb) without attendant excess division delays, for example, by the application of high hydrostatic pressure after the transition point (see Section 111). In addition, the dissociability of microtubule formation from the cell cycle (see Section I1,D) argues against the notion of microtubules as central to the control of the cell cycle. Rather, the mechanisms of microtubule biogenesis and regression should perhaps be regarded as quasi-autonomous, showing integration with fundamental cell cycle control mechanisms only at certain times and in specific regions. ACKNOWLEDGMENTS The author thanks Dr. E. Marlo Nelsen, Dr. Joseph Frankel, and Ruth Jaeckel Williams for helpful discussions and a critical review of the article. I also thank Dr. John J. Ruffolo, Jr., for the micrographs reproduced in Fig. 1 and 2,and Dr. E. Marlo Nelsen for the micrograph reproduced in Fig. 4.

REFERENCES Allen, R. D. (1967).J . Protozool. 14,553. Allen, R. D. (1969).1. Cell Biol. 40,716. Auclair, W . ,and Siege], B. W.(1966).Science 154,913. Behnke, 0.(1970).Int. Reo. E x p . Pathol. 9,1. Bieber, R. W..and Stone, G. E. (1972)./.Cell B i d . 55,lQa. Borisy, G.G.,and Olmsted, J. B. (1972).Science 177,1196. Borisy, G . G.,and Taylor, E. W. (1967). J. Cell Biol. 34,535. Buhse, H. E., Stamler, S. J., and Corliss. J. 0. (1973).‘frans.Amer. Microsc. SOC. 92,

95.

Child, F. M. (1965).J. Cell Biol. 21, Ma. Coyne, B., and Rosenbaum, J. L.(1970).J. Cell Biol. 47,777. Elliott, A. M.,and Kennedy, J. R. (1973).In “The Biology of Tetruhymena” (A. M. Elliott, ed.), pp. 57-87.Dowden, Hutchinson, & Ross, Stroudsburg. Pennsylvania.

REGULATION

OF MICROTUBULES IN

Tetruhymena

85

Falk, H., Wunderlich, F., and Franck, W. W. (1968).J . Protozool. 15,776. Forer, A., Nilsson, J. R., and Zeuthen, E. (1970).C. R. Trau. Lab. Carlsberg 38, 67. Frankel, J. (1962).C. R. Trau. Lab. Carlsberg 33, 1. Frankel, J. (1967).J . E r p . Zool. 164,435. Frankel, J. (1970).J . E r p . Zool. 173,79. Frankel, J. (1973).J . E r p . Zool. 183,71. Frankel, J., and Williams, N. E. (1973). In “The Biology of Tetruhymena” (A. M. Elliott, ed.), pp. 375-409. Dowden, Hutchinson, & Ross, Stmudsberg, Pennsylvania. Furgason, W. H. (1940).Arch. Protistenk. 94,224. Granett, S., Dentler, W., Whitman, G. B., and Rosenbaum, J. L. (1973).J . Cell Biol. 59, 119, Heckmann, E, and Frankel, J. (1968).J . E r p . Zool. 168, 11. Hultin, T. (1961).Erperientia 17,410. Inoub, S . (1964). In “Primitive Motile Systems in Cell Biology” (R.D. Allen and N. Kamiya, eds.), pp. 549-594. Academic Press, New York. Ito, J., Lee, Y. C., and Scherbaum, 0. H. (1968).E x p . Cell Res. 53,85. Kennedy, J. R. (1969).In “The Cell Cycle: Gene Enzyme Interactions” (G. M. Padilla, I. L. Cameron, and G. L. Whitson, eds.), pp. 227-248. Academic Press, New York. Kennedy, J. R., and Zimmerman, A. M. (1970).J. Cell Biol. 47,568. King, J., and Mykolajewycz, N. (1973).J. Mol. Biol. 75,339. Mason, T. H., and Schak, G. (1973).J. Biol. Chem. 248,1355. Moore, K. C. (1972).J. Ultrastruct. Res. 41,499. Morgan, J. L., and Seeds, N. W. (1973).J.Cell Biol. 59,233a. Nanney, D. L. (1966).Genetics 54,955. Nanney, D. L. (1971a).D eu el o p . Biol. 26,296. Nanney, D. L. (1971b).J . E x p . Zool. 178, 177. Nanney, D. L., and Chow, M. (1974).Amer. Natur. 108,125. Nelsen, E. M. (1970).J . E r p . Zool. 175,69. Nelsen, E. M. (1974).Ph.D. Thesis, University of Iowa, Iowa City. Nilsson, J. R., and Williams, N. E. (1966).C. R. Trau. Lab. Carlsberg 35, 119. Olmsted, J. B., and Borisy, G. G. (1973).Annu. Reo. Bfochem. 42,507. Perlman, B. S. (1973).J . E x p . Zool. 184,365. Plapp, F. V., and Burchill, B. R. (1972).J. Protozool. 19,663. RaE, R. A., Greenhouse, G., Gross, K. W., and Gross, P. R. (1971).J . Cell Biol. 50,516. Rannestad, J. (1974).J . Cell Biol. In press. Rannestad, J., and Williams, N. E. (1971).J . Cell Biol. 50, 709. Rosenbaum, J. L., and Carlson, K. (1969).J . Cell Biol. 40,415. Rosenbaum, J. L., and Child, F. M. (1967).J . Cell Biol. 34,345. Rosenbaum, J. L., Moulder, J. E., and Ringo, D. L. (1969). J . Cell Biol. 41, 600. Roth, L. E., and Minick, 0. T. (1961).J . Protozool. 8, 12. Ruffolo, J. J., Jr. (1970).J . Protozool. 17, 115. Schubert, D., Humphreys, S.,DeVitry, F., and Jacob, F. (1971).D e v e l o p . Biol. 25,514. Shelanski, M. L., Gaskin, F., and Cantor, C. R. (1973).Proc. Nat. Acad. Sci. U.S. 70, 765. Simpson, R. E., and Williams, N. E. (1970).J. E r p . Zool. 175,85. Stephens, R. E. (1972). Biol. Bull. 142,489. Tamura, S., Tsuruhara, T., and Watanabe, Y. (1969).E x p . Cell Res. 55,351. Weisenberg, R. C. (1972).Science 177,1104. Williams, N. E., and Frankel, J. (1973).J . Cell Biol. 56,441. Williams, N . E., and Nelsen, E. M. (1973).J. Cell B i d . 56,458.

86

NORMAN E. WILLIAMS

Williams, N. E.,and Scherbaum, 0.(1950).J. Embryol. Erp. Morphol. 7,241. Williams, N. E.,Michelsen, O., and Zeuthen, E. (1969).J . Cell Sci. I, 143. Wilt, F. H.,Sakai, H.,and Mazia, D. (1867).J . Mol. Biol. 27, 1. Wunderlich, F.,and Speth, V. (1970).Protoplasma 70,139. Yamada, K M.,and Wessels, N. K. (1971).Erp. Cell Res. 66,346. Zeuthen, E.(1971).Aduon. Cell Biol. 5111. Zeuthen, E.,and Rasmussen, L. (1972).In “Research in Protozoology” (T. T. Chen, ed.), Vol IV,pp. 9-145. Pergamon, Oxford. Zeuthen, E.,and Williams, N. E. (1969).In “Nucleic Acid Metabolism, Cell Differentiation and Cancer Growth” (E.V. Cowdry and S. Seno, eds.), pp, 203-217. Pergamon, Oxford.

Cellular Receptors and Mechanisms of Action of Steroid Hormones SHUTSUNGLIAO The Ben May Laboratory for Cancer Research and the Department of Biochemist y.The Unioersity of Chicago. Chicago. Zllinofs

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . .

. . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

I. Introduction 87 90 I1 Steroid-Binding Proteins in Blood 92 I11. Steroid Receptors in Target Tissues A Estrogen Receptors . . . . . . . . . . . . . . . 92 B. Androgen Receptors. 101 C. Progestin Receptors 109 113 D. Glucocorticoid Receptors 120 E . Mineralocorticoid Receptors F. Steroid Receptors in Brains . . . . . . . . . . . . 123 125 C . Receptor and Steroid Dependency of Cancer IV. Cytoplasmic-Nuclear Interaction of Steroid Receptors . . . . 127 A . Transformation and Nuclear Retention of Cytoplasmic Receptors . . . . . . . . . . . . . . 127 131 B. Cytoplasm-Independent Nuclear Receptors 132 C. Chromatin Acceptor Sites for Receptors. D . Ribonucleoprotein Binding of Receptors . . . . . . . 137 E . Intracellular Recycling of Receptors . . . . . . . . . 139 V. Gene Expression and Steroid Receptor 139 A . RNA Synthesis and Protein Induction . . . . . . . . 139 B. Hypothetical Models . . . . . . . . . . . . . . 144 C In Vitro Experimental Approaches . . . . . . . . . . 147 VI . Concluding Remarks . . . . . . . . . . . . . . . . 151 A. ReceptorandUptakeofSteroidbyCells . . . . . . . . 151 B. Natural Forms of Steroid Receptors . . . . . . . . . 151 C . Nature of Receptor-Steroid Interaction . . . . . . . . 152 D. Insect Hormones and Vitamin D3 . . . . . . . . . . 154 E . Cyclic AMP and Steroid Hormones . . . . . . . . . 155 References . . . . . . . . . . . . . . . . . . . 157

.

.

. . . . . .

. . . . . . .

. . . . . . . . .

.

I

.

Introduction

One of the most significant developments in the study of the molecular process of steroid hormone action in the last decade was the discovery of proteins that can selectively bind active steroids in target cells . Although the role of these proteins in hormone action has not been clearly established. they exhibit various properties one 87

88

SHUTSUNG LIAO

can expect to be characteristic of the functional receptors in target tissues. Interest in the study of steroid receptor molecules in a variety of target tissues has gained momentum in the last 5 years, the annual number of publications on the subject increasing from a few dozen to a few hundred during that time. This article covers those aspects related to the action of estrogens, androgens, progestins, glucocorticoids, and mineralocorticoids in vertebrates. The search for receptor proteins for these steroid hormones in most instances follows a unique pattern which includes (1) study of the uptake and retention of a radioactive hormone; (2) identification of the hormone presumed to be the active form; (3) detection and isolation of a specific protein that binds an active steroid but not an inactive steroid with a high affinity, and exists in larger amounts in target cells than in insensitive cells; and (4) demonstration that steroid antagonists can interfere with receptor binding of an active steroid. Studies on steroid receptors have been recently extended to several more biodynamic aspects such as qualitative and quantitative analysis of receptor proteins and their relation to biological responses of target cells and to some clinical situations. Concentrated efforts, however, are being made to discover the molecular mechanism whereby a steroid receptor complex may participate in the regulation of gene expression of the target cells. For this reason extensive studies have been carried out on the interaction of the steroid-receptor complex and the nuclear components of target cells. Numerous articles and books are available for readers interested in more comprehensive descriptions of the effects of various steroid hormones on cellular metabolic activities, including gene transcription (RNA synthesis) and translation (protein synthesis) (WilliamsAshman, 1965; Tata, 1966; Litwack, 1970, 1972; Smellie, 1971; McKern, 1971, 1974; Rasp&, 1971; Pasqualini and Scholler, 1972; Karlson and Sekeris, 1973; O’Malley and Means, 1973; Pitot and Yatvin, 1973; Rickenberg, 1974; Niu and Segal, 1974). This article reviews only those aspects that have been discussed in relation to the function of steroid receptors. Since many methodological aspects of the hormone receptor study can be found in a forthcoming volume of Methods in Enzymology (Academic Press), they are not included. For convenience, the chemical structures of many representative steroid hormones, natural or synthetic, and their antagonists are shown in Fig. 1. Since steroid hormones are transported from their production sites to many target tissues, a brief summary of the steroid-binding proteins in blood is presented in the next section.

CELLULAR RECEPTORS FOR STEROID HORMONES NATURAL

SYNTHETIC

aP

0’-

5o-Dihydrotestosterone

89

ANTAGONISTIC

&-CH3

0

;

CI

q

Cyproterone

Z-Oxa-1Io-methyl-17~3-hydroxyestra-4.9.Il-trien-3-one

OCH~CH~N’J

HO

@p

HO @OH

CH30&

Diethylstilbestrol

iln-Eth~yl-I!l-nortestosterone

Progesterone

Nafoxidine

13-Ethyl-llo-ethynyl-178hydroxyyona-i.9.11trien-3-one

CHZOH

CH20H LO

0

Dexmethasone

Cortisol

HO

I

0-CH

Aldoste rone

0-

Cortexolone

CHZOH

1

CO

90- Fluorocortisol

SDirolactone

FIG. 1. Chemical structures of natural and synthetic steroid hormones and their antagonists often used in the study of steroid receptors in target tissues.

90

SHUTSUNG LIAO

11. Steroid-Binding Proteins in Blood

The dynamics of steroid hormone distribution in the body are very complex. The process involves biosynthesis, distribution by means of the blood, interaction with blood and tissue components, metabolic alterations, and excretion. Each of these factors in turn is also affected by many other factors, including individual and diurnal variations. Biochemically, these conditions ultimately govern the availability of the active forms of the steroid to the cellular receptor sites, and the duration of hormonal effects on the target cells. The blood steroid concentrations appear to be of the order of 10-100 nM for androgens (Liao and Fang, 1969), glucocorticoids, and progestins, and of the order of 1 nM or lower for estrogens and mineralocorticoids (Diczfalusy, 1970). Most of the blood steroids are disposed of continuously and rapidly, so that to maintain the hormonal status for long periods requires large amounts of steroids. For example, in young adult male mammals the plasma concentration of testosterone is of the order of 5 nglml, but the production rate for androgens is 2-10 mg per day in a human being and 25-160 mg per day in a bull (Mann, 1964). The binding of steroids to serum proteins has been reviewed in detail by Westphal(l970, 1971), Liao and Fang (1969),and others in published symposia (Pincus et al., 1966; Diczfalusy, 1970). Serum albumin, the most abundant protein constituent of plasma, has a low affinity and low steroid specificity, but a high capacity for steroid binding. More specific steroid-binding globulins are present in lesser amounts (-1%)but, because of their high affinity, large portions of blood steroids may be strongly bound to them if total blood steroid levels are low. Many blood steroid-binding proteins have been purified and characterized (Table I). Their affinity constant is in general of the order of lo* M-' or even higher. Many of them do not have very rigid steroid specificities and in many cases appear to recognize limited portions of the steroid molecules such as special functional groups on ring A or D. These properties distinguish them from the cellular receptors that have a high affinity for very specific groups of steroids. Their presence, however, often complicates the study of cellular steroid receptors. The biological function of the plasma steroid-binding globulins is not clear. To emphasize its possible importance in transport, a plasma globulin was called transcortin (Sandberg et al., 1966). In several cases the steroid-binding proteins appear to be responsible for slower metabolic clearance (Sandberg et al., 1966; Baird et

TABLE I RELATIVE STEROID BINDING AFFINITY Steroid Testosterone Dihydrotestosterone Estradiol Estrone Progesterone Cortisol Corticosterone Aldosterone K a ir M-I

Human SAb Human CBG'

34

-

100 28 56 4 6 2 1.6 x 105

+

-

+

(+)

70 60 100 (+)

1.0 x 109

FOR

Human SBGd Human AAG'

33 100 20

-

2 1 1

-

1.2 x 109

67

-

11

-

100 2 11 8.9 x 105

SERUM PROTEINS' Rabbit DBGf Guinea pig PBG'

Rat fetus EBGh

33 100 5 6

5 15 0

1

100

0 54 100 12

-

-

14 X loR

2.6 x loR

1

6.2 X

loR

0 0 0

9

-

15

Fr

s*

9

En B

cd

32

" For human SA, CBG, and AAG, the ratios of the affinity constants are compared using the strongest binder as 100. For others, they

Y

are standardized from the data shown in the references cited. L, SA, serum albumin (Westphal, 1970). CBG, Corticoid-binding globulin; +, moderately active; (+), very weakly bound. Existing in numerous species. Increased by estrogens and decreased by corticoids (Chader and Westphal, 1968a,b; Sandberg et al., 1966). SBG, Sex steroid-binding globulin. Elevated during pregnancy or by estrogens (Pearlman and Crepy, 1967; Vermeulen and Verdonck, 1970; Mercier-Bodard et al., 1970). AAG, &,-acid glycoprotein (Kerkay and Westphal, 1968). DBG, Dihydrotestosterone-binding globulin (Mahoudeau and Corval, 1973). PBG, Progesterone-binding globulin. Existing in pregnant guinea pig but not in unpregnant guinea pig. Cannot be induced by estrogen or progesterone. Not present in pregnant rats or women (Milgrom et al., 1973a; Lea, 1973a,b). EBG, Estrogen-binding globulin, but does not bind diethylstilbestrol (Swartz et al., 1973; Raynaud et al., 1971).A similar estrogenbinding protein is also found in the plasma of pregnant rats but not in normal and castrated adults. The estrogen binding is not inhibited by antiestrogens (Soloff et al., 1971). K,, Apparent affinity constants (4°C) for the strongest binders shown.

!a

0

2M 8 x

I?

0

3m

z

92

SHUTSUNG LIAO

al., 1969; Mahoudeau et al., 1973). This may partially explain, for example, the high level of blood progesterone in pregnant animals (Illingworth et al., 1970; Milgrom et al., 1973a), or high blood retention of estrogens in immature rats (Weisz and Gunsalus, 1973; Raynaud et al., 1971). In addition, blood steroid-binding globulins may protect steroids from degradation (Sandberg and Slaunwhite, 1963; DeHertogh et al., 1970; Corvol and Bardin, 1973). There is, however, no strong evidence that blood proteins can indeed act as a reservoir from which intact steroids can be fed gradually to target organs. In fact, the biological activity of steroids in plasma is probably exerted by the unbound fraction, which is only a small percentage of the total concentration. This view is supported by several experiments showing the suppression of hormonal activities by a prior globulin binding of the steroid hormones that are to be administered to experimental animals (Sandberg et al., 1966; Gala and Westphal, 1967). Whether any significant portions of globulin-bound steroids ever become active under normal circumstances in uioo, or whether steroid binding by blood proteins is involved in regulation of the ratios of various steroids with different biological activities (Burke and Anderson, 1973), still needs further study. The latter possibility is indicated by the fact that the blood content of many of these globulins is influenced by the endocrine status of the animals (Table I). Whatever the function of the steroid-binding proteins in blood may be, it should be noted that not all animal species contain similar sets of steroid-binding globulins in the blood. It is therefore apparent that these proteins are not directly involved in the action of steroid hormones as functional receptors at cellular sites. 111. Steroid Receptors in Target Tissues

A.

ESTROGENRECEPTORS

1. Identification and Properties of Uterine Receptors The existence of receptor molecules for steroid hormones in target tissues was first indicated in a study of the uptake and retention of radioactive estrogens by mammalian tissues (Jensen and Jacobson, 1962). For example, immediately after the injection of a physiological dose of radioactive estradiol (17 pestradiol) into immature rats, uptake of the radioactive steroid by various tissues appears to reflect blood estrogen concentrations. The radioactivity then disappears

93

CELLULAR RECEPTORS FOR STEROID HORMONES

rapidly from the blood and tissues less sensitive to estrogens (kidney, liver, muscle, and diaphragm), but only slowly from target tissues such as the uterus and vagina (Fig. 2). The retention, but not the total uptake of estrogen, can be saturated by increasing the estrogen doses to the animals, indicating the presence of a limiting number of specific binding sites with a high affinity for estrogens (Jensen et al., 1967a,b). The importance of estrogen retention is also emphasized by the finding that only estrogenic compounds, including diethylstilbestrol, but not nonestrogenic compounds can compete with estradioL3H for uterine binding, and that the process is antagonized by many antiestrogenic compounds (Callantine et al., 1968; Kahwanago et al., 1970; Geynet et al., 1972; Jensen et al., 1972a). The use of various estrogen antagonists such as ethamoxytriphetol, clomiphene, nafoxidine, and Parke-Davis (21-628 has been valuable in distinguishing specific estrogen binding from nonspecific steroid association by tissues in uiuo and in uitro. pM’MG

1

DRY TISSUE

/o

\

2

4

5

HRS

6

16

FIG.2. Uptake and retention of radioactive estradiol by various tissues of immature rats receiving a single subcutaneous injection of 0.098 pg (11.5pCi) of estradiol3H. Total radioactivity is expressed as disintegrations per minute per milligram of dry tissue or disintegrations per minute per 5 p1 of blood. Since blood contains a mixture of radioactive metabolites but uterus and vagina incorporate only estradiol, the ratio of estradiol concentration between uterus and blood is about 500:l. (Details in Jensen and Jacobson, 1960, 1962.)

94

SHUTSUNG LIAO

The uterus of the immature rat (Jensen and Jacobson, 1962) or mouse (Stone, 1964) can retain estradiol firmly, but not other free or conjugated metabolites which are readily formed in liver and present in blood. Hexestrol, diethylstilbestrol, 17a-methylestradiol, 17aethynylestradiol, and estriol can also be retained by target tissues, all without prior chemical transformation (Jensen et al., 1967a,b; Laumas et al., 1970). Free l‘lphydroxyl and phenolic groups appear to be needed for uterine retention of estradiol (Noteboom and Gorski, 1965).Estrone retention by target tissues is less distinct than retention of estradiol. A portion of the estrone administered is reduced to estradiol and is retained. Autoradiographic and cell fractionation studies have shown that e ~ t r o g e n - ~localizes H in the nucleus as well as in the cytoplasm. Most of the radioactive estrogen retained by the uterus of immature rats is normally found in the cell nuclei (Talwar et al., 1964, 1968; Noteboom and Gorski, 1965; Jensen et al., 1968).Autoradiographic pictures show that nucleoli contain very little radioactive estrogen (Stumpf, 1969), but Arnaud et al. (1971) used a fractionation technique to demonstrate that the nucleolus can be a major site of estradiol retention. The presence of estradiol-binding protein(s) in the high-speed supernatant (cytosol) was first indicated by the failure of Sephadex gel, which excludes proteins but retains free steroids, to retain radioactive estrogen (Talwar et al., 1964). Toft and Gorski (1966) then found that, on ultracentrifugation in a sucrose gradient, the estradiolprotein complex sediments as a discrete band with a sedimentation coefficient of 9.5s. The protein appeared to be identical with the 8s components reported later by other investigators (Erdos, 1968; Rochefort and Baulieu, 1968; Korenman and Rao, 1968). These workers found that, when KCl or NaCl is present in the sample and gradient at concentrations of 0.2 M or higher, the 8s complex is reversibly transformed into a lighter complex which migrates slightly more slowly than bovine plasma albumin, or at about 4s. Addition of calcium ions and salt to uterine cytosol, prepared in the presence of EDTA, yields a “stabilized” 4s binding unit which after removal of salt does not revert to the 8s or larger aggregates (DeSombre et al., 1969; Jensen et al., 1969).Puca et al. (1971, 1972) have reported that the 8.6s form actually sediments as 5.3s in high concentrations of KC1 (>0.2 M).They consider 8.6s to be a dimer of the 5.3s form which may have a molecular weight of about 118,000. They have also presented evidence (Puca et al., 1972; Bresciani et al., 1973) showing that the 4.5s calcium-stabilized form (MW 61,000)

CELLULAR RECEPTORS FOR STEROID HORMONES

95

is formed from the 5.3s complex by a Ca2+-activated “receptortransforming factor.” The factor appears to be a protease which promotes transverse cleavage of the 5.3s complex into two equimolecular fragments. Estradiol retained by the uterine nuclear fraction appears to be associated with a protein that can be solubilized by extraction with 0.3-0.4 M KCl at pH 7.4-8.5 (Jungblut et al., 1967; Puca and Bresciani, 1968). The solubilized estradiol-receptor complex sediments at about 5S, somewhat more rapidly than bovine plasma albumin, and therefore can be clearly distinguished from the 4 s cytosol complex (Jensen et al., 1969). The 5s complex, like the cytosol 4 s unit, can aggregate to the 8-9s form if salt is removed (Korenman and Rao, 1968). According to Steggles and King (1970), the uterus of mature rats contains a specific 4 s estrogen receptor that does not associate to form 8s. This 4 s complex is not found in ovariectomized rats, and therefore is not identical with the 4 s complex described by other workers. Such a qualitative difference in the uterine 4 s entity in mature, in ovariectomized, or in immature animals has not been detected in other laboratories (Chamness and McGuire, 1972; Vondehaar et al., 1970). Rochefort and Baulieu (1972), however, have reported the presence of the 4s form in uterine or nontarget tissue nuclei previously incubated with estradiol-labeled uterine cytosol. The nuclear 4s complex also does not revert to 8 s and differs from the cytosol 4 s and nuclear 5 s forms. DeHertogh et al. (1973a) have described a 3.5s estradiol-protein complex that can be extracted from crude uterine nuclei by tris-EDTA buffer (without KC1). The reported association constant ( K a ) for the cytosol complex from uteri of various species has been generally in the range of 10y-lO1o M-’ if analyzed by Sephadex gel filtration (Giannopoulos and Gorski, 1971b; Lee and Jacobson, 1971), charcoal absorption (Korenman, 1970),sucrose gradient centrifugation (Toft et al., 1967), equilibrium dialysis (Truong and Baulieu, 1971), or other techniques (Steggles and King, 1970; Notides, 1970) that separate bound from unbound steroid. Hahnel and Twadale (1973) have estimated the K, of the estradiol-receptor complex in human breast cancer to be 4 X 1OI2M-I. K, for the nuclear complex (by gel filtration method) has been shown to be about lo9M-’(Puca and Bresciani, 1969).Alberga et al. (1971), however, have claimed the presence of a nuclear receptor with an exceptionally high affinity for estradiol (K, l O I 4 A 4 - I ) in the nonhistone chromatin protein fraction. Techniques involving kinetic analyses for steroid association and dissociation have also indicated K, val-

96

SHUTSUNG LIAO

ues of 101’-1012M-’ (Best-Belpomme et al., 1970; Jensen and DeSombre, 1972). Such a high affinity has also been predicted from the study of estrogen uptake and retention in uiuo (DeHertogh et al., 1971) and in uitro (Alberga and Baulieu, 1968). The purification of receptor proteins for various steroid hormones generally faces two difficulties, instability and tendency toward aggregation. In addition, the quantities of receptor protein in target tissues are very low, and it is necessary to purify 100,000-fold to achieve a pure state. In earlier unsuccessful efforts, affinity chromatographic techniques involving estradiol-linked benzylcellulose (Jungblut et al., 1967; Jensen et al., 1967b) and to poly(viny1-Nphenylene maleimide) (Vonderhaar and Mueller, 1969) were employed. Sica et al. (1973a,b) also found that various agarose derivatives containing estradiol covalently bound through linkages with the A ring of the steroid molecules are ineffective as adsorbents for receptors. These investigators, however, found that agarose derivatives containing 19-nortestosterone 17-hemisuccinate can selectively bind estradiol receptors. The most useful adsorbents were prepared by attaching 17pestradiol 17-hemisuccinate to agarose derivatives containing albumin or the branched copolymer of poly(L-lysine) as a backbone and poly(DL-alanine) as side arms. These macromolecular “leashes” presumably allow the attached ligand to stay very distant from the agarose and also, by possessing many functional groups, permit a high degree of ligand substitution. By using this new group of adsorbents, the cytosol estradiol receptor can be purified between 10,OOO- and 100,000-fold with an overall yield of 30-50% in a single step. The purified Ca”+-stabilized receptor sedimented as the 4.5s form. The electrofocusing pattern showed two major peaks at pH 6.6 and 6.8. The estradiol-binding activity was destroyed by heating for 5 minutes at 65°C. Truong et al. (1973) also employed somewhat different affinity columns to purify partially the Ca2+-stabilized4 s estradiol-receptor complex of calf uteri. A specific activity of the purified material reached 167 pmoleslmg protein, but the recovery from the column (10-40%) was low. Using more conventional techniques involving ammonium sulfate precipitation, Sephadex G-200 gel filtration, and DEAE-cellulose chromatography, DeSombre et al. (1969, 1971) purified the Ca2+stabilized 4 s complex about 5000-fold (ca. 5% pure). Further purification by acrylamide gel electrophoresis yielded a material showing a single radioactive protein band by amido black staining (De Sombre et al., 1971).The nuclear form of the cytosol complex (or the transformed cytosol receptor; see Section IV,A) has also been puri-

CELLULAR RECEPTORS FOR STEROID HORMONES

97

fied in a similar manner to a homogeneous component by gel electrophoresis and analytical ultracentrifugation. The purified complex sediments at 4.8s in sucrose gradients, with or without salt, and shows an isoelectric point (PI) of 5.8 and a Stokes radius of 36.5,indicating a molecular weight of about 72,000. It separates on gel electrophoresis from the Ca2+-stabilizedform of the cytosol receptor. Its amino acid composition has been determined (Gore11 et al., 1974). 2. Biodynamic Aspects In rat uterus the receptor protein for estradiol is present as early as the first day of life. The estradiol-receptor complex also can be retained by uterine nuclei at this time. A sudden induction of the receptor proteins therefore is not the trigger for puberty (McGuire and Lisk, 1971). Somjen et al. (1973b) have shown that the receptor content reaches its peak at the age of 10 days (see also Clark and Gorski, 1970). The specific estrogen-induced protein found by Gorski (see Section V,A) is not found in 5-day old rats, but its rate of synthesis (after an injection of estradiol) at 10 days is about twothirds of that observed at 20 days. Since the rapid estrogen stimulation of amino acid incorporation into protein fractions does not occur until the fifteenth day, Somjen et al. (1973a) believe that the nuclear binding of the estradiol receptor and the synthesis of the induced protein may be necessary but not sufficient conditions for some trophic action of estradiol. In the immature rat uterus, there are about 100 fmoleslmg cytosol protein of 8s estradiol-receptor complex (Jensen et al., 1968). Estradiol can enhance the synthesis of its receptor as the general protein-synthesizing activity of the uterine cells increase. Such an effect is a relatively slow process compared to the more rapid effects of estradiol, which occur well within the first 30 minutes (see Section V,A). There is, furthermore, no rapid loss of receptor protein after the animals are deprived of estrogens. SarfF and Gorski (1971) estimated the half-life of the receptor proteins as about 138 hours. Estradiol bound to the receptors in the cytoplasm or nuclei appears to be able to exchange rapidly with the circulating hormone, as DeHertogh et al. (1973a,b,c) have indicated by a careful infusion technique. Whether this is secondary to a recycling of the receptor molecule is not clear. Estrogen treatment of the uterus in vivo or in vitro is known to promote rapid depletion of cytoplasmic receptors, which is accompanied by a parallel increase in the nuclear estradiol receptor content (Jensen et al., 1968; Giannopoulos and Gorski, 1971a). Ac-

98

SHUTSUNG LIAO

cording to Somjen et al. (1973a,b), cell nuclei of rat uterus during postnatal development contain about 3000 to 7000 receptor binding sites per diploid cell nucleus. Estradiol injection increases this number to as much as 38,000. The concentration of the nuclear estrogen-receptor complex in rat uterine cells closely parallels the level of ovarian estrogen during the estrus cycle (Lee and Jacobson, 1971). Calculation from the data of Clark et al. (1972) shows that in each diploid cell the number of cytoplasmic estradiol receptors increases moderately from 5580 sites in metestrus to 8100 sites in proestrus. The changes in the number of receptors retained by uterine nuclei are more dramatic: 1000 (estrus and metestrus), 3500 (diestrus), and 5OOO (proestrus) sites per diploid nucleus. (The number of molecules in each nucleus can be calculated by assuming 2.5 pg DNA per diploid nucleus for chick cells and 6 pg DNA per diploid nucleus for a mammalian cell. If there are a femtomoles per microgram DNA and each nucleus contains b picograms DNA, the number of molecules per nucleus is a X b X 600.) Since estrogenic responses are significantly higher in proestrus than in metestrus or diestrus, the fluctuations in the nuclear receptor contents appear to correlate well with biological activities. In the hypothalamus, pituitary, and uterus of rat and hamster (Lisk et al., 1972), estradiol retention is also low just prior to ovulation when the animal is producing much estrogen, and high early in the cycle when very little estrogen is present in the system.

3. Universality Estradiol receptors have been most extensively studied in the uterus of experimental animals such as the rat, calf, and pig (Little et al., 1972), and also in human beings (Siiteri et al., 1972, 1973). Similar proteins have also been found in other estrogen-sensitive tissues such as the vagina, ovary, mammary gland, pituitary, and hypothalamus, as well as in the kidney, chick liver, pancreas (Sandberg et al., 1973), and testis (Brinkman et al., 1972). As noted later, it is debatable whether all the known estrogenic actions involve the same receptor mechanism. In this connection, it is important to note that King and Thompson (1974),after reviewing many estrogen-sensitive systems, concluded that there is no reason to suspect that different end-organ responses to estrogens are due to different forms of active estrogens. They pointed out that in most systems estradiol is the most potent natural estrogen. Jensen and Jacobson (1962) suggested that the action of estrone is due to its conversion to estradiol. Other studies in uiuo also show

CELLULAR RECEPTORS FOR STEROID HORMONES

99

that estrone retained by the mouse uterus (Stone and Martin, 1964) or vagina (Martin and Baggett, 1964a) declines more rapidly than estradiol. In the vagina most of the radioactivity retained at 25 minutes after local application of e ~ t r o n e - ~ can H also be identified as estradioh3H (Martin and Baggett, 1964b). It has further been shown that e ~ t r o n e - ~can H bind to the same receptor that binds e~tradiol-~H, but that, at the low concentrations of steroids under which the estradi~l-~H-receptorcomplex can be transformed to the one retainable by uterine nuclei (see Section IV,A), the e~trone-~Hreceptor complex is not retained by the same nuclei. Whether estrone itself has estrogenic activity in target cells has been reinvestigated recently by Ruh et al. (1973).This study measured the relative activity of the three estrogens in uitm in stimulating the synthesis of an estrogen-inducible protein (IP) during incubation of uteri. Quantitatively, the induction for all three steroids closely paralleled the specific uptake of the steroid for cytoplasmic estrogen-binding protein (estradiol > estriol > estrone). Since during the incubation estradiol was not formed to any significant amount from estriol or estrone, it was concluded that each of these estrogens in its own right is a biologically active estrogen. Ruh et al. (1973) were not able to detect the formation of a nuclear estrone-protein complex, but this was thought to be due to rapid dissociation and not to the inability of estrone to form a biologically active complex. Geynet et al. (1972), however, have been able to find estrone and estriol-receptor complexes (both as 4 and 5s forms) in the extract of uterine nuclei previously incubated with cytosol and radioactive estrone or estriol. For human uterus, translocation of the cytoplasmic estradiol- or estrone-receptor complex to the nuclei can be demonstrated in uitro (Siiteri et al., 1972, 1973). One property often considered unique is that estriol, which is as active as estradiol in stimulating uterine water imbibition, is much less potent than estradiol in increasing uterine dry weight. Anderson et al. (1973a), however, believe that there is a good correlation between the amount of nuclear estrogen receptor in the uterine cells and the extent of water imbibition and increase in uterine weight, regardless of which estrogen (estradiol, estrone, or estriol) is used, Another interesting case is the glucose oxidation response, which can be elicited rapidly by quantities of estrone, estradiol, and estriol that are insufficient to produce maximal quantities of nuclear estrogen receptor in the uterus. This observation may indicate that the nuclear estrogen receptor complex is not involved, or that the quantity of the receptor molecules required is only a fraction of the total

100

SHUTSUNG LIAO

number available (Anderson et al., 1973a). Sodium retention occurs during pregnancy, while plasma estrogen levels are high and estrogen at high doses can decrease the daily sodium excretion rate in humans and dogs (Johnson et al., 1970) and in adrenalectomized male rats (DeVries et al., 1972). Based on a study of steroid uptake and exchange, rat kidney appears to have specific and high-affinity (K,6.1 X 1OloW1)receptors for estrogens. The binding capacity appears to be low, the amounts being less than 10%of that found in rat uterus. A unique type of estrogen retention has been considered by Szego’s group, who showed that lysosome organelles of target organs of rats can retain estradioL3H very rapidly in uivo or in cell-free systems. Since protein-bound estradiol can be liberated from lysosomes, it was suggested that the formation of specific estrogen-protein complexes occurs in lysosomes and is involved in certain actions of estrogens (Szego, 1972; Szego and Seeler, 1973). By acridine orange fluorescent microscopy, it was observed that lysosomal components were present in high concentrations in nuclear fractions of preputial glands (and uteri) of rats pretreated with estradiol, diethylstilbestrol, or testosterone, but essentially absent in preparations obtained from animals injected with control (no hormone) or with 17aestradiol (nonestrogenic) solutions. The estradiol-induced response was reported to occur within 1 minute of hormone administration and was not observed in preparations from lung. Enzymic analysis also showed an increased release of hydrolytic enzymes from lysosomes, and nuclear metachromasy, upon hormone injection, suggesting that the nuclear surface andlor the intranuclear constituents might be modified. These effects were reported to be inhibited when the animals were treated with cortisol. Szego (1971, 1972) has proposed that alteration of the lysosomal membrane and the subsequent release of active agents such as vasoactive amines and hydrolases are responsible for multiple cellular events. She also believes that the primary effect of the intracellular action of estrogen receptors is reorientation of the cytostructure of the involved membranes. Such an effect on the plasma membrane is presumed to cause alteration of the cellular levels of cyclic AMP (see Section V1,E). Another type of estrogen retention presented by Tchernitchin (1973) and Tchernitchin and co-workers (1973) suggests that estrogens are retained somewhat specifically by rat uterine eosinophils. These investigators believe that the binding of estrogens by receptor molecules in uterine eosinophils is responsible for some of the early estrogenic responses, such as water imbibition, histamine

CELLULAR RECEPTORS FOR STEROID HORMONES

101

release, and estrogen-priming effects, and is different from the cytosol-nuclear (8s-5s)system. Williams, Rabin, and their co-workers (Sunshine et al., 1971) have also described an unusual observation in which estradiol in males and testosterone in females promoted binding of polysomes to smooth microsomal membranes of rat liver in uitro. Such a sex-specific effect was reported to accompany selective binding of these steroid hormones to microsomal membranes (Blyth et al., 1971, 1972). The possibility that there are multiple sets of receptor systems involved in different types of steroid hormone action deserves more study. The interesting observations made with lysosomes, eosinophils, and microsomal membranes need confirmation and further characterization before their significance in estrogen action can be properly evaluated. B. ANDROGEN RECEPTORS 1. Zdenti$cation and Properties of Prostate Receptor Earlier attempts to show selective retention of androgen by target tissues was rather difficult, principally because of the need to inject into experimental animals large quantities of androgens with a low specific radioactivity (see a review by Liao and Fang, 1969). Nevertheless, studies in rats by Pearlman and Pearlman (1961) and Harding and Samuels (1962) had indicated clearly that, while large amounts of conjugated metabolites could be found in the blood and liver, the prostate could accumulate unconjugated androgen metabolites. The uptake and retention of radioactive androgen by various androgen-sensitive tissues were later reinvestigated most extensively by a Norwegian group (Tveter and Attramadal, 1968; Tveter and Aakvaag, 1969) and also by other groups (Anderson and Liao, 1968; Fang et al., 1969; Bruchovsky and Wilson, 1968a; Belham et al., 1969; Mainwaring, 1969a). In these studies rats were injected with highly radioactive testosterone, and clear retention was observed for ventral and dorsal prostates, seminal vesicles, and injeccoagulating gland ?h-3 hours from the time of test~sterone-~H tion. The radioactivity in the blood, spleen, lung, thymus, and diaphragm is rapidly cleared within the first hour. In the male sex accessory organs, radioactive steroids can be detected even 16 hours after injection. Two or three hours after the injection, however, essentially all the radioactivity retained can be identified as dihydrotestoset al., 1969). The prolonged retention of dihydrotest e r ~ n e - ~(Fang H tosterone by these tissues is due mainly to the selective retention of

102

SHUTSUNG LIAO

this androgen by the cell nuclei of prostate (but not liver, thymus, or diaphragm), a phenomenon detectable within minutes after testoster~ne-~H reaches the male accessory glands (Bruchovsky and Wilson, 1968a; Anderson and Liao, 1968). These findings made by biochemical fractionation techniques agree well with autoradiographic observations by Tveter and Attramadal (1969)and Sar et al. (1970). In addition, the selective retention of dihydrotestosterone by cell nuclei of target tissues can be demonstrated by incubating minced tissue and radioactive androgens (Anderson and Liao, 1968; Fang et al., 1969), which provides a convenient system for in uitro study. The discovery of selective dihydrotestosterone retention by prostate cell nuclei immediately raised the possibility that dihydrotestosterone may be an active form of cellular androgen (Wilson and Gloyna, 1970; Liao and Fang, 1969). Earlier data showing that dihydrotestosterone is more active than testosterone in several bioassay systems, including the growth of rat prostates (cf. Liao and Fang, 1969), is also consistent with this proposal. Strong support comes from the finding that potent antiandrogens such as cyproterone (Fang and Liao, 1969) and flutamide (Peets et al., 1974; Liao et al., 1974a) can inhibit the retention of dihydrotestosterone by prostate tissues or nuclei in uiuo or in uitro. The dihydrotest~sterone-~H associated with prostate cell nuclei is tightly bound to nuclear proteins (Bruchovsky and Wilson, 1968b; Liao, 1968) and can be extracted with 0.4-1.0 M KCl solutions (Bruchovsky and Wilson, 1968b; Fang and Liao, 1969; Fang et al., 1969; Mainwaring, 1969a). The nuclear dihydrotestosterone-protein complex migrates as 3s in 0.4 M KC1 media (Fang and Liao, 1969; Fang et al., 1969). At low concentrations of radioactive androgens, cytosol dihydrotestosterone is exclusively bound to a high-affinity, lowcapacity protein and forms a complex which migrates as 3.5s either in 0.4 M KCI or in a medium with low ionic strength. An 8-9s complex can be also identified in the cytosol fraction exposed to dihydrotest~sterone-~H in uitro or obtained from rats injected with test~sterone-~H (Unhjem et al., 1969; Mainwaring, 1969b; Fang et al., 1969; Tveter et al., 1971; Baulieu and Jung, 1970). The 8s complex dissociates into components of about 3-4s in 0.4 M KCl. In the cytosol of rat ventral prostate, there are at least two proteins that bind dihydrotestosterone. One of them can be precipitated by the addition of ammonium sulfate to 40% saturation in respect to the salt (Liao and Fang, 1970). The protein ( p protein) exhibits an extremely high affinity (our recent data show K, 10l2 M - l ) and

CELLULAR RECEPTORS FOR STEROID HORMONES

103

specificity toward dihydrotestosterone and several synthetic androgens such as 7a,l7a-dimethyl-19-nortestosteroneand 2-oxa-17amethyl-l7/3-hydroxyestra-4,9,1 l-triene3-one, but not toward inactive steroids with related structures (5pdihydrotestosterone, 17a-dihydrote~tosterone)~ estrogens, progesterone, or corticosteroids Other natural androgens such as andros(Liao et al., 1972, 1973~). tenedione and 3p, 17P-dihydroxy-5a-androstanealso do not bind firmly to p protein, but they can be converted to dihydrotestosterone in the prostate (Bruchovsky, 1971). Another protein (called a protein) in the prostate cytosol fraction sediments at the ammonium sulfate concentration of 55-70% saturation. This protein binds dihydrotestosterone as well as testosterone, progesterone, and estradiol, but not cortisol (Fang and Liao, 1971). For convenience, the complexes of dihydrotestosterone and a or fl protein are designated complex I or complex I1 (Liao and Fang, 1970). In 0.4 M KCl solution, both cytosol complexes have sedimentation coefficients of about 3.5s. In the absence of KCl, complex I1 (but not complex I), gradually aggregates to larger forms. The aggregation of the small binding unit (3.5s)appears to involve other cellular components, since recentrifugation or partial purification of complex I1 tends to minimize the extent of aggregation (Liao et al., 1971b). In whole cytosol or in a crude ammonium sulfate preparation, the sedimentation properties of complex I1 vary in media with different salt concentrations but, more importantly, with different pH. Our recent study shows that, if the pH of the medium at 2°C is raised from 7.0 to 9.0, the amount of aggregates decreases and large 8s and small 3.5s forms emerge. At pH 9.5 only a 7s form is clearly present. At a higher pH most of the steroid is released from the proteins, whereas if the pH is lowered from 7.5 to 6.0 more aggregates appear with nearly complete loss of the 8 and 3.5s peaks. As shown later, only complex I1 (but not complex I) can be the precursor of the 3s steroid-receptor complex retained by prostate nuclei. The formation of complex I1 and nuclear retention of the complex in an in vivo or in vitro cell-free system can be diminished by antiandrogens such as cyproterone (Fang and Liao, 1969,1971) or by flutamide (Peets et al., 1974; Liao et al., 1974a). Other workers also have found that other antiandrogens can interfere with the binding of dihydrotestosterone to specific androgen-binding protein (Tveter and Aakvaag, 1969; Baulieu and Jung, 1970; Mangan and Mainwaring, 1972). These observations, steroid specificity and highaffinity binding as well as tissue specificity (e.g., complex I1 is not present in the liver, spleen, thymus, diaphragm, or blood, where

104

SHUTSUNG LIAO

androgens have very limited effects), suggest strongly that complex I1 is the specific androgen-receptor complex possibly functioning in the cells. Whether complex I is a precursor or a degradation product of complex I1 has not been determined. Various analytical properties of the dihydrotestosterone receptor proteins of rat ventral prostate have been well studied by Mainwaring and his associates. Using DNA-cellulose chromatography and isoelectric focusing, Mainwaring and Irving (1973) purified the cytosol receptor protein about 2000-fold or higher. The purified dihydrotestosterone-3H-receptor complex still had a sedimentation coefficient of 8s and, more importantly, was retained by prostate nuclear chromatin (see Section IV,C). The same investigators also employed the same procedure for the partial purification of the nuclear dihydrotestosterone-receptor complex. 2. Biodynamic Aspects The amount of receptor protein for dihydrotestosterone in the prostate cells diminishes after rats are castrated (Mainwaring, 1970; Baulieu and Jung, 1970; Mainwaring and Mangan, 1973).This loss is gradual (Liao et al., 1971b) and closely follows the rate of regression of the prostate after animals have been deprived of androgens (Sullivan and Strott, 1973). The immediate action of dihydrotestosterone after it reaches a target cell is therefore likely to depend on the existing cellular receptor rather than on a specific dihydrotestosteronedependent induction of the receptor protein. Our estimate of the half-life of the prostate receptor protein for dihydrotestosterone is about 3-4 days, which is similar to that for estradiol receptor in rat uterus (see Section III,A,2) and progesterone receptor in the uterus of the guinea pig (see Section III,C,2). The receptor contents of the ventral prostate of rats of different ages have been studied by Shain and Axelrod (1973). In 60-day-old rats, the amount of cytosol receptor (based on the appearance of the 10-11s form) was estimated to be of the order of 30 fmoleslmg protein, whereas in 14- to 23-month-old rats no receptor was detected. Our estimate of dihydrotestosterone receptor (complex 11) in the cytoplasm, including that bound to microsomal fractions (30%), in many groups of adult rats is 100 h o l e s +40 fmoleslmg protein, corresponding to about 4000 to 10,OOO binding sites per cell. The prostate nuclei, in oioo and in oitro, have the ability to retain 2000 to 6000 receptors per cell nucleus (Liao and Fang, 1970; Liao et al., 1971b, 1972; Mainwaring and Peterken, 1971; Rennie and Bruchovsky, 1972). Since not all the cell nuclei of rat ventral prostate re-

CELLULAR RECEPTORS FOR STEROID HORMONES

105

tain equal amounts of dihydrotestosterone, the maximum capacity of certain nuclei to retain receptor may be considerably higher than the values shown. Because testosterone, the major testicular androgen in blood, may be converted to dihydrotestosterone before receptor binding of the androgen (and thus function) can occur in the prostate, the levels of both the androgens and the converting enzyme (including NADPHdependent A4-3-ketosteroid-5~oxidoreductase; Shimazaki et al., 1965; Ofner, 1968; Moore and Wilson, 1972), at the local areas of target tissues can affect androgenic activity. Wilson and Gloyna (1970)concluded, from an assay of reductase activity in the accessory sexual tissues of many species and in human skin from a variety of anatomical sites, that dihydrotestosterone-forming activity is correlated with androgenic activity in many cases but is not clearly an obligatory feature of all androgen actions. In humans the dihydrotestosterone content has been found to be significantly greater in hypertrophic than in normal prostates, but the differences are not directly related to variations in the rate of dihydrotestosterone-forming activity. In the periurethral urea, where prostatic hypertrophy usually commences, the dihydrotestosterone content is reported to be two or three times greater than in the outer regions of the gland. In some cases of prostatic carcinoma in humans (Giorgi et al., 1971, 1972) and in the dog (Gloyna et al., 1970), the conversion of testosterone to dihydrotestosterone is indeed found to be small. It should be pointed out that the blood level of dihydrotestosterone has been shown to be significant, being 10-30% of that of blood testosterone in adult humans (It0 and Horton, 1970; Tremblay et al., 1970). Therefore in many target tissues the local concentration of dihydrotestosterone may be determined by the receptor proteins; and even in the tissues that lack reductase, significant androgenic action can occur. In a series of very careful studies, Bruchovsky and his associates examined the cellular concentrations in rat ventral prostate of dihydrotestosterone originating from various natural androgens. It appears that in these animals, dihydrotestosterone is formed from several androgenic precursors (Bruchovsky, 1971), and that more dihydrotestosterone is found in the prostate with androgens of high potency than with androgens of low potency. The intracellular distribution kinetics of androgens in the prostate have been studied by a pulse-chase method (Rennie and Bruchovsky, 1973). The relationship between androgen-binding and androgen insensitivity in the testicular feminization (Tfm) syndrome has been studied in recent years. In male pseudohermaphroditic rats, the cell nuclei of

106

SHUTSUNG LIAO

the preputial gland (Bullock and Bardin, 1970), liver, and kidney ( R i t z h et al., 1972) appear to have reduced ability to concentrate dihydrotestosterone. In mice the amount of dihydrotestosteronebinding protein in the cytoplasm and nuclei of the kidney has been found to be distinctly less in Tfm than in normal mice (Gehring et al., 1971). However, the uptake of radioactive androgen by the kidney (Bullock et al., 1971) and by the submandibular gland (Goldstein and Wilson, 1972) in the Tfm animals appears to be normal. In fact, there may be more androgen-binding protein in the submandibular glands of Tfm animals than in the same glands of normal male mice (Wilson and Goldstein, 1972; Dunn et al., 1973). Nevertheless, a defect in the cytosol androgen receptor in the kidney of Tfm mice is possible, since 7.5s is found in normal male or female mice but not in Tfm animals. Such a defect may be responsible for the decreased nuclear uptake of androgens. The androgen-dependent mouse mammary carcinoma (Shionogi 115 tumor grown in males; Yamaguchi et aZ., 1971) also contains androgen-binding proteins. By transferring the androgen-dependent cells to female mice, Bruchovsky and Meakin (1973) obtained androgen-insensitive cells. They found that four times more androgen (mainly testosterone) was bound to the receptor in the cytosol of androgen-dependent cells than in the autonomous tumor. An identical finding was made by Mainwaring and Mangan (1973),who studied the dihydrotestosterone-binding component with a sedimentation coefficient of 8s. Both groups indicated that the reduced concentration of cytoplasmic receptors appeared to be responsible for the impaired incorporation of androgens into nuclei of the insensitive cells. 3. Universality The suggestion that dihydrotestosterone receptor protein is involved in the actions of androgens in wide ranges of target tissues is supported by the finding of similar proteins in many androgen-sensitive tissues such as seminal vesicles (Tveter and Unhjem, 1969; Stem and Eisenfeld, 1969; Liao et al., 1971b), hair follicles (Fazekas and Sandor, 1973), sebaceous and preputial glands (Adachi and Kano, 1972; Eppenberger and Hsia, 1972; Bullock and Bardin, 1970; Mainwaring and Mangan, 1973), uterus (Jungblut et al., 1971), kidney (Gehring et aZ., 1971; RitzCn et al., 1972), submandibular gland (Goldstein and Wilson, 1972; Dunn et al., 1973), brain (see Section II1,F) and androgen-sensitive tumors (see Section III,B,2). The uptake of radioactive androgens (Gorgi et al., 1971; Hansson and

CELLULAR RECEPTORS FOR STEROID HORMONES

107

Tveter, 1971)and the presence of a specific receptor for dihydrotestosterone in human prostate have been indicated (Hansson et al., 1971, 1972; Geller and Worthman, 1973; Mainwaring and Milroy, 1973; Fang, 1973). The dihydrotestosterone-receptor complex of human prostate nuclei sediments as 3s in the medium containing 0.4 M KCl (Castaiieda and Liao, 1974b). Rat epididymis cytosol contains at least two dihydrotestosteronebinding proteins ( R i t z h et al., 1971).One of these appears to originate in the testis and is transported to the epididymis in efferent duct fluid. The other component appears to be formed in the epididymis. The nuclear uptake and binding of dihydrotest~sterone-~H can occur at a time when the testicular binding component is no longer present in the epididymis (Tindall et al., 1972).The cytosol dihydrotestoster~ne-~H-binding protein has been described as a 3-5s unit by Hansson and Djoseland (1972)and Tindall et al. (1972),but the presence of a 8.5s form has been detected (Blaquier and Calandra, 1973). Rat testis contains, in addition to the androgen-binding protein that binds both testosterone and dihydrotestosterone, a specific dihydrotestosterone-binding protein in the cytosol and nuclei. The protein is similar to that reported for ventral prostate (Hansson et al., 1973). No clear-cut evidence has yet been obtained for the existence of a dihydrotestosterone receptor in muscles which are also androgen target tissues. Jung and Baulieu (1972)found an 8s testosteronebinding protein that also binds dihydrotestosterone to some extent in the rat levato ani muscle. They speculate that the testosteronereceptor complex may be associated with myotrophic activities, whereas the dihydrotestosterone-receptor complex is involved in androgenic activities. Giannopoulos (1973a) also detected testosterone-binding protein(s) in the cytosol and nuclei of immature rat uterus. The protein(s) also bind dihydrotestosterone, but not progesterone or cortisol. The binding is antagonized slightly by estradiol, but not by antiandrogen (cyproterone acetate). Cellular testosteronebinding proteins that also bind dihydrotestosterone well have been found in the ventral prostate (Fang et al., 1969; Rennie and Bruchovsky, 1972) and the kidney (RitzBn et al., 1972). Testosteronebinding proteins have also been detected in rat anterior hypophysis and bovine spermatozoan preparations (Wester and Foote, 1972) as well as chick oviduct (Harrison and Toft, 1973). Further study is needed before these proteins can be identified as the specific cellular receptors for active androgens. Arguing from the retention-competition data for rabbit prostate, Kasuya and Wolff (1973)suggested that each of the three androgens 17P-hydroxy-5a-androstan-17a-

108

SHUTSUNG LIAO

TABLE I1 ANDROGEN ACTIONS NOT ATTRIBUTED TO DIHYDROTESTOSTERONE (DHT)" System Vagina, epithelial cells

Uterus

Prostate organ culture Seminal vesicle

Muscle Sexual behavior

Observation 3a-Hydroxy and 3-keto-androstanes stimulate the production of mucus by the superficial cells; 3p-hydroxy steroids affect deeper layers* Testosterone, but not DHT, stimulates glandular secretion and increases the height of the luminal epithelium' DHT and not 3p-androstanediol increases cell proliferation, but both androgens maintain epithelial cell growth The secretory output of fructose and citric acid is stimulated more by testosterone than by DHT No DHT receptor protein has been detectede.d Testosterone is active, but DHT is not in some speciesf

Blood cell

Erythropoiesis may be affected by 5p-dihydrotestosterone

Anovulatory sterility

Testosterone is active, but DHT is not Differentiation is induced by testosterone when 5a-reductase activity is absent

Wolffian ducts

Reference Huggins et al. (1954)

Gonzalez-Diddi et al. (1972) Baulieu et al. (1968); Gittinger and Lasnitzki (1972) Mann et al. (1971)

Aakvaag et al. (1972) Whalen and Luttge (1971a); Beyer et al. (1973) Kappas and Granick (1968); Gordon et al. (1970) Whalen and Luttge (1971b) Wilson (1973)

Modified from Liao and Liang (1974). proteins for 3-hydroxy androstanes have been isolated from rat vagina (Shao et al., 1975). Receptorlike proteins which bind testosterone preferentially over DHT have been found in these tissues (see text). The testosterone action may be due to its aromatization to phenolic estrogens that can activate male sexual behavior in male (.Feder et al., 1974).

* Binding

methyZJ4C, 17~-hydro~y-5a-androst-2-en-17a-methyZ-~~C, and dihydrotestosterone, binds to a different receptor site, although no androgen receptor protein has been isolated from this tissue, The possibility that some androgen actions may not be attributed to the receptor mechanism involving dihydrotestosterone alone is in fact strong. Some of the observations supporting, but not necessarily proving, this view are summarized in Table 11.

CELLULAR RECEPTORS FOR STEROID HORMONES

109

Further studies on these systems may show that varieties of androgen receptor molecules function either in the same or different cells in somewhat different manners. C. PROGESTINRECEPTORS 1. Identification, Properties, and Universality

The biological effects of progestational compounds are often dependent on prior estrogenization. In line with this fact, estrogen has been reported to increase the uptake of progesterone in the vagina of the ovariectomized mouse (Podratz and Katzman, 1968), and in the uterus of the hamster (Reuter et al., 1970; Leavitt and Blaha, 1972), guinea pig (Milgrom et al., 1970; Falk and Bardin, 1970), rabbit (Wiest and Rao, 1971), and rat (Davies and Ryan, 1972; Safian et al., 1973), and in chick oviduct (O’Malley et al., 1969; Toft and O’Malley, 1972). Administered pr~gesterone-~H undergoes transformation in the animals to 5a-pregnane-3,20-dione and other polar metabolites (Falk and Bardin, 1970; Wichmann, 1967; Armstrong and King, 1970) but, 1-3 hours after the injection of proge~terone-~H, 80% or more of the radioactivity retained by the uterus of an estrogen-primed guinea pig (Falk and Bardin, 1970) and rabbit (Wiest and Rao, 1971) can be identified as progesterone. In ovariectomized rats, however, estrogen-insensitive accumulation of a polar metabolite has been observed (McGuire and DeDella, 1971). Wiest (1971) has carefully reviewed the distribution and metabolism of progesterone in the uterus and their possible relationship to the progestational response. He concludes that the capacity to elicit the response resides in the progesterone molecule itself and supports the suggestion (Munck, 1968) that progesterone metabolism represents a destruction of potency. Although Armstrong and King (1970)have suggested that specific intranuclear reduction of progesterone may be a possible reaction mechanism, such a proposal remains unsubstantiated. Whether the broad spectrum of biological activity can be said to be exclusively due to progesterone cannot be determined from the limited information available so far. Nevertheless, the suggestion that progesterone is a universal progestin that can act without metabolic transformation is supported by the finding of specific progesterone-binding proteins in a variety of target tissues. a. Chick Ouiduct. Progesterone receptor molecules have been studied most extensively in the chick oviduct by O’Malley and his associates (1971a,b, 1972). The binding components of oviduct cytosol in uitro and in uiuo have been isolated that bind pr~gesterone-~H

110

SHUTSUNG LIAO

and characterized by sucrose gradient centrifugation, polyacrylamide gel electrophoresis, and gel filtration on agarose. The radioactive steroid in the isolated complex was identified as progesterone. The steroid affinity appears in the order: progesterone > testosterone > 20a-hydroxy-4-pregnene-3-one > estradiol > cortisol > estrone > androstenedione. It is thus closely related to their relative potencies (Sherman et al., 1970). Progesterone-binding components can be clearly separated from the corticosteroid-binding globulin of chick plasma. In the absence of KCI, the major components sediment as 5 and 8S, and have molecular weights of about 100,OOO and 360,000, respectively. In the prescomplex sediments ence of 0.3 M KCl the pr~gesterone-~H-receptor as a single peak at about 4 s (Sherman et al., 1970). On a DEAEcellulose ion-exchange column, two progesterone-binding components (A and B) were identified. They have identical steroid specificity and binding kinetics (dissociation constant Kd 0.8 nM). It has been suggested that they may be subunits of the progesteronereceptor complex. However, A and B alone, or together, are not able to reconstruct the 8s components under a variety of conditions (Schrader and O'Malley, 1972). The progesterone receptor formed in uiuo or in oitro can be extracted from nuclei by 0.3 M KCI. The nuclear binding complex on gradient centrifugation (with 0.3 M KCI) sediments closely with the cytosol complex as 4s. b. Guinea Pig Uterus. A progesterone-binding protein that does not bind nonprogestogenic steroids has been found in the estrogenprimed guinea pig uterus but not in nontarget organs. In low salt solutions it sediments as a 7 s unit which is converted to a 3.5-4.5s form in high-salt media (Milgrom et al., 1970; Corvol et al., 1972). By competition and on gradient centrifugation, the progesterone binder can be distinguished from plasma proteins that bind progesterone. MacLaughlin et al. (1972) also found a 5.5s progesterone-binding unit that sediments considerably more heavily than the progesterone-binding globulin. Faber et al. (1972a,b) also reported a 7.6s receptor for the (presumably) same progesterone receptor; but in 0.3 M KCl, the complex is said to be inactivated rather than dissociated into a 3.5-5.08 subunit. According to Corvol et al. (1972), the affinity constant of the progesterone-binding proteins is 2.3 x 10" M-I. Kontula et al. (1972) described a progesterone-binding protein from pregnant guinea pig uterus, which was different from that found in the blood plasma of pregnant animals and from that induced by estrogen priming in uteri of nonpregnant guinea pigs. The protein binds progesterone and 5a-pregnane-3, 20-dione equally well, but

CELLULAR RECEPTORS FOR STEROID HORMONES

111

not pregnenolone, 20(a or P)-hydroxy-4-pregnan-3-one,or cortisol. Significant binding was, however, observed with deoxycorticosterone. The progesterone-protein complex sediments as 5 s in a lowsalt medium as well as in high ionic media. c. Rabbit Uterus. The cytosol and nuclear fractions of rabbit uterus also contain specific progesterone-binding proteins (McGuire and DeDella, 1971; Wiest and Rao, 1971). Prior to estrogen priming, bound progesterone in cytosol is associated only with a 4s unit, but estrogen pretreatment results in the appearance of an 8s form which dissociates into a 4 s unit in 0.4 M KC1. The receptor molecule of rabbit uterus (and human endometrium) does not bind testosterone, estradiol, corticosterone, or progesterone metabolites firmly, and can be separated from corticosteroid-binding globulin by ammonium sulfate fractionation. The protein has a K d of about 0.2 nM. The mass ratio (an indication of the degree of purification needed to obtain completely pure material) of progesterone receptor to other proteins in the cytosol of an atrophic rabbit uterus has been estimated to be 1 :320,000 when the receptor molecular weight is taken as 60,000 (4s). Estrogen treatment increases the relative mass concentration of the receptor to 1 :35,000 (Rao e t al., 1973). McGuire et al. (1974) recently studied the structural requirement for steroids to bind to rabbit uterine progestogen receptor. d. Other Target Tissues. The cytosol of rat uteri, like that of rabbit uteri, contains a progesterone-binding protein which binds progesterone and cortisol well and has properties similar to plasma corticosteroid-binding globulin (Milgrom and Baulieu, 1970).This protein is not present in appreciable amounts in the cytosol of kidney, intestine, or muscle, and appears to be an intracellular component of the uterus rather than an experimental contaminant of the plasma or interstitial fluid. A similar protein is also found in human myometrium (Kontula e t al., 1973). Rat uteri, however, contain a different protein, which binds progesterone firmly (Kd 0.1 nM) but not nonprogestational compounds, and is not present in either blood or the cytosol fraction of nontarget tissues (McGuire and DeDella, 1971; McGuire and Bariso, 1972). Feil et al. (1972)and Philibert and Raynaud (1973) showed that the protein migrates as 4 and 7s. The progesterone-binding protein of mouse uterus has been studied by several investigators (Stone and Baggett, 1965; Smith et al., 1970; Feil et al., 1972; Philibert and Raynaud, 1973). Similar proteins, probably related to corticosteroid-binding globulins have been also found in rat lymphosarcoma (Hollander and Chiu, 1966) and hepatoma tissue culture cells (Gardner and Tomkins, 1969).

112

SHUTSUNG LIAO

Pearlman et al. (1973) found a specific progesterone-binding component in human breast cyst fluid. The protein has a low binding affinity for estradiol, testosterone, cortisol, and corticosterone, and is distinctly different from corticosteroid-binding globulins. Verma and Laumas (1973)also have described a progesterone-binding protein in human endometrium and myometrium. 2. Biodynamic Aspects In the chick oviduct or uterine systems described above, the amounts of specific progesterone-binding proteins in the target tissues of immature or ovariectomized animals are increased by estrogen injection, which also enhances the biological effects of progesterone, In a thorough study of this aspect, Milgrom et al. (1973b) showed that, 6 hours after estradiol injection in guinea pigs, the content of 4.5s uterine progesterone-binding receptor alone (either per cell or per milligram of protein) is doubled. The heavy form (6.7s) is found at about 20 hours after injection. Since the effects of the estrogen are prevented by inhibitors of RNA and protein synthesis, a new synthesis of the receptor or protein factors may be involved. It is of interest that progesterone administered 20 hours after estrogen provokes a rapid fall in receptor concentration to less than 20% within 1 day. Freifeld et al. (1973) have reported that such an effect can be observed within 3 hours of progesterone administration. These effects of estradiol and progesterone or progesterone receptor content in the uterus are compatible with the cyclic changes observed physiologically in intact animals. For example, progesterone receptors in guinea pig uterus vary during the estrous cycle (Milgrom et al., 1972a,b). Their concentration peaks to about 40,000 binding sites per cell at proestrus, and then falls rapidly during estrus and postestrus to about 2500 sites at diestrus. At proestrus, the heavy form (6.7s) predominates, whereas only the light form (4.5s) is observed at diestrus. Davies and Ryan (1972, 1973) showed that in rat uterus there is a causal relationship between the appearance in pregnancy or disappearance prior to parturition of the uptake of progesterone and of the progesterone-binding protein. The progesterone-binding sites in myometrical cytosol increase (20,000 fmolelmg protein) at the beginning of pregnancy, but decrease to a minimum (5000 fmoleslmg protein) during the last week of gestation. These changes also appear to play a major role in regulating progesterone concentration in the myometrium. The half-life of the progesterone receptor in the uterus of the

CELLULAR RECEPTORS FOR STEROID HORMONES

113

guinea pig has been estimated to be about 5 days (Milgrom et al., 1973b). The rat uterine receptor probably also has such a long halflife since, after ovariectomy, the 7s receptor diminishes slowly and lasts a week or longer. The complete recovery of the receptor level is seen 4-9 days after estrogen injection (Feil et al., 1972). In the castrated rabbit uterine progesterone-binding sites have been shown to increase from 550 sites to about 3500 sites per cell after estrogen treatment. The progesterone-binding sites in uterine cytosol of the guinea pig and rabbit have been estimated to be about 5000 fmoles (Corvol et al., 1972) and 10,000 fmoles (McGuire and Bariso, 1972) per milligram of protein. In the chick oviduct the progesterone receptor binds various steroids. The binding affinity appears to have the same order of effectiveness in potency as avidin inducer for these steroids. The fact that both avidin induction and progesterone-binding sites are stimulated by a prior treatment of the chicks with estrogens indicates a functional role for the progesterone-binding protein.

D. GLUCOCORTICOID RECEPTORS 1. Identification, Properties, and Universality Demonstration of the existence of a specific glucocorticoid receptor in animal tissues has often been complicated by extensive metabolism of the hormone in the animal (especially in the liver) and the presence of the multiple species of proteins that bind natural corticosteroids. One of the most practical techniques is to study the protein binding of potent synthetic glucocorticoids, since many corticosteroid-binding proteins that apparently are not cellular receptors do not bind to some of them. a. Liver. As a major organ of steroid catabolism, the liver, a glucocorticoid target tissue, contains large amounts (as much as 90% of the hepatic radioactivity) of the metabolites formed from injected cortisoL3H. However, 20 minutes after the injection of c~rtisol-~H, only the macromolecular fraction of the liver cytosol and the nucleus contain unmetabolized cortisoL3H (Beato et al., 1969; Morey and Litwack, 1969; Singer and Litwack, 1971a; Litwack et al., 1971). In adH to the dition, c ~ r t i s o l - ~itself H is virtually the only ~ t e r o i d - ~bound cell nucleus (Beato et al., 1969). The relative unimportance of metabolites of cortisol is also apparent from the fact that although there is a marked sex difference in the metabolic patterns of steroid in liver, the corticosteroid induction of tyrosine aminotransferase in

114

SHUTSUNG LIAO

this tissue is equally effective in females and in males (Singer and Litwack, 1971b). Two cortisol-binding proteins were originally identified in liver. One of them (B protein, MW 64,000) is very similar if not identical to the corticosteroid-binding globulin (transcortin) of the serum, and can be precipitated with antibodies to serum proteins. Another binder (A protein, MW 51,000) appears to be present only in the liver cytosol and has a higher affinity for natural glucocorticoids than B protein or purified transcortin (Beato et al., 1972). Kinetic experiments indicate that in uiuo cortisol may be first bound to B protein and then transferred to A protein, but other explanations are possible. The biological importance of A and B proteins is not obvious, since neither of them binds potent synthetic glucocorticoids such as 9wfluorocortiso1, triamcinolone, or dexamethasone. Later, a third binder (G protein), which can be clearly separated by column chromatography on Sephadex gels and distinguished from A and B protein, was found in liver cytosol (Beato and Feigelson, 1972; Koblinsky et aZ., 1972). G protein binds cortisol as well as synthetic glucocorticoids including dexamethasone. Unlike A and B proteins, which in sucrose gradients sediment as 4 s units independently of ionic strength, G protein sediments in low-salt sucrose gradients mostly as a heavy complex, 7s (MW 200,000), but as the 4 s form (MW 66,000) in the presence of 0.3 M KCl. The K, of G protein for dexamethasone at 0°C (2.7-7.3 X lo* M-') is somewhat lower than the binding of cortisol by A or B protein, but at physiological temperature the K, of G protein ( lo8A 4 - I ) is one order of magnitude higher than the K, for A or B proteins. Because of its proper steroid specificity, affinity, and resemblance in sedimentation properties to other steroid receptors, G protein is considered the glucocorticoid receptor in liver. Using chromatographic separation on DEAE-Sephadex gels, Litwack and co-workers (1972) also isolated four macromolecules in rat liver cytosol, which bind corticosteroids and their anionic metabolites. Binder I (MW 20,000), called ligandin (Litwack et al., 1971; Singer and Litwack, 1971a), binds different groups of steroids, their anionic metabolites, and bilirubin. It also binds azo dye carcinogens and 3-methylcholanthrene covalently as well as noncovalently. Binder I11 (MW 7000) has a high 260 nm/280 nm absorbance ratio, appears to contain nucleotides, and binds preferentially in uiuo a corticosteroid anion which is possibly a monosulfate derivative of a reduced metabolite (Morey and Litwack, 1969). Binder IV (50,000) has been identified as transcortin (CBG). From the ligand

CELLULAR RECEPTORS FOR STEROID HORMONES

115

specificity, the three binders do not appear to be the specific glucocorticoid receptor protein. Binder I1 (MW 67,000) has high-affinity binding and appropriate glucocorticoid-binding specificity: dexamethasone > corticosterone > cortisol S- cortisone + deoxycorticosterone. But it also binds progesterone and other sex hormones to some extent. The Kd for dexamethasone is about 0.6 nM, and for cortisol, in the range of 10 nM. An antibody has been prepared to binder 11. The antibody does not cross-react with other corticosteroid-binding proteins (Litwack et al., 1973). Binder I1 apparently can transfer to nuclear sites. From the liver nuclei exposed to corticosteroid in uiuo or in uitm, a macromolecular complex having characteristics (including PI 6.7) similar to cytosol binder I1 can be extracted. G proteins, which is obviously identical with binder 11, can also carry glucocorticoid-3H into nuclei where it is retained. The nuclear complex in 0.4 M KCl migrates as a 4 s form (Beato et al., 1973; Kalimi et al., 1973). b. Thymocytes. The most clear-cut evidence that cortisol can act without chemical alteration is seen in experiments with isolated thymus cells that do not metabolize cortisol (Munck and BrinckJohnsen, 1968). Specific binding of cortisol by thymus cells was originally characterized by the fact that cortisol bound to the cells dissociates relatively slowly compared to the more rapid steroid dissociation from nonspecific binding sites (Munck and Brinck-Johnsen, 1968). Glucocorticoid-binding proteins can be isolated from both the nucleus and the cytosol of disrupted cells. They have a higher affinity for dexamethasone than for cortisol and therefore are different from corticosteroid-binding globulins (Wira and Munck, 1970). Bell and Munck (1973) thoroughly studied the steroid-binding properties, especially those related to ligand-binding kinetics and stability of the cytosol glucocorticoid receptors of rat thymus cells. The K, for complexes with cortisol, dexamethasone, and triamicinolone acetonide at 3°C were estimated to be 3.3 X lo7W *1.3 ,x lo8M-I, and 2.7 x loyM - l , respectively. Kaiser et al. (1973) showed that the triam~inolone-~H acetonideprotein complexes of rat thymocyte cytosol sediment as 3.5 and 7s. The 7 s unit is transformed to the 4 s form by incubation at 37°C or by increasing the salt concentration. In 0.15 M KC1 only the 4 s form can be seen. The nuclear pellet fraction also contains the 4 s form which can be extracted by 0.15-0.4 M KC1. It is of interest that the cytosol of mouse thymocytes also contains the 7 s complex but not the 3.5s

116

SHUTSUNG LIAO

form. Antiglucocorticoid, corte~olone-~H, which competes for the specific binding of corticoids by thymus (Munck e t al., 1972), forms a 3.5scomplex which is unaffected by changes in temperature and salt concentration. c. Hepatoma Cells. Rat hepatoma tissue culture (HTC) cells provide an excellent system for the study of steroid hormone actions, since glucocorticoids induce several proteins, including tyrosine aminotransferase. Baxter and Tomkins (1971a,b) showed that, after incubation of HTC cells with de~amethasone-~H at 3 7 C , radioactivity is specifically bound to macromolecules in both the nuclear and cytoplasmic fractions. Since, in cell-free incubations, most of the binding activity is in the cytoplasmic fraction, nuclear localization of the radioactive steroid appears to occur only after initial binding in the cytoplasm. The receptor proteins isolated by these investigators sediment near 4 s at 0.5 M KC1 and near 8s at low ionic strength. The & has been estimated as 0.74 nM. The occasional finding of multiple components that migrate between 2 and 10s has also been described. The specific HTC cell receptor differs from transcortin in its steroid specificity and binding affinity. The binding reaction is reversible and is characteristic of second-order kinetics of binding and firstorder kinetics of dissociation. d. Lung. Glucocorticoids accelerate lung development and precocious appearance of pulmonary surfactant in fetal rabbit and lamb. Ballard and Ballard (1972), in a study of de~amethasone-~H binding to various tissues of the fetal animal, found proteins that bind glucocorticoid with a specificity that fits well with that shown by biological activity in HTC cells. The finding strongly suggests that glucocorticoid acts directly on lung. The & of the cytosol steroid-protein complex has been estimated as 2.7 nM. Nuclei exposed to the dexametha~one-~H-receptor complex retain the complex firmly. acetonideToft and Chytil (1973)found that the triam~inolone-~H receptor complex in the cytosol of rabbit fetal lung migrates as 7s in low salt. Giannopoulos (1973b) also has shown specific 7s dexamethasone-binding protein in rabbit fetal lung. In 0.4 M KCl, most of the binding is at the 4 s region. Cortisol inhibits formation of the 7s dexamethasone-protein complex, but by itself binds to the 4 s unit and does not form a 7s complex. It is suggested that synthetic and natural corticosteroids may bind to different conformational forms of the same binding protein. This indicates that the 7s form is not required for cortisol action. e. Other Target Cells. There is evidence that glucocorticoids act

CELLULAR RECEPTORS FOR STEROID HORMONES

117

directly in the uterus by inhibiting certain estrogen actions (fluid imbibition, growth). Giannopoulos (1973~)has described an 8s dexamethasone-binding macromolecule (Kd 0.27 nM) in the cytosol fraction of rabbit uteri. Cortisol and corticosterone, but not progesterone, testosterone, or estradiol, compete for binding. The KC1-extractable nuclear complex sediments at 4s. Specific dexamethasone binding, however, was not found by the same investigator in the rat uterus. Cellular proteins that specifically bind corticosteroids have been found in the placenta (Wong and Burton, 1973), brain (McEwen et al., 1972; Chytil and Toft, 1972), muscle, small intestine, skin (Giannopoulos et al., 1973a), kidney, heart (Funder et al., 1973c), fibroblasts (Hackney and Pratt, 1971), lymphoid cells, leukemic cells, and lymphosarcoma (Schaumburg, 1970; Wira and Munck, 1970; Kirkpatrick et al., 1972; Gailani et al., 1973; Werthamer et al., 1973; Simonsson, 1972; Lippman et al., 1973). If one assumes that these binders are indeed functional receptors, their existence may indicate that glucocorticoids act directly at many sites not normally considered target tissues. 2. Biodynamic Aspects Several lines of evidence support the suggestion that dexamethasone-binding proteins are indeed the cellular receptors involved in glucocorticoid action. One of the best studies of this subject was carried out with HTC cells. In this system the correlation between biological activity and receptor binding of steroids was examined in detail (Baxter and Tomkins, 1971a; Rousseau et al., 1972a). First, the kinetics of dexamethasone-binding and dissociation at 37°C is found to be rapid enough to account for the rapid kinetics of induction and reinduction of tyrosine aminotransferase. Second, the extent of protein induction and of steroid binding as a function of dexamethasone concentrations are very similar. This is apparently also true for other corticosteroids (cortisol, aldosterone) if their metabolic rates are taken into consideration. Third, the binding characteristics of several steroids (including inducers, antiinducers, and inactive steroids) parallel their biological activity. It is of interest that these steroid specificities for biological activities in turn are very similar to those reported for steroid binding by de~amethasone-~Hbinding proteins in liver and HTC cells (Beato et al., 1972; Rousseau et al., 1972a; Van Der Meulen and Sekeris, 1973), thymus cells (Munck et al., 1972), lungs (Ballard and Ballard, 1972; Giannopoulos, 1973b), and most of the other tissues described above. Strong support for the involvement of specific binding proteins in

118

SHUTSUNG LIAO

glucocorticoid action in target cells comes from the work of Rosenau et al. (1972), who studied cultured mouse lymphoid cells killed by exposure to glucocorticoids. These cells also contain cytosol dexamethasone-binding protein (430 fmoleslmg protein; 3000 sites per cell) which can be retained by isolated nuclei. In variant cell lines that resist the steroid killing effect, dexamethasone-binding capacity is found to be only 10%of that of sensitive cells. The apparent Kd for cytosol receptors of sensitive cells (20 nM) also differs from that of resistant cells (4.8 nM). The nuclei from both cell types are equally effective in retaining cytosol receptors in a cell-free system, but in intact cells only the nuclei of the sensitive line retain the dexamethasone-receptor complex, It is therefore concluded that the defect in the cytoplasmic receptor is responsible for the insensitivity of resistant cell lines to glucocorticoids. The importance of glucocorticoid-receptor complexes in the earliest steps in the action of glucocorticoids on thymus cells has been elegantly elaborated on by Munck et al. (1972). Cortisol and other glucocorticoids added at physiological concentrations to rat thymus cells in uitro at 37°C begin to inhibit glucose transport after about 15 minutes. Using antiglucocorticoid (cortexolone), actinomycin D, and cycloheximide to block receptor binding and RNA and protein synthesis, and to analyze their effects on glucose transport or protein synthesis, they visualized the time course of the steroid effect on thymus cells at 37°C as shown in Fig. 3. Within the first minute, cortisol is bound to the cytoplasmic receptor. The complex is then transferred to the cell nuclei, a process that occurs more effectively at 37°C than at 3°C. Within 5 minutes sufficient specific RNA for the hormonal effect is apparently made, and thereafter removal of cortisol (and steroid receptor) from the cell nuclei (andlor inhibition of RNA synthesis) does not abolish the hormonal effect. Next, there is a temperature-sensitive step, which is blocked at 20°C. After this gap the synthesis of specific proteins and inhibition of glucose metabolism start to appear at 15-20 minutes. General inhibition of protein synthesis then becomes clear at 40 minutes and is followed by cell lysis. Van Der Meulen and Sekeris (1973)studied glucocorticoid activity in stimulating RNA synthesis in vivo and dexamethasone-binding activity in the liver of postnatal rats. Both activities appear to be low during the first 10 days after birth, but increase nearly 10 times in the next 10 days and reach a maximal value at 20 days. Singer and Litwack (1971b),however, showed high corticosteroid binding by a receptor (binder 11) at the time of birth, but also that there is binding

119

CELLULAR RECEPTORS FOR STEROID HORMONES SPECIFIC TEMR RNA-SENSITIV&,

c

g ,. -'H

SPEC1 F I C PROTEIN

- ..* .n .S c

-10

5

~

Cell lysis

,

Cortexoione - : / / 7 i

Y

Act. D

-/ / / / / I

C

c

200

y

.....- ..

l

o

h

r

r

i

m

i

d

e

/

~

zzzzzz blocks

0foils to block

cortisol effect

FIG.3. Time course in rat thymus cell suspensions at 37°C of cortisol-receptor complex formation, cortisol-induced inhibition of glucose transport, and inhibition of protein synthesis. Cross-hatched segments of the horizontal bars in the lower part of the figure indicate roughly the periods (on the time scale above) during which emergence of the cortisol effect on glucose metabolism can be blocked by treatment with cortexolone (which displaces cortisol from glucocorticoid receptors), actinomycin D, and cyclohexirnide, and delayed by lowering the temperature. Open bars indicate periods during which these treatments have no effect. At the top of the figure is given the sequence of steps by which it is hypothesized that the cortisol-receptor complex leads to synthesis of a specific protein which inhibits glucose transport. (From Munck et al., 1972.)

activity by another small, nucleotide-containing, steroid-binding protein (binder 111),which increases as the enzyme-inducing ability of the animals increases between 15 and 40 days. Whether binder I11 is needed for the function of binder I1 (considered to be the cellular receptor) or whether binder I11 alone is involved in enzyme induction is not clear. In the lung the cytoplasmic receptor is present at 18 days of gestation. The concentration of receptor sites for de~amethasone-~H in fetal rabbit lung is about 430 fmoleslmg of cytosol protein, which is two to five times greater than that in fetal skin, kidney, heart, liver, thymus, and other tissues. It is estimated that there are 9500 nuclear binding sites and 12,000 cytoplasmic receptor sites per fetal lung cell. Giannopoulos (197313) detected the same receptor molecule in the fetal lung of rat, guinea pig, and human, but not in the adult lungs of these species. He suggests therefore that in some animals glucocorticoid action in the lung may be limited to certain periods of

120

SHUTSUNG LIAO

the development. Toft and Chytil (1973), however, found the receptor molecule in the lung of adult adrenalectomized rats. E. MINEFULOCORTICOIDRECEPTORS Aldosterone is the principal and most potent mineralocorticoid in vertebrates, although many glucocorticoids also show mineralocorticoid activity. Mineralocorticoids are normally characterized by their ability to regulate salt balance in vertebrates by increasing active Na+ transport across many epithelial tissues such as urinary, kidney, intestinal tract, and anuran skin and heart. Studies showing the counteractive effects of a variety of inhibitors of RNA and protein synthesis strongly support the induction hypothesis according to which mineralocorticoid presumably increases the concentration of induced proteins, which may alter the rate-limiting steps in Na+ transport (DeWeer and Crabbd, 1968; Sharp and Komack, 1971; Lahav et al., 1973; Lifschitz et aZ., 1973). Additional support for this view comes from the finding that aldosterone increases ribosomal capacity for protein synthesis (Trachewsky et al., 1972) and RNA polymerase activity (Liew et al., 1972). One direct action of aldosterone in the isolated toad bladder is indicated by its recovery as the unmetabolized steroid after the hormonal effect on Na+ transport is completed (Crabbd, 1963). Aldosterone obviously binds noncovalently to cellular binding sites, since the unchanged steroid is extracted readily from target tissues with organic solvents (Edelman et al., 1963; Sharp et al., 1966). Using a displacement binding technique, it is possible to show a saturable aldosterone-binding site in cytosol and nuclei of toad bladder epithelial cells (2700 sites diploid per nucleus) (Sharp and Alberti, 1971) and rat kidney (Edelman, 1971). Specific displacement of aldost e r ~ n e - ~is H seen in these systems, with nonradioactive d-aldosterone or other mineralocorticoids such as 9a-fluorocortisol, deoxycorticosterone, and cortisol in the approximate order predicted from their activity in the tissue. Very weak displacement was found with estradiol, but no interference was seen with inactive steroids like cholesterol and testosterone. The importance of the binding phenomena is underlined by the ability of mineralocorticoid antagonists including spirolactone. Autoradiographic techniques have also been used to substantiate the cellular localization of aldosterone-3H in both cytoplasm (perinuclear areas) and nuclei of toad bladder epithelium (Porter et al., 1964). Data from ald~sterone-~H uptake-retention studies indicate the presence of heterogenous aldosterone-binding sites in the bladder

CELLULAR RECEPTORS FOR STEROID HORMONES

121

and renal systems (Fanestil and Edelman, 1966; Alberti and Sharp, 1969; Swaneck et al., 1970). This is best illustrated by the study of Edelman and his co-workers on rat kidney corticosteroid-binding proteins. Using tritiated aldosterone, dexamethasone, and corticosterone, and analyzing the protein binding of each of the radioactive steroids in the presence of other nonradioactive competitors in vivo and during in vitro tissue slice incubation, they identified three different proteins showing different binding affinity for various steroids. Type I appears to be the mineralocorticoid receptor, since it has a high affinity for ald~sterone-~H (& at 3 7 T , 0.5 nM; 30 fmoleslmg cytosol protein), and desoxycorticosterone (a potent steroid), but not for corticosterone which has very low mineralocorticoid activity in the rat (Funder et al., 1973a). Type I1 has a high affinity for dexamethasone-3H (& at 25T, 5 nM; 160 fmoles/mg cytosol protein). The affinity for various steroids is in the order: dexamethasone > corticosterone > desoxycorticosterone 2 aldosterone 3 cortisol > progesterone >>> estradiol and dihydrotestosterone. At 25"C, but not at O T , de~amethasone-~H receptor is transferred from the cytoplasm to the nuclei (Funder et al., 1973b). Type I11 has a high affinity for corticosterone (& at 2 5 T , 3 nM; 800 holeslmg cytosol protein), and its affinity for other steroids is: corticosterone > cortisol > deoxycorticosterone > progesterone > aldosterone > dexamethasone. The order of steroid affinity exhibited by type I11 is identical to that of corticosteroid-binding globulin (CBG). However, type I11 sediments as 8s and also as 4s forms in low-Ca2+media, but CBG sediments only as a 4s unit. Type I11 also serves as a donor for nuclear cortic~sterone-~H uptake; the KC1-extractable nuclear complex sediments as 3s (Feldman et al., 1973). Although the affinity of the mineralocorticoid receptor (type I) for aldosterone is much higher than for corticosterone in rat kidney, the physiological plasma concentration of corticosterone is much higher than that of aldosterone. The net effect of the differences in affinities and plasma concentrations indicates that cellular mineralocorticoid receptor sites may be predominantly and inappropriately occupied by corticosterone. Funder et al. (1973a), however, on the basis of a series of in vivo infusion experiments, found that the corticosterone effect on mineralocorticoid binding of aldosterone was in fact smaller than was expected. They have suggested that the ability of aldosterone to occupy the receptors under physiological conditions is a function of the extensive preferential plasma binding of corticosterone. Specific aldosterone-protein complexes have been detected in the

122

SHUTSUNG LIAO

cytosol fraction of rat kidney (Herman et al., 1968; Robinson and Fanestil, 1970),duodenal mucosa, spleen, liver and brain (Swaneck et al., 1969),and salivary gland (Funder et al., 1972).The kidney cytosol receptor bound to aldo~terone-~H sediments as two broad peaks at 4.5 and 8.5s. Addition of 0.4 M KCl shifted the 8.5s peak to 4.5S, and addition of 6 mM CaC1, resulted in a peak at 3.5s with a shoulder at 4.5s (Marver et al., 1972). Two forms of the aldosterone receptor have been characterized from the renal nuclei of rats. The first, a 0.1 M tris-extractable complex, has a sedimentation coefficient of 3s; the second, a 0.4 M KC1-extractable complex, sediments at 4 s and is tightly bound to chromatin. The nuclear binding of both receptors appears to require intact DNA. Kinetic studies and mixing experiments with ald~sterone-~H-labeledcytosol and unlabeled nuclei suggest that, in uiuo, cytoplasmic receptor entering the nucleus (probably in the 4.5s form) is first converted to the 3s unit and then bound to chromatin as the 4 s form (Swaneck et al., 1970; Funder et al., 1972; Marver et al., 1972). Funder et al. (1974) recently studied the ability of a series of 24 spirolactone analogs to compete for aldosterone-3H binding to the cytosol receptor during the incubation of rat kidney slices. Among some interesting effects of structural modification are a decrease in the affinity by ring-B unsaturation at the C-6 and C-7 positions, and an increase in esterification or thioesterification at the C-7 position, also in ring B. Forte (1972) studied the effect of mineralocorticoid agonists and antagonists on the binding of aldosterone-gH to rat kidney plasma membranes. Mineralocorticoid receptors can be found in fetal kidney of the guinea pig at 25-40 days of gestation (Pasqualini et al., 1972). More than 50% of the total aldosterone receptors were found in the nucleus. Most of them are tightly bound to the chromatin fraction and can be extracted by 1 M NaCl but not by 0.1 M tris buffer. The nuclear complexes sediment as 2.5-3.5s and 4-5S, whereas a 8-9s form can be detected in the cytosol fraction. Pasqualini et al. (1972) also provide evidence that aldosterone bound to macromolecules may be metabolized to a tetrahydroaldosterone-macromolecule complex without dissociation. It should be noted again that the glucocorticoid receptor also binds aldosterone to some extent. In many cells, such as HTC cells (Rousseau et al., 1972b),aldosterone and dexamethasone appear to bind to a single class of sites with affinities that correspond to their potencies in glucocorticoid activity rather than in mineralocorticoid activity.

CELLULAR RECEPTORS FOR STEROID HORMONES

123

F. STEROIDRECEPTORS IN BRAINS Many reports are now available to show that estrogen in the female or androgen in the male can be preferentially retained by the hypothalamus, an apparent site of hormone action in the regulation of gonadotrophin production or sexual behavior (Eisenfeld and Axelrod, 1966; Kato and Villee, 1967; McEwen et al., 1970; Tuohimaa, 1971; Maurer and Woolley, 1971; Pkrez-Palacios et al., 1973; Plapinger and McEwen, 1973). No such sex difference in the selective retention of sex steroids appears to exist in pre- and neonatal rats before sexual differentiation of the hypothalamus (Tuohimaa and Niemi, 1972a). Androgenization of the female rat also results in loss of the capacity of the hypothalamus and anterior pituitary to accumulate estradiol (Flerkb et al., 1969; Tuohimaa and Johansson, 1971; Vertes et aZ., 1973), possibly explaining the loss of estrogen action in these animals. The feminization of male rats by neonatal antiandrogen (cyproterone) treatment (Neumann et al ., 1967) induces a cyclic output of pituitary gonadotrophins and a female type of sex steroid (estrogen) retention by the hypothalamus (Tuohimaa and Niemi, 1972b). The distribution of androgen-concentrating neurons can be seen in specific areas of the brain by autoradiographic techniques (Sar and Stumpf, 1972). Selective retention of androgen occurs in areas of the preoptic parolfactory region, the hypothalamus, the hippocampus, and the amygdala, as well as in cells of the anterior pituitary. The topographic distribution of androgen in the brain agrees well with areas that have been associated with the regulation of gonadotropin secretion and male sexual behavior (Sar and Stumpf, 1973a,b). In the pituitary the nuclei of a small number of anterior lobe cells (about 15%) concentrate radioactivity 1 hour after te~tosterone-~H injection. The cells of the intermediate and posterior lobes did not retain radioactivity. The labeling of anterior pituitary cells is confined to a 60% area of gonadotrophs (Sar and Stumpf, 1973a,b). However, estradioPH in male and female rats showed nuclear estrogen concentration not only in gonadotrophs but also in other areas of anterior pituitary cells (Stumpf, 1971). The differences in the topographic distribution of radioactivity in the brain and pituitary indicate that not all the action of androgen in the pituitary is due to the conversion of testosterone to estrogen (see Table I1 footnote). These studies are in accord with other findings, showing that estrogen and androgen can act directly on the pituitary gland in addition to a negative feedback effect on the hypothalamus (Debeljuk et al., 1972). The quantity of

124

SHUTSUNC LIAO

receptor in the pituitary has been estimated to be severalfold greater than in the hypothalamus. Hypothalamic receptors, however, are confined to the basal medial region, and the concentration in this region (1.1 fmolelpg DNA) is similar to the concentration in the pituitary (1.3 finolelpg DNA) (Anderson et al., 1973b; cf. Leavitt et al., 1973). Studies on isolated sex steroid receptors in the hypothalamus and pituitary have been very limited, but both the cytoplasmic and nuclear fractions of these tissues appear to have estrogen receptor molecules similar to those in uterus (Kahwanago et al., 1970; Clark et al., 1972; Payne et al., 1973; Kato et al., 1974).Testosterone-binding proteins are also found in the cytoplasm and in nuclear preparations of the anterior pituitary (Jouan et al., 1971). It has been shown that most of the protein-bound androgens can be identified as testosterone (Jouan et al., 1973). The formation of dihydrotestosterone from testosterone, however, occurs in the pituitary (Kniewald et al., 1969) or hypothalamus (Kniewald et al., 1971), and direct action of the dihydrotestosterone-receptor complex is also possible. Kato and Onouchi (1973) recently isolated an 8.6s dihydrotestosteronebinding component from rat hypothalamic cytosol. The number of binding sites is estimated to be 9 fmoleslmg cytosol protein. The dissociation constant of the complex is reported to be approximately 0.7 nM. An autoradiographic study in songbirds has shown that androgenconcentrating cells are in a midbrain area from which vocalizations can be electrically stimulated, suggesting that androgen acts on this particular site to affect avian vocal behavior (Zigmond et al., 1973). The possibility that progesterone inhibits the expression of an androgen-dependent courtship display of the male ring dove by blocking the hypothalamic accumulation of testosterone has been discussed by Stem (1972). Corticosteroids affect ACTH secretion and behavior in animals, apparently by functioning through the hypothalamus and other areas of brain (Bohus, 1970; Pfaff et al., 1971). The uptake and binding of radio-active corticosteroids, as seen by autoradiographic techniques and by biochemical analysis of rat tissue fractions, appear to occur preferentially in the neuron cell bodies of the hippocampus. Much of the labeled steroid appears in the cell nuclei, and the steroid-protein complex can be extracted from them by 0.4 M salt solutions (McEwen et al., 1972).Rat brain cytosol also contains corticosterone3H-binding macromolecules which are different from serum-binding

CELLULAR RECEPTORS FOR STEROID HORMONES

125

proteins. The protein binds natural and synthetic corticosteroids (cortisol, triamicinolone, dexamethasone, and corticosterone) and progesterone, but not estradiol, testosterone, or dihydrotestosterone (Chytil and Toft, 1972; McEwen et al., 1972; Grosser et al., 1973; McEwen and Wallach, 1973).The greatest concentration of the binding protein appears to be in the hippocampus. Watanabe et al. (1974) recently showed that mouse pituitary tumor cells contain glucocorticoid receptors in the cytosol and nuclear fractions. The cytosol 4-5s complex appears to translocate to the cell nuclei by a temperature-dependent process. The uptake of progestins by the brain has been studied, but specific or selective retention has not been clearly demonstrated (Whalen and Luttge, 1971c; Wade and Feder, 1972). G. RECEPTORAND STEROID DEPENDENCY O F CANCER Lactating and tumorous mammary tissues in experimental animals also contain 8 and 4S, high-affinity (K, lo9 M - l ) cytosol receptors, and 4-5s nuclear receptors for estrogens (Jensen et al., 1971a; Harris et al., 1971; McGuire and Julian, 1971; Wittliff et al., 1972; Boylan and Wittliff, 1973; Bresciani et al., 1973). The concentration of cytoplasmic receptor in mammary tissue appears to vary with differentiation of the mammary gland (Wittliff et al., 1972) and increases markedly during lactation. At the tenth day of lactation, there are about 5000 receptor sites per cell. This increase is not accompanied by a corresponding increase in the concentration of the nuclear estrogen-receptor complex. Since in many cases an injection of exogenous estradiol can result in the appearance of the nuclear receptor, the low nuclear receptor content may be simply due to the low blood level of estrogen during lactation. (Shyamala and Nandi, 1972; Hsueh et al., 1973; Gardner and Wittliff, 1973). However, in some DMBA-induced rat or mouse mammary tumors that fail to regress after ovariectomy, estradiol is unable to accumulate in the nuclei (McGuire and Julian, 1971; Shyamala, 1972; McGuire and Chamness, 1973). McGuire et al. (1972)believe that this difference is not due to the inability of nuclear chromatin to retain the cytoplasmic estradiol-receptor complex. Binding analysis revealed (McGuire et al., 1971) a similar affinity for cytoplasmic estradiol binding (Kd 0.29 nM) in both, but the number of binding sites in the hormone-dependent tumor (26 fmoleslmg cytosol protein) was much higher than that in the hormone-independent tumor (2.4 fmoles/mg cytosol protein).

126

SHUTSUNG LIAO

Human mammary carcinoma also contains estradiol-binding protein (8 and 4 s ) of high affinity (K,109-1012M-'),The content of estradiol binding sites in patients varies greatly, with a majority at 30-100 fmoleslmg protein, but some patients have as much as 600 finoles/mg protein (Hahnel and Twaddle, 1973; Spaeren et al., 1973; McGuire, 1973; McGuire and Delagarza, 1973; Hilf et al., 1973; Jensen et aZ., 1973; Brooks et aZ., 1973). The possibility that estrogen- or progesterone-insensitive mammary cancers of humans contain very few or no receptors for sex steroids has been studied in many laboratories, since this may provide an important means of screening patients who may not respond to hormone therapy and therefore can be spared unnecessary major surgery such as adrenalectomy. In an urgent need to clarify the feasibility of the method for therapeutic guidance, a Breast Cancer Cooperative Group has decided on a standard assay based on a charcoal adsorption technique and agar gel electrophoresis as an optional method (Heuson, 1973). The most reliable although more laborious method, however, is to assay for the formation of 8s receptor. Recently we developed a simple and rapid method of estimating a specific steroid receptor in the presence of nonspecific steroid-binding proteins. [The technique involves the use of an antibody to a steroid to remove the 3H-labeled steroid bound to blood and nonreceptor cellular steroid-binding proteins. Since 3H-labeled steroid bound to receptors is not dissociated in the presence of the antibody, it can be separated from the antibody-bound 3H-labeled steroid by gradient centrifugation or simply by the use of insolubilized antibody. The latter method can be used to assay multiple samples within 3-4 hours (Castaiieda and Liao, 1974a).] From the limited data available, encouraging results have been reported to show that there is a good correlation between the cytosol receptor content of tumors and the success or failure of endocrine treatment (Jensen et al., 1971a, 1973; Maass et al., 1972). More information is needed to establish whether quantitative estimation of estrogen receptors in mammary cancers can be used for the definite prediction of individual response to endocrine treatment. Factors such as age of the patient, histological type of tumor (Spaeren et al., 1973),and receptors for progesterone (Terenius, 1973) and other hormones may have to be considered. As described in Section III,B, specific androgen receptors have been demonstrated in human prostate tissue. Whether or not the receptor assay can assist in selecting the appropriate therapy for patients with prostate carcinoma is being studied (Mobbs et d., 1974).

CELLULAR RECEPTORS FOR STEROID HORMONES

127

IV. Cytoplasmic-Nuclear Interaction of Steroid Receptors AND NUCLEARRETENTION OF CYTOPLASMIC A. TRANSFORMATION

RECEPTORS It is now clear that in most steroid target tissues (with the exceptions noted below), nuclear retention of steroids and receptors is dependent on prior formation of steroid-receptor complexes in the cytoplasm. The “two-step” mechanism was put forth independently by Jensen and Gorski. An indication for this came from autoradiographic and cell fractionation studies showing that, when uterine tissue is incubated with estradioL3H at Oo-2”C, more than 70% of the radioactive steroid is seen in the extranuclear region on autoradiography but, when the tissue is warmed to 3 7 T , the nuclear bound steroid becomes predominant (Jensen et al., 1968). During such redistribution in uitro, the level of cytosol receptor identifiable as 8s rapidly diminishes, with a concomitant appearance of the nuclear complex (Jensen et al., 1968, 1969; Gorski et al., 1968; Shyamala and Gorski, 1969; Giannopoulos and Gorski, 1971a). Such a process appears to operate in uiuo, since injection of a physiological dose of estradiol can result in a progressive fall in the level of cytosol receptor for about 4 hours, after which the receptor content is gradually restored (Jensen et al., 1969; Sarff and Gorski, 1971). The requirement for cytosol in formation of the 5s nuclear complex is also evident from the fact that no 5s complex can be obtained if estradioL3H is incubated directly with uterine nuclei or a nuclear extract of estrogen-deprived animals. In the presence of a cytosol fraction, however, the nuclear estrogen receptor complex is readily detected in the extract of reisolated nuclei. The fact that the receptor proteins for estradiol are not found in the cell nuclei of uteri deprived of estrogens implies that nuclear retention of the cytosol receptor protein is dependent on prior interaction of steroids and receptors. Initially, it was assumed that the 5s nuclear complex was formed in the nucleus from the 4s cytosol complex (Jensen et al., 1969), but subsequent studies showed that incubation of uterine cytosol alone in the presence of (but, importantly, not in the absence of) estradiol yields a 5s complex which can be retained by the uterine nuclei. The transformation takes place only slowly in the cold, proceeds readily at 25 to 3 7 C , and is accelerated with increasing pH over the range 6.5-8.5 and by the presence of salt. EDTA, Ca2+,Mgz+,or Mn2+retards the transformation. Besides estradiols, other estrogens such as diethylstilbestrol, estriol, and hex-

128

SHUTSUNG LIAO

estrol also promote the 4s-to-5S transformation. Estrone binds to the 4 s unit but does not promote the process (Jensen et al., 1971b, 1972b) unless the concentration is raised to a higher level than that required for estradiol (cf. Ruh et al., 1973). It is suggested that the estrogendependent transformation of the 4 s complex to the 5s form involves a steroid-dependent change in the conformation of the receptor protein. [Jensen’s group (Gore11 et d . , 1974) recently named the untransformed cytosol (extranuclear) receptor estrophilin-I, and the transformed (the active form that can be retained by the nucleus) estrophilin-11.1 It should be noted that in 4 M urea both the cytosol and the nuclear estradiol receptor complexes have sedimentation coefficients (3.6s) identical with ovalbumin (Stance1 et al., 1973b), suggesting that no gross alteration of the basic chemical nature of the receptor protein occurs during the transformation. A somewhat different view has been presented by an Italian group (Puca et al., 1972; Bresciani et al., 1973). As described in Section III,A, they believe that estradiol binds to a 5.3s cytosol receptor, but that the complex is cleaved by a proteolytic factor to 4.5s forms retained by the uterine cell nucleus. Besides the receptor transformation, a temperature-dependent alteration of the cell nuclei may also facilitate receptor binding. In a study of the ability of the nuclear chromatin fraction of mammary tumors to retain the uterine estradiol-receptor complex, McGuire et al. (1972) showed that preincubation of the estrogen-receptor complex (but not the chromatin) alone at 21°C can result in a significant but small increase in binding, whereas much more binding occurs when the complex and chromatin are incubated together at 21°C. Chatkoff and Julian (1973) have also suggested that the chromatin acceptor” sites may be under dynamic control. They found that progesterone treatment (3 days) of ovariectomized rabbits increases the number of uterine chromatin acceptor sites and the binding affinity for the retention of estradiol receptor complexes. The temperature-dependent increase in retention of the dihydrotest~sterone-~H-receptorcomplex by prostate nuclei was originally observed in tissue incubation experiments using cell fractionation (Fang et al., 1969) and in autoradiographic study (Sar et al., 1970). Indication that specific dihydrotestosterone binding by prostate cell nuclei is dependent on the cytosol fraction is clear, since the salt extracts of the prostate nuclei of castrated rats contain very few proteins that bind dihydrotestosterone tightly. Incubation of isolated nuclei with radioactive androgen also does not result in significant formation of the nuclear androgen receptor complex. The nuclear complex, 66

CELLULAR RECEPTORS FOR STEROID HORMONES

129

however, can be obtained readily if the whole homogenate is used or a cytosol fraction is incubated with the nuclei (Fang et al., 1969; Liao and Fang, 1969). The need for prior interaction of dihydrotestosterone with the cytosol receptor (p protein) in nuclear retention of the receptor protein has been shown by Fang and Liao (1971). In the absence of androgen, p protein is not retained by isolated prostate cell nuclei. Another dihydrotestosterone-binding protein (aprotein) binds androgen but is not retained by nuclei. The levels of nuclear retention of other proteins that bind testosterone, estradiol, progesterone, and cortisol are not significant (Liao and Fang, 1970). In a cell-free system, the binding of dihydrotestosterone-receptor complex by prostate nuclei proceeds better at 20°C than at 4°C (Fang and Liao, 1971; Liao and Liang, 1974). Mainwaring and Peterken (1971) also observed the same temperature effect using prostate nuclear chromatin. A brief warming of the 8s cytosol dihydrotestoster~ne-~H-receptor complex before its incubation with nuclei markedly accelerated the rate but not the overall extent of the transfer of the complex into prostate chromatin. The temperature-dependent 4.6s

100-

.-cx

> .* c

3.6s

5.

J.

80.

'

2

60-

B

.

L

s

40-

n

c

.

.

.

13

Fraction number

. 20

Fraction number

FIG. 4. Temperature-dependent transformation of steroid-receptor complexes. The estradiol-receptor complex in the cytosol(3.8S) of calf uterus can be transformed, by incubation at 25"C, to a 5.3s form which can be retained by uterine cell nuclei (left). Similarly, the cytosol dihydrotestosterone-receptor complex (3.8s) of rat ventral prostate can be transformed, by incubation at 20°C for 20 minutes, to a 2.9s form which is indistinguishable from the one retained by prostate cell nuclei (right). Gradient solutions containing 0.4 M KCI were used in the sedimentation studies. Arrows show where bovine serum albumin (4.6s) and ovalbumin (3.6s) sediment. The left figure is taken from DeSombre et al. (1972).The right figure is based on our experiment.

130

SHUTSUNG LIAO

alteration of the androgen-receptor complex is accompanied by a pronounced decrease in sedimentation coefficient from 8 to 4.2s and, more interestingly, the isoelectric point declines from 5.8 to 6.5 (Mainwaring and Irving, 1973). Our study indicates that the 8s-to-4S transformation is not dihydrotestosterone-dependent and thus may differ in essence from the estradiol-dependent transformation of the 4 s unit to a 5 s form which enters uterine nuclei. The temperaturedependent transformation of the cytosol receptor to a nuclear form is probably best shown by one of our recent observations. As shown earlier, the prostate cytosol receptor (3.5s) sediments slightly faster than the nuclear receptor (3s)in 0.4 M KCl (Fang and Liao, 1969) or in 2 M urea (Liao, unpublished observation). Under conditions favoring the nuclear retention, the cytosol 3.5s form can be transformed to the 3s form by incubating at 20°C for 10-20 minutes (Fig. 4). As in the case of the estradiol-uterus and dihydrotestosteroneprostate systems, little or no salt-extractable progesterone-binding activity can be detected in oviduct nuclei prior to progesterone administration. Following injection of pr~gesterone-~H in vivo or incubation of oviduct slices in vitro with pr~gesterone-~H, a progressive increase in nuclear binding is observed. This is accompanied by a concomitant depletion of cytoplasmic receptor, also supporting the hypothesis that the nuclear steroid receptor arises by a hormonedependent transfer of the cytoplasmic receptor complex to the nucleus (O’Malley and Toft, 1971; O’Malley et al., 1971, 1972). Specific association of glucocorticoid with cell nuclei in the presence of cytosol has been studied in many systems. With thymus cells, Munck et al. (1972) showed that cytosol receptor complex prepared at 3°C is transferred to isolated nuclei at 3°C only if it has been prewarmed to 25°C. Bell and Munck (1973) present two possible mechanisms involved in the temperature-dependent transformation of the complex. One is an equilibrium mechanism in which the effect of the steroid is to change the position of equilibrium of the two forms of the receptor so as to favor the form with high affinity for the nucleus. The other mechanism is a kinetic one in which the steroid accelerates the rate of the temperature-sensitive transformation. In the equilibrium mechanism the association constant for the interaction of steroid with the transformed receptor must be greater, whereas in the kinetic mechanism the association constant is not expected to change. In rat liver the activation process also appears to occur only if the receptor is bound and is accelerated by an increase in temperature

CELLULAR RECEPTORS FOR STEROID HORMONES

131

and ionic strength (Kalimi et al., 1973; Beato et al., 1974). It is interesting that Milgrom et al. (1973~)found that the affinity of activated dexamethasone-receptor complex is raised not only for liver nuclei, but also for various polyanions (homologous and heterologous DNA, RNA, and even carboxymethyl and sulfopropyl Sephadex). Caz+ markedly inhibited the “acidiophilic activation” of the receptors. Rousseau et al. (1973)showed that binding of dexamethasone-3H by the HTC cells results in a loss of most of the cytoplasmic receptor, and retention of an equivalent number of steroid molecules in the nucleus. When the steroid is removed from the culture, it dissociates from the nucleus, while the level of cytoplasmic receptor returns to normal, even if protein or RNA synthesis is inhibited. Evidence showing retention of cytosol glucocorticoid-receptor complexes by nuclei has also been presented for rabbit fetal lung (Ballard and Ballard, 1972; Giannopoulos et al., 1973b), rat thymocytes (Kaiser e t al., 1973), and rat brain (McEwen and Wallach, 1973). The nuclear retention of mineralocorticosteroid in kidney also proceeds in a similar manner; the nuclear uptake of aldo~terone-~Hreceptor in rat kidney can amount to 60% of the receptor content lost from the cytosol fraction. Data from the time course of generation are consistent with the mechanism: cytosol (8.5 or 4.5s) + tris-soluble nuclear (3s)+ chromatin-bound (4s). However, the 3 and 4 s nuclear complexes may bind to independent nuclear sites. The formation of the chromatin-bound species is seen at 37°C but not at 0°C (Marver et al., 1972; Funder et al., 1972). Numerous examples are available to suggest that the two-step mechanism originally proposed for the nuclear retention of estrogen appears to be a general phenomenon. The mechanism involves a key step: steroid-dependent activation of steroid-receptor complexes to forms that can bind to nuclear sites of target cells. Whether the activation involves enzymic modification or conformational change in the complex is yet to be proven but, as Jensen and DeSombre (1973) believe, the step may represent the transformation of specific steroid-receptor proteins to biochemically functional forms. NUCLEAR RECEPTORS B. CYTOPLASM-INDEPENDENT In some studies formation of the specific steroid-receptor complex is possible with nuclei incubated with radioactive steroids without the addition of cytosol proteins. For example, cortisol-protein complexes can be extracted from purified rat thymus nuclei incubated with ~ortisol-~H (Munck and Wira, 1971; Abraham and Sekeris, 1973). Pasqualini et al. (1972) found that nuclear ald~sterone-~H-

132

SHUTSUNG LIAO

receptor complexes can be obtained after direct incubation of radioactive steroid with purified kidney nuclei. This differs from the suggestion that in the kidney of adult adrenalectomized rats the nuclear aldosterone-receptor complexes are derived from the cytosol complexes (see Section IV,A). In our laboratory, Shao and Castaiieda found tissue- and steroidand 3p, specific formation of 3/3,17p-dihydro~y-A~-androstene-~H 17~-dihydro~yandrostane-~H-protein complexes in vaginal nuclei in vioo. These complexes migrate as 4s complexes in 0.4 M KCl, but similar complexes in the cytosol fraction have not been detected. Vaginal cytosol is not required for formation of the nuclear complex during the incubation of purified nuclei and radioactive steroids. Cell nuclei from liver, prostate, brain, thymus, and kidney are not able to form the complex. The complex is distinguishable from the estradiol receptor complex in the vagina (Shao et al., 1975). Mester and Baulieu (1972) also suggested the presence of an estrogen-binding protein in the KC1-insoluble nuclear aggregate of chick liver. Ozon and Belle (1973) detected a 4s complex in liver nuclear extracts of chicken and toad. Although estrogen stimulates phosphoprotein synthesis in chick liver, no cytosol receptor could be detected in this organ by either group of workers. Arias and Warren (1971), however, have described an estrophilic macromolecule in chick liver cytosol. More studies must be made before the above examples can be regarded as exceptional cases in which hormone action is not dependent on the cytoplasmic receptor. For example, the apparent lack of steroid-protein complex in the cytosol fractions may be due to instability of the receptor protein and to insensitivity of the detection methods. Moreover, in hormone-deprived animals there may be some endogenous steroids that can facilitate the binding of the cytosol receptor to the cell nuclei. Nevertheless, it is possible that some receptors are bound to cellular components in the cytoplasm as well as in the nuclei and simply await activation by the arrival of steroid molecules. During such a process the interaction of a steroid hormone with a receptor protein may weaken the binding of the receptor with the cellular sites and thus facilitate its translocation.

c.

CHROMATIN ACCEPTOR SITES FOR RECEPTORS The number of receptor binding sites in target cell nuclei has been estimated to be in the order of 2000 to 10,000 per diploid genome, assuming one site for one receptor containing one steroid molecule. Whether all these bindings represent biologically important interac-

CELLULAR RECEPTORS FOR STEROID HORMONES

133

tions is open to question, It is also not clear whether more than one type of specific binding is involved in different functions in U ~ U O . Studies in this area have been hindered by difficulty in distinguishing between specific and nonspecific binding. One of the earliest observations showing the possibility that nuclear retention of the cytosol-receptor complex is nuclear-specific came from the study of nuclear retention of the dihydrotestosteronereceptor complex in cell-free systems (Liao and Fang, 1970; Fang and Liao, 1971; Liao et al., 1971b). To eliminate nonspecific association the amount of dihydrotestosterone-receptor complex retained in the specific manner was estimated by measuring the portion of the retained complex that could be extracted from the nuclei by 0.4 M KCl and sedimented as 3s (as is the complex retained in uiuo). By this technique it was shown that nuclei from prostate, but not liver, thymus, or diaphragm, can retain significant amounts of the dihydrotestosterone-receptor complex, and that nuclear binding sites can be saturated by the androgen-receptor complex. Specific retention was not observed with other steroid-protein complexes. Heating of nuclei at temperatures higher than 40T,or treatment of nuclei by proteolytic enzymes resulted in loss of specific binding, but also in an increase in nonspecific association of the steroid-receptor complex with the aggregated chromatin. It was suggested therefore that prostate nuclei contain limiting numbers of tissue and steroid receptor specific binding sites and that certain heat-labile protein factors (defined as acceptor proteins) are necessary for formation of the ternary complex. Subsequent studies showed that tissue and receptor specificity could be demonstrated by using chromatin fractions prepared from prostate nuclei (Mainwaring and Peterken, 1971; Steggles et a1., 1971), or with deoxyribonucleoprotein complexes reconstructed with purified DNA and salt-extractable proteins of prostate nuclei (Tymoczko and Liao, 1971). In a reconstructed system employing Millipore membrane filtration, the acceptor proteins appear to be heat-labile nonhistone proteins and require native calf thymus DNA for its receptor binding activity. Heat-denatured DNA does not appear to be functional, but binds the steroid-receptor complex in a nonspecific manner. Poly A and poly G, but not poly U or poly C, can substitute for native DNA in showing acceptor activity. A similar study using a reconstructed system for the retention of estradiol-receptor and progesteronereceptor complexes by uterine nuclei (Liang and Liao, 1972) has indicated the presence of similar heat-labile acceptor proteins in the 0.4 M KCl extract of female target tissue. In support of this view,

134

SHUTSUNG LIAO

Thompson and King (1974) recently showed that uterine chromatin exposed to 0.4 M KCI binds less estradiol-receptor complex than that not exposed to 0.4 A4 KCl. In contrast, preextraction of rat kidney nuclei with KC1 did not destroy their ability to accept the aldosterone-,H-receptor complex, indicating that the acceptor factor is not solubilized in 0.4 M KCl (Marver et al., 1972). The presence of acceptor proteins was also proposed independently by O’Malley and co-workers in chick oviduct systems. They first demonstrated that the progesterone-receptor complex can be retained by cell nuclei of the oviduct but not of the spleen, lung, intestine, or liver (O’Malley et al., 1971b). Similar results were obtained when nuclear chromatin fractions were compared (Steggles et al., 1971; Spelsberg et al., 1971). Using chromatin from which histones were removed selectively by 2 M NaCl and 5 M urea and reconstructed by the addition of various nuclear protein fractions, they showed that an acidic protein fraction (AP,) was necessary for chromatin to bind the progesterone-receptor complex (O’Malley et al., 1972). It is interesting that, of the two progesterone-protein complexes (Schrader and O’Malley, 1972), only receptor component B binds to chromatin containing AP3. Component A, however, binds to DNA but not to chromatin (Schrader et al., 1972).These observations are analogous to the finding that in prostate only dihydrotestosterone coupling with @protein (but not with a-protein) can be retained by prostate cell nuclei (Fang and Liao, 1971). AP, appears to contain an acceptor protein very similar to that of rat ventral prostate. Whether the oviduct protein is heat-labile is not known, however. A similar view on the nature of the nuclear binding of receptor has been expressed by King et al. (1971).Since all types of DNA bind estradiol receptor without showing specificity, but nuclei or chromatin of uterine origin bind more receptor than those of spleen or liver, they believe that tissue specificity is the result of different histonenonhistone protein compositions which determine the accessibility of DNA (the primary acceptor) to the receptor in different cells. The involvement of DNA is shown by the fact that DNase can release bound estradiol from uterine nuclei (Shyamala, 1971; King and Gordon, 1972). Since the removal of histone or other chromatin proteins enhances DNA binding of receptor, King and Gordon (1972) have warned that any attempt to purify acceptor protein factors must prove that the sites involved in the reconstructed system are identical with those in the intact chromatin. Furthermore, in the aforementioned studies involving rat ventral prostate and chick oviduct, there are many reports showing that the nuclear acceptor sites that bind a steroid-receptor complex in a

CELLULAR RECEPTORS FOR STEROID HORMONES

135

defined manner can be saturated if the steroid-receptor complex is present in excess. From such studies the apparent Kd values for ternary complexes of the nuclear acceptor and estradiol receptor in rat uterus (King and Gordon, 1972; Higgins et al., 1973b), dexamethasone receptor in HTC cells (Higgins et al., 1973a) and lung (Ballard and Ballard, 1972), and dihydrotestosterone receptor in rat ventral prostate (our data) have been estimated to be about 0.2-1 nM. A careful study by Higgins et al. (197313) has shown that different steroid-receptor complexes may bind to different types of nuclear acceptor sites. In this study hepatoma nuclei were found to have a fixed number of glucocorticoid receptor binding sites (3850 per haploid genome), but did not have specific sites for the estrogenreceptor complex of rat uterus. In the rat uterus it has been estimated that there are about 2150 nuclear sites for the hepatoma glucocorticoid-receptor complex and 3350 sites for the uterine estradiol receptor per haploid genome. The binding of one class of the radioactive steroid-receptor complex is shown to be inhibited only by the same class of nonradioactive complex and not by the other type, suggesting that the nuclear sites for the two types of complexes have different specificities. As additional evidence for the difference, Baxter et al. (1972) showed that DNase can inhibit nuclear binding activity for the glucocorticoid receptor but not for the estrogen receptor. They also estimated that purified DNA contains 6.7 pmoleslmg DNA binding sites for binding of glucocorticoid-receptor complex; in isolated nuclei there are only 0.6-2.2 pmoleslmg DNA available for binding of the same complex. This observation is not in agreement with other reports that nuclear bound estradiol-receptor complex in the uterus can be released by DNase and that the glucocorticoid receptor of rat liver does not bind purified DNA absorbed to cellulose if dexamethasone is also present (Beato et al., 1973). In correlating steroid induction of the synthesis of ovalbumin, conalbumin, ovomucoid, and lysozyme (Palmiter, 1972; Palmiter and Haines, 1973) with steroid receptor binding, Palmiter et al. (1973) studied the retention of dihydrotestosterone, progesterone, and estradiol-receptor complexes by the chromatin of magnum explants during tissue incubation. Since these steroids have synergistic effects on the production of egg-white proteins but do not compete for chromatin binding, it is suggested that they have distinct acceptor sites. The number of acceptor sites in the magnum from withdrawn chicks has been estimated as 1070, 2320, and 6400 sites per cell, respectively, for receptors of dihydrotestosterone, progrsterone, and estradiol. Mainwaring and Mangan (1971) were the first to use DNA-

136

SHUTSUNG LIAO

cellulose column chromatography to study the specificity of DNA binding of the dihydrotestosterone-receptor complex of rat ventral prostate. DNA prepared from rat prostate was shown to be more effective than DNA of rat liver, kidney, or Escherichia coli in retaining the androgen-receptor complex (see also Mainwaring and Peterken, 1971). Using the same technique, Clemens and Kleinsmith (1972) and Toft (1973) reported that the estradiol-receptor complex of rat uterus was able to bind more efficiently to rat DNA than to salmon DNA or E . coli DNA. Component A of the chick oviduct progesterone receptor was also shown to bind to chick oviduct DNA (Kd 0.3 nM) more firmly than to BacilZus subtilis DNA (O’Malley et al., 1973). In most of these studies, whether the observed differences are due to artifactual alteration of DNA during its preparation has not been carefully examined. It would be premature therefore to conclude that in steroid-sensitive cells there are tissue- or speciesspecific DNA sequences that bind specific steroid-receptor complexes. The binding of steroid hormones to DNA has been examined in the past decade. Huggins and Yang (1962) and Dannenberg (1963) pointed out the similarity in the molecular geometry of steroids and the base pairs of the DNA helix. T’so and Lu (1964), however, found that single-stranded DNA could bind steroids much more firmly than duplex DNA. More recently, Kidson et al. (1971)showed that the relative affinity of various steroids toward denatured bovine spleen DNA is in the order: progesterone = dihydrotestosterone > testosterone > estradiol > corticosterone = cis-testosterone (17a-hydroxy) = aldosterone. Of many polynucleotides tested, high affinity was observed with dG- or G-containing polymers (but not dG:dC), suggesting the involvement of 2-amino groups of guanine (see also Conen et al., 1969). It is interesting to note that Goldberg and Atchley (1966) indicated that certain steroid hormones could weaken the DNA intrastrand bonds holding the double helix. This action, however, has not been reproducible in our laboratory. Gottfried (1972) has proposed a model in which steroid-carbohydrate (carbosteroid) polymers can interact with DNA strands rich in certain bases. However, steroid receptors in uiuo bind unconjugated steroids, and there is no evidence for the existence of carbosteroidlike substances in the cell nuclei of target tissues. An interesting observation in this regard was made by Sluyser (1966a,b,c), who showed that, in rats injected with radioactive testosterone, radioactive steroid binds preferentially to the lysine-rich histone fraction of ventral prostate rather than to the lysine-poor histone

CELLULAR RECEPTORS FOR STEROID HORMONES

137

fractions. He claimed that testosterone could diminish the ability of prostatic lysine-rich histone to keep the two DNA chains from separating at raised temperatures. Cortisol, however, was shown to associate mostly with the lysine-poor or arginine-rich histones of rat liver (Sluyser, 1966c, 1969). The suggestion that there is a limited number of specific acceptor sites on the nucleus of a target cell has not been generally accepted. For example, Bresciani et al. (1973) estimated that as many as 100,000 estradiol-receptor complexes of either the native or Ca2+stabilized form can bind to certain basic proteins in the uterine cell nucleus. From a careful examination of the distribution of the estradiol-receptor complex in the uterine cytoplasm and the nuclei, Williams and Gorski (1972a,b) concluded that receptor distribution behaves like a unimolecular system, not in accord with the concept that nuclear interaction involves a limited number of acceptor sites. Chamness et al. (1974) also showed that uterine estradioL3Hreceptor complex binds to cell nuclei of the uterus as well as to nontarget tissues, and that the extent of binding is proportional to the concentration of the receptor, no saturation occurring when total radioactivity bound to nuclei is measured. This finding is consistent with the observations of Clark and Gorski (1969), Fang and Liao (1971), and Milgrom et al. (1973~) that nonspecific nuclear association of the steroid-receptor complex occurs readily and that the specific binding sites cannot be estimated unless an additional selective process is employed (cf. Clark et al., 1973).

D. RIBONUCLEOPROTEIN BINDINGOF RECEPTORS In mammalian cell nuclei, 20-30% of the total nucleic acid content is RNA. Nuclear RNAs (see a review by Weinberg, 1973) include precursors of cytoplasmic RNA (rRNA, tRNA, and mRNA), as well as those that remain in the nuclei. A major portion of the nuclear RNA appears to form complexes with nuclear proteins and exists as ribonucleoprotein (RNP) particles. Besides RNP particles that eventually become ribosomes (Burdon, 1971; Kumar and Warner, 1972), some of them may participate in the regulation of gene transcription (Paul, 1971; Britten and Davidson, 1969)or form informosomelike (or informoferlike) particles and become involved in transport of mRNA (Spirin, 1969; Samarina et al., 1968). To explore the possibility that some steroid hormones may be involved in the functions of various RNP particles, Liao, Liang, and Tymoczko (1973a) initiated a study on the nature and properties of the binding in estradiol-receptor and dihydrotestosterone-receptor

138

SHUTSUNG LIAO

complexes to nuclear and cytoplasmic RNP particles of the uterus and prostate. By using gradient centrifugation these steroidreceptor complexes were found to associate readily with the nuclear particles from the respective target tissues but to a lesser extent with those from liver or other less responsive tissues. The ternary complexes of steroid-receptor-RNP sediment at 60-80S, although in the uterine system distinct 50 and 80s peaks can be observed. The binding phenomena can be abolished by prior heating of RNP particles or by treatment with proteases or RNases (but not with DNase I), suggesting the involvement of RNA and heat-sensitive protein components of the particles. The steroid-receptor complex can be dissociated reversibly from the ternary complex by 0.4 M KCl. The receptor-binding sites on the particles can be saturated, and under such conditions less than 10-20% of the isolated nuclear RNP particles can bind to the steroid-receptor complex. It is possible that only those RNP particles with heat-labile acceptor factors can associate with the steroid-receptor complex. About 30-50% of the total cytosol steroid-receptor complexes present can bind to RNP particles, indicating that other steroid-protein complexes may be structurally not compatible with the RNP acceptor sites. The cytoplasmic fractions of uterus and prostate also contain RNP particles (40-60s) which bind estradiol-receptor and dihydrotestosterone-receptor complexes in a cell-free system. However, cytoplasmic polysomes or 80s monosome forms of ribosomes do not bind either of the steroid-receptor complexes. Since, with rats injected with 3H-labeled sex steroids, a radioactive steroid-protein complex can be found to associate with RNP particles, RNP binding of the steroid receptor may indeed occur in uioo. 3H-Labeled sex steroids added in a free form or as steroid-receptor complexes but inactivated by heating (SOT, 10 minutes) do not associate with either the nuclear or the cytoplasmic RNP particles of uterus or prostate. In addition, in the absence of steroid hormones, RNP particles do not appear to have the ability to retain receptor proteins which can later be released and bind radioactive steroids. This is analogous to the current belief that receptor proteins are not retained by cell nuclei without prior binding of the specific steroids (Tymoczko and Liao, 1974; Liang and Liao, 1974a). Binding of the dihydrotestosterone-receptor complex to various polyribonucleotides has been studied recently (Tymoczko and Liao, 1974). With polymers having a sedimentation coefficient of about 4 f l S , poly A, poly U, and to some extent poly G, gave distinct radioactive peaks at 50-70s. Such a significant shift in the sedimen-

CELLULAR RECEPTORS FOR STEROID HORMONES

139

tation coefficient was not observed with poly C, or tFtNA (liver or E. coli.).

E. INTRACELLULAR RECYCLING OF RECEPTORS In a study of glucocorticoid receptor of thymus cells in culture, Munck and Brinck-Johnson (1968) found that there is a correlation between ATP concentration and magnitude of specific cortisol binding. Thus cytoplasmic fractions from anaerobically incubated cells have reduced steroid-binding capacity. Since ATP added to the cytosol cannot increase the binding, and also the possibility of receptor degradation and resynthesis can be excluded, the energy-sensitive step is considered to be at the generation of active receptor. It was suggested (Munck et al., 1972) that ATP directly or indirectly is necessary for the conversion of the receptor from a form in which it is unable to bind cortisol into an active form. Bell and Munck (1973) suggested that ATP may supply the free energy for maintenance of a cycle in which the receptor forms a complex with the hormone, becomes inactivated after reaching the nucleus, and emerges to be reactivated and form a new complex. Ishii et al. (1972), from studies of the energy dependence of the binding of triamcinolone acetonides to cultured mouse fibroblasts (L cells), also have proposed that release of the receptor from the particulate fraction (nucleus) and its regeneration ( t l l z30 minutes) to a form capable of binding steroid again are energy-dependent. The possibility that nuclear RNP particles may be involved in steroid receptor recycling in target cells, and that such a recycling may be functionally related to gene expression, is being considered (Liao and Fang, 1969; Liao et al., 1973a,b), and is discussed in Section V,B. The fact that various steroid hormones can promote rapid transfer of cytoplasmic receptors to the cell nuclei (within minutes), and that the receptors appear to have a long half-life of about 3-5 days, also suggests that receptors are not rapidly consumed or destroyed during their action and very likely are shuttled between the nucleus and cytoplasm in target cells. V. Gene Expression and Steroid Receptor A. RNA SYNTHESISAND PROTEININDUCTION

Numerous studies have shown the importance of RNA and protein production in the early stages of steroid hormone action. They include the demonstration of (1)specific enzyme or protein induction

140

SHUTSUNG LIAO

by steroid hormones; (2) prevention of the effects of steroid hormones by inhibitors of RNA or protein synthesis; (3) rapid stimulation (or inhibition in thymocytes) of RNA synthesis in uiuo or nuclear RNA polymerase activities; and (4) ability of RNA extracted from steroid hormone-treated cells to elicite certain hormone responses in hormone-deficient target cells (see references cited in Section I). It is not clear, however, whether or not steroid hormones actually act directly on the main machinery of genetic transcription or translation. RNA synthesis in mammalian nuclei has often been studied as two separate entities: nucleolar RNA synthesis, which generally represents rRNA synthesis; and nucleoplasmic RNA synthesis, which presumably includes mRNA production. Polymerase complexes I and I1 are assumed, respectively, to be responsible for the types of RNA synthesized in nucleolar and nucleoplasmic areas of the nuclei. Since actinomycin D at low concentrations selectively inhibits nucleolar RNA synthesis by polymerase I, whereas a-amanitin inhibits polymerase I1 action of nucleoplasm, these inhibitors have been used effectively in the elucidation of hormonal action on nuclear RNA synthesis. The involvement of rRNA synthesis during the early effect of steroid hormones has been demonstrated for androgens in prostate (Liao et al., 1965, 1966; Liao and Fang, 1969), estrogens in uterus (see below) glucocorticoids in liver (Yu and Feigelson, 1969), and aldosterone in kidney (Chu and Edelman, 1972). There are strong indications that an increase in rRNA (or nucleolar RNA) synthesis in liver (Yu and Feigelson, 1971, 1972) and in prostate (Mainwaring et al., 1971) systems is due to an increase in active enzyme content rather than a change in overall template activity. Whether there is direct action of the steroid-receptor complex on nucleolar RNA synthesis is not known but, in several cases steroid-dependent induction of enzymes appears to be independent of rRNA synthesis (Gelehrter and Tomkins, 1967; Wicks, 1968; Jost et al., 1973). The manner in which a steroid hormone may affect rRNA synthesis by regulating nucleoplasmic polymerase activity is best illustrated in the studies by many investigators who traced the early steps of estradiol action in uterus. Within 1-4 hours after estradiol injection in estrogen-deprived rats, nuclear RNA polymerase activity is stimulated (Mueller et al., 1958; Gorski, 1964; Hamilton et al., 1965), predominantly at nucleolar sites (Barton and Liao, 1967; Tata, 1968; Hamilton, 1968; Billing et al., 1969). Since inhibitors (puromycin or cycloheximide) of protein synthesis suppress the increase in RNA

CELLULAR RECEPTORS FOR STEROID HORMONES

141

polymerase activity, continuous protein synthesis appears to be needed for the estrogen effect (Mueller et al., 1961; Gorski et al., 1965; Nicolette and Mueller, 1966). In fact, a group of proteins, including one acidic protein which can be identified electrophoretically, is formed in response to estrogen injection (Notides and Gorski, 1966) or in uitro estrogen treatment (DeAngelo and Gorski, 1970; Katzenellenbogen and Gorski, 1972) of uteri. The acidic protein, often called induced protein (IP) can be detected 30 minutes after estradiol enters uterine cells. Whether IP is responsible for the more general stimulation of RNA polymerase activity detectable later has not been proved. The synthesis of IP, however, is inhibited by actinomycin D and cr-amanitin (Baulieu et al., 1972a), suggesting a new synthesis of mRNA for IP. By pulse-labeling techniques an increase in in uiuo synthesis of labeled RNA can be observed within 30 minutes after estrogen administration (Hamilton, 1968; Wira and Baulieu, 1972). Glasser et al. (1972) also showed that a transient increase in a-amanitin-sensitive RNA polymerase activity can be detected within 10-15 minutes after estrogen treatment of animals. The simplest and currently accepted conclusion that can be drawn from the above studies is that estrogen (with or without receptor) enhances the synthesis of an mRNA that codes for specific protein(s). The protein(s) is then required for the subsequent increase in production of rRNA, and maybe also for other mRNA. In rat liver the need for a supply of protein(s) for rRNA synthesis is shown by the finding of Yu and Feigelson (1972) that in the nucleolus the turnover of RNA polymerase activity is rapid (tllP1.3 hours). Since rat liver cytosol and nuclei contain large amounts of wamanitin-insensitive (our recent observation) polymerase in soluble forms (Liao et al., 1968), the protein(s) required are probably regulatory factor(s) rather than catalytic portions of the enzyme complex. An alternative suggestion that the RNA product of a-amanitin-sensitive polymerase action may be directly involved in nucleolar RNA synthesis (Sekeris and Schmid, 1973) has not been explored. Several workers studied the effect of hormones on the template activity of nuclear chromatin of target tissues. Nuclear chromatins were prepared from hormone-deficient and also from hormonetreated animals and assayed for their capacity to support RNA synthesis in the presence of excess amounts of purified RNA polymerase. Cortisone (for liver of adrenalectomized rats, see Dahmus and Bonner, 1965; Beato et al., 1970a), estradiol (for uterus of ovariectomized rats, see Barker and Warren, 1966; Teng and Hamilton, 1968; Glasser et ul., 1972; for rabbit uterus, see Church and

142

SHUTSUNG LIAO

McCarthy, 1970; for chick oviduct, see Cox et al., 1973), thyroxine (for tadpole liver, see Kim and Cohen, 1966), and progesterone (for chick oviduct, see O’Malley et al., 1969) were found to enhance the template activity of isolated chromatin to a significant level in a matter of a few hours to several days. These observations are in accord with the finding that template activities of nuclear chromatin isolated from cells highly active in RNA synthesis in general are higher than those from cells showing a low rate of RNA synthesis (Bonner et al., 1968). Under the conditions that stimulate RNA synthesis, however, no significant increase in chromatin template activity is observed with androgens in rat ventral prostate (Liao and Lin, 1967), with estrogens in mouse uterus (Dati and Maurer, 1971), and with aldosterone in rat kidney (Trachewsky and Cheah, 1971). Caution in interpreting chromatin template activity as assayed in uitro has been expressed by many workers (Liao and Fang, 1969; Dati and Maurer, 1971; Glasser et al., 1972; Cox et al., 1973). Factors such as hydrolytic enzymes (nucleases, nucleotidases, proteolytic enzymes), which may affect template activity, have not been studied carefully in most instances. Whether the chromatin fractionated from target nuclei still represents the native status of all or selective regions of chromatin can also be questioned. More seriously, the RNA polymerase (mostly bacterial) used in these studies may not produce the same type of RNA as in hormone-stimulated nuclei in uiuo or in an isolated nuclear system (Liao and Lin, 1967; Butterworth et al., 1971). In addition, if only a few genes are activated, changes in template activity or an increase in incorporation of radioactive precursor into the RNA fraction probably may be very small and may not be detected. Alteration of DNA-associated chromatin components by hormones can be probed by measuring the DNA sites available for a c tinomycin-D binding in intact nuclei. The method initially employed for rat prostate suggests that there is no androgen-induced gross unmasking of DNA (Liao and Lin, 1967; Barton, 1967; cf. Seligy and Lurquin, 1973). Aldosterone, in rat kidney, results in lower actinomycin-D binding which appears to be correlated with an alteration in chromatin components (Trachewsky et al., 1972). Cortisol, within 10-20 minutes, enhances about 15% of actinomycin-D binding capacity and template activity of rat liver nuclei in uiuo and in uitro, suggesting that some chromosomal proteins are removed from the DNA surface (Beato et al., 1970a). Estradiol (Teng and Hamilton, 1970) and cortisol (Shelton and Allfrey, 1970) have been shown to

CELLULAR RECEPTORS FOR STEROID HORMONES

143

play a regulatory role in the synthesis of nonhistone nuclear protein in target tissues. Similarly, steroid-induced changes in nuclear protein components in rat prostate by androgen (Chung and Coffey, 1971; Couch and Anderson, 1973; Anderson et al., 1973), and in rat kidney and heart by aldosterone (Liew et al., 1973), have been seen. In some instances chromatin fractions with high template activities have been found to have high concentrations of acetyl (Pogo et al., 1968) or phosphate esters (Allfrey et al., 1966). An increase in histone acetylation has been reported in liver and thymus of adrenalectomized rats after treatment with cortisol (Allfrey et al., 1966), in kidney of adrenalectomized rats treated with aldosterone (Libby, 1972a; Liew et al., 1973; Trachewsky and Lawrence, 1972), and in rat uterus treated with estradiol (Libby, 1972b). In rat uterus the effect was seen between 5 and 10 minutes after estrogen administration, but not at a later period (20 minutes). Since actinomycin D and cycloheximide were not inhibitors, the synthesis of new proteins or RNA was not considered to be required (Libby, 1972b). A confirmation of this important study is still lacking (cf. Anderson and Gorski, 1971). Hormone specificity has often been considered to be due to the specific activation of certain genes. This is supported by the fact that many hormones can selectively increase production of the amounts of specific enzymes or proteins in target cells. The most exciting achievements in this regard came from the laboratories of Shimke and of O’Malley and co-workers, who by isolating mRNA and translating it in uitro supplied clear evidence showing an estrogen-dependent increase in the mRNA for ovalbumin in chick oviduct (Rhoads et al., 1971; Palmiter and Schimke, 1973; Comstock et al., 1972; Chan et al., 1973). A similar study was made in Feigelson’s laboratory on the glucocorticoid control of mRNA for hepatic trytophan oxygenase (Schutz et al., 1973). One of the interesting features of the oviduct system is that the induction of ovalbumin mRNA is seen only with oviduct previously treated with estrogens for several days. During the primary period estrogens may be considered to act at cellular sites other than those directly involved in ovalbumin mRNA production. Whether or not different estrogen receptors are involved in these two stages of estrogen action is not known. Elucidation of the roles of steroid hormones in such a system and in others in which cell differentiation is involved may be tied directly to a better understanding of the mechanisms by which a cell can switch from one type of gene program to another.

144

SHUTSUNG LIAO

B. HYPOTHETICALMODELS Immediately after our first realization that genetic formation is transmitted from DNA by mRNA to direct the synthesis of a specific protein, the “hormone-gene theory” was proposed (Zalokar, 1961; Liao and Williams-Ashman, 1962; Williams-Ashman, 1965). With the appearance of the original operon theory initially postulated for bacterial systems, and the suggestion that induction of some bacterial enzymes is the result of the removal of gene repressors to allow mRNA synthesis, it was proposed that hormones might act by incapacitating certain gene repressor molecules (Karlson, 1963). While this is still one of the most attractive hypotheses, it is now clear that the regulation of genetic expression can be achieved in a variety of ways at different sites: synthesis of RNA on DNA template (transcriptional control), processing of RNA, protein synthesis (translational control), or degradation of RNA and proteins. Taking into account some of the newer information, several hypothetical models (Fig. 5) have been proposed to show how steroid hormones may act on the molecular processes of gene expression. The incapacitation of a gene (or transcriptional) repressor may be achieved by the interaction of a steroid with a repressor. The repressor itself then is the ultimate receptor for the steroid. If this is so, 1. NEOATNE CONTROL A . Transcrptional Repressor

N-{t-+y,PRESSOR N-&+y,PRESSOR

,

I

STFROID RECEPTOR STFRAID RECEPTOR

’I

INACTIVATION

&ACTIVATION

STEROID RECEPTOR [\RNA

FIG.5. Hypothetical models showing how steroid hormones and their receptors may regulate gene expression in target cells. See text for explanation. Double helix at the left side of each model represents DNA.

CELLULAR RECEPTORS FOR STEROID HORMONES

145

the need for nuclear retention of the cytosol steroid-receptor complex is puzzling. It is possible that the cytosol steroid-binding protein is merely a transport protein which supplies a specific steroid to a specific nuclear locus where it functions (Beato et al., 1970b), possibly by transferring the steroid to another nuclear receptor (Baulieu et aZ.,197213). Alternatively, the steroid receptor and not the steroid alone may interact with the repressor. Repressor activity can also be abolished if its synthesis is inhibited by a steroid or steroid-receptor complex (Fig. 5, 1.A). Tomkins et aZ. (1970, 1972) in a study of the HTC cell system, described a “paradoxic” increase in the rate of tyrosine aminotransferase activity (superinduction) following actinomycin-D addition to glucocorticoid-treated cells. They proposed the existence of a labile posttranscriptional repressor which inhibits translation of specific mRNA and enhances mRNA degradation. Superinduction is thought to be produced by actinomycin-D inhibition of the synthesis of a translational repressor, although a more traditional model based on transcriptional control has been offered to explain similar results (Reel and Kenny, 1968; Palmiter and Schimke, 1973). From these studies, Tomkins et al. (1970) suggested that the steroid-receptor complex inhibits translational repressor activity by direct interaction with the repressor or by inhibiting its synthesis (Fig. 5, 1.B left). In a study of enzyme induction in prostate, Ohno has suggested that the receptor protein for an androgen may act as a “translational block” by binding to certain mRNAs and preventing them from being translated by ribosomes (Fig. 5, 1.B right). Androgen is assumed to bind to the suppressive protein. As a result, the mRNAs are released and utilized for the production of specific enzymes. Ohno has speculated further that the same androgen-receptor complex enters the prostate nuclei and activates nucleolar RNA polymerase (Ohno, 1971). It has been suggested that RNA polymerase may be present in excess in mammalian cells (Liao and Lin, 1967; Liao et aZ., 1968), and that the restricted distribution of certain nucleotide sequences may serve an important function in directing polymerase to a specific area of the DNA template or in regulating the rate of RNA synthesis. It was suggested that the action of steroid hormones may involve a confonnational fit of molecules required for the initiation of RNA synthesis. In recent years regulatory factors of this type that are necessary for specific synthesis of RNAs by polymerase have been described in bacterial systems. As distinct from “negative control” involving repressor-depression action, the role of these factors is to

146

SHUTSUNG LIAO

specify the DNA sites where RNA synthesis is to occur and therefore can be categorized as “positive control” (Fig. 5, I1 left). The possibility that steroid hormones are involved in posttranscriptional events has not been well investigated. Among some of the pertinent observations are that cortisol can affect the processing of 45s RNA in rat liver (Jacob et al., 1969) and that, in the uterus, the percentage of nuclear RNA transported to the cytoplasm (Church and McCarthy, 1970) and the maturation of ribosomal precursor RNA or particles (Luck and Hamilton, 1972) can be stimulated by estrogen treatment of the animals. Whether or not these occurrences are due to a direct effect of steroid hormones has not been determined. From the finding that androgen- and estrogen-receptor complexes can associate with certain nuclear RNP particles in target cells, Liao et al. (1973a) have recently proposed that the steroidreceptor complexes may be involved in the processing of nuclear RNA or in the protein synthesis (possibly initiation). In the hypothetical model (Fig. 5, I1 right), a steroid hormone forms a complex with a receptor protein in the cytoplasm. After a conformational change, the complex enters the cell nucleus and becomes involved in the regulation (possibly initiation) of RNA synthesis. During such a process the steroid-receptor complex and other protein factors (including an acceptor) may recognize and bind to certain sequences of DNA initially, but then to the specific RNA product (Liao and Fang, 1969). Steroid-receptor-bound RNP may be processed to a mature form and enter the cytoplasm and participate in protein synthesis. One can visualize the role of the steroid-receptor complex as providing structural specificity needed for the formation (i.e., selecting specific RNA species from a large RNA pool), processing, and/or functioning of RNP (cf. Kwan and Brawerman, 1972; Blobel, 1973; Weinberg, 1973). In the above model the receptor protein may lose its ability to bind to RNP at various stages of processing and utilization, especially if the steroid hormone of the cell is depleted. Both the receptor proteins and the acceptor factors may reassociate with these RNP particles when the steroid hormone is replenished. Thus the recycling process and its functions may be reinitiated by the steroid hormone at many different points (in the nucleus or in the cytoplasm) in the receptor cycle (Liao et al., 1973a,b; Liao, 1974). This suggests that the importance of gene transcription (RNA synthesis) in relation to gene translation (protein synthesis) for the overall functioning of a steroid hormone in target cells may be dependent on the amount of RNP particles at different stages of processing, and on their RNA and

CELLULAR RECEPTORS FOR STEROID HORMONES

147

protein constituents in the target cells at the time the hormone is supplied. If the target cells contain sufficient amounts of RNA and protein constituents of RNP, the early actions of the hormone may be simply dependent on the processing and utilization (or the activity) of RNP, and not upon RNA synthesis. If this is so, steroid hormone actions under some conditions may not be actinomycin-D-sensitive (cf. Talwar et al., 1965; Frieden et al., 1968). Recently, Whelly and Barker (1974)showed that 1 hour after estradiol administration in the ovariectomized mature rat there is a transient increase in the rate of peptide elongation in isolated uterine ribosomes. This effect was reported not to be inhibited by actinomycin D. Liang and Liao (1974b) also found that androgen injected into castrated rats can, within 10 minutes, stimulate the ability of the prostate cytosol fraction to sustain GTP-dependent binding of an initiator tRNA labeled with m e t h i ~ n i n e - ~ % to prostate ribosomal particles. The finding indicates the possibility that androgen may be involved in the regulation of the activities of the initiation factors in protein synthesis, C. In Vitro EXPERIMENTAL APPROACHES A clear understanding of the molecular mechanism involved in steroid hormone action must be dependent on demonstration of the a hormonal effect in a cell-free system. Attempts by numerous investigators to show a change in RNA-synthesizing activity by in vitro addition of physiological levels of steroids to isolated cell nuclei generally have failed. At high concentrations (> 10 testosterone and cortisol were reported to enhance (by 10%) RNA synthesis of isolated liver cell nuclei (Lukacs and Sekeris, 1967). Seshadri and H Warren (1968)also reported that RNA synthesis from g ~ a n i n e - ~by rat uterine nuclei was enhanced by 2 nM to 1 pM estrone. Estradiol was reported to be less active. Barker and Warren (1966, 1968) also reported that DNA template activity for RNA synthesis of uterine estrone. Beato chromatin can be increased by incubation with 5 et al. (1968) reported a similar effect of cortisol on liver nuclear systems. It has been also claimed that dihydrotestosterone but not testosterone, in the absence of added cytoplasmic protein, can stimulate the incorporation of radioactive nucleosides into RNA fractions by cell nuclei of rat ventral prostate (Bashirelahi et al., 1969; Bashirelahi and Villee, 1970). Most of these observations, however, have not been confirmed by other investigators. Probably the first experimental report to describe a steroid hormone as an inactivator of the repressor activity of steroid-binding

a),

148

SHUTSUNG LIAO

protein came from Talwar et al. (1964), who found that in the absence of estradiol the cytosol fraction of rat uterus was inhibitory for RNA synthesis directed by purified calf thymus DNA and E. cold RNA polymerase. It was claimed that the injection of estradiol 1 hour prior to sacrifice, or direct addition of a minute amount of estradiol, abolished the inhibitory action of the cytosol preparation. DeSombre et al. (1966), after extensive and careful studies, obtained a similar cytosol preparation which inhibited in uiuo RNA synthesis, but they could not demonstrate reversal of the inhibition by estradiol in uitro or in uiuo. In close accord with Talwar’s report, Wacker (1965) has reported that a soluble macromolecular protein fraction obtained from extracts of Pseudomonas testosteroni also inhibited RNA synthesis in the presence of sperm DNA and purified E . coli RNA polymerase. The protein fraction was less inhibitory when it was isolated from cells grown in the presence of testosterone, an inducer of several steroid-transforming enzymes. They also claimed to observe a partial reversal of the inhibition by addition of low levels of testosterone in uitro. Similar observations were reported by the same group on Streptomyces hydrogenans (Wacker et al., 1965b) and E. coli (Wacker et al., 1965a). Shikita and Talalay (1967) reinvestigated the P . testosteroni system in great detail. They found that the inhibitory effect resides largely in a heat-stable component, but they were unable to confirm that inducer steroids had any direct effect on the inhibitory properties, or any quantitative differences in the inhibitory power of fractions derived from steroid-induced or noninduced cells. Raynaud-Jammet and Baulieu (1969) reported that the ability of the nuclei from heifer endometrium to incorporate radioactive nucleotides into RNA is enhanced manyfold when they are first incubated with a mixture of estradiol and uterine cytosol. Since estradiol or cytosol alone was not effective, interaction of the hormone and receptor protein in the cytosol was considered to be responsible for the stimulative activity. Mousseron-Canet and her co-workers also reported that an increase in RNA polymerase activity of heifer endometrium nuclei, or of the enzyme preparation from these nuclei, could be produced by the direct addition of certain uterine fractions to the polymerase assay system (Amaud et d.,1971; Andress et al., 1973). Studies by Jensen and co-workers (1972b) showed that nuclei from immature rat uteri have much less ability to incorporate labeled nucleotide into RNA than do kidney or liver nuclei. After incubation

CELLULAR RECEPTORS FOR STEROID HORMONES

149

(25°C for 30 minutes) with uterine cytosol containing estradiol, the RNA polymerase activity of the reisolated uterine nuclei was reported to be increased nearly three-fold. Cytosol or estradiol alone was not effective. There was no enhancement of the already high activity of liver or kidney nuclei, whether incubation was in the uterine cytosol or in their own cytosols. Most importantly, these investigators observed that nuclei from calf endometrium were activated by incubation with estradiol and endometrial cytosol at 25"C, but not at 0°C. However, the stimulation can be seen with nuclei incubated at 0°C if the estradiol-cytosol mixture is first warmed to 25°C. Since the receptor transformation from the 4 s to the 5s form (see Section IV,A) takes place readily at 25°C but not at O"C, and the 5s form, but not the 4 s complex, can associate with the uterine nuclei, it was suggested that the 5s complex is responsible for the effect. In accord with this view, estrone, which binds to the 4 s form but does not readily induce its transformation to 5s at the concentration used, did not cause nuclear stimulation. Since estradiol cytosol-dependent stimulation was also seen with RNA polymerase activity solubilized from treated nuclei and assayed with purified DNA, the steroid receptor effect was thought to be at least in part on the enzyme itself (Mohla et a2., 1972; DeSombre et al., 1972; Jensen and DeSombre, 1973). According to Beato et al. (1970b), liver nuclei, in the presence of cytosol, responded with increased RNA synthesis to cortisol in the range of 0.1-0.01 pM, which is far lower than the amounts needed for glucocorticoids to stimulate the nuclei (10 /AM). They believe that steroid-binding proteins facilitate the transport of hormones into the nucleus, indicating support for the traditional mechanism (Fig. 5, 1.A) involving the interaction of a steroid, rather than the receptor protein, with RNA-synthesizing machinery. Abraham and Sekeris (1971) also reported that cortisol (10 /AM) could inhibit rRNA synthesis of isolated rat thymus nuclei. However, when receptor protein was extracted from the nuclei, no cortisol effect was observed. The addition of the receptor back into the system was effective in restoring sensitivity to cortisol (Van Der Meulen et al., 1972). The effects of cortisol on liver and thymus described above were also shown by Bottoms et a2. (1972). Using a fractionated cortisolbinding protein from rat liver cytosol, Gopalakrishnan and Sadgopal (1972) and Beato et al. (1970b) detected changes in the template activity of liver chromatin due to the presence of cortisol hemisuccinate (10 pM) or the cortisol-protein complex. Ribarac-Stepic et al.

150

SHUTSUNG LIAO

(1973) also obtained a cytosol preparation of cortisol-protein complex which could enhance the incorporation of radioactivity into the RNA fraction during the incubation of liver cell nuclei and orthop h ~ s p h a t e - ~ ~While P . Gopalakrishnan and Sadgopal (1972) showed that the active cortisol-protein complex was salted out at 50-70% ammonium sulfate saturation, Beato et al. (1970b) and RibaracStepic et al. (1973) reported that the active complex is salted out at 20-40% ammonium sulfate saturation. Davies et al. (1972), working on rat and dog prostate nuclei, observed an increase in the incorporation of radioactive isotope from %-labeled ATP into the RNA fraction by 40 pA4 of dihydrotestosterone and 5a-androstane-3p7 l7/3-diol in the absence of cytosol. Several other related steroids were inactive, while many estrogens were inhibitory. At 4 pA4, the effect was observed only in the presence of a cytosol preparation (Davies and Griffiths, 1973a,b). The steroid effect was observed with fractionated cytosol receptor (8 or 3s) or nuclear receptor (4.5s). Both dihydrotestosterone-protein complexes (I and 11)as originally described by Fang and Liao (1971) appear to be active. When RNA polymerase was isolated from prostate nuclei and assayed in the presence of prostate chromatin as exogenous template, the extent of stimulation was most distinct when the polymerase was the nucleolar (polymerase I) form (50 to 140%increase). The nucleoplasmic (polymerase 11)form gave a marginal effect of about 10%.An insignificant increase was observed when calf thymus DNA or liver chromatin was employed. The direct effect on nucleolar enzyme activity is interesting, since nucleolar RNA polymerase activity is particularly sensitive to androgen action in uivo (Liao and Lin, 1967; Liao and Stumpf, 1968).This view, however, does not conform with the previous suggestion that the enhancement of nucleolar RNA synthesis is secondary to a stimulation of nucleoplasmic RNA polymerase activity. It is possible, however, that the RNA polymerase activity purported to be nucleolar may indeed be responsible for the synthesis of rRNA as well as some mRNA (Liao and Fang, 1969). An in uitro demonstration that RNA synthesis can be affected by steroid hormones or the steroid-receptor complex appears strongly to support the current concept that nuclear chromatin is the ultimate cellular site for hormone action. Although our understanding of the chromatin activity involved in gene transcription is very limited and the results of the in uitro effects described so far can be subjected to severe criticism for their ambiguity, further study in this area may unfold the mystery ofthe molecular processes involved in steroid function. Many other possibilities, however, should also be considered.

CELLULAR RECEPTORS FOR STEROID HORMONES

151

VI. Concluding Remarks A. RECEPTORAND UPTAKE

OF

STEROID

BY

CELLS

Peck et al. (1973) examined the uptake and retention of estradiol3H by uterine and diaphragm tissue under initial velocity and equilibrium conditions, respectively. The rate of uptake of estradiol by the nontarget tissue, diaphragm, which possesses no estrogen receptor, is the same as that for uterus. Inhibitors of estrogen binding to receptor also do not alter the rate of uptake of estradiol in either tissue. These results also indicate that the estrogen receptor is not involved in the movement of estradiol into uterine tissue. The cell membranes of thymus cells are also freely permeable to cortisol, since the rate of dissociation of selectively retained cortisol from thymus cells with a time constant of 3 minutes at 37°C is essentially identical to that of glucocorticoid from the isolated cortisol-receptor complex (Munck and Wira, 1971; Bell and Munck, 1973). The possibility that a protein-mediated process is involved in estrogen entrance into rat uterine cells is suggested by Milgrom et al. (1973d), who showed that sulfhydryl-blocking agents inhibit the uptake of radioactive estrogens by target cells. The estrogen receptor is not considered to be the protein, since the uptake of diethylstilbestrol, which binds to the receptor, is not blocked significantly.

B. NATURAL FORMS OF STEROIDRECEPTORS Essentially, in all the systems described in this article, cytosol receptors carefully prepared at low temperatures (0°-2"C) can form complexes that sediment as 7-12s and 3-5s units. The larger forms can be transformed to the smaller ones by incubating at 20°-37"C or by making the salt concentrations 0.4 M KC1. Since many conditions are clearly not physiological, it is not clear whether these forms reflect an in viuo situation or are products of the factitious association of receptor themselves or receptors with other cellular materials, To mimic natural physiological conditions, cytosol estradiol-receptor complexes have been studied in 0.15 M KCl and found to be about 6s (Giannapoulos and Gorski, 1971b; Chamness and McGuire, 1972). It is interesting that, when the estrogen-receptor complex was pressed directly from the uteri with no added buffer and sedimented in sucrose gradients prepared in the same press juice (but deproteinized), only the 6s form was detectable (Reti and Erdos, 1971). These observations indicate, but do not prove, the possibility that the 6s form is the predominate one in the target cytoplasm.

152

SHUTSUNG LIAO

Several investigators have also pointed out the danger in the misuse of gradient centrifugation in identifying natural forms of receptor molecules. For example, Harris (1971) showed that by adjusting the concentration of the polyanion heparin, Polytak-RNA, or dextran sulfate, the 4-5s estradiol-receptor complex of rat uterine nuclei can sediment at 8s in a low-salt sucrose gradient. A similar study made by Chamness and McGuire (1972) also demonstrated that one can deliberately cause the cytosol or nuclear estrogenreceptor complex to assume any form sedimenting between 4 and 9s by altering the heparin concentration. In salt concentrations ranging from no KC1 to 0.4 M KC1, the estradiol-receptor complex has been observed in forms with sedimentation coefficients from 3.8 to 9s. These values appear to result from a slow aggregation of estradiolbinding proteins following tissue homogenization. Since the timedependent aggregation can be minimized by working with dilute solutions, Stance1 et al. (1973a) concluded that in uivo the uterine estradiol receptor may exist as a 3.8-4.8s species, rather than as 8s as normally assumed. The extreme instability of 8s complexes at temperatures between 10" and 37°C may indicate that such a complex, if formed in uiuo, is very rapidly altered to the 3-5s form. Unless the process is tied directly to receptor function, the usefulness of forming and degrading the 8s complex can be questioned. Nevertheless, the fact that 8s forms are found in the extracts of virtually all steroid target cells remains an interesting puzzle deserving more study. As described above, different receptors for different groups of steroid hormones may exist in the same target cell. In various responsive tissues, steroid action may be dependent on heterogeneous receptors specific for different steroids. The presence of heterogeneous receptors for the same steroid hormone in a target cell, and thus different ways of functioning, is a distinct possibility but has not been clearly proven. In rat uterus a majority (80%) of the cytosol estradiol-receptor complexes can enter the cell nuclei. In rat ventral prostate only about 50% of the cytosol receptor can be retained by nuclei or the nuclear RNP particles (Fang and Liao, 1971; Liao et al., 1973a). Whether the remaining portion of the receptor represents another type or altered form of receptor is not clear. Similar considerations can be given to other steroid-receptor complexes described.

c.

NATURE O F RECEPTOR-STEROID INTERACTION

In the past attempts have been made to use a semiempirical approach by comparing chemical structure and end-point activity to

CELLULAR RECEPTORS FOR STEROID HORMONES

153

predict the way androgens might interact with hypothetical receptors (see reviews by Liao and Fang, 1969).Suggestions were made for the binding of androgens by receptors from &face, @face, and peripheral attachments. It was also indicated that the steric and not the electronic characteristics of the steroid are the most important in eliciting a biological response (Wolff and Zanati, 1970, and references cited therein). These earlier predictions are well in line with the conclusions of Liao et al. (1972, 1973b), who studied the structural requirements for steroids to bind to the defined androgen receptor of rat ventral prostate. It is suggested that the bulkiness and flatness of the steroid molecule play a more important role in receptor binding than the detailed electronic structure of the steroid nucleus. The role of the A4-3-keto-5a-oxidoreductaseis apparently to convert testosterone to a flatter molecule which fits better to the receptor binding site and not simply to eliminate the double bond on ring A of the steroid. In fact, potent androgens with conjugated double bonds extending from rings A and B to ring C (such as 2-oxa-17~-hydroxyestra-4,9,1l-trien%one) are indeed very flat molecules and bind to the androgen receptor very firmly. These and other studies using methylated androgens suggest that the receptor binds simultaneously at multiple sides of an androgenic steroid as if the steroid molecule were being enveloped. This is in marked contrast to steroid-metabolizing enzymes or blood steroid-binding proteins which generally recognize only a portion of the steroid molecule. This conclusion is in line with our recent observation that steroids bound to various blood proteins can be removed by various steroid antibodies, while steroids bound to various cellular receptors are not affected (Section IV,A). The localization of steroid-binding sites well inside receptor proteins may be responsible for the very high affinity constants for receptor binding of steroids, the extremely slow rates of association and dissociation of steroids and receptor proteins (many hours) at low temperatures, the acceleration of rates of exchange of unbound steroids with bound steroids by freezing and thawing (Fang and Liao, 1971), and the inability of ethanol (30%)and detergents (2% Triton X-100 or desoxycholate) to free steroids from receptors in the cold. Some careful analyses of glucorticoid-binding properties have also enabled several workers to suggest the nature of the interaction of steroids and receptors. Using thymus cells, Munck et al. (1972) showed that the association rates of cortisol, dexamethasone, and cortexolone are very similar, and that differences in their binding constants are largely determined by dissociation rates (Bell and Munck, 1973). They suggest that the groups that distinguish these steroids,

154

SHUTSUNG LIAO

particularly the llp-hydroxyl and the 9CY-fluor0, do not come into play until the steroid has entered the steroid-binding site of a receptor. These and other findings on structural requirements for receptor binding also allow these investigators to conclude that both the a and psides of the steroid interact with the binding sites, which are probably located in a hydrophobic pocket in the receptor protein. Koblinsky et al. (1972) also showed that, while the interaction of cortisol with liver A and B proteins, and transcortin-binding of corticoids, are accompanied by a negative entropy change dexamethasone interaction with the G protein gives a positive entropy change (cf. Westphal, 1971). The finding is interesting in that the entropy change may be due to the displacement of water molecules (Warner, 1965) as well as to a local relaxation in the structure of the protein, which leads to an increased degree of freedom. This is probably necessary if the “enveloping” of a steroid by a hydrophobic region of a protein inside a protein (cf. Engle, 1967) is to occur. The enveloping of a steroid in the hydrophobic pockets of a receptor protein is likely to invoke the reorientation of certain flexible polypeptide chains. This may result in a conformational change in the protein, which is necessary before the steroid-receptor complex can interact with acceptor complexes and trigger hormone action. It is interesting that antiglucocorticoids such as progesterone-,H in HTC cells (Rousseau et al., 1973) and corte~olone-~H in rat thymocytes (Kaiser et al., 1972) bind to cytosol glucocorticoid receptors, but the complexes are not retained by the nuclei of these target cells, apparently because of their structural incompatibility. The concept of steroid enveloping suggests that receptor proteins, and not specific functional groups of steroids, participate in the key events leading to hormone action. This may explain why many different steroid hormones exhibit crossover hormonal activities.

D. INSECTHORMONES AND VITAMIN D, It is now apparent that cellular receptor proteins for steroid hormones in vertebrate animals have many properties in common and may function in similar ways in target cells. This generalization can also be extended to some steroidal insect hormones (Gilbert and King, 1973) and, most interestingly, to vitamin D. Vitamin D3 appears to be metabolized in the liver and other tissues to 25-hydroxycholecalciferol, and subsequently in the kidney to 1,25-dihydroxycholecalciferol, before it can act in initiating intestinal calcium transport (Haussler et al., 1971; Boyle et al., 1972; Wong et al., 1972). There is evidence that the active (hormonal) form of vitamin D,

CELLULAR RECEPTORS FOR STEROID HORMONES

155

binds to the chromatin protein of target cells (Tsai et al., 1972; Chen and DeLuca, 1973),and that nuclear binding is dependent on cytosol receptor protein (Tsai and Norman, 1973; Brumbaugh and Haussler, 1973).

E. CYCLICAMP AND STEROIDHORMONES Many peptide hormones employ cyclic AMP as a “second messenger” for their actions (Robison et al., 1971).The suggestion that some steroid hormones may use cyclic nucleotides as second messengers has not gained general support (Hechter and Soifer, 1971; Major and Kilpatrick, 1972). In the uterus, Szego and Davis (1967) reported acute elevation of cyclic AMP level by estradiol in vivo in 15 seconds, and cyclic AMP was seen to mimic certain effects of estrogen including amino acid and nucleotide incorporation into macromolecular fractions (Hechter et al., 1967; Griffin and Szego, 1968; Sharma and Talwar, 1967). Korenman et al. (1973) reported recently that estradiol did not stimulate cyclic AMP in vivo or in vitro. Singhal et al. (1971) showed that the administration of cyclic AMP produces testosteronelike induction of certain enzyme activities in rat prostate. The adenyl cyclase activities of various prostate preparations, however, are not affected by castration or androgens in vivo or in vitro (Rosenfeld and O’Malley, 1970; Liao et al., 1971a). Similarly, cortisol does not appear to influence the adenyl cyclase activity of rat liver cells (Soifer and Hechter, 1971). Although Ahmed (1971) has reported that androgens in vivo enhance the ability of prostate nuclei to phosphorylate nuclear proteins in vitro by “P-labeled ATP, Ichii et al. (1973)showed that androgens can decrease cyclic AMP-dependent protein kinase activity in prostate. Besides the well-known action of cyclic AMP on protein kinase, it is known that in bacterial systems cyclic AMP can bind to a specific protein and that the complex formed can promote the production (by binding to genome) of specific mRNA which codes for specific enzymes (Zubay et d., 1970; Pastan and Perlman, 1970). Since peptide hormone-sensitive adenyl cyclase appears to be present in the cell nuclei of rat liver (Soifer and Hechter, 1971) and rat ventral prostate (Liao et al., 1971a), a similar mechanism may be operating in mammalian cells. Mangan et al. (1973), however, reported that nuclear RNA polymerase activities that are enhanced by androgens are not stimulated by administration of cyclic AMP. It is intriguing to point out that cyclic AMP and steroid hormones are similar in many aspects. For example, their production is under the influence of some peptide hormones. Each of them can conform

156

SHUTSUNG LIAO

FIG. 6. Molecular models of steroid hormones and cyclic AMP. Corey-Pauling atomic models with Koltun connectors were used to construct estradiol (top left), cortisol (top right), dihydrotestosterone (bottom left) and cyclic AMP (bottom right). For steroids, they are side views with ring-A oxygens at the left and the a faces of the steroid facing downward. For cyclic AMP, the adenine base is at the left and the phosphate group at the far right. White balls are hydrogens, caged balls are hydroxyl oxygen, and black blocks are carbon atoms. For construction of models, the following articles were consulted: Cooper et uZ. (1969), Norton (1965), Cooper and Duax (1969), and Watenpaugh et aZ. (1968).

to rather compact structures which are very similar in general geometric dimensions (Fig. 6). Both appear to function in target cells by binding noncovalently to specific receptor proteins. In a bacterial system cyclic AMP, like steroids, appears to enhance the affinity of the receptor protein (receptor transformation) for genome (Riggs et al., 1971; Anderson et al., 1972) and to promote the production of specific species of RNA. In calf ovary cells there is now experimental evidence indicating a cyclic AMP-dependent translocation of cytoplasmic cyclic AMP-binding protein and possibly of cytoplasmic protein kinase to nuclear acceptor sites (Jungmann et aZ., 1974). In rat ventral prostate cyclic AMP-binding protein can be clearly distinguished from dihydrotestosterone receptor protein by an isoelectric focusing technique (Mangan et al., 1973). Cyclic AMP and androgen do not compete for the protein-binding sites. The complexes resemble each other in that they are found in the cytosol as well as in the nuclei (extracted by 0.4 M KCl), and they sediment in

CELLULAR RECEPTORS FOR STEROID HORMONES

157

the vicinity of 5s and 3s (Liao and Liang, 1974; Liao et aZ., 1974b). Our study has revealed that the cytosol cyclic AMP-binding protein can be retained by prostate cell nuclei, but this transfer is not dependent on formation of the cytosol dihydrotestosterone-receptor complex. The similarity between the corticosteroid-binding protein (binder 11) and cyclic AMP-binding proteins in rat liver is more obvious. They have several similar properties including molecular weight, isoelectric point, and comigration as one distinct band on polyacrylamide disc gel electrophoresis. These binding proteins in rat liver, however, can be distinguished from the regulating subunit or the protein inhibitor of phosphoprotein kinase (Filler and Litwack, 1973). ACKNOWLEDGMENTS

Research carried out in the author’s laboratory was supported by grants from the U S . National Institute of Health and the American Cancer Society, Inc. The author thanks Ms. Cassandra M. Black, Ms. Pamela A. Chudzinski, and Ms. Diane K. Howell

for their help in preparation of the manuscript. REFERENCES Aakvaag, A., Tveter, K. J., Unhjem, O., and Attramadal, A. (1972).J . Steroid Biochem. 3,375. Abraham, A. D., and Sekeris, C. E. (1971).Biochim. Biophys. Acta 247,562. Abraham, A. D., and Sekeris, C. E. (1973).Biochim. Biophys. Acta 297, 142. Adachi, K., and Kano, M. (1972).Steroids 19,567. Ahmed, K. (1971).Biochim. Biophys. Acta 243,38. Alberga, A., and Baulieu, E. E. (1968).Mol. Pharmacol. 4,311. Alberga, A., Massol, N., Raynaud, J.-P., and Baulieu, E. E. (1971).Biochemistry 10, 3835. Alberti, K. G. M. M., and Sharp, G. W. G. (1969).Biochim. Biophys. Acta 192,335. Allfrey, V. G., Pogo, B. G. T., Pogo, A. O., Kleinsmith, L. J., and Mirsky, A. E. (1966). In “Histones, Their Role in the Transfer of Genetic Information” (A. V. S . De Reuck, ed.), p. 50.Little, Brown, Boston, Massachusetts. Anderson, J. N., Peck, E. J., Jr., and Clark, J. H. (1973a).Endocrinology 92, 1488. Anderson, J. N.,Peck, E. J., Jr., and Clark, J. H. (197313).Endocrinology 93, 711. Anderson, K. M., and Liao, S . (1968).Nature (London)219,277. Anderson, K. M., Slavik, M., Evans, A. K., and Couch, R. M. (1973).E x p . Cell Res. 77, 143. Anderson, N. O., and Gorski, J. (1971).Fed. Proc., Fed. Amer. Soc. E x p . Biol. 30, 361. Anderson, W. B., Perlman, R. L., and Pastan, I. (1972)../. B i d . Chem. 247, 2717. Andress, D., Mousseron-Canet, M., and Raynal, F. (1973).C . R . Acad. Sci., Ser. D 276,97. Arias, F., and Warren, J. C. (1971).Biochim. Biophys. Acta 230,550. Armstrong, D.T., and King, E. R. (1970).Fed. Proc., Fed. Amer. Soc. E x p . B i d . 29, 250.

158

SHUTSUNG LlAO

Amaud, M., Beziat, Y.,Guilleux, J. C., Hough, A., Hough, D., and Mousseron-Canet, M. (1971).Biochim. Biophys. Acta 232, 117. Baird, D. T., Horton, R., Longcope, C., and Tait, J. F. (1969).Recent Progr. Horm. Res. 25.61 1. Ballard, P. L., and Ballard, R. A. (1972).Proc. Nat. Acad. Scl. U.S . 69,2668. Barker, K., and Warren, J. C. (1966).Proc. Nat. Acad. Sci. U.S. 56, 1299. Barker, K., and Warren, J. C. (1968).Endocrinology 80,536. Barton, R. W. (1967).Ph.D. Thesis, Univ. of Chicago, Chicago, Illinois. Barton, R. W., and Liao, S. (1967).Endocrinology 81,409. Bashirelahi, N., and Villee, C. A. (1970).Biochim. Biophys. Acta 202, 192. Bashirelahi, N., Chader, G. J., and Villee, C. A. (1969). Biochem. Biophys. Res. Commun. 37,976. Baulieu, E. E., and Jung, I. (1970).Biochem. Biophys. Res. Commun. 38,599. Baulieu, E. E., Lasnitzki, I., and Robel, P. (1968).Nature (London)219, 1155. Baulieu, E. E., Wira, C. R., Milgrom, E., and Raynaud-Jammet, C. (1972a).Acta Endocrinol. (Copenhagen),Suppl. 168,396. Baulieu, E. E., Alberga, A., Raynaud-Jammet, C., and Wira, C. R. (1972b). Nature (London),New 'Biol. 236,236. Baxter, J. D., and Tomkins, G. M. (1971a).Adoan. Biosci. 7,349. Baxter, J. D., and Tomkins, G . M. (1971b).Proc. Nat. Acad. Sci, U.S . 68,932. Baxter, J. D., Rousseau, G. G., Benson, M. C., Garcea, R. L., Ito, J., and Tomkins, G. M. (1972).Proc. Nat. Acad. Sci. U.S . 69, 1892. Beato, M., and Feigelson, P. (1972).1.Biol. Chem. 247,7890. Beato, M., Homoki, J., Lukacs, I., and Sekeris, C. E. (1968).Hoppe-Seyler's Z. Physiol. Chem. 349,1099. Beato, M.,Biesewig, D., Braendle, W., and Sekeris, C. E. (1969). Biochim. Biophys. Acta 192,494. Beato, M., Seifart, K. H., and Sekeris, C. E. (1970a).Arch. Biochem. Biophys. 138,272. Beato, M., Brandle, W., Bieoweig, D., and Sekeris, D. E. (1970b).Biochim. Biophys. Acta 208, 125. Beato, M., Schmid, W., and Sekeris, C. E. (1972).Biochim. Biophys. Acta 263, 764. Beato, M., Kalimi, M., Konstam, M., and Feigelson, P. (1973). Biochemistry 12, 3372. Beato, M., Kalimi, M., Beato, W., and Feigelson, P. (1974). Endocrinology 94, 377. Belham, J. E., Neal, G. E., and Williams, D. C. (1969).Biochim. Biophys. Acta 187, 159. Bell, P. A., and Munck, A. (1973).Biochem. J . 136,97. Best-Belpomme, M., Fries, J., and Erdos, T. (1970).Eur. J . Biochem. 17,425. Beyer, C., Larsson, K., PBrez-Palacios, G., and Morali, G. (1973).Horm. Behao. 4, 99. Billing, R. J., Barbiroli, B., and Smellie, R. M. S. (1969).Biochem. J . 112,563. Blaquier, J. A., and Calandra, R. S. (1973).Endocrinology.93,51. Blobel, G. (1973).Proc. Nat. Acad. Sci. U.S . 70,924. Blyth, C. A., Freedman, R. B., and Rabin, B. R. (1971).Nature (London),New Biol. 230, 137. Blyth, C. A,, Cooper, M. B., Roobol, A., and Rabin, B. R. (1972).Biochemistry 29,293. Bohus, B. (1970).Progr. Brain Res. 32, 171. Bonner, J., Dahmus, M. E., Fambrough, D., Huang, R. C., Marushige, K., and Tuan, D. Y. H. (1968).Science 159,47. Bottoms, G . D., Stith, R. D., and Roesel, 0. F. (1972).Proc. SOC. E x p . Biol. Med. 140, 946. Boylan, E. S., and Wittliff, J. L. (1973).Cancer Res. 33,2903.

CELLULAR RECEPTORS FOR STEROID HORMONES

159

Boyle, I. T., Miravet, L., Gray,R. W., Holick, M. F., and DeLuca, H. F. (1972). Endocrinology 90,605. Bresciani, F., Nola, E., Sica, V., and Puca, G. A. (1973).Fed. Proc., Fed. Amer. SOC. E x p . Biol. 32,2126. Brinkmann, A. O., Mulder, E., Lamers-Stahlhofen, G . J., Mechielsen, M. J., and Van Der Molen, H. J. (1972).FEBS (Fed. Eur. Biochem. SOC.), Lett. 26, 301. Britten, R. H., and Davidson, E. H. (1969).Science 165,349. Brooks, S . C., Lock, E. R., and Soule, H. D. (1973).J. Biol. Chem. 248,6251. Bruchovsky, N., (1971).Endocrinology 89, 1212. Bruchovsky, N., and Meakin, J. W. (1973).Cancer Res. 33, 1689. Bruchovsky, N., and Wilson, J. D. (1968a).J. Biol. Chem. 243,2012. Bruchovsky, N., and Wilson, J. D. (1968b).J.Biol. Chem. 243,5953. Brumbaugh, P. F., and Haussler, M. R. (1973). Biochem. Biophys. Res. Commun. 51,74. Bullock, L. P., and Bardin, C. W. (1970).J. Clin. Endocrinol. Metab. 31, 113. Bullock, L. P., Bardin, C. W., and Ohno, S . (1971).Biochem. Biophys. Res. Commun. 44, 1537. Burdon, R. H. (1971).Progr. Nucl. Acid. Res. Mol. Biol. 11, 33. Burke, C. W., and Anderson, D. C. (1973).Nature (London)240,38. Butterworth, P. H. W., Cox, R. F., and Chesterton, C. J. (1971).Eur. J. Biochem. 23, 229. Callantine, M. R., Clemens, L. E., and Shih, Y. (1968).Proc. SOC. E x p . Biol. Med. 128, 382. Castatieda, E., and Liao, S . (1974a).Methods in Enzymology (in press). Castatieda, E., and Liao, S . (1974b).J. Biol. Chem. (in press). Chader, G. J ., and Westphal, U. (1968a).J.Biol. Chem. 248,928. Chader, G . J., and Westphal, U. (1968b).Biochemistry 7,4272. Chamness, G . C., and McGuire, W. L. (1972).Biochemistry 11,2466. Chamness, G . C., Jennings, A. W., and McGuire, W. L. (1974).Biochemistry 13,327. Chan, L., Means, A. R., and O’Malley, B. W. (1973). Proc. Nut. Acad. Sci. U . S . 70, 1870. Chatkoff, M. L., and Julian, J. A. (1973). Biochem. Biophys. Res. Commun. 51, 1015. Chen, T. C., and DeLuca, H. F. (1973).1.Biol. Chem. 248,4890. Chu, L. L. H., and Edelman, I. S. (1972).J. Membrane Biol. 10,291. Chung, L. W. K., and Coffey, D. S . (1971).Biochim. Biophys. Acta 247,570. Church, R. B., and McCarthy, B. J. (1970).Biochim. Biophys. Acta 199, 103. Chytil, F., and Toft, D. (1972).J. Neurochem. 19,2877. Clark, J. H., and Gorski, J. (1969). Biochim. Biophys. Acta 192,508. Clark, J. H., and Gorski, J. (1970).Science 169,76. Clark, J. H., Anderson, J., and Peck, E. J., Jr. (1972).Science 176, 528. Clark, J. H., Anderson, J. N., and Peck, E. J., Jr. (1973).Adoan. E x p . Med. Biol. 36, 15. Clemens, L. E., and Kleinsmith, L. J. (1972).Nature (London),New Biol. 237, 204. Cohen, P., Chin, R. C., and Kidson, C. (1969).Biochemistry 8,3603. Comstock, J. P., Rosenfeld, G. C., O’Malley, B. W., and Means, A. R. (1972).Proc. Nut. Acad. Sci. U . S . 69,2377. Cooper, A., and Duax, W. L. (1969).J. Pharm. Sci. 58,1159. Cooper, A., Norton, D. A., and Hauptman, H. (1969).Acta Crystallogr., Sect. B 25,814. Corvol, P., and Bardin, W. (1973).Biol. Reprod. 8,277. Corvol, P., Falk, R., Freifeld, M., and Baulieu, E. E. (1972).Endocrinology 90, 1464. Couch, R. M., and Anderson, K. M. (1973).Biochem. Biophys. Res. Commun. 50,478.

160

SHUTSUNG LIAO

Cox, R. F., Haines, M. E., and Carey, N. H. (1973).Eur. J . Biochem. 32,513. CrabbB, J. (1963).“The Sodium-Retaining Action of Aldosterone.” Arscia, Brussels. Dahmus, M., and Bonner. J. (1965).Proc. Nut. Acad. Sci. U.S . 54, 1370. Dannenberg, V. H. (1963).Deut. Med. Wochenschr. 88,605. Dati, F. A., and Maurer, H. R. (1971).Biochim. Biophys. Acta 246,589. Davies, I. J., and Ryan, K. J. (1973).Endocrinology 92,394. Davies, I. J., and Ryan, K. J. (1972).Endocrinology 90,507. Davies, P., and Griffiths, K. (1973a).Biochem. Biophys. Res. Commun. 53,373. Davies, P., and Griffiths, K. (1973b).Biochem. J . 136,611. Davies, P., Fahmy, A. R.,Pierrepoint, C. G., and Griffiths, K. (1972).Biochem. J . 129, 1167. DeAngelo, A. B., and Gorski, J. (1970).Proc. Nut. Acad. Sci. U.S.66,693. Debeljuk, L., Arimura, A., and Schally, A. V. (1972).Endocrinology 90, 1579. DeHertogh, R.,Ekka, R. E.,Vanderheyden, I., and Hoet, J. (1970).Endocrinology 87, 874. DeHertogh, R.,Ekka, E., Vanderheyden, I., and Hoet, J. J. (1971).Endocrinology 88, 165. DeHertogh, R.,Ekka, E., Vanderheyden, I., and Hoet, J. J. (1973a).J. Steroid Biochem. 4,289. DeHertogh, R., Ekka, E., Vanderheyden, I., and Hoet, J. J. (1973b). J . Steroid Biochem. 4,301. DeHertogh, R.,Ekka, E., Vanderheyden, I., and Hoet, J. J. (1973~). J . Steroid Biochem. 4,313. DeSombre, E. R., Feldacker, B., Jungblut, P. W., and Jensen, E. V. (1966).Fed. Proc., Fed. Amer. Soc. E x p . Biol. 25,286. DeSombre, E. R.,Puca, G. A,, and Jensen, E. V. (1969).Proc. Nut. Acad. Sci. U.S . 64, 148. DeSombre, E. R., Chabaud, J. P., Puca, G. A., and Jensen, E. V. (1971).J. Steroid Biochem. 2,95. DeSombre, E. R., Mohla, S., and Jensen, E. V. (1972).Biochem. Biophys. Res. Commun. 48,1601. DeVries, J. R.,Ludens, J. H., and Fanestil, D. D. (1972).Kidney I n t . 2,95. DeWeer, P., and CrabbC, J. (1968).Biochim. Biophys. Acta 155,280. Diczfalusy, A., ed. (1970).Acta Endocrinol. (Copenhagen),S u p p l . 147. Dunn, J. F., Goldstein, J. L., and Wilson, J. D. (1973).J . Biol. Chem. 248, 7819. Edelman, I. S. (1971).Adoan. Biosci. 7,267. Edelman, I. S.,Bogoroch, R., and Porter, G. A. (1963).Proc. Nut. Acad. Sci. U . S . 50, 1169. Eisenfeld, A. J., and Axelrod, J. (1966).Endocrinology 79,38. Engel, L. L. (1967).Proc. Znt. Congr. H o m . Steroids, 2nd, Milan, 1966.Ser. No. 132, p. 52. Eppenberger, U., and Hsia, S . L. (1972). J . B i d . Chem. 247,5463. Erdos, T. (1968).Biochem. Biophys. Res. Commun. 32,338. Faber, L. E., Sandmann, M. L., and Stavely, H. E. (1972a).J . B i d . Chem. 247,5648. Faber, L. E.,Sandmann, M. L., and Stavely, H. E. (1972b).J. Biol. Chem. 247,8000. Falk, R. J., and Bardin, C. W. (1970).Endocrinology 86, 1059. Fanestil, D. D., and Edelman, I. S. (1966).Proc. Nut. Acad. Sci. U.S . 56,872. Fang, S. (1973).Personal communication. Fang, S . , and Liao, S . (1969).Mol. Phamacol. 5,428. Fang, S.,and Liao, S . (1971). J . B i d . Chem. 246, 16.

CELLULAR RECEPTORS FOR STEROID HORMONES

161

Fang, S.,Anderson, K. M., and Liao, S . (1969).J . Biol. Chem. 2p4,6584. Fazekas, A. G.. and Sandor, T. (1973).Proc. Znt. Congr. Endocrinol., 4th, Washington, D.C., 1972 p. 80 (Abstr.). Feder, H. H., Naftolin, F., and Ryan, K. J. (1974).Endocrinology 94,136. Feil, P. D., Glasser, S. D., Toft, D. O., and O’Malley, B. W. (1972).Endocrinology 91, 738. Feldman, D., Funder, J. W., and Edelman, I. S. (1973).Endocrinology 92,1429. Filler, R., and Litwack, G. (1973).Fed. Proc., Fed. Amer. SOC. E x p . Biol. 32, 453. Flerkb, B., Mess, G., and Illei-Dunhoffer, A. (1969).Neuroendocrinology 4, 164. Forte, L. R. (1972).Life Sci. 11,Part 1,461. Freifeld, M. L., Feil, P. D., and Bardin, C. W. (1973).Personal communication. Frieden, E. H., and Fishel, S. S. (1968).Biochem. Biophys. Res. Commun. 31, 515. Funder, J. W., Feldman, D., and Edelman, I. S. (1972).J . Steroid Biochem. 3, 201. Funder, J. W., Feldman, D., and Edelman, I. S. (1973a).Endocrinology 92,994. Funder, J. W., Feldman, D., and Edelman, I. S. (1973b).Endocrinology 92, 1005. Funder, J. W., Duval, D., and Meyer, P. (1973~). Endocrinology 93,1300. Funder, J. W., Feldman, D., Highland, E., and Edelman, I. S. (1974).Biochem. Pharmacol. 23, 1493. Gailani, S . , Minowada, J., Silvemail, P., Nussbaum, A., Kaiser, N., Rosen, F., and Shimaoka, K. (1973).Cancer Res. 33,2653. Gala, R. R., and Westphal, U. (1967).Acta Endocrinol. (Copenhagen)55,47. Gardner, D. G., and Wittliff, J. L. (1973).Biochemistry 12,3090. Gardner, R. S . , and Tomkins, G. M. (1969).J . Biol. Chem. 244,4761. Gehring, U.,Tomkins, G. M., and Ohno, S. (1971).Nature (London),New Biol. 232, 106. Gelehrter, T. D., and Tomkins, G. M. (1967).J . M o l . Biol. 29,59. Geller, J., and Worthman, C. (1973).Acta Endocrinol. (Copenhagen) Suppl. 177, 4. Geynet, C.,Millet, C., Truong, H., and Baulieu, E. E. (1972).Gynecol. Znoest. 3, 2. Giannopoulos, G. (1973a).J . Biol. Chem. 248, 1004. Giannopoulos, G. (197313).J . Biol. Chem. 248,3876. Giannopoulos, G. (1973~). Biochem. Biophys. Res. Commun. 54,600. Giannopoulos, G., and Gorski, J. (1971a).J . Biol. Chem. 246,2524. Giannopoulos, G., and Gorski, J. (1971b).J . Biol. Chem. 246,2530. Giannopoulos, G., Hassan, Z., and Solomon, S. (1973a).J . Biol. Chem. 248, 5016. Giannopoulos, G., Mulay, S., and Solomon, S. (1973b).J . Biol. Chem. 248,5016. Gilbert, L. I., and King, D. S. (1973).Physiol. Znsecta 1,250. Giorgi, E. P., Stewart, J. C., Grant, J . K., and Scott, R. (1971).Biochem. J . 123, 41. Giorgi, E. P.,Stewart, J. C., Grant, J. K., and Shirley, I. M. (1972).Biochem. J . 126, 107. Gittinger, J. W., and Lasnitzki, I. (1972).J . Endocrinol. 52,459. Glasser, S. R., Chytil, F., and Spelsberg, T. C. (1972).Biochem. J . 130,947. Gloyna, R. E., Siiteri. P. K., and Wilson, J. D. (1970). J . Clin. Znoest. 49, 1746. Goldberg, M. L., and Atchley, W. A. (1966).Proc. Nut. Acad. Sci. U. S . 55, 989. Goldstein, J. L., and Wilson, J. D. (1972).J . Clin. Znoest. 51, 1647. Gonzalez-Diddi, M., Komisaruk, B., and Beyer. C. (1972).Endocrinology 91, 1130. Gopalakrishnan, T. V.,and Sadgopal, A. (1972).Biochim. Biophys. Acta 287, 164. Gordon, A. S., Zanjani, E. D., Levere, R. D., and Kappas, A. (1970).Proc. Nut. Acad. Sci. U.S . 65,919. Gorell, T. A., DeSombre, E. R., and Jensen, E. V. (1974).Fed. Proc., Fed. Amer. SOC. Erp. Biol. 33,1511.

162

SHUTSUNG LIAO

Gorski, J. (1964).J . B i d . Chem. 239,889. Gorski, J,, Noteboom, W. D., and Nicolette, J. A. (1965). J . Cell. Comp. Physiol., 66, Suppl. 1,91. Gorski, J., Toft, D. O., Shyamala, G., Smith, D., and Notides, A. (1968). Recent Progr. Horm. Res. 24,45. Gottfried, H. (1972).J . Theor. Biol. 37,447. Griffin, D. M., and Szego, C. M. (1968). Life Sci. 7, 1017. Grosser, B. I., Stevens, W., and Reed, D. J. (1973). Brain Res. 57,387. Hackney, J. F., and Pratt, W. B. (1971). Biochemistty 10,3002. Hahnel, R.,and Twaddle, E. (1973). Cancer Res. 33,559. Hamilton, T. (1968). Science 161,649. Hamilton, T., Widnell, C. C., and Tata, J. R. (1965). Biochim. Biophys. Acta 108, 168. Hansson, V., and Djoseland, 0. (1972). Acta Endocrinol. (Copenhagen)71,614. Hansson, V., and Tveter, K. J. (1971). Acta Endocrinol. 68,69. Hansson, V., Tveter, K. J., Attramadal, A., and Torgersen, 0. (1971).Acta Endocrinol. (Copenhagen)68,79. Hansson, V., Tveter, K. J., Unhjem, O., and Djoseland, 0. (1972).J . Steroid Biochem. 3,427.

Hansson, V., Reusch, E., RitzCn, E.M., French, F. S . , Trygstad, O., and Torgersen, 0. (1973). Personal communication. Harding. B. W., and Samuels, L. T. (1962). Endocrinology 70, 109. Hams, G . (1971). Nature (London),New Biol. 231,246. Hams, G. S., Guandalini, S., Mehta, R.,and Nandi, S. (1971). Endocrine Soc., Program 53rd Meet., San Francisco p. 327. Harrison, R. W., and Toft, D. 0. (1973). Biochem. Biophys. Res. Commun. 55, 857. Haussler, M. R., Boyce, D. W., Littledike, E. T., and Rasmussen, H. (1971). Proc. Nat. Acad. Sci. U.S . 68, 177. Hechter, O., and Soifer, D. (1971). In “Basic Actions of Sex Steroids on Target Organs” (P. 0. Hubinont, F. Leroy, and P. Galand, eds.), p. 93. Academic Press, New York. Hechter, O., Yoshinaga, K., Halkerston, I. D. K., and Birchall, K. (1967). Arch. Biochem. Biophys. 122,449. Herman, T. S., Fimognair, G. M., and Edelman, I. S. (1968).J . Biol. Chem. 243,3849. Heuson, J. C. (1973). Eur. J . Cancer 9,379. Higgins, S . J., Rousseau, G. G., Baxter, J. D., and Tomkins, G. M. (1973a).J . Biol. Chem. 248,5866. Higgins, S . J., Rousseau, G. G., Baxter, J. D., and Tomkins, G. M. (1973b). J . Biol. Chem. 248,5873. Hilf, R., Wittliff, J. L., Rector, W. D., Savlov, E. D., Hall, T. C., and Orlando, R. A. (1973). Cancer Res. 33,2054. Hollander, N., and Chiu, Y. W. (1966). Biochem. Biophys. Res. Commun. 25, 291. Hsueh, A. J. W., Peck, E. J., Jr., and Clark, J. H. (1973). J . Endocrinol. 58, 503. Huggins, C., and Yang, N. C. (1962). Science 137,257. Huggins, C., Jensen, E.V., and Cleveland, A. S. (1954).J . E r p . Med. 100,225. Ichii, S . , Iwanaga, Y.,and Ikeda, A. (1973). Endocrinol. l a p . u),33. Illingworth, D. V., Heap, R. B., and Perry, J. S. (1970).J . Endocrinol. 48,409. Ishii, D. N., Pratt, W. B., and Aronow, L. (1972). Biochemistty 11,3896. Ito, T., and Horton, R. (1970).J . Clin. Endocrinol. Metab. 31,362. Jacob, S. T., Sajdel, E. M., and Munro, H. N. (1969). Eur. J . Biochem. 7,449. Jensen, E. V., and DeSombre, E. R. (1972). Annu. Reo. Biochem. 41,203.

CELLULAR RECEPTORS FOR STEROID HORMONES

163

Jensen, E. V., and DeSombre, E. R. (1973).Science 182,126. Jensen, E. V., and Jacobson, H. I. (1960).In “Biological Activities of Steroids in Relation to Cancer” (G. Pincus and E. P. Vollmer, eds.), p. 161.Academic Press, New York. Jensen, E. V., and Jacobson, H. I. (1962).Recent Progr. Horn. Res. 18,387. Jensen, E. V., DeSombre, E. R.,and Jungblut, P. W. (1967a).In “Endogenous Factors Influencing Host-Tumor Balance” (R.W. Wissler, J. L. Dao, and S . Wood, eds.), p. 15. Univ. of Chicago Press, Chicago, Illinois. Jensen, E. V., DeSombre, E. R., Hurst, D. J., Kawashima, T., and Jungblut, P. W. (196713).Arch. Anat. Microsc. Morphol. E r p . 56,547. Jensen, E. V., Suzuki, T., Kawashima, T., Stumpf, W. E., Jungblut, P. W., and DeSombre, E. R. (1968).Proc. Nut. Acad. Sci. U. S . 59,632. Jensen, E.V., Numata, M., Smith, S., Suzuki, T., Brecher, P. I., and DeSombre, E. R. (1969).Deoelop. Biol. 3, 151. Jensen, E. V., Block, G. E., Smith, S., Kyser, K., and DeSombre, E. R. (1971a).Nat. Cancer Inst., Monogr. 34,55. Jensen, E. V., Numata, M., Brecher, P. I., and DeSombre, E. R. (1971b).In “The Biochemistry of Steroid Hormone Action” (R. M. S. Smellie, ed.), p. 133. Academic Press, New York. Jensen, E. V., Jacobson, H. I., Smith, S., Jungblut, P. W., and DeSombre, E. R. (1972a).Gynecol. Inoest. 3, 108. Jensen, E. V., Mohla, S., Gorell, T., Tanaka, S., and DeSombre, E. R. (1972b).J. Steroid Biochem. 3,455. Jensen, E. V., Block, G. E., Smith, S., and DeSombre, E. R. (1973).In “Breast Cancer: A Challenging Problem” (M. L. Griem, E. V. Jensen, J. E. Ultmann, and R. W. Wissler, eds.), p. 55.Springer-Verlag, Berlin and New York. Johnson, J. A., Davis, J. O., Baumber, J. S., and Schneider, E. G. (1970).Amer. J. Physiol. 219,1691. Jost, J. P., Keller, R.,and Dierles-Ventling, C. (1973).J. Biol. Chem. 248,5262. Jouan, P., Samperez, S., Thieulant, M. L., and Mercier, L. (1971).J. Steroid Biochem. 2,223. Jouan, P., Samperez. S., and Thieulant, M. L. (1973).J . Steroid Biochem. 4,65. Jung, I., and Baulieu, E. E. (1972).Nature (London),New Biol. 237,24. Jungblut, P. W., Hatzel, I., DeSombre, E. R.,and Jensen, E. V. (1967).Colloq. Ges. Physiol. Chem. 18,58. Jungblut, P. W., Hughes, S. F., Gorlich, L., Gowers, U., and Wagner, R. K. (1971). Hoppe-Seyler’s 2.Physiol. Chem. 352, 1603. Jungmann, R. A., Hiestand, P. C., and Schweppe, J. S. (1974).Endocrinology 94, 168. Kahwanago, I., Heinrichs, W. L., and Herrmann, W. L. (1970).Endocrinology 86,1319. Kaiser, N . , Milholland, R. J., Tumell, R. W., and Rosen, F. (1972).Biochem. Biophys. Res. Commun. 49,516. Kaiser, N., Milholland, R. J., and Rosen, F. (1973).J. Biol. Chem. 248,478. Kalimi, M., Beato, M., and Feigelson, P. (1973).Biochemistry 12,3365. Kappas, A., and Granick, S. (1968).J. Biol. Chem. 243,346. Karlson, P. (1963).Perspect. Biol. Med. 6,203. Karlson, P., and Sekeris, C. E. (1973).Deut. Med. Wochenschr. 98,831. Kasuya, Y., and Wolff, M. E. (1973). J. Med. Chem. 16,832. Kato, J., and Onouchi, T. (1973).Endocrinol. / u p . 20,429. Kato, J., and Villee, C. A. (1967).Endocrinology 80,567. Kato, J., Atsumi, Y., and Inaba, M. (1974).Endocrinology 94,309.

164

SHUTSUNG LIAO

Katzenellenbogen, B. S., and Gorski, J. (1972).J. Biol. Chem. 247, 1299. Kerkay, J., and Westphal, U. (1968).Biochim. Biophys. Acta 170,324. Kidson, C., Cohen, P., and Chin, R . 4 . (1971).In “The Sex Steroids, Molecular Mechanisms” (K. W. McKems, ed.), p. 421.Appleton, New York. Kim, K. H., and Cohen, P. P. (1966).Proc. Nut. Acad. Sci. U.S . 55, 1251. King, R. J. B., and Gordon, J. (1972).Nature (London), New Biol. 240, 185. King, R. J. B., and Thompson, J. (1974).Adoan. E x p . Med. Biol. In press. King, R. J. B., Beard, V., Gordon, J., Pooley, A. S., Smith, J. A., Steggles, A. W., and Vertes, M. (1971).Adoan. Biosci. 7,21. Kirkpatrick, A. F., Kaiser, N., Milholland, R. J., and Rosen, F. (1972).J . Biol. Chem. 247,70. Kniewald, Z . , Massa, R.,and Martini, L. (1969). Excerpta Med., Sect. 3 23,59. Kniewald, Z., Massa, R., and Martini, L. (1971).In “Hormonal Steroids” (V. H. T. James and L. Martini, eds.), p. 784.Excerpta Medica, Amsterdam. Koblinsky, M., Beato, M., Kalimi, M., and Feigelson, P. (1972).J . B i d . Chem. 247, 7897. Kontula, K., Jiinne, O., Janne, J., and Vihko, R. (1972).Biochem. Biophys. Res. Commun. 47,596. Kontula, L., Janne, O., Luukkainen, T., and Vihko, R. (1973).Biochim. Biophys. Acta 328,145. Korenman, S. G. (1970).Endocrinology 87,1119. Korenman, S. G., and Rao, B. R. (1968).Proc. Nat. Acad. Sci. U.S . 61,1028. Korenman, S. G., Sanborn, B. M., and Bhalla, R. C. (1973).Adoan. E x p . Med. Biol. 36, 241. Kumar, A., and Warner, J. R. (1972). J . MoZ. B i d . 63,233. Kwan, S. W., and Brawerman, G. (1972).Proc. Nat. Acad. Sci. U . S . 69,3247. Lahav, M., Dietz, T., and Edelman, I. S. (1973).Endocrinology 92, 1685. Laumas, K. R.,Uniyal, J. P., Krishnan, A. R., Murugesan, K., and Koshti, G. S. (1970). In “Research on Steroids” (C. Conti, ed.), Vol. IV, p. 145.Pergamon-Vieweg, Brunswick, Germany. Lea, 0. A. (1973a).Biochim. Biophys. Acta 317,351. Lea, 0 .A. (1973b).Biochim. Biophys. Acta 322,68. Leavitt, W. W., and Blaha, G. C. (1972).Steroids 19,263. Leavitt, W. W., Kimmel, G. L., and Friend, J. P. (1973).Endocrinology 92,94. Lee, C., and Jacobson, H. I. (1971).Endocrinology 88,596. Liang, T., and Liao, S. (1972).Biochim. Biophys. Acta 277,590. Liang, T., and Liao, S. (1974a). J . Biol. Chem. 249,4671. Liang, T., and Liao, S. (1974b).Proc. Int. Congr. Horm. Steroid., 4 t h Mexico (in press). Liao, S. (1968).Amer. 2001. 8,233. Liao, S. (1974).In “Biochemistry of Hormones” (H. V. Rickenberg, ed.), p. 153.Med. Tech. Publ. Co., Oxford, England. Liao, S., and Fang, S. (1969).Vitam. Horm. (New York) 27, 17. Liao, S., and Fang, S. (1970).In “Some Aspects of Aetiology and Biochemistry of Prostate Cancer” (K. Griffiths and C. G. Pierrepoint, eds.), p. 105.Alpha Omega Alpha Publ., Cardiff. Liao, S., and Liang, T. (1974).In “Hormones and Cancer” (K. W. McKerns, ed.), p. 229. Academic Press, New York. Liao, S., and Lin, A. H. (1967).Proc. Nat. Acad. Sci. U.S . 57,379. Liao, S.,and Stumpf, W. E. (1968).Endocrinology 83,629. Liao, S., and Williams-Ashman, H. G. (1962).Proc. Nat. Acad. Sci. U. S . 48, 1956.

CELLULAR RECEPTORS FOR STEROID HORMONES

165

Liao, S., Leinninger, K. R., Sagher, D., and Barton, R. W. (1965). Endocrinology 77, 763. Liao, S., Barton, R. W., and Lin, A. H. (1966).Proc. Nat. Acad. Sci. U.S . 55, 1593. Liao, S., Sagher, D., and Fang, S. (1968). Nature (London) 220, 1336. Liao, S., Lin, A. H., and Tymoczko, J. L. (1971a). Biochim. Biophys. Acta 230, 535. Liao, S., Tymoczko, J. L., Liang, T., Anderson, K. M., and Fang, S. (1971b). Adoan. Biosci. 7 , 155. Liao, S., Liang, T., and Tymoczko, J. L. (1972).j . Steroid Biochem. 3,401. Liao, S., Liang, T., and Tymoczko, J. L. (1973a).Nature (London),New B i d . 241, 211. Liao, S., Liang, T., Shao, T. C., and Tymoczko, J. L. (1973b). Adoan. E x p . Med. Biol. 36, 232. Liao, S . , Liang, T., Fang, S., Castaneda, E., and Shao, T.-C. (1973~). j . Biol. Chem. 248, 6154. Liao, S . , Howell, D. K., and Chang, T. M. (1974a).Endocrinology 94, 1205. Liao, S., Fang, S., Tymoczko, J. L., and Liang, T. (1974b). In “Structure and Function of Male Sex Accessory Organs” (D. Brandes, ed.). Academic Press, New York (in press). Libby, P. R. (1972a).Fed. Proc., Fed. Amer. SOC. E x p . Biol. 31, 294. Libby, P. R. (1972b). Biochem. j . 130,663. Liew, C. C., Liu, D. K., and Gornall, A. G. (1972).Endocrinology 90,488. Liew, C. C., Suria, D., and Gornall, A. G. (1973). Endocrinology 93,1025. Lifschitz, M. D., Schrier, R. W., and Edelman, I. S. (1973).Amer. J. Physiol. 224,376. Lippman, M. E., Halterman, R. H., Leventhal, B. G., Perry, S. and Thompson, E. B. (1973).J . Clin. Znoest. 22, 1715. Lisk, R. D., Ciaccio, L.A., and Reuter, L. A. (1972).Gen. Comp. Endocrinol., Suppl. 3, 553. Little, M., Rosenfeld, G. C., and Jungblut, P. W. (1972). Hoppe-Seyler’s Z . Physiol. Chem. 353,231. Litwack, G., ed. (1970). “Biochemical Actions of Hormones,” Vol. 1. Academic Press, New York. Litwack, G., ed. (1972). “Biochemical Actions of Hormones,” Vol. 2. Academic Press, New York. Litwack, G., Ketterer, B., and Arias, I. W. (1971). Nature (London) 234,466. Litwack, G., Morey, K. S., and Ketterer, B. (1972). In “Effects of Drugs on Cellular Control Mechanisms,” (B. R. Rabin and R. B. Freedman, eds.), pp. 105-130. Macmillan, New York. Litwack, G., Fillert, R., Rosenfield, S. A., Lichtash, N., Wishman, C. A., and Singer, S. (1973).j . Biol. Chem. 248,7481. Luck, D., and Hamilton, T. H. (1972). Proc. Nat. Acad. Sci. U . S . 69, 157. Lukacs, I., and Sekeris, C. E. (1967). Biochim. Biophys. Acta 134,85. Maass, H., Engel, B., Hohmeister, H., Lehmann, F., and Trams, C . (1972). Amer. j . Obstet. Cynecol. 113,377. McEwen, B. S., and Wallach, G. (1973). Brain Res. 57,373. McEwen, B. S., Pfaff, D. W., and Zigmond, R. E. (1970). Brain Res. 21,29. McEwen, B. S., Magnus, C., and Wallach, C . (1972). Endocrinology 90,217. McGuire, J. L., and Bariso, C. D. (1972).Endocrinology 90,496. McGuire, J . L., and DeDella, C. E. (1971). Endocrinology 88, 1099. McGuire, J. L., and Lisk, R. D. (1971).In “The Sex Steroids, Molecular Mechanisms” (K. W. McKerns, ed.), p. 53. Appleton, New York.

166

SHUTSUNG LIAO

McGuire, J. L., Bariso, C. D., and Shroff, A. P. (1974).Biochemistry 13,319. McGuire, W. L. (1973).J . Clin. Znoest. 52,73. McGuire, W. L., and Chamness, G. C. (1973). In “Receptors for Reproductive Hormones” (B. W. O’Malley and A. R. Means, eds.), p. 113. Plenum, New York. McGuire, W. L., and Delagarza, M. (1973).J . Clin. Endocrinol. Metab. 36,548. McGuire, W. L., and Julian, J. A. (1971).Cancer Res. 31, 1440. McGuire, W. L., Julian, J. A., and Chamness, G . C. (1971).Endocrinology 89, 969. McGuire, W. L., Huff, K., and Chamness, G. C. (1972).Biochemistry 11,4562. McKems, K. M., ed. (1971). “The Sex Steroids, Molecular Mechanisms.” Appleton, New York. McKems, K. M., ed. (1974).“Hormones and Cancer.” Academic Press, New York. MacLaughlin, D. T., Harding, G. B., and Westphal, U. (1972).Amer. J . Anat. 135,179. Mahoudeau, J. A., and Corvol, P. (1973).Endocrinology 92, 1113. Mahoudeau, J. A., Corvol, P., and Bricaire, H. (1973). Endocrinology 92, 1120. Mainwaring, W. I. P. (1969a).J. Endocrinol. 44,323. Mainwaring, W. I. P. (196913).J . Endocrinol. 45,531. Mainwaring, W. I. P. (1970).In “Some Aspects of Aetiology and Biochemistry of Prostate Cancer” (K. Griffiths and C. G. Pierrepoint, eds.), p. 109. Alpha Omega Alpha Publ., Cardiff. Mainwaring, W. I. P., and Irving, R. (1973).Biochem. J . 134,113. Mainwaring, W. I. P., and Mangan, F. R. (1971).Adoan. Biosci. 7, 165. Mainwaring, W. I. P., and Mangan, F. R. (1973).J . Endocrinol. 59, 121. Mainwaring, W. I. P., and Milroy, E. J. G . (1973).J . Endocrinol. 57,371. Mainwaring, W. I. P., and Peterken, B. M. (1971).Biochem. J . 125,285. Major, P. W., and Kilpatrick, R. (1972).J . Endocrinol. 52,593. Mangan, F. R., and Mainwaring, W. I. P. (1972).Steroids 20,331. Mangan, F. R., Pegg, A. E., and Mainwaring, W. I. P. (1973).Biochem. J. 134, 129. Mann, T. (1964). “The Biochemistry of Semen and of the Male Reproductive Tract.” Methuen, London. Mann, T., Rowson, L. E. A., Baronos, S., and Karagiannidis, A. (1971).J. Endocrinol. 51, 707. Martin, L., and Baggett, B. (1964a).J. Endocrinol. 30,41. Martin, L., and Baggett, B. (1964b).J. Endocrinol. 30,337. Marver, D., Goodman, D., and Edelman, I. S. (1972).Kidney Znt. 1,210. Maurer, R., and Woolley, D. (1971).Endocrinology 88, 1281. Mercier-Bodard, C., Alfsen, A,, and Baulieu, E. E. (1970). Karolinska Symp. Res. Methods Reprod. Endocrinol., 2nd Symp.; Steroid Assay Protein Binding, p. 204. Mester, J,, and Baulieu, E. E. (1972).Biochim. Biophys. Acta 261,236. Milgrom, E., and Baulieu, E. E. (1970).Endocrinology 87,276. Milgrom, E., Atger, M., and Baulieu, E. (1970).Steroids 16,741. Milgrom, E., Perrot, M., Atger, M., and Baulieu, E.-E. (1972a).Endocrinology 90, 1064. Milgrom, E.,Atger, M., Perrot, M., and Baulieu, E.-E. (1972b).Endocrinology 90, 1071. Milgrom, E., Allouch, P., Atger, M., and Baulieu, E.-E. (1973a).J . Biol. Chem. 248, 1106. Milgrom, E., Thi, L., Atger, M., and Baulieu, E.-E. (1973b).J . Biol. Chem. 248, 6366. Milgrom, E., Atger, M., and Baulieu, E. E. (1973~). Biochemistry 12,5198. Milgrom, E., Atger, M., and Baulieu, E. E. (1973d).Biochim. Biophys. Acta 320, 267. Mobbs, B. G., Johnson, I. E., and Connolly, J. G. (1974).Urology 3, 105.

CELLULAR RECEPTORS FOR STEROID HORMONES

167

Mohla, S., DeSombre, E. R., and Jensen, E. V. (1972).Biochem. Biophys. Res. Commun. 46,661. Moore, R. J., and Wilson, J. D. (1972).J. Biol. Chem. 247,958. Morey, K. S., and Litwack, G. (1969).Biochemistry 8,4813. Mueller, G. C., Herranen, A., and Jervell, K. (1958).Recent Progr. Horm. Res. 8,95. Mueller, G. C., Gorski, J., and Aizawa, Y. (1961). Proc. Nut. Acad. Sci. U . S .

47,164. Munck, A. (1968). I n “Recent Advances in Endocrinology” (V. H. T. James, ed.). 8th Ed., p. 139.Churchill, London. Munck, A., and Brinck-Johnsen, T. (1968).J . Biol. Chem. 243,5556. Munck, A.,and Wira, C. (1971). Adoan. Biosci. 7,301. Munck, A,, Wira, C., Young, D. A., Mosher, K. M., Hallaham, C., and Bell, P. A. (1972). J . Steroid Biochem. 3,567. Neumann, F., Hahn, J. D., and Kramer, M. (1967). Acta Endocrinol. (Copenhagen) 54,

227.

Nicolette, J. A., and Mueller, G. C. (1966).Biochem. Biophys. Res. Commun. 24,851. Niu, M. C., and Segal, S . J., eds. (1974).“The Role of RNA in Reproduction.” NorthHolland Publ., Amsterdam. Norton, D. A. (1965).Bi0phys.J. 5,425. Noteboom, W.D., and Gorski, J. (1965).Arch. Biochem. Biophys. 111,559. Notides, A. C.(1970).Endocrinology 87,987. Notides, A. C., and Gorski, J. (1966).Proc. Nut. Acad. Sci. U . S . 56,230. Ofner, P. (1968).Vitam. Horm. (New York) 26,271. Ohno, S . (1971).Nature (London) 234, 134. O’Malley, B. W., and Toft, D. 0. (1971).J. Biol. Chem. 246, 1117. O’Malley, B. W., and Means, A. R., eds. (1973).“Receptors for Reproductive Hormones.” Vol. 36. Plenum, New York. O’Malley, B. W., McGuire, W. L.,Kohler, P. O., and Korenman, S. G. (1969).Recent Progr. Horm. Res. 25, 105. O’Malley, B. W., Sherman, M. R., Toft, D. 0.. Spelsberg, T. C., Schrader, W. T., and Steggles, A. W. (1971). Adoan. Biosci. 7,213. O’Malley, B. W., Spelsberg, T. C., Schrader, W. T., Chytil, F., and Steggles, A. W. (1972).Nature (London) 235, 141. O’Malley, B. W., Schrader, W. T., and Spelsberg, T. C. (1973).Adoan. E x p . Med. Biol.

36, 174.

Ozon, R., and Belle, R. (1973).Biochim. Biophys. Acta 297, 155. Palmiter, R. D. (1972). J . Biol. Chem. 247,6450. Palmiter, R. D., and Haines, M. E. (1973).J . Biol. Chem. 248,2107. Palmiter, R. D., and Schimke, R. T. (1973).J. Biol. Chem. 248, 1502. Palmiter, R. D.,Catlin, G. H., and Cox, R. F. (1973).Cell Differentiation 2, 163. Pasqualini, J. R., and Scholler, R., eds. (1972).J . Steroid Biochem. 3. Pasqualini, J. R., Sumida, C., and Gelly, C. (1972).J . Steroid Biochem. 3,543. Pastan, I., and Perlman, R. L. (1970).Science 169, 339. Paul, J. (1971).C u m Top. Deoelop. Biol. 5,317. Payne, A. H., Lawrence, C. C., Foster, D. L., and Jaffe, R. B. (1973).J . Biol. Chem.

248, 1598.

Pearlman, W. H., and Crepy, 0. (1967).J . Biol. Chem. 242, 182. Pearlman, W. H., and Pearlman, M. R. J. (1961).J . Biol. Chem. 236, 1321. Pearlman, W. H., Gubriguian, J. L., and Sawyer, M. E. (1973).J . Biol. Chem. 248,

5736.

Peck, E. J., Jr., Burgner, J., and Clark, J. H. (1973).Biochemistry 12,4596.

168

SHUTSUNG LIAO

Peets, E. A., Henson, M. F., and Neri, R. (1974).Endocrinology 94,532. PBrez-Palacios, G., PBrez, A. E., Cruz, M. L., and Beyer, C. (1973).Biol. Reprod. 8, 395. Pfaff, D. W., Silva, M. T. A., and Weiss, J. M. (1971).Science 172,394. Philibert, D., and Raynaud, J. P. (1973).Steroids 22,89. Pincus, G., Nakao, T., and Tait, J. F., eds. (1966).“Steroid Dynamics.” Academic Press, New York. Pitot, H. C., and Yatvin, M. B. (1973).Physiol. Reo. 53,228. Plapinger, L., and McEwen, B. S. (1973).Endocrinology 93, 1119. Podratz, K. C., and Katzman, P. A. (1968).Fed. Proc., Fed. Amer. SOC. E x p . Biol. 27, 497. Pogo, B. G. T., Pogo, A. 0.. Allfrey, V. G., and Mirsky. A. E. (1968).Proc. Nut. Acad. Sci. U.S . 59, 1337. Porter, G. A., Bogoroch, R., and Edelman, I. S. (1964).Proc. Nut. Acad. Sci. U.S. 52, 1326. Puca, G.A., and Bresciani, F. (1968).Nature (London) 218,967. Puca, G. A., and Bresciani, F. (1969).Nature (London) 223,745. Puca, G . A., Nola, E., Sica, V., and Bresciani, F. (1971).Biochemistry 10,3769. Puca, G . A., Nola, E., Sica, V., and Bresciani, F. (1972).Biochemistry 11, 4157. Rao, B. R.,Wiest, W.G., and Allen, W. M. (1973).Endocrinology 92,1229. Rasp&,G.,ed. (1971).Adoan. Biosci. 7. Raynaud, J . , Mercier-Bodard, C., and Baulieu, E. E. (1971).Steroids 18,767. Raynaud-Jammet, C., and Baulieu, E. E. (1969).C . R. Acad. Sci., Ser. D 268, 3211. Reel, J. R., and Kenney, F. T. (1968).Proc. Nut. Acad. Sci. U.S . 61,200. Rennie, P., and Bruchovsky, N. (1972).J . B i d . Chem. 247, 1546. Rennie, P., and Bruchovsky, N. (1973). J . B i d . Chem. 248,3288. Reti, I., and Erdos, T. (1971).Biochimie 53,435. Reuter, L. A,, Ciaccio, L. A., and Lisk, R. D. (1970).Fed. Proc., Fed. Amer. Soc. EXJJ. Biol. 29,250. Rhoads, R. E., McKnight, G. S.,and Schimke, R. T. (1971).J . Biol. Chem. 246, 7407. Ribarac-Stepic, N., Trajkovic, D., and Kanazir, D. (1973).Steroids 22, 155. Rickenberg, H.V. (1974).“The Biochemistry of Hormones.” Med. Tech. Publ. Co. Buttenvorths, London. Riggs, A. D., Reiness, G., and Zubay, G. (1971).Proc. Nut. Acad. Sci. U.S. 68, 1222. Ritzkn, E. M., Nayfeh, S. N., French, F. S., and Dobbins, M. C. (1971).Endocrinology 89, 143. RitzBn, E. M., Nayfeh, S. N., French, F. S.,and Aronin, P. A. (1972).Endocrinology 91, 116. Robinson, R. G., and Fanestil, D. D. (1970).Acta Endocrinol. (Copenhagen)Suppl. 147, 275. Robison, G . A., Butcher, R. W., and Sutherland, E. W. (1971).“Cyclic AMP.” Academic Press, New York. Rochefort, H., and Baulieu, E. E. (1968).C . R . Acad. Sci., Ser. D 267,662. Rochefort, H., and Baulieu, E. E. (1972).Biochimie 54, 1303. Rosenau, W., Baxter, J. D., Rousseau, G. G., and Tomkins, G. M. (1972).Nature (London),New Biol. 237,20. Rosenfeld, M. G., and O’Malley, B. W. (1970).Science 168,253. Rousseau, G . G.,Baxter, J. D., and Tomkins, G. M. (1972a).J . Mol. Biol. 67, 99. Rousseau, G. G., Baxter, J. D., Funder, J. W., Edelman, I. S., and Tomkins, G . M. (1972b).J . Steroid Btochem. 3,219.

CELLULAR RECEPTORS FOR STEROID HORMONES

169

Rousseau, G. G., Baxter, J. D., Higgins, S . J., and Tornkins, G. M. (1973).J . Mol. Biol. 79,539. Ruh, T. S . , Katzenellenbogen, B. S., Katzenellenbogen, J. A., and Gorski, J. (1973).Endocrinology 92, 125. S a h n , J., Loeser, R. K., Haas, B. M., and Stavely, H. E. (1973).Biochem. Biophys. Res. Commun. 53,202. Sarnarina, 0 . P., Lukanidin, E. M., Molnar, J., and Georgiev, G. P. (1968).J . Mol. Biol. 33,251. Sandberg, A. A., and Slaunwhite, W. R., Jr. (1963).J . Clin. Inoest. 42,51. Sandberg, A. A., Rosenthal, H., Schneider, S. L., and Slaunwhite, W. R., Jr. (1966).In “Steroid Dynamics’.’ (G. Pincus, T. Nakao, and J. F. Tait, eds.), p. 1. Academic Press, New York. Sandberg, A. A., Kirdani, R. Y.,Varkarakis, M. J., and Murphy, G. P. (1973).Steroids 22,259. Sar, M., and Sturnpf, W. E. (1972).Experientia 28, 1364. Sar, M., and Sturnpf, W. E. (1973a).Endocrinology 92,251. Sar, M., and Sturnpf, W. E. (1973b).Science 179, 389. Sar, M., Liao, S., and Sturnpf, W. E. (1970).Endocrinology 86, 1008. Sarff, M.,and Gorski, J. (1971).Biochemistry 10,2557. Schaumburg, B. P. (1970).Biochim. Biophys. Acta 214,520. Schrader, W. T., and O’Malley, B. W. (1972).J . Biol. Chem. 247,51. Schrader, W. T., Toft, D. O., and O’Malley, B. W. (1972).I. Biol. Chem. 247, 2401. Schutz, G.,Beato, M., and Feigelson, P. (1973).Proc. Nut. Acad. Sci. U . S . 70, 1218. Sekeris, C.E., and Schrnid, W. (1973).Biochim. Biophys. Acta 312,549. Seligy, V. L., and Lurquin, P. F. (1973).Nature (London),New Biol. 243,ZO. Seshadri, B.,and Warren, J. C. (1968).Proc. Int. Congr. Endocrinol., 3rd Mexico City p. 11. Shain, S. A., and Axelrod, L. R. (1973).Steroids 21,801. Shao, T. C.,Castafieda, E., Rosenfield, R. L., and Liao, S . (1975).J . Biol. Chem. (in press). Sharma, S . K.,and Talwar, G. P. (1967).Arch. Biochem. Biophys. 122,449. Sharp, G. W. G., and Alberti, K. G. M. (1971).Adoan. Biosci. 7,281. Sharp, G. W. G., and Kornack, C. L. (1971).Biochim. Biophys. Acta 247,66. Sharp, G . W. G., Komack, C. L., and Leaf, A. (1966).J . Clin. Inoest. 45,450. Shelton, K. R., and Allfrey, V. G. (1970). Nature (London)228, 132. Sherman, M.R.,Corvol, P. L., and O’Malley, B. W. (1970). J . B i d . Chem. 245,6085. Shikita, M., and Talalay, P. (1967). J . Biol. Chem. 242,5658. Shimazaki, J., Kurihara, H., Ito, Y., and Shida, K. (1965).GummaJ.Med. Sci. 14, 313. Shyarnala, G.(1971).Nature (London),New Biol. 231,246. Shyarnala, G . (1972).Biochem. Biophys. Res. Commun. 46, 1623. Shyarnala, G.,and Gorski, J. (1969).J . Biol. Chem. 244, 1097. Shyamala, G., and Nandi, S. (1972).Endocrinology 91,861. Sica, V., Nola, E., Parikh, I., Puca, G. A., and Cuatrecasas, P. (1973a).Nature (London), New Biol. 244,36. Sica, V., Parikh, I., Nola, E., Puca, G. A., and Cuatrecasas, P. (1973b). 1. Biol. Chem. 248,6543. Siiteri, P. K., Ashby, R., Schwarz, B., and MacDonald, P. C. (1972).J. Steroid Biochem. 3,459. Siiteri, P. K., Schwarz, B. E., Moriyama, I., Ashby, R., Linkie, D., and MacDonald, P. C.(1973).Adoan. E x p . Med. Biol. 36,97. Simonsson, B. (1972).Steroids 20,23.

170

SHUTSUNG LIAO

Singer, S., and Litwack, G . (1971a).Cancer Res. 31,1364. Singer, S., and Litwack, G. (1971b).Endocrinology 88, 1448. Singhal, R. L., Parulekar, M. R., and Vijayvargia, R. (1971). Biochem. J . 125, 329. Sluyser, M. (1966a).Biochem. Biophys. Res. Commun. 22,336. Sluyser, M. (1966b).J . Mol. Biol. 22,411. Sluyser, M. (1966~). J . Mol. B i d . 19,591. Sluyser, M. (1969).Biochim. Biophys. Acta 182,235. Smellie, R. M. S., ed. (1971). “The Biochemistry of Steroid Hormone Action.” Academic Press, New York. Smith, J. A., Martin, L., King, R. J. B., and Vertes, M. (1970). Biochem. J . 119, 773. Siimjen, D., Kaye, A. M., and Lindner, H. R. (1973a).Deoelop. Biol. 31,409. Siimjen, D., Siimjen, G., King, R. J. B., Kaye, A. M., and Lindner H. R. (1973b). Biochem. J . 136,25. Soifer, D., and Hechter, 0. (1971).Biochim. Biophys. Acta 230,539. Soloff, M. S., Creange, J. E., and Potts, G. 0. (1971).Endocrinology 88,427. Spaeren, U.,Olsnes, S., Brennhovd, I., Efskind, J., and Pihl, A. (1973).Eur.J.Cancer 9, 353. Spelsberg, T. C., Steggles, A. W., and O’Malley, B. W. (1971).J.B i d . Chem. 246,4188. Spirin, A. S. (1969).Eur. J . Biochem. 10,20. Stancel, G. M., Leung, K. M. T., and Gorski, J. (1973a).Biochemistry 12, 2130. Stancel, G . M., Leung, K. M. T., and Gorski, J. (1973b).Biochemistry 12,2137. Steggles, A. W., and King, R. J. B. (1970).Biochem. J. 118,695. Steggles, A. W., Spelsberg, T. C., Glasser, S. R., and O’Malley, B. W. (1971).Proc. Nat. Acad. Sci. U . S . 68, 1479. Stem, J. M. (1972).Cen. Comp. Endocrinol. 18,439. Stem, J. M., and Eisenfeld, A. J. (1969).Science 166,233. Stone, G. M. (1964).Acta Endocrinol. (Copenhagen)41,433. Stone, G . M., and Baggett, B. (1965).Steroids 6,277. Stone, G . M., and Martin, L. (1964).Steroids 3,699. Stumpf, W. E. (1969).Endocrinology 85,31. Stumpf, W. E. (1971).Amer. 2001.11, 725. Sullivan, J . N., and Strott, C. A. (1973).J . B i d . Chem. 218,3202. Sunshine, G . H., Williams, D. J., and Rabin, B. R. (1971).Nature (London),New B i d . 230, 133. Swaneck, G. E., Highland, E., and Edelman, I. S. (1969).Nephron 6,297. Swaneck, G . E., Chu, L. L. H., and Edelman, I. S. (1970).J . Biol. Chem. 245, 5382. Swartz, S. K., Soloff, M. S., and Suriamo, J. R. (1973).Biochim. Biophys. Acta 338,480. Szego, C. M. (1971). In “The Sex Steroids, Molecular Mechanisms” (K. W. McKems, ed.), p. 1. Appleton, New York. Szego, C. M. (1972).Horm. Antagonists Gynecol. Invest. 3,63. Szego, C. M., and Davis, J . S. (1967).Proc. Nut. Acad. Sci. U . S . 58, 1711. Szego, C. M., and Seeler, B. J. (1973).J . Endocrinol. 56,347. Talwar, G . P., Segal, S. J., Evans, A,, and Davidson, 0.W. (1964).Proc. Nut. Acad. Sci. U . S. 52, 1059. Talwar, G. P., Modi, S., and Rao, K. N. (1965).Science 150, 1315. Talwar, G . P., Sopori, M. L., Biswas, D. K., and Segal, S. J. (1968).Biochem. J . 107, 765. Tata, J. R. (1966).Progr. Nucl. Acid Res. Mol. Biol. 5, 191. T a b , J. R. (1968).Nature (London)219,331. Tchemitchin, A. (1973).J . Steroid Biochem. 4,277. Tchemitchin, A., Tchemitchin, X.,and Bongiovanni, A. M. (1973).J . Steroid Biochem. 4,401.

CELLULAR RECEPTORS FOR STEROID HORMONES

171

Teng, C. S., and Hamilton, T. H. (1968). Proc. Nut. Acad. Sci. U.S. 60, 1410. Teng, C. S., and Hamilton, T. H. (1970). Biochem. Biophys. Res. Commun. 40, 1231. Terenius, L. (1973).Eur. J. Cancer 9,291. Thompson, J., and King, R. J. B. (1974). Biochem. SOC. Trans. (in press). Tindall, D. J., French, F. S., and Nayfeh, S. N. (1972). Biochem. Biophys. Res. Commun. 49,1391. Toft, D. (1973). Adoan. Exp. Med. Biol. 36,85. Toft, D., and Chytil, F. (1973). Arch. Biochem. Biophys. 157,464. Toft, D., and Gorski, J. (1966).Proc. Nut. Acad. Sci. U.S . 55, 1574. Toft, D., and O’Malley, B. W. (1972). Endocrinology 90, 1041. Toft, D., Shyamala, G., and Gorski, J. (1967). Proc. Nut. Acad. Sci. U. S . 57, 1740. Tomkins, G. M., Martin, D. W., Jr., Stellwagen, R. H., Baxter, J. D., Mamont, P., and Levinson, B. B. (1970). Cold Spring Harbor Symp. Quant. Biol. 35,635. Tomkins, G. M., Levinson, B. B., Baxter, J. D., and Dethlefsen, L. (1972). Nature (London),New Biol. 239,9. Trachewsky, D., and Cheah, A. M. (1971). Can. 1.Biochem. 49,496. Trachewsky, D., and Lawrence, S. (1972). Proc. SOC. Exp. Biol. Med. 141, 14. Trachewsky, D., Nandi-Majumdar, A. P., and Congote, L. F. (1972). Eur. J. Biochem. 26,543.

Tremblay, R. R., Beitins, I. Z., Kowarski, A., and Migeon, C. J. (1970). Steroids 16,29. Truong, H., and Baulieu, E. E. (1971). Biochim. Biophys. Acta 237, 162. Truong, H., Geynet, C., Millet, C., Soulignac, O., Bucourt, R., Vignau, M., Torelli, V., and Baulieu, E. E. (1973). FEBS (Fed. Eur. Biochem. SOC.), Lett. 35,289. Tsai, H. C., and Norman, A. W. (1973).1. Biol. Chem. 248, 5967. Tsai, H. C., Wong, R. G., and Norman, A. W. (1972).J. Biol. Chem. 247,5728. Ts’o, P. 0. P., and Lu, P. (1964). Proc. Nut. Acad. Sci. U.S . 51, 17. Tuohimaa, P. (1971). In “Basic Actions of Sex Steroids on Target Organs” (M. Hubinot, and F. Leroy, eds.), p. 208. Karger, Basel. Tuohimaa, P., and Johansson, R. (1971). Endocrinology 88, 1159. Tuohimaa, P., and Niemi, M. (1972a).Acta Endocrinol. (Copenhagen)71,37. Tuohimaa, P., and Niemi, M. (1972b).Acta Endocrinol. (Copenhagen)71,45. Tveter, K. J., and Aakvaag, A. (1969). Endocrinology 85,683. Tveter, K. J., and Attramadal, A. (1968). Acta Endocrinol. (Copenhagen)59,218. Tveter, K. J.. and Attramadal, A. (1969). Endocrinology 85,350. Tveter, K. J., and Unhjem, 0. (1969). Endocrinology 84,963. Tveter, K. J., Unhjem, O., Attramadal, A., Aakvaag, A., and Hansson, V. (1971). Aduan. Biosci. 7, 193. Tymoczko, J. L., and Liao, S. (1971). Biochim. Biophys. Acta 252,607. Tymoczko, J. L., and Liao, S. (1974). In preparation. Unhjem, O., Tveter, K. J., and Aakvaag, A. (1969). Acta Endocrinol. (Copenhagen)62, 153.

Van Der Meulen, N., and Sekeris, C. E. (1973). FEBS (Fed. Eur. Biochem. SOC.), Lett. 33, 184.

Van Der Meulen, N., Abraham, A. D., and Sekeris, C. E. (1972). FEBS (Fed. Ertr. Biochem. SOC.),Lett. 25, 116. Verma, U., and Laumas, K. R. (1973). Biochim. Biophys. Acta 317,403. Vermeulen, A., and Verdonck, L. (1970). Karolinska Symp. Res. Methods Reprod. Endocrinol., 2nd Symp.; Steroid Assay Protein Binding p. 239. Vertes, M., Barnea, A., Lindner, H. R., and King, R. J. B. (1973). Aduan. Exp. Med. Biol. 36, 137.

172

SHUTSUNG LIAO

Vonderhaar, B. K., and Mueller, G. C. (1969).Biochim. Biophys. Acta 176,626. Vonderhaar, B. K., Kim, U. H., and Mueller, C. C. (1970).Biochim. Biophys. Acta 208, 517. Wacker, A. (1965).J . Cell. Comp. Physiol. 66, Suppl. 1, 104. Wacker, A., Trager, L., Chandra, P., and Feller, H. (1965a). Biochem. 2.342, 108. Wacker, A., Trager, L., and Chandra, P. (1965b).Natunoissenschaften 52,134. Wade, G . N., and Feder, H. H. (1972).Brain Res. 45,545. Wamer, D. T. (1965).Ann. N. Y. Acad. Sci. 129,605. Watanabe, H., Orth, D. N., and Toft, D. 0. (1974). Biochemistry 13,332. Watenpaugh, K., Don, J., Jensen, L. H., and Furberg, S. (1968). Science 159, 206. Weinberg, R. A. (1973).Annu. Reo. Biochem. 42,329. Weisz, J., and Gunsalus, P. (1973).Endocrinology 93, 1057. Werthamer, S., Samuels, A. J., and A m a d , L. (1973).J . Biol. Chem. 248,6398. Wester, R. C., and Foote, R. H. (1972).Proc. SOC. E r p . Biol. Med. 141,26. Westphal, U. (1970).In “Biochemical Actions of Hormones” (C. Litwack, ed.), Val. 1, p. 209. Academic Press, New York. Westphal, U. (1971).“Steroid-Protein Interactions.” Springer-Verlag, Berlin and New York. Whalen, R. E., and Luttge, W. G. (1971a).Honn. Behau. 2,117. Whalen, R. E., and Luttge, W. C. (1971b).Endocrinology 89,1320. Whalen, R. E., and Luttge, W. G. (1971~). Brain Res. 33, 147. Whelly, S. M., and Barker, K. L. (1974).Biochemistry 13,341. Wichmann, K. (1967).Acta Endocrinol. (Copenhagen),Suppl. 116,l. Wicks, W. D. (1968).J . Biol. Chem. 243,900. Wiest, W. C. (1971). In “The Sex Steroids, Molecular Mechanism” (K. W. McKerns, ed.), p. 295. Appleton, New York. Wiest, W. C., and Rao, B. R. (1971).Aduan. Biosci. 7,251. Williams, C., and Gorski, J. (1972a).Acta Endocrinol. (Copenhagen),Suppl. 168,420. Williams, G., and Gorski, J. (1972b).Proc. Nut. Acad. Sci. U.S . 69,3464. Williams-Ashman, H. C. (1965).Cancer Res. 25,1096. Wilson, J. D. (1973).Endocrinology 92, 1192. Wilson, J. D., and Gloyna, R. E. (1970).Recent Progr. Honn. Res. 26,309. Wilson, J. D., and Goldstein, J. L. (1972).J. Biol. Chem. 247,7342. Wira, C. R., and Baulieu, E. (1972).C. R . Acad. Sci., Ser. D . 274, 73. Wira, C. R., and Munck, A. (1970).J. Biol. Chem. 245,3436. Wittliff, J. L., Cardner, D. G., Battema, W. L., and Gilbert, P. J. (1972). Proc. Znt. Congr. Endocrinol., 4th, Washington, D.C . p. 252. Wolff, M. E., and Zanati, C. (1970).Erperientia 26,1115. Wong, M. D., and Burton, A. F. (1973). Biochem. Biophys. Res. Commun. 50, 71. Wong, R. G., Norman, A. W., Reddy, C. R., and Cobum. J. W. (1972).J.Clin. Znuest. 51, 1287. Yamaguchi, K., Mineseta, T., Kasai, K., Kurachi, K., and Matsumoto, K. (1971).Steroids 17,345. Yu, F. L., and Feigelson, P. (1969).Biochem. Biophys. Res. Commun. 39,499. Yu, F. L., and Feigelson, P. (1971).Proc. Nut. Acad. Sci. U.S . 68,2177. Yu, F. L., and Feigelson, P. (1972).Proc. Nut. Acad. Sci. U.S . 69,2833. Zalokar, M. (1961). In “Control Mechanisms in Cellular Processes” (D. M. Bonner, ed.), p. 87. Ronald Press, New York. Zigmond, R. E., Nottehohm, F., and Phff, D. W. (1973).Science 179, 1005. Zubay, C., Schwartz, D., and Beckwith, J. (1970). Proc. Nut. Acad. Sci. U.S . 66, 104.

A Cell Culture Approach to the Study of Anterior Pituitary Cells A. TIXIER-VIDAL,D. GOURDJI,AND C. TOUGARD Groupe de Neuroendocrinologie Cellulaire, Laboratoire de Physiologie Cellulaire, CollBge de France, Paris, France

I. Introduction . . . . . . . . . . . . . . . . . . . 11. Characteristics of Anterior Pituitary Cells Grown in Vitro. . . . . . . . . . . . . . . . . . . . . A. Hormonal' Secretion . . . . . . . . . . . . . . . B. Morphological Features . . . . . . . . . . . . . 111. Reactivity to Specific Regulating Agents . . . . . . . . . A. Primary Cultures. . . . . . . . . . . . . . . . B. Effects of Regulating Agents on Continuous Cell Lines . IV. Conclusion . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . .

.

173 177 177 184 205 205 218 234 235

I. Introduction The anterior pituitary cells offer fascinating material for the cell biologist. They belong to an endocrine gland, the adenohypophysis, which plays a prominent role in numerous endocrine as well as nonendocrine functions of the organism, and whose activity is closely regulated by substances originating either from the central nervous system or from target endocrine glands. At the level of the whole endocrine gland, several biochemical functions are expressed which correspond to the well-known adenohypophysial hormones. Three of them are polypeptides: corticotropic hormone (ACTH, 39 amino acids), the melanophorotropic hormones (a-MSH, 13amino acids, and P-MSH, 18 amino acids), and lipotropic hormone (PLPH); all three possess important common amino acid sequences. Two are proteins: somatotropic or growth hormone (STH or GH) and mammotropic hormone or prolactin (MTH or PRL), whose primary structure only is known (MW from 20,000 to 25,000) and which possess some common amino acid sequences. Three are glycoproteins (MW -30,000): thyrotropic hormone (TSH) and the two gonadotropic hormones -follicle-stimulating hormone (FSH) and luteinizing hormone (LH). These three glycoprotein hormones share important common features. They have a quaternary structure, and each of them consists of two subunits, a and P. Peptide sequences of these two subunits have been established in several mammalian species. These studies revealed that the a subunits are 173

174

A. TIXIER-VIDAL, D. GOURDJI AND C. TOUGARD

similar for the three hormones, and carry the zoological specificity, while the p subunits have more distinct primary structures, and carry the biological specificity (see review by Jutisz and D e La Llosa,

1972). To this biochemical heterogeneity of function, corresponds the cellular heterogeneity of the glandular tissue. This has been demonstrated by 50 years of morphological research using more and more elaborate techniques (see reviews by Herlant, 1964; Purves, 1966), from light microscopy to electron microscopy, and from tinctorial to cytochemical and immunocytochemical methods. As in many other endocrine cells, the fundamental feature of pituitary cells is the ability to store secretory products within secretory granules. Two types of consequences result from this feature. From a cytological point of view, this allowed workers to distinguish several glandular cell types according to the chemical or immunochemical nature of their secretory products. Finally, this led to the notion that one glandular cell type. corresponds to each pituitary hormone. In some cases, nevertheless, for example, LH and FSH, ACTH, and MSH, the existence of either a common or a dual cell type is still debated. Important physiological features of the pituitary cell are also related to its storage ability. One may assume on the basis of autoradiographic and biochemical studies that pituitary protein hormones are synthesized on the ribosomes and then follow a migratory pathway within the endoplasmic reticulum, through the Golgi zone to the secretory granules ,(see reviews by Farquhar, 1971; Farquhar et al., 1974; Labrie et al., 1973). The hormones are then released out of the cell mainly, if not only, by exocytosis. There is therefore a space and time discontinuity between synthesis and release of the hormone. In addition, an intracellular mechanism allows degradation of undischarged secretory granules, via lysosomes (Farquhar, 1971). This brief survey of the fundamental properties of anterior pituitary cells raises the problem of mechanisms that regulate (1)the rate of their secretory activity, and (2) the biochemical nature of their secretory product, in other words, that control both quantitatively and qualitatively the expression of their genetic differentiation. An impressive amount of physiological and biochemical research deals with regulation of the secretion of anterior pituitary hormone (see review by Kraicer, 1974). Two types of agents play this role: neurohormones and hormones from pituitary target glands, that is, gonads, thyroid, and adrenals. The neurohormones are polypeptides synthesized within the hypothalamus and carried to the anterior pituitary cells via the hypothalamohypophysial portal vessel system,

CELL CULTURE OF ANTERIOR PITUITARY CELLS

175

They have been isolated and chemically characterized on the basis of their specific stimulating effect on the release of one anterior pituitary hormone. Therefore one hypothalamic releasing factor (RF) or hormone (RH) corresponds to each pituitary hormone. The structure of several of these neurohormones has been elucidated and their synthesis effected, for example: MSH-RH, GH-RH, TSH-RH, and LH-RH. Moreover, progress in research led to a discovery at the same time of the existence of a dual hypothalamic control, stimulating and inhibiting, for several pituitary hormones such as MSH, GH, PRL, the inhibiting control prevailing in uivo for PRL. In addition, some synthetic hypothalamic hormones now appear to be less specific in their effects than initially presumed, since TSH-RH (TRH) acts not only on TSH release but also on PRL, and LH-RH (LRH) stimulates the release of both LH and FSH. These findings might question the specificity of hypothalamic hormones as regards anterior pituitary hormones. Sex steroids, corticosteroids, and thyroid hormones (thyroxine and triiodothyronine) form the other class of regulating agents known to act directly at the pituitary cell level in addition to their action through the hypothalamus (see review by Stumpf et al., 1974). Biochemically, they evidently differ from hypothalamic polypeptides. They inhibit the secretion of their corresponding pituitary hormone by a well-known feedback effect. But again the specificity of this action may be questioned since, for example, thyroxine and estradiol have been found to exert a direct stimulating action on the activity of PRL cells (Nicoll and Meites, 1963; Meites et al., 1963). The mechanism that controls differentiation of the anterior pituitary cells has been far less studied than those related to the control of their secretory activity. If it is assumed that all cells of a multicellular organism possess an identical genome, their differentiation within functionally specialized cell populations appears as the result of epigenetic events leading to the expression of one function only. In the case of the anterior pituitary gland, which contains an heterogeneous population of glandular cells, two steps must be distinguished in the course of their differentiation: (1) their being “programmed” as glandular cells able to synthesize and to secrete a group of seven proteins only, and (2) their being specialized within several cell lines, each one being able to synthesize only one of the seven proteins. The first step seems to occur during the first stages of fetal life: the pituitary anlage derived from the Rathke pouch follows its differentiation as adenohypophysial tissue when isolated at an early stage of the nervous infundibular floor, that is, 48 hours in the

176

A. TIXIER-VIDAL, D. GOURDJI AND C. TOUGARD

chick embryo (Ferrand, 1972) and 10 days in the mouse embryo (Ferrand and Nanot, 1968). The mechanisms involved in this first step are still unknown, and cell cultures do not seem to be a useful approach to their study. The second step may be considered during a later stage of fetal life, as well as during adult life. During fetal life electron microscope studies in the chick embryo (Guedenet et al., 1970; Mikami et al., 1973) and in the fetal rat (Yoshida, 1966; Daikoku et aZ., 1973; Dupouy and Magre, 1973) show at the beginning small primordial and agranular cells. Thereafter membrane-bound secretory granules appear progressively within an increasing number of cells. Finally, several cell types may be identified which look like adult pituitary cell types, although with fewer secretory granules. At the same time, some primitive cells still persist, together with follicular cells which seem to constitute an independent stem line (see review by Olivier et aZ., 1974). Immunocytochemical methods allowed GH cells, PRL cells, LH cells, and TSH cells to be observed at different stages in bovine fetus (Dubois, 1971a,b) and in rat fetus (S6td6 and Nakane, 1972). These morphological data agree with physiological studies demonstrating the appearance of several pituitary functions during fetal life (see review by Jost, 1966). Is differentiation of the pituitary cells definitely achieved after birth? In fact, several facts observed in adult pituitary do not fit in with this statement. Although the adenohypophysis is known as an organ of low mitosis (Leblond and Walker, 1956), dividing pituitary cells are seen even in steady-state conditions (Nouet and Kujas, 1973). Besides, when the production of one hormone is triggered by a modification of endocrine homeostasis, rapid changes in the cell population occur: e.g., cellular hypertrophy and hyperplasy of gonadotropic cells after castration, and of thyrotropic cells after thyroidectomy. The mechanism of such an adaptation of the glandular tissue is still unknown, although it has fascinated many pituitary cytologists. In their recent review, Olivier et al. (1974) discuss three main hypothesis which have been proposed and examined by several workers. The first one stipulates the persistence in the adult gland of stem cells, that is, multipotent cells that have retained the ability to divide and to differentiate into one specialized type when the production of one hormone is stimulated. The second hypothesis involves the mitosis of fully differentiated cells which have been observed in some cases and some species. The last hypothesis involves cellular transformation of one cell type to another. On the whole, the present studies do not allow one to decide definitely among these

CELL CULTURE OF ANTERIOR PITUITARY CELLS

177

three hypotheses, mainly because of difficulties in determining if one cell is functionally differentiated by morphological criteria only. Indeed, several features of the adenohypophysis hamper analysis at the cellular level of mechanisms involved in the control of differentiation and of secretory activity of pituitary cells. The main difficulties reside in the cellular heterogeneity of the gland and its anatomical position in one of the less accessible regions of the head. In this respect pituitary cell cultures offer an invaluable approach which has been used increasingly in recent years. In this chapter we concentrate our attention on pituitary cell cultures which allow work on dispersed pituitary cells maintained in uitro for months or years. They differ from pituitary organ cultures in which tissue organization is maintained, and from tissue culture starting from small fragments of tissue which grow and can display selection in some cell populations (see review by Tixier-Vidal, 1974). Among cell cultures we must distinguish (1) primary cultures that start from normal adult fully differentiated anterior pituitary cells, and (2) continuous cell lines that consist of homogeneous populations of pituitary glandular cells which continuously grow in uitro. In Section 11 we report the functional and morphological features of pituitary cells grown in these two situations. In Section 111 the effects on these models of factors that regulate the secretion of anterior pituitary hormones are analyzed and discussed with a view toward answering the main question raised by pituitary cells. 11. Characteristics of Anterior Pituitary Cells Grown in Vitro

A.

HORMONALSECRETION

1. Primary Cultures of Normal Pituitary Cells Primary cultures are initiated from normal anterior pituitary cells (man, monkey, rat) previously dispersed by enzymic treatment according to a large variety of techniques proposed within recent years (Portanova et aZ., 1970; Vale et al., 1972a; Hymer et al., 1973; TixierVidal et aZ., 1973; Hopkins and Farquhar, 1973). The cell suspension is then inoculated into plastic petri dishes or bottles, using an appropriate medium always enriched with adult and fetal serums. Such cultures can be maintained for weeks or months. Young adult pituitaries were generally used, with the exception of human fetal glands (Reusser et al., 1962; Gailani et al., 1970). This means that generally monolayers are started from adult, fully dif-

178

A. TIXIER-VIDAL, D. COURDJI AND C. TOUCARD

ferentiated cells. The major problem to be studied therefore concerns the evolution in culture of the biochemical and morphological parameters of their differentiation. Hormonal secretion has been followed as a function of culture time by measuring the hormonal content of the medium more frequently and the hormonal content of the cells less frequently. The sensitivity of the hormonal assay limits conclusions concerning the maintenance or disappearance of hormonal activity in a culture. In this respect great progress has been obtained as a result of radioimmunoassays, which have been applied to cell cultures only for the last 3 or 4 years. Two major features characterize the evolution in cell culture of the secretion of anterior pituitary hormones.

1. There is an important rise in medium PRL content. In fact such an increase in PRL secretion in uitro was first observed in pituitary tissue cultures (man and rat: Pasteels, 1961, 1963) and organ cultures (rat: Meites et al., 1961, 1963), and then widely confirmed in mammals (see review by Tixier-Vidal, 1974). Our observations (see Table I) show that the same phenomenon also occurs in cell culture. A very high level of medium PRL content is already observed after 5 days of culture (see Table I). This high level is maintained for 3 or 4 weeks, and thereafter a very slight decrease is observed. Since a similar hyperactivity of PRL cells has been observed in uivo after interruption of the hypothalamohypophysial anatomical link (see reviews by Pasteels, 1963; Meites et al., 1963), this cannot be considered an artifact of culture methods and represents an original feature of PRL cells. 2. For the other pituitary hormones, the medium hormonal content decreases with time in culture. This has been found for ACTH in rat (Stark et al., 1965), for GH in human adult (Batzdorf et al., 1971) or fetus (Gailani et al., 1970), for LH and FSH in rat (Vale et al., 1972a; Steinberger e t al., 1973; Tixier-Vidal et al., 1973) and in human fetus (Gailani et al., 1970), and for TSH in rat (Vale et al., 1972a). Depending on the investigator, this decrease leads either to the maintenance of a low level or to a more-or-less precocious disappearance, that is, an undetectable level. In our experiments on LH and FSH secretion, which involved more than 30 cultures, there was in this respect intrinsic variability from one experiment to another. In some cases gonadotropic hormones disappeared after 15 days, but in numerous series they were maintained for 50 days or more. In all cases the FSH/LH ratio increased with culture time,

179

CELL CULTURE O F ANTERIOR PITUITARY CELLS

TABLE I SD 26

AGAINST

Age of monolayer

5 days

14 days

21 days

28 days

35 days

56 days

ANTISERA OVINELHan

SERIES- NUMBER OF CELLS IMMUNOREACTIVE WITH

OVINEFSH, OVINELH, OVINELHP,

Positive cells (a)

FSH LH LHP LHa FSH LH LHP LHa FSH LH LHP LHa FSH LH LHP LHa FSH LH LHP LHa FSH LH LHP LHa

1.7 4.4 <1 <1 0 1 0.7 0 <1 1.8 0.9 <1 <1 1 1.3 2.2 <1 1.6
Medium LH content in same dish (ng/ml)*

27.7 31.2 22.7 24.7 1.6 1.7 1.3 1.5 <0.1
AND

Medium FSH content in same dish (nglmUb

80 85.5 70 20 22 9 22.5 20

-

(1

-

i l

-

8.5 13 3.1 3.5

-

7 6.2 <1

Medium PRL content in same dish (ng/ml)c

3300 3450 2600 2700 2600 3000 3200 3600 6000 4200 7200 3700 3500 2800 2100 2700 2400 2000 2350 1600 1960 1930 2080

" As compared to the medium LH, FSH, and PRL contents in the same culture dish, in monolayers of rat anterior pituitary cells of increasing age. Samples were assayed for rat LH and for rat FSH by B. Kerdelhue and M. Jutisz using radioimmunoassay methods as described in Tixier-Vidal et ~ l (1973). . PRL was measured by radioimmunoassay by B. Kerdelhue and D. Grouselle using rat PRL and corresponding antisera kindly provided by the NIAMDD program.

and FSH activity was consistently retained for a longer time than LH activity. The decrease in medium hormonal content is accompanied by a decrease in tissue hormonal content as shown for LH and TSH in 4-, 9-, 15- and 21-day rat pituitary cell cultures (Vale et al., 1972a). Again, in tissue culture a similar decrease in medium hormonal content has been observed for the same hormones (see review by

180

A. TIXIER-VIDAL, D. GOURDJI AND C. TOUGARD

Tixier-Vidal, 1974). Surprisingly enough, Ishikawa and Nagayama (1973), using a tissue culture from the pituitary of a 25-year-old woman noted almost no decrease in medium LH content within the first 30 days. Organ cultures, although generally used for short-term experiments, also allowed a similar decrease in medium hormonal content to be observed, followed by a constant low level, for example, LH in 2-week rat organ cultures (Tixier-Vidal et al., 1970). Using fetal human pituitary and long-term organ culture (4 or 9 months), Franchimont and Pasteels (1972) also observed, after a rapid decrease, a very low level of LH, FSH, and TSH but a high level of the a subunit, which is common to these three glycoprotein hormones. It is not known whether a similar autonomy of secretion of the a subunit also exists in cultures of adult rat pituitary cells. Results on hormonal secretion in cell culture suggest that pituitary cells can maintain their functional differentiation for several weeks and sometimes months. With the exception of PRL, their secretory level decreases in culture, Whether this decline results from a decrease in the number of differentiated cells or from a decrease in cell secretory ability is examined further by morphological methods (see Section 11,B). 2. Continuous Cell Lines Several continuous anterior pituitary cell lines performing secretory activities in uitro have been obtained from pituitary tumors by Sato and his colleagues, using the technique of alternate animal and cell culture passage (Buonassisi et aZ., 1962). Rat pituitary strains have been established starting from a tumor induced by chronic estrogen treatment in a female rat of the Wistar/Furth strain. This tumor, Mt/W5, is transplantable and secretes three hormones: ACTH, GH, and PRL (Takemoto et al., 1962). The primary culture of the tumor cells was morphologically heterogeneous. Two types of cells were distinguished: epithelial cells, and flat, spindle-shaped cells. These two morphologically different cell types were isolated and gave two clonal lines which secreted either ACTH (flat cell line) or GH (epithelial cell line) (Yasumura et al., 1966; Tashjian et al., 1968).Three clones have been separated from this last line-GH1, G H l 2C1, and GH3-all of which synthesize and secrete GH (Bancroft et al., 1969). Later, these three clonal strains were found to produce a second protein hormone, PRL (Tashjian et al., 1970). Among the various cell strains, the rate of hormonal production, as well as the relative proportions of GH and PRL within the culture medium, differ widely. Some subclones- the

CELL CULTURE OF ANTERIOR PITUITARY CELLS

181

GH1 2C, and two GH3 spinner cells-secrete GH only (Tashjian and Hoyt, 1972). Another, GH4, secretes only PRL (Tashjian et al., 1973). The hormones secreted in culture are biologically and immunologically undistinguishable from rat pituitary hormones. One functional cell line has been isolated from a transplantable mouse pituitary tumor, the At T-20 ACTH-secreting adenocarcinoma (Cohen and Furth, 1959; Buonassisi et al., 1962). It secretes ACTH as well as the subclonal cell line At-20/D, which was established by Yasumura (1968). Further studies on this subclone (Orth et al., 1973) have shown that it produces, in addition to biologically and immunoreactive ACTH, a bioreactive MSH which is immunologically related to p-MSH but not to a-MSH. Several important features distinguish these functional continuous cell lines from monolayers of normal anterior pituitary cells: 1. They divide continuously. For example, when GH3 cells are plated to a low density, after a lag phase of about 36 hours they grow exponentially (population doubling time about 50-60 hours) and then reach an early stationary phase (Bancroft et al., 1969). In contrast, normal anterior pituitary cells have an extremely low division rate as observed after a 6-hour treatment with colchicine followed by fixation and Herlant’s tetrachrome staining. With such a method we found during the first 8 days of culture an increasing number of mitotic figures within typical fibroblasts, but only a very few mitotic figures in glandular cells, mainly in PRL cells (personal unpublished observation). After a short pulse of thymidine-3H, Rappay et al. (1973) found, in 5- and 6-day monolayers, that 19 and 28%, respectively, of the nuclei were labeled, but they did not determine the exact nature of the cells with labeled nuclei. 2. They continuously secrete one or two specific protein hormones, while cultured normal pituitary cells secrete at a high rate, although not indefinitely, only PRL. It has been shown for GH3 cells that hormones are secreted during the exponential as well as the early stationary phases of growth (Bancroft et ul., 1969; Tashjian et al., 1970). The secretion rate increases linearly as a function of time in culture. In contrast, the intracellular hormonal content is very low. If it is assumed that the rate of production is constant during the collection period, the intracellular level of hormone is equal to the amount produced in about 15-30 minutes for GH (Bancroft et al., 1969), and in about 1-2 hours for PRL (Tashjian et al., 1970). This low concentration of intracellular hormone may reflect rapid secretion into the medium, or a very high rate of intracellular degradation.

182

A. TIXIER-VIDAL, D. GOURDJI AND C. TOUGARD

Although the second eventuality cannot be conclusively discarded, the first one is probably predominate. Cell capacity for hormonal synthesis has been directly observed by studying the incorporation of labeled amino acid into both GH and PRL. According to Bancroft (1973), labeled GH is secreted 15 minutes after a 12-minute exposure of labeled amino acids in GH3 cells, The newly synthesized hormone represents 8% of the total labeled cell protein. Tashjian and colleagues (Tashjian et al., 1973; Tashjian and Hoyt, 1972; Dannies and Tashjian, 1973) cultured for increasing periods of time GH3 cells in the presence of leucine-". They found that the accumulation of labeled PRL in the cells was linear for at least 1 hour, and that labeled PRL was released into the medium after 1 hour. In our laboratory we examined the turnover of PRL in a strain of rat PRL cells we called SD1 (Tixier-Vidal et al., unpublished observation). This strain appeared spontaneously in a long-term primary culture of normal Sprague-Dawley rat pituitary cells. It has been maintained in our laboratory for 2 years. Since we also have GH3 cells in culture, the possibility of contamination by this strain has not yet been completely excluded. Their karyotype is aneuploid (modal number, 66 k 2; limits, 58-70) and not very different from that of GH3 cells. They produce tumors in both SpragueDawley and WistarlFurth female rats. They produce a high level of PRL and a very low level of GH. Their ability to incorporate leucine3H into PRL has been examined by culturing them for 5 hours in the presence of l e ~ c i n e - ~and H comparing the results to those from GH3 cells. Labeled PRL was separated by gel electrophoresis in the presence of a rat PRL reference preparation kindly provided by the National Institute of Arthritis, Metabolism and Digestive diseases (rat pituitary hormone program, Bethesda, USA). We found that medium PRL radioactivity was lower for SD1 cells than for GH3 cells, although SD1 cells secreted at the same time more PRL (measured by radioimmunoassay) than GH3 cells (Gourdji et al., 1973a). This suggested that these cells may have, relative to GH3 cells a lower turnover rate, that is, a greater ability to store PRL. This interpretation has been checked by further studies involving chase experiments (in the presence of 5 mM L-leucine) after a short pulse (10 minutes) of l e ~ c i n e - ~ (33 H Cilmmoles, 250 pCilml). The results showed that labeled PRL appeared in the medium as soon as 15 minutes after 0 time of chase. The evolution as a function of the chase time (from 0 to 24 hours) of the specific radioactivity of the PRL released into the medium suggests the existence of two PRL in-

CELL CULTURE OF ANTERIOR PITUITARY CELLS

183

tracellular pools having a similar half-life (10 and 12 hours) but differing in size (Fig. 1) (Morin, 1974). 3. The maintenance of differentiated function together with a high rate of cell division cannot be related to a modification of the karyotype, since the ACTH AtT-20 mouse cell line is euploid (Orth et al., 1973) and the six GH rat pituitary cells are strongly aneuploid (modal chromosome number between 69 and 76) (Sonnenschein et al., 1970). 4. One of the main problems raised by the continuous pituitary cell line is the production of two hormones within clones and serial dpm/ng prolactin in medium

CHASE TIME (hours)

FIG.1. Evolution of medium PRL specific radioactivity (disintegrations per nanogram of PRL) as a function of the chase time in the absence ( 0 )or in the presence (x) of 60 nM TRH after a 10-minute pulse with L-leucine-:’H (250 pCi/ml) and without TRH. SDI cells adapted to grow in F 10 medium enriched with 2.5% horse serum and 2.5% fetal calf serum were used in quadruplicate for each time period and for each series. The decline in specific activity of PRL in the control medium can be resolved into two components, A and B. Curve A corresponds to the largest PRL pool with an half-life of 12 hours, and curve B to the smallest pool with an half-life of 10 hours.

184

A. TIXIER-VIDAL, D. GOURDJr AND C. TOUGARD

subclones. This is particularly striking for PRL and GH, which in uiuo are undoubtedly stored within separate cell types (Nakane, 1970). Such a finding strongly suggests that in continuous cell lines both hormones are synthesized and secreted by an individual cell. However, since the production of both hormones by a single cell has not yet been shown, the possibility that, in mass culture, two types of cells can differentiate from a parental cell cannot be excluded (Tashjian and Hoyt, 1972). In addition, the possibility of a secretory cycle with two successive phases, each involving one hormone, can also be considered. The simultaneous secretion of both ACTH and MSH by a single cloned mouse pituitary tumor cell line has also been found, but in that case the contrast with the in uiuo situation is less striking than for GH and PRL. According to Orth et al. (1973),these two hormones originate from a single cell type in man, and recent immunocytochemical studies have found the two hormones within the same cell type in rat (Moriarty, 1973) and in mouse (Naik, 1973). It is perhaps of interest to note that the ACTH mouse cell line is euploid, whereas the GH cell lines are strongly aneuploid. Finally, it must be noted that no continuous cell line producing glycoprotein hormones has been obtained. No LH, FSH, or TSH has been found in culture media of GH3 cells (Tixier-Vidal et al., personal observations). Mouse TSH tumor cells have been cultured (Posner et al., 1973) and “secrete TSH at constant levels for periods of subculture as long as 3 months.” But to our knowledge, these cultures did not produce a continuous strain as compared to the GH and ACTH cell lines. B. MORPHOLOGICALFEATURES 1. Primary Cultures of Normal Pituitary Cells Immediately after their inoculation into culture dishes, the dispersed anterior pituitary cells aggregated in small clumps. They then attached very slowly to the plastic dishes, which took at least 3 days. Attempts to select different cell populations by harvesting and replating the nonattached cells at different intervals after the inoculation were unsuccessful. a. Morphological Heterogeneity of the Monolayer i. Phase-contrast microscopy. When attached, the cell clumps flattened and stretched or grew until about the seventh day. They were made of small round or polyhedric cells which were refractile and had very distinct outlines. After the fifth day another cell population

CELL CULTURE OF ANTERIOR PITUITARY CELLS

185

appeared around the cell clumps. They displayed a fibroblastic aspect, were elongate and transparent (Fig. 2), and grew very quickly. They finally produced a continuous monolayer surrounding the glandular cell clumps, which remained unchanged after the first week. According to Steinberger et al. (1973), the total number of cells increased eight times from 4 to 14 days and then remained constant at least up to 41 days. ii. Specific staining. The cellular heterogenity of the monolayer, as observed in phase-contrast microscopy, is strongly confirmed after fixation and specific staining with Herlant’s tetrachrome (Herlant, 1960) (Fig. 2). The clumps of refractile small cells represent a typical glandular cell islet which contains at the beginning all types of rat anterior pituitary cells. In contrast, the transparent elongated cells have typical features of fibroblasts, bluish cytoplasm and a voluminous, clear nucleus. iii. Electron microscopy. Because of the morphological heterogeneity of the monolayer, a specific method was developed in view of its ultrastructural analysis. After in situ fixation (1% glutaraldehyde in 0.1 M cacodylate buffer) and in situ embedding with Epon ac-

FIG.2. Five-day monolayer of normal rat pituitary cells previously treated 6 hours before fixation by colchicine for 6 hours. Fixation: Bouin Hollande plus 10% mercuric chloride. Staining: Herlant’s tetrachrome. A clump of glandular cells is surrounded by some fibroblasts. Note mitosis (arrows). Green filter. X300.

186

A. TIXIER-VIDAL, D. GOURDJI AND C. TOUGARD

FIG.3. Two-month monolayer of normal rat pituitary cells. In situ fixation and embedding. Horizontal section within a clump of PRL cells. The glandular cells are tightly associated. One notices the extension of the Golgi zone and important variations among cells in their secretory granule content. ~ 4 0 0 0 .

cording to Brinkley et al. (1967), selected zones of the monolayers were cut with a warm punch. The plastic culture dish, which was cut at the same time, was then easily discarded. The remaining small Epon disc contained the monolayer. It was then either mounted on an Araldite block for sectioning in a horizontal plane, or reembedded in Epon for sectioning in a vertical plane (Picart and TixierVidal, 1974). Ultrastructural study confirms that the clumps of small refractile cells are composed of typical glandular cells characterized by the presence of secretory granules and a very well-defined cell membrane. In horizontal as well as vertical sections, the cells appeared associated as in parenchyme (Fig. 3). Specialized junctions between them were, however, never seen, and the presence of basement membrane was seldom observed. The spatial distribution of the cy-

CELL CULTURE OF ANTERIOR PITUITARY CELLS

187

toplasmic organelles, particularly as observed in secretory granule exocytosis, was the same in the two planes, which suggests that within monolayers the glandular cells have no morphological polarity. The exact nature of the glandular cells and their evolution with time in culture is described in Section II,B,b, and c. The other cells of the monolayer are not of a glandular type. They consist mainly of fibroblasts which can be easily identified by their ultrastructural features (Fig. 4). In perpendicular sections these cells are elongated and narrow; they have dark cytoplasm and a flat nucleus which is often lobulated. They form a continuous sheet over the plastic and under the glandular cells. In horizontal sections their cytoplasm contains stacks of cytofilaments and sometimes heterogeneous dense bodies. The cell membrane is ill-defined and has either abundant pinocytic vesicles or fibrillar structures which extend out of the cell. The nucleus appears larger in a horizontal plane and has an assymetric distribution of chromatin and abundant nuclear pores. With increasing time in culture, these fibroblasts become more and more numerous. They can already be seen after 5 days under glandular cells in vertical sections. After 1 month they represent 80-90% of the total cell population as determined by counting within cell pellets obtained from a whole monolayer dispersed by mild trypsinization. Besides the fibroblasts, other nonglandular cells are found within the monolayer. Some of them may be identified as real follicular cells by their tendency to form microvilli and large follicles (Fig. 5). It is known that follicular and stellate cells represent a normal nonglandular component of anterior pituitary tissue (Farquhar, 1957; Farquhar et al., 1974; Vila Porcile, 1973). These cells can survive, and perhaps divide, within monolayers for a very long time. Other voluminous nonglandular cells within the monolayer are characterized by large droplets, probably lipoid in nature (Fig. 4). These cells take up horseradish peroxidase at a very high rate (Fig. 4), a property that has also been found in freshly dispersed follicular cells (Farquhar et al., 1974). They may be modified follicular cells, the phagocytic activity of which is stimulated in culture. In conclusion, the ultrastructural study allowed interpretation of the morphological heterogeneity of the monolayer of normal rat pituitary cells. All the cellular components - glandular and nonglandular-of normal anterior pituitary tissue survive in vitro but, because of differences in their respective mitotic rate in culture, their relative proportions are strongly modified. Moreover, a tissuelike reorganization occurs precociously in culture, which indicates interesting cellular interrelations. The glandular cells recognize each

188

A. TWER-VIDAL, D. COURDJI AND C. TOUGARD

CELL CULTURE OF ANTERIOR PITUITARY CELLS

189

FIG. 5. Same material as in Fig. 4. Follicular cells are organized around a cavity and display abundant microvilli. X6000. FIG. 4. Seven-day monolayer of normal rat pituitary cells. In situ fixation and embedding. The cells were fixed after a 30-minute exposure to horseradish peroxidase (2 mglml). These are two types of nonglandular cells. The fibroblast (Fb) is characterized by abundant fibrillar structures, mainly at the cell periphery. The follicular cell (Fc) is characterized by a strong uptake of horseradish peroxidase as seen in numerous labeled droplets (black) and by abundant lipoid vacuoles. x 12,000.

190

A. TIMER-VIDAL, D. COURDJI AND C. TOUGARD

other and reassociate. At the beginning they seem to need the contact of fibroblasts for their attachment and survival in culture. Later, the overgrowth of fibroblasts limits their survival, probably because of competition for nutrients. b. Evolution of Glandular Cell Types within Monolayers. In order to find an explanation at the cellular level for the increase in PRL secretion and the decline in secretion of the other pituitary hormones, we followed the ultrastructural evolution of glandular cell types with time in culture. Particular attention was paid to PRL cells and to gonadotropic cells. The identification of the gonadotropic cells was made by direct application of the immunocytochemical technique to monolayers, allowing both quantitative and ultrastructural studies of these cells (Tougard et al., 1974). Antibodies against ovine FSH, ovine LH, ovine LHa, and ovine LHP were used for this purpose. It must first be remembered that in the rat anterior pituitary the relative proportion of the six different cell types is well defined. After dispersion by enzymic treatment, their ultrastructure as well as their relative proportions remain unchanged (Ishikawa, 1969; Malamed et al., 1970; Hymer et ul., 1973; Hymer, 1974 Farquhar et al., 1974). In our material (180-gm Sprague-Dawley male rats) we found, by counting in an electron microscope, pellets of freshly dispersed cells: GH cells, 45%; PRL cells, 6%; LH and FSH cells, 15%;TSH cells; 4%; ACTH cells: 3%; chromophobes or nonglandular cells 28%. In 5- to 7-day monolayers the different pituitary cell types retained their typical ultrastructural features and could be easily identified. For example, the different types of gonadotropic cells previously identified in uivo by electron microscope immunocytochemistry using antisera against ovine LH and ovine LHP (Tougard et al., 1973) were found: (1) cells with two types of round secretory granules (500 and 200 nm) and dilated ergastoplasmic cistemae (Fig. 6) (the FSH cell type described by Kurosumi and Oota, 1968); (2) cells with one class of small secretory granules and flattened ergastoplasmic cistemae (Fig. 7) (the LH cell type described by Kurosumi and Oota, 1968); (3) numerous intermediate forms between these two types. As in viuo, gonadotropic cells were generally found associated in small clumps. GH cells, which are the most FIG.6. Seven-day monolayer in control medium. A gonadotropic cell (GT) of the FSH Kurosumi cell type. It contains two types of round secretory granules and dilated ergastoplasmic cisternae. x 10,OOO.

CELL CULTURE OF ANTERIOR PITUITARY CELLS

191

FIG. 7. Same material as in Fig. 6. A gonadotropic cell (GT) of the LH Kurosumi cell type. It contains one class of secretory granules and flattened ergastoplasmic cisternae. x 10,OOO.

CELL CULTURE OF ANTERIOR PITUITARY CELLS

193

FIG.8. Twelve-day monolayer in control medium. The monolayer was treated for localization of acid phosphatase activity. A gonadotropic cell with rounded ergastoplasmic cistemae is seen. Several droplets display a positive reaction (arrows). Xu),OoO.

194

A. TIXIER-VIDAL, D. GOURDJI AND C. TOUGARD

numerous in vivo, still were very frequent at that stage of the monolayer. In contrast, PRL cells appeared more numerous than in vivo and were hyperactive. They were recognized by their large irregularly shaped secretory granules and displayed important development of the ergastoplasm and enlargement of the Golgi zone. Within monolayers of increasing age (up to 2 months), progressive modifications of the glandular cells occurred: (1) PRL cells increased in number and became the prominent component of the glandular cell population. They still displayed the ultrastructural features of high secretory activity, namely, numerous parallel ergastoplasmic cisternae and a large Golgi zone with condensing secretory granules. Their content of large or small secretory granules varied greatly (Fig. 3). (2) Other cell types became apparently less abundant and underwent ultrastructural modifications which made them difficult to identify without the use of specific methods. Thus far such a study has been made only for gonadotropic ceZZs (Tougard et aZ., 1974). Gonadotropic cells with two types of secretory granules and rounded ergastoplasmic cisternae progressively disappeared. At the same time, their large immunochemically negative droplets became positive for Gomori’s acid phosphatase method (Fig. 8). This suggests intracellular reorganization of the cells. Finally, a single gonadotropic cell type remained, which was characterized by low electron density of the cytoplasm, a single class of small, round secretory granules (125-150 nm), large dense bodies (600 nm), a few linear ergastoplasmic cisternae, and a small Golgi zone (Fig. 9). Immunocytochemical staining with antiserum to ovine LH confirms the gonadotropic nature of such a cell type (Fig. 10). It must be noted that similar immunoreactive LH or FSH cells can be seen in vivo (Tixier-Vidal et aZ., 1974). In parallel with this ultrastructural analysis, a quantitative study of the relative number of gonadotropic cells immunoreactive for A-LH, A-LHP, A-LHa, and A-FSH was made by direct count of the culture dishes (Tougard et aZ., 1974) (Tables 1-111). It was found that the number of LH and FSH cells did not dramatically decrease during time in culture (Table 11, SD 11 and SD 16; Table 111, SD 24). In some experimental series it increased, as well as the number of cells positive for A-LHP and for A-LHa (Table 111, SD 24). FIG.9. Two-month monolayer in control medium. A presumed gonadotropic cell (GT) displays typical features: light cytoplasm, one class of small secretory granules, and large dense bodies. G , Golgi zone; P, PRL cell. x11,OOO.

CELL CULTURE OF ANTERIOR PITUITARY CELLS

195

196

A. TIXIER-VIDAL, D. GOURDJI AND C. TOUGARD

FIG.10. Thirty-five-day monolayer (SD 26 series, see Table I), in control medium. Immunocytochemical staining with A-oLH. A gonadotropic cell (GT) displays a positive reaction on the secretory granules and on the cytoplasm. The dense bodies are negative (arrows), as well as the nucleus (N) and mitochondria (m). G, Golgi zone. The ultrastructural organization is similar to that of the presumed gonadotropic cell in Fig. Q. X11,OOO.

NUMBEROF

CELLS IMMUNOREACTIVE WITH

TABLE I1 ANTISERA AGAINST

OVINE

LH,

OVINE

SD 11 series

Age of monolayer

5 days 12 days 18 days 26 days 35 days

Medium LH content in same dish (ndml)

Positive cells (%)

LH LHB LHa LH LHP LHa LH LHP LHa

-

LHP

2 7.5 13 5.5 12 7 3.5 6.5 13.5

-

0

185 129 107 7.75 6.5 7.3 7.4 9.3 7.1

-

4.5

LHP

AND OVlNE

n

LHaaSb

M r r

SD 16 series Medium FSH content in same dish (ndml)

45

-

48 7.3 11.8 6.8 11 13 10.5

-

9.9

Positive cells (%)

Medium LH content in same dish (ng/ml)

Medium FSH content in same dish (ndml)

37.5 28.5

82 82

LH LHP

5 5

LH LHP

2.3 4

5.5 5.2

123 44.8

LH LHP

2.2 9

4.5 4.1

32.2 30.8

LH LHP LHa

3.7 4.5 12

2.6 2.4 2.2

33.8 27.2 30.8

n M

As compared to medium LH and FSH content in the same culture dish in monolayers of rat anterior pituitary cells of increasing age in two experimental series. For LH and FSH assays, see Table I. a

E rn

TABLE 111 NUMBER OF CELLSIMMUNOREACTIVE WITH ANTISERA AGAINST OVINEFSH, OVINE LH, OVINELHp, ~~

12 days

18 days

26 days

35 days

-

-

-

00

~~

Positive cells (%)

FSH LH LHP LHa FSH LH LHP LHa

(D

SD 24 series

SD 21 series

5 days

OWE L H e b

~

~~~

Age of monolayer

c-l

AND

3.5 2 4.3 5 2 1 4.5 14

Medium LH content in same dish (ndml) 23

25 22 24 0.4 1.2 0.8 0.8

-

-

Medium FSH content in same dish (nglml)

38 46 40 34 12.2 11.8 14 18.4

-

-

Medium LH content in same dish (ng/mU

Positive cells (%)

Medium FSH content in same dish (ns/ml)

?

8

M

FSH LH LHB LHa FSH LH LHB LHa FSH LH LHB LHa FSH LH LHB

LHa FSH LH

LHB LHa

0 0 0

<1 0 0 0

19.7 4.9 15.7 22.2 19.9 3 13.1 12.6 18 3.1 18.5

5.2 2.8 2.8 2 <0.5 (0.5 <0.5 1.2 <0.5 <0.5 <0.5 <0.5 i 0.5 <0.5 <0.5 <0.5

-

42 34 40 42

19 21 25 15 14

7

"E U Q

i

3

11

5U

10.5 13

P

15

0

16

10

-

24 24.2

a As compared to medium LH and FSH contents in the same culture dish in monolayers of rat anterior pituitary cells of increasing age in two experimental series. For LH and FSH assays see Table I.

cl

C

Q

CELL CULTURE OF ANTERIOR PITUITARY CELLS

199

c. Comparison of Morphological Data and Hormonal Secretion The rapid rise in PRL secretion is consistent with the hyperactivity of PRL cells already seen after 5 days in culture (Table I). Similar morphological findings have been found in tissue culture, as well as in organ culture (Pasteels, 1963, 1969). Whether the apparent increase in PRL cell number results from mitosis of preexistent PRL cells or from the differentiation of new PRL cells cannot be concluded from the present work on cell cultures. The rapid decline in LH and FSH secretion in culture does not correspond to a similar decline in the number of immunoreactive gonadotropic cells. In contrast, their ultrastructural evolution strongly suggests a decrease in cell secretory ability. The decline in hormonal secretion therefore does not reflect a dedifferentiation of gonadotropic cells. The persistence in these cells of a well-defined Golgi zone, together with large dense bodies, suggests both continuous synthesis and intracellular degradation of the hormone. The evolution of the gonadotropic cells toward a single cell type is of particular interest if one considers that at the same time only FSH remains detectable within the medium. Does this form represent a primitive gonadotropic cell type able to synthesize both LH and FSH but to release only FSH? For the other hormones whose secretion declines in culture there are no similar studies of corresponding glandular cell types, at least using a specific method. The conventional electron microscope method does not allow enough accuracy for identifying cell types having low secretory activity. Immunochemical studies would be desirable for GH, TSH, and ACTH cells. But one may presume that there is no dedifferentiation of glandular cell types, since there is no ultrastructural evidence for undifferentiated glandular cells within the monolayers. 2. Continuous Cell Lines To our knowledge, a few studies deal with the ultrastructure of a continuous cell line: Sato’s GH3 cell line (Gourdji et al., 1972) and our SD1 cell line (Tixier-Vidal et al., unpublished observations). GH3 cells appear as small cells with a high nucleus/cytoplasm ratio (Fig. 11). Secretory granules are often sparse or absent. In a small proportion of cells they may nevertheless be more numerous (Fig. 12a). Their diameter varies from 50 to 150 nm, and they generally are localized near the cell membrane. They are either polymorphic and small, or rounded and often large. The Golgi zone is made of several units scattered within the cytoplasm. Each unit consists of dilated

200

A. TIXIER-VIDAL, D. GOURDJI AND C. TOUGARD

FIG. 11. GH3 cell cultivated in control medium (HAM F 10 medium plus 15% horse serum and 2.5% fetal calf serum). Note high nuclear cytoplasmic ratio, the several units of the Golgi (G) zone, and very few secretory granules. x 10,OOO.

FIG.12. (a) GH3 cells, cultivated as in Fig. 10, display an accumulation of secretory granules and dense bodies. (b) “C” type virus outside a GH3 cell. x54,OOO. Inset: Budding of a virus particule from the plasma membrane. x 110,OOO.

CELL CULTURE OF ANTERIOR PITUITARY CELLS

201

202

A. TIXIER-VIDAL, D. GOURDJI AND C. TOUCARD

FIG. 13. A GH3 cell cultivated as in Fig. 10.Detail of the Golgi zone showing dilated sacculae and microfilaments (Q.sg, Secretory granule. X40,OOO.

saccules and small, smooth vesicles; condensing secretory granules are infrequent (Fig. 13). Multivesicular bodies and bundles of cytofilaments are often associated with the Golgi zone. The cytoplasm contains abundant free ribosomes and polysomes and a few linear ergastoplasmic cisternae. The development of these cytoplasmic

FIG.14. SD1 cell cultivated in control medium (HAM F 10 plus 2.5% horse serum and 2.5% fetal calf serum). As compared to GH3 cells, the secretory granules and dense bodies are more numerous and the saccules of the Golgi zone (G) more flattened. ~10,OOo.

204

A. TIXIER-VIDAL, D. GOURDJI AND C. TOUGARD

organelles varies greatly from one cell to another, which may indicate that the cells are not synchronized. The ultrastmcture of SD1 cells does not differ very much from that of GH3 cells, except that they contain more dense bodies and lysosomes (Fig. 14). The organization of the Golgi zone is also more similar to that of normal rat PRL cells. Such ultrastructural organization differs markedly from that of PRL cells in monolayers of normal pituitary cells, mainly as concerns the size of the secretory granules and the organization of the ergastoplasm. Such differences are to be related to the tumoral origin of GH3 cells. Similar modification has been described in estrogen-induced pituitary tumors (Waelbroeck-Van Gaver and Potvlihge, 1969), and in human pituitary adenomas (Olivier et al., 1974). The same explanation applies to other features such as some abnormal mitochondria, Moreover, GH3 cells, as well as SD1 cells, release by budding from the plasma membrane viral particles morphologically similar to those observed in murine and avian leukemias and sarcomas (Fig. 12b). These ultrastructural data are an important addition to hormonal secretion data as concerns the differentiation of continuous cell lines. The loss of several typical features of the morphological phenotype of PRL or GH cells contrasts with the maintenance of corresponding biochemical differentiation. Nevertheless, the indisputable presence of secretory granules suggests the existence of a storage compartment, which is consistent with the conclusions of our chase experiment showing two intracellular pools of PRL (see Fig. 1).The difference in normal pituitary cells could be more quantitative than qualitative, the storage compartment being larger in normal cells than in continuous cell lines. The paucity of secretory granules does not allow a definite conclusion as to whether PRL and GH are secreted by the same cells or by two morphologically different cell types in mass cultures of GH3 cells. In conclusion, there are several important differences between hormonal secretion and morphological features of primary cultures of normal pituitary cells on the one hand, and of continuous pituitary cell lines on other hand. The first are more similar to in uiuo pituitary tissue than the second, from both morphological and biochemical points of view, at least during the first days of culture. With increasing time their secretory activity decreases, except for PRL, and they become a very heterogeneous cell population with the relative number of glandular cells decreasing. They therefore do not offer at

CELL CULTURE OF ANTERIOR PITUITARY CELLS

205

that stage a good model for biochemical studies. In contrast, continuous pituitary cell lines, which consist of an homogeneous population of glandular cells which may secrete one hormone only, represent a very promising model for biochemical studies of mechanisms involved in the regulation of secretion of some pituitary hormones. Nevertheless, it must be kept in mind that their morphological organization is modified in regard to normal pituitary cells. We now compare in these two models the effects of factors that regulate the secretion of anterior pituitary hormones. 111. Reactivity to Specific Regulating Agents

A.

PRIMARY CULTURES

Since the work of Kobayashi et al. (1961,1963, 1971) showing that rat hypothalamic extracts enhanced and maintained gonadotropic activity of cultured anterior pituitary cells, the effects of several purified or synthetic hypothalamic factors have been analyzed using monolayers of normal anterior pituitary cells as models. Modifications of hormonal secretion were measured, but morphological studies are far less numerous. 1. Effects of Hypothalamic Factors on Hormonal Secretion a. Synthetic TSH-Releasing Hormone (TRH). Synthetic TRH is a tripeptide (pyro-Glu-His-Pro-NH,) originally purified from ovine (Burgus et al., 1970) and porcine (Nair et al., 1970) hypothalamus on the basis of its ability to stimulate TSH secretion. Later, it was also found to enhance PRL secretion, first in uitro in the GH3 rat pituitary cell line (Tashjian et al., 1971), and then in uiuo in man (Bowers et al., 1971) as well as in other mammals. Both these effects of TRH were observed in primary cultures of normal rat pituitary cells used for this purpose on the fourth day following plating, and in short-term experiments (2.5-4 hours). TSH release increased in amounts directly related to the TRH M. After a dose, the half-maximum response occurring at 2 X 3-day exposure a stimulation of TSH synthesis was indicated by an increase in total TSH (medium plus cell), although the cell content decreased. As in uiuo, the stimulating effect of TRH was depressed by thyroid hormone, A similar increase in TSH secretion was observed under the influence of other stimuli known to mimic the effects of hypothalamic factors, such as an increase in potassium concentration, theophylline, or prostaglandin PGEz (Vale et al., 1972a).

A. TIXIER-VIDAL, D. COURDJI AND C. TOUCARD

206

The magnitude of the stimulating effect, again examined in 4-hour experiments, decreased with time in culture and disappeared in a 21-day monolayer. This was accompanied by a decrease in TSH cell content (Vale et al., 1972a). A severalfold increase in PRL release was induced by TRH when cultured cells were obtained from hypothyroid (treated with propylthiouracil) rats. Pretreatment of such cultured cells with thyroid hormone depressed both the spontaneous release of PRL and the stimulating effect of TRH. When the cultured cells were obtained from normal rats, the effect of TRH was lower (less than 50%) and was inconsistent (Vale et al., 1973). No morphological studies have been made under such conditions. b. LH-Releasing Hormone (LRH).Both native (Amoss et al., 1971) and synthetic (Matsuo et al., 1971; Burgus et al., 1971) LRH have been used and have had similar effects on monolayers. i. Short-term experiments. With 4-days monolayers and after a 2.5-hour exposure, the amount of both LH and FSH released increased as a function of the dose with an apparent affinity constant of ca. lo+’M for LH. For FSH, the magnitude of the secretory response was nevertheless considerably lower. After a 3- or 6-day exposure to LRH, stimulation of LH synthesis was indicated by an increase in total LH (cell plus medium), and the LH cell content was decreased. The magnitude of the LH secretory response decreased with time in culture before exposure to LRH (from 4 to 21 days) (Vale et aZ., 1972a).These features of LRH effects on LH secretion are very similar to those of TRH on TSH. i i . Kinetics of secretion in short-term experiments. We followed (Tougard and Tixier-Vidal, 1974) the kinetics of LH and FSH secretion during the first 2 hours following a medium change and the introduction of 10 nglml LRH (Tables IV and V). In series SD 25 a highly significant stimulation of both LH and FSH release occurred after 30 TABLE IV SD 25 SERIES-LH AND FSH MEDIUM CONTENT 30 MINUTES AFTER MEDIUM CHANCE IN 7-DAY MONOLAYEROF NORMALRAT PITUITARYCELLS. EFFECTOF LRH (10 NG/ML) Time Last 24-hour period before change First 30 minutes, control medium First 30 minutes plus 10 nglml LRF

Medium LH (nglml) 2.14

* 0.13 (24 dishes)

1.33 k 0.12 (8 dishes) 4.00

* 0.32 (16 dishes)

Medium FSH (nglml) 92.55 f 7.03 (22 dishes) 5.08 f 0.77 (8 dishes) 14.30 & 0.40 (13 dishes)

CELL CULTURE OF ANTERIOR PITUITARY CELLS

207

TABLE V SD 29 SERIES-LH AND FSH MEDIUM CONTENT WITHIN 2 HOURSAFTER A MEDIUM CHANCE IN 7-DAY MONOLAYER OF NORMALRAT PITUITARY CELL (ONE DISH A T A TIME).EFFECT OF LRH (10 NG/ML)" Time

Medium LH (ng/ml)O

5 minutes

HRP HRP LRH

15 minutes

HRP

1 hour

2 hours

+ LRH

3.1 3.3 1.8

(2.5) (3.1) (1.5)

n.d.r n.d. n.d.

(23) (26.5) (30)

3.6 2.6

(3.1) (2.6)

0.6 0.3

(27.5) (27)

2.6 2.6

<0.1 <0.1 0.3

(22) (20.5)

LRH

2.4

(2.5) (3.3) (2.5)

HRP

2.8 3.2

(2.8) (2.8)

3.1 n.d.

HRP + LRH

2.7 3.2

(4.5) (3.7)

4.2 1.5

(29) (22) (21.5) (27.5)

LRH HRP HRP + LRH LRH

2.8 2.1 3.7 3.9

(2.5) (3.1) (5) (3.1)

2.4 1.1 2.2 6

(27) (27.5) (21.5) (25.5)

HRP HRP LRH

2.2 4.9 4

(3.1) (2.5) (2.5)

n.d. 0.9 12.8

(21.1) (8.5) (21)

HRP

30 minutes

+ LRH

Medium FSH (ng/rnl)O

+ LRH

(31)

In view of cytological studies, horseradish peroxidase (1 mglml) was introduced into both control medium and LRH-enriched medium. bValue of medium LH or FSH content for the 24 hours before the last medium change are shown in parentheses. n.d., Not detectable, below sensitivity of the assay.

minutes. In series SD 29 stimulation was observed after only 1hour. In the two series it appears, moreover, that after a medium change the 24-hour LH medium content was restored within the first 30 or 60 minutes, while at the same time the FSH medium content represented a very small fraction of the 24-hour medium content. This suggests important differences in the cellular mechanisms involved in the release of LH and of FSH. As compared to in vivo experiments in which an acute increase in serum LH occurred 1or 3 minutes after intracarotid injection of LRH (Shiino et al., 1972; Mendoza et al., 1973), the cultured gonadotropic cells seem to react more slowly. iii. Chronic treatment in long-term experiments. The effects of chronic treatment with LRH (2 nglml) on the kinetics of LH and FSH release in long-term (up to 2-month) primary cultures of rat pituitary cells has been followed (Tixier-Vidal et al., 1973). LRH in-

208

A. TIXIER-VIDAL, D. GOURDJI AND C. TOUGARD

duced an increase in both LH and FSH release. This increase appeared only after the first 5 days in culture. The stirnulatory effect on LH release reached a maximum on days 11 to 13, and thereafter progressively decreased to the level of the controls. The stimulatory effect on FSH release reached its maximum later than did that on LH, and generally remained significant for a longer time. The maximum of the FSH response was close to or less than that of the LH response. Depending upon the experiments, the maximum magnitude of the stimulatory effect greatly varied, as well as its maximum duration. These results are consistent with those of Vale et al. (1972a) in showing that cell ability to respond to LRH decreases with time in culture. In addition, they indicate that LRH itself is not able to maintain cell responsiveness, since in our experiments the cells were continuously exposed to LRH. Surprisingly enough, different results have been recently reported by Ishikawa and Nagayama, who found a marked increase in LH secretion after addition of a very high dose (150 nglml) of synthetic LRH to a 30-day monolayer of pituitary cells taken from a young woman. This increase was maintained at the same level for at least 11 days. Such differences in results obtained from rat pituitary monolayers may be related either to the species or to the physiological state of the donor pituitaries. In addition, these investigators found that LRH-induced LH release was rapidly and markedly inhibited by cycloheximide, but later and less acutely depressed by actinomycin D. They concluded that LRH stimulates LH release through LH synthesis and that new mRNA synthesis does not occur during the first 4 days of exposure to LRH. Previous results with cell culture, as well as with organ or tissue culture (see review by TixierVidal, 1974), also suggested a stimulatory effect of LRH upon LH synthesis. c. Effects of Total Hypothalamic Extracts on LH and F S H secretion. Crude hypothalamic extract produced the same effects as LRH on the kinetics of LH and FSH secretion in long-term primary cultures of rat pituitary cells (Steinberger et al., 1973).A similar decline in the stimulatory effect with time in culture was observed. This decline might therefore reflect an intrinsic feature of the cultured cells and not a lack within the LRH molecule of a possible component that would be able to maintain long-term stimulation of gonadotropic secretion. d. Effect of Somatostatin. Somatostatin or somatotropin release inhibiting factor (SRIF) is a tetradecapeptide which was isolated

CELL CULTURE OF ANTERIOR PITUITARY CELLS

209

from ovine hypothalamus on the basis of its ability to inhibit the secretion of radioimmunoassayable GH in primary cultures of normal rat pituitary cells (Vale et al., 1972b). A similar effect of SRIF was later found in uiuo (Brazeau et al., 1974). During a 3-hour experiment with a 4-day monolayer, the secretion rate and the synthesis of GH was reduced 80%. SRIF inhibits the enhancement of GH secretion induced by theophylline and dibutyryl-3'4' cyclic AMP. In conclusion, it appears that the responsiveness of monolayers of normal anterior pituitary cells to hypothalamic factors is quite comparable in magnitude and specificity to that of anterior pituitary gland in uiuo and anterior pituitary slices in uitro. In addition, monolayers have several advantages over the two last systems. First, they offer homogeneous samples of a population of reactive pituitary cells which give consistent responses with low standard errors, since "randomized hemipituitaries and pituitary quarters respond with unacceptable heterogeneity" (Vale et al., 1972a). In regard to freshly dispersed anterior pituitary cells, they give, after 4 or 5 days in culture, a far better response to TRH (Vale et al., 1972a) and to LRH (Tixier-Vidal et al., 1973). The reason for such delay is not well understood. One may easily presume that the acute enzymic treatment necessary for cell dispersion alters some components of the cell membrane that are involved in cell responsiveness. In this respect we found that the binding of 3H-labeled TRH by recently dispersed cells was very low, but increased 5 to 10 times after 5 days TABLE VI 3H-LABELED TRH BINDINGAFFINITY BY DISPERSEDNORMALRAT ANTERIOR PITUITARY CELLS PREVIOUSLY EITHERINCUBATED FOR 20 HOURSOR CULTURED F O R 5 DAYS" Bound radioactivity

Pretreatment

Experiment

20-hour preincubation 5-day preculture

A B C D

"-Labeled TRH concentration in the medium (nM)

135 135 135 135 27

cpml 5 x lo5cells

Percent of bound TRH per 5 x lo5cells

84.9 f 3.2 44.8 1.3 348.9 23.6 383.7 3.9 167.0 5.4

4.7 x 10-9 2.5 x lo-:$ 19.2 x lo-" 21.1 x lo-:$ 37.8 x 10-3

* * * *

Cells were incubated with "-labeled TRH (29 Cilmmole) for 30 minutes at 37°C. Each experiment was done in triplicate. Binding assay and radioactivity determination were performed as indicated in Gourdji et al. (1973b).

210

A. TIXIER-VIDAL, D. GOURDJI AND C. TOUGARD

of culture (Table VI) (A. Morin, unpublished observation). Another interesting possibility offered by monolayers is related to their ability to be maintained for a long time in culture. This allows an indirect approach to the study of mechanisms involved in the differentiation of pituitary cells. In this respect two conclusions appear clearly from the results reported above: (1)The reactivity of pituitary cells to their specific hypothalamic factor is closely associated with their ability to store and to secrete their specific hormone. (2) Hypothalamic factors, at least TRH and LRH, are not able to enhance indefinitely the secretion of an anterior pituitary hormone in uitro. These assumptions are discussed further after an examination of morphological data.

2. Morphological Effects of Hypothalamic Factors on Monolayers a. Efiects of LRH. The phase-contrast observation of living monolayers did not reveal any modification of their appearance after the introduction of LRH into cultures. The ultrastructural modifications induced within gonadotropic cells were studied in both short-term and long-term experiments. i . Short-term eflects on a 7-day monolayer. Modifications of gonadotropic cells were studied as a function of time during the first 2 hours following a medium change and the introduction of LRH (see Tables I11 and IV) (Tougard and Tixier-Vidal, 1974). Ultrastructural modifications were evident at 30 minutes in gonadotropic cells that contained either no or few dilated cisternae. Gonadotropic cells that had a high development of such cisternae were not modified at any time (Fig. 6). In the other gonadotropic cells, important changes were observed near the plasma membrane. The cell outline became infolded. In some places the cytoplasm seemed to stretch, and the plasma membrane became ill-defined (Fig. 15). In other places the plasma membrane showed numerous small invaginations. At the same time numerous secretory granules were lined up along the plasma membrane. We did not find classic profiles FIG. 15. Seven-day monolayer exposed to LRH (10 ng/ml) for 30 minutes. A cytoplasmic extension of a stimulated gonadotropic cell is shown. Several features of the LRH-induced modifications can be seen. In some places (+) the cell membrane is tangentially sectioned and several small canaliculli appear between the secretory granules. In other places, the plasma membrane display small invaginations (double arrows), the structure and the shape of the secretory granules are modified, and empty cavities appear (see inset). The reticulum which appears outside the cell represents “collagen fibrils” (cf) which presumably come out from fibroblasts. X36,OOO. Inset, X72,OOO.

CELL CULTURE OF ANTERIOR PITUITARY CELLS

211

212

A. TIXIER-VIDAL, D. GOURDJI AND C. TOUGARD

of granule extrusions, but important modifications of both secretory granule membrane and plasma membrane which sometimes were confounded (Figs. 15 and 17d). The secretory granules themselves presented a structural modification, namely, a peripheral or total lost of density, which lead to apparently empty cavities (Fig. 15). Other granules seemed to be elongated in the direction of the plasma membrane. Within the peripheral cytoplasm numerous small vesicles or flexuous canaliculi appeared, which were sometimes labeled with horse radish peroxidase, indicating increased pinocytic activity. No microtubules were found near the plasma membrane. Similar small vesicles or canaliculi were also seen within the Golgi zone which sometimes also contained microtubules (Fig. 16). After 1 and 2 hours, the modifications of the Golgi zone became prominent. Besides the increase in small vesicles, numerous multivesicular bodies, sometimes labeled with horseradish peroxidase, were found. Mature secretory granules which are often stored inside the Golgi zone in nonstimulated cells became unfrequent. Simultaneously, the linear cisternae became more numerous, thinner, and shorter, and condensing secretory material was found in some of them. After 4 hours the secretory granules decreased in size and were mainly localized within the peripheral cytoplasm. The Golgi zone was considerably enlarged, and had several stacks of saccules and abundant condensing secretory granules. Immunochemical staining of such stimulated monolayers, when examined with a light microscope, revealed a transitory decrease in LH immunoreactivity of the gonadotropic cells after 15 minutes of contact with LRH. The respective number of cells that react to each of the three anti-sera (A-oLH, A-oLHP, and A-oLHa) was not modified in comparison to controls. In the electron microscope, the main antigenic sites were the same for the three anti-sera, as previously observed in control cultures as well as in uitro, i.e., secretory granules and ground cytoplasm. In regard to results obtained with classic methods, immunochemical staining confirmed the identification of the gonadotropic cells. It also strongly confirmed the important modifications induced in secretory granules by LRH treatment, namely, their rapid migration along the plasma membrane and their FIG.16. Same material as in Fig. 14. A gonadotropic cell which has been exposed to LRH (10 nglml) plus horseradish peroxidase for 30 minutes. The Golgi zone shows an important increase in number of small vacuoles and canaliculi. A condensing secretory granule can be seen inside the innermost cisternae (arrow). X36,OOO.

CELL CULTURE OF ANTERIOR PITUITARY CELLS

213

214

A. TIXIER-VIDAL, D. GOURDJI AND C. TOUGARD

structural changes (Fig. 17). They shrank inside their membranes giving the impression of granules with halo or empty cavities. At the same time, the plasma membrane showed small invaginations which could be interpreted as granule exocytosis. Nevertheless, we observed that with better fixation a single membrane generally remained present. As compared to ultrastructural observations of pituitaries taken from rats sacrificed within minutes following the intracarotid injection of LRH, granule extrusions, which were not clearly seen in culture, were striking in d u o , at least in Kurosumi LH cell types. In contrast they were not seen in Kurosumi FSH cell types (Shiino et al., 1972; Mendoza et al., 1973). In addition, castration cells with highly dilated cisternae were totally unmodified by LRH (Rennels et al., 1971). The modifications of the cell membrane as described in cultured gonadotropic cells were not seen in uiuo. This may result from flattening of the cultured cells, which allows tangential sections of the plasma membrane to be made and reveals more details of its structural organization. ii. Chronic treatment. As already pointed out, at late stages of the monolayer only one gonadotropic cell type remained (Figs. 9 and lo), which had one class of small rounded secretory granules. In a 27-day monolayer chronically treated with LRH, the gonadotropic cells displayed features similar to those appearing after a 4-hour stimulation, but the diameter of secretory granules became very small (100 nm) and numerous parallel and flattened ergastoplasmic cisternae were found (Fig. 18)(Tixier-Vidal et al., 1974). In addition, LRH did not induce reappearance of the other gonadotropic cell types. It is of interest to note particularly that the Kurosumi FSH cell ~

~

FIG.17. Seven-day monolayer treated with LRH (10 nglml) for 15 minutes (a) 30 minutes (b and d), or 5 hours (c). Details of plasma membrane and neighboring secretory granules of stimulated gonadotropic cells are shown. In a, b, and c, immunocytochemical detection of binding sites of an antiovine LH serum was made. (a) The positive reaction is found on secretory granules which are lined up along the plasma membrane. One notes moreover a positive reaction in some places on the plasma membrane (single arrow) and modifications of the structure and of the repartition of the antigenicity in some secretory granules (double arrow). X30,OOO. (b) The positive reaction is found on secretory granules and on cytoplasm. Numerous pockets or empty cavities are found at the cell periphery (arrows). x50,OOO. (c) Same observations as in (b), but the structure of the plasma membrane is better seen. Profiles of granule extrusion are presumed (arrows). X60,OOO. (d) No immunocytochemical staining. With better fixation (1% glutaraldehyde, 0.1 M cacodylate buffer) and classic staining (uranyl acetate, lead citrate), several stages of the relations between plasma membrane and secretory granule membrane can be seen (arrows). x 100,OOO.

CELL CULTURE OF ANTERIOR PITUITARY CELLS

215

FIG. 18. Twenty-seven-day monolayer chronically treated with LRH (2 nglml). Section through a pellet of trypsinized cells. A stimulated gonadotropic cell (GT) displays a striking extension of the Colgi zone ( G )and of the ergastoplasmic cistemae. Secretory granules are scarce. x 12,000.

CELL CULTURE OF ANTERIOR PITUITARY CELLS

217

type did not reappear although at the same time only FSH secretion remained stimulated by LRH. ’ Another important feature of chronic treatment with LRH concerns its apparent inability to increase the number of gonadotropic cells. In addition, the PRL cells remained unmodified by LRH, which agrees with the specificity of its action on hormonal secretion. Morphological observations on chronically treated monolayers allowed us to confirm that LRH is able to induce a high development of organelles involved in synthesis and segregation of secretory product, namely, rough endoplasmic reticulum and Golgi apparatus. They revealed, moreover, its inability to induce differentiation of new gonadotropic cells either by mitosis of preexisting gonadotropic cells or by modification of other types of anterior pituitary cells. These long-term effects of LRH on monolayers may be compared to those observed at the level of pituitary grafts in hypophysectomized rats chronically treated for 2 months with LRH (Debeljuk et al., 1973). In the absence of treatment, the pituitary grafts contained the two types of gonadotropic cells (the LH and FSH cells described by Kurosumi). In the presence of LRH, the FSH cells were the most stimulated; theybwere enlarged, and had an extension of the Golgi zone and small pockets in the cell membrane. The features of the stimulation are identical to those observed in culture, but the cell type is different. This suggests that factors other than LRH may be necessary for the maintenance of different gonadotropic cell types. But in grafts as well as in monolayers, LRH does not seem to be able to induce an increased number of gonadotropic cells. b. Effects of Corticotropic Releasing Factor (CRF).Ohtsuka et al. (1971, 1972) studied the effect of purified CRF on morphological and functional differentiation of cultured chromophobe cells isolated from rat anterior pituitaries. In fact, these investigators did not use monolayers, although they started with dispersed pituitary cells. They isolated a “pure chromophobe” pellet which was cut into small pieces 2 x 2 x 2 mm; the pieces were then cultured as “explants” in roller tubes for up to 1 month. In control cultures (50 pg/liter Lthyroxine), chromophobes divided during the first 6 days and partly differentiated into “immature ambiguous cells.” Such explants contained small amounts of ACTH and GH, but a large amount of PRL. In the presence of CRF plus thyroxine combined with colchicine, chromophobe cell division was arrested and many cells became acidophils. The explants contained a larger amount of ACTH, but neither GH nor PRL. These investigators concluded that CRF in-

218

A. TIXIER-VIDAL, D. GOURDJI AND C. TOUGARD

duced the transformation of the chromophobes into acidophils which produce ACTH. Surprisingly enough, ACTH activity was found associated with small (100-200 nm in diameter) as well as large (300to 500-nm) and “huge” irregularly shaped granules. The latter results are not consistent with in uiuo immunochemical observations (Nakane, 1970), indicating that the rounded large granules contain GH and the huge irregularly shaped granules contain PRL. To our knowledge this is the only example of a possible functional transformation of one cell type into another without modification of the morphological features of the secretory granules.

B. EFFECTSOF REGULATING AGENTS ON CONTINUOUS CELL LINES Continuous cell lines that secrete PRL and GH have been widely used to study the mechanism of action of several agents known to regulate in uiuo the secretion of these two hormones, namely, hypothalamic factors, steroids (glucocorticoids, estrogens, testosterone), and thyroid hormones.

1. Hypothalamic Factors - TRH Stimulation of PRL release by GH3 cells was observed as an effect of crude hypothalamic extract as well as of other bovine tissues (brain cortex, liver, kidney) (Tashjian et al., 1970; Tashjian and Hoyt, 1972) which put in doubt the specificity of the hypothalamic extract. However, further studies have shown that synthetic TRH also stimulates PRL release in GH3 cells as well as in uiuo (see Section II,A,l,a). This effect can be considered specific to hormonal secreTABLE VII AUGMENTATIONOF PRL RELEASE INDUCED BY

TRH

IN

cr-13

INCREASING DOSES OF

CELLS“

TRF concentration (nM) Source Tixier-Vidal et ul. (unpublished data) Tashjian et ul.

Exposure time with TRH

0.27

2.7

27

270

2700

2 hours, 30 minutes

14%

47%

119%

139%

-

24 hours

70%

230%

365%

370%

375%

(1971)*

‘j

Values expressed as percentage of augmentation as compared to control. Values as calculated from published figures or tables.

CELL CULTURE OF ANTERIOR PITUITARY CELLS

219

TABLE VIII PERCENTAGE OF TRH-INDUCEDPRL RELEASE AUGMENTATIONIN GH3 CELLS DURING 24 Houns“

VARIATION IN THE

TRH concentration (nM) Source Tashjian et al. (1971)” Hinkle and Tashjian (1973)” Tashjian and Hoyt (1972)* Tixier-Vidal et al. (unpublished data) ‘I

0.27

2.7

27

70%

230%

38%

131% 83 %

365% 165% 98 %

-

-

70%

Values expressed as percentage of augmentation as compared to control. Values as calculated from published figures or tables.

tion, since TRH modified neither cellular growth nor total protein synthesis (Dannies and Tashjian, 1973). Thus rat pituitary PRL cell lines appeared to be a promising model for studies on the mechanism of action of TRH on PRL release. a. Efiect of TRH on Hormonal Secretion. Stimulation of PRL release was dose-dependent (Table VII and VIII), the maximum response generally being obtained with 27 nM. Nevertheless, depending upon the experiment, it was observed at 2.7-270 nM. In addition, the magnitude of the effect varied strongly from one experiment to another. The time course of the TRH-induced PRL release was examined in both short-term and long-term experiments. Within the first minutes and hours following the introduction of TRH, an increase in PRL release occurred as soon as after 15 minutes. The percentage of the augmentation was maximum between 15 and 30 minutes and thereafter remained approximately steady up to 3 hours. This was shown with two experimental schedules: (1) GH3 cells previously cultivated 1 week were exposed to TRH (27 nM), and small aliquots of medium were withdrawn after increasing time intervals (Gourdji et al., 1973b) ( 2 ) SD1 cells previously cultivated for 5 days were exposed to TRH (27 nM), and the medium was withdrawn and renewed after increasing time intervals (15 and 30 minutes, and 1 , 2 , 4,6,8,10,12, 16,20, and 24 hours) (Morin et al., 1974). According to Tashjian and colleagues who used GH3 cells with different experimental schedules, the stimulating effect was scarcely detectable from 3 to 8 hours and then increased regularly up to 24 hours and 48 hours (Tashjian et al., 1971; Tashjian and Hoyt, 1972). In long-term experiments (from 1 to 20 days), important experi-

?

EVOLUTION OF

TRH-INDUCED

PRL RELEASE IN

THE

TABLE IX MEDIUM AS

M A

FUNCTION OF

THE

ACE OF

THE CULTURE

5

(GH3)

B

Age of culture (days) Source Tashjian et al. (1971)b

1

2

3

4

5

6

-

-

-

-

-

180%

-

5570 165%

26% 34%

130%

96%

-

400%

Gourdji et al. (1972)

-

Tixier-Vidal et al. (unpublished data)

55%

a

-

37% 75%

147% 183%

83% 357% 172%

260%

181%

-

-

-

Values expressed as percentage of augmentation as compared to control.

* Values as calculated from published figures.

-

-

7

8

13

-

-

420%

-

195% 52%

-

-

2o

$ p

140%

0

- 2 -- Ea

- t U? -

22 1

CELL CULTURE O F ANTERIOR PITUITARY CELLS

INFLUENCE OF

TRH

ON

Source Tashjian and Hoyt (1972) Gautvik and Tashjian (1973) Gautvik et al. (1973)

TABLE X PRL CELLULAR

AND

MEDIUM CONTENT"

Treatment

Cells

Media

27 nM TRH, 72 hours 270 nM TRH, 72 hours 50 nM TRH, 72 hours 50 nM TRH, 48 hours

+200% +200% +370% +220%

+210% +210%

+45%

-

" Expressed as percent of modification induced as compared to control. Values as calculated from the published figures or tables.

mental variations of TRH cell responsiveness appeared (Table IX). Such variability may be related to interference with cell divisions which significantly modify the number of reactive cells during the course of long-term experiments. There are no data concerning an eventual relationship between cell cycle and cell responsiveness to TRF. According to Tashjian and colleagues, after a 6-hour exposure to 27 nM TRH the stimulatory effect persists, despite washing of the cells and the addition of fresh medium lacking TRH (Tashjian et al., 1971, 1973). Beside its action on PRL release, TRH also increased PRL cell content in long-term experiments from 200 to 370%(Table X) (Tashjian and Hoyt, 1972; Gautvik and Tashjian, 1973). The percentage of this augmentation does not seem to be closely related to that of the TRH-induced increase of PRL. With SD1 cells, and after 24 hours of TABLE XI PRL

CELL CONTENT OF SD1 CELLS AFTER A 24 HOUR EXPOSURE TO 60 nM TRH AS COMPARED TO CONTROL CULTUHESIN

T w o DIFFEHENT EXPERIMENTS~ Treatment

PRL cell content (ng/ml protein) ~~~

Experiment I Control TRH (60 nM) Experiment I1 Control TRH (60 nM)

~

*

102.6 5 92.1 & 3.6 671.2 k 59.4 271.0 f 48.8

Results were obtained in quadruplicate (experiment I) or triplicate (experiment 11). For experimental conditions see text. PRL was measured by radioimmunoassay as mentioned in Table I. (I

222

A. TIXIER-VIDAL, D. COURDJI AND C. TOUCARD

exposure, we found either no increase or a decrease (Table XI) (Morin, 1974). In contrast to its stimulating effect on PRL secretion, TRH decreased GH secretion (Tashjian et al., 1971; Tashjian and Hoyt, 1972). Such an effect of TRH on GH secretion in uitro by a rat continuous cell line has not been observed with normal pituitaries in viuo or in uitro. TRH did not induce TSH secretion in GH3 or in SD1 cell lines (Tixier-Vidal, unpublished observations). b. Binding of Tritium-Labeled TRH. The significance of radioactive hormone-binding studies in investigating hormonal control mechanisms implies that the labeled compound is as similar as possible to its nonlabeled counterpart. The 3H-labeled TRH prepared by Pradelles et al. (1972) meets this condition. It is undistinguishable from the nonradioactive peptide by both chemical analytical characteristics and biological activities. In contrast, although the biological activity of the "-labeled TRH labeled by Abbott Laboratories is equivalent to that of nonradioactive TRH, it contains 10-25% radioactive impurities (Hinkle and Tashjian, 1973). i . Kinetics of the binding. The binding of 3H-labeled TRH to GH3 cells was time-dependent, increasing linearly during the first 15 minutes (Gourdji et al., 1973b) or 30 minutes (Tashjian and Hoyt, 1972; Hinkle and Tashjian, 1973) and thereafter reaching a plateau stable for at least 1 hour (Gourdji et al., 1973b) or 4 hours (Tashjian and Hoyt, 1972). This time course did not vary with the 3H-labeled TRH concentration. It was closely similar to that of the TRH-induced increase in PRL (Gourdji et al., 1973b). The specificity of the binding has been tested by comparing the affinity of GH3 pituitary cells with that of 3T3 fibroblasts and C6 glial cells, and by competition experiments with unlabeled TRH and other peptides (Gourdji et al., 1973b). ii. Characteristics of the binding. After 30 minutes (Gourdji et al., 1973b) or 90 minutes (Hinkle and Tashjian, 1973), the number of bound molecules per cell was dose-dependent. In the range of biological doses and depending on the experiment, the number of bound molecules per cell varied from 10,000 to 130,000. It increased with TRH concentration following a complex law. At 3 7 C , intact cells possess at least two binding components as revealed by a Lineweaver and Burk plot. One (KA 4 X M )with high affinity corresponds to biologically active concentrations, the other (KA 3 x M ) with low affinity corresponds to high and unphysiological doses (155-1080 nM). With GH3 cell homogenate at O'C, only one binding component

CELL CULTURE OF ANTERIOR PITUITARY CELLS

223

for "-labeled TRH was found by Hinkle and Tashjian (1973) (Kd 25 x 10+ M). In this system the binding did not increase between 100 and 150 nM. Since these investigators did not report data with higher doses, they did not demonstrate real saturation. Comparative studies of "-labeled TRH binding to mouse TSHsecreting tumor cells and to GH3 cells revealed similar binding affinities in the two systems, as shown by the value of the affinity constant and competition with TRH structural analogs (Vale et al., 1973). In cultured mouse TSH-secreting tumor cells, two TRHbinding sites have also been demonstrated (Grant et al., 1973). One may presume a relationship between the number of bound molecules and the induced stimulation. However, this has not yet been carefully investigated. The only available data are those of Hinkle and Tashjian (1973). Half-maximum binding occurred at 11 nM TRH (1%-hour incubation) and the half-maximum biological effect occurred at 2nM TRH (3-day incubation). iii. Stability ofthe binding. After 3H-labeled TRH binding by intact GH3 cells at 37°C and washing at 4"C, the radioactivity of the cells was extremely stable at 0°C (Gourdji et al., 1973b; Hinkle and Tashjian, 1973). In contrast, the cell radioactivity rapidly decreased at 37°C. After a 30-minute binding at 37'32, addition of a 100-fold excess of nonradioactive TRH did not markedly increase the kinetics of this release. The nature of the released radioactive material was not analyzed (Gourdji et al. 197313). According to Hinkle and Tashjian (1973), nevertheless, the stability of 3H-labeled TRH binding to GH3 cells is time-dependent. When cells were incubated for increasing time intervals with 3H-labeled TRH, the ability to displace bound radioactivity with a 200-fold excess of unlabeled TRH declined from 10 minutes to 1 hour and disappeared after 24 hours. These investigators concluded that the binding "becomes irreversible over the ensuing 24 hour period." Such a conclusion, however, requires chemical identification of the radioactive material bound to the cell after a 24-hour exposure. iu. Subcellular localization. An autoradiographic study of GH3 cells incubated 30 minutes with 200 nM 3H-labeled TRH revealed intracellular and intranuclear radioactive material, with no concentration at the plasma membrane level (Gourdji et al., 1973b). A similar localization was found by dry-mount autoradiography in normal rat pituitary cells 1 hour after the injection of "-labeled TRH (Stumpf et al., 1974). A chemical analysis, by thin-layer chromatography in the presence of TRH, of the radioactive material bound to intact GH3 cells after a

224

A. TIXIER-VIDAL, D. GOURDJI AND C. TOUGARD

TABLE XI1 REPARTITION OF “H-LABELED TRH ASSOCIATED WITH CH3 CELLS AND SUBFRACTIONS OBTAINED AFTER 30-MINUTE INCUBATION WITH “-LABELED TRH (37‘42, 200 nM) Cell Fraction

Bound 3H-labeled TRH (nM/mg protein)

Intact cells or total homogenate Nuclear fraction “Organite” fraction Cytosol

590* 59 305 2 129 1217 2 268 1829

% of Total cell

radioactivity 100% 17.5 f 1.8% 25.6 & 6.5% 63.5 6.0%

*

30-minute incubation with ”-labeled TRH was performed within cell homogenate and subcellular fractions. The results (Table XII) demonstrated the existence of chemically unmodified TRH (90-95% of the bound radioactivity) within the GH3 cells at three levels: nucleus, cytoplasmic organelles, and cytosol (Brunet et al., 1974). As far as the intracellular binding of TRH is concerned, Poirier et al. (1972), studying the binding of 3H-labeled TRH by several subfractions of bovine pituitary, mentioned that, besides the “pure plasma membrane” fraction, the nucleus and the “total microsomes” are also able to bind TRH to a lower but significant level. In their opinion this binding, however, represents a minor component. Grant et al. (1973) reported that no significant intracellular uptake of 3Hlabeled TRH was detected during the first hour of incubation at 37°C of mice tumor pituitary thyrotropic cells. In contrast, our results support the hypothesis that, in addition to its possible interaction with a plasma membrane receptor (Labrie et al., 1972), TRH enters the cell and is then tightly bound to some intracellular sites or molecules. The eventual physiological significance of these intracellular events is examined further. c. Zntetference of Other Hormones or Drugs with TRH Action on PRL Secretion. In order to further analyze the mechanisms of action of TRH, its interaction with several hormones and drugs that have a direct effect on PRL release was examined. Hydrocortisone, CB 154 Sandoz (an ergot drug alkaloid), and colchicine have an inhibiting effect, while estrogens, 3‘-5‘-CAMP,prostaglandin, and high potassium concentration are potent stimulators (Tashjian et al., 1970; Tashjian and Hoyt, 1972; Gourdji et al., 1973a; Gautvik and Tashjian, 1973). After 48 and 96 hours of exposure, CB 154 decreased TRH-induced PRL release (Gourdji et al., 1973a), as well as TRH-induced cell spreading (Fig. 19). It did not modify the kinetics of 3H-labeled

FIG. 19. (a) GH3 cells cultivated in control medium. (b) GH3 cells cultivated 4 days in medium containing 270 niU TRH. Note the enlargement and spreading of cell cytoplasm and nucleus. (c) GH3 cells cultivated for 4 days in medium containing CB 154 (500 nglml). Note the cell retraction as compared to the control. (d) GH3 cells culNote tivated for 4 days in medium containing CB 154 (500nglml) and TRH (270 a). the relative inhibition of spreading compared to (b). ~ 8 0 0 .

226

A. TIXIER-VLDAL, D. GOURDJI AND C. TOUGARD

TRH binding, whether the cells were or were not exposed to the drug 24 hours before the binding experiment. It appears therefore that CB 154 acts on a later phase of the intracellular events induced by TRH. When cultures were pretreated with TRH, CB 154 increased the intracellular PRL concentration by about 80%, as it did in untreated cultures (Gautvik et al., 1973). Since the effect of CB 154 on the storage of secretory granules and of PRL has been found in organ cultures of normal pituitary (Pasteels et al., 1971; Ectors et aZ., 1972), its effects on GH3 cells suggest that these cells also have some ability to store PRL. Nevertheless, preliminary ultrastructural examination of CB-154-treated cells did not reveal significant storage of secretory granules within cytoplasm (TixierVidal, unpublished observation). Other substances known to act on secretory phenomenon in general also act on TRH-induced PRL release; colchicine increased the PRL cell content (70%) in TRH-stimulated GH3 cells (Gautvik et al., 1973). A high potassium concentration (50 mM) decreased intracellular PRL and increased extracellular PRL in TRH-stimulated cells (Gautvik and Tashjian, 1973). d. Effect of TRH on Synthesis and Turnover of PRL. Assuming that GH3 PRL cell content is low and that the duration and magnitude of its effect are too great, Tashjian and colleagues concluded that there was “an increase of PRL production which could be a c complished in two ways, either by increasing synthesis of prolactin or by decreasing degradation of the hormone” (Tashjian et al., 1971, 1973). These two hypothesis were examined by studying the incorporation of l e ~ c i n e - ~into H PRL in GH3 cells which were previously treated for 3 days with TRH or not treated. The increase in leucine3H incorporation into PRL caused by TRH was identical to the increase in medium PRL (Dannies and Tashjian, 1973).The increase in medium PRL in the presence of TRH therefore corresponds to an increase in PRL synthesis and not to a change in the rate of its degradation in the presence of TRH. In contrast with Tashjian and colleagues we found in our experiments on GH3 cells, as well as on SD1 cells, a more rapid effect of TRH on PRL release, which reached a maximum after 15-30 minutes of exposure (see Section III,B,l,a). We therefore examined by pulsechase experiments using SD1 cells the effects of TRH on both synthesis and turnover of PRL (Morin, 1974). Unstimulated cells previously maintained for 24 hours on a medium lacking free leucine molecules (except those contained in the 5 % serum) were submitted to a 10-minute pulse with ~ - l e u c i n e - ~(30 H Ci/mmole, 250 pCi/ml).

CELL CULTURE OF ANTERIOR PITUITARY CELLS

227

After three washings in the presence of 5 mM L-leucine, chase experiments were performed again in the presence of 5 mM L-leucine and in four different media: (1) control without TRH, (2) with 60 nM TRH, (3) with 3.6 X M cycloheximide, (4) with 60 nM TRH plus 3.6 X M cycloheximide. In some experiments proline-IC (5 pCi/ml) was introduced during the chase. At increasing time intervals (15 and 30 minutes, and 1, 2, 4, 6, 8, 10, 12, 16, 20, and 24 hours) the medium was withdrawn and renewed. Five experiments were performed using three to five dishes each time. The evolution of the specific radioactivity of the PRL released into the medium was determined as reported above (Section II,A,2). The results revealed that the specific radioactivity of medium PRL was lowered in the presence of TRH and for the first hour of chase; thereafter it reached the level of the controls (Fig. 1). This indicates that TRH released an old PRL pool which was synthesized before the pulse. In addition, cycloheximide, which inhibited protein synthesis 95% after 15 minutes of contact, did not inhibit the stimulatory effect of TRH on medium prolactin(ng/ml)

1

0

I 1

1

I

I

1

I

I

1

I

I

I

FIG.20. Evolution of medium PRL content (nanograms per milliliter) of SD1 cells as a function of time under different conditions: control (0-O), 60 nM TRH M cycloheximide (L -x), 3.6 X (.---OX 60 nM TRH plus 3.6, M cycloheximide (x- - -x). Each time represents triplicate dishes in each series. At each time the medium was withdrawn and renewed. PRL was measured by radioimmuncassay using the rat PRL kit kindly provided by NIAMDD program.

228

A. TWER-VIDAL, D. GOURDJI AND C. TOUGARD

PRL release as observed during the first 30 minutes, although it reduced PRL release in both series (control and TRH) (Fig. 20). This observation confirms that the rapid increase in PRL release induced by TRH was not mediated through new protein synthesis. It is consistent with those concerning the specific radioactivity of PRL. Besides its rapid effect on release of a previously synthesized PRL pool TRH was also shown to enhance PRL synthesis, but after a lag phase of 4 or 6 hours, which is consistent with Tashjian’s results. These results show that TRH has two effects: (1) a rapid effect on the release of preformed PRL; (2) a delayed effect on its synthesis. Whether these two effects are independent or not remains to be examined. Such a biphasic effect of TRH is consistent with current concepts of the mechanism of action of hypothalamic factor on pituitary cells (McCann, 1971; Jutisz et aZ., 1973). It appears therefore that continuous pituitary cell lines do not differ fundamentally in this respect from normal pituitary cells. e. MorphoZogicaZ Effects of TRH. Biologically active doses of TRH induced an important spreading of GH3 cells (Gourdji et d.,1972) (Fig. 19). This effect, as seen by phase-contrast microscopy was already evident after a 2 to 4-hour treatment with high doses of TRH (270 IN),but showed a 1-day lag with lower doses. TRH-treated cells appeared more tightly juxtaposed, and finally formed more continuous and regular monolayers than did control cells. In addition, the nucleus appeared more distinct, and the nucleolus was prominent (Fig. 19). Scanning microscope study of TRH-treated cells revealed fine-branching cytoplasmic processes which firmly bound the cells to the culture dishes (Tashjian and Hoyt, 1972). The cell ultrastructure was modified by TRH k a t m e n t , but to a variable extent from one cell to another. The relative number of modified cells increased with time of exposure. Newly divided cells were always unmodified. The more strongly altered organelle was the whole smooth membrane system, including the Golgi apparatus. TRH induced a conspicuous spreading of the Golgi zone, with flattening of the Golgi saccules and drastic multiplication of small, smooth vesicles and canaliculi (Fig. 21). Moreover, bundles of microfilaments, microtubules, and “annulate lamellae” appeared more frequently than in control cells. Maturing secretory granules were observed more often than in control cells. In addition, mature secreFIG.21. GH3 cell exposed to TRH (100 nM) for 30 minutes. Detail of the Golgi zone. Note important modifications as compared to control (Fig. 12): development and flattening of saccules, and increase in number of small vesicles and canaliculi. X45,000.

CELL CULTURE OF ANTERIOR PITUITARY CELLS

229

230

A. TIXIER-VIDAL, D. GOURDJI AND C. TOUGARD

CELL CULTURE OF ANTERIOR PITUITARY CELLS

231

tory granules increased in number and migrated toward the cell periphery. In that region of the cytoplasm, small vesicles and canaliculi were frequent, as seen in the Golgi zone (Fig. 22). Multivesicular bodies also appeared more numerous, as well as dense bodies. In contrast to the smooth endoplasmic reticulum, the rough endoplasmic reticulum was not strongly modified, although polysome arrangements were extremely conspicuous. Important modifications of the plasma membrane occurred from place to place. They began soon after a 15-minute exposure as an irregular and spreading wave (Fig. 22a). Simultaneously, numerous small invaginations were observed, which were sometimes coated (Fig. 22b). Despite the abundance of small invaginations, profiles of granule extrusion were infrequent. In addition, bunches of cytoplasmic processes together with microvilli became striking. These cytoplasmic processes always contained ribosomes, and sometimes rough endoplasmic reticulum and mitochondria (Fig. 23). Similar precocious spreading of the peripheral cytoplasm and modifications of the plasma membrane were found in gonadotropic cells of LRH-treated primary cultures (see Section IIIA,2,a). It appears therefore that, despite their tumoral origin, GH3 cells share common features with gonadotropic cells in their response to the appropriate hypothalamic releasing hormone, that is, TRH or LRH. This also suggests that there is no fundamental difference between the mechanism of action of these two hypothalamic hormones. 2. Steroid Hormones a. Glucocorticoids. Effects of glucocorticoids were studied in different cell lines: GH rat pituitary cell lines (GH3 and related strains) and an ACTH mouse pituitary cell line (AtT-20). i. GH cell lines. Hydrocortisone (HC) stimulated four- to eight-fold the rate of production of GH by GH3 cells, while it depressed the rate of PRL production to less than 25% of that in control cultures (Bancroft et al., 1969; Tashjian et al., 1970). This effect on GH secretion occurred at low doses (5 x 10+ M).It was accompanied by a FIG.22. TRH-stimulated GH3 cells. Modification of the plasma membrane and of the peripheral cytoplasm. (a) In a 15-minute-treated GH3 cell (27 nM) one can see irregular invaginations and spreading of the plasma membrane and small vesicles and cannaliculi in the cytoplasm. A multivesicular body labeled with horseradish peroxidase introduced into the medium at the same time as TRH. X60,OOO. (b)A 30-minutetreated GH3 cell (27 nM). One can see, in addition to features similar to those in (a), accumulation of secretory granules near the plasma membrane, coated invaginations of the cell membrane, coated small vesicles, and thin canaliculi (arrows). X45,OOO.

FIG.23. GH3 cell treated with TRH (100 nM) for 30 minutes. A bunch of cytoplasmic processes which contain classic cytoplasmic organelles. x 10,OOO.

CELL CULTURE OF ANTERIOR PITUITARY CELLS

233

40-50% decrease in the rate of cell growth and a 30-40% decrease in total protein synthesis. In addition, HC decreased the incorporation of l e ~ c i n e - ~within H PRL (Dannies and Tashjian, 1973). Other corticosteroids had either the same potencies as HC (corticosterone) or no effect (cortisone, tetrahydrocortisol)(Tashjian and Hoyt, 1972). The effect of HC was antagonized by testosterone and 17-methyltestosterone (Tashjian and Hoyt, 1972). It must be noted that such complex interactions of corticosteroids with GH secretion in GH3 cells have no equivalent in U ~ U O . ii. ACTH cell line. In contrast to GH cell lines, AtT-20 cells that secrete ACTH are physiological target cells for glucocorticoids. Indeed, glucocorticoid hormones inhibit ACTH production in these cells without significantly altering their growth (Watanabe et al., 1973a). Investigation of the binding capacity of a 3H-labeled synthetic glucocorticoid (triamecinolone acetonide) by cytosol and by a nuclear fraction obtained from AtT-20 previously incubated with the labeled glucocorticoid were performed (Watanabe et al., 1973b, 1974). Maximum cytosol uptake was reached after 5 minutes of incubation, while binding in the nuclear fraction continued to increase up to 30 minutes. In addition, cytosol binding capacity rapidly declined when that of the nuclear fraction increased. Cell-free nuclei required the presence of a cytosol hormone receptor complex for binding triamecinolone-3H acetonide. In both fractions the bound radioactivity consisted of undegraded hormone. Some chemical properties of the two receptors could be established. They both had a protein moiety with functional sulfhydryl groups, and they had similar sedimentation coefficients. A nuclear receptor was extracted from the chromatin of incubated cells (Watanabe et al., 1974). It appears that the AtT-20 cell line represents another example of an heuristic model for analysis at the cellular and molecular level of mechanisms that control another pituitary hormone, ACTH. b. 17pEstradiol. Since GH3 cells were obtained from an estradiol-induced rat pituitary tumor, its interaction with these cells is of physiological interest. First, the ability of GH3 cells to produce tumors when injected into Wistar/Furth rats depends on the estradiol circulating level of the recipient (Sonnenschein et al., 1973). The continuous cell line therefore retains the initial estradiol dependence of the tumor. In adM )increases dition, estradiol at a very low concentration (1O-li, PRL production by about twofold, while decreasing GH production by 30% (Tashjian and Hoyt, 1972). A similar action on PRL release, although not on GH secretion, has been observed in organ culture of

234

A. TIXIER-VIDAL, D. GOURDJI AND C. TOUGARD

rat pituitary (Meites et al., 1963; Lu et al., 1971). A selective uptake of 17pestradiol by GH3 cells has been shown, although for higher doses (lo+ and lo-* M )than those that stimulated PRL production according to Tashjian and Hoyt (Sonnenschein et al., 1973). Three high-affinity receptors for estradiol have been found within subcellular fractions of GH3 cells, one in the cytosol and two in the nucleus (Mester et al., 1973). These binding macromolecules displayed some similarities with estradiol receptors found in other estrogen target tissues. c. Thyroid hormones. Triiodotyronine (T,) and thyroxine (T,) induced a threefold enhancement of GH release in a GH1 rat clonal cell line (Samuels et al., 1973). This effect was detected within the first 2-3 hours of incubation of intact cells with thyroid hormones (Samuels and Tsai, 1973). The capacity of different subcellular fractions obtained from preincubated intact GH1 cells at 37°C to bind 1251-labeledT3 and T, was investigated (Samuels and Tsai, 1973). Only the nuclear fraction bound specifically Iz5I-labeled T3or T4 The nuclear receptor had less affinity for T4than for T,. The apparent Kd for these two hormones, as determined with the Scatchard method, were 29 nM for T3 and 260 nM for T,. In contrast, the maximum binding capacity was similar for both hormones at saturation (5000 molecules of hormone per nucleus). In addition, preincubation of G H l cells with an increasing concentration of nonlabeled T, or T, induced an augmentation of the binding capacity of their nuclear fraction. These data were interpreted as indicating a possible transfer to the nucleus of an unstable cytoplasmic receptor. The chemical analysis of the radioactive-bound material extracted from the nuclear fraction revealed that the thyroid hormones were not significantly degraded or metabolized in the hormone-receptor complex (Samuels and Tsai, 1973). The nuclear binding characteristics have been correlated with hormone-induced biological effects in intact cells. Hormone concentrations inducing a half-maximal biological effect are slightly lower than the hormone concentration giving a half-maximal binding. Nevertheless, they are similar enough to suggest that the nuclear binding sites are involved in the thyroid hormone regulation of GH1 cells (Samuels and Tsai, 1973).

IV. Conclusion In conclusion, anterior pituitary continuous cell lines were found to be a useful model for the analysis of mechanisms of action of

CELL CULTURE OF ANTERIOR PITUITARY CELLS

235

agents that quantitatively regulate their hormonal secretion. In several cases they appear to be very close to the in uiuo situation, e.g., control of PRL secretion by TRH and estrogens, control of GH secretion by thyroid hormone, and control of ACTH secretion by glucocorticoids. Because of their cellular homogeneity, they have the advantage in regard to primary culture of allowing investigations at the molecular level, which are undergoing rapid progress. They nevertheless display some “abnormalities” which can be related to their tumoral origin, mainly secretion of both PRL and GH by clonal strains and control of GH secretion by glucocorticoid. In addition, as concerns gonadotropic function, no continuous cell lines have been proposed for the time being. In this case, young primary cultures still offer a good model. Finally, as a result of the cell culture approach, great progress has already been obtained and can still be expected in our knowledge of mechanisms that control anterior pituitary secretion. In contrast, only negative results have been found on the mechanisms that control their differentiation. Studies with cultures of fully differentiated cells suggest that the differentiation of pituitary cells is irreversible, and that hypothalamic hormone does not act on pituitary cell differentiation. ACKNOWLEDGMENTS

We greatly appreciate the excellent technical assistance of Mrs. Picart, Miss M. F. Mormiche, Mr. D. Grouselle, and Mr. C. Pennarun, and the help of Miss A. Bayon in typing the manuscript. Part of our work was supported by grants from the Centre Nation de la Recherche Scientifique (E. R. 89,ATP) and from the Direction Gknkrale B la Recherche Scientifique et Technique (Contract 72 7 0100).

REFERENCES Amoss, M., Burgus, R., Blackwell, R., Vale, W., Fellows, R., and Guillemin, R. (1971). Biochem. Biophys. Res. Commun. 44,205. Bancroft, F. C. (1973).Endocrinology 92,1014-1021. Bancroft, F. C., Levine, L., and Tashjian, A. H. (1969).J. Cell Biol. 43, 432-441. Batzdorf, U.,Gold, V., Matthews, N., and Brown, J. (1971).J. Neurosurg. 34,741-748. Bowers, C. Y., Friesen, H. G., Hwang, P., Guyda, H. J., and Folkers, K. (1971).

Biochem. Biophys. Res. Commun. 45,1033-1041. Brazeau, P., Rivier, J., Vale, W., and Guillemin, R. (1974).Endocrinology 94,184-187. Brinkley, B. R., Murphy, P., and Richardson, C. (1967).J. Cell Biol. 35, 279-283. Brunet, N.,Gourdji, D., Tixier-Vidal, A., Pradelles, P., Morgat, J. L., and Fromageot, P.(1974).FEES (Fed. Eur. Biochem. SOC.), Lett. 38, 129-133.

236

A. TIXIER-VIDAL, D. GOURDJI AND C . TOUGARD

Buonassisi, V., Sato, G., and Cohen, A, I. (1962). Proc. Nut. Acad. Sci. US. 48, 1184-1190.

Burgus, A., Butcher, M., Ling, N., Monahan, M., Rivier, J., Fellows, R., Amoss, M., Blackwell, R., Vale, W., and Guillemin, R. (1971). C. R. Acad. Sci. 273, 1611-1614. Burgus, R., Dunn, T. F., Desiderio, D., Ward, D. N., Vale, W., and Guillemin, R. (1970). Nature (London) 226,321-324. Cohen, A. I., and Furth, J. (1959). Cancer Res. 19,72-78. Daikoku, S., Kinutani, M., and Watanabe, Y. G. (1973). Neuroendocrinology 11, 284-305.

Dannies, P., and Tashjian, A. H., Jr. (1973)./. B i d . Chem. 248,6170-6173. Debeljuk, L., Arimura, A., Shiino, M., Rennels, E. G., and Schally, A. V. (1973). Endocrinology 92,921-930. Dubois, M. P. (1971a). C . R. Acad. Sci., Ser. D 272,433-435. Dubois, M. P. (1971b). C. A. Acad. Sci., Ser. D 272,1793-1795. Dupouy, J. P., and Magre, S. (1973).Arch. Anat. Microsc. Morphol. Exp. 62, 185-206. Ectors, F., Danguy, A., and Pasteels. J. L. (1972).J . Endocrinol. 52,211-212. Farquhar, M. G. (1957). Anat. Rec. 127,29. (Abstr.) Farquhar, M. G. (1971). In “Subcellular Structure and Function in Endocrine Organs” (H. Heller and K. Lederis, eds.), pp. 79-124. University Press, Cambridge. Farquhar, M. G., Skutelsky, E. H., and Hopkins, C. R. (1974). I n “The Anterior Pituitary Gland” (M.G. Farquhar and A. Tixier-Vidal, eds.). Academic Press, New York. In press. Ferrand, R. (1972).Arch. Biol. 83,297-371. Ferrand, R., and Nanot, J. (1968). C. R. SOC. Biol. 162,983-985. Franchimont, P., and Pasteels, J. L. (1972). C. R. Acad. Sci., Ser. D 275, 1799-1802. Gailani, S., Nussbaum, A., McDongall, J., and McLimans, W. (1970). Proc. SOC. Exp. B i d . Med. 134,27-32 Gautvik, K. M., and Tashjian, A. H., Jr. (1973). Endocrinology 93,793-799. Gautvik, K. M., Hoyt, R. F., and Tashjian, A. H., Jr. (1973).J . Cell. Physiol. 82,401-410. Gourdji, D., Kerdelhub, B., and Tixier-Vidal, A. (1972). C . R . Acad. Sci. Ser. D 274, 437-440.

Gourdji, D., Morin, A., and Tixier-Vidal, A. (1973a).Human Prolactin, Excerpta Med. Found. Int. Congr. Ser. No. 308, pp. 163-166. Gourdji, D., Tixier-Vidal, A., Morin, A., Pradelles, P., Morgat, J. L., Fromageot, P., and Kerdelhuk, B. (19731-3).Exp. Cell Res. 82,39-46. Grant, G., Vale, W., and Guillemin, R. (1973). Endocrinology 42, 1629-1633. Guedenet, J. C., Grignon, G., and Franco, N. (1970). Int. Conf. Electron. Microsc. Proc., 7th Grenoble 3,565. Herlant, M. (1960). Bull. Microsc. Appl. 10,37-44. Herlant, M. (1964). Reu. Cytol. 17,299-381. Hinkle, P. M., and Tashjian, A. H., Jr. (1973)./. Bid. Chem. 248,6174-6179. Hopkins, C. R., and Farquhar, M. G . (1973)./. Cell Biol. 59 (2), part 1,276-303. Hymer, W. C. (1974). In “The Anterior Pituitary Gland” (M. G. Farquharand A. TixierVidal, eds.). Academic Press, New York. In press. Hymer, W. C., Evans, W. H., Kraicer, J., Mastro, A., Davis, J., and Criswold, E. (1973). Endocrinology 92,275-287. Ishikawa, H. (1969).Endocrinol. lap. 16,517. Ishikawa, H., and Nagayama, T. (1973). Biochern. Biophys. Res. Commun. 55, 492-498.

Jost, A. (1966). In “The Pituitary Gland” (G. W.Hams and B. T. Donovan, eds.) Vol. 2, pp. 299-323. Butterworth, London.

CELL CULTURE O F ANTERIOR PITUITARY CELLS

237

Jutisz, M., and D e La Llosa, P. (1972).In “Glycoproteins, Their Composition, S t r u c ture and Function,” Part B. (A. Gottschalk, ed.), 2nd Ed. p. 1019. Elsevier, Amsterdam. Jutisz, M.,Berault, A,, Kerdelhui., B., and Theoleyre, N. 1973. Some Aspects Hypothalamic Regul. Endocrine Funct. Symp. Med. Hoechst, Vienna. Kobayashi, T., Kobayashi, T., Kigawa, T., Mizuno, M., and Amenomori, Y. (1961).Endocrinol. Jap. 8,223-226. Kobayashi, T., Kobayashi, T., Kigawa, T., Mizuno M., and Amenomori, Y. (1963).Endocrinol. lap. 10, 16-24. Kobayashi, T., Kigawa, T., Mizuno, M., and Watanabe, T. (1971).Karolinsku Symp. Res. Methods Reprod. Endocrinol., 3rd Symp.; In Vitro Methods Reprod. Endocrinol. Stockholm pp. 27-40. Kraicer, J. (1974). In “The Anterior Pituitary Gland” (M. G. Farquhar and A. TixierVidal, eds.). Academic Press, New York. In press. Kurosumi, K., and Oota, Y. (1968).2. Zellforsch. Mikrosk. Anat. 85,34. Labrie, F., Barden, N., Poirier, G., and De Lean, A. (1972).Proc. Nut. Acad. Sci. U.S. 69,283-287. Labrie, F., Pelletier, G., Lemay, A., Borgeat, P., Barden, N., Dupont, A., Savary, M., Cote, J., and Boucher, R. (1973). Karolinsku Symp. Res. Methods Reprod. Endocrinol. Protein Syn. Reprod. Tissue, 6th, Stockholm pp. 301-340. Leblond, C. P., and Walker, B. E. (1956).Physiol. Rev. 36, 255-276. Lu, K. H., Koch, Y., and Meites, J. (1971).Endocrinology 89,229-233. McCann, S. (1971). In “Frontiers in Neuroendocrinology” (L. Martini and W. F. Ganong, ed.), pp. 209-236. Oxford Univ. Press, London and New York. Melamed, S., Portanova, R.,and Sayers, G. (1970.)J.Cell Biol. 47, p. 128a Abstr. No. 336. Matsuo, H., Baba, Y., Nair, R. M. G., Arimura, A., and Schally, A. V. (1971).Biochem. Biophys. Res. Commun. 43, 1134. Meites, J., Kahn, R. H., and Nicoll, C. S. (1961). Proc. Soc. E x p . Biol. Med. 108, 440-443. Meites, J., Nicoll, C. S., and Talwalker, P. K. (1963).Aduan. Neuroendocrinob Proc. Symp., Miami, 1961 pp. 238-277. Mendoza D., Arimura, A., and Schally, A. V. (1973). Endocrinology 92, 1153-1160. Mester, J., Brunelle, R.,Jung, I., and Sonnenchein, C. (1973).E x p . Cell Res. 81, 447. Mikami,-S. I., Hashikawa, T.. and Farner, D. S. (1973).2. Zellforsch. Mikrosk. Anat. 138,299-314. Moriarty, G. C. (1973).J . Histochem. Cytochem. 21,855-894. Morin, A. (1974).Thesis Dr. 3dcycle, Paris. Naik, D. V. (1973).2. Zellforsch. Mikrosk. Anat. 142,289-304. Nair, R. M., Barrett, J. F., Bowers, C. Y., and Schally, A. V. (1970). Biochemistry 9, 1103. Nakane, P. K. (1970).J . Histochem. Cytochem. 18,9-20. Nicoll, C. S., and Meites, J. (1963).Endocrinology 72,544-551. Nouet, J. C., and Kujas, M. (1973).2. Zellforsch. Mikrosk. Anat. 143,535-547. Ohtsuka, Y., Tshikawa, H., and Ohoto, T. (1971).Endocrinol. lap. 18,133. Ohtsuka, Y., Ishikawa, H., Watanabe, T., and Yoshimura, F. (1972).Endocrinol. Jar,. 19,237-249. Olivier, L., Vila-Porcile, E., Racadot, O., Peillon, F., and Racadot, J. (1974).In “The Anterior Pituitary Gland” (M. G. Farquhar and A. Tixier-Vidal, eds.). Academic Press, New York. In press.

238

A. TIXIER-VIDAL, D. GOURDJI AND C. TOUGARD

Orth, D. N., Nicholson, W. E., Mitchell, W. M., Island, D. P., Shapiro, M., and Byyny, R. L. (1973). Endocrinology 92,385-393. Pasteels, J. L. (1961).C . R. Acad. Sci. 253,3074-3075. Pasteels, J. L. (1963). Arch. Biol. 74,439-553. Pasteels, J. L. (1969). Mem. Acad. Roy. Med. Belg. 7, 1-45. Pasteels, J. L., Danguy, A., Frerotte, M., and Ectors, F. (1971). Ann. Endocrinol. 32, 188-192. Picart, R., and Tixier-Vidal, A. (1974).J . Microscop. 20,80a. Poirier, G., Labrie, F., Barden, N., and Lemaire, S. (1972) FEBS (Fed. Eur. Biochem. Soc.), Lett. 20,283-286. Portanova, R., Smith, D., and Sayers, G. (1970). Proc. SOC. E x p . Biol. Med. 133, 573-576. Posner, M., Chin, W., Weinberger, S., Palotta, J., and Sherwood, L. (1973). Endocrinology 92, Suppl., Abstr. No. 184. Pradelles, P., Morgdt, J. L., Fromageot, P., Oliver, C., Jacquet, P., Gourdji, D., and Tixier-Vidal, A. (1972). FEBS (Fed. Eur. Biochem. Soc.), Lett. 22, 19-21. Purves, H. D. (1966). In “The Pituitary Gland” (G. W. Harris and B. T. Donovan, eds.), Vol. 2, pp. 147-232. Butterworth, London. Rappay, G., Gyevai, A., Kondics, L., and Stark, E. (1973). In Vitro, 4,301-306. Rennels, E. G., Bogdanove, E. M., Arimura, A., Saito, M., and Schally, A. V. (1971). Endocrinology 88,1318-1326. Reusser, F., Smith, C. G., and Smith, C. L. (1962). Proc. SOC. Exp. Biol. Med. 109, 375-378. Samuels, H. H., and Tsai, J. S. (1973). Proc. Nut. Acad. Sci. U.S. 70, 3488-3492. Samuels, H. H., Tsai, J. S., and Cintron, R. (1973).Science 181, 1253-1256. St.tillb, G., and Nakane, P. K. (1972).Anut. Rec. 172,403. Shiino, M., Arimura, A. M., Schally, A. V., and Rennels, E. G. (1972). 2. Zellforsch. Mikrosk. Anut. 128,152-161. Sonnenschein, C., Richardson, I., and Tashjian, A. H., Jr. (1970). E x p . Cell Res. 61, 121-128. Sonnenschein, C., Posner, M., Saududdin, S., and Krasnay, S. (1973). E x p . Cell Res. 78,41-46. Stark, E., Gyevai, A., Szalay, K., and Posalaky, Z. (1965). J . Endocrinol. 31, 291-292. Steinberger, A., Chowdhury, M., and Steinberger, E. (1973). Endocrinology 92, 12-17. Stumpf, W. E., Sar, M., and Keefer, D. (1974). In “The Anterior Pituitary Gland” (M. G. Farquhar and A. Tixier-Vidal, eds.). Academic Press, New York. In press. Takemoto, H., Yokoro, K., Furth, J., and Cohen, A. I. (1962). Cancer Res. 22,917-924. Tashjian, A. H., Jr., and Hoyt, R. F., Jr. (1972). In “Molecular Genetics and Develop mental Biology” (M. Sussman, ed.), pp. 353-387. Prentice-Hall, Englewood Cliffs, New Jersey. Tashjian, A. H., Jr., Yasumura, Y., Levine, R., Sato, G. H., and Parker, M. L. (1968). Endocrinology 82,342-352. Tashjian, A. H., Jr., Bancroft, F. C., and Levine, L. (1970). J . Cell Biol. 47, 61-71. Tashjian, A. H., Jr., Barowsky, N. J., and Jensen, D. K. (1971). Biochem. Biophys. Res. Commun. 43,516-523. Tashjian, A. H., Jr., Hinkle, P. M., and Dannies, P. S. (1973). Proc. Int. Congr. Endocrinol.. 4th, Washington, D.C., I972 pp. 648-654. Tixier-Vidal, A. (1974). In “The Anterior Pituitary Gland” (M. G. Farquhar and A. Tixier-Vidal, eds.). Academic Press, New York. In press.

CELL CULTURE OF ANTERIOR PITUITARY CELLS

239

Tixier-Vidal, A., KerdelhuB, B., BBrault, A., and Jutisz, M. (1970).C.R. Acad. Sci., Ser. D 271,523-526. Tixier-Vidal, A., KerdelhuB, B., and Jutisz, M. (1973). Life Sci. 12, Part 1, 499-509. Tixier-Vidal, A., Tougard, C., KerdelhuB, B., and Jutisz, M. (1974). 5th Znt. Con5 Zmmunojluorescence Related Staining Tech. Ann. N.Y. Acad. Sci. In press. Tougard, C., and Tixier-Vidal, A. (1974).J . Physiol. 68, 11B. Tougard, C., KerdelhuB, B., Tixier-Vidal, A., and Jutisz, M. (1973).J . Cell Biol. 58, 503-521. Tougard, C . , Picart, R., Tixier-Vidal, A., KerdelhuB, B., and Jutisz, M. (1974). Znt. Symp. 2nd, Electron Microsc. Cytochem. Drienerlo, Neth. 1973 p. 163. Vale, W., Grant, G., Amoss, M., Blackwell, R., and Guillemin, R. (1972a). Endocrinology 91,562-572. Vale, W., Brazeau, P., Grant, G., Nussey, A., Burgus, R., Rivier, J.. Ling, N., and Guillemin, R. (1972b). C.R. Acad. Sci., Ser. D 275,2913-2916. Vale, W., Blackwell, R., Grant, G., and Guillemin, R.(1973).Endocrinology 93,2633. Vila Porcile, E. (1973).Ann. Sci. Natur. 2001.Biol. Anim. 15,61-138. Waelbroeck-Van Gaver, C., and Potvlihge, P. (1969).Eur. J . Cancer 5,99-117. Watanabe, H., Nicholson, W., and Orth, D. N. (1973a).Endocrinology 93,411. Watanabe, H., Orth, D., and Toft, D. 0. (1973b). J . Biol. Chem. 248, 7625-7630. Watanabe, H., Orth, D. N., and Toft, D. 0. (1974). Biochemistry 13,332-337. Yasumura, Y. (1968).Amer. 2001. 8,285-291. Yasumura, Y., Tashjian, A. H., Jr., and Sato, G. H. (1966). Science 154, 1186-1189. Yoshida, Y. (1966).Methods Achievements E x p . Pathol. 1, 439-454.

This Page Intentionally Left Blank

Immunohistochemical Demonstration of Neurophysin in the Hypothalamoneurohypophysial Systern W . B . WATKINS Postgraduate School of Obstetrics and Cynaecology. University of Auckland. Auckland. New Zealand

1. Introduction

.

. . . . . . . . . . . . . . . .. .

241

11 Relationship between Neurosecretory Material

and Neurophysin . . . . . . . . . . . . . . . I11. Methods of Extraction of Neurophysin . . . . . . . A . Extraction from Posterior Pituitary Glands . . . . B. Extraction from Whole Pituitary Glands . . . . . IV Purification of Neurophysin Antigens . . . . . . . A . Molecular Sieve and Ion-Exchange Chromatography B Preparative Polyacrylamide-Gel Electrophoresis . C . Preparative Isoelectric Focusing . . . . . . . . D . Extraction of Neurophysin from Electrophoretic Gels V. Production of Antibodies against Neurophysin A Cross-Species Reactive Antibodies . . . . . . . B Species-Specific Antibodies . . . . . . . . . VI . General Considerations of Antibody Production and Detection . . . . . . . . . . . . . . . . VII Immunohistochemical Techniques . . . . . . . . A . Light Microscope Level . . . . . . . . . . . B. Electron Microscope Level . . . . . . . C Photographic Procedures . . . . . . . D Preparation of Tissues . . . . . . . . . . . VIII . Demonstration of Neurophysin in the Hypothalamoneurohypophysial System Using Cross-Species Reactive Antineurophysin A Magnocellular Nuclei . . . . . . . . . . . . B. Other Areas of the Hypothalamus and Brain . . . C . Proximal Neurohypophysis . . . . . . . D . Pituitary Stalk . . . . . . . . . . . E . Posterior Pituitary Gland . . . . . . . . . . Ix. Use of Species-Specific Antisera for the Demonstration of Neurophysin . . . . . . . . . . A . Pig . . . . . . . . . . . . . . . B. Ox . . . . . . . . . . . . . . . x. Conclusions . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . .

.

.

. .

. .

. .

. . .

. . . .

. . . . . .

. .

.

. .

. . .

.

. .

. .

. .

. . . .

. . . . . .

. . . .

. .

. . . . .

I

.

. . . .

. .

. . . .

. .

243 244 244 246 246 246 249 249 250 250 250 254 255 256 256 257 257 257 260 260 266 269 275 277 279 279 280 280 281

Introduction

Osborne and Vincent (1900)were the first to report on the association of a high-molecular-weight material with the hormones in the 24 I

242

W. B. WATKINS

posterior pituitary gland. These workers demonstrated that treatment of a saline extract of the ox infundibulum, with either excess alcohol or by the addition of ammonium sulfate, resulted in precipitation of the pressor activity found in the tissue. In an extension of this work, Kamm et al. (1928)showed that both the oxytocic and pressor principals of the gland extract were salted out with 30% sodium chloride, but that the efficiency of the precipitation depended upon the dilution of the initial extract. By using relatively concentrated solutions, approximately 100% of the pressor activity was salted out and, furthermore, the ratio of pressor activity to oxytocic activity approximated that found in the fresh gland. Addition of sodium chloride to a dilute solution containing 10 IU of both oxytocic and pressor activities per milliliter gave incomplete precipitation of the biological activities. Rosenfeld (1940)subsequently found that the oxytocic and pressor activities sedimented in an ultracentrifuge with a sedimentation coefficient of about one-half to one-third that of egg albumin. Partial purification and biochemical characterization of the proteinaceous material isolated by Kamm et al. (1928)and Rosenfeld (1940)was begun by van Dyke et al. (1942).Employing the techniques available at that time, for example, constant solubility, electrophoresis, and differential centrifugation, it was concluded that the “van Dyke protein” was homogeneous and had a molecular weight of approximately 30,000. The protein is rich in residues of cystine, glutamic acid, and glycine as determined by amino acid analysis (Block and van Dyke, 1952).The finding that electrodialysis of the van Dyke protein can result in the dissociation of the oxytocic principal led Haselbach and Piquet (1952)to suggest that the polypeptide hormone is electrostatically bound to the proteinaceous component. Other procedures such as dialysis, ultrafiltration, electrolysis, treatment with trichloroacetic acid, and countercurrent distribution (Acher et al., 1955;Acher and Fromageot, 1957) further confirmed the noncovalent binding of hormone to protein. The protein component of the van Dyke protein was given the name neurophysin by Acher et al. (1955). When the van Dyke protein was initially isolated, it was considered a single protein, but more recent studies into the homogeneity of the protein by Frankland et al. (1966)established the presence of four discrete protein components. The presence of these multiple components has subsequently been attributed to the action of proteolytic enzymes during the extraction procedure (see Section I1I ,A,1).

IMMUNOHISTOCHEMISTRY OF NEUROPHYSIN

243

11. Relationship between Neurosecretory Material and Neurophysin

Histologically, the concept of neurosecretion is characterized by the presence of neurosecretory material (NSM) within the magnocellular hypothalamic nuclei and neurosecretory axons of the hypothalamoneurohypophysial system (HNS) (for reviews, see Bargmann, 1966; Bern and Knowles, 1966; Rinne, 1966; Sachs, 1969; Scharrer and Scharrer, 1954; Sloper, 1958). This material was first demonstrated by Bargmann (1949a,b, 1950, 1951) by the application of the acid permanganate chrome alum hematoxylin stain (CAH) which had been previously used by Gomori (1941) to stain the /3 cells of the endocrine pancreas. The basis of action of CAH and the more sensitive aldehyde-fuchsin (AF) (Gomori, 1950; Landing et al., 1956) is the oxidation of disulfide bonds in the tissue to sulfonic acid residues which then combine with hematoxylin and fuchsin residues, respectively. A more sensitive reagent reacting with sulfonic acid groups is pseudoisocyanin chloride which gives rise to an intense fluorescence (Sterba, 1964). NSM can also be demonstrated by using methods detecting protein-bound cystine such as performic acid-alcian blue (Adams and Sloper, 1956) and thioglycolate-ferric ferricyanide (Sloper, 1955). The thioglycolate-dihydroxy-dinaphthyl disulfide (Barnett, 1954; Sloper, 1955) and alkaline tetrazolium (Sloper, 1955; Howe and Pearse, 1956) techniques reveal NSM through reactions with protein-bound sulfhydryl groups formed by the reduction of disulfide bonds. Material containing arginine residues was detected in the infundibulum of the ox and rat by Howe (1959,1962), using a modification of the Sakaguchi reaction. An improved method for the detection of arginine in NSM was used by Bock and Schluter (1971). This method is based on the fact that arginine reacts with a derivative of ninhydrin to give a fluorescent compound which can be visualized in tissue slices (Rosselet, 1967). It was initially proposed (Schiebler, 1952) that NSM is a glycoprotein since it gives a positive reaction with periodic acid-Schiff, Millon, and sudanophilic lipid stains. However, later studies using the periodic acid-Schiff reaction gave inconclusive results (Gabe, 1960) which were attributed to species variation. It was implied that NSM is a lipid-protein because it was not present in sections fixed with lipid-dissolving solvents such as alcohol and acetone (Schiebler, 1951; Hild and Zetler, 1953). However, subsequent work

244

W. B. WATKINS

showed that, when these tissue sections were postfixed with formalin or floated on Bouin’s fluid instead of water, NSM was retained in the tissue (Sloper, 1955). Further studies on the dog (Sloper, 1955) and rat (Howe and Pearse, 1956) confirmed that NSM does not contain lipid or carbohydrate residues but is a protein rich in cystine. It is now being recognized that NSM present in the mdmmalian HNS and neurophysin, may be one and the same species. This correlation is based on the following observations.

1. The positive histological reactions performed on the HNS with CAH, AF, and pseudoisocyanin chloride stains can be considered to be due to the interaction with the large number of cystine residues present in neurophysin. It is interesting to note that neurohypophysial hormones, like neurophysin, contain approximately 16%cystine and are also capable of giving a positive reaction with pseudoisocyanin chloride (Gutierrez and Sloper, 1969). 2. The performic acid-pseudoisocyanin chloride reagent reacts with the neurophysin-hormone complex as prepared by the method of Acher, Light, and du Vigneaud (1958; also see Sterba, 1964). 3. Neurophysins obtained from the species so far studied contain between four and six arginine molecules per molecule of protein, which would be capable of reacting with the ninhydrin fluorescent system described by Bock and Schliiter (1971). 4. Osmotic stimulation of an animal reduces the amount of NSM in the posterior pituitary lobe concomitantly with neurophysin (Watkins and Evans, 1972) (see Section VII1,E). 111. Methods of Extraction of Neurophysin

A.

EXTRACTION FROM POSTERIOR PITUITARY GLANDS

1. Use of Dilute Acids In their initial preparation of the protein-neurohypophysial hormone complex, van Dyke et al. (1942)extracted fresh-frozen ox posterior pituitary glands with 0.01 N sulfuric acid overnight at 4°C. Subsequent fractionation of the van Dyke protein by gel exclusion and ion-exchange chromatography (Hollenberg and Hope, 1967) produced six discrete proteins as revealed by starch gel electrophoresis at pH 8.1. Two of the proteins possessed the ability to bind oxytocin and vasopressin, confirming their identity as neurophysins. Proteolytic enzymes present in acetone-dried posterior lobe

IMMUNOHISTOCHEMISTRY OF NEUROPHYSIN

245

powder have optimum activity at pH 3-4 using hemoglobin as a substrate (Dean et al., 1967; Pickup and Hope, 1971,1972). If the extraction of ox tissue is carried out at a pH that irreversibly destroys the catheptic acitvity, the number of neurophysin components subsequently identified is reduced (Hollenberg and Hope, 1968). Despite the degradation of neurophysins by the action of proteolytic enzymes, many workers have persisted in using dilute acid for the initial extraction procedure. Breslow and Abrash (1966), Fawcett et al. (1968), and Legros et al. (1969) extracted bovine neural lobes with 0.01 N sulfuric acid according to the method of Acher et al. (1958), while 0.1 M acetic acid was used to extract neurophysins from ox (Ginsburg and Ireland, 1965; Ginsburg et al., 1966) and pig (Ginsburg et al., 1966)posterior pituitary glands. Pickering (1968) extracted neurophysins from cod pituitaries with 0.05 N acetic acid. 2. Use of Strong Mineral Acids The degradation of neurophysins by the action of proteolytic enzymes during extraction with dilute acids is avoided by using 0.1 N hydrochloric acid for the initial extraction procedure. Neurophysins from ox (Rauch et al., 1969; Pligka et al., 1972; Breslow et al., 1971; Robinson et al., 1971), pig (Uttenthal and Hope, 1970), sheep (Watkins, 1972a, 1973a), rat (Burford and Moens, 1971; Coy and Wuu, 1972; Watkins, 1972b), guinea pig (Watkins and Ellis, 1973), human (Watkins, 1971), and dog (Watkins, unpublished results) have been isolated using 0.1 N hydrochloric acid. It can be argued that use of hydrochloric acid at a pH of 1.5-2 may cause chemical disruption of the “native” neurophysin molecule. There is however, some evidence to suggest that this phenomenon does not occur to any marked degree: (1) Proteins extracted from either fresh or acetone-dried posterior pituitary glands give a similar electrophoretic pattern on starch. gel irrespective of whether hydrochloric acid (pH 1.5) or electrophoresis buffer (pH 8.1) is used for their extraction (Hope and Uttenthal, 1969; Watkins, 1972b, 1973a; Watkins and Ellis, 1973). (2) Furthermore, ox neurophysins extracted with 0.1 N hydrochloric acid have an electrophoretic mobility on starch gel similar to those of proteins obtained by lysis of neurosecretory granules from bovine neurohypophysis (Dean et al., 1967). It is believed that the proteins present in the neurosecretory granules represent native molecules. (3)Pig neurophysins extracted with hydrochloric acid (Uttenthal and Hope, 1970) give amino acid analyses similar to those of pig neurophysins obtained by Wuu and SaEran (1969) and Cheng and Friesen (1971b) using mild extraction

246

W. B. WATKINS

procedures (see Section III,A,3). (4) All the neurophysins so far characterized possess alanine as the N-terminal residue. This is observed irrespective of the method used for extraction. 3. Extraction by the Percolation Method In their purification of porcine neurophysin, Wuu and SafFran (1969) extracted acetone-dried posterior pituitary lobe powder by percolation with a discontinuous gradient of water and acetic acid in ethanol. The fraction extracted with 70% aqueous ethanol containing 0.25 M acetic acid was then subjected to further purification by molecular sieve and ion-exchange chromatography (see Section IV). A similar extraction procedure was used by Martin et al. (1972) for the purification of bovine neurophysin. Neurophysins of the pig (Friesen and Astwood, 1967; Cheng and Friesen, 1971a,b) and human (Cheng and Friesen, 1972, 1973) were obtained by initially stirring acetone-dried neural lobes powders with 10 volumes of 40% acetone or ethanol followed by percolation with further acetone. Under these conditions neurophysin is extractable into the liquid phase (Martin et al., 1972). Addition of further organic solvent to a concentration of 90% (vlv) causes precipitation of the neurophysins.

B. EXTRACTION FROM WHOLEPITUITARYGLANDS

In his isolation of neurophysin from the cod, Pickering (1968)used whole pituitary glands as starting material. More recent work from this laboratory (W. B. Watkins and J. J. Evans, unpublished work) has shown that it is also possible to extract and purify sheep neurophysins from intact pituitary glands employing the standard methods developed for the isolation of neurophysins from neural lobes.

IV. Purification of Neurophysin Antigens CHROMATOGRAPHY A. MOLECULAR SIEVE AND ION-EXCHANGE 1. Method Used by Hope Neurophysin present in posterior pituitary lobe extracts (adjusted to pH 4) can be precipitated out of solution by the addition of sodium chloride to a final concentration of 10 gml100 ml (Acher et al., 1958). Dissociation of the neurohypophysial hormones, oxytocin and vaso-

IMMUNOHISTOCHEMISTRY OF NEUROPHYSIN

247

pressin, from the protein-hormone complex can be achieved by chromatography on a column of Sephadex G-25 using an acidic eluting solvent (Frankland et al., 1966). The chromatographic elution profile for such a column is shown in Fig. l a which represents the

voknrat.IRUII1M

-0c-w)

FIG. 1. Elution profiles obtained during column chromatographic purification of sheep neurophysin from the protein-neurohypophysial hormone complex. Circles, Ultraviolet absorption. (a) Chromatography on a column (2.2 x 146 cm) of Sephadex G-25 to separate oxytocin and vasopressin from high-molecular-weight material. The sample (103 mg) was applied to the column in 0.1 M formic acid and eluted with the same solvent at a rate of 25 ml per hour. Squares, Oxytocic activity; triangles, pressor activity. (b) Chromatography of the hormone-free protein (obtained from the void volume in Fig. la) (93 mg) on a column (2.2 x 146 cm) of Sephadex G-75. The column was eluted at a flow rate of 20 ml per hour to give albumin in peak A and a mixture of sheep neurophysin in peak B. (c) Purification of ovine neurophysin I11 from the material in peak B (b) by ion-exchange chromatography on a column (2.8 X 45 cm) of DEAE Sephadex A-50. The sample of neurophysins (93 mg) was applied to the column in tris-hydrochloric acid buffer (pH 8.1), and the column then eluted with tris-hydrochloric acid buffer (pH 8.1)with increasing concentrations of sodium chloride from 0 M to 0.3 M sodium chloride over a gradient of 700 ml. Open circles, sodium chloride concentration in milliequivalents per liter. The flow rate was 8 ml per hour. A large fraction of ovine neurophysin 111 (peak C) was separated from the neurophysin mixture, while peak D contained ovine neurophysins I and I1 and residual amounts of neurophysin 111. (d) Purification of neurophysins in peak D (Fig. lc) by chromatography on a column (2.2 x 19 cm) of C M Sephadex C-50. The neurophysins (33.9 mg) were applied to the column in 0.1 M acetic acid, and the column eluted with a gradient of sodium acetate buffer (pH 4.40 + 5.0; I = 0.1)over a total volume of 250 ml at a flow rate of 4 ml per hour. Open circles, pH. Peaks E, F, and G correspond to ovine neurophysins I, 11, and 111, respectively. (Adapted from Watkins, 1973a.)

248

W. B. WATKINS

first stage in the purification of sheep neurophysins (Watkins, 1973a). Although oxytocin and 8-arginine-vasopressin have similar molecular weights, the latter peptide is retarded further on Sephadex G-25 because of the electrostatic interaction between the carbonyl groups on the Sephadex and the basic residues on the peptide (Gelotte, 1960). The protein eluted in the void volume from Sephadex G-25 is rechromatographed on a column of Sephadex G-75 (Fig. lb). Of the two peaks that are resolved, the first consists mainly of serum albumin. The profile of the second peak, with its slowly rising leading edge and rapidly falling tailing edge is characteristic of the presence of neurophysin. During the purification of pig neurophysins, however, pig neurophysin I1 is eluted prior to pig neurophysins I and I11 because its molecular weight is substantially greater than that of either neurophysin I or I11 (Uttenthal and Hope, 1970). Subsequent purification of individual neurophysins is achieved by ion-exchange chromatography. The choice of type of ion-exchange resin to be used is best determined empirically. The three ox neurophysins are successfully resolved on a column of DEAE Sephadex A-50 (Rauch et al., 1969), while a column of CM-Sephadex C-50 is employed in the separation of pig neurophysin I from neurophysin I11 (Uttenthal and Hope, 1970). Because the relative distribution of sheep neurophysins I, 11, and I11 is approximately 1:1:5, it has been found convenient to use both anion- and cation-exchange columns for their isolation (Watkins, 1973a). The major proportion of neurophysin I11 is isolated by the use of a DEAE-Sephadex A-50 column (peak A in Fig. lc). A CM-Sephadex C-50 column is then used to separate neurophysin I from neurophysin I1 (Fig. Id). 2. Method Used by Saffran The percolate fraction containing pressor activity was applied to a column of Sephadex G-25, and the material eluted in the void volume rechromatographed on a column of Sephadex G-50. Fractionation of the major protein peak from this column was carried out on DEAE-cellulose from which pig neurophysin was isolated (Wuu and SafFran, 1969).

3. Method Used by Friesen Neurophysins isolated in H. G. Friesen’s laboratory have been extracted by methods similar to that described above (Section IV,A,2).

IMMUNOHISTOCHEMISTRY OF NEUROPHY SIN

249

In the purification of pig neurophysins, the percolate material was first passed through a DEAE-cellulose column and protein fractions subsequently resolved by chromatography on a Sephadex G-75 column (Cheng and Friesen, 1971b). A side fraction obtained during the purification of gonadotrophins was used as starting material for the isolation of human neurophysins (Cheng and Friesen, 1972). An extract of this fraction was applied to a Sephadex G-100 column, and the peak containing material immunologically cross-reactive against antineurophysin serum was purified by pH gradient chromatography on DEAE-cellulose. Two proteins with hormone-binding properties were obtained and named human neurophysins I and 11. 4. Method Used by Ginsburg The proteins present in the supernatant of Ginsburg and Ireland’s (1965) acid extract of ox neurohypophysis were freed from oxytocin and 8-arginine-vasopressin by gel exclusion chromatography on Sephadex G-25. Separation of the neurophysins from other highmolecular-weight material was carried out by ion-exchange chromatography on CM-cellulose.

5. Method Used by Pickering The hormone-free protein mixture extracted from cod pituitaries was passed through a Sephadex G-75 column, and the peak containing proteins with hormone-binding properties collected. Further purification of the neurophysin was performed on DEAE Sephadex (Pickering, 1968). B. PREPARATIVE POLYACRYLAMIDE GEL ELECTROPHORESIS Coy and Wuu (1971) demonstrated that a mixture of pig neurophysins could be purified by means of a preparative polyacrylamide gel electrophoresis system. This method was subsequently used for the isolation of three proteins from rat posterior pituitary glands (Coy and Wuu, 1972), which were later identified as neurophysins (Watkins, 1972b). C. PREPARATIVE ISOELECTFUC FOCUSING Bovine neurophysins I and I1 as obtained by isoelectric focusing, in a 0-40% sucrose gradient, were of purity comparable to that of

250

W. B. WATKINS

ion-exchange chromatography preparations (PliBka et al., 1972). Although there are current problems associated with the quantitative removal of proteins from the column matrix, further advances in the development of ampholites will undoubtedly make the technique of preparative isoelectric focusing invaluable for the separation of molecules with similar isoelectric points such as the neurophysins.

D. EXTRACTION OF NEUROPHYSIN FROM ELECTROPHORETIC GELS Large-scale column chromatographic separation of neurophysins can be achieved conveniently only with species in which it is practical to collect sufficient quantities of posterior pituitary gland tissue. When small amounts of tissue are available, soluble proteins can be resolved by either starch or polyacrylamide gel electrophoresis and neurophysin components tentatively identified by immunological techniques (Cheng and Friesen, 1971a; Ellis et al., 1972). Microgram quantities of neurophysin extracted from these gels can then be used for the production of antibodies (Vaitukaitus et al., 1971). V. Production of Antibodies against Neurophysin A. CROSS-SPECIES REACTIVE ANTIBODIES

1. Ox Neurophysin a. Crude Preparation of Neurophysins as Antigen. The first report of the raising of cross-species reactive antibodies to neurophysins was by Fawcett et al. (1968). These workers gave a total of 42 mg of a crude mixture of ox neurophysin to each of a series of rabbits over a period of 6 months. Antibodies collected after this period crossreacted with beef neurophysin and dog posterior pituitary lobe homogenates as demonstrated by microimmunodihsion on agar gels. However, the antisera also reacted with a high-molecular-weight protein fraction (fraction A), indicative of the presence of serum albumin in the initial neurophysin preparation. It must also be pointed out that the protein-hormone complex used by these workers was prepared according to the method of Acher and co-workers (1958), and therefore the neurophysin was subjected to partial degradation by tissue enzymes. Legros et al. (1969), using a total bovine neurophysin mixture as

IMMUNOHISTOCHEMISTRY OF NEUROPHYSIN

25 1

prepared by the method of Fawcett et a2. (1968), raised antibodies in rabbits according to the protocol in Table I. Their antisera crossreacted with components in human serum and extracts of rat neurohypophysis (Legros et al., 1971a,b). b. Purified Ox Neurophysin Antigen. The purified ox neurophysin preparation used by Martin et al. (1972) as a source of antigen was extracted by the method suggested by Wuu and Saffran (1969). Antibodies raised against ox neurophysins cross-reacted immunologically with extracts from the posterior pituitaries of humans, sheep, guinea pigs, and rats. It is worthy of note that nasal insufflation of acetone-dried posterior pituitary tissue (pituitary snuE), used in the treatment of diabetes insipidus, can result in the production of antibodies to the various protein constituents of the preparation (Pepys et al., 1965; Mahon et al., 1967). Using an antiserum raised against a purified mixture of ox neurophysins, Martin (1971) determined the antibody titer of the peripheral serum of patients who had been treated, for periods up to 45 years, with a commercial preparation of a mixture of bovine and porcine posterior pituitary lobe powder (Pitressin, Parke Davis). In view of the finding that aqueous Pitressin contains neurophysin at a level of approximately 10 pg/ml, it is therefore not unexpected that prolonged administration of the preparation elicits antibodies to neurophysins in humans. 2. Pig Neurophysin a. Purified Pig Neurophysin Antigens. The high degree of antigenicity of porcine neurophysin I1 is reflected in the relative ease with which this protein elicits the antibody response in rabbits. Four weeks after a single injection of this antigen, precipitating antibodies can be detected (see Section V,B,2). Multiple injections of porcine neurophysin I1 give rise to an antiserum which cross-reacts with neurophysins extracted from the posterior pituitary glands of rats (Norstrom et al., 1971; Watkins and Evans, 1972) and sheep (Livett and Parry, 1971; Watkins, 1973b). The cross-species reactivity of the antineurophysin serum was further demonstrated by its ability to form insoluble antibody-antigen complexes with neural lobe proteins obtained from a selection of domestic and exotic mammalian species (Ellis et al., 1972). In a similar study antibodies raised against peptide I1 (porcine neurophysin I) cross-reacted with homogenates from the neural lobes of several mammals (Cheng and Friesen, 1971a).

TABLE I METHODS USED Neurophysin antigen

FOR THE RAISING OF ANTIBODIES IN RABBITS AGAINST

Protocol'

MAMMALIAN NEUROPHYSINS Cross-reactivity

Reference

Bovine mixture

6 mg in Freund's complete adjuvant into toepads, 6 mg i.v. 1month later, 9 mg i.m. at second and third month, and 6 mg i.v. at fourth and sixth month; bled 5 days after each injection

Ox, dog

Fawcett et al. (1968)

Bovine mixture

Antigen in Freund's complete adjuvant injected twice a week for 3 weeks into toepads, followed by monthly i.p. injections

Man, ox, dog, rat, pig

Legros et al. (1969, 1971a,b)

Bovine mixture

1 mg antigen in Freund's complete adjuvant injected into neck every 2 weeks for 6 months; bled 10-14 days after each injection

Ox, man, pig, sheep, rat, guinea pig

Martin et al. (1972)

Bovine I

1 mg antigen in Freund's complete adjuvant injected into toepads and back weekly for six weeks and then monthly for six times; bled 1 week after each monthly injection

Bovine I, monkey

Robinson et al. (1971); Zimmerman et al. (197313)

Bovine I1

As for bovine I

Bovine I1

Robinson et al. (1971)

Porcine mixture

20 mg antigen in Freund's complete adjuvant injected into back each fortnight for three times and boosters given at intervals of 3-4 months

Pig

Ginsburg and Jayasena (1968)

Rat, rabbit, guinea pig, dog, sheep, ox, monkey, man

Cheng and Friesen (1971a)

E;; N

9 m 4

Porcine I 2 mg of either antigen in Freund's adjuvant injected weekly (peptide 11) on four occasions into dorsal surface, followed by an i.v. and 11 injection of 2 mg; blood collected 10 days later (peptide 111)

v)

Porcine I1

1.5 mg of neurophysin in Freund's complete adjuvant injected S.C. twice at fortnight intervals; blood collected 1 month after the final injection

Porcine I

0.5 mg neurophysin in Freund's adjuvant injected S.C. into a single site; 5 weeks later 0.5 mg antigen bound to

Giraffe, muntjac deer, coatimundi, pig, hippopotamus, ox, sheep, fallow deer, dog, cat, mouse, rat, guinea pig, dama wallaby, pig-tailed macaque, woolly monkey, chacma baboon, man, rabbit, hedgehog

Watkins and Evans (1972)

Ellis et al. (1972)

Porcine I, 1% against porcine I1

polymethylmethacrylate particles injected i.v.; blood collected 1 week later Porcine I1

As for porcine I

Porcine 11, 0.1% against porcine N-I

Pickup et al.

(1973);

z

Livett et al.

(1971)

Ovine 111

1.5 mg antigen in Freund's complete adjuvant injected S.C. twice at fortnightly intervals; blood collected 1 month

Warthog

Uttenthaland Hope (1972)

Sheep, rat

Watkins and Evans (1972)

after final injection Human I

2 mg of neurophysin suspended in Freund's adjuvant injected weekly for four times into back, followed by i.v. booster of 2 mg of antigen; blood collected 10 days later

Human I and I1

3c z

zce

U

E

Cheng and Friesen (1973) Q .t

"

i.v., Intravenously; i.m., intramuscularly; i.p., intraperitoneally; s.c., subcutaneously.

w

u(

W. B. WATKINS

254

3. Sheep Neurophysins a. Purified Sheep Neurophysin Antigen. Antisera raised in rabbits by using multiple doses of the major sheep neurophysin, neurophysin 111, as antigen cross-reacted with neurophysins from the rat (Watkins and Evans, 1972) and guinea pig (Evans and Watkins,

1973). 4. Human Neurophysins a. Purified Human Neurophysin Antigen. In the development of a homologous radioimmunoassay for human neurophysin, Cheng and Friesen (1973) used purified human neurophysins I and I1 as antigens for antibody production. Both antisera cross-reacted with each antigen. ANTIBODIES B. SPECIES-SPECIFIC

1. Ox Neurophysin A radioimmunoassay specific for bovine neurophysins I and I1 has recently been developed by Robinson et al. (1971).The specific antisera were raised by giving six weekly injections of the antigen into the back and footpads of rabbits. Although there was minimal crossreactivity of antineurophysin I serum with neurophysin 11, the antiserum cross-reacted with human and monkey neurophysin (Robinson and Zimmerman, 1973). Earlier work by Livett et al. (1971) failed to produce precipitating antibodies against either bovine neurophysin I or I1 after the use of various injection regimes.

2. Pig Neurophysin Antibodies produced against the pig neurophysin isolated by Ginsburg and Jayasena (1968) were found to be species-specific and did not cross-react with proteins extracted from the rat, ox, or guinea pig neurohypophysis. In view of the massive doses required for antibody production, and the method used for preparation of the antigen, it is questionable whether the antibodies raised were in fact against pig neurophysin. Furthermore, these workers also obtained cross-reactivity against extracts of kidney, uterus, and mammary gland. The presence of neurophysin in these tissues has not been confirmed by other workers (Livett et al., 1971; Robinson et al., 1971). If 0.5 mg of purified pig neurophysin I1 is injected into a single site in rabbits, the antisera collected after 1 month cross-react specifically with the antigen as determined by microimmunodihsion on

IMMUNOHISTOCHEMISTRY OF NEUROPHYSIN

255

agarose (W. B. Watkins, unpublished results). A week later the titer of the antiserum can be increased by an intravenous injection of the neurophysin bound to polymethylmethacrylate particles (Livett et al., 1971). The warthog, which has the same phylogeny (suborder Suiformes) as the domestic pig, is the only other species so far studied that has a neurophysin possessing immunological determinants sufficiently similar to those in porcine neurophysin I1 to form a precipitin line with this antiserum (Uttenthal and Hope, 1972). Using the injection regime developed for the production of specific antibodies against porcine neurophysin 11, Pickup and colleagues (1973) obtained an antiserum against porcine neurophysin I with only 1% cross-reactivity against porcine neurophysin 11.

VI. General Considerations of Antibody Production and Detection The success of any protocol employed in the raising of antibodies is only as good as the techniques available for detection of the antibodies. The most convenient method for determining the presence of many antibodies is by microimmunodihsion or microimmunoelectrophoresis on a supporting matrix such as agar, agarose, or cellulose acetate (Ouchterlony, 1968).This method, however, is suitable only for those antibodies that form insoluble complexes with the antigens. During an extensive study carried out in our laboratory on the raising of antibodies to neurophysins, it became apparent that not all the antibodies produced give immunoprecipitates with the antigen. In fact, we showed (H. K. Ellis and W. B. Watkins, unpublished work) that an antiserum that did not give a precipitin line possessed a higher titer (as determined by radioimmunoassay) than another antiserum that gives a strong precipitin line in the Ouchterlony system. The most sensitive method for antibody detection is undoubtedly the displacement of iodinated neurophysin by the addition of “cold” antigen. The failure of Livett et al. (1971) to raise antibodies against ox neurophysins probably lay in their inability to detect such species. This is supported by the finding that the antibodies to bovine neurophysin I and I1 (Robinson et al., 1971) did not readily give precipitin lines with the corresponding antigen (A. G. Robinson, personal communication). In order to raise the titer of circulating antibodies, Cheng and Friesen (1971a)gave a booster intravenous injection a week prior to the bleeding of the animal. Since intravenously administered neuro-

256

W. B. WATKINS

physin is rapidly cleared by the kidney and has a half-life of only approximately 3 minutes (Forsling et al., 1973), the effect of such a booster may be expected to be minimal.

VII. Immunohistochemical Techniques A. LIGHT MICROSCOPELEVEL

1. lmmunofluorescence Histochemisty The technique of immunofluorescence histology is a powerful tool in the detection of trace amounts of tissue antigen (for a review, see Nairn, 1969). Two approaches are available for fluorescent protein tracing. (1) The direct method involves the conjugation of antineurophysin serum with a fluorochrome such as l-dimethylaminonapthalene-5-sulfonic acid or fluorescein isothiocyanate. This conjugate is then applied to the tissue section, and the fluorescence observed using a fluorescence microscope. (2) The difficulties often encountered in conjugation of the antisera to the fluorochrome can be overcome by using the indirect or “sandwich technique.” The tissue containing neurophysin antigens is first treated with antineurophysin serum which acts as the middle layer of the “sandwich.” After 30 minutes excess antiserum is washed from the tissue, and then commercially available sheep antirabbit yglobulin coupled to fluorescein isothiocyanate is applied (at a dilution of one-fifth to one-tenth) to the section. Excess stain is removed after 30 minutes, and after the sections are mounted in a glycerolsaline mixture (Kawamura, 1969) they are ready for observation. Areas containing neurophysin appear yellowish green against a paleblue background. Serum components, which often give rise to nonspecific absorption onto the tissue, can often be eliminated by treatment of the antiserum with acetone-dried liver powders. The specificity of the antiserum reaction is confirmed using preimmune serum and fluorochrome alone. 2. lmmunoperoxidase Histochemistry The immunoperoxidase technique described by Nakane (1968)has been used, with several modifications by Zimmerman et al. (1973b), for the cellular localization of neurophysin. In the two-layer method

IMMUNOHISTOCHEMISTRY O F NEUROPHYSIN

257

the tissue is first treated with antineurophysin serum followed by horseradish peroxidase-labeled sheep antirabbit yglobulin. Treatment of the peroxidase conjugate with 3,3’-diaminobenzidine in the presence of hydrogen peroxide gives a brown deposit representing the position of neurophysin. With the three-layer technique the tissue is treated sequentially with the following reactants: antineurophysin serum, sheep antirabbit y-globulin, rabbit antiperoxidase, peroxidase, and finally 3,3’-diaminobenzidine in hydrogen peroxide. In contrast to immunofluorescence histochemistry, immunoperoxidase sections are permanent and can be counterstained.

B. ELECTRONMICROSCOPE LEVEL Although the immunoperoxidase method has been successfully applied to the detection of tissue antigens in ultrathin sections (Kawarai and Nakane, 1970; Nakane, 1971), the localization of neurophysin in the HNS has not been reported. C. PHOTOGRAPHICPROCEDURES The fluorescent micrographs reported from this laboratory were all taken through a Leitz Laborlux fluorescent microscope fitted with UG-1 and BG-38 primary filters and a K430 secondary filter. Photographs were taken with either Kodak High Speed Ektachrome at exposures of approximately 2-3 minutes and developed at 160 ASA, or Kodak Tri-X film exposed at 10-15 seconds. A major disadvantage of the immunofluorescence technique is the difficulty in recording, on photographic emulsion, areas of weak fluorescence. This problem is not encountered when the immunoperoxidase procedure is used.

D. PREPARATION

OF

TISSUES

1. Cyostat Sections Neurophysin was first demonstrated by immunofluorescence histochemistry on fresh-frozen tissue of pig hypothalamus and posterior pituitary gland (Livett et al., 1971). Subsequent work from this laboratory on the distribution of neurophysin in the rat (Watkins and Evans, 1972) and guinea pig (Evans and Watkins, 1973) was carried out on cryostat sections from fresh-frozen tissues. The freezing of tissues by immersion in a slurry of dry ice-acetone often results in

258

W. B. WATKINS

FIG.2. Immunofluorescence localization of neurophysin-containing cells close to the third ventricle of the guinea pig hypothalamus. (a) Transverse sections were cut from snap-frozen tissues. (b) Hypothalamic tissue was fixed in saline-formalin prior to paraffin embedding. The tissues were stained with rabbit antiovine neurophysin 111 and fluorescein-labeled sheep antirabbit yglobulin. x 170. (From Watkins and Evans, 1974.)

IMMUNOHISTOCHEMISTRY OF NEUROPHYSIN

259

damage due to ice crystal formation. These artifacts are more likely to occur in the freezing of large blocks of tissue such as hypothalamus from large animals. It is our experience that even faster freezing in liquid nitrogen causes the tissue to become brittle and fragment. Immunofluorescence staining of neurophysin in cryostat sections of hypothalamus often appears diffuse and lacks discrete morphological integrity (Fig. 2a). Postfixation of cryostat sections with 86-92% ethanol does not destroy the antigenicity of neurophysin antigens (Livett et al., 1971; Watkins and Evans, 1972; Evans and Watkins, 1973). Livett et al. (1971), however, found that fixation in formalin-saline, Bouin’s fluid, chloroform-methanol (2:l), and 70 and 100% ethanol gave unsatisfactory results. 2. Paraffin-Embedded Tissues Livett and Parry (1971) reported the successful use of 95% ethanol-fixed tissues embedded in paraffin for immunofluorescent localization of sheep neurophysin. Fixation of tissue in either salineformalin (pH 7.2) or Bouin’s reagent (pH 1.8) prior to embedding in paraffin dramatically increases the resolution and histological detail of the neurophysin-containing structures as compared with that of fresh-frozen tissues (Fig. 2b) (Watkins and Evans, 1974). Immunoperoxidase techniques are equally successful when applied to sections fixed in saline-formalin or Bouin’s fluid (Zimmerman et al., 1973b).There is now a growing realization that many proteins retain their immunoreactivity after treatment with formalin. Beck et al. (196913) demonstrated the presence of human placental lactogen by immunofluorescence in formalin-fixed placentas, and Pasteels et al, (1972) reported excellent results when studying growth hormone and prolactin by immunofluorescence in formalin-treated pituitary tissues that had been stored for extended periods in paraffin blocks. These findings are in contrast to Pearse’s (1961) statement: “Some antigens can withstand prolonged formalin fixation and remain antigenic enough to be demonstrated in routine paraffin embedded sections. There are exceptions, however, and for most work short precipitation of the protein in the section with ethanol is all that is required.” One advantage of the fluorescence technique over the immunoperoxidase system is that staining for NSM with A F can be performed on the same slide after the immunohistochemical reaction has been carried out.

260

W. B. WATKINS

VIII. Demonstration of Neurophysin in the Hypothalamoneurohypophysial System Using Cross-Species Reactive Antineurophysin

A. MAGNOCELLULAR NUCLEI

1. Normal Animals Paraventricular (PVN) and supraoptic (SON) nuclei have generally been recognized to be associated with the synthesis of oxytocin and vasopressin, respectively (Olivecrona, 1957; Aulsebrook and Holland, 1969; Sokol, 1970; Bisset et al., 1971). Sokol and Valtin (1967) presented evidence for synthesis of neurohypophysial hormones within separate neurons of the rat. Later it was shown that NSM was still present in the posterior pituitary lobes of Brattleboro rats after destruction of the PVN, indicating a limited synthesis of oxytocin in the SON (Sokol, 1970). Data available on the vasopressinloxytocin ratio in the PVN and SON (Lederis, 1962; Bisset et al., 1973) indicate that the SON are more specific for the production of vasopressin than the PVN are for the production of oxytocin. This has been demonstrated in the guinea pig (Tindal et al., 1968) and cat (Bisset et al., 1967, 1970, 1971), in which electrical stimulation of the hypothalamic nuclei caused the release of both oxytocin and vasopressin from the PVN and only vasopressin from the SON. Since neurophysins are closely associated with the synthesis, storage, and release of neurohypophysial hormones (for reviews, see Ginsburg, 1968; Sachs, 1969),it is expected that neurons in the brain containing the hormones also represent sites of neurophysin. Vice versa, the detection of neurophysinlike antigens, by sensitive immunohistochemical techniques, offers a method for the identification of the neuronal elements that should contain oxytocin and/or vasopressin. The apparent specificity of the SON for the production of vasopressin might be reflected in their association with a specific neurophysin (see Section IX,A,B). The concept of one neurophysin being specifically associated with one hormone was proposed when it was found that the ox posterior pituitary gland contains two major neurophysins present in ratios approximating the ratios of the hormones (Hollenberg and Hope, 1968). Furthermore, evidence has been presented for the storage of oxytocin with neurophysin I and of vasopressin with neurophysin I1 in separate ox neurosecretory granules (Dean et al., 1968a). The pig also has two major neurophysins (Uttenthal and Hope, 1970), each of which may be as-

IMMUNOHISTOCHEMISTRY OF NEUROPHYSIN

26 1

FIG.3. Immunofluorescence in the cell bodies and processes of sheep PVN (a) and SON (b). The sections were taken from saline-formalin-fixed tissues, and the immunofluorescence carried out by the “sandwich” technique using antiporcine neurophysin I1 serum. x 170. (From Watkins, 1973b.)

262

W. B. WATKINS

sociated with one or the other of the neurohypophysial hormones (Pickup et al., 1973).Both of the two major neurophysins in the rat also appear to have functional relationships with the hormones (Burford et al., 1971). More recent studies on the relative proportion of neurophysins present in other species (Ellis et al., 1972) suggest that the simple stoichiometric relationship between molar ratios of oxytocin and vasopressin and of the two major neurophysins found in the ox (Dean et al., 196813)and rat (Burford et al., 1971)may not be common among all mammals. Magnocellular nuclei can readily be identified by the intense fluorescent reaction obtained with acridine orange (Bertalanffy, 1960; Evans and Watkins, 1973) as a result of its interaction with cellular DNA and RNA. Intense immunofluorescence in the magnocellular cells of the PVN and SON of the rat (Watkins and Evans, 1972), dog (Alvarez-Buylla et al., 1973), guinea pig (Evans and Watkins, 1973), and sheep (Watkins, 1973b)(Fig. 3a and b), indicates the presence of neurophysinlike proteins. Occasionally, sections of nuclei were obtained that possessed a distinct speckled appearance (Fig. 4). Resolution of the cell nuclei and processes emanating from the perikaryon of the sheep SON (Fig. 3b) are clearly seen after formalin fixation of

FIG.4. Speckled nature of the structures containing neurophysin in guinea pig PVN as revealed by immunofluorescence histology. Cryostat section from a freshfrozen tissue block. x 170.

FIG.5. Transverse section of a rhesus monkey hypothalamus treated by the threelayer immunoperoxidase bridge method for the demonstration of neurophysin using antibody to bovine neurophysin I. Neurophysin is present in the PVN and SON concentrated near blood vessels (arrows) and in neurons of the paraventricular tract (PVT) optic tract (OT), and third ventricle (111). X42. (From Zimmerman et al., 1973b.)

FIG.6, Higher magnification of the PVN in Fig. 5. Note the varying amounts of immunoreactive neurophysin present in the magnocellular nuclei. x575. (From Zimmerman et al., 197313).

264

W. B. WATKINS

the tissue. The antisera used for these studies cross-reacted with each of neurophysins present in the various animals, Neurophysin was also demonstrated in the cells of monkey PVN and SON by Zimmerman et al. (1973b) using the three-layer immunoperoxidase bridge method (Figs. 5 and 6) in association with an antiserum raised against bovine neurophysin I. Beaded segments of nerve fibers can be seen leaving the PVN and SON (Fig. 6). Areas around several blood vessels also stain for neurophysin. Certain magnocellular cells in both PVN and SON of ox and monkey were found to contain only small amounts of neurophysin, while a few others were almost completely devoid of immunoreactive protein (Zimmerman et al., 1973b). The significance of these cells, which have also been observed in the guinea pig (Watkins and Evans, 1974), is not known. Do they represent cells with an inherent reduced level of protein synthesis? Or are they associated with highly active secretion of their contents? Could they be cells in a “resting phase”?

2. Dehydrated Animals In our studies on the distribution of neurophysin in rats (Watkins and Evans, 1972) and guinea pigs (Evans and Watkins, 1973), the intensity of neuophysin immunofluorescence in the PVN and SON of normal animals did not differ markedly from that observed in an animal deprived of water for extended periods. The effect of osmotic stimulation on the content of NSM and neurohypophysial hormones in the hypothalamus appears to be in conflict, Dehydration of dogs (Ortmann, 1951; Hild and Zetler, 1953) causes depletion of NSM in the magnocellular nuclei. These findings are in contrast with those reported by Andersson and Jewel1 (1957), who showed that in excessively hydrated dogs a high proportion of the cell bodies in the PVN and SON are also depleted of NSM. Fendler et al. (1968) failed to observe a depletion of NSM in the rat magnocellular system after water deprivation, nor was Diamond (1956) able to demonstrate a loss of antidiuretic hormone activity in dehydrated rat hypothalamus. In rats dehydrated for 10 days (Vilhardt, 1970), there was approximately an 80% reduction in vasopressin content of the hypothalamus. However, Cheng et al. (1972) measured a fivefold increase in neurophysin content of the PVN after dehydration. There was no significant increase in the neurophysin content of the PVN. Dehydration of rats has been shown to have the effect of increasing the in vivo uptake of cystineand cysteine-=S into the PVN and SON (Wells, 1963; Talanti,

IMMUNOHISTOCHEMISTRY OF NEUROPHYSIN

265

1971).There is also an increase in the amount of c ~ s t e i n e - ~incor~S porated into vasopressin during the in uitro incubation of hypothalamic slices obtained from dehydrated guinea pigs 4 days (Takabatake and Sachs, 1964). After an 8-day dehydration, however, the incorporation of radioactive label was less than normal. The inability to observe a marked change in the intensity of neurophysin immunofluorescence in the main body of the PVN and SON on dehydration might be due in part to the nature of the immunologically active components present in the cytoplasm of the perikaryon. Since the chemical nature of neurophysin antigens in the cytoplasm is unknown, a direct correlation with the neurophysin ex-

b

C

d

FIG. 7. Diagrammatic representation of transverse sections of the guinea pig hypothalamus showing areas of cells (represented by hatching) which contain neurophysin demonstrated by immunofluorescence histochemistry using antiserum raised against porcine neurophysin 11. OC, Optic chiasma; SOC, supraoptic complex; ASO, accessory supraoptic nucleus; SO, supraoptic nucleus; PV, paraventricular nucleus; APV, accessory paraventricular nucleus; OT, optic tract; 111, third ventricle; F, fomix. (a) and (e) represent the most rostra1 and caudal positions, respectively. Tracts of neurosecretory fibers are also represented in (c) and (d). (From Evans and Watkins, 1973.)

266

W. B. WATKINS

tracted from the neural lobe may not be valid. It has been our experience in guinea pigs, however, that immunoflurorescent cells situated away from the main areas of the PVN and SON (see Section VIII,B,l) are more easily visualized in the dehydrated animal. Beaded fibers also become more apparent in the stimulated animal. These findings are suggestive of increased protein biosynthesis as a result of osmotic stress. B. OTHER AREAS OF

THE

HYPOTHALAMUS AND BRAIN

1. Normal Animals In a recent study we cut transverse sections of guinea pig hypothalamus from extreme rostral to caudal positions and observed those regions outside the main area of the PVN and SON that contained neurophysin (Evans and Watkins, 1973). Neurophysinlike material was seen at the extreme rostral part of the hypothalamus and in a few cells positioned ventral and closely dorsal to the third ventricle (Fig. 7a). Advancing caudally, the cells of the SON became apparent. The anterior portion of the SON observed in several animals had a fi-

FIG. 8. Beaded fibers in the anterior SON of the dehydrated guinea pig. Neurophysin is demonstrated using antiovine neurophysin I11 serum in association with fluorescein-labeled sheep antirabbit y-globulin. Section from fresh-frozen tissue. X 170. (From Evans and Watkins, 1973.)

IMMUNOHISTOCHEMISTRY OF NEUROPHYSIN

267

FIG.9. Specific neurophysin immunofluorescence in cells of the guinea pig. (a) Supraoptic complex. (b) Posterior SON in a position medial to the optic tract. (c) Anterioventral portion of PVN. Sections were from snap-frozen tissue. X130. (From Evans and Watkins, 1973.)

268

W. B. WATKINS

FIG.10. Immunofhorescent staining of cells extending from the third ventricle to an area dorsal to the fomix (a), of accessory PVN (b) in a position ventral to the accessory PVN and between the fornix and optic tract of the guinea pig (c).The tissue was snapfrozen prior to sectioning. x 130. (From Evans and Watkins, 1973.)

IMMUNOHISTOCHEMISTRY OF NEUROPHY SIN

269

brous appearance due to the presence of bundles of nerve fibers (Fig. 8). Other cells lying dorsal to the optic chiasma between the SON and the third ventricle (Figs. 7b and 9a) often gave rise to immunofluorescence and corresponded to the “supraoptic complex” described in the rabbit by Ford and Kantounis (1957). The posterior SON medial to the optic track also gave rise to fluorescence (Figs. 7e and 9b). In the region of the accessory SON (Bandaranayake, 1971) (Fig. 7b), a few fluorescent cells were seen. Neurophysin-containing cells of the anterioventral portion of the PVN (Fig. 7c) often stained in a diffuse manner, as shown in Fig. 9c. Fibers spread laterally from the PVN and around the fornix in a similar manner, as demonstrated in the vole (Clarke and Kennedy, 1967) and guinea pig (Knaggs et al., 1971) by using A F stain. Proceeding further posteriorly into the hypothalamus, immunofluorescent cells were seen extending from an area close to the third ventricle to a position just dorsal to the fornix (Figs. 7d and 10a); the group of cells regarded as the accessory paraventricular nucleus (Fig. 7d) were also immunoreactive against antineurophysin (Fig. lob). A few cells ventral to the accessory paraventricular nucleus between the fornix and the optic tract were sometimes seen to contain neurophysin (Fig. 1Oc). 2. Neurophysin in the Brain of the Scrapie Sheep In sheep affected with natural scrapie there is degeneration of the mossy terminals within the granular layer of the cerebellar cortex (Beck et al., 1969a). Livett and Parry (1971) observed neurophysinspecific immunofluorescence in this part of the cerebellum, and also in an area of the third cranial nerve. It is uncertain whether this material is identical to the neurophysin present in the magnocellular nuclei or even how the neurophysinlike protein reaches the cerebellum. Does the protein result from local synthesis in the hindbrain by neurons with synthetic capabilities similar to those of the PVN or SON? Or does it accumulate as a result of axonal transport from the hypothalamic nuclei? The loss of neurons in the PVN and SON of scrapie sheep (Beck et al., 1964) is associated with reduced levels of neurophysin as demonstrated by immunofluorescence histology (Livett and Parry, 1971). C. PROXIMALNEUROHYPOPHYSIS

1. Median Eminence and Internal Infundibulum The hypothalamo-distal neurohypophysial system originates in the magnocellular cells of the hypothalamus. This neurosecretory path-

270

W. B. WATKINS

FIG.11. Coronal sections of sheep median eminence. Neurosecretory fibers containing neurophysin emanating from cells of the supraoptic nucleus (a) and proceeding down the median eminence (b). Herring bodies are marked with arrows. ~ 1 7 0 . (From Watkins, 1973b.)

way, which is also referred to as the supraopticoneurohypophysial system, consists of nerve fibers which descend from the hypothalamus via the internal zone of the infundibulum and terminate in the infundibular process (posterior pituitary gland). It is by this main neurosecretory pathway that the neurohypophysial hormones, together with neurophysin, enter the posterior pituitary lobe. Nerve fibers in the hypothalamus leaving the SON readily stain for neurophysin (Fig. lla). Herring bodies associated with these axons become more apparent as the fibers enter the median eminence (Fig. llb).

IMMUNOHISTOCHEMISTRY OF NEUROPHYSIN

271

FIG.12. (a) Neurophysin present on the ventricular surface of the monkey third ventricle and in the cytoplasm of tanyacte perikarya as demonstrated by the immunoperoxidase technique. x504. (From Robinson and Zimmerman, 1973.) (b) Immunofluorescence localization of neurophysin in tanyacte processes running laterally between the third ventricle (111) and the supraopticohypophysial tract (soh) of the sheep. X 170. (From Watkins, 1973b.)

272

W. B. WATKINS

Kobayashi and his colleagues (1970) demonstrated the uptake of horseradish peroxidase into ependymal cells of the third ventricle when the tracer was injected into the ventricular spaces of the rat and quail. Since neurophysin has a molecular weight of about onequarter that of horseradish peroxidase, it too would be expected to be absorbed from the cerebrospinal fluid. Using the Gomori stain, Talanti and Kivalo (1961) showed the presence of NSM in the ependymal cellular layer of the infundibular recess and extending to the third ventricle of the camel. This work was extended by Robinson and Zimmerman (1973) in the monkey (Fig 12a) and by Watkins (197313) in the sheep (Fig. 12b), and confirmed the presence of neurophysin in the cytoplasm of tanyacte perikarya and in tanyacte processes running perpendicular to the ventricle surface and lateral to the supraopticoneurohypophysial secretory system. The role of the tanyactes is still in question, but several workers (Anand Kumar and Knowles, 1967; Knowles and Anand Kumar, 1969; Porter et al., 1970) have suggested that they may play an endocrine role in providing a pathway from the cerebrospinal fluid to the portal system of the hypothalamus. In fact, Knowles and Anand Kumar (1969) traced tanyacte processes from the third ventricle to positions abutting onto the pericapillary spaces of the pituitary portal plexus blood vessels. These terminal processes are part of the external palisade layer of the median eminence-a region directly involved in the release of hypothalamic releasing factors (see Section VIII,C,2). 2. External Znfundibulum Nerve fibers of the hypothalamo-proximal neurohypophysial system originate mainly in the infundibular (acuate) nucleus of the hypothalamus. The tuberoinfundibular tract initially proceeds along the internal zone of the infundibulum, together with the supraopticoneurohypophysial tract, and then turns into the external zone of the median eminence where the fibers terminate close to the mantle capillary plexus. Under normal conditions the external infundibulum gives a negative reaction for NSM using histochemical stains. However, there is an increase in the NSM following bilateral adrenalectomy and hypophysectomy (Bock and Muhlen, 1968; Bock et al., 1969; Bock and Forstner, 1969; Brinkmann and Bock, 1970). This accumulation of Gomori-positive granules can be reduced in the hypophysectomized animal by administration of adrenocorticotrophin hormone, and in the adrenalectomized animal by the application of corticosteroids (Bock et al., 1969), which implies that the amount of NSM in the external infundibulum is directly

IMMUNOHISTOCHEMISTRY OF NEUROPHYSIN

273

FIG. 13. Sagittal sections of sheep infundibulum stained for neurosecretory material and neurophysin. (a) AF stain for neurosecretory material in the infundibulum. ~250.(b) Immunofluorescence of neurophysin present in the external zone (ze) and internal zone (zi). pt, Pars tuberalis; ir, infundibular recess. X170. (From Watkins, 1973b.)

274

W. B. WATKINS

related to the presence of corticotrophin-releasing factor (CRF). While the external zone of the sheep infundibulum is devoid of NSM as demonstrated by AF (Fig. 13a), specific neurophysin immunofluorescence extends from an area near the infundibular recess to the external infundibulum (Fig. 13b) (Watkins, 1973b). In sheep affected with scrapie, there is an estimated twofold increase in the intensity of the immunofluorescence in the external infundibulum as compared with that in the normal animal (Livett and Parry, 1973; Parry and Livett, 1973). The presence of neurophysin in both regions of the infundibulum has been confirmed in the normal monkey by Zimmerman et al. (1973a) using the immunoperoxidase method (Fig. 14). These workers also found that the high concentration of immunoreactive neurophysin present around the monkey portal capillaries of the median eminence was reflected in the portal vessel levels of neurophysin being approximately 25 times greater than that in the systemic circulation. With the now confirmed presence of neurophysinlike molecules in the external infundibulum of normal animals, it is tempting at this stage to speculate upon a functional significance for the proteins.

FIG.14. Neurophysin concentrated around the portal capillaries in the external zone of the median eminence (ze) of the monkey as revealed by immunoperoxidase techniques. X504. (From Zimmerman et al., 1973b.)

IMMUNOHISTOCHEMISTRY OF NEUROPHYSIN

275

Although CRF has so far eluded isolation and biochemical characterization, it is believed to possess a structure similar to that of vasopressin (Schally et al., 1962; Schally and Bowers, 1964). It is therefore conceivable that the synthesis of CRF may also be associated with the elaboration of higher-molecular-weight protein. This protein may be sufficiently similar to neurophysin to be able to cross-react immunologically with antineurophysin serum.

D. PITUITARYSTALK The neurosecretory fibers from the PVN and SON converge as they enter the region of the pituitary stalk and appear as a mass of immunofluorescent material (Fig. 15). The concept of the hypothalamic neurosecretory cells providing the posterior pituitary lobe with neurophysin by somatofugal transport has recently been verified by Alvarez-Buylla et al. (1973). Immunofluorescence histology was applied to parasagittal sections cut from a block of tissue containing the hypothalamus attached to the posterior pituitary gland obtained from a dog 20 hours after the pituitary stalk had been constricted. There was an accumulation of neurophysin proximal to the constriction, in contrast to the region of the stalk immediately distal to the crush, which was devoid of im-

FIG.15. Convergence of the fibers of the supraopticoneurohypophysialtract in the pituitary stalk of the sheep. Saline-formalin-fixed tissue. x 170. (From Watkins, 1973b.)

276

W. B. WATKINS

FIG.16. Montage showing the distribution of neurophysin immunofluorescence in the pituitary stalk of a dog 20 hours after the stalk had been constricted (arrows). Snapfrozen tissue. (From Alvarez-Buylla et al., 1973.)

munofluorescent material (Fig. 16). Six days after stalk constriction, the amount of neurophysin in the hypothalamus and stalk proximal to the crush was twofold greater than that in the sham-operated animal, clearly indicating a proximal-distal flow of neurophysin.

IMMUNOHISTOCHEMISTRY OF NEUROPHYSIN

277

E. POSTERIORPITUITARYGLAND The distribution of neurophysin within the posterior pituitary gland does not appear to be the same for all species studied. In the guinea pig neurophysin distribution had a distinct lobular appearance (Fig. 17a) similar to the pattern of NSM (Fig. 17b) revealed with AF. In this respect the guinea pig resembles the opposum (Bodian, 1951; Roth and Luse, 1964), whose NSM also has a lobular appearance. This is in contrast to the closely packed neurophysin found in the rat (Watkins and Evans, 1972) (Fig. 18a) and NSM present in the dog (Ortmann, 1951) posterior pituitary gland. Osmotic stimulation of an animal by either removing the drinking water or by salt loading, results in a progressive decrease in the amount of neurophysin concomitant with loss of the neurohypophy-

FIG. 17. Distribution of immunofluorescent neurophysin (X130) (a) and NSM (b) 150) in the posterior pituitary lobe of the guinea pig. The immunofluorescent study was carried out on fresh-frozen tissue, while the NSM was demonstrated on salineformalin-fixed tissue. (X

278

W. B. WATKINS

FIG.18. Effect of dehydration on the distribution of neurophysin [(a) and (b)] and neurosecretory material [(c)and (d)] in the posterior pituitary lobe of the rat. Normal animal [(a) and (c)] and animal dehydrated for 7 days [(b) and (d)], immunofluorescence [(a) and (c)]. (X238). AF stain [(b) and (d)] (X224). (Adapted from Watkins and Evans, 1972.)

sial hormones and AF material (Watkins and Evans, 1972; Evans and Watkins, 1973) (Fig. 18a-d).

IX. Use of Species-Specific Antisera for the Demonstration of Neurophysin A. PIG Using an antiserum that specifically cross-reacted against porcine neurophysin 11, Livett et al. (1971) showed that immunofluorescence

IMMUNOHISTOCHEMISTRY OF NEUROPHYSIN

279

was localized principally in the neurosecretory neurons arising from magnocellular SON. A few scattered cells close to the third ventricle in the region of the PVN also reacted. Since the bulk of the neurosecretory cells of the PVN did not cross-react with the antiserum, it was suggested by these workers that porcine neurophysin I1 may be associated with 8-lysine-vasopressin which was thought at the time to be the main hormone of the SON. However, more recent findings by Johnston et al. (1972) have confused this simple association of neurophysin I1 with vasopressin. Dissection of SON and PVN from cryostat sections of pig hypothalamus was performed, and the levels of oxytocin, 8-lysine-vasopressin, neurophysin I, and neurophysin I1 determined. It was found that both neurophysins and the neurohypophysial hormones were present only in the SON, and it is therefore difficult in the case of the pig to associate specifically one neurophysin with one hormone solely based on immunofluorescent histochemical studies.

B. Ox Antisera raised against bovine neurophysin I has been used to investigate the cellular distribution of neurophysin in ox brains using the immunoperoxidase technique. Neurophysin I was localized in both the SON and PVN (Fig. 19a and b), although results from radioimmunoassay suggest that there is more neurophysin I1 than neurophysin I in the SON (Zimmerman et al., 1973b). The latter results, together with those obtained from the subcellular distribu-

FIG. 19. Immunoperoxidase localization of bovine neurophysin I in the ox. (a) SON ~ 3 4 4(b) . PVN ~ 3 4 5 (From Zimmerman et al., 1973b.)

280

W. B. WATKINS

tion work of Dean et al. (1968a), partially confirm that the SON are more specialized in the elaboration of vasopressin than oxytocin, in accord with other species (Bisset et al., 1973).

X. Conclusions In this article the techniques of immunohistochemistry have been applied to study tissue distribution of neurophysin throughout the HNS. The sensitivity of the method enables one to visualize immunoreactive material not only in the perikarya of hypothalamic nuclei but also in the fine axons leading from the hypothalamus to the posterior pituitary lobe. This method has provided positive evidence for Bargmann’s initial concept of neurosecretion, in which protein material synthesized in the hypothalamic nuclei is conveyed by axonal transport to nerve terminals in the posterior pituitary gland. The presence of immunoreactive neuroph ysin in areas that are apparently devoid of NSM, as revealed by AF, clearly demonstrates the insensitivity of the traditional histochemical stains as compared with immunofluorescent techniques. Although much information has been obtained during the short 3-year history of the localization of neurophysin by immunofluorescence methods, several avenues are still open for investigation:

1. Application of immunoperoxidase techniques to thin sections using antisera specific to one neurophysin or another within a particular species will establish whether or not individual nerve fibers are specialized in the transport of specific neurophysins. 2. The results obtained from studying the ontogeny of neurophysins may be correlated with the presence of arginine vasotocin in certain fetal mammals (Vizsolyi and Perks, 1969). 3. The significance of neurophysinlike material in the hindbrain of neurologically disturbed animals is as yet unclear. 4. Perhaps the most intriguing aspect of all the neurophysin work published is the presence of neurophysin or molecules structurally similar to neurophysin in the external zone of the infundibulum. The confirmation of the presence of a carrier molecule for CRF will revitalize the study of the elusive releasing factor. ACKNOWLEDGMENTS

I thank the Medical Research Council of New Zealand and the Auckland Medical Research Foundation for financing this work. Miss V. Bailey is thanked for assistance

IMMUNOHISTOCHEMISTRY O F NEUROPHYSIN

28 1

in the photography. I also thank Drs.Earl A. Zimmerman and Bruce G. Livett for providing me with copies of photographs of their immunohistochemical results. REFERENCES Acher, R.,and Fromageot, C. (1957).In “The Neurohypophysis” (H. Heller, ed.), pp. 39-50.Butterworth, London. Acher, R., Manoussos, G., and Olivry, G. 1955.Biochim. Biophys. Acta 16, 155. Acher, R., Light, A., and du Vigneaud, V. (1958).J. Biol. Chem. 233, 116. Adams, C. W.M., and Sloper, J. C. (1956).J. Endocrinol. 13,221. Alvarez-Buylla, R., Livett, B. G., Uttenthal, L. O., Hope, D. B., and Milton, S . H. (1973).Mikrosk. Anat. Z . Zellforsch. 137,435. Anand Kumar, T. C., and Knowles, F. G. W. (1967).Nature (London)215,54. Anderson, B., and Jewel], P. A. (1957).J . Endocrinol. 15,332. Aulsebrook, L.H., and Holland, R. C. (1969).Amer. J . Physiol. 216,818. Bandaranayake, R. C. (1971).Acta Anat. 80, 14. Bargmann, W.(1949a).Z. Zellforsch. Mikrosk. Anat. 34,610. Bargmann, W.(1949b).Klin. Wochenschr. 27,617. Bargmann, W.(1950).Mikroscopie 5,239. Bargmann, W.(1951).Med. Monatsschr. 5,466. Bargmann, W. (1966).Int. Reo. Cytol. 19, 183. Barnett, R. J. (1954).Endocrinology 55,484. Beck, E., Daniel, P. M., and Parry, H. B. (1964).Brain 87, 153. Beck, E., Daniel, P. M., Gajdusek, D. C., and Gibbs, C. J., Jr. (1969a).“Virus Disease and the Nervous System” (C. W. M. Whitty, J. T. Hughes, and F. 0. MacCallum, eds.) pp. 107-120.Blackwell, Oxford. Pathol. 97,545. Beck, J. S., Gordon, R. L., Donald, D., and Melvin, J. M. 0. (1969b).J. Bern, H.A,, and Knowles, F. G. W. (1966).In “Neuroendocrinology” (L. Martini and W. F. Gangong, eds.), Vol. 1, pp. 139-186.Academic Press, New York. Mikroskopie 15,67. Bertalanffy, F. D. (1960). Bisset, G. W., Hilton, S . M., and Poisner, A. M. (1967).Proc. Roy Soc., Ser. B 166,422. Bisset, G. W.,Clark, B. J., and Errington, M. L. (1970).J . Physiol. (London) 207, 21P. Bisset, G. W., Clark, B. J., and Emngton, M. L. (1971). J . Physiol. (London)217, 111. Bisset, G. W., Errington, M. L., and Richards, C. D. (1973).Brit.J.Phannacol. 48,263. Block, R. J., and van Dyke, H. B. (1952).Arch. Biochem. Biophys. 36, 1. Bock, R., and Forstner, R. V. (1969).Z . Zellforsch. Mikrosk. Anat. 94,434. Bock, R., and Miihlen, K. (1968).Z . Zellforsch. Mikrosk. Anat. 92, 130. Bock, R., and Schliiter, G. (1971).Histochemie 25, 152. Bock, R., Forstner, R. V., Miihlen, K., and Stohr, P. A. (1969). Z. Zellforsch. Mikrosk. Anat. 96, 142. Bodian, D. (1951).Bull. Johns Hopkins Hosp. 89,354. Breslow, E., and Abrash, L. (1966).Proc. Nut. Acad. Sci. U . S . 56,640. Breslow, E., Aanning, H. L., Abrash, L., and Schmir, M. (1971).J . Biol. Chem. 246, 5179. Brinkmann, H., and Bock, R. (1970). 1. Neuro-Visc. Relt. 32,48. Burford, G. D., and Moens, L. (1971). J . Endocrinol. 51,609. Burford, G. D., Jones, C. W., and Pickering, B. T. (1971).Biochem.J. 124,809. Cheng, K. W., and Friesen, H. G. (1971a).Endocrinology 88,608. Cheng, K. W., and Friesen, H. G. (1971b).J . Biol. Chem. 246,7656. Cheng, K. W., and Friesen, H. G. (1972).J. Clin. Endocrinol. Metab. 34, 165. Cheng, K. W., and Friesen, H. G. (1973).J . Clin. Endocrinol. Metab. 36,553. Cheng, K.W., Friesen, H. G., and Martin, J. B. (1972).Endocrinology 90, 1055.

282

W. B. WATKINS

Clarke, J. R.,and Kennedy, J. P. (1967).Gen. Comp. Endocrinol. 8,455. Coy, D. H., and Wuu, T.-C. (1971).A n d . Biochem. 44,174. Coy, D. H.,and Wuu, T.-C. (1972).Biochim. Biophys. Acta 263,125. Dean, C. R.,Hollenberg, M.D., and Hope, D. B. (1967).Biochem. J . 104,8C. Dean, C. R.,Hope, D. B., and Wzi6, T. (1968a).Brit. 1. Pharmucol. 34, 192P. Dean, C. R.,Hope, D. B., and KAzib, T. (1968b).Brit. J . Pharmacol. 34, 193P. Diamond, M. C.(1956).Endocrinology 58,461. Ellis, H. K., Watkins, W. B., and Evans, J. J. (1972).J . Endocrinol. 55,565. Evans, J. J., and Watkins, W. B. (1973). Z . Zellforsch. Mikrosk. Anut. 145, 39. Fawcett, C. P., Powell, A. E., and Sachs, H. (1968).Endocrinology 83, 1299. Fendler, K.,Hefco, H., and Lissak, K. (1968).Acta Physiol. 34,285. Ford, D. H., and Kantounis, S. (1957).J . Comp. Neurol. 108,91. Forsling, M. L.,Martin, M. J., Sturdy, J. C., and Burton, A. M. (1973).J . Endocrinol. 57,307. Frankland, B. T. B., Hollenberg, M. D., Hope, D. B., and Schacter, B. A. (1966). Brit.]. Pharmacol. 26,502. Friesen, H. G., and Astwood, E. B. (1967).Endocrinology 80,278. Gabe, M. (1960).C. R. Acad. Sci. 250,937. Gelotte, B. (1960).J . Chromatogr. 3,330. Ginsburg, M. (1968).I n “Hancll,ook of Experimental Pharmi~cology”(13. Rerde. ed.), Vol. XXIII, pp. 286-371.Springer-Verlag, Berlin and New York. Ginsburg, M., and Ireland, M. (1965).J. Endocrinol. 22, 187. Ginsburg, M., and Jayasena. K. (1968).J . Physiol. (London) 197,53. Ginsburg. M., Jayasena, K., and Thomas, P. J. (1966).J . Physiol. (London) 184, 387. Gomori, C . (1941).Amer. J . Puthol. 17,395. Gomori, G. (1950). Amer. J . Clin. Puthol. 20,665. Gutierrez, M., and Sloper, J. C. (1969).Histochemie 17,73. Haselbach, C . H.,and Piquet, A. R. (1952).Helo. Chim. Acta 35,2131. Hild, W., and Zetler, G. (1953). Pfluegers Arch. Cesumte Physiol. Menschen Tiere 257, 169. Holleiiberg, M. D., and Hope, D. B. (1967).Biochem. 1. 104, 122. Hollenberg, M. D., and Hope, D. B. (1968).Biochem. J . 106,557. Hope, D.B., and Uttenthal, L. 0. (1969). Colloq. Znt. Cent. Nat. Rech. Sci. 177, 25. Howe, A. (1959).J . Physiol. (London) 149,519. Howe, A. (1962).In “Neurosecretion” (H. Heller and R. B. Clark, eds.), pp. 241-245. Academic Press, New York. Howe, A., and Pearse, A. G. E. (1956).].Histochem. Cytochem. 4,561. Johnston, C. I., Pickup, J. C., Uttenthal, L. O., and Hope, D. 13. (1972).Proc. Aust. Endocrine Soc. 15,41. Kamm, 0.. Aldrich, T. B., Grote, 1. W., Rowe, L.W., and Bugbee, E. P. (1928).J. Amer. Chem. Soc. 50,573. Ibwamura. A. (1969).“Fluorescent Antibody Techniques and their Application.” Univ. Park Press, Baltimore, Maryland. J . Histochern. Cytochem. 18, 161. Kawardi, Y.,and Nakane, P. K. (1970). Knaggs, G. S., Tindal, J. S., and Turvey, A. (1971).J . Endocrinol. 50, 153. Knowles, F., and Anand Kumar, T. C. (1969).Phil. Trans. Roy. Soc. London, Ser. B 256,357. Kobayashi, H., Matsui, T., and Ishii, S. (1970).Znt. Rev. Cytol. 29,281. Landing, B. H., Hall, H. E., and West, C. D. (1956). Lob. Znoest. 5,256. Lederis, K. (1962).In “Neurosecretion” (H. Heller and R. B. Clark, eds.), pp. 227-239. Academic Press. New York.

IMMUNOHISTOCHEMISTRY OF NEUROPHYSIN

283

Legros, J. J., Franchimont, P., and Hendrick, J. C. (1969).C. R. Soc. Biol. 163,2773. Legros, J. J., Stewart, U., Nordmann, J. J., Dreihss, J. J., and Franchimont, P. (1971a). C.R. Soc. B i d . 165,2443. Legros, J. J., Franchimont, P., and Barbe, L. (1971b).C. R. SOC. B i d . 105,203. Livett, B. G., and Parry, H. B. (1971).Brit. J . Pharmacol. 43,423P. Livett, B. G., and Parry, H. B. (1973).J . Physiol. (London)230,20P. Livett, B. G., Uttenthal, L. O., and Hope, D. B. (1971).Phil. Trans. Roy. SOC. London, Ser. B 261,371. Mahon, W. E., Scott, D. J., Ansell, G., Manson, G. L., and Fraser, R. (1967).Thorax 22,13. Martin, M. J. (1971). J . Endocrind. 49,553. Martin, M. J., Chard, T., and Landon, J. (1972).J . Endocrinol. 52,481. Nairn, R. C. (1969).“Fluorescent Protein Tracing,” 3rd Ed. Livingston, Edinburgh. J . Histochem. Cytochem. 16,557. Nakane, P. K. (1968). Nakane, P. K. (1971).Acta Endocrinol. (Copenhagen), S u p p l . 153, 190. Norstrom, A., Sjostrand, J., Livett, B. G., Uttenthal, L. O., and Hope, D. B. (1971). Biochem. J . 122,671. Olivecrona, H. (1957).Acta Physiol. Scand., S t c p p l . 40, 136. Ortmann, R. (1951). 2. Zellforsch. Mikrosk. Anat. 36,92. Osborne, W.A., and Vincent, S. (1900).Brit. Med. J . i, 502. Ouchterlony, 0. (1968). “Handbook of ImmunodifFusion and Immunoelectrophoresis.” Humphreys Sci. Publ., Ann Arbor, Michigan. Parry, H. B., and Livett, B. G. (1973).Nature (London) 242,63. Pasteels, J. A., Gausett, P.,Danguy, A., Ectors, F., Nicoll, C. B., and Varavudhi, T. (1972).J. Clin. Endocrinol. Metab. 34,959. Pepys, J., Jenkins, P. A., Lachman, P. J., and Mahon, W. E. (1965).J . Endocrind. 33, viii. Pickering, B. T. (1968). J . Endocrinol. 42, 143. Pickup, J. C., and Hope, D. B. (1971).Biochem. J . 123, 153. Pickup, J. C., and Hope, D. B. (1972).J. Neurochem. 19,1049. Pickup, J. C.,Johnston, C. I., Nakamura, S., Uttenthal, L. O., and Hope, D. B. (1973). Biochem. J . 132,361. Pearse, A. G . E. (1961). “Histochemistry Theoretical and Applied.” Churchill, London. Plixka, V., McKelvy, J. F., and Sachs, H. (1972).Eur.J.Biochem. 28, 110. Porter, J. C., Miscal, R. S., Tippit, P. R., and Drane, J. W. (1970).Endocrinology 86, 590. Rauch, R., Hollenberg, M. D., and Hope, D. B. (1969).Biochem.J. 115,473. Rinne, U. K. (1966).In “Methods and Achievements in Experimental Pathology” (E. Bajusz and G. Jasmin, eds.), Vol. 1, pp. 169-205.Karger, Basel. Robinson, A. G., and Zimmerman, E. A. (1973). J . Clin. Znuest. 52, 1260. Robinson, A. G.,Zimmerman, E. A., Engleman, E. G., and Frantz, A. G . (1971).Clin. Exp. Metab. 20, 1138. Rosenfeld, M.(1940).Bull. Johns Hopkins Hosp. 66,398. Rosselet, A. (1967).2. Wiss. Mikrosk. 68,22. Roth, L. M., and Luse, S. A. (1964).J. Cell B i d . 20,459. Sachs, H. (1969).Adoan. Enzymol. Relat. Areas. Mol. Biol. 32,327. Schally, A. V., and Bowers, C. Y. (1964).Clin. E x p . Metab. 13, 1190. Schally, A. V.,Lipscomb, H. S., and Guillemin, R. (1962).Endocrinology 71, 164. Scharrer, E., and Scharrer, B. (1954).Recent Progr. Horm. Res. 10, 183. Schiebler, T. H.(1951).Acta Anat. 13,233. Schiebler, T.H. (1952).Acta Anat. 15,393.

284

W. B. WATKINS

Sloper, J. C. (1955).J. Anat. 89,301. Sloper, J. C. (1958). Znt. Reo. Cytol. 7,337. Sokol, H. W. (1970). Neuroendocrinology 6 , W . Sokol, H. W., and Valtin, H. (1967). Nature (London) 214,314. Sterba, G. (1964). Acta Histochem. 17,268. Takabatake, Y., and Sachs, H. (1964). Endocrinology 75,934. Talanti, S. (1971). Z. Zellforsch. Mikrosk. Anat. 115, 110. Talanti, S., and Kivalo, E. (1961). Experientia 17,470. Tindal, J. S., Knaggs, G. S.,and Turvey, A. (1968).J. Endocrinol. 40,205. Uttenthal, L. O., and Hope, D. B. (1970). Biochem. J . 116,899. Uttenthal, L. O., and Hope, D. B. (1972). Proc. Roy. Soc., Ser. B 182,73. Vaitukaitis, J., Robbins, J. B., Nieschlag, E., and Ross, G . T. (1971). J . Clin. Endocrinol. 33,988. van Dyke, H. B., Chow, B. F., Greep, R. O., and Rothen, A. (1942). 1. Pharmucol. Exp. Ther. 74, 190. Vilhardt, H. (1970). Acta Endocrinol. (Copenhagen)63,585. Vizsolyi, E., and Perks, A. M. (1969). Nature (London) 223, 1169. Watkins, W. B. (1971).J . Endocrinol. 51,595. Watkins, W. B. (1972a). Biochem. J . 126,759. Watkins, W. B. (1972b).J . Endocrinol. 55,577. Watkins, W. B. (1973a).J . Endocrinol. 59, 17. Watkins, W. B. (197313).Z. Zellforsch. Mikrosk. Anat. 145,471. Watkins, W. B., and Ellis, H. K. (1973).J . Endocrinol. 59,30. Watkins, W. B., and Evans, J. J. (1972). Z. Zellforsch. Mikrosk. Anat. 131, 149. Watkins, W. B., and Evans, J. J. (1974).J . Histochem. Cytochem. 22,128. Wells, J. (1963). Exp. Neurol. 8,470. Wuu, T.-C., and Saffran, M. (1969).J . Biol. Chem. 244,482. Zimmerman, E. A., Camel, P. W., Kazim Husain, M., Ferrin, M., Tannenbaum, M., and Frantz, A. G. (1973a). Science 182,925. Zimmerman, E. A., Hsu, K. C., Robinson, A. G., Carmel, P. W., Frantz, A. G., and Tannenbaum, M. (1973b). Endocrinology 92,931. Note added in proof: Pelletier, G., Leclerc, R., Labrie, F., and Puviani, R. (1974). (Molec. Cell. Endocrinol. 1,157) reported the ultrastructural localization of neurophysin in the neurosecretory axons of the SON, PVN, internal layer of the median eminence, and in the nerve terminals of the posterior pituitary gland of the rat using the immunoperoxidase technique in association with rabbit anti-human neurophysin-I1 serum. After bilateral adrenalectomy of rats, neurophysin-like material appeared in the external layer of the median eminence (Vandesande, F., D e ‘May, M.,and Dierickx, K. (1974) Cell Tissue Res. 151, 187) and was augmented by the administration of steroids (Watkins, W. B., Schwabedal, P., and Bock, R. (1974) Cell Tissue Res. In press. Furthermore, the external median eminence of the physiologically normal cat and dog also contained material immunoreactive against rabbit anti-porcine neurophysin-I1 serum Watkins, W. B. (1975) Cell Tissue Res. In press. Three short reviews concerned with the immunohistochemical localization of neurophysin in the mammalian HNS have been submitted for publication in Ann. N. Y. Acad. Sci. (1975) by Watkins, W. B.; Livett, B. G.; and Zimmerman, E. A., Defendini, R., Sokol, H. W., and Robinson, A. G . An extensive review on the methods of isolation of neurophysin proteins has also been reported by Walter, R., and Breslow, E. (1974). In “Research Methods in Neurochemistry” (N. Marks and R. Rodnight, eds.), Vol. 2, pp. 247-279. Plenum Press, New York.

The Visual System of the Horseshoe Crab Limulus polyphemus WOLF H . FAHRENBACH hboratory of Electron Microscopy. Oregon Regional Primate Research Center. Beaverton. Oregon

I . Introduction . . I1. Dioptric Structures

. . . . . . . . . . .

. . . . A . Cuticular Cones and Lenses . B. Lens Epidermis . . . . C . ConeCells . . . . . 111. Pigment Cells . . . . .

.

IV . V.

VI . VII .

VIII .

IX. X.

XI.

A Guanophores . . . . B. Distal Pigment Cells . . C . Ommatidial Pigment Cells Neuroglial Cells . . . . Receptor Cells . . . . A . Retinula Cells . . . B. Eccentric Cells . . . C Arhabdomeric Cells . . Basal Lamina and Hemocoel Axons and Plexus . . . A . Afferent Axons . . . B. Efferent Axons . . . C Plexus . . . . . OpticNerves . . . . . Optic Centers . . . . Miscellaneous Aspects . . A . Abnormalities . . . . B. Pathology . . . . . Vision and Behavior . . . Addendum . . . . . References . . . . .

.

.

I

.

. . . . . . .

. . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . . .

.

. . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . .

. . . . . .

.

. . . . .

.

. . . . .

.

. . . . .

.

. . . . .

.

. . . . .

.

. . . . .

285 287 287 292 293 296 296 298 301 304 306 306 314 317 318 321 321 324 327 333 335 338 338 340 341 344 344

Introduction

Two noteworthy events were recorded for the year 1782. to wit. the last execution of a witch and the first microscopic study of the lateral eye of Limulus (AndrC. 1782). This auspicious start of a rational age was followed after about a century by two histological studies (Grenacher. 1879; Lankester and Bourne. 1883). which to this day are a pleasure to read . Later. WatasC (1890).by painstaking maceration and teasing of eyes. provided insight into the cellular composition of the ommatidium. including the presence of a ganglion (i.e., eccentric) cell . An analysis of the cuticular dioptrics of the compound 285

286

WOLF H. FAHRENBACH

eye (Exner, 1891) and a description of the ventral eye (Patten, 1893) laid the foundation for the only previous review of eye structure in the horseshoe crab (Demoll, 1914). [The correct name of the American horseshoe crab is Limulus polyphemus. The name Xiphosura pertains to its order within the Merostomata and has been suppressed as a generic name by the International Commission on Zoological Nomenclature (Stpjrmer, 1952; Levi, 1968).] The continuing interest in Limulus eyes arises principally from their neurophysiological accessibility. Studies of visual physiology in Limulus, complete with a Nobel prize and a voluminous bibliography, have contributed considerably to our understanding of the function of photoreceptors but are treated in this article only in relation to definitive structural elements. Two recent physiological reviews, one by Wolbarsht and Yeandle (1967),which also covers some pertinent systematics and zoogeography of the xiphosurans, and the other, a comprehensive study of ommatidial electrophysiology by Smith and Baumann (1969),furnish the reader ready access to the literature on the function of the Limulus eye. To a biologist accustomed to vertebrate tissues, the fine structure of the Limulus visual system is a somewhat bewildering and alien landscape. Consequently, the primary aim of this article is to supply a diagnostic guide to the various cells and tissues of the eyes and to review the current status of information on their structural-functional correlations. The anatomical orientation of most previous publications is acknowledged by means of diagrams and survey illustrations, but a cytological emphasis is considered more useful for future experimental work with Limulus. As far as is practical in the context of a review, my own work has been updated by new observations, and various small informational gaps in the literature have been closed. The most prominent eyes of the horseshoe crab are the lateral compound eyes, bean-shaped, convex structures set into the cuticle under the lateral spines of the prosoma. The two ocelli are centrally placed on the prosoma on either side of the median dorsal spine. The lateral rudimentary eyes, rather unorganized masses of photoreceptor cells and guanophores, lie at the posterior border of each compound eye. A similar mass, the median rudimentary (endoparietal) eye, accompanies the two ocelli. It is bipartite but fused into a Y-shaped organ, An additional photoreceptor, the ventral eye, consists of two groups of cells at the base of the so-called olfactory tubercle, a small protuberance anterior to the mouth. The two nerves from this eye contain additional visual cells scattered along their length. A few

THE VISUAL SYSTEM OF T H E HORSESHOE CRAB

287

receptor cells, associated with the ventral eye nerve but positioned at the surface of the second optic ganglion, complete the inventory of the Lirnulus photoreceptors. The phylogenetic relationship between these eyes and those of other arthropods is best appreciated by reference to Eakin’s (1970) comprehensive review of invertebrate eyes. 11. Dioptric Structures

A.

CUTICULAR CONES AND

LENSES

Only the compound and ocellar eyes contain sculptural specializations of the overlying cuticle, which is composed of tanned proteins and about 25% chitin (Hock, 1940) and is densely lamellated (Richter, 1969). The lenses of the ocelli, quite similar to those of spider ocelli (Carricaburu, 1970) and first described by Demo11 (1914), are each formed by a subspherical, internal projection of the transparent cuticle over the eye, with a diameter of about 0.5 mm in an animal of 5-cm width (Figs. 1 and 2). [Data relating molting stage, age, size, and eye size of Lirnulus have been compiled by Waterman (1954a) and Shuster (1955).]The cuticular lamellae dip into the lens

FIG. 1. Internal view of cleaned cuticle with the two ocellar lenses. The deep recess is the median frontal spine of the animal. x60.

288

WOLF H. FAHRENBACH

FIG. 2. Vertical section through a median ocellus. The spherical lens (L) is lined by lens epidermis, which changes to a pigmented epithelium laterally (P). Receptor cells (R)with contorted rhabdomeres sandwich guanophores (C).Optic fibers (A) and vascular sinusoids (S) border the base of the micrograph. x 150.

as they do into the cuticular cones of the compound eye. The lens proper has a refractive index gradient of 1.591at its proximal periphery to a lower one (1.555) at its center; but the overlying cuticle, although heterogeneous in its refractive properties, incorporates a weak converging lens (center refractive index 1.567; periphery 1.553) above the ocellar sphere (Carricaburu, 1968). However, the lens was tested at a wavelength of 585 nm, at which the ocellus is not a particularly sensitive photoreceptor. Although not image-forming,

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

289

the shape of the lens and the immediately subjacent array of retinulae suggest that the ocellus can at least convey directional information. Its large aperture in relation to that of an ommatidium makes it particularly adapted to seeing at low light levels. The cuticle of the compound eye is smoothly curved with hardly any sculptural indication of facets. However, its internal surface is covered with blunt-tipped cuticular cones (Figs. 4 and 5), first and excellently illustrated by And& (1782). Despite the initial impression of regularity, the precision of rows is maintained only over short stretches, and the “squeezing out” of lattice rows is common (these features are best observed by viewing Fig. 4 at a glancing angle). The array of cones gains its adult number of about 850 rapidly (Wa-

FIG.3. Detail of the relationship between ocellar lens and lens epidermis (Le), retinula cells (R) and their rhabdomeres, guanophores (G), optic fibers (A), and adjacent pigmented epidermis (P). X330.

290

WOLF H. FAHRENBACH

terman, 1954a)and is practically complete in an animal 5-10 cm wide. The cones are added peripherally and grow both in size and spacing with successive molts. Only a few at the center of the eye are at right angles to the surface; the remainder deviate from normal orientation to a maximal inclination of 55" at the periphery of the eye. Discounting the acceptance angle of individual ommatidia, this slanted insertion of the cones (k55") added to the curvature of the cuticle provides a 180" horizontal field of vision for each over the eye (70") eye, The vertical field of view is about 90". Any experimentation that involves the directional sensitivity of ommatidia (Waterman, 195413) or of individual retinula cells (Ratliff, 1966) is profoundly affected by the slanted insertion of most of the cones. The system of Gemperlein (1969) for encoding the location of specific facets, although potentially an ideal method for locating the few vertical cones, may be difficult to apply to the irregular array of the Limulus compound eye. Exner (1891) originally interpreted the cones as cylinder lenses, that is, structures that bend the light toward the optical axis by a gradient of refractive indices. Such cases have since been observed in fireflies, and their properties have been discussed (Seitz, 1969; Horridge, 1969), but the Limulus cuticular cone appears to have a

FIG.4. Internal view of the compound eye cuticle cleaned of adhering tissue. The surrounding cuticle is perforated by gland ducts. X60.

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

29 1

rather homogeneous refractive index (Carricaburu, 1967). Hence the function of the cone can best be understood as that of a wave guide. Light entering the cone at an angle to its axis undergoes total internal reflection, the limiting angle depending on the refractive index of the cone, the surrounding epidermal cells, and the concentration of pigment near the cone. This concentration would be expected to increase the angle of total reflection during light adaptation and would therefore allow more light to be absorbed by the pigment than in a dark-adapted eye (Seitz, 1969). Ray tracing on the basis of estimated refractive indices yields an acceptance angle of 70" for the Limulus ommatidium (Fahrenbach, 1968), which is reasonably close to the neurophysiologically derived result of 80"(Waterman, 1954b). This unusually high angle is probably not caused by leakage from

FIG.5. Vertical section through a cuticular cone and apical ommatidium. Vascular spaces (S) are bordered by guanophores (G) and distal pigment cells (P). C, Cone cells; R, retinula cells E, eccentric cell dendrite. The cuticle and subcuticular space have been invaded by fungus (F). X350.

292

WOLF H. FAHRENBACH

adjacent ommatidia because (1)the pigment surrounding the cones never leaves the region between the cones, and (2) little optical interaction between ommatidia could be detected by physiological means (Scholes and Hartline, in Shaw, 1969). Lateral inhibition probably provides a compensating mechanism to correct the overlap between the visual fields of adjacent ommatidia (Reichardt, 1961).A combination of eye curvature, angle of cone insertion, and angle of ommatidial acceptance provides for virtually omnidirectional vision in Limulus. The compound eye does not discriminate polarized light at normal incidence to the surface, but differentiates to an increasing degree with oblique illumination (Waterman, 1954~).However, the numerous concentric cuticular lamellae of the cone, first noted by Grenacher (1879),are made up of 150-A, swirling filaments which, like those of the ocellar lens, have enough form birefringence to effect complete depolarization of light transmitted through the cone (Fahrenbach, 1968). Therefore sensitivity to polarized light at slanting incidence must be attributed to spurious effects of reflection at the cuticular surface, a process that yields different intensities for different vectors of polarization. Furthermore, the cytological prerequisites for polarized light perception are lacking because of irregular rhabdomeres in the ocelli and electrotonic coupling of rhabdomeres in the ommatidia, a feature that effectively abolishes polarization sensitivity (Snyder et al., 1973).

B. LENS EPIDERMIS The epidermis underlying the nonocular cuticle is usually lightly pigmented in Limulus. The epidermis of the eyes, however, although retaining distinctive characteristics and general functions associated with molting, becomes variously modified into lens epidermis, cone cells, guanophores, and distal pigment cells. Since it is the most generalized type, the epidermis of the ocellar lens can serve here as the standard of comparison for the more deviant cell types (Figs. 2 and 3). This internal lining of the lens consists of highly interlaced, unpigmented cells with a plethora of junctions which emphasize the mechanical role of the tissue. At the apical surface, dense plaques resembling hemidesmosomes mediate the epitheliocuticular bond in conjunction with bundles of microtubules inserted on the plaques. This distinctive arthropodan junction (Bassot and Martoja, 1965, 1966; Noirot-TimothCe and Noirot, 1966) is often associated with an anchoring filament deeply embedded in the cuticle, as it is in other

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

293

FIG. 6. Cross section of dark-adapted ommatidia. The stellate rhabdornes display a few branched or looped fins. The vascular interstices contain hemocytes. ~ 1 0 0 .

arthropods (Bouligand, 1966). During the intermolt stage the epidermal surface forms a flush apposition to the cuticle. The predominantly microvillous surface frequently found in young animals and illustrated by Whitehead et al. (1969) presumably indicates an early molting stage, when a molting space is formed (Locke, 1969). Laterally, the cells are attached by an apical macula adhaerens and by extensive septate desmosomes. The basal surface is profusely fluted and covered by a thick basal lamina. Large quantities of glycogen are a distinguishing feature of these cells, but probably not to the extent of affecting the dioptrics of the system as in some other invertebrate eyes (Wolken and Florida, 1969; Eakin and Kuda, 1972). Longitudinal bundles of microtubules bypass the small basal nucleus and are inserted on dense plaques adjacent to the basal lamina. C. CONECELLS The cone cells of the compound eye, equivalent to the Semper cells of insects, were mentioned by Grenacher (1879) and Exner (1891)and have been described in detail by Fahrenbach (1968,1969) and by Whitehead et al. (1969) (Fig. 5). About 100 of these underlie the flat tip of the cuticular cone. A dozen of the most peripheral cells, their exact number corresponding to the rhabdomal fins, have long

CuticIe

Guanophore

Distal pigment cell

Cuticular cone

Basal lamina

. Basal infolding5 . Cone cells

Cone cell process

. Eccentric

cell dendrite

. Pigment cell partition Palisade

. Rhabdome . Relinula cell

Rhabdome

- Eccentric cell - Proximal pigment cells

- Neuroglial sheath

- Retinula cell axon

-

Eccentric cell axon

FIG.7. Semidiagrammaticview of a cutaway ommatidium. (Redrawn after Fahrenbach, 1969.)

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

295

attenuated processes which project to the bottom of the rhabdome at its periphery. The cone cells are even more featureless than the epidermis of the lens, containing few scattered organelles and some glycogen but primarily conveying an impression of transparency. Their apical junctions are similar to those described for the epidermal cell but lack the conspicuous participation of microtubules or cuticular anchoring filaments. The interlocked cells are attached laterally by occluding junctions rather than by the septate or gap junctions that prevail in other regions of the ommatidium. Basally, the cone cells directly abut the retinula cells and eccentric cell dendrite and are attached to both by adhering junctions (Fig. 18). These, however, do not effect a complete seal, because hemocyanin from adjacent blood spaces frequently enters the seam and thereby gains access to the apical region of the rhabdome. The cone cell processes, called crystalline threads in insects, lie at the edge of each rhabdomal fin and are bonded by uninterrupted fasciae adhaerentes to the two adjacent retinula cells. They are flattened, 0.1-0.4 pm thick, up to 200 pm long, and contain only sparse inclusions (Figs. 7 and 14). Several functional possibilities have been advanced for these crystalline threads, primarily in insects. Exner (1891)suggested that they act as light guides, a possibility that has been explored by Horridge (1969) in the firefly Photuris uersicolor, in which the processes bridge a gap of more than 100 pm between the cuticular cone and the rhabdome and in fact act as wave guides. In Limulus, however, the cone cell processes are located at the periphery of the rhabdome and are therefore not in the proper location to direct light into it. The small dimension of the processes in relation to the wavelength of light in addition to the surrounding pigment would, furthermore, produce rapid attenuation of the conducted light. The overall optical contribution of the cone cell processes is probably negligible. A second suggestion concerns the influence of the extensive adhering junctions on current flow in the ommatidium during light-induced depolarization (Fahrenbach, 1969). The possibility that these junctions introduce an impediment to ionic flux along the intercellular cleft and increase the current flow through the eccentric cell has been clarified by tracer studies. Perrelet and Baumann (1969) exposed bees’ eyes to lanthanum and ferritin and found that both electron-opaque substances traversed the fasciae adhaerentes and entered the intermicrovillous spaces of the rhabdome. Parenthetically, the rhabdomes of all investigated arthropods are peripherally

296

WOLF H. FAHRENBACH

cemented by adhering junctions between the constituent retinula cells, but not necessarily with participation of the cone cell processes. Hence the last and simplest interpretation remains: In Limulus, the cone cell processes with their extensive junctions subserve the mechanical role of assuring cellular cohesion at the periphery of the rhabdome. At the base of the ommatidium, where retinula cells touch directly without the intervention of a cone cell process, they nevertheless display similar junctions around the rhabdome. 111. Pigment Cells A. GUANOPHORES

Guanophores are widely distributed throughout the visual system. They occur between cuticular cones of the compound eyes and in partitions between retinula cells in the ocelli, in large numbers in the rudimentary eyes (Figs. 5, 7, 8, and 9),and even under the cap-

FIG.8. Part of the lateral rudimentary eye. The mass consists of guanophores. Adjacent axons (A) are those of nearby rudimentary eye retinula cells. x 100.

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

297

FIG.9. A cluster of receptor cells embedded in the guanophores (G) of the lateral rudimentary eye. Large inclusion masses are a conspicuous feature of these cells. Rh, rhabdomere; A, axons; N, small nerve. ~ 2 5 0 .

sule of the brain in association with common ectopic retinula cells near the second optic ganglion. Unlike the reflecting pigment in most arthropods, that in guanophores is in fact guanine (Kleinholz, 1959).Birefringent guanine crystals in the shape of polygonal blocks or thick platelets, up to 1.5 pm in diameter, are formed in opaque Golgi-derived droplets and maintain a limiting membrane. The crystals are resistant to sectioning and normally leave angular holes in the section (Fig. 15). The epidermal guanophores of the compound eye line the deep cuticular recesses between cones (Fig. 7) and thus form the white, refractile surround of the compound eye facets that accounts to a large degree for the pseudopupil of the Limulus eye. An overall impression of an active cell type is supported by ample endoplasmic reticulum, free ribosomes, Golgi bodies, and various lysosomal structures. Microtubules are inserted as previously described and often exceed 100 in a single bundle, In these cells, as well as in the distal pigment cells, two functions can be attributed to these fascicles, the static one of mediating cellular attachment to the cuticle, and the dynamic one of governing the rather rapid photomechanical pigment migration. After light adaptation the guanine platelets are concen-

298

WOLF H. FAHRENBACH

trated under the cuticle, but are distributed throughout the cell in the dark-adapted eye. The guanophores of the ocelli (Fig. 3) have no connection with the cuticle but clothe the sides of the receptor cells. Nevertheless, they are remarkably similar to the guanophores of the compound eye in their extensive Golgi cisternae, glycogen, and scattered microtubules running longitudinally in these slender and interlocked cells. The contained guanine platelets have not been observed to migrate; neurosecretory terminals and microtubules, similar to those in cells with mobile pigment, provide positive circumstantial evidence that migration may occur. The remaining category of guanophores, which incidentally are devoid of any innervation, consists of cells filled to capacity with guanine crystals at the expense of practically all other organelles except a small nucleus. These cells comprise virtually the entire mass of lateral and median rudimentary eyes (Fig. 9), and are associated with ectopic retinula cells, such as those in the brain. Their engorged contents, lack of requisite filaments or tubules, and random disposition around photoreceptor cells minimize any possibility of pigment migration. Guanophores of this type are also widely scattered in the connective tissue of Limulus, far from any photoreceptor site, and may function as storage cells of an excretory product which has found secondary utility because of its refractile properties. Rare guanophores in the ocelli and rudimentary eyes contain an approximately equal percentage of guanine platelets and ommochrome pigment droplets. These pigment droplets occur as an occasional admixture in mature guanophores, particularly those of the ocelli. The converse condition, that is, guanine “contamination” of ommatidial pigment cells, has not been found.

B. DISTALPIGMENTCELLS Each ocellus is surrounded by a broad zone of pigmented cells (Figs. 2, 3, and 10) that mirror the distal pigment cells of the compound eye in practically all cytological details except that they form an uncomplicated, pseudostratified epithelium. In the compound eye the stray light between cuticular cones is screened out by the heavily pigmented distal pigment cells (Fahrenbach, 1968; Whitehead et aZ., 1969)that are attached to the sides of the cone and form a cap over the apical part of the ommatidium. Although the cells are slender and up to 300 pm in length, they correspond closely to the previously described basic epidermal type in their apical and lateral junctional specializations, changed appearance of the apical surface

THE VISUAL SYSTEM O F THE HORSESHOE CRAB

299

FIG. 10. Pigment cells rimming the cuticle around the ocellar lens (see Fig. 2). This pseudostratified epithelium shows apical adhesive modification (J), longitudinal microtubule bundles (B), mitochondria (M), and large pigment droplets. The subjacent vascular tissue is bordered by reserve cells (Re) and contains a hemocyte (H), a discharged hemocyte (D), and two cyanoblasts (Cy). x 1500.

300

WOLF H. FAHRENBACH

during molting, deep basal flutings with an underlying basal lamina, and fascicles of microtubules which are particularly aggregated just below the cuticle. Apical and basal insertions of microtubules on hemidesmosomes are most easily demonstrated in these cells. The pigment, deep violet to black, is contained in 0.2 to 0.8-pm membrane-bounded droplets which are totally electron-opaque. Butenandt et al. (1958) identified the substance with chromatography and ultraviolet spectroscopy as an ommin, that is, a high-molecularweight ommochrome (contra Wasserman, 1967). A second ommochrome, an ommidin, which is a low-molecular-weight derivative of tryptophan and methionine, has been described in Limulus by Linzen (1966). The natural screening properties of the pigment were found by Wasserman (1967)to be those of a neutral absorber; but action spectra of retinula cells with and without shielding pigment, the latter in albino Limulus, demonstrate that the difference can be attributed to a red-transmitting pigment (Nolte and Brown, 1970). During dark adaptation (Fig. ll),the pigment is mostly retracted to

FIG.11. Diagrammatic representation of pigment distribution in the dark-adapted (left) and light-adapted (right) ommatidium in longitudinal and cross section.

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

301

the base of the cells, that is, around the distal third of the ommatidium; however, a scattering of granules remains elsewhere. Light adaptation, which is complete in about 10 minutes, moves the pigment distad to a subcuticular position, leaving an apical zone of about l p m pigment-free. At the level of the cone cells, the ascent of pigment in the surrounding distal pigment cells forms an iridal aperture which restricts the exit pupil of the cone to a circle equal to or smaller than the diameter of the flattened tip of the cone (20-50 pm). The cone and apex of the ommatidium become shielded thereby to a degree that precludes light leakage to adjacent units. The distal pigment cells, like the adjacent guanophores with mobile pigment, are supplied with efferent neurosecretory fibers. C. OMMATIDIAL PIGMENTCELLS Two groups of these cells have been differentiated on the basis of location (Fahrenbach, 1969): intraommatidial pigment cells, whose nuclei and cell processes lie in the partitions between retinula cells of an ommatidium (Figs. 12 and 14); and about 100 proximal pigment

FIG. 12. Cross-sectional view of a dark-adapted ommatidium, illustrating the eccentric cell dendrite (E) in the center of the rhabdome, the nucleus of the unsymmetrical retinula cell (U), pigment cell partitions (P), and adjacent vascular space. The bridging tissue between the adjacent ommatidia carries efferent fibers. x 1OOO.

302

WOLF H. FAHRENBACH

cells which lie as a cup-shaped mass at the base of the ommatidium but extend for some distance distally. The two cell types are distinguishable only by their nuclei. Those of the intraommatidial cells are predominantly spherical and 4-5 pm in diameter, although some are pyknotic and quite small, depending perhaps on the stage of development. The nuclei of the proximal pigment cells, in contrast, are 10-12 pm long and located at or below the base of the retinula cells. The intraommatidial pigment cells, also described as pigmented glial cells by Lasansky (1967), form sheaths one to six cells deep around all the visual cells and partitions between them, and invaginate to a variable degree into the visual cells, particularly the eccentric cell, but not to tlie same degree as in invertebrate central neurons. The cells are filled with pigment droplets (ommochrome) and contain longitudinally oriented bundles of microtubules which lie for the most part in the peripheral folds of the deeply fluted cells. The cytoplasm abounds with glycogen in the rosette-shaped a form and has a normal set of organelles, but these cells do not appear to metabolize pigment with any intensity. Intraommatidial cells have very few junctions of any kind. At the periphery of the rhabdome, where the cone cell processes are bonded to adjacent retinula cells by adhering junctions, pigment cells are occasionally seen to participate. More commonly, proximal pigment cells provide such junctions at the base of the ommatidium proximal to the termination of the cone cell processes. Extensive quintuple-layered junctions which have been described between pigment and receptor cells (Lasansky, 1967) can be produced at will as an artifact of high sucrose concentration in the fixative (Fahrenbach, 1969). The periphery of ommatidia varies between a compact sheath of pigment cells with no interstices to speak of to an open cancellous architecture (Figs. 6 and 12), especially in older animals. Although fixation has an undeniable influence, the feature may be mostly a function of age, stage of nutrition, and degree of dark adaptation, the last of which involves drastic changes in the distribution of pigment cell volume. The proximal pigment cells (Fig. 24) form a compact, cup-shaped mass whose cellular details show no noteworthy differences from the intraommatidial pigment cells except for their larger nuclei. The pigment in both cell types, which is withdrawn to the base of the ommatidium during dark adaptation (Fig. ll), starts migrating upward in long, pearl necklace-like columns into the ommatidium within minutes of the onset of light. This movement, parallel and adjacent to microtubule sheaves, leads to a pronounced increase in the pig-

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

303

FIG. 13. Detail of the center of the rhabdome. The eccentric cell dendrite (E) abounds with microtubules and agranular cisternae. It has a few peripheral microvilli and a subsurface filamentous zone. Rhabdomeral microvilli from two adjacent retinula cells abut along the line indicated by the arrow. The fully formed palisade (Pa) indicates the dark-adapted state. X15,OOO.

mentation and thickness of the partitions between retinula cells, particularly at the edge of the rhabdomal fins. Both cell types have a conspicuous supply of neurosecretory fibers which terminate on them. Scattered cells containing ommochrome can be found in the septa between the retinula cells of the ocelli, in the plexus, and occasionally in the ventral eye and ocellar nerves. These are usually rather compact and give no indication of pigment migration; they do not appear to receive efferent fibers. Lipetz (1960) determined that the ommatidial pigment cells have an approximately 40-fold higher electrical resistivity than the adjacent retinula cells or body fluid and suggested that they therefore

304

WOLF H. FAHRENBACH

form an insulating sleeve to confine light-induced current flow to the center of the ommatidium, that is, to the eccentric cell dendrite. Given the more recent findings of patent intercellular pathways from the rhabdome to the periphery (Lasansky, 1970; Perrelet and Baumann, 1969)and the tight coupling between receptor cells (Behrens and Wulff, 1965; Borsellino et al., 1965; Smith et al., 1965),the electrical properties of the pigment cells probably have only a marginal effect on the performance of the ommatidium. IV. Neuroglial Cells The distribution of neuroglial cells in the visual system is inversely related to the abundance of pigment cells. The ommatidia contain relatively few glial cells; most of them serve as sheaths for the eccentric cells, and some are Schwann cells carrying efferent fibers. Guanophores act as glial cells in the ocelli and rudimentary eyes, but frequently a glial cell layer is inserted between the pigment and retinula cells in these receptors. The ventral eye, which is not in intimate contact with pigment cells, displays an elaborate layering of neuroglia (Clark et al., 1969b). All axons and synaptic regions are surrounded by glial sheaths of different complexities (Figs. 22, 23, and 28). Several criteria can be adduced to identify neuroglial cells. Their condensed nuclei, with average diameters of 2-6 pm, are uniformly the smallest among those of neighboring cell types. Generally, their cytoplasm is distinctly less opaque than that of other cells and contains some lipofuscin (gliosomes) but no shielding pigment, scattered minute mitochondria, and often considerable granular vesicles and tubules which should not be confused with more regular synaptic vesicles in the plexus. Often glial cytoplasm appears so vacuous as to require visual reference to rare twists of endoplasmic reticulum or seemingly free-floating Golgi bodies to ascertain its cellular nature. Since this transparency has been observed in many animals and with many fixatives, it is probably a true characteristic. Other types of neuroglial cells in the Limulus brain display a wealth of organelles and inclusions under identical conditions of fixation. Glial cells are capable of secreting basement lamina material, as is unambiguously illustrated by the external lamina of all branches of the plexus (Fig. 27) and within the optic nerve, or more strikingly by the thick, multilayered sheath which occasionally surrounds small efferent nerves (Fig. 23). Few if any junctions are found between glial cells or adjacent to

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

305

receptor cells. Rare five-layered junctions should probably be viewed with reservations for previously mentioned reasons. In a few places, for example, in the ventral eye, glial cells directly abut the rhabdomeral microvilli, where they form gap junctions. Glial cell membranes facing adjacent hemocoel, that is, the intervening basal lamina, are decorated with abundant dense plaques, a presumptive adhesive modification vis-8vis the basal lamina. Another characteristic of neuroglial cells, shared with both pigment and reserve cells, is their extremely attenuated and filmy shape which finds its most extreme expression in the lateral rudimentary eye, where 10 to 30 complex folds occasionally produce a solid

FIG.14. The periphery of the rhabdome. Pa, Palisade; C, cone cell process; R, retinula cell; A, efferent axon. Intraommatidial pigment cells form partitions between (arrows)and infoldings (I) into retinula cells. ~ 7 0 0 0 .

306

WOLF H. FAHFWNBACH

peripheral glial mass (Fahrenbach, 1970a). In the more orthodox situations, one to several layers of glial cells invest parts of the eccentric cell, the receptor cells in the ocellar, ventral, and rudimentary eyes, and all axons. Glial folds invaginate into receptor cells to some degree but not to the extent of breaking up the periphery of the neuron into the configuration associated with a trophospongium. Large axons, like those of eccentric cells, and rudimentary or ocellar receptors, receive an individual sheath of a few neuroglial cells, whereas retinula cell axons of the compound eye travel in fascicles of two to six fibers (the number decreasing with age) in juxtaposition to one another and are usually enveloped by a single glial cell (Dumont et al., 1965; Nunnemacher and Davis, 1968; Fahrenbach, 1970a, 1971) (Fig. 22).

V. Receptor Cells An unusual attribute of Limulus photoreceptors is the presence of second-order visual neurons in juxtaposition to primary visual cells, that is, retinula cells. Eccentric cells play this role in the ommatidia and arhabdomeric cells in the ocelli. Much of the neurophysiological utility of the compound eye lies in the fact that the eccentric cell provides an accessible neural output in the form of modulated spike trains, which can be observed simultaneously with its analog input from the retinula cells. Conversely, recent interest in the ventral eye derives from its large, accessible retinula cells devoid of complicating secondary neurons or synaptic regions.

A. RETINULA CELLS The primary receptor neuron in the visual organs is the retinula cell, distinguished by a large soma, a proximal axon, and superficial, more-or-less continuous brush borders called rhabdomeres. Despite assorted structural differences in retinula cells from various locales, they are similar enough in their cytology to obviate the need for separate descriptions. Details on these cells have been published in conjunction with reports on the ventral eye (Clark et al., 1969b), the ocelli (Jones et al., 1971), the lateral rudimentary eye (Fahrenbach, 1970a), and the compound eye (Miller, 1957; Lasansky, 1967; Fahrenbach, 1969). In the simplest case (in rudimentary eyes), retinula cells are rounded, 30- to 50-pm cells indented into recesses in the accompanying mass of guanophores (Figs. 9 and 15). They assume an elongate shape in the ocelli and ventral eyes (up to 200 pm), and a

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

307

quite precise trilateral, orange-segment form in the compound eye, where a cell can also reach a length of 200 pm (Figs. 7 and 12).

1. Rhabdomeres and Rhabdome Rhabdomeral microvilli, whose orderly disposition largely hinges on fixation, are 1-2.5 pm long and reach their greatest number in the ommatidial retinula cells which have been estimated to bear about 5 x lo5microvilli per cell. Hence an ommatidium with an average of 12 retinula cells contains about 4 mm2 in microvillous membrane, equivalent to 30-40 cm2 in one compound eye (Fahrenbach, 1969). Although the ommatidial rhabdomeres are rather precisely oriented, those on other retinula cells are highly irregular and curvilinear. The beaded appearance of microvillous membranes both in Limulus (Fahrenbach, 1969) and in other invertebrates and vertebrates (Fernandez-Morin, 1962; White, 1967) may indicate visual pigment molecules intimately incorporated into the membrane. The integrity of an ordered array of rhodopsin molecules in the photoreceptor membrane is thought to be essential in the production of an early receptor potential (Brown et al., 1967). Rhodopsin, which constitutes about 10% of the structural protein of the membrane (Hubbard and Wald, 1960), is extractable from rhabdomeres of Limulus; incorporation of the vitamin-A moiety of rhodopsin into Helix rhabdomeres has been demonstrated with autoradiography

FIG.15. Irregular rhabdome in the lateral rudimentary eye with five retinula cells, a guanophore (lower right), a possible neuroglial cell process (N), and a profile that is suggestive of an arhabdomeric cell dendrite (D). X4800.

308

WOLF H. FAHRENBACH

(Brandenburger and Eakin, 1970). Electrophysiological evidence to support the functional role of the rhabdomeral membrane has been obtained in Bombyx mod, whose retinula cells develop an electrical response to light only when the rhabdome becomes differentiated (Eguchi et al., 1962). Other evidence of the localization of photopigments in rhabdomeres has been discussed by Eakin (1972). A small amount of extracellular space is contained in the prismatic interstices between and at the bases of microvilli, a volume that has been estimated at a minimum of 7OOO-10,OOO pm3 per rhabdome (Fahrenbach, 1969). This space has free access to the hemocoel at the apex of the ommatidium-between retinula and cone cells-and probably only slightly restricted pathways through the adhering junctions that surround the sides and base of the ommatidium. Shunt pathways, which would ordinarily be of physiological interest in the context of electrotonic transmission between retinula and eccentric cells, probably play a minor role in view of the tight electrical coupling between these cells. Retinula cells contact one another only at their rhabdomeres; all other surfaces are almost invariably sheathed by cone or pigment cells (Figs. 13 and 14). Rhabdomeral microvilli form quintuplelayered gap junctions with all adjacent surfaces, a feature that is not affected by fixation (Lasansky, 1967; Fahrenbach, 1969). In ventral, ocellar, and rudimentary eyes, where rhabdomeres are commonly folded upon themselves (“self-rhabdome” of Jones et al., 1971), the tips of microvilli make contact with those of the same cell or, rarely, abut the glial cells. Microvilli in all photoreceptors are attached by gap junctions side by side or tip to tip. They form the sole contacting elements between retinula and eccentric cells. The function of these junctions between the surfaces of the same cell is admittedly enigmatic, particularly since all other arthropod rhabdomes, except those of the locust (Shaw, 1967a) and the drone honeybee (Shaw, 1967b), manage to function and cohere without additional junctions. In the compound eye, however, the gap junctions provide for electrotonic coupling in basically the same way as they do in various epithelia and neurons (Bennett et al., 1963, 1967; Loewenstein and Kanno, 1967; Pappas et al., 1971; Hudspeth and Yee, 1973).The ommatidial receptor cells can be represented as an electrotonic syncytium with all retinula cells connected to their nearest neighbors as well as to the eccentric cell (Behrens and Wulff, 1965; Smith et al., 1965; Smith and Baumann, 1969). The physiological effects are a faithful transmission of light-induced depolarization from the retinula cells to the eccentric cell without the delays or potential fatigue encountered in a chemical synapse. The complex elec-

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

309

trophysiological interactions have been discussed in detail by Smith and Baumann (1969). 2. Palisade In the dark-adapted state, the base of all rhabdomeres is occupied by a more-or-less regular system of distended, agranular cisternae (Figs. 13 and 14) that have been referred to in various arthropods as Schaltzone (Hesse, 1901), perirhabdomal vacuoles (Eguchi and Waterman, 1966), subrhabdomere cisternae (Lasansky, 1967), or the palisade (Horridge and Barnard, 1965). This palisade has no continuity with the surface membrane; in fact, its membrane is considerably thinner than that of the microvilli (Fahrenbach, 1969; Jones et d., 1971). Its degree of regularity is very easily altered by fixation, and it is structurally transient. On light adaptation, the palisade is gradually displaced by pigment (in the ommatidium), or at least dispersed in the receptor cell cytoplasm in other locations, a process first explored by Horridge and Barnard (1965) in the locust eye and illustrated in Limulus by Miller (1958). The dynamics of the palisade make the hypothesis of Horridge and Barnard (1965) an attractive one. They suggest that the palisade, by its low refractive index and its position next to the refractile rhabdome, serves to confine light to the inside of the rhabdome in the dark-adapted state. As in the cone, the angle of internal reflection in the rhabdome increases with light adaptation as a consequence of pigment aggregation and thereby increases the light loss out of the rhabdome. A second functional possibility emerges from studies with intracellularly injected aequorin (Lisman and Brown, 1972; Brown and Blinks, 1972; Brown, personal communication). Retinula cells of the ventral eye, which also have a palisade, release calcium ions from an intracellular compartment during light-induced depolarization. The calcium ions, which seem to throttle the sodium influx across the receptor membrane to a steady-state level, are rapidly cleared out of the cytoplasm when stimulation ceases. By analogy to the sarcoplasmic reticulum, calcium pumping by the palisade is a possibility favored by its opportune juxtaposition to the rhabdomere and by the absence of other capacious intracellular compartments. 3. Cytoplasm The cytoplasm of the receptor cells is replete with such diverse and abundant contents that it can support various functional claims, including neurosecretion (Waterman and Enami, 1954). In particular, the cells are filled with large amounts of endoplasmic reticulum (ER) and free ribosomes (Fig. 16). Commonly, the nucleus is surrounded by a wide band of ER (Clark et al., 1969b) in addition to the regular

310

WOLF H. FAHRENBACH

FIG.16. Retinula cell cytoplasm with extensive ER, pigment, and residual bodies. Y. 6300.

peripheral stacks of ER in the ommatidia. A second type of retinula cell in the compound eye (Fig. 12) has been distinguished on the basis of several criteria, particularly a flared rhabdomere, less ER, and more mitochondria than the other retinula cells of the same ommatidium (Fahrenbach, 1969); but no experimental evidence is available to equate this cell type with, say, the 10% of retinula cells that have a distinctive action spectrum (Wasserman, 1969). Two major synthetic processes presumably proceed at the same time: the production of rhodopsin (and rhabdomeral membrane protein) and the synthesis of ommochrome. The first has been investigated by the autoradiographic tracing of tritiated leucine incorporation into the compound eye of Limulus (Bumel et al., 1970). These investigators found a conspicuous light-dark effect. Animals kept in the light for 12 hours, but held in the dark after injection, incorporated 10 to 20 times more radioactivity into the rhabdome than animals exposed to the reverse regimen. This effect was attributed to a possible competition for available ATP between energy demands

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

31 1

of the transduction process in illuminated cells on the one hand, and to metabolic claims of protein synthesis on the other. Much of the synthetic activity of the primary visual cells seems to be directed toward the metabolism of the ommochrome pigment. Membrane-bounded pigment granules are produced by the accretion or growth of small, opaque vesicles, seemingly in the same manner as has been related for type I (ommochrome) pigment in the eye of Drosophila melanogaster (Shoup, 1966; Fuge, 1967). There is a subtle distinction in size, but not in structure, between the pigment of sensory cells (ca. 0.5 pm maximum) and pigment cells (ca. 1.5 pm maximum) (however, see Section IX,A). Degradation of the pigment by polymorphic autophagosomes (type IV granule of Shoup, 1966) commences with the incorporation of one to three pigment droplets in a large vesicle with granular, opaque contents and continues through diverse and bizarre lysosomal forms with granular and myeloid content to a presumptive residual stage characterized by irregular stacks or whorls of membranes. Except for the lack of any oversized structures in this sequence, the entire course mirrors that found during granulolysis in the retinula cells of the mantis shrimp Squilla mantis (Perrelet et al., 1971).The abundance of residual bodies, like that in neurons of the central nervous system, appears to increase with the age of the animal. In the absence of direct experimental investigations, any discussion of the remaining, often copious, inclusions, leans toward educated extrapolation. Various small Golgi-derived dense bodies appear to be lysosomes and enter into the larger cytophagosomes. Coated vesicles are produced to such a degree that, particularly in the rudimentary eyes, they frequently form concentrated aggregates which simulate glycogen (Fahrenbach, 1970a). They are also frequently found in surface continuity with the base of rhabdomeral microvilli, a juxtaposition that has been interpreted as light-dependent pinocytotic uptake in the eye of the crab Libinia emarginata (Eguchi and Waterman, 1967) and, conversely, as a sign of mucopolysaccharide secretion into the periodic acid-Schiff (PAS)-positive rhabdomeral space in Limulus (Fahrenbach, 1969). Large multivesicular bodies are a common attribute of all retinula cells. Exploration of these structures in the eye of a cricket, Pteronemobius heydeni, has shown some continuity between multivesicular bodies and lamellar bodies, which supports their customary link to degradative processes (Wachmann, 1969); however, Eguchi and Waterman (1967) found a striking increase in multivesicular bodies with light adaptation and a progressive decrease in

3 12

WOLF H. FAHRENBACH

the dark in the crab Libinia. In addition to the generally favored lysosomal hypothesis, the apparently inverse relationship between these structures and the existence of the palisade warrants consideration. The multivesicular bodies in retinula cells could form a membrane reservoir of palisade cisternae dispersed into the cytoplasm during light adaptation, possibly in conjunction with lamellar bodies. Because of the irregularities of its adaptation processes, Limulus does not provide a promising model in which to study these phenomena. Retinula cells of the ocelli and lateral rudimentary eyes have an added cytological peculiarity in the form of large homogeneous pools of a featureless, PAS-positive, amylase-resistant substance, evidently a glycoprotein, mixed with a variable amount of glycogen (Fahrenbach, 1970a) (Fig. 19).This storage material, commonly in juxtaposition to masses of coated vesicles, also includes some lipid and lipofuscin and is transported along the large axons of these cells toward the brain. The fluctuating quantities and transport of these inclusions have given rise to two reports that mistakenly attribute neurosecretory function to the lateral rudimentary eye (Waterman and Enami, 1954; Nunnemacher and Davis, 1968). 4. Light Adaptation Only a brief resume of the structural changes during light and dark adaptation is appropriate because the subject has not been explored to any extent in Limulus, nor are its ommatidia, the most active organs in this respect, particularly predictable (Fig. 11). The usual process of aggregation and dispersion of pigment cannot truly be said to exist in Limulus ocular pigment cells and is demonstrable only in modified form in the retinula cells. Rather, pigment granules reciprocate within the attenuated cells as they move to a distal (i.e., dispersed) position in the light and collect basally in the dark. Intrinsic changes in the photosensitivity of retinula cells do not necessarily effect observable changes in their ultrastructure. Light adaptation in the compound eye is demonstrable after 5-10 minutes and consists of a general distal pigment migration which results in a subcuticular position of guanine and a dense pigmentary layer around the cuticular cone. A cufflike concentration of pigment at the level of the cone cells provides, in effect, a diaphragm at the exit pupil of the cone. Pigment and some cytoplasm of the proximal pigment cells move distad into the ommatidium and thereby produce more substantial and heavily pigmented partitions between the retinula cells. The palisade of the retinula cells is displaced by pigment which aggregates in the wedge of cytoplasm between rhabdomeres, a phe-

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

313

nomenon illustrated by Miller (1958) and Bass and Moore (1970). The palisade forms and is dispersed in ocellar, ventral, and rudimentary eye cells, but no pigmentary changes analogous to those in the compound eye are seen. Dark adaptation is a more gradual process fully executed after an hour or more. The systematic changes that may occur in the cytological components, and ultrastructural corollaries of light-dependent protein synthesis (Burnel et al., 1970) in retinula cells, have not been explored.

5. Spectral Response Eyes of the horseshoe crab have diverse receptor cells with different action spectra, but to date no morphological distinctions have been established. Graham and Hartline (1935) were the first to record that about 90% of the lateral eye retinula cells (a cells) are maximally sensitive at 520-530 nm (Wasserman, 1969). This action spectrum coincides closely with the absorption spectrum (520 nm) of rhodopsin extracted from Limulus compound eyes (Hubbard and Wald, 1960). Both the ventral eye (Murray, 1966; Nolte and Brown, 1970) and the “visible” cells of the ocellus (Nolte and Brown, 1969, 1970) have a spectral characteristic similar to that of the a cell except for a slight deviation at long wavelengths (Wald and Krainin, 1963). The second class of ommatidial retinula cells, the /? cell (Wasserman, 1969), was also first noted by Graham and Hartline (1935). It has a broad spectral response with one peak at 525 nm and a second, slightly (1dB) lower one in the far-violet region (357-400 nm) (Wasserman, 1969).About 10%of the retinula cells in the compound e y e display this spectral characteristic, a feature that has been ascertained by single-cell recordings and, hence, is not due to overlapping responses of several cells. A third spectral type is restricted to the ocelli and is primarily an ultraviolet receptor (Wald and Krainin, 1963; Chapman and Lall, 1967). These retinula cells are most sensitive at 360 nm, their sharp peak being about 2 log units higher than that of the “visible” cells (Nolte and Brown, 1969, 1970; Lall, 1970). Visible light suppresses the response, that is, hyperpolarizes the ultraviolet-sensitive cells; the converse happens in the “visible” cells (Nolte et al., 1968). La11 and Chapman (1973) calculated that the median eye of Limulus can detect the near-ultraviolet component of moonlight at a water depth of 20 meters. The photosensitivity of intact lateral eyes extends over 10 log units (Kaplan et al., 1973). Most of this range-equal to the psychophysical perception of the human eye-can be attributed to the retinula cells themselves, since only 1-2 log units variation could be ascribed to

314

WOLF H. FAHRENBACH

pigment migration in a study of screening pigment effects in the moth Deilephila elpenor (Hoglund and Struwe, 1971).

B. ECCENTRICCELLS The eccentric cell of the ommatidium, first clearly discerned by Watas6 (1890), represents a secondary neuron of the optic pathway. Two eccentric cells per ommatidium are common in some animals, but instances of three or none are on record (Hartline and Ratliff, 1972). From its offset position at the base of the ommatidium, the soma sends a long, tapering dendrite through the center of the rhabdome (Figs. 7, 12, and 13). The bulbous tip of the dendrite abuts the cone cells, to which it is attached by adhering junctions (Fig. 18). All adjoining surfaces between the rhabdome are bounded by gap junctions (Lasansky, 1967). In addition, the surface of the dendrite is studded with an estimated 30,000 to 50,000 microvilli (Fahrenbach, 1969), which make similar contacts with the interlocking microvilli of the retinula cell and increase the surface area of the dendrites by

FIG. 17. Eccentric cell cytoplasm with abundant microtubules, clustered ER, and large Golgi bodies. x 15,000.

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

315

FIG. 18. Tip of the eccentric cell dendrite. Its apex abuts cone cells (C) and hemocyanin-filled extracellular space (H). The cytoplasm is filled with faintly stained glycogen, mitochondria, and some ER. A thin terminal web lines the internal surface of the dendrite. x 15,000.

316

WOLF H. FAHRENBACH

FIG. 19. Cytoplasm of rudimentary eye cell with residual bodies and accumulations of mixed glycogen and glycoprotein. x 13,000.

10-50% (Cohen, 1973). The physiological result of these junctions is an electrotonic, recti&ing synapse which transmits the receptor potential of the retinula cells to the dendrite (Smith et d., 1965; Smith and Baumann, 1969). Despite its microvilli, a condition akin to that found in the eccentric cell of the isopod Porcellio scaber (Eakin, 1972), the eccentric cell has no palisade and is not intrinsically photosensitive (Waterman and Wiersma, 1954; Smith and Baumann, 1969). Hence, in addition to providing an extensive junctional surface, the microvilli probably aid in anchoring the dendrite within the rhabdome, a possibility strengthened by the presence of a filamentous terminal web lining the internal surface of the dendrite (Lasansky , 1967). Neuronal attributes dominate the cytoplasmic constitution of the eccentric cell (Fig. 17). At first sight, numerous large Golgi bodies, scattered small arrays of ER cisternae, abundant mitochondria, and

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

317

ubiquitous microtubules differentiate the cell from retinula cells. The dendrite (Figs. 13 and 18)is characterized by compact mixed accumulations of glycogen, mitochondria, and lipid droplets, mostly peripheral in position; evenly distributed microtubules; and agranular cisternae and tubules arranged in a transversely repeating pattern. Pigment is virtually absent but various residual bodies are in evidence, particularly in older animals. The eccentric cell soma is not particularly insulated by its surrounding pigment and glial cells, since hemocyanin often diffuses along intercellular clefts up to the soma in the same manner as it does at the tip of the dendrite. The receptor potential of the retinula cells is conveyed electrotonically via gap junctions to the eccentric cell dendrite. The resultant generator potential of the eccentric cell produces a propagated spike either some distance along the axon (Fuortes, 1959) or among the collaterals of the eccentric cell axon (Purple and Dodge, 1965). The spike, however, does not invade the soma and dendrite except by passive spread. Although several examples of eccentric cells or seemingly similar structures exist, namely in Bombyx mori, the silkworm (Eguchi, 1962), Macrocyclops albidus, a copepod (Fahrenbach, 1964), and Porcellio scaber, an isopod (Eakin, 1972), these cases probably represent separate evolutionary trends without relationship to the condition found in Limulus. C. ARHABDOMERICCELLS Available information on the arhabdomeric cell suggests that this cell of the ocellus is the equivalent of the eccentric cell, hence a secondary receptor neuron. It is located in the receptor layer of the ocellus at a level just below the rhabdomeral zone, or else near the base of the retinula cells. A 100- to 150-pm dendrite ascends into the region of rhabdomeres, where it branches repeatedly. Its terminals are surrounded by rhabdomeral microvilli of adjacent retinula cells. The termini, which like the eccentric cell dendrite contain microtubules, glycogen, and mitochondria, have the usual gap junctions to the adjacent microvilli. The surface of the branches also has a small number of microvilli but no hint of a palisade. The cytoplasm of the soma appears to be identical with that of the eccentric cell, that is, it contains ample glycogen, some ER, large Golgi complexes, and no pigment. Recordings from these cells (Jones et al., 1971) show that they produce spikes in response to ultraviolet light, a characteristic that sets them clearly apart from the two types of retinula cells in the ocellus.

318

WOLF H . FAHRENBACH

Some profiles, suggestive of arhabdomeric cells, are occasionally found in the rudimentary eyes (Fig. 15).Even a few such cells with their presumably spiking axons would largely negate the need for unusual length constants of retinula cell axons, a requirement without which rudimentary eyes would be incapable of transmitting any electrical message to the brain. VI. Basal Lamina and Hemocoel The 0.5- to 3-pm thick basal lamina that clothes the general epidermis follows the surface of each ommatidium and the ocellus, covers the surface of the rudimentary eyes, and is continuous with the laminae of adjacent neural and vascular tissues (Figs. 7,20, and 22). The layer, previously illustrated in various contexts of the visual system (Clark et al., 196913; Fahrenbach, 1969, 1970a,b, 1971; Whitehead et al., 1969), is partly composed of a faintl$ fibrous ground substance, presumably mucopolysaccharide, and embedded fibrils of invertebrate collagen (Harper et al., 1967) which average 100 A in diameter and have a periodicity of about 510 A. The basal lamina does not pose an impediment to the diffusion of even large molecules, since hernocyanin is found in most cellular interstices in the photoreceptors. Some small nerves, particularly efferent ones, are surrounded by 30 or more roughly concentric laminae of faintly fibrous glycocalyx without an admixture of collagen (Fig. 23). The hemolymph contains at least 1% (wlv) of the dissolved respiratory protein, hernocyanin, as toroidal granules 190 A in diameter ( F e m b d e z - M o r h et al., 1966; Fahrenbach, 1970b). In many instances the hemocyanin serves as a natural marker for patent diffusion channels in photoreceptors. Thus it enters ommatidia through the basal lamina and diffuses up to the adhering junctions at the periphery of the rhabdome and often partly into the rhabdome at the apex and base of the ommatidium. That this infiltration is not artifactitious can be ascertained by the fact that hemocyanin frequently stagnates and becomes more concentrated in the interstices than in the general circulation and thereby gives rise to the “granular ground substance” described by Lasansky (1967) (Fig. 18). Rare, free-floating crystals of hemocyanin resembling bundles of microtubules result from the normal breakdown of cyanocytes. A second large (100-A) molecular species, a hemagglutinin (Femhdez-Morhn et al., 1968), cannot be differentiated from hernocyanin in sectioned material, although it is distinct after negative staining. It constitutes about 5 % of the hemolymph protein.

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

3 19

FIG.20. Small blood vessel near the ocellus. A thick lamina coats the endothelial cells, which are surrounded by several connective tissue layers. A small nerve (N; see Fig. 23) accompanies the vessel. X 1200. FIG.21. A nearly 100-pm long cyanocyte in a circulatory space. Beyond this stage the cell breaks up and liberates the contained hemocyanin. x 1200.

Two principal cell types circulate in the bloodstream and are frequently intimately associated with various parts of the visual system, particularly the lateral eye plexus, even though they remain separated from adjacent tissue by a basal lamina. Most numerous (92-99%) are the granular, ameboid hemocytes responsible for the clotting of blood by aggregation and release of their granules (Loeb, 1928; Holme and Solum, 1973; Dumont et al., 1966) (Figs. 10 and 22). Within a minute after the blood is exposed to air or seawater, the

320

WOLF H. FAHRENBACH

homogeneous content of the dense, large granules is changed to a microtubular and subsequently granular one, which is discharged from the cell. Simultaneous ameboid activity results within 3 minutes in a seemingly new cell type (type H2of Herman and Preus, 1972), devoid of specific granules but full of swollen cisternae and a variety of inclusions (Clark et al., 1969a) (Fig. 10). The mode and speed of dissection determine the abundance of degranulated hemocytes. The granules of hemocytes, surrounded by a conspicuously asymmetric membrane, have been mistaken for pigment and neurosecretory material where these cells have squeezed themselves into narrow interstices of adjacent tissues. The second blood cell type, occurring with a frequency of less than 1-8%, depending on molting stage, forms hemocyanin (Fahrenbach, 1970b). The youngest identifiable stage, the cyanoblast, is distinguishable from a degranulated hemocyte by its conspicuous content of free ribosomes. With maturation, the cell accumulates hemocyanin in crystalline form until ultimately, as a cyanocyte, it resembles a large crystal, up to 100 pm long, with an appended nucleus and minimal cytoplasm (Fig. 21). The breakup of this cell results in the previously mentioned crystalline hemocyanin fragments which gradually disperse into individual molecules. Cancellous vascular spaces with extremely thin partitions, the expression of a true open hemocoelic system (Figs. 2 and lo), surround the eyes and fill the spaces between photoreceptors and adjacent organs. The partitions are commonly composed of a thin central fibrous lamina and a cellular lining on both sides. The highly attenuated lining cells are probably identical with the tissue reserve cells (R cells) described by Schlottke (1934, 1935) and Herman and Preus (1972) in connection with the hepatopancreas in Limulus. Depending on the nutritional state of the animal, the reserve cells are more or less filled with the conspicuous rosettes of *glycogen, variable numbers of large homogeneous lipid droplets, numerous mitochondria, and an assortment of dense bodies and vacuoles (Figs. 10 and 27). Despite the hemocoelic nature of the circulatory system, progressively smaller branches of the sturdy muscular arteries can be traced between sinusoidal spaces. Only thin-walled “capillaries” are found in immediate proximity to the eyes, often in the space between adjacent ommatidia. These vessels, mentioned by Whitehead et al. (1969),have a wall one to two cells thick and a ridged internal surface covered by a deep lamina similar to the fibrous lamella lining the lumen of large arteries (Dumont et al., 1965) (Fig. 20).

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

32 1

In the peculiar internal anatomy of Limulus, its optic nerves penetrate the hepatopancreas and gonad. Several articles describe the unusual histology and cytology of these organs (hepatopancreas: Schlottke, 1935; Herman and Preus, 1972; Johnson et al., 1973; ovary: Munson, 1898; Dumont and Anderson, 1967; testis: Benham, 1883; Fahrenbach, 1973a). VII. Axons and Plexus A. AFFERENT AXONS

With few exceptions the cytology of the visual axons is remarkably uniform (Nunnemacher and Davis, 1968; Clark et al., 1969b; Fahrenbach, 1970a, 1971). Their cytoplasm is filled with quite regularly spaced, longitudinal microtubules with a center-to-center spacing of 1000-1500 A. Mitochondria, mostly in a size range of less than 0.2 pm, are distributed toward the periphery of the fiber. Scarce agranular reticulum occurs in some axons (eccentric cell) in a transverse disposition of cisternae, which is also seen in the eccentric cell dendrite (Gur et al., 1972). Several of the axonal types contain further specialized contents, some of which have diagnostic value but are usually enigmatic as to function. Eccentric cell axons often carry small, compact masses of a PAS-positive, diffusely fibrillar material, apparently a glycoprotein. Chains of lipid droplets lie in an occasional retinula cell axon of the lateral optic nerve, virtually occluding the entire cross section of the fiber. An extremely rare fiber type in the lateral optic nerve, intermediate in size between eccentric and retinula cell axons, is filled with mitochondria and dense bodies and may represent the tip of a growing retinular axon (Fahrenbach, 1971). Profuse quantities of aglycogen and glycoprotein, produced in the lateral rudimentary eyes, travel along the large axons of these cells and have contributed to a mistaken diagnosis of neurosecretion (Waterman and Enami, 1954; Pannesi, 1964; Nunnemacher and Davis, 1968). The diameter of visual fibers, although typical for a given cell type of origin, is unfortunately not related to its functional properties (Fig. 22). Action potentials have been recorded from eccentric cells, arhabdomeric cells, and scattered retinula cells in the ocellar nerve (Waterman, 1953). The axons of the latter two cells have a 5-pm diameter, whereas those of eccentric cells increase from a minimum of 1 pm in a young animal to an average of 10-15 p m in an adult. A similarly large diameter has been recorded for the axons of retinula

322

WOLF H. FAHRENBACH

FIG.22. Juxtaposition of eccentric cell (E) and retinula cell (A) axons in the lateral optic nerve. Both contain microtubules and are enveloped by sheaths of glia and basal lamina. Two neuroglial nuclei are shown. Hemocytes lie in the intervening circulatory space. X 17,000.

cells in the median and ventral eyes (Jones et al., 1971; Clark et al., 196913). The fibers of the lateral rudimentary eye are initially the largest in the lateral optic nerve (3-10 pm) but increase more slowly in diameter than eccentric cell axons to a diameter of 10-25 pm (Waterman and Wiersma, 1954; Fahrenbach, 1971). Since the initially segregated grouping of rudimentary eye fibers in juvenile animals is not maintained in the adult nerve, the axons can be distinguished from eccentric cell fibers only by unusually large size or by the previously mentioned inclusions. All these fibers are enveloped in one or more neuroglial cells, and

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

323

each is surrounded by a separate thick external lamina. In contrast, the retinula cell fibers of the compound eye, measuring only 0.5-1.5 prn in diameter, are bundled into fascicles of two or three directly adjacent fibers, invested by one to several glial cells. This arrangement develops from larger bundles of 6 to 10 or more fibers in a young animal. The one-to-one correspondence between spiking axons in the lateral optic nerve and eccentric cells was established by Waterman and Wiersma (1954). However, identification of retinular fibers in the optic nerve depends on their correct numerical ratio in relation to eccentric cell fibers and their size, given the absence of detectable electrical activity in them. The small admixture of efferent and rudimentary eye axons can be singled out only by optic nerve section and pileup of secretory products in the proximal and distal stumps, respectively (Fahrenbach, 1971; Pannesi, 1964). Spiking in the nerve of the ocellus, first observed by Waterman (1953), is mostly attributable to arhabdomeric cell fibers, although retinula cells have not been ruled out (Jones et al., 1971) and ectopic retinula cells within the nerve seem to produce propagated action potentials (Waterman, 1953).Despite intensive studies on the physiology of the ventral eye (Millecchia and Mauro, 1969a,b; Lisman and Brown, 1972; Yeandle and Spiegler, 1973), no action potentials have been detected in its nerve. The ventral eye plays a definite role in visual behavior (Wasserman, 1973a) notwithstanding the 2- to 3-cm-long nerve, a length that would drastically attenuate signal transmission by electrotonic spread. Both lateral and median rudimentary eyes produce receptor potentials (Millecchia et al., 1966)but do not give rise to propagated spikes. Their functional role is unknown. Numerical axonal input to the brain increases with age but can be approximated for an animal of 10-cm prosomal width as follows (Nunnemacher and Davis, 1968; Fahrenbach, 1971) (both sides combined) : Compound eyes Eccentric cells Retinula cells Lateral rudimentary eyes Ocelli Retinula cells Arhabdomeric cells Endoparietal eye Ventral eyes

1,200 16,000 200 300 50 100 600

324

WOLF H. FAHRENBACH

B. EFFERENTAXONS All Limulus photoreceptors are supplied with efferent, neurosecretory fibers which originate in the central nervous system (FahrenThis peculiar innervation appears to be unique bach, 1970c, 1973~). to Limulus with the possible and unconfirmed exception of neurosecretory fibers seen in the eye of the honeybee by Baumann (personal communication, cited in Clark et al., 1969b). Less than 10 small (1-to 2 - p ) axons course in the lateral optic nerve to supply the compound and lateral rudimentary eyes; the numerical complement of efferent fibers in the other nerves has not been determined. Only rarely does one encounter the characteristic neurosecretory granules -opaque, cylindrical bodies with a period substructure - in the nerve, but ligature or transsection produces rapid proximal pileup of the secretory product (Fahrenbach, 1971) (Fig. 25). This procedure also causes degranulation and gradual degeneration of the terminals. In the plexus of the lateral eye, efferent fibers are conspicuous by

FIG. 23. Small nerve containing efferent axons (A). The enveloping neuroglial cells have numerous adhesion plaques which face the surrounding, lamellated sheath. x21,Ooo.

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

325

their increased number, presumably the result of progressive branching, and by their content of elementary granules (Fig. 27). Each ommatidium receives about 70 minute efferent axons which terminate in approximately equal numbers on the proximal pigment cells, intraommatidial pigment cells, and adjacent epidermal cells, including guanophores. The fibers travel to their destination in the partitions between retinula cells and at the periphery of the ommatidium and occasionally form short commissural tracts between adjacent ommatidia. The terminals, basically identical with the “synaptoid” neurosecretory endings of insects (Scharrer, 1968),end on the surface of ommatidial pigment cells and on receptor cells of rudimentary, median, and ventral eyes; often, however, they indent or deeply invaginate into the target cell (Clark et al., 1969b; Fahrenbach, 1969, 1970a, 1973c) (Fig. 24). In all cases the basal lamina is penetrated. The synaptoid terminal is distinguished by a presynaptic, submembranous density surrounded by some clear vesicles and many elementary granules, but no postsynaptic specialization is present. In retinula cells of the ventral and, in extremely rare instances, the compound eye, efferent axons end in immediate juxtaposition to the rhabdomere. The region of the axon hillock in retinula cells of the lateral rudimentary eye is occasionally invaded by terminals. Cells that seem to be morphologically unresponsive to changes in illumination, that is, guanophores outside the compound eye, epidermal cells of the ocellus, and eccentric cells, lack an efferent supply. The origin of these fibers is uncertain, but circumstantial evidence points to a group of large neurosecretory cells in the central body, the protocerebral center on which all visual inputs converge (Fahrenbach, 1973c) (Figs. 30 and 32); however, neurosecretory cells elsewhere in the nervous system (Herman, 1970, 1972; Herman and Preus, 1973) cannot be excluded. Efferent innervation probably governs the rapid phase of pigment migration, which occurs during light adaptation. Chromatophorotropic hormones in crustaceans are generally responsible for ocular pigment migration (Kleinholz, 1966)and have, in fact, been extracted from the central nervous system of Limulus (Brown and Cunningham, 1941; Fingerman et al., 1971); however, the brain is only one-third as active as the subesophageal ganglion. In any event, by combining attributes of both neurohormonal and neurohumoral control systems, the efferent supply to the eyes is a typical example of the intermediate mode of neuroendocrine interaction (Scharrer, 1972).

326

WOLF H. FAHRENBACH

FIG.24. Several efferent fibers near their terminations in proximal pigment cells. The largest terminal illustrates several views of the granules and a synaptoid junction (arrow). x 19,OOO. FIG.25. Proximal stump of a severed optic nerve (3days after operation), showing the nerve sheath, degenerating axons and glia, and pileup of granules in an efferent fiber. C, Collagen, F, fibrocyte. X8000.

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

327

C. PLEXUS

1. Architecture Both the median ocelli and the compound eyes have a plexus, that is, a zone of interconnecting and synapsing receptor axon collaterals below the layer of retinula cells. Since the median eye plexus has not been described, the following discussion deals exclusively with the plexus of the compound eye. The open architecture of this region combined with its volume, 75 mm3 or more in a large animal, provides a severe impediment to any structural studies in which precise fiber tracing is contemplated. The important process of lateral inhibition, discovered by Hartline (1949) and subsequently extensively explored by Hartline and his collaborators (see reviews in Ratliff, 1965; Wolbarsht and Yeandle, 1967; Hartline, 1969; Knight et aZ., 1970; Hartline and Ratliff, 1972), takes place within the plexus and involves the reciprocal inhibition of impulse rate of one ommatidium by the activity of its neighbors. Axons descend from the ommatidium singly at first and lie in a partly delimited vascular space corresponding to the periphery of the ommatidium before becoming grouped into small bundles of a few fibers (Hartline et aZ., 1961; Miller, 1965). Synaptic neuropile appears slightly below the ommatidia and is especially conspicuous at the intersections of the predominantly vertical visual fibers and the more or less horizontal fiber tracts that link neighboring ommatidia (Fig. 26). The interconnections, 10-30 pm in diameter, may lie anywhere between the base of the ommatidia to a millimeter proximal to that. Beyond that level the optic fibers aggregate into several nerves, which pass through a cribriform, cuticular plate separating the inside of the adult compound eye from the body cavity. Below the plate all fibers coalesce into the lateral optic nerve. The branching pattern of the visual axons has been investigated by Procion Yellow injection (Schwartz, 1971) and by serial sectioning for electron microscopy (Gur et aZ., 1972). The value of the former technique is limited by the uncertainty of whether the finest axonal branches are indeed labeled. Serial sectioning, although theoretically the ultimate method of ascertaining pre- and postsynaptic contributions to the plexus, is impeded by the practical barriers of continuous tracing through an uninterrupted series. For example, Gur et al. (1972) collected 2000 to 3000 sections from the plexus region underlying each of several ommatidia, lost about 20% of the series, and assembled in montage about 5%. Large collaterals can be traced by eye through incomplete stretches of the series, but the continuity of

328

WOLF H. FAHRENBACH

,

FIG. 26. Horizontal section through the plexus of the compound eye. The dark knots of neuropile (see Fig. 27) correspond approximately to the position of overlying ommatidia. The thick interconnecting tracts of axons are not necessarily as unidirectional as shown here. X280.

the smallest branches becomes ambiguous with the loss of even one or two sections. Injected preparations of the retinula cell axons show only small (ca. 5-pm) spines and no long collaterals. From serial sections it appears that a short distance below the ommatidium the retinula cell axons of that ommatidium enter into a mutual synaptic plexus by way of collaterals, without the apparent participation of the eccentric cell or contributions from adjacent ommatidia. This region has only a very limited vertical extent. The eccentric cell axons, in contrast, contribute the main substance of the plexus by long and short collaterals. The long collaterals, mainly traceable in Procion Yellow-injected preparations (Schwartz, 1971), extend through the previously mentioned interconnections and terminate in proximity to axons from adjacent ommatidia. They usually measure less than 1 pm in diameter and have been traced over a distance of 200 pm, which just about corresponds to the in-

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

329

FIG.27. Cross section of a synaptic column in the plexus underlying an ommatidium. A, Efferent fiber; N, neuroglial cell; R, reserve cell; H, hemocytes. ~10,OOo.

terommatidial spacing in adult animals. However, much longer collaterals probably exist. In a study of inhibitory fields, which are presumably a direct reflection of long eccentric cell collaterals, Barlow (1969)found that inhibitory interaction reaches a peak at a distance of 3 to 5 ommatidia and approaches zero at about 12 ommatidia. These inhibitory fields are at their widest in a horizontal plane (anteroposterior on the eye) and are compressed vertically. No morphological information over a comparable spatial extent is available. The short eccentric cell axon collaterals (Miller, 1965; Schwartz, 1971; Gur et al., 1972) form a more-or-less continuous neuropile sleeve around the axon through the depth of the plexus (Figs. 27 and 28). This neuropile generally starts just beyond the level of the axon hillock and receives contributions from sizable branches of the eccentric cell axon. These give off clusters of smaller offshoots, which are highly irregular in course and diameter. Apparently, many collat-

330

WOLF H. FAHRENBACH

erals synapse with others of the same eccentric cell, particularly in the distal part of the plexus, that is, close to the ommatidia. Contributions of the retinula cell collaterals to the eccentric cell plexus are minimal, if present at all. The distribution of the interacting collaterals supports the contention that the inhibition is the recurrent type, that is, the inhibitory influence is exerted at or near the site of impulse generation (Ratliff, 1965). Thus impulse reduction in an inhibited ommatidium is effected by a lowered level of impulse generation rather than by the suppression of impulses in an existing spike train. The contents of the collaterals of either type consist of microtubules, abundant mitochondria, various synaptic and related vesicles and minor amounts of glycogen. Golgi bodies are found only in adjacent glial cells; hence, they cannot be presumed to play a role in the neuropharmacology of the synaptic regions (Adolph and Tuan,

1972). 2. Synapses The synapses in the plexus were first studied by Miller (1965), who described peculiarly thickened membranes at presumptive sites of inhibitory synaptic interaction without any evidence of subsynaptic specializations. This peculiarity, although apparently a differentiated feature of limited areas, can be demonstrated only with the fixative used (10%glutaraldehyde) and has not been observed by subsequent investigators. In any event the tight junctions described by Miller (1965) as presumptive sites of excitatory interaction must be viewed with some reservation in view of the ease with which they can be produced artifactitiously. A more orthodox picture, akin to that of primary receptor synapses in insects (Trujillo-Cenbz, 1965; Trujillo-Cenbz and Melamed, 1967), emerges from the study of Whitehead and Purple (1970).An opaque hillock and a diffuse synaptic ribbon are attached to the presynaptic membrane at the trigonal point of one presynaptic and two postsynaptic profiles. A zone clear of synaptic vesicles surrounds the ribbon. Quite commonly, two reciprocal synaptic elements face each other. This particular specialization, unusual if not unique for a chemical synapse, can be construed as forming the substratum for (1) the process of self-inhibition of an ommatidium (Purple and Dodge, 1966), provided the two contributing collaterals originate from the same ommatidium, or (2) reciprocal lateral inhibition, if components from two different ommatidia are involved (Gur et al., 1972). Not only do synapses lie between small fibers, but large (2-pm) axons can

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

33 1

FIG. 28. Synaptic detail of the plexus. One axon collateral (A) has conspicuously larger synaptic vesicles than average. This and the preceding figure are indicative of the relative sparsity of dense vesicles. Arrows point at synapses. N, Neuroglial cells. x 25,000.

332

WOLF H. FAHRENBACH

provide en passant synapses to adjacent small collaterals or, conversely, small branches may synapse on the eccentric cell axons, even occasionally near the axon hillock. The great majority of synaptic terminals are filled with round, clear vesicles about 400-500 A in diameter (Miller, 1965) (Fig. 28). Conspicuous efferent fibers, characterized by the typical angular granules described for efferent terminals, transverse the plexus without synapsing. In addition to these vesicles, many terminals contain a variable admixture of dense or dense-cored vesicles (Adolph and Tuan, 1972). A rather rare type of synapse is populated by large and irregular vesicles which average about 800 A in diameter, Subtle differences in synaptic morphology may exist but have not been explored. Whether the lateral excitatory interactions that have been detected in the eye (Tomita e t al., 1960; Purple and Dodge, 1965) can be attributed to some of the less common synapses is unknown.

3. Pharmacology The neuropharmacology of the plexus has been approached by micropipet injection of active substances into the lateral eye (Behrens and Wulff, 1970; Adolph, 1966), as well as by a combination of these and biochemical techniques (Adolph and Tuan, 1972). Unless certain experimental conditions are observed, for example, simultaneous recording of the output of interacting ommatidia or inhibition of a test ommatidium by antidromic stimulation of neighboring fibers, it is difficult to sort out synaptic influences from relatively unspecific membrane effects that derive from unphysiological, topical concentrations of an applied pharmacological agent. Depending on their specificity, metabolic poisons of known steps in the synthetic pathway of a putative transmitter are less likely to confuse the picture with direct neurophysiological sequelae. Both y-aminobutyric acid (GABA) and the GABA inhibitor picrotoxin affect the spike action potentials of an ommatidium (Adolph, 1966; Behrens and Wulff, 1970) by depressing or only slightly increasing frequency, respectively. However, inhibition by antidromic stimulation is blocked by the protracted action of 0.1 mM picrotoxin. Brief (30-minute) exposure of the eye to aminooxyacetic acid, an inhibitor of GABA metabolism, does not produce marked effects on lateral inhibition. Serotonin (5-HT) also depresses ommatidial spike activity (Behrens and Wulff, 1970; Adolph and Tuan, 1972). The increase in spike activity observed by Adolph (1966) has been attributed to the

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

333

phosphate buffer vehicle (Behrens and Wulff, 1970). After the application of 5-HT, the resultant depression of lateral inhibition recovers more slowly than the firing rate of the affected ommatidia, an indication of a specific synaptic effect in addition to any nonspecific influence on the cell membrane. The eye does not respond to monoamine oxidase inhibitors nor does it contain the normal (vertebrate) 5-HT metabolite 5-hydroxyindole acetic acid, according to Adolph and Tuan (1972). These investigators, however, found a seemingly higher concentration of melatonin than 5-HT, its metabolic precursor. In physiological concentrations melatonin has no effect but may be associated with ocular pigment migration. With either GABA or 5-HT, the inhibitory effect is most pronounced if the region of application lies proximal to the test ommatidium by 0.5-1 mm. Epinephrine, norepinephrine, and various amino acids affect the spike output of ommatidia (Adolph, 1966; Behrens and Wulff, 1970), but their participation in any snyaptic mechanism in the plexus is unconfirmed. Dilute solutions of ethanol abolish lateral inhibition (MacNichol and Benolken, 1956) without depressing the spike activity, an observation that sheds little light on synaptic pharmacology, useful though it is to neurophysiological studies. For a convincing solution of this problem, ultrastructural studies will have to be combined with autoradiographic, histochemical, and biochemical investigations. VIII. Optic Nerves The lateral optic nerves, the only ones studied in any detail (Pannesi, 1964; Nunnemacher and Davis, 1968; Fahrenbach, 1971), contain fibers from the lateral rudimentary and compound eyes. They travel forward, then turn ventrad and backward to enter the optic laminae at the dorsal anterolateral bulges of the brain. The ocellar nerves loop together usually to the right side of the underlying digestive tract and enter the brain in the dorsal midline. The ventral eye nerves course directly backward and join the ganglionic mass of the lateral eyes at the level of the medulla (Fig. 29). Limulus nerves have the peculiar attribute of being enclosed by an arterial wall. Hence the nerve sheath is more complex and substantial than would be expected of an epineurium. Its structure, detailed by Dumont et al. (1965), consists of alternating layers of fibrocytes and extracellular material (Fig. 25). The interdigitated cells have conspicuous Golgi cistemae but are otherwise rather undistinguished. The intervening acellular layer, termed the fibrous la-

334

WOLF H. FAHRENBACH

mella, is made up of a medial layer of fine collagen fibrils and a peripheral thick external lamina of the adjacent fibrocytes. Scattered muscle fibers in the wall of optic nerves display the characteristics of arthropod visceral muscles (Fahrenbach, 1967; Dewey et al., 1973), that is, a sarcomere length of about 10 pm and a myofilament array of one myosin filament being surrounded by 10 to 12 actin filaments. Both outer and inner surfaces of the vessel wall are covered by a basal lamina which faces the circulatory space. A deviant nerve type is found near the lateral rudimentary and ocellar eyes, often containing only efferent axons (Fahrenbach, 1970a). Here a small number of fibers are carried in an irregular, fluted, and interlocked array of neuroglial cells (Fig. 23). These secrete a cylindrical sheath, up to 10 pm thick, composed of multiple

FIG. 29. Dissection of the brain in an oblique, ventral view. Almost the entire mass consists of corpora pedunculata. The circumesophageal ring starts at the upper left and right. A, Lateral optic nerve; B, ventral eye nerve; C, frontal organ nerve; D, ocellar nerves; E, stomodeal nerve. X40.

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

335

external laminae without any participation of collagen. The fluted glial cells embedded in this sheath exhibit numerous conspicuous hemidesmosomes.

IX. Optic Centers The various optic ganglia have received nowhere near the same attention as the eyes. The principal publications outlining the internal anatomy of the protocerebrum, that is, the purely sensory anterior region of the brain, are those of Packard (1891), Viallanes

FIG.30. Representative neurons of the optic ganglia and central body in frontal section. 1, Ocellar nerves; 2, ventral eye nerve; 3,compound eye nerve; 4, lamina; 5, chiasma; 6, medulla; 7,corpora pedunculata; 8, optic tract; 9, central body; 10, ocellar ganglia; 11, ectopic retinula cells; 12,globuli cells; 13, neurosecretory celIs; 14,protocerebral neurons with axons into both sides of circumesophageal ring; 15, medullary axon to contralateral medulla; 16,fibers to circumesophageal ring and corpora pedunculata. [Modified after Hanstrom (1926a) and Patten (1912).]

336

WOLF H. FAHRENBACH

(1893), Patten (1893), Patten and Redenbaugh (1900), Hanstrom (1926a,b), and Johansson (1933).A detailed study has been directed at the projection of the lateral eye mosaic onto the first visual ganglion and the subsequent processing of visual information (Snodderly and Barlow, 1970; Snodderly, 1971). Only a brief coverage is given here (Fig. 30). The lateral optic nerve commonly breaks up into a number of bundles before entering into the first optic ganglion (lamina) (Fig. 31).Visual fibers, presumably of eccentric cell origin, form optic cartridges (Hanstriim, 1926a), and some continue after synapsing into the second ganglion (Patten, 1912; Snodderly, 1971). Horizontal bands of ommatidia are projected in virtually continuous fashion onto the lamina, their relative dorsoventral positions coinciding between ommatidial origin and laminar termination (Snodderly and Barlow, 1970).Anteroposterior projection may also occur but has not been explored in detail. Postsynaptic off-responses, akin to those in vertebrate ganglion cells, are generated in the optic lamina (Wilska and Hartline, 1941; Snodderly, 1971).

FIG. 31. Horizontal section through the lateral optic ganglia. Orientation of the midline is indicated by the long arrow. N, Lateral optic nerve; L, lamina neuropile; C, chiasma; M, medulla neuropile; small lobe of corpora pedunculata is at upper left. x 110.

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

337

The lamina is connected to the medulla (second optic ganglion) by way of a chiasma. The crossing of fibers has been illustrated by Hanstrom (1926a),but occurs in a plane tilted from exact frontal; hence it is not particularly striking in most sections (Figs. 30 and 31). The neuropile of the medulla is stratified into four dense layers which alternate with three intervening less compact ones (Hanstrom, 1926a). Ganglion cells cover part of the surface of the medulla and range from the size of minute (ca. 10 pm) association neurons, called globuli cells, to a cluster of cells up to 70 pm in diameter. The ventral eye nerves enter the nervous system at the level of the medulla,

FIG.32. Horizontal section of the central body. The vascular spaces around the central body contain hemocytes and a trabecular meshwork of neuroglial cells. M, Optic medulla; N, ocellar nerve; 0, ocellar ganglion; C1, central body neurons (globuli cells); C2, protocerebral neurons; C3, neurosecretory cells. ~ 9 0 .

338

WOLF H. FAHRENBACH

which is connected to the central body by a sizable tract. Neurophysiological recordings from the medulla (Snodderly, 1971) have produced on-units which display spatial summation and a receptive field of about half the compound eye; off-units firing for at most 1-2 seconds after the cessation of a light stimulus; extended-response units which fire both during and after the stimulus; delayed-off units responding for up to a minute after termination of light; and neurons with various rhythms subject to modification by light stimulation. Fibers from the medulla penetrate into both sides of the circumesophageal ring and the contralateral medulla. The corpora pedunculata, which contain most of the neurons (5 x log in a 10-cm animal) and occupy 80% of the volume of the brain, also have connections with the second optic ganglion (Hanstrom, 1926b; Fahrenbach, 1973b). Ectopic retinula cells, associated with the ventral eye nerve, frequently lie underneath the sheath of the brain in immediate proximity to the medulla. Presumably, these cells are responsible for light responses recorded from the medulla severed from its laminar input (Snodderly, 1971). Fibers from the ocellar complex terminate in two small ganglia located between the anterior horns of the central body (Figs. 30 and 32). Small secondary neurons of these ganglia synapse within the central body. This structure is a horseshoe-shaped mass of neuropile, dorsal and superficial in location, its two free arms curving ventrad and approaching each other in the midline (Fig. 32). It is covered with globuli cells on its dorsal and lateral surfaces, receives the previously mentioned visual inputs, and has ventrally broad connections with the protocerebral neuropile, the stalks of the corpora pedunculata, and the circumesophageal ring.

X. Miscellaneous Aspects A. ABNORMALITIES

Limulus appears to have an unusual propensity toward developing supernumerary eyes. Among 190 animals, Hanstrom (1926a) found six specimens with three ocelli. The third ocellus was always located behind the normal median eyes in the midline, had its own lens, and gave rise to two optic nerves. This condition, reminiscent of the nauplius eyes of crustaceans and the ocelli of insects, may represent a not so “rudimentary” condition of the endoparietal eye, which is normally bipartite, including its nerve. In young Tuchgpleus tridentutus, an Asiatic xiphosuran, a transparent cuticular region overlies the endoparietal eye (Waterman, 1953).Four ventral eyes in a side-

T H E VISUAL SYSTEM OF T H E HORSESHOE CRAB

339

FIG.33. Pigment cell partition in the ommatidium of an albino Limulus (compare to Fig. 14). The retinula cells are normally pigmented, but the pigment cells contain virtually no pigment and are thickened. Their content is apparently of lysosomal nature. x6500. FIG.34. An encysted metacercaria of Microphallus limuli surrounded by globuli cells of the corpora pedunculata. ~ 4 5 0 .

340

WOLF H. FAHRENBACH

by-side position with four separate nerves have been observed (J. E. Brown, personal communication). A far more exotic variant has been reported in the form of an animal with an extra pair of compound eyes, located on a single, 2cm-high solid stalk on the dorsal carapace, 3 cm off the midline (Barlow and Kaplan, 1972). These eyes, about two-thirds the size of the lateral eyes, produced pronounced activity in their optic nerves, and as in the lateral eyes, a changed level of illumination affected the heart rate. Two types of rare white-eyed mutants have been found on the western coast of Florida. The eyes of the first, totally unpigmented type have been explored in considerable physiological detail by Nolte and Brown (1970) because the absence of shielding pigment makes it possible to determine pure action spectra unaffected by the passage of light through ommochromes. The eyes of the second mutant (Fahrenbach, unpublished observations; live animal provided by courtesy of J. E. Brown) were light brown and showed normal, possibly increased pigmentation of the retinula cells, and normal guanophores, but highly aberrant ommatidial pigment cells. These had very few, generally small, ommochrome droplets but an abundance of cytophagosomes of the type described in normal eyes. Pigment cell partitions between retinula cells (Fig. 33) were much thicker than in control animals, an observation that suggests increased cytoplasmic migration of proximal pigment cells into the ommatidium in response to sustained photic stress unalleviated by the usual adaptive pigment migration. The overall architecture of the ommatidia and the efferent innervation appeared normal.

B. PATHOLOGY Most horseshoe crabs larger than 20 mm in width (about the fourth molting stage) are infested with the metacercariae of a digenetic trematode, Microphallus ZimuZi (Fig. 34). The life history of this fluke, unraveled by Stunkard (1951, 1953, 1968), starts in the snail, Hydrobia rninuta, which is invaded by the miracidium and harbors two generations of sporocysts in its hernolymphatic sinuses. The cercariae released by the snail invade young LimuZus (even animals of 3to 4-mm width in an experimental situation) and encyst as metacercariae. These ovoid bodies, 150-200 pm in length, are found mainly in interstitial connective tissue but also commonly in the lateral eye plexus and brain. Encapsulation of the cysts by host cells is minimal, and no foreign body reaction or tissue disruption except physical displacement has been observed. A foreign body reaction, however, has

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

341

been observed in Limulus in another context (Loeb, 1902). After consuming horseshoe crabs, the herring gull Lams argentatus becomes the final, natural host to the fluke. In old, infrequently molting animals and in aquarium-stored specimens, the cuticle, particularly that over the compound eyes, is attacked by bacteria and fungi. The former generally erode small pits in the cuticle (Fig. 2), whereas a fungal infestation can penetrate the generally intact cuticle, invade the ommatidia (Fig. 5), and cause profound local destruction and possible secondary bacteremia, which would cause death by generalized intravascular clotting (Bang, 1956).Chitinolytic fungi of arthropods have been studied primarily in connection with the recent outbreak of the European crayfish plague (Unestam, 1965, 1968). A single instance of an abnormal cuticular concretion at the anterior border of the brain has been described by Hanstrom (1926~). Since injury responses can frequently be put to use in neuroanatomical tracing, several largely unpublished observations may provide some useful information. Individual ommatidia or small groups of them can be extirpated by laser irradiation through a suitable aperture. Affected ommatidia are marked by a rapid invasion of hemocytes, and subjacent synapses show degenerative changes by an increase in electron opacity. Although this approach might lend itself to the morphological tracing of the inhibitory field, it appears unsuitable to the study of projection to the optic ganglia since severed axons maintain an essentially normal appearance for many months except for an increase in subsurface cistemae and some degeneration near the cut (Fahrenbach, 1971). Synapses of severed efferent optic fibers become degranulated within a few days but involute at an extremely slow rate (at 13°C) (Fahrenbach, 1971). Similarly, a retrograde injury reaction in peripheral visual or central neurons has not been detected despite considerable effort (Fahrenbach, 1973c) and may unfortunately occur to an experimentally useful degree only in insects among the arthropods (Cohen and Jacklett, 1967; Boulton, 1969; Young et al., 1970).

XI. Vision and Behavior To prevent the further dissemination of apocryphal tales about the total or partial blindness of Limulus, its light-induced responses and behavior should be discussed here. The earliest studies were concerned with phototaxis and demonstrated, even from a retrospective standpoint, a perplexing array of factors that influence these seem-

342

WOLF H. FAHRENBACH

ingly simple reactions. Newly hatched animals are positively phototactic when swimming (ventral eye up), but negatively so when crawling or in warm water (Loeb, 1893). Older animals are positively phototactic and exhibit circus movements with unilaterally occluded eyes (Cole, 1923; Wolf and Zerrahn-Wolf, 1937), but their behavior is easily reversed by such diverse influences as hunger, being tied by the tail, protracted darkness, “fright,” or no apparent reason at all (Cole, 1924; Northrop and Loeb, 1923; von Campenhausen, 1967). Unrestrained animals moving on the ocean bottom at a shallow depth execute rapid turns toward the light when they are being shaded on one side (Adolph, 1971). The effect of ocelli versus lateral eyes in positive phototaxis was explored by La11 and Chapman (1973), who found that one ocellus or one lateral eye alone is sufficient to mediate the response. As might be expected from the neurophysiological characteristics of ocellar receptor cells, the ocelli influence the behavior most markedly in ultraviolet light but less so under full sunlight, in which the ultraviolet-sensitive cells might be partially inhibited. La11 and Chapman (1973) suggested that the extraordinary sensitivity of the ocelli to ultraviolet light and the rapid attenuation of ultraviolet light in water serve as a basis for depth detection when horseshoe crabs ascend to the beaches for breeding. A corollary observation to phototactic turning is an increase in leg movement and closing of the terminal flaps of the fifth walking leg, both primarily on the contralateral side (Corning and Von Burg, 1968; Lahue, 1973). Attempts to exploit these turning tendencies to establish an optomotor response gave results on the border of statistical significance (von Campenhausen, 1967); the principal impression generated is that of unremediable “disobedience” of the animal. A more tractable and reliable unconditioned response is that of a downward tail movement in response to illumination (Wasserman and Patton, 1970; Wasserman, 1973a,b).The discovery of this behavior was the offshoot of unsatisfactory attempts to produce conditioned responses in Limulus (Smith and Baker, 1960; Wasserman and Patton, 1969; Makous, 1969; Wasserman, 1970). The reaction has a latency of between 2.5 and 4 seconds, in keeping with the generally phlegmatic deportment of the animal, but in contrast to a low winter response reaches almost 100% in frequency in a group tested during the spring. Over a total of about 1300 trials, 33% reacted to stimulation of the lateral eyes, 30% to ocellar illumination, and an unexpected 91% to ventral eye stimulation. Although Wasserman (1973a) did not positively exclude the participation of retinula cells at the surface of the brain-the deafferented medulla responds to light

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

343

stimulation (Snodderly, 1971)- this reliable behavioral response invites reconsideration of “the apparently senseless waste of photoreceptor cells strung out along the silent ventral eye nerve where they transduce light with great elegance to absolutely no purpose” (Clark et al., 1969b). Permanently implanted electrodes (Coming et al., 1965; Adolph, 1971) to monitor activity in the optic nerves, heart, and abdominal ganglia have revealed that the heartbeat is considerably influenced by visual input (Coming and Von Burg, 1968; Coming et al., 1971). The principal effect is a brief acceleration of heart rate at the onset of illumination and the converse at “off.” In addition, short- and long-term cardiac periodicities (3 minutes versus 15-20 minutes) are affected by illumination as well as by lesions to the visual area of the central nervous system (Coming and Von Burg, 1970). Even the ectopic compound eyes found by Barlow and Kaplan (1972)had a similar effect on the heart rate, although the relative influence of the other photoreceptors has not been explored. The potential usefulness of such investigations is enhanced by the various studies on the neurophysiology (Tanaka et al., 1966; Von Burg and Coming, 1969; Corning and Von Burg, 1970; Palese et al., 1970; Lang, 1971; Rulon et al., 1971), morphology (Bursey and Pax, 1970; Leyton and Sonnenblick, 1971; Sperelakis, 1971; Lang, 1972), and pharmacology of the Limulus heart (Pax and Sanbom, 1967a,b;Von Burg and Coming, 1971). Various other aspects of the life of Limulus are undoubtedly guided by visual stimuli but have not been explored. The animals have been dredged from depths of 20 meters several miles offshore (Shuster, 1960), yet are said to adjust their breeding time to the phase of the moon (Lockwood, 1870). In Cape Cod Bay, Massachusetts, tagged animals have been observed to travel an average of 1.1 km per day over a 2-week period and to migrate as far as 34 km in 2 months (Shuster, 1950).Given their apparent maximum speed of about 2.5 km per day (extrapolated from Cole, 1923),this means that, despite a complex and changing coastline, some animals maintain a stubborn directional orientation over a period of weeks. ACKNOWLEDGMENTS

I am grateful to Drs. N. J. Alexander, G . F. Gwilliam, J. W. Hawkes, L. H. Kleinholz, and C. J. Russell for constructive criticism of the manuscript and to Mr. J. H. It0 for his artistry in rendering the diagrams. The personal research cited in this article has received painstaking and patient technical attention from Ms. A. J. Gri5n. I am indebted to Ms. M. T. Barss and S. E. Maher for editorial and clerical assistance.

344

WOLF H. FAHRENBACH

This paper constitutes Publication No. 717 of the Oregon Regional Primate Research Center. The research of the author has been supported by Grants RR00163 and EY00392 from the National Institutes of Health and by a Bob Hope Grant in Aid by Fight-for-Sight, Inc., New York City. Figures 9, 12, 15, 18, and 32 are reproduced by permission of Springer-Verlag, Berlin and New York.

ADDENDUM Two recent articles [M. E. Behrens, J . Comp. Physiol. 89, 45 (1974);W.H.Miller and D. F. Cawthon, Znuest. Ophthulmol. 13,401 (1974)l provide new details about the process of light adaptation in the Limulus compound eye. Behrens (1974)illustrates (in material fixed in boiling water) an increase in the depth of the cone cells, length of the eccentric cell dendrite, and radial extent of the rhabdomal rays with light adaptation. Maintenance of adaptation after short-term denervation or partial occlusion of the eye points to direct photomechanical responses of the pertinent cells. An apparent circadian rhythm of morphological changes that persists during 24 hours of constant darkness may be attributable to neurosecretory influences. Miller and Cawthon (1974)confirm the changes in the shape of the rhabdome and dendrite and show that high concentrations of topically injected colchicine, dissolved in dilute dimethyl sulfoxide, cause pigment migration in retinula cells to the light-adapted position. Both studies mention the variable response of ommatidia to light and dark adaptation. REFERENCES Adolph, A. R. (1966).In “The Functional Organization of the Compound Eye” (C. G. Bernhard, ed.), pp, 465-482.Pergamon, Oxford. Adolph, A. R. (1971).Vision Res. 11,979. Adolph, A. R.,and Tuan, F. J. (1972).J . Gen. Physiol. 60,679. AndrC A. (1782). Phil. Trans. Roy. SOC. London 72,440. Bang, F. (1956).Bull.Johns Hopkins Hosp. 98,325. Barlow, R. B., Jr. (1969).J. Gen. Physiol. 54,383. Barlow, R. B., Jr., and Kaplan, E.(1972).Biol. Bull. 143,454. Bass, L., and Moore, W. J. (1970).Biophys. J . 10, 1. Bassot, J.-M., and Martoja, R. (1965).J. Microsc. (Paris) 4,87. Bassot, J.-M., and Martoja, R. (1966).2. Zellforsch. Mikrosk. Anat. 74, 145. Behrens, M. E.,and Wulff, V. J. (1965). J . Gen. Physiol. 48, 1081. Behrens, M. E.,and Wulff, V. J. (1970).Vision Res. 10,679. Benham, W. B. S. (1883).Trans. Linn. SOC. London 2,363. Bennett, M. V. L., Aljure, E.,Nakajima, Y., and Pappas, G.(1963).Science 141,262. Bennett, M. V. L., Nakajima, Y., and Pappas, G. D. (1967).J . Neurophysiol. 30, 161.

T H E VISUAL SYSTEM OF T H E HORSESHOE CRAB

345

Borsellino, A., Fuortes, M. G. F., and Smith, T. G. (1965). Cold Spring Harbor Symp. Quant. Biol. 30,429. Bouligand, Y. (1966). Mem. Mus. Nut. Hist. Natur. (Paris) Ser. A 40, 189. Boulton, P. S. (1969). Z . Zellforsch. Mikrosk. Anat. 101,98. Brandenburger, J. L., and Eakin, R. M. (1970). Vision Res. 10,639. Brown, F. A., Jr., and Cunningham, 0. (1941). Biol. Bull. 81,80. Brown, J. E., and Blinks, J. R. (1972). Biol. Bull. 143,456. Brown, J. E., Murray, J. R., and Smith, T. G. (1967). Science 158,665. Bumel, M., Mahler, H. T., and Moore, W. J. (1970). J . Neurochem. 17, 1493. Bursey, C. R., and Pax, R. A. (1970).J . Morphol. 130,385. Butenandt, A., Biekert, E., and Linzen, B. (1958). Hoppe-Seyler‘s Z . Physiol. Chem. 313, 251.

Carricaburu, P. (1967). C . R. Acad. Sci., Ser. D.264,1476. Camcaburu, P. (1968). C . R. Acad. Sci., Ser. D.267, 1630. Carricaburu, P. (1970). Vision Res. 10,943. Chapman, R. M., and Lall, A. B. (1967). J. Gen. Physiol. 50,2267. Clark, A., Levin, J., and Allen, R. D. (1969a). Biol. Bull. 137,395. Clark, A. W., Millecchia, R., and Mauro, A. (1969b).J . Gen. Physiol. 54,289. Cohen, H. A. (1973). J. Neurocytol. 2,429. Cohen, M. J,, and Jacklet, J. W. (1967). Phil. Trans. Roy. SOC. London, Ser. B . 252,500. Cole, W. H. (1923).J. Gen. Physiol. 5,417. Cole, W. H. (1924).J . Gen. Physiol. 6,295. Coming, W. C., and Von Burg, R. (1968). In “Neurobiology of Invertebrates” (J. Salanki, ed.), pp. 463-477. Plenum, New York. Coming, W. C., and Von Burg, R. (1970). Can. J . Zool. 48, 1450. Coming, W. C., Feinstein, D. A,, and Haight, J. R. (1965). Science 148, 394. Corning, W. C., Lahue, R., and Von Burg, R. (1971). Can. J . Physiol. Phannacol. 49, 387.

Demoll, R. (1914). Zool. Jahrb.,Abt. Anat. Ontog. Tiere 38,443. Dewey, M. M., Levine, R. J. C., and Colflesh, D. E. (1973). J . Cell Biol. 58, 574. Dumont, J. N., and Anderson, E. (1967). J . Microsc. (Paris)6,791. Dumont, J. N., Anderson, E., and Chomyn, E. (1965). J . Ultrastruct. Res. 13, 38. Dumont, J. N., Anderson, E., and Winner, G. (1966).J . Morphol. 119, 181. Eakin, R. M. (1972). In “Handbook of Sensory Physiology” (H. J. A. Dartnall, ed.), Vol. 7, pp. 625-684. Springer-Verlag. Berlin and New York. Eakin, R. M., and Kuda, A. (1972).J . E r p . Zool. 180,267. Eguchi, E. (1962).J. Ultrastruct. Res. 7,328. Eguchi, E., and Waterman, T. H. (1966). In “The Functional Organization of the Compound Eye” (C. G. Bernhard, ed.), pp. 105-124. Pergamon, Oxford. Eguchi, E., and Waterman, T. H. (1967). Z . Zellforsch. Mikrosk. Anat. 79,209. Eguchi, E., Naka, K., and Kuwabara, M. (1962).J . Gen. Physiol. 46, 143. Exner, S. (1891). “Die Physiologie der facettirten Augen von Krebsen und Insecten.” Deuticke, Vienna. Fahrenbach, W. H. (1964). Z . Zellforsch. Mikrosk. Anat. 62, 182. Fahrenbach, W. H. (1967).J . Cell Biol. 35,69. Fahrenbach, W. H. (1968). Z . Zellforsch. Mikrosk. Anat. 87,278. Fahrenbach, W. H. (1969). Z . Zellforsch. Mikrosk. Anat. 93,451. Fahrenbach, W. H. (1970a). Z. Zellforsch. Mikrosk. Anat. 105,303. Fahrenbach, W. H. (1970b).J. Cell Biol. 44,445. Fahrenbach, W. H. (1970~). J . Cell Biol. 47,59a.

346

WOLF H. FAHRENBACH

Fahrenbach, W. H. (1971). Z . Zellforsch. Mikrosk. Anat. 114, 532. Fahrenbach, W. H. (1973a).J . Morphol. 140,31. Fahrenbach, W. H. (1973b).Anat. Rec. 175,316. Fahrenbach, W. H. (1973~).Z. Zellforsch. Mikrosk. Anat. 144, 153. Femhdez-MorBn, H. (1962). Circulation 26, 1039. Femhdez-Morin, H., van Bruggen, E. F. J., and Ohtsuki, M. (1966).J. Mol. B i d . 16, 191.

Femindez-MorBn, H. Marchalonis, J. J,, and Edelman, G. M. (1968).J. Mol. B i d . 32, 467.

Fingerman, M., Bartell, C. K., and Krasnow, R. A. (1971). B i d . Bull. 140,376. Fuge, H. (1967). Z . Zellforsch. Mikrosk. Anat. 83,468. Fuortes, M. G. F. (1959).J. Physiol. (London) 148, 14. Gemperlein, R. (1969). Z. Vergl. Physiol. 65,428. Graham, C. H., and Hartline, H. K. (1935).J. Gen. Physiol. 18,917. Grenacher, H. (1879). “Untersuchungen iiber das Sehorgan der Arthropoden, insbesondere der Spinnen, Insecten und Crustaceen.” Vandenhoeck & Ruprecht, Giittingen. Gur, M., Purple, R. L., and Whitehead, R. (1972).J. Gen. Physiol. 59,285. Hanstrom, B. (1926a). Lunds. Unio. Arsskr., Aod. 2 22, 1. Hanstrom, B. (1926b). Z . Mikrosk.-Anat. Forsch. 7, 139. Hanstrom, B. (1926~). Zool. Anz. 66,213. Harper, E., Seifter, S., and Scharrer, B. (1967).J . Cell B i d . 33,385. Hartline, H. K. (1949). Fed. Proc., Fed. Amer. SOC. Erp. B i d . 8,69. Hartline, H. K. (1969). Science 164,270. Hartline, H. K., and Ratliff, F. (1972). In “Physiology of Photoreceptor Organs” (M. G. F. Fuortes, ed.), pp. 3 8 1 4 7 . Springer-Verlag, Berlin and New York. Hartline, H. K., Ratliff, F., and Miller, W. M. (1961). In “Nervous Inhibition” (E. Florey, ed.), pp. 241-284. Pergamon, Oxford. Herman, W. S. (1970).Amer. Zool. 10,497. Herman, W. S. (1972). Gen. Comp. Endocrinol. 18,301. Herman, W. S., and Preus, D. M. (1972). Z . Zellforsch. Mikrosk. Anat. 134,255. Herman, W. S., and Preus, D. M. (1973).J . Morphol. 140, 53. Hesse, R. (1901). Z . Wiss. Zool. 70,347. Hock, C . W. (1940). B i d . Bull. 79, 199. Hoglund, G., and Struwe, G. (1971). Z . Vergl. Physiol. 74,336. Holme, R., and Solum, N. 0. (1973). J. Ultrastruct. Res. 44,329. Horridge, G . A. (1969). Proc. Roy. Soc., Ser. B . 171,445. Homdge, G. A,, and Barnard, P. B. T. (1965). Quart. J. Microsc. Sci. 106, 135. Hubbard, R., and Wald, G. (1960). Nature (London)186,212. Hudspeth, A. J., and Yee, A. G. (1973). Inoest. Ophthalrnol. 12,366. Johansson, G. (1933). Acta Zool. (Stockholm) 14, 1. Johnson, R. G., Herman, W. S., and Preus, D. M. (1973). J. Ultrastruct. Res. 43, 298. Jones, C., Nolte, J., and Brown, J. E. (1971). Z. Zellforsch. Mikrosk. Anat. 118, 297. Kaplan, E., Bayer, D. S., and Barlow, R. B., Jr. (1973). B i d . Bull. 145,442. Kleinholz, L. H. (1959). B i d . Bull. 116, 125. Kleinholz, L. H. (1966). Amer. Zool. 6, 161. Knight, B. W., Toyoda, J.-I., and Dodge, F. A., Jr. (1970). J . Gen. Physiol. 56, 421. Lahue, R. (1973). In “Invertebrate Learning,” Vol. 2. (W. C. Coming, J. A. Dyal, and A. 0. D. Willows, eds.), pp. 1-48. Plenum, New York. Lall, A. B. (1970). Vision Res. 10,905.

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

347

Lall, A. B., and Chapman, R. M. (1973).J . Exp. B i d . 58,213. Lang, F. (1971).B i d . Bull. 141,269. Lang, F. (1972).2. Zellforsch. Mikrosk. Anat. 130,481. Lankester, E. R., and Bourne, A. G. (1883).Quart. J . Microsc. Sci. 23, 177. Lasansky, A. (1967).J. Cell B i d . 33,365. Lasansky, A. (1970).Neurosci. Res. Program, Bull. 8,467. Levi, H.W.(1968).Turtox News 46, 183. Leyton, R. A., and Sonnenblick, E. H. (1971). J . Cell Biol. 48, 101. Linzen, B. (1966).Z. Naturforsch. B . 21, 1038. Lipetz, L. E. (1960).Adoan. Biol. Med. Phys. 7 , 131. Lisman, J. E., and Brown, J. E. (1972).J . Gen. Physiol. 59,701. Locke, M. (1969). J . Morphol. 127,7. Lockwood, S . (1870).Amer. Natur. 4,257. Loeb, J. (1893).Arch. Gesamte Physiol. Menschen Tiere 53,81. Loeb, L. (1902).J . Med. Res. 7 , 145. Loeb, L. (1928).Protoplasma 4,596. Loewenstein, W.R., and Kanno, Y. (1967).J. Cell Biol. 33,225. MacNichol, E. F.,Jr., and Benolken, R. (1956).Science 124,681. Makous, W.I. (1969).Psychon. Sci. 14,4. Millecchia, R., and Mauro, A. (1969a).J . Gen. Physiol. 54,310. Millecchia, R., and Mauro, A. (1969b).J. Gen. Physiol. 54,331. Millecchia, R.,Bradbury, J., and Mauro, A. (1966).Science 154,1199. Miller, W.H.(1957).J . Biophys. Biochem. Cytol. 3,421. Miller, W.H.(1958).Ann. N . Y. Acad. Sci. 74,204. Miller, W.H.(1965).In “The Structure of the Eye” (J. W. Rohen, ed.), pp. 159-169. Schattauer, Stuttgart. Munson, J. (1898).J.Morphol. 15, 111. Murray, G.C.(1966).Science 154,1182. Noirot-Timothbe, C., and Noirot, C. (1966).J . Microsc. (Paris)5,715. Nolte, J., and Brown, J. E. (1969).J . Gen. Physiol. 54,636. Nolte, J . , and Brown, J. E. (1970).J. Gen. Physiol. 55,787. Nolte, J., Brown, J. E., and Smith, T. G., Jr. (1968).Science 162,677. Northrop, J. H.,and Loeb, J. (1923).J . Gen. Physiol. 5,581. Nunnemacher, R. F., and Davis, P. P. (1968).J . Morphol. 125,61. Packard, A. S. (1893).Mem. Nut. Acad. Sci. U.S . 6,289. Palese, V. J., Jr., Becker, J. L., and Pax, R. A. (1970).J. Exp. Biol. 53,411. Pannesi, P. A. F. (1964).M. A. Thesis, Clark Univ., Worcester, Massachusetts. Pappas, G. D., Asada, Y.,and Bennett, M. V. L. (1971).J . Cell Biol. 49, 173. Patten, W.(1893).Quart. J . Microsc. Sci. 35, 1. Patten, W. (1912).“The Evolution of the Vertebrates and Their Kin.” Blakiston, Philadelphia, Pennsylvania. Morphol. 16,91. Patten, W., and Redenbaugh, W. A. (19OO).J. Pax, R. A., and Sanbom, R. C. (1967a).Biol. Bull. 132,381. Pax, R. A., and Sanbom, R. C. (1967b).Biol. Bull. 132,392. Perrelet, A.,and Baumann, F. (1969).J. Cell Biol. 40,825. Perrelet, A., Orci, L., and Baumann, F. (1971).J. Cell Biol. 48,684. Purple, R. L., and Dodge, F. A. (1965).Cold Spring Harbor Syrnp. Quant. Biol. 30, 529. Purple, R. L., and Dodge, F. A. (1966).In “The Functional Organization of the Compound Eye” (C. G. Bemhard, ed.), pp. 451-464.Pergamon, Oxford.

348

WOLF H. FAHRENBACH

Ratliff, F. (1965). “Mach Bands: Quantitative Studies on Neural Networks in the Retina.” Holden-Day, San Francisco, California. Ratliff, F. (1966). In “The Functional Organization of the Compound Eye” (C. G. Bemhard, ed.), pp. 187-191. Pergamon, Oxford. Reichardt, W. (1961).Kybernetik 1,57. Richter, 1.-E. (1969).Z . Motphol. Tiere 64,85. Rulon, R., Hermsmeyer, K., and Sperelakis, N. (1971). Comp. Biochem. Physiol. 39, 333. Scharrer, B. (1968).Z . Zellforsch. Mikrosk. Anat. 89, 1. Scharrer, B. (1972). Progr. Brain Res. 38,7. Schlottke, E. (1934). Z . Mikrosk.-Anat. Forsch. 35,57. Schlottke, E. (1935).Z. Vergl. Physiol. 22,359. Schwartz, E. A. (1971).J . Neurobiol. 2, 129. Seitz, G. (1969).Z. Vergl. Physiol. 62,61. Shaw, S . R. (1967a). Z. Vergl. Physiol. 55, 183. Shaw, S. R. (1967b).J . Gen. Physiol. 50,2480. Shaw, S. R. (1969).Science 165,88. Shoup, J. R. (1966).J . Cell Biol. 29,223. Shuster, C. N., Jr. (1950). “Investigations of Methods of Improving the Shellfish Resources of Massachusetts,” pp. 18-23. Woods Hole Oceanogr. Inst., Woods Hole, Massachusetts. Shuster, C. N., Jr. (1955).Ph.D. Thesis, New York Univ., New York. Shuster, C. N., Jr. (1960). Estuarine Bull. 5,3. Smith, J. C., and Baker, H. (1960).J . Comp. Physiol. Psychol. 53,279. Smith, T. G., and Baumann, F. (1969). Progr. Brain Res. 31,313. Smith, T. G., Baumann, F., and Fuortes, M. G . F. (1965). Science 147, 1446. Snodderly, D. M., Jr. (1971).J . Neurophysiol. 34,588. Snodderly, D. M., Jr., and Barlow, R. B., Jr. (1970). Nature (London) 227,284. Snyder, A. W., Menzel, R., and Laughlin, S. B. (1973). J . Comp. Physiol. 87, 99. Sperelakis, N. (1971). Z. Zellforsch. Mikrosk. Anat. 116,443. St@rmer,L. (1952).J . Palaeontol. 26,630. Stunkard, H. W. (1951).Biol. Bull. 101,397. Stunkard, H. W. (1953).J . Parasitol. 39,225. Stunkard, H. W. (1968).B i d . Bull. 134,332. Tanaka, I., Sasaki, Y., and Shin-mura, H. (1966).Jap.J . Physiol. 16, 142. Tomita, T., Kikuchi, R., and Tanaka, I. (1960). In “Electrical Activity of Single Cells” (Y.Katsuki, ed.), pp. 11-23. Igaku Shoin, Tokyo. Trujillo-Cenbz, 0. (1965). Cold Spring Harbor Symp. Quant. Biol. 30,371. Trujillo-Cenbz, O., and Melamed, J. (1967). Z. Zellforsch. Mikrosk. Anat. 76, 377. Unestam, T. (1965).Physiol. Plant. 18,483. Unestam, T. (1968).Physiol. Plant. 21, 137. Viallanes, A. (1893).Ann. Sci. Nut., Zool. 7,405. Von Burg, R., and Coming, W. C. (1969).Can. J . Zool. 47, 735. Von Burg, R., and Coming, W. C. (1971). Can. J . 2001.49, 1044. von Campenhausen, C. (1967).J. Exp. Biol. 46,557. Wachmann, E. (1969).Z. Zellforsch. Mikrosk. Anat. 99,263. Wald, G., and Krainin, J. M. (1963).Proc. Not. Acad. Sci. U.S . 50, 1011. Wasserman, G. S. (1967).J . Gen. Physiol. 50, 1075. Wasserman, G. S. (1969).Vision Res. 9,611. Wasserman, G. S. (1970). Psychon. Sci. 19, 184.

THE VISUAL SYSTEM OF THE HORSESHOE CRAB

349

Wasserman, G . S. (1973a).Vision Res. 13,95. Wasserman, G . S. (1973b).Vision Res. 13,1203. Wasserman, G . S.,and Patton, D. G . (1969).Psychon. Sci. 15, 143. Wasserman, G . S.,and Patton, D. G . (1970). J. Comp. Physiol. Psychol. 73, 11. WatasB, S. (1890).Stud. Biol. Lab., Johns Hopkins Unio. 4,287. Waterman, T.H. (1953).Proc. Nut. Acad. Sci. U . S . 39,687. Waterman, T.H. (1954a).J. Morphol. 95,125. Waterman, T.H. (1954b).Proc. Nut. Acad. Sci. U.S. 40,252. Waterman, T. H. (1954~). Proc. Nut. Acad. Sci. U.S . 40,258. Waterman, T. H., and Enami, M. (1954).Pubbl. Sta. Zool. Napoli 24, Suppl., 81. Waterman, T. H., and Wiersma, C. A. G . (1954).J . Exp. Zool. 126,59. White, R. H. (1967). J . E x p . 2001. 166,405. Whitehead, R. A.,and Purple, R. L. (1970).Vision Res. 10, 129. Whitehead, R. A.,Purple, R. L., and Hopper, K. (1969).J. Morphol. 129,31. Wilska, A.,and Hartline, H. K. (1941).Amer. J . Physiol. 133,491. Wolbarsht, M.L.,and Yeandle, S. S. (1967).Annu. Reo. Physiol. 29,513. Wolf, E., and Zerrahn-Wolf, G. (1937).J . Cen. Physiol. 20,767. Wolken, J. J., and Florida, R. G. (1969). J. Cell Biol. 40,279. Yeandle, S., and Spiegler, J. B. (1973).J . Gen. Physiol. 61,552. Young, D.,Ashhurst, D. E., and Cohen, M.J. (1970).Tissue Cell 2,387.

This Page Intentionally Left Blank

Subject Index A Androgen, receptors in tissues, 101-109 Anterior pituitary cells, characteristics of cells in oitro, hormonal secretion, 177-184 morphology, 184-205 specific regulatory agents in oitro, continuous cell lines, 218-234 primary cultures, 205-218 Antibodies, neuroph ysin, cross-species, 250-254 general considerations, 255-256 species-specific, 254-255

B Bacterial chromosome, attachment to membrane, 19 morphological considerations, association between nucleoid and membrane, 5-13 membrane, 4-5 nucleoid, 3-4 nature of binding to membrane, 23-24 polarization and segregation, 15-17 sites of attachment, 22-23 Blood, steroid-binding proteins in, 90-92 Brain, neurophysin in, 266-269 steroid receptors in, 123-125

C Cancer, steroid dependency of, 125-126 Cell(s), fractionation, methodological considerations, 17-19 Cell cycle, 83-84 regulatory patterns, cycle-independent, 66-68 nucleus, 65-66 oral apparatus, 63-65 somatic ciliature, 60-63 Cell membrane, bacterial chromosome and, association between nucleoid and membrane. 5-13

membrane, 4-5 nucleoid, 3-4 DNA replication on, 20-22 mode of segregation, 13-15 morphology, 4-5 origin of DNA replication and, 20 role in compactness of nucleoid, 24-25 special region, chromosome attachment and, 19 Cyclic adenosine monophosphate, action in oitro, pgalactosidase synthesis, 46-47 initiation of lac on mRNA synthesis, 51-54 mediation of, 47-49 properties of receptor protein, 49-51 catabolite repression of pgalactosidase, 36-37 control of, 42-43 cellular concentrations, 37-39 in oiuo evidence of site of action, lac promoter mutants and, 45-46 banscriptional or translational control, 43-45 metabolism, enzymes involved, 39-42 role in lactose operon, 30-31 steroid hormones and, 155-157

D Deoxyribonucleic acid, origin of replication, cell membrane and, 20 replication, cell membrane and, 20-22

E Estrogen, receptors in tissues, 92-101

G pGalactosidase, repression, CAMP and, 36-37 control of, 42-43 synthesis in oitro, 46-47 Glucocorticoids, receptors in tissues, 113-120 351

352

SUBJECT INDEX

H

Hormones, insect, vitamin D3and, 154-155 Horseshoe crab visual system, abnormalities, 338-340 axons, afferent, 321-323 efferent, 324-326 basal lamina and hemocoel, 318-321 dioptric structures, cone cells, 293-296 cuticular cones and lenses, 287-292 lens epidermis, 292-293 neuroglial cells, 304-306 optic centers, 335-338 optic nerves, 333-335 pathology, 340-341 pigment cells, distal, 298-301 guanophores, 296-298 ommatidial, 301-304 plexus, 327-333 receptor cells, arhabdomeric, 317-318 eccentric, 314-317 retinula, 306-314 vision and behavior, 341-343 Hypothalamus, neurophysin in, 266-269

1

Lactose operon, inducer, nature of, 34-35 organization of, 31-34 regulation, additional aspects, 54-55 role of CAMP,30-31

M

Magnocellular nuclei, neurophysin in, 260-266

Messenger ribonucleic acid, lactose operon, synthesis in oitro, 51-54

M icrotu bules , control of formation, synthesis-dependent assembly, 79-83

tubulin synthesis and, 77-79 dynamic nature, evidence in Tetrahymena,, 74-75 turnover and subunit exchange, 75-77

regulatory patterns and cell cycle, cycle-independent, 66-68 nuclear, 65-66 oral apparatus, 63-65 somatic ciliature. 60-63 stability and regression, 68-69 conclusion, 73-74 nuclear, 72-73 somatic and oral, 69-72 Mineralocorticoids. receptors in tissues, 120-122

N Neurohypophysis, proximal, neurophysin in, 269-275 Neurophysin, antibody preparation, cross-species, 250-254 general considerations, 255-256 species-specific, 254-255 antigen purification, extraction from gels, 250 isoelectric focusing, 249-250 molecular seive and ion-exchange chromatography, 246-249 polyacrylamide-gel electrophoresis, 249

demonstration using antineurophysin, magnocellular nuclei, 260-266 other areas of hypothalamus and brain, 266-269 pituitary stalk, 275,276 posterior pituitary gland, 277, 278 proximal neurohypophysis, 269-275 demonstration with antisera, ox, 279,280 pig, 278,279 immunohistochemical techniques, electron microscopy, 257 light microscopy, 256-257 photography, 257 preparation of tissues, 257-259 methods of extraction,

353

SUBJECT INDEX

posterior pituitary gland, 244-246 whole glands, 246 relationship to neurosecretory material, 243-244

Nucleoid, compactness, membrane and, 24-25 morphology, 3-4 Nucleus, microtubules, 65-66 stability and regression, 72-73 P

Pituitary stalk, neurophysin in, 275,276 Posterior pituitary gland, neurophysin in, 277,278

Progestin, receptors in tissues, 109-113

R Ribonucleic acid, synthesis, steroid receptors and, 139-143

S

Steroid(s), binding proteins in blood, 90-92

Steroid hormones, CAMP and, 155-157 Steroid-receptors, brain and, 123-125 cancer and, 125-126 cellular steroid uptake and, 151 cytoplasmic-nuclear interaction, chromatin acceptor sites, 132-137 cytoplasm-independent nuclear receptors, 131-132 intracellular recycling, 139 ribonucleoprotein binding, 137-139 transformation of cytoplasmic receptors, 127-131 gene expression and, hypothetical models, 144-147 in oitro approaches, 147-150 RNA synthesis and protein induction, 139-143 natural forms, 151-152 nature of interaction, 152-154

Contents of Previous Volumes Ascorbic Acid and Its Intracellular Localization, with Special Reference Some Historical Features in Cell Biology to Plants-J. CHAYEN -ARTHUR HUGHES Aspects of Bacteria as Cells and as OrNuclear Reproduction4. LEONARD ganisms-SmmT MUDD AND EDWARD HUSKINS D. DELAMATER Enzymic Capacities and Their Relation Ion Secretion in Plants-J. F. SUTCLIFFE to Cell Nutrition in AnimalsC'EORGE Multienzyme Sequences in Soluble W. KIDDER Extracts-HENRY R. MAHLER The Application of Freezing and Drying The Nature and Specificity of the FeulTechniques in Cytology-L. G. E. gen Nucleal Reaction-M. A. LESSLER BELL Quantitative Histochemistry of PhosphaEnzymatic Processes in Cell Membrane taSeS-wILLXAM L. DOYLE Penetration-TH. ROSENBERGAND w. Alkaline Phosphatase of the NucleusWILBRANDT M. C ~ V R E M O N AND T H. FIRKET Bacterial Cytology-K. A. BISSET Gustatory and Olfactory Epithelia-A. F. Protoplast Surface Enzymes and AbsorpBARADIAND G. H. BOURNE tion of Sugar-R. BROWN Growth and Differentiation of Explanted Reproduction of Bacteriophage-A. D. Tissues-P. J. GAIL,LARD HERSHEY Electron Microscopy of Tissue SectionsThe Folding and Unfolding of Protein A. J. DALTON Molecules as a Basis of Osmotic Work A Redox Pump for the Biological Per-R. J. GOLDACFLE formance of Osmotic Work, and Its Nucleo-Cytoplasmic Relations in AmphibRelation to the Kinetics of Free Ion ian Development-G. FRANK-HAUSER Diffusion across Membranes-E. J. Structural Agents in Mitosis-M. M. CONWAY SWA" A Critical Survey of Current Approaches Factors Which Control the Staining of in Quantitative Histo- and CytochemTissue Sections with Acid and Basic istry-DAvm GLICK Dyes-Mucus SINGER Nucleo-cytoplasmic Relationships in the The Behavior of Spermatozoa in the Development of Acetabularia-J. HAMNeighborhood of Eggs-hm ROTHSMERLINC

Volume 1

CHILD Report of Conference of Tissue Culture The Cytology of Mammalian Epidermis Workers Held at Cooperstown, New and Sebaceous Glands-wzLIAM York-D. J. HETHERINGTON MONTAGN A AUTHOR INDEX-SUBJECT INDEX The Electron-Microscopic Investigation H. BRETSCHof Tissue Sections-L. NEIDER Volume 3 The Histochemistry of Esterases-G. The Nutrition of Animal Cells-CwmY ChMORI

AUTHOR INDEX-SUB

WAYMOUTH

JECT INDEX

Volume 2 Quantitative Aspects of Nuclear Nucleoproteins--HEwsoN SWIFT

Caryometric Studies of Tissue CulturesOTTO BUCHER The Properties of Urethan Considered in Relation to Its Action on MitosisIVOR CORNMAN

354

co"Ts OF PREVIOUS VOLUMES

355

Composition and Structure of Giant Evidence for a Redox Pump in the Active Transpoxt of Cations-E. J. CONWAY ALFERT Chromosomes-Ma How Many Chromosomes in Mammalian AUTHOR INDEX-SUB JECT INDEX Somatic Cells?-R. A. BEATTY The Significance of Enzyme Studies on Volume 5 Isolated Cell Nuc1ei-ALEXANDER L. Histochemistry with Labeled Antibody DOUNCE -ALBERT H. COONS The Use of Differential Centrifugation The Chemical Composition of the Bacin the Study of Tissue Enzymesterial Cell Wall-C. S. CUMMINS CHR.DE DUVEAND J. BERTHET Theories of Enzyme Adaptation in MicroEnzymatic Aspects of Embryonic Differorganisms-J. MANDELSTAM entiation-TRYccvE GUSTAFSON The Cytochondria of Cardiac and Azo Dye Methods in Enzyme HistochemSkeletal M U S C ~ ~ J OW. H NHARMON istry-A. G. EVERSONPEARSE The Mitochondria of the NeuronMicroscopic Studies in Living MamWARREN ANDREW mals with Transparent Chamber The Results of Cytophotometry in the Methods-Roy G. WILLIAMS Study of the Deoxyribonucleic Acid The Mast Cell-G. ASBOE-HANSEN (DNA) Content of the NucleusElastic Tissue-EDwmm W. DEMPSEY R. VENDRELY AND C. VENDRELY AND ALBERT I. LANSING Protoplasmic Contractility in Relation to The Composition of the Nerve Cell Gel Structure: Temperature-Pressure Studied with New Methods-SvENExperiments on Cytokinesis and OLOEB R A T T G ~AND D HOLCERHYDEN Amoeboid Movement - DOUGLAS MARSLAND AUTHOR INDEX-SUBJECT INDEX Intracellular pH-PETER C. CALDWEU The Activity of Enzymes in Metabolism Volume 4 and Transport in the Red Cell-T. A. J. PRANKERD Cytochemical Micrurgy-M. J. KOPAC Uptake and Transfer of Macromolecules Amoebocytes-L. E . WACGE by Cells with Special Reference to Problems of Fixation in Cytology, HisGrowth and Development-A. M. tology, and Histochemistry-M. WOLMAN

~CHECHTMAN

Bacterial Cytology-ALFRED MARSHAK Cell Secretion: A Study of Pancreas and C. J. J U N Q U E ~ Histochemistry of Bacteria-R. VENDRELY Salivary Glands-L. AND G. C. HIRSCH Recent Studies on Plant MitochondriaThe Acrosome Reaction-JEAN c. DAN DAVIDP. HACKETT cytology of spermatogenesis-VISHWA The Structure of Chloroplasts-K. NATH M~~HLETHALER The Ultrastructure of Cells, as Revealed Histochemistry of Nucleic Acids-N. B. by the Electron Microscope--OF KURNICK s. SJOSTRAND Structure and Chemistry of NucleoliAUTHOR INDEX-SUB JECT INDEX W. S. VINCENT On Goblet Cells, Especially of the InVolume 6 testine of Some Mammalian SpeciesHAFIALD MOE The Antigen System of Paramecium Localization of Cholinesterases at aurelio-G. H. BWE COU- The Chromosome Cytology of the Ascites Neuromuscular Junctions-R. TEAUX Tumors of Rats, with Special Ref-

356

CoNTENTS OF PREVIOUS VOLVMES

erence to the Concept of the Stemline Cell-Snjmo MAIUN~ The Structure of the Gold ApparatusARTHUIt w. POLLISTER AND PRISCHIA F. P O L ~ T E R An Analysis of the Process of Fertilization and Activation of the EggA. MONROY The Role of the Electron Microscope in Virus Research-ROBLEY c. WILLIAMS The Histochemistry of PolysaccharidesARTHURJ. HALE The Dynamic Cytology of the Thyroid Gland-J. GROSS Recent Histochemical Results of Studies on Embryos of Some Birds and Mammals-ELIo BORGHESE Carbohydrate Metabolism and Embryonic Determination-R. J. O'CONNOR Enzymatic and Metabolic Studies on Isolated Nuc1ei-G. SIEBERTAND R. M. S.

The Structure and Innervation of Lamellibranch Muscle-J. BOWDEN Hypothalamo-neurohypophysial Neurosecretion-J. C. SLOPER Cell Contact-PmL WEISS The Ergastoplasm: Its History, Ultrastructure, and Biochemistry--COISE HAGUENAU Anatomy of Kidney Tubules-Jow"Es &IODIN

Structure and Innervation of the Inner Ear Sensory Epithelia-Hme ENGSTROM AND JANWERSKLC The Isolation of Living Cells from Animal Tissues-L. M. RINALDINI AUTHOR INDEX-SUBJECT

INDEX

Volume 8

The Structure of C y t o p l a s d u x u s s OBERLING Wall Organization in Plant Cells-R. D. Recent Approaches of the Cytochemical PRESTON Study of Mammalian TissuedEoRGE H. HOGEBOOM, EDWARD L. KUFF, AND Submicroscopic Morphology of the Synapse-EDuARDo DE ROBERTIS WALTER c. SCHNEDER The Cell Surface of Pusumecium--C. F. The Kinetics of the Penetration of NonE m AND E. L. POelectrolytes into the Mammalian ErythThe Mammalian Reticulocyte-LEAH rocyte-FREDA BOwyeR MIRIAM LOWENSTEIN AUTHOR INDEX-SUB JECT INDEX The Physiology of ChromatophoresSMELLIE

CUMULATIVE SUBJECT INDEX (VOLUMES 1-5)

MILTON

FINGERMAN

The Fibrous Components of Connective Tissue with Special Reference to the Elastic Fiber-DAvm A. HALL Volume 7 Experimental Heterotopic OssificationJ. B. BRIDGES Some Biological Aspects of Experimental A Survey of Metabolic Studies on IsoRadiology: A Historical Review-F. G. lated Mammalian Nuclei-D. B. SPEAR ROODYN The Effect of Carcinogens, Hormones, Trace Elements in Cellular Functionand Vitamins on Organ Cultures-ILsE BERT L. VALLEE A N D FREDERIC L. hSNXTZE3

Recent Advances in the Study of the Kinetochore-A. LIMA-DE-FARM Autoradiographic Studies with S"-Sulfate -D. D. D~IEWIATKOWSKI The Structure of the Mammalian Spermatozoon-DoN W.FAWCEIT The Lymphocyte-4. A. TROWELL

HOCK3

Osmotic Properties of Living CellsD. A. T. DICK Sodium and Potassium Movements in M. Nerve, Muscle, and Red Cells-I. GLYNN Pinocytosis-H. HOLT~R AUTHOR INDEX--

JECT INDEX

357

CONTENTS OF PREVIOUS VOLUMES

Volume 11

Volume Q

The Influence of Cultural Conditions on Electron Microscopic Analysis of the F. WILKINSON Secretion Mechanism-K. KUROSUMI Bacterial Cytology-J. The Fine Structure of Insect Sense AND J. P. DUGWID Organs-ELEmOR H. SLIFER Organizational Patterns within ChromoP. KAUFMANN, Cytology of the Developing Eyesomes-BmwmD ALFRED J. COULOMBRE HELENGAY, AND MARGARETR. The Photoreceptor Structures-J. J. MCDONALD Enzymic Processes in Celk-JAY BOYD WOLKEN Use of Inhibiting Agents in Studies on BEST Fertilization Mechanisms-C-s B. The Adhesion of Cells-LEoNARD WEBS METZ Physiological and Pathological Changes The Growth-Duplication Cycle of the in Mitochondria1 Morphology-&. Cell-D. M. F'RESCOTT ROUILLER Histochemistry of Ossification-Row0 The Study of Drug Effects at the CyL. C A B B. WILSON tological Level-G. Cinematography, Indispensable Tool for Histochemistry of Lipids in OogenesisCytology-C. M. POMERAT VISHWANATH AUTHOR INDEX-SUBJECT INDEX Cyto-Embryology of Echinoderms and DAN Amphibia-Kmsum The Cytochemistry of Nonenzyme Pro- Volume 12 teins-RONALD R. COWDEN Sex Chromatin and Human ChromoAUTHOR INDEX-SUB JECT INDEX somes-Jom L. HAMERTON Chromosomal Evolution in Cell Populations-T. C. Hsu Volume 10 Chromosome Structure with Special Reference to the Role of Metal IonsThe Chemistry of Shiff's ReagentDALE M. STEFFENSEN FREDERICK H. KASTEN Electron Microscopy of Human White Spontaneous and Chemically Induced Blood Cells and Their Stem CellsKUMAR Chromosome Breaks-hm MARCELBESSXSAND JEAN-PAULT ~ E R Y SHARMAAND ARCHANAS u m In Vivo Implantation as a Technique in The Ultrastructure of the Nucleus and Skeletal Biology-WmmM J. L. Nucleocytoplasmic RelationsPAuL FELTS WISCHNITZER The Nature and Stability of Nerve The Mechanics and Mechanism of CleavMyelin-J. B. FINEAN age-LEwrs WOLPERT Fertilization of Mammalian Eggs in The Growth of the Liver with Special Vitro-C. R. AUSTIN Reference to Mammals-F. DOLJANSKI Physiology of Fertilization in Fish Eggs Cytology Studies on the Affinity of the -TOKI-O YAMAMOTO Carcinogenic Azo Dyes for Cyto- AUTHOR INDEX-SUB JECT INDEX plasmic Components-YosHP*u NAGATAN1

Epidermal Cells in Culture-A. GEDEON Volume 13 MATOLTSY The Coding AUTHOR INDEX-SUB

JECr INDEX

CUMULATIVE SUBJECT (VOLUMES

1-9)

INDEX

Hypothesis-MARTYNAS

YEAS

Chromosome Reproduction-J. TAYLOR

HERBERT

358

CONTENTS OF PREVIOUS VOLUMES

Sequential Gene Action, Protein Synthesis, and Cellular DifferentiationREED A. FLICKINGER The Composition of the Mitochondria1 Membrane in Relation to Its Structure and Function-Ewc G. BALL AND CLIFFE D. JOEL Pathways of Metabolism in Nucleate and Anucleate Erythrocytes-H. A. SCHWEIGER

Some Recent Developments in the Field of Alkali Cation Transport-W. WILBRANDT

Chromosome Aberrations Induced by Ionizing Radiations-H. J. EVANS Cytochemistry of Protozoa, with Particular Reference to the Golgi Apparatus and the MitwhondriaVISHWANATR AND G. P. DUTTA cell Renewal-FEux BERTALANFFY AND CHOSENLAU AUTHOR INDEX-SUBJECT

Volume 14

INDEX

The Tissue Mast Wall-Doucus SMITH AUTHOR INDEX-SUB

E.

JECT INDEX

Volume 15 The Nature of Lampbrush Chromosomes -H. G. CALIAN The Intracellular Transfer of Genetic Information-J. L. SIRLIN Mechanisms of Gametic Approach in Plants-LEONMACHLU~ AND E R I ~ A RAWITSCHER-KUNICEL The Cellular Basis of Morphogenesis and Sea Urchin Development-T. CUSTAFSON AND L. WOLPERT Plant Tissue Culture in Relation to Development C y t 0 1 o g y - C ~ R. ~ ~ PARTANEN

Regeneration of Mammalian LiverNANCYL. R. BUCHER Collagen Formation and Fibrogenesis with Special Reference to the Role of Ascorbic Acid-BEwS. COULD The Behavior of Mast Cells in Anaphyl a x i s - 1 ~ ~MOTA ~ Lipid Absorption-ROBERT M. WOT~ON

Inhibition of Cell Division: A Critical and Experimental Analysis-SEYMOUR AUTHOR INDEX-SUB JECT INDEX GELFANT Electron Microscopy of Plant Protoplasm Volume 16 -R. BUVAT Cytophysiology and Cytochemistry of the Ribosomal Functions Related to Protein Synthesis-Tom HULTIN Organ of Corti: A Cytochemical Theory of Hearing-J. A. VINNJKOV Physiology and Cytology of Chloroplast Formation and “Loss” in EuglenaAND L. K. TITOVA M. GRENSON Connective Tissue and Serum ProteinsCell Structures and Their Significance R. E. MANCINI for Ameboid Movement-K. E. WOHLThe Biology and Chemistry of the Cell FARTH-BOTFERMAN Walls of Higher Plants, Algae, and Microbeam and Partial Cell Irradiation Fungi-D. H. NORTHCOTE 4.L. ShllTIi Development of Drug Resistance b y Nuclear-Cytoplasmic Interaction with Staphylococci in Vitro and in Viu+ MARYBARBER Ionizing Radiation-M. A. LESSLER Cytological and Cytochemical Effects of In Vivo Studies of Myelinated Nerve Fibers--cuU. CASKEYSPEIDEL Agents Implicated in Various Pathological Conditions: The Effect of Respiratory Tissue: Structure, Histophysiology, Cytodynamics. Part I: Viruses and of Cigarette Smoke on the Cell and Its Nucleic Acid4cxL.m Review and Basic CytomorphologyAND RUDOLF LEUCHLEUCATENBERGER FELIX D. BERTALANFFY TENBERCER

AUTHOR INDEX-SUBJECT

INDEX

359

CONTENTS OF PREVIOUS VOLUMES

Volume 17

Volume 19

The Growth of Plant Cell Walls-K. WILSON Reproduction and Heredity in Trypanosomes: A Critical Review Dealing Mainly with the African Species in the Mammalian Host-P. J. WALKER The Blood Platelet: Electron Microscopic Studies-J. F. DAVID-FERREIRA The Histochemistry of Mucopolysaccharides-RoeERT c. CURRAN Respiratory Tissue Structure, Histophysiology, Cytodynamics. Part 11. New Approaches and Interpretations -FEW D. BERTALANFFY The Cells of the Adenohypophysis and Their Functional Significance-Mac HERLANT

“Metabolic” DNA: A Cytochemical Study-H. ROELS The Significance of the Sex ChromatinMURRAY L. BARR Some Functions of the Nucleus-J. M. MITCHISON Synaptic Morphology on the Normal and Degenerating Nervous System-E. G. GRAYAND R. W. GUILLERY Neurosecretion-W. BARGMANN Some Aspects of Muscle RegenerationE. H. BETZ, H. FIRKET,AND M.

AUTHOR INDEX-SUB

JECT INDEX

kZNIK

W. The Gibberellins as Hormones-P. BRIAN Phototaxis in Plants-WOLFGANG HAUPT Phosphorus Metabolism in Plants-K. S. ROWAN AUTHOR INDEX-SUBJECT

INDEX

Volume 18 The Cell of Langerhans-A.

S. BREATH-

NACH

The Structure of the Mammalian EggROBERTHADEK Cytoplasmic Inclusions in OogenesisM. D. L. SRIVASTAVA The Classification and Partial Tabulation of Enzyme Studies on Subcellular Fractions Isolated by Differential Centrifuging-D. B. ROODYN Histochemical Localization of Enzyme Activities by Substrate Film Methods: Ribonucleases, Deoxyribonucleases, Proteases, Amylase, and Hyaluronidase -R. DAOUST Cytoplasmic Deoxyribonucleic Acid-P. B. GAHANAND J. CHAYEN Malignant Transformation of Cells in VitfO-KATHERINE K. SANFORD Deuterium Isotope Effects in CytologyS. BOSE, H. I. E. FLAUMENHAFT, CRESPI,AND J. J. KATz The Use of Heavy Metal Salts as Electron Stains-C. RICHARDZOBEL AND MICHAELBEER AUTHOR INDEX-SUB

JECT INDEX

Volume 20 The Chemical Organization of the Plasma Membrane of Animal Cells-A. H.

MADDY

Subunits of Chloroplast Structure and Quantum Conversion in Photosynthesis-RODERIc B. PARK Control of Chloroplast Structure by Light -LESTER PACKERAND PAUL-AND& SIEGENTHALER The Role of Potassium and Sodium Ions as Studied in Mammalian BrainH. HILLMAN Triggering of Ovulation by Coitus in the Rat-CLAUDE hON, GITTAASCH,AND JAQUELINE Roos Cytology and Cytophysiology of NonMelanophore Pigment Celk-JOsEPH T. BAGNARA The Fine Structure and Histochemistry of Prostatic Glands in Relation to Sex Hormones-DAm BRANDE8 Cerebellar Enzymology-Lucm ARVY AUTHOR INDEX-SUBJECT

INDEX

360

CON”S

OF PREVIOUS VOLUMES

Volume 23

Volume 21

Histochemistry of Lysosomes-P. B. Transformationlike Phenomena in Somatic Cells-J. M. OLENOV GAHAN L. BRAHM- Recent Developments in the Theory of Physiological Clocks-R. Control and Regulation of Cellular ACHARY PrOCeSSeS-ROBERT ROSEN Ciliary Movement and Coordination in Ciliates-BELA PARDUCA Contractile Properties of Protein Threads from Sea Urchin Eggs in Relation to Electromyography: Its Structural and Cell Division-HIKoIcHI SAKAI Neural Basis-Jom V. BASMA J I ~ Cytochemical Studies with Acridine Electron Microscopic Morphology of Oogenesis-ARNE N#RREVANG Orange and the Influence of Dye Contaminants in the Staining of Dynamic Aspects of Phospholipids during Nucleic Acids-FREDERICK H. KASTEN Protein Secretion-LowELL E. HOKIN Experimental Cytology of the Shoot The Golgi Apparatus: Structure and Function-H. W. BEAMSAND R. G. Apical Cells during Vegetative Growth and Flowering-A. NOUKESSEL G ~ D E The Chromosomal Basis of Sex DeterNature and Origin of Perisynaptic Cells R. LEWIS AND mination--KENmm of the Motor End Plate-T. R. SHANBERNARD JOHN THAVEERAPPA AND G. H. BOURNE AUTHOR INDEX-SUBJECT INDEX AUTHOR INDEX-SUB

JECT INDEX

Volume 24 Volume 22 Synchronous Cell DifferentiationGEORGEM. PADILLAAND IVANL. Current Techniques in Biomedical Electron Microscopy-SAUL WISCHNI~ZER CAMERON The Cellular Morphology of Tissue Re- Mast Cells in the Nervous Systempair--R. M. H. McMI” YNGVE OLSON Structural Organization and Embryonic Development Phases in Intermitosis and DifferentiationxaJANAN V. SHERBET the Preparation for Mitosis of Mammalian Cells in Vitro-BLAcoJE A. AND M. s. LAKSHMI NEBKOVI~ The Dynamism of Cell Division during Early Cleavage Stages of the EggAntimitotic Substances-Guy DEYSSON N. FAUTREZ-FIRLEFYN AND J. FAUTREZThe Form and Function of the Sieve Lymphopoiesis in the Thymus and Other Tube: A Problem in ReconciliationTissues: Functional Implications-N. P. E. WEATHERLEY AND R. P. C . B. EVERETT AND RUTH W. TYLER JOHNSON ( CAFFREY ) Analysis of Antibody Staining Patterns Structure and Organization of the MyoObtained with Striated Myofibrils in Fluorescence Microscopy and Electron neural Junction-C. C O ~ Microscopy-FRANK A. PEPE The Ecdysial Glands of ArthropodsWILLIAM S. HERMAN Cytology of Intestinal Epithelial CellsCytokinins in Plants-B. 1. SAHAISRIVAS- PETERG. TONER TAVA Liquid Junction Potentials and Their Effects on Potential Measurements in AUTHOR INDEX-SUB JECT INDEX Biology Systems-P. C. CALDWELL CUMULATIVE SUBJECT INDEX AUTHOR INDEX-SUB JECT INDEX ( VOLUMES 1-21 )

C 0 " T S

36 1

OF PREVIOUS VOLUMES

Volume 25

Volume 27

Cytoplasmic Control over the Nuclear Events of Cell Reproduction-NOEL DE TERRA Coordination of the Rhythm of Beat in Some Ciliary Systems-M. A. SLEIGH The Significance of the Structural and Functional Similarities of Bacteria and Mitochondria-Snvm NASS The Effects of Steroid Hormones on Macrophage Activity-B. VERNONROBERTS The Fine Structure of Malaria Parasites -MARIA A. RUDZINSE~ The Growth of Liver Parenchymal Nuclei and Its Endocrine Regulation -RITA CARRIERE Strandedness of Chromosomes-!hmLDoN

Wound-Healing in Higher PlantsJACQUESLIPETZ Chloroplasts as Symbiotic OrganellesDENNIS L. TAYLOR The Annulate Lamellae-SAuL WISCHNIlZER

Gametogenesis and Egg Fertilization in Planarians-G. BENAZZI LENTATI Ultrastructure of the Mammalian Adrenal COrteX-sIMON IDELMAN The Fine Structure of the Mammalian Lymphoreticular System-1C~RR Immunoenzyme Technique: Enzymes as Markers for the Localization of Antigens and Antibodies-SmTm AVRAMEAS AUTHOR INDEX-SUB

JECT INDEX

WOLFF

Isozymes : Classification, Frequency, and S i g d i c a n c e - C m n s R. SHAW The Enzymes of the Embryonic Nephron -LUCIE AFWY Protein Metabolism in Nerve Cells-B. DROZ Freeze-Etching-Hms MOOR AUTHOR INDEX-SUB

JECT INDEX

Volume 26 A New for the Living A Summary of the Theory and Recent Experimental Evidence in Its support -GILBERT N. LING The Periphery-LEoNm Mitochondria1 DNA: Physicochemical Properties, Replication, and Genetic Function-p* BoRsT AND AKRooN Metabolism and Enucleated Cells-KoNF~AD KECK Stereological Principles for Morphometry in Electron Microscopic CytologyEWALDR. WEIBEL Some Possible Roles for Isozymic Substitutions during Cold Hardening in Plants-D. W. A. ROBERTS AUTHOR INDEX-SUB

JECT INDEX

Volume 28 The Cortical and Subcortical Cytoplasm of Lymnuea E g g - c m w r w P. RAVEN The Environment and Function of Invertebrate Nerve Cells-J. E. TREHERNE AND R. B. MORETON Virus Uptake, Cell Wall Regeneration, and Virus Multiplication in Isolated Plant Protoplasts-E. C. COCKING The Meiotic Behavior of the Drosovhila OOCYt+ROBERT c. KING The Nucleus: Action of Chemical and physical Agents--RENlk smm The Origin of Bone Cells-MAumEN OWEN

Regeneration and Differentiation of sieve Tube ElementswmLrAM p. jAmBs

Cells, Solutes, and Growth: Salt Accumulation in Plants ReexaminedF. C. STEWARDAND R. L. M o m AUTHOR INDEX--SUB

JECX INDEX

Volume 29 Gram Staining and Its Molecular Mechanism-B. B. BISWAS,P. S. BASU,AND M. K. PAL

362

CONTENTS OF PREVIOUS VOLUMES

The Surface Coats of Animal Cells-A. MART~EZ-PAUIMO Carbohydrates in Cell Surfaces-P.”m J. WINZLER Diflerential Gene Activation in Isolated Chromosomes-Mmus LEZZI Intraribosomal Environment of the Nascent Peptide chain-HIDEKo nJ1 Location and Measurement of Enzymes in Single Cells by Isotopic Methods Part I-E. A. BARNARD Location and Measurement of Enzymes in Single Cells by Isotopic Methods Part 1 1 4 . C. BUDD Neuronal and Glial Perikarya Preparations: An Appraisal of Present Methods AND BETTY -PATRICIA V. JOHNSTON I. ROOTS Functional Electron Microscopy of the Hypothalamic Median EminenceHIDESHIKOBAYASHI, TOKUZOMATSUI,

Morphological and Histochemical Aspects of Glycoproteins at the Surface of Animal Cells-A. RAMBOURC DNA Biosynthesis-H. S. JANSZ,D. VAN DER MEI, AND G. M. ZANDVLXET Cytokinesis in Animal Cells-R. RAPPAPORT

The Control of Cell Division in Ocular Lens-C. V. HARDINC,J. R. REDDAN, N. J. UNAKAR, AND M. BAGCHI The Cytokinins-Ham KENDE Cytophysiology of the Teleost Pituitary -MARTIN SAGE AND HOWARDA. BERN AUTHOR INDEX-SUBJECT

INDEX

Volume 32

Highly Repetitive Sequences of DNA in Chromosomes-W. G. FLAMM The Origin of the Wide Species Variation AND SUSUMI h i U in Nuclear DNA Content-H. REES Early Development in Callus CulturesAND R. N. JONES MICHAELM. YEOMAN Polarized Intracellular Particle Transport: AUTHOR INDEX-SUB JECT INDEX Saltatory Movements and Cytoplasmic Streaming--LroNEL I. REBHUN The Kinetoplast of the HemoflagellatesVolume 30 LARRY SIMPSON High-pressure Studies in Cell BiologyTransport across the Intestinal Mucosal ARTHUR M. ZIMMERMAN Cell: Hierarchies of Function-D. S. Micrurgical Studies with Large FreePARSONS AND C. A. R. BOYD Living Amebas-K. W. JEON AND Wound Healing and Regeneration in the J. F. DANIELLI Crab Paratelphusa h&odrmousThe Practice and Application of Electron RITA G. ADNODI Microscope Autoradiography-J. JACOB The Use of Ferritin-Conjugated AntiApplications of Scanning Electron bodies in Electron MicroscopyMicroscopy in Biology-K. E. CARR COUNCILMAN MORGAN Acid Mucopolysaccharides in Calcified Metabolic DNA in Ciliated Protozoa, Tissues-~mjmo KOBAYASHI Salivary Gland Chromosomes, and AUTHOR INDEX-SUBJECT INDEX Mammalian C e l l s C . R. PELC CUMULATIVE SUBJECT INDEX

( VOLUMES 1-29 )

AUTHOR INDEX-SUB

JECT INDEX

Volume 31

Volume 33

Studies on Freeze-Etching of Cell Membranes-KvRT MUHLETHALER Recent Developments in Light and Electron Microscope Radioautography -C. C. BUDD

Visualization of RNA Synthesis on Chromosomes-0. L. MILLER,JR. AND BAREAFIA A. HAMKALO Cell Disjunction ( “Mitosis”) in Somatic cell Reproduction-ELAINE G . DIA-

363

CONTENTS OF PREVIOUS VOLUMES

SCOTT HOLLAND, AND PAULINE PFCORA Neuronal Microtubles, Neurofilaments, and Microfilaments-RAYMOND B. WUERKERAND JOEL B. KIRKPATRICK Lymphocyte Interactions in Antibody Responses-J. F. A. P. MILLER Laser Microbeams for Partial Cell Irradiation-Mrcwm w. BERNS AND CHRISTIANSALET Mechanisms of Virus-Induced Cell Fusion-GEORcE POSTE Freeze-Etching of Bacteria-CMRLEs C. REMSENAND STANLEYW. WATSON The Cytophysiology of Mammalian Adipose Celk-BERNARD G. SLAVIN CUMAKOS,

AUTHOR INDEX-SUB

JECT INDEX

Volume 34

Synthetic Activity of Polytene Chromosomes-Hms D. BERENDES Mechanisms of Chromosome Synapsis at Meiotic Prophase-PETER B. MOENS Structural Aspects of Ribosomes-N. NANNINGA Comparative Ultrastructure of the Cerebrospinal Fluid-Contacting NeuronsB. VICH AND I. VICH-TEICHMANN Maturation-Inducing Substance in Starfishes-Hmuo KANATANI The Limonium Salt Gland: A Biophysical and Structural Study-A, E. HILL AND B. S. HILL Toxic Oxygen Effects-Hmom M. Swmm AUTHOR INDEX-SUB

JECT INDEX

Volume 36

Molecular Hybridization of DNA and RNA in SitlGWOLFCANG HENNIC The Relationship between the PlasmaNmER lemma and Plant Cell Wall-JEANThe Energy State and Structure of IsoCLAUDEROLAND lated Chloroplasts: The Oxidative Recent Advances in the Cytochemistry Reactions Involving the Water-Splitand Ultrastructure of Cytoplasmic ting Step of Photosynthesis-ROBERT Inclusions in Mastigophora and L. HEATH Opalinata (Protozoa)-G. P. DWI-CA A. Transport in Neurospora-GENE Chloroplasts and Algae as Symbionts in SCARBOROUGH MOIIUSCS-LEONARDMUSCATINEAND Mechanisms of Ion Transport through RICHARDW. CRJZENE Plant Cell Membranes-EhuNmL The Macrophage-SamoN GORDONAND ERSTEIN ZANVIL A. COHN Cell Motility: Mechanisms in ProtoDegeneration and Regeneration of Neuroplasmic Streaming and Ameboid secretory Systems-Howr-DIETER Movement-H. KOMNICK,W. STOCDELLMANN KEM, AND K. E. WOHLEFARTHAUTHOR INDEX-SUB JECT INDEX BOTTERMANN The Gliointerstitial System of MolluscsGHISLAINNICAISE Volume 37 Colchicine-Sensitive Microtubles-Lm Units of DNA Replication in ChromoMARGULIB somes of Eukaroytes-J. HERBERT AUTHOR INDEX-SUB JECT INDEX The Submicroscopic Morphology of the Interphase Nucleus-!huL WISCH-

TAYLOR

Volume 35 The Structure of Mammalian Chromosomes-ELTON STUBBLEFIELD

Viruses and Evolution-D. C. REANNEY Electron Microscope Studies on Spermiogenesis in Various Animal SpeciesGONPACHIRO YASuzuMI Morphology, Histochemistry, and Bio-

364

CONTENTS OF PREVIOUS VOLUMES

chemistry of Human Oogenesis and Ovulation--Smm. S. GURAYA Functional Morphology of the Distal Lung-KAm H. KILBuRN Comparative Studies of the Jwtaglomerular Apparatus-Hmomrm SOWE AND MIZUHOOGAWA The Ultrastructure of the Local Cellular CIRR Reaction to Neoplasia-h AND J. C. E. U ~ ~ w w o o ~ Scanning Electron Microscopy in the Ultrastructural Analysis of &e M m malian Cerebral Ventricular System~, D. E. S m , G. p. K ~ L O W S AND M. N. SHERIDAN AUTHOR INDEX-SUBJECT

INDEX

Volume 38

Volume 39 Androgen Receptors in the Nonhistone Protein Fractions of Prostatic Chromatin-TUNG YUE WANC AND LEROY M. NYBERG Nucleocytoplasmic Interactions in Development of Amphibian HybridsSTEPHEN SUBTELNY The Interactions of Lectins with Animal Cell SurfaCeS-GARTH L. NICOLSON Structure and Function of Intercellular Juncti0ns-L. ANDREW STAEHELIN Recent Advances in Cytochemistry and Ultrastructure of Cytoplasmic Inclusions in Ciliophora (Protozoa)-G. P. DUTTA Structure and Development of the Renal Glomerulus as Revealed by Scanning Electron Microscopy-FRANC0 SplNELLI

Genetic Engineering and Life Synthesis: An Introduction to the Review by R. Widdus and C. Ault-jAms F. DANIELLI Progress in Research Related to Genetic Engineering and Life Synthesis-Roy WIDOWS AND CHARLES R. AULT The Genetics of C-Type RNA Tumor Viruses-J. A. W m Three-Dimensional Reconstruction from Projections: A Review of AlgorithmRICHARD GOWN AND GABOR T. HERMAN The Cytophysiology of Thyroid CellsVLADIMIR R. PANTIk The Mechanisms of Neural Tube Formation-Prmw KARFUNJCEL The Behavior of the XY Pair in Mamrnak-ALBmm J. soFine-Structural Aspects of Morphogenesis in Acetabu~~d~-G.WE= Cell Separation by Gradient Centrifugation-R. HARWOOD

Recent Progress with Laser Microbeams -MICHAEL W. BERNS The Problem of Germ Cell Determinants -H. W. BEAMSAND R. G. KESSEL SUBJECT INDEX

B-Chromosome Systems in Flowering Plants and Animal Species- R. N. JONES The Intracellular Neutral SH-Dependent Protease Associated with Inflammatory Reactions - HIDEOHAYASHI The Specificity of Pituitary Cells and Regulation of Their ACtiVitieS-vLADIMlR R. PANTIC Fine Structure of the Thyroid GlandHISAOFUJITA Postnatal Gliogenesis in the Mammalian Brain -A. PRIVAT Three-Dimensional Reconstruction from Serial Sections-UNDLE W. WAREAND VINCENT LOPREST1

SUBJECT INDEX

SUBJECT INDEX

40

A B C 0 E F G H 1

5 6 7 B 9 O 1 2

J 3

International Review Of Cytology: A Survey of Cell Biology

Read more

International Review of Cytology: A Survey of Cell Biology, Volume 205

Read more

International Review of Cytology: A Survey of Cell Biology, Volume 254

Read more

International Review of Cytology: A Survey of Cell Biology, Volume 198

Read more

International Review of Cytology: A Survey of Cell Biology, Volume 139

Read more

International Review of Cytology, Volume 185: A Survey of Cell Biology

Read more

International Review of Cytology: A Survey of Cell Biology, Volume 222

Read more

Sexual Reproduction in Flowering Plants (International Review of Cytology: A Survey of Cell Biology, Volume 140)

Read more

International Review of Cytology: A Survey of Cell Biology, Volume 11

Read more

International Review of Cytology: A Survey of Cell Biology, Volume 209

Read more

International Review of Cytology: A Survey of Cell Biology, Volume 105

Read more

International Review of Cytology: A Survey of Cell Biology, Volume 26

Read more

International Review of Cytology: A Survey of Cell Biology, Volume 239

Read more

International Review of Cytology: A Survey of Cell Biology, Volume 57

Read more

International Review of Cytology: A Survey of Cell Biology, Volume 95

Read more

International Review of Cytology: A Survey of Cell Biology, Volume 35

Read more

International Review of Cytology: A Survey of Cell Biology, Volume 114

Read more

International Review of Cytology: A Survey of Cell Biology, Volume 233

Read more

International Review of Cytology: A Survey of Cell Biology, Volume 113

Read more

International Review of Cytology: A Survey of Cell Biology, Volume 90

Read more

International Review of Cytology: A Survey of Cell Biology, Volume 195

Read more

International Review of Cytology: A Survey of Cell Biology, Volume 144

Read more

International Review of Cytology: A Survey of Cell Biology, Volume 118

Read more

International Review of Cytology: A Survey of Cell Biology, Volume 106

Read more

International Review of Cytology: A Survey of Cell Biology, Volume 48

Read more

International Review of Cytology: A Survey of Cell Biology, Volume 264

Read more

International Review of Cytology, Volume 195: A Survey of Cell Biology

Read more

International Review of Cytology: A Survey of Cell Biology, Volume 8

Read more

International Review of Cytology: A Survey of Cell Biology, Volume 128

Read more

International Review of Cytology: A Survey of Cell Biology, Volume 263

Read more

Recommend Documents

International Review Of Cytology: A Survey of Cell Biology

International Review of A Survey of Cytology Cell Biology VOLUME 257 SERIES EDITORS Geoffrey H. Bourne James F. Dan...

International Review of Cytology: A Survey of Cell Biology, Volume 205

International Review of A Survey of Cytology Cell Biology VOLUME 205 SERIES EDITORS Geoffrey H. Bourne James F. Dan...

International Review of Cytology: A Survey of Cell Biology, Volume 254

International Review of A Survey of Cytology Cell Biology VOLUME 254 SERIES EDITORS Geoffrey H. Bourne James F. Dan...

International Review of Cytology: A Survey of Cell Biology, Volume 198

International Review of A Survey of Cytology Cell Biology VOLUME 198 SERIES EDITORS Geoffrey H. Bourne James F. Dan...

International Review of Cytology: A Survey of Cell Biology, Volume 139

VOLUME 139 SERIES EDITORS Geoffrey H. Bourne James F. Danielli Kwang W. Jeon Martin Friedlander 1949-1 988 1949-1 98...

International Review of Cytology, Volume 185: A Survey of Cell Biology

VOLUME 185 SERIES EDITORS Geoffrey H. Bourne James F. Danielli Kwang W. Jeon Martin Friedlander Jonathan Jarvik 1949...

International Review of Cytology: A Survey of Cell Biology, Volume 222

International Review of A Survey of Cytology Cell Biology VOLUME 222 SERIES EDITORS Geoffrey H. Bourne James F. Dan...

Sexual Reproduction in Flowering Plants (International Review of Cytology: A Survey of Cell Biology, Volume 140)

International Review of cyt oI ogy SEXUAL REPRODUCTION IN FLOWERING PLANTS VOLUME 140 Cyto I og y Geoffrey H. Bourn...

International Review of Cytology: A Survey of Cell Biology, Volume 11

INTERNATIONAL REVIEW OF CYTOLOGY VOLUME11 This Page Intentionally Left Blank INTERNATIONAL Review of Cytology EDI...

International Review of Cytology: A Survey of Cell Biology, Volume 209

International Review of A Survey of Cytology Cell Biology VOLUME 209 SERIES EDITORS Geoffrey H. Bourne James F. Dan...