International Review of Cytology: A Survey of Cell Biology, Volume 128

VOLUME 128

SERIES EDITORS Geoffrey H. Bourne James F. Danielli Kwang W. Jeon Martin Friedlander

1949-1988 1949-1984 19671984-

ADVISORY EDITORS H. W. Beams Howard A. Bern Dean Bok Gary G. Borisy Bharat 6. Chattoo Stanley Cohen Rene Couteaux Marie A. DiBerardino Donald K. Dougall Charles J. Flickinger Nicholas Gillham M. Nelly Golarz De Bourne Elizabeth D. Hay Mark Hogarth Keith E. Mostov Audrey Muggleton-Harris

Andreas Oksche Muriel J. Ord Valdimir R. Pantic M. V. Parthasarathy Lionel 1. Rebhun Jean-Paul Revel L. Evans Roth Jozef St. Schell Hiroh Shibaoka Joan SmithSonneborn Wilfred Stein Ralph M. Steinman Hewson Swift Masatoshi Takeichi M. Tazawa Alexander L. Yudin

Edited by Kwang W. Jeon Department of Zoology The University of Tennessee Knoxville, Tennessee

Martin Friedlander Jules Stein Eye Institute UCLA School of Medicine Los Angeles, California

VOLUME 128

Academic Press, Inc. Harcourt Brace Jovanovich, Publishers San Diego New York Boston London

Sydney Tokyo Toronto

This book is printed on acid-free paper. @ Copyright 0 1991 BY ACADEMIC PRESS, INC. All Rights Reserved. No 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. San Diego, California 92101 United Kingdom Edition published by ACADEMIC PRESS LIMITED 24-28 Oval Road, London NW1 7DX

Libraty of Congress Catalog Card Number: 52-5203 ISBN 0-12-364528-X (alk. paper)

PRINTED IN THE UNITED STATES OF AMERICA

91 92 93 94

9 8

7 6 5 4 3 2 1

Contributors ................................................................................................................

ix

The Replication. Differentiation. and Inheritance of Plastids with Emphasis on the Concept of Organelle Nuclei Tsuneyoshi Kuroiwa I.

11. Ill. IV . V. VI . VII .

Introduction ......................................... ..... ............ ........................... Location of the Plastid Nuclei in Plastid Organization of the Plastid Nucleus ................................................................ Division of Plastids ...................................................... ......... Differentiation of Plastids .. ......................................... Cytoplasmic Inheritance of Plastids ............................................................... .......... Summary ................................................... References ................... ..........................................................

1 3 6 9 25 36 55 58

Cell Division in Diatoms Jeremy Pickett-Heaps .................................................................................. I . Introduction .......... I I . Early Work ................................................. .......................................................... Ill. The Microtubule Center (MC) IV. The Central Spindle ...................................................... ................................................................. V . Kinetochores ................ VI . The "Collar": A Spindle Matrix? ...................................................................... .......... VII . Spindle Elongation (Anaphase B) by Sliding of Half Spindles . VIII . Control of Microtubule Stability ....................................................................

63 64 64 70 76 83 89 94 v

vi

CONTENTS

IX. Cleavage ......................................................................................................... X . Summary ....................................................................................................... References .....................................................................................................

98 98 104

Nerve Growth Factor Synthesis and Nerve Growth Factor Receptor Expression in Neural Development Alun M. Davies I. I1. 111. IV. V. VI . VII .

Introduction ........................................................ ..................................... The Neurotrophic Theory ............................................................................... Molecular Biology. Biochemistry. and Biosynthesis of NGF ........................... NGF Synthesis ................................................................................ Biochemistry and Molecular Biology of NGF Receptors ................................. NGF Receptor Expression . ........................................................................ Conclusions .......................................................... References ............................... ....................................................

109 109 112 113 123 125 132 133

Control of Calcium Regulating Hormones in the Vertebrates: Parathyroid Hormone. Calcitonin. Prolactin. and Stanniocalcin S . E. Wendelaar Bonga and P. K . T . Pang I. I1. Ill. IV.

Introduction ................................................................................................... Terrestrial Vertebrates .................................................................................... Aquatic Vertebrates ........................................................................................ Conclusions ................................................................................................... References .....................................................................................................

139 142 177 194 203

lmmunoarchitecture of Regenerated Splenic and Lymph Node Transplants R. Pabst. J . Westermann. and H. J. Rothkotter 1. Introduction ................................................................................................... I1. Functional Anatomy of the Spleen .................................................................. 111. Phases of Regeneration of Splenic Autotransplants .......................................

215 217 223

CONTENTS IV. V. VI . VII .

Which Factors Influence Splenic Regeneration? ............................................ Function of Regenerated Splenic Tissue ........................................................ Regeneration of Autotransplanted Lymph Nodes ........................................... Conclusions ................................................................................................... References .....................................................................................................

vii 232 239 246 250 253

Molecular Biology in Studies of Ocean Processes Paul G . Falkowski and Julie LaRoche 1. II. Ill. IV.

Introduction ................................................................................................... Marine Organisms and Oceanographic Processes ......................................... Molecular Approaches to Oceanography ........................................................ The Application of Molecular Techniques for Identification. Enumeration. and the Study of Genetics of Marine Organisms ............................................ V . Application of Molecular Techniques to Studies of Organism Function .......... VI . Potential for Biotechnological Exploitation ..................................................... VII . Conclusions ................................................................................................... References .....................................................................................................

261 262 265

Index .........................................................................................................................

305

273 284 296 296 297

This Page Intentionally Left Blank

CONTRIBUTORS

Numbers in parentheses indicate the pages on which the authors’ contributions begin

S. E. Wendelaar Bonga (139),Department of Animal Physiology, Faculty of Science, University of Nijmegen, 6525 ED Nijmegen, The Netherlands Alun M. Davies (109),Department of Anatomy, St. George’s Hospital Medical School, London SW17 ORE England Paul G. Falkowski (261), Oceanographic and Atmospheric Sciences ~ivision, Brookhaven National Laboratory, Upton, New York I 1973 Tsuneyoshi Kuroiwa (l),Department of Biology, Faculty of Science, Division of Developmental Biology, University of Tokyo, Hongo 113,Japan Julie LaRoche (261), Oceanographic and Atmospheric Sciences Division, Brookhaven National Laboratory, Upton, New York 11973

R. Pabst (215), Center ofAnatomy, Medical School of Hannover,0-3000 Hannover 61, Germany P. K. T.Pang (139),Depattment of Physiolog~Schoul of MedicineJUniversi~of AlbertaJEdmonton, Alberta, Canada T6G 2H7 Jeremy Pickett-Heaps (63), School of Botany, University of Melbourne, Parkville, Victoria305ZJAustralia H. J. Rothkotter (215), Center of Anatomy, Medical School of Hannover, 0-3UOU Hannover 61, Germany

J. Westermann (215), Center of Anatomy, Medical School of Hannover, D-3UUU Hannover 61, Germany

ix

This Page Intentionally Left Blank

INTERNATIONAL REVIEW OF CYTOLOGY, VOL. 128

The Replication, Differentiation, and Inheritance of Plastids with Emahasis on the Concept of Organelle Nuclei TSUNEYOSHI KUROIWA Department of Biology, Faculty of Science, Division of Developmental Biology, University of Tokyo, Hongo 113, Japan

I. Introduction During the past 15 years, there have been remarkable advances in at least three areas related to the chloroplast genome. One of the most active areas involves the molecular biology of the organization of the genes, which has been based on sequencing of the chloroplast genome. Higherplant chloroplast (cp) DNA can be isolated as a covalently, closed circular molecule with a molecular mass of 85-95 x lo3 kDa (Kolodner and Tewari, 1975). Denaturation mapping (Kolodner and Tewari, 1975) and restriction endonuclease analysis (Bedbrook and Bogorad, 1976) have shown that the majority of the circular molecules in the chloroplasts of a given species are identical in sequence. Since the construction of the physical maps of the cpDNA from Zea mays (Bedbrook et al., 1977) and Chlamydomonas reinhardtii (Rochaix, 1978), the maps of cpDNA from various plants have been reported (see Palmer, 1985). Shinozaki et al. (1986) and Ohyama et al. (1986), respectively, sequenced the entire cpgenome from the chloroplasts of Nicotiana tabacum and Marcantia polymorpha. The cpDNA from N . tabacum and M . polymorpha contains 155,844and 121,024bp, respectively. Each contains about 80 genes which encode a complete set of 30 tRNAs, four rRNAs (23 S, 16 S, 5 S, and 4.5 S), 20 ribosomal proteins, and 22 proteins of thylakoid membrane complexes. In addition there are about 30 open reading frames for which the functions remain to be determined. Thus, each of the chloroplasts contains a specific genome that is essential for the semiautonomy of the organelle. A second area that has seen great progress is related to the biogenesis of the chloroplast membrane system. ATPase and the complexes of photosystems I and 11 are supramolecular complexes of enzymes located in the membrane of chloroplasts. The ATPase consists of a set of different subunits designated a,p, y , 6, and E (Nelson et al., 1980). Three subunits (a,p, and 6) are synthesized within chloroplasts and the remaining two in the cytoplasm (Watanabe and Price, 1982).Close cooperation between the two genetic systems and the two compartments of the cell is necessary for 1 Copyright 0 1991 by Academic Press. Inc. All rights of reproduction in any form reserved.

2

TSUNEYOSHI KUROlWA

the biosynthesis of the ATPase. The complex molecular and cellular mechanism of formation of the enzyme involves: (1) synthesis of different components in the cytoplasm and in the plastids; (2) transport and integration of the enzyme precursors; (3) organization of precursor molecules in an assembly process; and (4) arrangement of the final supramolecular complex within the membrane layer. Coordination of these individual processes is required for the formation of a functional supramolecular complex, such as the ATPase and photosystems I and 11. It is suggested that the molecular chaperone is related to the assembly or formation of supramolecular complexes of the proteins, such as RuBisCO, ATPase, LHCP 11. etc. in chloroplasts (Lubben et al., 1989). As described above, research related to the organization of the cpgenome and the biogenesis of plastids, including assembly of subunit proteins, developed as a result of experiments with a population of DNA or proteins isolated from whole tissues or organs of plants. However, it must be remembered that one leaf is composed of a variety of tissues such as spongy parenchyma, palisade parenchyma, epidermis, etc. These tissues contain large numbers of cells. One cell contains many plastids which do not divide synchronously. Accordingly, it should be noted that results obtained from a population do not always reflect events in an individual plastid . Progress in cell biology related to the distribution, organization, separation, differentiation, and inheritance of plastids has also been considerable. One remarkable development has added new dimensions to our concepts of the distribution, organization, separation, and inheritance of the pt-genome within the last 10 years. Results of electron microscopy suggested that plastids arise from the division of preexisting organelles and can differentiate into various types of plastid, such as amyloplasts, chromoplasts, chloroplasts, leucoplasts, etc. However, the following questions remain to be answered. Do all of these plastids (pt) contain DNA? How are the ptDNAs organized into organelles? How are pt-chromosomes separated into daughter organelles? How does plastidkinesis occur? How are ptDNAs transmitted to a cell's descendants? How many copies of ptDNA does each plastid contain? What interactions are there among the cell nuclear, mitochondrial (mt), and pt-genomes? The answeres to these basic questions appear to offer a key to the understanding of the timing of expression of genes on the pt-genome and the insertion of newly synthesized proteins for biogenesis occurring during the division cycle and differentiation of plastids. A DNA-binding fluorochrome, 4'-6-diamidino-2-phenylindole(DAPI), which emits stronger fluorescence than do conventional fluorochromes, was synthesized by Dann et al. (1971) and applied to observations of organelle DNA. The use of DAPI combined with epifluorescence micro-

PLASTIDS AND ORGANELLE NUCLEI

3

scopy makes it possible to visualize extremely low levels of DNA in various organelles to analyze their behavior. The results of such analyses indicate that cp- and mt-DNA are not naked but are organized, with proteins, to form organelle nuclei. Therefore, all of organelle DNA can be observed under DAPI-epifluorescencemicroscopy. A description of the molecular organization of the plastid genome and the biogenesis of plastids is omitted here since each has been reviewed exhaustively elsewhere (Rochaix, 1985; Sugiura, 1987; Lubben et al., 1989). However, since only the rough outlines are known of the organization, separation, and inheritance of the pt-genome, as revealed by various cell-biological techniques (including DAPI-epifluorescencemicroscopy) it seems appropriate to review these issues at this time. 11. Location of the Plastid Nuclei in Plastids of Various Plants

Ris and Plaut (1962) described DNA-like fibers within the chloroplastsof C. reinhardtii. These fine fibrils appeared to become clumped in an electron-transparent area, the “nucleoid,” in the matrix of chloroplasts after conventional fixation. Such DNA-containing regions have also been reported in the chloroplasts of many other plants, such as red alga, brown alga, green alga (Werz, 1966; Yokomura, 1967), and higher plants (Kislev et al., 1965; Gunning, 1965; Yokomura, 1967). In numerous plants, a small number of DNA-like fibrils appear in an electron-transparent spherical area, 0.1-0.5 pm in diameter, within individual chloroplasts. However, even within a single species, such an area (the DNA-like fibers containing the area) is not always visible in all chloroplasts under an electron microscope because a very small amount of DNA is embedded in the semielectron-dense matrix in chloroplasts under standard physiological conditions and with conventional fixation techniques. Even when the DNA-like fibers are visible in electron-transparent areas, the amount of DNA fibers is much lower than anticipated. Probably, some parts of the DNA-like fibers are embedded in a somewhat electron-dense matrix around the electron-transparent regions. Although the DNA-containing regions in the electron-transparent area (ETA) of plastids are conventionally called “plastid nucleoids,” they do not contain the entire DNA of the genome. By contrast, after staining with DAPI, the fluorescent spots and intact isolated ptDNA regions show all pt-chromosomes in plastids. Therefore, we consider it preferable to designate these compact isolated structures found in situ as “pt-nuclei” (Kuroiwa et al., 1981; Kuroiwa, 1982; Nemoto et d., 1988), analogous to “bacterial nuclei’’ (Robinow, 1956) and “mitochondria1 nuclei”.(Kuroiwa et al., 1976; Kuroiwa, 1982). Based on this definition, the nuclei are called “cell nuclei.” As described above, electron microscopic examination is

4

TSUNEYOSHI KUROIWA

not always a suitable technique for monitoring the distribution of all the DNA in chloroplasts. DAPI was used for the first time for staining mtDNA in yeast by Wiliamson and Fennel1 (1973, cpDNA in a higher plant by James and Jope (1978), and in green algae by Coleman (1978). The blue-white fluorescence of DAPI is stronger than that of other fluorochromes such as ethidium bromide, Hoechst 33258, acridine orange, etc., which were previously used to stain DNA in organelles. In addition, epifluorescence microscopes have been improved to generate strong fluorescence as a result of strong excitation light. Furthermore, DAPI epifluorescence microscopy and an apparatus that combines it with conventional microscopic fluorimetry, or videointensified photon counting system (VIMPICS), made it easy to observe the distribution or behavior of cpDNA and to estimate extremely low and to estimate extremely low levels of DNA per plastid. Kuroiwa el al. (1981) examined the distribution and behavior of ptnuclei during the development of chloroplasts and the cycle of plastid division in many plants by staining with DAPI. The number, size, shape, and distribution of pt-nuclei were found to change during chloroplast development and during the division of plastids, and these parameters differed among various plants. The small proplastids of early embryonic cells of Brassicaljuncea and of cultured cells of N. tabacum (Figs. l a and 2) contain only one small, spherical proplastid (pp) nucleus, 0.2 pm in diameter, whereas the pp-nucleus in the proplastids of dormant embryonic cells is ovoid, 0.5-1 .Opm in diameter and, thus, several times larger than the pp-nucleus of the early embryonic cells. Since the proplastids divide actively, the increase in volume of pt-nuclei seems to be due to endoduplication of ptDNA. When proplastids develop into etioplasts in the dark, the size of the pp-nucleus increases 2- to 4-fold, the pp-nucleus becomes cup shaped, and is often found near starch grains or a prolamellar body. Once etioplasts are illuminated, the pt-nuclei begin dividing into tiny spherical structures and their numbers increase markedly to reach more than 20 cp-nuclei in fully mature chloroplasts, which emit red autofluorescence (Figs. lb and 2). In most land plants and algae examined extensively by Kuroiwa et al. (1981, 1989b) and Coleman (1989, the patterns of distribution of pt-nuclei in mature chloroplasts were peculiar to plant groups, although they changed during the division cycle (Kuroiwa et al., 1981; Zachler and Cepfik, 1987) and the development of plastids. FIG. 1 . Photomicrographs of cell nuclei (CN), a proplastid nucleus (large arrow in a), mitochondria1nuclei (small arrows in a and b), and a chloroplast nucleus (large arrow in b) in a Nicoriunu rubacum cultured cell (line BY-2) (a); and a mature leaf cell (line BY-2) (b) after staining with DAPI. Bar = 10 pm.(Photographs courtesy of Dr. Y. Nemoto, Tokyo Agriculture and Technology University, Japan.)

PLASTIDS AND ORGANELLE NUCLEI Proplastid division

Proplastid division

Etioplast division

5

Chloroplast division

FIG.2. A diagram of plastid nuclear events during chloroplast development and senescence. The chloroplasts of the SN, CN, CL, PS, and SP type are differentiated from proplastids throughout several plastid divisions. The term plastid nucleus indicates the cp-nucleaus and the terms plastid genome and chloroplast genome mean the same thing. Pt-M, Pt-S, and CDC show the plastid-division stage, the plastid DNA-synthetic stage, and the chloroplast division cycle, respectively. (From Kuroiwa ef al., 1981; reproduced by permission of the Japanese Society of Plant Physiologists.)

Most eukaryotic plants can be classified into five types according to differences in the shape, size, and distribution of the cp-nuclei in their mature chloroplasts (Fig. 2; Kuroiwa er al., 1981). The first, the SN type (scattered pt-nuclei), is characterized by chloroplasts with small, uniformly dispersed cp-nuclei in the matrix between thylakoid membranes and/or granas. The land plants and algal groups, such as Chlorophyceae, Prasinophyceae, Chloroohyceae, Euglenophyceae, Cryptophyceae, Eustigmatophyceae, and Dinophyceae, are of the SN type (Fig. 2; Kuroiwa er al., 1981; 1989b). The second, the CN type (centrally located pt-nuclei), is characterized by chloroplasts with one or a few cp-nuclei located in the central area surrounded by lamellae. Cyanidiurn caldarium RK-1, which was reported to be a red alga, is a typical example of this type (Fig. 2). Recently, Miyamura and Hori (1989) found an unusual type of chloroplast

6

TSUNEYOSHI KUROIWA

in Caulerpa okamurae, the pyrenoid of which contained one large cpnucleus during a particular phase of the life cycle. Sodmergen et al. (1989) reported that the cells of the coleoptile of Oryza satiua contain mature chloroplasts with 1-3 centrally located cp-nuclei of the CN type while the cells in the first and second leaves contain chloroplasts with dispersed cp-nuclei of the SN type. The third, the CL type (circular pt-nuclei), has chloroplasts with a large, ring-shaped cp-nucleus inside the girdle lamellae (Fig. 2). Such a circular pt-nucleus isolated from the brown alga Ectocarpus indicus appears to be a chain of small spherical particles which may correspond to the small cp-nuclei of the SN type (Kuroiwa and Suzuki, 1981). Algae, such as the Chrysophyceae, Xanthophyceae, Bacillariophyceae, Phaeophyceae, Rhaphidphyceae, and Haptophyceae, are of the CL type. The shape of the circle changes with the shape of the chloroplast in different species. In some diatoms and brown algae, the chloroplasts are disc shaped and the cp-nucleus is a ring; in other diatoms and in C. caldurium M-8,which has irregularly shaped chloroplasts, the ptnucleus forms an irregular circle that lies along the periphery of the chloroplast (Nagashima et al., 1986; Kuroiwa et al., 1989). The fourth, the PS type (peripherally scattered pt-nuclei), and the fifth type, the SP type (spread pyrenoid pt-nucleus), are modifications of the CL type and the SN type, respectively. The PS type is characterized by chloroplasts with pt-nuclei scattered along their peripheries beneath an inner limiting membrane. The rhodophycean algae, such as Gellkidium amansii and Symphyocfadia latiusculu, are of this type. The last SP type has numerous small pt-nuclei which form a shell around a pyrenoid in the chloroplast (Fig. 2). The green alga Bryopsis plumos is of this type. The pt-nuclei tend frequently to be located near the pyrenoid. It is interesting that the CN type, which can be found among the lower eukaryotes, was also observed in undifferentiated proplastids and chloroplasts, and the CL type, to which chloroplasts in some red algae and almost all brown algae belong, can be observed in etioplasts before the development of chloroplasts in monocots (Sellden and Leech, 1981). Such diversity in the distribution, number, size, and shape of pt-nuclei in various plants must depend basically on the development of the thylakoid membrane during the evolution of membrane systems and the development of chloroplasts (Kuroiwa, 1982; Rose, 1988). III. Organization of the Plastid Nucleus As described above, almost all of the mature chloroplasts of land plants and algae contain a pt-genome which is located in the specific regions of chloroplasts that develop from tiny proplastids. The proplastids also can

PLASTIDS AND ORGANELLE NUCLEI

7

differentiate both directly and indirectly from other plastids, such as etioplasts, chloroplasts, chromoplasts, leucoplasts, amyloplasts, etc., in a tissue-specific manner and/or depending upon such environmentalfactors as light and temperature. These plastids also contain pt-genomes within a specific area of the plastids. The functional and structural changes in plastids during leaf development are accompanied by an accumulation of plastid proteins, many of which are encoded by pt-genes (Klein and Mullet, 1986). Pt-genes are located on a circular strand of DNA which is 1.2 X Id to 1.8 x 10’ bp in length (Figs. 4e and 4h). This DNA contains up to 137 genes which encode tRNA, rRNA (16 S, 23 S, 4.5 S, and 5 S), and numerous proteins (Fig. 4h). It has been pointed out that the packing mode, i. e., the dispersion and condensation of the cp-nuclei, is intimately related to the photosynthetic oxygen-evolvingactivity of the cells in Chfamydornonas(Nakamura et al., 1986). Therefore, the three-dimensional structure of pt-nuclei may be important for understandingthe function of ptDNA during the division and differentiation of plastids. To elucidate the organizations of pt-nuclei, the intact pt-nuclei must be isolated from the plastids of various plants. Several groups have tried to isolate pt-nuclei from chloroplasts or chromoplasts (Kuroiwa and Suzuki, 1981; Briat et af., 1982;Reiss and Link, 1985; Hansmann et al., 1985). It is, however, difficult to purify the isolated pt-nuclei because the thylakoid membrane system in chloroplasts and chromoplasts is highly developed. By contrast, proplastids have poorly developed membrane systems and they contain the smallest numbers of copies of pt-genome. Thus, they are useful for the isolation of intact pt-nuclei. Nemoto et al. (1988) developed a method for isolating morphologically intact proplastids in large quantities from protoplasts of N. tabacum by a method that involves disruption of cells by forcing them through a layer of nylon mesh (Fig. 3). Isolated proplastids contain one to several pp-nuclei which are similar in appearance to pp-nuclei in viuo (Figs. 3d-3n). After treatment with the detergent Nonidet P-40, the pp-nuclei remain as small, spherical particles which are composed of fine fibrils (Figs. 4a-4e). The isolated pp-nuclei contain a number of polypeptides, only four of which (69, 31, 30, and 14 kDa) are bound to the ppDNA. After treatment with proteinase K or deproteinization, a number of loops of DNA fibrils are also released from the cp-nuclei and the circular DNA is observed (Fig. 4f, g). It is possible to reconstruct the pt-nucleus from the pt-DNA and the relevant polypeptides after dialysis (Nemoto et af., 1989). Therefore, it appears that one to several copies of the ptDNA molecule, which is approximately 53 pm in length, are packed into a pt-nucleus of approximately 0.5 p m in diameter (Fig. 4h). Since the packing ratio of DNA in the metaphase chromosomes in animals is about 140 (DuPraw, 1970),that of the pt-nuclei

8

TSUNEYOSHI KUROIWA

PLASTIDS AND ORGANELLE NUCLEI

9

may be higher than those of the cell nucleus. This hypothesis is supported by the results of fluorimetry of cellular DNA in the higher plant (Kuroiwa et al., 1990a)and of immunogold electron microscopy in which the number of gold particles per unit area in the cp-nuclei was higher than that in the cell nucleus of the algae (Johnson and Rosenbaum, 1990; Scheer et al., 1987). From these results, it is concluded that the ppDNAs are not naked in situ but are organized by interactions with some basic proteins to form compact structures called the “pt-nuclei.” This concept should be applicable to the pt-nuclei of other types of differentiated plastid.

IV. Division of Plastids The concept of organelle nuclei has changed our ideas about the division of organelles themselves. It seems clear that the process of organelle division must be composed of two main events: division of the organelle nucleus and organellekinesis (division of the other components of the mitochondrion or plastid). The latter term has been adopted as an appropriate analogue of cytokinesis.

A. DIVISION OF THE PLASTID NUCLEUS In general, when the cell volume becomes approximately double during metaphase, the cell divides into daughter cells with an equal division of chromosomes. Then the cell cycle is repeated with a doubling of both DNA and cell volume during the next cell cycle. Similar events may occur in the case of plastids and one might assume that a round of replication of cpDNA occurs during each cycle of chloroplast division in higher plants (Possingham and Rose, 1976; Szmidt et al., 1983;Rose, 1988). However, it

FIG. 3. Isolation of proplastids. (a) Living cells. (b) Protoplasts observed by phasecontrast microscopy (arrowhead indicates a cell nucleus). (c) Fluorescence photomicrograph of a protoplast fixed with glutaraldehyde and stained with DAPI, showing a cell nucleus (CN) and many tiny fluorescent spots from pt-nuclei (large arrow) and mitochondrial nuclei (small arrow) in the cytoplasm (d-i). Part of a protoplast prepared by squashing and observed by DAPI-fluorescence (d, g), DAPI-fluorescence and phase-contrast (e, h), and phase-contrast microscopy (f, i). (d, e, f ) and (g, h, i) are the same fields, respectively. Large and small arrow in (d) indicate a pt-nucleus and a mt-nucleus, respectively. (i-n) Isolated proplastids observed by DAPI-fluorescence(i),fluorescence and phase-contrast (k),phase-contrast (l), and electron microscopy (m, n).(i-l) are the same fields. Large arrows indicate pp-nuclei. a and b, x 330. Bars in c, d, m, and n represent 100,100,land 0.1 pm,respectively. (From Nemoto er al. with minor modifications, 1988; reproduced by permission of the Japanese Society of Plant Physiologists.)

10

TSUNEYOSHI KUROIWA

PLASTIDS AND ORGANELLE NUCLEI

11

is difficult to obtain direct evidence to prove this hypothesis because the chloroplast contains amounts of DNA that are too small to examine quantitatively by pulse-labeling autoradiography and, furthermore, the chloroplasts do not divide synchronously even within a single cell. Therefore, various phases of the cycle such as plastid GI, plastid S, plastid G2, and plastid M have not been clearly identified. Division cycles of three different types, namely, CL type, SN type, and CN type, have been studied in detail in plastids. In the CL type, the ring-shaped cp-nucleus segregates into two daughter loops, each of which is transmitted to a daughter chloroplast (Kuroiwa et al., 1981). In the SN type, each of 20-30 small pt-nuclei, which are dispersed throughout the entire chloroplast, divides and the number of pt-nuclei doubles. Subsequently, equal numbers of pt-nuclei appear to be packaged into daughter chloroplasts (Kuroiwa et al., 1981; Miyamura el al., 1986). In spite of the difference in the location of ptDNA, the pt-chromosomes, which are synthesized during the plastid S phase, must be separated equally into daughter plastids. The CL and SN types of division cycle are observed in chloroplasts of brown algae and in the etioplasts of higher plants, and in chloroplasts of green algae and green land plants, respectively. Since plastid divisions of the CL and SN types have been observed in multiplastidic cells of plants and are often not synchronized, it is difficult to study these division cycles in detail. By contrast, the typical CN type of division cycle is observed in proplastids in the young thalli of algae, in young meristems of leaves (Miyamura et al., 1990),and in chloroplasts of the red alga C. caldarium RK-1. Since C. caldarium RK-I contains one cell nucleus, one mitochondrion, and one chloroplast (Nagashimaand Fukuda, 1981),it is useful for analysis of plastid division cycles. The basic life cycle of C. caldariurn RK-1 is shown in Fig. 5 . If we start with an examination of young cells, we see that each cell contains a cell nucleus of about 0.1 pm3 in volume, a mitochondrion of about 0.06 pm3 in volume, and a chloroplast of about 1 pm3 in volume. Growth of mother cells takes place for up to 50 hours after the FIG.4. (a) Pp-nuclei isolated from N. tabacum cultured ceUs (line BY-2) observed by DAPI-fluorescence; (b) DAPI-fluorescenceplus phase-contrast; (c) phase-contrast; (d, e) and negative-staining electron microscopy. (f, g) Extracted bauquet of pp-nuclei (f) and circular ptDNA (g) observed by electron microscopy. (h) Circular gene map of the tobacco chloroplast genome and pt-nucleus (arrow). Inverted repeats, IRA and IRBare shown by bold lines. JLA. JKB and JsA and JSB are junctions between a large (LSC) and a small single-copy region (SSC). Genes shown outside the circle are on the A strand and are transcribed counterclockwise. Genes shown inside the circle are on the B strand and are transcribed clockwise. Asterisks indicate split genes. Major open reading frames (ORFs) are included. Bars in a, d, and f represent 10, 1 and 1 pm, respectively. [Gene map (h) (from Shinozaki et al., with minor changes, 1986; reproduced by permission of M. Sugiura.)]

12

TSUNEYOSHI KUROIWA

FIG. 5 . Epifluorescence photomicrographs of a cell nucleus (large arrows in a, c), a mt-nucleus (small arrows in a-c), a pt-nucleus (middle arrows in a, c), and a chloroplast (b) in a cell of the red alga Cyunidium cddurium RK-1 cell after staining with DAPI. (a) and (b) are the same field. (c) Diagram of the life cycle of the alga depicting a possible sequence of events based on the observations of cells fixed at various times (hours) after ISC. The chloroplast growth stage: the small, spherical chloroplast increases in volume and becomes a football-like structure. The stage at which formation of the plastid-dividing (PD) ring occurs: the somewhat electron-dense body (SEB) is fragmented into many small somewhat electron-dense granules (SEGs), which are aligned along the equatorial region of the chloroplasts and fine filaments are formed from the SEGs in the equatorial region of the chloroplasts. The fine filaments of the plastid-dividing ring align themselves according to the longest axis of their overall domain. Constriction stage: a bundle of fine filaments begins to contract and generates deep furrow. PD ring conversion stage: after chloroplast division, the remnants of the PD ring are converted into SEGs. Similar events occur during the second cycle of chloroplast division. The divisions of organelle nuclei occurjust before organellekinesis. Cp-S and CN-S indicate the chloroplast DNA-synthetic stage and the cell nuclear DNA synthetic stage, respectively (Mita and Kuroiwa, 1988; reproduced by permission of Springer-Verlag.)

PLASTIDS AND ORGANELLE NUCLEI

13

initiation of a synchronous culture (ISC) and these growing cells are mostly four-endospore cells after the second endospore divisions. The divisions of the chloroplast, the cell nucleus, the mitochondrion, and the cell itself occur in that order. The cp-nucleus is located in the central area of the chloroplast; it increases in volume with the growth of the chloroplast and it divides just before chloroplast division. The levels of cpDNA in the chloroplast increase soon after ISC and reach four times the initial value before the first division of the chloroplast. The amount of DNA in the chloroplast decreases stepwise after each endospore division, while the DNA in the cell nucleus is duplicated during each cycle of endospore division (Fig. 5). Division of the cp-nucleus as well as of the mt-nucleus precedes organellekinesis. The duration of the period of cpDNA synthesis appears to be about 8 hours and the DNA synthesis occurs between 6 and 22 hours after ISC. Similar events have been observed in the chloroplast of another red alga, namely, C. caldarium M-8of the CL type (Kuroiwa et al., 1989c), in the proplastids in tobacco-cultured cells (Yasuda et al., 1988), and in the leucoplasts in Allium cepa (Nishibayashi and Kuroiwa, 1982).Therefore, we consider that, compared with the division cycle of the cell nucleus, there are two distinguishing features of plastid division: one is an endoduplication of DNA and a greater than 2-fold increase in volume during a single plastid division cycle; the other is the presence of a division cycle without the synthesis of DNA or increase in volume. Stepwise reduction in the volume and DNA content of plastids by division has been observed in the plastids of spermatocytes, during spermatogenesis,in the green alga Bryopsis maxima (Kuroiwa and Hori, 1986),in Charu australis (Sun et ul., 1988), and in the fern Pteris uittata (Figs. 17 and 18; H. Kuroiwa et al., 1988). In spite of differences in patterns of timing of the synthesis of DNA during the plastid division cycle, pt-chromosomes are basically divided equally among daughter plastids during the division of plastids. However, the molecular mechanism for separation of pt-chromosomes is unclear. An intensive search has been made for the mechanism of segregation of pt-chromosomes since Jacob et al., (1963) first proposed a hypothetical mechanism for the segregation of bacterial chromosomes, namely, that the chromosome is attached to the cell membrane in the region of the replicating fork; indeed, in some species, there is even evidence for additional attachment at the point of origin of replication. The chromosomes may be aided in their separation by the growth of a membrane between these two points of attachment. Rose and Possingham (1976)and Rose (1988) emphasized the role of the association between DNA and membranes in the segregation of pt-chromosomes. Plastid chromosomes are attached to the thylakoid membrane and, thus, can be separated equally into daughter plastids by an elongation of the binding sites that is accompanied by

14

TSUNEYOSHI KUROIWA

growth of the thylakoid membrane system. Rose (1988) proposed an interesting model in which segregation of pt-chromosomes is related to thylakoid membrane systems in dividing chloroplasts, etioplast. and proplastids with various types of cp-nucleus, for example, the SN and PS types. However, there is no direct evidence showing such a specific association for the segregation. Kawano and Kuroiwa (1985) isolated a membrane-DNA complex by centrifugation of a sheared lysate of isolated mitochondria from Physarum polycephalum. Analyses by Hoechst 33258/CsC1 density-gradient centrifugation and restriction-endonuclease treatment of the complex showed that DNA in the complex was richer in A-T base pairs than the total mtDNA, and contained the specific EcoRI fragment (E-8), which was localized on the right-hand side of the Physarum mt-genome. The sequence of the 1 kbp from the E-8 showed that the region contains an attachment site for topoisomerase, inverted repeats, stem loops, and tandem repeats (S. Kawano, personal communication). In an attempt to examine whether or not specific regions of cpDNA are involved in interaction with spinach thylakoids by use of restriction endonucleases (Lindbeck and Rose, 1987), cpDNAs from vesicles and chloroplasts were found to have similar restriction patterns and all cpDNA sequences were represented in the vesicle-bound cpDNA. This result suggests that all cpDNA sequences are represented in the vesicle-bound cpDNA. The simplest explanation for these data is that random DNA sequences are responsible for the DNA-membrane interactions. By contrast, Nemoto et al. (1991) recently obtained evidence that specific sites on the cp-chromosomes in two species of N. tabacum were bound to a complex that consists of thylakoid membrane and proteins. When isolated, intact cp-nuclei were digested by the restriction enzyme, EcoRI. The pattern of the restriction fragments was different from that obtained by direct digestion of the purified DNA: a few fragments containing the specific EcoRI fragment designated E-2 were preferentially deleted (Fig. 6a). The result suggests that, when specific regions were associated to form a complex which was composed of the membrane system and specific proteins, they were not digested by the restriction enzyme. This hypothesis was confirmed by the following experiments. When the complexes were treated with SDS or proteinase K, their restriction patterns were similar to those of total purified DNA. Nemoto et al. (1991) showed that at least four regions (the region between IRF 168 and rpo B in LSC; the region between rps 16andpsb A in LSC; and the regions between rps 7 and 23 S in inverted repeats) in the circular genome (Figs. 4h and 6b) bind to membranes. The region containing psa A is stronger in binding capacity than the other regions. However, when intact pp-nuclei isolated from tobacco-cultured cells were treated with restriction enzymes, their frag-

PLASTIDS AND ORGANELLE NUCLEI

15

FIG. 6. Analysis of DNAs by agarose gel electrophoresis after cpDNA, cp-nuclei, and pp-nuclei were digested with EcoRI. Lane M, Hind111 fragments of lamda phage DNA as molecular-size markers; Lane I , cpDNA; lane 2, cp-nuclei; lane 3, cp-nuclei treated with SDSlproteinase K;lane 4, cp-nuclei treated with SDS; lane 5 , cp-nuclei treated with proteinase K; lane 6, cp-nuclei treated with RNase A; lane 7,cp-nuclei; lane 8, pp-nuclei treated with SDS/proteinase K; lane 9, pp-nuclei; lane 10, pp-nuclei treated with SDS/proteinase K. Cp-nuclei were isolated from N. rabacum (line BY-Z)(lanes 2-6) or from Xanthi Nc (lanes 8,9). Pp-nuclei were isolated from N. tabacum cultured cells (line BY-2). Arrowhead shows the position E-2 of the restriction fragment that appears to bind tightly to protein in cp-nuclei but not in pp-nuclei. (From Nemoto et al., 1990; reproduced by permission of the Japanese Society of Plant Physiologists.)

ment patterns were similar to those of purified DNA (Fig. 6). There is, as yet, no direct evidence for an association between a specific part of the pt-chromosome and a membrane system that is a prerequisite for the separation of pt-chromosomes. If there is such an association, the weak binding at one or all of three binding regions may be related to the separation of pt-chromosomes.

16

TSUNEYOSHI KUROIWA

B. PLASTIDKINESIS BY THE PLASTID-DIVIDING RING The mitochondria of P. polycephalum contain electron-dense, rodshaped mt-nuclei, and they are therefore particularly suitable for investigations of mitochondrial division (Kuroiwa et al., 1977). When a small explant of a plasmodium was incubated in a solution that contained cytochalasin B, a large number of mitochondria did not exhibit the dumbbell shape but developed a large spherical or ovoid configuration (Kuroiwa and Kuroiwa, 1980). It is likely that cytochalasin B disrupted the function of actin such that mitochondria failed to form dumbbells. However, we have been unable, as yet, to discern any fine structures that are intimately related to mitochondriokinesis. Since plastids are considerably larger than mitochondria and contain more highly developed membrane systems, a bulkier apparatus may be required for plastidkinesis than is required for mitochondriokinesis. In the plastids of land plants, it is difficult to clarify whether or not the dumbbellshaped plastids in higher plants are dividing plastids, because dumbbellshaped plastids are not always dividing plastids in living cells (Whatley, 1980). Thus, it is difficult to observe the behavior of the plastid-dividing (PD) ring during the division of plastids in multiplastidic land plants. It has proved advantageous for analyses of the apparatus involved in chloroplast division to use synchronous cultures of the monoplastidic cells of C. caldariurn RK-1. Mita et al., (1986) were the first to identify PD ring located in the cytoplasm of the alga. Synchronous cultures of C. caldarium RK-1 can be initiated from young cells. The number of cells increases stepwise after ISC and finally reaches a value of about 4 times the initial number. Growth of mother cells takes place up to 50 hours after ISC and these growing cells are mostly fourendospore cells after the second endospore divisions (Fig. 5 ) . In control preparations, when the mother cells are fixed at 36 hours after ISC and excited with UV after staining with DAPI, four spherical cell nuclei and four irregularly shaped ct-nuclei, emitting blue-white fluorescence, can be seen (Fig. 7a). When the same field of such cells is excited with green light, four chloroplasts, emitting red autofluorescence, can clearly be seen in the areas where the ct-nuclei are located (Fig. 7a). When cells are exposed throughout two sequential endospore divisions to cremart, an inhibitor of the assembly of tubulin, each mother cell can be seen to contain one cell nucleus and four chloroplasts (Fig. 7c, d). By contrast, when the cells are exposed throughout two sequential endospore divisions to cytochalasin B, an inhibitor of the polymerization of actin into filaments, each mother cell can be seen to contain four cell nuclei and one large chloroplast (Fig. 7e, f). These results suggest that microtubules are not involved in the division of the chloroplast but are involved in the division of the cell nucleus, while

PLASTIDS AND ORGANELLE NUCLEI

17

FIG. 7. Epifluorescence photomicrographs illustrating cell nuclei (CN 1-4) and the chloroplast (CP) in mother cells C. caldarium RK-1 fixed immediately after incubation for 30 hours without any inhibitor (a, b); and after incubation for 30 hours with cremart at 10 pg/ml (c,d). Fluorescence photomicrographs with UV light (a, c, e) and green tight (b, d, f ) were taken in the same field for each treatment. In the control (a,b), one mother cell is composed of four discrete endospore cells, each of which contains a cell nucleus and a chloroplast. When cells are treated with cremart through two sequential endospore divisions, the mother cell contains one cell nucleus and four chloroplasts (c, d). By contrast, when the cells are treated with cytochalasin B, the mother cell contains four cell nuclei and one large chloroplast (e, f). Bar = 1 pm. (Mita and Kuroiwa, 1988; reproduced by permission of Springer-Verlag.)

18

TSUNEYOSHI KUROIWA

polymerized actin filaments are not involved in division of the cell nucleus but play an important role in the division of chloroplasts and in cytokinesis. When cells of C. caldarium are stained with rhodamin-conjugated phalloidin, a fluorescent dye that is specific for actin, a ring that emits orangecolored fluorescence appears faintly in the equatorial region of the dividing chloroplast. However, such a result does not exclude the possibility that the ring may correspond to the cytoplasmic contractile ring involved in cytokinesis. In synchronized cells of C. caldarium RK-1, it is possible to observe plastidkinetic events and fine structures in detail, from the formation of the PD ring during the early stages to the disapperance of the PD ring during the late stage, as summarized in Fig. 5. When the cell, the cell nucleus, and the chloroplast increase in volume about 3, 2, and 3.5-fold, respectively, the shape of the chloroplast changes from a spherule to a football-like structure and concentric, circular, thylakoid membranes in the chloroplast begin to separate into two parts. At that time, many somewhat electrondense granules (SEG), each 40-90 nm in diameter, and electron-dense deposits appear in the cytoplasm close to the outer envelope membrane and begin to be distributed at the equatorial region of chloroplast (Mita and Kuroiwa, 1988). The PD ring is made up of SEG. A portion of the PD ring can be seen as a bar, about 60 nm in width, at the edge of the PD ring. The bar consists of fine filaments, each about 5 nm in diameter, which are aligned parallel to the longitudinal direction of the bar. The arrangement indicates that the PD ring is a bundle of fine filaments. By the time the PD ring starts to contract, the small SEGs have completely disappeared. When sequential thin sections are cut through the constricted isthmus of a dividing chloroplast of the alga during the middle phase of division, it appears that the electron-dense deposits at the bridge between daughter chloroplasts are distributed as a close ring or beltlike structure, 60 nm wide and 50 nm thick, lying in close apposition to the outside of the outer envelope of the chloroplast. At the final stage, the width of the deposits that make up the constricted PD ring appears to be somewhat greater than that of the deposits at the early and middle stages of chloroplast division (Fig. 8a,b). However, the width of the PD ring does not deviate very much from the cited value of 60 nm, regardless of the stage of division or the steepness of the walls of the furrow between the daughter chloroplasts, nor does it vary much among chloroplasts of vastly different volumes. At higher magnification, the cross sections of the PD ring clearly reveal that the PD ring is located on the cytoplasmic side of the outer envelope (Fig. 8b,c). Inside the inner envelope, some electron-dense deposits can also be seen, but their width and thickness do not change from the early stage of

FIG.8. Electron micrographs of sections cut in directions perpendicular to (a, b) and parallel to (c), the plane of division at the constricted isthmus of dividing chloroplast in C. culdurium RK-I. When the chloroplast is progressively pinched, the PD ring becomes more electron dense and is seen to increase in thickness (arrow in a). At higher magnification, the PD rings can be seen on the cytoplasmic side of the outer envelope membrane (arrows in b and c), while the inner electron-dense belt (arrowheads in b and c) does not change from what is observed at the early phase of the division. Bars = 0.1 pm. (Mita and Kuroiwa, 1988; reproduced by permission of Springer-Verlag.)

20

TSUNEYOSHI KUROIWA

chloroplast division until it is complete. Similar events can be seen in the second division of the chloroplasts. The PD ring, when visualized in sections cut in a direction parallel to the plane of division, is clearly made up of a circular belt that appears to be composed of tightly packed, fine filaments. When the chloroplast has ceased dividing, a centriolelikeplaque with microtubules develops outside the cell nucleus and electron-dense deposits, which consist of actinlike filaments and are related to cytokinesis, appear beneath the cell membrane. Mita and Kuroiwa (1988) proposed that the main components of the PD ring are actinlike filaments on the basis of the following pieces of evidence: (i) Cytochalasin B inhibits the division of chloroplasts without inhibiting division of the cell nucleus; (ii) the ringlike structure around the chloroplasts can be stained with rhodamin-conjugated phalloidin; and (iii) fine filaments, observed in the PD ring, are very similar in diameter to actin filaments. However, such a proposal is not supported by the results of immunogold-staining experiments with actin-specific antibodies: a few gold particles, which showed the localization of actin, were found on the SEGs and none were found on the PD ring (Mita and Kuroiwa, 1988). A cytoplasmic PD ring, like that in C . caldarium, was also observed around the plastids of a green alga that was the green alga Trebouxia (Senda and Ueda, IW),in Pyrarninornonas uirginica (T. Hori, personal communication), and in the moss Funaria hygrometrica during the division of plastids, but no inner matrix ring was observed in any of these cases. In the green alga P . uirginica, the division of the chloroplast with a large pyrenoid occurs after the cell nucleus division and immediately before cytokinesis (Fig. 9a, b, d). The PD ring appears to generate the contractile force that is involved in the division of the chloroplast with a pyrenoid (Fig. 9). In the chloronema and caulonema of the moss F. hygrometrica, Tewinkel and Volkmann (1987) observed a distinct filamentous structure similar to the PD ring in the plane of division outside the plastids, but close to the envelope, in three-dimensional reconstructions prepared from electron micrographs. The PD ring was also visible around the narrow isthmus of dividing chloroplasts and amylopiasts during the late phase of plastid division. The cross-sectioned filamentous structures were 10-40 nm in width and 10-15 nm thick and ran parallel to the outer envelope at a distance of about 10 nm. In higher plants, Suzuki and Ueda (1975) and Luck and Jordan (1980) reported the appearance of electron-dense material, which they considered to be evidence of “buffles” and a “septum,” respectively, at the constricted isthmus between daughter proplastids and daughter amyloplasts in Pisum satiuum and Hyacinthiodes nonscripta. Similar electrondense deposits have been observed at the narrow neck of dumbbell-shaped

PLASTIDS AND ORGANELLE NUCLEI

21

FIG. 9. Electron micrographs of sections cut in directions perpendicular to the plane of division at the constricted isthmus of a dividing chloroplast in the green alga Pyraminomonas virginica (a, b) and in A. satiuum (c). d and e show two models of a single PD ring of the C . caldarium type ( arrow in d) and a PD ring doublet with an outer ring (large arrow) and an inner ring (small arrow) of the A. satiuurn type (e), respectively. Bars = 0.1 p n . [Photographs (a, b) courtesy of Dr. T. Hori, Tsukuba University, Japan; (c) reproduced by permission of H. Hashimoto and Springer-Verlag.]

chloroplasts and have been described as "fuzzy plaques" in Triticum aestivum, Atriplex semibaccata, and Sesamum indicum var. glauca (Leech et al., 1981). Fuzzy plaques of electron-opaque material were frequently, but not always, seen covering or displacing the membranes of the isthmus (Leech et al., 1981). Such early observations were made from individual thin sections. Subsequently, Chaly and Possingham (1981) sur-

22

TSUNEYOSHI KUROIWA

veyed the deposits located at the constricted isthmus between daughter chloroplasts in various plants, such as P . sativum, Phaseolus vulgaris, Lycopersicon esculentum, Lactuca sative, Citrullus lanatus, Hordeum vulgare, Z . mays, T. aestivum, and Pisum radiata, using serial sections and they concluded that electron-dense deposits, which were parts of an annulus, were located in the interspace between the outer and inner envelope membranes at the constricted isthmus of dumbbell-shaped proplastids of these plants. To explain the formation of the “fuzzy plaque” or “annulus,” Leech et al. (1981) and Possingham and Lawrence (1983) applied to plastids the theoretical model of cell division proposed by Greenspan (1977), namely, that the internal fluid flow generated by surface changes leads to a concentration of material in the equatorial region and to formation of an annulus. They suggested that the electron-opaque material might be present, but diffuse at earlier stages of division and only become visible when sufficiently concentrated within a narrow constriction. However, in earlier observations of higher plants, there is no definite evidence to indicate that dumbbell-shaped plastids are dividing plastids. By contrast, Hashimoto (1986)observed the presence of an electron-dense double ring structure (PD ring doublet) around the constricted isthmus of dumbbell-shaped plastids of Avena sativum, using a serial thin-sectioning technique. The inner and outer rings of the doublet were reported to coat the inside (stromal side) of the inner envelope membrane and the outside (cytoplasmic side), respectively (Fig. 9c, e). There are discrepancies in the interpretations of the localization of the PD ring at the constricted isthmus, as observed in different species and in the same species of higher plants. Therefore, Kuroiwa (1989a) examined the localization of the PD ring of higher plants using the serial thinsectioning technique. The localization of the PD ring in proplastids of N . tabacum was found to be similar to that of the PD ring doublet reported by Hashimoto (1986). A similar PD ring doublet was reported in spinach, bean, tobacco, and wheat by Oross and Possingham (1989). T. Hori (personal communication, 1990) observed a PD ring in plastids of the gymnosperm Ginkgo biloba. It is likely that the cytoplasmic PD ring is to be found in plastids of red, brown, and green algae, in mosses and ferns, and in gymnosperm, while the clear PD ring doublet is to be found in all angiosperms if a careful search is made. The PD ring doublet has also been seen in a mutant deficient of the monocot that is deficient in plastid ribosomes (Hashimoto and Possinham, 1989). The observation does not conflict with the hypothesis that the PD ring consists of an actinlike protein, which must be encoded in the cell nuclear chromosomes. The plants in which electrondense deposits, an annulus, or a PD ring have been observed to date are summarized in Table I.

TABLE 1 OBSERVATIONS OF ELECTRON-DENSE MATERIALSAT CONSTRICTED ISTHMUSOF DUMBBELL-SHAPED PLASTIDS AND PLASTID-DIVIDING RINGSIN VARIOUS PREPARATIONS OF PLANT MATERIAL ~~

Species

Red algae Cyanidium caldarium RK-1

Tissue

Plastid

Stage

~

Name or type

~~

Distribution

~

~~

Ref.

Single cell

Chloroplast

Early Middle Late

PDring

Cytoplasm

Mita ef al. (1986) Mita and Kuroiwa (1988) Kuroiwa (1989a)

Trebouxia sp.

Single cell

Chloroplast

PDring

Cytoplasm

Senda and Ueda (1990)

Pyrarninomonas virginica

Single cell

Chloroplast

Middle Late Middle Late

PD ring

Cytoplasm

T. Hori (1989)"

Moss Funaria hygrometrica

Protonema

Chloroplast, amyloplast

Late

PD ring

Cytoplasm

Tewinkel and Volkmann (1987)

Proplastid Proplastid, chloroplast Proplastid Chloroplasts Proplastid Chloroplasts

Late Late Late Late

PDring PD ring doublet PD ring doublet PD ring doublet

Cytoplasm Cytoplasm, matrix Cytoplasm Cytoplasm

Phaseolus vulgaris

Sperm First leaf Cultured cell Leaf Root Leaf

Late

PD ring doublet

Cytoplasm

Nicotiana tabacum

Leaf

Chloroplasts

Late

PD ring doublet

Cytoplasm

Triticum aestivum

Leaf

Chloroplasts

Late

PD ring doublet

Cytoplasm

T. Hori (1990)" Hashimoto (1986) Kuroiwa (1989a) Oross and Possingham (1989) Oross and Possingham ( 1989) Oross and Possingham (1989) Oross and Possingham ( 1989)

Green algae

Higher plats Ginkgo biloba Avena sativa Nicotiana tabacum Spinacia oleracea

(continued)

TABLE I Continued Species

Tissue

Plastid

Stage

Name or type

Distribution

Ref. Suzuki and Ueda (1975) Luck and Jordan (1980) Leech et al. (1981) Chaly and Possingham (1981) Chaly and Possingham (1981) Chaly and Possingham (1981) Chaly and Possingham (1981) Chaly and Possingham (1981) Chdy and Possingham (1981) Chaly and Possingham (1981) Chaly and Possingham (1981) Chaly and Possingham (1981) Chaly and Possingham (1981)

Pisum sativum Hyacinthiodes non-scripta Triticum aestivum Sesamum indicum

Root tip Pollen Root tip Root tip

Proplastid amyloplast Plastid (proplastid) Proplastid Proplastid

Late Late Late Late

Septum Septum Fuzzy plaque Annulus

Plastids Plastids Plastids Plastids (interspace)

Pisum sativum

Root tip

Proplastid

Late

Annulus

Plastids

Phaseolus vulgaris

Root tip

Proplastid

Late

Annulus

Plastids

Lycopersicon esculentum

Root tip

Proplastid

Late

Annulus

Plastids

Lactuca saliva

Root tip

Proplastid

Late

Annulus

Plastids

Citrullus lanatus

Root tip

Proplastid

Late

Annulus

Plastids

Hordeum vulgare

Root tip

Proplastid

Late

Annulus

Plastids

Zea mays

Root tip

Proplastid

Late

Annulus

Plastids

Triticum aestivum

Root tip

Proplastid

Late

Annulus

Plastids

Pinus radiata

Root tip

Proplastid

Late

Annulus

Plastids

a

Personal communication.

PLASTIDS AND ORGANELLE NUCLEI

25

When does the cell nucleus generate the information that induces the division of plastids? Kamata et al. (1989) caused matured mesophyll protoplasts of tobacco to fuse with protoplasts from cultured cells by electric fusion. When the fusion products were cultured for 2 days, the division of chloroplasts was observed in the heterokaryocytes, while such division of chloroplasts was not observed when mesophyll protoplasts alone were cultured under the same conditions. Since the matured mesophyll cells never divide in the leaf (whereas the cultured cells multiply regularly), these results suggest that proliferating cells may synthesize an unknown substance that induces the division of chloroplasts. Many cytologists accept the hypothesis that plastidkinesis occurs by partition of the inner limiting membrane (Modrusan and Wrischer, 1990). However, more data must be gathered before this mode of division is confirmed. (i) Each step of plastid division has not yet been studied by the serial thin-sectioning technique. (ii) The complete step-by-stepdivision of a plastid, accompanied by division of pt-nuclei by partition, has never been observed. (iii) A quantitative analysis of the number of plastids before and after the division of plastids by partition has not yet been made in detail. In cells of algae and higher plants, the majority of plastids divide by constriction, but about 1% or even fewer of the total dumbbell-shaped plastids examined appear to have been partitioned at the center by the inner limiting membrane. However, three-dimensional reconstructions created from serial sections have indicated that, in alI the partitioned plastids examined, the two daughter progeny were actually still connected to each other by a channel of plastid matrix. Then the invagination of the inner limiting membrane was cut in a direction perpendicular to the invagination, an image of the partition was seen at the equatorial region of the dumbbell-shaped plastids. As the invagination of inner limiting membrane often occurs in a direction perpendicular to a cross section of the stack of grma and thylakoid membranes in the chloroplast, the direction of invagination can be judged easily by reference to the morphological characteristics of the grana.

V, Differentiation of Plastids ‘The electron microscopic approach to differentiation of plastids was initiated in the late 1950s. Miihlethaler and Frey-Wyssling (1959) suggested that proplastids, etioplasts, and chloroplasts arise from smaller organelles termed “plastid initials” (0.002-0.05 pm in diameter) which are present in leaf buds and leaf meristems of a number of higher plants. With new and improved methods for fixation for electron microscopy, it is now possible

26

TSUNEYOSHI KUROIWA

to distinguish between most of the organelles in meristematic tissue and there have been no recent observations of “proplastid initials” (Possingham and Lawrence, 1983). With the advances in DAPI staining and epifluorescence microscopy, the number of copies of ptDNA per plastid is easily counted. Such studies have shown that there are variations among proplastids, which can be conveniently classified into at least two types (Fig. 2; Kuroiwa et ul., 1981). One type of proplastid is characterized by the presence of one to two copies of the plastid genome per plastid of about 1 p m in diameter, and by the presence of a single pt-nucleus per plastid. The other type is characterized by several copies of the pt-genome per plastid of about 2-3 p m in diameter and about 2-5 pt-nuclei per plastid. Herein, the latter will be referred to as “proplastids” (pp) to distinguish them from the morphologically less complex and smaller “pp-precursors” characteristic of the former type. Often these two types of proplastids were mixed in one cell as seen in tobacco culture cells (Yasuda et a/., 1988). Typical morphological changes in pt-nuclei during the development of chloroplasts of monocots and dicots are shown in Figs. I and 2. The pp-precursor in higher plants contains only one small, spherical pt-nucleus and can divide according to the pp-division cycle 1 (Fig. 2). When the pp-precursors develop into proplastids, and then into etioplasts, if growth takes place in the dark the pt-nucleus becomes cup shaped with concomitant endoduplication of ptDNA, and is often found to be associated with starch grains or the prolamellar body (Kuroiwa et al., 1981). Since the association between the pt-nucleus and starch grains occurs commonly in various plastids, such as etioplasts, chloroplasts, and amyloplasts, it must play an important role in an as yet unknown way. The proplastids and etioplasts can also divide according to the proplastid division cycle 2 and the etioplast division cycle, respectively. In etioplasts of monocots, the pt-nucleus becomes a ring-shaped structure (Sellden and Leech, 1981; Hashimoto, 1985; Miyamura et al., 1986). Once etioplasts have been illuminated, the pt-nuclei begin dividing into 20-30 small, spherical ptnuclei, which are distributed individually into mature chloroplasts. During greening of pea leaves, the synthesis of proteins in the plastids may be a prerequisite for the dispersion of pt-nuclei into an entire plastid since such dispersion does not occur after addition of the inhibitors chloramphenicol and lincomycin (Fig. 13; Sasaki and Kuroiwa, 1989). The changes in numbers of copies of ptDNA have been examined by biochemical techniques and cytochemical fluorimetry. In Brussica, Kuroiwa et af. (1981) showed qualitatively, by staining with DAPI, that the DNA content per plastid increased markedly during the division cycle of proplastids and during the development from proplastids to etioplasts, but

PLASTIDS AND ORGANELLE NUCLEI

27

the DNA content increased only slightly after illumination, even though the cp-nucleus divided into small, scattered cp-nuclei. Miyamura et al. (19:36) examined the fluorescence intensity of each cp-nucleus by use of VIlvlPICS and showed clearly that the number of copies of ptDNA increased approximately 8-fold during the division cycle of plastids and during the differentiation of the proplastid to the etioplast; the number of copies reached a plateau when the chloroplasts began to engage in photosynthesis. As a result of the division of young chloroplastsaccording to the chloroplast division cycle, a 10-fold increase in the number of chloroplasts per cell occurs in spinach leaves and a 2- to 3-fold increase occurs in wheat leaves. This process is, therefore, important in determining the photosynthetic potential of the mature leaf(Leech, 1976).Plastids of both epidermal and palisade cells of P.vulgaris also divide at all stages of plastid development, but division ceases soon after the plastids become mature (Whatley, 19110).The matured chlorophyll-containingchloroplastsof red, brown, and green algae and the green leaves of land plants exist in a specialized form for photosynthesis and the fixation of carbon, using the energy of the sun, and synthesize amino acids as precursors of various proteins. However, when &hedisks of young green leaves of spinach are dissected and grown in sterile culture, the division of chloroplasts continues and the number of chloroplasts per cell increases 5- to 10-fold over a 7-day culture period in the light. There is virtually no cell division during this time (Rose er a/., 1974). Since the synthesis of cpDNA occurs in mutants that are deficient in chloroplast ribosomes (Hermann et al., 1974; Scott etal., 1982),in chloroplasts without substantial portions of the cp-genome (Day and Ellis, 1984) as well as in plastids that have ceased to synthesize proteins in the presence of inhibitors (Sasaki and Kuroiwa, 1989), the synthesis of cpDNA may not be related to the genome of chloroplasts themselves. By contrast, the synthesis of cpRNA is dependent on the genome of chloroplasts (Cozens and Walker, 1986; Shinozaki et al., 1986; Zaitlin et al., 19139; Hu and Bogorad, 1990).The mature chloroplasts of the CL, PS, and SP type can also develop from proplastids that contain only one ppnucleus throughout several divisions (Fig. 2). [n addition to differentiation from proplastids to chloroplasts (Figs. 1 an'd 2), proplastids are also able to differentiate into functional plastids of various forms, such as amyloplasts, elaioplasts, and chromoplasts, according to the differentiation of tissues. Small amyloplasts occupied by starch grains can be seen during the differentiation from proplastids to etioplasts in leaves of angiosperms, during embroygenesis, and during the formation of pollen grains, egg cells, central cells, and root caps. Specialized large amyloplasts are filled with large starch grains and can be

28

TSUNEYOSHI KUROIWA

observed in the cells of storage tissues, such as the cotyledons of dicots, endosperms of monocots, and tubers of potato. Their function is to accumulate starch as a reserve material. Elaioplasts, which are found in the epidermal cells of some monocots, are plastids that are largely filled with oil. Chromoplasts are carotenoid-containing plastids responsible for the colors of fruits and flower petals. As seen in the fruits of tomato and pepper, the chloroplasts are transformed into the chromoplasts during maturation of fruits. In general, the numbers of pt-nuclei and the DNA content per plastid are smaller in amyloplasts, chromoplasts, and elaioplasts than in chloroplasts. In addition to the cytological studies, some researchers have tried to elucidate the molecular mechanism of plastid differentiation.The proteins encoded by ptDNA are involved in transcription, translation, and photosynthesis. Therefore, the differentiation of chloroplasts requires selective activation of pt-gene expression and selective translation of pt-genes. The modes of chromoplast gene expression during the development of fruit and of gene expression in amyloplasts have been examined in comparisons with those operating in chloroplasts (Bathgate et a / . , 1986;Piechulla et al., 1986;Ngernprasirtsiri et al., 1988a,b). For example, Ngernprasirtsiri et al. ( 1988a) proposed that the amyloplast genome is mostly inactive with the exception or the gene for 16 S rRNA and p s b A, which are presumably regulated at the transcriptional level. Several workers (Sasaki and Kuroiwa, 1989;Sasaki et d.,1990) have also shown that the photogenespsb A, psb B, and rbc L are active in green tissues, such as leaves and stems, but are inactive in noncolored root tissue, while the housekeeping genes are active in many tissues including noncolored roots (Fig. 10a). Sasaki et al. (1990) showed that variations in the relative levels of transcripts of the photogenes in different organs were similar to the variations in gene dosage, but those in levels of the transcripts of ribosomal protein L2 were not. They proposed that gene dosage affects the organ-specific expression of photogenes. Baumgartner et al. (1989) showed that plastid transcriptional activity and numbers of copies of ptDNA increase early in chloroplast development, and they suggested that transcriptional activity per DNA template varied up to 5-fold during the biogenesis of barley leaf. Since the physical restriction map of cpDNA is basically similar to those of other differentiated plastids such as amyloplasts (Macherel et ai., 1985; Scott et al., 1984; Ngernprasirtsiri et al., 1988a),chromoplasts (Hansmann et al., 1985),and proplastids (Nemoto et al., 1988; Fig. lob), it is difficult to explain the mechanism of plastid differentiation at the structural levels of DNA molecules. The transcriptional and posttranscriptional regulations are considered to control protein synthesis in plastids (Gruissem et al., 1988). If there are at least two regulatory systems, the system based on

PLASTIDS AND ORGANELLE NUCLEI

29

FIG. 10. (a) Northern analyses of the total RNA in various tissues from P.satiuum plants. Three-microgram samples of total RNA were loaded on gels from which each autoradiogram was prepared. Sp, specific activity of the probe (10' cpm/pg); Ex, exposure time for autoradiography (hours); Se, seeds; L, leaves; S, stems; P, petals. 16 S rDNA, gene for 16 S ribosomal RNA; psb A, gene for the p700 apoprotein Al of photosystem I; psb B, gene for p700 apoprotein A2 of photosystem I; rbc L, gene for the large subunit of RuBisCO; psb E, gene for the 8-kDa subunit of cytochrome b 559; atp A, gene for a subunit of H'-ATPase subunit; rpl2, gene for the 50 ribosome subunit CL 12; pet A, gene for cytochromefin the cyiochrome blfcomplex; rbc S, gene for the small subunit of RuBisCO; Cab, gene for LHCP 11. (From Sasaki et al., 1990; reproduced by permission of the Japanese Society of Plant Physiologists); (b) Patterns after electrophoresis on agarose gels of DNAs from cp-nuclei (lanes C) and pp-nuclei (lanes P) isolated from N.tabacum BY-2 digested by three restriction enzymes, namely, EcoRI, HindIII, and BamH I. (From Nemoto et al., 1990; reproduced by permission of Japanese Society of Plant Physiologists.)

30

TSUNEYOSHI KUROIWA

transcriptional regulation must be more active than that based on posttranscriptional regulation during the early and middle period of chloroplast development, while the posttranscriptional regulation may be more important after maturation of chloroplasts, since cpDNAs are digested soon after maturation (Sodmergen et al., 1989; 1990). Two approaches have been developed to elucidate the molecular mechanisms of regulation at the level of transcription, as related to plastid differentiation. One is based on the physical changes in ptDNA itself and the other on changes in proteins that bind to ptDNA. A. METHYLATION OF PLASTID DNA

Ngernprasirtsiri et al. (1988b) analyzed methylation of ptDNA from fully ripened red fruits, green mature fruits, and green leaves of tomato. They found from Southen blots that no methylation could be detected of DNA fragments that contained certain genes that are actively expressed in chloroplasts, whereas DNA fragments of genes that are barely transcribed in chromoplasts were methylated. In addition Ngernprasirtsiri et al. ( 1988a)also examined gene expression of amyloplast DNA in the heterotrophically grown white cells of sycamore as compared with expression of the cpDNAs isolated from the green mutant cells. They demonstrated that the cp-genes in which methylation was not detectable were active as templates for transcription in uitro by RNA polymerase from Escherichia coli, but the methylated amyloplast genes were apparently inactive. They proposed that methylation of DNA is a likely mechanism for the regulation of expression of amyloplast DNA in sycamore cells. By contrast, it has been reported that both amyloplast and chloroplast genes in P.satiuum are methylated and chromoplast DNA during tomato fruit ripening (Marano and Carriilo, 1991) is not methylated (N. Ohta et al., unpublished data). We cannot yet explain the discrepancy between methylation of cytosines in ptDNA and gene expression in plastid. B. PLASTID DNA-BINDING PROTEINS In the early 1980s some workers began to consider the role of proteins that bind to cpDNA, and the concept of “organelle nuclei” (Kuroiwa, 1982) increased in importance. Based on an analogy with mt-nuclei (Kuroiwa, 1973; 1974; 1982)and the isolation of pt-nuclei, the possibility that the ptDNA is not naked but is organized in situ by some proteins to form a compact structure has been claimed by many investigators (Kuroiwa et al., 1981; Kuroiwa and Suzuki, 1981; Briat et al., 1982; Reiss and Link, 1985; Hansmann et al., 1985). If this is indeed the case, then intact pt-nuclei, and not ptDNA, must be

PLASTIDS AND ORGANELLE NUCLEI

31

isolated and their components must be characterized in terms of both structure and function. Hallick et al. (1976)and Briat et al. (1979)isolated a DNA-protein complex from chloroplasts of Euglena and spinach, respectively, and showed that they retained their transcriptional activity. Yoshida et al. (1978)isolated a looped and folded structure of cpDNA from spinach which remained even after a vigorous deproteinization. These results conflict with the later observation that the cp-nuclei isolated from several plants become swollen and are easily dispersed by deproteinization. Kuroiwa et al. (1981)found by DAPI-fluorescencemicroscopy that the cp-nuclei of Nitella become dispersed in situ upon treatment with proteinase K, a result which suggests that the cpDNAs are organized into cp-nuclei by some proteinaceous components. Furthermore, in an attempt to avoid contamination from spherical fragments of celI nuclei and mtnuclei, Kuroiwa and Suzuki (1981) succeeded in isolated ring-shaped cpnuclei, with their structures intact, from the brown alga E. indicus. Results of enzymatic treatment again indicated that the ring-shaped structure of the cp-nuclei was maintained by some proteins. However, none of these earlier studies included a biochemical analysis of any proteinaceous components associated with the ptDNAs. Briat et al. (1982) examined the transcriptionally active DNA-protein coniplexes from spinach chloroplasts by electron microscopy and also analyzed the acid-soluble polypeptides contained in the complexes. They showed that some basic and low-molecular-weight proteins were present in the preparation. Reiss and Link (1985) compared similar DNA-protein coniplexes from etioplasts and chloroplasts of Sinapis alba and showed that they contained common as well as unique polypeptides. Meanwhile, Hatismann et al. (1985) succeeded in isolating cp- and chromoplast nuclei in a more condensed form from leaves and coronae, respectively, of Narcissus pseudonarcissus. They demonstrated characteristic differences between the polypeptide patterns of the two types of pt-nucleus as well as several polypeptides common to both. However, these studies did not reveal which polypeptides actually bind to ptDNA. It is likely that isolation of pt-nuclei (with their compact structures intact) by using a mild detergent might solve the problem of contamination by the membrane frac:tion which is not solubilized under such conditions. Therefore, methods must be developed to distinguish the proteins that are directly involved in the organization of ptDNA from membrane proteins that are unrelated to the pt-nuclei. Proplastids are very suitable for the isolation of pt-nuclei from organelles. Nemoto et al. (1988) succeeded in identifying four polypeptides (69, 31,30, and 14 kDa) that were bound to ptDNA in the pt-nuclei by combining DAPI-fluorescence microscopy with biochemical techniques. For

32

TSUNEYOSHI KUROIWA

studies of the interactions between the ppDNA and the DNA-binding proteins in more detail, neutral salts seem to be the most useful among various agents, since they can disrupt the DNA-protein interaction so gently that their effects can be reversed by removal of salt by dialysis. The organization of cell nuclear chromatin has been studied intensively by reconstituting the structure of nucleosomes by dialysis of purified DNA with purified histones, which had previously been dissolved in 2 M NaCI, against decreasing concentrations of neutral salt (McGhee and Felsenfeld, 1980; Igo-Kemenes et al., 1982). Nemoto et a f .(1989) examined the effects of various concentrations of NaCl on the compact structure of pt-nuclei using DAPI-fluorescence microscopy and, at the same time, they investigated the behavior of the four proplastid DNA-binding proteins by SDSPAGE after centrifugal fractionation. They found that the pp-nuclei, after dispersion by 2 M NaCl, could be organized once again into compact structures similar to those of the original pp-nuclei by a reduction in the salt concentration. Furthermore, the four DNA-binding proteins from proplastids, even though they had been dissociated from ppDNA by the treatment with 2 M NaC1, were found to have reassociated with ppDNA in the reassembled pp-nuclei. These results strongly suggest that the DNAbinding proteins in proplastids play a critical role in the organization of the compact pp-nuclei. Recently, Nemoto et al. (1989, 1990) isolated cp-nuclei from the chloroplasts in N . tabacum (Fig. 11) and compared their DNAbinding proteins with those of proplastids (Fig. 12). Major polypeptides, which were present in pp-nuclei, were not found in cp-nuclei, a result that suggests that the population of DNA-binding proteins changes during the differentiation of plastids (Figs. 12 and 13). It seems feasible that certain special regions of the mt- and ptchromosomes are involved in interactions with the respective membrane systems. There must be more than three roles for membrane-bound ptDNA. In addition to a role in DNA replication, a role in the separation of daughter pt-chromosomes, and a role in RNA transcription, there must be a role for such interactions in the dispersion of the clumped ptchromosomes during greening (Figs. 2 and 13). It has been shown in chloroplasts of N . tabacum that binding of two types occurs: there are weak and strong associations between ptDNA and the membrane system (Figs. 6 and 13). The weak association between the regions that contain psb A or between genes for rRNA and the thylakoid system may be a prerequisite for the DNA synthesis and for the separation of ptchromosomes into daughter plastids. The strong association between the region that contains the psa A, psa B, psb C, and psb D genes of ptchromosomes and the membranous systems does not seem to occur in

FIG. 11. Three types of nuclei in mesophyll protoplasts of N. tabacum cv. Bright Yellow-2. (a) Living protoplasts and (b) fluorescence photomicrograph of a protoplast fixed with glutaraldehyde and stained with DAPI, showing a cell nucleus and many tiny fluorescent spots from cp-nuclei (large arrow) and mt-nuclei (small mow) in the cytoplasm; (c,d, e) Part of a protoplast, observed by DAPI-fluorescence (c), DAPI-fluorescence and phase-contrast (d), and phase-contrast (e). (c, d, e) are the same field. (f, g, h) Isolated chloroplasts containing cp-nuclei (arrow) observed by DAPI-fluorescence (f), DAFT-fluorescence and phase-contrast (g), and phase-contrast (h); (i-k) Isolated cp-nuclei (arrow) observed by DAPI-fluorescence (i), DAPI-fluorescence and phase-contrast (i),and phase-contrast (k). (i, :#, k) are the same fields. Bars = 10 wm. (From Nemoto et al., 1990; reproduced by permission of the Japanese Society of Plant Physiologists.)

34

TSUNEYOSHI KUROI WA

FIG. 12. SDS-PAGE patterns of polypeptides from the pallet (a) and supernatant (b) of isolated pp-nuclei (lane PN) and isolated cpnuclei (lane CN) treated with 2 M NaC1. Polypeptides for which molelcular masses are shown are DNA-binding proteins. The DNAbinding proteins (35.28. and 26 kDa) of cpDNA differ markedly in molecular mass and in their binding sites from those of ppDNA (69, 31, 30, and 14 m a ) . (From Nemoto et al., 1990; reproduced by permission of the Japanese Society of Plant Physiologists.)

proplastids, but occurs during the development of chloroplasts (Nemoto et al., 1991). Since the pt-nucleus is divided and distributed into many tiny spherical cp-nuclei throughout the entire chloroplast during greening, and since the transcriptional activity of the region which contains the photogenes related to the photosynthesis is preferentially higher in the chloroplasts than in other areas, the membranous complex must play an important role in the control of both the dispersion of the clumps of copies and pt-chromosomes and the expression of pt-genes (Fig. 13). It has proved possible to isolate pt-nuclei from a variety of plastids, such as proplastids, etioplasts, chloroplasts, amyloplasts, and chromoplasts, from various plants. Since the physical structure of ptDNA in proplastids does not differ from that of the DNA in the other differentiated plastids, the DNA-binding proteins found in the various types of differentiated

35

PLASTIDS AND ORGANELLE NUCLEI

\

Cell nucleus

/

proplastid

Etioplast

/ ,

Chloroplast

+

Dispersion of plastid chromosome ___)

Greening

; 1 I

',

Etioplast I

I I I

I

I

I

I

t

iPlastid chromosome

I

111

, ,

I 1 '

I

I

I

I

I I

I I

I I I

-tL

I

I

DNA-binding proteins 'Weak binding

'Strong binding@

Membrane system

FIG. 13. Schematic representationof the changes in the molecular architecture of plastid nuclei during the differentiation of plastidsfrom proplastids to chloroplasts. Signalsfrom the cell nucleus are prerequisite for the endoduplication and dispersionof pt-chromosomes.

plastid must be important contributors to the regulation of the transcription that is essential for the differentiation of plastids.

c. PREFERENTIAL D~GESTION OF PLASTlD NUCLEI PRIOR TO

LEAFSENESCENCE

In the mature chloroplasts of higher plants, small spherical cp-nuclei are dispersed throughout the entire chloroplast and the ptDNAs appear to be actively transcribed. It has been shown by biochemical analysis (Scott and Possingham, 1980) and conventional fluorimetry (Lawrence and Possingham, 1986) that the DNA content of chloroplasts in spinach increases markedly during the development of the chloroplasts but is reduced in mature leaves. A similar decrease in the number of copies of ptDNA during the latter phase of chloroplast development has been observed in several other plants (Boffy and Leech, 1982), barley (Baumgartner et al., 1989), and pea (Lamppa and Bendich, 1979; Lamppa et al., 1980). Possingham and Lawrence (1983) suggested that the decrease in number of copies of DNA per mature chloroplast was due to the division of chloro-

36

TSUNEYOSHI KUROIWA

plasts that was not accompanied by cpDNA synthesis. This suggestion does not conflict with the hypothesis that the DNA content of the chloroplasts of pea leaves (Lamppa er al., 1980) is inversely related to the number of chloroplasts per cell. As green leaves begin to turn yellow in Prunus persica, the cell nuclear DNA and the ptDNA in palisade cells are degraded completely (Nii et al., 1988). The levels of protein, DNA, and RNA per cell all decrease significantly during aging of the leaf. During the decrease in levels of DNA in senescing peach leaves, a nuclease that requires Zn2+for full activation is gradually activated, while the activity of a nuclease that requires Ca2+ does not change during leaf senescence (Nii et al., 1988). The mechanism for the conversion from an inactive state of the Zn2+-dependent nuclease to an active state at the time when digestion of cpDNA starts is of considerable interest. Sodmergen et al. (1989) found in the coleoptile of 0. sariue that when the grana and thylakoid membrane system had developed sufficiently in the mature chloroplast, the digestion of each ptDNA had already begun but the cell nucleus remained intact. Since the number of plastids per cell remained constant throughout the subsequent growth and aging of the coleoptile, the preferential reduction in the amount of cpDNA per plastid was not due to the division of plastids but might, perhaps, be associated directly with the aging of the cells of the coleoptile, which precedes senescence of the coleoptiles. Similar degradation of cpDNA occurred before the first and the second leaves of 0. sariua turned yellow (Sodmergen et al., 1990). It is likely that such degradation occurs in the third leaves and in the other leaves. Since the nuclease that requires Zn2+ is activated in the chloroplasts with concomitant degradation of cpDNA, the nuclease may be involved in the digestion of cpDNA in the coleoptile, and in the first and second leaves (Fig. 14).

VI. Cytoplasmic Inheritance of Plastids Uniparental and biparental patterns of transmission of cp-genes were found in studies of the non-Mendelian inheritance of leaf color in angiosperm plants by Correns (1909) and Baur (1909), respectively. Correns (1909) showed that the green variegated, and white patterns in leaves of Mirabilis jalapa could be inherited in a uniparental fashion while, at the same time, Baur (1909) found that the green and variegated patterns of leaves of Pefargoniurntonale were inherited in a biparental fashion. The literature on transmission of plastids has been summarized in a number of reviews (Connett, 1987; Gillham, 1978; Gillham et al., 1985; Hagemann, 1976, 1983; Kirk and Tilney-Bassett, 1978; Sears, 1980; Smith, 1988;

PLASTIDS AND ORGANELLE NUCLEI

37

FIG. 14. Electrophoretic patterns ofcoiE1 plasmid DNA digested by the Ca2+-(a), Mg2+(b) , Zn2+-(c), and Mn*+-(d) dependent nucleases in crude extracts of chloroplast preparations. The chloroplasts were isolated from leaf blades %hours (lane 2) and 144 hours (lane 3) after imbibition. Lane 1 shows the pattern of colEl plasmid DNA incubated without any chloroplast extract. Lane 4 shows the pattern of colEl plasmid DNA incubated with a crude extract of chloroplasts isolated from leaf blades 144 hours after imbibition but without any added metal ions. All the incubations were carried out at 37°C for 80 minutes. 0, opencircular; L, linear; and C, closed circular form of colEl plasmid DNA. (From Sodmergen et a / . 1991; reproduced by permission of Springer-Verlag.)

Tilney-Bassett, 1975; Whatley, 1982; Boffey and Lloyd, 1988) and, most recently, in “Transmission of Plastid Genes’’ by Gillham et a f . (1990). These reviews summarized biochemical, genetic, and electron microscopic approaches to this problem and their results. Mechanisms of nonMendelian inheritance of ptDNA in algae and higher plants have been explained in terms of the specific behavior of plastids, such as the exclusion of plastids from one parent in the gamete, destruction of plastids in one parent in the zygote, and plastid fusion in the zygote followed by destruction of ptDNA from one parent (Gillham et al., 1990). Eukaryotic organisms have been classified by differences in the sizes of female and male gametes into at least three different groups: isogamous, arusogamous, and oogamous species. When gametes are alike in appearance they are called isogametes, as in the case of Chfamydomonas, Ufothrix, and Ectocarpus, and the plants exhibit isogamy. Gametes having only a slight difference in size are said to be anisogametes and such species exhibit anisogamy. Bryopsis maxima is classified as a number of this group. When gametes differ in size and activity, as they do in algae such as Vdsluox,Fucus, and Polysiphonia, in mosses, ferns, and higher plants, the plants are said to exhibit oogamy. In higher plants, genetic analyses of mechanisms of maternal inheritance have proved possible, but biochemical analyses have been dimcult since double fertilization must be performed in cells embedded in various tissues. Many workers prefer to use

38

TSUNEYOSHI KUROIWA

microorganisms with a short life cycle in which genetic and biochemical analyses are relatively easy. Sager (1954) showed for the first time that uniparental inheritance of cytoplasmic genes for resistance to streptomycin occurred in C. reinhardtii. Since then, many studies on mechanisms of plastid inheritance at the genetic and biochemical level have been performed using C. reinhardtii (Gillham, 1974). A. ISOGAMOUS ALGAE The findings of uniparental inheritance and DNA in chloroplasts (Sager and Ishida, 1963) in the isogamous alga C. reinhardtii led to the important hypothesis that the mechanisms of maternal inheritance were based on enzymatic reactions, and maternal transmission of cp-genes was connected to the behavior of organelle DNA. Sager and Lane (1972) showed that cpDNA of female origin remained in zygotes for at least 6 hours after mating, during which time cpDNA from the male parent disappeared. This result suggested the existence of a mechanism of maternal inheritance based on the preferential degradation of male cpDNA in young zygotes. To explain the preferential degradation of male cpDNA, it has been proposed that the maternal transmission of cpDNA is governed by a methylation-restriction system analogous to that found in bacteria: after gametic fusion, the male cpDNA is degraded by a restriction enzyme while the modified female cpDNA remains intact. Modification is assumed to occur as a result of methylation. Studies using CsCl density gradients (Burton et af., 1979), high-pressure liquid chromatography (Royer and Sager, 1979), and restriction endonucleases (Sano et al., 1981)have shown that female cpDNAs are methylated but male cpDNAs are not. Furthermore, the isolation of DNA methyltransferases from C. reinhardtii with molecular weights of 60,OOOand 200,000 has been reported. Methylation of cpDNA from both parents has been reported in a mutant of C. reinhardtii in which the cpDNAs of both mating types are heavily methylated but maternal inheritance occurs as in the wild-type strain (Bolen et af., 1980). It was confirmed subsequently that the methylation of female cpDNA is at its lowest level during the vegetative stage and that this basal level of methylation of cpDNA increases more than 20-fold after gametogenesis. A large number of methyl groups are incorporated into cpDNA at the 7-h-zygote stage in C. reinhardtii. However, results of experiments using inhibitors of the methylation of DNA suggest that extensive methylation of female gametic cpDNA during gametogenesis is not required for the maternal inheritance of cp-genes (Feng and Chiang, 1984). The methylation at the 7-h-zygote stage is probably not related to the maternal inheritance of cpDNA because, in most of the zygotes, preferen-

PLASTIDS AND ORGANELLE NUCLEI

39

tial degradation of male cpDNA occurs within 60 minutes after mating (Kuroiwa et al., 1982). The methylation-restriction hypothesis is associated with some problems which must be solved. Cp-nuclei can be easily observed in DAPI-stained cells. Therefore, if male cpDNAs in a zygote of C. reinhardtii are preferentially degraded within 6 hours after mating, as suggested by the results of biochemical experiments (Sager and Lane, 1972), the disappearance of cp-nuclei of male origin should be observable by epifluorescence microscopy during formation of zygotes. The accurate timing and morphological pattern of the preferential destruction of cp-nuclei in each zygote must also be identified in relation to biochemical reactions, such as methylation of cpDNA and activation of as yet unidentified nucleases. Ten minutes after the mating of gametes of two wild-type strains, the newly formed zygotes are spherical, have four flagella, and contain two cell nuclei and two discrete chloroplasts. Each chloroplast contains 8-10 dispersed cp nuclei. About 40 minutes after mating, the cp-nuclei in the chloroplast from one parent disappear completely, prior to cell nuclear fusion (Fig. 15). Various crosses were arranged between morphologically different gametes by using cells that differed in size, in the size of ct-nuclei, in the length of the flagellum, and in chlorophyll content (Kuroiwa et al., 1982; Tsubo and Matsuda, 1984; Kuroiwa, 1985) and in BUdR label (Munaul. et al., 1990). The cp-nuclei that disappeared were confirmed to be of male origin. Since similar uniparental degradation of cp-nuclei has been observed in C. moewusii (Coleman and Maguire, 1983) and in multicellular green algae such as Monostroma nitidum, Dictiosphaeria calvernosa, and Acetabularia calyculus (Kuroiwa et al., 1985a),preferential degradation of organelle DNA of uniparental origin may occur in many isogamous algae, such as Caulerpa brachypus and C. okamurai (Kuroiwa and Hori, 1990). On the other hand, in a few green algae, such as Ulva arasaki and Enteromorpha intestinalis, the preferential digestion of organelle DNA was not observed 10 hours after mating (Kuroiwa and Hori, 1990). In each case, therefore, it is essential to examine whether or not the uniparental transmission of cpDNA occurs. I3oynton et al. (1990) and Hams (1989) pointed out the possibility that the: disappearance of male cp-nuclei may be due to preferential dispersion of cp-nuclei throughout the entire chloroplast and, thus, cpDNA of male origin may still be present. However, this possibility has been completely excluded. The total population of cp-nuclei contained 80-100 copies of cpDNA and occupied more than 10% of the matrix, exluding starch grains anld thylakoid membranes. Therefore, if the total cp-nuclei had been dispersed throughout the entire matrix of chloroplast of male origin, the chloroplasts that contained the dispersed DNAs would be visualized as

40

TSUNEYOSHI KUROIWA

Time ( h ) a f t e r mating 0

2

1 1

Control

mt

.

a

Chloropla8t nuclei

FIG. 1.5. Diagrammatic summary of representativeevents of cell nuclei(N)and chloroplast nuclei in young zygotes during the first 2 hours (a) and epifluorescencephotomicrographs (b, c ) of zygotes 10 (b) and 40 minutes (c) after mixing of mt' (male) and mt+ (females)gametes of C. reinhardtii. cn and p show ceU nucleus and pt-nucleus. respectively. Bar = 1 pm.

bright areas by high-resolution epifluorescence microscopy and the fluorescence intensity would be easily measurable by VIMPICS. VIMPICS has the ability to identify one gene in uifro and one molecule of ptDNA. In fact, the fluorescence intensity in the matrix of chloroplasts of female origin did not change either before or after the disappearance of cp-nuclei of male origin (Kuroiwa and Nakamura, 1986). Immunofluorescence patterns obtained with DNA-specific antibodies corresponded to the pattern of staining by DAPI (Sun et al., 1988). Addition of inhibitors of nucleases soon after mating completely inhibited the preferential destruction of cpnuclei of male origin. These results strongly suggest that the disappearance of cp-nuclei is due to nucleolytic reactions. Therefore, DAPI staining

PLASTIDS AND ORGANELLE NUCLEI

41

appears to be a very simple and rapid method for making direct estimation of the extent of degradation of cpDNA in young zygotes. ‘To identify the factors involved in the preferential destruction of cpnuclei of male origin, young zygotes of C. reinhardtii were treated with specific inhibitors of translation in chloroplasts(chloramphenicol,erythromycin) and in the cytoplasm (cyclohexirnide),with inhibitors of nucleases (anrintricarboxylicacid, ethidium bromide), with inhibitors of transcription in chloroplasts (rifampicin) and in the cytoplasm (actinornycin D and a-amanitin), at various temperatures, with chelating agents (EDTA, EGTA), and with UV light (Kuroiwa et al., 1983a,b). The degradation of cp-nuclei was then examined by staining with DAPI and epifluorescence microscopy. The nucleolytic reaction in uiuo displays the following characteristics: (i) digestion of about eight cp-nuclei of male origin occurs synchronously within a single chloroplast; (ii) each cp-nucleus deflates and dissolves from its periphery toward its center; and (iii) the cp-nuclei are completely digested within 20 minutes after the initiation of degradation. These characteristics suggest that DNases similar to DNase I and micrococcal nuclease may be involved in the digestion of the cp-nuclei. If cpDNAs are digested only by a restriction endonuclease, as described in the restriction-modificationhypothesis, the cp-nuclei should swell slightly and would not be expected to disappear rapidly. Attempts to isolate nucleases from young zygotes of C. reinhardtii revealed that extracted nucleases required the presence of Ca2+for full activation. Thus this class of nucleases is called nuclease C (Ogawaand Kuroiwa, 1985).The nuclease is pcllymorphic and includes six molecules (16, 18, 20, 22, 25, and 26 kDa). Nucleases with similar properties have been found in all plants examined showing uniparental transmission of ptDNA, such as A. calyculus, Nitella axillformis, M . polymorpha, Neurospora crassa, N . tabacum, and P . ptrsica (Nakamura and Kuroiwa, 1987; Nii et al., 1988). Since preferential destruction of cp-nuclei of male origin is inhibited by EGTA, nuclease C may be responsible for the preferential digestion of cpDNA in uiuo. However, nuclease C is found in zygotes and in female and male vegetative cells and gametes (Ogawa and Kuroiwa, 1985a). When cpDNAs isolated from female and male vegetative cells and gametes were infabated in a solution of nuclease C, cpDNAs of both male and female origin were completely digested. Therefore, the preferential destruction of cp-nuclei of male origin is probably not controlled at the DNA level. The presence of nuclease C in both female and male gametes led to the hypothesis that nuclease C is presynthesized prior to mating, exists in an inactive form in pregametic cells, and is activated by enzymatic systems in young zygotes. Preferential destruction of cp-nuclei of male origin can be completely inhibited by addition of cycloheximide soon after mating but not by addi-

42

TS UNEYOSHI KUROl WA

tion of chloramphenicol and erythromycin. This result suggested that proteins which were synthesized de nouo in the cytoplasm after mating were required for the preferential destruction (Kuroiwa ef al., 1983a, b). An examination was made of the polypeptides that are synthesized in the cytoplasm and related to the preferential destruction of cpDNA of male origin by incorporation of [S35]methionineand two-dimensional gel electrophoresis. About 200 polypetides were synthesized within 30 minutes 94 (PI, 94 ( y ) , 52,50, and after mating and six of these polypetides [94 (a), 38 kDa] appeared to be essential for the preferential destruction of cpnuclei of male origin (Nakamura and Kuroiwa, 1989; Nakamura et al., 1988).These polypeptides are probably involved in the activation of nuclease C, perhaps via alterations in biochemical or biophysical parameters such as the concentration of intracellular Ca2', the permeability of chloroplast membrane, the structure of lysosomes, etc. In Monosfroma, the preferential degradation of cpDNA from one parent occurs within 30 minutes. Just before this degradation the fluorescence intensity of the chlorophyll in the chloroplast in question decreases while that of the chlorophyll in the other chloroplast does not change (Kuroiwa and Hori, 1990). Under the electron microscope, the entire chloroplast can be seen to become electron dense while the other chloroplast does not change, suggesting that some parameter, probably the permeability of the membrane systems of the chloroplast from one parent, changes markedly, promoting the action of nuclease C. The six polypeptides may be encoded by mRNAs synthesized just after mating since the protein synthesis, RNA synthesis, and the preferential destruction of cpnuclei are inhibited completely by treatment with actinomycin Dafter mating (Kuroiwa etal., 1983b; 1985b). The question, then, is whether cell nuclei of male or female origin contribute to the synthesis of relevant mRNAs. When female gametes of C . reinhardtii which have been irradiated with UV light mate with unirradiated gametes, the preferential destruction of cp-nuclei is inhibited, but when male gametes are irradiated with UV light, destruction of cp-nuclei occurs normally. The effects of UV irradiation suggest that the target of the irradiation is the DNA molecule. The important mRNAs are probably synthesized in the cell nucleus from the female parent just after mating. 5-Fluorodeoxyuridine (5-FdUrd) selectively decreases the amount of cpDNA in plastids by inhibiting the synthesis of organelle DNA but it does not affect the amount of the cell nuclear DNA, the rate of proliferation of cells, or the cell density at stationary phase (Wurtz ef al., 1977). When the female cells are incubated in 5FdUrd, the cpDNAs in some chloroplasts of female gametes are completely lost (Nakamura and Kuroiwa, 1989). When female gametes without cpDNA are crossed with normal male gametes, the preferential digestion of cp-nuclei of male origin occurs and, as a result, the male zygotes,

PLASTIDS AND ORGANELLE NUCLEI

43

containing chloroplasts without their cpDNAs, are generated. When the preferential digestion of cpDNA of male origin occurs after mating of male gametes treated with 5-FdUrd and normal female gamete, the six polypeptides mentioned above are synthesized. These results indicate that mRNA thiit encodes the information of the preferential digestion of cp-nuclei of male origin is synthesized in the cell nucleus of female origin. These results do not conflict with the finding by Sager and Ramanis (1973) that irradiation of female gametes with UV light just before mating induces the biparental inheritance of nonchromosomal genes. However, Wurtz e f al. (1!)77) reported that when the mt' gamete is treated with 5-FdUrd, the decrease in levels of cpDNA appears to perturb the normal maternal transmission of cp-genes, with a dramatic increase in the proportion of exceptional zygotes. The discrepancy between these cytological and genetic experiments can be explained if 5-FdUrd strongly inhibits synthesis of organelle DNA but only weakly inhibits the expression of the cell nuclear gene(s) essential for the preferential digestion of cpDNA of male origin in young zygotes. Protection of female cpDNA from degradation can be clearly explained by methylation in Sager's model but not in the present nuclease C model. Hlowever,Ogawa and Kuroiwa (1985b) obtained interesting data using cell msodels prepared by treating vegetative and gametic cells of C. reinhardtii wicth a medium that included EDTA, 2-mercaptoethanol, and spermin. When the model cells were incubated with the nuclease C fraction, cpnuclei in female and male vegetative cells and male gametes disappeared completely, but cp-nuclei in female gametes did not. In each case, cell nuclei were not markedly affected. Numbers of female gametes with cpnuclei that were not digested by nuclease C increased as gametogenesis advanced. These results suggest that there is an as yet unidentified mechanism that protects cpDNA of female origin from nuclease C in domains such as the chloroplast matrix and the membrane that surrounds female cpDNAs. Such a hypothesis is presented schematically in Fig. 16 (Kuroiwa, 1985). For further elucidation of the molecular mechanisms of maternal inheritance, we need to identify the substances that protect female cpDNA and to characterize the proteins that activate nuclease C, which may be encoded by mRNA synthesized de nouo in young zygotes just after mating. Detailed analyses of mutants that do not display preferential destruction should provide important clues.

B. ANISOGAMOUS AND OOGAMOUS ALGAEAND FERN The preferential degradation of male cp-nuclei occurs in young zygotes after mating in several isogamous algae and it may be responsible for

44

TSUNEYOSHI KUROIWA

maternal inheritance of cp-genes. However, little information is currently available on the behavior of ptDNA during gametogenesis and zygote formation in anisogamous algae. In B . maxima, 12 hours before the release of gametes, male and female green lateral branches of thalli became orange-brown and dark green in color, respectively. Large chloroplasts in the male gametangia divide to form about 32 or more small daughter chloroplasts which contain a few cp-nuclei, and the large chloroplasts in the female divide to form eight or more daughter chloroplasts that contain about 20 ct-nuclei. At about 10 hours before the release of gametes, when cleavage of cytoplasm into gametes begins, cp-nuclei and mt-nuclei disap-

PLASTIDS AND ORGANELLE NUCLEI

45

pear completely from most male gametangia, though about 10% of the gametes still contain one to three cp-nuclei and one or two mt-nuclei. By contrast, cp-nuclei and mt-nuclei in organelles of female gametangia persist. The disappearance of cpDNA in male gametes has been confirmed by CsCl density-gradient centrifugation (Kuroiwa et al., 1991). After the onset of the light period, male and female gametes are markedly anisogamous and contain one spherical cell nucleus, one chloroplast, and one mitochondrion. The chloroplasts in both male and female gametes are spherical and cup-shaped, respectively, and occupy a major part of the posterior hemisphere of the gametes; the mitochondrion lies at the base of the flagella. Only a small percentage of the total population of gametes contains cp- and mt-nuclei, but female gametes contain about 20 cp-nuclei per chloroplast and 3-6 mt-nuclei per mitochondrion. Two minutes after mating of female and male gametes, the anisogamous gametes form an irregular heart-shaped union, in which cp-nuclei and mt-nuclei that originated in the male gamete completely disappear, while the cp-nuclei and mt-nuclei of female origin remain. Twenty-five minutes after mating, both cell nuclei fuse in the young zygote, but the chloroplasts of male and female origin are never seen to fuse. After mating, the Golgi vesicles and ly sosomes develop markedly. When the lysosome is associated preferentially with only the small chloroplast without DNA, the small chloroplast begins to break down and finally is fragmented to small membranes. By contrast, the large chloroplast of female origin remains and increases in volume. The small chloroplast of male origin is digested within 12 hours after mating (T. Hori and T. Kuroiwa, unpublished data). The preferential degradation of ptDNA in male gametes before mating occurs in B. plurnosa (Ogawa, 1988; Saito et al., 1989). In many oogamous algal species and in some anisogamous species, the preferential degradation of cpDNA during spermatogenesis must be responsible for the disintegration of the plastids that are contributed by the male parent to the egg cell after gametic fusion. Elimination of plastids before mating has been reported for the spermatozoids of a few oogamous Xanthophytes and brown algae (Whatley, 1982). In the Characeae, the behavior of cpDNA has been examined after double-staining with DAPI and FITC subsequent to treatment with DNAspecific antibodies. The cp-nuclei, which were present in internode cells and cells at the early spermatid stage, disappeared during spermatogenesis (Sun et al., 1988). Similar events also occurred during spermatogenesis in the fern P. uittata, as summarized in Figs. 17 and 18 (H. Kuroiwa et al., 1'988). Spores of P. uittata were grown on a solid medium that contained antheridiogen and an antheridium initial formed on each protonema cell. The antheridium initial divided and produced 16 spermatocytes and three surrounding cells. The chloroplast in the spermatocytes decreased in vol-

46

TSUNEYOSHI KUROIWA

FIG. 17. Phase-contrast (a, d, g) and epifluorescence images (b, c, e, f, h) showing cell nuclei (N), plastids (arrows in a, d, and g), pt-nuclei (arrows in c and f), and flagella (F in g) in 16-cell spermatocytes (a-c), the transformed cells derived from 16-cell spermatocytes (d-f ), and sperm (g, h) in the fern P. virrora. In the I k e l l spermatocyctes that are being transformed into sperm, the chlorophyll of the plastids has disappeared completely (arrows in b, e) but plastids (arrows in a, d) and pt-nuclei (DNA) (arrows in c, f ) remain. When the sperm have matured, pt-nuclei also disappear (h). Thus, the plastids in sperm do not contain DNA. Bar = 10 pm. (From H.Kuroiwa er ol., 1988; reproduced by permission of Springer-Verlag.)

PLASTIDS AND ORGANELLE NUCLEI

47

Spore 1

ChlOrOPlESt

Cp-nuclei

8

ca

@

0

ume as cell division occurred and this process was repeated until the final volume of each chloroplast was 1/15 that of the primary chloroplast, The DNA content of the chloroplasts was also reduced to 1/5 of the original value. When the sperm matured and the shape of its cell nucleus began to transform into a spiral, the ptDNA disappeared. The plastids without DNA remained visible until the final stage of development of the sperm (Figs. 17 and 18). These results indicate that, in anisogomous algae and fern, the preferential reduction in the amount of cpDNA per chloroplast occurs by division of chloroplasts without DNA synthesis, as the first steps, and then the preferential digestion of cpDNA in the chloroplast or plastid occurs during the transformation of the cell nucleus. Thus, the fertilization of an egg that contains a large amount of ptDNA by a sperm that contains a very small number of small plastids without

48

TSUNEYOSHI KUROIWA

DNA seems to occur in the anisogamous algae, in mosses, and in ferns. It was reported that, in motile spermatozoids of lower plants, a cytoplasmic vesicle which includes the plastids is discarded before the male gamete reaches the egg (Whatley, 1982). C. HIGHERPLANTS Correns (1909) and Baur (1909) identified two basic patterns of maternal and biparental transmission of cytoplasmic genes in plants. In angiosperms the pattern of plastid transmission has been examined genetically in crosses of more than 50 genera of dicotyledons and monocotyledons (Smith, 1988), and the pattern of transmission of ptDNA has been observed cytologically in more than 100 genera (Miyamura et al., 1987; Corriveau and Coleman, 1988; Kuroiwa and Hori, 1990; Kuroiwa, T. et al., 1990b). Such studies have shown that plastid transmission is most often maternal, but biparental inheritance is not uncommon and has been demonstrated in about 20 genera (Sager, 1972; Corriveau and Coleman, 1988; Gillham et al., 1990; Sears, 1980; Smith, 1988; Tilney-Bassett, 1975). The progeny of certain species regularly contain at least some paternally derived plastids (Smith, 1988; Gillham er al., 1990). The mechanism of maternal inheritance of plastids in angiosperms has been explained in terms of elimination of plastids from the pollen parent during the time between the formation of pollen grains and fertilization of the egg by one of the two sperm cells. Therefore, before a description is provided of the mechanism of cytoplasmic inheritance in angiosperms, the process of formation of sperm cells must be summarized (Fig. 20). The haploid microspore cells, which are formed after meiosis, can develop into four pollen grains. During the formation of pollen grains, each microspore divides mitotically and unequally to yield a large vegetative cell that contains large amounts of cytoplasm and a small generative cell that contains a small amount of cytoplasm. The generative cell moves to the internal region of the vegetative cell and, thus, is surrounded by the vegetative cell. The generative cell divides a second time to form two daughter sperm cells. The division of the generative cell can occur either in pollen grains or in pollen tubes. In the pollen tube, the two sperm cells are surrounded by the cytoplasm of the vegetative cell. At fertilization, one of the sperm cells fuses with the egg cell, forming a diploid zygote from which the embryo will develop, while the other fuses with the polar cells to yield a triploid cell from which the endosperm of the seed develops (Fig. 20). From observations by both light and electron microscopy, Hagemann (1983) recognized four patterns of plastid transmission in angiosperms. In the Lycopersicon type, which is represented by the tomato Lycopersicon

PLASTIDS AND ORGANELLE NUCLEI

49

esculentum, all microspore plastids segregate to the vegetative cell, and only chloroplasts from the maternal parent are transmitted to the progeny. In the Solanum (potato) type, the first partitioning of the cytoplasm during pollen mitosis is more equal, with the generative cell as well as the vegetative: cell receiving some plastids. However, in the course of further developrnent of pollen, the plastids of the generative cell are lost, so the sperm cells do not contain plastids. In the Triticum type, which so far appears to be restricted to grasses, plastids are found in the generative and sperm cells. However, when the sperm cell enters the egg cell, enucleated cytoplasmic bodies containing mitochondria and plastids are left outside (Mogensen, 1988; Mogensen and Rusche, 1985). In the Pelargonium patten1 of plastid transmission, found in geraniums, the distribution of plastids to the generative and vegetative cells during the first pollen mitosis results in biparental transmission of plastids. Patterns of plastid transmission can vary within a single taxon such as the Liliaceae (Schroder, 1984; Schroder and Hagemann, 1985; Vaughn et al., 1980). As described above, mechanisms of maternal inheritance in angiosperms have been explained by the absence of plastids (Sears, 1980) and mitochondria in the generative cells. However, in Lillium longifioem, in which organelle DNA shows maternal inheritance, a few plastids and many small mitochondria are present in sperm cells in pollen tubes (MikiHiroshige and Nakamura, 1977). However, when the organelles in the vegetative cells are stained with DAPI, fluorescent spots are never seen, suggesting that the organelles do not themselves contain DNA (Miyamura et d . ,1987). Since the generative cell contains organelle DNA at the early stages of the formation of pollen grains, the preferential elimination of organelle DNA from the organelles in the generative cell occurs during forination of sperm cells and may be responsible for the maternal inheritance, as described in the case of the algae and the fern. Miyamura et al. (1987) selected 18 representative plants in which the two different types of inheritance occurred and examined cytologically whether or not the preferential elimination of the ptDNA and the mtDNA in the generative cellls occurred during formation of pollen grains and in pollen tubes. They cornpared their cytological results with the results of classical genetic studies and found that the plants could be classified into three types according to the behavior of the organelle DNA in the generative cells during the formation of pollen grains. The first type, exemplified by N . tabucum and Lycoris radiata, is characterized by the digestion of organelle DNA in the generative cells immediately after the first mitosis in pollen grains (Figs. 19a and 20). The second type, exemplified by T. aestiuum, is Characterized by the digestion of organelle DNA between the second pollen mitosis and the initiation of the transformation of sperm nuclei (Fig.

50

TSUNEYOSHI KUROIWA

FIG. 19. Fluorescence photomicrographs of vegetative nuclei (VN), generative nuclei

(GN),a sperm nucleus (SN), pt-nuclei (large arrows) and mt-nuclei (small arrows) in pollen grains of four angiosperms such as Lycoris radiata (a), Opuntiaficus-indica (b), Rhododendron indicurn (c), and Pelargonium zonale (d) after staining with DAPI. Bar = 1 Fm.

51

PLASTIDS AND ORGANELLE NUCLEI Maternal inheritance

Biparental inheritance

a@

(0

Qermination of pollen grain

4

Trlcellular pollen

2nd pollen grain m i t ori r

Blcellular pollen grain Migration of generative cel l

Y

Mlrabiiir type

m l t orl r

Pe1argonium.type

FIG. 20. Schematic representation explaining mechanisms of maternal inheritance and biparental inheritance of organelle DNA in angiosperms on the basis of whether or not organelle DNAs are present. Maternal inheritance of plastids and mitochondria of Mirabilis type may result from three steps, as follows: a reduction in the number of organellesjust after the first pollen mitosis (three arrowheads); the preferential digestion of organelle nuclei during formation of sperm cells; and the digestion of organelles that do not contain DNA during fertilization. Biparental inheritance of organelles of Pelargonium type may be the results of the protection of organelle DNA from nucleases or the absence of nucleases and the separation of organelles from both parents during formation of the embryo. GN, generative nucleus; VN, vegetative nucleus; PN, plastid nuclei; MN, mitochondrial nuclei; EN, egg nucleus; EPN, egg plastid nucleus; EMN, egg mitochondrial nucleus.

52

TSUNEYOSHI KUROIWA

20). In general, those plants that are known for the maternal inheritance of ptDNAs belong to the first and second types. The third type is characterized by the presence of organelle DNA in sperm cells in pollen tubes. Plants such as Opuntia Ficus-indica (Fig. 19b), Rhododendron indicum (Fig. 19c), and P . zonale (Fig. 19d), which show biparental transmission, can be classified as the third type. The cytological observations are in basic agreement with the classical genetic data and the results have been confirmed by the extensive observations of Comveau and Coleman (1988)and Kuroiwa and Hori (1990; Table II), and by the technique of molecular biology (Comveau, et al., 1990). Kuroiwa and Hori (1990) also identified a number of additional species in Aozpaceae, Cactaceae, Campanulaceae, Ericaceae, Fabaceae, etc. in which ptDNA can be found in generative or sperm cells of pollen (Kuroiwa and Hori, lw),suggestingthat these have the potential for biparental inheritance. These results indicate that the preferential digestion of organelle DNA during formation of sperm cells must be responsible for the maternal inheritance in higher plants as well as in anisogamous algae and ferns. How does the preferential digestion of the organelle DNA in the generative cells occur in higher plants? In many species an examination has been made of whether or not the volume of cytoplasm in the generative cell, formed by the first pollen mitosis, affects the preferential destruction of organelle DNA. In general, in plants such as the Geraniaceae, in which the ratio of the volume of cytoplasm to that of the cell nucleus is high, the organelles remain and the preferential destruction of the organelle DNA does not occur. By contrast, in plants such as the Compositae, in which the ratio is low, the preferential destruction of the organelle DNA does indeed occur. However, in some cases, such as the Liliaceae, in spite of the ratio not being low, the preferential destruction of organelle DNA does occur. Accordingly, it is not always the case that the decrease in the volume of the generative cell by the first pollen mitosis or unequal division is directly responsible for the destruction of the organelle DNA in the generative cells. It is commonly held that the elimination of ptDNA from sperm cells (sperms) in anisogamous algae, ferns, and angiosperms involves the elimination of the organelle DNA in two steps, as follows. The number of organelles in the generative cells decrease markedly as a result of unequal division at the first step and then the organelle DNA is digested by Caz+ -dependent or Z2+ -dependent nucleases while the degradation of organelles without organelle DNA may be induced by lysosomes (Fig. 20). The activation of such nucleases may not operate in plants such as Pelargonium zonale and Schlumbergera russellienum that show biparental transmission of cpDNA (Fig. 20; Sodomergen and Kuroiwa, unpublished

53

PLASTIDS AND ORGANELLE NUCLEI

TABLE I1 BETWEEN THE BEHAVIOR OF Pt-NUCLEI AND CYTOPLASMIC INHERITANCE RELATIONSHIP IN VARIOUS SPECIES OF ANGIOSPERM

Taxon

Helianthus annuus Impatiens capenicis Arabidopsis thaliana Arabis albida Brassica campestris Bem vulgaris Chenopodium album Glycine max Trijblium pratense Hytirophyllum virginianum Go::sypiumhirsutum Mirabilis jalapa Pleantago major Avena sativa Corx lacryma-jobi Hordeum vulgare Oryza sativa Sorghum vulgare Triiicum aestivum Zeci mays Capsicum annuum Lycopersicon esculentum Nicotiana tabacum Petunia hybrida Solanum tuberosum Viola tricolor Hy,vericum perforathum Rhododendron maximum Rhododendron indicum Medicago sativa Geranium maculata Peragonium hortorum Ch forophytum elatum Oenothera biennis Phmeoius vulgaris Pisum sativum Ipomoea nil

a

Absence (-) or presence (+) of pt-nuclei

Genetic evidence for maternal inheritance (M) or biparental inheritance (B)"

-

M' M' M' M' M' M'

-

M'

-

+ + + + + + + + -

+ +

MI

M1.2

M'

M' M MI MI M1.2 MI

'

M' MI

M1.2

M' MIB'

M2

B' B' B2 B' B' B' B' B1.2 B' M' MI

(1) Based on Comveau and Coleman (1988);(2) based on Miyamura er 01. (1987).

54

TSUNEYOSHI KUROIWA

data). The transmission of paternal plastids at low frequency has been documented in both Nicotiana (Medgyesy et al., 1986)and Petunia. Such cases may be due to some accidents of the control system that induces digestion of ptDNA during the formation of pollen and of the separation system of plastids during embryogenesis. The amount of cpDNA in the sperm cells of the biparental type, in which cpDNA remains until the formation of pollen tubes, has been examined by VIMPICS. In the plants tested, the number of copies of cpDNA per cell varied from 24 to 550 (Table HI). It was surprising to us that the number of copies of cpDNA in P . tonale was the highest among all plants examined because Baur (1909) seemed to have already chosen Pelargonium as the most suitable material for research into non-Mendelian and biparental inheritance of leaf color. Triform repens contained a few copies of ptDNA in sperm cells (Miyamura et al., 1987). It is likely that in T. repens, maternal transmission of cpDNA occurs at lower frequency, as shown by Corriveau and Coleman (1988). Among biparental progeny of Oenothera, maternally derived plastids predominate, whereas in the alfalfa, Medicago satiua, paternally derived plastids predominate. These phenomena also can be explained by the number of copies of cpDNA in sperm cells. Both Oenotheru and M . satiua show biparental transmission of ptDNA and their sperm cells contain ptDNA. Since the DNA content is lower than that in other plants, such as Pelargonium and Rhododendron, the frequency of maternal transmission of ptDNA is high. Of course, there is the possibility that additional mechanisms influence the inheritance of plastids (Lee et al., 1988; Smith, 1989). Spatial distribution of the plastids in the zygote

TABLE 111 LEVELS OF ptDNA CONTENT PER SPERM CELLI N VARIOUSPLANTS WITH PI-NUCLEI

Taxon Trifolium repens Oenorhera sp. Pelargonium zonale Pelargonium inquinans Geranium yesoense Geranium roberrianum Rhododendron indicum Triticum aestivum

Number of plastids/cell 11 6 55 51

19 21 12 14

Number of pt-nuclei/plastid

T value/pt-nuclei" 4 5 10 9 8 9 10 2

Number of copies of ptDNA 44 30 550 459 152

189 120 24

" Fluorescence intensity of T4 phage, fured and stained with DAPI in the same way as plant cells, was used as a standard for measurements ( 7 = 1).

PLASTIDS AND ORGANELLE NUCLEI

55

prior to its asymmetric division into a suspensor and a terminal cell may play a critical role, with paternal plastids in alfalfa being more favorably situated for entry into the terminal cell (Tilney-Bassett and Almouslem, 1989).

The asymmetric dispersion of plastids during embryogenesis provides a useful explanation of the biparental pattern of inheritance in the case of Pelargonium, which yields green, white, and variegated plants (Fig. 20). Since the plants showing biparental inheritance of ptDNA are composed of cells that contain organelles from both parents, they will probably be important materials for “organelle technology.” Paternal inheritance and transmission of cpDNA have been shown by genetic experiments (Ohba et al., 1971) and through the use of restriction fragment polymorphisms (RFLPs) in a number of conifers (Neale et al., 1986; Neale and Sederoff, 1989; Wagner et al., 1987; Szmidt et al., 1983, 1987). Paternal transmission in conifers can be explained by the observation that paternal plastids enter the egg cell and maternal plastids are often degraded (Whatley, 1982). Apparent cpDNA recombinants have been observed in pine, and Pinus banksiana and P. contorta are known to hybridize naturally (Govindaraju et al., 1988). In the green algaEuglena, Ehara et al. (1984) reported that, at a certain stage of the cell cycle, several spherical chloroplasts conjoin to form a single giant body that contains a threadlike cp-nuclei, which may be formed by the fusion of small, spherical cp-nuclei. Recently, Kawano et al. (1991) found in Physarum polycephalum that mitochondrial fusion, the mt-nuclear fusion, and the recombination of mtDNA were controlled by a plasmid in the mitochondria. Plasmids associated with the fusion of chloroplasts and recombination of ptDNA may be present in many organisms.

w.summary Our present understanding of the replication, differentiation, and inheritance of plastids can be summarized as follows: 1. Most plants can be classified as being one of five types: the SN, CN, CL, PS, or SP type, based on differences in the shape, size, and distribution of the cp-nuclei in their mature chloroplasts. The differences in patterns of chloroplast nuclei can be explained by the pattern of distribution of the replicated chloroplast genomes in mature chloroplasts. They may be distributed randomly throughout the chloroplast (SN type); gathered around the pyrenoid in the chloroplast (SP type); distributed along the periphery of the chloroplast (PS type; fused to form spherules in the

56

TSUNEYOSHI KUROIWA

central area of the chloroplast (CN type); or fused to form a circle along the periphery of the chloroplast (CL type). 2. Proplastids contain one to a few copies of their own unique circular genome, of which the size varies from 12 X lo5 bp to 18 X lo5 bp. The genome is organized to form a pp-nucleus approximately 0.5 p m in diameter in the central area of a proplastid by interactions with specific proteins. The packing ratio is higher than that of cell nuclei. Proplastids can divide into daughter proplastids by binary fission with concomitant separation of pp-genomes after endoduplication of ppDN A. 3. The concept of organelle nuclei, as described above, has changed our ideas about the division of organelles. Thus, the process of organelle division must be composed of two main events: division of the organelle nucleus and organellekinesis (analogue of cytokinesis). The pt-nuclear division occurs after endoduplication of ptDNA or in the absence of prior DNA synthesis. The association between pt-DNA and membrane systems has not been clearly proven in proplastids. The strong binding between specific sites of ptDNA and membrane systems is observed in chloroplasts and may be related to the dispersion of pt-chromosomes and the activation of photogenes after greening. The plastidkinesis is mediated by the plastiddividing (PD) ring in lower eukaryotes, in moss, and in gymnosperm which is located on the cytoplasm outside plastids. In angiosperm, the plastiddividing ring is a doublet which is composed of an outer ring and inner ring. The PD ring is composed of a bundle of actinlike filaments. Plastidkinesis occurs by contraction of this bundle. 4. Proplastids can differentiate with a concomitant increase in DNA content and volume of more than 10-fold, and with dispersion of pt-nuclei into etioplasts, chloroplasts, amyloplasts, and chromoplasts, which are very specific organelles in cells in differentiated tissues. Proplastids and all differentiated plastids are called plastids as a general term and all plastids contain a genome (DNA) with basically identical physical characteristics. The genome forms a pt-nucleus by association with different proteins, and it can divide by constriction rather than partition. There are two hypotheses to explain the mechanism of plastid development: one is that the development of plastids is induced by demethylation of ptDNA; the other is that the development is controlled by the proteins that bind to the ptDNA. 5. In higher plants, the DNA content per chloroplast is reduced in mature leaves. Some workers have explained the decrease in number of copies of DNA per mature chloroplast as being due to the division of chloroplasts that was not accompanied by synthesis of ptDNA. As green leaves begin to turn yellow in P.persica, the cell nuclear DNA and ptDNA

PLASTIDS AND ORGANELLE NUCLEI

57

in palisade cells are degraded completely. During the decrease in levels of DNA in senescing peach leaves, the nucleolytic activity that requires Zn2+ for full activation develops gradually. In 0. sariua, when the grana and thylakoid membrane system are adequately developed in the mature ch1,oroplastsin the coleoptile, the first leaf, and the second leaf, the digestion of each ptDNA has already begun but the cell nucleus remains intact. The nuclease that requires Zn2+is activated with concomitant degradation of cpDNA, and, thus, is involved in relation to the digestion of cpDNA in the coleoptile, the first, and the second to the other leaves. 6. The preferential digestion of cpDNA of male origin occurs in the young zygotes of isogamous algae and seems to be responsible for the maternal inheritance of cpDNA. In a hypothetical mechanism for the regulation of maternal inheritance of cpDNA, female gametes have the ability to protect their cpDNA against Ca2' -dependent nuclease C during gametogenesis by changing the domains that surround female cpDNA. Soon after mating of male and female gametes, specific mRNAs are synthesized in a cell nucleus of female origin in the newly formed zygote. Then the proteins encoded by the mRNAs are synthesized de nouo in the cytoplasm and directly or indirectly activate nuclease C. After a selective change in the permeability of the membrane systems of chloroplasts of male origin, nuclease C preferentially enters the chloroplasts of male origin and digests the cpDNA. Since cpDNA of female origin remains an'd is transmitted into the progeny, maternal inheritance occurs. In the case of maternal inheritance in anisogamous algae, ferns, and higher plants, the preferential digestion of cpDNA occurs during spermatogenesis (formation of sperm cells) and is responsible for maternal inheritance. The degradation of plastids of male origin occurs in at least three steps: the reduction of the number of plastids per sperm is reduced by unequal division; the ptDNA in sperms is preferentially digested during spermatogenesis; and the degradation of the plastids without DNA occurs during or after fertilization. In plants which show biparental transmission of ptDNA, the preferential degradation of ptDNA does not occur.

ACKNOWLEDGMENTS 1 wish to thank Drs. S. Kawano and H. Duroiwa for their valuable advice and encouragement and Drs. T. Hori, Y. Sasaki, S. Nakamura, H. Hashimoto, Sodomergan,Y. Nemoto, and S. Miyamura for providing photographs, and Miss K. Mori for providing samples. This work was supported by grant nos. 63440003 and 02242206 from the Japanese Ministry of Education, Science, and Culture and a grant pioneering research project in biotechnology from Ministry of Agriculture, Forestry, and Fisheries of Japan.

58

TSUNEYOSHI KUROIWA

REFERENCES Bathgote, B., Goodenough, P. W., and Grierson, D. (1986). Plant Physiol. W, 223-233. Baumgartner, B. J., Rapp, J. C., and Mullet, J. E. (1989). Plant Physiol. 89, 1011-1018. Baur, E. (1909). Z. Vererbungs 1,330-351. Bedbrook, J . R., and Bogorad, L. (1976). Proc. Natl. Acad. Sci. U.S . A . 73,4309-4313. Bedbrook, J . R., Kolodner, R., and Bogorad, L. (1977) Cell (Cambridge, Mass.) 11,739-749. Boffey, S. A., and Leech, R. M. (1982) Plant Physiol. 69,1287-1391. Boffey, S. A., and Lloyd, D. (1988). “The Division and Segregation of Organelles.” Cambridge UNv. Press, Cambridge. Boffey, S. A., Ellis, J. R., Sellden, G., and Leech, R. M. (1979). Plant Physiol. 64,502-505. Bolen, P. L., Gillham, N. W., and Boynton, J. E. (1980). Curr. Genet. 2, 159-167. Boynton, J. E., Gillham, N. M., and Hams, E. H.(1990).Adu. Plunt Gene Res. 6, (in press). Briat, J. -F., Laulhtre, J. -P, and Mache, R. (1979). Eur. J. Biochem. 98,285-292. Briat, J. -F., Gigot, C., Laulhtre, J. -P., and Mache, R. (1982). Plant Physiol. 69,1205-1211. Burton, W. G., Grabowy, C. T., and Sager, R. (1979). Proc. Natl. Acad. Sci. U.S.A. 76, 1390- 1394. Chaly, N., and Possingham, J. V. (1981). Biol. Cell. 41,203-210. Coleman, A. W. (1978). Exp. Cell Res. 114,95-100. Coleman, A. W. (1985). J. Phycol. 21, 1-16. Coleman, A. W., and Maguire, M. J. (1983). Curr. Genet. 7, 211-218. Connett, M. B. (1987). Plant Mol. Biol. Rep. 4, 103-205. Correns, C. (1909). 2. Vererbungs 1,291-329. Comveau, J. S., and Coleman, A. W. (1988). Am. J. Bot. 75, 1443-1458. Coniveau, J. L., Goff, L. J., and Coleman, A. W. (1990). Curr. Genet. 17,439-444. Cozens, A. L., and Walker, J. E. (1986). Biochem. J. 236,453-460. Dann, O., Bergen, G., Demant, E., and Volz, G. (1971). Justus Liebigs Ann. Chem. 749,68. Day, A . , and EUis, T. H. N. (1984). Cell (Cambridge, Mass.) 39,359-368. DuPraw, E. J. (1970). “DNA and Chromosomes.” Holt, Reinhart, & Winston, New York. Ehara, T., Sumida, S., Osafune, T., and Hase, E. (1984). Plant Cell Physiol. 25,1133-1146. Feng, T. Y.,and Chiang, K. S. (1984). Proc. Natl. Acad. Sci. U.S.A. 81,3438-3442. Gillham, N. W. (1974). Annu. Rev. Gene 8,347-391. Gillham, N. W. (1978). “Organelle Heredity.” Raven Press, New York. Gillham, N. W., Boynton, J. E., and Hams, E. H. (1985). I n “Evolution of Genome Size” (T. Cavalier-Smith, ed.), pp. 299-351. Wiely, New York. Gillham, N. W., Boynton, E., and Haris, E. H. (1990). I n “Cell Culture and Somatic Cell Genetics of Plants” (L. Bogorad and I. K. Vasil, eds.), Vol. 7. Academic Press, San Diego, California (in press). Govindaraju, D. R., Wagner, D. B., Smith, G. P., and Dancik, B. P. (1988). Can. J. For. Res. 18, 1347-1350. Greenspan, H. P. (1977). J. Theor. Biol. 65,79-99. Gruissem, W.. Barkan, A., Deng, X.W., and Stem, D. (1988). Trends Genet. 4,258-263. Gunning, B. E. S. (1%5). J . CeN Biol. 24,79-85. Hagemann, R. (1976). I n “Genetics and Biogenesis of Chloroplasts and Mitochondria” (T. Bucher, W. Neupert, W. Sebald, and S . Werner, eds.), pp. 331-338. North-Holland Publ., Amsterdam and New York. Hagemann, R. (1983). I n “Fertilization and Embryogenesis of Ovulated Plants” (0. Erdelska, ed.), pp. 97-99. Bratislava Slov. Acad. Sci., Bratislava. Hallick, R. B., Lipper, C., Richards, 0. C., and Rutter, W. J. (1976). Biochemistry 15, 3039-3045.

PLASTIDS AND ORGANELLE NUCLEI

59

Hansmann, P., Falk, H., Ronai, K., and Sitte, P. (1985). Planta 164,459-472. Hams, El. H. (1989). “The Chlamydomonas Sourcebook.” Academic Press, San Diego, California. Hashimoto, H. (1985). Protoplasma 127, 119-127. Hashimoto, H. (1986). Protoplasma l35, 166-172. Hashimoto, H. (1989). Protoplasma 149,20-23. Hashimoto, H., and Possingham, J. V. (1989). Plant Physiol. 89, 1178-1183. Hermann, R. G., Kowallik, K. V.,and Bohnert, H. J. (1974). Port. Acta Biol. Ser. A 14, 91-110. Hu, J., and Bogorad, L., (1990). Proc. Natl. Acad. Sci. U.S.A. 87, 1531-1535. Igo-Kemcnes, T., Hoerz, W., and Zachau, H. G. (1982). Annu. Rev. Biochem. 38, 14-19. Jacob, F , Brenner, S., and Cuzin, F. (1963). Cold Spring Harbor Symp. Quant. Biol. 28, 329-348, James, T. W., and Jope, C. (1978). J . Cell Biol. 79,623-630. Johnson, K. A., and Rosenbaum, J. L. (1990). Cell (Cambridge, Mass.) 62,615-619. Kamata, Y . , Kondoh, K., Kuroiwa, T., and Nagata, N. (1989). Plant Cell Physiol. 30, 151- 156. Kawano, S., and Kuroiwa, T. (1985). Exp. Cell Res. 161,460-472. Kawano. S., Takano, H., Mori, K., and Kuroiwa, T. (1991). Protoplasma In press. Kirk, J. ‘r. O., and Tilney-Bassett, R. A. E. (1978). “The Plastids: Their Chemistry, Structure Growth and Inheritance,” 2nd ed. North-Holland Publ., Amsterdam and New Yorl;. Kislev, Pi., Swift, H., and Bogorad, L. (1%5). J. Cell Biol. 25, 327-344. Klein, R R., andMullet, J. E. (1986). J . Biol. Chem. 261, 11138-11145. Kolodnei-, R., and Tewari, K. K. (1975). Biochim. Biophys. Acta 402,372-390. Kuroiwa, H., Sugai, M., and Kuroiwa, T. (1988). Protoplasma 146,89-100. Kuroiwa, T. (1973). Exp. Cell Res. 78,351-359. Kuroiwa, T. (1974). J. Cell Biol. 63,299-306. Kuroiwa, T. (1982). Int. Rev. Cytol. 75, 1-69. Kuroiwa, T. (1985). Microbiol. Sci. 2,267-272. Kuroiwa, T. (1989a). Bot. Mag. (Tokyo) 102,231-329. Kuroiwa, T. (1989b). In ‘‘Analysis of Systematics of Chlorophyll a, c Plants’’ (M. Chihara, ed.). pp. 119-136. Tsukuba Univ. Press, Tsukuba. Kuroiwa, T., and Hizume, M. (1974). Exp. Cell Res. 87,407-412. Kuroiwa, T., and Hori, T. (1986). Protoplasma l33,85-87. Kuroiwa, T., and Hori, T. (1990). In “Dynamic Analysis of Regulatory Mechanisms in Life Cycle ofHigherPlants”(N. Takahashi, ed.), pp. 149-188. UNv. ofTokyoPress, Tokyo. Kuroiwa, T., and Kuroiwa, H. (1980). Experientia 36, 193-194. Kuroiwa, T., and Nakamura, S. (1986). Histochem. Cytochem. Acta 19,95-102. Kuroiwa, T., and Suzuki, T. (1981). Exp. Cell Res. W, 457-461. Kuroiwa, T., Kawano, T., and Hizume, M. (1976). Exp. Cell Res. 97,435-445. Kuroiwa, T., Kawano, T., and Hizume, M. (1977). J . Cell Biol. 72,687-697. Kuroiwa, T., Suzuki, T., Ogawa, K., and Kawano, S. (1981).Plant Cell Physiol. 22,322-3%. Kuroiwa, T., Kawano, S., Nishibayashi, S., and Sato, C. (1982). Nature (London) 298, 48 1-483. Kuroiwa, T., Kawano, S., and Sato, C. (1983a). Proc. Jpn. Acad. 59,177-181. Kuroiwa, T., Kawano, S., and Sato, C. (1983b). Proc. Jpn. Acad. 59,182-185. Kuroiwa, T., Enomoto, S., and Shihira-Ishikawa, I. (1985a). Experientia 41, 1171-1179. Kuroiwa, T., Nakamura, S., Sato, C., and Tsubo, Y. (1985b). Protoplasma 125,43-52. Kuroiwa, T., Nagashima, H., and Fukuda, I. (1989~).Protoplasma 149, 120-129.

60

TSUNEYOSHI KUROIWA

Kuroiwa, T., Kuroiwa. H.,Mita, T., and Fujie, M. (1990a). Protoplasma 158, 191-194. Kuroiwa. T., Kuroiwa, H., Mita,T., and Fujie, M. (1990b). Fluoresc. Technol. (in press). Kuroiwa, T., Kawano, S., Watanabe, M.,and Hori, T . (1991). Protoplasma (in press). Lamppa, G . G., and Bendich, A. J. (1979).Plant Physiol. 64, 126-130. Lamppa, G. K., Elliot, L. V., and Bendich, A. J. (1980). Planta 148,437-443. Lawrence, M.. and Possingham, J. V. (1986). Plant Physiol. 81,708-710. Lee, D. J., Blake, T. K., and Smith, S. E. (1988). Theor. Appl. Genet. 76,545-549. Leech, R. M. (1976).i n “Cell Division in Higher Plants” (M.Yoneman, ed.), pp. 135-159. Academic Press, New York. Leech, R. M., Thomson, W. W., and Platt-Alola, K. A. (1981). New Phytol. 87, 1-9. Lindbeck, A. G. C., and Rose, R. J (1987).Protoplasma w9,92-99. Lubben, T., Donaldson, G. K., Viitanen, P. V., and Gatenby, A. A. (1989). Plant Cell 1, 1223-1230. Luck, B. T., and Jordan, E. G. (1980). Ann. Bot. (London)45,511-514. Macherel, D., Kobayashi, H., Akazawa, T., Kawano, S., and Kuroiwa, T. (1985).Biochem. Biophys. Res. Commun. l33, 140-146. Marano, M. R., and Canillo, N. (1991). Plant Mol. Biol. 16, 11-19. McGhee, J . D., and Felsenfeld, G. (1980). Annu. Reo. Biochem. 49, I 1 15-1 156. Medgyesy. P., Pay, A., and Marton, L. (1986). Mol. Gen. Genet. 204, 195-198. Miki-Hiroshige, H . , and Nakamura, S. (1977). Jpn. J . Palynol. 19, 11-19. Mita, T., and Kuroiwa, T. (1988). Protoplasma, Suppl. 1, 133-152. Mita, T., Kanbe, T., Tanaka, K., and Kuroiwa, T. (1986). Protoplasma 130,211-213. Miyamura. S . , and Hori, T. (1989).Plant Morphol. 1, 19-22. Miyarnura, S., Nagata, T., and Kuroiwa, T. (1986). Protoplasma l33,66-72. Miyamura, S . , Kuroiwa, T., and Nagata, T. (1987). Protoplasma 141, 149-159. Miyamura. S.. Kuroiwa, T., and Nagata, T. (1990). Plant Cell Physiol. 31,597-607. Modrusan, Z . , and Wrischer, M.(1990). Protoplasma l54, 1-7. Mogensen, H. L. (1988).Proc. Nod. Acad. Sci. U.S.A. 85,2594-2597. Mogensen, H. L., and Rusche, M. L. (1985).Protoplasma US, 1-13. Munaut, C., Dombrowigz, D., and Matagune, R. F. (1990). Curr. Genet. 18,259-263. Muhlethaler, K., and Frey-Wyssling, A. (1959).J. Biophys. Biochem. Cytol. 6, 509-512. Nagashima, H., and Fukuda, I. (1981).Jpn. J . Phycol. 29,237-242. Nagashima, H., Kuoriwa, T., and Fukuda, 1. (1986). Plant Cell Physiol. 28, 315-321. Nakamura, S., and Kuroiwa, T. (1987). Plant Cell Physiol. 28,545-548. Nakamura, S . , and Kuroiwa, T. (1989). Eur. J . CellBiol. 48, 165-173. Nakamura, S., Itoh, S., and Kuroiwa, T. (1986). Plant Cell Physiol. 27,775-784. Nakamura, S . , Sato, C., and Kuroiwa, T. (1988). Plant Sci. 56, 129-136. Neale, D. B., and Sederoff,R. R. (1989). Theor. Appl. Genet. 77,212-216. Neale, D. B., Wheeler, N. C., and Allard, R. W. (1986). Can. J . For. Res. 16, 1152-1154. Nelson, N.. Nelson. H., and Schatz, G. (1980).Proc. Natl. Acad. Sci. U.S.A. 77,1361-1369. Nemoto, Y .. Kawano. S., Nakamura, S., Mita, T., Nagata, T., and Kuroiwa, T. (1988).Plant Cell Physiol. 29, 167-177. Nemoto, Y., Nagata, T., and Kuroiwa, T. (1989). Plant Cell Physiol. 30,445-454. Nemoto, Y., Kawano, S., Kondoh, K., Nagata, T., and Kuroiwa, T. (1990). Plant Cell Physiol. 31,767-776. Nemoto, Y., Kawano, S., Nagata, T., and Kuroiwa, T. (1991). Plant Cell Physiol. 32, 131-141. Ngemprasirtsiri, J . , Kobayashi, H., and Akazawa, T. (1988a).Proc. Natl. Acad. Sci. U.S.A. 85,4750-4754. Ngemprasirtsiri. J.. Kobayashi, H., and Akazawa, T. (1988b).Plant Physiol. 88, 16-20.

PLASTIDS AND ORGANELLE NUCLEI

61

Nii, N., Kawano, S., Nakamura, S., and Kuroiwa, T. (1988).J. Jpn. Hortic. Sci. 57,390-398. Nicihibayashi, S . , and Kuroiwa, T. (1982). Protoplasma 110, 177-184. Ogiiwa, K., and Kuroiwa, T. (1985a). Plant Cell Physiol. 26,481-491. Ogitwa, K., and Kuroiwa, T. (1985b). Plant Cell Physiol. 26,493-503. Ogitwa, S . (1988). Eot. Gaz. (Chicago) 149,25-29. Ohlba, K., Iwakawa, M., Ohada, Y.,and Murai, M. (1971). Siluae Genet. 20, 101-107. Ohyama, K., Fukuzawa, H., Kohchi, T., Shirai, H., Sano,T., Sano, S., Umesono, K., Shiki, Y., Takeuchi, M., Chang, Z., Aota, S., Inokuchi, H., and Izeki, H. (1986). Nature (London) 322,572-574. Oross, J. W., and Possingham, J. V. (1989). Protoplasma 150, 131-138. Palmer, J. D. (1985). Annu. Rev. Genet. 18,325-354. Piechulla, B., Pichershy, E., Cashaeore, A. R., and Gruissem, W. (1986). Plant Mol. Eiol. 5, 367-376. Possingham, J. V., and Lawrence, M. E. (1983). In?. Rev. Cytol. 84, 1-56. Pocisingham, J. V., and Rose, R. J. (1976). Proc. R. SOC.London, Ser. E 193, 295-305. Potisingham, J. V., Chaly, N., Robertson, M., and Cain, P. (1983). Eiol. Cell. 47,205-212. Rerss, T., and Link, G. (1985). Eur. J . Biochem. 148,207-212. Ris, H., and Plaut, W. (1%2). J. Cell Biol. l3,383-391. Robinow, C, G. (1956). Bacteriol. Rev. 20,207-242. Rochaix, J. -D. (1978). J . Mol. Eiol. 126,597-617. Rochaix, J. -D. (1985). Int. Rev. Cytol. 93,57-91. Rose, R. J. (1988). I n “The Division and Segregation of Organelles” (S. A. Boffey and D. Lloyd, eds.), pp. 171-195. Cambridge Univ. Press, Cambridge. Rose, R. J., and Possingham, J. V. (1976). J . Cell Sci. 20,341-355. Rose, R. J., Cran, D. G., and Possingham, J. V. (1974). Nature (London) 251,641-642. Royer, H. D., and Sager, R. (1979). Proc. Natl. Acad. Sci. U.S.A. 76,5794-5798. Sager, R. (1954). Proc. Narl. Acad. Sci. U.S.A. 40,356-363. Sager, R. (1972). “Cytoplasmic Genes and Organelles.” Academic Press, New York. Sager, R., and Ishida, M. (1%3). Proc. Natl. Acad. Sci. U.S.A. 50,725-730. Sager, R., and Lane, D. (1972). Proc. Natl. Acad. Sci. U.S.A. 69,2410-2413. Saj:er, R., and Ramanis, Z. (1973). Theor. Appl. Genet. 43,101-108. Saito, A., Ogawa, S., and Wada, S. (1989). Eot. Gaz. (Chicago) 150,25-29. Sano, H., Grabowy, C., and Sager, R. (1981). Proc. Natl. Acad. Sci. U.S.A. 78,3 118-3 122. Sasaki, Y., and Kuroiwa, T. (1989). Plant Mol. Sci. 11,585-588. Sasaki, Y., Morioka, S., and Matsuno R. (1990). Plant Cell Physiol. 31,925-931. Scliroder, M. -B. (1984). Eiol. Zentrulbl. 103,547-555. Schrcder, M. -B., and Hagemann, R. (1985) I n “Reproduction in Seed Plants, Ferns, and Mosses” (M. T. M. Willemse and J. L. Van Went, eds.), p. 52. Pudoc, Wageningen. Scott, N. S., and Possingham, J. V. (1980). J. Exp. Eot. 123, 1081-1092. Scott, N. S.. Cain, P., and Possingham, J. V. (1982). Z. Pjlanzentphysiol. lOS, 187-191. Scott, N. S., Tymmis, M. J., and Possingham, J. V. (1984). Planta 161, 12-19. Sears, B.B. (1980). Plasmid 4,233-255. Sellden, G . , and Leech, R. M. (1981). Plant Physiol. 68,731-734. Senda, Y., and Ueda, K. (1990). Proc. 55th Annu. Meet. Eot. SOC.Jpn, p. 247. Sheer, U., Messner, K., Raska, I., Hansmann, P., Falk, H., Spiess, E., and Franke, W. W. (1987). Eur. J . Cell Biol. 43, 358-371. Shinozaki, K., Ohme, M., Tanaka, M., Wakasugi, T., Hayashida, N., Obokata, J., Y Shinozaki, K., Ohto, C., Deno, H., Kamogashira, T., Yamada, K., Kusuda, J., Takaiwa, F., Kato, A., Tohdoh, N., Schimada, H., and Sugiura. M. (1986). EMBOJ. 5, 2043-2049.

62

TSUNEYOSHI KUROIWA

Smith, S. E. (1988). Plant Breed. Rev. 6,361-393. Smith, S . E. (1989). J . Hered. 80.214-217. Sodmergen, Kawano, S., Tano. S., and Kuroiwa, T. (1989). Protoplasma l52,65-68. Sodmergen, Kawano, S., Tano,S., and Kuroiwa, T. (1990). Protoplasma (in press). Sugiura, M. (1987). Bot. Mag. (Tokyo) 100,407-436. Sun, G. -H., Ueda, T. Q. P., and Kuroiwa, T. (1988). Protoplasma 144, 185-188. Suzuki, T., and Ueda, R. (1975). Bot. Mag. (Tokyo)88,319-321. Szmidt, A. E., El-Kassaby, Y. A., Sigugeirsson, A., Alden, T., Tymms, M. J., Scott, N. S., and Possingham, J. V. (1983). Plant Physiol. 71,785-788. Szmidt, A. E., Alden, T., and HaUgren, J. -E. (1987). Plant Mol. Biol. 9,59-64. Tewinkel, M., and Volkmann, D. (1987). Planta 172,309-320. Tilney-Bassett, R. A. E. (1975) In "Genetics and Biogenesis of Mitochondria and Chloroplasts" (C. W. Birky, Jr.. P. S. Perman, and T. J. Byers, eds.), pp. 268-308. Ohio State U N v . Press, Columbus. They-Bassett, R. A. E.. and Almouslem, A. B. (1989). Heredify 63, 145-153. Tsubo, Y., and Matsuda, Y. (1984). Current Genet. 8,223-229. Vaughn, K. C., DeBonte, L. R., Wilson, K. G., and Schaeffer, G. W. (1980). Science u)8, 1916- 1918. Wagner, D. B., Fumier, G. R., Saghai-Maroof, M. A., Williams, S. M., Dancik, B. P., and Allard, R. W. (1987). Proc. Natl. Acad. Sci. U.S.A. 84,2097-2100. Watanabe, A., and Price, C. A. (1982). Proc. Natl. Acad. Sci. U.S.A. 79,6304-6308. Werz, G . (1966). Plantu 68,256-268. Whatley, J. M. (1980). New Phytol. 86, 1-16. Whatley, J . M.(1982). Biol. Rev. Cambridge Philos. Soc. 57,527-569. Whitfeld. P.R . , and Bottomley, W. (1983). Annu. Rev. Plant Physiol. 34,279-310. Williamson, D. H., and Fennell, D. J. (1975). Methods Cell Physiol. 12,335. Wurtz, E. A., Boynton, J. E., and Gillham. N. W. (1977). Proc. Narl. Acad. Sci. U.S.A.74, 4552-4556. Yasuda, T., Kuroiwa, T., and Nagata, T. (1988). Planta 74,235-241. Yokomura, E. (1%7). Cytologia 32, 361-377. Yoshida, Y., Laulh&re,J. -P., Rozier. C., and Mache, R. (1978). Biol. Cell. 32, 187-190. Zachleder, V., and Cepiik, V. (1987). Proroplasrnn l38,37-44. Zaitlin, D., Hu, J., and Bogorad, L. (1989). Proc. Natl. Acad. Sci. U.S.A. 86,876-880.

INTERNATIONALREVIEW OF CYTOLOGY, VOL. 128

Cell Division in Diatoms JEREMYPICKETT-HEAPS School ($Botany, University of Melbourne, Parkville, Victoria 3052, Australia

I. Introduction Diatoins are among the most widespread, prolific organisms on this planet and they are enormously important in global ecology. The existence of vast deposits of diatomaceous earth, hundreds or thousands of feet thick, attest to their continuing abundance over the last 50 million years. The many thousands of species have long fascinated biologists because of their exquisitely sculptured, symmetrical cell walls or valves, fashioned from silica. A new valve is created for each daughter cell. Every cell division and morphogenesisof the various features of the valve involves an unusual diversity of organelles and cytoplasmic activities (reviewed in Pickett-Heaps et al., 1991). Thus, evolution of the valve morphogenetic machinery displays a virtuosity and opportunism in using cellular systems and organelles unmatched by any other group of organisms. Diatoms also provide the biologist with valuable material for studying the still enigmatic process of mitosis. Many diatoms are easy to isolate and culture. Some are very large, offering excellent microcopic objects for light microscopy in uiuo, with the advantage that chromosome movement is rapid enough to be detectable with the naked eye. Other species vary down to the minute, facilitating precise, fine structural analysis of the spindle. This review is specifically concerned with mitosis in diatoms and how the 0bse:rvations and inferences from these cells relate to mitotic mechanisms in general. Mitosis in diatoms is reviewed briefly in Pickett-Heaps and Tippit (1978) and Pickett-Heaps et al., (1982). Many of the observations on, and conclusions derived from, diatom spindle structure and function were unexpected and controversial. The relevance of these observations to understanding mitosis in more conventional cell types has often been questioned and, on occasion, dismissed on the grounds that diatoms are “primitive.” However, as will be shown below, almost all of these observations have proved to be directly relevant to more conventional spindle types. The diatom spindle has turned out to be exceptional for demonstrating the relationship of structure and function; in contrast, the spindles of most other cells are much more difficult to define structurally and functionally. Hyams and Brinkley (1989) have edited an excellent summary of recent work in the field. 63 Copyright 0 1991 by Acadermc Press, Inc.

AU nghts of reproduction in any form reserved

64

JEREMY PICKETT-HEAPS

Before proceeding further, nonbotanists are reminded that there are two major groups of diatoms, the centrics and the pennates, both diploid in the vegetative state. Centrics and pennates are distinguished (in most species) by their different symmetry as well as other significant differences. Both undergo two meiotic divisions preceding sexual reproduction. The centrics are oogamous, differentiating uniflagellate sperm that fertilize a large, inert oogonium. In contrast, pennates have (usually)isogamous amoeboid gametes. In both centrics and pennates, sexual reproduction is utilized via a complex series of cytoplasmic maneuvers to recreate large vegetative individuals in a population that steadily diminishes in size as a consequence of the cells being enclosed in the rigid cell wall. Such matters are reviewed in textbooks of phycology.

II. Early Work The course of nuclear division, or mitosis, was first elucidated by Strasburger and Flemming in the 1880s. A few years later, Lauterborn (1896) published a remarkable treatise on the cytology of diatoms which included a long chapter on cell division. He worked with a variety of cells, but concentrated on a large Surirella which is still one of the best species to follow mitosis in uiuo (Figs. 1-3). His descriptions of mitosis, from stained preparations and observations in uiuo, are outstanding even by modern standards and they are accurate in every respect, including his clear descriptions of features such as spindle fibers and Golgi bodies. This work has been an inspiration, and we have published a translation of his chapter (Pickett-Heaps et al., 1984a) and included Nomarski and ultrastructural images for comparison with the paintings he made. Even Lauterborn’s discussion is remarkably modem in context, but his book has been largely overlooked in the literature on mitosis. Modern ultrastructural methods were first applied to dividing diatoms by Manton and colleagues (Manton et al., 1969a, b, 1970a, b). About 5 years later, we started intensively investigating spindle structure and function. Several films showing mitosis and cytokinesis in uiuo are available (Pickett-Heaps, 1980, 1982, 1983b, 1984a). Cande and co-workers have been using the diatom spindle to great advantage for reactivation of anaphase B in uitro (Section VI1,B).

HI. The Microtubule Center (MC) A. INTERPHASE

Most diatoms studied possess a prominent granule tightly associated with the interphase nucleus. Lauterborn (18%) drew cytoplasmic fibrils

FIGS. 1-3. Spindle formation, live cells of Surirello. Figs. 1a.b. Tiny biconcave spindle (a) inmediately underneath MC (b). Figs. lc,d. Growth of this spindle over several minutes. Bar = 10 pm. Figs. 2a-c. Rapid growth of central spindle at end of cell (n, prophase nucleus) as nuclear envelope breaks down; central overlap (arrow) obvious by prometaphase (Fig. 2c). Bar = 10 pm. Figs. 3a-d. Metaphase-telophase, Chromosomes, gathered in donut-shaped mas:. around hollow central spindle (Fig. 3a), separate rapidly (Fig. 3b) in anaphase A. Spindle elongates (anaphase B) as overlap diminishes (Fig. 3c); cleavage furrow (arrows) passes between separated chromosomes and half spindles (Fig. 3d). Bar = 10 pm. Reproduced with permission from Pickett-Heaps ef al. (1984a).

66

JEREMY PICKE‘IT-HEAPS

emanating from this “centrosome” which was intimately involved with the appearance of the spindle during earliest prophase. Electron microscopy has confirmed that it is the focus of interphase microtubules (MTs) and therefore a microtubule organizing center (MTOC: Pickett-Heaps, 1969). However, the possibility that it nucleates MTs has yet to be directly demonstrated, so we have conservatively labeled it the microtubule center (MC) in many publications. In some larger diatoms, the MC is very conspicuous in uiuo (e.g., Fig. 1b). Ultrastructurally, the MC varies considerably in appearance, although it is similar in closely related organisms (e.g., Surirella and Cymatopleura). It usually is spherical during interphase but may change during the cell cycle. In centrics, it often is quite small and amorphous (e.g., Dirylum: Pickett-Heaps et al., 1988b). However, in Melosira, it is a very dense osmiophilic granule (Crawford, 1973,Tippit et al., 1975),as it is in many pennates (Pinnularia: Pickett-Heaps et al., 1978a; Achnanthes: Boyle et al., 1984; Hanrzschia: Pickett-Heaps et al., 1980b; Nirzschia: Pickett-Heaps, 1983a). In Surirella and Cyrnatopfeura,it has a characteristic finely tubular substructure (Figs. 4,5; Drum and Pankratz, 1963; Tippit and Pickett-Heaps, 1977; Pickett-Heaps et al., 1984a; Pickett-Heaps, 1991);in Nuuicula, it is rather ill defined (Edgar and Pickett-Heaps, 1984).

B. THEMC AND THE CELLCYCLE:ITS ROLEIN SPINDLE FORMATION Lauterborn’s paintings show the primordial spindle arising closely adjacent to the MC, although the two are quite separate structures. This has been confirmed in the pennates examined so far (Figs. I , 4; described below). As the primorial spindle grows, the MT cytoskeleton associated with the MC becomes particularly striking. Then the nucleus moves to a predetermined position in the cell for mitosis, a movement orchestrated by the MT cytoskeleton enlarging around the MC. In centrics, the nucleus divides while adpressed to the side against what are called the “girdle bands” of the wall or “valve,” where it usually resides during interphase; Wordeman et al. (1986)have demonstrated with anti-MT drugs that the MC/MT cytoskeleton positions the nucleus before mitosis and the spindle during cytokinesis. Ditylurn is unusual in that the interphase nucleus is suspended centrally in the vacuole and so this lateral movement always indicates impending mitosis (Pickett-Heaps et al., 1988b). In pennates too, the premitotic nuclear movement is to one side of the valve while the large chloroplast(s) usually become(s) rearranged, a morphological change that signals the imminence of mitosis. Thus, in Hantzschia, the central nuclear opening between the chloroplasts enlarges markedly as the forming spindle appears just under the valve (Fig. 6;

FIGS.4 and 5. Relationship between spindle and MC, Scrrirella. Fig. 4a. Striated primordial spindle with one polar vacuole (v) visible, close to MC. (n, nucleus; t, MTs). Bar = 0.5 pm. Fig. 4b. Growth of MTs between outer layers (polar plate) of primordial spindle; MC and nucleus at end of cell, MC beginning to disperse (ne, nuclear envelope). Bar = 0.5 fim. Fig. 5a. Polar plate (p) detached from central spindle at telophase, with two small MCs (arrows) at edge; these later fuse. Bar = 0.5 pm. Fig. 5b. Valve initiation with newly formed MC moving to new Silicia Deposition Vesicle (s); circular polar plate (p) dispersing. Bar = 0.5 pm. Reproduced with permission from Tippit and Pickett-Heaps (1977).

68

JEREMY PICKE'IT-HEAPS

Pickett-Heaps et al., 1980a). In Pinnularia, the chloroplasts rotate so that when the cell is observed from the side, the two chloroplasts appear edge on instead of having their interphase orientation in face view. In Surirella and Cymatopleura, the central interphase nucleus moves to one particular end of the cell during prophase (Lauterborn, 1896; Tippit and PickettHeaps, 1977; Pickett-Heaps et al., 1984a; Pickett-Heaps, 1991);this movement, too, is toward the girdle bands in a cell whose symmetry has been derived from that of naviculoid pennates (see Pickett-Heaps et al., 1991). These cytoplasmic changes in the larger diatoms greatly facilitate the selection of live or fixed dividing cells for examination and experimentation. The activity of the MT cytoskeleton during this premitotic movement is particularly striking in live Surirella (Lauterborn, 1896; Pickett-Heaps, 1983b), while in Cyrnatopleura, the movement is very sensitive to illumination (Pickett-Heaps, 1984a, 1991). The cytoplasm around the MC becomes linearly striated, and straight protrusions project from it into the vacuole, clearly revealing the MT activity. Once the nucleus has reached its appointed position and the prophase spindle grows (Figs. la-d), the MC invariably shrinks and disappears (Lauterborn, 1896);this can be followed in uiuo (e.g., Pickett-Heaps, 1983b). Later, new MCs appear adjacent to the spindle polar structures as mitosis goes to completion. In Surirella, they arise as small coalescing granules immediately adjacent to the polar plate which detaches from the telophase spindle (Figs. 5a, b; Lauterborn, 18%; Tippit and Pickett-Heaps, 1977). In other diatoms, the morphology of the forming MCs undergoes remarkable changes. For example, in Pinnularia, the new MCs first appear by prometaphase in the form of fibrillar, fan-shaped structures on each polar plate (Fig. 14; Pickett-Heaps er al., 1978a, b). One must follow the cells into valve initiation to demonstrate that these are indeed transformed in the new MC in the daughter cells (Pickett-Heaps et al., 1979a,b). In Hantzschia, the MCs arise from spiderlike lamellae near the telophase nucleus and again, following the cells into valve morphogenesis, confirms their transformation into the new MCs (Pickett-Heaps and Kowalski, 1981). In summary, the MC is intimately involved with the creation of the spindle, after which it dissolves. In turn, new MCs arise close to the spindle poles. This sequence is diagrammatically summarized in Tippit and Pickett-Heaps (1977). Whether the material of each is interconvertible is not clear; both appear to be complex MTOCs, but the form of each is variable bet ween different diatoms and pronounced, predictable morphological changes may occur in the MC in individual species. The MC is not directly equivalent behaviorly to the centrosome (with or without cen-

CELL DIVISION IN DIATOMS

69

trioles) characteristic of many conventional cells; centrosomes replicate and separate to form the poles of the mitotic spindle. Such a sequence of events has not yet been recorded for the MC in diatoms. C. THEMC AND VALVEMORPHOGENESIS Once reformed at telophase, the MC becomes the focus of an extensive MT cytoskeleton. The MCfMT complex now takes up a precise position with respect to the symmetry of the forming valve (Fig. 5b; in uiuo: Pickett-Heaps, 1980, 1982, 1983b; Pickett-Heaps et al., 1979a, 1988a). MTs from it appear centrally involved in various aspects of valve differentiation (particularly of the raphe or labiate processes); this aspect of diatoms is reviewed in Pickett-Heaps et al., (1991; see also Section X, B). D. De Nouo FORMATION OF BASAL BODIESDURING MEIOSIS Centric diatoms differentiate simple uniflagellated sperm for sexual reproduction (beautifully documented in uiuo by Drebes, 1%7). Centrioles are absent from the interphase cell (and from all pennates). Sexual differentiation in the male cells commences with a variable number (depending on the species) of mitotic cell divisions in the spermatogonia, eventually creating two, four, or eight masses of cytoplasm in the parental valves (e.g., Manton et al. 1969a,b, 1970a,b). Gametogenesis is completed as the singje nucleus in each mass undergoes one meiotic division without cleavage; four flagella are extruded and then the two nuclei divide meiotically again with a single flagellum defining the pole of the two spindles. Finally, the cytoplasm cleaves around the four nuclei, leaving each with the single flagellum. Interestingly, although this flagellum is normal in function, it does not possess the central two MTs found in other flagella (Manton and voni Stosch, 1966; Heath and Darley, 1972). Manton et al. (1969b, 1970b) have followed these events ultrastructurailly in Lithodesrnium. The basal bodies arise de nouo between meiosis I and I1 from a mass of dense material, termed the “paracentrosome,” accumulating at the poles of the first meiotic spindle; the poles of the second meiotic division are then significantly smaller than those of the first, as if some of their material has been “used up” (sic)during creation of the basal bodies. The polar plates of diatoms are thus ontogenetically related to the more diffuse, centric polar structures in animal cells types and also those plant cells (e.g., algae, lower land plants, etc.) that can create centrioles de nouo as flagella are required for reproduction.

70

JEREMY PICKEn-HEAPS

IV. The Central Spindle The most striking feature of the mitotic spindle in pennate diatoms-and the feature that, more than any other, appears to make the diatom mitosis unusual-is the massive “central spindle” (Lauterborn, 18%). Analysis of the structure, behavior, and function of this central spindle has proved particularly rewarding as many of its features are now known to apply to more conventional spindle types. It is extranuclear in origin, rapidly growing and sinking into the nucleus during prometaphase (Figs. l , 2); it defines the spindle axis, and its ends represent the poles; the chromosomes cluster around it, becoming stretched over it by metaphase and they later move over it during anaphase. The three attributes that make this central spindle so interesting to the microscopist are: (i) the high degree of structural order it displays; (ii) the clarity with which it can be seen in large pennates, where it is a brilliantly birefringent, rectangular bar or cylinder; and (iii) the clarity with which phenomena associated with anaphase A (chromosome-to-pole movement) can be distinguished from those involved in anaphase B (spindle elongation and separation of the poles). That anaphase consists of these two phenomena has been known for some time (Ris, 1949). The central spindle consists of two tightly interdigitated half spindles. In large pennates, it is either solid (e.g., Hantzschia: Figs. 11,21) or tubular (Pinnularia and Surirella: Figs. 2, 3, 10, 15). The appearance of chromosomes attached to these spindles can also vary, particularly when viewed with the light microscope in uiuo. In many cells, the chromosomes are stretched over the centrd spindle (e.g., Huntzschia: Fig. 13) while in others, the chromosomes at metaphase form a compact, donut-shaped mass around the center of the spindle (e.g., Surirella and Pinnularia: Fig. 3). In some centric diatoms, the prophase central spindle remains a single structure during mitosis (e.g., Melosira: Tippit et al., 1975; Stephanophyxis: Wordeman et al., 1986) while in other species, it splits up into discrete fibers during prometaphase. These latter spindles are more conventional in appearance but the fibers maintain the structural organization of the single central spindle (Manton et al., 1%9a,b, 1970a,b; PickettHeaps et al., 1988b). Smaller but probably equivalent central spindles have been reported elsewhere, although their detailed organization is not usually resolved. The persistent extranuclear spindle in Dientoamoeba displays square close-packed MTs (see below) like those in diatoms (Camp et ul., 1974), while various protozoa such as Trichomonas (Brugerolle, 1975), acantharia (Febvre, 1977), and dinoflagellates (Kubai and Ris, 1969; Oakley and Dodge, 1976;Ris and Kubai, 1974)generally display intranuclear or extra-

CELL DIVISION IN DIATOMS

71

nucl.ear central spindles. These are quite massive in the large protozoa ~ar~~~anym and p h~ a~ i c ~ o n (Kubai, y m ~ ~1973; a Inoue and Ritter, 1978; Rittcr et al., 1978). The extensive literature on fungal mitosis reveals numerous MT bundles that could be equated with the central spindle (reviewed by Heath, 1976). In more conventional spindles, the large number (ofMTs and their more irregular arrangement makes structural analysis almost impossible. However, at telophase, the appearance of the spindle indicates that it, too, is composed of two half spindles with the interdigitation forming the midbody. This conclusion was drawn from counts of MT:; (McIntosh and Landis, 1971; Brinkley and Cartwright, 1971), although the presence of pole-to-pole MTs could not be ruled out. Generalizing, the diatom central spindle probably shares its major features with many other types of cells. BEFORE PROPHASE A. THEPRIMORDIAL SPINDLE

In the pennate species most closely examined, the primordial spindle arises immediately adjacent to the MC, very early in prophase (Fig. 4a; Lauterborn 1896). In some species, it appears to be present throughout interphase and so it was first called the “Persistent Polar Complex” in Diatorna (Pickett-Heaps et al., 1975). A similar tiny bipartite spindle precursor was found in Hantzschia long before mitosis (Figs. 53, 54 in Pickett-Heaps and Kowalski, 1981), and Manton et al. (1969a) imply it is present during interphase in Lithodesrnium. Whether this is the norm is undetermined. The primordial spindle in preprophase pennate diatoms is a complex of symmetrically disposed layers traversed by fine cross striations, almost always with one or more prominent “polar vacuoles” embedded in each outer surface (Figs. 4a,b, 8b; Diarorna: Pickett-Heaps et al., 1975; Surirella: Tippit and Pickett-Heaps, 1977; Pickett-Heaps et al., 1984a; Pinrzularia: Pickett-Heaps et al., 1978a; Hantzschia: Pickett-Heaps er al., 1980b; Stephanophyxis: Wordeman et al., 1986). In centrics, while the primordial spindle is similar (Lithodesmium: Manton et al., 1969a; Stephanophyxis: Wordeman et al., 1986), its relationship with the MC is not as well documented except in Mefosira (Tippit et al., 1975). B. POLARPLATES AND ANCILLARY FEATURES: CHANGES DURING PROPHASE The polar plates and ancillary features vary in morphology between species and are their most elaborate during prophase. The polar vacuoles diminish in prominence and are gone by metaphase. MT assembly is

72

JEREMY PICKETT-HEAPS

initiated at the edges of the plates, pushing them apart (Fig. 4b) while briefly deforming them (e.g., Pickett-Heaps et al., 1980b; Pickett-Heaps et al., 1984b). The MTs terminate just under the plate (as in Fig. 8a). In Hantzschia (Pickett-Heaps ef a/., 1980b),each MT ends in a tiny granule which may be its nucleating center, the granules apparently being layered by the plate. Tippit et al. (1978) and McIntosh et al. (1979) discuss where tubulin subunits might be added at this stage since MT elongation and sliding are probably occurring concurrently. The outer layers of the primordial spindle transform into the polar plate that caps each end of the central spindle by metaphase. This plate may become a ring in cylindrical spindles (e.g., Surirella: Fig. 5b). There may be other changes around the poles; for example, large Surirellas display a mass of membranous material capping each pole which disperses during mitosis (Pickett-Heaps e f al., 1984a). At telophase, the polar plateslrings in some pennates detach from the central spindle (beautifully documented by Lauterborn) and are briefly free in the cytoplasm before they disperse (Fig. 5b; Tippit and Pickett-Heaps, 1977). C. THEPROPHASE/PROMETAPHASE TRANSITION The development of the early prophase spindle is quite slow and delicate, often ceasing if the cell is handled and/or illuminated for cinematography. As in other cells, breakdown of the nuclear envelope triggers the transition from prophase to prometaphase, marked by rapid spindle growth in uiuo. Prophase spindles in Hantzschia are brilliantly birefringent rectangles (Fig. 7a; Pickett-Heaps et al., 1980a); during prometaphase, these elongate rapidly and double their length in minutes (Figs. 6, 7a-d; Pickett-Heaps, 1982). The formation of the cylindrical central spindle in Surirella and Pinnularia is no less dramatic (Figs. 2a-c; Pickett-Heaps, 1980, 1983b).Central spindle formation is also rapid in large centrics (e.g., Odontellu: Pickett-Heaps, film in preparation). Ultrastructurally, at the end of prophase, the central spindle consists of tightly packed MTs that extend between the separating polar plates (Figs. 4b, 8a). Numerous other polar MTs radiate outwards from the edges of the plates over the surface of the nucleus (Fig. 8b). By early prometaphase, the elongating spindle is transformed into two interdigitated half spindles (Figs. 7a-d) while the other polar MTs penetrate extensively among the chromosomes (Fig. 12a). D. THECENTRAL OVERLAP: INTERACTION AND PACKING OF OPPOSITELY POLARIZED MICROTUBULES (MTs)

The high degree of organization of the central spindle was first revealed by Manton et al. (1969a, b, 1970a, b) in the square/hexagonal packing of its

CELL DIVISION IN DIATOMS

73

FIGS. 6 and 7. Early mitosis, spindle visualized with polarization optics, Hanrzschia amphioxys. Fig. 6. Elongating prometaphase spindle in enlarged central gap in chloroplasts.

Bar = 10 pm. Fig. 7a. Prophase spindle. Fig. 7b. Initiation of rapid growth. Figs. 7c,d. Appearence of overlap (arrows)in the next few minutes. Bar = 10 pm. Reproduced with pennission from Pickett-Heaps et al. (1980a).

clusters of MTs. The forming spindle in Melosira (Tippit et al., 1975) vivitdly demonstrated this order in a sequence of serial sections (Figs. 9a-d) which showed: (i) the MTs arranged in precise square close packing in the central overlap of the two half spindles (Fig. 9c); and (ii) as the sections moved out of the overlap into each half spindle, alternate MTs along any row dropped out of the square array (Figs. 9a, b, d). The way the MTs drop out of the pattern demonstrate that in the ovwlap, each MT in one half spindle has as its four closest neighbors, four Ml's in the other half spindle. If the MTs in each half spindle are of uniform molecular polarity (Section V,B), this arrangement suggests that during prophase, MTs growing from each polar plate pack so as to maximize lateral interaction between oppositely polarized MTs. In other diatom spiindles, the domains of the square packing are interrupted by slight disllocations and skewing of the lattices (Figs. 10 and 11). When these observations were made, there was considerable theoretical interest in spindle MT polarity, stemming particularly from the model of mitosis proposed by McIntosh et al. (1969). The parallel arrangement of its MTs lends the central spindle to analysis by serial section reconstruction, an exercise that allows the unambiguous determination of spindle structure. From counts of MTs, Manton et al.

FIG. 8. Two of a set of serial sections, prophase spindle of Hantzschia arnphioxys. (a) Close-packedMTs run between polar plates (n, nucleus). (b) At edge of spindle, numerous long MTs (arrows) radiate out over surface of nucleus (v, polar vacuole). Bar = 0.5 pm. Reproduced with permission from Pickett-Heaps et al. (1980b).

74

FIGS.9-1 1. Four of a set of serial sections through prophase spindle of Melosira. Fig. 9c. Through middle of overlap region, showing MTs in precise square, close-packed array. Figs. 9a,b,d. Sections on either side of Fig. 9c, moving from overlap into the two half spindles. MTs drop out of the central square array alternately along any row (e.g., arrow). Bar = 0.5 pm. Reproduced with permission from Tippit et al. (1975). Fig. 10. Transverse section, overlap of the liollow central spindle of Pinnularia, showing domains of square (arrow) and hexagonal (i.e., skewed square) packed arrays of MTs. Bar = 1 pm. Reproduced with permission from Pickett-Heaps et al. (1978b). Fig. 1I . Same as for Fig. 10, but through overlap of Hantzschia amphioxys. Bar = 0.5 pm. Reproduced with permission from Pickett-Heaps et al. (1980b). 75

76

JEREMY PICKET"-HEAPS

(197Oa) surmised that some MTs run pole-to-pole and a few of these were detected by serial section reconstructions in Diatoma (McDonald et al., 1977). Counts of MTs in animal spindles (Brinkley and Cartwright, 1971; McIntosh and Landis, 1971) indicated the same conclusion. In the small central spindle of Fragiflaria (Tippit et af., 1978),MTs extend pole-to-pole at prophase but by metaphase, they have differentiated into two half spindles. No pole-to-pole MTs remain, so these when present are probably unimportant in spindle function. These apparent pole-to-pole MTs may also be a result of tracking errors, since occasionally MTs from one pole end immediately adjacent to MTs from the other (e.g., Figs. 14a-c in Tippit et al., 1980b). The tracking data in Fragiffaria confirmed the existence of square packing/lateral interaction of MTs from opposite poles although the pattern suffered local distortion, often to the extent of appearing hexagonal. McDonald et a f . (1979) analyzed MT spacing statistically in the Diatoma spindle; MTs from opposite poles were spaced on average 40 nm centerto-center, strong evidence for a preferred interaction of oppositely polarized (antiparallel) MTs. They also published stereo reconstructions of this central spindle (McIntosh et a f . , 1979). Similar parallel/antiparallel MT interactions have since been reported elsewhere. The continuous spindle fibers in Ochromonas are composed of overlapping MTs (Tippit et a f . , 1980b) displaying this lateral interaction (Tippit et al., 1983). The central spindle of Dictyostefium (Roos and Camenzind, 1981)and Puccinia are composed of two half spindles with a preferred antiparallel MT spacing (McIntosh et al., 1985; Tippit et al., 1984).

-

V. Kinetochores Initially (e.g., Pickett-Heaps et af., 1975), we could not recognize kinetochores or kinetochore fibers in diatom spindles, a matter that Lauterborn (18%) also discussed. Manton et al. (1970a) mentioned merely that chromatin was associated with MTs. Later, kinetochores were identified in Pinnufaria (Pickett-Heaps et al., 1978a) and Hantzschia (Fig. 17; Pickett-Heaps et al., 1980b). The dimculty of identifying them stemmed from three characteristics of these spindles: (i) as in other organisms, kinetochores vary in structure and in many diatoms (particularlycentrics) they have no distinctively recognizable morphology; (ii) the nature of the kinetochore fiber was not as expected (i.e., with MTs terminating at a specific kinetochore structure); and iii) kinetochores, obvious in some diatoms on unattached or monopolar chromosomes (e.g., Figs. 12-14), become invisible when chromosomes are stretched over the metaphase spindle.

FIG.12. Early prometaphase spindle, Hanfzschia amphioxys. (a) MTs (t) from poles penetrating among chromosomes and beginning to interact laterally with kinetochores (arrow). Bar = 1 pm. (b,c) Lateral interaction of kinetochores with MTs from poles whose direction is indicated by arrows. Bar = 0.5 fim. (d) A chromosome having moved near pole with one kinetochore (black arrow) stretched along polar MTs, while other kinetochore bare (white arrow). Bar = 0.5 pm. (e) Lateral interaction of one kinetochore (large arrow) with MT, two other bare kinetochores nearby (small arrows). Bar = 0.5 pm. Reproduced with perniission from Tippit er al. (1980a).

FIG. 13. Midmetaphase, Huntzschiu umphioxys; chromosomes (ch) stretched tightly over central spindle with large overlap ( 0 ) . One chromosome still not yet bipolarly attached; kinetochore (k) laterally interacting with MTs from upper pole. Bar = 1 Fm. Reproduced with permission from Pickett-Heaps el al. (1980b).

78

FIG. 14. Same as for Fig. 13, but in Pinnularia whose less-stretched chromosomes form a donut-shaped mass (like Surirella: Figs. 3a-d) around the hollow central spindle (cf. Fig. 13). Dense collar material (large arrows) runs between attachment of chromosomes and poles. Fan-shaped features (small arrows) on polar caps will become the new MCs at telophase. Kinetochore (k) visible on unattached chromosome (ch). Bar = 1 pm. Reproduced with pennission from Pickett-Heaps et al. (1978a). 79

80

JEREMY PICKETT-HEAPS

Kinetochores in Hantzschia (Pickett-Heaps et al., 1980b) and Pinnularia (Pickett-Heaps et al., 1978a)consist of a single, densely stained layer attached to chromatin by fine fibers; a fan-shaped brush of other fine filaments extends outward (Fig. 17). During their initial interaction with spindle MTs (Tippit et al., 1980a), kinetochores are visible on the tips of stretching chromatin (Fig. 12d) but they are unrecognizable once the chromosomes are bipolarly attached and fully stretched over the spindle. The filaments from these kinetochores are probably analogous to the corona of some mammalian kinetochores (discussed in Section X,C). Two aspects of the diatom spindle are of considerable relevance to understanding kinetochore function, still a contentious, unresolved issue (Section X,C). A. BEHAVIOR OF KINETOCHORES: SLIDING ON POLAR MICROTUBULES DURING

PROMETAPHASE

Prometaphase activity in uiuo (Pickett-Heaps, 1980, 1982; PickettHeaps et al., 198Oa; Tippit et al., 1980a)in several large diatoms appeared at first unusual and unconventional. As the growing central spindle sinks into the nucleus, chromosomes commence vigorous movement to either pole, led by a pointed projection that each chromosome develops; each then oscillates irregularly along invisible linear tracks focused on that pole while maintaining attachment to it. Soon, sometimes within seconds, each chromosome suddenly develops a second pointed end which moves to the other pole along another linear or slightly curved track; very quickly, the chromosome becomes stretched over the central spindle. Once in this bipolar attachment, the chromosomes become entirely quiescent until anaphase. In Surirella and Pinnularia, in which the chromosomes collect in a mass around the central spindle (Fig. 3a), this oscillating movement in the late-attaching chromosomes takes place over the surface of chromosomes already bipolarly attached (Figs. 1,2 in Pickett-Heaps et al., 1979b). This behavior in viuo, in conjunction with ultrastructural analysis (Section V,B), indicated that kinetochores are sliding over polar MTs during prometaphase. Thus they display active motility, and the prometaphase oscillations represent the mechanism by which each chromosome finds the second pole while maintaining attachment to the first. We proposed that this type of sliding activity is the norm in conventional spindles (Tippitt et al., 1980a). However, the importance of sliding in mitotic movement generally has not been taken seriously (see also Section X,B); this consensus is partly derived from the large number of published electron micrographs showing MTs terminating in kinetochores, an image evocative of MT nucleation (Section V,B) and not sliding activity.

CELL DIVISION IN DIATOMS

81

There are, in fact, two general issues related to this sliding activity, one very recently resolved, the other remaining contentious. Let us consider first the issue of initial chromosome interaction with the prometaphase spindle. It is now clear that in many conventional cells in uiuo, early proinetaphase chromosomes move initially to either pole and then oscillate just as diatom chromosomes do, although not as rapidly (e.g., Roos, 1976; Bajer, 1982; Church and Lin, 1985; Rieder et al., 1986). Furthermore, carc:ful examination of cells fixed at this stage reveals lateral association of kinetochores with spindle MTs even in conventional cell types (Section V,B). Only recently has the reality of this lateral association and its significance been unambiguously resolved by the elegant work of Rieder and Alexander (1990). They demonstrated that the initial movement of a chromosome poleward in uiuo in the newt spindle coincided with the lateral association of its kinetochore with a single astral MT that unequivocally extended past the kinetochore. Thus, as in diatoms, the MT does not have to terminate in the kinetochorefor the latter to moue. In a subsequent paper, Hayden et al. (1990) recorded the capture of an MT by a kinetochore and its subsequent sliding in uiuo,a remarkable technical achievement. Thus, kinetochores in conventional spindle types, as well as diatomis, can move poleward by sliding at prometaphase. This observation refiites the possibility proposed by Mitchison (1988) that MT disassembly by the kinetochore powers prometaphase movement (Section X,C). The second issue is this. Even if sliding occurs during prometaphase, is it relevant to anaphase? Discussion of this issue will be deferred to Section X,C, since it constitutes an important area of speculation at the time of writing. B. BEHAVIOR OF KINETOCHORES: THEORIGINOF KINETOCHORE FIBERS DURING PROMETAPHASE The observations of live diatoms described above were accompanied by an ultrastructural analysis of prometaphase cells. Two central points emerged. First, during spindle formation, all the MTs from one pole appeared equivalent; specifically, those MTs that formed the central spindle were indistinguishable from those that initially ran over the surface of the nucleus (Figs. 8a,b; Section IV,C) and soon penetrated among and interacted with the chromosomes (Fig. 12a). Thus, we predicted that all the MTs in each half spindle must be equivalent, and therefore had the same polarity (Pickett-Heaps et al., 1980b). Second, the prometaphase kinetochores were always laterally associated with polar MTs which extended past them (Figs. 12b-d); few if any MTs terminated in the kinetochore (Tippit et al., 1980b). These polar MTs must be the invisible linear tracks

82

JEREMY PICKETT-HEAPS

followed by the kinetochore (i.e., the pointed end of chromosomes) in uiuo. Thus, a primary function of the kinetochore is to laterally associate with polar MTs, capturing and sliding over them, and we suggested that this interpretation is also valid for kinetochores generally (Pickett-Heaps and Tippit, 1978; Pickett-Heaps et al., 1982). When first made, this suggestion was in direct conflict with the then current consensus of opinion derived from interpretations of ultrastructural and biochemical data. In the early literature of mitosis, two likely alternatives (summarized by Schrader, 1953) could not be resolved: that kinetochores either form their own fibers of else they attach to fibers originating from the pole. But by the 1970s, opinion firmly favored the concept that the kinetochore is an MTOC, nucleating the MTs of the kinetochore fiber (see any review on mitosis at that time). This consensus (reviewed by Brinkley et al., 1989) stemmed mostly from: (i) the observation that MTs terminate in kinetochore; (ii) in uitro work demonstrating MT growth from kinetochores (e.g., Gould and Borisy, 1978; McGill and Brinkley, 1975; Pepper and Brinkley, 1979; Snyder and McIntosh, 1975; Telzer et al., 1975); (iu) recovery experiments in which cells released from anti-MT blocks showed ‘‘nucleation’’ of MTs at kinetochores (e.g., DeBrabender et al., 1979, 1980, 1981; Ris and Witt, 1981; Witt e? al., 1980); and finally (iv) the report that both poles and kinetochores generate MTs with the fast growing, [+I end distal (Bergen et al., 1980). Thus, the half spindle would contain antiparallel MTs; MTs from kinetochores and poles would be in opposing polarity, an arrangement which suggested the basis of anaphase chromosome movement by sliding (e.g., as proposed by McIntosh et al., 1969). Only in a few cases (e.g., Weisenberg and Rosenfeld, 1975) was no such in uitro nucleating activity of kinetochores reported. This matter was (somewhat unexpectedly) resolved by the development of two different labeling techniques that were able to reveal the intrinsic molecular polarity of MTs (Heidemann et al., 1980; Telzer and Haimo, 1981). Both methods demonstrated that the MTs of each half spindle of conventional spindles are predominantly of one polarity (Eutenauer and McIntosh, 1980, 1981; Telzer and Haimo, 1981). The status of the earlier biochemical data now is in doubt; the results can be explained with hindsight as a consequence of adjusting the experimental conditions (e.g., tubulin concentrations) to force the kinetochores to act in a way not representative of their behavior in uiuo. Whatever the explanation for the in uitro results, the polarity data forced a reexamination of opinion regarding the origin of the kinetochore fiber. Ultrastructurally, several papers showed lateral association of kinetochores and MTs (e.g., Fig. 8 in Roos, 1973a; Figs. 5, 6 in Roos, 1973b;

CELL DIVISION IN DIATOMS

83

Nic:klas et al., 1979), and the possibility that kinetochores capture polar MTs was apparent from and/or discussed in several papers (Kubai, 1973; Rieder and Borisy, 1981; Church and Lin, 1982; Rieder, 1982). Current corisensus now firmly favors the view that kinetochores capture polar MTs (e.g., Salmon, 1989); capture in uitro has been carefully explored by Milchison and Kirschner (1985a,b) and Huitorel and Kirschner (1988). The behavior of kinetochores is complex since after capture, incorporation of tubulin into these MTs can occur at this captured end during prometaphase, metaphase (Mitchison et ul., 1986; Geuens et al., 1989; Gorbsky and Borisy, 1989), and even early anaphase (Wadsworth et al., 1989).

M. The “Collar”: A Spindle Matrix? A. MORPHOLOGICAL EVIDENCE FOR A SECOND COMPONENT ATTACHINGCHROMOSOMES TO THE SPINDLE A N D POLES

By metaphase, chromosomes in diatoms are either stretched over the central spindle (Fig. 13) or grouped in a compact donut-shaped mass around the center of the central spindle (Figs. 3a,14); even in the latter case, some chromatin is stretched to the pole (Fig. 14). These stretched regions did not look like the kinetochore/MT attachment of conventional spindles. As in prometaphase, polar MTs, closely associated with the chromatin, ran past the point where the kinetochores were presumed to lie, out into the nucleus and cytoplasm. On careful examination, the outer Ml’s of the central spindle were seen to be permeated by an ill-defined but distinctly recognizable matrix (Fig. 14) we called the “collar” and it is most clearly revealed in transverse sections of the spindle (Figs. 15a,b). Provocatively, it extended precisely from the poles to the leading points of attachment of the chromosomes (Tippit and Pickett-Heaps, 1977; PickettHeaps et ul., 1978a,b). In later experiments, metaphase spindles of Huntzschiu were treated with colchicine, which disassembles the polar MTs associated with stretched bipolar chromosomes while leaving the very stable central spindle intact. When these cells were treated with cochicine in uiuo, the doubled chromosomes release irregularly, springing to either pole (Pickett-Heaps, 1983~).In similarly treated cells examined under the electron microscope, the collar becomes visible as a cloud around the poles (Fig. 16a) with numerous kinetochores (Fig. 17) enmeshed in it (Fig. 16b; Pickett-Heaps and Spurck, 1982a). Thus, the kinetochores which had become invisible where stretched (Section V) reappeared when tension in the: chromosomes was released while they apparently maintained their connection to the collar. (The corona of the kinetochore in PtK cells

FIG. 15. Transverse sections of Pinnularia spindle. (a) The region of chromosome attachment (other chromosomes are present out of the plane of this section, at bottom). The ring is the collar material. Bar = 1 pm. (b) Section between chromosomes and pole; matrix of collar material permeates outer MTs of central spindle. Bar = 1 pm. Reproduced with permission from Pickett-Heaps er al. (1978b).

84

FIG. 16 and 17. Metaphase spindle offfantzschiaamphioxys treated with O.l%colchicine (wlv) for 20 minutes before fixation. Fig. 16a. Polar MTs have disappeared while central spindle remained intact. As polar MTs disassemble, stretched bipolar chromosomes release irregularly and elastically spring to either pole. At poles, collar matrix (large arrows) associatcd with kinetochores (small arrows) have both become visible when tension is removed (cf Fig. 13). Bar = 2 pm. Fig. 16b. Detail from another section of this spindle with two kinetochores (arrows) associated with collar material. Bar = 1 pm. Reproduced with permission from Pickett-Heaps and Spurck (1982a). Fig. 17. High magnification image of unattached kinetochore (i.e., one not under tension) displaying typical brush of coronalike filaments. Rr,r = 0 5 urn.

86

JEREMY PICKETT-HEAPS

behaves similarly, disappearing in stretched kinetochores and reappearing when the tension is released-Section X,C). A further observation may be linked to the presence of this component. At telophase, the chromosomes move so as to cluster tightly over the ends of the central spindle (Pickett-Heaps et al., 1980b); thus, they move pas? the ends of the spindle MTs (Fig. 20). This movement appears to be a continuation of anaphase with the chromosomes drawn together, presumably by the collar since there are no MTs nearby. I believe that an equivalent movement of chromosomes also happens in conventional spindles at telophase when the centrioles end up on one side of the reforming nucleus with the midbody MTs on the other. The possibility that material equivalent to the collar might be present in conventional spindles has been discussed by Pickett-Heaps et al. (1984b). Special fixation protocols used on the green alga Oedogonium showed fine filaments attached to kinetochores and extending poleward among kinetochore MTs, with both filaments and MTs embedded in an amorphous matrix (Schibler and Pickett-Heaps, 1980); the filaments terminated in kinetochores before and after MT capture, and so could represent an extended, highly visible corona that interacted with the surface of polar MTs. However, while some matrix is occasionally detected, for example, in cold-treated spindles (Rieder, 1981), there has been no evidence of the existence of collar material or these filaments in most conventional spindles, even when specifically sought (Fuge, 1981). Whether the absence of a collar matrix is due to fixation/straining problems or represents a real difference between diatoms and conventional spindles is unknown. The issue may be central if there does exist a second component involved in moving chromosomes poleward, as has also been argued by a few other workers on the basis of other experimental data (Forer, 1966, 1988).

B. THEPHYSIOLOGY OF ANAPHASE A During normal mitosis, metaphase chromosomes stretched over the central spindle appear to spring poleward at the onset of anaphase. That there is elastic energy stored in the metaphase spindle was shown by cutting it partially or wholly with a microbeam, whereupon it immediately buckled or collapsed (Leslie and Pickett-Heaps, 1983). This does not happen once the chromatids split at anaphase. Thus, the metaphase central spindle is apparently being compressed by the stretched chromosomes/collar material running between the poles. To probe the physiology of chromosome movement and where this elastic energy comes from, live diatoms were treated with various metabolic inhibitors so as to deplete the cells’ ATP levels and energy reserves.

87

CELL DIVISION IN DIATOMS

The inhibitors included dinitrophenol (DNP), cyanide, azide, rotenone, antimycin A, and oligiomycin (Pickett-Heapsand Spurck, 1982b). All had the same cytological effects, although the cells’ ability to recover reversibly varied; DNP, the inhibitor of choice, could be used up to six times to stop mitosis which still proceeded to completion normally as long as the cells were not left longer than 5-7 minutes in the DNP each time. Upon treatment with the inhibitors, all cellular activity, including cleavage, ceased within about 45 seconds; it restarted within about 1 minute of warnshout (Pickett-Heaps, 1983~).The threshold of effectiveness was marked, DNP working at 1 mM but almost without effect at 0.1 mM. These inhibitors had little effect on metaphase spindles, while anaphase spindle elongation ceased (cf. Section VI1,B). The most interesting result was unexpected and at first, made no sense: oscillating prometaphase chromosomes always moved up to the pole, clustering there as the cell became energy depleted. Upon removal of the drug, the cell initiated recovery and prometaphase oscillations with a marked movement of clustered chromosomes away from the poles. These observations (summarized in Fig. 18) were interpreted to show: (i) movement to the poles (P-movement)is different than movement away from the poles (AP move-

+

Ronwtapharc

Mtaphar.

Anrph.w

FIG. 18. Diagrammatic summary of spindle structure, chromosome attachment, and mobement in diatoms. The MTs in each half spindle are of uniform polarity (arrowheads, prornetaphase). During initial chromosome attachment, kinetochores slide toward and become attached to poles (P, poleward force). An ATP-requiring system generates AP (antipoleward) force, moving chromosomes irregularly and elastically away from pole, creating prornetaphase oscillations. Bipolar attachment occurs when second kinetochore encounters MT.5from opposite pole; movement of this second kinetochore poleward creates stretched, stable metaphase configuration. Chromatid splitting initiates anaphase A movement from energy stored elastically in chromosome attachments, and anaphase B by sliding of half spindles apart. Reproduced with permission from Pickett-Heaps et al. (1982).

88

JEREMY PICKETT-HEAPS

ment); and (ii) AP movement requires ATP while P movement does not (or else requires a lower level of ATP). Paradoxically, therefore, the most obvious ATP-requiring motility system detected by these experiments moves chromosomes away from the pole. Mitchison and Kirschner (1985b) recorded a similar result in uifro when ATP addition to their model system appeared to cause kinetochore movement toward the [+I end of the MT. We found it easiest to explain these results on the basis that for monopolar chromosomes, one kinetochore becomes attached elastically to the pole (presumably involving the collar material), and that energy is fed into the spindle by stretching the attachment away from that pole to the other. One test of this hypothesis is that DNP treatment should not stop anaphase A. In Hanfzschia,poleward movement is initially so rapid that it is finished before the DNP, introduced as the chromatids separated, can become effective. But in a large Surirefla, where the chromosomes are arranged in two donut-shaped masses around the central spindle (Fig. 3a), DNP treatment did not prevent anaphase A while cellular activity ceased (Cohn and Pickett-Heaps, 1988). These results (Fig. 18 and Pickett-Heaps e f al., 1982b; Pickett-Heaps, 1986, 1987) suggested an unusual view of mitosis in which the metaphase spindle is elastically loaded, and that anaphase A results from the release of this energy upon chromatid splitting. This concept also suggests why the traction system powering anaphase A, long sought in vain by cell physiologists, has proved so elusive-it may not exist. To test this model more generally, we tried equivalent experiments with animal cells. At early prometaphase, PtK cells responded to DNPjust like diatoms (Spurck et al., 1986a)with chromosomes clustering tightly around the poles (Pickett-Heaps, 1984b). This result was as expected if the mechanism of initial chromosome interaction with polar fibers is the same as in diatoms (Section V,B). However, when anaphase cells were treated with DNP, chromosome movement ceased completely (and reversibly), as also reported by Hepler and Palevitz (1986). Thus, it appeared at first that the anaphase traction system in diatoms might be different than that in conventional cells. However, there is a major structural difference between the two spindle types. While MTs do not terminate in kinetochores of diatoms, they do in kinetochores of conventional spindles and it is well established that these MTs have to shorten to allow anaphase A to proceed (Section X,B; Salmon, 1989). We therefore tested the possibility that it is MT disassembly at the kinetochore, rather than an anaphase traction system, that requires the ATP needed for anaphase A (Spurck ef al., 1986b). We found that: (i) MTs themselves are sensitive to ATP. MT cytoskeletons break down quite rapidly in ATP. Interestingly, these same cytoskeletons

CELL DIVISION IN DIATOMS

89

proved not to be sensitive to nocodazole; neither do anti-MT drugs break down MTs in ATP-depleted cells in uiuo, an unexpected result that confinns that ATP is needed for MT disassembly (even in the presence of anti-MT drugs), as with the diatom half spindle (Hinz et al., 1986). This result also suggests that spindle MTs are not in a simple dynamic equilibrium with tubulin (InouC and Sato, 1%7). (ii) In a permeabilized cell system, ATP or certain analogues such as ATP-y-S were needed for anaphase A. However, MT disassembly conditions alone (low temperatures or increased calcium levels) also promoted anaphase A. Thus,in PtK ce!Us too, anaphase A could apparently proceed from energy stored in the spindle but MT disassembly is necessary for this stored energy to be released (Section X,C). ‘These experiments support what we believe is a consistent theme for explaining anaphase A. However, this model based on diatom spindles does not currently enjoy much favor, a matter discussed in Section X. ,One final point needs mentioning here. AP-movement seems to occur over the same linear tracks that guides P movement and from the appearance of live spindles, we suspected that movement may have been brought about by the second kinetochore sliding away from the pole (PickettHeaps et al., 1982). However, Reider et al., (1986) have suggested that AP movement is due to ejection forces generated by growth of MTs from the poles. This theory has several attractive features, including its ability to explain an important property of the spindle: that the force it exerts on a chromosome is proportional to the distance of that chromosome from the pole (Hays et al., 1982). This property is what moves monopolar chromosomes from the pole to the metaphase plate as they achieve bipolar attachment. While this behavior is consistent with the chromosomal attachment being elastic, the proposition by Reider et al. (1986) also explains this proportionality very well, since the closer to the pole any chromatid is, the more polar MTs, and hence the greater the ejection force, acting on it (Salmon, 1989).

VII. Spindle Elongation (Anaphase B) by Sliding of Half Spindles

A. ULTRASTRUCTURAL AND in Viuo OBSERVATIONS From inspection of the first central spindles sectioned longitudinally (Pickett-Heaps et al., 1975), spindle elongation (anaphase B) could obviously be explained on the basis of the two half spindles sliding apart (Figs. 13,19,20,21 ,a,b). This interpretation, apparent in all subsequent electron microscopy (e.g., Pickett-Heaps et al., 1980b), has been confirmed by

90

JEREMY PICKETT-HEAPS

FIGS. 19 and 20. Late mitosis, Hantzschia amphioxys. Fig. 19. Kinetochores at poles; elongating spindle has diminished overlap ( 0 ; cf. Figs. 2, 13,211. MTs clumping in half spindles. Bar = 5 fim. Fig. 20. Overlap gone as anaphase B completed. Chromosomes have movedpast the ends of the central spindle, between sets ofarrows. Bar = 5 pm. Reproduced with permission from Pickett-Heaps et al. (1980b).

CELL DIVISION IN DIATOMS

91

1:IG. 21. Anaphase B (spindle sliding)in live cells of Hantzschia amphioxys, polarization opt~cs.(a) Metaphase,prominent overlap (0).(b) Elongated spindle, overlapalmostgone and eloilgation equal to sum of half spindle lengths. Bar = 5 pm. Reproduced with permission from Pickett-Heapset al. (1980a).

delailed analysis and reconstruction of central spindles (Tippit et al., 1978; McIntosh et ul., 1979). Sliding is easy to follow in uiuo (Figs. 21a,b; Pickett-Heaps, 1980,1982,1983b)and the extent ofelongation matches the size of overlap at metaphase (Pickett-Heaps et al., 1979b, 1980a). Since the packing arrangement in the central overlap maximizes interaction of MTs from opposite poles (Section IV,D), the force driving this separation could clearly be generated in the overlap by a motility system thart translocates one set of MTs against the other, similar to what happens in flagella with translocation of MTs mediated by dynein. Schulz and Jarosch (1980) have suggested that the movement could be generated by contrarotating sets of interacting MTs, but there is no evidence yet to support this possibility. ‘To demonstrate that the two half spindles are pushed apart by force generated in their overlap and not at the poles, a UV-microbeam was used to sever one pole from a Hantzschia spindle. Although the chromosomes collapsed around the spindle, later it still underwent elongation during aniiphase, with the shortened half spindle ejected from the other uncut half spindle (Leslie and Pickett-Heaps, 1983). With a more versatile microbeam, one can cut both half spindles, and again, the isolated overlap repjon, now floating free, underwent elongation as normal (Figs. 22a-i; Stonington et al., 1989).

FIG.22. Double cuts with UV-microbeam to demonstrate spindle sliding in overlap, Hantzschia amphioxys. Taken from frames of videodisc recorder [frame no.-to left; time in seconds-bottom left, zeroed at (b)]. (a) Anaphase spindle positioned for UV-microirradiation; photographic graticule (arrows) indicates site of future. cut. (b) First 3-second cut, time zeroed at 0 seconds. (c) Spindle immediately repositioned with respect to graticule. (d) Second cut completed. (e) Polar sections of cut half spindles disassembling poleward (arrows). (f) Polar segments almost gone, overlap free between chromosomes. (g,h) Elongation (arrows) as two remnants of the half spindles slide apart while cleavage furrow (f) approaches. (i) Disassembly of extruded remnants (cf. Fig. 23) as cleavage finished, 16 minutes after inadiation. Bar = 10 pm. Reproduced with permissionfrom Stonington er al. (1989).

CELL DIVISION IN DIATOMS

93

Is this type of spindle sliding typical of other dividing cells? This question remains unresolved. Animal cells show evidence of half spindles sliding apart (Saxton and McIntosh, 1987),but if they do, the sliding never proceeds to separation as in diatoms, and so the terminal stages of cell division are greatly prolonged. Other phenomena are also involved in ana.phase B. For example, microbeam experiments by Aist and Berns (1981) indicate that in the fungus Fusarium, the central spindle retards an elongation apparently generated by polar arrays of MTs. Also, half spindle elongation (accompanied and perhaps driven by MT assembly) may occur during anaphase B. Such elongation is seen to some extent in diatoms, particularly in centrics (Wordeman et al., 1986). In the fungus Puccinia (Tippit et al., 1984), the metaphase spindle elongates 3-fold, far too much to be explained on the basis of sliding of the two half spindles; the decrease in the overlap is accompanied by a considerable elongation in the dwindling number of overlapped MTs forming the remaining continuous fibers. Lik.ewise, rapid spindle elongation in Dictyostelium cannot be accounted for by sliding alone (Roos and Camenzind, 1981; McIntosh et al., 1985). Thus, in many cells, elongation of the half spindles contributes to anaphase B. Unlike animal cells where the midbody prolongs terminal stages of cleavage and MT sliding does not go to completion (Section VIII,A), smaller cells such as Ochromonas finish cell separation rapidly. Since the number of overlapped pole-to-pole fibers decreases steadily as they elongate (e.g., Tippit et al., 1980b, 1984), sliding is accompaniedby half spindle elongation in these cells, but the MTs of the fibers still rapidly disassemble once they have separated completely (Section VII1,A). As for the site of subunit addition, we (Pickett-Heaps et al., 1986) and some others (Brinkley and Cartwright, 1971; Manton et al., 1970b) propo5ied that the tubulin might be added at the [-] polar ends, while others (e.1:. McIntosh et al., 1985)argue for addition at the [+I end. There is now good experimental evidence that tubulin is in fact added at the [ +] ends in u i w (Saxton and McIntosh, 1987) and in uitro (Section VI1,B). B. PHYSIOLOGY OF ANAPHASE B Cande and co-workers (summarized by Cande, 1989) have used the central spindle isolated from the centric Stephanophyxis as an elegant in uitro system for reactivating the analyzing this elongation. There isolation prccedure is described in Cande and McDonald (1986) and active sliding apart of the half spindles was induced and followed by video microscopy (Baskin and Cande, 1988) upon addition of ATP in an appropriate buffer. Elongation is inhibited by vanadate and sulphydryl blocking reagents. In addition, they could mimic the further elongation of the half spindles that

94

JEREMY PICKE'IT-HEAPS

accompanies sliding in uiuo,by adding exogenous neurotubulin (Masuda and Cande, 1987). In the presence of ATP, the newly added tubulin is incorporated into the ends of MTs in the overlap, that end from which disassembly occurs in uiuo (SectionVII1,A). Thus, sliding is seen as the primary event in spindle elongation, augmented by MT assembly which does not, in itself, generate elongation. This system offers the strong possibility of allowing the molecules involved in these processes to be isolated and characterized (Cande, 1989). VIII. Control of Microtubule Stability

The MTs in late-division half spindles clump into coarse aggregates (Figs. 13,19,20, 21; Pickett-Heaps et al., 1978b. 1980a); since these are from the one pole, the lateral interactions that clump them must be different from the interactions in the overlap. Similar experimentally induced clumping of spindle MTs is reported by Rieder and Bajer (1977). SEPARATION OF THE HALFSPINDLES: STABILITY A N D DISASSEMBLY OF SP~NDLE MICROTUBULES

In large pennates, following anaphase spindle elongation, the spindle buckles as its two half spindles separate completely (Pickett-Heaps et al., 1979b; Pickett-Heaps 1980,1982). We originally thought that the cleavage furrow was breaking the central spindle (Pickett-Heaps et al., 1979b). However, if sliding was inhibited (e.g., inadvertently by illumination during filming), the cleavage furrow impinged upon the spindle without effect. Conversely, if cleavage was inhibited by cytochalasin, the central spindle still separated, buckled, and disassembled as normal (Pickett-Heaps, 1983~;Soranno and Pickett-Heaps, 1982). The central spindle at metaphase is very stable; for example, when released from a ruptured cell, it remains visible for many minutes. Once teleophase half spindles are separated in uiuo,there may be a brief pause, and then they shorten rapidly from the overlap end back to the pole (Fig. 23a-d; Pickett-Heaps, 1980, 1982; Soranno and Pickett-Heaps, 1982). Very soon after the spindle has disappeared, a new MT cytoskeleton arises which is involved in valve morphogenesis (Section 111,C). This pattern of spindle breakdown is significantly different than what happens in animal cells; sliding of half spindles (if it occurs) does not go to separation and the result is the midbody comprised of two sets of densely packed interdigitated MTs. The midbody is remarkably persistent and stable, prolonging the terminal stages of cell separation considerably (e.g.,

CELL DIVISION IN DIATOMS

95

1:IG. 23. Half spindle disassembly in Pinnularia, visualized by polarization optics; from 16 rnm film. (a) Half spindles separate (arrows),buckle. (b-d) Disappearence of birefringence from overlap edge (arrows) back to poles. Bar = 5 pm. Reproduced with permission from Soranno and Pickett-Heaps (1982).

Byers and Abramson, 1968).This delay in cell separation is not apparent in most unicellular organisms that also have central spindle fibers (Section VI1,A) so it might be unusual rather than the norm. Furthermore, tracking data on late-division spindles in Ochromonas (Tippit et al., 1980b) shows that in the rapidly dwindling continuous fibers, the remaining MTs are almost all overlapped; presumably they slide until they cease being overlapped, whereupon they rapidly disappear. ‘Two conclusions were drawn from this behavior of the half spindles, both now of general validity: (i) the [+I ends of the MTs are stabilized by their lateral association with MTs from the other pole in the overlap, but once removed from this configuration, they are able to rapidly disassemble; and (ii) since disassembly is unidirectional, the direction of disassembly is dependant on the intrinsic polarity of the constituent MTs (Soranno an,d Pickett-Heaps, 1982). This disassembly was inhibited reversibly by metabolic inhibitors and therefore, MT disassembly appears to need ATP (Hinz et al., 1986). *Sincethis unidirectional disassembly could be due to the polar end of the half spindles being capped in some way, we (Leslie and Pickett-Heaps, 1934)severed MTs in the half spindle by notching it with a UV-microbeam; this procedure creates new [+I and [-I ends at the two edges of the cut. Ellectron microscopy confirmed that MTs were severed by the microbeam. In all irradiations, the [ -1 end of MTs appeared stable over many minutes while birefringence disappeared poleward from the other edge; thus the [+] and [-I ends at the cut behave the same as the ends of the separated half spindle (summarized in Fig. 24). Furthermore, the rate of this loss of

JEREMY PICKETT-HEAPS

FIG. 24. Diagrammatic representation of control over spindle MT disassembly via intrinsic MT polarity. (a) In normal cells during late mitosis, sliding separates half spindles completely, whereupon MT disassembly proceeds unidirectionally from distal [+] end back to poles. (b) Half spindle has been notched by a UV-microbeam, creating new [+] and [-I ends at the notch. The new [-] ends of MTs remain stable, while unidirectional disassembly poleward occurs at the new [+] ends. Reproduced with permission from Leslie and PickettHeaps (1984).

birefringence from the [ +] ends increased during anaphase and telophase, indicating an overall transition towards MT disassembly conditions as the cell nears the end of mitosis. These results on diatoms were in conflict with the behavior of areas of reduced birefringence (ARBs) induced in crane fly spindles by Forer (1965) using UV-microbeams; he reported that his ARBs moved poleward at metaphase and anaphase (i.e., the severed fibers from the kinetochores and midregion of the spindle elongated while the polar cut fibers shortened). To investigate this dicrepancy, newt and R K spindles were irradiated with a UV-microbeam (Spurck et al., 1%). The effects were consistent with the observations on diatoms; the cut edge of the ARB nearest the kinetochores remained stable while birefringence rapidly faded from the

CELL DIVISION IN DIATOMS

97

other edge. This pattern of MT stability is also observed in mechanically cut spindles (Nicklas et al., 1989) and single MTs severed by a UVmicrobeam in uitro (Walker et al., 1989). It is not clear why our irradiation experiments gave results different from those of Forer and his collegues. A third, more indirect conclusion was drawn from these observations. InouC and Sat0 (1967) and others had suggested that kinetochore fibers shorten at the pole, pulling the chromosomes during anaphase. Once the MTs in the half spindle were known to be of uniform polarity, it was unclear how in this model the pole could discriminate between those MTs attached to kinetochores and those forming the central fibers, selectively shortening the former at their [-] end. The unidirectional disassembly of MI’Sobserved in the diatom half spindle suggested that kinetochore MTs of the same polarity in conventional spindles should behave similarly and disassemble from the [+I end back to the pole; if so, a major function of the kinetochore would be to disassemble MTs and thereby control the fate of its liber (Pickett-Heapsand Tippit, 1978; Pickett-Heaps et al., 1982). Forer and collegues (e.g., Forer, 1976; Schapp and Forer, 1984) have also proposed this alternative although some of their other observations with UV-microbeams indicate MT disassembly at the poles (Forer, 1965; see previous paragraph). The suggestion that the kinetochore disassembles MTs was, when first made, a provocative contrast to the kinetochore’s supposed role in nucleatirig MTs (Section V,B) but it could, for example, explain the experiments of McNeill and Berns (1981), which showed that kinetochores act individually in shortening their fiber. That the kinetochore does in fact disassemble MTs has been shown by the elegant work of Gorbsky et al. (1987,1988) and Mitchison et al. (1986). Many lines of evidence now suggest that the [+] end of MTs is where both diassembly and assembly predominantly occ:ur in uitro (Heidemann et at., 1980; Johnson and Borisy, 1977; Summers and Kirschner, 1979) and in uiuo (e.g., Soltys and Borisy, 1985; refs. cited in Cassimeris et al., 1988;reviewed by Salmon, 1989). The dynamics of MTs in both situations has turned out to be unexpectedly complex, involving the phenomenon of dynamic instability (Mitchison and Kirschner, 1984; Kirschner and Mitchison, 1986; Horio and Hotani, 1986; Sarnmak and Borisy, 1988; Cassimeris et al., 1988; reviewed by Salmon, 1989). Irubulin addition to elongating half spindles during anaphase seems to occur at the [+I ends of MTs (Section VI1,B); if so, the cell must have a subtle method of controlling MT dynamics, since this is also the site of MT disassembly once the half spindles separate. To accomplish this concurrent assembly/disassembly,McIntosh et al. (1985) suggest that the overlap contains a factor that associates only with antiparallel MTs and that biases the system toward MT assembly.

98

JEREMY PICKETT-HEAPS

IX. Cleavage Cleavage in centrics and pennates is rapid in uiuo. In large pennates, cleavage is initiated by prometaphase but halted by metaphase. During anaphase, it restarts and cleavage may cut the chloroplast (Hantzschia: Pickett-Heaps, 1982; Pickett-Heaps et al., 1980a; Surirella and Cymaropleura: refs. below). The positioning of the furrow is often asymmetric. It grows fastest distal to the spindle in some cells (e.g., Surirella: PickettHeaps, 1983b; Cymaropleura: Pickett-Heaps, 1984a). In flat centrics such as the very large Coscinodiscus wailesii, one can observe the furrow in face view, an unusual aspect; it grows in asymmetrically, this time fastest from between the daughter nuclei. Cleavage finishes near the opposite side of the cell while linear striations from near the nuclei extend across the ingrowing membrane (Pickett-Heaps, film in preparation). In all pennates examined, the ingrowing edge of the furrow is lined by a restricted zone of fine filaments (e.g., Pickett-Heaps et al., 1975 and others). The sensitivity of cleavage to cytochalasin (Pickett-Heaps, 1983c; Pickett-Heaps and Spurck, 1982a) supports their identification as actin filaments. Similar filaments probably line the cleavage furrow of centrics, but they appear less well preserved.

x. summary In this review, I have stressed how the diatom spindle has clarified various aspects of mitosis although several inferences derived from these cells (e.g., uniform polarity of MTs in the half spindle, role of kinetochores in capturing polar MTs, ability of kinetochores to slide over the surface of MTs, role of kinetochores in disassembling, rather than assembling MTs, etc.) at first conflicted with ideas prevailing at the time. There remains another instance where the diatom spindle suggests a scenario that at present commands little support, the mechanism of anaphase A. Before discussing this matter, it is appropriate to reflect on the differences that may exist in mitosis between diatoms and more conventional cells. A. SIGNIFICANT DIFFERENCES BETWEEN MITOSISIN DIATOMS AND CONVENTIONAL SPINDLE TYPES

Several unusual features of mitosis in diatoms might be related. One such feature is that MTs do not terminate in kinetochores. The closest relatives to the diatoms are the chrysophytes (some authorities include

CELL DIVISION IN DIATOMS

99

diatoms in the division Chrysophycophyta: e.g., Bold and Wynne, 1978). Since chrysophytes appear to have MTs terminating in kinetochores like all other cells examined, I conclude that the situation in diatoms must be derived, not primitive. Once the feverish activity of prometaphase is over, bipolarly attached chromosomes in diatoms are immobile. In some conventional spindles, however, metaphase chromosomes move about a great deal; in PtK cells, all ,chromosomesexhibit individual, almost rhythmic oscillations about the metaphase plate and these stop quite abruptly about halfway through anaphase A (e.g., see movie by Pickett-Heaps, 1984b). Newt cells are similar, with these oscillations also continuing briefly into anaphase (Bajer, 1982; Rieder et al., 1986). The metaphase oscillations could reflect MT dynamics even when the MTs are stabilized by insertion at the kinetochore (reviewed by Salmon, 1989; Cassimeris et al., 1990). There is almost no evidence that supports the existence of the collar material (Section V1,A) in conventional cells. Perhaps the significant difference is that the diatom’s traction system (possibly the collar/corona) aided by extensible chromosomes, attaches the metaphase kinetochores directly to the poles (Sections X,B; X,C). Polar MTs maintain the integrity of this connection (since MT disassembly results in eventual chromosome release: Pickett-Heaps and Spurck, 1982a; Pickett-Heaps, 1983c)but MT dyriamics do not directly affect metaphase chromosome behavior. This possibility is supported by microbeam experiments: when the central spindle is severed, the poles collapse together (Leslie and Pickett-Heaps, 1983); the elastic linkage between poles and chromosomes must remain intact when the MTs are cut. In contrast, this collapse does not happen when the half spindle of conventional cells is similarly cut (Spurck et at., 19910).

B. EVOLUTION OF THE DIATOM SPINDLE On occasion, the spindle of diatoms has been termed “primitive” as an explanation of its apparently unusual characteristics (Section I). In contrast, I believe that the spindle of pennate diatoms represents a highly evolved mitotic system and argue that most, and perhaps all, of its unusual features are equivalent to those in more conventional spindles. There are several reasons I draw this conclusion. First, diatoms as a group display a remarkable range of cytoplasmic systems involved in morphogenesis of their valves (reviewed in PickettHeaps et al., 1991). Different species use various organelles (including MTs, microfilaments, mitochondria, and other poorly defined cytoplasmic entities) in conjuction with a variety of cellular activities (e.g., controlled

100

JEREMY PICKETT-HEAPS

plasmolysis, interaction of daughter cells, localized secretion of mucilage, site-specificadhesionsto the parental valves, etc.) in various combinations to create the marvelously diverse morphology characteristic of their thousands of species. Thus, they display a virtuosity in the evolution of their cytoplasmic abilities unmatched in any other group of organisms. An example of this rapid evolution of cytoplasmic systems is suggested by comparison of the closely related Cymatopleura and Surirella. During cell division in both genera, the MT/MC cytoskeleton moves the premitotic nucleus to the end of the cell, and later the daughter nuclei back to the center of the cell &er cytokinesis (Section 111,B).Next, the single chloroplast undergoes precisely orchestrated morphogenetic movement. One end curls and moves over the valve until it sandwiches the central nucleus; then the fold is stretched back until both valve surfaces are covered by lobes of the chloroplast. In Cymatopleura (Pickett-Heaps, 1991),the initial folding takes about 7-10 minutes while the reverse stretching takes about 30 minutes. Both colchicine-sensitive events involve the active participation of the MC/MT cytoskeleton. The same movements of the chloroplast in Surirella, however, take far longer (about 1.5 and 2.5 hours, respectively) and do not apparently involve the MC/MT cytoskeleton (work in progress). I conclude that the utilization of the MC/MT system in Cymatopleura represents a nonessential but more rapid and efficient means of achieving cellular reorganization after mitosis-and one that has evolved in isolation, since I know of no record of equivalent behavior in any other algae. These examples serve to illustrate how diatoms modify and evolve their cytoplasmic systems. The differences between diatom and other spindles (Section X,A) can be accounted for in terms of such evolutionary modifications. Centric diatoms appeared first in the fossil record and they are generally agreed upon to be the more primitive group. The spindle of many centric diatoms appears superficially quite normal in structure; specifically, the splitting of the prophase central spindle into continuous (i.e., pole-to-pole) fibers during prometaphase creates a spindle structuraily similar to that in many other organisms (Section IV,D). Maintaining the central spindle as a single entity instead of permitting it to split up would not involve the evolution of significant new capabilities, and the rigidity thus acquired would then allow the pennates to evolve a spindle that holds considerable elastic energy. Perhaps the major structural innovation in the pennates is their ability to attach chromosomes directly to the pole via stretchable chromosomes and the collar material, using the polar MTs as a passive support, a function less obvious in conventional spindle types. This discussion is relevant to the mechanism(s)of anaphase A.

CELL DIVISION IN DIATOMS

101

c. THEMECHANISMOF ANAPHASEA: THEUNRESOLVED ISSUES Chromosomes in diatoms seem to spring elastically to the pole once they split; disassembly of spindle MTdshortening of kinetochore fibers is not needed for anaphase A. Thus, the sliding of prometaphase also occurs during anaphase A and several experiments suggest that energy for anaphase movement is elastically stored in the metaphase spindle. We suggest (Pickett-Heaps et al., 1982, 1984b) that this interpretation is relevant to anaphase in conventional cells too, but now necessarily complicated by the kinetics of disassembly of kinetochore MTs. I[n conventional spindles, MTs terminate in the kinetochore and disassembly of these MTs is necessary and rate-limiting for anaphase A movement (Fuseler, 1975; Inout5 et d . ,1975; Salmon and Begg, 1980). Koshland et al. (1988) now propose that the energy for poleward motion is derived solely from this MT disassembly. The proposal stems from their report of “polewards chromosome movement” (sic) in an in uitro model in which kirietochores were bound to MTs. Their proposal has received widespread endorsement (even being cited as the mechanism of anaphase A: Darnel1et al., 1990, p. 851) and has been incorporated into the popular “PAC-MAN” model (e.g., Salmon, 1989). However, it is not pedantic to point out that movement was inferred, not observed, in those experiments, nor was there a pole present. What Koshland et al. measured was the length of isolated MTs attached to chromosomes; over time, these lengths decreased to a limited extent. Mitchison (1988) and Mitchison and Kirschner (1985b) have elaborated upon the mechanisms involved in MT disassembly/ translocation at the kinetochore. While Mitchison suggests that the poleward force at prometaphase is probably produced by the same mechanism as during anaphase, his proposal that MT disassembly at the kinetochore powers prometaphase motion is now suspect (Section V,A). If this proposal by Koshland et al. is valid, anaphase A motion also should proceed to live cells depleted of ATP (since the energy required for movement is derived from disassembly of MTs already attached to the kinetochore). While anaphase A in diatoms does proceed in metabolic inhibitors where no MT disassembly is needed for anaphase A, poleward movement is m fact stopped completely in conventional cells treated with metabolic inhibitors (Section IV,B). Our results suggest a contrary view; the ATP required for anaphase A is actually for kinetochore MT disassembly, but again the experimental system used in those experiments, the permeabilized cell model (Spurck and Pickett-Heaps, 1987),is one with potential for generating misleading results. However, the possibility that energy released by MT disassembly drives anaphase A movement is supported by

102

JEREMY PICKETT-HEAPS

another evocative in uitro model. Due to the courtesy of Drs. Vivan Lombillo and Richard McIntosh (University of Colorado), the author has viewed recordings of isolated chromosomes attached to MTs in uitro; as these MTs are made to shorten, their attached chromosomes are visibly moved. While these experiments are promising, history (e.g., MT nucleation experiments with kinetochores) should alert us to the potential problems of applying in uitro observations directly to the living cell. Fuge (1989) also outlines other salient objections to this PAC-MAN model. That the kinetochore is a MT-based motor (Salmon, 1989), but not one dependent upon MT disassembly, is obvious from the behavior of diatom (Section V,A; reviewed in Pickett-Heaps et al., 1982) and newt (Rieder and Alexander, 1990) kinetochores in uiuo. I believe that the form and behavior of the corona (including the possibility that it interacts with or becomes part of an extended spindle matrix or collar equivalent) is central to the mechanism of chromosome movement (see also Rieder, 1982; Rieder and Alexander, 1990). Morphologically, the corona of diatom kinetochores and other cells are similar. Furthermore, just as the whole diatom kinetochore plus its corona become unrecognizable when fully stretched (Section V,A) and can reappear once tension is artificially removed (Pickett-Heaps and Spurck, 1982a), so the corona of PtK cells becomes invisible upon the kinetochore’s attachment to MTs, and will reappear if the kinetochore is released from these MTs, either by osmotic shock (when extensive amorphous material is associated with the kinetochore: Pover et al., 1985; Snyder, 1988) or anti-MT drugs (Cassimeris et al., 1990). In pairs of kinetochores, Ris and Witt (1981) also illustrate one with MTs inserted that has no corona while the other, bare of MTs, has a corona. My conclusion is that the ultrastructural appearance of structures fixed while under stress may be deceptive as this stress must be relieved at some stage during processing of the specimen for electron microscopy. If the corona, with or without additional matrix material, were to be stretched over kinetochore MTs, its form could well be disrupted during processing. I suggest that the images of the kinetochore fiber in Oedogonium (Schibler and Pickett-Heaps, 1980) provide a useful structural basis for visualizing the kinetochore and its filaments (Le., corona) as an MT-based sliding system that operates during prometaphase and anaphase, the latter in conjunction with kinetochore-MT disassembly. Furthermore, MT tracking analysis of Oedogonium reveals another important point: the kinetochore MTs are quite variable in length, some being short (Schibler and Pickett-Heaps, 1987). Clearly, all or even a few MTs do not have to run the full distance to the pole for the chromosome to move. The traction

CELL DIVISION IN DIATOMS

103

system of each kinetochore almost certainly operates indiscriminately ove:r intermingled polar and kinetochore MTs of uniform polarity that includes those forming continuous fibers. I suggest that to the kinetochore, these MTs are functionally equivalent; to move, it would only need to disassemble those MTs inserted directly in it. All that is needed is an array or continuum of MTs that reaches the polar region, as indeed is the case in higher plant spindles which have no discrete polar structure. In the tracking work on Oedogonium,it proved technically impossible to determine the length of the filaments that lie parallel to and among the kinetochore MTs. They appear long enough to store some elastic energy (discussed in Pickett-Heaps et al., 1986), perhaps in conjunction with the dense matrix permeating the kinetochore fiber; the latter could be the equivalent of the collar material of diatoms. Due to the elusive nature of this matrix, one cannot estimate whether it extends all the way to the pole. My personal model for the traction mechanism in conventional spindles is one in which the corona extends over the surface of MTs toward the pole, growing in extent and becoming tensioned as it does so. While the extended form of this complex might well reach the pole, it does not have to do so to be functional in moving the chromosomes since it can slide over polar MTs. If the MT continuum over which it operates is disrupted rapidly, it is concurrently broken down while if MTs are more slowly disassembled (e.g., with anti-MT drugs), it will function, for example, by pulling the poles in toward metaphase chromosomes. In summary, I see no compelling reason why cells with a demonstrably effective poleward traction mechanism, sliding, at prometaphase would need and utilize a second mechanism involving MT disassembly at anaphase as suggested by Salmon (1989), Koshland et al. (1988), Gorbsky et d. (1987, 1988), Kirschner and Mitchison (1986), Mitchison (1988, 1989), and others. I believe that the diatom spindle with its apparent unusual attachment and functioning of kinetochores allows us to see cleaxly what is happening at this deceptive and enigmatic stage of mitosis. The rate at which new information has been uncovered in the last 10 years makes it likely that we will soon have the resolution of this issue, of considerable significance in understanding how the spindle functions. ACKNOWLEDGMENTS The author acknowledges the support of grants from the SystematicsBiology and the Cell Biology Sections, NSF over many years, and a recent grant from the Australian Research Couiicil (Grant A18830947) and the NH and MRC (Grant 890871) which all enabled this work to be camed out.

104

JEREMY PICKETT-HEAPS

REFERENCES Aist, J. R., and Berns, M. W. (1981). J. Cell Biol. 91,446-458. Bajer, A. S. (1982). J. Cell Biol. !?3,33-48. Baskin, T. I., and Cande, W. Z. (1988). Cell Motil. Cytoskeleron 10,210-216. Bergen, L. G., Kuriyama, R., and Borisy, G. G. (1980). J . Cell Biol. 84, 151-1.59. Bold, H. C., and Wynne, M.J. (1978). “Introduction to the Algae.” Prentice-Hall, Englewood Cliffs, New Jersey. Boyle, J. A., Pickett-Heaps, J. D., and Czamecki, D. (1984). J. Phycol. 20,563-573. Brinkley, B. R., and Cartwright, J. (1971). J. Cell Biol. 50,416-431. Brinkley, B, R., Valdivia, M. M.,Tousson, A., and Balczon, R. D. (1989). I n “Mitosis Molecules and Mechanisms” ( J . S. Hyams and 8. R. Brinkley, eds.), pp. 77-118. Academic Press, San Diego. Brugerolle, G. (1975). Protistologica 11,457-468. Byers, B., and Abramson, D. H. (1968). Proroplsma 66,413-435. Camp, R. R., Mattem. C. F. T., and Honigsberg, B. M.(1974). 1.Protozool. 21,69-82. Cande, W. Z. (1989). I n “Mitosis Molecules and Mechanisms” (J. S. Hyams and B. R. Brinkley, eds.), pp. 303-326. Academic Press, San Diego. Cande, W. Z., and McDonald, K. L. (1986). 1. Cell Biol. 103,593-604. Cassimeris, L., Pryer, N. K., and Salmon, E. D. (1988). J. Cell Biol. 107,2223-2231. Cassimeris, L., Rieder, C. L., Rupp, G., and Salmon, E. D. (1990). J. Cell Sci. %, 9-15. Church, K., and Lin, H. -P. P. (1982). J. Cell Biol. 93,365-373. Church, K., and Lin, H. -P. P. (1985). Chromosoma 92,273-282. Cohn, S. A., and Pickett-Heaps, J. D. (1988). Eur. J . Cell Biol. 46, 523-530. Crawford, R. M.(1973). Br. Phycol. J . 6, 175-186. Darnell, J., Lodish, H., and Baltimore, D. (1990). “Molecular Cell Biology,” 2nd ed.. Scientific American Books, New York. DeBrabander, M.,Geuens, G., DeMey, J., and Joniau, J. (1979). Biol. Cell. 34,213-226. DeBrabander, M.,Geuens, G., Nuydens, R., Willebrords, R., and DeMey, J. (1980). I n “Microtubules and Microtubule Inhibitors 1980” (M. DeBrabender and J. DeMey, eds.), pp. 255-268. Elsevier, Amsterdam. DeBrabander, M.,Geuens, G., DeMey, J., and Joniau, J. (1981). Cell Motil. 1,469-483. Drebes, G . (1967). “Stephanophyxis turris (Centrales)-Geschlechtliche Fortpflanzung; Differenzierung der Spermien”. B/W 16 nun silent film, 8.5 min.; Cat # El343 from I. W. F., Gottingen, W. Germany. Drum, R. W., and Pankratz, H. S. (1963). Science 142,61-62. Edgar, L. A., and Pickett-Heaps, J. D. (1984). J. Phycol. 20, 47-61. Euteneuer, U . , and McIntosh, J. R. (1980). J . Cell Biol. 87,509-515. Euteneuer, U., and McIntosh, J. R. (1981). J. Cell Biol. 89,338-345. Febvre, J. (1977). J. Ulrrastruct. Res. 60,279-295. Forer, A. (1%5). J . Cell Biol. 25,95-117. Forer, A. (1966). Chromosoma 19,44-98. Forer, A. (1976). I n “Cell Motility. C. Microtubules and Related Proteins” (R. Goldman, T. Pollard, and J. Rosenberg, eds.), pp. 1273-1293. Cold Spring Harbor Lab, Cold Spring Harbor, New York. Forer, A. (1988). J. Cell Sci. 91,449-453. Fuge, H. (1981). Eur. J. Cell Biol. 25,9044. Fuge, H. (1989). Bio. Cell 66,209-213. Fuseler, J. W. (1975). J. Cell Biol. 67, 789-800.

CELL DIVISION IN DIATOMS

105

Geucns, G., Hill, A. M., Levilliers, N., Adoutte, A., and DeBrabander, M. (1989). J. Cell Biol. 108,939-953. Gorksky, G. J., and Borisy, G. G. (1989). J. Cell Biol. 109,653-662. Gorkeky, G. J., Sammak, P. J., and Borisy, G. G. (1987). J. Cell Biol. lM,9-18. Gortsky, G. J., Sammak, P. J., and Borisy, G. G. (1988). J. Cell Biol. 106, 1185-1192. Gould, R. R.,and Borisy, G. G. (1978). Exp. Cell Res. lW,369-374. Hayden, J., Bowser, S. S., and Rieder, C. L. (1990). J. Cell Biol. 111, 1039-1046. Hay:,, T. S., Wise, D., and Salmon, E. D. (1982). J. Cell Biol. 93,374-382. Heath, I. B. (1976). “Nuclear Division in the Fungi.” Academic Press, New York. Heath, I. B., and Darley, W. M.(1972). Eur. J . Phycol. 72,5149. Heidemam, S . R.,Zeive, G. W.,and McIntosh, J. R. (1980). J. Cell Biol. 87, 152-159. Hepler, P. K., and Palevitz, B. A. (1986). J. Cell Biol. 102, 1995-2005. Him, I., Spurck, T. P., and Pickett-Heaps, J. D. (1986). Protoplasma W2,85-89. Horio, T., and Hotani, H. (1986). Nature (London) 321,605-607. Huitorel, P., and Kirschner, M.W. (1988). J . Cell Biol. 106, 151-159. Hyruns, J. S., and Brinkley, B. R. (1989). “Mitosis Molecules and Mechanisms.” Academic Press, San Diego. InouC, S., and Ritter, H. (1978). J. Cell Biol. 77,655-684. InouC, S., and Sato, H. (1%7). J. Gen. Physiol. 50,259-292. InouC, S., Fuseler, J., Salmon, E. D., and ENS, G. W. (1975). Biophys. J . 15,725-744. Johnson, K. A., and Borisy, G. G., (1977). J. Mol. Biol. 117, 11-31. Kirschner, M.,and Mitchison, T.J. (1986). Cell#, 329-342. Kosliland, D. E., Mitchison, T. M., and Kirschner, M. W. (1988). Nature (London) 331, 499-504.

Kubai. D. F. (1973). J. Cell Sci. W, 511-552. Kubai, D. F.,and Ris, H. (1%9). J. Cell Biol. 40,508-528. Lauferborn, R. (1986). “Untersuchungen iiber Bau, Kernteilung und Bewegung der Diatomeen”. Wilhelm Engelmann, Leipzig. Leslie, R. J., and Pickett-Heaps, J. D. (1983). J. Cell Biol. 96,548-561 Leslie, R. J., and Pickett-Heaps, J. D. (1984). Cell 36, 717-727. Manton, I., and Von Stosch, H. A. (1966). J. R . Microsc. SOC.85, 119-134. Manton, I., Kowallik, K.,and Von Stosch, H. A. (1%9a). J Microsc. 89,295-320. Manton, I., Kowallik, K., and Von Stosch, H. A. (1%9b). J. Cell Sci. 5,271-298. Manton, I, Kowallik, K., and Von Stosch, H. A. (1970a). J. Cell Sci. 6, 131-157. Manton, I., Kowallik, K., and Von Stosch, H. A. (1970b). J. Cell Sci.7,407-443. Masuda, H., and Cande, W. 2.(1987). Cell, 49, 193-202. McDonald, K. L., Pickett-Heaps, J. D., McIntosh, J. R., and Tippit, D. H. (1977). J . Cell Biol. 74,377-388. McIbnald, K.L., Edwards, M.K., and McIntosh, J. R. (1979). J. Cell Biol, 83,443-461. McCiill, M., and Brinkley, B. R. (1975). J. Cell Biol. 67, 189-199. McIntosh, J. R., and Landis, S. C. (1971). J. Cell Biol. 49,468-497. McIntosh, J. R., Hepler, P. K., and Van Wie, D. G. (1%9). Nature (London) 224,659-663. McIntosh, J. R.,McDonald. K. L., Edwards, M. K., and Ross, B. M. (1979). J. Cell Biol. 83, 428-442.

McIntosh, J. R.,Roos, U.-P., Neighbours, B., and McDonald, K. L. (1985). J. Cell Sci. 75, 93-129.

McPleeill, P. A., and Berns, M.W. (1981). J . Cell Biol. 88,543-553. Mitchison, T. (1988). Annu. Rev. Cell Biol. 4,527-549. Mitchison, T. (1989). Curr. Opinion Cell Biol. 1,67-74.

106

JEREMY PICKETT-HEAPS

Mitchison, T., and Kirschner, M. (1984). Nature (London)3l2,237-242. Mitchison, T., and Kirschner, M. (1985a). J. Cell Biol. 101,755-765. Mitchison, T., and Kirschner, M. (1985b). J . Cell Biol. 101,766-777. Mitchison, T., Evans, L., Schulze, E., and Kirschner, M. (1986). Cell&, 515-527. Nicklas, R. B., Brinkley, B. R., Pepper, D. A., Kubai, D. F., and Rickards, G. K. (1979). J . Cell Sci. 35,87-104. Nicklas. R. B., Lee, G. M, Rieder, C. L., and Rupp, G. (1989). J . Cell Sci. 94,415-423. Oakley, B. R., and Dodge, J. D. (1976). Cytobios 1 7 , 3 5 4 . Pepper, D. A., and Brinkley, B. R. (1979). J . Cell Biol. 82,585-591. Pickett-Heaps, J. D. (1969). Cytobios 3,257-280. Pickett-Heaps, J. D. (1980). “Cell Division in the Diatom Pinnularia.” 16mm sound, color film; 16 min; available through CYTOGRAPHICS, School of Botany, University of Melbourne, Parkville, Vic. 3052, Australia. Pickett-Heaps, J. D. (1982). “Cell Division in the Diatom Huntzschiu amphioxys,” 16mm sound, color film; 19.5 min; avaiable through CYTOGRAPHICS, School of Botany University of Melbourne, Parkville, Vic. 3052, Australia. Pickett-Heaps, J. D. 91983a). J . Phycol. 19,269-281. Pickett-Heaps, J. D. (1983b). “Cell Division in Surirellu. A tribute to Robert Lauterborn.” 16mm sound, color film; 14 min; available through CYTOGRAPHICS, School of Botany, University of Melbourne, Parkville, Vic. 3052, Australia. Pickett-Heaps, J. D. (1983~).“Cell Division in Huntzschiu amphioxys; Effects of Various Drugs.” 16mm sound, color film;12 min; available through CYTOGRAPHICS, School of Botany, University of Melbourne, Parkville, Vic. 3052, Australia. Pickett-Heaps, J. D. 91984a). “Cell Division in the Diatom Cymatopleuru,” 16mm sound, color, film; 12 min; available through CYTOGRAPHICS, School of Botany, University of Melbourne, Parkville, Vic. 3052, Australia. Pickett-Heaps, J. D. (1984b). “Experimental Investigations into Mitosis in R K Cells,” 16mm sound, color, film; 13 min; available through CYTOGRAPHICS, School of Botany, University of Melbourne, Parkville, Vic. 3052, Australia. Pickett-Heaps, J. D. (1986). Trends Biochem. Sci. 11,504-507. Pickett-Heaps, J. D. (1987). In “Algal Development; Molecular and Cellular Aspects” (W. Wiessner, D. G. Robisson, and R. C. Starr, eds.), pp. 28-33. Springer-VerlagBerlin-Heidelberg-New York. Pickett-Heaps, J. D. (1991). Cell Motil. Cyfoskel.In press. Pickett-Heaps, J. D., and Kowalski, S. E. (1981). Eur. J. CellBiol. 25, 150-170. Pickett-Heaps, J. D., and Spurck, T. P. (1982a). Eur. J. Cell Biol. 28,77-82. Pickett-Heaps, J. D., and Spurck, T. P. (1982b). Eur. J. Cell Biol. 28,83-91. Pickett-Heaps, J. D.. and Tippit, D. H. (1978). Cell 14,455467. Pickett-Heaps, J. D., McDonald, K.,and Tippit, D. H. (1975). Protoplusma 86, 205-242. Pickett-Heaps, J. D., Tippit, D. H., and Andreozzi, J. A. (1978a). Biol. Cell. 33,71-78. Pickett-Heaps, J. D., Tippit, D. H., and Andreozzi, J. A. (1978b). Biol. Cell. 33,79-84. Pickett-Heaps. J. D., Tippit, D. H., and Andreozzi, J. A. (1979a). Biol. Cell. 35, 199-206. Pickett-Heaps, J. D., Tippit, D. H., and Andreozzi, J. A. (1979b) Biol. Cell. 35,295-304. Pickett-Heaps, J. D., Tippit, D. H., and Leslie, R. (1980a). Eur. J. Cell Biol. 21, 1-1 1. Pickett-Heaps, J. D., Tippit, D. H., and Leslie, R. (1980b) Eur. J . Cell Biol. 21, 12-27. Pickett-Heaps, J. D., Tippit. D. H., and Porter, K. R. (1982). Cell, 29,729-744. Pickett-Heaps, I. D., Schmid, A. M., and Tippit, D. H. (1984a). Protoplusmu 120, 132-154. Pickett-Heaps, J. D., Spurck, T., and Tippit, D. H. (1984b). J. Cell Biol. 99, 137s-143s. Pickett-Heaps, J. D., Tippit, D. H., Cohn, S. A.. and Spurck, T. (1986). J. Theoret. Biol. 118, 153-169.

CELL DIVISION IN DIATOMS

107

Pickett-Heaps, J. D., Cohn. S., Schmid, A. M., and Tippit, D. H. (1988a). J. Phycol. 24, 35-49. Pickett-Heaps, J. D., Wetherbee, R., and Hill, D. R. A. (1988b).Protoplasma 143, 139-149. Pickett-Heaps, J. D., Schmid, A. M., and Edgar, L. A. (1991). In “Progress in Phycological Research.” In press. Pover, N. K., Golub, R. J., McLelland, S. L., and Snyder, J. A. (1985). Eur. J. CellBiol. 39, 366-372. Riedcr, C. L. (1981). Chromosoma 84, 145-158. Rieder, C. L. (1982). Int. Rev. Cytol. 79, 1-58. Rieder, C. L., and Alexander, S. P. (1990). J. CeNBiol. 110,81-85. Rieder, C. L., and Bajer. A. (1977). J. Cell Biol. 74,717-725. Rieder, C. L., and Borisy, G. G. (1981). Chromosoma 82,693-716. Rieder, C. L., Davison, E. A., Jensen, L. C. W., Cassimeris, L., and Salmon, E. D. (1986). .I. CellBiol. 103,581-591. Ris, H. (1949). Biol. Bull. %, 90-106. Ris, H., and Kubai, D. H. (1974). J. Cell Biol. 60,702-720. Ris, H., and Witt, P. L. (1981). Chromosoma 82,153-170. Ritter, H., Inouk, S., and Kubai, D. (1978). J . Cell Biol. 77,638-654. Roos, U. -P. (1973a). Chromosoma 40,43-82. Rooa, U. -P. (1973b). Chromosorna 41, 195-220. Roos, U. -P. (1976). Chromosoma 54,363-385. Rooa, U. -P., and Camenzind, R. (1981). Eur. J. Cell Biol. 25,248-257. Salmon, E. D. (1989). In “Mitosis Molecules and Mechanisms” (J. S. Hyams and B. R. Brinkley, eds.), pp. 119-182. Academic Press, San Deigo. Salmon, E. D., and Begg, D. A. (1980). J. Cell Biol. 85,853-865. Saminak, P. J., and Borisy, G. G. (1988). Nature (London) 332,724-726. Saxton, W.M., and McIntosh, J. R. (1987). J. CellBiol. 105,875-886. Schaap, C. J., and Forer, A. (1984). J. Cell Sci. 65,21-40. Schibler, M. J., and Pickett-Heaps, J. D. (1980). Eur. J . Cell Biol. 22,687-698. Schibler, M. J., and Pickett-Heaps, J. D. (1987). Protoplasma l37,29-44. Schrader, F. (1953). “Mitosis. The Movement of Chromosomes in Cell Division,” 2nd ed. Columbia University Press, New York. Schulz, D., and Jarosch, R. (1980). Eur. J . Cell Biol. M ,249-253. Snyder, J. A. (1988). Cell Motil. Cytoskeleton 11,291-302. Snyder, J. A., and Mclntosh, J. R. (1975). J. Cell Biol. 67,744-760. Soltys, B. J., and Borisy, G. G. (1985). J . Cell Biol. 100, 1682-1689. Soranno, T., and Pickett-Heaps, J. D. (1982). Eur. J. Cell Biol. 26,234-243. Spurck, T. P., and Pickett-Heaps, J. D. (1987). J. Cell Biol. 105, 1691-1705. Spurck, T. P., Pickett-Heaps, J. D., and Klymkowsky, M. W. (1986a). Protoplasma 131, 17-59. Spurck, T. P., Pickett-Heaps, J. D., and Klymkowsky, M. W. (1986b). Protoplasma 131, 60-74. Spurck, T. P., Stonington, 0. G., Snyder, J. A., Pickett-Heaps, J. D., Bajer, A., and Mole-Bajer, J. (1990). J . Cell Biol. 111, 1505-1518. Stonington, 0. G., Spurck, T. P., Snyder, J. A., and Pickett-Heaps, J. D. (1989). Protoplasma 153,62-70. Summers, K., and Kirschner, M. W. (1979). J. CellBiol. 83,205-217. Telzer, B. R., and Haimo, L. T. (1981). J . CellBiol. 89,373-378. Telzer, B. R., Moses, M. J., and Rosenbaum, J. L. (1975). Proc. Natl. Acad. Sci. U.S.A. 72, 4023-4027.

108

JEREMY PICKEIT-HEAPS

Tippit, D. H., and Pickett-Heaps. J. D. (1977).J . Cell Eiol. 73,705-727. Tippit, D. H., McDonald, K. L., and Pickett-Heaps, J. D. (1975). Cytobiologie l2,52-73. Tippit. D . H., Schultz, D., and Pickett-Heaps, J. D. (1978). J . Cell Biol.79,737-763. Tippit, D. H., Pickett-Heaps, J . D., and Leslie, R. (1980a).J . Cell Biol. 86,402-416. Tippit, D. H., Pillus, L., and Pickett-Heaps, J. D.(1980b).J . Cell Biol. 87,531-545. Tippit, D. H.,Pillus, L., and Pickett-Heaps, J. D. (1983). Eur. J . Cell Eiol. 30,9-17. Tippit, D. H.,Fields, C. T., O'Donnell, K. L., Pickett-Heaps J. D., and McLaughlin, D. J. (1984). E m . J . CeN Biol. 34, 34-44. Wadsworth, P., Shelden, E., Rupp, G., and Rieder, C. L. (1989). J. CellBiol. 109,2257-2265. Walker, R. A., Inouk, S., and Salmon, E. D. (1989). J . Cell Eiol. 108,931-937. Weisenberg, R. C., and Rosenfeld, A. C. (1975).J. Cell Eiol. 64, 146-158. Witt, P. L., Ris. H., and Borisy, G. G. (1980). Chromosoma 81,483-505. Wordeman, L. McDonald, K. R., and Cande, W. Z. (1986).J. Cell Biol. 102, 1688-1698.

INTERNATIONAL REVIEW OF CYTOLOGY.VOL. 128

Nerve Growth Factor Synthesis and Nerve Growth Factor Receptor Expression in Neural Development ALUNM.DAVIES Department of Anatomy, St. George’s Hospital Medical School, London SW17 ORE, England

I. Introduction Since its discovery in the early 1950s, nerve growth factor (NGF) has been the prototype of a class of proteins termed neurotrophic factors that regulate the survival and morphology of neurons during development and growth of the vertebrate nervous system. In the past decade, the establishment of sensitive assays for NGF and its receptor together with the cloning of the genes encoding these molecules have provided the necessary tools for ascertaining the normal patterns of NGF synthesis and NGF receptor expression during development and for studying the factors that regulate the expression of the NGF and NGF receptor genes. These studies have led to considerable progress in clarifying the role of NGF in development and are the subject of this review article. I shall present this recent work on NGF in the context of the neurotrophic theory and discuss its implications for understanding the molecular basis of neuron-target interactions in the developing nervous system.

11. The Neurotrophic Theory

The neurotrophic theory addresses the molecular basis of two aspects of the dependency of neurons on the cells they innervate: the survival of neurons and the growth and arborization of their axons and dendrites. The ear!liest and most striking manifestation of this dependency is the occurrence of a phase of neuronal death which begins shortly after neurons start innlervating their targets (Oppenheim, 1981). This eliminates superfluous andl inappropriately connected neurons and adjusts the number of neurons to the size and requirements of their target fields (Cowan el al., 1984). After the phase of neuronal death, the continued dependency of neurons on their targets is manifested by the ongoing adjustment of axonal branching and dendritic arborizationto changes in the size and shape of the 109 Copyright Q 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.

110

ALUN M. DAVIES

target field during the growth of the animal (Purves, 1988; Purves et al., 1988). The principal tenet of the neurotrophic theory is that the survival of developing neurons is dependent on the supply of a neurotrophic factor which is synthesized in their target field. The number of neurons that survive during the phase of target field innervation is controlled by regulating the supply of this neurotrophic factor in the target field (for recent reviews see Thoenen and Barde, 1980; Davies, 1988a, 1988b; Purves, 1988; Purves er al., 1988; Barde, 1989; Oppenheim, 1989). In addition to regulating neuronal survival, there is evidence that the supply and distribution of neurotrophic factors also mediate the long-term trophic effects of the target field on neuronal morphology (Purves, 1988). In the following section I shall outline the principal experimental evidence for the neurotrophic theory. This theory has been substantiated primarily by work on NGF, the first neurotrophic factor to identified, and has been corroborated by studies of the more recently identified neurotrophic factor, brain-derived neurotrophic factor (BDNF). This brief outline will serve as an introduction and framework for the more detailed discussion of the developmental importance of the prototypical neurotrophic factor, NGF. It should be pointed out, however, that the role of NGF is not restricted to its effects on neuronal survival and morphology. For example, NGF has been shown to regulate the synthesis of neurotransmitters and neuropeptides in NGF-dependent neurons (Thoenen er al., 1971; Kessler and Black, 1980).These and other possible roles of NGF will be discussed where appropriate. A. NEURONALSURVIVAL

The most important direct evidence for the neurotrophic theory is the demonstration that populations of developing neurons that are supported by NGF in uirro, namely sympathetic neurons and certain kinds of sensory neurons, are also dependent on NGF in uiuo. Anti-NGF antibodies administered during the phase of target field innervation eliminate these neurons, whereas exogenous NGF rescues neurons that would otherwise die (LeviMontalcini and Angeletti, 1968;Johnson er al., 1980; Hamburger and Yip, 1984)Substantial indirect support for the role of neurotrophic factors in regulating neuronal survival comes from studies of NGF synthesis, uptake, and transport. Nerve growth factor is synthesized in the target fields of sensory and sympathetic neurons during development (Davies et al., 1987a; Korsching and Thoenen, 1988; Harper and Davies, 1990). Specific cell surface receptors mediate the uptake of NGF by axons in the target field

NGF SYNTHESIS AND RECEPTOR EXPRESSION

111

(Sutter et al., 1979). Fast axonal transport conveys the receptor-ligand com.plex from the target field to the cell bodies of the innervating neurons where it exerts its survival-promoting effects (Korsching and Thoenen, 1983; Palmatier et al., 1984; Davies et al., 1987a). Preventing the uptake of NGF from the target field by destroying adrenergic terminals with 6hydroxy dopamine or by interrupting axonal transport with vinblastine or axoi:omy leads to the death of developing sympathetic neurons which can be prevented by the concomitant administration of NGF (Levi-Montalcini et d., 1975; Hendry & Campbell, 1976; Menescini-Chen et af., 1978). In addition to the overwhelming evidence that NGF regulates the survivail of sympathetic neurons and certain kinds of sensory neurons during development, several findings suggest that NGF may play a similar role in the development of cholinergic neurons of the basal forebrain. NGF is synthesized in the regions of the brain innervated by these neurons (Korsching et al., 1985) and is conveyed by axonal transport from these regions to the neuronal cell bodies (Korsching et al., 1986). NGF promotes the ,survivalof basal forebrain cholinergic neurons in culture (Hartikka and Hefti, 1988; Hatanaka et al., 1988) and prevents these neurons from degenerating in uiuo after axotomy (Hefti, 1986; Williams et al., 1986; Kromer, 1987; Montero and Hefti, 1988). It has yet to be demonstrated, however, whether administration of NGF or anti-NGF during developmerit influences the survival of these neurons.

B. AXONALAND DENDRITIC GROWTH One of the earliest demonstrations that NGF promotes axonal branching in addition to neuronal survival in uiuo was the finding that NGF administered to neonatal rodents results in profuse, generalized branching of sym,pathetic axons (Olson, 1967). A recent elegant in uivo demonstration that NGF influences axonal branching locally comes from a study of transgenic mice in which NGF was selectively overexpressed in pancreatic p cells (Edwards et al., 1989). In a transgenic line where NGF was overexpressed in only a fraction of the p cells, profuse branching of symipathetic axons was restricted to the vicinity of these cells. This finding agrees with the earlier demonstration that NGF preferentially maintains the growth of sympathetic neurites selectively exposed to it in uitro (Caimpenot, 1982a, 1982b). Localized injections of NGF in neonatal mice also cause prolific branching of NGF-responsive sensory axons in the vicinity (Hopkins and Slack, 1984), and anti-NGF antiserum prevents the formation of sensory axon collaterals from nerves adjoining a region of surgically denervated skin (Diamond et af., 1987; Owen et al., 1989). In addition to a local effect on axonal branching, elevated levels of NGF in

112

ALUN M. DAVIES

the target field also result in increased dendritic arborization in the corresponding sympathetic ganglia (Snider, 1988).

C. THEGENERALITY OF THE NEUROTROPHIC THEORY The generality of the neurotrophic theory has become apparent from work on BDNF, a protein originally purified from pig brain (Barde et at., 1982) that shares extensive homology with NGF (Leibrock et al., 1989). BDNF promotes the survival of embryonic retinal ganglion cells (Johnson et al., 1986; Rodriguez-Tebar et al., 1989) and NGF-independent sensory neurons in culture (Lindsay et al., 1985; Davies et al., 1986a, 1986b) and prevents cell death in populations of BDNF-responsive sensory neurons when administered to chick embryos (Hofer and Barde, 1988). Although antibodies that block the activity of BDNF have not been available, it has been possible to use silastic barriers to prevent access of developing sensory neurons to their normal supply of BDNF in vivo,a manipulation which results in a considerable reduction in the number of surviving neurons (Kalcheim et al., 1987). Another neurotrophic factor with extensive homology to NGF and BDNF has recently been cloned (Hohn et al., 1990; Maisonpierre et al., 1990; Rosenthal et al., 1990). This factor, termed neurotrophin-3, has a distinct neuronal specificity to NGF and BDNF.

III. Molecular Biology, Biochemistry, and Biosynthesis of NGF

The fortuitous discovery of exceptionally large amounts of NGF in the submandibular salivary glands of adult male mice has led to substantial progress in understanding the structure and synthesis of this protein. Mouse NGF is a dimer consisting of two identical, 118 amino acid chains each possessing three disulphide bridges (Angeletti and Bradshaw, 1971). The primary amino acid sequences of NGF currently available from other species (human, guinea pig, cow, chicken, and snake) reveal a high degree of homology (Ullrich et al., 1983; Schwarz et al., 1989; Meier et al., 1986; Ebendal et al., 1986; Hogue-Angeletti et al., 1976). This is particularly the case with four highly conserved domains that contain the histidine and tryptophan residues whose biochemical modification abolishes the biological activity of NGF (Frazier et al., 1973; Cohen et al., 1980; Dunbar et al., 1984). In terms of the structure-function relationships of NGF and other neurotrophic factors, it is of considerable interest that at least part of each of these domains is completely homologous in porcine BDNF (Leibrock et

NGF SYNTHESIS AND RECEPTOR EXPRESSION

113

al., 1989) and in mouse neurotrophin-3 (Hohn et al., 1990). Although NGF has been successfully crystallized (Wlodawer et al., 1975), little progress has been made in the analysis of its tertiary structure (Williams et al., 1982). Studies of NGF biosynthesis (Berger and Shooter, 1977; Darling et al., 1983; Edwards et al., 1988a) together with the isolation of NGF cDNA clones (Scott et al., 1983; Ullrich et al., 1983)have demonstrated that NGF is derived from larger precursors by proteolytic cleavage. The mature NGF protein is encoded by a single 3' exon in the NGF gene (Ullrich et al., 198.3; Edwards et al., 1986). There are several smaller 5' exons, and alternative splicing produces at least four different mRNA transcripts that predict precursors that differ in their amino termini (Selby et al., 1987). The ratio of these different mRNA transcripts is unchanged in development and is similar in virtually all tissues except the salivary gland, where the large quantities of NGF secreted play no role in neural development or function. The two major transcripts are translated into precursors that differ in size (27 and 34 kDa) and in the location of the putative signal peptide (located at the amino terminus in the 27 kDa precursor and 67 amino acids internal to the amino terminus in the 34 kDa precursor). The significance of these different precursors is unclear, as both are rapidly cleaved at the signal peptide concomitantly with translocation into the endoplasmic reticulum to give the same protein (Edwards et al., 1988a). This protein becomes glycosylated and is further cleaved to produce the biologically active, mature NGF protein (Edwards et al., 1988b). Studies using recombinant vaccinia virus vectors to transfect cells with the two major NGF-mRNA transcripts have demonstrated that many mammalian cell lines possess the necessary machinery to process and secrete NGF (Edwards et al., 1988a).

IV. NGF Synthesis Two major advances have facilitated studies of the site, timing, and regulation of NGF synthesis during development. The first is a very sensitive:, specific, two-site enzyme immunoassay for NGF that is capable of detecting the very low levels of this protein present in developing neuronal target fields (Korsching and Thoenen, 1983). The second is the isolation of NG,F cDNAs (Scott et al., 1983; Ullrich et al., 1983) and the subsequent utilization of sensitive hybridization techniques to detect the extremely low levels of NGF mRNA in RNA extracted from tissues (Heumann el al., 1984; Shelton and Reichardt, 1984)and in tissue sections by in situ hybridization (Bandtlow et al., 1987).

114

ALUN M. DAVIES

A. SITEOF NGF SYNTHESIS

I . NGF Synthesis by Neuronal Target Cells Although careful measurements of NGF and NGF mRNA have confirmed that NGF is synthesized in regions innervated by NGF-dependent neurons (Korsching and Thoenen, 1983; Heumann et al., 1984; Shelton and Reichardt, 1984; Korsching et al., 1985), the identity of the cells that synthesize NGF in these regions during normal development has only recently been determined. This was first ascertained in a detailed developmental study of the whisker pad of the mouse embryo, which receives a rich sensory innervation from the trigeminal ganglion (Davies et al., 1987a). The feasibility of enzymatically separating the early whisker pad into its two principal cellular components, the epithelium (presumptive epidermis) and mesenchyme (presumptive dermis and subcutaneous tissue), permitted measurements of NGF mRNA in these components during the early stages of their innervation. Both components contained NGF mRNA, the concentration being 5-fold higher in the epithelium. In situ hybridization additionally revealed that NGF-mRNA expression in the mesenchyme is concentrated in the region beneath the epithelium (presumptive dermis). These findings suggest that NGF is synthesized in regions of the target field where the great majority of axons terminate. An unresolved issue is whether or not NGF synthesis occurs in all epithelial cells or is restricted to a subset of these cells such as Merkel cells (which make synapses with certain sensory axons) or Merkel cell progenitors. This issue may have important implications for governing which neurons survive and which die during the phase of neuronal death. Generalized synthesis would argue in favor of a stochastic process; neurons are lost at random until a balance between NGF supply and removal is reached. Restriction of synthesis to a defined subset of cells would suggest that neurons are selectively maintained in accordance with the distribution of their axon terminals in the target field, with neurons whose axons terminate in relation to NGF-producing cells being selectively supported. At a later stage in development, the restriction of NGF synthesis to a subset of cells in the target field may play an important role in remodeling axon terminations within the target field. The demonstration of NGF mRNA in cutaneous epithelium (in which there are no Schwann cells) also invalidates the view that NGF is synthesized exclusively by Schwann cells in the peripheral nervous system (Rush, 1984; Finn et al., 1986). This view was based on the observation that Schwann cells in the denervated iris are intensely labeled by anti-NGF antibodies. It is likely, however, that his observation was due to de n o w synthesis of NGF by Schwann cells and expression of NGF receptors on

NGF SYNTHESIS AND RECEPTOR EXPRESSION

115

these cells which occur as a consequence of axonal injury (Taniuchi et al., 198,6; Heumann et al., 1987a). Furthermore, the distribution of NGF mR.NA in the intact normal iris is not consistent with exclusive expression in Schwann cells (Bandtlow et al., 1987). The most direct demonstration that NGF is synthesized in the cells innervated by NGF-responsive neurons comes from recent in situ hybridization studies of the distribution of NGF mRNA in the brain of adult rodlents (Ayer-LeLievre et al., 1988; Whittemore et al., 1988). Using 3H-labeled NGF cDNA probes, NGF mRNA was unambiguously localizecd to the cell bodies of pyramidal neurons in the hippocampus which are innervated by the NGF-responsive cholinergic neurons of the basal forebrain. Whether NGF synthesis also occurs in target neurons during the early stages of hippocampal innervation has yet to be confirmed by in situ hybridization studies carried out at this stage. 2 . NGF Synthesis by Other Cell Types In addition to the growing body of evidence that NGF is normally synthesized by the cells directly innervated by NGF-dependent neurons in uiuo, there is some evidence that glial cells and fibroblasts may synthesize NGF. A variety of biochemical and immunochemical techniques have been used to show that primary cultures of fibroblasts (Houlgatte et al., 1989), Schwann cells (Bandtlow et al., 1987), and astrocytes (Lindsay, 1979; Furukawa et al., 1986; Tarris et al., 1986; Assouline et al., 1987) together with glial cell lines (Longo, 1978; Norrgren et al., 1980) synthesize and release NGF into the culture medium. Although it has been proposed that these cells normally provide a supply of NGF to neurons in uiuo,direct evidence for this view is lacking. Indeed several observations suggest that NGF synthesis in these cultured cells occurs as a consequence of removing them from their normal environment. It has been shown that the: level of NGF mRNA is very low in the Schwann cells and fibroblasts of intact adult sciatic nerve in uiuo but increases markedly when these cells are grown in culture (Heumann et al., 1987a, 1987b; Lindholm et al., 1987; Bandtlow et al., 1987).Studies of the localization of NGF mRNA by in situ hybridization in the adult rat brain revealed no labeling in regions rich in oligodendrocytes, such as the corpus callosum, in contrast to the clear labeling of pyramidal neurons in the hippocampus (Ayer-LeLievre et al., 1988). :it is likely that the de nouo synthesis of NGF by glial cells placed in culture represents part of the standard response of these cells to disruption of their normal association with neurons. The best evidence in support of this proposal comes from in uiuo and in uitro studies of NGF synthesis in Schwann cells. When the sciatic nerve of adult rats is cut or crushed, NGF

116

ALUN M. DAVIES

synthesis is rapidly induced in the Schwann cells that lie adjacent and distal to the lesion (Heumann et al., 1987a). A similar rapid increase in NGF synthesis is observed in sciatic nerves placed in culture (Heumann et a f . , 1987b). After the initial increase in NGF synthesis in uiuo there is a further rise in synthesis which is stimulated by interleukin-I released from macrophages that migrate into the damaged nerve (Lindholm et al., 1987). Although this second rise in NGF synthesis is not observed in cultured sciatic nerves (Heumann et al., 1987b), there is some evidence for less potent serum factors that stimulate and maintain NGF synthesis in cultured cells (Houlgatte et al., 1989). It is likely that de nouo NGF synthesis in damaged peripheral nerves plays an important role in regeneration. NGF enhances the survival of adult sensory neurons following peripheral nerve injury (Otto et al., 1987; Rich et al., 1987, 1989) and increases the rate of regeneration of neurites from adult sensory neurons in culture (Lindsay, 1988). Although the level of NGF synthesis in adult Schwann cells is normally insignificant, the level of NGF mRNA in newborn rat sciatic nerve is an order of magnitude higher than in the adult nerve and only falls to the very low adult levels by 3 weeks (Heumann et al., 1987b). A plausible explanation for the high level of NGF synthesis at this stage of development is that it too results from the consequences of axonal degeneration. The fetal sciatic nerve contains large numbers of degenerating axons associated with naturally occurring cell death in the respective sensory ganglia and motoneuron pools. This degeneration may explain the large numbers of macrophages present in the newborn sciatic nerve (Stoll and Muller, 1986) which may themselves contribute to the stimulation of NGF synthesis in Schwann cells. To confirm or refute this proposal it will be essential to determine when NGF synthesis begins in the developing nerve and whether this is related to the onset of axonal degeneration, and to determine if the Schwann cells that synthesize NGF are related to degenerating axons. Given the indubitable role of the target field in regulating its innervation, it seems unlikely that NGF synthesis in Schwann cells plays any role in regulating neuronal survival or morphology during normal development. 3 . Regional Differences in NGF Synthesis

One of the accepted tenets of the neurotrophic theory is that regional differences in innervation density are regulated, at least in part, by the local supply of neurotrophic factors during development; target fields that provide a rich supply of these factors are able to support more neurons than target fields that provide a poor supply. It is a matter of controversy,

NGF SYNTHESIS AND RECEPTOR EXPRESSION

117

however, whether differences in the supply of these factors from one target field to another are regulated by controlling neurotrophic factor production or by controlling the access of neurons to these factors. (Olppenheim, 1989). !Support for the access hypothesis comes from studies of pharmacologically paralyzed chick embryos in which motoneurons do not die. Soluble extracts of the muscle of these embryos were found to be no more effective than muscle extracts of normal embryos in promoting the survival of motoneurons both in uitro (Tanaka, 1987) and in uiuo (Houenou et al., 1989), and this has been taken to indicate that the level of neurotrophic factor synthesis in hyperinnervated and normally innervated muscle is the same (Oppenheim, 1989). Experiments using muscle extracts are, however, difficult to interpret because extracts are likely to contain a mixture of neurotrophic factors, not all of which may normally play a role in regulating the survival of a particular class of neurons in uiuo. In contrast to the bona fide neurotrophic factors NGF and BDNF, which are encoded by very low-abundance mRNA molecules that possess leader sequences for secretion (Scott et al., 1983; Leibrock et al., 1989), fibroblast growth factor (FGF) and ciliary neurotrophic factor (CNF), which have been shown to support the survival of neurons in uitro (Morrison et al., 1986; Walicke, 1988; Barbin et al., 1984), are relatively abundant proteins that lack leader sequences (Abraham et al., 1986; Stockli et al., 19r39). Thus, proteins that may not normally be efficiently secreted from celils may be present in relatively large amounts in tissue extracts and overshadow the effects of developmentally relevant neurotrophic factors. Clmrly, the relative importance of the production and availability of neuror.rophic factors in regulating neuronal survival can only be adequately resolved by studying the synthesis of the relevant neurotrophic factors in different target fields during the stage of n a t d y occurring neuronal death. Until recently the only comparative studies of NGF synthesis in tissues and organs that differ in innervation density had been carried out in adult animals. These studies revealed a good correlation between sympathetic innervation density and the levels of both NGF and NGF mRNA (Korsching and Thoenen, 1983; Heumann et al., 1984; Shelton and Reichardt, 1984). However, because the number of sympathetic neurons remains unchanged in the mature nervous system and NGF is no longer required for their survival (Chun and Patterson, 1977; Goedert et al., 1978), the significance of these findings for understanding the regulation of the NGF supply during development is unclear. Regional differences in the expression of the NGF gene during the early stages of target field innervation have recently been studied in the embryonic mouse trigeminal system (Harper and Davies, 1990). The cutaneous

118

ALUN M.DAVIES

territory of the trigeminal ganglion is particularly suitable for this kind of investigation because it is divided into three well-defined regions that differ markedly in their innervation density. Measurement of the concentration of NGF mRNA in epithelia isolated from each region at the onset of neuronal death revealed a clear correlation with innervation density. Although the relationship between the level of NGF and the level of NGF mRNA is not precisely known in developing epithelium, these findings nonetheless argue in favor of the level of production of NGF as being a major factor in regulating the NGF supply in different target fields during development. In addition to addressing fundamental issues such as the regulation of the NGF supply duringdevelopment, studies of the site of NGF synthesis, particularly in the central nervous system, may provide an indication of further populations of neurons that depend on NGF for survival. Indeed, NGF mRNA has been demonstrated in many parts of the brain in addition to those innervated by cholinergic neurons (Shelton and Reichardt, 1986).

B. TIME-COURSE OF NGF SYNTHESIS The application of sensitive assays for NGF and its mRNA to wellcharacterized developing neuronal systems has provided valuable data on the onset and changes in the level of NGF synthesis in relation to the time-course of target field innervation and neuronal death. These data have not only clarified the role of NGF in development but have also improved our understanding of the trophic interactions that regulate neuronal survival. I . Onset of NGF Synthesis The most precise data on the timing of NGF synthesis has been obtained from developmental studies of the whisker pad of the mouse embryo, a cutaneous target field that receives a rich sensory innervation from the trigeminal ganglion. This clearly defined target field, whose time-course of innervation has been precisely documented (Davies and Lumsden, 1984), can easily be dissected from the embryo at stages prior to and throughout the phase of innervation. Neither NGF nor NGF mRNA are detectable prior to the onset of innervation but are first detected as the earliest axons reach the target field (Davies et al., 1987a). The conclusion that NGF synthesis commences with the onset of target field innervation was also reached in a developmental study of NGF levels in the heart ventricle and submandibuiar gland of the mouse embryo (Korsching and Thoenen, 1988). Nerve growth factor is first detectable in these tissues around the

NGF SYNTHESIS AND RECEPTOR EXPRESSION

119

time of initial innervation by sympathetic neurons, as deduced from anatoniical studies of the development of sympathetic innervation in the rat embryo (Rubin, 1985). The demonstration that NGF synthesis does not begin until the earliest sensory and sympathetic axons have reached their target fields invalidates the view that NGF attracts sensory and sympathetic axons to their target fields by chemotropism (Levi-Montalcini, 1982). This view was based on in d u o and in uitro observations of the effects of very high concentrations of NGF on the gowth of late-embryonic or neonatal sensory and sympathetic axons (Menesini-Chen et al., 1978; Letourneau, 1978; Gundersen and Barrett, 1980). For a detailed discussion of this work and its interpretation in terms of the trophic effects of NGF and the physical properties of this molecule at the high concentrations used, see Davies (1987a). In contrast to developing sensory and sympathetic target fields where the onset of NGF synthesis is coordinated with the arrival of axons, the appearance of NGF and NGF mRNA in the hippocampus precedes by several days the onset of its innervation by the NGF-responsive cholinergic neurons of the basal forebrain (Large et al., 1986; Auburger et al., 1987; Lu et al., 1989). The interpretation of data from central nervous system systems is, however, not as straightforward as that from the peripheral nervous system systems described above. The hippocampus contains not only the target neurons of basal forebrain cholinergic neurons but also interneurons that make local connections. The possibility that the early synthesis of NGF in the hippocampus may play a role in the development of 1.hese interneurons is suggested by the coexpression of the NGF receptor gene in the hippocampus at this time (Lu et al., 1989). Further evidence that NGF may exert local actions on neuronal development comes from studies of the cholinergic neurons of the striatum which consist exclusively of nonprojecting interneurons (Bolam et al., 1884; Mesulam et al., 1984). These neurons respond to NGF both in uiuo and in uitro with increases in choline acetyl transferase (ChAT) activity (Martinez et al., 1985). Significantly, NGF and its receptor together with NGF mRNA and NGF receptor mRNA are coexpressed in the striatum with a very similar developmental time-course (Mobley et al., 1989; Lu et al., 1989). Congruent patterns of NGF mRNA and NGF receptor mRNA expression are also observed in the developing cerebellum and olfactory bulb, although the identity of the possible NGF-responsive neurons in these regions is unlcnown (Lu et al., 1989). There is, however, a caveat in accepting that temporally congruent patterns of NGF and NGF receptor expression in the: same region indicate local actions of NGF on neurons. NGF mRNA and NGF receptor mRNA are coexpressed with the same time-course in developing cutaneous target fields (Wyatt et al., 1990). Although the func-

I20

ALUN M. DAViES

tion of NGF receptors in this location is unclear, it is possible that they may play a role in restricting the distribution of NGF in the target field (see below). That NGF receptors may have a similar function in certain parts of the developing CNS cannot be excluded. In situ hybridization studies of the regions of the developing CNS in which NGF mRNA and NGF receptor mRNA are coexpressed may help to resolve this issue. 2 . Developmental Changes in the Synthesis and Level of NGF Similar developmental changes in NGF concentration are observed in sensory and sympathetic target fields during the early stages of their innervation. In the mouse whisker pad the NGF concentration increases steadily for the first 2 to 3 days after the onset of innervation and then falls 4-fold (Davies et al., 1987a). Likewise, in the mouse heart ventricle and submandibular gland the concentration of NGF increases for the first 4 to 5 days after the onset of innervation and then falls over 2-fold (Korsching and Thoenen, 1988). In the whisker pad the fall in NGF concentration coincides with the onset of neuronal death in the trigeminal ganglion. The fall in NGF concentration in the heart ventricle and submandibular gland also appears to be closely related to the onset of sympathetic neuron death as estimated from developmental studies of the rat superior sympathetic ganglion (Wright et al., 1983). As numerous in uitro studies have shown that the survival of NGFsensitive sensory and sympathetic neurons is dependent on the concentration of NGF in the culture medium, it is possible that the faU in target field NGF Concentration during development contributes to the onset of neuronal death. Furthermore, it is conceivable that the relatively high level of NGF in the target field prior to the fall is able to delay the onset of neuronal death until a large proportion of axons have reached the target field, thereby ensuring that the majority of neurons compete for survival during the same period of development (i.e., after the fall in the level of NGF). If the level of NGF were limiting from the earliest stages of target field innervation it is possible that too many neurons would be eliminated before the capacity of the growing target field increased to support the required number of neurons. In speculating about the significance of changes in NGF concentration it is important to bear in mind that measurement of the overall concentration of NGF in a target field does not neccessarily reflect the actual concentration of NGF available to axon terminals. NGF may have a restricted distribution within the target field, and this would have a marked effect on the local concentration of NGF. In the developing mouse whisker pad NGF mRNA appears just before

NGF SYNTHESIS AND RECEPTOR EXPRESSION

121

NGF, and the levels of these molecules increase in parallel during the early stages of innervation (Davies et al., 1987a).Likewise, in the cultured iris a rapid increase in the level of NGF mRNA is followed within hours by a corresponding increase in NGF synthesis (Heumann and Thoenen, 1986). Thlese findings together with the constitutive pathway of NGF secretion (Birth et al., 1984) suggest that the level of NGF synthesis in the peripheral nervous system is governed by the level of NGF mRNA. Thus, the continuous increase in the total amount of NGF mRNA in both the whisker pad (Davies et al., 1987a) and heart ventricle (Clegg et al., 1989) throughout the period axons reach these target fields during development suggests that the fall in the concentration of NGF occurring at the onset of neiironal death is unlikely to be due to a reduction in the level of NGF synthesis. The most likely explanation is removal of NGF from these target fields by axonal transport. Increasing quantities of NGF unaccompanied by its mRNA appear in the trigeminal ganglion (Davies et al., 1987a) and superior cervical sympathetic ganglion (Korsching and Thoenen, 1988; Heumann et al., 1984) shortly after the onset of NGF synthesis in the periphery. The relationship between the developmental changes in the levels of NGF mRNA and NGF protein in the perinatal CNS target fields is not as straightforward as in peripheral nervous system target fields. During the first three postnatal weeks in the rat hippocampus there is a steady increase in the level of NGF mRNA concomitant with the arrival of cholinergic axons in this target field (Auburger et al., 1987).In contrast, the level of NGF remains fairly low initially and increases over 4-fold within the space of a few days at the end of the second postnatal week. Likewise, there is a delay of several days between the appearance of NGF mRNA in the neonatal caudate-putamenand the appearance of NGF protein (Mobley et al., 1989). These findings raise the possibility that in the hippocampus and caudate-putamen the level of NGF mRNA is not rate limiting for the production of NGF. It is possible that other steps such as the proteolytic processing of NGF from its precursor are rate limiting in these locations. It should be borne in mind, however, that developmental changes in the level of NGF in target regions are influenced not only by the rate of NGF synthesis but also by the magnitude of NGF uptake and degradation. Local uptake and degradation of NGF occur not only in the cholinergic interneuroils in the caudate-putamen but may also occur in cholinergic interneuroils in the hippocampus. For these reasons, careful studies of NGF synthesis in cultured hippocampal pyramidal neurons will be required to ascertain whether the rate-limiting step in NGF synthesis differs from that in other NGF-producing cells.

122

ALUN M. DAViES

3. Regulation of NGF Synthesis Knowledge of the factors that regulate the onset and level of NGF synthesis is essential for understanding how trophic interactions are coordinated in development and how regional differences in innervation density are established. Although the arrival of sensory axons in developing mouse skin coincides with the onset of NGF synthesis (Davies et al., 1987a), several studies suggest that the onset of NGF synthesis is not triggered by the arrival of axons in the target field. First, the level of NGF mRNA is the same in normally innervated embryonic chick hindlimbs as in hindlimbs where innervation has been prevented by removal of the corresponding neural primodia prior to sensory and motor axon ingrowth (Rohrer et al., 1988). Second, the level and developmental pattern of expression of NGF mRNA in the heart ventricle of chemically sympathectomized neonatal rats is not significantly different from normal animals (Clegg et al., 1989). Third, when twice the number of sensory axons are compelled to innervate a single hindlimb in juvenile frogs, the total number of surviving sensory neurons supported by this limb is not significantly greater than the number it normally supports (Lamb et al., 19891, suggesting that the level of NGF synthesis is unchanged. Fourth, in reaggregating embryonic hippocampal cell cultures established prior to the onset of innervation, NGF synthesis begins at the same time as it does in uivo (Roback et al., 1990). Several agents have been shown to influence the level of NGF synthesis in certain cell types. Interleukin-l causes a rapid increase in the level of N G F mRNA in primary cultures of adult rat fibroblasts and Schwann cells (Lindholm et al., 1987) This effect of interleukin-1 may be important for increasing NGF synthesis in damaged peripheral nerves as this molecule is released by marcophages that invade nerves undergoing wallerian degeneration (Heumann et al., 1987b). Despite these clear and pronounced effects of interleukin-1 on NGF synthesis, there is no direct evidence that it plays any role in regulating the timing and level of NGF synthesis during development. Indeed, at least in the developing rat brain, the timing of interleukin-I expression is not correlated with that of NGF (Giulian et al., 1988). Activation of the P-adrenergic receptor by isoproterenol in cultured astrocytoma cells also leads to a rapid increase in NGF mRNA and NGF synthesis (Schwartz et al., 1977;Mocchetti et al., 1989). Interestingly, this effect appears to be mediated via elevation of CAMP and induction of c-fos (Quarless and Heinrich, 1987; Mocchetti et al., 1989). Here again, it is unclear whether this response has any relevance for the regulation of NGF synthesis in development,

NGF SYNTHESIS AND RECEPTOR EXPRESSION

123

V. Biochemistry and Molecular Biology of NGF Receptors

Elinding studies have revealed that NGF-dependent neurons possess two types of specific, saturable NGF receptors (Sutter et al., 1979; Riclpelle et al., 1980; Oleander and Stach, 1980; Godfrey and Shooter, 1986). The predominant type, designated the low-affinity NGF receptor, has a dissociation constant of 2 x 10-9M and releases NGF rapidly after binding ( t 1 / 2 = 10 seconds). The minor type, designated the high-affinity NGF receptor, has a dissociation constant of 2 x 10-"M and releases NGF slowly after binding ( t 1 / 2 = 30 minutes). Photoaffinity crosslinking and immunoprecipitation studies have shown that the molecular weight of the low-affinity receptor is about 80,000 while that of the high-affinity receptor is 130,000 to 140,000 (Puma et al., 1983; Burser et al., 1983; Ross et al., 1984; Hosang and Shooter, 1985). Further biochemical characterization of the low-affinity receptor (Grob et al., 1985) and the isolation of cDNAs encoding this receptor (Heuer et al., 1990; Large et al., 1989; Radeke et al., 1987;Johnson et al., 1986; Chao et al., 1986) indicate that it is a single glycosylated protein. Several observations suggest that this glycoprotein is also a component of the high-affinity receptor. First, the low-affinity receptor is converted to the high-affinity receptor by the binding of NGF (Landreth and Shooter, 1980) and by agents that cluster receptors in the plane of the cell membrane such as wheat germ agglutinin and anti-NGF antibodies (Vale and Shooter, 1982, 1983; Grob and Bothwell, 1983). Second, the low-affinity NGF receptor mE.NA is present in cell lines that express exclusively either the lowaffinity receptor or the high-affinity receptor (Chao et al., 1986). Third, introduction of a recombinant retrovirus containing the low-affinity NGF receptor cDNA into NGF receptor-deficient PC 12 cells restores both the low-affinity and high-affinity receptors in these cells (Hempstead et al., 1989). These studies suggest that the high-affinity NGF receptor is comprised of the low-affinity receptor associated with an additional protein or proteins. In this respect NGF receptors resemble interleukin-2 receptors where the high-affinity binding state is also dependent on the expression of two distinct binding proteins (Teshigawara et al., 1987). The amino acid sequences of the low-affinity NGF receptor in chicken, rat, and man have been deduced from the nucleotide sequences of corresponding cDNAs (Heuer et al., 1990; Large et al., 1989; Radeke et al., 19117; Johnson et al., 1986). These cDNAs encode an integral membrane protein of 396 to 399 amino acids comprised of an extracellular domain of 220 to 222 amino acids, a single transmembrane domain of 22 to 24 amino acids, and a cytoplasmic domain of 151 to 155 amino acids. In addition, there is a signal peptide of 19 to 29 amino acids at the amino terminal end of

124

ALUN M. DAVIES

the coding sequence. The extracellular domain contains two putative glycosylation sites and four cysteine-rich repeat elements of 35 to 42 amino acids in which the positions of the cysteines are highly conserved. In contrast to other growth factor receptors, the cytoplasmic domain lacks the consensus sequence of an ATP-binding domain, suggesting that NGF does not activate an endogenous kinase in the receptor. Both the transmembrane domain and the four cysteine-rich elements of the extracellular domain are highly conserved between species. The low-affinity NGF receptor gene is present as a single copy located on chromosome 17 of the human genome (Chao et af., 1986; Huebner et al., 1986) and consists of six exons spanning 23 kb (Sehgal et al., 1988). Interestingly, the promoter region is very GC rich and consensus transcription sequences such as TATA and CAAT are absent. In this respect the low-affinity NGF receptor gene resembles genes that are involved in growth regulation, such as c-Harvey ras, c-Kirsten ras, and the EGF receptor, and genes that encode enzymes with housekeeping functions. The binding of BDNF to BDNF-responsive neurons is also mediated by low-affinity and high-affinityreceptors that have remarkably similar dissociation constants to those of the corresponding NGF receptors (Rodriguez-Tebar and Barde, 1988). The demonstration that cells transfected with the low-affinity NGF receptor gene develop low-affinity binding for BDNF as well as NGF and that BDNF and NGF are equally effective in competing for binding at this receptor suggests that the lowaffinity NGF and BDNF receptors are the same molecule (RodriguezTebar et al., 1990). In contrast, a 1000-fold excess of the heterologous ligand is required to reduce the binding of either BDNF or NGF to their respective high-affinity receptors. This suggests that the proteins that associate with the low-affinity receptor to confer the high-affinity binding state also confer ligand specificity. in this respect, neurotrophic factor receptors resemble integrins where ligand specificity is determined by the particular a- and /3-subunit types in the receptor heterodimer (Hynes, 1987). As discussed below, the evidence for a common low-affinity receptor for NGF and BDNF and possibly other related neurotrophic factors such as neurotrophin-3 has important implications in the reinterpretation of studies of the distribution of low-affinity NGF receptors. Several findings suggest that the survival-promotingand neurite growthpromoting effects of NGF are mediated exclusively via the high-affinity receptor. First, these effects are observed in cultured sensory neurons at concentrations of NGF which result in preferential occupancy of the high-aflinity receptor (Greene, 1977; Griffin and Letourneau, 1980). Second, E4 chick dorsal root ganglion neurons, which possess low-affinity but not high-aflinity NGF receptors (Sutter et al., 1979), are unresponsive to

NGF SYNTHESIS AND RECEPTOR EXPRESSION

125

NCiF (Davies et al., 1986b), but by E6, when high-affinity receptors are first detectable, a large number of these neurons also respond to NGF. Sinlilar studies with BDNF-responsive neurons also suggest that the survival-promoting effects of this factor are also mediated via its highaffinity receptor (Rodriguez-Tebar and Barde, 1988). Furthermore, there is evidence that the induction of c-fos in embryonic neurons by either NGF or :BDNFis also mediated by the corresponding high-affinity receptors for these ligands (Davies, unpublished data). In these respects neurotrophic factor receptors resemble the interleukin-6 receptor where an additional membrane glycoprotein is required for signal transduction (Taga et al., 1989). A further important functional difference between the two kinds of NCiF receptors is that the high-affinity receptor is internalized when bound by NGF whereas the low-affinity receptor does not appear to be internalized when bound (DiStefano and Johnson, 1988). The importance of this difference for interpreting data on the site and timing of NGF receptor expression will be discussed below.

VI. NGF Receptor Expression Several techniques have been used to study the expression of NGF receptors and the NGF receptor gene. Autoradiography has been used to defect the binding of iodinated NGF to cells in culture and in tissue sections. Antibodies against the low-a!XnityNGF receptor have been used to measure NGF receptor levels by immunoassay and to localize NGF receptors in tissue sections by immunohistochemistry. DNA and RNA probes have been used to measure NGF receptor mRNA levels in isolated tissues by Northern blotting and to localize NGF receptor mRNA in tissue sections by in situ hybridization. It is important to bear in mind that im munochemical and nucleic acid hybridization techniques detect only the lowaffinity NGF receptor and its mRNA and provide no indication of whiether or not high-affinity receptors are expressed. Furthermore, although the concentration of iodinated NGF used in autoradiographic studies can be adjusted to provide some indication of whether or not highaffinity receptors are present, the majority of studies that have used this technique to localize NGF receptors have not incorporated this refinement. Thus, caution is required in interpretatingthe functional significance of NGF receptor expression as detected by the above methods given the evidence that the low-affinity NGF receptor is also the lowaffinity receptor for BDNF and possibly other related neurotrophic factors (Rodriguez-Tebar et al., 1990).

126

ALUN M.DAVIES

In addition to the expression of low-affinity NGF receptors on neurons, these receptors are also expressed on a variety of nonneuronal cell types. In the following sections I shall describe in turn the time-course of NGF receptor expression on NGF-dependent and NGF-independent neurons and on other cell types and discuss the functional and developmental significance of these patterns of receptor expression.

A. NGF-DEPENDENT NEURONS I . Onset of NGF Receptor Expression As with studies of the timing of NGF synthesis, the most detailed information on the relationship between the pattern of NGF receptor expression and the early phases of neuronal development (axonal outgrowth, target encounter, and neuronal death) has been obtained from studies of the embryonic mouse trigeminal system (Davies et al., 1987a; Wyatt et al., 1990). Trigeminal neurons were unlabeled by iodinated NGF when cultured before innervating their targets, and the proportion of labeled neurons increased to reach a maximum in cultures established at intervals throughout the period axons reach their targets (Davies et al., 1987a). Although this finding suggests that NGF receptors are first expressed on developing sensory neurons when their axons reach their targets, autoradiography is a relatively insensitive, nonquantitative method of detecting NGF receptors. In a recent study of developmentalchanges in the level of NGF receptor mRNA in the trigeminal ganglion (Wyatt et al., 1990),a low level of NGF receptor mRNA expression in the ganglion was detected prior to the onset of target field innervation. From a knowledge of the number of neurons in the trigeminal ganglion throughout development and the demonstration that NGF receptor mRNA expression in the developing ganglion is restricted to neurons it has been possible to calculate the mean level of NGF receptor mRNA per neuron prior to and throughout the phase of target field innervation. When trigeminal axons are growing to their targets and for about 1 day after the earliest axons have reached their targets the mean level of NGF receptor mRNA per neuron remains constant at about 30 molecules per neuron. The mean level of NGF receptor mRNA per neuron then increases 5-fold throughout the period axons arrive in the target field to reach a second plateau about 1 day after the last axons have reached their targets. Because NGF receptor expression appears to be correlated with the level of NGF receptor mRNA in development (Escandon and Chao, 1989),these findings suggest that the number of NGF receptors on developing sensory neurons increases markedly shortly

NGF SYNTHESIS AND RECEPTOR EXPRESSION

127

after their axons reach the peripheral target field-the stage at which the neurons become labeled by iodinated NGF. The detection of NGF receptor mRNA by in situ hybridization in embryonic chick dorsal root ganglia (DRG) during the earliest stages of axonal outgrowth (Heuer et al., 1990)is also consistent with the view that developing sensory neurons express NGrF receptors prior to innervating their targets. Indeed, the detection of NGrF receptor mRNA in premigratory neural crest by in situ hybridization suggests that at least some undifferentiated neural crest cells may also express NGF receptors (Heuer et al., 1990). The demonstration that the survival of trigeminal neurons is independent of NGF when cultured prior to innervating their targets and becomes dependent on NGF after the onset of target field innervation (Davies and Lumsden, 1984; Davies, 1987a) suggests that the increase of NGF receptors related to target field innervation coincides with the onset of NGF dependency. Likewise, a phase of NGF-independent survival has also been described in newly differentiated DRG neurons (Ernsburger and Rohrer, 1988) and sympathetic neurons (Coughlin and Collins, 1985; Rolirer and Thoenen, 1987; Ernsburger ef al., 1989) grown in culture. It will be of interest to ascertain whether the onset of NGF dependency in these neurons is also related to an increase in the expression of NGF receptors on these neurons. The presence of a low level of NGF receptors on developing neurons prior to the onset of NGF dependency raises the question of the function of these receptors. One possibility is that they may play a role in coordinating the onset of NGF dependency with the arrival of axons in the target field. The: binding of NGF to these receptors when the axons reach their targets may lead to an increase in receptor expression and the onset of dependency. The likelihood that target field NGF increases the level of NGF receptors on the innervating neurons is supported by the finding that NGF upregulates its receptor in PC12 cells (Landreth and Shooter, 1980; Dolierty et af., 1988) and in adult DRG neurons (Verge et al., 1989; Lindsay et al., 1990). Further evidence that target-derived neurotrophic factors regulate the onset of neurotrophic factor dependency in early neurons comes from the finding that transient exposure of newly differentiated nodose ganglion neurons to BDNF in uitro accelerates the acquisition of IBDNF dependency (K.S.Voge1 and A.M. Davies, unpublished findings). The binding of target-derived neurotrophic factors to receptors on early neurons is not, however, the only mechanism for controlling the onset of neurotrophic factor dependency. Early cranial sensory neurons display differences in the duration of BDNF-independent survival that are correlated with the distance their axons have to grow to reach their targets (K.S. Vogel and A.M. Davies, unpublished findings). It is possible that the

128

ALUN M. DAVIES

early expression of neurotrophic factor receptors on developing neurons may be involved in fine-tuning the onset of neurotrophic factor dependency with the arrival of axons in the target field. Although studies of the embryonic mouse trigeminal system have demonstrated changes in the level of NGF receptor expression related to target field innervation, the issue of whether changes occur in the kinds of NGF receptors expressed during development has not been addressed in this system. Receptor binding studies using embryonic chick DRG neurons suggest that only low-affinity receptors are expressed at E4, whereas low-affinity and high-affinity receptors are present by E6 (Sutter et al., 1979). The finding that DRG neurons are unresponsive to NGF at E4 (Davies et al., 1986b) and have begun to respond to NGF by E6 raises the possibility that the onset of NGF responsiveness is additionally related to the appearance of high-affinity NGF receptors in developing sensory neurons. This does not appear to be the case for embryonic chick sympathetic neurons, however, Receptor-binding studies have shown that both highaffinity and low-affinity receptors are present from E6.5 to E20 (Godfrey and Shooter, 1986). At E7 sympathetic neurons survive and grow in culture independently of NGF (Ernsberger et al., 1989). Thus, the onset of NGF responsiveness in these neurons at E8 is apparently not correlated with the de nouo expression of high-affinity NGF receptors.

2 . Developmen fa1 Changes in NGF Receptor Expression During and beyond the stage of target field innervation there are marked developmental changes in the levels of NGF receptor mRNA and NGF receptors in sensory ganglia. The interpretation of such changes is complicated by the expression of low-affinity NGF receptors on Schwann cells at these later stages (Yan and Johnson, 1987; Scarpini et al., 1988). Studies on dissociated DRG cell suspensions (containing both neurons and nonneuronal cells) have revealed that between El4 and El6 in chick embryos there is a 4- to 6-fold decrease in capacity of the cells to bind iodinated NGF specifically (Herrup and Shooter, 1975). Likewise, the intensity with which chick DRG and neural crest-derived cranial sensory ganglia are labeled with iodinated NGF in tissue sections also decreases over this same period of development (Raivich et al., 1985, 1987). NGF receptor mRNA levels are also markedly lower in 3-month-oldrat DRG than at birth (Buch et al., 1987) and immunoprecipitation studies have shown that the level of the NGF receptor per milligram protein is also appreciably lower in adult rat DRG than in late fetal DRG (Yan and Johnson, 1987). The finding that the proportion of embryonic chick DRG neurons specifically labeled

NGF SYNTHESIS AND RECEPTOR EXPRESSION

129

with iodinated is lower in El6 cultures compared with El0 cultures and that the proportion of labeled neurons in El0 cultures decreases after several days (Rohrer and Barde, 1982)suggests that at least a proportion of sensory neurons lose NGF receptors during development. This apparent decrease in the level of NGF receptors on developing DRG neurons correlates with the reduction in the magnitude of neurite outgrowth elicited by NGF from DRG explants (Hemp and Shooter, 1975). Although several in uiuo and in uitro studies suggest that the majority of adult DRG neurons are not dependent on NGF for survival (LeviMontalcini and Angeletti, 1968; Johnson et al., 1980; Rich et al., 1984; Johnson and Yip, 1985; Lindsay, 1988), a proportion of DRG neurons continue to express high-affinity NGF receptors (Verge et al., 1989). In mature DRG neurons these receptors probably mediate the effects of NGF on neuropeptide synthesis (Lindsay and Harmer, 1989). In contrast to the decrease in the level of NGF receptors on maturing sensory neurons, several observations suggest that the level of NGF receptors increases on sympathetic neurons from birth to adulthood. The level of NGF receptor mRNA in rat sympathetic ganglia increases 5- to 10-foldfrom birth (Bucket al., 1987)and the level of NGF receptor protein increases 4- to 5-fold from birth (Yan and Johnson, 1987). NGF continues to play an important role in adult sympathetic neurons in regulating the synthesis of noradrenaline (Thoenen et al., 1971). It is unclear, however, why NGF receptor levels should continue to increase in these neurons in contrast to the decrease in NGF receptor expression on sensory neurons. Developmental patterns of NGF receptor expression have also been described in the cholinergic neurons of the basal forebrain and caudateputamen. NGF binding studies on membrane fractions obtained from the rat caudate-putamen (Mobley et al., 1989)indicate that high-anity NGF receptors are present from at least birth onwards. There is 2-fold increase in the level of these receptors during the third post-natal week concomitant with a marked rise in the level of choline acetyl transferase in the caudateputamen. The level then remains substantially unchanged into adulthood. Although NGF receptor mRNA is also detectable in the adult caudateputamen (Ernfors et al., 1988), it has not been possible to detect NGF receptors by immunohistochemistry in this location (Yan and Johnson, 1988; Schatteman et al., 1988; Eckenstein, 1988; Gage et al., 1989). This may simply be due to the relatively low density of NGF receptors in the caudate-putamencompared with the basal forebrain. That NGF may continue to influence these neurons in the adult is suggested by the demonstration that intraventricular infusion of NGF or local injections of NGF in the caudate-putamen increases the size and choline acetyl transferase immunoreactivity of these neurons (Gage et al., 1989; Hagg et al., 1989).

130

ALUN M.DAVIES

B. NGF-INDEPENDENT NEURONS The recent experimental evidence that the low-affinity NGF and BDNF receptors are the same molecule (Rodriguez-Tebar er al., 1990)provides a satisfying explanation for the previously perplexing observation that BDNF-dependent neurons that do not depend on NGF for survival bind radiolabeled NGF and express the NGF receptor gene. For example, proprioceptive neurons of the trigeminal mesencephalic nucleus, which are supported in culture by BDNF but not NGF, bind iodinated NGF specifically (Davies et af., 1987b). Likewise, developing nodose ganglion neurons express the NGF receptor gene (Heuer et al., 1990; Hallbook et al., 1990) but are not dependent on NGF for survival either in uiuo (Pearson et al., 1983; Hofer and Barde, 1988) or in uitro (Davies and Lindsay, 1985; Lindsay and Rohrer, 1985)and fail to take up and transport iodinated NGF injected into their peripheral target field (Stoeckel et al., 1976; Johnson et al., 1978). BDNF, however, promotes the survival of these neurons both in uiuo (Hofer and Barde, 1988)and in uitro (Lindsay et al., 1985; Davies et al., 1986b). The weak labeling of nodose ganglion neurons by iodinated NGF in autoradiographic studies compared with more intense labeling of NGF-dependent neurons (Raivich et al., 1987; Hallbook et al., 1990) suggests that nodose ganglion neurons express the low-affinity but not the high-affinity NGF receptor. The transient expression of the low-aflinity NGF receptor gene and low-affinity NGF receptors has also been demonstrated by in situ hybridization and immunocytochemistry in embryonic motoneurons (Ernfors et al., 1988,1989; Heuer et af.,1990; Hallbook er al., 1990; Yan and Johnson, 1988). In addition, competition assays of the binding of iodinated NGF to frozen sections indicate that motoneurons bind NGF with the same aflinity as DRG and sympathetic neurons, which suggests that motoneurons also express high-affinity NGF receptors (Raivich et al., 1985, 1987). Consistent with the expression of high-aflinity receptors is the finding that at the time of peak receptor expression, iodinated NGF can be retrogradely transported in a specific receptor-mediated manner that appears identical to that seen in sympathetic, sensory, and basal forebrain cholinergic neurons (Yan et af., 1988; Wayne and Heaton, 1990a). Although NGF receptor expression in developing motoneurons is maximal around the time of natural neuronal death and declines to undetectable levels shortly afterward, the available evidence suggests that motoneurons do not depend on NGF for survival. Exogenous NGF does not prevent either naturally occurring death of motoneurons in chick embryos (Oppenheim et al., 1982) or axotomy-induced death of motoneurons in newborn rats (Miyata er a!., 1986; Yan et al., 1988). NGF fails to promote the

NGF SYNTHESIS AND RECEFTOR EXPRESSION

131

survival of identified chick motoneurons in dissociated culture and does not potentiate the survival-promoting effect of muscle extract on these neurons (Dohrmann et al., 1987;Wayne and Heaton, 1990b. Furthermore, in contrast to the effects of NGF on basal forebrain cholinergic neurons, NCiF does not increase choline acetyltransferase activity in motoneurons (Yam et al., 1988).Although NGF does not appear to promote the survival of motoneurons, it has been shown to enhance the extent of neurite outgrowth from cultured embryonic motoneurons (Wayne and Heaton, 19SQb). The significance of this in uitru observation for normal development is as yet unclear.

C. OTHERCELLTYPES Several recent studies have revealed that during development the lowaffinity NGF receptor gene and low-affinity NGF receptors are expressed in several nonneural tissues (Yan and Johnson, 1988; Ernfors et al., 1988; Wyatt et al., 1990). In some locations this has no obvious connection with neural development; for example, NGF receptors are expressed in lymphoid tissue and in somites. In other locations low-affinity NGF receptors are expressed in tissues that are innervated by NGF-responsive neurons. For example, NGF receptors are expressed in developing skin and teeth which are innervated in part by NGF-dependent sensory neurons (Yan and Johnson, 1988). The observation that the levels of NGF receptor mRNA and NGF mR.NA increase in parallel throughout the period axons arrive in cutaneous target fields and reach a plateau after the arrival of the last axons (Wyatt et al., 1990) is circumstantial evidence for a role of NGF receptor expression in the target field regulating some aspect of innervation . One possibility is that the expression of NGF receptors on certain target field cells is a way of restricting the distribution and availability of NGF to particular cells or regions in the target field. Low-affinity NGF receptors, which are not internalized when bound by NGF (DiStefano and Johnson, 1988),would provide a means of holding NGF on the surface of these cells until sequestered by the high-affinity receptors on the axons of NGFdependent neurons. Such restricted availability of NGF within the target field is an attractive mechanism for selectively supporting the survival of neurons in accordance with the location of their axon terminals in the target field (Davies, 1987a) and also for refining the distribution of axon tenminals in the target field. The expression of NGF mRNA in the presumptive dermal mesenchyme of developing skin (Wyatt et al., 1990) enables NGF, which is primarily synthesized in the overlying epithelium

132

ALUN M. DAVIES

(Davies et al., 1987a), to be concentrated in the region where large numbers of axons terminate. A similar mechanism has also been proposed for presenting NGF to growing axons in regenerating peripheral nerves (Taniuchi et al., 1986)because Schwann cells synthesize NGF and express low-affinity NGF receptors after axotomy. Low-affinity NGF receptors, which bind BDNF and possibly also neurotrophin-3 (Rodriguez-Tebar et a / . , 19!30), might play a similar role in restricting the distribution of these latter neurotrophic factors in the relevant target fields. For example, NGF receptors are also transiently expressed in developing muscle (Raivich et al., 1985;Yan and Johnson, 1988; Ernfors et al., 1988) which is innervated in part by neurotrophin-3dependent proprioceptive neurons (Hohn et al., 1990). In addition to the putative role of low-affinity receptors in restricting the distribution of neurotrophic factors in the target field, it is possible that these receptors might also mediate some direct effect of neurotrophic factors on nonneuronal cells.

VII. Conclusions

Studies of NGF synthesis and NGF receptor expression have made a major contribution to our understanding of the molecular basis of trophic interactions in the developing vertebrate nervous system and have provided a theoretical framework for investigations of other neurotrophic factors and their receptors. Although the role of NGF and related factors in mediating trophic interactions is firmly established, the regulation of these interactions remains an enigma. To understand how neuronal populations become matched in size and appropriately interconnected during development, it is necessary to determine what controls the site, timing, and level of neurotrophic factor synthesis and the expression of the appropriate receptors. It is anticipated that the imaginative application of extremely sensitive methods for measuring mRNA (such as those employing the PCR technique) to the appropriate in vivo and in vitro developing neural systems will provide answers to these pressing questions.

ACKNOWLEDGMENTS Research in the author's laboratory is supported by grants from the Medical Research Council, Wellcome Trust, Cancer Research Campaign, Science and Engineering Research Council, Royal Society, and Glaxo USA. My thanks to Kris Vogel for helpful discussions of the manuscript.

NGF SYNTHESIS AND RECEF'TOR EXPRESSION

133

REFERENCES Abraham, J. A., Mergia, A., Whang, J. L., Tumolo, A., Friedman, J., Hjenild, K. A., Gospodarowicz, D., and Fiddes, J. C. (1986). Science 233,545-548. Angeletti, R. H., and Bradshaw, R. A. (1971). Proc. Natf.Acad. Sci. U.S.A. 68,2417-2420. Assouline, J. G., Bosch, P., Lim, R., Kim, I. S., Jensen, R., and Pantazis, N. J. (1987). Deu. Brain Res. 31, 103-1 18. Autwger, G., Heumann, R., Hellweg, R., Korsching, S., and Thoenen, H. (1987). Deu. Biol. 1u),322-328. Ayer-LeLievre, C., Olson, L., Ebendal, T., Seiger, A., and Persson, H. (1988). Science 240, 1339-1341. Bandtlow, C. E., Heumann, R., Schwab, M., and Thoenen, H. (1987). EMBO J. 6,891-899. Barbin, G., Manthorpe, M., and Varon, S. (1984). J. Neurochem. 43,1468-1478. Bade, Y.- A. (1989). Neuron 2, 1525-1534. Barde, Y .-A., Edgar, D., and Thoenen, H. (1982). EMBO J. 1,549-553. Barth, E.- M., Korsching, S., and Thoenen, H. (1984). J. CeffBiof. 99,839-843. Berger, E. A., and Shooter, E. M. (1977). Proc. Natl. Acad. Sci. U.S.A. 74,3647-3651. Bolim, J. P., Wauner, B. H., and Smith, A. D. (1984). Neuroscience 12,711-718. Buck, C. R., Martinez, H. J., Black, I. B., and Chao, M. V. (1987). Proc. Natf. Acad. Sci. U.S.A. 84,3060-3063. Buxser, S. E., Kelleher, D. J., Watson, L., Puma, P., and Johnson, G. L. (1983). J. B i d . Chem. 258,3741-3749. Canipenot, R. B. (1982a). Deu. Biol. 93, 1-12. Canipenot, R. B. (1982b). Deu. Biof. 93, 13-21. Chao, M. V., Bothwell, M.A., Ross, A. H., Koprowski, H., Lanahan, A. A., Buck, C. R., and Sehgal, A. (1986). Science 232,518-521. Chun, L. L. Y.,and Patterson, P. H. (1977). J . Cell Biof. 5,705-711. Clelgg, D. O., Large, T. H., Bodary, S. C., and Reichardt, L. F. (1989).Deu. Biof.W, 30-37. Cohen, P., Sutter, A., Landreth, G., Zimmermann, A., and Shooter, E. M. (1980). J. Biof. Chem. 255,2949-2954. Coughlin, M. D., and Collins, M. B. (1985). Deu. Biol. ll0,392-401. Cowan, W. M., Fawcett, J. W., O'Leary, D. D. M., and Stanfield, B. B. (1984). Science 225, 1258-1265. Darting, T. L., Petrides, P. E., Beguin, E., Frey, P., Shooter, E. M., Selby, M., and Rutter, W. J. (1983). Cold Spring Habor Symp. Quant. Biof. 48,427434. Dawies, A. M. (1987a). Development 101, 185-208. Dawies, A. M. (1987b). Development 100,307-311. Dawies, A. M. (1988a). Trends Genet. 4, 139-143. Dawies, A. M. (1988b). Trends Neurosci. 11,243-244. Dawies, A. M., and Lindsay, R. M. (1985). Deu. Biol. 111,62-72. Davies, A. M., and Lumsden, A. G. S. (1984). J . Comp. Neurol. 253,13-24. Dawies, A. M., and Lumsden, A. G. S. (1986). J . Comp. Neurol. 253,13-24. Dawies, A. M., Thoenen, H., and Barde, Y. A. (1986a). Nature (London)319,497-499. Dawies, A. M., Thoenen, H., and Barde, Y. A. (1986b). J. Neurosci. 6,1897-1904. Dawies, A. M., Bandtlow, C., Heumann, R., Korsching, S., Rohrer, H., and Thoenen, H. (1987a). Nature (London)326,353-358. Dawies, A. M., Lumsden, A. G. S., and Rohrer, H. (1987b). Neuroscience 20,3746. Diamond, J., Coughlin, M., MacIntyre, L., Holmes, M., and Visheau, B. (1987). Proc. Natf. Acad. Sci. U.S.A. 84,6956-6600.

134

ALUN M. DAVIES

DiStefano, P. S., and Johnson, E. M.Jr. (1988). J. Neurosci. 8,231-241. Doherty, P., Seaton, P., Elanigan, T. P., and Walsh, F. S. (1988). Neurosci. Lerr. 92, 222-227. Dohrmann, U., Edgar, D., and Thoenen, H. (1987). Deu. Biol. W, 145-152. Dunbar, J. C., Tregear, G . W., and Bradshaw, R. A. (1984). J. Protein Chem. 3,349-356. Ebendal, T., Larharnmer, D., and Pearson, H. (1986). EMBO J. 5, 1483-1487. Eckenstein, F. (1988). Brain Res. 446,149-154. Edwards, R. H., Selby, M.J., and Rutter, W. J. (1986). Nature (London)319,784-787. Edwards, R. H., Selby, M.J., Mobley, W. C., Weinreich, L., Hruby, D. E., and Rutter, W. J. (1988a). Mol. Cell Biol. 8,2456-2461. Edwards, R. H., Selby, M.J., Garcia, P. D., and Rutter, W. J. (1988b). J. Cell Biol. 263, 68 10-68 15. Edwards. R. H., Rutter, W. J., and Hanahan, D. (1989). Cell58, 161-170. Emfors, P., HallWk, F., Ebendal, T., Shooter, E. M.,Radeke, M.J., Misko, T. P., and Persson, H. (1988). Neuron 1,983-996. Emfors, P., Henschen, A., Olson, L.. and Persson, H. (1989). Neuron 2, 1605-1613. Ernsberger, U.,and Rohrer, H. (1988). Deu. Biol. u6,420-432. Ernsberger, U., Sendtler, M.,and Rohrer, H. (1989). Neuron 2, 1275-1284. Escandon, E., and Chao, V. C. (1989). Deu. Brain Res. 47, 187-1%. Finn. P. J., Ferguson, I. A., Renton, F. J., and Rush, R. A. (1986). J . Neurocyrol. 15, 169- 176. Frazier, W. A., Angeletti, R. A. H., Sherman, R., and Bradshaw, R. A. (1973).Biochemistry 12,3281-3293. Furukawa, S., Furukawa, Y., Satoyoshi, E., and Hayashi, K. (1986). Biochem. Biophys. Res. Commun. l36,57-63. Gage, F. H., Batchelor, P.. Chen, K. S., Chin, D., Higgins, G. A., Koh, S., Deputy, S., Rosenberg, M.B.. Fisher, W., and BjorMund, A. (1989). Neuron 2, 1177-1184. Giulian, D., Young, D., Woodward, J., Brown, D., and Lachman, L. ( 1988). J. Neurosci. 8, 709-714. Godfrey, E. W., and Shooter, E. M.(1986). J. Neurosci. 6,2543-2550. Goedert, M.,Otten. U., and Thoenen, H. (1978). Brain Res. 148,264-268. Greene, L. A. (1977). Deu. Biol. 58, 106-113. Grimn, C. G., and Letourneau, P. C. (1980). J. Cell Biol. 86, 156-161. Grob, P. M.,and Bothwell, M. A. (1983). J. Biol. Chem. 258, 14136-14143. Grob, P. M.,Ross, A. H., Koprowski, H.,and Bothwell, M.A. (1985). J. Biol. Chem. 260, 8044-8049. Gundersen, R. W.,and Barrett, J. N. (1980). J. Cell Biol. 87,546-554. HaUbMk, F., Ayer-Lelievre, C.. Ebendal, T., and Persson, H. (1990). Development 108, 693-704. Hamburger, V., and Yip, H. W. (1984). J. Neurosci. 4,767-774. Harper, S. J., and Davies, A. M.(1990). Development (in press). Hartikka, J . , and Hefti, F. (1988). J. Neurosci. 8,2967-2985. Hatanaka. H., Tsukui, H., and Nihonmatsu, 1. (1988). Deu. Brain Res. 39, 85-95. Hefti, F. (1986). J. Neurosci. 6,2155-2162. Hempstead, B. L., Schleifer, L. S., and Chao, M.V. (1989). Science 243,373-375. Hendry. I. A., and Campbell, J. (1976). J. Neurocytol. 5,351-360. H e m p , K., and Shooter, E. M.(1975). J. Cell Biol. 67, 118-125. Heuer, J. G . , Fatemie-Nainie. S., Wheeler. E. F., and Bothwell, M.(1990). Deu. Biol. 137, 287-304.

NGF SYNTHESIS AND RECEPTOR EXPRESSION

135

Heuinann, R., and Thoenen, H. (1986). J. Biol. Chem. 261,9246-9249. Heumann, R., Korsching, S., Scott, J., and Thoenen, H. (1984). EMBO J. 3,3183-3189. HeumaM, R., Korsching, S., Bandtlow, C., and Thoenen, H. (1987a). J. Cell Biol. 104, 1623- 1631. Heumann, R., Lindholm, D., Bandtlow, C., Meyer, M., Radeke, M. J., Misko, T. P., Shooter, E. M., andThoenen, H. (1987b).Proc. Natl. Acad. Sci. U.S.A. 84,8735-8739. Hofer, M. M., and Barde, Y. -A. (1988) Nature (London) 331,261-262. Hogue-Angeletti, R. A., Frazier, W. A., Jacobs, J. W., Niall, H. D., and Bradshaw, R. A. 1:1976). Biochemistry 15,26-34. Hohn, A., Leibrock, J., Bailey, K., and Barde, Y. -A. (1990). Nature 344,339-341. Hopldns, W. G., and Slack, J. R. (1984). Neuroscience W, 951-956. Hosang, M., and Shooter, E. M. (1985). J. Biol. Chem. 260,655-662. Houenou, L., Prevette, D., and Oppenheim, R. (1989). SOC.Neurosci. Abstr. 15,436. Houlgatte, R., Wion, D., and Brachet, P. (1989). Dev. Brain Res. 47, 171-179. Huebner, K., Isobe, M., Chao, M., Bothwell, M., Ross, A. H., Finan, J., Hoxie, J. A., Sehgal, A., Buck, C. R., Lanahan, A., Nowell, P. C., Koprowski, H., and Groce, C. 111986).Proc. Natl. Acad. Sci. U.S.A. 83, 1403-1408. Hynm, R. 0. (1987). Cell 48,549-554. Johnson, D., Lanahan, A., Buck, C. R., Sehgal, A., Morgan, C., Mercer, E., Bothwell, M., md Chao, M. (1986). Cell 47,545-554. Johnson, E. M., and Yip, H. K. (1985). Nature (London) 314,751-752. Johnson, E. M., Andres, R. Y., and Bradshaw, R. A. (1978). Brain Res. 150,319-331. Johnson, E. M., Gorin, P., Brandeis, L. D., and Pearson, J. (1980). Science 210,916-918. Johnson, J. E., Barde, Y. -A., Schwab, M., and Thoenen, H. (1986). J. Nenrosci. 6, 303 1-3038. Kalcheim, C., Barde, Y. -A., Thoenen, H., and Le Douarin, N. M. (1987). EMBO J. 6, 287 1-2873. Kessler, J. A., and Black, I. B. (1980). Proc. Natl. Acad. Sci. U.S.A. 77,649-652. Korsching, S., and Thoenen, H. (1983). Proc. Natl. Acad. Sci. U.S.A. 80,3513-3516. Korsching, S . , and Thoenen, H. (1988). Dev. Biol. U6,40-46. Korsching, S . , Auburger, G., Heumann, R., Scott, J., and Thoenen, H. (1985). EMBO J. 4, 1389-1393. Korsching, S., Heumann, R., Thoenen, H., and Hefti, F. (1986).Neurosci. Lett. 66,175-180. Krorner, L. F. (1987). Science 235,214-216. Lamb, A. H., Ferns, M. J., and Klose, K. (1989). Dev. Brain Res. 45, 149-153. Lanclreth, G. E., and Shooter, E. M. (1980). Proc. Natl. Acad. Sci. U.S.A. 77,4751-4755. Large, T. H., Bodary, S. C., Clegg, D. O., Weskamp, G., Otten, U., and Reichardt, L. F. 1.1986).Science 234,352-355. Large, T. H., Weskamp, G., Helder, J. C., Radeke, M. J., Misko, T. P., Shooter, E. M., and :Reichardt, L. F. (1989). Neuron 2,1123-1134. Leibrock, J., Lottspeich, F., Hohn, A., Hofer, M., Hengerer, B., Masiakowski, P., 'Thoenen, H., and Barde, Y. -A. (1989). Nature (London) 341, 149-152. Letourneau, P. C. (1978). Deu. Biol. 66, 183-1%. Levi-Montalcini, R. (1982). Ann. Rev. Neurosci. 5,341-362. Levi.Montalcini, R., and Angeletti, P. U. (1968). Physiol. Rev. 48,534-569. LeviMontalcini, R., Aloe, L., Mugnani, E., Oesch, F., and Thoenen, H. (1975). Proc. Natl. .4cad. Sci. U.S.A. 72,595-599. Lindholm, D., Heumann, R., Meyer, M., and Thoenen, H. (1987). Nature (London) 330, 658-659.

136

ALUN M.DAVIES

Lindsay, R. M.(1979). Nature (London)282,80-82. Lindsay, R. M.(1988). J. Neurosci. 8,2394-2405. Lindsay, R. M.,and Harmar, A. J. (1989). Nature (London) 337,362-364. Lindsay, R. M., and Rohrer, H. (1985). Deu. Biol. 112,3048. Lindsay, R. M.,Thoenen, H., and Barde, Y. -A. (1985). Deu. Biol. lU, 319-328. Lindsay, R. M.,Shooter, E. M., Radeke, M.J., Misko, T. P., Dechant, G., Thoenen, H.,and Lindholm, D. (1990). Eur. J. Neurosci. 2,389-3%. Longo, A. M.(1978). Deu. Biol. 65,260-210. Lu, B., Buck, C. R., Dreyfus, C., and Black, 1. B. (1989). Exp. Neurol. 104, 191-199. Maisonpierre, P. Beluscio. L., Squinto, S., Ip, N., Furth, M.,Lindsay, R., and Yancopoulos, G. (1990). Science 247, 1446-1451. Martinez, H.J., Dreyfus, C. F., Miller-Jonakait, G.,and Black, I. B. (1985). Proc. Natl. Acad. Sci. U.S.A. 82,7117-1781. Meier, R., Becker-Andre, M.,Gots, R., Heumann, R., Shaw, A., and Thoenen, H. (1986). EMBO J . 5, 1489-1493. Menesini-Chen. M. G . , Chen, J. S., and Levi-Montalcini, R. (1978). Arch. Irul. Biol. 116, 53-84. Mesulam. M. -M., Mufson, E. J., Levey, A. I., and Wainer, B. H. (1984).Neuroscience 12, 669-686. Miyata, Y., Kashihara, Y ., Homma, S., and Kuno, M. (1986).J. Neurosci. 6,2012-2018. Mobley, W.C., Woo, J. E., Edwards, R. H., Riopelle, R. J., Longo, F. M.,Weskampe, G., Otten, U., Valletta, J. S., and Johnson, M. V. (1989).Neuron 3,655-664. Mocchetti, I., De Bernardi, M. A., Szekely, A. M.,Alho, H., Brooker, G., and Costa. E. (1989). Proc. Natl. Acad. Sci. U.S.A. 86,3891-3895. Montero, C. N., and Hefti, F. (1988). J. Neurosci. 8,2986-2999. Momson, R. S., Shanna. A., De Vellis, J., and Bradshaw, R. (1986). Proc. Natl. Acad. Sci. U.S.A. 83,7537-1541. Nongren, G . , Evendal, T.,Belew, M., Jacobson, C. -O., and Porath, J. (1 980). Exp. Cell Res. WO, 31-39. Oleander, E. J., and Stach, R. W. (1980). J . Biol. Chem. 255,9338-9343. Olson, L. (1967). Z . Zellforsch. Mikrosk. Anat. 81, 155-173. Oppenheim, R. (1981). In "Studies in Developmental Biology" (W.M. Cowan, ed.), pp. 14-133. Oxford University Press, New York. Oppenheim, R. W. (1989). Trends Neurosci. 12,252-255. Oppenheim, R. W., Maderdrut, J. L., and Wells, D. J. (1982). J. Comp. Neurol. 210, 114-189. Otto, D., Unsicker, K.. and Grothe, C. (1987). Neurosci. Lett. 83, 156-160. Owen, D. J., Logan, A., and Robinson, P. P. (1989).Brain Res. 476,248-255. Paimatier, M.A., Hartman, B. K.,and Johnson, E. M. (1984). J . Neurosci. 4,751-756. Pearson, J., Johnson, E. M.,and Brandeis, L. (1983). Deuel. Biol. %,32-36. Puma, P., Bwtser, S. E.. Watson, L., Kelleher, D. J., and Johnson, G. L. (1983). J . Biol. Chem. 258,3310-3315. Purves, D. (1988). "Body and Brain: A Trophic Theory of Neural Connections." Harvard University Press, Cambridge, MA. Purves, D., Snider, W. D., and Voyvodic, J. T. (1988). Nature (London) 336, 123-128. Quarless, S. A., and Heinrich, G. (1987). Mol. Brain Res. 2,235-242. Radeke, M.J., Misko, T. J., Hsu, C., Herzenberg, L. A., and Shooter, E. M. (1987). Nature (London) 325,593-597. Raivich, G . , Zimmerman, A., and Sutter, A. (1985). EMBO J . 4,637-644.

NGF SYNTHESIS AND RECEPTOR EXPRESSION

137

Raivich, G., Zimmerman, A., and Sutter, A. (1987). J. Comp. Neurol. 256,229-245. Rich, K. M., Yip, H. K., Osborne, P. A., Schmidt, R. E., and Johnson, E. M. (1984). J . Comp. Neurol. 230,110-1 18. Rich, K. M., Luszcynski, J. R., Osborne, P. A., and Johnson, E. M. (1987). J. Neurocytol. 16,265-268. Rich, K. M., Disch, S. P., and Eichler, M. E. (1989). J. Neurocytol. 18,569-576. Riopclle, R. J., Klearman, M., and Sutter, A. (1980). Brain Res. 199,63-77. Robick, J. D., Large, T. H., Otten, U., and Wainer, B. H. (1990). Deu. Biol. 137,451-455. Rodriguez-Tebar, A., and Barde, Y. -A. (1988). J. Neurosci. 8,3337-3342. Rodriguez-Tebar, A., Jeffrey, P. L.,Thoenen, H., andBarde, Y. -A. (1989). Deuel. Biol. l36, 2%-303. Rodriguez-Tebar, A., Dechant, G., and Barde, Y. -A. (1990). Neuron 4,487-492. Rohrer, H., and Barde, Y. -A. (1982). Deu. Biol. 89,309-315. Rohrer, H., and Thoenen, H. (1987). J. Neurosci. 7,3739-3748. Rohrer, H., Heumann. R., and Thoenen, H. (1988). Deu. Biol. 128,240-244. Rosenthal, A., Goeddel, D. V.,Nguyen, T., Lewis, M., Shih, A., Laramee, G. R., Nikolics, K., and Winslow, J. W. (1990). Neuron 4,767-773. Ross, A. H., Grob, P., Bothwell, M., Elder, D. E., Emst, C. S., Marano, N., Ghrist, B. F. D., Slemp, C. C., Hermlyn, M., Atkinson, B., and Koprowski, H. (1984). Proc. Natl. Acad. Sci. U.S.A. 81,6681-6685. Rubin, E. (1985). J. Neurosci. 5,673-684. Rush, R. A. (1984). Nature (London)312,364-367. Scarpini, E., Ross, A. H., Rosen, J. L., Brown, M. J., Rostami, A., Koprowski, H., and Lisak, R. P. (1988). Deu. Biol. 125,301-310. Schatteman, G. C., Gibbs, L., Lanahan, A. A., Claude, P., and Bothwell, M. (1988). J . Neurosci. 8,860-873. Schwartz, J. P., Chuang, De-Waw, and Costa, E. (1977). Brain Res. l37,369-375. Schwarz, M. A., Fisher, D., Bradshaw, R. A., and Isackson, P. J. (1989).J. Neurochem. 52, 1203- 1209. Scott, J., Selby, M., Urdea, M., Quiroga, M., Bell G. I., and Rutter, W. J. (1983). Nature (London)302,538-540. Sehgal, A., Patil, N., and Chao, M. (1988). Mol. Cell Biol. 8,3160-3167. Selby, M. J., Edwards, R., Sharp, F., and Rutter, W. J. (1987). Mol. Cell Biol. 7,3057-3064. Shelton, D. L., and Reichardt, L. F. (1984). Proc. Natl. Acad. Sci. U.S.A. 81,7951-7955. Shelton, D. L., and Reichardt, L. F. (1986). Proc. Natl. Acad. Sci. U.S.A. 83,2714-2718. Snider, W. D. (1988). J. Neurosci. 8,2628-2634. Stoeckel, K., Solomon, F., Paravicini, U., and Thoenen, H. (1975). Nature (London) 250, 150- 15 1. Stoeckel, K., Gurroff, G., Schwab, M., and Thoenen, H. (1976). Brain Res. 109,261-284. Stockli, K. A., Lottspeich, F., Sendtner, M., Masiakowski, P., Carroll, P., Gotz, R., Lindholm, D., and Thoenen, H. (1989). Nature (London)342,920-923. Stoll, G., and Muller, H. W. (1986). Neurosci. Lett. 72,233-238. Sutter, A., Riopelle, R. J., Hams-Warrick, R. M., and Shooter, E. M. (1979). J. Biol. Chem. 254,5972-5982. Taga., T., Hibi, M., Hirata, Y., Yamasaki, K., Yasukawa, K., Matsuda, T., Hirano, T., and Kishimoto, T. (1989). Cell 58,573-581. Tanztka, H. (1987). Deu. Biol. W, 347-357. Taniuchi, M., Clark, H. B., and Johnson, E. M. (1986). Proc. Natl. Acad. Sci. U.S.A. 83, 4094-4098.

138

ALUN M.DAVIES

Tams, R. H., Weichsel, M. E. Jr., and Fisher, D. A. (1986). Pediatr. Res. 20, 367-372. Teshigawara, K., Wang, H. -M., Kato, K.,and Smith, K. A. (1987). J. Exp. Med. 165, 223-238. Thoenen, H., and Barde, Y. -A. (1980). Physiol. Rev. 60,1284-1335. Thoenen, H., Angeletti, P. U., Levi-Montalcini, R., and Kettler, R. (1971). Proc. Natl. Acad. Sci. U.S.A. 68, 1598-1602. Ullrich, A., Gray, A., Berman C., and Dull, T. J. (1983). Nature (London) 303,821-825. Vale, R. D., and Shooter, E. M. (1982). J . Cell Biol. 94,710-717. Vale, R. D., and Shooter, E. M. (1983). Biochemistry22,5022-5028. Verge, K. M. K., Riopelle, R. J., and Richardson, P. M. (1989). J . Neurosci. 9,914-922. Wdicke, P. A. (1988). J. Neurosci. 8,2618-2627. Wayne, D. B., and Heaton, M. B. (1990a). Deu. Biol. 138,473-483. Wayne, D. B., and Heaton, M. B. (1990b). Deu. Biol. l38,484-498. Whittemore, S . R., Friedman, P. L.. Larhammar. D., Persson, H., Gonzalez-Carvajal, M., and Holets, V. R. (1988). J. Neurosci. Res. u),403-410. Williams, L. R., Varon, S., Peterson, G. M.,Wictorin, K., Fischer, W., Bjorklund, A., and Gage, F. H . (1986). Proc. Natl. Acad. Sci. U . S . A . 83,923 1-9235. Williams, R.,Gaber, B., and Gunning, J. (1982). J . Biol. Chem. 257,3321-3323. Wtodawer, A., Hodgson, K.O., and Shooter, E. M. (1975). Proc. Natl. Acad. Sci. U.S.A. 72, 777-779. Wright, L. L., Cunningham, T. J., and Smolen, A. J. (1983). J. Neurocyrol. l2,727-738. Wyatt, S . , Shooter, E. M..and Davies, A. M. (1990). Neuron 4,421-427. Yan, Q.,and Johnson, E. G. Jr. (1987). Deu. Biol. 121, 139-148. Yan, Q.. and Johnson, E. G. Jr. (1988). J . Neurosci. 8,3481-3498. Yan, Q., Snider. W. D., Pinzone. J. J., and Johnson, E. M.(1988). Neuron 1,335-343.

INTERNATIONAL REVIEW OF CYTOLOGY, VOL. 128

Control of Calcium Regulating Hormones in the Vertebrates: Parathyroid Hormone, Calcitonin, Prolactin, and Stanniocalcin S. E. WENDELAARBONGA*AND P. K. T. PANG?

* Department of Animal Physiology, Faculty of Science, University of Nijmegen, 6525 ED Nijmegen, The Netherlands f Department of Physiology, School of Medicine, University of Alberta, Edmonton, Alberta, Canada T6G 2H7

I. Introduction Control of ionic calcium levels of the extracellular fluids is mandatory for all vertebrates, since only minor changes of calcium, in particular the ionized calcium fraction, have pronounced effects on the permeability of cell membranes to ions and, consequently, on important physiological processes such as muscle contraction, nerve signal transduction, and control of cellular metabolism. Hypocalcemia leads to increased excitability of cellular membranes, which may result in tetany and seizures. Hypercalcemia reduces neuromuscular transmission and induces myocardial dysfunction and lethargy. Both conditions can be lethal, at least for mammals and birds. Lower vertebrates, in particular fish, seem more resistant to chronic changes in extracellular calcium, although they eventually show the same symptoms (Pang, 1971; Peacock, 1980). It is likely that the dramatic effects of extracellularCa2' perturbations are the reason that Ca2' concentrations are more tightly controlled than the concentrations of other extracellular ions. The total of extracellular calcium represents less than 0.1% of total body calcium. About 0.5% is located intracellulairly and 99.4% in the skeleton. The large amounts of calcium in the skeleton not only give strength to the body, but also represent an easily acctssible supply of calcium that can be mobilized when the calcium concentration of the extracellular fluid is low, and a large reservoir in which calcium can be deposited when the extracellular calcium level tends to me. The external sources of calcium are either exclusively dietary in origin, as in the terrestrial tetrapods, or are represented by diet and the ambient water, as in the aquatic vertebrates that possess gills. The gut, or the gut and the gills, respectively, are the organs involved in calcium uptake in terrestrial and aquatic vertebrates. In terrestrial animals calcium is taken up with the food-an episodic process-and transported from the gut to the blood. In general calcium 139 Copyright 8 1991 by Academic Press. Inc. All rights of reproduction in any form reserved.

140

S . E.WENDELAAR BONGA AND P. K.T.PANG

uptake is balanced by renal calcium excretion, with exception of periods of body growth or, in females, periods of reproduction and parental care. During growth, large amounts of calcium are taken up, transported via the blood, and deposited in the skeleton. In female animals large amounts of calcium are accumulated in the eggs or, in mammals, transferred to the offspring via placenta or milk glands. In the aquatic vertebrates the uptake of calcium from the environment is essentially a continuous process and takes place via the gut as well as the gills, for which the water serves as a practically limitless supply. Similar to the terrestrial vertebrates, during periods of growth and, in females, reproduction, substantial amounts of calcium are taken up, transported to the blood, and deposited in the skeleton or the ovaries, respectively. Thus, under all conditions animals experience influx and efflux of calcium into the body circulation, and since this extracellular calcium concentration is maintained at a fairly constant level or “setpoint” (Brown et al., 1987), effective homeostatic control mechanisms must be present. During periods of growth and during the female reproductive periods the influxes and effluxes are highly increased, and consequently the demand on these control mechanisms is much higher than usual. As will be discussed later in this review, the setpoint for the concentration of ionic calcium of the extracellular fluid is obviously not fixed for a lifetime in an individual, and we feel that the homeostatic control of calcium may be more flexible than often has been assumed, in particular in the lower vertebrates. Nevertheless, it is beyond doubt that the ionic calcium concentration generally is controlled within narrow limits and that this control is effected by complex hormonal mechanisms. Many hormones have effects on the handling of calcium at the cellular and organismal levels, which is obviously connected with the circumstance that calcium has so many diverse functions. Traditionally, parathyroid hormone (FTH) and calcitonin (CT) are considered the main calcium-regulating hormones for hypercalcemic and hypocalcemic control, respectively, by regulating calcium influx and efflux to the blood at the level of the gut, the bone and the kidneys. Vitamin D, in particular its metabolite l125(OH)2D3,is the third important calcium-controllingfactor, but its action concerns total body calcium balance rather than homeostasis of the extracellular fluid. The steroid will therefore not be considered in this review. PTH rather than CT is considered the dominating hormone in calcium homeostasis, mainly because the plasma calcium level has the tendency to fall in the absence of PTH, whereas CT deficiency is often without serious consequences. It may be surprising, given the importance of FTH for calcium homeostasis in mammals and birds, that true parathyroid glands are present only

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

141

in the terrestrial vertebrates and in those amphibians that spend most of their lifetime outside the water (see Section 11,A). These glands have not been found in typical aquatic vertebrates: fishes and amphibians with a predominantly aquatic lifestyle, such as many salamandrids (Fig. 1). The absence of PTH in the aquatic vertebrates contrasts with the presence of calcitonin in all vertebrates except the jawless fishes (see Sections I1,B and 111,IB). A hypercalcemic hormone homologous with PTH has not been identified in animals without parathyroid glands, although prolactin, a hormone that occurs in all major vertebrate groups, exerts effects in several amphibians and fish that show some similarities with the action of PTH. The characteristics of prolactin as a calcium-regulating hormone will be discussed in Section II1,A. In this review we want to stress that there is what we consider a fundamental difference between the homeostatic control of plasma calcium levels between terrestrial and aquatic vertebrates. We conclude from the available literature that in the aquatic vertebrates the plasma calcium 1eve:ls do not fall in the absence of regulatory hormones, but may remain constant or tend to rise. The evidence will be discussed in Section IV. This can explain why a predominating hypercalcemic hormone such as PTH is not found in the purely aquatic animals, and, consequently, why the parathyroid glands may have evolved during the water-to-land transition. This new concept also provides the explanation for the presence of a potent hypocalcemic or antihypercalcemic hormone in bony fishes that is unique for these animals: stanniocalcin, the hormone of the Stannius corpuscles (Section 111,C; Fig. 1). PTH PRL

CT

STC

mamma 1s birds TETIUPODS

reptiles amphibians

anurans urodeles

osteichthyans (most modern fishes; lung fishes) FIS:ES

elasmobranchs (e.g.

sharks, rays)

agnathans (jawless fishes)

FIG. 1 . The Occurrence among the vertebrate classes of the four calcium-regulating hormones discussed in this paper.

142

S. E. WENDELAAR BONGA AND P. K. T. PANG

These differences are sufficientreason for us to treat the terrestrial and aquatic vertebrates separately in this review, which is aimed at defining the function of PTH, calcitonin, prolactin, and stanniocalcin in the control of extracellular Ca2+,by analyzing the factors controlling biosynthesis and release of these hormones.

II. Terrestrial Vertebrates A. PARATHYROID HORMONE 1 . Introduction

PTH dominates the control of the Ca” concentration of the extracellular fluid by its calcium-mobilizing action on bone and, indirectly, via its control of the synthesis of I ,25(OH)&, intestinal calcium absorption. PTH further regulates renal phosphate excretion and has anabolic effects on bone tissue. The physiology of the hormone has been reviewed recently by Aurbach (1988) and Hurwitz (1989). PTH is a single-chain nonglycosylated protein. In the few mammalian species in which the primary structure has been elucidated, native PTH contains 84 amino acid residues and has a molecular weight of %oo. Analysis of the PTH sequences of human, rat, cow, pig, and dog has shown a great deal of structural homology (Cohn et al., 1986; Kemper, 1986). The PTH molecule of chicken, the only nonmammalian species of which the hormone has been sequenced so far, consists of 88 amino acids, with striking homology with the mammalian hormones in the 1-32 region (Khosla et al., 1988). Recently a factor sharing homology with the 1-13 region of FTH has been isolated that has affinity for the PTH receptor and is held responsible for humoral hypercalcemia of malignancy (Broadus ez al., 1988; Donahue ei al., 1990).This PTH-related peptide has so far not been isolated from normal PTH tissue and will not be considered further. The structure of the parathyroid glands of the terrestrial vertebrates has been reviewed by Clark et al. (1986). PTH is produced by the predominant cell type of the glands, the chief cells, which occur in a light or inactive phase, and a dark or secretory phase. In many species another cell type of unknown function, the oxyphil, is present. The ultrastructure of the dark cells is remarkably similar in amphibians to mammals and shows the general structure of peptide-producing secretory cells: extensive granular endoplasmic reticulum, Golgi areas, and a varying number of electrondense secretory granules. In general, the parathyroid glands are well vascularized and have a rich nervous supply. In mammals the nervous innervation is mainly adrenergic; in birds both adrenergic and cholinergic.

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

143

In both groups most nerves are associated with the vascular system, indicating that they regulate the blood flow through the gland (Clark et al., 1986). For man and some birds, direct nervous contacts with parathyroid cell:$have been reported. In many avian species the parathyroid gland is associated with a carotid body. The function of this association is unknown (Clark et al., 1986). 2 . Nosynthesis and Release of PTH The synthesis of PTH follows the general pattern of protein hormones. Since any of these steps represents a potential target for extracellular regulatory factors, this process will be discussed briefly. It has been reviewed in detail by Cohn and MacGregor (1981), Habener et al. (1984) and, more recently, by Kemper (1986) and Aurbach (1988). After formation of prepro-mRNA by transcription and posttranscriptional processing of mRNA, the message is translated in the prepro form of the hormone by ribosomes bound to the endoplasmic reticulum. The signal- or presequence typically consists of 25 amino acid residues, and is removed rapidly from the prohormone upon its entering the cisternal space of the endoplasmic reticulum. The prohormone is subsequently transported via vesicles to the Golgi compartment, where additional processing and packaging into secretory granules takes place. The prosequence typically consist!; of six amino acids for all mammalian and avian species examined. The conversion of proPTH to PTH takes place in the Golgi compartment by prolPTH-converting enzyme. Autoradiography of the parathyroid glands in uiuo or in uitro with labeled amino acids has shown that within 2 minutes after administration of the label it is found primarily in the endoplasmic reticulum, after 10 minutes in the Golgi area, and by 20 to 30 minutes in the secretory granules. After packaging of the hormone in the secretory granules, the hormone is either stored in the cytoplasm, released into the circulation by exocytosis, or degraded intracellulary .

3. lntracellular PTH Degradation The blood circulation contains many PTH fragments in addition to the native molecule, as has been reported first by Berson and Yalow (1968). Most of these are carboxy-terminal (C) fragments, and only a very small amount may represent amino-terminal (N) fragments. The high percentage of C-terminal fragments may partially be related to their low turnover rate in circulation when compared to the intact hormone (Silverman and Yalow, 1973). The production of these fragments is located in the Kupffer cell:$in the liver, the kidneys (Hruska et al., 1977; SegrC et al., 1981), and, surprisingly, the parathyroid glands. The intracellular degradation of PTH was first demonstrated for rat parathyroid glands in uitro (Chu et al., 1973).

144

S. E. WENDELAAR BONGA AND P. K. T. PANG

Flueck et al. (1977) identified both C-terminal and N-terminal PTH fragments in addition to intact PTH in the venous effluent blood of parathyroid glands of patients suffering from parathyroid adenomas. The release of intact PTH and PTH fragments from human parathyroid cells in uitro was shown by Hanley and Ayer (1986), in a study on tissue of hyperplastic glands. Evidence that PTH fragments were indeed, and in significant amounts, secreted into the circulation by normal functioning parathyroid glands was reported by Mayer et al. (1979), who showed that venous effluent plasma samples of the parathyroid glands of young calves contained C-terminal and N-terminal fragments. Morissey and Cohn (1979b) showed that, although both newly synthesized and stored PTH are affected, the stored hormone is degraded preferentially. Degradation of newly synthesized PTH did not start until 20 minutes after the formation of proPTH, indicating that PTH and not proPTH is degraded and that this process takes place after packaging in the secretory granules. The degradation has been assumed to take place after fusion of secretory granules with lysosomes (Cohn and Elting, 1983). The primary enzymes responsible for PTH degradation are cathepsins B and D, which, at least in bovine parathyroid cells, yield different C-terminal and N-terminal fragments. The biological activity for most functions of PTH is located in the first (N-terminal) 34 amino acids, which form one of the two domains of the molecule. However, the presence of the carboxy-terminal region 37-84, although in itself without biological activity, ensures that the intact hormone has a higher potency than the 1-34 terminal in several assay systems. Moreover, only 1-84 PTH but not 1-34 PTH is taken up by the liver and stimulates hepatic glucose mobilization, as has been shown in dogs. The C-terminal fragments are considered as biologically inert (Cohn and MacGregor, 1981;Habener et al., 1984). From these structure-activity studies it can therefore be concluded that the intracellular degradation pathway substantially reduces the ratio of active to inactive PTH molecules that are released. The degradation pathway of the parathyroid cells seems part of a remarkable mechanism of control of hormone release that probably is characteristic of the parathyroid glands and which will be discussed in Section II,A,6. 4 . Control of PTH Secretion

a. Extracellular Calcium. Extracellular calcium has been considered the primary regulator of PTH secretion since the pioneering studies of Patt and Luckhardt (1942) about 50 years ago. Following sodium oxalate injection in normal and thyroparathyroidectomized dogs, they observed that

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

145

senim calcium was rapidly reduced and returned to normal only in the intact animals. When they perfused decalcified blood through a thyroid/ parathyroid preparation and injected the perfusate into normal dogs they observed a rise in serum calcium. This effect did not occur when normai blood was used. They concluded that low blood calcium is a direct stimulus for the parathyroid glands to secrete more hormone into circulation. This conclusion has been confirmed frequently by later studies, and the results have shown a close and inverse relationship between extracellular calcium, in particular the ionic calcium fraction, and hormone release by the parathyroid glands, in uiuo and in uitro (Copp and Davidson, 1961; Sherwood et al., 1966; Kemper et al., 1974; Brown, 1983). Whereas in earlier reports this relationship has been described as inversely linear over a wide range of Ca2+ concentrations, more detailed analysis has shown a sigmoid relationship, with PTH release stimulated markedly by small decreases in the physiological range, around the setpoint for Ca2+ (the extracellular Ca2+ concentration resulting in 50% inhibition of the maximum rate of PTH release; 2 1.5 mM), and with a basal release not suppressible by high Ca2+ levels (Habener and Potts, 19786; Brown, 1982, 1983). Around the setpoint, a change in extracellular Ca2+of about 1 mM causes a 5-fold change in PTH release, as has been shown both in uiuo and in uitro (Fig. 2). Not only are FTH cells very sensitive to minute changes in extracellular Ca2+around its setpoint, their response is also extremely rapid. Increased PTH release follows a drop in extracellular CaZ+within seconds in uiuo and in uitro (Brown et al., 1987). In newborn animals plasma calcium may be maintained at a higher level than at an older age. Keaton et al. (1978) and LeBoff et al. (1983) have shown, in uiuo and in uitro, respectively, that the hypercalcemia of newborn calves is most likely caused by a higher setpoint for calcium (Fig. 2). Within a few weeks the setpoint is reduced to the level typical for adult animals. Increased setpoints have also been reported for patients with primary hyperparathyroidism (Brown et al., 1987). The precise mechanism underlying the control of PTH secretion by CaZ+has long been unclear. The observation that the above mentioned effects of extracellular calcium on PTH release are also shown by dispersed parathyroid cells in uitro implies that Ca2+ exerts its effects directly on the cell membrane and not indirectly via hormonal or nervous mechanisms. Bruce and Anderson (1979) have shown that the resting potential of mouse parathyroid cells has a unique sensitivity to changes in the extracellular Ca2+ concentration when compared to other endocrine tissues such as the thyroid gland or the pancreatic beta cells. The authors connected the striking changes in the membrane potential of the parathyroid cells over a small range of ionic calcium concentrations (1 3 - 2 3 mM)

S. E. W E N D E L M BONGA AND P. K.T. PANG

146

C .-

T

T

I 5

6

7

a

0

10

1

PLASMA Ca, mgi lOOml

FIG.2. Comparison of the relationship between immunoreactive PTH release and plasma total calcium concentration in neonatal and older calves, based on measurements of blood from veins draining the parathyroid glands. The values for PTH release in the neonates were grouped according to their associated plasma calcium concentration. Each group spanned a calcium concentration range of 0.5 mg/100 m (0.125 mM). Mean release rate f SE of each group. The relationship between FTH release and plasma calcium was sigmoid in nature in both groups, with the curve for the neonates shifted to the right of the curve of older calves. PTH release was greater at all plasma calcium concentrations in the neonates (Keaton et al., 1978). [O. Neonatal calves (age, 1-3 days); 0,older calves (age, 2-14 weeks).]

with the high calcium sensitivity of the PTH release of these cells in the same concentration range. Calcium ions depolarize the cells by altering the plasma membrane conductance to K+,a capacity that is shared with other divalent ions and with La3+. In contrast to other gland cells, stimulated secretion is associated with membrane hyperpolarization instead of depolarization. This difference emphasizes the uniqueness of the PTH releasecontrolling mechanism. The unusual and prompt response of the PTH cells to extracellular Ca2+ led to the assumption of a Ca2+ sensor or receptor on the outer cell membranes (Lopez-Barneo and Armstrong, 1983; Wallace and Scarpa, 1983). This was supported by Juhlin et al. (19871, who raised monoclonal antibodies to cell surface components of these cells that could block the response of the cells to extracellular calcium. They further demonstrated that the parathyroid cells of patients with hyperparathyroidism and hypercalcemia displayed reduced binding of these antibodies, indicating that the

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

147

reduced sensitivity of these cells to extracellular Ca2+is associated with reduced expression of the antigen. Recently, patch clamp studies by Jia et al. (1988) on bovine PTH cells revealed the presence of a specific type of cakium-activated potassium channel in the outer membrane that might represent the calcium-sensing mechanism. The authors found potassiumselective channels dependent on internal Ca2+ to open. Different from similar channels in other cells, these channels close when the internal Ca2+ coricentration rises above 160 pM. This concentration corresponds to extracellular Ca2+concentrations between 0.5 and 1.0 mM. This unusual behavior of the calcium-dependent potassium channels of the parathyroid cells may explain the depolarization of their outer membranes with increased extracellular Ca2+levels. It is still unclear, however, whether the membrane hyperpolarization of the cells mediates the stimulatory effect of low extracellular Ca2+on PTH secretion, since increased PTH release by external K+ ions is associated with membrane depolarization. Jia et al. (1988) and Pocotte and Ehrenstein (1989) suggested that the specific potassium channels are also present in the secretory granules, and that closing of these channels in response to increased intracellular Ca2+could account for stimulated secretion rather than the hyperpolarization of the ouier cell membrane. On the other hand, Fitzpatrick et al. (1986a,c) have suggested that the interaction of extracellular Ca2+ with voltage-gated Ca:!+ channels mediates the effects of extracellular Ca2+on PTH release (me Section II,A,S,b). Thus, there is no unanimity about the nature of the C$+-sensing mechanism of the outer membrane of the PTH cells. The above data deal with the control by extracellular Ca2+ of PTH release. How does extracellular Ca2+ affect synthesis, processing, and intracellular degradation of the hormone? Whereas the studies mentioned point to prominent and prompt effects of Ca2+on PTH release, extracellular Ca2+appears to have little more than marginal effects on PTH synthesis.. at least in the short term. Heinrich et al. (1983) have determined the coricentration of mRNA in bovine parathyroid slices in uitro. After incubation of the tissue for 5 to 7 hours in the presence of different extracellular Ca'+ concentrations no differences were noticed in total mRNA, but poly(A) containing PTH-mRNA levels were changed. Unexpectedly, in the presence of nonphysiologically high external Ca2+ (5 mM), which inhibit PTH release, poly(A) containing PTH-mRNA amounted to 30% of total mRNA, substantially more than the 10% found in tissue slices incubated in low-Ca2+ media. At variance with these results, Russell et al. (1983) showed that in dispersed bovine parathyroid cells in uitro the concentration of PTH-mRNA was reduced gradually in 96 hours by almost 7056 at high extracellular Ca2+ (2.5 mM) when compared to low Ca2+ (0.5mM). A change from high Ca2+to low Ca2+after 36 hours reversed

148

S. E. WENDELAAR BONGA AND P. K. T. PANG

the decrease. These observations were confirmed by Brookman et al. (1986): in similar experiments on cultured parathyroid cells PTH-mRNA levels were unchanged over 72 hours at normal extracellular Ca2+( 1 mM), and during the first 4 hours at 3 mM Ca2+.At the latter concentration a 50% suppression was observed after 24 and 48 hours. Low external Ca2+ (0.5 mM) was reported to have no effect (Russell et al., 1983) or a slightly stimulating effect by 48 hours (Brookman et al., 1986). These in uitro results were not fully consistent with in uiuo observations in rats. Reduction of serum calcium by sodium phosphate injection resulted in a marked increase of PTH-mRNA at 6 hours. Calcium infusion, causing frank hypercalcemia for 6 hours, did not change PTH-mRNA levels (Naveh-Many et al., 1989). With pulse chase experiments of porcine parathyroid tissue in vitro with labeled amino acids it was found that the rates of synthesis and conversion of proPTH were the same irrespective of the Ca2+concentration of the incubation medium (Morissey and Cohn, 1979b). The absence of consistent effects of extracellular Ca2+on PTH synthesis or processing in short-term experiments is in contrast with more longterm ultrastructural observations in uiuo, that invariably point to an inverse relation between extracellular Ca2+and the secretory capacity of the parathyroid glands. As reviewed by Clark et al. (1986), hypercalcemia induced by exogenous Ca2+ caused degenerative changes in parathyroid cells of different mammalian species, which strongly suggested decreased secretory activity. Hypertrophy and hyperplasia of the glands have been reported in response to experimentally induced hypocalcemia, for instance in cows, cats, dogs, chickens, iguanas, and newts. These studies indicate that prolonged hypercalcemia or hypocalcemia profoundly influence the capacity for hormone synthesis of the glands. They further demonstrate that the ultrastructural responses of the parathyroids to prolonged hypocalcemia and hypercalcemia are similar in mammals to amphibians. Nevertheless, the effects of extracellular Ca" on PTH biosynthesis become noticeable in days rather than hours, and this contrasts with the immediate effects on PTH release. How then are the parathyroid glands, which usually contain only very small hormone stores, capable of maintaining PTH release during the first hours of a prolonged hypocalcemic stimulus?Chu et al. (1973) and Habener et al. (1975) were the first to reveal what seems to be a unique regulatory mechanism by showing that extracellular Ca2+ can effectively modulate the degradation pathway of bovine PTH cells and in uitro, with more degradation and thus less bioactive PTH secreted at increasing extracellular Ca2+levels. This mechanism enables the cells to persistently increase the output of intact hormone without changing the rate of gene transcription or translation. It will be discussed in Section II,A,6.

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

149

b . Extracellular Magnesium. High concentrations inhibit FTH secretion in uitro and in uiuo. It was shown that Mg2+is two-and-a-half to three times less potent than Ca” in inhibiting PTH release in uitro (Habener and Potts, 1976). The effects of Ca2+and Mg2+were reported as independent and additive. In a systematic study on the relative potencies of Ca2+and Mg2+ to inhibit PTH secretion from dispersed bovine parathyroid cells, Brown et al. (1984a) have shown that the relationship between both cations is more complex. At a physiological free calcium concentration (1.0 mM Ca2+),the M 2 + concentration giving half maximal inhibition of PT;H release was 1.8 mM, which was only slightly higher than the corresponding Ca2+ concentration (1.25 m).At a Ca2+ concentration of 0.5 mM, similar to that used by Habener and Potts (1976), the Mg2+concentration giving half maximal inhibition (3 mM) was almost two-and-ahalf times that of Ca2+(1.25 mM) with a Mg2+concentration of 0.5 mM, thereby confirming Habener and Potts’ data. However, at low Ca2+concentrations, Mg2+was markedly less potent as an inhibitor of PTH release than at physiological Ca” levels. Conversely, low Mg2+concentrations had little effect on the responsiveness of PTH release to Ca2+(Brown et al., 1984a). Thus, at low Ca2+concentrations Mg2+was less potent than predicted by the 2.5 :1 ratio suggested by Habener and Potts (1976), and the effect of Mg2+was dependent on the presence of extracellular Ca2+. Shoback et al. (1983) have shown that the enhanced PTH secretion induced by lowering of the Mg2+concentration in vitro is associated with a decrease in cytosolic Ca2+as measured with quin2. In uitro studies of Mahaffee et al. (1982) have shown that in Mg-free meldia PTH secretion is inhibited. Anast et al. (1972, 1976)reported hypocalcemia and reduced immunoreactive plasma PTH levels in patients with abriormally low plasma Mg2+levels, caused by impaired intestinal Mg2+ uptake. Although at fist these observations may seem paradoxical in view of the above mentioned inhibitory effects of high external M$+ levels, these effects probably are not specific for PTH release but rather a reflection of the general dependency on magnesium of many cellular functions. Magnesium is indispensable as a stabilizer of membrane structures, as a chelator of intracellular ATP, and as a cofactor of many enzyme reactions. It is also essential for hormonal activation of the receptor adenylate cyclase complex of many cell types: there are specific magnesium-binding sites. Using homogenates of rat parathyroid glands, Maha€fee et al. (1982) have indeed shown that magnesium ions are essential for adenylate cyclase activation. This dependency probably reflects the action of intracellular M$’ rather than extracellular Mg2+(see Section II,A,S,a). In uiuo data of the effect of administration of magnesium on PTH secretion are scarce, but seem to support the inhibitory effects of high extracel-

150

S. E. WENDELAAR BONGA AND P. K. T. PANG

lular Mg2+concentrations as observed in uitro. Cholst et al. (1984) showed that administration of very high doses of magnesium to pregnant women induced a sharp decrease of plasma immunoreactive PTH (iPTH). In a study on hemodialysed patients, Pletka et al. (1971) demonstrated a 20% decrease of plasma iPTH after increasing the magnesium concentration of the dialysis fluid from 0.75 to I .25 mM for 2 months. In the only study on healthy man known to us, Ferment et al. (1987) compared the effects of magnesium and calcium injections on plasma PTH levels, using a homologous radioimmunoassay (RIA) directed toward the 53-84 C-terminal PTH fragment. After a single dose of magnesium sulphate which resulted in a 50% increase of plasma M$+ after 45 minutes, a 30% reduction of PTH levels was measured. Injection of magnesium pyrrolidone carboxylate had no significant effect on plasma PTH although the rise in plasma M 2 + was identical to that following administration of magnesium sulphate. A lower dose of calcium gluconate produced a similar effect on plasma PTH. The authors concluded that Mg2+,on a molar basis, was less potent than Ca2+in inhibiting PTH secretion. The effect of M$+ was also less sustained than that of Ca2+. We conclude from the above data that changes in extracellular Mg2+ influence PTH secretion. However, there is no clear evidence that Mg2+is of importance as a regulatory factor in the physiological control of PTH secretion. The effects of Mg2+on parathyroid secretion become evident only under conditions of extreme hypo- and hypermagnesemia. As far as hypomagnesemia is concerned, this condition probably leads to magnesium depletion of the parathyroid cells, and the effects on secretion are, therefore, mainly indirect and/or nonspecific. These effects reflect that the parathyroid cells are dependent for proper functioning on adequate extracellular Mg2+ levels, a dependency shared with other cell types. This inhibitory effect of low plasma Mg2+levels on FTH secretion will therefore only become noticeable under pathological conditions. The inhibitory effects of high external Mg2+ levels on PTH secretion are also unlikely to become noticeable, given the normal ionic levels of Ca2+and Mg2+(1.5 and 1.O mM,respectively) and the low potency of Mg2+at normal plasma Ca2+ levels. Thus, variations in extracellular Mg2+ at most will have a small modulating effect on the control of PTH secretion by Ca2+ under normal physiological conditions. c. Hormones and Neurotransmitters. i. Proteins and peptides. In older literature the pituitary gland has been implicated in the control of the parathyroid glands. In a number of studies in the 1930s, which have been surveyed by Campbell and Turner (1942) and Latman (1980), degeneration of the parathyroid glands was reported for hypophysectomized toads,

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

151

pigeons, rabbits, dogs, monkeys, and rats. Anterior pituitary extracts increased serum calcium and induced hypertrophy of the parathyroids of intact dogs, rats, and guinea pigs. Campbell and Turner (1942) concluded from these studies that the stimulatory effects of the pituitary gland on the parathyroids were indirect. More recently, Latman (1980) tested the effects of a pituitary gland extract on parathyroid hormone release by bovine parathyroid glands in uitro and found a stimulatory action on C-terminal PTH and, in particular, on N-terminal PTH. Injection of the extract had no effect on the serum iPTH level in calves, which the author ascribed to the relative insensitivity of this parameter to changes in PTH release. Since TSH, LH, GH, ADH, oxytocin, prolactin, ACTH, MSH, and lipotropin did not effect PTH release, he concluded that there was an unknown pituitary factor that can directly stimulate PTH secretion. Injection of the pituitary extract had no effect on the serum iPTH level in calves, which the author ascribed to the relative insensitivity of this parameter to rapid changes in PTH release (Latman, 1980; Fig. 3). Recently, we have reviewed the literature on the role of growth hormone and prolactin in calcium regulation in mammals and concluded that these hoimones have important direct effects on calcium uptake via the gut during periods of high calcium demand, viz. growth, gestation, and lactation, but are not implicated in the control of calcium homeostasis of the extracellular fluid (Wendelaar Bonga and Pang, 1986). Also indirect effects via modulation of the PTH release are unlikely. Data on iPTH levels in hyperprolactinemic patients are conflicting (Raymond er al., 1982; Scldechte et al., 1983; Fiore et al., 1984). Injection of GH has been reported to increase plasma calcium levels, an effect that may be mediated by stimulation of PTH secretion. However, the doses used were excessive and the effect might be relevant only for explaining the hyperparathyroidism and hypercalcemia in acromegalic patients (Lancer et a / . , 1975).

As will be dealt with in more detail in Section II,B,2,b,i, gastrointestinal hormones are capable of influencing plasma calcium levels, and therefore interactions with the parathyroid glands are indicated. With respect to secretin, stimulatory effects have been reported on bovine PTH release in uitro (Windeck et al., 1978; Sethi et al., 1981). Increases in PTH levels following secretin administration were reported for rats and normal human subjects (Sethi et al., 1983). The doses were rather high, however. Infusion of acid in rat duodenum, known to stimulate specifically the release of endogenous secretion, caused a significant rise in PTH release. Somatostatin caused a dose-related decrease of plasma PTH in rats and monkeys. It reduced PTH release from normal bovine and adenomatous human parathyroid tissue in uitro (Hargis et al., 1978). Somatostatin antiserum

152

S. E. WENDELAAR BONGA AND P. K. T. PANG

120

t

loo 2

p80W

a

0 W

cnwI

c Q.

zww

8

z

9 20 -

3

ap

-r

0 -

-20

-

TIME (MIN) FIG. 3. Stimulation of N-terminal imrnunoreactive PTH release under slightly hypercalcemk conditions (1.6 mM Caz+)by bovine parathyroidglands perfused in uitro with anterior pituitary gland extract (black bar). Perfusion with all known pituitary hormones had no effect; mean percentage change ? SE compared to the level before perfusion (Latman, 1980).

increased plasma FTH in rats, and inhibited FTH release from bovine parathyroid tissue in uitro. The hormone was attributed with a function of local regulator of the parathyroid glands (Williams et al., 1979). ii. Bioarnines. Dopamine stimulates FTH secretion of bovine parathyroid cells in uiuo (Brown et al., 1977) and in uitro (Attie et al., 1980), but its functional role is unknown. Injections of epinephrine and norepinephrine in cows (Blum er al., 1978) and men (Kukreja et al., 1975) increase serum PTH immediately, similar to administration of the P-adrenergic agonist

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

153

isoproterenol. The stimulatory effects are to some extent additive with those of hypocalcemia and are prevented by the p-blocker propranolol. This indicates that these effects are mediated by /3-adrenergicreceptors. In uitro experiments have shown that the p-adrenergic stimulation is exerted directly on the parathyroid cells, and associated with cAMP production (Williams et al., 1979), via a receptor coupled with adenylate cyclase through a stimulatory guanine nucleotide regulatory protein (Fitzpatrick et al., 1986b,c). a-adrenergic agents inhibit cAMP production and hormone release from bovine parathyroid glands in uitro (Brown et al., 1978a). This explains the biphasic effect of epinephrine, a mixed a-adrenergic and p-adrenergic agonist. The p-adrenergic stimulation is observed in vitro at low concentrations (up to 10-7 M )whereas at higher concentrations an inhibition becomes apparent. This inhibition is an a-adrenergic response since it is prevented by phentolamine (Brown et al., 1978a). The observation that the inhibitory effect of high doses of epinephrine is blocked by pertussis toxin indicates that the a-adrenergic inhibition of PTH secretion is mediated by the inhibitory guanine nucleotide regulatory protein (Fitzpalrick et al., 1986b). The physiological significance of the a-adrenergic inhibition on PTH secretion is unknown. Conversely, the p-adrenergic influence may be physiologically important since propranolol reduced basal serum levels of PTH in man (Kukreja et al., 1975). Epinephrine was suggested to modulate PTH secretion in uiuo under stress conditions: the rise of serum PTH during insulin-induced hypoglycemic stress has been attributed to the accompanying rise of (adreno-medullary) plasma epinephrine (Shah ef al., 1975). This has not been confirmed by Body et al. (1983). The marked rise in plasma epinephrine they recorded following injection in healthy man was paralleled by a reduction of plasma PTH, which was interpreted as an effect of hypercalcemia induced by the expenmental treatment. Moreover, epinephrine injections resulting in plasma epinephrine concentration up to the physiological maximum did not influence plasma PTH (or CT). They concluded therefore that /3-adrenergic stimulation requires nonphysiologically high adrenomedullary epinephrine levels. Heath et al. (1980) examined PTH secretion in adrenalectonnized rats and found no indication of important adrenomedullary effects. This seems to limit the function of the p-adrenergic control to catecholaminesoriginatingfrom the sympathetic nerves in the parathyroid glands (Body et al., 1983; Heath et al., 1980). However, chemical sympalhectomy by 6-OHDA treatment did not interfere with the homeostatic control of plasma calcium by PTH, and thus the physiological importance of 'the p-adrenergic control remains unclear. iii. Steroids. PTH as well as 1,25 dihydroxycholecalcirol (1 ,25(OH)2D3)are both important hormones regulating calcium homeosta-

154

S.E. WENDELAAR BONGA AND P. K . T. PANG

sis of the extracellular fluids. There is, however, a clear difference between the actions of both hormones: PTH dominates the short-term regulation of plasma calcium and can become effective within minutes; the action of 1,25(OH)2D3 becomes noticeable after hours or days, and, as mentioned earlier, relates to the total body calcium balance rather than to the calcium homeostasis of the extracellular fluid. PTH controls 1 ,25(OH)2D3 formation by regulating 25 hydroxycholecalciferol-1 ahydroxylase, a key enzyme in the synthesis of the steroid (Fraser and Kodicek, 1973). Conversely, 1,25(OH)2D3 influences PTH secretion. It acts directly on the parathyroid cells. High affinity receptors for 1 ,25(OH)2D3have been demonstrated in the cytosol of human and chicken parathyroid cells (Oldham et al., 1974; Hughes and Haussler, 1978; Wecksler et al., 1980). Significant binding of tritiated 1,25(OH)2D3was demonstrated autoradiographically in nuclei of chick and rat parathyroid cells (Henry and Norman, 1975; Wecksler et al., 1977). Early observations on the effects of 1,25(OH)$3 on PTH secretion were contradictory. Inhibitory as well as stimulatory effects or no changes have been reported. The effects were predominantly inhibitory, however. Observations on rachitic children suggested a direct inhibitory effect on PTH secretion (Fischer et al., 1973). Chertow et al. (1975) reported a reduction of serum FTH by 1,25(OH)2D3in rats and of PTH secretion in uitro in bovine parathyroid slices. Henry e? al. (1977) showed a reduction of the size of chick parathyroid glands after treatment with vitamin D3 or 1,25(OH)2D3.The steroid significantly inhibited PTH secretion of normal porcine parathyroid glands and human parathyroid adenomas in uitro within 4 hours. This was followed by ultrastructural signs of decreased hormone synthesis during continued incubation (Dietel et al., 1979). Conversely, Canterbury e? al. (1978) reported a rise in PTH in the thyroid venous blood of dogs given 1 ,25(OH)2D3into the thyroid artery. Pharmacological doses were required and no change in peripheral iPTH or plasma Ca2’ occurred. The physiological significance of the effect was therefore doubted by the authors. In rachitic or normal dogs no clear effect of 1,25(OH)~D3on serum iPTH or serum calcium could be demonstrated (Oldham et al., 19791, but the authors found evidence that the steroid increased the responsiveness of the parathyroid glands to the suppressive effects of normal or increased serum calcium levels. An attenuating effect on the secretory response of the parathyroid glands to hypocalcemia was suggested by Hurst and Mayer (1977). Using carboxyterminal N-terminal and C-terminal RIAs for PTH, Golden et al. (1980) were unable to find an effect of 1,25(OH)zD3on PTH release during 4 hours of incubation of bovine parathyroid slices or isolated cells, at variable Ca” concentrations. An effect could be demonstrated on the patterns of neither secreted

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

155

intact hormone nor hormone fragments. Llach et al. (1977)found no acute effect of 1,25(OH)2D3 on serum iPTH in humans. Silver et al. (1985) showed that 1,25(OH)2D3 inhibits PTH secretion in isolated bovine parathyroid cells. The steroid reversibly and selectively reduced preproPTH-mRNA, as well as the cellular content of the transcript. No effect on total RNA synthesis or total RNA content was found. The effect developed slowly. It became noticeable around 24 hours after the start of the incubation of the parathyroid cells with 1125(OH)2D3and was maximal after 48 hours. Similar results were reported for rats. PTH-mRNA levels were reduced about 50% in 6 hours and 75-90% in 24 hours after a dose of 100 pmol without change in serum Ca (Silver et al., 1986;Naveh-Many and Silver, 1988; Naveh-Many et al., 1989; Fig. 4). In other studies of this group the effects of 1,25(OH)2D3on PTH secretion during short-term (30- 120 minutes) and long-term (24-96 hours) incubation of bovine parathyroid cells were compared. Whereas in short-term incubations no effect was found, PTH secretion was inhibited dose-dependently during long-term incubation with the steroid, as was established with both Nterminal and C-terminal PTH assays. At 48 hours a strong correlation was

a

0 0

5

0

10

Serum

15 20 25 C a l c t u m (mg/dl)

30

FIG. 4. Effect of changes in serum calcium on mRNA levels for PTH, CT, and action of rat parathyroidglands at 6 hours. Rats were injected intrapentoneally with different amounts of calcium gluconate or monosodium phosphate at 0 and 3 hours. The results at each point are the mean for four rats and expressed as percentage of basal (Naveh-Many et al., 1989). 0 , PTH; A, actin; 0, calcitonin.

156

S. E. WENDELAAR BONGA AND P. K. T. PANG

found between the decrease in PTH release and the lowered concentration of preproPTH-mRNA in these cells (Cantley et al., 1985). These studies indicate that 1,25(0H)2D3 inhibits PTH secretion by suppressing gene transcription. The effect is specific. Vitamin D3 or the metabolites 25(OH)D3 and 24,25(OH)*D3had no effect at physiologically relevant concentrations in primary parathyroid cell cultures (Cantley et al., 1987). The relatively slow response may explain some of the contradictory results reported earlier, although species-dependent differences in the effects of 1,25(OH)2D3cannot be excluded. At least in the bovine species a negative feedback relationship at the level of hormone synthesis has now been established between PTH and 1,25(OH)2D3. The suppression of secondary hyperparathyroidism by intravenous administration of 1 ,25(OH)2D3in uremic patients indicated a similar inhibitory effect in humans (Slatopolsky et al., 1984). Okazaki et al. (1988) reported that the steroid decreased the transcription of the human PTH gene introduced in cultured rat pituitary cells, possibly via binding of the 1,25(OH)2& to the promotor region of the gene. The physiological significance of the effect of I ,25(OH)2D3on PTH secretion will be discussed in Section IV. Gonadal and interrenal hormones may also influence PTH secretion. Estrogens have a profound effect on bone. Bone loss increases markedly at menopause, a process that can lead to osteoporosis, and this can be reduced by estrogen treatment. Serum PTH levels are significantly elevated in osteoporotic women after long-term estrogen treatment, an effect related to the reduced loss of bone calcium in these patients (Gallagher et al., 1980). Although an indirect effect could not be excluded since the treatment induced mild hypocalcemia, Greenberg et al. (1987) have shown recently that 17P-estradiol and progesterone induced a specific and doserelated stimulation of PTH release from bovine parathyroid tissue in vitro. These results were confirmed by the same group for human parathyroid adenomas in vitro. The concentrations used were in the range occurring in physiological situations (Duarte et al., 1988). Similar effects may occur in vivo. In parathyroid cells of male rats estradiol induced the same morphometrical changes as decreasing serum calcium levels, even though no changes in serum calcium could be demonstrated: cell surface area and surface area of granular endoplasmic reticulum and Golgi complex were significantly increased within 6 hours after the start of estradiol treatment, indicating stimulated PTH synthesis (Wild et al., 1989). Cortisol increases serum PTH levels in rats (Williams et al., 1974) and man (Fucik et al., 1975). Adrenalectomy in rats decreased thyroidal CT-mRNA concentration. This could be partially corrected by injection of the synthetic glucocorticoid dexamethasone. Plasma CT levels were unchanged (Besnard et al., 1989).

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

157

5 . Second Messengers ai. Cyclic AMP. In the early 1970s the first evidence was reported for the implication of adenylate cyclase and its product cAMP in the regulation of PTH release. Studying bovine parathyroid gland slices in uitro, Abe and Sherwood (1972) demonstrated that dibutyryl cAMP and the phosphodiesterase inhibitor theophylline caused significant stimulation of PTH secretion. Similar observations were reported by Williams et al. (1973). Many following studies, in which (usually) bovine parathyroid slices or dispersed cells were employed, have indicated that intracellular cAMP levels change in parallel with receptor-mediated PTH release in uitro, irrespective of the agonists tested, e.g., epinephrine, norepinephrine, isopralterenol, dopamine, prostaglandin El and E2, glucagon, calcitonin, secretin, or NaF (see Cohn et al., 1986). The catecholamines act via a-adrenergic receptors. Concomitant stimulation of cAMP and PTH release with a-adrenergic agonists, dopamine, and secretin has also been demonstrated in uiuo (Blum et al., 1980; Fischer et al., 1982). Furthermore, the stimulatory action on PTH release of reduced extracellularCa2+ and Mg2+ levels is associated with increased cAMP concentrations (Brown et al., 1978b). When studied in homogenates of dog and horse parathyroid glands high calcium levels still inhibit, but high Mg2+levels stimulate, adenylate cyclase activity (Dufresne and Gitelman, 1972; Rodriguez et al., 1978). This effect is only seemingly in contrast with the stimulating effect of low extracellular Mg2+levels, since the observations on homogenates probably reflect the action of intracellular M$+. Sufficient intracellular Mg2+is a prerequisite for adenylate cyclase activity. More recently, Mahaffee et al.. (1982) concluded from similar studies on rat parathyroid homogenates that Ca2+has a direct inhibitory action on the guanine nucleotide adenylate cyclase complex, whereas intracellular M g + may enhance the activity of adenylate cyclase by guanine nucleotides or by competingwith Ca" for binding to a distinct intracellular regulatory site. While PTH stimulation is related to elevated cAMP levels, suppression of PTH release is typically associated with low CAMP, e.g., following exposure to high extracellular Ca2+,M g + , Mn*+,a-adrenergic catecholamines, or prostaglandin F2, (Brown et al., 1978a;Rodriguez et al., 1978). Parathyroid cells contain CAMP-dependent protein kinase activity, and this can be reduced by high extracellular Ca2+(Pines and Hurwitz, 1981). Thre effects of low extracellular Ca2+on adenylate cyclase are not limited to mere stimulation of the enzyme. Whereas a log-linear relationship between cAMP and PTH release has been established for many agonists for PTH release, including extracellular Ca2+(Fig. 5 ) , extracellular Ca"

158

S. E. WENDELAAR BONGA AND P. K. T. PANG

Y

0.8 y

3-

1 5

Ly --I

V

CAMP-

n

0

0.7

2 2-

>

h

f

a.

1-

I 0

40.5 I

1 .o

I

2 .o

I

3.0

I

CALCIUM mM

FIG. 5. PTH release and intracellular cAMP as a function of extracellular Ca2' concentration in porcine parathyroid cells in culture (Cohn and Eking, 1983).

appeared to be able to modify this relationship. At high external Ca2' concentrations PTH release induced by a-adrenergic agonists was associated with higher cAMP levels than at low Ca2+ concentrations. Thus, variations in external Ca2+apparently modulate the response of the parathyroid cells to a given level of intracellular CAMP. It was further shown that the inhibition of PTH release by high external Ca2+cannot be explained on a quantitative basis by the observed reduction of cAMP levels (Brown e? al., 1978a). At very high external Ca2' concentrations PTH release can be suppressed irrespective of the cAMP levels. Moreover, whereas several secretagogues stimulate CAMPdependent protein kinase activity in parallel with PTH release, the inhibitory effects of high extracellular Ca2+and Mg2+levels are not mediated by changes in the activity of such enzymes (Brown and Thatcher, 1982). Morissey and Cohn (1979a,b) had concluded earlier, on the basis of pulselabeling experiments, that PTH release by low extracellular Ca2+ was different than that effected by P-adrenergic agonists: whereas low Ca2+ stimulated the release of both newly synthesized and stored hormone, with P-adrenergic agonists only stored hormone was released. The release of newly synthesized hormone, in contrast to that of stored hormone, appeared to be independent of the cAMP level in the cells. These data stimulated the search for other intracellular mediators of the effect of extracellular Ca2+on PTH release. b. Intracellular Calcium. Whereas in earlier studies cAMP has been considered the primary second messenger involved in mediating the effects of secretagogues on PTH secretion, more recent data indicate an important role for intracellular Ca2+.In striking contrast to many other

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

159

gland cells, in which secretion is stimulated by a rise in intracellular Ca2+, PTH release is associated with a decrease in intracellular Ca2'. Evidence for the implication of intracellular Ca2' was presented by Habener et al. (1977), who showed that the calcium ionophores A23187 and X537A inhibited PTH release in the presence, but not in the absence, of extracellular Ca:!+ . The inhibitory effect of Ca2+ionophores has been frequently confirnied afterward, although not under all conditions (see below). In both bovine and human parathyroid cells the rise in external Ca2' that suppresses PTH release is accompanied by a rise in intracellular Ca" (Fig. 6). An increase in intracellular CaZ+is also associated with the effects of other divalent cations that inhibit PTH release (albeit at higher concentrations than Ca2'), such as Mg2+, Sr2+,and Mn2' (Shoback ef al., 1983, 1984a; Larsson et al., 1984; Nemeth and Scarpa, 1987). It was further shown that the parathyroid cells of patients with hyperparathyroidism not

OL

0.'5

IlO 1.15 115 l.:5 EXTRACELLULAR [Co] , mM 0.k

210

FIG. 6. Relationship between the extracellular Ca2+ concentration and PTH release or cytosolic Caz' in quin2-loadedbovine parathyroid cells in uirro. Note the inverse relationship between PTH release and cytosolic CaZ+(Shoback et al., 1983).

160

S. E. WENDELAAR BONGA AND P. K . T. PANG

only displayed a reduced response to a rise in extracellular Ca2+,but also a chronically reduced level of intracellular Ca2+(Larsson et al., 1984). The question presents itself whether the rise in intracellular Ca2+ is of extracellular or of intracellular origin. In many endocrine cells hormone release is initiated by the influx of Ca2+into the cell through voltage-gated Ca2+channels. A role for such channels in the inhibition of PTH release has been suggested on the basis of results obtained with Ca2+ channel blockers such as Verapamil and D600. The results were inconsistent, however. Inhibitory (Hove and Sand, 1981) as well as stimulatory effects (Wallace and Scarpa, 1983) have been reported. These blockers did not influence cytoplasmic Ca2+in the parathyroid cells (Wallace and Scarpa, 1983)and this finding, together with the observation of Wallfelt er al. (1985) that membrane depolarization by potassium ions did not influence Ca2+ fluxes across the cellular membrane, pointed to the absence of voltagegated Ca" channels in parathyroid cells. Nevertheless, the inhibitory effect of extracellular Ca2+on FTH release appeared to be associated with increased influx of extracellular Cat+ as was shown by Wallfelt et al. ( 1985). They suggested the presence of a specific type of Ca2+channel in the outer cell membrane, activated by extracellular Ca2+.A similar suggestion was made by Larsson et al. (1984). Fitzpatrick et al. (1986a) concluded the presence of dihydropyridine-sensitive voltage-gated Ca2+ channels because an antagonist of this type of channel stimulated, and an agonist inhibited, FTH secretion in bovine parathyroid cells in uitro. The presence of such channels could not be confirmed by Muff et al. (1988), even though they used the same type of channel agonist and antagonist. Fitzpatrick et al. (1988) subsequently supported their claim by demonstrating that FTH release could be inhibited by exposing these cells to antibodies specific for the a-subunit of the skeletal muscle dihydropyridine-sensitive Ca2+ channel. The presence of dihydropyridinesensitive Ca2+ channels was subsequently confirmed by demonstrating specific high-affinity binding to FTH membranes of a specific radioligand (Jones and Fitzpatrick, 1990). The same group showed earlier that the Ca2+inhibition of PTH release is mediated by a pertussis toxin sensitive G protein (Fitzpatrick et al., 1986b,c).Their results indicate the presence in bovine parathyroid cells of voltage-gated Ca2+channels linked to signaltransducing G proteins. They suggested that extracellular Ca2+can inhibit PTH secretion by binding to the antagonist receptor of the channel itself (Jones and Fitzpatrick, 1990). It was further demonstrated that high external Ca2+induces a biphasic increase of intracellular Ca2+ in bovine parathyroid cells: a sustained increase from extracellular origin was preceded by a transient rise of intracellular Ca2+ from internal, nonmitochondrial stores (Nemeth and

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

161

Scatrpa, 1987; Hawkins et al., 1989). In the same studies it was shown that inositol phosphates were involved in the transduction of the extracellular signals to these stores. Thus, as in other secretory cells, Ca ions of extracellular and intracellular origin contribute to the rise of cytoplasmic Ca2+ asssociated with hormone release in the PTH cells, albeit that changes in intracellular Ca2' had effects different from those in other cells. In most secretory cells release increases sigmoidally with a rise in intracellular Ca2', with half-maximum stimulation at about lo-' M intracellular Ca". In bovine PTH cells maximum release was observed at loe7M Ca2+, with secretion at both higher and lower Ca2' concentrations (Pocotte and Ehrenstein, 1989). However, maximum release at much higher intracellular levels (4 X M )was reported by Oetting et al. (1987) for calf PTH cells. The difference was related to the difference in the setpoint for extracellular Ca2+ between adult and juvenile bovine PTH cells, but remains difficult to interpret. Whereas changes in PTH release are frequently associated with changes in cytosolic Ca", both responses are not necessarily coupled. In dispersed bovine parathyroid cells stimulation of PTH release by secretagogues such as dopamine (Shoback et al., 1984b)or lithium ions (Nemeth et al., 1986),that increase CAMPlevels, do not notably influence cytosolic Ca2'. In similar cells incubated at extracellular Ca2' levels (0.5 mM) that stimulate PTH release, the addition of ionomycin increased cytosolic Ca2' witlhout inhibiting PTH release (Nemeth et al., 1986). Similarly, exposure of such cells to 1,25(OH)2D3,which does not immediately affect PTH release (see Section II,A,4,a,iii), induced a prompt and dose-dependent rise in intracellular Ca" . These observations indicate that receptormediated effects that inhibit PTH release are not, or not exclusively, dependent on a rise in cytosolic Ca2+.It has been suggested that a localized change in Ca2' concentration may be critical for PTH release rather than a change in total cytoplasmic Ca2' (Dean et al., 1986). c. Inositol Phosphates and Diacylglycerol. In many secretory cells the increase of cytoplasmic Ca2+ from internal stores is mediated by inositol 1,4,5-triphosphate (IP3), produced in the cellular membrane by phclsphoinositide hydrolysis, whereas inositol 1,3,4,54etrakisphosphate (IP',), produced by phosphorylation of IP3, may act as a stimulator of the influx of extracellular Ca2' (Bemdge and Irvine, 1984,1989).Epstein et al. (1985) and subsequently Brown et al. (1987) and Hawkins et al. (1989) demonstrated that the primary secretion-controllingfactor of the parathyroid cells, the external Ca" concentration, also influenced the formation of several inositol phosphates in these cells, in particular IP, and IP4. In permeabilized bovine parathyroid cells (Epstein et al., 1983, IP3 pro-

162

S. E. WENDELAAR BONGA AND P. K. T. PANG

moted the release of intracellular Ca2+stores. A relationship between IP3 and the initial intracellular Ca2+transient, and of IP4 with the subsequent more sustained elevation of intracellular Ca2+,was indicated. The accumulation of inositol phosphates likely was dependent on guanine nucleotide regulatory protein (G protein) since it could be promoted by fluoride (Chen et al., 1988; Hawkins et al., 1989), which is known to activate the G proteins that stimulate inositol phosphate formation. The Gi blocker pertussis toxin prevented the inhibitory effects of Ca2+ on PTH release (Fitzpatrick et al., 1986c) and CAMPaccumulation (Chen et al., 1988). However, pertussis toxin did not suppress the effect of high extracellular Ca2+on IP3 and IP4 accumulation, indicating that the formation of these phosphates is mediated by fluoride-stimulated but pertussis toxininsensitive G protein. The above data showed that the inhibition of PTH release by extracellular Ca2+could be mediated by IP3. This is unlikely for IP4. Substantial formation of IP4 was only observed at external Ca2+ concentrations ( 5 mM) that were well above those producing maximal inhibition of PTH release (Hawkins et al., 1989). Recently it was suggested that another phospholipid metabolite, phosphatidic acid, may be involved in signal transduction in PTH cells (McGhee and Shoback, 1990), but its role has still to be defined. More data are available on the role of diacylglycerol (DG), a recognized second messenger in many cell types (Berridge, 1987). As in other cells, in the parathyroid cells the formation of IP3is stimulated concomitantly with the generation of DG. Diacylglycerol generally stimulates secretion by activating protein kinase C, a group of enzymes connected with secretion and other cell functions, and a stimulator of the plasma membrane calcium pump (Furukawa et al., 1989). Because high extracellular Ca2+stimulates IP3 as well as DG, one would expect that activation of protein kinase C is associated with inhibition of PTH release. However, in bovine parathyroid cells the relation between DG and hormone release probably is complicated, as appeared in studies on the effects of the activation of protein kinase C by phorbol esters. In most studies FTH release was stimulated (Brown et al., 1984b; Nemeth et al., 1986; Kobayashi et ul., 1988), even at high extracellular Ca2+ levels. Intracellular Ca2+ levels were unaltered (Brown et al., 1984b)or decreased (Nemeth et al., 1986). Kobayashi et al. (1988) showed that acute exposure of the cells to low extracellular Ca2+ (0.5 mM) increased, and high Ca2+ (1.75 and 2.5 mM) decreased cell membrane protein kinase activity. The authors concluded that the effects of extracellular Ca2+are at least partly mediated by regulation of protein kinase C activity. Membreiio et al. (1989) came to a different conclusion. They examined the effects of several phorbol esters at high as well as low

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

163

extracellular Ca2+ concentrations. They could not confirm that these protein kinase C agonists stimulated PTH release in the presence of high extracellular Ca2+.They further showed that the rise in intracellular Ca2+ that normally occurred in the presence of high extracellular Cat' was suppressed in the presence of phorbol esters, which indicated that the stimulation of PTH by this treatment was mediated by reduced intracellular Ca2+ levels. Surprisingly, the PTH release normally induced by low extracellular Ca2+ was inhibited by these substances. Membreiio and co-workers concluded that protein kinase C agonists suppress PTH release at low extracellular Ca2+ and enhance PTH release at high extracellular ca::+ . The function of protein kinase C in the regulation of PTH release needs further clarification, although at present the route leading to activation of this enzyme seems the most important one for stimulus-secretion coupling in the PTH cells. 6 . Control of Intracellular PTH Degradation Pis discussed above (Section II,A,3), a substantial part of the newly synthesized and stored PTH is not secreted as intact molecules but as fragments produced by intracellular proteolysis. Chu et al. (1973), studying rats fed normal or low calcium diets, concluded that extracellular Ca2+ may control the amounts of intracellular PTH through regulation of the rate of intracellular degradation rather than the rate of synthesis. Habener et 01. (1975) showed that release of intact PTH from bovine parathyroid tissue slices was stimulated 5-fold when extracellular Ca2+was reduced from 2 to 1 mM, with little change in the rate of PTH biosynthesis. At 2 mM Ca2+intracellular stores of PTH were higher by only 30-40%, much less than predicted on the basis of the observed reduction of PTH release and estimated new formation. They concluded that at high extracellular Ca2+ levels up to 50% of the newly synthesized PTH was degraded intracellularly, whereas at low extracellular Ca2+ less than 10% was degraded. These results were confirmed for porcine parathyroid tissue in uitro by Moirissey and Cohn (1979b), who further demonstrated preferential degradation of stored PTH. Interestingly, they demonstrated that only the degradation of newly synthesized PTH was controlled by external Ca2+. Evidence for extracellular Ca" as a factor controlling the catabolic pathway in the parathyroid cells was also obtained in uiuo. Starting from the notion that the degradation products are predominantly C-terminal fragments, Mayer et al. (1979) determined the ratio of C-terminal and Nterminal fragment immunoreactivity in the venous effluent plasma of neonatal calves. The C : N ratio increased from about 1.3 in hypocalcemic animals to more than 3 during hypercalcemia. They concluded that a relatively high percentage of intact hormone molecules was released dur-

S.E. WENDELAAR BONGA AND P. K. T. PANG

164

ing hypocalcemia, whereas mostly inactive fragments were secreted during hypercalcemia. These results have been confirmed for human parathyroid tissue in v i m by Hanley and Ayer (1986). They showed that at high Ca2+ levels proportionally more carboxy-terminal PTH fragments were released than intact FTH, whereas the reverse was found at low Ca2' (Fig. 7).

. I

40

45

50

55 60 65 FRACTION (2 ml)

70

75

80

FIG. 7. Chromatographic profile of perifusates of hyperplastic human parathyroid tissue perfused with low calcium (A, 0.5 mM Ca2+)or high-calcium (B, 2.0 mM Ca2+)fluid. (A) Note the prominent first peak of PTH coeluting with bovine PTH and smaller peaks coeluting with human PTH (39-84) and human PTH (1-34). (B) The first peak is smaller than the second peak, indicating that proportionally more fragments than intact PTH were released from the tissue when exposed to the high Ca2' concentration (Hanley and Ayer, 1986).

165

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

Concomitantly with suppression of PTH release the number of lysosomelike bodies in the cytoplasm increases (Capen, 1971; Roth et af., 1974; Setoguti et al., 1981). These bodies may be involved in hormone degradation, as has been described for other endocrine gland cells after suplpression of hormone release (Farquhar, 1976). However, many of the hormone degradation products are released from the parathyroid cells, although in this respect species differences seem to occur. Since the contents of lysosomes in general are not released from the cell-with the lysosomes of the liver as notable exception-a nonlysosomal pathway cannot be excluded. 5.5

r

1251 ti PTH 39-84

4.5 0

2 p %

4.0

Y

I-

z

2

<

5

$

3.5

3.0

I

c a. m 2.5 2.0

1.5

1.o

0.5

c

40

I

1

45

50

I

I

1

55 60 85 FRACTION (2ml)

FIG.I

(continued)

I

I

1

70

75

80

166

S. E. WENDELAAR BONGA AND P. K . T. PANG

Whatever the mechanism involved, the above data strongly suggest that there is a constant overproduction of PTH by the parathyroid glands, and that the physiological demand, reflected by the calcium concentration of the extracellular fluid, determines the ratio of intact to degraded hormone that is released into circulation. Regulation of hormone degradation rather than hormone production as a mechanism of control of hormone secretion seems unique for the parathyroid glands. Although hormone degradation is commonly observed in other endocrine glands, this process concerns a substantial part of hormone production only when the physiological demand is suddenly reduced, e.g., in prolactin cells of lactating rats after removal of the pups (Farquhar, 1976). In these cells it is also a transient phenomenon, accompanied by rapid reduction of hormone synthesis. Thus, in these cells it is not a mechanism for controlling hormone output, as in the parathyroid glands, but a mechanism for the incidental removal of superfluous hormone stores.

B.

CALCITONIN

1. Introduction

The existence of a calcium-regulatinghormone with a function originally supposed to be antagonistic to that of PTH was first demonstrated by Copp et al. (1962). It was isolated soon after, and sequenced a few years later. Calcitonin (CT) reduces the extracellular calcium concentration by inhibiting osteoclastic bone resorption and by promoting calcium deposition in the skeleton. In mammals it is produced in the parafollicular or C cells of the thyroid glands. These cells are derived from the neural crest. During embryogenesis they migrate downward, and merge with tissue from the most caudal branchial pouch and become incorporated in the primordial thyroid. In the nonmammalian vertebrates the C cells do not merge with the thyroid tissue, but form distinct organs, the ultimobranchial glands. In birds C cells may occur in the thyroid gland as well as in the ultimobranchials (Robertson, 1986). Since the homology of the parafollicular or C cells with the ultimobranchial cells is beyond dispute, we will designate them as C cells, irrespective of their thyroidal or ultimobranchial origin. C cells have been demonstrated in all vertebrate groups with the exception of the jawless fishes. They belong to the family of APUD cells (Pearse, 1968): C cells are capable of uptake and decarboxylation of the precursors of bioamines, in particular 5-hydroxytryptamine and dopamine. The cells have the ultrastructural characteristics of protein-producing endocrine cells: granular endoplasmic reticulum, Golgi regions, and dense-cored vesicles that release their contents by exocytosis. The ultrastructure of the

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

167

C cells is remarkably similar in fish to mammal (Nuiiez and Gershon, 1978; Ericson and Sundler, 1984; Robertson, 1986). Thyroid glands and ultimobrarichial bodies have in general a rich innervation of autonomic origin. Adrenergic, cholinergic, and peptidergic fibers have been demonstrated (Munson, 1976; Nuiiez and Gershon, 1978). As reviewed by Robertson (1986), there is no structural evidence for synaptic innervation of the C cells in mammals, with the exception of dolphins. Synaptic contacts with C cells has also been established for some birds and amphibians, whereas in rcptiles and fish no direct neural contacts have been reported, although nerve fibers are common in the connective tissue around the cells. In addition to CT and varying quantities of bioamines, the C cells frequently produce other peptides, in particular somatostatin and calcitonin gene related peptide (CGRP). The gene coding for CT is also coding for CGIIP. In man and rat in fact two differentially regulated genes have been demonstrated: CT/aCGRP and CTIPCGRP. Both are expressed in various parts of the body. Whether CT or CGRP is the main product depends on the mode of splicing of the transcript of the gene. In normal C cells predominantly CT mRNA is produced, which yields a precursor molecule of about 135 amino acids. Posttranscriptionalprocessing includes cleavage of the N terminal and C terminal of this molecule and results in the formation of CT. From mammal to fish CT consists of 32 amino acids, although the amino acid composition may vary considerably (Zaidi et al., 1987).

Calcitonin gene related peptide is the main product expressed in many regions of the central and peripheral nervous system and cardiovascular system. It has effects on cardiovascular, respiratory, and sensory functions. CGRP is also produced in large amounts in transformed C cells and, generally in small quantities, in normal C cells (Zaidi et al., 1987; Breimer et al., 1988). It is colocalized with CT in the secretory granules of the C cells. In these granules further varying amounts of somatostatin are present. This location implies the cosecretion of CGRP and somatostatin with CT (Zabel and Schiifer, 1988). For the CGRP and somatostatin of the C cdls a paracrine function has been suggested (Williams et al., 1979). Calcitonin is also produced in the central nervous system, but in these regions it may act as a neurotransmitter or neuromodulator, and is unlikely to contribute significantly to the total amounts of CT present in the general circulation. We will confine this review to the control of the secretion of CT of thyroidal and ultimobranchial origin that is released into the systemic circulation. CT of ultimobranchialorigin was isolated and sequenced in birds and fish, and CT bioactivity was demonstrated for the ultimobranchial glands of all vertebrate groups in which these glands are present (Copp and Mine, 1989). Originally the main function of the hormone was

168

S. E. WENDELAAR BONGA AND P. I(.T. PANG

defined as the control of the extracellular calcium level, in an antagonistic action with FTH. In young mammals, birds, and reptiles, the hypocalcemic effects of CT were generally clear and undisputed (Copp and Kline, 1989). The hypocalcemic response of young rats has frequently been used as a reliable and sensitive bioassay for CT, and salmon and human CT preparations are effective therapeutic agents in treating hypercalcemia in man (Wisneski, 1990). However, the function of CT as a hypocalcemic hormone was reevaluated in the early 1970s, because chronically elevated or reduced circulating levels in mammals or birds were only occasionally associated with substantial disregulation of plasma calcium levels. The cycle in annual plasma calcium concentration in frogs was unchanged by removal of the ultimobranchial glands (Fig. 8) in contrast to removal of the parathyroid glands, which had a depressive effect (Robertson, 1977). Administration of CT, although effective in young animals, frequently did not reduce plasma calcium levels in adult mammals, birds, and reptiles (Copp and Kline, 1989). Observations on plasma calcium levels in amphibians were contradictory (Srivastav and Rani, 1989). In these animals CT stimulated the accumulation of calcium carbonate in the lime sacs (Oguro er al.,

Month

FIG. 8. Annual plasma calcium concentration in sham control and ultimobranchialectomized (UBX)adult male frogs (Rana pipiens). The periodic change in calcium levels of intact frogs (solid line) (N = 247) throughout the year is expressed: Cap = 7.186 + 1.079 cos (30r - 315); r = 0.635; p = .001 and not significantly different from controls. Both groups display a significant ( p = .001) minimum in April-June, while the maximum plasma calcium levels are attained in October-November. Each point is mean 2 SD; N = 15-30 frogs (Robertson, 1977). 0, Control; 0 ,UBX.

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

169

1984.; Srivastav and Rani, 1989) and prevented the PTH-induced mobilization of these calcium stores (Robertson, 1972). Fish will be discussed separately (see Section 111,B). Thyroidectomy in mammals or ultimobranchialectomy in birds or amphibians (Robertson, 1977; Kapoor and Chhabra, 1981) were normally not followed by a rise in plasma calcium. Although such observations could be explained by the assumption that potential hypocalcemic actions of CT are obscured by efficient hypercalcemic control mechanisms, they have raised doubts about the importance of CT as a hypocalcemic hormone. The role of CT in reducing plasma calcium may be limited to redressing the hypercalcemia resulting from intestinal calcium resorption during and after a meal. This explains the well-known postprandial surge in CT secretion (Gray and Munson, 1969). It is also consistent with observations that calcium infusions produced hypercalcemia in ultimobranchialectomized birds or amphibians, but had no or small effect on intact controls (Calamy and Barlet, 1970; Copp et al., 1970; Robertson, 1971a; Sasayama and Oguro, 1976; Sasayama, 1978). CT is further essential for the conservation of ingested calcium by promoting calcium deposition in the bone. It further protects the skeleton against demineralization, by inhibition of osteoclastic activity and osteocytic osteolysis, during periods of potential calcium shortage, such as growth, pregnancy, and lactation and, in the submammalian vertebrates, ovarian growth and yolk formation (Lewis ef al., 1971; Stevenson et al., 1979). CT also stimulates phosphate deposition into bone and prevents its loss from bone and bone fluid (Talmage and Van der Wiel, 1979). With respect to plasma calcium, CT may have an antihypercalcemic function rather than being implicated in the minute-to-minute control of plasma calcium homemtasis. The function of CT will be discussed further in Section IV, after we have reviewed the factors regulating CT secretion.

2 . Control of CT Secretion In mammals the control of CT secretion has received less attention than that of PTH. This is understandable for two reasons. First, the modes of control turned out to follow the general pattern known from other cell types, whereas the control of PTH secretion is unique for the hormone. Second, the number of CT-producing cells is relatively small (in humans about 0.1% of thyroid mass), and, because of the location of these cells between the thyroid follicles, rather inaccessible for experimental manipulation or in uitro studies. To overcome the latter difficulties,several studies concern medullary thyroid carcinoma tissue, which produces large amounts of CT. However, data obtained with transformed cells require verification in normal cells and this has been done only occasionally.

170

S. E. WENDELAAR BONGA AND P. K . T. PANG

a. Extracellular Calcium. Extracellular calcium is considered an important regulator of CT secretion in mammals as well as in the submammalian vertebrates. In general, plasma CT levels parallel daily fluctuation in plasma calcium concentrations, as has been demonstrated for mammals (Milhaud et al., 1972), buds (March and McKeown, 1977), and amphibians (Robertson, 1978). The C cells become activated by experimental hypercalcemia induced by, e.g., injections of calcium salts or vitamin D preparations. Following such treatments, in mammals hypertrophy of the C cells, enlargement of granular endoplasmic reticulum and Golgi regions, degranulation of the cytoplasm, and release of secretory granules by exocytosis were reported (Zabel and Schiifer, 1988; Nuiiez and Gershon, 1978; Ericson and Sundler, 1984; Srivastav and Rani, 1988). Such structural changes were accompanied by decreased intracellular CT contents and elevated CT levels (Cooper et al., 1971a; Rix et al., 1984; Zabel and Schhfer, 1988). Drinking of calcium-containing water resulting in rises of plasma calcium within the normal range induced dose-related increases in circulating CT levels in young adult men. This was consistent with the hypothesis mentioned above that CT prevents excessive postprandial hypercalcemia (Austin et al., 1979; Fig. 9). The rise in CT following calcium administration can be a rapid response: a surge of CT occurred in healthy persons within 2 minutes. Plasma calcium returned to normal within 30 minutes after the start of calcium infusion (Rude and Singer, 1977; Parthemore and Deftos, 1978). Ingestion of CaC12 by newborn rats, which induced a rapid rise of plasma calcium levels, resulted in a biphasic response: within 5 minutes a peak in plasma CT occurred that was followed by a prolonged rise starting after 1 hour (Garel and Besnard, 1980). The effects of extracellular calcium on CT secretion are, at least partially, direct. This was first indicated by the now classical experiment that led to the discovery of CT (Copp et af., 1962). Perfusion of the thyroid glands of cannulated dogs with calcium-enriched blood produced hypocalcemia in the systemic circulation. In a similar perfusion experiment in pigs, increased CT release was demonstrated by direct measurement in the thyroid effluent blood (Care et af., 1W).This was confirmed for perfused ultimobranchial glands in chicken (Ziegler et al., 1969) and goose (Bates et at., 1969). Rat thyroid slices incubated in uitro responded with increased CT release to increased ionic calcium levels in the extracellular fluid (Feinblatt and Raisz, 1971); this was confirmed for different mammalian species including man (Bell, 1975; Cooper et al., 1977b), and for ultimobranchial glands of birds (Feinblatt et al., 1975; Nieto et al., 1975), and fish (see Section III,B,2). Relatively few studies are available on the effects of extracellular Ca2+ on CT synthesis. As reported earlier, experimental manipulation of plasma

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

171

calcium for 6 hours in rats revealed that low calcium markedly stimulated PTH-mRNA in the parathyroid cells. However, neither increased nor decreased plasma calcium had any effect on CT-mRNA in the thyroid tissue of these animals (Fig. 4; Naveh-Many et al., 1989). This confirmed the result of an earlier in uitro experiment showing that high calcium increased CT release but had no effect on CT-mRNA in rat thyroid slices over 3- and 6-hour periods (Jacobs et al., 1983). However, Segond et al. (1984) reported a rapid but transient stimulation of CT-mRNA levels 2-8 minutes after acutely induced hypercalcemia. Administration of depolarizing potassium concentrations in the rat increased plasma CT levels as well as CT-mRNA extracted from the thyroid gland within 30 minutes. (Jousset et al., 1988). The response of C cells to hypocalcemia is puzzling as far as mammals are concerned. Whereas reduction of secretory activity might have been expected for a hypocalcemic hormone, frequently the reverse has been indicated. In chronically parathyroidectomized rats, which experience hypocalcemia for at least 1 or 2 months, serum CT as measured by bioassay increased, and CT accumulated in the thyroid gland (Gittes et al., 1968). Using a homologous RIA, Peng and Gamer (1979) confirmed this observation. After several months they observed C-ceI1 hyperplasia, increased thyroid CT-content and plasma CT levels, and a rise of plasma calcium to normal concentrations. Increased mitotic activity, hypertrophy, and hyperplasia during hypocalcemia were reported several times (Nufiez and Gershon, 1978; Srivastav and Rani, 1988). In addition vacuolar degeneration of C cells was observed (Peng and Gamer, 1979), suggestive of physiologically controlled cell death by apoptosis (Wyllie et al., 1980). These responses are difficult to interpret when starting from the hypothesis that CT is involved in the homeostatic control of plasma calcium. They can be explained by the assumption that CT protects the skeleton against demineralization that may occur during prolonged hypocalcemia. For submammalian vertebrates, data on the response of C cells to hypercalcemia or hypocalcemia are scarce, but in general they point to activation in the former, and inactivation in the latter, condition. Light microscopic indications for C cell activation were reported for birds, lizards, and frogs. Young chickens on a calcium-deficient diet showed cellular atrophy, and granule accumulation was reported during experimental hypocalcemia in iguanas (Copp and Kline, 1989). For fish ultimobranchial cells, similar responses were reported (see Section 111,B). b. Hormones and Neurotransmitters. i. Proteins and peptides. Factors originating from the gastrointestinal system frequently have been

172

S. E. WENDELAAR BONGA AND P. K . T. PANG

reported to influence CT secretion. This is of interest with respect to the antihypercalcemic function ascribed to CT during and following dietary calcium uptake. Attention was drawn to the gastrointestinal hormones by the observation that in pig thyroid venous blood CT levels rose after oral calcium administration even without detectable changes in plasma calcium (Cooper and Deftos, 1970). Subsequently, Cooper et al. (1971b) and Care et al. (1971a) demonstrated the potent stimulatory effects of gastrin and related peptides on CT secretion, such as the structural analogue pentagastrin and cholecystokinin. The latter hormone has some structural Cterminal homology with gastrin. Infusion of low doses of gastrin into the thyroid artery of pigs stimulated CT secretion, whereas intravenous infusion had no effect, indicating that the hypocalcemic action reported for the peptide was mediated by CT (Cooper et al., 1972). Oral administration of meat or calcium at doses that did not noticeably increase plasma calcium levels increased both plasma gastrin and glucose in pigs (Care et al., 1971b; Cooper et al., 1974). The physiological importance of the gastrin-CT connection therefore seems well established in these animals. Whether it is of significance for other species is unclear. For adult men, McGuire er al. ( 1972) reported that gastrin infusion of physiologically relevant doses induced hypocalcemia. However, a rise in plasma gastrin levels induced by low or high calcium meals was not associated with increased CT levels, whereas plasma CT could not be increased with intravenous pentagastrin infusion, unless at very high doses. At least in adult men the physiological significance of the gastrin-CT relationship is doubtful. In rats it is unlikely that any effects of feeding on CT release are mediated by gastrin. In this species CT release was stimulated by feeding, even when plasma calcium was slightly below normal (Talmage et al., 1975; Toverud and Munson, 1976). In particular CT release was increased in young rats when suckling and in their mothers when eating (Cooper et al., 1977a). Nevertheless, gastrin injections were ineffective in stimulating plasma CT levels in testing and suckling newborn rats or their mothers (Garel and Jullienne, 1977; Cooper et al., 1977b), and attempts to promote CT secretion by thyroid tissue of suckling rats were unsuccessful (Cooper et al., 1977b). Secretin is another potential secretagogue. In normal human subjects as well as in 8-day-old rats secretin stimulated serum CT levels. The effective doses were about three times those observed in the postprandial state. A dose-related increase in CT release was found during incubation with secretin of rat thyroid tissue (Sethi et al., 1981). Interestingly, in rats, duodenal infusion of acid (known to stimulate secretin but not gastrin secretion) caused a slight but significant rise in serum CT concentration (Sethi et af., 1983).

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

173

Another intestinal hormone that could be a mediator of intestinal influence on the C cells is glucagon. Pancreatic glucagon was reported to induce hypocalcemia in several mammalian species (Munson, 1976). For dogs it was concluded that the glucagon-induced hypocalcemia was mediated by CT, because glucagon was ineffective in thyroidectomized animals (Avioli et al., 1969). Glucagon had no effect on plasma CT in 2-hour-old fasted newborn rats, although a slight elevation was noted in 14-day-old fasted rats (Garel and Jullienne, 1977). However, in general no clear effects of glucagon on plasma CT-levels have been demonstrated and its physiological importance as a secretagogue for CT has not been established. Another pancreatic peptide, somatostatin, caused a dose-related decrease of serum CT in rat and monkey (Hargis et al., 1978) and rats receiving somatostatin antiserum had elevated CT levels (Williams et al., 1979). The authors suggested that somatostatin, which can be cosecreted with CT (see Section II,B,l) is a local regulator of CT release. The autonomic nerves innervating the thyroid of several mammalian species contain neuropeptides such as vasoactive intestinal peptide (VIP), gastrin/cholecystokinin(CCK), and substance P. Only substance P could be demonstrated in nerves innervating the chicken ultimobranchialglands (Ahrdn et al., 1980,1983). With the exception of VIP, these neuropeptides did not stimulate thyronine secretion after intravenous injection in rats. However, VIP, CCK-4, CCK-8, and substance P all stimulated CT secretion in these animals (AhrCn et al., 1983), indicating that nerves ending on or around the C cells could influence CT secretion. ii. Bioamines. The literature on this subject before 1980 has been reviewed by Heath (1980) and will be summarized briefly. Evidence for catecholaminergic effects on CT secretion was first obtained by perfusion of thyroid glands of young pigs in situ with high-calcium blood (Bates el al., 1970). Epinephrine and the p-agonist isoproterenol stimulated CT release, although only in the presence of the a-adrenergic blocker phentolamine. The effect was inhibited by the P-blocker propranolol, confirming the p-adrenergic nature of the stimulation (Bates et al., 1970). The stimulatory effect of epinephrine was confirmed for sheep. However, in these experiments both isoproterenol and phentolamine prevented the stimulation of CT secretion (Philippo et al., 1970). In similar experiments on dogs, epinephrine was ineffective when given alone, but was stimulatory in combination with phentolamine (Avioli et al., 1971). Care et al. (1970) demonstrated for pigs that the effect of epinephrine was much more pronounced than that of norepinephrine, an observation that has been confirmed in uitro for porcine thyroid gland slices (Bell, 1975). These and other studies established a p-adrenergic stimulation and a-adrenergic inhibition

174

S . E. WENDELAAR BONGA AND P. K . T. PANG

of CT release in several mammalian species, including man (Munson, 1976; Heath, 1980). The physiological significance of these effects is still unclear, however. Pathologically elevated plasma catecholamine concentrations in man were not generally associated with increased CT levels (Miller et al., 1975),and adrenalectomy in rats did not impair CT secretion or the response of the C cells to hypercalcemia (Heath et al., 1980). Evidence for a role of sympathetic nerves is poor (see below). Since dopamine is produced by the C cells and can be cosecreted with CT, effects on C cells might be anticipated. Dopamine indeed stimulated CT release in uitro from porcine thyroid tissue (Bell, 1975). A similar inhibitory effect on CT secretion was reported for the precursor L-dopa when administered to patients with medullary thyroid carcinoma. The substance reduced the stimulatory response to infusion of calcium or pentagastrin. It was also effective on cultured carcinoma tissue (Baylin et al., 1979). iii. Steroids. For several mammalian species a clear sex difference has been reported in plasma CT levels. In rats, lower basal values were found for males than for females (Peng et al., 1978; Roos et al., 1978) and this difference increased after feeding (Peng and Gamer, 1979; Fig. 9). Women had lower basal as well as calcium-stimulated or gastrin-stimulated CT levels than men (Heath and Sizemore, 1977; Hillyard et al., 1978). Nevertheless, in humans as well, exogenous estrogens (but not androgens) stimulated CT secretion (Stevenson et al., 1981, 1983). This effect probably was attributed to the reduction in plasma calcium and reduced bone resorption following estrogen treatment (Riggs et al., 1969; Aitken ef al., 1971). More recently receptor-mediated direct effects of estrogens were reported on bone cells, showing that there were also interactions with CT and PTH at the target level (Eriksen er al., 1988; Colston et al., 1989; Takano-Yamamoto and Rodan, 1989). While Morimoto et al. (1980) did not observe a change in plasma CT levels in elderly women treated with estrogens for several weeks, possibly because of the time of the day of blood sampling, the increment in plasma CT in response to calcium infusion was significantly higher after estrogen treatment than before, indicating increased sensitivity to CT. There was a close relationship between plasma estrogens and CT in adult women, and the postmenopausal reduction in estrogen levels was associated with lowered CT secretion (Stevenson et al., 1983). Since plasma PTH and 1,25(OH)~D3were in general unchanged (Stevenson et al., 1981) the low CT levels were the most likely cause of the increased loss of bone minerals in postmenopausal women, which may result in osteoporosis. Very low CT levels were reported for osteoporotic patients (Milhaud ef al., 1978). Bone resorption could be prevented or even slightly reversed by administration of estrogens and/or CT, in particular in combination with calcium supplements

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

p< 001

175

1

1

0

--5 HRS BEFORE FEEDING

I HR BEFORE FEEDING

2 HRS FEEDING

( NOON1 (4pm) Opm) FIG. 9. Sex difference in the rise in CT after feeding in rats on a fixed feeding schedule. Food was given at 5 PM. There were 12 rats in each group. The height of each bar represents the mean values, and the brackets show SE. The p value refers to the comparison of means between 6-month-old female (unshaded area) and male (shaded area) rats that were fed for 2 hours (Peng and Gamer, 1979).

(Jowsey et al., 1978), although the effect might be transient (Aitken et al., 1971). Variations in plasma CT were reported during the menstrual cycle (Pitkin et al., 1978), but this observation was not confirmed (Baran et al., 1980). Other ovarian factors might be implicated in addition to estrogens. The role of estrogens in control of CT could be species specific. In adult rats, with higher CT levels in males than in females, ovariectomy induced a transient fall in CT secretion, which could not be corrected by exogenous estrogens (Cressent et al., 1981). After studying CT levels during the estrus cycle in rats (Cressent et al., 1983) a minimum was reported during estrus followed by a peak during diestrus. The relationship with estrogens was not clear, and no correlation was found with plasma calcium levels. Recently 1,25(OH)*D3was reported to influence CT-gene transcription rate in rat C cells, at physiologically relevant concentrations(Naveh-Many

176

S. E. WENDELAAR BONGA AND P. K . T. PANG

and Silver, 1988). Surprisingly, the effect was inhibitory, similar to that of the steroid on PTH gene transcription (see Section II,A,4,c,iii). A reduction to 6% of the basal level was noticed at 6 hours after injection of 100 pmol of 1,25(OH)2D3per animal, a dose that did not change serum calcium. The effect may have physiological significance, because cytoplasmic receptors to the steroid have been demonstrated in medullary thyroid carcinomas (Freake and Maclntyre, 1982)and thus may be present in normal C cells. In an earlier study Segond et a/. (1985) reported a transient increase in CT-mRNA levels, 1-2 hours after 1,25(OH)zD3administration to rats. However, these authors used a much higher dose (500 pmol per animal), and no effect was found at lower doses (Legendre et a / . , 1989). Data on the effects of the steroid on CT release are scarce. The stimulatory action of high doses of 1,25(OH)zD3or vitamin D2 on the secretory activity of the C cells (Munson, 1976; Nuiiez and Gershon, 1978) or on plasma CT (Raue et al., 1983) have been considered as indirect, and mediated by hypercalcemia. However, administration of physiologically more relevant doses caused a rise in CT levels in rats, despite a transient hypocalcemia, indicating a direct effect on CT release (Legendre et al., 1989). 3 . Second Messengers Cyclic AMP, cy tosolic Ca" , and phosphatidylinositides have all been implicated in the external signal transduction leading to increased CT synthesis and release. The stimulating effect of CAMP, in contrast to that of cGMP, was established by the early studies of Bell (1970) and Bell and Queener (1974) on porcine tissue slices, that were soon confirmed and extended for other mammals and for birds (Feinblatt and Raisz, 1971; Nieto et al., 1975). In cultured human medullary thyroid carcinoma (MTC) cells, CAMPdecreased cell proliferation, an indication of maturation in these cells, and increased CT gene transcription and CT release (De Bustros et al., 1985). The stimulating effect of CAMP was observed only in the presence of extracellular Ca2+.The suppressive effect of Ca2+ entry blockers on the in uitro release of CT from rat thyroparathyroid gland complexes was consistent with this observation and indicated the importance of Ca2+influx as a signal for CT release (Ramp et al., 1979). During treatment with the calcium channel agonist Bay K 8644 (Hishikawa et al., 1985a) or after a rise in extracellular Ca2+ (Haller-Brem et al., 1987) cytosolic Ca2+and CT release were raised concomitantly in rat MTC cells. Fried and Tashyan (1986) have drawn attention to the unusual calcium sensitivity of cytosolic Ca2+in such cells to changes in extracellular Cazf. HaUer-Brem et al. (1987) studied a human MTC cell line that, through an

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

177

apparent defect in the outer cell membrane, was unresponsive to changes in extracellularCa2+.Calcitonin release could be stimulated by extracellular Ca2’ after electropermeabilizationof the cells, as well as by ionomycin, a Ca” ionophore acting on intracellular Ca” stores (Haller-Brem et al., 1987). Cytosolic Ca2+ and CT release were raised in rat MTC cells by voltage-dependent calcium channel agonists Bay K 8644 and (+)202-791 (Hishikawa et al., 1985a; Muff et al., 1988). Involvement of voltageindependent calcium channels was indicated by the stimulatory effect of maiotoxin on CT secretion from cultured rat fetal thyroid C cells (Nishiyama et al., 1990). Involvement of the phosphoinositide pathway was first indicated by the results of studies with phorbolesters. The phorbolester 12-0-tetradecanoylphorbol-13-acetate (TPA) had effects on rat (Hishikawa et al., 1985a,b)and human (De Bustros et al., 1986) MTC cells that were similar to those of CAMP decreased cellular proliferation, and stimulated CT gene transcription and CT release. In the activation of protein kinase C, TPA may act as a substitute for diacylglycerol. In the rat MTC cells synthetic diacylglycerol also stimulated release. The effect of both TPA and diacylglycerol was synergistically enhanced with the calcium ionophore A23 187 (Hishikawa et al., 1985b). TPA did not alter cAMP levels: the effects of TPA and cAMP on CT release were additive (Hishikawa et al., 1985b; De Bustros et al., 1986),similar to the effects of ionomycin and TPA in human MTC cells (Haller-Brem et al., 1988). Thus, similar to many other cells, including the parathyroid cells, both the CAMP-dependent protein kinase A route and the Caz+-dependentphosphatidyl inositol pathway resulting in activation of protein kinase C might be involved in CT secretion. However, in contrast to most other cells, in the C cells the effect of both pathways was stimulatory and closely linked. This was further demonstrated by the observation that forskolin, an adenylate cyclase agonist, increased cAMP levels dramatically in human MTC cells, without changing cytosolic Ca2+or increasing CT levels more than marginally. Ionomycin and TPA did not change cAMP levels but promoted CT release. However, forskolin synergistically enhanced the release-stimulating effects of ionomycin and TPA (Haller-Brem et al., 1988). 111. Aquatic Vertebrates

Plasma calcium levels in amphibians and fish are directly influenced by the calcium concentration of the environment, because of the intimate contact between the blood and the water via the skin and, in particular, the gill surface. Changes in water Ca2+concentration as well as, for instance,

178

S. E. WENDELAAR BONGA AND P. K.T. PANG

periods of rapid growth may have pronounced effects on the plasma calcium level (Fig. 10). This indicates that extracellular calcium is less precisely controlled in these animals than in the higher vertebrates, even though the aquatic animals are very well equipped to rapidly reduce experimental hypercalcemia (Ma and Copp, 1978; Lafeber and Perry, 1988). As discussed in Section I, control of calcium homeostasis and calcium metabolism in the purely aquatic vertekates is different from that of the terrestrial vertebrates: parathyroid glands are missing, the pituitary glands may act as a source of hypercalcemic hormones, in particular prolactin, and some bony fishes possess a unique hypocalcemic hormone in addition to CT- stanniocalcin. The group of amphibians includes terrestrially oriented as well as purely aquatic species, and this difference is reflected in the control of their

3-

OJ,

,\

0

1

+*

2

3

4

5

warel calcium ImMI

FIG. 10. Effect of the water calcium concentration of freshwater (FW) on plasma total and ultrafiltrable calcium and phosphate levels of male teleost fish (Oreochromis mossambicus). Fish from water with 0.8 mM calcium (controls) were exposed for 5 weeks to water with higher or lower calcium concentrations; means 5 SD of eight fish per group; *, significantly different from controls, p < .05; ** p < .01. The peak at left corresponds to the calcium concentration (0.2 m M )that gives the highest growth rate (Urasa and Wendelaar Bonga, 1987). 0 , Total calcium; 0, ultratiltrable calcium; D, total phosphate; 0, ultrafiltrable phosphate.

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

179

calcium metabolism. Parathyroid glands are found in the more advanced and partially or fully terrestrial anurans (frogs and toads), and in some of the urodeles (newts and salamanders). Interestingly, in these animals the glands develop around the time of metamorphosis of the purely aquatic larvae into the adult life form. In many purely aquatic urodeles and in the fishes, parathyroid glands are missing. Of the systematically highly diverse group of fishes only the modem rayfinned fish, the teleosts, have been studied to an extent that allows some general, albeit tentative, conclusions with respect to the control of calcium metabolism. A. PROLACTIN

I. Introduction The absence of parathyroid glands in some lower, purely aquatic amphibians and in fish stimulated the search for a hormone in these animals that exerts the function of PTH as a hypercalcemic endocrine factor. Several studies on the aquatic vertebrates have indicated the pituitary gland as the endocrine organ dominating hypercalcemic control, with prolactin (PRL) as the active principle. The involvement of PRL in the control of calcium metabolism is certainly not limited to the aquatic animals. In the terrestrial vertebrates PRL has effects on calcium metabolism that may become functional during specific life periods, such as pregnancy and lactation in mammals and yolk formation in the lower vertebrates. During these periods prolactin secretion may be markedly increased. During lactation in rats, suppression by bromocriptine of the high rate of PRL release reduced the elevated intestinal calcium uptake to the levels found in nonlactating animals. The effect may be mediated by 1,25(OH)*D3(Robinson et al., 1982).During lactation, prolactin may stimulate calcium mobilization from the maternal skeleton via a vitamin D-independent mechanism (Brommage and DeLuca, 1983, and promote the transfer of calcium to the fetus (Barlet, 1985). Although there are several reports stating that PRL has hypercalcemic effects and that the rate of PRL secretion is influenced by the circulated calcium level, the literature is inconsistent. In a recent review we concluded that the available information does not support a role for PRL in the homeostatic control of extracellular calcium, and that its involvement in calcium metabolism is limited to the regulation of the transfer of calcium to the offspring (Wendelaar Bonga and Pang, 1989). Interestingly, in some amphibians the role of the pituitary gland in calcium control seems more important the more time the animals spend in the water. In frogs and toads the extracellular calcium concentration seems mainly under control of PTH and CT. Although the available infor-

180

S. E. WENDELAAR BONGA AND P. K . T. PANG

mation does not allow definite conclusions, the function of both hormones in these animals is comparable to that of the higher vertebrates (see previous sections). Nevertheless, about 60 years ago the first evidence for the implication of the pituitary gland in the control of extracellular calcium was obtained in the toad Xenopus laeuis, one of the few permanently aquatic anurans. Removal of the pituitary gland induced hypocalcemia that could be partially corrected by extracts of the anterior lobe (Shapiro and Zwarenstein, 1933). These observations were confirmed afterward. In the mainly aquatic salamanders and newts, there is more evidence for a role in calcium control of the pituitary gland, although its importance is variable and related to the presence or absence of the parathyroid glands. In red-spotted newts, which possess parathyroids, removal of the pituitary gland had no significant effect on plasma calcium, whereas parathyroidectomy resulted in marked hypocalcemia (Wittle, 1984). In some other newt species removal of either the parathyroids or the pituitary gland caused hypocalcemia (Oguro and Uchiyama, 1975; Matsuda et al., 1991). Conversely, in the giant water salamander the parathyroidectomy had no effect on plasma calcium whereas significant hypocalcemia followed the removal of the pituitary gland (Oguro, 1973). In some urodeles that are lacking parathyroid glands, removal of the pituitary gland produced hypocalcemia (Oguro and Uchiyama, 1975; Oguro et al., 1978). That PRL was the active hypophyseal principle was indicated by studies showing that the hypocalcemia followingremoval of the pituitary gland can be compensated by injections of preparations of mammalian PRL (Oguro et al., 1978;Pang, 1981b). In Cynops pyrrhogaster, a newt species that exhibited a transient hypocalcemia after removal of its parathyroid glands, the recovery following this operation could be blocked by hypophysectomy (Oguro et ai., 1978)and by injection of an antiserum against homologous prolactin (Matsuda et al., 1990).Thus, in the amphibians some species seem to depend on the parathyroid glands for their hypercalcemic control, some on the pituitary gland, and others on both. However, the function of prolactin in this group in the control of calcium homeostasis of the extracellular fluid still has to be defined. Studies on the importance of the extracellular Ca2+level on prolactin secretion in amphibians are badly needed. In fish the evidence for a role of PRL in calcium metabolism is more complete than in amphibians. Removal of the pituitary gland in freshwater fish frequently, although not invariably, leads to hypocalcemia. The reported difference in the responses may be related to the species studied and to the calcium concentration of the ambient water (see review by Wendelaar Bonga and Pang, 1989). The reduced calcium levels caused by removal of the pituitary gland of killifish from low-calcium freshwater

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

181

could be corrected by pituitary homogenates or mammalian prolactin (Pang, 1981a), an observation that was confirmed for other species. Treatment with mammalian as well as teleost PRL induced hypercalcemia in intact freshwater fish, although negative results were also reported (Wendelaar Bonga and Pang, 1989). Studies by Flik et al. (1984, 1989a) showed that cells specialized for ion transport in the gills of several fish species contain a specific Ca2+-ATPaseactivity that is the driving force for branchial calcium uptake from the water and that is stimulated by PRL. Administration of PRL increased the calcium density of the skeleton in tilapia (Wendelaar Bonga and Flik, 1982), and thus the hypercalcemic effects of PRL are unlikely to be produced in a PTH-like fashion, i.e., by stimulating bone calcium release. Pituitary factors other than PRL that were reported to increase plasma calcium levels in fish are ACTH and hypercalcin, the hypercalcemic factor postulated by Parsons et al. (1978)to explain the fast-acting hypercalcemic effects of pituitary extracts of cod and eel. These reports need confirmation, however (Wendelaar Bonga and Pang, 1989). 2 . Control of PRL Secretion

a. Calcium and Magnesium Concentration of the Water. The calcium status of a fish is dependent on the calcium concentration of the ambient water to such an extent that fish and water have even been considered as an “open system” in terms of calcium exchange. Whether this is defendable or not, changes in water calcium concentration are reflected, although often only transiently, in changes in calcium concentration in the blood plasma and in the bony compartment, in particular the scales. For several freshwater teleost species an inverse relationship was established between the water calcium concentration and PRL cell activity (Wendelaar Bonga and Pang, 1989). Species differences do exist, however. Whereas in the African cichlid Oreochromis mossambicus PRL secretion could be suppressed almost completely by high water calcium concentrations (Wendelaar Bonga et al., 1985), in European eels only a modest decrease occurred after a similar treatment (Olivereau and Olivereau, 1983). In seawater a relationship between the water calcium concentration and PRL cell activity is questionable for most species. Whereas in sticklebacks the PRL cells became clearly activated after reduction of the high calcium concentration typical for seawater, only moderate effects or no effects at all were reported for other species (Wendelaar Bonga and Pang, 1989) and circulating PRL levels were not influenced by reducing the calcium concentration of the ambient seawater (Nicoll et al., 1981;Hirano, 1986). Thus, a specific relationship between the water calcium concentra-

182

S. E. WENDELAAR BONGA A N D P. K. T. PANG

tion and PRL cell activity was only demonstrated for a few teleost species, and was not found for fish in general. Changes in water magnesium levels were also reported to influence prolactin secretion. In sticklebacks and tilapia, fish with PRL cells responsive to changes in water calcium, PRL secretion could be suppressed by high water magnesium levels, although magnesium ions were less effective than calcium ions in these species (Wendelaar Bonga, 1978; Wendelaar Bonga et a / . , 1985). b. Extracellular Calcium. In their studies of tilapia PRL cells in uitro, Grau and Helms (1989, 1990) found that PRL release was independent of the extracellular Ca2+ concentration, although a basal extracellular calcium concentration was required. MacDonald and McKeown (1983) examined the Ca2+ dependency of PRL cells of coho salmon in uitro and established a 3-fold rise when Ca2+increased from 0 to 1.3 mM, which is close to plasma Ca2+in salmon, and a decrease at higher concentrations. The authors suggested that PRL might have some role in homeostatic Ca2' regulation. This suggestion was not supported by in uiuo observations on the relationship between prolactin cell activity and plasma total or free calcium. Although during experimental manipulation of water calcium concentration in fish an inverse relationship between changes in plasma calcium and PRL cell activity was reported, a clear dependency of PRL on plasma calcium, whether total or ionic calcium, was not established for any one species (Olivereau et al., 1982; Wendelaar Bonga et a / . , 1985, 1988). Strongly reduced or increased Ca2+levels are inhibitory for endocrine cells in general, and do not allow conclusions with respect to prolactin's function in fish. Osmolarity has more pronounced effects on PRL release in uitro than Ca2+,but the physiological significance of this phenomenon is also unclear because there is some discrepancy with in uivo observations (Wendelaar Bonga and Pang, 1989). c. Hormones and Neurotransmitters. Control of PRL cells seems to be dominated by hypothalamic centers, as is usual for pituitary cells. Teleost PRL cells in v i m respond to TRH with increased secretion, whereas dopamine, somatostatin, and vasoactive intestinal peptides are inhibitory. The response of the PRL cells to these factors in uiuo may be modulated by the sensitivity of the cells to, e.g., osmolarity of the extracellular fluid, cortisol, or estrogens (Nishioka et a f . , 1988; Grau and Helms, 1989; Wendelaar Bonga and Pang, 1989). No clear relationship with calcium regulatory hormones can be deduced from literature on the control of PRL secretion (Wendelaar Bonga and Pang, 1989).

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

183

B. CALCITONIN 1 . Introduction

a. Aquatic Amphibians. In the aquatic amphibians, i.e., some adult anurans, anuran tadpoles, and the urodeles, the function of CT seems comparable to that of the more terrestrial amphibians (see Section II,B,l). After surgical removal of the ultimobranchial bodies of the frog Rana pipiens (Robertson, 1970) or bullfrog tadpoles (Sasayama and Oguro, 1976), hypercalcemia developed when the animals were exposed to calcium-enriched water. The hypercalcemia could be reduced or prevented by transplantation of ultimobranchial tissue or by injection of salmon CT (Sasayama, 1978). These data are comparable with a role of CT in suppressing hypercalcemia in these animals. b. Bony Fishes. The physiological function of CT in fish is still under debate, partially as a result of the relative scarcity of the experimental data when compared with the higher vertebrates, and partially as a result of the inconsistency of many of these data. In conformity with the first studies on CT in mammals and birds, several investigators tried to demonstrate a hypocalcemic effect of CT in fish (Dacke, 1979). The results were variable. No effect on plasma total calcium was observed after injection of mammalian CT preparations or synthetic salmon CT in many different freshwater or seawater fish (Dacke, 1979; Copp and Kline, 1989). Infusion of Japanese eels with a homologous antiserum did not influence plasma Ca2+ levels (Hirano et al., 1981). Conversely, hypocalcemia following CT treatment was reported for freshwater European eel (Chan et al., 1968; Lopez et al., 1976), sticklebacks adapted to low-calcium freshwater (Wendelaar Bonga, 1981; only the ultrafiltrable calcium fraction was reduced), and for mudskippers that had been made hypercalcemic by taking them out of the water (Fenwick and Lam, 1988). Hypophosphatemia accompanied the reduction in plasma calcium in European eels (Chan et al., 1968) and hyperphosphatemiawas reported for mudskippers kept in, as well as out, of the water (Fenwick and Lam, 1988). The absence of an effect in most experiments may not only have been caused by the absence of any hypocalcemic potency of the CT preparations used, but also by the effective counteraction of hypercalcemic control mechanisms. In this respect it is of interest that in sticklebacks CT injections stimulated prolactin cell activity (Wendelaar Bonga, 1980). Prolactin has hypercalcemic effects in this species. This would indicate that CT has at least some calcium-lowering activity in fish; this was supported by the results of removal of the ultimobranchial glands. Partial ultimobranchialectomy in European and Ameri-

184

S. E. WENDELAAR BONGA AND P. K . T. PANG

can eels led to transient, although moderate, hypercalcemia, which became more pronounced when the fish were exposed to high calcium water (Lopez et al., 1976; Fenwick, 1978). The transient nature of the effect might have been caused by regeneration of the gland or by increased secretion of stanniocalcin: the corpuscles of Stannius became enlarged after ultimobranchialectomy (see Section I11,C). With respect to plasma calcium levels the effect of ultimobranchialectomy was to some extent similar to that of removal of the Stannius bodies. The latter operation led to marked but transient hypercalcemia (Fenwick, 1974; Hanssen et al., 1989) and was followed by gradual hypertrophy and hyperplasia of the ultimobranchial gland (Lopez et al., 1968). We conclude that CT probably has some hypocalcemic activity in fish. This potency is certainly less than that of stanniocalcin, the hormone of the Stannius bodies. This was illustrated by the observation on different species that the hypercalcemia following removal of the Stannius bodies activated the ultimobranchial cells but still remained excessive for at least 4-6 weeks. It could be rapidly reduced by the hormone of the Stannius bodies, stanniocalcin (see Section 111,C). Although CT may have hypocalcemic effects under some conditions, it is doubtful whether it has a hypocalcemic function in fish. The doses necessary to demonstrate an effect on plasma calcium are high and nonphysiological. In another important aspect the effect of CT is comparable to that described for higher vertebrates: activation of osteoblasts and promotion of bone mineral deposition has been reported after injection of salmon CT in tilapia (Wendelaar Bonga and Lammers, 1982). It is very possible that CT has a role in skeleton protection during periods of high calcium demand, such as during ovarian growth (see below).

c. Elasmobranchs. In contrast to the observations on bony fishes, salmon CT injections produced hypercalcemia in three sharks and rays (Glowacki et al., 1985). 2. Control of CT Secretion

a. Environmental and Extracellular Ca” . i. Aquatic amphibians. Only a few studies prevail on the control of CT secretion in the aquatic vertebrates. In bullfrog tadpoles, the ultimobranchial bodies became hyperactive when the ambient water was enriched with calcium. Since ultimobranchialectomy led to hypercalcemia under these conditions, the conclusion is indicated that the C cells responded with enhanced secretion to a rise in extracellular calcium (Robertson, 1970, 1971b). ii. Bonyfishes. Data are available only on teleost fish. When euryhaline fish are transferred from freshwater to seawater, usually a mild

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

185

hypercalcemia develops. If CT would operate as a hypocalcemic or antihypercalcemic hormone, one would expect that the ultimobranchials are highly active under the latter condition. However, the results of such experiments are contradictory. Pang (1971) could not find histological evidence for increased cellular activity in killifish from seawater as compared to freshwater. A similar observation was reported by PeignouxDeville et al. (1975) for European eels. Hirano et al. (1981) could not find any difference in plasma CT levels between freshwater and seawater Japanese eels, and in salmon migrating from seawater to freshwater, no clear relationship between plasma CT and plasma caIcium was found (Watts et al., 1975). Conversely, a study on Japanese eels transferred to seawater (Orimo et al., 1972) showed that plasma CT levels increased rapidly after transfer. Suryawanshi and Mahajan (1976) reported for catfish that hypercalcemia caused by calcium enrichment of the water induced hypertrophy of the ultimobranchial cells. The rise in plasma calcium in these fish was relatively high. Transfer of sticklebacks from normal seawater to low-calcium seawater, a procedure followed by a substantial drop in plasma calcium, was followed by regression of the ultimobranchial cells (Wendelaar Bonga, 1980). The extreme hypercalcemia that developed after removal of the Stannius bodies led to hyperactivity of the ultimobranchials (Lopez et al., 1968; Wendelaar Bonga and Greven, 1978). Ultimobranchial hyperactivity was also reported for parrotfish, which ingest extremely large amounts of calcium feeding on coral (Fontaine et al., see Peignoux-Deville et a f . , 1975). One might conclude that the glands only respond with increased CT secretion to marked hypercalcemia. This could explain why CT levels were not notably changed during migration, when plasma calcium levels do hardly change, but were affected after abrupt transfer to high- or low-calcium water, which is usually associated with transient plasma calcium imbalance. Our tentative conclusion is further supported by the results of the only in uitro study known to us on fish ultimobranchial cells, which showed that cultured trout C cells were far less responsive to changes in extracellular Ca” than the C cells of the terrestrial vertebrates (Roos et al., 1974), and also less responsive than the cells of the Stannius bodies (see Section 111,C). Thus the available data, albeit limited, on the control of CT secretion in fish do not indicate a dominant role for CT in the homeostatic control of extracellular calcium, although some effect cannot be excluded under extreme conditions. b. Hormones. i. Pituitary gland. Removal of the pituitary gland of Japanese eels did not influence circulating CT levels, although the animals developed a significant hypocalcemia (Hirano et al., 1981).

186

S. E. W E N D E L M BONGA AND P. K . T. PANG

ii. Estrogens. In fish there is a relationship between gonadal activity and plasma CT levels comparable to that of the terrestrial vertebrates. The ultimobranchial glands show histological signs of increased secretory activity during natural or experimentally induced sexual maturation, in particular in female fish (Peignoux-Devilleet d . ,1975). Yamauchi et a / .(1978) reported significantly increased CT levels in male eels during spawning. Serum calcium was unchanged. In another study on eels these authors determined immunoreactiveCT during ovarian maturation induced in eels by chum salmon pituitary homogenate. Serum CT levels increased parallel with the gonadosomatic index, and peaked during ovulation. No changes in serum calcium levels were observed (Yamauchi et al., 1978). Watts et al. (1975) determined circulating CT levels in male and female sockeye salmon during their spawning migration from the sea to freshwater. Plasma CT was higher in female fish and increased during migration until spawning. Afterward it decreased, below the concentration in males. Bjornsson et al. (1986) and Fouchereau-Peron et al. (1990) determined plasma CT levels in rainbow trout during the yearly sexual cycle. In female fish the CT levels were elevated during the 3 months of the reproductive period. They increased until ovulation and decreased sharply thereafter. The authors could not find a close correlation between circulating CT levels and free or protein-bound calcium concentrations. In male fish, plasma CT levels were constant throughout the year (Bjornsson et al., 1986), or fluctuated in relation to the gonadosomatic index (Fouchereau-Peron et al., 1990). In female fish, protein-bound calcium increased during the period of oogenesis. Free calcium was unchanged. It was concluded that free plasma calcium is not the primary controlling factor for CT secretion, because of the lack of a clear correlation between CT levels and plasma free calcium (Bjornsson et al., 1986). Recently Bjornsson et al. (1989) showed that injections of 17P-estradiol increased plasma CT levels in several salmonid species. They considered a reproductive role for CT in fish more likely than a calcium regulatory role (Bjornsson et al., 1989). Direct evidence for a reproductive role is still missing, however. iii. Gastrointestinal hormones. In their study on cultured trout ultimobranchial cells Roos et al. (1974) reported that pancreozymin had a marked stimulatory effect on CT release. However, relatively high conM). centrations were required (around

c. Second Messengers. Roos er al. (1974) found that dibutyryl CAMP stimulates CT release by cultured trout ultimobranchial cells.

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

187

C. STANNIOCALCIN 1 . Introduction

The corpuscles of Stannius (CS) are found exclusively in teleostean and holostean fishes. They produce a hormone (stanniocalcin) that seems unique for these animals and that prevents hypercalcemia. The glands are associated with the kidney. In most fishes one pair is found. Higher numbers are present in more primitive teleosteans, whereas several hundreds have been described in the holostean Amia calua. The cytophysiology of these glands has been reviewed by Krishnamurthy and Bern (1971), Wendelaar Bonga and Pang (1986), and Hirano (1989). Briefly, the bodies contain either one or two structurally different types of gland cells (type-1 and type-2 cells), both showing the ultrastructure of proteinproducing endocrine cells. They may produce different hormones or may represent different phases of the same cell type, although each responds differently to experimental treatments. In typical marine fish so far only type-1 cells have been described, whereas two types occur in freshwater fish or euryhaline fish spending part of their life cycle in freshwater. The type-1 cells predominate and are rounded with extensive granular endoplasmic reticulum and large secretory granules. The small granules are present in the type-2 cells, which are slender, often with cytoplasmic extensions. The CS are well vascularized. The blood vessels form an intricate capillary network that is drained by veins that join the renal portal system. The CS are richly innervated by autonomic nerves that penetrate the gland along with the blood vessels. Small nerves follow the venules and arterioles running through the connective tissue septa that separate the lobes of gland cells. No synaptic contacts with the gland cells have been reported, and the nerves and axons seem to end at a distance from these cells. The larger nerves may be associated with ganglionlike groups of neurons at the periphery of, or outside, the glands (Wendelaar Bonga and Pang, 1986). In a detailed study on trout CS, Unsicker et al. (1977) demonstrated the presence in the nerves of considerable amounts of catecholamines, 5-hydroxytryptamine, and acetylcholine. Although nervous control of secretion of the CS cannot be excluded, most authors consider it likely that the nerves are mainly controlling the blood flow through the glands. Removal of the glands (stanniectomy) leads to marked hypercalcemia (Fig. 11; Hirano, 1989). The hypercalcemia is caused by increased Ca2+ influx from the water (So and Fenwick, 1979; Milet et al., 1979). It persists for 5 to 7 weeks (Fenwick, 1978; Hanssen et al., 1989), which indicates that endogenous calcitonin, probably secreted at higher levels in

188

S. E. WENDELAAR BONGA AND P. K. T. PANG

h

4

v

V

-

zE

Y

7

N

m

$2

3

2 1 0

10

20

30

10

20

30

40

8

7

6

5 4

3

2 1

0

40

Time (days) FIG. 11. Effect of removal of the Stannius corpuscles (STX)or sham operation (control) on the blood ionic calcium concentration (A) and plasma total calcium concentration (B) in cannulated eels. Values are represented as means ? SEM; N = 8 (Hanssen ef al., 1989).

stanniectomized fish (Lopez et ai., 1%8), is unable to restore normal calcium levels in these fish, in contrast to injections of the hormone of the CS. Corpuscles of Stannius extracts reduce Ca” intlux from the water by acting on the gills (So and Fenwick, 1979; Milet et al., 1979) and on the gut (Takagi et al.. 1985). 2 . Identity, Biosynthesis, and Action of Stanniocalcin There has been some discussion on the identity of the active principle of the CS. Pang et al. (1973) characterized it as a heat stable product of high

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

I89

molecular weight and called it hypocalcin. A small glycoprotein with a molecular weight between 3000 and 4000 kD was isolated from salmon by Ma and Copp (1978) and was named teleocalcin. In the CS of European eels a product was identified that disappeared from the gland after calcium infusion. It cross reacted with an antiserum against PTH and therefore the name parathyrin of the CS was suggested (Milet et al., 1980; Lopez et al., 1984). More recently Butkus et al. (1987, 1989) established the primary structure of the main secretory protein of the CS of Australian eels from the cDNA sequence. It appeared to be a preprotein of 263 amino acids: a 17 amino acid signal peptide, a prosegment of 15amino acids, and the mature form of the hormone consisting of 231 amino acids. The N terminal 1-20 fragment seems to contain most biological activity, since it inhibits Ca2+ uptake in fingerling trout. The N-terminal sequence of eel proved to be almost identical to the N terminal of hypocalcemic glycoproteins isolated from salmon and trout (Wagner et al., 1986; Lafeber et al., 1988a). It had no true homology with PTH. Most authors agreed on stanniocalcin as the new name of the hormone (Flik et al., 1989b), a glycoprotein with a molecular radius that seems to vary slightly between species and that probably occurs in a labile dimeric form. In salmon the monomeric form of the hormone is a 26-kDa product, with a 32-kDa precursor (Wagner et al., 1988); in trout and goldfish these values are 28 and 32 kDa, and in eels 30 and 34 kDa, respectively (Flik et al., 1989b). Stanniocalcin appeared to be a relatively fast-acting hormone with the capacity to reduce effectively experimentally increased plasma calcium levels, by inhibiting the Ca” influx from the water (Wagner et al., 1986; Lafeber and Perry, 1988; Hanssen et al., 1989; Perry et al., 19891, and leaving the Ca2+efflux unaffected (Lafeber et al., 1988~;Fig. 12). As indicated earlier by using CS extracts (Bailey and Fenwick, 1975) the purified hormone specifically reduced the ionized calcium fraction (Lafeber et al., 1988b; Hanssen et al., 1989). A model for the mode of action of stanniocalcin is presented in Fig. 13. 3. Control of Stanniocalcin Secretion a. Calcium Concentration of the Water. Although fish are able to regulate their plasma calcium concentration efficiently, there is a close relationship between the calcium concentration of the ambient water and the secretory activity of the gland cells of the CS,in particular the predominant type-1 cells. As established with light and electron microscopy, these cells become activated upon transfer of fish from freshwater, with to seawater (full moderate to low Ca2+ levels (generally below 1 a), strength seawater contains about 10 mM Ca2+). This has been demonstrated for, e.g., goldfish, killifish, sticklebacks, and trout (Wendelaar

190

S. E. WENDELAAR BONGA AND P. K. T. PANG

FIG. 12. Effects of Stannius corpuscles crude homogenate (CH), purified stanniocalcin (SC),and residue proteins (RP)on Ca2+intlux and efflux across the gills in trout. Control (C), sham injected fish; hatched bars: net flux; *, significantly different from controls; means ? SE; numbers of fish per group are indicated in the bars (Lafeber et ol., 1988~).

Bonga and Pang, 1986). The CS were also activated in Ca-enriched freshwater (Wendelaar Bonga and Pang, 1986; Urasa and Wendelaar Bonga, 1987) but not in Ca-deficient seawater (Cohen et al., 1975). Increased activity of the CS in seawater was further indicated by the high immunoreactive stanniocalcin levels as observed in the blood plasma of seawater eels. b. Extracellular Calcium. Because changes in water Ca2+concentration are frequently reflected in extracellular calcium levels, the effects of the external Ca2+concentration on CS activity could well be mediated by extracellular calcium. This is in fact indicated by several reports demonstrating that increases in plasma calcium levels by injection of CaC12 result in degranulation of the gland cells of the CS. This was shown for goldfish, eel and, trout (Yamada e f al., 1982; Lopez et al., 1984;Lafeber and Perry, 1988; Hanssen et al., 1991). The degranulation was associated with the disappearance of stanniocalcin from the glands and stimulation of synthesis of the hormone (Flik et al., 1989b), elevation of plasma stanniocalcin levels (Hanssen et al., 1991), and inhibition of Cat+ influx from the water (Hanssen et al., 1989; Lafeber and Perry, 1988; Perry et a f . , 1989). The hypercalcemia resulting from the injection of CaC12was reduced rapidly in

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

191

U

FIG. 13. Model for the action of stanniocalcin on calcium uptake by chloride cells in the bronchial epithelium (BE), based on observations of eels and trout. Ca2+from the ambient water (W)enters the chloride cells in the gills passively through Ca2+channels (ch) in the apical cell membrane (am).It is transported via Ca2+binding proteins (Ca2+bp)and translocated across the basolateral cell membrane (bm) by Ca2+-ATPaseactivity into the extracellular fluid (ef) and in the blood (B). Stanniocalcin (SC) blocks the Ca2+entry into the chloride cells, an effect that may be mediated by an unknown second messenger; tj, tight junction (G. Flik, F. P. J. G. Lafeber, P. K.T. Pang, S. F. Perry, and S. E. Wendelaar Bonga, unpublished).

intact fish. When the CS were removed, the CaClt-inducedhypercalcemia persisted, demonstrating that the restoration of preinjection plasma Ca2+ levels was affected by the CS and not by the ultimobranchialbodies (Perry et al., 1989). The above data show that there is a negative feedback relationship between plasma ionic calcium and stanniocalcin secretion by the CS. The question presents itself whether the plasma Ca2+concentration influences the stanniocalcin secretion directly, at the level of the gland cells, or indirectly, i.e., via normal or neural pathways. This problem has been studied by analysis of the effects of varying Ca” concentrations on CS secretion in uitro. Aida et al. (1980) showed that incubation of salmon CS in high-Ca2+media (3 and 6 mM Ca2+)led to degranulation of the gland cells. At a concentration of 1.5 mM the upper end of the physiological range of plasma Ca2+ degranulation was absent, although it could be induced by the Ca2+ ionophore calimycin. In line with these results,

192

S. E. WENDELAAR BONGA AND P. K. T. PANG

Hanssen et a f . (1991) found for eel CS that variations in Ca2+concentration of the incubation medium between 0 and 2.00 mM- thus including the physiological range of plasma Ca2'-did not significantly affect the basal release of newly synthesized or total stanniocalcin. Stimulated release and degranulation of the gland cells was observed only at concentrations of 2.5 and higher. This is well above the physiological plasma Ca2+concentrations. Such high concentrations were obtained experimentally by CaC12 injections, as reported above. Thus, these results do not support the hypothesis that changes of plasma Ca2+ in the physiological range have a direct effect on the rate of secretion of stanniocalcin. However, Wagner et a/. (1989) demonstrated for trout CS that stanniocalcin release is concentration-dependently stimulated in uitro by Ca2+ between 1 and 2.5 mM.This result was obtained with CS cells in primary culture. This fact, or species difference, may account for the discrepancy between these observations. c. Hormones and Neurotransmitters. Several hormones may influence the secretory activity of the CS, although the effects may be mainly indirect. The evidence is exclusively based on histological observations. No relationship between thyroid glands and CS has been demonstrated. The pituitary gland may be implicated in the control of CS activity, but the observations are inconclusive. After hypophysectomy, no effect, reduced granulation, or hypertrophy and hyperplasia have been reported (Hirano, 1989). Ovine prolactin caused hypertrophy of the CS as well as hypercalcemia, indicating that the effect of prolactin was mediated by plasma calcium (Olivereau and Olivereau, 1978).The absence of direct effects of prolactin has been substantiated by the finding that ovine prolactin had no direct effects on stanniocalcin secretion of trout CS in primary culture. Direct effects of neither salmon growth hormone nor salmon gonadotropic hormone could be demonstrated (Wagner et al., 1989). Removal of the adrenocortical homologue of freshwater eels resulted in regression of the CS, whereas after the same operation in seawater eels the glands showed hypertrophy. These effects were also most likely mediated by plasma calcium, since after the operation the freshwater eels became hypocalcemic and the seawater eels hypercalcemic (Hanke et al., 1967; Olivereau and Olivereau, 1%8). Injection of mammalian calcitonin stimulated the type-I cells of the CS of sticklebacks (Wendelaar Bonga, 1980). Removal of the ultimobranchial glands in eels led to regression of the CS (Chan, 1972), similar to the removal of the ovaries in sexually mature female catfish (Pandey, 1988). The latter effect could be prevented by estradiol injections. These injections stimulated the CS in intact fish (Pandey, 1988). In sexually mature female fish, the secretory activity of the CS is generally

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

193

elevated during the period of rapid ovarian growth (Subhedar and Prasado Rao, 1979; Urasa and Wendelaar Bonga, 1985). Then the liver produces calcium phospholipoproteins (yolk proteins or vitellogenins) under the influence of estrogens. These yolk proteins are secreted into the blood, transported to the ovaries, and incorporated in growing oocytes. This process causes markedly elevated plasma levels of protein-bound calcium. The activation of the CS during reproduction may be a response to these high protein-bound calcium levels since they may tend to increase the plasma ionic calcium concentration (Urasa and Wendelaar Bonga, 1985). Multiple injections of vitamin D3 and 1,25(OH)zD3induced hyperactivity of the CS in catfish (Srivastav et al., 1985; Srivastav and Srivastav, 1988). This effect probably is also indirect because vitamin Dj and several of its metabolites, including 1,25(OH)&, cause hypercalcemia in these fish (Swarup and Srivastav, 1982; Srivastav et al., 1985; Srivastav and Srivastav, 1988). Although the CS are richly innervated, the nerve fibers are associated with the blood vessels rather than the gland cells, and neural control of the secretory activity seems therefore not indicated. However, recently R. G. J. M. Hanssen (unpublished) showed that the acetylcholine agonist carbachol stimulates stanniocalcin synthesis and release by eel CS in vitro. The effect was dose related and could be blocked by the atropine. This observation opens the possibility that the secretory activity of the CS is modulated or even controlled by the nervous system. 4. Second Messengers

Only a few studies have dealt with extracellular signal transduction in the CS cells. Aida et al. (1980) showed that the calcium ionophore A23187 caused degranulation of the CS of coho salmon if some extracellular Ca2+ was present, thereby mimicking the effects of high extracellularCa2+.This observation was confirmed by Wagner et al. (1989) for rainbow trout CS cells in primary culture. EDTA or cobalt chloride blocked the effect of high calcium on hormone release. These observations indicate that Ca2+entry into the cell, possibly via voltage-gated Ca2+channels, can initiate stanniocalcin release. However, the voltage-gated channel blockers verapamil and nifedipine gave inconclusive results. The authors concluded that either the binding sites for these compounds were unrelated to voltagedependent CaZ+channels or more than one type of Ca” channel was present in the trout CS cells (Wagner et al., 1989).Different concentrations of dibutyryl cyclic AMP were ineffective in changing STC release in uitro from the CS of the cichlid fish Oreochromis mossambicus (S. E. Wendelaar Bonga, unpublished).

I94

S. E. WENDELAAR BONGA AND P. K. T. PANG

IV. Conclusions A. PARATHYROID HORMONE

Parathyroid hormone cells have at least three properties that make these cells unique among the endocrine cells, and optimally equipped for the homeostatic control of extracellular Ca2+. First is the intimate inverse relationship between PTH release and the extracellular Ca2+ level that enables these cells to respond rapidly and efficiently to very small changes in Ca2+.This property of the cells is based on the extreme sensitivity of a Ca2+-receptormechanism on the outer cell membrane. Although in the last 10 years there has been much progress in our understanding of the functioning of this mechanism, we are still waiting for its definite identification. Dihydropyridine-sensitive Ca2+ channels coupled to guanine nucleotide regulatory proteins are good candidates. The relationship between intracellular Ca2+and PTH release represents another unique feature of the PTH cells: it is the reverse of what is found in other gland cells in that an increase in cytoplasmic CaZ+concentration is associated with inhibition of hormone release. The studies on permeabilized cells have revealed the complexity of the relationship between intracellular CaZ+and PTH secretion. Cytoplasmic Ca2+is certainly not the common denominator of stimulus secretion coupling. At least the inhibitory effect of high extracellular Ca2+on FTH release may be mediated by local fluxes across the outer cell membrane rather than by an increase in cytoplasmic Ca" . Most stimulatory agonists of PTH release increase cAMP levels. However, for at least the stimulatory actions of low extracellular Ca2+-the dominating factor in the control of PTH release-and of other cations a rise in cAMP is not a prerequisite. Regulation of protein kinase C by diacyl glycerol could be the common key mechanism for stimulus-secretioncoupling in the PTH cells. The third unique property of the PTH cells is their capacity to increase hormone release without adapting the rate of hormone synthesis during the first hours of stimulation. This phenomenon is effected by maintaining a high rate of PTH synthesis combined with inactivation of up to 90% of the newly found products under unstimulated conditions. Although uneconomical, this capacity enables the FTH cells to respond more readily and efficiently to a persistent release-promoting stimulus than other gland cells. Other cells, after depletion of their hormone stores, have to reactivate hormone synthesis if not gene translation and processing of gene products. The process of control by degradation of the output of the PTH cells bears only a slight resemblance to the process of crinophagy that has been reported for PRL cells (Farquhar, 1976).In the latter cells lysosomal

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

195

breakdown of secretory granules occurs after sudden inhibiton of hormone release and synthesis, whereas in the PTH cells synthesis proceeds unabated for at least several hours. Only more prolonged action of inhibitory or stimulatory agonists leads to adjustment of the rate of gene transcription and eventually of cell size and numbers. Thus, typical for the PTH cells are the high basal rate of hormone synthesis and the delay of at least several hours between the start of the stimulus and the onset of changes at the level of gene transcription. In addition to extracellularCa2+,many hormones and neurotransmitters have the capacity to influence PTH release. With few exceptions, the physiological significanceof the many nonionic FTH agonists is still poorly defined. For hormones such as calcitonin, glucagon, secretin, and cortisol the concentrations required to evoke a response seem too high to be of physiological importance. A potentiating role of these hormones when acting in concert with other factors cannot be excluded. Growth hormone, PRL, and estrogens, although implicated in the control of Ca2+-dependent processes such as growth, placental calcium transfer, and egg formation seem to have mainly indirect effects on PTH secretion. These hormones promote the transfer of calcium via the blood to their respective target organs, and this transport will affect the parathyroid glands. Since growth hormone and PRL are capable of changing vitamin D hydroxylase activity of the kidneys (Wendelaar Bonga and Pang, 1989), there is a fair possibility that these hormones also influence FTH secretion via 1,25(OH)& To date this steroid seems to be the only serious candidate for a direct and physiologically important modulator of PTH secretion. It would be of interest, however, to examine the effects of PRL and growth hormone on the setpoint for Ca2' of the PTH cells. Given the dominant position of extracellular Ca2+in the control of PTH secretion, what then might be the role of these hormones? The close relationship between Ca2+and PTH release seems to leave at most a modulating influence to these factors, at least for short-term regulation. Nevertheless, modulation may have important physiological functions in the long term, and several possibilities present themselves such as: 1. Adjustment of calcium homeostasis to specijic requirements of d$ferent life stages. Fetal and newborn mammals maintain their extracellular calcium at a higher level than older animals. This may be connected to the

rapid growth and calcification of the skeleton during early life. There is good evidence now, although limited to a few species, that the high extracellular calcium level is associated with a higher setpoint for Ca2+and thus of lower sensitivity to Ca2+of the PTH cells than in older animals. This suggests that the setpoint is a physiological variable that is under hormonal

I96

S. E. WENDELAAR BONGA AND P. K. T. PANG

or nervous control. Setpoint regulation needs more attention also, because of the recent report that the setpoint in vitro may be lower in individual isolated cells than in a cell population (Fitzpatrick and Leong, 1990). It should further be examined whether the setpoint is changed during other periods of high calcium transfer through the body circulation, in particular during the egg-laying cycle in birds and reptiles, and during ovarian growth in amphibians. Hormones to be considered for this type of adjustment of PTH secretion are growth hormone, PRL, and estrogens, possibly via stimulation of 1,25(OH)zDssynthesis, as mentioned above. 2. Control of daily and annual cycles in extracellular calcium levels. Daily cycles have been reported for most vertebrate groups, and annual cycles have been established in some lower vertebrates (Fig. 8). These cycles run parallel with changes in PTH levels or in other parameters for parathyroid gland activity. At least for the daily rhythmicity an endogenous origin is likely. Both the daily and annual cycles indicate that extracellular calcium is not rigidly controlled at a fixed level, and that most likely parathyroid secretion is under autonomic control of the PTH cells for the short term only. Hormones that might be involved in this type of modulation of PTH secretion are not known. 3. Adjustment to calcium and phosphate shortage. PTH release in response to short-term hypocalcemia seems an adequate response to restore normocalcemia. However, high and prolonged release rate evoked by chronic hypocalcemia might be uneconomical and injurious to the skeleton. Chronic hypocalcemia may be caused by dietary calcium shortage or by a high rate of bone mineralization, placental calcium transfer, or egg formation. It would lead to progressive demineralization of the bone and to phosphaturia. Under these conditions adjustment of PTH release to total body calcium and phosphate balance is necessary to limit the potentially injurious side effects of PTH action. The main hormone regulating total body calcium and phosphate balance, 1,25(OH)*D3,could be involved in this type of adjustment of PTH secretion. The understanding of the significance of the role of 1,25(OH)2D3in PTH control has been complicated by the apparently biphasic action on the parathyroid glands. The short-term inhibitory effects reported in early literature are likely indirect and mediated by the elevation of extracellular Ca2+following administration of the steroid. Its receptor-mediated inhibitory effect on PTH gene transcription, although indicated by histological observations of cellular inactivation, has long escaped the attention of biochemists, probably because it is a long-term phenomenon. Both the short-term and long-term effects occur at physiological concentrations. Total body calcium and phosphate balance are controlled by 1,25(OH)*D3

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

197

through its stimulating effect on intestinal calcium and phosphate uptake. We tentatively conclude that during hypocalcemia caused by intense skeletal calcification, placental calcium transfer, or egg formation, the potentially harmful effects of PTH on the skeleton are reduced or prevented by enhanced formation of 1,25(OH)zD3.This steroid subsequently stimulates intestinal calcium uptake and suppresses PTH synthesis. The formation of 1,25(0H)*D3 in the kidneys may be promoted by one or more of the hormones known to influence this process: growth hormone during growth, PRL during placental calcium transfer, estrogens during egg formation, and PTH during hypocalcemia in general.

B. PROLACTIN In the terrestrial vertebrates the evidence for the involvement of PRL in calcium metabolism is limited to calcium handling during gestation and lactation in mammals. It would be of interest to investigate whether PRL would have a comparable function in the promotion of calcium from parent to offspring in the lower vertebrates. The available data indicate that only estrogens are involved. In the aquatic vertebrates the role of PRL in calcium regulation is definitely more comprehensive, although still not well defined and often confusing because of species-related differences. The data on amphibians show that PRL has a higher hypercalcemic potency in animals lacking parathyroid glands and with a more aquatic lifestyle. This would be interpreted as indicating that PRL has a function more or less equivalent to PTH in these animals. However, as will be discussed in the section on stanniocalcin,there is a fundamental difference in calcium metabolism between terrestrial vertebrates and fishes, and this may also apply to the lower amphibians and the purely aquatic amphibian larvae. For these animals the water might be an additional alternative source of calcium, as it is in fish. Any definite conclusion with respect to the role of PRL in the homeostatic control of calcium in the lower amphibians is premature because of the scarcity of data. For several fish species it has now been established that PRL has hypercalcemic actions, and in this respect the hormone might be compared with PTH. However, there are several differences with the latter hormone. First, PRL raises plasma Ca” by stimulatingthe uptake of Ca” from the water and not from the bone. Second, PRL is certainly not primarily implicated in the control of Ca2’ homeostasis of the extracellular fluid. The control of PRL secretion is not dominated by changes in extracellular Ca2+, but by hypothalamic factors and, possibly, factors such as plasma osmolarity. As has been concluded in recent reviews on PRL in fish, its main function seems to be the control of the integumental permeability to

198

S. E. WENDELAAR BONGA AND P. K . T. PANG

water and ions (Hirano, 1986; Wendelaar Bonga and Pang, 1989). The involvement of prolactin in calcium metabolism may be derived from this function, because calcium is important for structural integrity and permeability control of cellular membranes. In the absence of other known serious candidates for a €TH-like homeostatic function in fish one could tentatively consider the possiblity that a true hypercalcemic homeostatic hormone is missing. Such a function might be dispensable in these animals for reasons discussed in the section on stanniocalcin. C. CALCITONIN Whereas the regulation of PTH secretion is fully consistent with a homeostatic function of the hormone in the control of extracellular Ca2+, several arguments point against a comparable antagonistic role for CT. First, the extracellular Ca2+concentration is less prominent as a regulatory factor for CT secretion than it is for PTH secretion. The C cells are missing the exquisite sensitivity to changes in extracellular Ca2+ that characterizes the PTH cells, and the response is less prompt. Conversely, the involvement of hormones, in particular factors of gastric and intestinal origin and estrogens, is more prominent in the control of CT than of PTH. Second, whereas injection of PTH in adult animals, from mammals to amphibians, produces a dose-dependent hypercalcemia, exogenous CT usually decreases plasma calcium only mildly or not at all. The results are rather consistent from mammals to fish and indicate that CT is not essential for the minute-to-minute control of extracellular Ca” . This inference is further supported by the absence of serious consequences for plasma calcium following the removal of the C cells. Such observations have even been interpreted as a lack of hypocalcemic potency of CT. This interpretation seems untenable because there are many reports dealing with marked hypocalcemia after CT administration. In addition, in studies on the effect of exogenous CT on experimentally induced hypercalcemia the hormone almost invariably is effective in reducing plasma calcium levels to normal. An exception must be made for teleost fish (see below). How then can it be explained that CT proved to be ineffective in many experiments? Several explanations are plausible: 1. The response ofthe bone cells. Not only the properties of the hormone but also the condition of its main target organ, the bony tissue, determines the capacity of CT to reduce plasma calcium. Because the CT-induced hypocalcemia is mainly effected by the deposition, via cellular

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

199

interaction, of calcium and phosphate compounds, the effect of CT on plasma calcium is dependent on the presence of active bone-salt depositing cells as well as of adequate amounts of phosphate. In general, osteoblastic and osteocytic activity is highest in young animals and decreases with age. Only in female birds and female reptiles a particular type of metabolically highly active bony tissue develops during the egg-laying period. Interestingly, reports on marked hypocalcemia following CT administration are mainly restricted to young animals, from mammals to amphibians (reports on young fish are not available), and to egg-laying birds. 2. Lack of adequate phosphate supplies. There is evidence that the response to exogenous CT is enhanced by the concurrent injection of phosphate (Talmage et al., 1981). The few reports of CT-induced hypocalcemia in fish include experiments with sticklebacks from low-calcium water (Wendelaar Bonga, 1980),a condition that is usually associated with hyperphosphatemia. The possibility of phosphate as a limiting factor for calcium deposition into the bone should receive more attention. 3. Counteraction by PTH. Any reduction of plasma Ca2' by CT will be counteracted intensely by PTH, as soon as the Ca2+ level tends to fall below the setpoint for Ca2+of the PTH cells. Of the above three explanations, the last one most likely is of major importance. The calcium-lowering effects of CT during experimental hypercalcemia clearly show that the calcium-depositing capacity of the bone cells or the availability of circulating phosphate are in general not limiting, although they may modulate the intensity of the response to CT. Only in young animals, with highly active bone forming cells, the rate of CTdependent calcium deposition may exceed the calcium-mobilizing capacity of PTH. In adult animals, the apparent predominance of PTH over CT in manipulating extracellular Ca2+seems to warrant that stimulation of CT release may induce harmful hypocalcemia. This is indicated by the readily induced hypocalcemia by CT in parathyroidectomized animals. If the hypocalcemic potency of CT cannot be denied, what then is the function of this hormone? The evidence points against a homeostatic function under steady state conditions of the animal. The extracellular Cat+ level has a natural tendency to fall, as becomes clear after parathyroidectomy. Apparently, renal excretion and exchange of Ca2' with the bone result in a negative balance for extracellular Ca2+.Thus, when there is no net uptake from the environment, maintenance of extracellular Ca2' homeostasis requires the continuous action of PTH, and not of a hypocalcemic hormone. Hypercalcemia may develop mainly during the episodic uptake of dietary calcium. Restoration of plasma Ca2' needs an antihypercalcemic hormone. For the terrestrial vertebrates this clearly is CT, which

200

S. E. WENDELAAR BONGA AND P. K. T. PANG

is in line with the postprandial surge in CT secretion that has frequently been reported. Another process that may disturb calcium homeostasis is the female reproductive activity. This not only requires increased intestinal uptake of calcium, which is transported and stored temporarily in the skeleton or, in amphibians, in lime sacs. It further involves remobilization of calcium and transfer through the blood, either for placental transfer to the offspring or, in the submammalian vertebrates, for incorporation into egg shells and yolk protein. These proteins are transported from the liver to the ovarium. Part of the calcium seems to be loosely bound to these proteins and may represent an additional load for the plasma Ca2+ level. Thus, the high circulating CT levels reported for female mammals at the end of gestation, for birds during egg laying, and for fish during the period of ovarian growth can be explained as an antihypercalcemicresponse. That plasma Ca2+levels are not, or only slightly, elevated under these conditions does not point against this interpretation, but indicates that also during periods of increased calcium transport via the circulation plasma Ca2+ is controlled efficiently. High CT secretion during female reproduction has been interpreted as a response aimed at protection of the skeleton against excessive demineralization. Antihypercalcemic and skeleton protective effects of CT are hard to separate because they represent interdependent aspects of CT action. We tentatively suggest, however, that the hypercalcemic stimulus predominates over any stimulus from the bony tissue in the control of CT secretion, since placental transfer or yolk formation proceed to the detriment of the skeleton when female animals are facing lack of exogenous calcium during gestation or ovarian growth. The importance of CT for controlling plasma Ca2' following episodic intestinal calcium uptake and during female reproduction is reflected in the hormonal control of CT cell function: important stimulatory effects have been described for gastric and intestinal factors, in particular gastrin, and for estrogen. Marked species-specific differences have been reported for gastrin, varying from prominent direct stimulation at physiological concentrations as in pigs, to no effect at all in rats. The reports for estrogens are contradictory for mammals, but more consistent for the lower vertebrates, including fish. This is understandable since egg formation and ovarian growth are under estrogenic control, whereas placental transfer of calcium to the offspring is controlled by P I U and other hormones rather than estrogens. A specific but undefined function for CT during reproduction as has been suggested by others cannot be excluded, but in our opinion the assumption of such a function is not necessary to explain the present data on CT secretion in reproducing female vertebrates.

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

201

D. STANNIOCALCIN The corpuscles of Stannius are now recognized as the source of an important hypocalcemic or, preferably, antihypercalcemic hormone of the holostean and teleostean fishes, and recently some consensus has been achieved about the identity of this hormone in teleosts. Caution is indicated, however, because the corpuscles have been studied in only very few of the more than 20,000 teleostean species. In these species the antihypercalcemic capacity of stanniocalcin clearly surpasses that of calcitonin, and its mode of action differs essentially from that of the latter. Stanniocalcin's main targets are not the bone but the entry routes of calcium: the intestine and, in particular, the gills. Thus, antihypercalcemic control in fish seems to be effected by reduction of calcium uptake from the environment and not by calcium deposition in the skeleton. Although some exchange of calcium and phosphate occurs between the skeleton-in particular the scales-and the extracellular fluid, fish bone seems to represent a phosphate store rather than a calcium store. At present we consider stanniocalcin the primary hormone for the homeostatic control of extracellular Ca2' in fish, and in this respect it is equivalent to FTH in the terrestrial vertebrates. However, whereas PTH has a hypercalcemic function, stanniocalcin is antihypercalcemic. This difference reflects a fundamental difference in calcium economy between terrestrial and aquatic vertebrates: in the water, calcium is present as a virtually inexhaustible source that can be tapped almost instantaneously. Terrestrial animals are dependent for their calcium on intermittent and variable dietary uptake, and therefore have to rely on internal stores during times when external calcium supplies are insufficient or not available at all. In seawater, with an ionic calcium concentration about eight times that of the extracellular fluid, hypercalcemia is a continuous threat for fish. But also in freshwater, with Ca2' concentrations similar to or up to ten times lower than that of the extracellular fluid, calcium can passively enter the body, across the outer apical membrane of specialized cells in the gills that are in close contact with the water. Even in very soft water the Ca2+concentration is at least 100-fold higher than typical cytoplasmic Ca" levels. We consider this gradient the driving force for Ca2' entry in fish. According to the model presented in Fig. 12, the passive entry of Ca2' is under inhibitory control of stanniocalcin. Then it can be explained why extracellular Ca2' increases to extreme values when the corpuscles of Stannius are removed. This rise contrasts with the fall in extracellular Ca2' that follows removal of the source of the primary Ca2' regulating hormone in the terrestrial vertebrates. The model further explains why the predominant homeostatic

202

S. E. WENDELAAR BONGA AND P. K. T. PANG

regulator in fish could be an antihypercalcemic hormone rather than a hypercalcemic hormone, even in low-calcium fresh water. Reduction of Ca2+entry into the blood by stanniocalcin will not necessarily lead to reduction of plasma Ca2+.This is effected as soon as the infiux of Ca2' falls below the losses of Ca2+ that occur through passive efflux across the gills or through renal excretion. In particular the hormonal control of Ca2+ excretion in fish needs more attention. Calcitonin might be involved (Fenwick, 1978). Because stanniocalcin seems to have no prominent effects on bone, the homeostatic regulation of Ca2+ in fish probably is less closely linked to phosphate metabolism than it is in the terrestrial vertebrates. Osteoclasts and encapsulated osteocytes are scarce or even absent in many fish, and this phenomenon indicates that the hormonal and skeletal mechanisms for the terrestrial type of control of Ca2+homeostasis have developed during or after the water-to-land transition in the vertebrate evolution. The control of stanniocalcin seems to be dominated by the Ca2+concentration of the extracellular fluid. The cells producing this hormone do respond to changes in external Ca2+ in v i m , but whether or not the sensitivity of these cells to Ca2+is sufficient to be of physiological importance still has to be demonstrated. A role for hormones and neural factors in the control of stanniocalcin secretion is indicated but its significance is unclear. Given the indispensability of stanniocalcin for Ca2+ homeostasis in teleost fish, the question arises how extracellular Ca2+ is controlled in other groups of fish, such as lungfishes, elasmobranchs, and jawless fish in which no Stannius corpuscles have been found. Both latter groups have no calcified skeleton. Another question is how amphibian tadpoles and newts maintain Cat+ homeostasis in the absence of stanniocalcin as well as parathyroid glands. Could calcitonin have a function comparable to stanniocalcin in these animals? Although in recent years there has been rapid progress in the understanding of calcium regulation in fish, these questions illustrate that our knowledge of aquatic vertebrates is still superficial and fragmentary. Because the development of the vertebrates started in the water, studies on fish and amphibians could provide data that are indispensable for our understanding of the evolution of calcium regulation in the higher vertebrates, including man.

ACKNOWLEDGMENT The authors are grateful to Mrs. Elizabeth Jansen-Hoorweg for her assistance during preparation of the manuscript.

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

203

REFERENCES Abe, M., and Shenvood, L. M. (1972). Biochem. Biophys. Res. Commun. 48,396. Ahren, B., Alumets, J., Ericsson, M., Fahrenkrug, J., Fahrenkrug, L., H&anson, R., Hedner, P., Lorkn, I., Melander, A., Rerup, C., and Sundler, F. (1980). Nature (London) 287,343. Ahren, B., Grunditz, T., Ekman, R., Htikanson, R., Sundler, F., and Uddman, R. (1983). Endocrinology (Baltimore) 113,379. Aida, K., Nishioka, R. S., and Bern, H. A. (1980). Cen. Comp. Endocrinol. 41,305. Aitken, J. M., Hart, D. M., and Smith, D. A. (1971). Clin. Sci. 41,233. Anast, C. S., Mohs, J. M., Kaplan, S. L., and Burns, T. W. (1972). Science 177,606. Anast, C. S., Winnacker, J. L.,Forte, L. R., and Bums, T. W. (1976). J. Clin. Endocrinol. Metab. 42,707. Attie, M. F., Brown, E. M., Gardner, D. G., Spiegel, A. M., and Aurbach, G. D. (1980). Endocrinology (Baltimore) 107, 1776. Aurbach, G. D. (1988). In “Calcium in Human Biology” (B.E.C. Nordin, ed.), p. 43. Springer-Verlag, London. Austin, L. A., Heath, H., 111, and Go, V. L. W. (1979). J. Clin. Invest. 64, 1721. Avioli, L. V., Birge, S. J., Scott, S., and Shieber, W. (1969). Am. J. Physiol. 216,939. Avioli, L. V., Shieber, W., and Kipnis, D. M. (1971). Endocrinology (Baltimore) 88, 1337. Bailey, J. R., and Fenwick, J. C. (1975). Can. J . Zool. 53,630. Baran, D. T., Whyte, M. P., Haussler, M. R., Deftos, L. J., Slatopolsky, E., and Avioli, L. V. (1980). J. Clin. Endocrinol. Metab. 50,377. Barlet, J.-P. (1985). J. Endocrinol. 107, 171. Bates, R. F. L., Bruce, J., and Care, A. D. (1969). J . Endocrinol. 45, XIV. Bates, R. F.L ., Bruce, J. B., and Care, A. D. (1970). J. Endocrinol. 46, XI. Baylin, S. B., Hsu, T. H., Stevens, S. A., Kalman, C. H., Trump, D. L., and Beaven, M. A. (1979). J . Clin. Endocrinol. Metab. 48,408. Bell, N. H. (1970). J. Clin. Invest. 49, 1368. Bell, N. H. (1975). Horm. Metab. Res. 7,77. Bell, N. H., and Queener, S. (1974). Nature (London)248,343. Benidge, M. J. (1987). Annu. Rev. Biochem. 56, 159. Benidge, M. J., and Irvine, R. F. (1984). Nature (London)312,315. Bemdge, M. J., and Irvine, R. F. (1989). Nature (London)341,197. Berson, S . A., and Yalow, R. S. (1968). J . Clin. Endocrinol. Metab. 28, 1037. Besnard, P., Jousset, V., and Garel, J.-M. (1989). FEBS Lett. 258,293. Bjornsson, B. T., Haux, C., Forlin, L., and Deftos, L. J. (1986). J. Endocrinol. 108, 17. Bjornsson, B. T., Haux, C., Bern, H. A., and Deftos, L. J. (1989). Endocrinology (Baltimore)

u5,1754, Blum, J. W., Fischer, J. A., Hunziker, W. H., Binswanger, U., Picotti, G. B., DaPrada, M., and Guilebeau, A. (1978). J. Clin. Invest. 61, 1113. Blum, J. W., Kunz, P., Fischer, J. A., Binswanger, U., Lichtensteiger, W., and Prada, M. D. (1980). Am. J . Physiol. 239, E255. Body, J. J., Cryer, P. E., offord, K. P., and Heath, H. (1983). J. Clin. Invest. 71,572. Breimer, L. H., MacIntyre, I., and Zaidi, M. (1988). Biochem. J . 255,377. Broadus, A. E., Mangin, M., Ikeda, K., Insogna, K. L., Weir, E. C., Burtis, W. J., and Stewart, A. F. (1988). N . Engl. J . Med. 319,556. Brommage, R., and DeLuca, H. F. (1985). Am. J. Physiol. 248, E182. Brookman, J. J., Farrow, S. M., Nicholson, L., O’Riordan, J. L. H., and Hendy, G. N. (1986). J . Bone Miner. Res. 1,529.

204

S. E. WENDELAAR BONGA AND P. K. T. PANG

Brown, E. M.(1982). Miner. Electr. Metab. 8, 130. Brown, E. M.(1983). J. Clin. Endocrinol. Metab. 56,572. Brown, E. M.,and Thatcher, J. G. (1982). Endocrinology (Baltimore)110, 1374. Brown, E. M.,Carroll, R. J., and Aurbach, G. D. (1977). Proc. Nail. Acad. Sci. U . S . A . 74, 4210. Brown, E. M.,Hunvitz, S. H., and Aurbach, G. D. (1978a). Endocrinology (Baltimore)103, 893. Brown, E. M.,Gardner, D. G., Windeck, R. A., and Aurbach, G. D. (1978b). Endocrinology (Baltimore) 103,2323. Brown. E. M.,Thatcher, J. G., Watson, E. J., and Leombruno, R. (1984a).Metab. Clin. Exp. 33, 171. Brown, E. M.,Redgrave, J., and Thatcher, J. (1984b). FEBS Lett. 175,72. Brown, E. M.,Leboff, M. S., Oetting, M.,Posillico, J. T., and Chen, C. (1987). Recent Prog. Horm. Res. 43,337. Bruce, B. R., and Anderson, N. C. (1979). A m . J. Physiol. 236,CIS. Butkus. A.. Roche, P. J., Fernley, R. T., Haralambidis, J., Penschow, J. D., Ryan, J. B., Trahair, J. F., Tregdar, G. W., and Coghlan, J. P. (1987).Mol. Cell. Endocrinol. 54,123. Butkus, A., Yates, N. A., Copp, H. A., Miliken, C., McDougall, J. G., Roche, P. J., Tregkar, G. W., and Coghlan, J. P. (1989). Fish Physiol. Biochem. 7,359. Calamy, H., and Bariet, J. P. (1970). C . R . Hebd. Seances Acad. Sci. 271,2153. Campbell, 1. L.. and Turner, C. W.(1942). Res. Bull.-M., Agric. Exp. Stn. 352, 1. Canterbury, J. M., Lerman, S., Cldin, A. J., Henry, H., Norman, A. W.,and Reiss, E. (1978). J . Clin. Invest. 61, 1375. Cantley, L. K., Russell, J., Lettien, D., and Sherwood, L. M.(1985). Endocrinology (Baltimore) 117,2114. Cantley. L. K., Russell, J. B., Lettieri, D. S.,and Sherwood, L. M . (1987). Calcif. Tissuelnt. 41,48. Capen, C. C. ( I97 1). A m . J . Med. 50,598. Care, A. D., Cooper, D. W., Duncan, T., and Onmo, H. (1968). Endocrinology (Baltimore) 83, 161. Care, A. D., Bates, R. F. L., and Gitelman, H. J. (1970). 1. Endocrinol. 48, 1. Care, A. D., Bruce, J. B., Boelhuis. J., Kenny, A. D., Conaway, H., and Anast, C. S. ( 197 I a). Endocrinology (Baltimore)89,262. Care, A. D., Bates, R. F. L., Swaminathan, R., andGanguli, P. C. (1971b).J. Endocrinol. 51, 735. Chan, D. K. 0. (1972) Gen. Comp. Endocrinol., Suppl. 3,411. Chart, II K.O . ,Chtsici Junes, I., nhd Smith, R N.(IW).Lien. Comp. Endocrind. 11.243. Chen, C. J., Anast, C.S . , and Brown, E. M.(1988). J. Bone Miner. Res. 3,279. Chertow. B. S.. Baylink, D. J., Wergedal, J. E., Su, M. H. H., and Norman, A. W. (1975). J . Clin. Invest. 56,668. Cholst, I. N., Stemberg, S. F., Tropper, P. J., Fox, H. E., Segie, G. V., and Bilezikian, J. P. (1984). N. Engl. 1. a d . 310, 1221. Chu, L. L. H., MacGregor, R. R., Anast, C. S., Hamilton, J. W., and Cohn, D. V. (1973). Endocrinology (Baltimore)93,915. Clark, N. B., Kaul, K., and Roth, S. I. (1986).I n “Vertebrate Endocrinology: Fundamentals and Biomedical Implications” (P. K. T. Pang and M. P. Schreibman, eds.), p. 207. Academic Press, Orlando, Florida. Cohen, R. S., Pang, P. K. T.. and Clark, N. R. (1975). Gen. Comp. Endocrinol. 27,413. Cohn, D. V., and Eking, J. (1983). Recent Prog. Horm. Res. 39, 181. Cohn, D. V., and MacGregor, R. R. (1981). Endocr. Reu. 2, 1.

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

205

Cohn, D. V., Kumarasamy, R., andRamp, W. K. (1986). Vitam. Horm. ( N . Y.)43,283. Colston, K. W., King, R. J. B., Hayward, J., Fraser, D. I., Horton, M. A., Stevenson, J. C., and Arnett, T. R. (1989). J. Bone Miner. Res. 4,625. Cooper, C. W., and Deftos, L. J. (1970). Fed. Proc., Fed. Am. SOC.Exp. Biol. 29,253. Cooper,C. W., Deftos, L. J., andPotts, J. T., Jr. (1971a).Endocrinology(Baltimore)88,747. Cooper, C. W., Schwesinger, W. H., Mahgoub, A. M., and Ontjes, D. A. (1971b). Science 172, 1238. Cooper, C. W., Schwesinger, W. H., Ontjes, D. A., Mahgoub, A. H., and Munson, P. L. (1972). Endocrinology (Baltimore) 91, 1079. Cooper, C. W., McGuigan, J. E., Schwesinger, W. H., Brubaker, R. L., and Munson, P. L. (1974). Endocrinology (Baltimore) 95,302. Cooper, C. W., Ramp, W. K., Becker, D. I., and Ontjes, D. A. (1977a). Endocrinology (Baltimore)101,304. Cooper, C. W., Obie, J. F., Toverud, S. E., and Munson, P. L. (1977b). Endocrinology (Baltimore) 101, 1655. Copp, D. H.,and Davidson, A. G. F. (l%l). Proc. SOC.Exp. Biol. Med. 107,342. Copp, D. H., and Kline, L. (1989). I n “Vertebrate Endocrinology: Fundamentals and Biomedical Implications” (P. K. T. Pang and M. P. Schreibman, eds.), Vol. 3, p. 79. Academic Press, San Diego, California. Copp, D. H., Cameron, E. C., Cheney, B., Davidson, A. G. F., and Henze, K. (1%2). Endocrinology (Battimore) 70,638. Copp, D. H., Brooks, C. E., Low, B. S., Newsome, F., O’Dor, R. K., Parkes,C. O., Walker, V., and Watts, E. G. (1970). I n “Calcitonin” (S. Taylor, ed.), p. 281. Heinemann Medical Books, London. Cressent, M., Bonizar, Z., Moukhtar, M. S., and Milhaud, G. (1981). Proc. SOC.Exp. Biol. Med. 166,92. Cressent, M.,Elie, C., Taboulet, J., Moukhtar, M. S., and Milhaud, G. (1983). Proc. SOC. Exp. Biol. Med. 172, 158. Dacke, C. G. (1979). “Calcium Regulation in Sub-mammalian Vertebrates.” Academic Press, New York and London. Dean, W. L., Adunyah, S. E., and Cohn, D. V. (1986). Bone Miner. 1,59. De Bustros, A., Baylin, S. B., Berger, C. L., Roos, B. A., Leong, S. S., and Nelkin, B. D. (1985). J. Biol. Chem. 260,98. De Bustros, A., Baylin, S. B., Levine, M. A., and Nelkin, B. D. (1986). J. Biol. Chem. 261, 8036. Dietel, M.,Dorn, R., Montz, R., and Altenahr, E. (1979). Endocrinology (Baltimore) 105, 237. Donahue,H. J.,Fryer,M. J.,andHeath,H., lII(1990). Endocrinology(Baltimore)126,1471. Duarte, B., Hargis, G. K., and Kukreja, S. C. (1988). J. Clin. Endocrinol. Metab. 66,584. Dufresne, L. R., and Gitelman, H: J. (1972). I n “Calcium, Parathyroid Hormone and the Calcitonins” (R.V. Talmage and P.L. Munson, eds.), p 202. Excerpta Medica, Amsterdam. Epstein, P. A., Prentki, M., and Attie, F. (1985). FEES Lett. 188, 141. Ericson, L. E., and Sundler, F. (1984). In “Ultrastructure of Endocrine Cells and Tissues” (P. M. Motta, ed.), p. 276. Martinus Nijhoff, Boston, Massachusetts. Eriksen, E. F., Colvard, D. S., Berg, N. J., Graham, M. L., Mann, K. G., Spelsberg, T. C., and Riggs, B. L. (1988). Science 241,84. Farquhar, M. G. (1976). In “Comparative Endocrinology of Prolactin” (H. D. Dellmann, J. A. Johnson, and D. M. Klachko, eds.), p. 37. Plenum, New York. Feinblatt, J. D., and Raisz, L. G. (1971). Endocrinology (Baltimore) 88,797.

206

S. E. WENDELAAR BONGA AND P. K. T. PANG

Feinhlatt, J. B.,Tai. L.-R., and Kcnny, A. D. (1975). E n h ~ r i n u l u g y(Bulrlmcrrr) 96,282. Fenwick, J. C. (1974). Gen. Comp. Endocrinol. 23, 127. Fenwick, J. C. (1978). In “Comparative Endocrinology” (P. Y. Gaillard and H. H. Boer, eds.), p. 255. Elsevier, Amsterdam. Fenwick, J. C., and Lam, T. J. (1988). Gen. Comp. Endocrinol. 70,224. Ferment, O . ,Gamier, P. E., and Touitou, Y. (1987). J . Endocrinol. 1l3, 117. Fiore, C. E., Dagata, R., Clementie, G., and Malatino, L. S. (1984).J. Endocrinol. Invest. 7, 647. Fischer, J. A., Binswanger, U., Fanconi, A., Illig, R., Baerlocher, K., and Prader, A. (1973). Horm. Metab. Res. 5,381. Fischer, J. A., Blum, J. W., Born, W., Dambacher, M. A., and Dempster, D. W. (1982). Calc$. Tissue I n t . 34,313. Fitzpatrick, L. A., and Leong, D. A. (1990). Endocrinology (Baltimore)126, 1720. Fitzpatrick, L. A., Brandi, M. L., and Aurbach, G. D. (1986a). Biochern. Biophys. Res. Commun. l38,960. Fitzpatrick, L. A., Brandi, M. L., and Aurbach, G. D. (1986b). Endocrinology (Baltimore) 118,2115. Fitzpatrick, L. A., Brandi, M. L., and Aurbach, G. D. (1986~).Endocrinology (Baltimore) 119,2700. Fitzpatrick, L. A., Chin, H., Nirenberg. M.,and Aurbach, G. D. (1988).Proc. Natl. Acad. Sci. U.S.A. 85,2115. Flik, G., Wendelaar Bonga, S. E., and Fenwick, J. C. (1984). Comp. Biochem. Physiol. B . BB, 521. Flik, G.. Fenwick, J. C., and Wendelaar Bonga, S . E. (1989a).A m . J . Physiol. 257, R74. Flik, G., Labedz, T., Lafeber, F. P. J. G., Wendelaar Bonga, S. E., and Pang, P. K. T. (1989b).Fish Physiol. Biochem. 7,343. Flueck, J. A., DiBella, F. P., Edis, A. J., Kehnvald, J. M.,and Arnand, C.D . (1977).J. Clin. Inuest. 60, 1367. Fouchereau-Peron, hi., Arlot-Bounemains, Y., Maubras, L., Milhaud, G., and Moukhtar, M.S. (1990). Gen. Comp. Endocrinol. 78, 159. Fraser, D. R., and Kodicek, E. (1973). Nature (London), New Biol. 241, 163. Freake, H. C., and Maclntyre, I. (1982).Biochem. J . u)6, 181. Fried, R. M.,and Tashyan, A. H. (1986).J. Biol. Chem. 261,7669. Fucik, R. F., Kukreja, S. C., Hargis, G. K., Bowser, E. N., Henderson, W. J., and Williams, G. A. (1975).J. Clin. Endocrinol. Metab. 40, 152. Fumkawa, KA., Tawada, Y., and Shigekawa, M. (1989). J. Biol. Chem. 264,4844. Gallagher, J. C., Riggs. B. L., and DeLuca, H. F. (1980). J . Clin. Endocrinol. Metab. 51, 1359. Garel, J.-M., and Besnard, P. (1980). Biomedicine 33, 124. Garel, J. -M., and Jultienne, A. J. (1977). Endocrinology (Baltimore)75, 373. Gittes, R. F., Toverud, S. U.. and Cooper, C. W. (1968). Endocrinology (Baltimore) 82, 83. Glowacki, J., O’Sullivan, J.. Miller, M.,Wilkie, D. W., and Deftos, L. J. (1985).Endocrinology (Baltimore)116,827. Golden, P., Greenwalt, A., Martin, K., Bellorin-Font, E., Mazey, R., Klahr. S., and Slatopolsky, E. (1980). Endocrinology (Baltimore)107,602. Gmu, E. G., and Helms, L. M. H. (1989). Fish Biochem. Physiol. 7, 1 1 . Grau, E. G., and Helms, L. M. H. (1990). In “Progress in Comparative Endocrinology” (A. Epple, C. G. Seames, and M. H. Stetson, eds.), p. 534, Wiley-Liss, New York. Gray, T.K., and Munson, P.L . (1%9). Science 166,512.

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

207

Greenberg, C., Kukreja, S. C., Bowser, E. N., Hargis, G. K., Henderson, W. J., and Williams, G. A. (1987). FEES Lett. 36,151. Habener, J. F., and Potts, J. T., Jr. (1976). Endocrinology (Baltimore) 98, 197. Habener, J. F., Kemper, B., and Potts, J. T., Jr. (1975). Endocrinology (Baltimore) 97,431. Habener, J. F., Stevens, T. D., Ravazzola, M., Orci, L., and Potts, J. T., Jr. (1977). Endocrinology (Balrimore) 101, 1524. Habener, J. F., Rosenblatt, M., and Potts, J. T., Jr. (1984). Physiol. Reu. 64,985-1053. Haller-Brem, S . , Muff, R., Petermann, J. B., Born, W., Roos, B. A., and Fischer, J. A. (1987). Endocrinology (Baltimore) 121, 1272. Haller-Brem, S., Muff, R., and Fischer, J. A. (1988). J . Endocrinol. 119, 147. Hanke, W., Bergerhoff, K., and Chan, D. K. 0. (1967). Gen. Comp. Endocrinof. 9,64. Hanley, D. A., and Ayer, L. M. (1986). J . Clin. Endocrinol. Metab. 63, 1075. Hanssen, R. G. J. M., Lafeber, F. P. J. G., Flik, G., and Wendelaar Bonga, S. E. (1989). J . Exp. Biol. 141, 177. Hanssen, R. G. J. M., Aarden, E. M., Van der Venne, W. P. H. G., Pang, P. K. T., and Wendelaar Bonga, S. E. (1991). Gen. Comp. Endocrinol. (in press). Hargis, G. K., Williams, G. A., and Reynolds, W. A. (1978). Endocrinology (Baltimore) 102, 745. Hawkins, D., Enyedi, P., and Brown, E. (1989). Endocrinology (Baltimore) W, 834. Heath, H. (1980). Endocr. Rev. 1,319. Heath, H., and Sizemore, G. W. (1977). J . Clin. Inuest. 60, 1135. Heath, H., Larson, J. M.,and Laakso, K. (1980). Endocrinology (Baltimore) 107,977. Heinrich, G., Kronenberg, H. M., Potts, T. J., Jr., and Habener, J. F. (1983). Endocrinology (Baltimore) 112,449. Henry, H. L., and Norman, A. W. (1975). Biochem. Biophys. Res. Commun. 62,781. Henry, H. L., Taylor, A. N., and Norman, A. W. (1977). J . Nurr. 107, 1918. Hillyard, C. J., Stevenson, J. C., and MacIntyre, I. (1978). Lancer 1,961. Hirano, T. (1986). In “Comparative Endocrinology: Development and Directions” (C. L. Ralph, ed.), p. 53. Liss, New York. Hirano, T. (1989). In “Vertebrate Endocrinology: Fundamentals and Biomedical Implications’’ (P. K. T. Pang and M. P. Schreibman, eds.), Vol. 3, p. 139. Academic Press, San Diego, California. Hirano, T., Hasegawa, S., Yamauchi, H., and Orimo, H. (1981). Gen. Comp. Endocrinol. 43,42. Hishikawa, R., Fukase, M., Takenaka, M., and Fujita, T. (1985a). Biochem. Biophys. Res. Commun. WO, 454. Hishikawa, R., Fukase, M.,Yamatani, T., Kadowaki, S., and Fujita, T. (198Sb). Biochem. Biophys. Res. Commun. l32,424. Hove, K., and Sand, 0. (1981). Acta Physiol. Scand. ll3,87. Hruska, K. A.. Martin, K., Mennes, P., Greenwalt, A., Anderson, C., Klahr, S., and Slatopolsky, E. (1977). 1.Clin. Invest. 60,501. Hughes, M. R., and Haussler, M. R. (1978). J . Biol. Chem. 253, 1065. Hurst, J. G., and Mayer, G. P. (1977). I n “Vitamin D: Biochemical, Chemical and Clinical Aspects Related to Calcium Metabolism” (A. W. Norman, ed.), p. 139. de Gruyter, New York. Hurwitz, S. (1989). In “Vertebrate Endocrinology: Fundamentals and Biomedical Implications (P. K. T. Pang and M. P. Schreibman, eds.), Vol. 3, p. 45. Academic Press, San Diego, California. Jacobs, J. W., Simpson, E., Penschow, J., Hudson, P., CoghIan, J., and Niall, H. (1983). Endocrinology (Baltimore) 1l3, 1616.

208

S. E. WENDELAAR BONGA AND P. K. T. PANG

Jia, M.. Ehrenstein, G., and Iwasa, K. (1988). Proc. Natl. Acad. Sci. U.S.A. 85,7236. Jones, J. I., and Fitzpatrick, L. A. (1990). Endocrinology (Baltimore) W, 2015. Jousset, V . , Besnard, P., Segond. P., Jullienne, A., and Garel, J.-M. (1988). Mol. Cell. Endocrinol. 59, 165. Jowsey, J . , Riggs, B. L., Kelly, P. J., and Hoffman, D. L. (1978). J . Clin. Endocrinol. Metab. 47,633. JuhIin, C . , Johansson, H., Holmdahi, R., Gylfe, E., Larsson. R., Rastad, J., Akerstrom, G., and Klareskog, L. (1987). Biochem. Biophys. Res. Commun. 143,570. Kapoor, A. S., and Chhabra, C. (1981). Gen. Comp. Endocrinol. 44,307. Keaton, J. A., Barto, J. A., Moore. M. P., Gruel, J. B., and Mayer, G. P. (1978).Endocrinology (Baltimore) 103,2161. Kemper, B. (1986). CRC Crir. Reu. Biochem. 19,353. Kemper, B., Habener, J. F., Rich, A., and Potts, J.T., Jr. (1974). Science 184, 167. Khosla. S., Demay, M., Pines, M., Hurwitz, S., Potts. J. T., Jr., and Kronenberg, H. (1988). J . Bone Miner. Res. 3,689. Kobayashi, N., Russell, J., Lettieri, D., and Sherwood, L. M. (1988).Proc. Natl. Acad. Sci. U.S.A. 85,4857. Krishnamurthy, V. G., and Bern, H. A. (1971). Gen. Comp. Endocrinol. 16, 162. Kukreja, S. C., Hargis, G. K.,Bowser, E. N., Henderson, W. J., Fisherman, E. W., and Williams, G. A. (1975). J. Clin. Endocrinol. Metab. 40,478. Lafeber, F. P. J. G., and Perry, S. F. (1988). Gen. Comp. Endocrinol. 72,136. Lafeber, F. P. J. G., Hanssen, R.G. J. M., Choy, Y.M., Fhk, G., Hermann-Erlee, M. P. M., Pang, P. K. T., and Wendelaar Bonga, S. E. (1988a). Gen. Comp. Endocrinol. 69, 19. Lafeber, F. P. J. G., Hanssen, R. G. J. M., and Wendelaar Bonga, S. E. (1988b). J . Exp. Bid. 140,199. Lafebet, F. P. J. G., Flik, G., Wendelaar Bonga, S. E., and Perry, S. F. (1988~).Am. J. Physiol. 254, R891. Lancer, S. R., Bowser, E. N., Hargis, G. K., and Williams, G. A. (1975). Endocrinology (Baltimore) 98, 1289. Larsson, R., Wallfelt, C . , Abrahamsson, H., Gylfe, E., Ljunghall, S., Rastad, J., Rorsman, P., Wide, L., and AkerstrGm, G. (1984). Biosci. Rep. 4,909. Latman, N. S. (1980). J. Exp. Zool. 212,313. LeBoff, M.S., Rennke, H.G., and Brown, E. M. (1983). Endocrinology (Baltimore)1l3,277. Legendre, B., Besnard, P., Tahri, E. H., Tahraoui, A., Segond, N., Jullienne, A., and Garel, J.-M. (1989). J . Physiol. (Paris) 83,74. Lewis, P., Raf€erty, B., Shelley, M., and Robinson, C. J. (1971). J. Endocrinol. 49, IX. Llach, F., Coburn, J. W., Brickman, A. S., Kurokawa, K., Norman, A. W., Canterbury, J. M., and Reiss, E. (1977). J . Clin. Endocrinol. Metab.. 44, 1054. Lopez, E., Deville, J., and Bagot, E. (1968). C . R. Hebd. Seances Acad. Sci. 267, 1531. Lopez, E., Peignoux-Deville,J., Lallier, F., Martelly, E., and Milet, C. (1976).Calcif. Tissue Res. 20, 173. Lopez, E., Tisserand-Jochem, E. M., Eyquem, A.. Milet, C., Hillyard, C., Lallier, F.,Vidal, B., and Maclntyre, I. (1984). Gen. Comp. Endocrinol. 53,28. Lopez-Barneo, J., and Armstrong, C. M. (1983).J . Gen. Physiol. 82,269. Ma, S . W. Y., and Copp, D. H. (1978). In "Comparative Endocrinology" (P. J. Gaillard and H. H. Boer, eds.), p. 283. Elsevier, Amsterdam. MacDonald, D. J., and McKeown, B. A. (1983). Can. J. Zool. 61,682. Mahaffee, D. D., Cooper, C. W.,Ramp, W.K.,and Ontjes, D. A. (1982). Endocrinology (Baltimore) 110,487. March, G . L., and McKeown, B. A. (1977). Endocrinol. Exp. 11,263.

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

209

Matsuda, K., Oguro, C., Sasayama, Y., and Kikuyama, S. (1991). Gen. Comp. Endocrinol. (in press). Mayer, G . P., Keaton, J. A., Hurst, J. G. and Habener, J. F. (1979). Endocrinology (Baltimore) 104, 1778. McGhee, J. G., and Shoback, D. M. (1990). Endocrinology (Baltimore)u6,899. McGuire, A., Cohen, S., and Brooks, F. B. (1972). Clin. Res. 20,734. Membreiio, L.,Chen, T.-H., Woodley, S., Gagucas, R., and Shoback, D. (1989). Endocrinology (Baltimore) W, 789. Milet, C., Peignoux-Deville, J., and Martelly, E. (1979). Comp. Biochem. Physiol. A . 63A, 63. Milet, C., Hillyard, C. J., Martelly, E., Girgis, G., MacIntyre, I., and Lopez, E. (1980). C.R. Hebd. Seances Acad. Sci. 291,977. Milhaud, G., Perault-Staub, A.-M., and Staub, J.-F. (1972). J . Physiol. (London)222,559. Milhaud, G., Benezeck-Lefevre, M., and Moukhtar, M. S. (1978). Biomedicine 29,272. Miller, S. S., Sizemore, G. W., Sheps, S. G., and Tyce, G. M. (1975).Ann. Intern. Med. 82, 372. Morimoto, S . , Tsuji., M., Okada, Y., Onishi, T., and Kumahara, Y. (1980). Clin. Endocrinol. W, 135. Morissey, J. J., and Cohn, D. V. (1979a). J. CellBiol. 82,93. Morissey, J. J., and Cohn, D. V. (1979b). J. Cell Biol. 83,521. Muff, R., Nemeth, E. F., Halter-Brem, S., and Fischer, J. A. (1988). Arch. Biochem. Biophys. 265,128. Munson, P. L. (1976). In “Handbook of Physiology” (G.D. Aurbach, ed.), Vol. 7, Sect. 7, p. 443. Am. Physiol. Soc.,Washington, D. C. Naveh-Many, T., and Silver, J. (1988). J. Clin. Invest. 81,270. Naveh-Many, T., Friedlaender, M. M., Mayer, H., and Silver, J. (1989). Endocrinolugy (Baltimore) 125, 275. Nemeth, E. F., and Scarpa, A. (1987). J. Biol. Chem. 262, 1518. Nemeth, E. F., Wallace, J., and Scarpa, A. (1986). J. Biol. Chem. 261,2668. Nicoll, C. S., Wilson, W.S.,Nishioka, R. S., and Bern, H. A. (1981). Gen. Comp. Endocrinol. 44,356. Nieto, A., Fando, J. J. L., and Candela, J. L. R. (1975). Gen. Comp. Endocrinol. 25, 259. Nishioka, R. S., Kelley, K. M.,and Bern, H. A. (1988). Zool. Sci. 5,267. Nishiyama, I., Yasumoto, T., and Fujii, T. (1990). Horm. Metab. Res. 22,258. Nuiiez, E . A., and Gershon, M. D. (1978). Int. Rev. Cytol. 52, 1. Oetting, M., Leboff, M. S.,Levy, S., Swiston, L., Preston, J., Chen, C., and Brown, E. M. (1987). Endocrinology (Baltimore) 121, 1571. Oguro, C. (1973). Gen. Comp. Endocrinol. 21,565. Oguro, C . , and Uchiyama, M. (1975). Gen. Comp. Endocrinol. 27,531. Oguro, C., Uchiyama, M., Pang, P. K. T., and Sasayama, Y. (1978). In “Comparative Endocrinology” (P. J. Gaillard and H. H. Boer, eds.), p. 269. ElsevierlNorth-Holland Biomedical Press, Amsterdam. Oguro, C., Fujimoro, M., and Sasayama, Y. (1984). 2001.Sci. 1,82. Okazaki, T., Igarashi, T.,and Kronenberg, H. M. (1988). J . Biol. Chem. 263,2203. Oldham, S . B., Fischer, J. A., Shen, L. H., and Arnaud, C. D. (1974).Biochemistry W, 4790. Oldham, S. B., Smith, R., Hartenbower, D. L., Henry, H. L., Norman, A. W.,and Coburn, J. W. (1979). Endocrinology (Baltimore) 104,248. Olivereau, M., and Olivereau, J. (1968). Z . Zellforsch. Mikrosk. Anat. 84,44. Olivereau, M., and Olivereau, J. (1978). Cell Tissue Res. 186,81. Olivereau, M., and Olivereau, J. (1983). Cell Tissue Res. 229,243.

210

S. E. WENDELAAR BONGA AND P. K. T. PANG

Olivereau, M., Olivereau, J., and Aimar, C. (1982). Comp. Biochem. Physiol. A 71A, 1 1 . Orimo, H., Ohata, M., Fujita, T., Yoshikawa, M.. Higashi, T., Abe, J., Watanake, S., and Otani, K. (1972).In "Endocrinology 1971" (S. Taylor, ed.), p. 48. Heinemann, London. Pandey, A. C. (1988). Gen. Comp. Endocrinol. 69,467. Pang, P. K.T. (1971).Am. Zool. W, 775. Pang, P. K. T. (1981a). Gen. Comp. Endocrinol. 43,252. Pang, P. K.T. (1981b). Gen. Comp. Endocrinol. 44,524. Pang, P. K. T., Pang, R. K.,and Sawyer, W. H. (1973). Endocrinology (Baltimore) 93,705. Parsons, J. A., Gray, D., RatTert, B., and Zanelli, J. M. (1978).In "Endocrinology of Calcium Metabolism" (D.H.Copp and R.V. Talmage, eds.), p. 111. Excerpta Medica, Amsterdam. Parthemore, J. G., and Deftos, L. (1978).J. Clin. Endocrinol Metab. 47, 184. Patt, H. M., and Luckhardt, A. B. (1942). Endocrinology (Bafrimore)33,384. Peacock, M. (1980). Metab. Bone Dis. Relar. Res. 2, 143. Pearse, A. G. E. (1968). Proc. SOC. London, B 170,71. Peignoux-Deville, J., Lopez, E.. Lallier, F., Martelly-Bagot, E., and Milet, C. (1975). Cell Tissue Res. 164,73. Peng, T.-C., and Gamer, S. C. (1979). Endocrinology (Baltimore) 104, 1624. Peng, T.-C., Cooper, C. W., Gamer, S. C., and Volpert, E. M. (1978). 1.Pharmacol. Exp. Ther. 206.7 10. Perry, S. F., Seguin, D., Lafeber, F. P. 1. G., Wendelaar Bonga, S. E., and Fenwick, J. C. (1989).J . Exp. Biol. 147,249. Philippo, M., Bruce, J. B., and Lawrence, C. B. (1970).J. Endocrinol. 46, XII. pines, M., and Hurwitz, S. (1981). FEBS Lett. l33,27. Pitkin, R. M., Reynolds, W. A., Williams, G. A., and Hargis, G. K. (1978).J. Clin. Endocrino1 Metab. 47,626. Pletka, P., Bemstein, D. S., Hampers, C. L., Meml, J. L., and Shenvood, L. M. (1971). Lancet 2,462. Pocotte, S. L., and Ehrenstein. G. (1989). Endocrinology (Baltimore) 125, 1587. Ramp, W. K.,Cooper, C. W ., Ross, A. J.. and Wells, S. A. (1979).Mol. Cell. Endocrinol. 14, 205.

Raue, F., Deutschle, I., and Zeigler, R. (1983). Horm. Metab. Res.W, 208. Raymond, J. P., Isaac, R., Merceron, R. E., and Wahbe, F. (1982). J . Clin. Endocrinol. Metab. 55, 1222. aggs, B. L., Jowsey, J., Kelly, P. J., Jones, J. D., and Maker, G. T. (1%9). J. Clin. Inuest. 48,1065. a x , E., Raue, F., Deutschle, I., and Ziegler, R. (1984). Hisrochemistry 80, 503. Robertson, D.R. (1970). Endocrinology (Baltimore) 87,1041. Robertson, D .R. (1971a).J. Exp. Zool. 178, 101. Robertson, D. R. (1971b). Gen. Comp. Endocrinol. 16,329. Robertson, D. R. (1972). Gen. Comp. Endocrinol., Suppf.3,421. Robertson, D. R. (1977). Gen. Comp. Endocrinol. 33,336. Robertson, D. R. (1978). J . Endocrinol. 79, 167. Robertson, D. R. (1986). In "Vertebrate Endocrinology: Fundamentals and Biomedical Implications" (P.K.T. Pang and M.P. Schreibman, eds.), Vol. 1, p. 235. Academic Press, Orlando, Florida. Robinson, C. J., Spanos, E., James, M. F.,Pike, J. W.. Hanssler, M. R., Makeen, A. M., Hillyard, C. J., and Maclntyre, 1. (1982).J. Endocrinol. 94,443. Rodriguez. H. J., Momson, A., Slatopolsky, E., and Klahr, S. (1978). J. Clin. Endocrinol. Merab. 41,319.

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

21 1

Roos, B. A., Bundy, L. L., Bailey, R., and Deftos, L. J. (1974). Endocrinology (Baltimore) 95,1142.

Roos, B. A., Cooper, C. W., Frelinger. A. L., and Deftos, L. J. (1978). Endocrinology (Baltimore) 103,2180. Roth, S. I., Su, S. P., Segr6, G.V., Habener, J. F., and Potts, J. T., Jr. (1974). Fed. Proc., Fed. Am. SOC.Exp. Biol. 33,241. Rude, R. K., and Singer, F. R. (1977). J. Clin. Endocrinol. Metab. 44,980. Russell, J., Lettieri, D., and Sherwood, L. M. (1983). J. Clin. Invest. 72, 1851. Sasayama, Y.(1978). Gen. Comp. Endocrinol. 34,229. Sasayama, Y., and Oguro, C. (1976). Comp. Biochem. Physiol. A SA, 35. Schlechte, J. A., Sherman, B., and Martin, R. (1983). J. Clin. Endocrinol. Metab. 56, 1120. Segond, N., Jullienne, A., Lasmoles, F., Desplan, C., Milhaud, G., and Moukhtar, M. S. (1984). Eur. J . Biochem. l39,209. Segond, N., Legendre, B., Tahri, E. H., Besnard, P., Jullienne, A., Moukhtar, M. S., and Garel, J.-M. (1985). FEBS Lett. 184,268. Segr6, G. V., D’Amour, P., Hultman, A., and Potts, J. T., Jr.(1981). J. Clin. Invest. 67,439. Sethi, R., Kukreja, S. C., Bowser, E. N.,Hargis, G. K., and Williams, G. A. (1981). J. Clin. Endocrinol. Metab. 53, 153. Sethi, R., Kukreja, S. C., Bowser, E. N., Hargis, G. K., and Williams, G. A. (1983). Horm. Metab. Res. 15,362. Setoguti, T., Inoue, Y., and Kato, K. (1981). Cell Tissue Res. 219,457. Shah, J. H., Motto, G. S., Kukreja, S. C., Hargis, G. K., and Williams, G. A. (1975). J. Clin. Endocrinol. Metab. 41,692. Shapiro, H. A., and Zwarenstein, H. (1933). J. Exp. Biol. 10, 186. Sherwood, L. M., Potts, J. T., Jr., Care, A. D., Mayer, G. P., and Aurbach, G. D. (1966). Nature (London) u)9,52. Shoback, D. M.,Thatcher, J., Leombruno, R., and Brown, E. M. (1983). Endocrinology (Baltimore)1W,424. Shoback, D. M., Thatcher, J. G., and Brown, E. M. (1984a). Mol. Cell. Endocrinol. 38,179. Shoback, D., Thatcher, J. G., Leombruno, R., and Brown, E. M. (1984b). Proc. Natl. Acad. Sci. U.S.A. 81,3113. Silver, J., Russell, J., Lettieri, D., and Shenvood, L. M. (1985). Proc. Natl. Acad. Sci. U.S.A. 82,4270. Silver, J., Naveh, T., Mayer, H., Schmelzer, H., and Popovtzer, M. M . (1986). J. Clin. Invest. 78, 12%. Silvennan, R., and Yalow, R. S. (1973). J. Clin. Invest. 52, 1958. Slatopolsky, E., Weerts, C., Thielan, J., Horst, R., Harter, H., and Martin, K. J. (1984). J . Clin. Invest. 74,2136. So, Y. P., and Fenwick, J. C. (1979). Gen. Comp. Endocrinol. 377, 143. Srivastav, A. K., and Rani, L. (1988). Biol. Struct. Morphog. 1, 117. Srivastav, A. K., and Rani, L. (1989). Gen. Comp. Endocrinol. 74, 14. Srivastav, A. K., and Srivastav, S. P. (1988). 2001.Sci. 5, 197. Srivastav, S. P., Swamp, K., and Srivastav, A. K. (1985). Cell. Mol. Biol. 31, 1. Stevenson, J. C., Hillyard, C. J., MacIntyre, I., Cooper, H., and Whitehead, M. I. (1979). Lancet 2,769. Stevenson, J. C., Abeyasekera, G., Hillyard, C. J., Phang, K. G., and MacIntyre, I. (1981). Lancet 2,693. Stevenson, J. C., Abeyasekera, G., Hillyard, C. J., Phang, K. G., MacIntyre, I., Campbell, S., Lane, G., Townsend, P. T., Young, O., and Whitehead, M. I. (1983). Eur. J . Clin. Invest. W, 481.

212

S. E. WENDELAAR BONGA AND P. K. T. PANG

Subhedar, N., and Prasado Rao, P. D. (1979).Z. Mikrosk.-Anat. Forsch. 93,774. Suryawanshi. S. A., and Mahajan, S. M. (1976). Acta Biol. Acad. Sci. Hung. 27,269. Swamp, K . , and Srivastav, S. P. (t982). Gen. Comp. Endocrinol. 46,271. Takagi, Y . ,Nakamura, Y., and Yamada, J. (1985).Zool. Sci. 2,523. Takano-Yamamoto, T., and Rodan, G.A. (1989). J . Bone Miner. Res. 4, Suppl. I , 5235. Talmage, R. V., and Van der Wiel, C. J. (1979). CaIciJ Tissue Int. 28,113. Talmage, R. V., Doppelt, S. H., and Cooper,C. W. (1975).Proc. SOC.Exp. Biol. Med. 149, 855. Talmage, R. V., Van der Wiel, C. J., and Matthews, J. L. (1981). Mol. Cell. Endocrinol. 24, 235. Toverud, S. U.,and Munson, P. L. (1976). Abstr., 58th Annu. Meet., A m . Endocr. Soc., p. 104. Unsicker, K.. Polonius, T., Lindmar, R., Lbffelholz, K., and Wolf, U. (1977). Gen. Comp. Endocrinol. 31, 121. Urasa, F. M., and Wendelaar Bonga, S. E. (1985). Cell Tissue Res. 241,219. Urasa, F. M., and Wendelaar Bonga, S. E. (1987). Cell Tissue Res. 249,681. Wagner, G . F., Hampong, M.,Park, C. M., and Copp, D. H. (1986).Gen. Comp. Endocrinol. 63,481.

Wagner, G. F., Copp, D. H.,and Friesen, H. G. (1988).Endocrinology (Baltimore)UZ,2064. Wagner, G. F., Gellersen, B., and Friesen, H. G. (1989). Mol. Cell. Endocrinol. 62, 31. Wallace, J., and Scarpa, A. (1983).J. Biol. Chem. 258,6288. WaUfelt, C., Larsson, R., Johansson, H., Rastad, J., Akerstrom, G., Ljunghall, S., and Gytfe, E. (1985).Acta Physiol. Scand. W, 239. Watts, E. G . , Copp, D. H., and Deftos. L. J. (1975). Endocrinology (Baltimore)%, 214. Wecksler, W. R., Henry, H. L., and Norman, A. W. (1977).Arch. Biochem. Biophys. 183, 168. Wecksler, W. R., Ross, F. P., Mason, R. S., Posen, S., and Norman, A. W. (1980). Arch. Biochem. Eiophys. u)1,95. Wendelaar Bonga, S . E. (1978). Gen. Comp. Endocrinol. 34,265. Wendelaar Bonga, S. E. (1980). Gen. Comp. Endocrinol. 40,99. Wendelaar Bonga, S. E. (1981). Gen. Comp. Endocrinol. 43, 123. Wendelaar Bonga, S. E., and Flik, G. (1982). In “Comparative Endocrinology of Calcium Regulation” (C. Oguro and P. K.T. Pan, eds.), p. 19. Jpn. Sci. SOC.Press, Tokyo. Wendelaar Bonga, S. E., and Greven, J. A. A. (1978). Gen. Comp. Endocrinol. 36,W. Wendelaar Bonga, S. E., and Lammers, P. I. (1982). Gen. Comp. Endocrinol. 48,60. Wendelaar Bonga, S. E., and Pang, P. K. T. (1986). In “Vertebrate Endocrinology: Fundamentals and Biomedical Implications” (P. K.T. Pang and M. P. Schreibman, eds.), Vol. 1, p. 439. Academic Press, New York. Wendelaar Bonga. S. E., and Pang, P. K . T. (1989). I n “Vertebrate Endocrinology: Fundamentals and Biomedical Implications” (P. K. T. Pang and M. P. Schreibman, eds.), Vol. 3, p 207. Academic Press, New York. Wendelaar Bonga, S. E., Flik, G., Lowik, C. W., and Van Eys, G. J. (1985). Gen. Comp. Endocrinol. 57,352. Wendelaar Bonga, S. E., Balm. P. H.M., and Flik, G. (1988).Gen. Comp. Endocrinol. 72,l. Wild, P., Schraner, E. M.,and Santini-Withes. P. (1989). Experientiu 45, 1121. Williams, G. A., Hargis. G. K.,Bowser, E. N., Henderson, W. J., and Martinez, N. J. (1973). Endocrinology (Baltimore)92,687. Williams, G. A., Peterson, W. C., Bowser, E. N., Henderson, W. J., Hargis, G. K., and Martinez, N. J. (1974). Endocrinology (Baltimore)95,707.

CALCIUM REGULATING HORMONES IN THE VERTEBRATES

213

Williams, G.A., Hargis, G. K., Ensinck, J. W., Kukreja, S. C., Bowser, E. N., Chertow, B. S.,and Henderson, W. J. (1979). FEBS Lett. 28,950. Windeck, R., Brown, E., Gardner, D., and Aurbach, G. (1978). I n “Endocrinology of Calcium Metabolism” (D.H. Copp and R.V. Talmage, eds.), p. 364. Excerpta Medica, Amsterdam. Wisneski, L. A. (1990). Calcif. Tissue Znt. 46, (Suppl.), 526. Wittle, L. W.(1984). Gen. Comp. Endocrinol. 54, 181. Wyllie, A. H., Kerr, J. F. R., and Cunie, A. R. (1980). In?. Rev. Cytol. 68,251. Yamada, J., Tomioka, J., Yamane, S.,Iguchi, M.,and Nakamura, Y.(1982). I n “Comparative Endocrinology of Calcium Regulation” (C. Oguro and P. K. T. Pang, eds.), p. 143. Jpn. Sci. SOC.Press, Tokyo. Yamauchi, H., Orimo, H., Yamauchi, K., Takano, K., and Takahashi, H. (1978). Gen. Comp. Endocrinol. 36,526. Zabel, M., and Schiifer, H. (1988). Histochemistry 88,623. Zaidi, M., Breimer, L. H., and MacIntyre, I. (1987). J. Exp. Physiol. 72,371. Ziegler, R.,Telib, M.,and Pfeiffer, E. F. (1969). Horm. Metub. Res. 1,39.

This Page Intentionally Left Blank

INTERNATIONAL REVIEW OF CYTOLOGY. VOL. 128

Immunoarchitecture of Regenerated Splenic and Lymph Node Transplants R. PABST,J. WESTERMANN, AND H.J. ROTHKOITER Center of Anatomy, Medical School of Hannouer, 0-3000 Hannover 61, Germany

I. Introduction A. GENERALASPECTS OF REGENERATION

The term regeneration is normally used to describe the natural renewal of a damaged tissue, e.g., the regeneration of the skin after a cutaneous wound. These mechanisms have been studied in detail. Hepatic regeneration after partial hepatectomy is another example of a well-regulated phenomenon where the original size and function of the organ are reestablished. Regeneration of splenic or lymph node tissue is somewhat different, as small segments of these organs are purposely placed at an unusual site where they regenerate by rebuilding the organ compartments. The term transplantation will often be used in this review, but the type of transplantation used for the spleen can hardly be compared to the wellknown transplantation of the kidney, liver, heart, or lung. The two main differences are: (1) that transplantation of lymphoid organs is always autotransplantation, i.e., the individual is given all or parts of its own organ and (2) that no connection to the vascular system is reestablished, i.e., the autotransplants are avascular. Therefore, the mechanisms of regeneration and its regulation are also different and can be considered of special interest not only as a general biological phenomenon but also for the clinical situation.

B. CHANGING INTERESTS IN THE FUNCTION AND CLINICAL RELEVANCE OF THE SPLEEN Interest in the function of the spleen reaches far back to ancient Jewish writings, and over the centuries a large variety of functions has been attributed to it. This historical background was elegantly summarized by Sherman in 1980. For a long time the importance of the spleen was at least questionable and as early as 1578 Balloni had addressed the basic issue by asking “Este igitur splenatam necessarius?”, i.e., “Is the spleen really essential for life?”. Because of this long-standing attitude enormous numbers of healthy spleens have been removed after traumatic laceration of 215 Copyright 0 1991 by Academic Press. Inc. All rights of reproductionin any form reserved.

R. PABST et al.

216

the splenic capsule following accidents. Also, comparable numbers of incidental and accidental splenectomies have been performed, e.g., the normal spleen was removed to facilitate operation on organs nearby or when splenic trauma occurred during an operation. This attitude underwent a complete change over the last 15 years after the increased rate of fatal overwhelming postsplenectomy infections (OPSI) was documented and generally accepted in children (e.g., Ein et al., 1977; Eraklis and Filler, 1972; Singer, 1973; Walker, 1976) and with some reservations in adults (Green et al., 1986; Malangoni et al., 1984; Sekikawa and Shatney, 1983; West and Grosfeid, 1985; Zarrabi and Rosner, 1984). It is obvious from Table I that morbidity and lethality are correlated to the indication for splenectomy in children. It is important to note that about 40% of children with OPSI after splenectomy for trauma die despite intensive care. All these data point to the important role of the spleen in resisting certain infections (Bohnsack and Brown, 1986). The changing concept of the function of the spleen has resulted in belief that the casual attitude of some surgeons to incidental splenectomy during other operations is no longer justified (Fellows et al., 1988). There are now several reviews available on the updated indications for splenectomy (Cooper and Williamson, 1984; Leonard er al., 1980; Perry, 1988; Seufert, 1986; Sherman, 1980) and on alternatives to splenectomy , which mainly involve salvage of the organ in situ (Buntain and Gould, 1985; Witte et al., 1983). Autotransplantation of splenic tissue gained clinical interest and research was stimulated after Pearson et al. (1978) documented that a certain degree of splenic function returned in children with splenosis after splenectomy for trauma. Thus, the spleen is no longer taken to be a superfluous organ. After trauma it should be salvaged in situ or, failing that, autotransplantation of splenic fragments can be performed as an alternative to provide some splenic function. More than 400 splenic autotransplantations have already been performed and described in the literature. However, neither the

INCIDENCE

OF

TABLE 1 POSTSPLENECTOMY SEPSISIN CHILDREN'

Indication for splenectomy

Morbidity

Mortality

Lethality

Trauma(n = 1042) Hemolytic anemia (n = 390) Ideopathic thrombopenic purpura (n = 235) Thalassemia, M. Hodgkm, etc. (n = 215) All indications ( n = 1882)

1.6%

2.8% 4.7% 9.8%

0.7% 1.5% 2.6% 3.7% I .4%

41% 55% 55% 38% 45%

3.2%

' Data p l e d from Ein (1977). Pedersen (1983). Wahlby and DomeUof (1981). and Walker (1976).

REGENERATED SPLENIC AND LYMPH NODE TRANSPLANTS

217

structure nor the function of the splenic transplants has been documented satisfactorily, largely due to the fact that biopsies or other invasive techniques on human splenic transplants can hardly be morally justified. Therefore, data from animal models are essential in understanding the basic mechanisms of splenic regeneration. 11. Functional Anatomy of the Spleen

A. BASICSTRUCTURE The spleen is not only the largest single lymphoid organ in man, but it also has an important clearance function for particles and antigens from the blood. The spleen contributes to hematopoiesis in fetal life in man and throughout life in many species. This capacity obviously exists lifelong in man, for in certain pathological situations such as osteomyelofibrosis(Falk et al., 1990), splenic hematopoiesis is reestablished. However, little is known about the microenvironmentalfactors which regulate the seeding of stem cells in the splenic stroma. In studies on splenic regeneration the immunological and clearance function of the spleen have mainly been investigated, rather than hematopoiesis. Here, only the histological elements of these functions will be described in more detail. The topographical and macroscopical anatomy of the spleen is generally known and the basic histological pattern has often been described (van Krieken and te Velde, 1988, 1991; Severin, 1989; Timens and Poppema, 1985; Weiss, 1988). In many aspects of splenic function and regeneration it has to be taken into account that accessory spleens are found in 10-20% of patients, which might obscure the effects of splenectomy (Rudowski, 1985). For surgery of the spleen after splenic trauma partial resection of splenic tissue has been advocated under the assumption that the remaining part will show compensatory growth. It is important to mention that the human spleen has a clear-cut segmentation, as documented by corrosion cast studies of the splenic arteries and veins (CortCs and Pellico, 1988; Dawson et al., 1986; Garcia-Porrero and Lemes, 1988; Redmond et al., 1989). There are major species differences in the amount of musculature in the capsule and trabeculae of the spleen. For example, in dogs there is a huge storage capacity in spleens, for -25% of all erythrocytes (Hartwig and Hartwig, 1985). In rodent and human spleens there is no comparable storage function for red cells, but there is an important storage function for platelets (Kotze et al., 1986). The histological structure of the spleen can either be described as a reflection of the three-dimensional network of the connective tissue (Faller, 1985) or more conventionally in its relationship to the vascular tree (Fig. 1). Recent publications should be referred to for

218

R. PABST et al.

FIG.1 . Schematic drawing of the compartments and the vascular supply of the spleen. The majority of the arterioles pass the white pulp and terminate in the marginal zone.

information on the phylogeny of the spleen as well as on its ontogeny (Timens er al., 1987, 1988; Tischendorf, 1985; Vellguth et al., 1985). Having entered the splenic pulp, the arteries and veins branch and run parallel in trabeculae. Thereafter, the arterial branches are no longer accompanied by veins. The reticular fibers in the pulp surrounding the arteries are partly parallel to the vessel and also connected to the reticulum of the marginal zone and red pulp. In contrast to many other organs in the spleen, arterioles branch at right angles and pass through lymphoid follicles terminating partly in the marginal zone and partly in the red pulp in a typical splenic structure, often called by different names, e.g., ellipsoids, periarteriolar macrophage sheaths, or sheaths capillaries (Buyssens et al., 1984; Weiss et al., 1985). There are major species differences in these structures. The functional significanceof the macrophages which embrace the terminating arteriole is not clear. There has been a long-standing debate on whether the spleen consists of a “closed circulation”, that is a continuous endothelid lining from the arterioles to the venules, or an “open circulation”, that means termination of the arterial vessels in the cords of the red pulp. Weiss (1988) stressed that anatomical and functional data are partially contradictory. Based on light and electron microscopic

REGENERATED SPLENIC AND LYMPH NODE TRANSPLANTS

219

evaluations of serial sections, there is ample evidence of an open circulation. Physiological studies have shown major species differences and the blood flow has been compartmentalized,with a major part in rapid flow but also having slow components. Studies using scanning electron microscopy on corrosion casts of human spleens (e.g. Fujita et al., 1982; Kashimura, 1985; Schmidt et al., 1988) revealed evidence of both the open and closed circulation, though more for the former. On route from the splenic cords into the sinus the cells have to squeeze through slits. Recent freeze-fracture investigations of the red pulp in the human spleen occasionally revealed desmosomes and maculae adherents as connections to sinus endothelial cells (Kaiserling et al., 1989). Using scanning electron microscopy, Fujita et al. (1982) convincingly documented that aging red cells which have lost plasticity, and pathological erythrocytes, as in hereditary spherocytosis, can no longer pass through these slits and are phagocytozed by macrophages. In addition, intracellular parasites like malaria or nuclear remnants are removed from erythrocytes. This is the so-called “pitting function” of the red pulp (Fig. 2), which can be used as an indicator of the presence and function of splenic tissue. The innervation of lymphoid organs in general and of the spleen in particular has long been neglected. Using classical histological and histochemical techniques the presence of nerves has recently been documented in the human spleen (Heusermann, 1988). Large numbers of nerves enter the splenic parenchyma with the vessels in the hilar area and follow the route of the arteries. These fibers seem to be important in regulating the blood flow and contraction of muscle cells. Other recent studies have provided evidence of an association of sympathetic postganglionic nerve fibers with lymphocytes and macrophages in the rat spleen (Felten, D. L. et al., 1987; Felten, S. Y.et al., 1988), which might be of major importance in regulating immune responses. Further evidence of this role has come from studies on the parallel postnatal development of the splenic compartments and the appearance of noradrenergic innervation (Ackerman et al., 1989).

B. CELLULAR COMPOSITION OF SPLENIC COMPARTMENTS The spleen can be compartmentalized into the red pulp, the marginal zone, and the white pulp, consisting of two further compartments: the periarteriolar lymphatic sheath (PALS)and the follicles with the surrounding corona (Fig. 3). All these compartments are composed of a typical pattern of cells (Eikelenboom et al., 1985; van Ewijk and Nieuwenhuis, 1985). Macrophages in the red pulp, marginal zone, and white pulp express

220

R. PABST et al.

FIG. 2. Schematic drawing of the red pulp, indicating the passage of red blood cells from the pulp cords into the sinus under normal and pathologic conditions.

different surface antigens in rats (Dijkstra et al., 1985a). In the marginal zone there is also another typical cell type, the marginal metallophiis, so named because they are blackened by silver impregnation techniques along the inner border of the marginal sinus. These cells are weakly phagocytic and show strong nonspecific esterase activity in mice (see Eikelenboom et al., 1985). A recently proposed function of these cells is to form a microenvironment with other macrophages for initiating plasmacellular reactions after trapping antigen-specificB cells, which then differentiate into antibody-formingcells (Matsunoet al., 1989).The preferential localization of lymphoid cell subsets in the different splenic compartments has not only been documented by demonstrating the effects of postnatal

FIG. 3. Serial cryostat sections of a normal rat spleen showing the distribution of the different cell populations within the various compartments. (a) B lymphocytes (His 14+)are found mainly in the follicles and the marginal zone. (b) T lymphocytes (Ox 19+)predominate in the PALS (c) Macrophages (ED29 are localized in the red pulp. The areas (d) and (e) are stained on consecutive sections for follicular dendritic cells, FDC (Ki-M4RT), and interdigitating dendritic cells, IDC (EDl+), respectively. The arrows indicate an FDC (d) and an IDC (e). F: follicle; MZ: marginal zone; PALS: penarteriolar lymphatic sheath; RP: red pulp. (Immunohistologicalstaining revealed by the alkaline phosphatase anti-alkaline phosphatase reaction). Magnification (a), (b), and (c) x 63; (d) and (e) x 378.

222

R. PABST et al.

thymectomy but also largely by using monoclonal antibodies against characteristic membrane antigens. While rat follicular lymphocytes mainly express 1 and S on their surface, marginal zone B cells are predominantly p + K and classified as sessile cells (Gray et al., 1982). Experiments on human spleens indicate that it is an oversimplification to state that the marginal zone contains nonrecirculating IgM+ I@- cells only (van Krieken and te Velde, 1986,1989; Timens and Poppema, 1985). T lymphocytes are mainly seen in the PALS with T helper lymphocytes outweighing T cytotoxic/suppressor lymphocytes which are also found dispersed in the red pulp. In a recent study on human spleens, memory cells (CD45RO+) by far outweighed virgin lymphocytes (CD45RA+). The memory lymphocytes were seen mainly in the PALS (Brede et at., 1991). The spleen seems to contain the largest number of NK cells of all lymphoid organs and the blood. CD56+ NK cells make up about 25% of all lymphoid cells in the human spleen. The majority of NK cells are localized in the red pulp and have a different phenotype from that of the small number of NK cells in the white pulp. These differences in surface markers could also be seen in functional tests, e.g., cytotoxicity or stimulation of isolated splenic B cells to produce immunoglobulins (Witte et al., 1989). A further cell family in the white pulp are the dendritic cells, so named because of their long cellular processes. Interdigitating dendritic cells (IDC) are localized in the PALS and in T cell areas of lymph nodes (Dijkstra and Dopp, 1983; van der Valk et al., 1984). The so-called follicular dendritic cells (FDC) are typically found in germinal centers. FDC are characterized by their ability to bind antigen-antibody complexes for long periods. The origin of these cells seems to be largely species dependent (see Eikelenboom et al., 1985). There is no doubt that the FDC play an essential role in the regulation of the immune response (reviewed by Klaus et af.,1980; Szakal et al. 1989). The preferential localization of the different lymphoid and nonlymphoid cells in the splenic compartments is summarized for the human spleen in Fig. 4 (Grogan et al., 1983, 1984; Tanaka et al., 1984a, b). In many studies on splenic functions or on the number of different cell types, etc., the size of the different compartments has to be taken into account. Morphometry on routine stains of paraffin sections is not easy in all species. On immunohistologically stained sections the compartments are much more definitely delineated (Fig. 3). As the spleen is responsible for initiating immune responses against antigens in the blood, the age and the microbial status of experimental animals largely influence the amount of white pulp. For morphometrical studies it is essential to know whether the spleen was fixed by perfusion, or in the clinical situation whether the arteries or veins were ligated first during splenectomy. Van Krieken et al.

REGENERATED SPLENIC AND LYMPH NODE TRANSPLANTS

223

FIG. 4. Distribution of lymphocyte subsets and dendritic cells in the human spleen. Data derived from several studies quoted in the text. NK,Natural killer cells.

(1983, 1985) studied the size of the red and white pulp of resin-embedded human spleens. These authors argued that traumatically ruptured spleens should not be used as controls; they hypothesized that these spleens might contain a stimulated white pulp. An obvious age-dependent finding concerns germinal centers in the spleens of patients younger that 20 years, as older patients had only few secondary follicles (van Krieken and te Velde, 1986), and children younger than 2 years did not show a normal marginal zone (Timens et al., 1989b).The normal histological structure and function of the spleen should always be used as a reference in studies on splenic regeneration. 111. Phases of Regeneration of Splenic Autotransplants

As early as 1917 Manley and Marine reported experiments in which splenic slices had been transplanted subcutaneously in rabbits. On histological sections 8 to 13 days after transplantation the central parts were

224

R. PABST et al.

necrotic but viable splenic parenchyma was found at the border. Although there was little evidence these authors stated in the summary “. . . autotransplantation had all the morphologicalcharacteristics of a fully differentiated and functionally active spleen”. Their final sentence is still true: “This method of transplantation would seem to offer a means of learning more of the central development, regeneration and function of this complex tissue.” In many following studies only a few figures of histological sections were included and statements were made on the “normal” microscopical structure of the splenic tissue. In this section experiments will preferably be quoted in which the splenic tissue was excised and studied at a number of time points and at short intervals. The methods used include light microscopy (Perla, 1936; Wolf, 1982a, b), enzyme histochemistry (Stutte et al., 1974),immunohistochemistry (Dijkstra et al., 1983; Westermann et al., 1988a), transmission electron microscopy (Johnson and Weiss, 1989; Schroder, 1985), and scanning electron microscopy (Sasaki, 1986, 1989). Only in one study did the author state that not even partial necrosis was seen after implanting very thin slices of rat splenic tissue (-150-300 pm thickness) on the mesotestis and checking between 13 and 389 days after operation (Knake, 1952). The reason for the absence of any necrosis was probably due to the fact that the earliest time point studied was 13 days, by which time the necrotic phase had ended. In all other studies known to us the splenic tissue was described as undergoing necrosis except for a very small rim at the outer border. The cells in avascular transplants of splenic tissue completely depend on the diffusion of oxygen and nutrients. Removal of the capsule before transplantation was therefore advocated to decrease the diffusion barrier at least at that border of the splenic segment. In rabbits, splenic particles (either decapsulated or control segments)were implanted into the greater omentum. The decapsulated particles did not regenerate any better than the others (Livaditis and Sandberg, 1980).Thus, there is no experimental evidence to recommend decapsulation, which would only extend the duration of splenic transplantation in man. A. EARLY EVENTS: NECROSIS AND

INGROWING VESSELS

The following sequence of regeneration has been repeatedly described, with no major differences noted between the species. Except for a very thin outer rim of viable tissue, various stages of pyknosis and karyorhexis have been documented. The splenic tissue is rapidly destroyed, leaving only a mixture of cellular debris and remaining erythrocytes. In this area macrophages full of cellular debris and often positive for iron staining are

REGENERATED SPLENIC AND LYMPH NODE TRANSPLANTS

225

found for a long time afterward. The peripheral parts of the autotransplants are of more interest, as many reticular-type cells are found here which form a syncytiallike structure during the first few days. Mitotic figures have often been described in these activated young mesenchymal cells (Dijkstra and Langevoort, 1982; Perla, 1936). In rodents, signs of hematopoiesis were localized here after a few days; e.g., in mice the hematopoietic zone after 10 days was largely erythropoietic but also contained granulocytic and megakaryocytic colonies (Johnson and Weiss, 1989). The many fibroblasts were long with large euchromatic nuclei, abundant RER, microfilaments, and dense subplasmalemmal zones as well as all signs of activated reticular cells (Johnson and Weiss, 1989). Reticulin fibers persisted in the central necrotic part for a few days but disappeared within 2 to 3 weeks. In the outer regenerating zone the typical pattern of reticulin fibers of follicles was described from 21 days onward in regenerating splenic tissue in the rat (Dijkstra and Langevoort, 1982). Johnson and Weiss (1989) noted reticular fibers by Gomori stain extending only as far into the regenerating part of the splenule as the activated fibroblasts. Using histochemistry ,acid phosphatase, and nonspecific esterase, positive cells were found at the periphery of the rat transplants during the first few days. After 6 weeks the positive cells were preferentially found at the inner border of the marginaI sinus. The typical weblike structure of ATPase was seen in the follicles as early as 5 weeks after transplantation (Dijkstra and Langevoort, 1982).

B. DEVELOPMENT OF SPLENIC COMPARTMENTS In rabbit transplants the new sinus-lining cells were identified by a-naphthol acetate esterase and naphthol-AS-acetate esterase reactivity, and from about 3 weeks onward acid phosphatase-positive macrophages were seen in the still narrow pulp cords (Stutte et al., 1974). In mice, by 2 weeks after transplantation white pulp was recognized ( Johnson and Weiss, 1989), while in pig spleens only a narrow rim of viable cells surrounded the central necrosis at this early time point (Pabst and Reilmann, 1980). In rats a typical sequence was seen when monoclonal antibodies were used to identify different cell subsets: development of the white pulp with the PALS at 3 weeks followed by the primary follicles, the marginal zone at about 4 weeks, and secondary follicles at about 8 weeks (Dijkstra et al., 1983). In the same study IDC were identified in splenic transplants by anti-Ia antigens. In nude mice lacking a thymus (and thus T lymphocytes) the typical T cell area, the PALS, was very poorly developed in transplanted splenic tissue (Sasaki, 1990b).

226

R. PABST et al.

c. REVASCULARIZATION OF THE TRANSPLANT A prerequisite of all regenerating tissues is a sufficient supply of oxygen and nutrients. Thus, the revascularization of the splenic transplant is an essential part of regeneration (Dijkstra et al., 1983; Sasaki, 1986, 1989; Schroder, 1985). It has been repeatedly described that during the first few days after transplantation red cells were seen in spacesjust underneath the surface. Some fibroblasts were located within these spaces imitating endothelial cells and these spaces did not constitute a part of the original splenic vasculature (Johnson and Weiss, 1989). The vascular space has been described as polarized, with endothelial cells rapidly developing along the outer surface, and later fingerlike projections of fibroblasts and red cells expanding deeper into the necrotic area (Johnson and Weiss, 1989). In mice, the original large vascular space was broken up into several smaller spaces at the end of the first week of regeneration. As early as 9 days after implantation the first small arteries in the regenerated zone were surrounded by lymphocytes, indicating the first signs of regeneration of the white pulp. A specialized area of the splenic vasculature in murine spleens is the marginal zone and the marginal sinus. This failed to develop in a normal fashion even after 10 weeks (Johnson and Weiss, 1989). Sasaki (1986, 1989) studied the neovascularization in splenic autotransplantation in rats using scanning and transmission electron microscopy. Some subcapsular sinus endothelial cells migrated to form a preliminary vascular network and obviously anastomosed with the ingrowing capillary sprouts. Later the interendothelial slits developed after vacuoli formation, fenestration, and condensation of intraendothelial microfilaments. In a further study rats were injected with phenylhydrazine after splenic autotransplantation into the greater omentum. This damages red cells which are filtered by the developing red pulp. After day 1 I the sinus endothelial cells contained normally developed microfilaments, the reticular cells also contained microfilaments, a normal basal lamina was obvious, and large numbers of erythrocytes passed through the slits (Sasaki, 1990a). After transplanting splenic fragments under the kidney capsule, surprisingly no capsulelike connective tissue developed at the border between splenic and kidney cells (Sasaki, 1988). This enabled experiments on the connections between vascular systems. Twelve weeks after splenic autotransplantation in rats the tissue was studied by scanning electron microscopy. There were two types of vessels: (1) the conglomerated complicated splenic vasculature and (2) a few fine vessels forming bridges between kidney and spleen vasculature. AU these experiments showed that the regeneration of autotransplanted splenic tissue is not only of interest in itself, but might also be used to study angiogenesis in general. It is note-

REGENERATED SPLENIC AND LYMPH NODE TRANSPLANTS

227

worthy that not only the complicated structure of the splenic vasculature, the localization of lymphocyte subsets, macrophages, and dendritic cells are reconstituted to an dmmt normal pattern ugder optimal conditions, but nerves also seem to grow into the regenerated splenic tissue (Felle and Folan, 1988; Weihe et al., 1991)similar to the innervation pattern in normal rat spleens (Felten, D.L. et al., 1987). Despite the development of splenic vessels, red pulp, and compartments of the white pulp, there were also areas, mainly localized in central parts, which contained unusual amounts of extracellularmaterial, such as granular matrix or collagenous bands, and also clusters of macrophages filled with cell detritus and pigments. There were strands of typical scars in all species, which were definitely different from normal trabeculae. The amount of scars was largely dependent on the age when splenic transplantation had been performed, i.e., least in newborn infants and most in adults (Westermann et al., 1988a). The time needed to reconstitute all splenic compartments is obviously dependent on species and age, but the general sequence is comparable in all species. This pattern is summarizedin Fig. 5. D. IMMUNOHISTOLOGY OF REGENERATED SPLENIC TISSUE Antibodies against lymphocyte subsets, macrophages, or other cell types have not yet been used to any great extent in immunohistological techniques on splenic transplants. In a case report on a biopsy taken 5 weeks after implantation an obvious depletion of T and B lymphocytes was described, while on tissue sections from a patient 4 years after implantation, T helper and T suppressor lymphocytes were stained and found to have a normal ratio, although the follicles were hypoplastic (Gudat et al., 1985). The immunoarchitecture of splenic transplants in adult rats revealed a sparsity of lymphocytes and an abnormal mixture of B and T lymphocyte subsets, with a marked reduction of the PALS and follicles (Dugan et al., 1986). In a more recent study a series of monoclonal antibodies against lymphocyte and macrophage subsets was used with immunohistological techniques. Three months after splenic autotransplantation a completely normal distribution of lymphocyte and macrophage subsets was found in newborn rats (Fig. 6),while in adults the splenic compartments were only poorly developed (Westermann et al., 1988a). These studies indicate that descriptions of a “normal splenic structure” in autotransplanted tissue in many studies using routine histology are probably due to the fact that the abnormal composition of lymphoid cells in the regenerated spleens had not been detected. Since monoclonals are now available for lymphocyte subsets and macrophages in most species of interest in splenic transplan-

Necrosis houn aflef implantation

Activation days afler implantation

New tissue formation days afler implantation

Red pulp formation weeks aner implantation

White pulp formation weeks afler implantation

Final structure months after implantation

FIG.5 . Schematic drawing summarizing the data on the process of splenic regeneration. The different phases of regeneration are illustrated and approximate time points are indicated. The relative size of the different compartments roughly corresponds to the respective area. It is not shown, however, that the mass of the final regenerated tissue is usually smaller than that of the autotransplanted tissue. CO: corona; F: follicle; GC: germinal center; lyrn. acc.: lymphocyte accumulation outside the capsule; MZ:marginal zone; PALS: penartenolar lymphatic sheath. 228

FIG.6. Serial cryostat sections showing the distribution of red pulp macrophages (ED2+, left panel) and T lymphocytes (Ox19+, right panel) in splenic regenerated tissue 12 months after autotransplantation. The operations were performed on (a) 24-week-, (b) 3-week-, and (c) 2-day-old rats. Note the abnormal structure when the operation was done on old rats and the apparently normal size and composition of the splenic compartments when young animals were operated on. F: follicle; FT: fibrotic tissue; MZ:marginal zone; PALS: periarteriolar lymphatic sheath; RP: red pulp. The asterisk indicates an accumulation of lymphocytes outside the capsule (immunohistologicalstaining revealed by the alkaline phosphatase antialkaline phosphatase reaction). x 378.

230

R. PABST et al.

tation, the use of immunohistologicaltechniques will be essential in future studies.

E. MORPHOMETRY OF REGENERATED SPLENIC COMPARTMENTS There are only few data available on the quantitative aspects of the regeneration of autotransplanted splenic tissue (Felle and Harding, 1988; Moore et al., 1986). As will be pointed out later, the regeneration of the different splenic compartments depends on the age at the time of operation. Moore et al. (1986) used male young adult rats, implanted a splenic slice of about 100 mg in a pouch of the small intestinal mesentery and studied the regenerated tissue 6 months later. Routinely stained sections revealed a reduction in the white pulp from 24.8% (control) to 5.5%, in the marginal zone from 16.6 to 3.5% and in the PALS from 9.9 to 2.3%. Another interesting aspect of this study was the tremendous variation; for example, the marginal zone in transplants ranged from 0 to 15.7%. The different compartments of the spleen can be excellently delineated in sections immunohistologically stained with the antisera-detecting macrophages and B and T lymphocytes. The results are summarized in Table I1 and show a decrease in the marginal zone and PALS of the splenic transplants. In comparing morphometric data from different groups it has to be taken into account that age, strain, and microbial status (conventional, SPF, etc.) all influence the size of splenic compartments. It can be concluded that when autotransplantation is performed in adult animals, the regenerated splenic tissue does not manage to produce a typical pattern qualitatively or morphometrically, but consists of an abnormal architectural organization. The few data published on the situation in man point in the same direction (Grogan et nl., 1984; Gudat et nl., 1985). TABLE I1 MOWHOMETRY OF THE RELATIVE SIZE OF THE A N D AUTOTRANSPLANTED COMPARTMENTS IN NORMAL RAT SPLEENSa

Normal spleens Red pulp Marginal zone Follicles PALS

55.4 ? 4.3% 27.6 2 3.1% 5.4 & 0.7% 11.1 +. 2.8%

Data from Willfilhr er al., (1990). This also included fibrotic tissue.

Regenerated sdeens 69.9 2 19.4 2 6.6 2 4.1 2

12.1%b 10.5% 1.2% 2.1%

REGENERATED SPLENIC AND LYMPH NODE TRANSPLANTS

231

F. Do SURVIVING SPLENIC CELLSREPOPULATE THE TRANSPLANT OR Is IT INVADED BY EXOGENOUS CELLS?

Various techniques have been used to elucidate this question. Metcalf and Wakonig-Vaartaja (1964) used the T6 chromosome as a marker for metaphases of regenerated splenic tissue at different times after implanting spleens from I-day-old AKR mice into (AKR x T6)Fl animals. Decreasing donor mitotic figures were identified with time after implantation, resulting in a complete host type from 35 days onward. SI/Sld mice with a genetic defect in erythropoiesis were used in other experiments. Implants of normal splenic stroma stimulated erythropoiesis in cells of recipient origin in the regenerated splenic tissue (Wolf, 1978). When splenic fragments were placed in millipore chambers within the abdominal cavity to avoid the entry of precursor cells into the avascular splenic tissue, no regeneration of any lymphoid elements was found (Willfuhr et a1.,1990). Connective tissue cells, at least at the outer border of the splenic fragments, have to survive and form the first structural elements or the microenvironment for the immigrating precursors. Splenic implants were irradiated before implantation with more than 1750 rads in rats (Tavassoli et al., 1975)or 10oO rads in mice (Wolf, 1982a)to destroy splenic cells. After this procedure, no regeneration was found. These data therefore indicate the importance of the viability of stromal elements in the implants. All these experiments show that no stem celllike elements survive the transplantation of splenic tissue. Consequently, precursor cells from the bone marrow or other lymphoid organs have to immigrate into the regenerating splenic tissue to reconstitute splenic hematopoiesis and the lymphoid elements.

G.

HOW

LONGDOESTHE REGENERATIVE PROCESS OF SPLENIC TISSUETAKE?

An important aspect of general biological and clinical significance is the length of time needed to reach the final size, structure, and function of the regenerated spleen. Obviously, for splenic tissue in small laboratory animals like mice and rats regeneration seems quicker than in pigs and man. The function and structure of autotransplants have been studied at different times after the operation making a comparison of published data difficult. In mice and rats regeneration has been described as complete after 2-3 months. In adult rats about 50% of the resected spleen was transplanted into the greater omentum and the regenerated splenic weight and blood !low were determined 3,6,12, and 18months later (Westermann et al., 1988b). There was no significant difference between the age groups

232

R. PABST et al.

in the regeneration after 3 months, when about 20% of the weight of control spleens and less than 1/6 of normal splenic blood flow was reached. As 18 months is about half the life span of a rat, this study can be regarded as a model for long-term regeneration. In young pigs there was less developed regenerated splenic tissue at 3 months than at 6 months after transplantation (Pabst and Reilmann, 1980; Reilmann er a/., 1983). In man, scintigraphy is usually the only possible technique to check the growth of splenic transplants. The tempo of regeneration seems to be as slow as in pigs. Nielsen er al. (1981) studied 90 patients with splenectomies after splenic ruptures and detected signs of splenosis in about 50%, with a time span between the operation and positive scans of between 3 months and 11 years. In several other case reports the regeneration of splenic tissue was detected after a few months in man, although no repeated checks were made at later time points (Bottcher and Seufert, 1985; Grogan er al., 1984; Holschneider et al., 1982; Liidtke er al., 1989; Moore et al., 1984; Nielsen er al., 1984; Patel er al., 1981; Seufert, 1986; Seufert et al., 1981; Traub er al., 1987; Velcek et al., 1982a). Diirig and Harder (1986) even described a decrease in splenic clearance function over a few years using perfusion scintigraphy in man. When Witte et al. (1988) documented data on a patient with scintiscans at 3, 9, and 17 months after autotransplantation, there was no evidence of growth over this time period.

IV. Which Factors Muence Splenic Regeneration? A. Is THEREAN OPTIMAL SITEFOR SPLENIC REGENERATION?

Splenic fragments or particles have been implanted at many different sites, e.g., subcutaneous, intramuscular, retroperitoneal, subperitoneal, at the mesentery of the small intestines, in the kidney, and most often in a pouch of the greater omentum. Some of these sites, although interesting for studying specific questions of splenic regeneration, are of no clinical relevance as a model for the situation in man. Sasaki (1988) implanted rat splenic tissue in the kidney in order to study the vascular anastomoses with scanning electron microscopy. This site was advantageous because there is no connective tissue like a scar surrounding the implant. An interesting finding was that splenic tissue seemed to invade the kidney, which is in contrast to another study in rats (Callery et al., 1989). Thus, the newly formed splenic capsule found at most sites is dependent on the bed of the transplant and is not a product only of the splenic tissue. In most studies on splenic autotransplantation only one site was used, and arguments were usually put forward as to why that site should be favored. Only few data

233

REGENERATED SPLENIC AND LYMPH NODE TRANSPLANTS

have been published in which different sites were compared using the same species, the same techniques, and the same time after transplantation. The intraomentaland subcutaneousor subfascial sites have been advocated for man most often and therefore the experimental evidence for both these sites will be described in more detail. In mice, subcutaneous implants contained fibrous tissue and were less developed than tissue implanted in the greater omentum (Hebert ef al., 1989). In rats, the hemolysin titer after i.v. injections of sheep red blood cells, the number of plaque-forming cells per lo6 spleen cells, and the spleen cell count were all significantly higher in omental as compared to subcutaneous transplants (Bowman et al., 1977). In a naturally occurring epidemic murine mycoplasma infection, the intraperitoneal implants provided protection but not the subcutaneous implants (Livingston et d., 1983b). In a further study Vega et al. (1981) quantitated the reticuloendothelial function of splenic transplants using radioactively labeled isologous red cell scans and revealed a significantly better function at intraomental than subcutaneous sites. The antibody response to i.v. injection of sheep red blood cells was significant in spleen transplanted rats, with higher levels in animals which-hadtransplants at intraomental sites as compared to subcutaneous sites (Schwartz et al., 1977). In a recent study of rabbits (Thalhamer et al., 1989b) a normal spleen was compared to intraomental and subperitonealtransplants with respect to the immune response and the protein composition. The immune response to i.v.-injected sheep red blood cells was measured by the plaque-forming assay. The capacity of subperitoneal transplants was only about 3% of that of control spleens in contrast to about 70% in omental transplants. Using the immunoblotting technique two antigenic determinants could not be detected in most regenerated splenic tissue. In addition, there were some specific alterations in protein composition between the peritoneum and the intraomental implants. As long as these proteins and the missing antigenic determinants have not been determined, the functional meaning of these findings is difficult to define. These experiments also showed differences at the molecular level between regenerated tissue at different sites. Intraomental transplants also regenerated in dogs and were identified by radionuclide imaging, which was not the case with subcutaneous implants (Krasna and Thompson, 1985). In pigs the intraomental grafts showed a prominent white pulp, while subfascial implants were much less developed and contained more fibrous tissue. The number of follicles per cm2was 17.0 8.2 at subfascial sites in contrast to 39.2 2 31.4 at intraomental sites (Reilmann ef al., 1983). In conclusion, whenever the subcutaneous and intraomental sites were compared in the same species and similar techniques used, the intraomen-

*

234

R. PABST et al.

tal transplants always showed better results. It has often been argued that venous draining into the portal system is the essential aspect, but more detailed studies are necessary to elucidate the factors which result in better splenic regeneration in the greater omentum.

B. D o ~ THE s SIZEAND AMOUNT OF IMPLANTED SPLENIC TISSUE THE MASSOF REGENERATED TISSUE? DETERMINE In a study which has often been quoted, Tavassoli et al. (1973)implanted autologous splenic fragments of different weights in subcutaneous pockets of adult male rats and determined the mass of tissue 5 weeks later. There was a linear relationship between the weight of the implanted tissue and that of the recovered splenic tissue up to 100 mg of the implanted tissue. It seems to be relevant that each rat received multiple implants at the same time because these different pieces might have influenced the regeneration of each other. After a period longer than 5 weeks larger particles might also have shown a linear relationship. Thus, the repeated extrapolation to the situation in man, that particles of less than 100 mg are optimal, is a misinterpretation of this study. Using a different site, (e.g., the greater omentum), other age groups of animals, or a later time for evaluation are all factors which might have altered the results. For the clinical situation it is not only important to define the size of each transplanted particle but also the total mass of transplanted tissue, as any increase will increase the amount of necrosis, resulting in fever and other complications. In young adult rats multiple splenic fragments were implanted subcutaneously and their regeneration studied 6 weeks later (Tavassoli, 1975). The weight of recovered tissue did not correlate with that of the implanted splenic mass, and animals receiving more than 10 spleens died within 3-6 days. This was probably due to massive necrosis. When rats received 2-3 spleens every 3 days up to a total of 15 spleens, only the first grafts survived (Tavassoli, 1975). However, when a number of thymus fragments were transplanted, a completely different pattern emerged, as up to 10 times the normal weight of thymus regenerated in a single animal (Metcalf, 1%3b).Thus, the regeneration of splenic fragments seems to be regulated by the total amount of implanted tissue, in contrast to the autonomous regeneration of each thymus graft. The role of the total mass of implanted autologous splenic tissue on the amount of regenerated tissue in the greater omentum has been studied in normal young pigs. The range was from -7% of the resected splenic tissue to the whole spleen, but the resulting weight at 6 months did not differ significantly (Pabst and Kamran, 1986; Pabst and Westermann, 1987). Thus, in this animal model there was no evidence to advocate the autotransplantation of large

REGENERATED SPLENIC AND LYMPH NODE TRANSPLANTS

235

amounts of splenic tissue. In fact, the more avascular fragments are transplanted, the more necrosis will develop. In the clinical situation different amounts of splenic tissue have been transplanted and, based on the mass extrapolated from scintigraphies, larger implants did not result in greater masses, although no comparative studies in man are known to us. A further important aspect is whether the regenerated tissue should be compared to the amount implanted or to the weight of an age-matched control animal. The latter alternative seems to be most sensible, especially when splenic autotransplantation is performed in young animals or children. For the clinical situation, an additional aspect might be of importance in the future. When a spleen has to be removed for diagnostic purposes but no pathological elements are then found, it could be of interest to freeze part of the spleen and transplant these fragments once a definite diagnosis has been established. In rats at least, frozen and thawed splenic tissue regenerated in the same way as fresh particles (Crosby and Benjamin, 1961).

C. INFLUENCE OF AGEON THE REGENERATION AND FUNCTION OF SPLENIC TRANSPLANTS In the foregoing sections we have seen that, although autotransplanted splenic fragments regenerate, the mass of regenerated tissue only attains a fraction of that of a normal spleen. In children the incidence of OPSI is much higher than in adults, and therefore the restoration of sufficient splenic tissue after splenectomy is an urgent clinical problem in this age group. The following experiments not only underline the influence of age on splenic regeneration, but are also of general interest. As early as 1920 Marine and Manley noted an influence of age on splenic regeneration in rabbits. In suckling and adult rats about 25% of the splenic mass was autotransplanted in homogenized form directly into the abdominal cavity. In all autotransplanted suckling rats splenic tissue was recovered 12 weeks later, while in the adult group only five out of ten animals had regenerated tissue, and mostly only tiny amounts (Kovacs et al., 1981). Although implantation directly into the peritoneal cavity is not an ideal technique, these experiments demonstrated a better regenerative capacity in young rats. The effect of age was studied recently in much more detail in a series of experiments (Westermann et al., 1988a, b; Willfuhr el al., 1991). There are two major reasons why age might influence splenic regeneration: (1) the regenerative capacity of the splenic tissue might be better the younger the donor is and (2) the age might influence the capacity of ingrowing vessels at the implantation site to restore a normal blood supply as the basis for splenic regeneration. To test both aspects inbred Lewis rats

236

R. PABST et a/.

were used and fetal, newborn, weanling, or adult splenic fragments were implanted in the greater omentum of newborn, weanling, or adult rats. The weight and blood flow of the regenerated tissue were measured 3 months later (Westermann et ol., 1988b). The pertinent finding was that the younger the recipient and the donor, the better the regeneration and perfusion of the regenerated splenic tissue, about 100% better in young as compared to adult animals. However, it was significant that no more than 40% of the total splenic mass or 1/3of the normal blood flow was attained. In a similar combination of splenic transplants the immunoarchitecture of the regenerated tissue was studied using monoclonal antibodies against macrophages and T and B lymphocyte subsets. A normal splenic structure with the typical compartments (red pulp, marginal zone, PALS, and follicles) was completely absent in fetal and newborn spleens, but it was these implants which were the only ones to regenerate to splenules with welldeveloped compartments within 3 months (Fig. 6) (Westermann et al., 1988a). Obviously the young spleen has a much better potential for regeneration. These findings suggest that splenic regeneration in children might result in splenic tissue with almost normal compartments, in contrast to the inadequately developed structures found in the few cases studied in man so far (Gudat et al., 1985; Moore et al., 1983). Only the histology of a transplant from a child removed 13 months after implantation seemed to show normal compartments (Henneking and Dobroschke, 1987), which is in line with the animal data. In a retrospective study on 120 splenectomized patients the frequency of splenosis was determined by 99mTcscans and revealed the highest rate of 80% in children of less than 10 years (Alvarez et al., 1987).

D. REGENERATIONFOLLOWING AVASCULAR TRANSPLANTATION OF WHOLESPLEENS OR L~GATION OF ALL SPLENIC VESSELS

In adult mice complete spleens were implanted subcutaneously in the sacrospinal area (Warner and Krueger, 1975). There was a continuous drop to about 1/3 of the implanted weight during the first 4 weeks. From day 38 to day 50 a linear increase was seen in splenic weight, but this only reached about 50% of the weight of control spleens. The regeneration started underneath the capsule and the new pulp was colonized by hematopoietic cells beginning with granulocytopoiesis and thrombocytopoiesis but no erythrocytopoiesis. Thus, a similar pattern to the regeneration of splenic fragments was seen after transplanting whole spleens in the mouse with not only lymphoid cells but also hematocytopoiesis recovering. In pedicle-ligated mouse spleens regeneration took place in a similar sequence to that in fragment transplantation (Wolf, 1982b). A small rim of surviving cells started to proliferate underneath the capsule and recon-

REGENERATED SPLENIC AND LYMPH NODE TRANSPLANTS

237

struction seemed to be completed after 40 days. Extramedullary hematopoiesis always preceded the restoration of lymphoid compartments. This might indicate that different stromal elements are needed for hematopoiesis and for homing of lymphoid cells. It has been stressed a number of times that the splenic capsule should be removed before transplanting splenic fragments. The experiments cited above, however, document comparable regeneration despite an intact capsule around the spleen. OF ORTHOTOPIC SPLENIC TISSUE ON THE E. INFLUENCE REGENERATION OF HETEROTOPIC TRANSPLANTS

This aspect is not only of interest as a factor in regeneration but is also of clinical relevance. If a human traumatized spleen is salvaged by segmental resection leaving a part of the spleen connected to the normal blood supply, is it necessary to transplant the resected part heterotopically? Metcalf( 1963a)transplanted splenic tissue subcutaneously in inbred mice. The recipients were either splenectomized or sham-splenectomizedbefore transplantation. Six weeks later the regenerated splenic mass was about 3 to 5 times greater in splenectomized mice and the amount of lymphoid tissue was also much more prominent in these animals. Splenectomy or sham operation 1 week after implantation still resulted in about twice the amount of regenerated tissue in splenectomized animals. When tiny splenic fragments were placed on the mouse mesentery, well-developed splenic implants were recovered in splenectomized animals, while only small atrophic nodules were found in normal recipients (Ambrus et al., 1964). In experiments on adult rats, spleen transplants grew larger and functioned more actively in animals splenectomized beforehand than in normal rats (Bradshaw and Thomas, 1982;Jacob etal., 1963). The effect of transplanting half of the spleen and either leaving the rest in situ or excising it was studied in normal young pigs (Pabst and Kamran, 1986). After 6 months a mean of 4.5 g of regenerated splenic tissue was found in splenectomized pigs but only 2.0 g when half of the spleen was left in situ. These studies document a suppressive effect of the orthotopic splenic tissue on the regeneration of splenic tissue at other sites. So far nothing is known about which substances of factors are responsible for this effect or whether a stimulatory factor is lacking.

F. ENHANCEMENT O F SPLENIC REGENERATION Autotransplanted splenic tissue does not attain the mass of a normal spleen. This is of clinical relevance in respect to the protective function, but also of general biological interest. Why does regeneration stop after reaching about 10-20% of the normal splenic weight in many experimental

238

R. PABST et al.

animals? After partial hepatectomy the normal size of the liver is reached after a short period of regeneration. Obviously different regulatory mechanisms are at work in splenic regeneration. One hypothesis was that by increasing the work load to regenerating splenic tissue, the size of the transplant might be improved, since earlier studies had indicated such effects (Ambrus et al., 1964; Jacob et al., 1%3). As the splenic compartments differ in their function, different stimuli were applied: (1) the red pulp was stimulated by damaging autologous red cells using repeated injections of phenylhydrazine; (2) the white pulp was stimulated by intravenously injecting particulate antigens; (3) the splenic macrophage system was stimulated by intraperitoneal injections of methylcellulose (Pabst et al., 1985a). In normal animals the size of the spleen was enhanced by a factor of two using phenylhydrazine and by about ten using methylcellulose. Histological sections revealed a predominance of the compartments red pulp, white pulp, or macrophages depending on the stimuli. The same was true for splenic transplants: while NaCI-injected control rats reached 20% of the implanted mass within 3 months, after stimulating the white pulp, the red pulp, and the macrophages, this increased to >50%, loo%, and about 2 W , respectively. Thus, when the work load to the splenic compartments was increased the weight of the normal and the regenerated spleen increased in response. A further interesting aspect was whether this effect would disappear once stimulation ceased. In spleen transplanted rats the red pulp was stimulated as described and the histology and weight were studied 1 and 11 weeks after the end of the stimulatory period. Within this period the weight of the splenic transplants dropped to -64% of the implanted mass, which was only half of that recorded at one week (- 12Wo), but still three times higher than in unstimulated animals (Pabst et al., 1985b). In pigs, however, repeated injections of phenylhydrazine or of lipopolysaccharideafter autotransplantation of splenic tissue did not result in a greater mass of regenerated tissue (Pabst et al., 198%). Thus, there seem to be important species differences. For the patient the most critical aspect is the protective function of the splenic transplant rather than its weight. Therefore, rats with stimulated splenic regeneration were infected with Plasmodium berghei, as the height of the parasitemia and the survival of malaria depend on functional splenic tissue. Despite the dramatic increase in the red pulp after phenylhydrazine treatment or stimulation of the white pulp or macrophages, the survival rate was no better than in NaCItreated rats. Thus the mass of regenerated tissue did not correlate with the protective function of the splenic tissue (Malangoni et al., 1985a; Pabst et al., 1988; Westermann and Pabst, 1988). It remains to be studied whether other protocols of stimulation or other organisms like pneumococci would provide more insight into the mechanisms governing spleen size and protective function.

REGENERATED SPLENIC AND LYMPH NODE TRANSPLANTS

239

G. SPLENIC REGENERATION IN DIFFERENT SPECIES

Splenic tissue has been transplanted in many different species, whereby the size of the different compartmentsof the spleen, the number and size of trabeculae, and the size of the capsule vary considerably (Tischendorf, 1969,1985).The peculiarities of the splenic structures have to be taken into account when splenic transplantation in animals is used as a model for the situation in man. For example, in dogs the red pulp is predominant and can store up to about 25% of all red blood cells, which can be rapidly mobilized. The experiments on splenic transplantation in dogs can therefore hardly be used as a model for man (e.g., Krasna and Thompson, 1985; Velcek el al., 1982b). Many studies have been performed in rodents, such as mice (e.g., Dickerman et al., 1979; Likhite, 1978; Metcalf, 1963a), rats (e.g., Church et al., 1981;Moxonand Schwartz, 1980;Nielsen et al., 1983; Scher et al., 1985b; Schwartz et al., 1977; Tavassoli et al., 1973), and rabbits (Manley and Marine, 1917; Marine and Manley, 1920; Schroder, 1985; Mizrahi and Barzilai, 1988; Stutte et al., 1974). In most studies using mice, splenic tissue regenerated rapidly to reach its final size in a few weeks. This is in contrast to the situation in man, with much slower regeneration. The pig has the advantage as an experimental model that its tempo of regeneration is more comparable to that in man. Furthermore, techniques used in man (e.g., the use of slices of the spleen or minute particles, fixation of the implants in the greater omentum, etc.) can easily be studied in the pig (Pabst and Kamran, 1986; Pimp1 and Wayand, 1983; Reilmann et al., 1983; Sakso et al., 1985; Thalhamer et al., 1991). Two studies in pigs have dealt with the question of whether thin slices of splenic tissue or minced splenic tissue would result in a larger mass of regenerated tissue. Sakso et al. (1985) found 33% of the implanted mass after implanting small particles and 21% after implanting slices. The operations were carried out in adult pigs, both techniques being performed in the same animals. In young minipigs either a simple slice or three small particles, both amounting to about 7% of the spleen, were implanted, and 6 months later no differences were seen in the regenerated mass (Pabst and Westermann, 1987).Thus, not only the species but also the age of the experimental animals are important parameters, which should always be considered in comparing data from different studies.

V. Function of Regenerated Splenic Tissue The normal spleen has many different functions, which can partly be attributed to specific splenic compartments. The parameters which can be determined in the normal spleen and regenerated splenic tissue in experi-

240

R. PABST et al.

mental animals and man will be described here: blood flow, clearance of particles, filter function of the red pulp, effects on the cellular composition of other lymphoid organs and peripheral blood, and finally the most important aspect, protection against sepsis. A. BLOODFLOW

Many functions of the spleen are largely dependent on the blood flow. The flow per gram of splenic regenerated tissue was dramatically reduced in pigs by up to 8Wo (Pabst and Kamran, 1986;Pimp1et ai., 1987; Reilmann et al., 1983). Similar data were published for rabbit splenic tissue (Horton et af., 1982). In rats, however, the perfusion per unit weight did not differ between normal and regenerated splenic tissue. Due to the small amount of regenerated tissue mass, however, the perfusion did not reach more than one third of the normal splenic blood flow (Westermann et al., 1988b).

B. CLEARANCE FUNCTION Intravenously injected particles are removed by phagocytic cells in the spleen. Tantalum particles were found in the marginal zone in regenerated splenic tissue in rats and later these phagocytosed particles were distributed throughout the red pulp (Dumont and Martelli, 1978).The intravascular clearance of a radioactively labeled gelatinized lipid emulsion (Chaudry et al., 1980)or *Tc sulfur colloid has been used in rats (Malangoni et al., 1988) and in patients (e.g., Millikan et al., 1982). As the blood stream clearance did not differ between control, splenectomized, and autotransplanted rats (Malangoni et al., 1988),this was therefore not a good functional parameter. It can, however, be used to visualize the transplant (e.g., Millikan et al., 1982). In several experiments bacterial clearance from the blood was tested after regeneration of splenic tissue. In rabbits no d a e r ence was found in the clearance kinetics of Escherichia coli between splenectomized and autotransplanted animals. However, the clearance was remarkably decreased in comparison to normal rabbits (Thalhamer e f al., 1986). In rats pneumococcal blood clearance was significantly reduced in autotransplanted compared to control and partially splenectomized animals (Scher et al., 1985a). Similar data were obtained by measuring radio-colloid uptake in rats (Malangoni er al., 1985b) or using comparable techniques in rabbits and dogs (Pate1et al., 1986; Velcek et al., 1982b). An interesting modification is the use of radioactively labeled human serum albumin millimicrospheres which are taken up by splenic macrophages and degraded later, indicating the uptake and digestion capacity of splenic macrophages (Bottcher and Seufert, 1985; Bottcher et al., 1981; Seufert et

REGENERATED SPLENIC AND LYMPH NODE TRANSPLANTS

241

al., 1981). Thus, the clearance of different particles from the blood is dependent on a functional spleen, and regenerated splenic tissue partly restores this function, although some data are not unequivocal.

C. FILTERFUNCTIONOF THE RED PULP The normal splenic red pulp filters red cells which are not able to squeeze through the pores of the red pulp cords. Autologous red cells can be mildly damaged by heating. These erythrocytes resemble spherocytes and, after labeling with a radioactive marker, they can be used as a functional test for the red pulp (Kiroff et al., 1983; Nielsen et al., 1984; Spencer et al., 1981; Witte et al., 1988). In all these studies regenerated splenic tissue had been found and the results showed that the percentage uptake per gram was higher in orthotopic remnants than in heterotopic, regenerated splenic tissue. When this technique was used repeatedly at 3, 9, and 17 months after implantation it was possible to show that in man, splenic remnants failed to increase with time and that red cell uptake remained poor (Witte et al., 1988). The red pulp of the spleen removes nuclear remnants called Howell Jolly bodies from erythrocytes. In normal man far less than 1% of red cells contain Howell Jolly bodies. In spleenless patients these numbers are much higher, but after spleen autotransplantation they are not quite as high as in splenectomized patients (Ludtke et al., 1989). The reappearance of some splenic function can be deduced from the reduction of Howell Jolly bodies with time after autotransplantation (Alvarez et al., 1987; Holschneider et al., 1982; Orda et al., 1981; Pate1 et al., 1981; Velcek etal., 1982a). By using interferencecontrast microscopy, tiny vacuoles or “pits” can be identified on the membrane of red cells in patients without a spleen. These pocked or pitted red cells are normally smoothed on passing through the red pulp (Pearson et al., 1978). The advantage of using pitted red cells as a parameter for splenic function is that patients without a spleen have more than 30% pitted red cells. However, the disadvantage is that there are no pitted red cells in some experimental animals, e.g., rats and pigs. This splenic function only commences after birth, as documented by the high incidence in premature and newborn infants and the decrease with age (Holroyde and Gardner, 1970; Holroyde et al., 1969; Kim et al., 1980). Using this parameter some function of the regenerated splenic tissue was shown, although normal values were not attained (Livingston et al., 1983a; Traub et al., 1987). The spleen plays a critical role in protozoal infections and in particular in malaria (Garnham, 1970; Todorovic et al., 1967; Wyler, 1983), with the red pulp being of particular importance (Weiss et al., 1986; Wyler et al., 1981). Therefore, malaria infection with P . berghei has been used to test the

242

R. PABST et al.

protective effect of regenerated splenic tissue (Pabst et al., 1988; Westermann and Pabst, 1988). Spleen autotransplanted rats showed comparable values to normal rats with respect to the peak parasitemia, the course of the parasitemia, and the extent of resulting anemia. The duration of the parasitemia was about 3 weeks in control, and about 10 days longer in transplanted animals. In contrast, splenectomizedrats were totally unable to clear the parasites. D. EFFECTSOF SPLENIC TISSUEON THE PERIPHERAL BLOODA N D OTHERLYMPHOID ORGANS Tarnuzi and Smiley (1967) have studied the effects of splenic implants on the blood in rats. After splenectomy an increased number of leucocytes and thrombocytes were found in the blood, but splenic implants, either in the peritoneal cavity or enclosed in a diffusion chamber, reduced the leucocytosis. These data point to a humoral substance produced by the normal and also by the autotransplanted spleen which influences the number of platelets. Granulocytes from patients splenectomized after splenic trauma displayed a normal function in several in uitro tests, but the important functions such as chemotaxis, bactericidal capacity, aggregation, and stimulated 02 and H202 production were impaired (Wysocki et al., 1989). It is noteworthy that the plasma of these patients induced increased adherence of autologous as well as control neutrophils. One particular study on blood lymphocytes in spleen autotransplanted patients (Durig et al., 1984) initiated a long-lasting discussion. The pertinent data gained from 14 splenectomized and nine splenic autotransplanted patients were as follows: after splenectomy there was a major increase in B cells, a minor increase in T cells, and an unchanged ratio of T helperlT suppressor cells. In autotransplanted patients, however, a selective rise in B cells and T suppressor cells was found, which led to an approximately 50% reduction in the T helper/T suppressor (CD4KD8) ratio. In other studies on splenectomized patients increased numbers of B and T suppressor cells (CD8+) were also documented (Brodsky et ai., 1988; Ferrante et al., 1987). However, in a recent paper no significant decrease in the CD4/CD8 cell ratio was reported in patients with splenic autotransplants (Bergmann et al., 1990). Thus, conflicting data were found for T subsets and the CD4/ CD8 ratio not only in splenectomized but also in spleen transplanted patients, although B lymphocytosis after splenectomy was a constant finding. The contradictory data are probably due to differences in age, indication for the splenectomy, or the interval between the operation and determination of lymphocyte subsets in the blood, which can in itself be influenced by a great number of factors (reviewed by Westermann and

REGENERATED SPLENIC AND LYMPH NODE TRANSPLANTS

243

Pabst, 1990). Blood B lymphocytes of splenectomized patients showed a defective polyclonal IgM synthesis (di Padova et al., 1983). In patients splenectomized for trauma, IG synthesis in uitro and in uiuo was different in comparison to control .subjects with spleens and this defect was not corrected by the generation of splenosis (Drew et al., 1984).Thus, not only are the B cell numbers in the peripheral blood abnormal, but also their function is irregular. In this respect regenerated splenic tissue lacks any corrective effect. Alterations in lymphocyte subsets in the blood have also been demonstrated in a long-term study on rats using repetitive determination of lymphocyte markers (Westermann et al., 1990). The differences in the blood might be due to the different migration kinetics of lymphocytes in splenectomized and spleen transplanted animals and man, as indicated by a different subset composition in the mesenteric lymph nodes of rats (Westermann and Pabst, 1986) and the migration pattern through the regenerated splenic tissue (Willfiihr et al., 1990). The effects of the presence of splenic tissue on lymphocyte numbers in the blood are not surprising, as lymphocyte migration through the spleen quantitatively outnumbers that using the recirculation pathway via high endothelial venules in lymph nodes and the thoracic duct severalfold (Pabst, 1988a,b).

E. IMMUNOLOGICAL AND BIOCHEMICAL DIFFERENCES BETWEEN NORMAL AND REGENERATED SPLENIC TISSUE In a recent study (Thalhamer et al., 1989a) autologous splenic particles were implanted in the greater omentum of normal pigs, and 6 to 12 months later various immunological and biochemical tests were performed on the lymphocytes of normal and regenerated splenic tissue. Several surprising data were obtained. The number of plaque-forming cells against sheep red blood cells, which had been used as an i.v. antigen 3 days before, was similar to the number of antibody-forming cells per lo6 spleen cells. When various mitogens such as the T and B cell stimulators Con A, PHA, PWM, and LPS were tested, a dramatically increased mitogen stimulation index for T cells was revealed in the autotransplanted tissue. Electrophoreses and immunochemical tests showed that lymphocytes from the normal spleen produced one protein which was absent in the transplants. Another protein was found to be enriched severalfold in the transplants. In further experiments the same authors found comparable differences in protein expression in rabbit splenic transplants (Thalhamer et al., 1989b). The functional meaning of these differences has yet to be studied, but these data document more subtle differences between normal spleens and splenic autotranspiants than were previously believed to exist.

244

R. PABST el al.

F. PROTECTIVE EFFECTOF REGENERATEDSPLENIC TISSUE

In the clinical situation OPSI is most often caused by pneumococci. Numerous studies on experimental infection with pneumococci after splenic transplantation have been published. The results, however, were contradictory due to a variety of reasons: different species at various ages were used, different periods after splenic transplantation, different strains and doses of pneumococci, and probably the most important variable, different routes for applying the pneumococci. In i.v. or i.p. injections high doses are suddenly injected into the blood and the splenic tissue is confronted with these specific antigens. In contrast, pneumococci introduced by an aerosol or intranasal inoculation imitate the normal route in man via the respiratory tract. The clearance of i.v.-injected pneumococci obviously depends on the presence of the spleen (Brown et al., 1981 ;Fasching and Cooney, 1980; Pouche et al., 1987). Autotransplanted splenic tissue showed a protective effect in some studies (Livingstone et al., 1983c; Patel et al., 1982; Steely et al., 1988) but not in others (Alwmark et al., 1983), and in some cases only marginal protection (Harding et al., 1987; Vega et al., 1981; van Wyck et al., 1986). It has been argued that protection would be achieved in experiments using the “normal” route via the respiratory tract, but not after i.v. injections (Bottcher et al., 1981). This thesis, however, seems to be an oversimplification, as shown in most recent experiments. Transplantation of splenic tissue was performed in newborn, weanling, and adult rats and 36 weeks later all animals were exposed to Streptococcus pneumoniae via nasal drops. A protective effect of the regenerated splenic tissue was seen only in the newborn rats (Willfuhr et al., 1991). In man a specific compositionof lymphocyte subsets has been described in the marginal zone of the spleen (van Krieken er al., 1985). This population has the capacity to be rapidly and easily activated in a primary immune response (Timens et al., 1989a). In rats and mice, macrophages in the marginal zone can be identified by specific monoclonal antibodies (Dijkstra et al., 1985b) and specific uptake of neutral polysaccharides (Humphrey and Grennan, 1981; Matsuno et al., 1986). These macrophages seem to play an essential role as antigen-presenting cells for thymusindependent type 2 antigens (Humphrey, 1985). The antibody response to this type of antigen has been shown to be dependent on the presence of splenic tissue (Amlot et al., 1985; Claassen et al., 1986). When this type of antigen was injected into splenectomized patients they only reached about 10% of antibody titers in controls (Amlot and Hayes, 1985). Specific subsets of memory B cells which are important in T cell-dependent antibody responses have been detected in the marginal zone of rats (Liu et al.,

REGENERATED SPLENIC AND LYMPH NODE TRANSPLANTS

245

1988; Zhang et al., 1988). In experiments on the influence of the marginal zone of splenic transplants on type 2 thymus-independent antigens, the restoration of the antibody titers correlated with the return of B cells and occurred before the macrophages in the marginal zone were restored (Claassen et al., 1989). Even more surprising data have been described recently. Marginal zone macrophages were destroyed by an anti-macrophage antibody coupled with a toxin (Kraal et al., 1989) or liposomes (Claassen et al., 1988), but despite the lack of marginal zone macrophages the thymus-independent type 2 reaction was still able to function. Thus, it is unclear which cells in the marginal zone are the prerequisite for an effective thymus-independent type 2 reaction (e.g., to pneumococci). However, there is no doubt that the marginal zone is the important compartment for this type of reaction. It is also noteworthy that the marginal zone is immature in the human spleen until the age of about 2 years (Timens et al., 1989b), which has been described as the most vulnerable time for contracting postsplenectomy sepsis. There are a number of case reports in which patients died of OPSI despite some remaining splenic tissue, splenosis, or autotransplanted splenic tissue (e.g., Moore et al., 1983; Sass et al., 1983). Eighteen fatal cases until 1985 were summarized by Pabst and Kamran (1986). It is therefore still important to study whether vaccination protocols can be helpful in preventing OPSI in man. G . THEEFFECTOF SPLENIC TISSUEON THE IMMUNIZATION AGAINST PNEUMOCOCCI

As mentioned before, pneumococci are the organisms which most frequently cause OPSI in man. Immunization with polyvalent pneumococcal vaccines has therefore been recommended in splenectomized patients (Chattopadhyay, 1989; Powell et al., 1988; Shaw and Print, 1989). Nevertheless, OPSI has also been described in vaccinated splenectomized patients (summarized in Hebert et al., 1989). It is therefore of great clinical relevance to define the role of splenic tissue in the antibody response to pneumococci injections. Peripheral blood mononuclear cells of splenectomized patients failed to produce IgM anti-pneumococcal capsular antibodies of an IgM type after pokeweed mitogen stimulation (di Padova et al., 1985), substantiatingother evidence of this defect previously found by the same group (di Padova et al., 1983). In another study only minor differences were found between splenectomized patients in class and subclass antipneumococcal antibody response (Oldfield et al., 1985). Kiroff et

246

R. PABST et a/.

al. (1985) compared the IgM and IgG antibody response in normal and splenectomized patients with and without splenosis after polyvalent pneumococcal immunization. All splenectomized patients showed decreased antibody responses. The presence of ectopic splenic tissue obviously did not function in this respect, which was in contrast to experiments in mice where better antibody responses were noted in intraperitoneally transplanted animals than in splenectomized animals (Hebert et al., 1989). in rats, immunization improved survival in splenectomized and autotransplanted animals after challenge with live pneumococci (Cooney et af., 1979; Dawes et af., 1985). Despite these discrepancies there is no doubt that the lack of splenic tissue means a considerable danger to patients, as they might succumb to overwhelming pneumococci septicemia. There are some indications that the spleen is not only important as a filter for pneumococci in the blood stream but also essential in initiating an effective immune response. Clearance of pneumococci from the mouse lung and transport to the draining peribronchial lymph nodes was dramatically impaired in splenectomized animals (Hebert, 1989a). This reduced pneumococcal clearance could partly be improved by pneumococcal vaccination (Hebert, 1989b). An important question is which compartment of the spleen is of most importance in dealing with pneumococci. In this respect it is of interest that the presence of splenic tissue also had an effect on the function of pulmonary macrophages (Lau et al., 1983; Shennib et al., 1983) and on the response of Kupffer cells to endotoxin (Billiar et al., 1988). Thus, unknown factors from the spleen seem to influence the function of the macrophage system in other organs.

VI. Regeneration ofAutotransplantedLymph Nodes A. PHASES OF REGENERATION

In contrast to the numerous studies on the regeneration of autotransplanted splenic tissue and also of thymic fragments (e.g., Dukor et af., 1965; Stutman and Zingale, 1964), only a few experiments have been published on the regeneration of lymph node (LN) fragments. More than 60 years ago, pieces of inguinal LN (Marmorston-Gottesman and Gottesman, 1928) or whole cervical nodes (Jaf€e and Richter, 1928) were autotranspianted into pockets of the abdominal muscles in adult rats. The results of both studies were comparable in describing the initial destruction of the nodal tissue with obvious survival of some reticular cells. These proliferated after a few days and lymphoid cells reappeared first at the outer border. After about a week the regeneration of the nodes seemed to

REGENERATED SPLENIC AND LYMPH NODE TRANSPLANTS

247

be complete. In these studies it was speculated that lymphocytes differentiate from the reticular cells of the stroma, which has since been completely negated. In two more recent experiments, normal young minipigs were used to study the regeneration of LN fragments in more detail (Pabst and Rothkotter, 1988; Rothkotter and Pabst, 1990). Superficial inguinal LNs were excised and thin slices placed in pockets in the subcutaneous tissue of the groin. The implants were removed at weekly to 1Cday intervals over 12 weeks. The central parts rapidly became necrotic, followed by fatty degeneration and areas of giant cells phagocytozing the cellular debris (Fig. 7). Only a thin rim of surviving reticular and lymphoid cells were seen initially, followed by many capillaries growing into the necrotic parts. From about 3 weeks onward follicular lymphoid aggregates were identified and the first postcapillary venules developed with a cuboidal endothelium (Rothkotter and Pabst, 1990). In experiments on splenic regeneration, the age at the time of autotransplantation greatly influenced the development of a spleen with normal compartments (see above and Westermann et al., 1988a). To test the effect of age on LN regeneration, slices of lymph node tissue were implanted in young and adult minipigs. The regeneration in both age groups was comparable (Rothkotter and Pabst, WO),resulting in small nodules consisting of all compartments of normal pig nodes, e.g., cortex, paracortex, and medulla (Binns and Pabst, 1988). The development of germinal centers after local antigen injections, of HEV in the paracortex as entry sites for recirculating lymphocytes, and of connections of afferent lymphaticsto the regenerated tissue, transporting antigens, and accessory cells, all documented the reestablishment of normal lymph node functions (Fig. 8).

B.

IMPORTANCE OF THE SITE OF TRANSPLANTATION

The site for implantation was found to be of major importance. Implantation of LN slices under the kidney capsule, a site documented to be adequate for thymic fragments (Dukor et al., 1965), resulted in few takes, with only a follicular structure and no paracortex or sinuses developing (Pabst and Rothkotter, 1988; Rothkotter and Pabst, 1990). Very little developed LN tissue was recovered underneath the fascia of the adductor muscles 6 months after the operation. There was no evidence of regeneration after implanting fragments of mesenteric or inguinal LN in the greater omentum. In subcutaneous pockets the number of takes was higher for superficial inguinal LN than for mesenteric LN tissue. It was interesting that the optimal site for the regeneration of splenic tissue, the greater omentum, was completely inadequate for LN fragments.

FIG.7. Stages of lymph node regeneration. (a) One week after autotransplantation of lymph node fragments: fatty degenerated tissue, containing only few lymphocytes. (b) After 2 weeks: new tissue formation, groups of lymphocytes surrounded by fibrotic tissue. Inset: Venule with a high endothelium. (c) After 4 weeks: areas repopulated with lymphocytes and also giant cells in necrotic areas (arrows). (d) After 6 weeks: fully developed compartments as in a normal lymph node. Inset: Lymphocytes in the wall of a high endothelial venule (arrows). F: follicle; FT:fibrotic tissue. x 54, insets x 367.

REGENERATED SPLENIC AND LYMPH NODE TRANSPLANTS

249

FIG.8. (a) Regenerated lymph node tissue 6 months after autotransplantation. Fibrotic tissue can still be detected. (b) Part of a follicle with mitotic figures (arrows). (c) High endothelial venule in the paracortex with lymphocytes in its wall (arrows). F: follicle; FT: fibrotic tissue; P paracortex. (a) x 60,(b) and (c) x 395.

250

R. PABST et al.

C. LYMPHATICS REGENERATE The hypothesis for the success of the subcutaneous site was that lymphatics which were dissected when creating the subcutaneous pouch would grow into the fragments and exert a stimulatory effect. There were several lines of evidence that afferent lymphatics reached the regenerating lymph node and affected its development: (1) subcutaneously injected Berlin blue was detected in newly developed sinuses (Fig. 9), (2) after S.C. injection of killed bacteria in the draining area prominent germinal centers were found in the regenerated lymph nodes, and (3) many postcapillary venules with a high cuboidal epithelium were seen in the paracortical area as in normal LN (Fig. 8). The latter finding can be taken as evidence of afferent lymphatics, since ligation of afferent lymphatics in rat LN resulted in a progressive decrease in the height of the endothelium in the paracortical postcapillary venules (Hendriks et al., 1980; Drayson and Ford, 1984). Recently, mesenteric and peripheral lymph nodes were transplanted in mice and the expression of typical antigenic markers on the HEV was studied. Regenerated lymph nodes from adult animals kept their type of vascular addressin, irrespective of whether they had been implanted orthotopically or heterotopically. In newborn animals this was different, e.g., when a peripheral lymph node was implanted in the mesentery, the HEV in the regenerated node expressed the addressin typical for mesenteric nodes (Mebius et al., 1991). These data indicate an effect of age and an influence of factors or cells in the afferent lymph on the structure and function of regenerated lymph nodes. Lymph edema is still a major clinical problem after radical dissection of regional lymph nodes, in particular in combination with radiation, e.g., in patients with mammary tumors. Despite tremendous efforts, surgical therapy of this type of lymph edema has been very disappointing so far (for review see Clodius and Gibson, 1983). If autologous LN fragments also regenerated in irradiated subcutaneous tissue and afferent lymphatics reached such LN, these lymph edemas might be drained by the newly developed LN. Many more experiments are needed to define factors influencing lymph node regeneration after autotransplantation, the role of cytokines, different cell types and the function of the regenerated lymph nodes in immune response and as a filter and drain for afferent lymphatics.

VII. Conclusions The process of splenic regeneration is surprisingly similar in different species and after different implantation techniques. It seems to depend on

REGENERATED SPLENIC AND LYMPH NODE TRANSPLANTS

25 1

FIG. 9. (a) Regenerated lymph node tissue, the sinuses filled with Berlin blue dye injected into the draining area of the lymph node. All typical lymph node compartments are developed. (b) An empty connective tissue network [lower left comer of (a) enlarged]. This was probably not repopulated by lymphocytes during the regenerative process. F: follicle; P: paracortex; S: sinus; T trabecula. (a) X 57, (b) x 152. S.C.

252

R. PABST et al.

a few surviving reticular cells which provide a suitable microenvironment for regeneration. During regeneration granulocytes, monocytes, and lymphocytes immigrate into the implanted tissue. Thus, both the surviving cells and the immigrating cells of the host play a role in regeneration, forming the stromal elements and the free-moving cells, respectively. It is still unclear whether the arterial and venous structures originate from the implanted tissue or from ingrowing vessels of the host. This question could be addressed immunohistologicallyby using animals with different alloantigens. In rodents after about 2 weeks all necrotic tissue has been replaced either by functioning tissue or by fibrotic material. Three to 6 months later the process of regeneration is complete and all structures of a normal spleen can be seen. The factors terminating this development are largely unknown. During regeneration the red pulp is visible first, after which the white pulp appears; the factors regulating this order have yet to be determined. In the white pulp the PALS is formed first, followed by follicles and the marginal zone, and finally germinal centers. This is very similar to the ontogenetic development of the spleen. However, the final mass of the regenerated splenic tissue is much smaller than that of a normal spleen and the relative size of the compartments does not resemble that in the normal spleen. The fact that the regenerated tissue contains all elements of a normai spleen is probably the reason for the misleading notion in so many publications which report “completely normal histology” in the regenerated splenic tissue. The final size of these compartments seems to be regulated independently from the time of appearance of the different compartments and to depend very much on the age at implantation. In young animals the structure of the regenerated tissue is similar to that of a normal spleen. Other factors apart from age (e.g., the presence of a normal spleen) iduence the process of regeneration. However, the regulating mechanisms are poorly understood. The regenerated splenic tissue performs most of the functions of a normal spleen but at a much reduced level. Therefore, the key question “Can it prevent infections?” has to be answered with caution. Protective function does not solely depend on the amount of regenerated tissue but rather on its histologic structure. The regenerated splenic tissue seems to be effective in malaria infections. Here, the protozoa are mainly removed in the red pulp, which regenerates early and attains a reasonable size. In contrast, less protection is afforded against lethal sepsis caused by pneumococci, probably because the marginal zone, known to be of paramount importance in this case, is poorly developed in most regenerated tissue. Thus, future studies should investigate whether it is possible to improve the regeneration of splenic implants. In addition, it would be of

REGENERATED SPLENIC AND LYMPH NODE TRANSPLANTS

253

great value to find parameters which are easy to determine in experimental animals and man for predicting the functional capacity of the transplants. Although fewer data are as yet available on lymph node regeneration, it is quite clear that this process is comparable to splenic regeneration. However, the implanted lymph node tissue shows fatty necrosis and the following regeneration depends very much on the site of implantation. After regeneration the tissue is not only connected to arteries and veins, but lymph flow is also reestablished and high endothelial venules develop. Thus, it is possible to demonstrate the functional capacity of the regenerated lymph node tissue by recording the development of germinal centers after subcutaneous injection of killed bacteria. It is hoped that the newly formed lymphatics might drain the lymph in lymph edema after lymphatics have been destroyed by surgical removal of axillary lymph nodes and irradiation, as in patients with mammary tumors. The regeneration of splenic and lymph node tissue provides interesting models for studying general biological phenomena in cellular and humoral interactions in regeneration. In addition, it is possible that these tissues might be used in the clinical situation to reduce the incidence of lethal infections and relieve lymph edema for the benefit of future patients.

ACKNOWLEDGMENTS We would like to thank I. Dressendorfer, M. Peter, S. Schlecht, P. Schwarz, and A. Skupin for their technical help in our own studies and for preparing the figures. We are particularly grateful to S. Fryk for meticulous secretarial help and correcting the English manuscript. Our own studies were supported by a number of grants from the Deutsche Forschungsgemeinschaft.

REFERENCES Ackerman, K. D., Felten, S. Y.,Dijkstra, C. D., Livnat, S., and Felten, D. L. (1989). Exp. Neurol. 103,239-255. Alvarez, S. R., Fernandez-Escalante, C., Rituerto, D. C., Lopez-Espadas, F. L., and Fernandez, J. M. (1987). Int. Surg.72, 149-153. Alwmark, A., Bengmark. S., Gullstrand, P., Idvall, I., and Schalen, C. (1983). Eur. Surg. Res. 15,217-222. Ambrus, J. L., Ambrus, C.M., Pickren, J. W., Amos, D. B., Neter, E., and Heim, J. (1964). Ann. N Y Acad. Sci. 113,898-914. Amlot, P. L., and Hayes, A. E. (1985). Lancet 1, 1008-101 1. Amlot, P. L., Grennan, D., and Humphrey, J. H. (1985). Eur. J . Immunol. 15,508-512, Bergmann, L., Boettcher, W., Seufert, R. M., and Mitrou, P. S. (1990). Arch. Orrhop. Trauma Surg. 109, 102-105. Billiar, T. R., West, M. A., Hyland, B. J., and Simmons, R. L. (1988). Arch Surg. 123, 327-332.

254

R. PABST et al.

Binns, R. M.,and Pabst R. (1988). In “Migration and Homing of Lymphoid Cells” (A. J. Husband, ed.), Vol. 11. pp 137-174. CRC Press, Boca Raton, Florida. Bohnsack, J. F., and Brown, E. J. (1986).Annu. Rev. Med. 37,4949. Bottcher, W., and Seufert, R. M.(1985). In “Die Splenektomie und ihre Alternativen” (M. Diirig, and F. Harder, eds.). pp. 64-71. Huber, Bern. B6ttcher. W., Seufert, R. M.,Heusermann, U.,and Munz, D. (1981). In “Chirurgisches Forum ‘81 fur Experimentelle und Klinische Forschung” (H. Junghanns, ed.), pp. 2 1 1-21 5. Springer-Verlag. Berlin-Heidelberg. Bowman, C. A., Wyck, D. B., Witte, M.H., and Witte, C. L. (1977). J . Am. Med. Women’s Assoc. 32,409-419. Bradshaw, P. H., and Thomas, C. G. (1982).J. Surg. Res. 32, 176-181. Brede, U., Tenner-Racz, K., Janossy, G., Letvin, N., and Racz, P. (1991). In “Lymphatic tissues and in Vivo Immune Responses” ( S . Ezine, S. Berrih-Aknin, and B. A. Imhof, eds.). M.Dekker, New York. Brodsky, E., Steiner, Z., Manor, Y., Schneider. M..Druker, I., and Klajman, A. (1988).Isr. J . Med. Sci. 24,702-705. Brown, E. J., Hosea, S. W., and Frank, M.M. (1981).J. CIin. Inuest. 67,975-982. Buntain, W. L., and Gould, H. R. (1985). WorldJ. Surg. 9,398-409. Buyssens, N., Paulus, G., and Bourgeois, N. (1984). Virchows Arch. (Pathol. Anat.) 403, 27-40. CaUery, M.P.,Bennett, J. A., and McKneally, M.F. (1989). Curr. Surg. 46,207-210. Chattopadhyay, B. (1989). Er. 1.Hosp. Med. 41, 172-174. Chaudry, I. H., Tabata, Y., Schleck, S., and Baue, A. E. (1980).J. Trauma 20,649-656. Church, J. A., Mahour, G. H.. and Lipsey, A. I. (1981).J. Surg. Res. 31,343-346. Claassen, E.. Kors, N., Dijkstra, C. D., and Rooijen, N. (1986). Immunology 57,399-403. Claassen, E., Westerhof, Y., Versluis, B.. Kors, N., Schellekens, M.,and Rooijen, N. (1988). Er. J . Exp. Purhol. 69,865-875. Claassen, E., Ott, A., Boersma, W. J. A., Deen, C., Schellekens, M.M.,Dijkstra, C. D., Kors, N., and Rooijen, N. (1989). Ciin. Exp. Immunol. 77,445-451. Clodius, L., and Gibson, T. (1983).In “Lymphangiology” (M. Foldi and J. R. Casley-Smith, eds.), pp. 683-706. Schattauer, Stuttgart. Cooney, D. R.. Dearth, J. C., Swanson, S. E., Dewanjee, M.K., andTelander, R. L. (1979). Surgery 86,561-569. Cooper, M.J . . and Williamson, R. C. N. (1984). Br. J . Surg. 71, 173-180. Cortes, J. A.. and PeUico, L. G. (1988). Surg. Radiol. Anat. 10,323-332. Crosby, W. H., and Benjamin, N. R. (l%l). Am. J. Pathol. 39, 119-127. Dawes, L. G . , Malangoni, M.A., Spiegel, C. A., and Schiffman, G. (1985).J. Surg. Res. 39, 53-58. Dawson, D. L., Molina, M. E., and Scott-Conner, C. E. H. (1986).Am. Surg. 52,253-256. Dickerman, J. D., Homer, S. R., Coil, J. A., and Gump, D. W. (1979). Blood 54,354-358. Dijkstra, C. D., and Diipp, E. A. (1983). Cell Tissue Res. 229,351-363. Dijkstra, C. D.. and Langevoort, H. L. (1982). Cell Tissue Res. 222.69-79. Dijkstra. C. D., Wpp, E. A.. and Langevoort, H. L. (1983). Cell Tissue Res. 229,97-107. Dijkstra, C. D., Diipp, E. A., Joling, P., and Kraal, G. (1985a) Immunology 54,589-599. Dijkstra, C. D., Vliet, E., Diipp, E. A., and Lelij. A. A. (1985b). Immunology 55,23-30. Drayson, M.T., and Ford, W. L. (1984). lmmunobiology 168,362-379. Drew, P. A., KUoff,G. K.,Fenante,A., andCohen, R. C. (1984).J. Immunol. l32,191-1%. Dugan, M.C., Grogan, T. M.,Witte, M.H., Rangel, C., Richter, L., Witte, C. L., and Wyck, D. 9. (1986). SEM 11,607613. Dukor, P.,Miller, J. F. A. P., House, W., and Allman, V. (1965). Transplantution3,639-668.

REGENERATED SPLENIC AND LYMPH NODE TRANSPLANTS

255

Dumont, A. E., and Martelli, A. B. (1978). Anat. Rec. 192,261-268. Diirig, M., and Harder, F. (1986). Chirurg 57, 189-193. Diirig, M., Landmann, R. M. A., andHarder, F. (1984). J. Lab. Clin. Med. 104,ll0-115. Eikelenboom, P., Dijkstra, C. D., Boorsma, D. M., and Rooijen, N. (1985). Experientia 41, 209-215. Ein, S. H., Shandling, B., and Simpson, J. S. (1977).Ann. Surg. 185,307-310. Eraklis, A. J., and Filler, R. M. (1972). J. Pediatr. Surg. 7,382-388. van Ewijk, W., and Nieuwenhuis, P. (1985). Experientia 41, 199-208. Falk, S., Mix, D., and Stutte, H. J. (1990). Virchows Arch. (Purhol. Anat.) 416,437-442. Faller, A. (1985). Experientia 41, 164-167. Fasching, M. C., and Cooney, D. R. (1980).J . Surg. Res. 28,449-459. Felle, P., and Folan, J. C. (1988). J. Anur. 161,236. Felle, P., and Harding, B. (1988). Eur. Surg. Res. 20,220-224. Fellows, I. W., Hart, S., and Toghill, P. J. (1988). Postgrad. Med. J . 64,267-270. Felten, D. L., Ackerman, K. D., Wiegand, S.J., and Felten, S. Y. (1987). J . Neurosci. Res. 18,28-36. Felten, S. Y., Felten, D. L., Bellinger, D. L., Carlson, S. L., Ackerman, K. D., Madden, K. S., Olschowka, J. A., and Livnat, S. (1988). Progr. Allergy 43, 14-36. Ferrante, A., Drew, P. A., Kiroff, G. K., and Zola, H. (1987). Clin. Exp. Immunol. 70, 158- 163. Fujita, T., Kashimura, M.,and Adachi, K. (1982). SEMI, 435-444. Garcia-Porrero, J. A., and Lemes, A. (1988).Acfa Anar. l31,276-283. Garnham, P. C. C. (1970).Acra Trop. 27, 1-14. Gray, D., MacLennan, I. C. M., Bazin, H.. and Khan, M. (1982). Eur. J . Immunol. U, 564-569. Green, J. B., Shackford, S. R., Sise, M.J., and Fridlund, P. (1986). J . Trauma 26,999-1004. Grogan, T . M.. Jolley, C. S., and Rangel, C. S. (1983). Lymphology 16,72-82. Grogan, T. M., Rangel, C. S., Richter, L. C., Wirt, D. P., and Villar, H. V. (1984).Lymphology 17961-68. Gudat, F., Dung, M.,Landmann, R., and Harder, F. (1985) 1n“Die Splenektomie und ihre Alternativen” (M. Diirig and F. Harder, eds.), pp. 93-101. Huber, Bern. Harding, B., Kenny, F., Given, F., Murphy, B., and Lavelle, S. (1987).Eur. Surg. Res. 19, 135-139. Hartwig, H., and Hartwig, H. G. (1985). Experienria 41, 159-163. Hebert, J. C. (1989a).J. Trauma 29,1217-1221. Hebert, J. C. (1989b). J. Surg. Res. 47,283-287. Hebert, J. C., Fisher, J. M.,and Ershler, W. B. (1989).J. Trauma 29,355-359. Hendriks, R. R., Eestermans, I. L., and Hoefsmit, E. C. M. (1980). Cell Tissue Res. 211, 375-389. Henneking, K., and Dobroschke, J. (1987). Chirurg 58, 124-126. Heusermann, U. (1988). “Morphology and Function of the Human Spleen. Enzyme Histochemical and Ultrastructural Analysis of the Splenic Lymphatic Vessels, Nerves and Connective Tissue.” Gustav Fischer, Stuttgart-New York. Holroyde, C. P., and Gardner, F. H. (1970). Blood 34,566-575. Holroyde, C. P., Oski, F. A., and Gardner, F. H. (1%9). N. Engl. J . Med. 281,516-520. Holschneider, A. M., Daeumling, S., Strasser, B., and Belohradsky, B. H. (1982). 2. Kinderchir. 35, 145-152. Horton, J., Ogden, M. E., Williams, S., and Coln, D. (1982). Ann. Surg. 195,172-176. Humphrey, J. H. (1985). Immunol. Leu. 11, 149-152. Humphrey, J. H., and Grennan, D. (1981). Eur. J. Zmmunof. 11,221-228.

256

R. PABST eta!.

Jacob, H. S., MacDonald, R. A.. and Jandl, J. H. (1963).J . Clin. Invest. 42, 1476-1490. Jaffe, H. L., and Richter, M. N. (1928).J. Erp. Med. 47,977-980. Johnson, J., and Weiss, L. (1989).Am. J . Anar. 185,89-100. Kaiserling, E.,Wolburg, H., and Banks, P. (1989).Virchows Arch. (Cell Pathol.) 58, 15-25. Kashimura, M. (1985).Arch. Histol. J p n . 48,279-291. Kim,K. Y.,Choi,J. W.,Sohn,Y.M.,andChung, K.S.(1980). YunseiMed.J.21,110-115. Kiroff, G.K.,Mangos, A., Cohen, R., Chatterton, B. E., and Jamieson, G. G. (1983).Aust. NZ 1.Surg. 53,431-434. Kiroff, G. K., Hodgen, A. N., Drew, P. A., and Jamieson, G. G. (1985).Clin. Exp. Immunol. 62948-56. Klaus, G . G . B., Humphrey, J. H., Kunkl. A., and Dongworth, D. W.(1980).Irnrnunol. Reu. 53, 3-28. Knake, E. (1952).Virchows Arch. 321,508-516. Kotze, H.F., Heyns, A., Wessels, P., Pieters, H., Badenhorst, P. N., and Loetter, M. G. (1986).Scand. J . Huematol. 37,259-264. Kovacs. K. F., Caride, V. J., and Touloukian. R. J. (1981).Arch. Surg. 116,335-336. Kraal, G.,Hart, H.T., Meelhuizen, C., Venneker, G., and Claassen, E. (1989).Eur. J . Immunol. 19,675-680. Krasna. 1. H., and Thompson, D. A. (1985).J . Pediutr. Surg. uI,30-33. van Krieken, J. H. J. M., and te Velde, J. (1986).Histopathology 10,285-294. van Krieken, I. H. J. M., and te Velde, J . (1988).Am. J . Surg. Pathol. U ,777-785. van Krieken, J. H. J. M., and te Velde, J. (1991). In “Histology for Pathologists” (S. S. Sternberg, ed.). Raven Press, New York. van Krieken, J. H. J. M., te Velde. J., Hermans, J., Cornelisse, C. J., Welvaart, C., and Ferrari, M. (1983).Histoparhology 7,767-782. van Krieken, J. H.1. M., te Velde, J., Hermans, J., and Welvaart, K. (1985).Histopathology 9,401-416. van Krieken, J. H. J. M., Schilling, C., Kluin, P. M.,and Lennert, K. (1989).Hum. Parhol. 20,320-325. Lau, H.T., Hardy, M. A., and Altman, R. P. (1983).J . Surg. Res. 34,568-571. Leonard, A. S., Giebink, G. S., Baesl, T. J., and Krivit, W. (1980).WorldJ. Surg. 4,423-432. Likhite, V. V. (1978).Exp. Hematol. 6,433-439. Liu, Y. J., Oldfield, S., and MacLennan, 1. C. M.(1988).Eur. J . Immunol. 18,355-362. Livaditis, A.,and Sandberg, G. (1980). Z. Kinderchir. 29, 148-157. Livingston, C.D., Levine, B. A., Lecklitner, M.L., and Sirinek, K. R. (1983a).Arch. Surg. 118,617-620. Livingston, C. D., Levine, B. A., and Sirinek, K. R. (1983b).Arch. Surg. 118,458-461. Livingston, C. D.. Levine, B. A., and Sirinek, K. R. ($983~). Surg. Gynecol. Obstet. 156, 761-766. Liidtke, F. E., Mack, S. C., Schuff-Werner, P., and Voth, E. (1989).Acta Chir. Scand. US, 533-539. Malangoni, M. A., DiUon, L. D., Klamer, T. W., and Condon, R. E. (1984).Surgery 96, 775-782. Malangoni, M. A., Dawes, L. G., Droege, E. A., and Almagro, U. A. (1985a). Ann. Surg. Un,323-328. Malangoni, M.A., Dawes, L. G., Droege, E. A., Rao, S. A., Collier, B. D., and Almagro, U. A. (1985b). Arch. Surg. 1u),275-278. Malangoni, M. A., Evers, B. M., Peyton, J. C., and Wellhausen, S. R. (1988).Am. J . Surg. 155,298-302. Manley, 0 .T., and Marine, D. (1917).J . Exp. Med. 25,619-628.

REGENERATED SPLENIC AND LYMPH NODE TRANSPLANTS

257

Marine, D., and Manley, 0. T. (1920). J. Exp. Med. 32,113-130. Marmorston-Gottesman, J., and Gottesman, J. (1928). Arch. Pathol. 6,406-417. Matsuno, K., Fujii, H., and Kotani, M. (1986). Cell Tissue Res. 246,263-269. Matsuno, K., Ezaki, T., and Kotani, M. (1989). Cell Tissue Res. 257,459-470. Mebius, R. E., Streeter, P. R., and Kraal, G. (1991). Zn “Lymphatic tissues and in Viuo Immune Responses” (S. Ezine, S. Bemh-Aknin, and B. A. Imhof, eds.). M. Dekker, New York. Metcalf, D. (1%3a). Austr. J. Exp. Biol. Med. Sci. 41,51-60. Metcalf, D. (1%3b). Austr. J . Exp. Biol. Med. Sci. 41,437-441. Metcalf, D., and Wakonig-Vaartaja, R. (1964). Lancet 1, 1012-1014. Millikan, J. S., Moore, E. E., Moore, G. E., and Stevens, R. E. (1982). Am. J. Surg. 144, 711-716. Mizrahi, S., and Barzilai, A. (1988). Zsr. J. Med. Sci. 24,706-709. Moore, F. A., Moore, E. E., Moore, G. E., and Millikan, J. S. (1984). Am. J. Surg. 148, 800-805.

Moore, G. E., Stevens, R. E., Moore, E. E., and Aragon, G. E. (1983). Am. J. Surg. 146, 4 13-414. Moore, M. T. F., Leong, A. S. Y.,Drew, P. A., Kiroff, G. K., and Jamieson, G. G. (1986). Virchows Arch. (Pathol. Anat.) 409,693-704. Moxon, E. R., and Schwartz, A, D. (1980). Blood 56,842-845. Nielsen, J. L., Ellegaard, J., Marquersen, J., and Hansen, H. H. (1981). Scand. J. Haematol. 27951-56. Nielsen, J. L., Andersen, H. M. K., Hansen, K. B., Sakso, P., Knstensen, E. S., and Sorensen, F. H. (1983). Scand. J. Haematol. 30,367-373. Nielsen, J. L., Sakso, P., Sorensen, F. H., and Hansen, H. H. (1984). Acta Chir. Scand. 150, 469-473, Oldfield, S . , Jenkins, S., Yeoman, H., Gray, D., andMacLennan, I. C. M.(1985). Clin. Exp. Immunol. 61,664-613. Orda, R., Barak, J., Baron, J., Spirer, Z., and Wiznitzer, T. (1981). Ann. Surg. 194,771-774. Pabst, R. (1988a). Zmmunol. Today 9,43-45. Pabst, R. (1988b). I n “Migration and Homing of Lymphoid Cells” (A. J. Husband, ed.), pp. 63-84. CRC Press, Boca Raton, Florida. Pabst, R., and Kamran, D. (1986). J. Pediatr. Surg. 21, 120-124. Pabst, R., and Reilmann, H. (1980). Cell Tissue Res. 209, 137-143. Pabst, R., and Rothkotter, H. J. (1988). Cell Tissue Res. 251,597-601. Pabst, R., and Westermann, J. (1987). Pediatr. Surg. I n t . 2,161-163. Pabst, R.. Hafke, R., and Hillebrand, J. (1985a). J. Trauma 25,326-328. Pabst, R.,Hafke, R., and Hillebrand, J. (1985b).Adu. Exp. Med. Biol. 186,565-569. Pabst, R., Peter, M.,and Westermann, J. (1985~).IRCS Med. Sci. U,360-361. Pabst, R., Westermann, J., Hillebrand, J., and Hatke, R. (1988). Pediatr. Surg. Znt. 3, 338-342. di Padova, F., Dung, M., Wadstrom, J., and Harder, F. (1983). Br. Med. . I 287,1829-1832. . di Padova, F., Dung, M., Harder, F., di Padova, C., and Zanussi, C. (1985). Br. Med. J. 290, 14-16. Patel, J., Williams, J. S., Shmigel, B., and Hinshaw, J. R. (1981). Surgery 90, 683-688. Patel. J., Williams, J. S., Naim, J. O., and Hinshaw, J. R. (1982). Surgery 91,638-641. Patel, J. M., Williams, J. S., Naim, J. 0.. and Hinshaw, J. R. (1986). J. Pediatr. Surg. 21, 877-880. Pearson, H. A., Johnston, D., Smith, K. A., and Touloukian, R. J. (1978). N . Engl. J. Med. 298,1389-1392.

R. PABST et al.

258

Pedersen, F. K. (1983). Acta Paediatr. Scand. 72,589-595. Perla, D. (1936). Am. J. Parhol. U,665-675. Peny, J. F. (1988). Curr. Probl. Surg. 25,749-832. Pirnpl, W., and Wayand, W. (1983). Acta Chir. Ausr. 1, 1-8. Pirnpl, W., Thalhamer, J., and Pattermann, M. (1987). Eur. Surg. Res. 19,53-61. Pouche, A., Savoldi, F., Colombi, A., Tiberio, G., and Turano, A. (1987). Eur. Surg. Res. 19, 86-90.

Powell, R. W., Blaylock, W. E., Hoff, C. J., and Chartrand, S. A. (1988a). J. Surg. Res. 45, 56-59.

Powell, R . W . , Blaylock, W . E., €€off, C . J . , and Chartrand, S . A . (1988).J . Surg. Res. 45, 56-59.

Redmond, H. P., Redmond, J. M., Rooney, B. P.. Duignan, J. P., and Bouchier-Hayes, D. J. (1989). Br. J . Surg. 76, 198-201. Reilmann, H., Pabst, R., and Creutzig, H. (193). Eur. Surg. Res. 15, 168-175. Rothkotter, H. J., and Pabst, R. (1990). Scand. J . Plasr. Reconsrr. HandSurg. 24,101-105. Rudowski, W. J. (1985). W o r f d I .Surg. 9,422-430. Sakso, P., Hansen, K. B.. Nielsen, J. L., and Sorensen, F. H. (1985). Acta Chir. Scand. 151, 409-411.

Sasaki, K. (1986). Virchows Arch. (Parhol. Anat.) 409,325-334. Sasaki, K.(1988). Anat. Anz. 166,297-304. Sasaki, K. (1989). Clin. Anar. 2, 175-1%. Sasaki, K. (199Oa). Acta Anor. l39,315-319. Sasaki, K. (199ob). Acto Pathot. Microbiol. Immunol. Scand. 98,507-513. Sass, W., Bergholz, M., Kehl, A., Seifert, J., and Hamelmann, H. (1983). Klin. Wochenschr. 61, 1075-1079.

Scher, K. S., Scott-Conner, C., Jones, C. W., and Wroczynski, A. F. (1985a). Ann. Surg. 202,595-599.

Scher, K. S., Wroczynski, A. F., and Scott-Conner, C. (1985b). Am. Surg. 51,269-271. Schmidt, E. E., MacDonald, I. C., and Groom, A. C. (1988). Am. J . Anat. 181,253-266. Schr6der. L. (1985). “Tierexperimentelle Untersuchungen zur Morphologie der autologen Milzimplantation Medizin.” Universitiit Kiel, Kiel, Germany. Schwartz, A. D., Dadash-Zadeh, M., Goldstein, R., Luck, S., and Conway, J. J. (1977). Blood 49,779-783. Sekikawa, T., and Shatney, C. H. (1983). A m . J . Surg. 145,667-673. Seufert, R. M. (1986). “Surgery of the Spleen.” Thierne, New York. Seufert, R. M., Bottcher, W., Munz, D.. and Heusermann, U. (1981). Chirurg 52,525-530. Severin, C. M. (1989). I n “Disorders of the Spleen” (C. Pochedly, R. H. Sills, and A. D. Schwartz, eds.), pp. 1-19. Dekker, New York. Shaw, J. H. F., and Print, C. G. (1989). Br. J. Surg. 76,1074-1081. Shennib, H., Chiu, R. C. J.. and Mulder, D. S. (1983). J . Trauma U,7-12. Sherman, R. (1980). 1. Trauma 25, 1-13. Singer, D. B. (1973). Perspecr. Pediarr. Parhol.1, 285-31 I. Spencer. G. R., Bud, C., httiero, D. L.,Br0wn.T. R., Mackenzie, F. A. F., andphiliips, M. J. (1981). Br. J. Surg. 68,412-414. Steely, W. M., Satava, R. M., Brigham, R. A., Setser, E. R., and Davies, R. S. (1988). J . Surg. Res. 45,327-332. Stutrnan, 0.. and Zingale, S. B. (1964). Proc. SOC. Exp. Med. Biol. 117, 389-393. Stutte, H. J., Panvaresch, M. R., and Rossius, H. (1974). Res. Exp. Med. 163,79-94. Szakai, A. K., Kosco, M. H., and Tew. J. G. (1989). Annu. Rev. Immunol. 7,91-109.

REGENERATED SPLENIC AND LYMPH NODE TRANSPLANTS

259

Tanaka, H., Takasaki, S., Muroya, T., Suzuki, T., and Ishikawa, E. (1984a). Acta Histochem. Cytochem. 17,339-358. Tanaka, H., Takasaki, S., Sakata, A., Muroya, T., Suzuki, T., and Ishikawa, E. (1984b). Acta Pathol. Jpn. 34,251-270. Tarnuzi, A., and Smiley, R. K. (1%7). Blood 29,373-384. Tavassoli, M. (1975). Blood 46,631-635. Tavassoli, M.,Ratzan, R. J., and Crosby, W. H. (1973). Blood 41,701-709. Tavassoli, M., Sacks, P. V., and Crosby, W. H. (1975). Proc. SOC. Exp. Biol. Med. 148, 780-783. Thalhamer, J., Pimpl, W., and Pattermann, M. (1986). Res. Exp. Med. 186,229-238. Thalhamer, J., Lenglachner, C., Grillenberger, W., and Pimpl, W. (1989a). Ann. Surg. 210, 630-636. Thalharner, J., Lenglachner, C., Grillenberger, W., and Pimpl, W. (1989b). Leuk. Res. l3, Suppl. 1,39. Thalhamer, J. et al. (1991). submitted. Timens, W., and Poppema, S. (1985). Am. J. Pathol. UO, 443-454. Timens, W., Rozeboom, T., and Poppema, S. (1987). immunology 60,603-609. Timens, W., Boes, A., Rozeboom-Uiterwijk, T., and Poppema, S. (1988). Virchows Arch. (Parhol. Anat.) 4l3,563-571. Timens, W., Boes, A., and Poppema, S. (1989a). Eur. J. Immunol. 19,2163-2166. Timens, W., Boes, A,, Rozeboom-Uitenvijk, T., and Poppema, S . (1989b). J. immunol. 143, 320-3206. Tischendorf. F. (1%9). in “Handbuch der Mikroskopischen Anatomie des Menschen” (W. Bargmann, ed.), pp. 1-968. Springer, Berlin. Tischendorf, F. (1985). Experienria 41, 145-152. Todorovic, R., Ferris, D., and Ristic, M. (1967). Exp. Parasitol. 21,354-372. Traub, A., Giebink, G. S., Smith, C., Kuni, C. C., Brekke, M. L., Edlund, D., and Perry, J. F. (1987). N . Engl. J . Med. 317,1559-1564. van der Valk, P., Loo, E. M., Jansen, J., Daha, M. R.,and Meijer, C. J. L. M. (1984). Virchows Arch. (Cell. Pathol.) 45, 169-185. Vega, A., Howell, C., Krasna, I., Campos, J., Heyman, S., Ziegler, M., and Koop, C.E. (1981). J. Pediatr. Surg. 16,898-904. Velcek, F. T., Jongco, B., Shaftan, G. W., Klotz, D. H., Rao, S. P., Schsman, G., and Kottmeier, P. K. (1982a). J. Pediatr Surg. 17,879-883. Velcek, F. T., Kugaczewski, J. T., Jongco, B., Shaftan, G. W., Rao, P. S.,Schsman, G., and Kottmeier, P. K. (1982b). J. Trauma 22,502-506. Vellguth, S., Gaudecker, B., and Mueller-Hermelink, H. K. (1985). Cell Tissue Res. 242, 579-592. Wahlby, L., and Domellof, L. (1981). Acta Chir. Scand. 141, 131-135. Walker, W. (1976). Br. J. Surg. 63,36-43. Warner, T. F. C. S., and Krueger, R.G. (1975). Br. J. Haematol. 31,405-409. Weihe, E., Westermann, J., and Pabst, R. (1991). in preparation. Weiss, L. (1988). I n “Cell and Tissue Biology. ATextbookof Histology” (L. Weiss, ed.), 6th ed., pp. 516-538. Urban and Schwarzenberg, Baltimore, Maryland. Weiss, L., Powell, R., and SchSman, F.J. (1985). Experientia 41,233-242. Weiss, L., Geduldig, U., and Weidanz, W. (1986). Am. J. Anat. 176,251-285. West, K. W., and Grosfeld, J. L. (1985). World J. Surg. 9,477-483. Westermann, J., and Pabst, R.(1986). Clin. Exp. Zmmunol. 64, 188-194. Westermann, J., and Pabst, R. (1988). Res. Exp. Med. 188,267-276.

260

R. PABST et at.

Westennann, J., and Pabst, R. (1990). Immunol. Today 11,406-410. Westennann, J., Peschel, P., and Pabst, R. (1988a). Cell Tissue Res. 254,403-413. Westermann, J.. Willfiihr, K. U., and Pabst, R. (1988b). J. Pediatr. Surg. 23,835-838. Westemann. J., Schwinzer, R., Jecker, P., and Pabst, R. (1990). Scand. J. Immunol. 31, 327-334. Wiilfiihr, K.U.,Westennann, J., and Pabst, R. (1991). J . Pediatr. Surg., submitted. Willfuhr, K. U., Westermann, J., and Pabst, R. (1990). Eur. J . Immunol. M,903-91 I. Witte, C. L., Witte, M. H., McNeill, G. C., Hall, J. N., Werf, G. P.. and Woolfenden, J.M. (1988). Am. J . Surg. 155, 303-310. Witte, M.H., Witte,C. L., Wyck, D. B., and Farrell, K.J. (1983). Lymphology 16,128-137. Witte, T., Wordelmann, K., and Schmidt, R.E. (1989). Immunology 69, 166-170. Wolf, N.S.(1978). Cell Tissue Kinet. 11,335-345. Wolf, N.S. (1982a). Exp. Hematol. 10, 108-1 18. Wolf, N . S. (1982b). Exp. Hemarol. 10,98-107. van Wyck, D. B., Witte, M.H., and Witte. C.L. (1986). Clin. Sci. 71,573-579. Wyler. D. J. (1983). Lymphology 16, 121-127. Wyler, D. J., Quinn, T. C., and Chen, L. T. (1981).J . Clin. Invest. 67, 1400-1404. Wysocki, H., Wierusz-Wysocka, B., Karon, H., Dotka, J., Wykretowicz, A., Szczepanik, A., and Klimas, R. (1989). Clin. Exp. Immunol. 75, 392-395. Zambi, M.H., and Rosner, F. (1984). Arrh. Intern. Med. 144, 1421-1424. Zhang, J.. Liu, Y.J., MacLennan, 1. C. M.,Gray, D., and Lane, P. J. L. (1988). Eur. J . Immunol. 18, 1417-1424.

INTERNATIONAL REVIEW OF CYTOLOGY, VOL.128

Molecular Biology in Studies of Ocean Processes PAULG. FALKOWSKI AND JULIE LAROCHE Oceanographic and Atmospheric Sciences Division, Brookhaven Na 1ional Laboratory, Upton, New York 11973

I. Introduction Throughout much of the 20th century, biological oceanographers have catalogued the spatial and temporal distributions of organisms in the oceans. Their goals have been to understand how biological and nonbiological processes affect these distributions, and to determine how individual species, communities, or trophic levels affect or are affected by changes in the physical or chemical environment of the ocean. Progress has advanced slowly and regularly over the years. However, large leaps in understanding have come from the application of new techniques and technologies. Three new tools became available to biological oceanographers in the 1980s: remote sensing, which provided satellite images of processes on an ocean-basin scale; moored instrumentation in situ,which provided time series of coupled biological and physical processes in a moving fluid; and molecular biological techniques, which allowed investigation of mechanisms of biological phenomena, gene structure, organization, and regulation. These three tools span 15 orders of spatial magnitude. Application of remote sensing techniques and in situ instruments has provided insight into large spatial and temporal distributions of phytoplankton and zooplankton in the upper ocean in relation to physical processes (Falkowski et al., 1991). While the application of molecular techniques is still in its infancy, it is hoped that these techniques will provide a basis for mechanistically understanding the biological basis for the distribution and function of oceanic organisms. Within this context we review some of the problems which biological oceanographers are addressing with molecular biological techniques and discuss the advances in understanding of ocean processes that the application of such techniques has brought. We broadly define molecular biology as the study of the structure, functional organization, and regulation of nucleic acids and proteins. We broadly define ocean processes to include biogeochemical cycles, the evolution of marine organisms, population dynamics, systematics, and the physiological adaptation of organisms to a fluctuatingocean environment. 26 1 Copyright 8 1991 by Academic Press. Inc. All rinhrs of reproduction in any form resewed.

'262

PAUL G. FALKOWSKI AND JULIE LaROCHE

The elucidation of the physical structure of DNA and its relationship to protein synthesis provided a rational basis for understanding many biological processes at fundamental levels. The application of molecular techniques to biological oceanography, however, has lagged behind other areas of biology. We identify at least three causes for this lag: (1) a difficulty in applying many standard molecular biological techniques to oceanic systems, where concentrations of nucleic acids and proteins may be extremely low and their sources poorly identified; (2) a small historical data base, combined with poorly characterized genetic systems; and (3) a lack of communication between biological oceanographersand molecular biologists. At present many of our colleagues in the biological oceanographic community are beginning to apply or adapt molecular techniques to biological oceanographic problems, while our more molecularly oriented colleagues are uninformed about important issues in biological oceanography. Therefore, we will identify some of the many biological oceanographic problems which are being addressed or are potentially amenable to molecular biological methods. Our major emphasis will be on planktonic organisms and invertebrates; Powers et al. (1990) have recently reviewed molecular biological aspects of fish and other marine animals. We address the broad question of the potential value of molecular biology to biological oceanography, highlighting the techniques that we believe will be appropriate for use in the marine environment. In so focusing our discussion, we will emphasize techniques that can be modified for use at sea, techniques that are simple and low cost, and those that are relevant to major problems in oceanography.

II. Marine Organisms and Oceanographic Processes A. QUESTIONS IN BIOLOGICAL OCEANOGRAPHY

The vast majority of the biomass in the Ocean is contained in organisms smaller than 1 mm (Sherr and Sherr, 1988; Sieburth ef al., 1978). Such small organisms cannot control their horizontal position in the water column in the face of ocean currents, and, therefore, are called plankton (Greek: planktos, meaning to wander or drift). Historically, plankton have been classified into three compartments: bacterioplankton, phytoplankton, and zooplankton (Sieburth et al., 1978). Operationally, however, plankton are usually separated by size (Fig. I), based, in part, on the ability to separate organisms by filtering water samples though matrices with appropriate pores. The plankton include virus particles <0.2 p m (Bergh et al., 1989), prokaryotes, including both gram-positive and gram-

MOLECULAR BIOLOGY IN STUDIES OF OCEAN PROCESSES

PLANKTON

FEYTOPICOn w - u~mPLANK?" PLANKTOW PLANKTON PLANK?" MESOSLAM?" O.O?-O.Zpm 0.1-2.0pm 2 . 0 - Z O p 10- 200pm 0.2-20rm

263

MACRO- YEGAPLANKTOW PLANK?" I-ZOa 2W00m

yy NEKTON

z-zomm

DM Metton Z-ZM

CY Nrkton 2-2Oa

Y Nekton 2-zon,

I I i I I I I I

I

-I

IV'R'o-l+l PLANKTON

YYCOPLANKTON PnvmPLANKTON

PLANKTON

NEKTON I

1;-8

lb-7

1;-6

A-5

lb-4

10-3

I

10-2

I

10-1

10 0

101

Size (m) FIG. 1. Size-functionclassificationscheme for marine organisms (Sieburth et al., 1978).

-

negative bacteria, ranging from submicrometer sizes to 2 pm (Pace et al., 1986), a wide variety of eukaryotic cells, including photoautotrophic phytoplankton and heterotrophic protozoa, ranging from 0.1 to lo00 pm, and multicellular zooplankton, which may be several millimeters. Size, however, does not describe biological function (Sieburth et a/., 1978). There are large, colonial cyanobacteria which can fix nitrogen. There are extremely small, eukaryotic, carbon-king phytoplankton which are the size of heterotrophic bacteria (Johnson and Sieburth, 1982). There are heterotrophic microzooplankton which squeeze through filter pores smaller than the equivalent spherical diameter of the organisms. Thus, a key focus in biological oceanography is the development of operational protocols to separate and enumerate organisms on the basis of their physiological or ecological functions. The plankton community mediates the fluxes of many biogeochemically important elements, including nitrogen, hydrogen, carbon, sulfur, silica, phosphorus, and oxygen. Quantifying the organisms which are most important in mediating such fluxes requires a knowledge of what organisms are present in the water column or sediments, their life cycles, how they

-

264

PAUL G . FALKOWSKI AND JULIE LaROCHE

are distributed in space and time, their potential and their actual biological functions, their rates of growth and mortality, and the factors which control these processes. Marine biologists and biological oceanographers are also interested in population genetics, gene flow, rates of speciation, and distinguishing between closely related species, cohorts, or ecotypes. Ecologists have pondered whether stability in the environment has led to increased or decreased genetic diversity (Dunbar, 1960; Sanders, 1%8; Woodwell and Smith, 1969). The problem of quantifying the rate of gene flow and drift is of special interest to fishery biologists, who are concerned with genetic diversity within fish populations, tracking migratory behavior of cohorts, larval recruitment, and the potential for genetically improving commercial fish species (Powers et al., 1990).

B. EARLY MEASUREMENTS OF NUCLEICACIDS Early measurements of nucleic acids in the ocean were bulk measurements, and workers attempted to develop a new technique for measuring either phytoplanktonic biomass or growth rates. Holm-Hansen et al. (I=) measured DNA in the water column at several stations in the Atlantic and Pacific oceans using a fluorometric method based on the reaction of 3,5-diaminobenzoic acid with DNA. Concentrations of DNA ranged from 0.1 to 30 pg/liter, and were poorly correlated with chlorophyll (a surrogate measure of phytoplankton biomass). The authors concluded that “there is a considerablequantity of living material that is high in DNA or that DNA is associated with particulate, nonliving material.” Falkowski and Owens (1982) attempted to extend such DNA measurements, using a modification of Brunk et al.’s (1979) DAPI-staining method, to estimate rates of cell division of phytoplankton in coastal waters of the northeast United States. Their findings also indicated that varying amounts of “detrital” DNA (DNA associated with nonliving particulate material) were present. They concluded that, “Because of the difficulty in determining the proportion of nonphytoplankton DNA to the total DNA in natural phytoplankton communities, the calculation of phytoplankton division rates (from the time dependent change in DNA) is not straightforward.” The work of Sieburth and co-workers (Sieburth et al., 1978) and others (Pace et al., 1986; Sherr and Sherr, 1988; Bergh et al., 1989; Proctor and Fuhrman, 1990), revealed that a large but variable fraction of particulate organic material is associated with bacteria, microzooplankton, and virus particles. Thus simply budgeting the nucleic acids in a natural water sample remains a challenge.

MOLECULAR BIOLOGY IN STUDIES OF OCEAN PROCESSES

265

Molecular biological techniques, including the use of antibodies, gene probes, restriction fragment analyses, and polymerase chain reactions enable biological oceanographers to overcome some of the earlier problems. Specifically, using molecular techniques, biological oceanographers can address the following questions: (1) What organisms are there in the sample? (2) Are two or more morphologically related organisms the same species, different species, or ecotypes? (3) Does an organism contain genes encoding enzymes important in mediating a particular biochemical transformation (e.g., can the cell fix inorganic nitrogen or carbon)? (4) If so, are the genes constitutivelyexpressed or are they inducable, and, if the latter, under what conditions? ( 5 ) How did organisms with similar functional biochemical processes evolve? (6) How has that evolution been related to the geological history of the physical circulation and chemistry of the ocean waters? Before examining the applications of molecular biology to such questions, we will briefly review some techniques that have potential in biological oceanography. For a comprehensive review of molecular biological methods, there are several introductory texts such as Freifelder (1983), Watson et af. (1987), or Darnell et af. (1986). For a detailed description of techniques and protocols, refer to Sambrook et af. (1989).

III. Molecular Approaches to Oceanography Molecular biological approaches can be conveniently divided into three major categories: (1) isolation and characterization of nucleic acids and proteins; (2) assay of either the pool size or rate of synthesis of specific proteins, genes, or RNA molecules; and (3) either direct or indirect experimental manipulation of the organism(s) to test specific hypotheses. Each of these categories has been the subject of numerous books and reviews, and we shall only briefly describe how the methods have been adapted for, or applied to, the study of marine organisms, while directing the interested reader to some source references. A. NUCLEIC ACIDISOLATION

Most nucleic acid isolation protocols (Sambrook et al. 1989) can be adapted to marine organisms. The protocol for the extraction of RNA is similar to that of DNA; however, more care must be taken to inhibit RNAses (Maniatis et al., 1982). Table I is a partial list of marine organisms from which nucleic acids have been isolated: the data include prokaryotic

266

PAUL G. FALKOWSKI AND JULIE LaROCHE TABLE I MARINEORGANISMS FROM WHICH NUCLEICACIDSHAVEBEENISOLATED Organism

Cylindrotheca sp. strain Ni

Nucleic acid isolated

Reference

Chloroplast (cp) DNA Total RNA Total RNA

Hwang and Tabita, 1989

Chapman and Powers, 1984

Green sea turtle Atlantic eels Salmon

Mitochondrial (mt) DNA mtDNA mtDNA mtDNA

Mussel Ascideans Costaria costata

mtDNA DNA, R N A DNA

Bacteroid symbiont of the Caribbean flash-light fish (Kryptophanaron alfredi] Artemia Uronena nigricans (hymenostome ciliate) Red algae

DNA

Fremy ella diplosiphon Strongylocentrotus purpuratus eggs and embryos (sea urchin) Anabaena sp. strain PCC HzO Cryptomonas Q, Olisthodiscus luteus Total bacterioplankton Skeletonema costaturn Cylindrotheca fusiformis Codium fragile Crypthecodium cohnii Deep-sea barophilic bacteria Acroporidae (scleractinian coral) Reftia pacttiptila vestinentiferan tube worm Fish

Conley e f al., 1986 Posakony et al.. 1983; Jacobs et al., 1988 DNA Mazur et al., 1980 RNA Golden et al., 1987 cp DNA Douglas, 1988 cp DNA Aldrich et al., 1982 Total DNA Fuhrman et al., 1988; Somerville et al., 1989 RNAIDNA Medlin et al., 1988 Polysomal RNA Reeves and Volcani, 1985 cp DNA Manhart et al., 1989 Total RNA Hinnebusch et al., 1981 Ribosomal RNA Deming et al.. 1984 Genomic DNA sperm McMillan et al., 1988 from coral Ribosomal RNA Stahl et of.. 1984

mtDNA DNA RNA Plastid DNA

Bowen et al., 1989 Avise et al., 1986 Berg and Fems, 1984; Birt et al., 1986 Skibinski, 1985 Kumar et al., 1988 Bhattacharya and Druehl, 1989 and references therein Haygood and Cohn, 1986 Batuecas et al., 1988 Soldo et al., 1981

Goff and Coleman, 1988

and eukaryotic organisms, as well as invertebrates, vertebrates, and macrophyte algae. Protocols for isolating nucleic acids vary slightly from organism to organism. For example, macroalgae and metazoans require rapid freezing in liquid Nz and grinding in mortar and pestle to promote cell tysis and reduce shearing DNA. The basic methods for bacterial or yeast

MOLECULAR BIOLOGY IN STUDIES OF OCEAN PROCESSES

267

DNA isolation (Maniatis et al., 1982) should work well for most unicellular marine organisms that can be grown in culture. There are three major difficulties in simply applying the commonly used DNA isolation protocols to biological oceanographic problems: (I) collecting a sufficient amount of material from seawater; (2) preserving the material so that it maintains its structural integrity; and (3) separating the material of interest from seawater, which usually contains a large number of species. Organisms from different taxonomic groups (e.g.,prokaryotes and eukaryotes) often overlap in size (Fig. I), thereby creating a problem when mechanical filtration is used to collect the samples. The isolation and culture of many marine microbes, unicellular algae, and heterotrophic flagellates from seawater samples is extremely labor intensive, impractical, and sometimes impossible. Molecular biologists usually work with well-defined organisms or clones, and therefore are certain of the source of their isolated nucleic acids. Unless biological oceanographers restrict their research to axenic cultures of organisms, or develop rapid methods for isolating single species from natural marine microbial communities (using, for example, flow cytometers with cell sorters), the nucleic acids isolated from natural seawater samples will be representative of the entire community. Consequently, such nucleic acids will have a low abundance of specific genes and transcripts, and large amounts of nucleic acids may be required to detect rare genes, although the application of polymerase chain reaction techniques (see below) may make detection of rare genes easier. Marine microbes are small (0.4-0.6 pm diameter) and there are few in a sample compared with commonly used laboratory bacterial strains (>1 pm). Therefore it is difficult, especially at sea, to harvest microbes by centrifugation. Ultrafiltration facilitates the rapid concentration of microorganisms from large (liter) volumes of seawater with relatively simple, cheap, and easily portable equipment. Several ultrafiltration methods have been developed to isolate nucleic acids from water samples. Sommerville et al. (1989) used cartridge-type filters, such as the Sterivex-GS filter (Millipore Corp.), that can be stored at -20°C for future extraction. Current protocols report DNA yields of 1-6 ng of DNA per lo6cells (Fuhrman et al., 1988; Paul et al., 1988; Somerville et al., 1989). Thus, DNA yields of approximately 100 ng/liter of seawater filtered (assuming lo5 cells/ml for oligotrophic areas) can be obtained, an amount of DNA that theoretically is sufficient to detect specific DNA sequences from organisms representing 0.003% or less of the microbes (i.e., organisms present at a density of 1 to 3 cells per ml), using Southern blots and radioactive probes (see below) (Holben et al., 1988). DNA isolation from zooplankton or benthic crustaceans requires extensive sorting of the desired species unless the environment is one of low

268

PAUL G . FALKOWSKI AND JULIE LaROCHE

diversity where only one or two species prevail. DNA can be isolated from larger metazoans and vertebrates using slight modifications of established techniques (Drouin et af., 1987). Mitochondria1 DNA of fish is often isolated from soft tissues such as liver or gonads (Avise et al., 1986; Bentzen et al., 1988) but a trial and error approach may be required for several marine invertebrates. For example, in scallops, DNA is best isolated from the adductor muscle (Snyder et al., 1987). Blood samples or small amounts of tissues from a biopsy are often used to extract genomic DNA from larger animals (Baker et al., 1990). DNA has been isolated from coral gametes with only minor modifications of established procedures (McMillan et af., 1988).

B. ORGANELLE DNA Organelle DNA is of interest for three reasons: (i) it encodes key proteins, which are relatively well characterized; (ii) organelle genomes are much smaller and less variable than nuclear genomes, making mapping and sequencing more tractable; and (iii) often there are numerous organelles per cell, each with multiple copies of the genome, making detection of specific genes easier. Organellar DNA can be isolated in two ways: (1) directly from the organelles in species where the latter are easily separated from other cell constituents by density gradient centrifugation or (2) by separating the organellar DNA by GC content or on the basis of its conformation, usually by CsCl gradient centrifugation, or by size, using pulsed-field electrophoresis. Isolation of the organelle DNA can present considerable problems for biological oceanographers. First, the sample must be well characterized, which limits research to larger metazoans or pure cultures of microorganisms. Second, organelles may be difficult to preserve intact in frozen samples; thus, it is preferable to isolate the organelle on fresh samples. Finally, these techniques are difficult to perform aboard a ship because (among other things) of the difficulty in running a centrifuge safely unless it is gimbled (which is feasible but not common). As we shall discuss later, mitochondria1 DNA has been analyzed from a variety of marine metazoans. Analysis of chloroplast DNA is feasible on macrophyte algae (Manhart et al., 1989), but has not, to our knowledge, been performed on natural phytoplankton samples, However, chloroplast DNA has been isolated and characterized from some eukaryotic algae, such as Cryptomonas sp. (Douglas, 1988), Ocramonas danica (Reith and Cattolico, 1986), Griffithisiapacifica (Li and Cattolico, 1987), Dunaliella tertiolecta (Brown and J. La Roche, unpublished), Skeletonema costaturn (Stabile and Gallagher,

MOLECULAR BIOLOGY IN STUDIES OF OCEAN PROCESSES

269

personal communication), and Cyfindrotheca sp. (Hwang and Tabita, 1989).

C. PROTEIN ISOLATION A N D ANALYSES Until recently, biological oceanographers were more or less content to assay the total protein in a sample; polyacrylamide, agarose, or starch gel electrophoresis was rarely performed on marine samples. The reasons for this are, as for nucleic acid work, the low concentration of particulate protein in the sea and the difficulty in interpreting the results from a heterogenous mix of organisms. Nonetheless, gel electrophoresisis sometimes useful if the protein can be obtained from a known source, such as fish and larger invertebrates; or if smaller organisms can be collected and concentrated sufficiently. Natural marine plankton communities can be concentrated by either filtration or centrifugation and the proteins can be precipitated with cold TCA or acetone. The proteins then may be resolubilized and separated by electrophoresisin polyacrylamide gels (Laemmli, 1970). Quantitation of specific proteins can be accomplished by Western blots using antibodies as a probe. Western blots have two major advantages over other immunoassays: (1) the proteins are separated by molecular weight; therefore nonspecific binding or cross reactions can be readily assessed, and (2) the blots are stable indefinitely. Thus if an antibody later becomes available to a protein of interest the blot can still be assayed. A major disadvantage of Western blots is the difficulty of quantitation. Dot blots, rocket electrophoresis, and enzyme linked immunoassays (ELISA) overcome the difficulties of quantitation but these techniques can be used only once the specificity of the antibody has been assessed (Ward, 1990). Studies of protein distributions have been used to infer the spatial and temporal distributions of marine organisms. For example, Bucklin et al. (1989) isolated soluble proteins from frozen individual copepods in a Trid sucrose mix and subjected the samples to nondenaturing electrophoresis to examine variability of allozymes within cohorts. This approach can also yield information on the function of marine organisms. Enzymes such a nitrogenase, nitrate and nitrite reductase, ribulose 1,5-bisphosphate carboxylase, carbonic anhydrase, and alkaline phosphatase all catalyze key reactions in biogeochemical cycles. Biological oceanographers have historically assayed enzymatic activity, almost always in uitro, to assess key rate processes. The use of immunoassays in conjunction with enzyme assays gives an understanding of induction and posttranslational regulation. Capone et al. (1990) and Currin et al. (1990), using SDS-PAGE, examined the spatial and temporal distribution of nitrogenase and total cell

270

PAUL G . FALKOWSKI AND JULIE LaROCHE

proteins in the cyanobacterium Trichodesmium spp. collected by net tows from the Caribbean and off the North Carolina coast. In addition, polyclonal antibodies raised against nitrogenase have been used to study a posttranslational modification of the enzyme, which is required for activity (Ohki et al., in press). Care must be taken in the interpretation of immunoassays of heterologous samples. For example, the large subunit of ribulose I ,5-bisphosphate carboxylase of higher plants will cross react with the enzyme from chlorophyte algae; however, the antibodies do not react well with the enzyme from chromophyte (chlorophyll c containing) algae (Newman and Cattolico, 1987). Further, polyclonal antibodies raised against the chromophyte enzyme do not always cross react with the green algal enzyme. This lack of cross reaction was a puzzle because the large subunit of ribulose 1,5-bisphosphate carboxylase, encoded in the chloroplast, is thought to be highly conserved (Glover, 1989). Amino acid sequencing has been used to establish differences in functionally identical ribulose 1,5-bisphosphate carboxylase from different organisms (Keen et al., 1988). Thus, it may be misleading to use a polyclonal antibody raised against a putatively highly conserved protein to quantitatively assay the target protein from a mixture of organisms. Furthermore, even if the protein is reported, absolute quantification of protein by immunoassay requires the isolation of the protein from the organism. Because of variability in the number and specificity of epitopes, there is inevitably variability in antibody reactions between organisms, This problem is not critical if knowledge of only the relative abundance of the target protein is desired. There may be some advantage to using monoclonal antibodies to quantify or semiquantify proteins, but making monoclonal antibodies is labor intensive and costly.

D. PROTEIN SYNTHESIS The incorporation of [35S]rnethionineinto proteins is seldom used with marine organisms to follow protein synthesis in uitro, because [35S]methionine uptake rates are often extremely low. [35S]sulfateincorporation was used to measure protein synthesis in phytoplankton (Bates, 1981), but because the concentration of SOs in natural seawater is high (25 mM), extremely high levels of 35S and long incubation times are required for autoradiography (Pick et al., 1987). To reduce the background, plankton may be preincubated in artificial seawater with low SO4 levels (Sukenik et al., 1990); however, that is not always feasible in the field. One alternative with photoautotrophic organisms is to use I4C as NaHI4CO3(Friedman and Alberte, 1986; Henry and Falkowski, unpublished). With heterotrophic bacteria (Kirchman et al., 1986) or invertebrate metazoans (Fal-

MOLECULAR BIOLOGY IN STUDIES OF OCEAN PROCESSES

271

kowski, unpublished), ['4C]leucine is often rapidly incorporated and may be a suitable tracer for protein synthesis.

E. POLYMERASE CHAINREACTION (PCR) The polymerase chain reaction (Saiki et al., 1985; 1988) amplifies specific sequences of DNA from picogram quantities of starting materials. Conceptually, the PCR is simple; upon heating DNA, the hydrogen bonds, which maintain the structure of double-stranded DNA, are broken. The PCR method requires two oligonucleotideprimers, one complementary to the 5' end and the other to the 3' end of the sequence to be amplified. In the presence of the four-nucleotide mixture and a heat-stable DNA polymerase, DNA complementary to each strand can be made. Upon cooling, the newly synthesized DNA hybridizes to its complementary sequence. The heating and cooling cycles are repeated, each cycle amplifying the sequence by a factor of two, eventually leading to an exponential accumulation of the specific target sequence. The specificity and efficiency of the technique was greatly improved with Taq DNA polymerase isolated from a thermophyllic bacterium, Thermus aquaticus (Chien et al., 1976) which is resistant to repeated cycles of heating and cooling. The current PCR techniques allow for a 4 x lo6-fold amplification of sequences as rare as one molecule in lo6 cells and can potentially overcome many of the problems concerning the heterogeneity and low abundance of organisms in natural marine samples. PCR is useful in several different applications. It is extremely powerful in cloning rare sequences (Scharfet al., 1986; Zehr and McReynolds, 1989) and has been used to sequence genomic (Medlin et al., 1988) and organellar DNAs (Wrischnik et al., 1987), and in studies of nucleotide sequence variability (Saiki et al., 1985). A list of useful protocols and applications has been published (Innis et al., 1990). The design of specific oligonucleotides to study the sequences of interest is a factor limitingthe application of PCR to biological oceanography. Because very few genes from marine organisms have been studied in detail, one often assumes that highly conserved amino acid sequences in well-characterized proteins are conserved in homologous proteins of marine organisms. This is not always the case. For example, the Rubisco enzyme, ubiquitous among autotrophs, is not highly conserved between the different algal groups (Keen et al., 1988). However, the knowledge of conserved regions in homologous proteins, in conjunction with PCR techniques, has already been useful in the isolation of genes from marine organisms (Zehr and McReynolds, 1989). A degenerate oligonucleotideprimer strategy was used by Zehr and McReynolds (1989) to isolate and characterize a DNA fragment encoding

272

PAUL G . FALKOWSKI AND JULIE LaROCHE

the nif D gene (which encodes for an iron protein in the nitrogenase enzyme complex) from the marine cyanobacterium, Trichodesrniurn sp. The techniques described above have been used in studies of marine organisms. We continue our list below, with techniques that have been successful in research in molecular biology, but have not yet been widely used in marine research. We believe that with modifications and with time they may become powerful tools in oceanography. PROFILES OR DNA FINGERPRINTING F. RESTRICTIONENDONUCLEASE This technique has been used to distinguish between species of corals (McMillan and Miller, 1990), and may be useful in following larval recruitment and intra- and interpopulation genetic variations. Using chloroplast and nuclear gene probes from higher plants, Stabile et al. (1990) detected restriction fragment length polymorphism (RFLP) among three strains of the marine diatom Skeletonemu cosrarum, thereby supporting an earler hypothesis that genetic variation is the basis for local population diversity (Gallagher, 1982).

G. SOUTHERN AND NORTHERN BLOTS One of the most powerful techniques used by molecular biologists is hybridization of specific probes to nucleic acids immobilized on filter blots such as Southern blotting (Southern, 1972) and Northern blotting. There are some drawbacks with Southern and Northern blots: they are not very sensitive to minor sequence changes, such as point mutations, and the specificity of the probes must be well established. Heterologous DNA probes are in general less successful than heterologous antibody reactions because of difference in species-specific codon biases and the degeneracy of the genetic code. For example, different codon bias in chlorophytic algae and higher plants prevents cross-hybridization of DNA sequences encoding the highly conserved LHC I1 apoproteins from these two groups (LaRoche er al., 199Oa).

H. DNA A N D cDNA LIBRARIES Two basic types of libraries may be constructed, genomic DNA (Kaiser and Murray, 1986) and cDNA (Huynh et al., 1986), the choice depending on the goals of the research and the organism(s) being investigated. cDNA libraries contain no introns and therefore are preferable for deducing amino acid sequences of proteins, for antibody screening, and if the gene of interest is highly expressed. A genomic library is useful for studying

MOLECULAR BIOLOGY IN STUDIES OF OCEAN PROCESSES

273

control regions of genes which may not be part of the reading frame and for investigating gene organization. The success of the screening method depends on the availability and specificity of probes, but few are available specifically for marine organisms; however, probes from other sources are frequently used to initially screen libraries.

I. SEQUENCING NUCLEIC ACIDS Nucleic acid sequencing has many attractions. It provides a means of rigorously assessing the similarity between two or more genes and of deducing amino acid sequences for the gene product. While relatively few sequences are available from marine organisms, there is considerable interest in sequencing specific genes or regions. For example, in an attempt to understand the evolution of Rubisco, Cattolico and co-workers (Reith and Cattolico, 1986; Li and Cattolico, 1987) sequenced the rbcL gene from a variety of sources. Pace and co-workers (1986) sequenced 5 S and 16 S rRNA in marine microbes in order to understand the evolution and taxonomic relationships among procaryotes. LaRoche et al. (19Wa) and Grossman (personal communication)have sequenced genes encoding light-harvesting chlorophyll protein complexes in unicellular algae in an effort to elucidate pigment-binding sites and evolutional origins. We will now discuss specific applications of these basic molecular biological techniques to biological oceanographic questions. IV. The Application of Molecular Techniques for Identification, Enumeration, and the Study of Genetics of Marine Organisms A significant effort in biological oceanography is directed toward identifying and quantifying organisms. This information is essential to our understanding of basic community structures and food chain dynamics of marine ecosystems. In many cases, especially for larger organisms, visual identification of morphological characteristics is sufficient. However, sometimes two or more closely related species cannot be definitively distinguished by morphological characteristics. Additionally, many benthic organisms have planktonic larval stages, which are morphologically different from the adult form; the literature is rife with misidentifications. Of all marine organisms, heterotrophic bacteria are the most difficult to identify and enumerate by traditional methods. Molecular biological techniques can distinguish closely related species and even populations.within a species. Such techniques may help biological oceanographers to under-

274

PAUL G . FALKOWSKI AND JULIE LaROCHE

stand phylogenetic relationships and population dynamics. We shall describe how some of the techniques are applied to specific questions. A. APPROACHES AT THE CELLULAR LEVEL

Only about 5 percent of the bacteria in the ocean are taxonomically or functionally identified. While that number may be higher for phytoplankton, the taxonomic identification of cyanobacteria and small eukaryotic cells is largely unsatisfactory. Moreover, because most of these organisms either do not reproduce sexually or have poorly uqderstood sexual cycles, determination of what constitutes a species is often problematic. Up until the late 1970s the major method of estimating the total viable bacteria in aquatic environments was by plate counts. Plate counts underestimate, often by several orders of magnitude, the total number of bacteria compared with epifluorescence microscopy and autoradiography (Roszak and Colwell, 1987a,b; Kurath and Morita, 1983). The epifluorescent microscopic method (Hobbie et al., 1977),is now widely used to enumerate marine bacterial biomass. Some obvious problems are that it cannot differentiate between dead and living bacteria nor identify bacterial species. However, immunofluorescence, in conjunction with specific inhibitors (Kogure and Taga, 1979), can distinguish between viable and nonviable cells (Brayton et af., 1987). The epifluorescent technique is often coupled with bulk measurements of heterotrophic activity from filtered or total sea water samples to obtain estimates of bacterial growth. This approach is adequate to estimate growth of the ensemble of bacteria in a sample; however, the method is not sensitive enough to study specific microenvironments which are important in microbial processes. Immunoassays of whole plankton communities, using polyclonal antibodies raised against specific algal classes or species (Campbell et al., 1983) allowed the identification and enumeration of microbes and phytoplankton that are difficult to see by conventional microscopic techniques. Particularly favored are immunofluorescence assays using a primary antibody conjugated with fluorescein isothiocyanate (FITC) (Ward, 1990). A secondary antibody enhances the number of fluorochromes which bind to the epitope. A seawater sample treated with FITC-labeled antibodies raised against whole cells of a single species, often shows crossreactivity with the antigenic species. This technique allows for the quantitation and identification of cells which may be too small to distinguish under a light microscope. Flow cytometry has greatly increased the biological oceanographer’s ability to count and size bacteria and phytoplanktonquickly, and, based on

MOLECULAR BIOLOGY IN STUDIES OF OCEAN PROCESSES

275

inherent (e.g., chlorophyll) or externally added fluorochromes (e.g., DAPI), to distinguish at phylum or class levels between particle types. In principle, flow cytometers should be able to sort species at sea, thereby greatly facilitating the isolation and characterization of protein and DNA from known sources (Burkill, 1987; Olsen et al., 1983; Robertson and Button, 1989). B. APPROACHES AT THE PROTEIN LEVEL I . Allozyrne Analyses One means of assessing genetic variability is allozyme analysis. Comparison of loci of particular allozymes between clones, populations, or organisms provides an objective means of distinguishing between morphologically related (or identical) organisms. Allozyme analyses have been used by fishery biologists to investigate cohort size, migration pattern, and recruitment (Powers et al., 1990). It has also been used in studies of plankton populations (Bucklin et al., 1989). For example, some species of phytoplankton are distributed thoughout the world’s oceans. The marine centric diatoms, Thalassiosira pseudonana and Skeletonerna costaturn extend from the nutrient-rich coastal waters of the North Atlantic to the nutrient-poor open ocean waters of the Sargasso Sea. Are morphologically similar species genetically identical? Murphy and Guillard (1976) examined the electrophoretic distribution of allozymes in two species of diatoms from the genus Thalassiosira and in 14 clones of Thalassiosira pseudonana. They found that most of the individuals were homozygous at all loci assayed, although loci with more than one allele were common. In a follow-up study with 12 new isolates of T. pseudonana, several heterozygotegenotypes were found for malate dehydrogenase and phosphohexose isomerase loci. Murphy (1978) suggested that occasional sexual recombination in culture led to heterozygosity and that in the natural environment it is likely that each oceanic water mass contains its own T.pseudonana homozygous genotype. Thus, although sexual recombination rarely occurs in cultured phytoplankton, its occurrence may lead to phenotypic heterogeneity that is then maintained by vegetative reproduction. In subsequent studies, Gallagher (1982) concluded that various isolates of a diatom Skeletonemu costaturn could be distinguished from each other by their allozyme banding pattern, and Gallagher and Alberte (1985) established that different isolates of S.costaturn display different physiological characteristics. Gallagher (1982) concluded that variability between populations (clones) of the same species can be as large as that between species.

276

PAUL G . FALKOWSKI AND JULIE LaROCHE

Patterns of allozyme variation have also been used to look at the population genetics of several copepod species (Bucklin et al., 1989; Sevigny et al., 1989; Sevigny and McLaren, 1988). Sevigny and McLaren (1988) used isoelectric focusing to determine the variability of glucose phosphate isomerase in individual copepods. The allozyme variations were used to confirm that seven species of Pseudocalanus, which had been reclassified on the basis of morphological differences (Frost, 1989), had diverged genetically. Bucklin et al. (1989) used allozyme analyses to follow variations in the populations of various copepods off the coast of California. The technique allowed them to distinguish between physical processes which separate and focus planktonic organisms and biological variability within patches of zooplankton. 2 . Protein Sequencing Protein sequencing is potentially useful for distinguishing between two different species but is unlikely to reveal differences at the population level. Compared with allozyme analysis in studies of populations, protein sequencingis more time consuming, more expensive, and more difKcult to interpret. However, protein sequencing is possible with marine organisms (LaRoche et al., 199Oa), especially since the introduction of polyvinylpyrrolidone (PVP) membranes and gas phase sequencers, which have increased the sensitivity and ease of sequencing the N termini. N-terminal sequence information may be used to design oligonucleotides which subsequently can be used as gene probes (Zehr and McReynolds, 1989).

C. APPROACHES AT

THE

NUCLEICACID LEVEL

In the field of oceanography, molecular biology has had its greatest impact thus far in the study of the phylogeny and evolution of marine organisms. The underlying concept is to analyze nucleic acids from various organisms and to compare the degree of similarity between populations or species. Phylogenetic relationships are inferred from statistical analyses of nucleic acid sequence similarity between organisms (Olsen et al., 1986, and references therein). In principle, the molecular approach provides an objective means of identifying and classifying organisms. Aside from the identification of closely related species, clonal variations within a single species, and the formation of hybrids, molecular techniques may be useful in identifying basic groups or organisms that play similar ecological roles in the marine environment. The majority of the techniques are based on sequence analysis of either ribosomal RNA (rRNA), mitochondrial DNA (mtDNA), or specific genes. We will discuss the usefulness of each of these approaches and give examples of recent work with marine organisms.

MOLECULAR BIOLOGY IN STUDIES OF OCEAN PROCESSES

277

I . Ribosomal RNA Ribosomal RNA has been used extensively to establish the phylogeny of prokaryotes (Fox et al., 1980; Olsen et al., 1986). More recently, this type of study has been extended to marine prokaryotes (Giovannoni et al., 1990) and to phytoplankton (Hinnebusch et al., 1981; Medlin et al., 1988). Based on molecular mass, there are three general types of rRNAs found in prokaryotes: a 23 S rRNA, a 16 S rRNA, and a 5 S rRNA. Eukaryotes may contain a fourth, 5.8 S rRNA. The 5 S and the 16 S rRNA are the most commonly used for phylogenic studies. The rRNAs are functionally and evolutionarily homologous (but not identical) in all organisms. They are readily identifiable by their sizes, constitute a significantpart of the cellular mass, and are easy to isolate from many types of organisms. The main advantage of studying rRNA is that it becomes possible to characterize the community structure of microorganisms, bypassing the problem of noncultivability of most of the naturally occumng microbes. The diverse environments in which marine microorganisms have adapted often make it difficult, if not impossible, to reconstruct adequate growth conditions (Roszak and Colwell, (1987b).

a. Analyses of5 S rRNA. The analysis of 5 S rRNA requires relatively small amounts of starting material. For microbes, 108-1@ cells are generally sufficient for the analysis. This is the average bacterial population found in one liter of seawater. For eukaryotes and multicellular organisms, a few milligrams of tissue is adequate. The extraction of nucleic acids may be complicated by the presence of extraneous materials such as sediments, which may adsorb organisms or nucleic acids. Most of the early information about rRNA from marine microorganisms came from 5 S rRNA sequences because its small size made sequencing more tractable (DonisKeller et al., 1977; Peattie, 1979). While 5 S rRNA are small and, therefore, contain limited information, nevertheless, the analysis of 5 S rRNA has contributed significantly to the understanding of the community structure. For example, 5 S RNA sequences have been used to investigate microbial diversity in deep-sea, submarine hydrothermal vents. Hydrothermal vents are found in areas of crustal spreading such as the mid-ocean ridges (Edmon and VonDamn, 1983) and are densely populated by chemolithotrophic bacteria which support large animal communities (Jannasch and Mott, 1985). The bacteria oxidize hydrogen sulfide, which is spewed out of the vents, and use the reducing power to fix inorganic carbon (Cavanaugh, 1983). The vent systems are characterized by symbiotic relationships between the bacteria and a variety of invertebrates, which harbor symbionts in their guts. The animals symbiotically associated with the chemolithotrophic bacteria all

278

PAUL G. FALKOWSKI AND JULIE LaROCHE

lack mouths and digestive systems and derive their reduced carbon entirely from the symbionts. Stahl et al. (1984) studied 5 S rRNA from symbiotic bacteria associated with these submarine hydrothermal vents, where the species diversity of the microbial population is relatively low. The 5 S rRNA data are available for two such invertebrate species and their symbionts: the giant vestimentarian tube worm, R i t i a pachyprila, and the giant clam, Calyptogena magnijica. Each invertebrate symbiont contains two distinct 5 S rRNA; one eukaryotic type which is closely related to the mollusc, and one which is eubacterial, related to the purple bacteria group. Analyses of the 5 S rRNA revealed that each symbiont is closely related to a different member of the gamma subdivision of the purple bacteria group (Stahl et al., 1984). b. Analyses of16 S rRNA. The 16 S rRNA are 13 times larger than 5 S rRNA, and therefore contain more genetic information; Lane et al. (1988) developed a method to sequence 16 S rRNA rapidly without previous isolation of the gene or the RNA species. The relative advantages and disadvantages of 5 S versus 16 S rDNA analysis have been reviewed extensively by Olsen et al. (1986), but with the development of more effective DNA sequencingtechniques, the need to identify closely related species, and the introduction of ECR amplification, the advantage of sequencing the 16 S rRNA genes outweighs the difficulties of working with the larger molecules. In fact, the major advantage of using 16 S rRNA is its larger size, which greatly improves the confidence limits of evolutionary distance estimates (Fig. 3 in Olson et al., 1986). The primary sequence information coupled with the secondary structure of the 16 S rRNA is a powerful tool for ecological and phylogenetic studies. A compilation of conserved secondary structure and primary sequences is available for the major taxonomic groups (Gutell et al., 1985) and is useful for the design of specific oligonucleotide probes. The greater sensitivity obtained from the 16 S rRNA allows the detection of a microorganism at the species or strain level. Ribosomal RNA sequences are usually obtained by sequencing the rDNA. Two methods are currently used to analyze the rRNA gene sequences: one is aimed at the complete sequence of the gene while the other requires only the sequence of large regions which are evolutionarily important (Lane et al., 1985). The latter method has the advantage of selecting the areas to be sequenced so as to reveal comparable homologous sequences in many organisms, and can be done without extensive manipulation of the cloned DNA. Two recent studies have demonstrated the potential impact that 16 S rRNA gene sequence information can have on our understanding of microbial community structure.

MOLECULAR BIOLOGY IN STUDIES OF OCEAN PROCESSES

279

Ward et al. (1990) sequenced 16 S rRNA genes to follow the population dynamics in a cyanobacterial mat of Octopus Spring in Yellowstone National Park.Their approach was to make cDNA from RNA extracted from natural populations (Weller and Ward, 1989). By using a primer specific to 16 S rRNA, the authors made libraries that contained only 16 S rRNA cDNA clones. None of the eight sequences isolated from their library was identical to sequences from microorganisms previously isolated from the Octopus Spring or other geothermal springs. Giovannoni et al. (1990) reached a similar conclusion in a study of the genetic diversity of bacterioplankton from the Sargasso Sea. They reported that none of the 16 S rRNA clones sequenced was identical to those of cultured microorganism species isolated from tropical or oligotrophic oceans. They found two major gene clusters: one related to oxygenic phototrophs, and a deep-branching cluster (SARII) closely related to the a-purple eubacterial phyla. The authors suggested that the new cluster SARII may represent a novel phylum which is primarily found in the Sargasso Sea and oligotrophic oceans. Evidence supporting this hypothesis was obtained by using oligonucleotide probes, specific to the SARII cluster, on bacterioplankton samples isolated from a wide range of environments. They observed that the SARII bacterial type, although predominant in the Sargasso Sea, was also found in coastal waters of Bermuda and Florida; however, it was absent from seawater samples collected along the coast of Oregon. While the collections of isolates from these two areas are not extensive, the works of Giovannoni et al. (1990) and Ward et al. (1990) imply what has long been suspected by biological oceanographers; either the natural species diversity is large or the species that can be isolated and cultured are not representative of the natural environment. The techniques employed by these two groups allow the construction of representative cDNA or genomic DNA libraries from natural microbial communities which, in turn, can be used to identify the species diversity of a particular environment. Well-tailored probes complementary to specific sequences of RNA or DNA could be used for the relatively rapid, parallel processing of a large number of samples (Giovannoni et al., 1988). It is clear that an equally important problem is the enumeration of the identified microbes. The current methods employed by marine microbiologists (i.e., DAPI-stained bacteria) only provide total bacterial counts (Hobbie et al., 1977).

The possibility of using 16 S rRNA probes or sequences to identify and enumerate specific or general groups of organisms was initially explored in the laboratory (Edwards et al., 1989; DeLong et al., 1989); these techniques have already proven useful in natural samples. The use of in uiuo hybridization techniques allows for the radioactive or fluorescent labeling of individual cells present in preserved field samples. The design of hybrid-

280

PAUL G . FALKOWSKI AND JULIE LaROCHE

ization probes depends primarily on their use; restriction fragments containing entire genes can be used. However, in practice, small oligonucleotide probes (20-30 bp long) give better results for in situ hybridization studies because small probes can better permeate intact, fixed cells. The specificity of the probes requires prior knowledge of the gene sequenceof a variety of organisms. Specific probes are currently available for eukaryotes, eubacteria, and archebacteria (Giovannoni et al., 1988) and these probes can be used either for screening genomic libraries of new marine isolates or in the enumeration of mixed microbial populations. Fluorescent probes, combined with light microscopy, are a potentially powerful tool in studies of natural microbial assemblages (DeLong et al., 1989). Fluorescent probes are sensitive, relatively safe, do not require autoradiography for visualization, and, unlike radioactive probes, do not decay with time (although some fluorochromes may degrade); therefore, they can be made in advance. The latter is attractive for research on board a ship. More than one probe can be used if they have different fluorescing properties (DeLong et al., 1989) so that more than one class of organisms in the same microscope field can be detected and enumerated.

2. Total Community DNA Hybridization Lee and Fuhrman (1990) developed a simple method, based on total community DNA hybridization, that measures the similarity between several different microbial communities. With this technique they demonstrated a temporal variation of the bacterial community in Long Island Sound. In addition, they observed a 20-50% similarity between samples from Long Island Sound, the Sargasso Sea, and the Caribbean Sea, while microbes from a coral reef were less than 10% similar to all other samples studied (Fig. 2). While this technique is relatively simple and undoubtedly will be useful in screening bacterial assemblages from various oceanic regions or seasons, which can then be studied in more detail, with a more time-consuming analysis of species composition such as 16 S rRNA, it also is relatively insensitive and requires relatively large amounts of nucleic acids. 3. Mitochondria1 and Chloroplast DNA

While rRNA analysis is applicable to all kingdoms, it is particularly useful in microbial studies primarily because of the large data base on microorganisms. In eukaryotic organisms, analysis of mitochondrial DNA (mtDNA) or chloroplast DNA (cpDNA), provides an alternative method to rRNA analysis for studies of phylogeny, evolution, and population genetics. In unicellular algae and macrophytes, the organelle genome most often studied is cpDNA (Boczar et al., 1989; Francis et al., 1990; Goff and

MOLECULAR BIOLOGY IN STUDIES OF OCEAN PROCESSES Lagoon Area Center Edge

Sargasoo Sea

Lagoon Area

Sergesso Sea

H

Depth 25m

I221

loom

281

FIG.2. An example of the use of community DNA hybridization to screen for similarities and differences between marine microbes. Similarities between species composition of bacterioplankton assemblages from a lagoon to open waters of the Sargasso Sea near Bermuda were calculated based on total community DNA hybridization analyses (Lee and Fuhrman, 1990). Numbers at the top of the box indicate sample numbers. Percent similarities are given in italic. Probes, defined as loo%, were samples (panel A) and 4 (panel B). Samples 1 and 4 and 3 and 7 were replicates collected a few days apart. (Data kindly provided by S.-H. Lee).

Coleman, 1988; Douglas and Durnford, 1989), which is larger and easier to isolate than mtDNA. For example, the mtDNA genome of the chlorophyte Chlamydornonas reinhardtii is linear and is 16 kb in size (Grant and Chiang, 1980), while cpDNA is circular and is 160 kb (Harris, 1989). Mitochondria1 DNA is maternally inherited and is generally a duplex, covalently closed, circular molecule. In animals, mtDNA varies in size

-

282

PAUL G . FALKOWSKl AND JULIE LaROCHE

from 14.2 to 4~ kb, but in the majority of the species studied, clusters between 14.2 and 19.5 kb. Animal mtDNA encodes for 2 rRNA, 22 tRNA, and 13 proteins involved in electron transport and ATP synthesis (Attardi, 1985). The maternal inheritance of mtDNA makes it especially useful for examining the evolutionary history of maternal gene flow in a species, independently of Mendelian genetics. The rate of mutation in mtDNA is relatively constant compared with single copy nuclear genes, and analysis of mtDNA provides information on the rate of gene flow, population structure, hybrid zones, biogeography, and phylogenetic relationships (Avise et al., 1987; Moritz et al., 1987). Three types of changes occur in mtDNA that are relevant to evolutionary studies: base substitution, length variation, and sequence rearrangement. Different parts of the mtDNA molecule mutate at different rates. Therefore, sections of the genome have to be selected on the basis of the relatedness between the organisms to be compared. Both restriction site polymorphism (Edward and Skibinski, 1987; Ovenden et al., 1988) and sequence data (Moir and Dixon, 1988) have been used in phylogenetic studies. Although sequence data offers more potential and is more sensitive, restriction maps are the predominant type of data collected (Graves et al., 1984; Bowen et al., 1989). Base substitution is useful for comparing the mtDNA sequence among closely related m a . mtDNA has several advantages over the study of single-copy nuclear genes. Being present in multiple copies, it is more readily detectable by Southern blot hybridization of total DNA. Distantly related organisms can be studied by the analysis of gene order and rearrangment of the secondary structure of tRNA and rRNA, while more closely related species or different populations within a species can be studied by differences in nucleic acid sequence of more specific regions. Vertebrate mtDNA is well characterized and does not vary much in size between species (Attardi, 1985). It has been proven to be a powerful tool in the elucidation of several zoogeographical problems involving marine vertebrate species. A few case studies are discussed here. Fisheries management requires an understanding of the population dynamics of the species. For example, the Skipjack tuna is present in both the Atlantic and Pacific Ocean basins. The tuna do not have specific spawning grounds and their pelagic larvae are found circumtropically. Mitochondria1 DNA sequence information was used by Graves et al. (1984) to establish that the tuna from the Atlantic and Pacific Oceans are not genetically differentiated and can be treated as a single breeding population. In contrast, migratory observations suggested that, although no obvious geographical barrier separates the major Humpback whale pop-

MOLECULAR BIOLOGY IN STUDIES OF OCEAN PROCESSES

283

ulations, they are divided in seasonal subpopulations that share a common winter breeding ground. Analyses of mtDNA have been used to confirm that the different populations of the whale are indeed genetically segregated (Baker et al., 1990) despite the lack of geographical barriers. mtDNA has also been useful in studying the migratory behavior of sea turtles. In particular, the origin of the isolated nesting colony established on Ascension Island has been highly debated over the years (Cam and Coleman, 1974; Gould, 1978). Green turtles nesting on Ascension Island undergo a remarkable seasonal migration of 2000 km to their feeding grounds located off the coast of Brazil. Carr and Coleman (1974) have hypothesized that these migrations could have developed over millions of years as the result of their strong natal homing behavior. As a result of sea-floor spreading in the Atlantic, islands which were originally close to the feeding grounds 70 million years ago are now very distant. Alternatively, Ascension Island may have been colonized by a rare but recent event. The low mtDNA variability found in sea turtle populations supports the hypothesis of recent colonization (Bowen et al. 1989).

A classical problem in zoogeography is whether or not the eel populations of North America and Europe are different species. The eels migrate from estuaries to spawn in the Sargasso Sea. Their larvae travel thousands of kilometers back to estuaries where they metamorphose. On the basis of this life history, one would expect a genetic uniformity between European and American eel populations. Avise et al. (1986) determined that the mtDNA of the North Atlantic eels is differentiated into two distinct types, strongly implying segregation of two populations. Although the analysis of vertebrate mtDNA has proven conclusive in the study of many zoogeographic problems, this tool should be used with caution when studying problems related to invertebrates. Gene rearrangement, large size variations, codon bias, duplications, and deletions are common in invertebrate mtDNA (Moritz et al., 1987; Snyder et al., 1987; La Roche et al., 1990~).In the invertebrates, an initial characterization of the size and variation in mtDNA will be required for each new species studied to assess the usefulness of the mtDNA as a genetic marker. An emerging pattern from the study of invertebrate mtDNA is that tremendous differences in the mtDNA can exist within a single family. For example, Snyder et af. (1987) found large size variations in the mtDNA from a deep-sea scallop, Plucopecten rnagellanicus, while another genus of the family Placopectinidae, the bay scallop Argopecten, has a small mtDNA genome of 16 kb which shows little size variation (B. Gjetnaj, personal communication).

284

PAUL G. FALKOWSKI AND JULIE LaROCHE

V. Application of Molecular Techniques to Studies of Organism Function

Molecular approaches also may be used to understand the underlying biological factors governing the spatial and temporal distribution of organisms in the sea and their role in biogeochemical cycles. The distribution of organisms in the oceans is affected by the physical environment and biological responses. While the ability of a species or population to genetically adapt, in a Darwinian sense, to variations in the physical environment determines the range, survival, and success of that species, on shorter time scales, the ability of an organism to physiologically acclimate to short-term variations in the physical environment may lead to phytoplankton blooms, survival of a scallop in the face of hypoxia or trace metal pollution, dormancy or cyst formation when adequate nutrition is not available, and so on. For more than a century, ecologists have described the physiological and biochemical responses to changes in temperature, salinity, irradiance, pressure, nutrient regime, and pH for a myriad of marine organisms (Hochachka and Somero, 1984). While these descriptions provide a framework for understanding the distribution of organisms in the ocean, they have not provided a mechanistic understanding of biological regulation of these marine processes. Molecular biological techniques may provide a basis for understanding the role of a species or a group of organisms involved in a specific process such as biogeochemical cycles; for determining the level of control, i.e., transcriptional, posttranscriptional, translational, posttranslational; and for determining if the process is inducible, and, if so, the signal transduction mechanisms. We shall address these issues by example. A. PHYSIOLOGICAL STATEA N D METABOLICACTIVITY OF MARINEMICROBES A typical seawater sample contains
MOLECULAR BIOLOGY IN STUDIES OF OCEAN PROCESSES

285

colonize sinking dead particulate materials, such as zooplankton fecal pellets. The microbes appear to induce surface-bound or extracellular enzymes which facilitate the hydrolysis of protein and carbohydrates, thereby effectively solubilizing the particles during their descent to the sea floor. The induction mechanisms and enzyme structures are as yet unknown (F. Azam, personal communication). Thus, an important issue in biological oceanography is: are the bacteria growing fast, dying rapidly, and, therefore, providing an ample flux of inorganic nutrients to the primary producers?; or are they growing slowly, and thereby potentially limiting the recycling of essential nutrients for primary producers? Roszak and Colwell (198%) defined several physiological states and characteristics of marine bacteria. They suggested that the “normal state” of native marine bacteria is a “starved state” which is characterized morphologically by a significant decrease in the size of the cells (Novitsky and Morita, 1978). Viability may be pragmatically defined as the ability of the bacteria to divide, but it is appropriate to consider that bacteria undergo a recovery period before they can divide again (Morita, 1982). Thus, bacteria can be viable but nonculturable and the quantitationof viable cells should be determined by methods that do not require culturing in enriched broth at temperatures higher than ambient (Torrella and Morita, 1981). The measurement of growth in starved bacteria presents a problem. The currently accepted method for measuring growth rates of marine bacteria relies on the specific incorporation of [methyl-3H]thymidineinto bacterial DNA (Fuhrman and Azam, 1982). Although this method has been extensively tested and widely used, (Pollard and Moriarty, 1984) it has some limitations that are intrinsic to biological oceanography studies in general (e.g., Coveney and Wetzel, 1988). Measurements of bacterial secondary productivity using the thymidine uptake method require the addition of trace amounts of substrate and incubation of the seawater samples in a bottle. Any manipulations of the sample such as filtration, confinement, and substrate addition may affect the true in situ growth rate of the microbes (Fergusson et al., 1984). Although autoradiography combined with thymidine uptake can give a qualitative estimate of the percentage of bacteria that are actively growing, it is dimcult to achieve quantitative measurements using this method (Fuhrman and Azam, 1982). An alternative approach to this problem is to measure a cellular product indicative of metabolic activity and growth, such as rRNA (DeLong et al., 1989). Laboratory studies have demonstrated that rRNA/cell (and rRNA/ DNA) is a linear function of growth rate and metabolic activity (Bremmer and Dennis, 1987). The sensitivity of such measurements is high because rRNA is the most abundant RNA species in the cell ( > W o of total RNA; 10‘‘ ribosomes/cell).This type of technique employs either radioactive or

-

286

PAUL G . FALKOWSKI AND JULIE LaROCHE

fluorescent probes (Giovannoni et al., 1988; DeLong et al., 1989) and is amenable to either whole water samples or individual cell measurements. Morphological observations and measurements of activity in marine bacteria have led to the conclusion that native marine bacteria live most of the time in a state of starvation (Roszak and Colwell, 1987b). However, little is known of the adaptation mechanisms by which marine bacteria living in a variable environment can downshift their metabolism and survive starvation. Kjelleberg et al. (1987) have reviewed this subject and suggested that in marine bacteria starvation leads to an increased potential for uptake and assimilation of exogenous substrates, and that this is achieved through the induction of specific proteins. Wrangstadh et al. (1986) used fluorescent antibodies to demonstrate the formation of specific carbohydrates during starvation, while Kjelleberg et al. (1987), using immunoblotting techniques and pulse labeling, demonstrated that three to four new proteins were synthesized in the periplasmic space, and one in the outer membrane during the initial stages of starvation. B. MICROBES IN

THE

NITROGEN CYCLE

We have already discussed cycling of dissolved organic matter by bacteria. However, the role of prokaryotes in the nitrogen cycle deserves special consideration. The marine nitrogen cycle is of interest because the supply of fixed inorganic nitrogen is thought to limit primary producers and hence the flux of carbon in the ocean (Dugdale and Goering, 1967; Eppley and Petersen, 1979).

I . Nitrification and Denitrification Nitrification and denitrification are important processes within the marine nitrogen cycle. These processes have been reviewed (Kaplan, 1983; Hattori, 1983) and are only briefly summarized here. The pool of dissolved inorganic nitrogen in the upper mixed layer of the ocean, in the form of nitrate, nitrite, and ammonium, is vanishingly low, in the nanomolar to micromolar range. Nitrate and nitrite are supplied from the deep ocean to the surface by physical mixing. These two compounds are generated by the oxidation of ammonium by nitrifying bacteria. Simultaneously, however, denitrifying bacteria reduce nitrate to NOz-, nitrous oxide, and eventually N2, the last two compounds being innaccessible to phytoplankton. The microbes responsible for denitrification are facultative or obligate anaerobes that use nitrate as a terminal electron acceptor in the place of oxygen. To further complicate matters, phytoplankton, using light energy, are involved in assimilatory nitrate reduction, a process which involves uptake and reduction of NO3- to NH4+. Ammonium is further

MOLECULAR BIOLOGY IN STUDIES OF OCEAN PROCESSES

287

assimilated for use in biosynthesis of cellular components. While nitrification and assimilatory nitrate reduction regenerate nitrate and transform it into particulate nitrogen, respectively, the process of denitrification has the opposite effect of removing nitrogen from the ocean into the atmosphere. The direct measurement of the rate of either nitrification or denitrification is dimcult. Traditionally, oceanographershave relied on the vertical distribution of several dissolved gases and nutrients such as oxygen, NO, NzO, NH4+, NO;, NO*-, and phosphate. Nitrogen budgets are constructed from vertical profiles of the individual nitrogen species in the water column, assuming that nitrification and assimilatory and dissimilatory nitrate reduction are vertically segregated; assimilation of nitrate occurs mainly in the euphotic zone, nitrification towards the base of the euphotic zone, and dissimilatory nitrate reduction below the euphotic zone and in regions low in dissolved oxygen. In addition, "N and 15Nhave been used as tracer in bottle incubations of seawater to follow the transformation of dissolved nitrogen species (Hattori, 1983; Kaplan, 1983; Capone et al., 1990). Kaplan (1983) pointed out that denitrificationand nitrification can occur in the same parcel of water and that some intermediates such as NO2 and N2O are common to both processes. In brief, there is no easy way to assess the impact of each pathway on the nitrogen cycle. The general methods used, such as counting DAPI-stained cells, do not allow the identification or enumeration of the species responsible for metabolism of specific substrates. Moreover, Ward (1984) pointed out that nitrification provides little energy (therefore a lot of nitrification equals little bacterial growth) and that the growth rate and abundance of these bacterial strains do not necessarily reflect their impact on the overall nitrogen cycle. Some progress has been made in elucidating the importance of specific metabolic pathways of microbes with immunological techniques (Ward et al., 1987). Using cultured ammonium- and nitrite-oxidizing bacteria, antibodies were raised which were specific to ammonium-and nitrite-oxidizing bacteria. Using an immunofluorescence technique, Ward et al. (1987) determined that ammonium- and nitrite-oxidizingbacteria were present at concentrations of approximately 10'' cells/liter. Autoradiography,coupled with a fluorescent antibody technique, was used to measure the number of active cells of a specific genus (Yentsch et al., 1988). However, this work failed to reveal a relationship between the species abundance and nitrification activity. The results, however, suggest that although nitrifying bacteria are distributed throughout the water column, induction of enzymes specific to the nitrification pathway may occur only under favorable environmental conditions. Thus, additional information may be obtained if

288

PAUL G . FALKOWSKI AND JULIE LaROCHE

antibodies were available for enzymes directly involved in the metabolic conversion of NH4+to NO3-. In addition, probes for the mRNA encoding of these enzymes might permit the study of the induction of the proteins by environmental stimuli.

2 . Nitrogen fiation In the steady state, nitrogen lost to the atmosphere by denitrification must be resupplied to the ocean. While fixed nitrogen is also replenished by runoff and by atmospheric electrical discharges, the contribution of biological nitrogen fixation (Capone and Carpenter, 1982) is likely to be more significant than previously anticipated and will undoubtedly need to be revised in the future (Codispoti, 1989). Nitrogen fixation is carried out exclusively by prokaryotes; in the ocean the process occurs in cyanobacteria. The major nitrogen-fixing organism is thought to be the colonial cyanobacterium, Trichodesmium spp. (Carpenter and Capone, 1983), although nitrogen fixation has also been detected in unicellular oxygenic cyanobacteria (Mitsui et al., 1986). Nitrogenase is a multiprotein enzyme complex which catalyzes the reduction of dinitrogen gas to ammonium. The complex has two major subunits, an Fe subunit 43 kDa in Trichodesmiurn and a Mo-Fe subunit (-- 55 m a ) . In freshwater cyanobacteria, the enzyme complex is induced when all fixed nitrogen sources (i.e., nitrate, ammonium, and urea) are depleted. The enzyme is rapidly inactivated by oxygen. Thus, many nitrogen-fixing cyanobacteria have evolved a mechanism for spatially segregating the nitrogen-fixing system from the oxygen-evolving complex in photosystem 11. Briefly, upon starvation for fixed nitrogen, vegetative cells differentiate, forming special cells called heterocysts. Heterocysts degrade photosystem 11 (and hence lose the ability to evolve oxygen photosynthetically), but retain photosystem I. In the process of differentiation, the genes for nitrogenase undergo a rearrangement, which is thought to be essential for expression of the proteins. This is one of the earliest forms of cellular differentiation known (Golden et at., 1985). Trichodesmiurn and other nitrogen-fixing cyanobacteria in the ocean do not contain heterocysts, yet fix nitrogen and evolve oxygen. The mechanism for oxygen protection in these organsisms is poorly understood. However, it is known that nitrogenase is regulated at three levels in Trichodesmium. First, in the presence of fixed nitrogen the enzyme is suppressed. Second, the Fe protein can be posttranslationally modified, which leads to reversible activation ( O M et al., 1991; Currin et al., 1990). Third, in the presence of light and absence of fixed nitrogen, the enzyme is synthesized and activated, requiring both transcriptional and posttranslational controls. The genes encoding for nitrogenase (the nifoperon) do not

-

MOLECULAR BIOLOGY IN STUDIES OF OCEAN PROCESSES

289

appear to undergo a rearrangement when Trichodesmium or single-celled cyanobacteria initiate the nitrogen fixation process (J. P. Zehr, personal communication).

C. PHOTOSYNTHESIS AND PRIMARY PRODUCTION Inorganic carbon fixation by photosynthetic organisms in the sea accounts for about half of the global primary production. By far, the vast majority of oceanic photosynthesis is carried out by single-celled organisms, the phytoplankton. Unlike terrestrial plants or marine macrophytes, phytoplankton store little carbon in inert biomass, and instead allocate most of their reduced carbon to proteins. Phytoplankton biomass turns over on the average of 10 times per annum, compared with a 4% per annum turnover for terrestrial plants. The rapid turnover and patchy distribution of phytoplankton make rate measurements difficult. Moreover, phytoplankton are extremely plastic physiologically; that is, they acclimate rapidly to changes in their environment. The physiological adaptations often affect carbon fixation rates. Although many of these adaptative phenomena have been described, little is understood about their molecular basis. We will briefly describe a few of the adaptation responses and what is known about the molecular bases of physiological adaptation. 1. Adaptation to Irtadiance Levels

Phytoplankton and macrophyte algae can physiologically acclimate to changes in irradiance. Photoadaptation maximizes light harvesting when photon densities are low and minimizes photooxidative damage to the photosynthetic machinery at high light levels (Falkowski, 1980). Adaptation is characterized by changes in cellular content of pigment protein complexes (Fig. 3), ratios of the photosynthetic pigments, activity of enzymes involved in carbon fixation, and changes in cell volume, respiration, and chemical composition. The molecular basis of these changes is poorly understood, yet photoadaptation is a major factor which determines the functional relationship between photosynthesis and irradiance (Falkowski, 1980; Perry et al., 1981). It is this latter relationship which, in turn, forms a basis for rationally modeling the spatial and temporal distribution of primary production from satellite images of the distribution of surface chlorophyll in the upper ocean (Platt and Sathyendranath, 1988). Most of the research on the mechanisms of photoadaptation upon changes in photon flux densities has focused on the photoregulation of the light-harvesting chlorophyll proteins (Sukenik et al., 1990). These pigment protein complexes serve as the major antenna for the reaction centers; pigment-binding proteins are encoded in the nucleus. The best-studied of these complexes is the light-harvesting chlorophyll alb protein complex

PAUL G. FALKOWSKI AND JULIE LaROCHE 12 10

8 6 4

2

0

40

ao

120

160

200

llME(h0urd FIG. 3. Die1 variability in cellular chlorophyll in a marine diatom Thalassiosira weisflogii. Cells grown on a 12 : 12 hour tight/dark cycle show an increase in cellular chlorophyll during midday and a decrease during the dark period. Upon transfer from a high-growth irradiance level (HL) to a low-growth irradiance level (LL), cellular chlorophyll increases. The photoadaptive increase is superimposed on the die1 cycle. The variations in chlorophyll reflect changes in the relative abundance of light-harvesting chlorophyll protein complexes (Post et 01..

1984).

from green algae. The amino acid sequence, deduced from gene or cDNA sequences (LaRoche et af., 1990a), is highly conserved between green algae and higher plants (50%). In both green algae and diatoms the protein is translated in the cytoplasm to a larger precursor, which, following import into the chloroplast, is posttranslationally modified. The mature protein, which binds pigments, is inserted into the photosynthetic membrane. The stability of pigment protein complexes is thought to depend on the binding of pigment to the apoprotein; hence turnover of the apoprotein occurs unless pigment synthesis keeps pace with protein synthesis (Falkowski, 1980). LaRoche et al. (1990b)found that a decrease in photon flux leads to a rapid increase in LHC I1 mRNA (Fig. 4). These results suggest that a light intensity-dependent transcriptional control mechanism may be superimposed on a posttranslational control. The nature of the signal transduction mechanism in response to changes in inadiance has not been elucidated. Photosynthesis may be inhibited at high irradiance levels (Neale, 1987). This phenomenon, known as photoinhibition, is well documented in natural marine and freshwater phytoplankton communities, and its molecular basis has received considerable attention. A key component in photoin-

MOLECULAR BIOLOGY IN STUDIES OF OCEAN PROCESSES

291

FIG. 4(A). Combination of Northern and Western blots showing the increase in LHC I1 on mRNA (cab) and apoproteins (LHCII). Dunaliella tertiolecta, grown in high light (700 pmol quanta/m%econd) at steady state, was transferred to low light (70 pmol quanta/ m2/second). Total RNA was hybridized to 32P-labeledcab cDNA (top) and whole cell protein extracts (bottom), were challenged with anti-LHCP raised against pea. (B) Changes in a-tubulin, and LHC I1 mRNA (cab), and LHCII apoproteins (bottom) after a shift from high to low light at 0 hours.

292

PAUL G . FALKOWSKI AND JULIE LaROCHE

hibition appears to be the destruction of D1, a 32-kDa protein (Kyle et al., 1984) which, together with a 34-kDa protein (D2) and cytochrome b559, comprise the reaction center of photosystem I1 in all oxygen-evolving photosynthetic organisms. Under normal conditions D1, which is encoded in the chloroplast genome, turns over with a half-time of 30 minutes (Ohad et al., 1986). When cells or macrophytes are exposed to high light, however, the rate of D1 synthesis cannot keep up with the photodestruction of the protein; hence, there is a net loss of D1 and the activity of the oxygen evolving system is reduced. The promoter for the gene encoding D1 is extremely strong (Sheen and Bogorad, 1988), and numerous copies of mRNA for the gene product are normally produced. The basis of the enhanced turnover of D1 at high photon flux densities remains unclear; however, it appears to involve the formation of an oxygen singlet and its interaction with a tyrosine residue on the DI protein. 2 . Chromatic Adaptation Cyanobacteria have light-harvesting complexes called phycobilisomes, which are super molecular structures consisting of proteins and porphyrinderived pigments (Glazer, 1986). Three major phycobiliproteins are well characterized: the green-absorbing phycoerythrin, the orange-absorbing phycocyanin, and the red-absorbing allophycocyanin. These pigment proteins are arranged with a central disc of allophycocyanin, from which arms of phycocyanin radiate. The phycoerythrin is thought to extend distally from the phycocyanin. Light absorbed by phycoerethryrin is transferred to phycocyanin and to allophycocyanin,which is energetically connected to photosystem I1 (Glazer, 1986). Not all cyanobacteria have phycoerythrin, however, and a closely related phycobilipigment,phycourobilin, is a major absorber and fluorescence emitter in the open ocean (Alberte et al., 1984). When phycoerythrin-containingorganisms are grown in red light, phycoerythrin synthesis is suppressed. Upon transfer to green light, however, phycoerythrin synthesis can be detected within a few minutes (Gendel et al., 1979). This light-quality-dependent synthesis of phycoerythrin is sometimes called chromatic adaptation. In freshwater cyanobacteria, Oelmueller et al. (1988) showed that synthesis of green-light-dependent phycoerythrin is regulated at the transcriptional level. The molecular mechanism has not been elucidated in marine algae, but for a review of the molecular bases of complementary chromatic adaptation in a fresh water cyanobacteria refer to Grossman et al. (1989). 3. UV-B Effects

Ultraviolet radiation is rapidly absorbed by water molecules in the upper ocean; however, UV-B radiation appears to penetrate to about 10 m.

MOLECULAR BIOLOGY IN STUDIES OF OCEAN PROCESSES

293

Recently, because of the erosion of ozone in the upper troposphere, there has been an increased interest in the dose-response characteristics of marine organisms to UV-B radiation. The action spectrum for UV-B radiation and its molecular interactions that cause damage to DNA are well documented (Setlow, 1974). However, no direct effect of UV-B radiation at natural levels has been observed on marine organisms. 4 . Circadian Rhythms

Marine organisms exhibit strong circadian rhythms in behavior such as feeding and vertical migration of zooplankton (Forward, 1988) and biochemical activities, such as pigment or photosynthetic rates (Prezelin and Sweeney, 1977; Post et al., 1984), cell cycles (Edmunds et al., 1984; Sweeney, 1982), and bioluminescence (Hastings and Sweeney, 1958). Although the diel cycle of marine organisms is largely entrained by the dayjnight cycle, a circadian rhythm usually persists for severalcycles after the organisms are transferred to constant conditions. Feeding patterns of zooplankton exhibit diel cycles in northern latitudes, where the light signal is minimal. Circadian rhythms are poorly understood at the biochemical and molecular level, but studies by Morse et al. (1989) on the circadian rhythm of bioluminescence in the dinoflagellate Gonyaulax may provide a useful system to understanding its molecular basis. Bioluminescence in Gonyaulax occurs mainly during scotophase (Hastings and Sweeney, 19581, and the circadian-controlled process of bioluminescencecorrelates with the absolute amounts of two proteins; the luciferase enzyme (Dunlap and Hastings, 1981; Johnson et al., 1984) and the luciferin-binding protein (LBP) (Morse et al., 1989). Morse et al. (1989) demonstrated that the changes in LBP must be translationally controlled since no variation in the amount of mRNA encoding for LBP is detectable during a diel cycle.

D. NUTRIENT LIMITATION IN PHYTOPLANKTON 1 . Nitrogen a. General Limitation. Dissolved fixed inorganic nitrogen has been identified as a major limiting resource for phytoplankton biomass in the ocean. On a molecular level, nitrogen limitation leads to a reduction in the intracellular concentration of free amino acids, especially Gln and Asp (Zehr et al., 1989). The reduction in amino acids affects protein synthesis at a translational level. Coleman et al. (1988) and Fakowski ef al. (1989) suggested that proteins encoded in the nucleus are preferentially synthesized over proteins encoded in the chloroplasts; however, the opposite was observed in Chlamydomonas reinhardtii by Plumley and Schmidt (1989). Moreover, the latter workers suggested that the regulation of

294

PAUL G . FALKOWSKI AND JULIE LaROCHE

nuclear-encoded proteins by nitrogen limitation affects either transcription and/or mRNA stability. Kolber et al. (1990) have shown that photosynthetic energy conversion efficiency in natural phytoplankton communities is directly related to nitrogen availability. They suggested that two chloroplast encoded proteins, CP43 and CP47, were less abundant in nitrogen-limited cells. The results of Kolber et al. (1990) indicate a molecular basis for understanding the effect of nitrogen limitation on carbon fixation in the sea. b. New Production and Nitrate Reduction. The concept of new and regenerated nitrogen (Dugdale and Goering,1967) has strongly iduenced the development of biological oceanography over the past two decades. Briefly, in the open ocean, dissolved inorganic nitrogen is available for eukaryotic photoautotrophs in two forms. Oxidized nitrogen, in the form of nitrate, can diffuseinto the upper mixed layer from the deep ocean. That nitrogen source, which was formed by nitrifying bacteria through the sequential oxidation of ammonical nitrogen, is called new nitrogen. Within the upper mixed layer, reduced nitrogen can be supplied as ammonium (or urea), which is produced from the excretion of microzooplankton, zooplankton, and fish, which either directly or indirectly consume phytoplankton. The latter source of nitrogen is called regenerated. In the steady state, the flux of particulate nitrogen from the upper ocean to the deep ocean interior (and the sea floor) must be balanced by the vertical diffusion of new nitrogen to the upper ocean. The measurement of new production in the ocean is problematic and contentious. A variety of ingenious methods have been introduced to infer the fraction of nitrogen which has been acquired as nitrate, including (but not limited to) 15Ntracer studies, direct measurement of nitrate utilization at submicromolar concentrations (Garside, 1982), sediment trap analyses, and models (Hayward, 1987). Two enzymes are required to utilize nitrate, namely nitrate reductase and nitrite reductase. Zehr et al. (1989) have shown that cells grown on ammonium with vanishingly small concentrations of nitrate can still reduce nitrate and incorporate nitrogen into cell protein. Cell culture has shown that both nitrate and nitrite reductases can be induced by the substrates and suppressed by ammonium. The data of Zehr et al. (1989) suggest, however, that part of the reducing capability is either constitutive, or is induced at vanishingly low concentrations of substrate in the presence of the presumed suppressor. Thus, attempts are being made to assess the presence of the enzyme in natural populations using Western blots (Balch et al., 1988). Polyclonal antibodies from the nitrate reductase of higher plants do not have a high affhity for phytoplankton enzyme (Falkowski, unpublished), and some effort is required

MOLECULAR BIOLOGY IN STUDIES OF OCEAN PROCESSES

295

to purify the unstable protein. Clearly, however, the molecular control of nitrate and nitrite reductases in phytoplankton needs further investigation. 2. lron Iron is essential for the development of the photosynthetic apparatus; it is found in cytochromes, nonheme iron sulfur complexes, and coordinated in membrane-bound proteins. Iron is especially important for the function of photosystem I. lron is supplied to the ocean from the atmospheric deposition of terrestrial particles (dust). In the late 1980s Martin and co-workers (Martin and Fitzwater, 1988; Martin ef al., 1989) found evidence of iron limitation of primary production in the subarctic Pacific Ocean. Little is known about the molecular basis of iron limitation. Riethman and Sherman (1988) showed that, in the freshwater cyanobacteria, Anacystis nudulans, iron limitation (starvation) leads to the synthesis of three polypeptides of 36-,35-, and 34-kDa. They have isolated, cloned, and sequenced the genes which encode these proteins (Sherman, personal communication; Sherman el al., 1987). To our knowledge, homologous genes or proteins have not been identified in eukaryotic algae, nor in cells taken from natural waters. It has been shown that bioluminescence in some marine Vibrio sp. is regulated by iron (Haygood and Nealson, 1985, and references therein). The presence of iron represses the expression of bioluminescence,and this response is not well understood because, unlike iron transport systems, luminescence does not appear to have an obvious function under conditions of iron limitation. The genes and proteins necessary for bioluminescence have been identified (Engelbrecht and Silverman, 1984). 3. Silica Diatoms have a strict requirement for dissolved amorphous silica to synthesize their silicious cell walls. Volcani and co-workers (Reeves and Volcani, 1985; Ludwig and Volcani, 1986)have also implicated Si in DNA synthesis in diatoms. This suggestion is based on the observation that when the diatom Cylidrofhecafusiformis is grown in synchronousculture, DNA synthesis is arrested if silicon is not provided. Moreover, Si limitation affects gene expression and mRNA stability in these organisms (Reeves and Volcani, 1985).

E. ZOOPLANKTON DEVELOPMENT AND GROWTH

Only a few studies have dealt with the molecular biology of zooplankton. The genome sizes of several copepod species have been characterized

2%

PAUL G . FALKOWSKI AND JULIE LaROCHE

by Feulgen microspectrophotometry (McLaren et al., 1988; 1989) and a correlation has been found between genome size, body size, and development time of embryos. McLaren et al. (1989) found intraspecific and interspecific variations of genome size within Pseudocalanus females. These authors remarked that although this relationship is well established for protists (Shuter et al., 1983) and plants (Bachmann et al., 1989, it has rarely been suggested for multicellular animals. McLaren ef al. (1989) suggest that the tendency to determinate nucleus sizes among copepods may be responsible for the apparent “nucleotypic” control of body size and development rates.

VI. Potential for Biotechaological Exploitation One motive for understanding molecular biological aspects of marine organisms is to exploit or enhance the production of the organism itself, or a product synthesized by the organism. For example, Powers (1989) has proposed genetically transforming trout with a gene encoding for growth hormone. Preliminary results indicate the resultant fish grows 25% larger than the wiid type. Genetic engineering of shellfish to improve yields has been proposed (Morse, 1984). Transformation systems in algae (Chlamydomonas) are at an early stage of development, and attempts are under way in several laboratories to use particle guns (Boynton et al., 1988) or other methods (Kindle, 1990) to insert foreign DNA into host organisms. Microalgae are potentially important sources of valuable lipids, pigments, and pharmaceuticals; however, their commercial value has not reached expectations primarily because of low yields and growth. It is hoped that genetic transformation will lead to increased commercialization;however, the potentials of genetic engineering have not been achieved in marine organisms to date.

VII. Conclusions We have described potential applications of molecular biological techniques to study specific biological oceanographic processes. Molecular techniques are particularly useful and undoubtedly will continue to be used to understand the evolution of organisms, gene flow, and population genetics, and to construct phylogenetic trees. We also have suggested the potential application of molecular techniques to understanding molecular mechanisms of key biogeochemcial processes, such as nitrification, carbon fixation, silica incorporation, and iron limitation. While an under-

MOLECULAR BIOLOGY IN STUDIES OF OCEAN PROCESSES

297

standing of mechanisms may not provide quantitative rate information, which is often desired by oceanographers, it provides a basis for understanding how physical and chemical ocean environments are related to complex biological phenomena. We hope this review fosters more crossfertilization between marine scientists and molecular biologists, to utilize molecular techniques, along with remote sensing, and in siru instrumentation, as a resource with which to address oceanographic problems.

ACKNOWLEDGMENTS The authors were supported by the U.S. Department of Energy, Office of Health and Environmental Research, and Office of Basic Energy Biosciences under Contract No. DEAC02-76CH00016. We thank Paul Kemp, Rose Anne Cattolico, Sang Hoon Lee, Jon Zehr, Linda Shapiro, Carl Price, Charlie Miller and Noel Cam for discussions and suggestions. We especially thank Avril Woodhead for her editorial assistance.

REFERENCES Alberte, R. S.,Wood, A. M., Kursar, T.A., and Guillard, R. R. L. (1984). Plant Physiol. 75, 732-739.

Aldrich, J., Gelvin, S., and Cattolico, R. A. (1982). Plant Physiol. 69, 1189-1 195. Attardi, G. (1985). Int. Rev. Cytol. 93,93-145. Avise, J. C., Helfman, J. S.,Saunders, N.C., and Hales, L. S. (1986). Proc. Natl. Acad. Sci. U.S.A. 83,4350-4354. Avise, J. C., Arnold, J., Ball, R. M.,Bermingham, E., Lamb, T.,Neigel, J. E., Reeb, C. A., and Saunders, N.C. (1987). Ann. Rev. Ecol. Syst. 18,489-522. Bachmann, K., Chambers, K. L., and Price, H. J. (1985). I n “The Evolution of Genome Size” (T. Cavalier-Smith, ed.), pp. 267-276. Wdey, New York. Baker, C. M., Palumbi, S. R., Lambertsen, R. H.,Weinrich, M. T., Calambokidis, J., and O’Brien, S. J. (1990). Nature (London) 344,238-240. Balch, W. M.,Yentch, C. M., Reguera, B., and McCue, F. C. (1988). I n “Immunochemical Approaches to Coastal, Estuarine, and Oceanographic Questions” (P .K. Horan ,ed.), pp. 263-276. Springer-Verlag, New York. Bates, S . S. (1981). J . Exp. Mar.Bid. Ecol. 51,219-240. Batuecas, B., Garesse, R., Calleja, M., Valverde, J. R., and Marco, R. (1988). Nucleic Acids Res. 16,6515-6529. Bentzen, P., Leggett, W. C.,and Brown, G. G. (1988). Genetics 118,509-518. Berg, W. J., and Fems, S. D. (1984). Can. J. Fish Aquat. Sci. 41, 1041-1047. Bergh, O., Borsheim, K. Y., Bratbak, G., and Heildal, G. (1989). Nature (London) 340, 467-468.

Bhattacharya, D., and Druehl, L. D. (1989). Mar. Biol. 1@2,15-23. Birt, T. P., Green, J. M.,and Davidson, W. S. (1986). Can.J . Zool. 64, 118-120. Boczar, B. A., Delaney, T. P.,and Cattolico, R. A. (1989). Proc. Natl. Acad. Sci. U.S.A.86, 4996-4999.

Bowen, B. W., Meylan, A. B., and Avise, J. C. (1989). Proc. Natl. Acad. Sci. U.S.A. 86, 573-576.

Boynton, J. E., Gillham, N.W., Harris, E. H., Hosler, J. P., Johnson, A. M., Jones, A. R.,

298

PAUL G. FALKOWSKI AND JULIE LaROCHE

Randolph-Anderson, 9. L., Robertson, D., Klein, T. M., Sharkj, K. B., and Sanford, J. C. (1988).Science 240, 1534-1538. Brayton, P. R., Tamplin, M. L., Huq, A., and Cofwell, R. R. (1987).53,2862-2865. Bremmer. H., and Dennis, P. P. (1987).I n “Escherichia coli and Salmonella typhimurium Cellular and Molecular Biology”(F. C. Neidhardt and J. Doe. eds.), pp. 1527-1542. Amer. Soc.Mirobiol., Washington, D. C. Brunk, C. F., Jones, K. C., and James. T. W. (1979).Anal. Biochem. 92,497-500. Bucklin, A., Rienecker, M. M., and Mooers, C. N. K. (1989).J. Geophys. Res. 94,82778288. Burkill. P.H.(1987).In “Microbes in the Sea” (M. Sleigh, ed.), pp. 139-166. Ellis Honvood Ltd., England. Campbell, L.. Carpenter, E. J., and Iacono, V. J. (1983).Appl. Enuiron. Microbiol. 46, 553-559. Capone, D. G., and Carpenter, E. J. (1982).Science 217, 1140-1 142. Capone, D. G.. O’Neil, J. M., Zehr, J., and Carpenter G. J. (1990).Appl. Enuiron. Microbiol. 56,3532-3536. Carpenter, E. J., and Capone, D. G., (1983). In “Nitrogen in the Marine Environment” (E. J. Carpenter and D. G. Capone, eds.),p. 900. Academic Press, San Diego. Carr, A., and Coleman, P. J. (1974).Nature (London)249, 128-130. Cavanaugh, C. M. (1983).Nature (London)302,5841. Chapman, R. W., and Powers, D. A. (1984).Tech. Rep. N o . UM-SG-TS-84-05.Maryland Sea Grant Program, College Park, Maryland. Chien, A., Edgar, D. 9..and Trela, J. M. (1976).J. Bacteriol. 127, 1550-1557. Codispoti, L. A. (1989).I n “Productivity of the Ocean; Present and Past” (W.H. Berger, V. S. Smetacek, and G. Weber, eds.), pp. 377-394. Wdey Ltd., London. Coleman, L. W., Rosen, B. H., and Schwartzbach, S. D. (1988).Plant Cell Physiol. 29, 1007-1014. Codey, P. B., Lemieux, P. G., Lomax, T. L., andGrossman, A. R. (1986).Proc. Natl. Acad. Sci. U.S.A. 83,3924-3928. Coveney, M. F., and Wetzel, R. G. (1988).Appl. Enuiron. Microbiol. 54,2018-2026. Cumn, C. A., Paerl, H. W., Suba, G. K.. and Alberte, R. S. (1990).Limnol. Oceanogr. 35, 59-7 1. Darnell, J., Lodish. H., and Baltimore, D. (1986). “Molecular Cell Biology,” p. 1187. American Books, Inc., New York. DeLong, E. F., Wickham, G. S., and Pace, N. R. (1989).Science 243, 1360-1363. Deming, J. W., Hada, H., Colwell, R. R., Luehrsen, K. R., and Fox, G. E. (1984).J . Gen. Microbiol. WO, 191 1-1920. Donis-Keller, H., Maxam, A.. and Gilbert, W. (1977).Nucleic Acids Res. 4,2527-2538. Douglas, S. E. (1988).Curr. Genet. 14,591-598. Douglas, S. E., and Durnford, D. (1989).Plant Mol. Biol. W, 13-20. Drouin, G . , Hofman, J. D., and Doolittle, W. F. (1987).J. Mol. Biol. 196,943-946. Ducklow, H. W. (1983).Biol. Sci. 33,494-501. Dugdale, R. C., and Goering, J. (l%7). Limnol. Oceanogr. l2,196-206. Dunbar, M. J. (1960).A m . N a t . 94,129-136. Dunlap, J. C., and Hastings, J. W. (1981).J. Biol. Chem. 256, 10509-10518. Edmon, J. M., and VonDamm, K. (1983).Sci. Am. 248,78-93. Edmunds, L. N., Jr., Tay, D. E.. and Laval-Martin, D. L. (1982).Plant Physiol. 70,297-302. Edward, C. A., and Skibinski, D. 0. F. (1987).Mar. Biol. 94,547-556. Edwards, U., Rogall, T., Blkker, H., Emde, M., and Bottger, C. (1989).Nucleic Acids Res. 17,7843-7853. Engelbrecht, J., and Silverman, M. (1984).Proc. Natl. Acad. Sci. U.S.A. 81,4154-4158.

MOLECULAR BIOLOGY IN STUDIES OF OCEAN PROCESSES

299

Eppley, R. W., and Petersen, B. J. (1979). Nature (London)282,677-680. Falkowski, P. G. (1980). I n “primary Productivity in the Sea” (P. G. Falkowski, ed.), pp. 99-119, Plenum Press, New York. Falkowski, P. G., and Owens, T. C. (1982). Limnol. Oceanogr. (1982)27,776782. Falkowski, P. G., Sukenik, A., and Herzig, R. (1989). J . Phycol. 25,471-478. Falkowski, P. G., Flagg, C. N., Kemp, P., Kolber, Z., Smith, S . L., Wallace, D., and Wirick, C. D. (1990). Oceanography 5, (in press). Fergusson, R. L., Buckley, E. N., and Palumbo, A. V. (1984). Appl. Enuiron. Microbiol. 47, 49-55.

Forward, R. B., Jr. (1988). Oceanogr. Mar. Biol. 26,361-393. Fox, G . E.,Stackebrandt, E., Hespell, R. B., Gibson, J., and Maniloff, J. (1980). Science 209,457-463.

Francis, M. A., Suh, E. R., and Dudock B. S. (1990). J . Biol. Chem. 264,17243-17249. Freifelder, D. (1983). “Molecular Biology,” Van Nostrand Reinhold, New York. Friedman, A. L., and Alberte, R. S. (1986). Plant Physiol. 80,43-51. Frost, B. W. (1989). Can. J. Zool. 67,525-551. Fuhrman, J. A., and Azam. F. (1982). Mar. Biol. 66, 109-120. Fuhrman, J. A., Hagstromand, A., and Chan, A. M. (1988). Appl. Environ. Microbiol. 54, 1426- 1429.

Gallagher, J. C. (1982). J . Phycol. 18, 148-162. Gallagher, J. C., and Alberte, R. S . (1985). J. Exp. Mar. Biol. Ecol. 94,233-250. Garside, C. (1982). Mar. Chem. 11, 159-167. Gendel, S., Ohad, I., and Bogorad, L. (1979). Plant Physiol. 64,786-790. Giovannoni, S . J., DeLong, E. F., Olsen, G. J., and Pace, N. R. (1988). J. Bacterial. 170, 720-726.

Giovannoni, S. J., Britschgi, T. B., Moyer, C. L., and Field, K. G. (1990). Nature (London) 345,60-63.

Glazer, A. N. (1986). Annu. Reu. Biochem. 52, 125-157. Glover, H. E. (1989). Int. Rev. Cytol. 115,67-138. Goff,L . J., and Coleman, A. W. (1988). J. Phycol. 24,357-368. Golden, J. W., Robinson, S. J., and Hasselkorn, R. (1985). Nature (London) 314, 419423.

Golden, S. S . , Brusslan, J., and Haselkom, R. (1987). EMBO J. 5,2789-2798. Gould, S. J. (1978). Nat. Hist. 87,22-28. Grant, D., and Chiang, K.-S. (1980). Plasmid 4,82-86. Graves, J. E., Ferris, S. D., and Dizon, A. E. (1984). Mar. Biol. 79,315-319. Grossman, A. R., Anderson, L. K., Conley, P. B., and Lemaux, P. G. (1989). I n “Algae as Experimental Systems,” (A. W. Coleman, L. J. Goff, and J. R. Stein-Taylor, eds.) pp. 269-288. Alan R. Liss, Inc. Gutell, R. R., Weiser, B:, Woese, C. R., and Noller, H.F. (1985). Prog. Nucleic Acid Res. Mol. Biol. 32, 155-216. Harris, E. H. (1989). I n “The Chlamydomonas Sourcebook A Comprehensive Guide to Biology and Laboratory Use.” Academic Press, San Diego, California. Hastings, J. W., and Sweeney, B. M. (1958). Biol. Bull. 115,440-458. Hattori, A. (1983). In “Nitrogen in the Marine Environment” (E. J. Carpenter and D. G. Capone, eds.), pp. 191-232. Academic Press, San Diego, California. Haygood, M. G., and Cohn, D. H. (1986). Gene 45,203-209. Haygood, M. G., and Nealson, K. H. (1985). J . Bacteriol. 162,209-216. Hayward, T. L. (1987). Deep-Sea Res. 34, 1593-1627. Hinnebusch, A. G., Klotz, L. C., Blanken, R. L., and Loeblich, A. R. 111. (1981). J. Mol. Evo~.17,334-347.

300

PAUL G. FALKOWSKI AND JULIE LaROCHE

Hobbie, J. E., Daley, R. J., and Jasper, S. (1977). Appl. Enuiron. Microbiol. 33, 1225-1228. Hochachka, P. W., and Somero, G. N. (1984). In “Biochemical Adaptation,” p. 537 Princeton UNv. Press., Princeton, NJ. Holben, W. E., Jansson, J. K., Chelm, 9. K., and Tiedje, J. M. (1988). Appl. Enuiron. Microbiol. 54,703-71 I . Holm-Hansen, O., Sutcliff, W. H. Jr., and Sharp, J. (1968). Limnol. Oceanogr. W, 507-514. Huynh, T. V.,Young, R.A., and Davis, R. W. (1986). In “DNA Cloning, Vol. I, A Practical Approach” (D.M. Glover, ed.), p. 49-78. IRL Press Oxford, Washington, D. C. Hwang, S., and Tabita, F. R. (1989). Plant Mol. Biol. W, 69-79. innis, M. A., Gelfand, D. H., Svinsky, J. J., and White, T. J. (1990):‘ PCR Protocols. A Guide to Methods and Applications.” p. 482. Academic Press, San Diego, California. Jacobs, H. T., Elliott, D. J., Math, V. B., and Farquharson, A. (1988). J. Mol. Biol. 202, 185-2 17.

Jannasch, H. W., and Mott, M. J. (1985). Science 22!3,717-725. Johnson, C. H.. Roebert, J. F., and Hastings, J. W. (1984). Science 223, 1428-1430. Johnson, P. W., and Sieburth, J. McN. (1982). J . Phycol. 18,318-327. Kaiser, K.,and Murray, N. E. (1986). In “DNA Cloning, Vol. I. A Practical Approach” (D. M.Glover, ed.), p. 1-47. IRL Press Oxford, Washington, D. C. Kaplan, W. A. (1983). I n “Nitrogen in the Marine Environment” (E.J. Carpenter and D.G. Capone, eds.). pp. 139-190. Academic Press, San Diego, California. Keen, J. N., Pappin, D. J. C., and Evans, L. V. (1988). 1.Phycol. 24,324-327. Kindle, K.L. (1990). Proc. Nat.Acad. Sci. U.S.A. 87, 1228-1232. Kirchman, D., Newell. S., and Hodson, R. (1986). Mar. Ecol. Frog. Ser. 32,47-59. Kjelleberg, S., Hermansson, M., Marden, P., and Jones, G . W. (1987). Annu. Rev. Microbiol. 41,2549. Kogure, K. V., and Taga, N. (1979). Can. J . Microbiol. 25,415-420. Kolber, Z., Wyman, K.D., and Falkowski, P. G. (1990). Limnol. Oceanogr. 35,72-79. Kumar, S., Degman, 9. M., Rosa, I. L., Hawkins, C. J., and Lavin, M. F. (1988). Mar. Biol. Lett. 98,95-100.

Kurath, G., and Morita, Y. (1983). Appl. Enuiron. Microbiol. 45, 1206-1211. Kyle. D. J ..Ohad, I., and Amtzen, C. J. (1984). Proc. Nail. Acad. Sci. U.S.A. 81,4O70-4O74. Laemmli, U. K. (1970). Nature (London)277,680-685. Lane, D. J., Pace, B., Olen, G. J., Stahl, D. A., Sogen, M. L., and Pace, N. R. (1985). Proc. Natl. Acad. Sci. U.S.A. 82,6955-6959. Lane, D. J., Field. K. G., Olsen, G. J., and Pace, N. R. (1988). I n “Methods in Enzymology” (L. Packer and A. Glazer, eds.), Vol. 167, pp. 134-144. Academic Press, San Diego, California. LaRoche, J., Bennett, J., and Falkowski, P. G. (199Oa). Gene%, 165-171. LaRoche, J., Mortain-Bertrand, A., Bennett, J., and Falkowski, P. G. (1990b). Proc. Vlllrh Int’l. Phorosyn. Cong., Stockholm, W. Junk, Amsterdam (in press). LaRoche, J., Snyder, M., Cook, D. I., Fuller, K., and Zouros, E. (1%). Mol. Biol. Euol. 7 , 45-64.

Lee, S., and Fuhrman. J. A. (1990). Appl. Enuiron. Microbiol. 56,739-746. Li, N., and Cattolico, R. A. (1987). Mol. Gen. Genet. u)9,343-351. Ludwig, J. R., and Volcani, B. E. (1986). I n “Biomineralization in Lower Plants and Animals” (9. S. C. Leadbeaker and R. Riding, eds.), p. 10. The Systematic Assoc., Special Vol. 30. Oxford Univ. Press, New York. Manhart, J. R., Kelly, K., Dudock, 8 . S., and Palmers, J. S. (1989). Mol. Gen. Genet. 216, 417-421.

Maniatis, T., Fritsch, E. F., and Sambrook, S. (1982). “Molecular Cloning: A Laboratory Manual,” Cold Spring Harbor Laboratory, Cold Spring Harbor, New York.

MOLECULAR BIOLOGY IN STUDIES OF OCEAN PROCESSES

301

Martin, J. H. and Fitzwater, S. E. (1988). Nature (London) 331,314-343. Martin, J. H., and Gordon, M.(1989). Deep-sea Res. 35, 177-1%. Mazur, B. J., Rice, D., and Haselkom, R. 1980. Proc. Natl. Acad. Sci. U.S.A. 77,186-190. McLaren, I. A.. Sevigny, J.-M., and Corkett, C. J. (1988).Hydrobiologia 167/168,275-284. McLaren, I. A., Sevigny, J.-M., and Corkett, C. J. (1989). Can. J. Zool. 67,559-569. McMillan, J., and Miller, D. J. (1990). Mar. Biol. Lett. 104,483-488. McMillan, J., Yellowlees, D., Heyward, A., Harrison, P., and Miller, D. J. (1988).Mar. Biol. Lett. 98,271-276. Medlin, L., Elwood, H. J., Sticekl, S.,and Sogin, M.L. (1988). Gene 71,491-499. Mitsui, A., Kumazawa, S., Takahashi, A., Ikemoto, H., Cao, S., and Arai, T. (1986). Nature (London) 323,720-722. Moir, R. D., and Dixon, G. H. (1988). J . Mol. Euol. 27,8-16. Morita, R. Y.(1982). Adu. Microbial Ecol. 6, 171-198. Moritz, C., Dowling. T. E., and Brown, W. M.(1987). Annu. Reu. Ecol. Syst. 18,269-292. Morse, D. E. (1984). Aquaculture 39,263-282. Morse, D., Mdos, P. M.,Roux, E.,and Hastings, J. W. (1989). Proc. Natl. Acad. Sci. U.S.A., 86,172-176. Murphy, L. S.(1978). J. P h y ~ ~14,272-250. l. Murphy, L. S.,andGuillard, R. R. L. (1976). J. Phycol. U,9-13. Neale, P. J . (1987). I n “Photoinhibition” (D. J. Kyle, C. B. Osmond, and C. J. Arntzen, eds.), pp. 39-65. Elsevier, Amsterdam. Newman, S.,and Cattolico, R. A. (1987). Plant Physiol. 84,483-490. Novitsky, J. A., and Morita,R. Y.(1978). Mar. Biol. Lett. 48,289-295. Oelmueller, R.,Conley, P. B. Federspiel, N., Briss, W. R., and Grossman, A. R. (1988). Plant Physiol. 88, 1077-1083. Ohad, I., Kyle, D. J., and Hirschberg, J. (1986). EMBOJ. 4, 1655-1659. Ohki, K., Zehr, J., Falkowski, P. G., and Fujita, Y. (1991). Arch. Microbiol. (in press). Olsen, R. J., Frankel, S. L., Chisholm, S. W., and Shapiro, H. M.(1983). Exp. Mar. Biol. E d . 68,129-144. Olsen, G. J., Lane, D. J., Giovannoni, J., and Pace, N. R. (1986). Annu. Rev. Microbiol. 40, 337-365. Ovenden, J. R., White, R. W.G., and Sauger, A. C. (1988). J. Fish Biol. 32,137-148. Pace, N. R., Stahl, D. A., Lane, D. J., andOlsen, G. T. (1986).Adu. Microbiol. Ecol. 9,1-55. Paul, J. H., Deflaun, M. F., Jeffrey, W. H., and David, A. W. (1988). Appl. Enuiron. Microbiol. 54,331-336. Peattie, D. A. (1979). Proc. Natl. Acad. Sci. U.S.A. 76, 1760-1764. Perry, N. J., Talbot, M.C., and Alberte, R. S.(1981). Mar. Biol. 62,91-101. Pick, U., Gounaris, K., and Barba, J. (1987). Plant Physiol. 85, 194-198. Platt, T., and Sathyendranath, S. (1988). Science 241, 1613-1620. Plumley, F. G., and Schmidt, G. W. (1989). Proc. Natl. Acad. Sci. U.S.A. 86,2678-2682. Pollard, P. C., and Moriarty, D. J. W. (1984). Appl. Enuiron. Microbiol. 48, 1076-1083. Posakony, J . W., Flytzanis, C. N., Britten, R. J., and Davidson, E. H.(1983). J . Mol. Biol. 167,361-389. Post, A. F., Dubinsky, Z., Wyman, K., and Falkowski, P. G. (1984). Mar. Biol. 83, 231238. Powers, D. A. (1989). Science 246,352-358. Powers, D., Allendorf, F., and Chen, T. (1990). In “Patterns, Processes and Large Maine Ecosystems” (K. Sherman, L. M.Alexander, and B. D. Gold, eds.), pp. 104-121. Amer. Assoc. Acad. Sci. Prezelin, B. B., and Sweeney, B. M.(1977). Plant Physiol. 60,388-392. Proctor, L. M.,and Fuhrman, J. A. (1990). Nature (London) 343,60-62.

302

PAUL G. FALKOWSKI AND JULIE LaROCHE

Reeves, C. D., and Volcani, B. E. (1985). 1.Gen. Microbiol. U1, 1735-1744. Reith, M.,and Cattolico, R. A. (1986). Proc. Natl. Acad. Sci. U.S.A. 83,8599-8603. Riethman, H. C., and Sherman, L. A. (1988). Plant Physiol. 88,497-505. Robertson, B. R., and Button, D. K. (1989). Cytometry 10,70-76. Roszak, D. B., and Colwell, R. R. (1987a). Appl. Environ. Microbiol. 53(12), 2889-2983. Roszak, D. B., and Colwell, R. R. (198%). Microbiol. Rev. 51, 365-379. Saiki, R. K., Scharf. S., Faloona, F.. Mullis, K. B., Horn, G. T., Erlich, H. A., and A h e i m , N. (1985). Science 230,1350-1354. Saiki, R. K., Gelfand, D. H., Stoffel, S.,Scharf, S. J., Higuchi, R., Horn, G. T., Mullis, K. B., and Erlich, H. A. (1988). Science 239,487-494. Sambrook, J., Fritsch, E. F., and Maniatis, J. (1989). “Mo;ecular Cloning: A Laboratory Manual,” 2nd Edition, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. Sanders, H. L. (1968). Am. Nut. 102,243-282. Scharf, S. J., Horn, G. T., and Erlich, H. A. (1986). Science 233, 1076-1078. Setlow, R. B. (1974). Proc. Notl. Acad. Sci. U.S.A. n,3363-3366. Sevigny, J.-M., and McLaren, I. A. (1988). Hydrobiologia 167/168,267-274. Sevigny, J.-M., McLaren, I. A., and Frost, B. W. (1989). Mar. Biol. 102,321-327. Sheen, J.-Y. and Bogorad, L. (1988). Plant Physiol. 86, 1020-1026. Sherr, E. B., and Sherr, B. F. (1988). Limnol. Oceanogr. 33, 1225-1227. Sherman, L. A., Reddy. K. J., Reithman, H. C., and Bullerjahn, G. S. (1987). I n “Progress in Photosynthesis Research” ( J . Biggins, ed.), Vol. 4, pp. 773-776. Martinus Nijhoff, Dordrecht. Shuter, B. J., Thomas, J. E., Taylor, W. D., and Zimmerman, A. M.(1983). Am. Nat. 122, 26-44. Sieburth, J. McN., Smetacek, V.,and Lenz, J. (1978). Limnol. Oceanogr. 23, 1256-1263. Skibinski, D. 0. F. (1985). J . Exp. Mar. Biol. Ecol. 92,251-258. Snyder, M., Fraser, A. R., LaRoche, J., Gartner-Kepkay, K. E., and Zouros, E. (1987). Proc. Nail. Acad. Sci. U.S.A. 84,7595-7599. Soldo. A. T., Merlin, E. J., and Godoy, G. A. (1981). J . Gen. Microbiol. 123,259-267. Somerville, C. C., Knight, 1. T., Straube, W. L., and Colwell, R. R. (1989). Appl. Enuiron. Microbiol. 55,548-554. Southern, E. M. (1972). 3. Mol. Biol. 98,503-517. Stabile, J. H., Gallagher, J. C., and Wurtzel, T. (1990). Biochem. Syst. Ecol. 18,5-9. Stahl, D. A., Lane, D. J., Olsen, G. J., and Pace, N. R. (1984). Science 224,409-41 I. Sukenik, A., Bennett, J., Mortain-Bertrand, A., and Falkowski, P. G. (1990). Plant Physiol. 92,891-898. Sweeney, B. M.(1982). Plant Physiol. 70,272-276. Torrella, F. and Morita, R. Y. (1981). Appl. Enuiron. Microbiol. 1u),587-603. Ward, B. B. (1984). Limnol. Oceanogr. 29(2), 402-410. Ward, B. B. (1990). Oceanography 3,30-35. Ward, B. B., Kilpatrick, K. A., Novelli, P. C.. and Scranton, M.1. (1987). Nature (London) 327(21), 226-229. Ward, D. M.,Weller, R., and Bateson, M. M. (1990). Nature (London)345.63-65. Watson, J. D., Hopkins, N. A., Roberts, J. W., Steitz, J. A., and Weiner, A. M. (1987). “Molecular Biology of the Gene.” Benjamin Cummings, Menlo Park, California. Weller, R., and Ward, D. M.(1989). Appl. Environ. Microbiol. 55, 1818-1822. WoodweU, G. M., and Smith, H. H. (1%9). Brookhaven Symp. Biol. 22,25-37. Wrangstradh, M.,Conway, P. L., and KjeUeberg, S. (1986). Arch. Microbiol. 145,220-227. Wrischnik, L. A., Higuchi, R. G., Stoneking, hi., Erlich, H. A., Amheim, N., and Wilson, A. C. (1987). Nucleic Acids Res. 2,529-541.

MOLECULAR BIOLOGY IN STUDIES OF OCEAN PROCESSES

303

Yentsch, C. M., Mague, F. C., and Horan, P. K. (1988). “Lecture Notes on Coastal and Estuarine Studies. 25. Immunochemical Approaches to Coastal, Estuarine and Oceanogmphic Questions.” Springer-Verlag, New York. Zehr, J. P., and McReynolds, L. A. (1989). Appl. Environ. Microbiol. 55,2522-2526. Zehr, J. P., Capone, D. G . , and Falkowski, P. G. (1989). Mar. Ecol. Prog. Ser. 51,237-241.

This Page Intentionally Left Blank

INDEX

A Actin cell division in diatoms and, 98 plastids and, 16, 18,20,22,56 Adaptation, ocean processes and, 261,284, 289-292 chromatic adaptation, 292 irradiance levels, 289-292 Addressin, regenerated lymph node transplants and, 250 Adenomas, calcium regulating hormones and, 144, 154 Adenylate cyclase, calcium regulating hormones and, 149, 153, 157-158, 177 Adrenalectomy , calcium regulating hormones and, 156, 174 Adrenergic effects, calcium regulating hormones and, 142, 153, 157-158, 167, 173 Affinity, nerve growth factor and, 123-126, 129-132 Age regenerated lymph node transplants and, 250 regenerated splenic transplants and, 230-231,235-236 function of regenerated tissues, 241-242,244 Algae cell division in diatoms and, 86, 100 ocean processes and, 2% application of molecular techniques, 274,280 molecular approaches, 266-268, 270-273

organism function, 289-290,292 plastids and, 57 cytoplasmic inheritance, 37,43,45, 48-49,52 differentiation, 27 division, 11, 13, 18,20,22-23 isogamy, 38-43 location of nuclei, 3-6 Allium cepa, plastids and, 13 Allophycocyanin, ocean processes and, 292 Allozymes, ocean processes and, 269, 275-276 Amia calua, calcium regulating hormones and, 187 Amino acids calcium regulating hormones and, 142-143, 148, 167, 189 nerve growth factor and, 112-1 13, 123-124 ocean processes and, 270-273,290,293 plastids and, 27 Amino terminal parathyroid hormone and, 143-144, 154-155, 163 stanniocalcin and, 189 Ammonium, ocean processes and, 286-288 Amphibians, calcium regulating hormones and, 141, 177 calcitonin, 167-171, 183-184, 199-200 parathyroid hormone, 142, 148, 1% prolactin, 179-180, 197 Amyloplasts, organelle nuclei and, 2, 56 differentiation, 26-28,30,34 division, 20,23-24 Anacystis nudulans, ocean processes and, 295 305

306

INDEX

Anaphase, cell division in diatoms and, 70, 98,101-102 collar, 86 kinetochores, 80-81 microtubule stability, 94.96-97 Anaphase A, cell division in diatoms and, 98-103 central spindle, 70 COIIW,86-89 Anaphase B, cell division in diatoms and, 64,70,82-83 physiology, 93-94 ultrastructure, 89-93 Ancillary features, cell division in diatoms and, 71 Angiosperms, plastids and. 48.52-53 Anisogamy, plastids and, 37,44-45, 47-48, 52,57 Annulus, plastids and, 22, 24 Antibodies monoclonal, see Monoclonal antibodies nerve growth factor and, 110, 114, 123, I25 ocean processes and, 265,270,272,274, 287,294 plastids and, 20,40,45 regenerated splenic transplants and, 222, 227,233.243-246 Antibody probes, ocean processes and, 269 Antigens calcium regulating hormones and, 147 ocean processes and, 274 regenerated lymph node transplants and, 247,250 regenerated splenic transplants and, 217, 220,222,225 function of regenerated tissues, 243-245 iduences on regeneration, 233,238 Apoprotein, ocean processes and, 290 Apoptosis, calcium regulating hormones and, 171 Aquatic vertebrates, calcium regulating hormones and, see Calcium regulating hormones Areas of reduced birefringence, cell division in diatoms and, % ATP calcium regulating hormones and, 149 cell division in diatoms and, 86.88-89, 93-95. 101

nerve growth factor and, 124 ocean processes and, 281, 293 ATPase calcium regulating hormones and, 181 plastids and, 1-2 regenerated splenic transplants and, 224 Atriplex semibaccata, plastids and, 21 Autofluorescence, plastids and, 4, 16 Autoradiography calcium regulating hormones and, 143, 154 nerve growth factor and, 125-126, 130 ocean processes and, 270,274,280, 285, 287 Autotransplant lymph node, see Regenerated lymph node transplants regenerated splenic, see Regenerated splenic transplants Autotrophs. ocean processes and, 271 Avena sativum, plastids and, 22 Axons, nerve growth factor and, 131-132 neurotrophic theory, 109-1 11 receptor expression, 126-128 synthesis, 114-116, 118-122

B B cells, regenerated splenic transplants and, 220,222,242-245 Bacteria ocean processes and application of molecular techniques, 274,277-280 marine organisms, 262-264 molecular approaches, 266,271 organism function. 284-287,294 plastids and, 4, 13, 38 Basal bodies, cell division in diatoms and, 69 Basal forebrain, nerve growth factor and, 1 1 1 , 115, 119, 129, 131 BDNF, see Brain-derived neurotrophic factor Bioamines, calcium regulating hormones and, 152-153, 166-167, 173-174 Bioluminescence, ocean processes and, 293,295 Birds, calcium regulating hormones and, 143, 168-171, 1%. 199-200

307

INDEX Birefringence, cell division in diatoms and, 95-96

Blood flow, regenerated splenic transplants and, 240 Bone, calcium regulating hormones and, 140

calcitonin, 166, 169, 174, 198-199 parathyroid hormone, 142, 156, 196 prolactin, 181, 197 resorption, 174 stanniocalcin, 201-202 Brain-derived neurotrophic factor, 110, 112 biochemistry, 124- 125 nerve growth factor receptor expression and, 125, 127, 130, 132 synthesis, 117 Brassica, plastids and, 26 Brassical juncea, plastids and, 4 Bryopsis maxima, plastids and, 13,37,44 Bryopsis pulumos, plastids and, 6

degradation, 163- 166 extracellular calcium, 144-148, 158, 160-163, 194-1%

intracellular calcium, 158-163, 194 second messengers, 157-163 secretion, 144-156 prolactin in aquatic vertebrates, 179- 182, 197- 198 extracellular calcium, 182 stanniocalcin in aquatic vertebrates, 187- 189,201-202

extracellular calcium, 190, 193 second messengers, 193 secretion, 189-193 Calyptogena rnagnifca, ocean processes and, 278 Carbohydrate, ocean processes and, 285-286

Carbon Ocean processes and, 263,2% application of molecular techniques, 277-278

C

C cells, calcium regulating hormones and aquatic vertebrates, 184-185 terrestrial vertebrates, 166-167, 170-171, 173-177, 198

Calcitonin, 140-142, 198-202 in aquatic vertebrates, 178-179, 183-187, 192, 198-202

in terrestrial vertebrates, 153, 156, 166-169

second messengers, 176-177 secretion, 169-176 Calcitonin gene related peptide, calcium regulation and, 167 Calcium, plastids and, 36,41-42,52,57 Calcium regulating hormones, 139-142 aquatic vertebrates, 177-179 calcitonin in aquatic vertebrates, 183-186, 198-202

extracellular calcium, 184-185 calcitonin in terrestrial vertebrates, 166-169, 198-200

extracellular calcium, 168, 170-171, 177, 198-199

second messengers, 176-177 secretion, 169-176 parathyroid hormone in terrestrial vertebrates, 142-144, 194-197

organism function, 286,289,294 plastids and, 27 Carboxy terminal calcitonin and, 172 parathyroid hormone and, 143-144, 151, 154-155, 163-164

Catecholamines, calcium regulating hormones and, 153, 157, 174, 187 Cathespins, calcium regulating hormones and, 144 Caulerpa okamurae, plastids and, 6 cDNA nerve growth factor and, 113, 115, 123 ocean processes and, 272-273, 279,290 Cell cycle, cell division in diatoms and, 66-69

Cell division in diatoms, see Diatoms, cell division in Central spindle, cell division in diatoms and, 70-76 Centrally located plastid nuclei, 5-6, 11, 55-56

Centrics, cell division in diatoms and, 64,98

Centrioles, cell division in diatoms and, 69,86

Centrosomes, cell division in diatoms and, 66,68-69 Chlamydomonas, plastids and, 7, 37

308

INDEX

Chlamydomonas reinhardtii, ocean processes and, 281,293 Chloramphenicol, plastids and, 26.42 Chlorophyll Ocean processes and, 264,273,289 plastids and, 39.42 Chloroplast DNA Ocean processes and, 268,280-283 plastids and, I. 4 cytoplasmic inheritance, 38-39,41-43, 45.47 differentiation, 27-28, 30-3 1, 36 division, 9-10, 13 higher plants, 52. 54-55 Chloroplasts cell division in diatoms and, 66.68.98, 100 ocean processes and, 270,272,290, 292-294 plastids and, 1-3,55-57 cytoplasmic inheritance, 3645.47, 49-55 differentiation, 25-28, 30-36 division, 9-23,25 location of nuclei, 3,5-6 organization of nuclei, 6-7,9 Cholecystokinin, calcium regulating hormones and, 172-173 Chotinergic effects, calcium regulating hormones and, 142, 167 Cholinergic neurons, nerve growth factor and, 111 receptor expression, 129, 131 synthesis, 115, 118-119, 121 Chromatids, cell division in diatoms and, 86,88-89 Chromatin cell division in diatoms and, 76,80,83 plastids and, 32 Chromoplasts, organelle nuclei and, 2, 7,56 differentiation, 27-28, 30-32 Chromosomes cell division in diatoms and, 63.97, 99-103 central spindle, 70,72 collar, 83-89 kinetochores, 76,80432 nerve growth factor and, 124 plastids and, 2,7,9,56

differentiation, 32, 34 division, 9, 11, 13-15, 22 Chrysophytes, cell division in diatoms and, 98-99 Circadian rhythms, Ocean processes and, 293 Circular plastid nucleus, 6, 11, 13,27, 55-56

Cleavage cell division in diatoms and, 94.98 nerve growth factor and, 113 Clones nerve growth factor and, 109, 112-1 13 Ocean processes and application of molecular techniques, 275-276.278-279 molecular approaches, 267,271 organism function, 295 Closed circulation, regenerated splenic transplants and, 218-219 Colchicine, cell division in diatoms and, 83, 100 Collar, cell division in diatoms and, 83-89,99 Compartments, regenerated splenic transplants and, 236,239 functional anatomy, 219-223 phases, 225,227, 230 Consensus sequences, nerve growth factor and, 124 Corpuscles of Stannius, see Stanniocalcin Coscinodiscus wailesii, cell division in diatoms and, 98 Cyanidium caldarium, organelle nuclei and division, 11, 13, 16, 18,20. 23 location of nuclei, 5-6 Cyanobacteria, ocean processes and, 263 application of molecular techniques, 274, 279 organism function, 288-289,292,295 Cyclic AMP calcium regulation and calcitonin, 176-177, 186 parathyroid hormone and, 153, 157-158

stanniocalcin, 193- 194 nerve growth factor and, 122 Cycloheximide, plastids and, 41 Cylidrotheca fusiformis, ocean processes and, 295

309

INDEX Cymatopleura, cell division in diatoms and, 66,68,98,100 Cynops pyrrhogaster, calcium regulating hormones and, 180 Cysteine, nerve growth factor and, 124 Cytochalasin, cell division in diatoms and, 94,98 Cytochalasin B,plastids and, 16,20 Cytokinesis cell division in diatoms and, 64,66,100 plastids and, 9, 18,20 Cytoplasm calcium regulating hormones and calcitonin, 170, 176 parathyroid hormone, 143, 160-161, 165, 194 stanniocalcin, 187,201 cell division in diatoms and, 63,99-100 central spindle, 72 microtubule center, 66,68-69 nerve growth factor and, 124 ocean processes and, 290 plastids and, 1-2,56-57 algae, 38-48 division, 18.23 higher plants, 48-55 inheritance, 36-38 Cytoskeleton, cell division in diatoms and, 66,68-69,88,94, 100

D DAPI ocean processes and, 264,287 plastids and, 2-4, 16 cytoplasmic inheritance, 39-41,45,49 differentiation, 26.31-32 Degradation, calcium regulating hormones and, 144, 147-148, 163-166, 194 Demineralization, calcium regulating hormones and, 169, 171,1%, 200 Denaturation mapping, plastids and, 1 Dendrites, nerve growth factor and, 109, 112

Dendritic cells, regenerated splenic transplants and, 222,227 Denitrification, ocean processes and, 286-289

Depolarization, calcium regulating hormones and, 146-147, 160, 171 Deproteinization, plastids and, 31 Diacylglycerol, calcium regulating hormones and, 162, 177, 194 Diatoma, cell division in diatoms and, 71,76

Diatoms, cell division in, 63-64,98-99 anaphase A, 101-103 central spindle, 70-76 cleavage, 98 COUX,83-89 early work, 64-65 kinetochores, 76-83 microtubule center basal bodies, 69 cell cycle, 66-69 interphase, 64.66-67 valve morphogenesis, 69 microtubule stability, 94-97 spindle anaphase B, physiology of, 93-94 evolution of, 99-100 ultrastructure, 89-93 Diatoms, Ocean processes and, 275,290,295 Dibutyryl CAMP,calcium regulating hormones and, 157,186, 193 Dictyosteliurn, cell division in diatoms and, 76,93 Dientoamoeba, cell division in diatoms and, 70 Dhydropyridine, calcium regulating hormones and, 160,194 1.25 Dihydroxycholecalciferol, parathyroid hormone and, 153 Dinitrophenol, cell division in diatoms and, 87-88 Ditylum, cell division in diatoms and, 66 DNA, see also Chloroplast DNA; Plastid

DNA nerve growth factor and, 125 ocean processes and, 262,264,2% application of molecular techniques, 275-276,278-283

molecular approaches, 265-268, 27 1-273

organism function, 285,293,295 plastids and, 1-4,56-57 cytoplasmic inheritance, 38-43,45-49, 52,54-55

310

INDEX

differentiation, 26-36 division, 9, 11, 14-15 organization of nuclei, 7, 9 DNA fingerprinting, ocean processes and, 272 DNA polymerase, Ocean processes and, 27 1 DNase, plastids and, 41 Dopamine, calcium regulating hormones and. 152. 157, 161, 166, 174. 182 Dorsal root ganglion (DRG), nerve growth factor and, 124, 127-130 Dot blots, ocean processes and, 269 Dynein, cell division in diatoms and, 91

E Ecology, cell division in diatoms and, 63 EcoRI, plastids and, 14 Ectocarpus, plastids and, 6, 31, 37 Elaioplasts, plastids and, 27-28 Elasmobranchs, calcium regulating hormones and, 184,202 Electron-dense deposits, plastids and, 18, 20, 22-24 Electron microscopy calcium regulating hormones and, 189 cell division in diatoms and, 66,80, 83, 89.95, 102 plastids and, 2-4.9, 20, 25 cytoplasmic inheritance, 37,42,48 regenerated splenic transplants and, 2 18 Electron-transport area, plastids and, 3 Electrophoresis ocean processes and, 268-269,275 plastids and, 42 ELISA, ocean processes and, 269 Endonuclease, plastids and, I, 14,38,41 Endoplasmic reticulum calcium regulating hormones and, 142-143, 156, 166, 170. 187 nerve growth factor and, 113 Endospores, plastids and, 13, 16 Endothelium regenerated lymph node transplants and, 247,250 regenerated splenic transplants and, 219, 226

Environment calcium regulating hormones and, 177, 184-185, 199,201 ocean processes and, 261,264, 267,297 application of molecular techniques, 275-277,279 organism function, 285-289 Enzyme histochernistry, regenerated splenic transplants and, 224 Enzymes calcium regulating hormones and, 143-144, 149, 154, 158, 162-163 nerve growth factor and, 113-1 14, 125 ocean processes and, 265 molecular approaches, 269-270 organism function, 285, 288. 293-294 plastids and, 1-2, 14-15, 38,41 Epidermal growth factor receptor, 124 Epifluorescence microscopy ocean processes and, 274 plastids and, 3-4,26, 39,41 Epinephrine, calcium regulating hormones and, 152-153, 173 Epithelium nerve growth factor and, 114, 118, 131 regenerated lymph node transplants and, 250 Epitopes, ocean processes and, 270 Erythrocytes, regenerated splenic transplants and, 217, 219, 224,226, 24 I Erythromycin, plastids and, 42 Escherichia coli plastids and, 30 regenerated splenic transplants and, 240 Esterase, regenerated splenic transplants and, 220,225 Estradiol, calcium regulating hormones and, 156, 186, 192 Estrogen calcitonin and, 174-175, 198, 200 parathyroid hormone and, 156, 195-197 prolactin and, 182, 197 stanniocalcin and, 193 Etioplasts, organelle nuclei and, 4, 6-7,56 differentiation, 25-27, 31-32 division, 11, 14 Eubacteria, ocean processes and, 278-280 Euglena, plastids and, 31, 55

31 1

INDEX Eukaryotes, ocean processes and, 263, 294 application of molecular techniques, 274, 277-278,280 molecular approaches, 266,268 Evolution, ocean processes and, 265,273, 296 application of molecular techniques, 276-278,280,282 Exocytosis, calcium regulating hormones and, 143, 166, 170 Exons, nerve growth factor and, 113, 124

F Ferns, organelle nuclei and, 57 cytoplasmic inheritance, 37,45, 48-49,52 division, 13, 22 Fibroblasts, nerve growth factor and, 115, 122 Filter function, regenerated splenic transplants and, 241-242 Fish calcium regulating hormones and, 139, 141, 177-179 calcitonin, 167, 170-171, 183-186, 198, 200 stanniocalcin, 187- 188, 191- 193,200, 202 ocean processes and, 262,268-269,215 Flagella cell division in diatoms and, 69,91 plastids and, 45 Flow cytometry, ocean processes and, 274-275 Fluorescence ocean processes and, 279-280,286-287, 292 plastids and, 2-4, 16, 18 cytoplasmic inheritance, 40,42,49 differentiation, 31-32 Fluoride, calcium regulating hormones and, 162 Fluorochromes ocean processes and, 274-275 plastids and, 2,4 5-Fluorodeoxyuridine, plastids and, 42-43

Follicular dendritic cells, regenerated splenic transplants and, 222 Fragillaria, cell division in diatoms and, 76 Funaria hygrometrica, plastids and, 20 Fusarium, cell division in diatoms and, 93 Fuzzy plaques, plastids and, 21-22,24

G G protein, calcium regulating hormones and, 160, 162 Gametogenesis cell division in diatoms and, 69 plastids and, 38,4344 Ganglia, nerve growth factor and biochemistry, 124 neurotrophic theory, 112 receptor expression, 126-130 synthesis, 114, 116, 118, 120-121 Gastrin, calcium regulating hormones and, 172-174,200 Gastrointestinal system, calcium regulating hormones and, 171-172, 186 Gel electrophoresis, ocean processes and, 269 Gellkidium amansii, plastids and, 6 Gene dosage, plastids and, 28 Gene expression, plastids and, 30 Gene probes, ocean processes and, 265, 272,276 Genes nerve growth factor and, 109, 124 ocean processes and, 265,278 molecular approaches, 265,267,273 organism function, 288,292, 295 plastids and, 1,32,37-38,43,52 Genetic variability, ocean processes and, 275 Genetics, ocean processes and, 262,264, 285,2% application of molecular techniques, 276, 280,282-283 Genomes nerve growth factor and, 124 ocean processes and application of molecular techniques, 279-281,283

312

INDEX

molecular approaches, 268,271-272 organism function, 292,295-2% plastids and, 1-3,7, 55-56 differentiation, 26 division, 13 Glucagon, calcium regulating hormones and, 173 Glucose, calcium regulating hormones and,

Histochemistry. regenerated splenic transplants and, 225 Histology, regenerated splenic transplants and, 236,238,252 Homeostasis, calcium regulating hormones and, 140, 178 calcitonin, 169, 171, 185, 199-200 parathyroid hormone, 151, 153-154, 194-1%

144, 172

Glycoprotein, nerve growth factor and, 122, 125

Glycosylation, nerve growth factor and, 113, 123-124

Golgi bodies, cell division in diatoms and, 64 Golgi region, calcium regulating hormones and, 142-143, 156, 166, 170 Gonyaulax, ocean processes and, 293 Growth hormone calcium regulating hormones and, 15 1, 195-197

ocean processes and, 2% Guanine nucleotides, calcium regulating hormones and, 153, 157. 194

prolactin, 180, 182, 197-198 stanniocalcin, 201-202 Homology calcium regulating hormones and, 141 aquatic vertebrates, 180, 183, 189 terrestrial vertebrates, 142, 150, 166, 171

nerve growth factor and, 112 ocean processes and, 271,277-278.295 Hormones, calcium regulating, see Calcium regulating hormones Hyacinthiodes nonscripta, plastids and, 20 Hybridization, see also in situ hybridization nerve growth factor and, 113, 125 ocean processes and, 271-272,276, 279-280.282

H Hantzschiu, cell division in diatoms and, 91 central spindle, 7 1-72 collar, 83.88 kinetochores, 76,80 microtubule center, 66,68 Hematopoiesis, regenerated splenic transplants and, 217, 225, 231, 236-237 Heterocysts, ocean processes and, 288 Heterotopic transplants, spleen and, 237, 241 Heterotrophs, ocean processes and, 263, 267,273-274 High endothelial venules. regenerated lymph node transplants and, 247. 250, 253 High-pressure liquid chromatography, plastids and, 38 Hippocampus. nerve growth factor and, 115, 119, 121-122

plastids and, 55 5-Hydroxytryptamine, calcium regulating hormones and, 166, 187 Hyperactivity, calcium regulating hormones and, 185, 193 Hypercalcemia. 139-141, 178 calcitonin and aquatic vertebrates, 183-185 prolactin and, 179-181, 197-198 stanniocalcin and, 187, 190, 192-193, 201-202

terrestrial vertebrates, 169-171, 174, 176, 198-200

parathyroid hormone and, 142, 146, 148, 151, 153, 163-164

stanniocalcin and, 187. 190, 192-193, 201-202

Hypercalcin, calcium regulating hormones and, 181 Hypermagnesemia, calcium regulating hormones and, 150 Hyperparathyroidism, calcium regulating hormones and, 146. 151, 156, 159

INDEX H yperphosphatemia, calcium regulating hormones and, 183, 199 Hyperplasia, calcium regulating hormones and, 171, 184 Hyperpolarization, calcium regulating hormones and, 146-147 Hypertrophy, calcium regulating hormones and, 170-171, 184, 192 Hypocalcemia, 139-14 1 calcitonin and, 168-173, 176, 183-185, 198-199 parathyroid hormone and, 148, 153-154, 156, 163-164, 1%-197 prolactin and, 180 stanniocalcin and, 189, 192,201 Hypocalcin, calcium regulating hormones and, 189 Hypoglycemia, calcium regulating hormones and, 153 Hypomagnesemia, calcium regulating hormones and, 150 H ypophosphatemia, calcium regulating hormones and, 183 Hypothalamus, calcium regulating hormones and, 182, 197

I Immune response, regenerated splenic transplants and, 233,244,246 Immunization, regenerated splenic transplants and, 245-246 Immunoassays, Ocean processes and, 269, 274 Immunofluorescence ocean processes and, 274,287 plastids and, 40 Immunoglobulins, regenerated splenic transplants and, 222,243,245-246 Immunohistochemistry, regenerated splenic transplants and, 224 Immunohistology, regenerated splenic transplants and, 227,229-230, 252 Immunoprecipitation, nerve growth factor and, 123, 128 in situ hybridization

313

nerve growth factor and, 113, 115, 120, 125, 127, 130 ocean processes and, 280 Infection, regenerated splenic transplants and, 241,252 Inheritance, cytoplasmic, plastids and, 36-38,57 algae, 38-48 higher plants, 48-55 Inhibitors calcium regulating hormones and calcitonin, 166, 169, 173-174, 176 parathyroid hormone, 149-150, 152-162,194-1% prolactin, 182 stanniocalcin, 189-190, 201 cell division in diatoms and, 86-87, 93-95,101 ocean processes and, 265,274 plastids and, 16,26-27,38,40-42 Initiation of a synchronous culture, plastids and, 13, 16 Innervation, nerve growth factor and density, 118 neurotrophic theory, 110-1 11 receptor expression, 126-127, 131 synthesis, 118, 120-122 Inositol phosphates, calcium regulating hormones and, 161-162 Insulin, calcium regulating hormones and, 153 Integrins, nerve growth factor and, 124 Interdigitating dendritic cells, regenerated splenic transplants and, 222, 225 Interference contrast microscopy, regenerated splenic transplants and, 24 1 Interleukin-1, nerve growth factor and, 116, 122 Interleukin-2, nerve growth factor and, 123 Interleukin-6, nerve growth factor and, 125 Interphase, cell division in diatoms and, 64,66-69,71 Iodination, nerve growth factor and, 125, 127-130 Ionomycin, calcium regulating hormones and, 161, 177 Iris, nerve growth factor and, 115, 121 Iron, Ocean processes and, 295-2%

314

INDEX

Irradiance levels, ocean processes and. 289-292 Isogamy, plastids and, 37-43, 57 lsoproterenol, calcium regulating hormones and, 153, 173

K Kidney regenerated lymph node transplants and, 247 regenerated splenic transplants and, 226. 232 Kinetochores, cell division in diatoms and. 76-83,97-99, 101-103 collar, 83,86,88-89

L Large single-copy region, plastids and, 14 Leaf senescence, plastids and, 36,57 Leucoplasts, plastids and, 2 . 7 , 13 Ligands, nerve growth factor and, 11 1, 124- 125 Light microscopy calcium regulating hormones and, 171, I89 cell division in diatoms and, 63,70 ocean processes and, 274,280 plastids and, 48 regenerated splenic transplants and, 218, 224 Lillium longifloem, plastids and, 49 Lirhodesmium, cell division in diatoms and, 69.7 1 Liver, calcium regulating hormones and, 143-144, 165, 193 Luciferin-binding protein, ocean processes and, 293 Lycopersicon. plastids and, 48 Lycoris radiara. plastids and, 49 Lymph node regenerated, transplants and, 215, 246-253 regenerated splenic transplants and, 222. 243.246

Lymphocytes regenerated lymph node transplants and, 247,252 regenerated splenic transplants and, 2 19, 222 function of regenerated tissues, 242-244,246 influences on regeneration, 236 phases, 226-227,230 Lysosomes calcium regulating hormones and, 144, 165 plastids and, 42,45,52

M Macrophages nerve growth factor and, 116, 122 regenerated splenic transplants and, 218-220 function of regenerated tissues, 244-246 influences on regeneration, 236, 238, 240 phases, 224-225, 227, 230 Magnesium, calcium regulating hormones and, 149-150, 157-159, 182 Malaria, regenerated splenic transplants and, 241.252 Maps ocean processes and, 268,282 plastids and, 1, 28 Marcanria polymorpha, plastids and, 1 Medicago saliva, plastids and, 54 Medullary thyroid carcinoma cells (MTC), calcium regulating hormones and, 176-177 Meiosis cell division in diatoms and, 64,69 plastids and, 48 Melosira, cell division in diatoms and, 71,73 Memory cells, regenerated splenic transplants and, 222,244 Merkel cells, nerve growth factor and, I14 Mesenchyme, nerve growth factor and, 114 Mesenteric nodes, regenerated lymph node transplants and, 250 Mesophyll cells, plastids and, 25

315

INDEX Messenger RNA calcium regulating hormones and calcitonin, 167, 171, 176 parathyroid hormone, 143, 147-148, 155-156 nerve growth factor and, 113, 132 receptor expression, 125-129, 131 synthesis, 113-1 15, 117-122 ocean processes and, 288,292-295 plastids and, 42-43,57 Metaphase, cell division in diatoms and, 91,94,98-99, 101, 103 central spindle, 70-72,76 collar, 87-89 kinetochores, 76,83 Metazoans, ocean processes and, 268,270 Methionine ocean processes and, 270 plastids and, 42 Methylation, plastids and, 30, 38-39 Microenvironment, regenerated splenic transplants and, 217,220,231 Microfilaments, regenerated splenic transplants and, 225-226 Microtubule center, cell division in diatoms and, 64-69,71,82, 100 Microtubule disassembly, cell division in diatoms and, 89,97, 101-102 Microtubule organizing center, cell division in diatoms and, 66,68 Microtubules cell division in diatoms and, 98-103 central spindle, 70-76 collar, 83.86, 88-89 kinetochores, 76, 80-83 microtubule stability, 94-97 spindle elongation, 91,93-94 plastids and, 16, 20 Microzooplankton, ocean processes and, 264,284,294 Mineralization, calcium regulating hormones and, 1% Mirabilis jalapa, plastids and, 36 Mitochondria, plastids and cytoplasmic inheritance, 43-45,49, 55 differentiation, 30-32 division, 11, 13-14, 16 Mitochondrial DNA, ocean processes and, 268,276,280-283 Mitochondriokinesis, plastids and, 16

Mitosis calcium regulating hormones and, 171 diatoms and, see Diatoms, cell division in plastids and, 48-49, 52 regenerated splenic transplants and, 225, 23 1 Molecular biology in ocean processes, see Ocean processes, molecular biology in Monoclonal antibodies calcium regulating hormones and, 146 ocean processes and, 270 regenerated splenic transplants and, 222, 225,227,236,244 Monostroma, plastids and, 42 Morphogenesis, cell division in diatoms and, 63,68-69,94,99-100 Morphology cell division in diatoms and, 100, 102 central spindle, 71 collar, 83-86 kinetochores, 76 microtubule center, 66.68 nerve growth factor and, 109-1 10, 116 ocean processes and, 265,273, 276, 285-286 plastids and, 7,25-26, 39 regenerated splenic transplants and, 224 Morphometry, regenerated splenic transplants and, 222,230 Moss, plastids and, 22-23,37,48 Motoneurons, nerve growth factor and, 117, 130-131 Mutation ocean processes and, 272,282 plastids and, 27,30,38,43

N Narcissus pseudonarcissus, plastids and, 31 Nauicula, cell division in diatoms and, 66 Necrosis regenerated lymph node transplants and, 247,253 regenerated splenic transplants and, 224-226,234-235,252 Nerve growth factor, 109, 132

316

INDEX

biochemistry, 123- 125 molecular biology, 1 12- 1 13 neurotrophic theory, 109-1 12 synthesis, I13 site, 114-1 18 time course, 118-122 Nerve growth factor receptor, 109, 119-120 expression, 125-126 NGF-dependent neurons, 126-129 NGF-independent neurons, 130- I3 1 Neuronal target cells, nerve growth factor and, 114-1 15 Neurons, see Nerve growth factor Neurotransmitters calcium regulating hormones and calcitonin, 166 parathyroid hormone, 150-156, 195 prolactin, 182 stanniocalcin, 192- 193 nerve growth factor and, 110 Neurotrophic factor, nerve growth factor and, 109, 117 biochemistry, 124-125 receptor expression, 125, 127-128, 132 Neurotrophic theory, nerve growth factor and, 109-1 12 Neurotrophin-3, nerve growth factor and, 112-113, 124, 132 Neurotubulin, cell division in diatoms and, 94 Nicotiana, plastids and, 54 Nicotiana tabacum, plastids and, 1,4, 7 cytoplasmic inheritance, 41,49 differentiation, 32 division, 14, 22 Nitella, plastids and, 31 Nitrate, ocean processes and, 286-287. 294 Nitrate reductase, ocean processes and, 294-295 Nitrification, Ocean processes and, 286-289,294,2% Nitrite, ocean processes and, 286-287 Nitrite reductase, Ocean processes and, 294-295 Nitrogen, ocean processes and, 263,266 limitation in phytoplankton, 293-295 organism function, 286-289 Nitrogen fixation, ocean processes and, 288-289

Nitrogenase, ocean processes and, 269-270,288 Nocodazole, cell division in diatoms and, 89 Noradrenaline. nerve growth factor and, 129 Norepinephrine, calcium regulating hormones and, 152 Northern blots, Ocean processes and, 272 Nuclease C, plastids and, 41-43,57 Nucleases, plastids and, 36,39,41,52,57 Nucleation, cell division in diatoms and, 82 Nuclei, organelle, plastids and, see Plastids Nucleic acids, Ocean processes and, 261-262,264-265 application of molecular techniques, 276-283 isolation, 265-268 sequencing, 273 Nucleoids. plastids and, 3 Nucleosomes, plastids and, 32 Nucleotides nerve growth factor and, 123 ocean processes and, 271

0 Ocean processes, molecular biology in, 261-262,2%-297 application of techniques, 273-214, 284 cellular level, 274-275 nucleic acid level, 276-283 protein level, 275-276 biotechnological exploitation, potential for, 2% marine organisms, 262-265 molecular approaches, 265,272-273 nucleic acid isolation, 265-268 organelle DNA,268-269 polymerase chain reaction, 271-272 protein analysis, 269-270 protein synthesis, 270-271 organism function, 284-286 microbes in nitrogen cycle, 286-289 photosynthesis, 289-293 phytoplankton, 293-295 zooplankton, 295-2% Ochromonas, cell division in diatoms and, 76,93,95

INDEX Oedogonium, cell division in diatoms and, 86, 102-103 Oenotheru, plastids and, 54 Olfactory bulb, nerve growth factor and, 119 Oligonucleotides, Ocean processes and, 271,276,278-280 Oogamy, plastids and, 37,45 Open circulation, regenerated splenic transplants and, 218-219 Oreochromis mossambicus, calcium regulating hormones and, 181, 193 Organelle DNA, ocean processes and, 268, 27 1 Organelle nuclei, plastids and, see Plastids Organellekinesis, plastids and, 9, 13 Organelles, cell division in diatoms and, 63,99 Orthotopic splenic tissue, regenerated splenic transplants and, 237,241 Oryza satiua, plastids and, 6,36,57 Osmolarity,calcium regulating hormones and, 182, 197 Osteoblasts, calcium regulating hormones and, 184,199 Osteoclasts, calcium regulating hormones and, 166, 169,202 Osteoporosis, calcium regulating hormones and, 156, 174 Overwhelming postsplenectomy infections (OPSI). regenerated splenic transplants and, 216,235,244-245 Ovulation, calcium regulating hormones and, 186 Oxygen ocean processes and, 286-288,292 plastids and, 7 regenerated splenic transplants and, 224, 242

P Pancreas, calcium regulation and, 173 Pancreatic beta cells calcium regulating hormones and, 145 nerve growth factor and, 111

317

Paracentrosomes, cell division in diatoms and, 69 Parasites, regenerated splenic transplants and, 242 Parathyroid gland, calcium regulating hormones and, 178-181 Parathyroid hormone, 140-142, 194-198 calcitonin and, 171, 174, 176, 199 prolactin and, 179 stanniocalcin and, 189 in terrestrial vertebrates, 142-144 degradation, 163-166 second messengers, 157-163 secretion, 144-156 Pea, plastids and, 35-36 Pelargonium, plastids and, 49,54-55 Petargonium zonale, plastids and, 36, 52,54 Pennates, cell division in diatoms and, 64, 66-70,72,98-100 Pentagastrin, calcium regulation and, 172, 174 Peptides calcium regulating hormones and, 142, 150- 152, 172- 173 nerve growth factor and, 113, 123 Periarteriolar lymphatic sheath (PAL$), regenerated splenic transplants and, 219,222,225,227,230,252 Peripheral nervous system, nerve growth factorand, 114, 116, 119, 121, I32 Pertussis toxin, calcium regulating hormones and, 153, 160, 162 Petunia, plastids and, 54 Phagocytosis regenerated lymph node transplants and, 247 regenerated splenic transplants and, 219-220,240 Phaseolus vulgaris, plastids and, 27 Phenotype, ocean processes and, 275 Phentolamine, calcium regulation and, 153, 173 Phorbol esters, calcium regulating hormones and, 162-163, 177 Phosphates calcium regulating hormones and, 169, 1%-197, 199,201-202 ocean processes and, 287

318

INDEX

Photoautotrophs, Ocean processes and, 263,270,294 Photogenes, plastids and, 28, 34 Photoinhibition, Ocean processes and, 280, 282 Photosynthesis ocean processes and, 289-293 plastids and. 7, 27-28. 32 Photosystems ocean processes and, 288,292,295 plastids and, 1-2 Phycobiliproteins, ocean processes and, 292 Phycocyanin, ocean processes and, 292 Phycoerythrin, ocean processes and, 292 Phycourobilin, ocean processes and, 292 Phylogeny, Ocean processes and, 276-278, 280.282,2% Physarum, plastids and, 14, 16.55 Phytoplankton, Ocean processes and, 26 I -263 application of molecular techniques, 274-275,277 molecular approaches, 268,270 nutrient limitation, 293-295 organism function. 284, 286,289 Pinnularia, cell division in diatoms and, 68,70-72,76,80 Pinus. plastids and, 55 Pisum sativum. plastids and, 20, 30 Pituitary gland, calcium regulating hormones and, 178 calcitonin. 185 parathyroid hormone, 150-151, 156 prolactin. 179-1 82 Piacopecien magellunicus, Ocean processes and, 283 Plankton, Ocean processes and, 262, 274-276 Plaque-forming cells, regenerated splenic transplants and, 233. 243 Plasma calcium regulating hormones and, 140-141, 178 calcitonin, 168-175, 183-186, 198-200 parathyroid hormone, 149-154, 156 prolactin, 180-182, 197 stanniocalcin, 189-193,202 regenerated splenic transplants and, 220, 242

Plasma membrane, calcium regulating hormones and, 146, 162 Plasmodium berghei, regenerated splenic transplants and, 238,241 Plastid-dividing ring, plastids and, 16-25.56 Plastid DNA, organelle nuclei and, 2-3,7, 56-57 cytoplasmic inheritance, 37,40-41, 44-48 differentiation, 26-28,30-36 higher plants, 48-49, 52.54-55 Plastidkinesis, 2, 16-25, 56 Plastids, 1-3, 55-57 cytoplasmic inheritance, 36-38 algae, 43-48 higher plants, 48-55 isogamous algae, 38-43 differentiation, 25-30 digestion of nuclei, 35-36 DNA-binding proteins, 30-35 DNA methylation, 30 division, 9 nucleus, 9-15 plastidkinesis, 16-25 location of nuclei, 3-6 organization of nuclei, 6-9 Plate counts, ocean processes and, 274 Platelets, regenerated splenic transplants and, 217,242 Pneumococci, regenerated splenic transplants and, 238,240,244-246, 252 Polar plates, cell division in diatoms and, 71-73 Polar vacuoles, cell division in diatoms and, 71 Polarity, cell division in diatoms and, 93, 95,97, 103 central spindle, 73, 76 collar, 83, 86,89 kinetochores, 80-83 Pollen, plastids and, 24, 48-49, 52,54 Polymerase chain reaction, ocean processes and, 265.267, 271-272,278 Polymerization, plastids and, 16, 18 Polymorphism, ocean processes and, 272, 282 Polypeptides, plastids and, 7, 31.42-43 Polyvinylpyrrolidone, ocean processes and, 276

3 19

INDEX Population dynamics, ocean processes and, 261,274-275,278-279,282-284 Population genetics, ocean processes and, 276,280,2% Porphyrin, ocean processes and, 292 Potassium, calcium regulating hormones and, 146-147, 160, 171 Primordial spindle, cell division in diatoms and, 71-72 Progesterone, calcium regulating hormones and, 156 Prokaryotes, ocean processes and, 262, 273,277,286,288 Prolactin, 141-142, 1%-197,200 in aquatic vertebrates, 178-182 calcitonin and, 183 parathyroid hormone and, 151, 166, 194-199 stanniocalcin and, 192 Prometaphase, cell division in diatoms and, 98-102 central spindle, 70,72-74 collar, 83,87-88 kinetochores, 80-83 Prophase, cell division in diatoms and, 66, 68,71-74 Proplastids, organelle nuclei and, 56 differentiation, 25-28,32, 34 division, 11, 14,20,22-24 location, 4 , 6 organization, 6-9 Propranolol, calcium regulating hormones and, 153 proPTH, calcium regulating hormones and, 143-144, 148, 155 Prostaglandins, calcium regulating hormones and, 157 Protein calcium regulating hormones and calcitonin, 166, 186, 200 parathyroid hormone, 142-143, 150, 153, 194 stanniocalcin, 193 nerve growth factor and, 112-1 13 biochemistry, 123-124 receptor expression, 128-129 synthesis, 113, 117, 121 ocean processes and, 261-262 application of molecular techniques, 275-276

isolation, 269-270 molecular approaches, 265,267, 271-273 organism function, 285-286, 288-290, 292-295 synthesis, 270-271 plastids and, 1,7,9,56-57 cytoplasmic inheritance, 42-43 differentiation, 27-28,30-36 division, 14,22 regenerated splenic transplants and, 233, 243 Protein kinase, calcium regulating hormones and, 157-158 Protein kinase C, calcium regulating hormones and, 162-163, 177 Proteolysis, nerve growth factor and, 113, 121 Protoplasts, plastids and, 7,25 Protozoa, ocean processes and, 263 Prunus persica, plastids and, 36 Pseudocalanus, ocean processes and, 276, 2% Pteris vittata, plastids and, 13,45 F'tK cells, cell division in diatoms and, 83, 88-89,96,99, 102 Puccinia, cell division in diatoms and, 76,93 Pyraminomonas virginica, plastids and, 20

R Radioactive labels, regenerated splenic transplants and, 240-241 Radioactive probes, ocean processes and, 267,279,285-286 Radioimmunoassay,calcium regulating hormones and, 150, 171 Rana pipiens, calcium regulating hormones and, 183 Red pulp, regenerated splenic transplants and, 252 function of regenerated tissues, 241-242 functional anatomy of spleen, 218-219, 223 influences on regeneration, 238-240 phases, 227,230 Regenerated lymph node transplants, 215, 246-253

320

INDEX

Regenerated splenic transplants, 250, 252-253 function of regenerated tissues, 239-240 blood OW, 240 clearance, 240-241 effects on lymphoid organs, 242-243 filtering by red pulp, 241-242 immunization against pneumococci, 245-246 normal tissues, 243 protective effect, 244-245 function of spleen, 215-217 functional anatomy of spleen compartments, 219-223 structure, 217-219 influences on regeneration, 236-237 age, 235-236 different species, 239 enhancement, 237-238 mass, 234-235 site, 232-234 phases of regeneration, 223-224 development of compartments, 225 early events, 224-225 immunohistology of tissue, 227, 229-230 length of process, 231-232 morphometry of compartments, 230 revascularization, 226-228 surviving cells, 231 Replication cell division in diatoms and, 69 plastids and, 9, 13, 32.55 Reproduction calcium regulating hormones and, 140, 186,200 Ocean processes and, 274-275 Reptiles, calcium regulating hormones and, 167-168, I%, 199 Restriction endonuclease ocean processes and, 272 plastids and, 1, 14.38.41 Restriction fragment analysis, ocean processes and, 265,280 Restriction fragment length polymorphism ocean processes and, 272 plastids and, 55 Restriction maps, Ocean processes and, 282

Reticular cells regenerated lymph node transplants and, 246-247 regenerated splenic transplants and, 218, 225-226,252 Reticulin fibers, regenerated splenic transplants and, 225 Revascularization, regenerated splenic transplants and, 226-227 Rhododendron, plastids and, 52,54 Ribosomal RNA ocean processes and, 273,276-280,282, 285 plastids and, 7,28,32 Ribosomes, plastids and, 1.22, 27-28 Ribulose 1,5-bisphosphate carboxylase, ocean processes and, 269-270 Rifiia pachyptila, ocean processes and, 278 RNA calcium regulating hormones and, 155 nerve growth factor and, 113, 125 ocean processes and, 265-266,278-279, 285 plastids and, 27, 36,42 RNA polymerase, plastids and, 30 RNase, ocean processes and, 265 Root tip, plastids and, 24

S Salivary gland, nerve growth factor and, 112-113 SARII, ocean processes and, 279 Scanning electron microscopy, regenerated splenic transplants and, 219,224,226, 232 Scattered plastid nuclei, 5-6, 11, 14,55 Schwann cells, nerve growth factor and, 114-116, 122, 128, 132 Sciatic nerve, nerve growth factor and, 115-1 16 Scintigraphy, regenerated splenic transplants and, 232,235 SDS-PAGE, plastids and, 32 Second messengers, calcium regulating hormones and calcitonin, 176-177, 186

32 1

INDEX parathyroid hormone, 157-163 stanniocalcin, 193 Secretin, calcium regulating hormones and, 151, 157, 172

Secretion, calcium regulating hormones and calcitonin, 169-176, 184-186, 198, 200

parathyroid hormone, 144-146, 161, 163, 194-195

prolactin, 181-182 stanniocalcin, 187, 189-193,202 Secretory granules, calcium regulating hormones and calcitonin, 167 parathyroid hormone, 142-144, 147, 195 stanniocalcin, 187, 189 Segregation, plastids and, 13-14 Sensory neurons, nerve growth factor and biochemistry, 124 neurotrophic theory, 110-1 12 receptor expression, 126, 128, 131 synthesis, 118 Septum, plastids and, 24 Sequences calcium regulating hormones and, 142, 189

nerve growth factor and, 112, 117, 124 plastids and, 1, 14 Setpoint, calcium regulating hormones and, 140, 145, 195-1%

Sheep red blood cells, regenerated splenic transplants and, 233, 243 Silica, ocean processes and, 295-2% Sinus regenerated lymph node transplants and, 247, 250

regenerated splenic transplants and, 219-220,225-226 Skeletonema costatum, ocean processes and, 212,275 Sliding, cell division in diatoms and, 80-82,93-94 Solanum, plastids and, 49 Somatostatin, calcium regulating hormones and, 151, 167, 173, 182 Somewhat electron-dense granules (SEG), plastids and, 18,20 Southern blots, ocean processes and, 267, 282

Sperm cell division in diatoms and, 64,69 plastids and, 57 cytoplasmic inheritance, 45,47-49, 52,54

division, 13,23 Spinach, plastids and, 31, 35 Spindles, cell division in diatoms and, 63-64,98-103

central spindle, 70-76 collar, 83-89 elongation, 89-94 kinetochores, 76,SO-81 microtubule center, 66-69 microtubule stability, 94-97 Spleen, see Regenerated splenic transplants Splenectomy, 216217,222,232 function of regenerated tissues, 240246

influences on splenic regeneration, 235-237

Splenic transplants, regenerated, see Regenerated splenic transplants Splenosis, regenerated splenic transplants and, 232,236,243,245 Stanniocalcin, 141-142, 197- 198,201 202

in aquatic vertebrates, 178, 184 second messengers, 193 secretion, 189-193 Stephanophyxis, cell division in diatoms and, 70-7 1,93 Steroids, calcium regulating hormones and, 153-156, 174-176, 1% Streptococcus pneumoniae, regenerated splenic transplants and, 244 Streptomycin, plastids and, 38 Stroma regenerated lymph node transplants and, 247 regenerated splenic transplants and, 217, 231,237, 252 Substance P, calcium regulation and, 173 Surirella, cell division in diatoms and, 64, 80,88,98, 100 central spindle, 70-72 microtubule center, 66,68 Symbiosis, ocean processes and, 278

322

INDEX

Sympathetic neurons, nerve growth factor and neurotrophic theory 110-1 11 receptor expression, 127-130 synthesis, 117, 119-120

T T cells, regenerated splenic transplants and, 242-244 Taxonomy, ocean processes and, 267, 273-274,278 Teleosts, calcium regulating hormones and, I79 calcitonin, 184, 198 prolactin, 181 - 182 stanniocalcin, 187,201 Telophase, cell division in diatoms and, 68, 71-72,86, % Thalassiosira pseudonana, ocean processes and, 275 Theophylline, calcium regulating hormones and, 157 Thermus aquaticus, ocean processes and, 27 1 Thylakoid membranes, plastids and, 1, 13-14,25. 32.39 Thymidine, ocean processes and, 285 Thymus regenerated lymph node transplants and, 247 regenerated splenic transplants and, 225, 234,244-245 Thyroid, calcium regulating hormones and calcitonin, 166-167, 169-174, 376177 parathyroid hormone, 145, 154 stanniocalcin, 192 Tobacco, plastids and, 13-14 Total community DNA hybridization, Ocean processes and, 280 Transcription calcium regulating hormones and calcitonin, 167, 175-177 parathyroid hormone, 143, 148, 155-156, 195-1% nerve growth factor and, 113, 124 ocean processes and, 284,288,290,292, 294 plastids and, 28,30-32,35 Transfer RNA

Ocean processes and, 282 plastids and, 1 , 7 Translation calcium regulating hormones and, 143, 148, 194 ocean processes and, 284,290, 293 plastids and, 28, 41 Translocation cell division in diatoms and, 91, 101 nerve growth factor and, 113 Transmission electron microscopy, regenerated splenic transplants and, 224,226 Transplant lymph node, see Regenerated lymph node transplants regenerated splenic, see Regenerated splenic transplants Trebouxia. plastids and, 20 Trichodesmium, ocean processes and, 270, 272,288-289 Trichomonas, cell division in diatoms and, 70 Trichonympha, cell division in diatoms and, 71 Triform repens, plastids and, 54 Triticum, plastids and, 49 Triticum aestiuum, plastids and, 21, 49 tRNA, see Transfer RNA Tubulin cell division in diatoms and, 72, 83.89, 93-94 plastids and, 16

U Ultimobranchid cells, calcium regulating hormones and, 166, 168, 183-186, 191 Ultrastructure, cell division in diatoms and, 64,69,81-82,89-93 Ultraviolet-B radiation, ocean processes and, 292-293 Ultraviolet light cell division in diatoms and, 91, 95-97 plastids and, 41-43 V Vaccination, regenerated splenic transplants and, 245

323

INDEX Valve morphogenesis, cell division in diatoms and, 69 Vasoactive intestinal peptide (VIP), calcium regulation and, 173, 182 Vertebrates, calcium regulating hormones in, see Calcium regulating hormones Vesicles, plastids and, 14 VIMPICS, plastids and, 4,27,40,54 Vitamin D, calcium regulating hormones and, 140 calcitonin, 170, 174-176 parathyroid hormone, 142, 144, 154-156, 161, 195-197 prolactin, 179 stanniocalcin. 193

W

influences on regeneration, 233,238 phases, 225,227, 230

X Xenopus laevis, calcium regulating hormones and. 180

Y Yeast ocean processes and, 266 plastids and, 4 Z Zea mays, plastids and, 1

Western blot, ocean processes and, 269, 294 White pulp, regenerated splenic transplants and. 219.222-223.252

Zinc, plastids and, 36,57 Zooplankton, ocean processes and, 261-263,267,276 organism function, 285,293,295-2%

This Page Intentionally Left Blank

This Page Intentionally Left Blank