A SEARCH FOR SYNAPTONEMAL COMPLEXES IN USTILAGO MAYDIS

Transcription

1 J. Cell Sri. 50, (1981) 171 Printed in Great Britain Company of Biologists Limited xg8i A SEARCH FOR SYNAPTONEMAL COMPLEXES IN USTILAGO MAYDIS H. L. FLETCHER* Genetics Department, National Institute for Medical Research, The Ridgeway, Mill Hill, London NW7 1AA, England SUMMARY Germinating spores, germ-tubes (promycelia), gall tissue and spores developing in the gall were examined. No synaptonemal complexes (SCs) were found in any of these cell types. There are 3 possible explanations for this: (1) Ustilago maydis does not have SCs. This is the case in Schizosaccharomyces pombe (Olson, Eden, Egel-Mitani & Egel, 1978) and in many strains of Saccharomyces cerevisiae (e.g. see Byers & Goetch, 1975). (2) The SC might occur in spores either in the gall or before germination when the spore wall was too solid to allow examination of the contents by the methods used. (3) The SC was present, but for a very short time, and most of the cells examined were in any case not undergoing meiosis, so the SCs were not seen. (Large numbers of spores were examined, but few spores germinate and undergo normal meiotic division; also a minority of the cells in the gall are forming spores at any one time.) The gall was found to contain uninucleate single cells (i.e. it is a yeast, as it is in artificial culture medium) and virtually all these cells were haploid, 1 in 3000 recovered cells was diploid. It appears that the haploids fuse to form a heterokaryotic or diploid cell immediately before spore formation. A heterokaryotic phase is presumed to exist to establish and maintain the infection. INTRODUCTION Ustilago maydis is a heterobasidiomycete that causes covered smut in maize. Numerous mutants are known, and U. maydis is amenable to genetic and biochemical analysis. The mechanisms that manipulate DNA during meiosis, especially those leading to hybrid DNA formation and genetic conversion or recombination, are of particular interest (Holliday, 1977). In order to facilitate the biochemical investigation of these processes it was desirable to examine microscopically the stages of meiosis to correlate the physical and biochemical events of meiosis. Ultrastructural investigation of the synaptonemal complex (SC) should indicate when the homologous bivalents were synapsed. Reconstruction of the SCs from serial sections should also reveal the number of bivalents for comparison with genetic linkage maps. The vegetative phase of U. maydis is a budding yeast. Both haploid and diploid cells are prototrophic and can be cultured in artificial medium. Haploids of opposite mating type will fuse on special medium or in the host tissue to form a heterokaryon. Diploids may be selectively recovered from gall tissue. The meiotic phase is only initiated within the host. Meristematic host tissue is induced to proliferate and form a gall, which is eventually digested by the U. maydis and becomes filled with teliospores. Present address: Genetics Subdepartment, David Keir Building, The Queen's University of Belfast, Stranmillis Road, Belfast BT7 inn, Northern Ireland.

2 172 H. L. Fletcher These can be germinated on complete medium. The spore grows a promycelium or germ-tube. The nucleus moves into this and then undergoes both meiotic divisions. Cell walls partition the germ-tube between the nuclei, and then the 4 cells bud off haploid, yeast-like basidiospores that reproduce by vegetative budding (Fischer & Holton, 1957). There is evidence that the temperature, ph and nutritional and ionic composition of the medium on which the teliospores germinate all affect the frequency of recombination between the mating type locus and the centromere (Hiittig, 1931, 1933). Recombination at other loci changes non-linearly with temperature (Holliday, personal communication). Thus it seemed that meiotic recombination occurred during germination and this was the first stage to examine for SCs. In many fungi (e.g. Neottiella and Neurospora) meiosis occurs immediately after fusion of 2 haploid nuclei to form a diploid. Developing teliospores of the Cedar Apple Rust fungus Gymnosporangiumjumperi-virgimanae were found to contain SCs (Mims, 1977). This species is also a heterobasidiomycete and the most closely related species to U. maydis in which SCs have been found. In Blastocladiella emersonii (Phycomycete) SCs are present in resting sporangia 28 to 52 h old (Olson & Reichle, 1978). It could be advantageous to the fungus to start meiosis while in a secure position and so reduce the time required for germination of the spore in a potentially hostile environment. The developing spores in the gall are thus the second place in which to look for SCs. Irradiation of diploid cells induces meiotic levels of recombination in surviving cells, presumably as a result of the repair mechanisms. This was also a reasonable situation in which to look for SCs. MATERIALS AND METHODS Spores The most important criterion was a high level of synchronous germination. Spores from cross Mi46 nari-12 niri-i inosi-3nici-2 + B2 b_2 nan-13 niri-i inosi-3 + tsdi-i ai bi were used initially, then spores from a wild-type cross. Spore suspension was dropped onto the surface of complete agar medium so that there was a spacing of around 30 /tm to 40 /im between spores. If the spores were too close together it was difficult to section them. The spores germinated in the plane of the surface, so sectioning in this plane consistently gave sections of both spores and complete germ-tubes with buds at appropriate stages. Germination was at temperatures between 25 and 35 C. Germinating spores were enrobed by pouring a thin layer of molten 2 % agar at 46 C over them. Enrobed cells continued to grow if left unfixed, so the few seconds at 46 C did not harm them. Fixation was in 4 % glutaraldehyde in o-i M-cacodylate or 0-05 M-phosphate buffer at ph 7 applied before, with, or after the agar. Few cells could be sectioned unless the cell walls were eroded enzymically, but after penetrating the cell wall the enzymes attacked the cytoplasm. The mercaptoethanol/55 % gluculase treatment of Zickler & Olson (1975) was used: o-i M-mercaptoethanol, 0-2 M-Tris, 0-02 M- EDTA at ph 9 for 10 min at 35 C; followed by o-i ml of o-i M-sodium citrate/phosphate buffer, ph 5-8,+ 0-3 ml 1 M-KC1 + O-OI ml 1 M-MgCl + O'S ml gluculase for 3 h at 35 C. Cellulase (Okazaki) was also at between 1 and 5 % concentrations in o-i M-citrate/phosphate at ph 5 (485 ml of 01 M-citric acid ml 01 M-NajPO*) either alone or with gluculase. Incubation times varied between 30 min and 4 h. Intracellular membranes did not stain

3 Synaptonemal complexes in Ustilago maydis 173 with osmium after gluculase treatment. Gluculase did not digest SCs in grasshopper testes. Spores were usually postfixed in 1 % osmium tetroxide in sodium barbitone buffer, ph 7-2 (379 g/1 sodium barbitone, 2-42 g/1 sodium acetate, M-HC1) f r 4 h or overnight. Dalton's (1955) chrome/osmium was also used overnight. Spores were thoroughly washed, dehydrated through an ethanol/water series or in acidified 2-2-dimethoxypropane for 5 min (Muller & Jacks, 1975). Some samples were stained in uranyl acetate in 75%, 90% or 98% ethanol for 1-2 h. Cells were infiltrated with Spurr' (1969) resin overnight and embedded. Sections were cut with a glass knife, and were poststained with uranyl acetate/lead citrate (Reynolds, 1963). Gall tissues Maize seedlings were inoculated with compatible wild-type strains of U. maydis. Galls formed within a week, and spores then started to accumulate. Segments of gall just beginning to produce spores were fixed in 4 % glutaraldehyde. The U. maydis cells were thin-walled but embedded in a gel that collapsed when dehydrated and was impossible to infiltrate with resin. The gel was removed by digestion with 2 % cellulase (Okazaki) for 10 min, a balance between insufficient removal of gel and damage to cells. The gall tissue was postfixed overnight in Dalton's osmium, washed, dehydrated in 2-2-dimethoxypropane for 5 min and embedded in Spurr's resin. Sections were poststained in uranyl acetate/lead citrate. Irradiated vegetative cells: diploid cells of strain d68 nari-12 nin-i inosi-3nici-2 + a2 b2 nari-13 niri-i inosi-3 + tsdi-i ai bi were irradiated with 150, 300 and 450 krad of y from a cobalt 60 source. These doses produced 8, 92 and 200 nar + recombinants per io 6 survivors, respectively. At i-h intervals from 1-4 h cells were fixed in 4 % glutaraldehyde for at least 24 h. A few drops of a paste of settled cells was put into a i-cm diameter round-bottomed test-tube and sufficient o-j-mm diameter glass beads were added to appear just above the surface. The whole was whirlimixed for up to 10 min, regularly monitoring the cell-wall removal with a phase-contrast microscope. Refractivity changed when the wall ruptured. This method avoided damage from enzymes. The cells were resuspended, decanted off the glass beads, stained with Dalton's (1955) chrome/ osmium, dehydrated for 3 min in 2-2-dimethoxypropane, infiltrated for 3-4 h with Spurr's (1969) resin and embedded. Light microscopy (1) Basic fuchsin: cells were fixed in Schaudin's solution (2 parts saturated mercuric chloride solution : 1 part absolute alcohol) for 1 h, rehydrated, washed, hydrolysed in 1 M-HC1 at 60 C for 10 min and stained in basic fuchsin. (2) Hoechst florochrome: cells were either fresh or fixed with 4% glutaraldehyde. Hoechst was used in aqueous solutions at concentrations between 1 /tg/ml and 50 fig/mx. Cells were examined in an ultraviolet fluorescence microscope. At low concentration only the nuclei fluoresced, and they were always blue. At progressively higher concentrations presumed cell-growth zones fluoresced yellow and fungal cell walls and cytoplasm fluoresced green. RESULTS Germinating spores No structures resembling SCs were seen at any stage in the germinating spores. The sectioned nuclei generally had a uniform granular appearance (Fig. 1). with the nucleolus showing up as a denser area. The germ-tube grows first, then the nucleus migrates into it from the spore (Fig. 2, 3). Microtubular spindles were frequently found with rather amorphous dense poles attached to the nuclear envelope (Fig. 4)

4 H. L. Fletcher

5 Synaptonemal complexes in Ustilago maydis 175 or to intranuclear intrusions of the nuclear membranes (Fig. 5). Spindles were found in nuclei while they were still in the spore. By analogy with all other known cases, the spindle forms after chromosome synapsis and formation of the SC. This suggests that only the final division stages of meiosis occur in the germinating spore and germtube. There is a resting phase between germ-tube growth and budding, during which meiosis was thought to occur. Examination of individual germinating spores revealed this to be a very variable stage, ranging from 15 min to more than 5 h, with a modal time of 75 min at 32 C and 100 min at 25 C for strain M146. The reason for this is not clear. The laboratory stocks are largely maintained by mitotic growth and may have lost some meiotic functions, causing failure of meiosis in some spores. In Saccharomyces cerevisiae, meiosis takes at least 4 h (7 h including induction), again suggesting that only the later stages are possible in germinating U. maydis spores. The major difficulty (apart from sectioning) was a low level of synchronous germination combined with the failure of a very large proportion of spores to complete meiosis and produce buds. Between 20 and 95 % of germinating spores were still plain germ-tubes when they were overgrown by colonies from the spores that germinated first. Calculation suggests that approximately 3 % of plated spores were in any particular i-h stage of meiosis at any time. Several thousand spores were examined and it is unlikely that normal SCs could have been overlooked if they had been there. In a test sample of yeast in which 3 % of cells were sporulating, these cells were easily found in every section (although SCs were not found). In a few germ-tubes arrested before nuclear division the chromosomes condensed and became visible by fluorescence microscopy after staining with Hoechst There were at least 6, and perhaps 10,fluorescent structures, although the smaller ones were nearly invisible. The larger ones were clearly double structures, and while they are presumed to have been bivalents they might have been chromosomes with visible chromatids. Similar bodies were found in electron microscope sections (Fig. 6) spread throughout a microtubular spindle. The sizes were from 0-25 /tm x 0-5 /fm to 0-5 /<m x 0-9 /im, and the smaller ones would have been unresolvable by transmitted light microscopy while inside the cell wall. They are visible as light sources when they fluoresce. Gall tissue Most histological studies of Ustilaginales were made early in this century when, in the absence of effective fungicides, they caused major crop diseases. The gall material Figs Spore germination. Voids and unstained membranes are present in most cells because of enzymic digestion. Fig. 1. Germinating spore showing a nucleus with a nucleolus (arrow). Figs. 2, 3. Spores with germ-tubes containing nuclei with a spindle (arrows), i.e. metaphase to telophase. Fig. 4. Spindle (M x ) showing dense amorphous bodies at the poles (arrows). Fig. 5. Spindle fibres apparently attached to intruded nuclear membrane (arrow). Fig. 6. Bodies presumed to be condensed chromosomes (arrows) in a nucleus arrested in M t from the germ-tube in Fig. 2.

6 H. L. Fletcher

7 Synaptonemal complexes in Ustilago maydis 177 is difficult to work with, and the observations of various workers on different species of Ustilago do not give a clear indication of the organization of a U. maydis gall (Ainsworth & Sampson, 1950; Fischer & Holton, 1957). No mycelium was found in any of the galls described here. Cells tended to form end-to-end chains and clusters of attached buds. Most cells were uninucleate (Figs. 7, 8); the binucleate cells could have been actively dividing cells or, in particular cases, haploids fusing prior to spore formation. Many cells looked similar in shape to those grown on artificial medium (Fig. 9), particularly during the early part of the gall's growth, and when proliferating around wounds in the gall. As the gall increased in size large nodules of U. maydis cells formed within it. The cells in the nodule were small and irregular. A gel was deposited between the cells, which apparently pushed them apart (Fig. 10), leaving separated cells with matching adjacent faces. The jelly collapsed and solidified when it was dehydrated, and it had to be digested away with cellulase before electron microscopy was possible. In more advanced nodules, spore walls started to form around some larger cells. Among these young spores were large, rounded, thinwalled cells, and a few of these contained 2 nuclei (Fig. 11). This suggests that fusion of haploids produces a diploid cell, which immediately forms a spore. This pattern of behaviour is common in fungi. All these observations were readily confirmed by incident light fluorescence with Hoechst staining, which displayed well-separated uninucleate cells in large undisturbed masses of gall tissue. Observations of such large pieces of gall was impossible using transmitted light because of refraction. Only the nuclei of the maize cells fluoresced (blue) but the U. maydis nuclei fluoresced blue while the cytoplasm and cell wall fluoresced green. The gelatinous coat around each cell looked like a continuation of the cytoplasm when viewed by visible light, and the interface of the gel with surrounding water looked like a cell wall, giving a misleading appearance. No definite synaptonemal complexes were seen. One large thin-walled cell (presumed to be a diploid spore-forming cell) had a short length of a structure like a very faint SC but this was not continued through adjacent sections. Nuclei could be seen in spores with substantially thickened walls, and the chromatin in these was partially condensed producing electron-dense regions similar to those that characterize condensed chromatin in higher plants and animals (Fig. 12). This was thought to be a preparation for the dormant, resistant phase of the teliospore. It is possible that SCs occurred in slightly older spores than those examined. The thickened spore wall prevented examination of the contents. Figs Gall cells. Figs. 7, 8. U. maydis cells within the gall are small and have only 1 nucleus (arrows). Fig. 9. U. maydis cells growing in part of the gall cut and damaged a day before. A mature spore (J) is visible. Fig. 10. Undisturbed gall showing cells widely spaced by a gelatinous matrix (#) and 2 partially formed spores (1). Fig. 11. Large binucleate cell (n,n) with a thin wall found in a spore-forming zone. This is thought to be a heterokaryon after fusion of haploids and before spore formation. Fig. 12. Incompletely formed spore (cf. Fig. 1) partially collapsed during preparation, with condensed chromatin in the nucleus (arrow).

8 178 H. L. Fletcher An attempt was made to monitor the course of cell fusion in the host as follows: two compatible haploids carrying complementary auxotrophic markers were inoculated into a maize seedling. One week later a small piece of gall was removed and incubated at 26 C in complete medium in a shaker to produce a suspension of single cells, which appeared after about 4 h and were then the typical shape of liquid-cultured cells. The U. maydis cells embedded in the gall are assumed to have grown there, they were not part of the original inoculation. No spores were visible; the first ones would have been expected a few days later. Cells from the suspension were plated on complete and minimal media to count all cells and diploids, respectively. The frequency of diploids was 288 in io 8 cells, less than 1 in Most gall cells obtained were haploid. While these could have budded off from heterokaryons breaking down in the medium, no such heterokaryons were seen and the simplest conclusion is that the uninucleate U. maydis cells in the gall were haploid. This method could be used to search for the occurrence of diploids and possible meiotic recombination if cells beginning meiosis could be reverted to mitosis as has been done in the yeast S. cerevisiae (Olson & Zimmermann, 1978 a). Irradiated cells Cells irradiated at 300 krad and 450 krad undergo recombination at a rate approaching that found in meiosis. This occurs within 4-5 h (Holliday, 1971) and appears to be directly related to repair of double-strand damage. Vegetative cells were easy to examine without treatment with enzyme but no ultrastructural differences were found between irradiated cells and unirradiated controls. It seems that the SC is not necessary for recombination repair of DNA broken by irradiation. DISCUSSION Dr P. B. Moens (personal communication) has previously examined meiosis in U. maydis. He did not find synaptonemal complexes either in germinating spores or in the gall, where hyphae, binucleate cells, nuclear fusions and spore maturation were traced from long series of sections. The results reported in this paper are in full accord with those of Dr Moens: no conclusive evidence of SCs was found at any stage of spore maturation or germination. The examination was difficult because of the nature of the organism, and the stage at which SCs would be expected is not clear. Classic SCs should have been found if they had been present in only 1 % of any of the cell types examined. The SC of S. cerevisiae offers an interesting comparison. Moens & Rapport (1971) said that the absence of conventional SCs was the most striking feature of meiotic prophase nuclei. Zickler & Olson (1975) found SCs and have continued to do so. Byers & Goetsch (1975) satisfactorily resolved the problem. They examined a temperature-sensitive cell-cycle mutant cdc-\. When homozygous it produces continuous SCs in meiotic cells at the permissive temperature, and intermittent SCs at the restrictive temperature. No SCs were found in the heterozygous a/r-4/wild-type strain. Thus the extent ot the SC may be a variable feature found only in particular strains. Olson et al. (1978) have also reported the absence

9 Synaptonemal complexes in Ustilago maydis 179 of SCs during meiosis in Schizosaccharomyces pombe and Olson & Zimmermann (1978 a) have shown that gene conversion (which requires chromosome pairing and hybrid DNA formation) occurs before SCs form during meiosis in S. cerevisiae. This also suggests that there may not be an absolute requirement for SCs during meiosis in these fungi. Some other fungi lack SCs for a more specialized reason: they have localized or restricted recombination. Zickler (1973) reported an absence of normal SCs in Podospora anserina and P. sertosa (Ascomycetes). P. anserina normally produces 4 binucleate spores, the nuclei being of opposite mating type. The restricted SCs are apparently related to the presence of a single crossover per arm, which causes the mating type loci to segregate at the second meiotic division, and presumably facilitates the inclusion of nuclei of opposite mating type into the same spore. The cultivated mushroom Argaricus bisporus (Basidiomycete) produces pairs of dikaryotic spores and 70 % of these produce homothallic mycelia, suggesting that they may be heterozygous for mating type alleles (see Evans, 1959) in a system analogous to that of P. anserina. A. bisporus also has several very short pieces of SC (personal observation) and this may be usual in species producing homothallic dikaryotic spores. No SCs were found in mitotic U. maydis cells undergoing meiotic levels of recombination to repair DNA damaged by radiation (see Holliday, 1971). S. cerevisiae can also undergo recombination repair of irradiated DNA without SCs (Olson & Zimmermann, 19786). In order for recombination repair of double-strand breaks in DNA it may be necessary for the homologous chromosomes to be paired as intimately as they are in meiosis. Diploid U. maydis suffers around 30 % mortality after 450 krad of y irradiation, and genes altered by the recombination repair process in survivors can be expressed by the production of active enzyme 4-5 h after irradiation (Holliday, 1971). G 2 diploid cells can repair around 100 double-strand breaks per haploid genome. It can be calculated that death only occurs when all 4 chromatids (cells are in G 2 phase) are broken in a short region approximately the length of a gene or operon (Fletcher, 1981). The rapid formation of hybrid DNA during repair suggests that the homologues are paired in mitotic cells. This close-pairing may be facilitated by the small size of the chromosomes, the average 5. cerevisiae chromosome is 1/6th the size of the genome of Escherichia coli. It seems to be possible for some fungi to undergo meiotic recombination and segregation without the complication of an SC. The discovery of almost entirely haploid single cells in the gall tissue was surprising. Infection with 2 haploids differing at 2 loci (mating type and infectivity) is necessary to produce a gall in the host. Diploid strains that are heterozygous at these 2 loci are solopathogenic: they do not need a complementary strain to become infectious. Compatible haploid cells will fuse on mating medium. Mass matings will produce multinucleate cells. Some singleton nuclei in these cells may move into buds and become haploid again. True heterokaryotic cells form very long 'infection' hyphae. These will slowly grow and divide on medium containing charcoal. They do not have clamp connections and readily revert to haploid yeast on other media (Day & Anagnostakis, 1971). The charcoal is thought to absorb a product of the U. maydis, which

10 180 H. L. Fletcher would promote the change to the yeast form. The observations reported here suggest that a limited amount of heterokaryon is formed, which infects the host and induces the galls. The heterokaryon may then break down, at least partially, to give equal numbers of the 2 haploid types, which may proliferate in the gall. Diploid sporeforming cells appear to be produced by the fusion of haploid cells. My thanks are due to Dr R. Holliday for guidance and discussion and to Dr P. R. Day for criticism of the manuscript. The work was supported by an M.R.C. Training Fellowship. REFERENCES AINSWORTH, G. C. & SAMPSON, K. (105 )- The British Smut Fungi. Kew, Surrey, England: The Commonwealth Mycological Institute. BYERS, B. & GOETSCH, I. (1975). Electron microscopic observations of the meiotic karyotype of diploid and tetraploid Saccharomyces cerevisiae. Proc. natn. Acad. Sci. U.S.A. 7a, DALTON, A. J. (1955). A chrome-osmium fixative for electron microscopy. Anat. Rec. 121, 281. DAY, P. R. & ANAGNOSTAKIS, S. L. (1971). Corn smut dikaryon in culture. Nature, new Biol. 231, EVANS, H. J. (1959). Nuclear behaviour in the cultivated mushroom. Chromosoma 10, FISCHER, G. W. & HOLTON, C. S. (1957). Biology and Control of the Smut Fungi. New York: Ronald Press Co. FLETCHER, H. L. (1981). Resistance to radiation, recombination repair of DNA and chromosome organisation. Mutat. Res. 80, HOLLIDAY, R. (1971). Biochemical measure of the time and frequency of radiation induced allelic recombination in Ustilago maydis. Nature, new Biol. 232, HOLLIDAY, R. (1977). Recombination and meiosis. Phil. Trans. R. Soc. Lond. B 277, HOTTIG, W. (1931). Ober den Einfluss der Temperatur auf die Keimung und Geschlechterverteilung bei Brandpilzen. Z. Bot. 24, HOTTIG, W. (1933). Ober den Physikalische und Chemische Beeinflussungen des Zeitpunktes der Chromosomenreduction bei Brandpilzen. Z. Bot. 26, MlMS, C. W. (1977). Ultrastructure of teliospore formation in the cedar-apple rust fungus Gymnosporangittm juniper-virginianae. Can. J. Bot. 55, MOENS, P. B. & RAPPORT, E. (1971). Synaptic structures in the nuclei of sporulating yeast, Saccharomyces cerevisiae (Hanson). J. Cell Sci. 9, MULLER, L. L. & JACKS, T. J. (1975). Rapid chemical dehydration of samples for electron microscopic examinations. J. Histochem. Cytochem. 23, OLSON, L. W., EDEN, U., EGEL-MITANI, M. & EGEL, R. (1978). Asynaptic meiosis in the fission yeast? Hereditas 89, OLSON, L. W. & REICHLE, R. (1978). Synaptonemal complex formation and meiosis in the resting sporangium of Blastocladiella emersonii. Protoplasma 97, OLSON, L. W. & ZIMMERMANN, F. K. (1978a). Meiotic recombination and synaptonemal complexes in Saccharomyces cerevisiae. Molec. gen. Genet. 166, OLSON, L. W. & ZIMMERMANN, F. K. (19786). Mitotic recombination in the absence of synaptonemal complexes in Saccharomyces cerevisiae. Molec. gen. Genet. 166, REYNOLDS, E. S. (1963). The use of lead citrate at high ph as an electron opaque stain in electron microscopy. J. Cell Biol. 17, SPURR, A. R. (1969). A low-viscosity epoxy resin embedding medium for electron microscopy. J. Ultrastruct. Res. 26, ZICKLER, D. (1973). Fine structure of chromosome pairing in ten Ascomycetes: meiotic and premeiotic (mitotic) synaptonemal complexes. Chromosoma 40, ZICKLER, D. & OLSON, L. W. (1975). The synaptonemal complex and the spindle plaque during meiosis in yeast. Chromosoma 50, (Received 12 January 1981)