Ascospore Development in the Fission Yeasts Schizosaccharomyces Pombe and S. Japonicus

J. Cell Sd. 56, 263-279 (1982) 263 Printed in Great Britain © Company of Biologists Limited 1982

ASCOSPORE DEVELOPMENT IN THE FISSION YEASTS SCHIZOSACCHAROMYCES POMBE AND S. JAPONICUS

KENJI TANAKA* AND AIKO HIRATA Institute of Applied Microbiology, The University of Tokyo, Bunkyo-ku, Tokyo 113, Japan

SUMMARY The fine structure of ascospore formation in the fission yeasts Schizosaccharomyces pombe and Schizosaccharomyces japonicus var. japonicus was studied by serial thin-sectioning and electron microscopy. The morphogenetic events were almost the same in both species. Ascospore development was initiated by the formation of the forespore membrane on the cytoplasmic side of the differentiated nucleus-associated organelle (NAO) in the interval between meiosis I and II in S. pombe, or during the post-meiotic nuclear division in S. japonicus, and the process proceeded almost synchronously through the two or four nuclei in the ascus. The forespore membrane developed by fusion of the cytoplasmic vesicles and this was clearly demonstrated in S. japonicus where the behaviour of vesicles involved in the forespore membrane development could be traced as they were marked by the presence of electron-dense granules. The staining technique, by phosphotungustic acid-chromic acid (PTA-CA) after treatment with periodic acid, was used to attempt to elucidate the origin and the nature of the forespore membrane. The method specific to plasmalemma-type membranes stained both ascus and ascospore plasmalemmas; the forespore membrane was not stained at first but developed the same affinity for stain as the plasma membrane in the course of ascospore development. The results suggest that the forespore membrane did not come directly from the ascus plasma membrane, but from another membrane system such as the endoplasmic reticulum. Spore wall material was deposited in the space between the inner and outer leaflets of the forespore membrane.

INTRODUCTION Ascospore development in yeasts seems to be one of the most advantageous systems for the study of cell differentiation. Ascospore formation in bakers' yeast involves zygote formation by conjugation between sexually compatible cells, meiotic divisions and ascospore development in the ascus. Ascospore delimitation in Saccharomyces cerevisiae has been observed to be initiated at the site of spindle-pole bodies or nucleus-associated organelles (NAOs) of the second meiotic spindles, by the development of a pair of unit membranes (Moens & Rapport, 1971; Peterson, Gray & Ris, 1972), between which the spore wall is subsequently laid down (Lynn & Magee, 1970; Beckett, Illingworth & Rose, 1973). However, the origin of the ascospore- delimiting membrane has not been clarified.

• Present address: Institute of Medical Mycology, Nagoya University School of Medicine, Showa-ku, Nagoya 466, Japan. 264 K. Tanaka and A. Hirata The fission yeast Schizosaccharomyces pombe has been used widely for cell-cycle and genetic studies, and the ultrastructure of ascospore formation also has been studied by several investigators (Conti & Naylor, i960; Yoo, Calleja & Johnson, 1973). They have established the basic outline of events in the process leading to the formation of ascospores, but have not revealed any details of the nuclear structures in the process, because they used fixation with permanganate exclusively for studies of the membrane structures. Olson, Eden, Egel-Mitani & Egel (1978) have provided electron micrographs of sections of S. pombe nuclei undergoing meiosis. Our own experience of early phases of this process has been similar (Hirata & Tanaka, 1982) and the conclusions have been incorporated into Fig. 1. This paper describes the ultrastructural details of the origin and formation of the ascospore-delimiting membranes of glutaraldehyde-osmium tetroxide fixed cells during ascosporogenesis in S. pombe and in Schizosaccharomyces japonicus var. japonicus as demonstrated by serial thin-sectioning and electron microscopy.

MATERIALS AND METHODS Organisms and culture for sporulation S. pombe strain I190 a homothallic strain and S. japonicus Yukawa et Maki var. japonicus IFO 1713 were used in this study. For the sporulation of S. pombe, cells were cultured in presporulation medium, which contained 3 % glucose and 0-5 % yeast extract, and harvested at the early stationary phase of growth. After washing once in distilled water, cells were inoculated into 100 ml of sporulation medium (SSL) in conical flasks at a density of about 5 x 10' cells per ml (Egel, 1971). They were shaken at 30 °C. In these conditions sporulation was completed within 24 h. Sporulation of S. japonicus was brought about by following the method described by Tsuboi, Ohashi, Takahara & Hayashibe (1978). Modified Burkholder medium (B medium) supplemented with 0^5 % yeast extract and 0-5 % malt extract (BYM medium) was prepared for the presporulation culture, which was kept in the dark to avoid light-induced sexual flocculation. Sporulation was carried out in B medium, from which (NH4),SO4 was omitted (B — N medium), in the light with shaking, at 26 °C. Electron microscopy For electron microscopy aliquots were taken at intervals from the sporulation culture, and cvi vol. of glutaraldehyde fixative was added before centrifugation. Centrifuged cells were resuspended in fresh fixative at room temperature for 2 to 3 h, and then stored in a refrigerator. The fixative was a 3 % glutaraldehyde solution in potassium phosphate buffer (pH 7-0). Fixed

Fig. 1. A diagrammatic representation of meiosis and the development of ascospores in S. pombe and S. japonicus var. japonicus. A to D are common to the two species. A. A zygote with two nuclei (dotted), each associated with a single NAO. B. Fusion of the two nuclei is preceded by fusion of both NAOs. c. A fused and elongated nucleus, with a fused NAO at one end containing a number of linear elements, which is thought to be a stage of meiotic prophase. D. Meiosis I nucleus with a spindle (parallel lines) between a pair of NAOs. E and F, from S. pombe. E. Forespore membranes of double membrane sacs originating on the cytoplasmic sides of the NAOs cover the half-spindles of meiosis II in the two nuclei. F. Four ascospores with thick walls. Spore nuclei carry their respective NAOs. c, H and 1, from S. japonicus. c. Two nuclei with spindles at meiosis II in the ascus. H. Forespore membranes developing on the differentiated NAOs of the post-meiotic spindles of four nuclei. Development seems to proceed synchronously through the four nuclei. 1. An eight-ascospore ascus. Ascospore development in fission yeast 265

Fig. 1 266 K. Tanaka and A. Hirata

V. 30L1

'•% c Ascospore development in fission yeast 267 cells were washed several times in buffer, and treated with Zymolyase 60000 (Kirin Brewery Co. Ltd, Takasaki, Japan) (o-i mg/ml) in the same buffer at 30 °C. The disintegration of the cell wall was examined with a phase-contrast microscope. The Zymolyase-treated cells were washed again in buffer and postfixed in 2 % OsO4 for 2 h at room temperature. After washing in distilled water, they were soaked in a 0-5 % (w/v) aqueous solution of uranyl acetate for 2 h and embedded in agar blocks. The cells were dehydrated by passing them through a series of increasing concentrations of ethanol and absolute acetone, and then they were embedded in Spurr's resin. Serial sections were obtained with a Dupont diamond knife on a Porter- Blum MT-2 ultramicrotome and treated by one of the following methods. (1) Sections were mounted on Formvar-coated single-slot grids and stained with uranyl acetate and lead citrate. (2) In order to apply specific staining to the plasmalemma, sections were collected using a small plastic sheet with a hole (3-0 mm diameter) punched through it at one end. Sections floating on a film of distilled water within the hole were treated with 1 % periodic acid and stained with phosphotungustic acid plus chromic acid (PTA-CA) (Roland, Lembi & Morr6, 1972). Sections were viewed with a JEOL 200 CX electron microscope at 100 kV.

RESULTS Ascospore formation in the fission yeast involves the developmental stages of cell flocculation, cell fusion, karyogany, meiotic nuclear divisions, ascospore delimitation by the development of double unit membranes or forespore membrane, and spore wall maturation. The morphological changes in the meiotic nuclei and the process leading to ascospore formation were almost the. same in 5. pombe (Hirata & Tanaka, 1982) and S. japonicus (Hirata & Tanaka, unpublished observations), except that the latter underwent one additional nuclear division before ascospore delimitation began. This is illustrated in Fig. 1. Conjugation of the two haploid cells resulted in a binucleate zygote (A) where karyogamy had occurred by fusion of the two nucleus-associated organelles (NAOs) (B). At the presumed prophase in meiosis I, the zygote nucleus with a fused NAO at one end was elongated and contained a number of filamentous linear elements (c) (Olson et al. 1978). Nuclear division was effected by a spindle consisting of microtubules connecting NAOs at opposite poles (D). In 5. pombe ascospore delimitation first became apparent at the start of meiosis II (E) and a four-spored ascus was produced (F). In 5. japonicus this event is delayed because in this species meiosis is followed by another round of nuclear division (G, H). The primary envelopes for the eight spores about to be created begin to be formed simultaneously near the NAOs at the spindle poles of the dividing haploid nuclei even before these mitoses have been completed (H, I).

Figs. 2-10. S. pombe. Fig. 2. A zygote at meiosis II. Two nuclei that are the products of meiosis I seem to be at almost the same stage of meiotic division as shown by the presence of spindle microtubules (arrow). This indicates that meiosis II is proceeding synchronously between the two nuclei, x 16coo. Fig. 3. Two consecutive sections (A, B) show the multi-plaque structure of the modified NAO (nao) at meiosis II and the forespore membrane (fin) developing on it. Traces of spindle microtubules can be seen (mt). A is an enlarged view of one of the two nuclei in the sections shown in Fig. 2. Associations of endoplasmic reticulum (er) and vacuolar membranes with forespore membrane can be seen, x 60000. 268 K. Tanaka and A. Hirata

Fig. 4. An ascus stained with PTA-CA in which the two nuclei at meioais II are associated with the developing forespore membrane. The ascus plasmalemma (pm) with large imaginations, and one segment of forespore membrane, are stained positively, x 12000. Fig. 5. The left-hand nucleus in Fig. 4 at higher magnification. Forespore membranes (Jm) in the upper part are not stained, while those accompanying a few vesicles in the lower part are stained to medium density. The nuclear envelope (ne) is not stained, pm, plasmalemma. x 42720. Fig. 6. The right-hand nucleus in Fig. 4, enlarged, where one segment of forespore membrane (Jm) is positively stained to the same extent as that of the ascus plasmalemma (pm), while the other is definitely unstained (lower part, arrow), ne, nuclear envelope, x 42720. Ascospore development in fission yeast 269

S. pombe The two products of meiosis I underwent meiosis II almost synchronously (Fig. 2). The NAO of the dense plaque embedded in the nuclear envelope of the meiosis I spindle has now elaborated several layers of dense plaques, and the forespore membrane is being developed immediately adjacent to the outermost plaque on the cytoplasmic side (Fig. 3 A, B). The forespore membranes at both NAOs might be initiated at the same time and developed at the same rate, because serial-section analysis has showed that the forespore membranes developed to a similar extent on both sides, concomitant with spindle elongation. The forespore membrane was sometimes connected to flattened vesicles or tubular reticulum, which was definitely connected to the smooth endoplasmic reticulum running along the plasmalemma (Fig. 3 A, B). Fusion of cytoplasmic vesicles with the forespore membrane can also be seen in Fig. 7. Fig. 8 shows a nucleus (meiosis being finished but the forespore membrane not being closed) where the layered plaques of the meiosis II NAO are no longer clear and have dedifferentiated into a single dense NAO on the nuclear envelope. When sections of the ascospore-forming ascus were subjected to PTA-CA staining, not all of the forespore membranes showed positive reactions. The ascus plasmalemma and membranes or vesicles derived from it were stained positively, but the nuclear envelope, endoplasmic reticulum and mitochondrial membrane were not stained (Fig. 4). The forespore membranes surrounding both nuclei at meiosis II seemed to be developed to a similar extent, but they were not uniformly stained by PTA-CA. The forespore membranes surrounding one of the meiotic nuclei were only faintly stained (Fig. 5), while those of the other nucleus were asymmetrically stained; one leaf of the membranes being stained positively and the other negatively (Fig. 6). Wall material was deposited between the double unit membranes; thus the inner membrane facing the nucleus constituted the plasmalemma of the ascospore proto- plasm. The nucleus, mitochondria, endoplasmic reticula, lipid granules and other granules were enclosed in the ascospore (Fig. 9). In mature asci with four ascospores, PTA-CA clearly stained the plasmalemmas of both ascus and ascospores as well as the outer membranes or spore coats of the ascospores, while membranes of the cytoplasmic organelles were left unstained (Fig. 10).

S. japonicus var. japonicus Ascospore formation in S. japonicus began in the ascus that contained four nuclei after two successive meiotic divisions (Fig. 1). The four-nuclei ascus contained a large number of dense granules in the cytoplasm (Fig. 11), which were enclosed in vesicles 50-80 nm in diameter and seen abundantly in the region of the smooth endoplasmic reticulum (Fig. 12). The nuclei bore differentiated NAOs of multi- layered plaques, in the proximity of which a few cytoplasmic vesicles were seen to aggregate to form incipient forespore membranes (Fig. 13). Fig. 14 shows the last or post-meiotic nuclear division by a spindle consisting of microtubules that connect opposite NAOs. No part of the nuclear envelope was disrupted. Forespore membranes 270 K. Tanaka and A. Hirata Ascospore development in fission yeast 271 formed on the outermost plaque of the modified NAOs at opposite poles seemed to envelop each half-spindle to a similar extent. Pairs of unit membranes of the forespore membrane were usually closely apposed to one another and each pair was separated by a space about 27 nm in diameter, in the middle of which electron-dense material formed a distinct layer. In places, particularly at the advancing or growing front of the forespore membrane, double membranes were separated giving the appearance of a flattened sac. Cytoplasmic vesicles carrying dense material were mostly found in the cytoplasm between the nuclear envelope and the forespore membrane (Figs. 14, 16). One such vesicles can be seen apparently in the process of fusion with the inner leaf of the forespore membrane, thus adding its share to the dense matter that at this time begins to accumulate in the space between the two leaves of the forespore membrane (Fig. 15). We conjecture that the numerous vesicles with dark centres seen in the vicinity of developing forespore membranes in Figs. 16 and 17 have the same function, since Johnson, Yoo & Calleja (1973) in their study of cell plate growth in S. pombe have shown examples of ' dense round bodies' that suggest participation in that process. In order to determine the origin and nature of the cytoplasmic vesicle membrane and the forespore membrane, sections were treated by the PTA-CA method. The results were the same as in S. pombe, and two serial sections were presented to demonstrate the reliability of the staining technique (Fig. 16A, B). Membranes of dense vesicles were in general not stained by PTA-CA, suggesting that their origin was in the cytoplasmic membrane system; the forespore membranes were not uniformly stained. Parts of the double membranes joined by a layer of dense material were not stained, while other parts with swollen intercisternal spaces were stained (Figs. 17, 18). The developing forespore membrane finally encircled the divided ascospore nucleus and a pair of unit membranes joined by a dense layer could be seen (Figs. 19, 20). Further maturation of the ascospore wall seemed to proceed by deposition of wall material between the inner leaf of the forespore membrane and the dense intermediate layer (Figs. 21, 22). A section of seven ascospores with a thick spore wall within a matured eight-spored ascus is shown in Fig. 23, in which the ascus and ascospore plasmalemmas and the spore coat were definitely stained by the PTA-CA method.

Fig. 7. Cytoplasmic vesicles (v) fusing with the forespore membrane (Jm). x 75coo. Fig. 8. The forespore membrane (Jm) fairly well developed, but not completely encircling the nucleus, which has completed meiosis II. x 35000. Fig. 9. The spore wall (no) has developed between the two unit membranes of the forespore membrane. Nucleus (n), mitochondria (mi), ER (er), lipid globule (/) and other vesicles are enclosed in the spore, x 35000. Fig. 10. A four-spore ascus. The ascospores are not fully mature. The section is stained with PTA-CA, by which method ascus and spore plasma membranes as well as the peripheries of the spore coat are stained, while nuclei and mitochondria within the spores are not stained, x 16200. 272 K. Tanaka and A. Hirata

13 *•% 12 Ascospore development in fission yeast 273 Figs. 11-23. S.japonicus var. japonicus. Fig. 11. A zygote after meiosis II, three of the four nuclei (n) can be seen. Note the occurrence of dense granules in the cytoplasm, x 35000. Fig. 12. Part of the cytoplasm of a cell at a stage similar to the one shown in Fig. 11. Dense granules are enclosed in vesicles (v) that seem to be derived from smooth endoplasmic reticulum («•). x 41000. Fig. 13. Part of the right-hand nucleus shown in Fig. n, enlarged. A few cytoplasmic vesicles forming the initial forespore membrane (Jm) are seen close to the NAO (nao) on the nuclear envelope (ne), from which multi-plaques are differentiating, x 80000. Fig. 14. A nucleus with a spindle at the post-meiotic division. The nuclear envelope (ne) is preserved and spindle microtubules (mi) connect the modified NAOs at both poles. A pair of forespore membranes (Jm) are enveloping each half- spindle and seem to have developed to a similar extent on both sides (indicated by arrows). The two adjacent membranes of the forespore membrane are in most regions joined by a layer of dense material and in places, particularly at the advancing fronts of the membranes, are swollen into flattened sacs. Vesicles carrying dense granules are seen exclusively in the cytoplasm enclosed by the developing forespore membrane, x 41000. Fig. 15. A cytoplasmic vesicle (v) earring dense material is fused with the inner leaf of the forespore membrane (fin), x 80000. Fig. I6A, B. TWO serial sections (but with four intervening sections skipped) of an ascus at the post-meiotic division stained by the PTA-CA method. Two sets of forespore membranes surrounding the two nuclei (n) and part of the forespore membrane of the third nucleus (arrows on the right-hand side) can be seen, x 9770. Fig. 17. The foreapore membranes (Jm) around the left-hand nucleus of Fig. I6B, enlarged. The regions of the forespore membrane joined to a dense layer are not stained, but those with a swollen cisternal space are stained, ne, nuclear envelope, x 27140. Fig. 18. The forespore membrane (Jm) around the right-hand nucleus of Fig. I6B, enlarged. The swollen parts of the forespore membrane are stained, while most of the membranes of vesicles with dense granules and the nuclear envelope are not stained. ne, nuclear envelope, x 27 140. Fig. 19. The foreapore membrane (Jm) is closed and includes a nucleus (n), mitochondria (m) and other organelles. Most of the forespore membrane is filled with material, x 40000. Fig. 20. Part of the forespore membrane in the section adjacent to that shown in Fig. 19, enlarged. Pairs of unit membranes of the forespore membrane closely apposed to one another; the space in the middle of each pair contains electron-dense material forming a distinct layer, x 120000. Fig. 21. A maturing ascospore. Wall material (no) seems to be deposited between a layer of dense material and the inner leaf of the forespore membrane, x 40000. Fig. 22. Higher magnification of part of Fig. 21. x 120000. Fig. 23. Seven ascospores in the eight-spore ascus are shown by PTA-CA staining. Ascus, ascospore plasmalemmas and spore coat are distinctly stained, x 7360. 274 K. Tanaka and A. Hirata

fmj

X-

For legend see p. 273. Ascospore development in fission yeast 275

n t

n n

17 18

For legend see p. 273. 276 K. Tanaka and A. Hirata

23 For legend see p. 273. Ascospore development in fission yeast 277

DISCUSSION Ultrastructural changes in the fission yeast S. pombe during ascospore development have been reported by seveial authors (Conti & Naylor, i960; Yoo et al. 1973), but they were all results obtained on cells fixed with permanganate, which impaired the nucleoplasm and therefore gave information only on the membrane structure. This paper reports the fine structure of ascospore development in the two species of Schizosaccharomyces fixedwit h glutaraldehyde and osmium tetroxide and attempts to clarify cytochemically the origin of the ascospore wall by using PTA-CA staining. Ascospore development began in the interval between meiosis I and II in S. pombe, producing four-spored asci, and during a post-meiotic division in S.japonicus, where eight-spored asci were produced. The earliest morphological event in ascospore development was the appearance of the forespore membrane initially on the differentiating nucleus-associated organelle (NAO) of a multi-plaque structure, and this was clearly demonstrated at the four-nuclei stage of S. japonicus (Fig. 13). The multi-plaque structure of the NAO during the last nuclear division in ascospore development in the ascus was commonly observed among the ascomycetes. Ashton & Moens (1979), in discussing the taxonomic significance of the ultrastructure of the NAO, concluded that NAO structure was relatively consistent in higher ascomycetes, while variety could be found in the hemiascomycetes. According to their categorization of NAO morphology at meiosis II, the NAO of S. pombe may belong to type II, which consist of three plaques. This was also in accord with their unpublished observations. It also seems to be a common feature among ascomycetes that the forespore membrane or prospore wall is initiated as a disk-shaped vesicle on the cytoplasmic side of the NAO. In ascospore development, it was established that ascospore wall material is deposited between the double membrane envelope or forespore membrane, which is derived from fusion of cytoplasmic vesicles. However, the origins of the vesicles and the wall material remain unknown. Some authors claim that vesicles are produced by the imagination of the ascus plasmalemma (Bandoni, Bisalputra & Bisalputra, 1967; Syrop & Beckett, 1972), while others have proposed that they are derived by blebbing from nuclear membrane (Carroll, 1967; Oso, 1969; Yoo et al. 1973) or endoplasmic reticulum (ER) (Lynn & Magee, 1970; Black & Gorman, 1971). The former presumed that the plasmalemma-type characteristics of the membrane would be conserved, while the latter assumed that there would be a change from ER- to plasmalemma-type during ascospore development, considering that the inner membrane of the double forespore membrane would constitute the ascospore plasmalemma. In this study we made observations from which we conclude that the forespore membrane was derived from the fusion of vesicles that originated from endoplasmic reticulum: in S. pombe the forespore membrane was observed in direct continuity with the ER and in S. japonicus the behaviour of vesicles carrying dense material presented more direct evidence for the origin of the forespore membrane. Syrop & Beckett (1972) demonstrated the membrane continuity, by showing that the spore-delimiting membranes were formed by the invagination of the ascus 278 K. Tanaka and A. Hirata plasmalemma, using the PTA-CA method, which preferentially stains plasmalemma- type membrane. When the same technique was applied in this study to ascospore development in the fission yeast, results were not consistent enough to conclude that the double membrane had a single origin. We found some vesicles stained with PTA-CA near the unstained forespore membrane, which was stained with medium density as development proceeded and eventually very densely as it formed the ascospore plasmalemma. The conclusion reached was that the forespore membrane with ER characteristics at the early stage of its development might change progressively to acquire plasmalemma characteristics, this process of membrane modification possibly being facilitated chemically or by adding plasmalemma- derived vesicles to the forespore membrane. This process of chemical modification seemed to progress asynchronously between the separate nuclei in an ascus, while morphological development went on synchronously. Ascospore wall material was thought to be deposited between the double leaflets of the forespore membrane, but where it was synthesized and how it was deposited are not known. It was clearly demonstrated, at least in S. japonicus, that part of the wall material was carried by the vesicles fused with the forespore membrane, as was shown by the behaviour of the dense material carried by the vesicles. Thickening of the ascospore wall seemed to occur after the delimitation of the ascospore was complete, and in S. japonicus this seemed to proceed between the layer of dense material and the inner leaf of forespore membrane. The synthesis and mode of deposition of the wall material require further investigation.

This work was supported in a part by grants nos. 344001 and 56470127 from the Ministry of Education, Science and Culture, Japan.

REFERENCES ASHTON, M. L. & MOENS, P. B. (1979). Ultrastructure of sporulation in the hemiascomycetes Atcoidea corymbosa, A. rubescent, Cephaloascus fragrans, and Saccharomycopsis capsularis. Can.J. Bot. 57, 1259-1284. BANDONI, R. J.( BISALPUTRA, A. A. & BISALPUTRA, T. (1967). Ascospore development in Hansenula anomala. Can. J. Bot. 45, 361-366. BECKETT, A., ILLINGWORTH, R. F. & ROSE, A. H. (1973). Ascospore wall development in Saccharomyces cerevisiae. J. Bad. 113, 1054-1057. BLACK, S. H. & GORMAN, C. (1971). The cytology of Hansenula. III. Nuclear segregation and development during ascosporogenesis in Hansenula taingei. Arch. Mikrobiol. 79, 231-248. CARROLL, G. C. (1967). The ultrastructure of ascospore delimitation in Saccobolus kerverni J. Cell Biol. 33, 218-224. CoNTi, S. F. & NAYLOR, H. B. (i960). Electron microscopy of ultrathin sections of Schizo- saccharomyces octosporus. III. Ascosporogenesis, ascospore structure, and germination. J. Bad. 79, 417-425. EGEL, R. (1971). Physiological aspects of conjugation in fission yeast. Planta 98, 89-96. HIRATA, A. & TANAKA, K. (1982). Nuclear behaviour during conjugation and meiosis in the fission yeast Schizosaccharomyces pombe. J. gen. appl. Microbiol. 28 (In Press.) JOHNSON, B. F., YOO, B. Y. & CALLEJA, G. B. (1973). Cell division in yeasts: Movement of organelles associated with cell plate growth of Schizosaccharomyces pombe. J. Bad. 115, 358-366. LYNN, R. R. & MAGEE, P. T. (1970). Development of the spore wall during ascospore formation in Saccharomyces cerevisiae. J. Cell Biol. 44, 688-692. Ascospore development in fission yeast 279 MOENS, P. B. & RAPPORT, E. (1971). Spindles, spindle plaques, and meiosis in the yeast Saccharomycet cerevisiae (Hansen). J. Cell Biol. 50, 344-361. OLSON, L. W., EDEN, U., EGEL-MITANI, M. & EGEL, R. (1978). Asynaptic meiosis in fission yeast? Hereditas 89, 189-199. Oso, B. A. (1969). Electron microscopy of ascus development in Ascobolus. Arm. Bot. 33, 205-209. PETERSON, J. B., GRAY, R. H. & Ris, H. (1972). Meiotic spindle plaques in Saccharomyces cerevisiae. J. Cell Biol. 53, 837-841. ROLAND, J.-C, LEMBI, C. A. & MORRE, D. J. (1972). Phosphotungustic acid-chromic acid as a selective electron dense stain for plasma membranes of plant cells. Stain Technol. 47, 159- 200. SYROP, M. J. & BECKETT, A. (1972). The origin of ascospore-delimiting membranes in Taphrina deformam. Arch. Mikrobiol. 86, 185-191. TSUBOI, M., OHASHI, K., TAKAHARA, M. & HAYASHIBE, M. (1978). Sexual process in eight- spored ascus-forming yeast, SMzosaccharomyees japonicus. I. Culture medium for the synchronous sexual process. Plant & Cell Pkytiol. 19, 1327-1332. Yoo, B. Y., CALLEJA, G. B. & JOHNSON, B. F. (1973). Ultrastructural changes of the fission yeast (Schizosaccharomycet pombt) during ascospore formation. Arch. Mikrobiol. 91, 1-10.

(Received 21 August 1981)