The role of sterility genes (ste and aft) in the initiation of sexual development in Schizosaccharomyces pombe

Transcription

1 Mol Gen Genet (1988) 213: Springer-Verlag 1988 The role of sterility genes (ste and aft) in the initiation of sexual development in Schizosaccharomyces pombe M. Sipiczki Department of Genetics, L.K. University, P.O. Box 56, H-4010 Debrecen, Hungary Summary. Haploid homothallic strains of Schizosaccharomyces pombe with mutations in any of nine "sterility genes" (ste) do not mate with wild-type fertile strains. Those defective in genes stel to ste4 and ste7 to ste9 are also deficient in meiosis and sporulation. I found that the stel, ste3 and ste8 genes act very early in the sexual development, presumably before the patl-controlled conjugation-specific event. ste5 and ste6 exert their function downstream of patl in the initiation of conjugation and do not play any role in the meiotic pathway, ste2, ste4, ste7 and ste9 are involved in both sexual pathways: they seem to act downstream of patl in conjugation but upstream of patl in the initiation of meiosis. A new gene, affl, whose defective allele suppresses the patl-114-provoked haploid sporulation and arrest of vegetative growth is also described. It is supposed that the affl + gene product participates in a cascade of regulatory events, as a factor antagonistic to patl. Key words: Fission yeast - Sterility - Conjugation - Meiosis - Protoplast fusion Introduction In the cell cycle of Schizosaccharomyces pombe (S. pombe vat. pombe, according to a recent taxonomic revision of the genus; see Sipiczki et al. 1982), a unicellular Ascomycete, three developmental pathways can be distinguished (see Gutz et al for a review). During most of its life cycle, the cells are haploid and grow vegetatively by cell elongation, accompanied by mitotic nuclear division and cell septation (fission). Starvation, particularly of a nitrogen source, arrests cells predominantly in the G1 phase of the cell cycle and causes cells of opposite mating type (h + and h-) to undergo sexual conjugation leading to the formation of an h-/h + diploid zygote. Conjugation is normally followed by meiosis and sporulation. However, there is a short period after completion of zygote formation during which the cell is able to return to the mitotic cell cycle when transferred to a nutritionally rich medium. This period, during which the cell is not yet committed to meiosis separates the conjugation-specific developmental pathway from the meiotic one. The initiation and regulation of the two sexual pathways (conjugation and meiosis-sporulation) seem to share certain features. Both are initiated by nutritional deprivation (Egel 1971) and both require the activity of genes stel to ste4 and ste7 to ste9 (class II ste genes) (Thuriaux et al. 1980; Girgsdies 1982; Michael and Gutz 1987; U. Leupold, personal communication). Partial inactivation of the ranl/patl gene product initiates conjugation and its complete loss triggers meiosis (Nurse 1985; Iino and Yamamoto 1985a; Beach et al. 1985). The ranl/patl gene is a negative regulator of both conjugation (Beach et al. 1985) and sporulation (Beach et al. 1985; Iino and Yamamoto 1985b) and was found to be essential for mitotic division (Beach et al. 1985). Two temperature-sensitive recessive mutant alleles, ran1-3 (Nurse 1985) and patl-114 (Iino and Yamamoto 1985a), of this gene were isolated and characterized independently. Since the strain used in this study was a gift from M. Yamamoto, whose laboratory described the patl-li4 allele, the ranl/patl gene will be referred to as patl throughout this paper. This is also in accordance with the convention proposed recently for the naming of genes of S. pombe (Kohli 1987). Strains of S. pombe carrying the recessive temperaturesensitive allele patl-ll4 bypass both normal meiotic requirements (heterozygosity at the mating-type locus and nutritional starvation) when shifted from a permissive to a restrictive temperature. Vegetative growth and mitotic division are inhibited and haploid cells undergo meiosis and subsequent sporulation (Nurse 1985; Iino and Yamamoto 1985a). Transfer of cells to a semi-permissive temperature does not initiate immediate meiosis; the actively dividing cells first slow their rate of growth, accumulate in the G1 phase of the cell cycle, conjugate with cells of opposite mating type and then sporulate (Beach et al. 1985). Thus, the normal requirement of starvation for the initiation of conjugation is bypassed at the semi-permissive temperature, but the partial inactivation ofpatl + cannot switch on meiosis as long as the cell is haploid. In the zygotes, however, a meiotic control gene, mei3, is activated which initiates meiosis and sporulation by further inactivation of patl + (McLeod and Beach 1987). Since the events which occur in the patl-ll4 haploid cells at semi-permissive and restrictive temperatures bear many similarities to true conjugation and meiosis, it has been proposed thatpatl + plays a central role in the sexual development of this organism. In addition to mei3, a further meiotic gene, mei2, is also involved in the initiation of meiosis, mei2 + is required for meiotic committment after full inhibition of pat1 + (Beach et al. 1985; Iino and Yamamoto 1985b). Destruction of its product suppresses the derepression of meiosis which is triggered

2 530 by the inactivation of the patl + gene product, indicating that mei2 acts downstream ofpatl. The activity of the so-called "sterility genes" is also required for accomplishing the sexual cycle (Girgsdies 1982; Michael and Gutz 1987; Thuriaux et al. 1980). Mutants defective in these genes confer a sterile phenotype regardless of the mating type of the cells. Although none of these mutants can conjugate, only those defective in genes stel to ste4 and ste7 to ste9 are also deficient in meiosis and sporulation (the so-called class II ste mutants). Since the product ofpatl is involved in the regulation of both sexual pathways, any information on the functional relationship between the ste genes and patl might contribute to a better understanding of the initiation of these pathways. In the present study I demonstrate that all ste genes act upstream of patl in the initiation of meiosis, whilst they can be divided into two categories as to their interaction with patl in the initiation of conjugation. The identification and characterization of a suppressor gene of the pat1--driven meiosis and sporulation which acts downstream of pat1 in the initiation of both sexual pathways is also discussed. TaMe 1. List of strains Strain OG15stel ura4-294 h- JM57ste2 leul h 9 JM66ste3 leul h 9 JM104 ste4 ade6-m216 h 9 JM70 ste5 leul h 9 JM75 ste6 leul h 9 JM83 ste7 leul h 9 JM86 ste8 leul h 9 B36 ste9 aden704 h 9 HS patl-ll4 ade6-m210 h 9 HS a patl-114 ura4-294 h 9 edc ade3-34 h- ura5-294 h ura4-d18 b- ade7-152 ural 171 cdc h + ade7-50 edc2-35 h 9 ade6-m210 spo18-b317 h 90 arg4-55 h- ura4-d18 leul-32 h- BCa-312 patl 114 ade6-m210 affl-312 h 9 Origin U. Leupold M. Yamamoto this study P. Schuchert P. Schuchert this study Materials and methods Strains. All the strains used were derived from the heterothallic wild-type strain L972 (mating type h-) or L975 (h +) and from the homothallic wild-type strain L968 (hg ). Table 1 lists the genetic markers used in the present study and the origin of strains. Growth media. Yeast-extract liquid (YEL), yeast-extract agar (YEA), minimal agar (MMA) and malt-extract agar (MEA) were as described elsewhere (Sipiczki and Ferenczy 1977). Minimal sorbitol agar (MMAS) was MMA supplemented to contain 1.2 M sorbitol. The liquid minimal medium (a modified EMM2) was prepared according to Castello et al. (1986). Protoplastfusion. The method used was a modified version of the polyethylene glycol-induced protoplast fusion procedure (Sipiczki and Ferenczy 1977; Sipiczki et al. 1985). Cells were grown in YEL to a titre of 107/ml, harvested, washed twice in sorbitol, resuspended in 0.65 M KC1 supplemented with 5 mg/ml Novozym SP234 (Novo Industries A/S, Denmark) so that the final concentration was l0 s cells/ml, and incubated at 25 C for 90 min. For fusion, the protoplasts were collected by careful centrifugation, washed twice in 1.2 M sorbitol, resuspended in 1 ml of 1.2 M sorbitol, and mixed with protoplasts of the other strain. After centrifugation the mixed pellet was gently resuspended in 2 ml of 27% (w/v) PEG (polyethylene glycol, mol. wt. 4000) in 10 mm CaC12. After a 30 rain treatment the suspension was mixed into MMAS at 43 C and transferred as a thin layer to the surface of solidified MMAS plates. The plates were incubated for 1 week at 30 C, and the fusion colonies were isolated and transferred onto MMA slants. Genetic techniques. Methods of random spore and tetrad analysis were essentially the same as those described by Kohli et al. (1977). To induce haploidization, the fusants were cultivated on YEA supplemented with p-fluorophenylalanine (Sipiczki and Ferenczy 1977). Fertility of the segregants was determined on MEA with fertile h and h + strains. Complementation was tested in somatic hybrids constructed by protoplast fusion. Results ste- mutations do not abolish the patl-driven meiosis As both the ste- and the patl-ll4 alleles are recessive, their interaction could be studied most conveniently in stepat- haploids at a temperature restrictive for patl-ll4. However, the inability of the ste- strains to conjugate precludes the construction of these haploids by a conventional cross. To overcome this obstacle, the sterile mutants were hybridized somatically with the pat1 114 strain (Table 2). The method used was a modified version of the polyethylene glycol-induced protoplast fusion procedure (see the Materials and methods). The diploid fusants (somatic hybrids) grown up on the nitrogen-rich selective medium were transferred to MEA where they first multiplied and then, after depletion of nitrogen from the medium, sporulated. Since the fusants were either heterozygous at the mating type locus or h9 /h9 their cells mostly converted to type asci. In addition, however, each fusant clone also formed "giant zygotic asci" due to accidental conjugations between diploid cells prior to meiosis and sporulation (Sipiczki and Kucsera 1983). The spores were spread on the nitrogen-rich YEA medium and the colonies appearing at 23 C (a permissive temperature for patl-ll4) were replicaplated onto MEA and YEA. The MEA plates were incubated at 23 C, whilst the YEA plates were shifted to 35 C (restrictive for pat1-114). After 5 days the plates were treated with iodine vapour to identify the sporulating colonies (colonies containing spores turn dark in iodine vapour; see Kohli et al. 1977). In the hybrids ste2- to ste9- xpatl- 114 where both fusion partners were homothallic, all ste + colonies were expected to sporulate on MEA. From the fusants stel- h- xpatl-114 h 9 two classes ofste + colonies could arise: non-sporulating h and sporulating h 9 ones. Their ratio was expected to be 1:1 because stel and the

3 531 Table 2. Somatic hybridization of ste- mutants with patl-ll4 Fusion Sporulation of fusants Sporulation of ste- pat- recombinants Azygotic asci Zygotic (giant) asci at 35 C at 29 C OG15 stel ura4-294 h- x HSl14-t32 pat1-114 ade6-m210 h 9 JM57 ste2 leui h 9 HSlt4-132 patl-ll4 ade6-m210 h 9 JM66 ste3 leul h 9 HS patl-ll4 ade6-m210 h 9 JM104 ste4 ade6-m216 h 9 x HSl14-t32a patl-l14 ura4-294 h 9 JM70 ste5 leul h 90 HS patl-ll4 ade6-m210 h 9 JM75 ste6 leul h 9 x HS patl-ll4 ade6-m210 h 9 JM83 ste7 leul h 90 x HS patl-ll4 ade6-m210 h 90 JM86 ste8 leul h 9 x HS patl-ll4 ade6-m210 h 9 B36 ste9 ade6~704 h 9 x HSl14-132apatl-l14 ura4-294 h 9 zygotic and zygotic and zygotic and mating-type region are not linked (Thuriaux et al. 1980). On YEA, however, only colonies bearing the patl-114 allele could sporulate because the nitrogen concentration in the medium would be above the inhibitory threshold level even at the time when the cells enter the stationary phase. Furthermore, the patl- cells cannot grow at 35 C, so copies of patl- colonies on YEA remained faint. As for the ste pat- recombinants, two different phenotypes are conceivable. If the lack of the ste + gene product suppresses pat! 114, the recombinants would not sporulate on either of the media. If the reverse is true, the recombinants would form spores on YEA at 35 C but be sterile on MEA at 23 C. Colonies sporulating only on YEA at the restrictive temperature were detected in each ste- xpatl-ll4 combination, indicating that none of the ste- mutations analysed could suppress the patl-ll4-driven haploid sporulation. This observation suggests that genes stel to ste9 all act upstream ofpatl in the initiation of the meiotic pathway. Effect of ste- mutations on the patl-induced conjugation Seven out of the nine ste genes analysed in this study are known to be involved in the initiation of both sexual pathways (see the Introduction). To test whether patl-ll4 also neutralizes their conjugation defect, the homothallic stepat- recombinants were cultivated on MEA for 20 h at 23 C and then shifted to 29 C (a semi-permissive temperature for patl-ll4). At this temperature most cells of the control ste + patl-ll4 h 9 culture conjugated after 24 h, but a fraction of the culture (10%-20%) underwent haploid sporulation. Table 2 shows that neither the ste5 patl-ll4 h 9 nor the ste6 patl 114 h 9 recombinants conjugated, which suggests that the class I ste genes act downstream ofpatl in the conjugation pathway. The recombinants of class II mutants could be grouped into two classes: fertile (class IIa) and infertile (class IIb). In conjunction with the finding that all ste mutants are capable ofpatl-driven sporulation, these data suggest that the genes belonging to the former category (stel, ste3 and ste8) exert their function in both sexual pathways before the patl-regulated event. The class IIb mutants suppress the ability of patl-ll4 to induce conjugation, suggesting that the corresponding genes act downstream of patl in the initiation of this pathway. Isolation of an extragenic suppressor ofpatl-114 The finding that the patl-driven haploid sporulation is not affected by defects in the ste genes suggests that the pat1 + gene product may perform the concluding event which commits the cell to meiosis. However, data are available that contradict this conclusion and indicate that there must be genes acting downstream of pat1 in the initiation of meiosis. It is known that met2- pat1- double mutants neither cease vegetative growth nor sporulate at the restrictive temperature (Iino and Yamamoto 1985a; Beach et al. 1985). In addition, Nurse (1985) observed frequent reversion of the temperature-sensitive ranl-3 mutant (a leaky allele of pat1) to pat +. This observation suggests that there must exist additional genes whose mutations can neutralize the effect of the patl-ll4 defect. Since little is known about the nature of the ran1 revertants and their relationship to met2, I decided to isolate extragenic patl-ll4 suppressors that are not allelic to met2. In patl-li4 all revertants studied to date were found to be met2- mutants (Iino and Yamamoto 1985a). To obtain such suppressors, cells of a pat1-114 strain (patl-114 ade6- M216 h 9 ) were streaked on YEA and plates were incubated at 35 C for 7-10 days. One hundred and fifty independent revertant colonies which evaded the temperature-sensitive growth defect ofpatl-ll4 were isolated and their stability was tested. The 79 stable ones fell into two classes : (i) fertile but sporulation negative and (it) infertile. In a complementation test with the known meiotic mutants, the fertile ones were found to be allelic to met2 and therefore were not studied further. Ten sterile revertants were hybridized by protoplast fusion with the wild-type strain ura4-d18 leul-3 h-. The fusants sporulated when cultivated on MEA and formed both and giant zygotic asci. When the asci were subjected to tetrad analysis, a segregation pattern of 2 sterile: 2 fertile was observed. These findings indicate that the infertility was due to a single chromosomal mutation, and this mutation was recessive. Two sterile spore clones, BCa-312 leul-3 and BCa-312 ura4-d18, isolated from a tetrad of the somatic hybrid constructed between the sterile isolate BCa-312 and the fertile ura4-d18 leul-3 h- were fused with one another. Since the fertile spores of the tetrad were h, the sterile ones had to be h 9. The fusants were sterile and did not sporu-

4 532 late. Thus, the sterility of the revertant BCa-312 resembles the class II ste- mutations, because it causes not only infertility but also inability to sporulate. However, it differs from the ste- mutants in a fundamental feature: BCa-312 is unable to respond to the patl-derepression of meiosis and sporulation while the ste- mutants can undergo the patldriven sexual development. This difference suggests that BCa-312 is defective in a gene acting downstream ofpati. Since little is known about the post-patl part of the initiation pathway and the test of the effect of a single mutant allele per each ste gene on the patl- phenotype (see above) alone cannot completely exclude the possibility of allelism between BCa-312 and one of the ste genes, segregation tests (random spore analysis) were performed. For this, somatic hybrids were constructed between BCa-312 and each ste- mutant. After sporulation, spores from all fusants were spread on MEA. The iodine treatment of the colonies formed at 23 C revealed sporulating ones for each combination (with frequencies ranging from 18% to 27%) indicating that the second mutation in BCa-312 is not allelic to any known ste genes. Thus, it represents a new class of sterile mutants in which the possibility to switch on sexual processes seems to be completely lost. As this deficiency is accompanied by a release from the patl- growth arrest which is necessary for the shift from proliferation to a sexual pathway, these mutants were called aff (affranchised). Four independently isolated aff- mutants were fused with BCa-312 and the fusants were cultivated on MEA. None of them could conjugate and sporulate. These results indicate that the mutants define a single gene, affl. The allele represented by BCa-312 will hereafter be called aff Mapping of affl-312 One of the sterile spore clones used above, affl-312 leul&, turned out to be of h- mating type because its somatic hybrids with affl + ura4-d18 h- did not sporulate. This finding was exploited for gross mapping of affl. Diploids were constructed between affl-312 leul-3 h- and affl + h strains carrying various markers on each chromosome. The fusants were then cultivated on YEA containing p- fluorophenylalanine to induce haploidization, and the phenotype of the mitotic segregants was determined, aff1-312 showed no linkage with any chromosomal markers (free recombination) except leul and ura5, suggesting that affl is on chromosome II (Table 3). Accurate mapping requires tetrad analysis. This was done by multifactorial fusions with h + or h 9 strains involving markers on chromosome II. The results revealed close linkage of affl with arg4 which is located on the right arm of the chromosome (Table 4). Since no ste, mei or spo genes have been mapped in this region (Kohli 1987), affl is apparently a hitherto unknown gene of the sexual development of S. pombe. aff]-312 does not cause sterility by rendering the cells unable to stop in G1 Nitrogen starvation snychronizes the cells in G1 and induces conjugation and/or meiosis. If the cells lose the ability to stop in Ga after exhaustion of nitrogen from the medium, they will not respond to the starvation by switching on the sexual processes. To test this possibility, affl + and affl cells were cultivated in a nitrogen-limited minimal medium (containing 5 mm NHgC1) and in a nitrogen-rich medium Table 3. Mapping of affl by haploidization List of fusions: cdc ade3-34 h- x aff1-132 leul-3 h- ura5-294 h- x aff1-312 leul-3 h- ura4-d18 h ~ x affl-312 leul-3 h- Segregation of aff1-312: Chromosomal marker Chromosome affl-312 (%) cdel6 I 43 ade3 I 50 ura5 II 18 leul II 76 ura4 III 37 Table 4. Mapping of affl by tetrad analysis List of fusions : ade7-152 ural-171 cdelo-129 h + x affl-312 leul-3 h ade7-50 cde2-35 h 9 x affl-312 leul-3 h- ade6-m210 spol8-b317 h 90 x affl-312 leul 3 h- arg4-55 h - x aff1-312 ura4-d18 h 9 Segregation of affi-312: Gene pair PD T NPD Distance in map units a ade7-affl cdc2-affl leul-affl spo18-affl arg4-affl PD, parental ditype; NPD, non-parental ditype; T, tetratype a Map distance = 50(T + 6NPD)/(PD + NPD + T) (Perkins 1949) (minimal medium supplemented with 0.5% yeast extract and 100 mm NH4C1) to stationary phase (for 72 h). In these media cells accumulate in G1 and G2, respectively (Castello et al. 1986). Then the cells of each culture were transferred into fresh nitrogen-rich media and monitored microscopically, affl cells grown in nitrogen-rich medium resumed cell division with a lag time of 5 h. In afff cultures this period was I h longer. In cultures grown in the nitrogen-limited medium, the affl- cells started dividing somewhat earlier. The lag period was 9 h and 10 h for affland affl, respectively. The good synchrony of division and the 3 h difference between the lag periods suggest that aff 1- cells are also able to arrest both in G0/1 and in G0/2. It is remarkable, however, that the difference between the lag phases is shorter in affl (3 h) than in the wild-type (5 h). The background of this phenomenon is unclear but it might mean that the afff cells arrest in the "late G1 phase". Since neither the patl-ll4 affl&12 cdc nor the pat1-114 affl-312 cdc2-35 strains synchronized by shifting the cells to 35 C (restrictive for the cdc- alleles) could sporulate, the possibility that the sterility of affi-312 is due to a defect in GI arrest can be excluded, cdc2 and cdclo are control genes whose mutations block cells in GI (Nurse and Bissett 1981).

5 533 Discussion The patl-ll4-induced conjugation and haploid sporulation provide a direct way to determine the role of sterility genes in the initiation of sexual development of S. pombe, patl- 114 h 9 haploids harbouring defective alleles of the ste genes are sporulation proficient, indicating that these genes act upstream ofpatl in the initiation of meiosis. However, most of them retain their sterility even on a sporulation medium, which suggests that in the initiation of conjugation most of them exert their function after the patl-controlled event. As for the class I mutants, ste5 and ste6, this finding is consistent with the observation that their homozygous somatic hybrids can sporulate even in a pat + background (Girgsdies 1982; Michael and Gutz 1987). Thus, ste5 and ste6 are conjugation-specific and do not play any role in the meiotic pathway. The phenotype of the class IIb mutants (ste2, ste4, ste7 and ste9) is also partially suppressed by patl-ll4: they undergo haploid sporulation but remain sterile. Presumably, these genes are involved in the post-pat1 part of the conjugation pathway. However, they are not strictly conjugation-specific, because the somatic hybrids homozygous for their defective alleles do not sporulate (Michael and Gutz 1987). The reason for this peculiarity is obscure but might indicate that these ste genes act between patl and the class I ste genes in the conjugation pathway and upstream of patl in the initiation of meiosis. The class IIa genes (stel, ste3 and ste8) whose defective mutants become fertile and sporulation proficient when combined with patl-ll4 must act very early in the sexual development, presumably before the patl-controlled conjugation-specific event. The existence of these confirms the idea that the patl + gene product is a negative regulator of both the conjugation and the meiosis (Beach et al. 1985). The genes belonging to this class might be engaged in the monitoring of the nutritional signal which triggers the sexual development and/or its transmission to the initiation mechanisms of the pathways. The patl-ll4 phenotype can also be exploited to isolate mutants defective in genes acting downstream of patl in the initiation of the sexual development. To date, only one gene, mei2, has been found to act after patl (Iino and Yamamoto 1985a; Beach et al. 1985). affl described in this work seems to be the second one. Their functions are, however, very different, as is apparent from the differences between the phenotypes of their defective alleles, mei2- abol- ishes the growth arrest and meiotic initiation caused by the inactivation of the patl-ll4 gene product, but the cells remain fertile. The affl- cells are also released from the inhibitory effects of patl- but, in addition, they are completely sterile. Elevation of the intracellular 3'5'cAMP concentration to abnormally high levels was previously found to have virtually the same effect (Beach et al. 1985). The camp effect was supposed to be an overstimulation of the camp-dependent protein kinase which is not itself involved in regulation of the sexual processes, but which may replace the patl + gene product (probably a protein kinase as well; see McLeod and Beach 1986) in patl- cells: the campdependent protein kinase might process the substrate of the patl + gene product. The obvious analogy between the affl- phenotype and the consequence of the increased intracellular camp concentration suggests that affl might be involved in camp-related cellular processes. There are numerous ways in which affl- could affect this system. It might increase the cellular camp level, either by stimulating adenylate cyclase activity (like the mutation IAC in Saccharomyces cerevisiae; see Uno et al. 1982) or by inhibiting camp degradation (as in pdel mutants of S. cerevisiae; see Uno et al. 1983). Giving support to this idea is the finding that the elevation of intracellular camp by introduction of the S. cerevisiae gene encoding adenylate cyclase (CYR1) into S. pombe also suppresses the patl phenotype (Beach et al. 1985). However, the CYRl-transformed cells were morphologically altered and highly elongated when grown to stationary phase, which is not the case in affl- mutants. Furthermore, a mutation resulting in hyperactive adenylate cyclase should be dominant over the wild-type allele as in the case of IAC (Uno et al. 1982). Another possibility is that affl- increases the activity of the camp-dependent protein kinase, e.g. by rendering it independent from camp. An analogous mutation, CYR3, is known in S. eerevisiae, but that is also dominant (Uno et al. 1982). Thus, it appears unlikely that affl encodes a regulator of camp-dependent protein kinase. Rather, it might be a gene whose product dephosphorylates certain proteins phosphorylated by protein kinases. In S. cerevisiae the ppdl mutation which causes a phosphoprotein phosphatase deficiency completely suppresses meiosis and sporulation. It is supposed that this mutation exerts its suppressor activity by the accumulation of certain phosphorylated proteins (Matsumoto et al. 1986). It is conceivable that in S. pombe a similar dephosphorylating system may play a regulatory role and affl might be involved in it. If the affl G 1 cell ctcle (n) --aft1 -- potl-~_ t N ste IIa ste lib ste ] conjugation ~ celt cycle (2n) offl,, I ste II. b N ste IIa il meiosis Fig. 1. A hypothetical model of the role of ste and affl genes in the initiation of conjugation and meiosis. The rectangular lines indicate negative regulation. N: nutritional signal (nutritional deprivation) triggering sexual development

6 534 gene product is missing, no dephosphorylation would occur and so even the low-efficiency phosphorylation ofpatl substrates by the normal level of camp-dependent protein kinase might abolish the shift from the vegetative proliferation to a sexual pathway. Summarising the results of this report and the findings published by other authors, I suggest a model for the role of the ste genes and affl in the mechanisms controlling the shift from cell proliferation to conjugation and/or meiosis-sporulation (Fig. 1). In this scheme, the affl gene product participates in a cascade of regulatory events resulting in cell-type specialization, probably as a factor antagonistic to patl. Acknowledgements. I wish to thank P. Nurse and for comments on this work. I also thank P. Fantes,, U. Leupold,, P. Schuchert and M. Yamamoto for supplying strains, and U. Leupold for communication of unpublished results. The technical assistance of Mrs. Lamfalusi and Mrs. Tasnadi is also gratefully acknowledged. References Beach D, Rodgers L, Gould J (1985) RAN1 + controls the transition from mitotic division to meiosis in fission yeast. Curr Genet 10: Castello G, Rodgers L, Beach D (1986) Fission yeast enters the stationary phase GO state from either mitotic G1 or G2. Curr Genet 11 : Egel R (1971) Physiological aspects of conjugation in fission yeast. Planta 98 : Girgsdies D (1982) Sterile mutants of Schizosaccharomyees pombe: Analysis by somatic hybridization. Curr Genet 6: Gutz H, Heslot H, Leupold U, Loprieno N (1974) Schizosaccharomyees pombe. In: King RC (ed) Handbook of genetics. Plenum Press, New York, pp Iino Y, Yamamoto M (1985a) Mutants of Schizosaccharomyces pombe which sporulate in the haploid state. Mol Gen Genet 198: Iino Y, Yamamoto M (1985b) Negative control for the initiation of meiosis in Schizosaccharomyees pombe. Proc Natl Acad Sci USA 82: Kohli J (1987) Genetic nomenclature and gene list of the fission yeast Sehizosaccharomyees pombe. Curr Genet 11 : Kohli J, Hottinger H, Munz P, Strauss A, Thuriaux P (1977) Genetic mapping in Schizosaccharomyces pombe by mitotic and meiotic analysis and induced haploidization. Genetics 87 : Matsumoto K, Uno I, Ishikawa T (1986) Role of cyclic AMP in cell division. In: Hicks J (ed) Yeast cell biology. Liss, New York, pp McLeod M, Beach D (1986) Homology between the ranl + gene of fission yeast and protein kinases. EMBO l 5 : Michael M, Gutz H (1987) Sterility (ste) genes of Schizosaecharomyees pombe. Yeast 3 : 5-9 Nurse P (1985) Mutants of the fission yeast Schizosaccharomyces pombe which alter the shift between cell proliferation and sporulation. Mol Gen Genet 198: Nurse P, Bissett Y (1981) Gene required in G1 for committment to cell cycle and in G2 for control of mitosis in fission yeast. Nature 292 : Perkins DD (1949) Biochemical mutants in the smut fungus Ustilago maydis. Genetics 34: Sipiczki M, Ferenczy L (1977) Protoplast fusion of Sehizosaeeharomyces pombe auxotrophic mutants of identical mating type. Mol Gen Genet 151:77-81 Sipiczki M, Kucsera J (1983) A possible control mechanism over conjugation and meiotic division in diploid cells of the fission yeast Schizosaecharomyces pombe var. malidevorans. In: Chaloupka J, Kotyk A, Streiblovfi E (eds) Progress in cell cycle controls. 6th European Cell Cycle Workshop, Czechoslovak Academy of Sciences, Prague, p 191 Sipiczki M, Kucsera J, Ulaszewski S, Zsolt J (1982) Hybridization studies by crossing and protoplast fusion within the genus Sehizosaecharomyces Lindner. J Gen Microbiol 128 : Sipiczki M, Heyer W-D, Kohli J (1985) Preparation and regeneration of protoplasts and spheroplasts for fusion and transformation of Sehizosaeeharomyces pombe. Curt Microbiol 12: Thuriaux P, Sipiczki M, Fantes P (1980) Genetical analysis of a sterile mutant by protoplast fusion in the yeast Schizosaccharomyces pombe. J Gen Microbiol 116: Uno I, Matsumoto K, Ishikawa T (1983) Characterization of a cyclic nucleotide phosphodiesterase-deficient mutant in yeast. J Biol Chem 258: Uno I, Matsumoto K, Ishikawa T (1982) Characterization of cyclic AMP-requiring yeast mutants altered in the regulatory subunit of protein kinase. J Biol Chem 257 : Communicated by W. Gajewski Received March 26, 1988