A Novel Suppressor of rasl in Fission Yeast , byr 4 , Is a Dosage - dependent Inhibitor of Cytokinesis

A novel gene, designated byr4, was identified in Schizosaccharomyces pombe that affects the mitotic cell cycle and shows genetic interactions with the rasl signaling pathways. Null alleles of byr4 cause cell cycle arrest in late mitosis and permit multiple rounds of septation. The multiple septa typically divide two nuclei, but the nuclei frequently do not stain equally with 4',6-diamidino-2-phenylindole (DAPI), suggesting that byr4 is required for proper karyokinesis. Overexpression of byr4 inhibits cytokinesis, but cell cycle progression continues leading to multinucleate cells. When byr4 is overexpressed, the early steps in the cytokinesis pathway, including formation of the medial F-actin ring, occur normally; however, the later steps in the pathway, including contraction of the F-actin ring, septation, and rearrangement of the medial F-actin following mitosis, rarely occur, byr4 shows two genetic interactions with rasl. The inhibition of cytokinesis by byr4 overexpression was exacerbated by null alleles of rasl and scdl, suggesting a link between pathways needed for cell polarity and cytokinesis. Overexpression of byr4 also partially bypasses the need for rasl for sporulation. The electrophoretic mobility of the byr4 protein varied in response to mutants that perturb cytokinesis and karyokinesis, suggesting interactions between byr4 and these gene products. A more rapidly migrating byr4 protein was found in cells with mutations in cdc16, which undergo repeated septation, and in cdcl5, which fail to form a medial F-actin ring in mitosis. A slower migrating byr4 protein was found in cells with a mutation in the 13-tubulin gene, which arrests cells at the metaphase-anaphase transition. T Hc~ ras proteins are GTPases that cycle between an ive, GTP-bound form and an inactive, GDP. I t . bound form (Boguski and McCormick, 1993). Depending on the cellular context, activated ras can stimulate the cell division cycle, alter cell shape, or cause cellular differentiation (Bourne et al., 1990). Several pathways are implicated in signaling downstream of mammalian ras proteins. The best characterized pathway activates the raf kinase. Activated ras recruits the raf kinase and activated raf, in turn, activates a mitogen-activated protein (MAP) 1 kinase cascade (Herskowitz, 1995; Marshall, 1995). A second ras effector may be phosphatidylinositol(PI)-3 kinase. Activated ras binds to PI-3 kinase and ras binding stimulates PI-3 kinase activity four-fold in vitro (RodriguezVinciana et al., 1994). An activated allele of PI-3 kinase, however, activates ras and requires ras for signaling, suggesting ras is an effector of PI-3 kinase (Hu et al., 1995). A third ras effector may be ralGDS, a positive regulator of the ral GTPase (Hofer et al., 1994; Kikuchi et al., 1994; Spaargaren and Bischoff, 1994). Activated ral GTPase Address all correspondence to Charles F. Albright, Department of Biochemistry, Vanderbilt University School of Medicine, Nashville, TN 37232-0146. Tel.: (615) 343-2174; Fax: (615) 343-0704. 1. Abbrevia t ions used in this paper: DAPI, 4',6-diamidino-2-phenylindole; MAP, mitogen-activated protein; MM, minimal media; YE, yeast extract. binds a protein with GTPase-activating protein (GAP) activity for the cdc42 GTPase, suggesting that this pathway may regulate the actin cytoskeleton using a GTPase cascade (Cantor et al., 1995). We are studying ras signal transduction pathways in the fission yeast Schizosaccharomyces pombe because the accumulated data suggest that these pathways are similar to those in mammals and because this system is amenable to genetics (Gotoh et al., 1993; Hughes et al., 1993; Xu et al., 1994). S. pombe are rod-shaped yeast that usually exist as haploids (Gutz et al., 1974). S. pombe has a single ras gene, rasl, that is necessary for cellular differentiation in response to mating pheromone or nutrients and helps maintain the normal rod shape, rasl is not an essential gene, but null alleles of rasl eliminate conjugation, greatly reduce sporulation, and change cells to round shaped (Fukui et al., 1986; Nadin-Davis et al., 1986a,b). rasl protein signals through at least two pathways. In one pathway, rasl binds to and presumably activates the byr2 kinase (van Aelst et al., 1993; Masuda et al., 1995). The byr2 kinase is part of a MAP kinase cascade that includes byrl and spkl (Nadin-Davis and Nasim, 1988; Toda et al., 1991; Wang et al., 1991). Null alleles of byr2, byrl, and spkl eliminate conjugation and sporulation without affecting cell viability or cell shape. Hence, as in mammals, ras in fission yeast signals through a MAP kinase cascade. In a second path© The Rockefeller University Press, 0021-9525/96/06/1307/13 $2.00 The Journal of Cell Biology, Volume 133, Number 6, June 1996 1307-1319 1307 on Jauary 7, 2018 jcb.rress.org D ow nladed fom way, rasl binds and presumably activates the scdl protein (Chang et al., 1994). scdl has sequence similarity to GDP dissociation stimulators for the cdc42 GTPase. Null alleles of scdl eliminate conjugation and change the cell shape to round, but do not affect sporulation rates or cell viability (Chang et al., 1994; Fukui and Yamamoto, 1988). The ras1-scd1-cdc42 pathway is very similar to the RSR1CDC24-CDC42 signaling pathway used to control cell polarity in Saccharomyces cerevisiae (Chant, 1994). To identify new components in the rasl signaling pathways, we looked for genes that when overexpressed could suppress the conjugation defects of strains with a rasl null allele. Previous attempts to isolate multicopy suppressors of rasl null alleles isolated the byrl, byr2, and byr3 genes (Nadin-Davis and Nasim, 1988; Wang et al., 1991; Xu et al., 1992). These genes partially bypass the need for rasl in sporulation, but do not affect cell shape and only weakly affect the conjugation rates of raslstrains (Xu et al., 1992). We reasoned that rasl might signal through two pathways to stimulate conjugation and we might, therefore, find multicopy suppressors of the raslconjugation defect if the MAP kinase cascade was constitutively activated. One gene found in this screen was byr4. This study shows that byr4 plays a critical role in the control of cytokinesis, in addition to showing genetic interactions with rasl. Fission yeast is a good model system to study the control of the eukaryotic cell division and cytokinesis. A number of S. pombe cell cycle mutants were identified whose terminal phenotype suggest they participate in the control of septum formation and cytokinesis. The "early septation" genes include cdc7, cdcll, cdc14, and cdc15 (Fankhauser and Simanis, 1994b). Cells with mutations in these genes arrest as highly elongated cells with multiple nuclei, since growth, DNA replication, and mitosis continue in the absence of cytokinesis (Nurse et al., 1976). The "late septation" genes include cdc3, cdc4, cdc8, and cdc12 (Fankhauser and Simanis, 1994b). Cells with mutations in these genes arrest with two to four nuclei and contain disorganized septal material (Nurse et al., 1976). A third class of mutants, defined only by cdc16 mutations, undergo repeated rounds of septation without cell cycle progression, leading to cells with two nuclei separated by multiple septa (Minet et al., 1979). Genetic studies suggest the gene products of cdc7, cdc11, cdc14, and cdc16 may interact (Marks et al., 1992). Since the phenotypes resulting from mutations in these genes have similarities to the phenotypes from mutations in the byr4 gene, byr4 may be part of this signaling pathway. Materials and Methods Strains and Growth Conditions The strains used in this study are listed in Table I. Yeast were grown in yeast extract (YE) or minimal media (MM) with required supplements at the levels of 75 mg/liter for adenine, uracil, leucine, and 0.4 mM thiamine (Moreno et al., 1991). Derivatives of MM media were used, as indicated, where ammonia was omitted (MM-N), or 10 mM glutamate was substituted for 100 mM ammonia (MM-glu). CA5 was derived from SP870 by transforming SP870 with a DNA fragment from the HpaI site upstream of the rasl coding sequence to the BsmI site at the 3' end of the rasl gene where the DNA sequence coding for amino acids 29 to 155 was replaced with the ura4 gene (Nadin-Davis et al., 1986a). Stable ura + transformants were isolated and the authenticity of the rasl disruption was tested by Southern analysis (Sambrook et al., 1989). CA28 and CA29 were isolated by tetrad dissection from the diploid CA5/SPSU. CA76 was made by transforming SPSU with a linear fragment containing the entire coding sequence of the byr2 gene, where serines 402 and 404 were changed to alanines (Wang et al., 1991). After overnight growth in 10 ml of YE media supplemented with adenine and uracil, 108 cells were plated on 10-cm YE plates with adenine, uracil, and 1 mg/ml 5-fluoroorotic acid (Sigma Chemical Co., St. Louis, MO) (Grimm et al., 1988). Uracolonies were screened by Southern analysis to test the authenticity of the gene replacement (Sambrook et al., 1989; Moreno et al., 1991). CA78 and CAl15 were made by tetrad dissection of the diploids CA76/CA28 and CA29/SPSCD1U, respectively. CA103 is described below. Isolation of the byr4 Gene Strain CA78 was transformed with a S. pombe genomic DNA library in pWH5 (P. Young, Queens University, Kingston, Kingston, Ontario, Canada). The transformants were plated on MM-glu plates at a density of about 5,000 colonies per 15-cm plate. The colonies were stained with iodine vapors for 3 min and those with positive staining reaction were streaked onto fresh MM-glu plates (Moreno et al., 1991). Approximately 100,000 transformants were screened. Positive clones were rescreened twice and only clones in which all of the colonies during each rescreen had a positive reaction were retained. The majority of the colonies from the primary screen arise from spontaneous diploids. Diploids containing byr2$402/4A sporulated at high levels in the absence of rasl, giving a positive iodine staining reaction. These false positives could be largely eliminated by the repeated screening of the primary isolates. False positives (independent of the plasmid) yielded haploid colonies that were iodine-negative while true positives (dependent on the plasmid) would contain only iodine-positive colonies. As a final screen, plasmids were rescued in E. coli, transformed into CA78, and tested again. Molecular Analysis of byr4 The genomic DNA insert contained in the plasmid isolated from the library, pS17, was subcloned using standard techniques (Sambrook et al., 1989). A deletion of a 2.7-kb PstI fragment yielding pS17APst retained full activity and was used as a starting point for subsequent work. A vector containing a 2-kb fragment starting at the remaining BamHI site retained partial suppressor activity. This fragment and I kb of adjacent DNA were sequenced revealing a 1995 base open reading frame that spanned the BamHI site. Database searches were done using the BLAST program (Altschul et al., 1990). A S. pombe cDNA library (J. Fikes, Massachusetts Institute of Technology, Cambridge, MA) was screened with a 2116-bp fragment (derived by PCR amplification) that encoded the entire open reading frame of the predicted byr4 protein (Becker et al., 1991). A partial cDNA clone was isolated and the ends of this clone were sequenced to establish the exact location of the cDNA relative to the genomic sequence. A series of PCR reactions was performed using the cDNA and genomic DNA as templates. The size of these PCR products revealed no difference in size between the cDNA and genomic products, suggesting that there are no introns in this region of the byr4 gene. To detect the byr4 mRNA in S. pombe, cells from SP870 strain were grown in MM media to a density of 5 × 106 cells/ml and harvested or shifted into MM-N for 6 h before harvesting (Watanabe and Yamamoto, 1994). 15 Ixg of total RNA prepared from these cells was electrophoresed in a 0.8% agarose gel containing 7.4% formaldehyde (EM Science, Gibbstown, NJ) and transferred to a nylon membrane (Hybond; Amersham Corp., Arlington Heights, IL) (Sambrook et al., 1989). This Northern blot was probed with the same fragment used to screen the cDNA library. Assay for Bypass of ras l in Sporulation Plasmids were transformed into CA5/CA7, a rasldiploid, and transformants were streaked on MM-glu plates (Moreno et al., 1991). After 4-5 d of growth at 29°C, single colonies were pooled in 300 t~l of water and the fraction of cells that sporulated was determined by counting. At least 1,500 cells were counted for each diploid using independent transformants and samples from different days. Little variation between transformants or days was observed. The Journal of Cell Biology, Volume 133, 1996 1308 on Jauary 7, 2018 jcb.rress.org D ow nladed fom Table L S. pombe Strains Used in This Study Strain Genotype Source SP870 h 9° leul-32 ura4-dl8 ade6-210 M. Wigler SPSU h 9° leul-32 ura4-dl8 ade6-210 byr2::ura4 M. Wigler SPSCD1U h 9° leul-32 ura4-d18 ade6-210 scdl::ura4 M. Wigler CA5 h 9° leu1-32 ura4-dl 8 ade6-216 ras l : :ura4 This study CA7 h 9° leul-32 ura4-d18 ade6-210 rasl : :ura4 This study CA21 h 9° leul-32 ura4-dl 8 ade6-216 This study CA28 h 9° leu l 3 2 ura4-d18 ade6216 by r2 : : ura4 ras l : : ura4 This study CA29 h 9° leu1-32 ura4-d18 ade6-216 byr2::ura4 This study CA76 h 9° leu1-32 ura4-dl8 ade6-210 byr2-S402/4A This study CA78 h 9° leu1-32 ura4-dl8 ade6-216 byr2-S402/4A rasl::ura4 This study CA103 h 9° leu1-32 ura4-dl8 ade6-210/ This study h 9° leu1-32 ura4-d18 ade6-216 byr4::ura4 KGY444 h + leu1-32 ura4-dl8 ade6-210 cdc25-22 K. Gould CA91 hura4-d18 ade6-216 nda3-KM311 This study CA117 hleu1-32 ade6-210 cdc16-116 This study CA110 h 9° leul-32 ura4-dl8 ade6-216 scdl::ura4 This study KGY247 h + leu1-32 ura4-d18 ade6-210 K. Gould CA115 h 9° leul-32 ura4-d18 ade6-216 byr2: :ura4 scdl : :ura4 This study KGY433 hcdc3-124 leu1-32 K. Gould KGY439 hcdc4-8 ura4-dl 8 leul-32 K. Gould MBY105 h ÷ cdc7-24 ade6-210 ura4-d l 8 M. Balasubramanian MBY106 h ÷ cdc11-123 ade6-210 M. Balasubramanian MBY97 h ÷ cdc14-118 ade6210 leu1-32 ura4-d l 8 M. Balasubramanian MBY 107 h ÷ cdc15-136 ade6-210 leu1-32 M. Balasubramanian KGY657 h ÷ cdc8-110 leul-32 ura4-d18 K. Gould MBY81 cdc12-112 mei l-102 lys l-131 M. Balasubramanian Construction of byr4 Null Allele To construct a null allele of byr4, a BclI site was added at the stop codon of pS17APst by PCR mutagenesis. The coding sequence and 426 bp of 5' untranslated region were then removed by deleting the 2.4-kb BclI fragment. The resulting 2.4-kb HindIII fragment from this vector was cloned into pBSK + (Stratagene, La Jolla, CA) and the ura4 gene containing BamHI ends was inserted into the BclI site (Grimm et al., 1988; LeesMiller et al., 1992). The 4.1-kb HindIII to SmaI fragment from this vector, where the entire coding region of byr4 was replaced with the ura4 gene, was transformed into the diploid strain CA21/SP870. Stable ura ÷ colonies were isolated and screened by Southern blot analysis to identify those strains in which one copy of the byr4 gene was disrupted. One such transformant was designated as CA103. Tetrads from CA103 were dissected to determine the effect of deletion of byr4. For germination of spores in liquid culture, CA103 was grown to midexponential phase in YE and then transferred to MM sporulation media (MM-glu with 1% glucose) (Fankhauser et al., 1993; Hagan and Hyams, 1988). After growth for 48 h in MM sporulation media, vegetative cells were lysed by treatment with 0.5% (vol/vol) glusulase (DuPont, Wilmington, DE) for 24 h at 29°C. The isolated spores were incubated at 3 × 106 cells/ml in MM supplemented with adenine and leucine but without uracil, so that only those spores which carry the byr4::ura4 allele could germinate. After incubation for 24 h at 29°C, cells were fixed and observed microscopically as described below. To isolate haploids containing the byr4 : :ura4 allele, the CA103 diploid strain was transformed with pbyr4/REP41 or pbyr4-409/REP41 and transformants were selected and maintained in MM media with thiamine (Basi et al., 1993). The resulting strain was allowed to sporulate and purified spores were prepared as described before (Moreno et al., 1991). The resulting spore preparation was germinated on plates containing adenine and thiamine but without uracil and leucine. Under these conditions, only haploids containing both the byr4::ura4 allele and pbyr4/REP41 plasmid can germinate and grow. To determine the effect of the byr4 null allele, this strain was grown in MM media with leucine, adenine, and thiamine for 16 h. The growth with leucine supplementation allows cells to lose the complementing plasmid and dilute the endogenous byr4 protein, resulting in some fraction of the cells exhibiting the null phenotype.