Our studies of the yeast Saccharomyces cerevisiae focus on the three cell types: the two haploid cell types (a and @) and the diploid cell type formed by mating (the a/@ cell). These cell types differ from each other in several respects, notably in the production of and response to these pheromones. Several facets of these cell types were studied to address a variety of problems in cell biology and gene control, including the pathway by which cells respond to the pheromones, identification of determinants of cell polarity and morphogenesis (choice of sites for budding and mating), control of sporulation, and molecular mechanisms for controlling gene expression. The corn smut fungus, Ustilago maydis, a dimorphic organism which infects corn, also was studied. The goal was to understand the mode of action of two regulatory loci (a and b) and other genes that determine the ability of cells to progress through the life cycle; that is, to mate, form filaments, induce tumors, and produce spores.


@-factor is a pheromone of 13 amino acids that is secreted by a cells and used in communication between the two mating partners during the mating process. @-factor elicits three main responses in a cells: transcriptional induction of genes involved in mating, arrest in the G1 phase of the cell cycle, and morphological changes causing the cell to become a pear-shaped shmoo. We are studying all of these aspects.

The response pathway begins with a serpentine receptor which communicates with a heterotrimeric G protein. Activation of the receptor causes release of the Gbg subunit, which then activates a MAP kinase cascade by an unknown mechanism. The STE20 gene codes for a putative protein kinase that is thought to be involved in activating the protein kinase, STE11. Aaron Neiman showed a couple of years ago that STE20 can phosphorylate STE11 in vitro, although the functional significance of this is not known. Matthias Peter (with input from Aaron, Maarten van Lohuizen, and Hay-Oak Park) provided some important new information on STE20. Because a mammalian homologue of STE20, p65-PAK, is activated by a human homologue of CDC42 (Manser et al., 1994), Matthias asked whether CDC42 is necessary for the pheromone response pathway and for action of STE20. He first showed that CDC42 does indeed bind to STE20 in vitro and that mutants of STE20 that lack the CDC42 binding site are defective in mating and signal transduction. Depleting the cells of CDC42 also blocks the pheromone response pathway. These and other findings reveal an unexpected role for CDC42: although discovered as an essential gene required for bud formation, we now see that it also plays an essential role in the pheromone response pathway. Matthias then tried determine how CDC42 controls STE20 activity. One possibility is that CDC42 localizes STE20 (analogous to the localization of Raf by Ras).

When yeast cells mate with each other, they reorient their cytoskeleton: actin and other components of the cytoskeleton are oriented towards the site of cell contact, presumably in order to facilitate cell fusion. In addition, the spindle pole bodies face each other as a prelude to nuclear fusion. Janet Chenevert carried out a mutant hunt to identify genes that may be involved in the reorientation process by isolating mutants that mate poorly to strains that are enfeebled for mating (Chenevert et al., 1994). These mutants are normal for cell cycle arrest and gene induction in response to mating pheromones. One class of mutants exhibits altered morphology and is described in last year's report and in Chenevert et al. (1994).

Wild-type cells use a gradient of mating pheromone to locate their mate spatially, much as higher eukaryotes use gradients for chemotaxis. This gradient apparently leads cells to polarize towards their mating partner. We have sought to identify the molecular mechanism by which cells choose the site for mating. Nicole Valtz studied a class of mutants from the Chenevert hunt which exhibits normal morphological response to mating factors and sought to test whether these mutants are defective in orienting towards their mating partner. Nicole has used two assays to determine if any of these mutants are defective in locating their mating partner. The first (the pheromone confusion assay) determines the ability of cells to mate in the presence of high levels of exogenously added @-factor. The second is a direct microscope assay for polarization towards a gradient of @-factor. These studies reveal that four mutants are unable to orient towards a source of mating pheromone: all carry mutations in the FAR1 gene, which was identified because it is necessary for cell cycle arrest. The FAR1 protein appears to carry out two functions: cell cycle arrest (mediated by the amino terminal segment) and mating-site selection (which requires the carboxy terminal 2/3 of the protein and at least some of the amino terminus). Matthias Peter and Nicole characterized point mutations affecting the mating activity of FAR1 and shown that three affect the carboxy terminus and one the amino terminus. These studies show that reorientation in response to mating factors is required for mating. Nicole and Matthias worked on determining the role of FAR1 in selection of the mating site.

Jennifer Philips studied a later step in mating, about which essentially nothing is known: cell fusion. Once cells have chosen a mating partner, several fusion events must occur to produce a zygote. First, cell fusion occurs, followed by fusion of organelles such as mitochondria and the endoplasmic reticulum, and then nuclear fusion. Cell fusion requires regulated degradation of cell wall material in the region of contact between two mating cells and subsequent plasma membrane fusion. She examined another class of mutants identified in the Chenevert screen which may be defective in cell fusion. The products of previously identified genes (FUS1, FUS2, and FUS3) are required for efficient cell wall degradation during mating. Cell fusion normally occurs rapidly, but mutants defective in these genes result in accumulation of "prezygotes," structures containing two adherent cells that have not fully degraded the cell wall or completed later events such as plasma membrane or nuclear fusion. Several mutants were identified which accumulate prezygotes that are not defective in FUS1 or FUS2. These mutants contain defects in single genes and are of two classes: specific to a cells or affecting both a and @ cells. She is currently cloning these genes and shall determine their role in cell fusion.

[page top]


Budding yeast grows vegetatively by a specialized pattern of oriented cell division. Yeast cells choose budding sites on their cell surface in a genetically programmed manner that depends on their cell type. a and @mother cells choose a bud site immediately adjacent to their last daughter, and daughter cells bud towards their mothers (the axial pattern). In a/a cells, mother cells choose bud sites adjacent to their daughter or at the opposite end, whereas daughter cells choose a bud site opposite the mother (the bipolar pattern). We have identified five genes required for proper bud site selection (BUD1-BUD5; Chant & Herskowitz, 1991; Chant et al., 1991) and are seeking to determine their roles in this process.

The BUD1, BUD2, and BUD5 proteins are required for bud site selection in all cell types. BUD1 is identical to RSR1, which encodes a protein with strong similarity to human H-ras protein. BUD5 encodes a GDP-GTP exchange factor for BUD1, converting it from the GDP- to the GTP-bound state. BUD2 is a GTPase-activating protein (GAP) for BUD1 (Park et al., 1993). These studies lead to the hypothesis that BUD1, BUD2, and BUD5 function as a general bud site selection machine. We have previously hypothesized that they function by recruiting other proteins necessary for cell polarity and bud formation to the bud site. Hay-Oak Park tested this hypothesis by direct biochemical analyses using proteins produced in E. coli. She showed that CDC24 binds to the GTP-bound form of BUD1, and BEM1 binds to the GDP-bound form of BUD1. Hay-Oak furthermore found that the GDP-bound form of CDC42 binds to the GTP-bound form of BUD1. These findings indicate that bud-site selection involves a GTPase cascade, involving the RAS relative BUD1 and the rho relative CDC42, and support the view that BUD1 is responsible for localizing cell polarity proteins.

Mutations in the BUD3 and BUD4 genes (as well as in the AXL1 gene identified by Fujita et al., 1994) cause cells to exhibit the bipolar budding pattern, which led to the hypothesis that these products are important for specifying the axial budding pattern. Sylvia Sanders cloned and sequenced BUD4 and found that it codes for a large novel protein with a potential GTP-binding motif at its carboxy terminus. As has been shown for BUD3 (Chant et al., 1995), BUD4 is produced as a ring in all cell types and is localized to the mother/bud neck. Most cells with large buds contain two distinct rings of BUD4 immunoreactive staining; approximately half of the unbudded cells contain a single ring; cells with a small bud lack localized BUD4. Analysis of synchronized cell populations indicates that unbudded cells first contain a BUD4 ring which then disappears. BUD4 protein is likewise absent from yeast cells at this time in the cell cycle. Chant et al. (1995) showed that BUD3 colocalizes with the neck fibers at the mother/bud junction; Sylvia found that maintenance of the BUD4 rings (as known for the BUD3 rings; Chant et al., 1995) requires integrity of the neck fibers. These observations suggest that BUD3 and BUD4 cooperate to recognize a spatial landmark in a and @cells, presumably the neck fibers. BUD3/BUD4 is then proposed to direct the site for assembly of the neck fibers in the next cell cycle.

In fission yeast, control of entry into mitosis is controlled by the phosphorylation state of the mitotic form of the cyclin- dependent kinase, Cdc2-cyclin B, which can be phosphorylated on tyrosine 15. The protein kinases, WEE1 and MIK1, and the phosphatase, CDC25, are responsible for phosphorylation and dephosphorylation of this residue. Although budding yeast has similar enzymes, the physiological role of these enzymes is unknown. Maarten van Lohuizen and Matthias Peter sought to determine the role of the WEE1 homologue, SWE1, in budding yeast by identifying synthetically lethal mutations of yeast which cause lethality in a swe1 mutant. Eleven mutants were obtained, of which six have been studied in some detail. They affect three different genes: BEM1 and BEM2, which are necessary for establishment of cell polarity and bud formation, and GAS1, which is a major GPI-linked cell surface glycoprotein of unknown function. Mutants defective in SWE1 and in any of these genes arrest as large unbudded cells with multiple nuclei. In contrast, mutants defective only in BEM1, BEM2, or GAS1 exhibit a delay in G2 when grown at 24°C but are uninuclear. Mutants defective in these genes that lack the SWE1 phosphorylation site on CDC28 behave like the SWE1-defective mutants: they are multinuclear. These results lead to the hypothesis that cells which encounter a problem in bud growth delay further cell cycle progression (in particular, nuclear division) until bud emergence has been restored. Activation of SWE1 (and thus inhibition of the CDC28 kinase) is responsible for this delay. This morphogenetic checkpoint was independently identified by Lew & Reed (1995). Maarten and Matthias continued working to understand the signal transduction pathway responsible for triggering the transient cell cycle arrest.

[page top]


Bruce Cree's goal was to identify genes that function around the time of the G1/S transition -- Start. He isolated several ts mutants with arrest points at this phase of the cell cycle and shown that one of these is the independently identified gene PTC1, which codes for a protein serine- threonine phosphatase. Bruce's original mutant, carrying his ptc1-1 mutation, arrests at high temperature with a high fraction of cells having two or more buds, indicating a defect in cytokinesis. Because the mutant cells are uninucleate and because the ptc1-1 mutation is partially suppressed by high copy plasmids carrying the G1 cyclin gene, CLN2, these mutants are apparently defective in progression from G1 to S. Preliminary studies indicate that the level of CLN2 protein is decreased in ptc1 mutants. The nature of the requirement of PTC1 for high- level synthesis of CLN2 is under study. Work by others suggests that PTC1 negatively regulates the HOG pathway (Maeda et al., 1994). In addition, Bruce made several observations that lead him to propose that PTC1 also negatively regulates the PKC1 pathway.


Joe Gray was interested in the molecular mechanisms of second-messenger signaling systems and their roles in controlling growth of the eukaryotic cell. As the result of a fiendishly clever, albeit circuitous, screen for mutants that may be defective in calcium signaling, Joe became enmeshed and fascinated with the protein kinase C (PKC) pathway. The in vivo role of the PKC MAP kinase pathway is not clear, but one function appears to be for integrity of the cell surface. In studies of mutants defective in the PKC pathway, Joe noted a variety of novel vegetative defects displayed by these mutants and a little characterized gene, HCS77, identified by Joe Ogas (Ogas et al., 1991) as a dose-dependent suppressor of mutants defective in the G1/S transition. Mutants defective in HCS77, like mutants defective in the PKC1 MAP kinase, exhibit temperature-dependent lysis. HCS77 encodes a putative type I transmembrane protein that may be localized to the cell surface (J. Ogas, Ph.D. thesis). Joe worked on testing the possibility that HCS77 functions upstream of PKC1 in this pathway and as a sensor of the state of the cell surface.

In other studies, Joe showed that yeast PKC, as has long been suspected for its mammalian relatives, is indeed an activator of proliferation in yeast. He found that the PKC pathway stimulates the G1/S transition, apparently by enhancing the activity of the G1 cyclins. Our working model is that the PKC pathway couples some cell sur-face signal (perhaps cell size) sensed by HCS77 to the decision to proliferate.

[page top]


During meiosis, diploid cells produce haploid gametes by undergoing two rounds of cell division preceded by a single round of DNA synthesis. In yeast, meiosis is coupled to ascospore development in a process known as sporulation. The four products of a single meiosis (a tetrad) are neatly packaged into an ascus.

Megan Grether studied the molecular mechanisms which permit two consecutive cell divisions to occur in the absence of an intervening S phase. Genes have been identified in which loss-of- function mutations causesporulating cells to form dyads which result from a single meiotic division. Such genes, which include SPO12 and SPO13, are excellent candidates for regulators of the meiotic cell cycle. Megan undertook a visual screen to identify other dyad-forming mutants, with plans to characterize the function of the SPO13 protein in hopes of establishing its role during meiosis

Shelley Chu was interested in identifying other components required for the process of cell differentiation during sporulation. These include genes whose products have functions either unique to the meiotic (rather than to the vegetative) cell cycle or are involved in ascospore development. Identification of recessive mutations in a process exhibited by diploid cells poses certain problems that must be circumvented. Towards these ends, Shelley used a haploid SPO13-deficient strain which, upon sporulation, forms viable dyads. Shelley planned to screen visually for mutants which can no longer sporulate in order to identify new sporulation-specific genes.

[page top]


Anita Sil studied lineage-specific regulation of mating- type interconversion. Mating-type interconversion is an asymmetrically regulated event: haploid cells that have experienced previous cell divisions (mothers) are able to switch mating type whereas their genetically identical, newly-born progeny (daughters) cannot. After daughter cells traverse one cell cycle, they become mother cells and are then competent to initiate the switching process. The basis of this asymmetric cell fate is asymmetric gene expression (Nasmyth, 1983), but ectopic expression of HO in daughter cells is sufficient to allow mating-type switching (Jensen & Herskowitz, 1984; Nasmyth, 1987).

The mechanism that prevents daughter cells from transcribing HO is unknown. Anita took a genetic approach to identify regulators of the asymmetry of HO expression. Using a novel microcolony screen, she isolated a mutant strain whose daughters switch mating-type at high frequency. These mutant daughters are able to transcribe the HO gene, indicating that HO transcription is no longer asymmetrically regulated. Mating-type switching in the mutant is dependent on SWI5, a known activator of HO transcription that is thought to be limiting in amount or activity in wild-type daughter cells. Anita cloned the gene that is defective in this mutant and will test the possibility that it encodes an inhibitor of HO transcription that is distributed asymmetrically to the daughter cell at division.

[page top]


Ustilago maydis is a dimorphic Basidiomycete fungus that infects maize, inducing tumors. The two forms that characterize its life cycle are a unicellular, haploid form that divides by budding to produce compact colonies on a variety of media and is nonpathogenic. The second form is the filamentous dikaryon, which results from fusion of two haploid cells. Two mating-type loci, a and b, govern dimorphism and life-cycle transitions. The a locus has two alleles, a1 and a2, which encode pheromones and receptors (reviewed in Banuett, 1992, 1995). This locus regulates cell fusion and filamentous growth but is not required for pathogenicity, fungal differentiation within the plant, or for meiosis. The b locus is multiallelic (25 alleles): any combination of different b alleles triggers filamentous growth and pathogenicity. Self- nonself recognition is now known to occur at the level of polypeptides: only polypeptides encoded by different b alleles are able to interact to form the active regulatory protein (Kämper et al., 1995).

Proliferation of the filamentous dikaryon in the plant leads to tumor formation. The fungus then differentiates within the tumor, a process that entails a number of morphological changes resulting in production of rounded cells in which karyogamy takes place and around which a specialized cell wall is deposited to form a teliospore. Flora Banuett identified genes (fuz genes) that alter filament formation on charcoal media. fuz1 mutants induce tumors but are unable to differentiate, and thus no teliospores are produced. This observation indicates that the fuz1 gene is necessary for differentiation in planta. The fuz1 gene is also necessary for normal cell and colony morphology. Its sequence has provided no clues to its functional role. The fuz7 gene encodes a MAP kinase activator that is required for the pheromone response pathway involved in conjugation and filamentous growth (Banuett & Herskowitz, 1994a). fuz7 is also required for tumor induction. Because tumor induction is independent of the a locus, Flora proposed that the Fuz7 kinase is activated in response to a plant signal. Other fuz genes are currently being characterized.

Flora and Wei Wei isolated a variety of temperature-sensitive mutants that exhibit altered cell and colony morphology. Analysis of such mutants (jacks, ginseng, beaded, etc.) may provide insights into mechanisms that determine morphology of this fungus.

Lab Home  |  People  |  Research Summary  |  Bibliography  |  Links  |  Protocols | page top