Cellular Differentiation in Response to Nutrient Availability: The Repressor of Meiosis, Rme1p, Positively Regulates Invasive Growth in Saccharomyces cerevisiae

Corresponding author: Florian F. Bauer, Department of Viticulture and Oenology, University of Stellenbosch, Matieland 7602, South Africa., fb2{at}sun.ac.za (E-mail)

Communicating editor: A. MITCHELL

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

In the yeast Saccharomyces cerevisiae, the transition from a nutrient-rich to a nutrient-limited growth medium typically leads to the implementation of a cellular adaptation program that results in invasive growth and/or the formation of pseudohyphae. Complete depletion of essential nutrients, on the other hand, leads either to entry into a nonbudding, metabolically quiescent state referred to as G₀ in haploid strains or to meiosis and sporulation in diploids. Entry into meiosis is repressed by the transcriptional regulator Rme1p, a zinc-finger-containing DNA-binding protein. In this article, we show that Rme1p positively regulates invasive growth and starch metabolism in both haploid and diploid strains by directly modifying the transcription of the FLO11 (also known as MUC1) and STA2 genes, which encode a cell wall-associated protein essential for invasive growth and a starch-degrading glucoamylase, respectively. Genetic evidence suggests that Rme1p functions independently of identified signaling modules that regulate invasive growth and of other transcription factors that regulate FLO11 and that the activation of FLO11 is dependent on the presence of a promoter sequence that shows significant homology to identified Rme1p response elements (RREs). The data suggest that Rme1p functions as a central switch between different cellular differentiation pathways.

IN many unicellular organisms, nutrient-rich environments support the rapid growth and multiplication of single cells, leading to an exponential increase in cell numbers. When essential nutrients become limiting or cannot be efficiently utilized, growth rate is reduced, and organisms use specific strategies to adapt to the changed environment. In some nonmotile species, in particular in numerous species of yeast, including Saccharomyces cerevisiae and Candida albicans, reduced availability of nitrogen and carbon sources may initiate a morphological differentiation process that is characterized by a dimorphic switch from an ovoid to an elongated cell shape. Cells stay attached to each other after budding, forming hyphae-like structures in a process that is also referred to as pseudohyphal differentiation. Under the same conditions, cells may also grow invasively into the growth substrate, a phenotype referred to as "invasive growth" (MADHANI and FINK 1998 ; BAUER and PRETORIUS 2001 ; GANCEDO 2001 ; GAGIANO et al. 2002 ). It has been suggested that these adaptations allow yeast cells to grow toward or into nutrient-rich environments (GIMENO et al. 1992 ).

While the shift from a rich to a limited supply of nutrients may lead to a change in growth patterns, a complete depletion of any of several essential nutrients may lead to a different set of adaptations. In haploid yeast, cells arrest in the G₁ phase of the cell cycle and enter a quiescent phase referred to as G₀. Diploid yeast strains, on the other hand, can initiate meiosis to form ascospores (KRON and GOW 1995 ). Meiosis is favored by the absence of nitrogen and, in addition, requires the absence of glucose and the presence of a nonfermentable carbon source.

Meiosis is a tightly regulated process and several transcriptional regulators play key roles in controlling the sequential expression of sets of genes (VERSHON and PIERCE 2000 ). Entry into meiosis is inhibited by Rme1p (Regulator of meiosis), a three-zinc-finger motif-containing DNA-binding protein (COVITZ and MITCHELL 1993 ), which can exert positive or negative effects on gene expression. Rme1p represses the transcription of the IME1 gene, which is pivotal to the induction of early meiosis-specific genes (KASSIR et al. 1988 ; COVITZ and MITCHELL 1993 ). The protein has been shown to directly bind to two binding sites, Rme1p response elements (RREs), within the IME1 promoter (COVITZ and MITCHELL 1993 ; SHIMIZU et al. 1997 ). In addition to repressing IME1, Rme1p has been shown to positively regulate the CLN2 gene (TOONE et al. 1995 ; FRENZ et al. 2001 ), which encodes a G₁ cyclin and controls cell cycle progression through the initializing phase of a new cell division cycle (CROSS 1995 ). Thus, Rme1p appears to be able to promote mitosis by inducing CLN2 transcription and to prevent meiosis by repressing IME1 (TOONE et al. 1995 ). It has been suggested that repression and activation by Rme1p are due to the exclusion of other factors from the promoter and that this exclusion can occur at large distances from the RRE (SHIMIZU et al. 1997 , SHIMIZU et al. 1998 ; BLUMENTAL-PERRY et al. 2002 ). Some evidence suggests that the Rme1p-dependent exclusion of transcription factors may be linked to chromatin condensation (COVITZ et al. 1994 ). Additional data suggest that Rme1p interacts with the yeast Mediator complex, required for various aspects of transcriptional regulation, and in particular with the subunits Rgr1p and Sin4p (BLUMENTAL-PERRY et al. 2002 ).

In haploid yeast, RME1 is constitutively expressed at relatively high levels. In these cells, nutrient depletion leads to a further induction of RME1 expression to ensure that haploids will not initiate meiosis under any circumstances (SHIMIZU et al. 1997 ). Compared to haploid strains, the expression of RME1 is repressed 10- to 20-fold in diploid strains by the MATa/ heterodimeric repressor (MITCHELL and HERSKOWITZ 1986 ). However, expression in both haploid and diploid strains is cell cycle dependent, with an observed increase in expression at the M/G₁ boundary of the cell cycle (FRENZ et al. 2001 ). The data suggest that Rme1p may contribute to some unknown cellular functions in diploid strains (FRENZ et al. 2001 ).

Invasive and pseudohyphal growth are controlled by a network of signaling modules and transcription factors that respond to the limited availability of nutrients (GAGIANO et al. 2002 ). Signaling modules include the nutrient-dependent mitogen-activated protein (MAP) kinase cascade (LIU et al. 1993 ; MOSCH et al. 1996 ; MADHANI et al. 1997 ) and the cAMP-protein kinase A (PKA) pathway (WARD et al. 1995 ; ROBERTSON and FINK 1998 ; PAN and HEITMAN 1999 ). Some evidence also implicates G₁ cyclins in the regulation of this cellular adaptation (OEHLEN and CROSS 1998 ; LOEB et al. 1999 ). Deletions of CLN1 and/or CLN2 result in a decrease in invasive growth, with the deletion of CLN2 leading to a less severe reduction.

All of the signaling pathways appear to converge on the promoter of the FLO11 (also known as MUC1) gene, the expression of which is essential for invasive growth and pseudohyphal differentiation to occur (LAMBRECHTS et al. 1996 ; LO and DRANGINIS 1998 ; GAGIANO et al. 1999B ; RUPP et al. 1999 ). FLO11 encodes a glycosyl-phosphatidylinositol-anchored cell wall protein and is coregulated with the STA2 gene, which encodes a starch-degrading glucoamylase (GAGIANO et al. 1999A , GAGIANO et al. 1999B ).

The promoters of FLO11 and STA2 are 97% identical and represent some of the largest promoters identified in S. cerevisiae. Indeed, sequences >2.5 kb upstream of the ATG translation start site have been shown to be required for proper regulation (GAGIANO et al. 1999A ; RUPP et al. 1999 ). The extensive size of the promoters appears to correlate with the complexity of the transcriptional control, since numerous regulators have been associated with FLO11 and STA2 expression. Transcription of FLO11 and/or STA2 has been shown to be negatively affected by the products of the NRG1, NRG2 (KUCHIN et al. 2002 ), SFL1 (ROBERTSON and FINK 1998 ; PAN and HEITMAN 2002 ), and SOK2 (WARD et al. 1995 ; PAN and HEITMAN 2000 ) genes, while the FLO8, MSN1, MSS11, PHD1, STE12, and TEC1 genes have all been shown to encode activating proteins (GAGIANO et al. 1999A , GAGIANO et al. 1999B , GAGIANO et al. 2003 ; RUPP et al. 1999 ; PAN and HEITMAN 2000 ; KOHLER et al. 2002 ).

Here we show that RME1 acts as a central switch between nutrient-induced cellular differentiation pathways. The data demonstrate that Rme1p activates invasive growth and starch degradation in haploid cells by inducing FLO11 and STA2. We furthermore show that the promoter of FLO11 contains a functional RRE and that mutations within this site render Rme1p incapable of exerting its effect. The activity of Rme1p appears independent of the identified signaling pathways that regulate invasive growth, including the cAMP-PKA pathway, the nutrient-sensing MAP kinase cascade, and the G₁ cyclins, as well as of other transcriptional regulators that affect FLO11 and STA2 transcription. The data therefore suggest the existence of an additional pathway that controls cellular adaptation to the nutritional status of the environment and that Rme1p may act as a central regulatory element of this pathway.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Strains and culture composition:
The yeast strains used in this study are listed in Table 1. Strains were cultivated at 30° using standard YPED medium prior to transformation or synthetic minimal medium lacking the appropriate amino acids for plasmid/knockout selection (SHERMAN et al. 1991 ). Yeast strains were transformed using the lithium acetate method according to AUSUBEL et al. 1994 . The YPED medium was supplemented with 300 mg/liter geneticin (Sigma-Aldrich, St. Louis) for the selection of geneticin-resistant transformants. Media used for starch degradation, invasive growth, and ß-galactosidase assays contained 2% starch (SCS), 3% glycerol and 3% ethanol (SCGE and SLAGE), 2% glucose (SCD and SLAD), or 0.1% glucose (SCLD X-gal) as carbon source. The SCS, SCGE, SCD, and SCLD X-gal media contained 0.67% yeast nitrogen base without amino acids (Difco Laboratories, Detroit), whereas SLAD and SLAGE media contained 50 µM ammonium sulfate as the sole nitrogen source and 0.17% YNB without ammonium sulfate and amino acids (Difco). The SCLD X-gal medium contained 40 mg/liter X-gal (Sigma-Aldrich ) and was prepared according to AUSUBEL et al. 1994 . Solid media contained 2% agar (Difco).

Plasmid DNA was amplified with Escherichia coli strain DH5 (GIBCO BRL/Life Technologies, Rockville, MD), which was cultivated in Luria-Bertani broth at 37°. Bacterial transformations and plasmid isolation were performed according to the procedures described by SAMBROOK et al. 1989 .

Plasmid construction and recombinant DNA techniques:
All the plasmids, constructs, and primers used in this investigation are listed in Table 2 and Table 3. RME1 was isolated from a genomic library (plasmid YEp24-MSS12) as a 1622-bp HpaI-SphI fragment and was subcloned into the HpaI-SphI sites of the YEpLac plasmids (GIETZ and SUGINO 1988 ) to generate YEplac112-RME1, YEplac181-RME1, and YEplac195-RME1. To construct the disruption cassette pgpa2, a 1774-bp SpeI-NruI fragment from pUC118-GPA2 (kindly provided by J. Winderickx), was replaced with the SmaI-NheI fragment containing the LEU2 marker of pJJ252 (JONES and PRAKASH 1990 ). The episomal plasmid YEpLac112-PHD1 and the disruption cassette pphd1 were constructed by digesting a 2792-bp PHD1 PCR product with BamHI-HindIII and cloning the obtained fragment into the corresponding sites of YEplac112 and subsequently a 2214-bp XbaI-BglII fragment of the resulting YEplac112-PHD1 was replaced with LEU2 (XbaI-BamHI) of pJJ252. For the disruption of RAS2, pras2 was constructed by replacing the 428-bp BalI-PstI fragment of YCplac22-RAS2 (GAGIANO et al. 1999B ) with the LEU2-containing SmaI-PstI fragment of YDp-L (BERBEN et al. 1991 ). YEplac112-TEC1 and ptec1 were constructed by cloning a PCR-amplified TEC1 fragment, containing primer-generated EcoRI sites, into the corresponding EcoRI sites of YEplac112 and pSPORT1 (Invitrogen Life Technologies). The resulting pSPORT-TEC1 plasmid was digested with XbaI, blunt-ended, and redigested with NheI, to replace 975 bp of the TEC1 open reading frame (ORF) with LEU2 (SmaI-NheI) of YDp-L. The disruption constructs cln1::HIS3 and cln2::LEU2 were supplied by B. Futcher. An rme1::URA3 disruption cassette was generated with RME1-DISR-F and RME1-DISR-R. Both primers contain 48 nucleotides homologous to upstream and downstream sequences of the RME1 ORF and 20 nucleotides homologous to flanking regions of the URA3-gene of YEp24 (BOTSTEIN et al. 1979 ). The construction of the additional disruption cassettes used in this study is described in GAGIANO et al. 1999A , GAGIANO et al. 1999B .

Reporter cassettes were constructed to determine FLO11 and STA2 expression. P_FLO11-lacZ and P_STA2-lacZ were isolated from pPMUC1-lacZ and pPSTA2-lacZ (GAGIANO et al. 1999A ) as XbaI-NcoI fragments, with 461 nucleotides of the respective promoters fused to lacZ, and ligated to the SpeI-NcoI sites of pGEM-T (Promega, Madison, WI). The resulting constructs were digested with NcoI, blunt-ended, and ligated to the HIS3 gene (BamHI digested and blunt-ended) from YDp-H (BERBEN et al. 1991 ). The integration cassettes were PCR amplified with Fp-PFLO11_BstEII, which binds 430 bp upstream of FLO11/STA2 ATGs, in combination with Rp-PFLO11-lacZ-pGEM-T and Rp-PSTA2-lacZ-pGEM-T, consisting of 60-nucleotide FLO11- and STA2-specific sequences and 20 nucleotides of pGEM-T situated immediately 3' of the reporter cassettes.

Yeast strain construction:
The wild-type yeast used to construct recombinant strains is from the ISP15 and 1278b genetic backgrounds. The laboratory strain, ISP15, carries the STA2 gene, which encodes a glucoamylase (LAMBRECHTS et al. 1996 ; GAGIANO et al. 1999A , GAGIANO et al. 1999B , GAGIANO et al. 2003 ). Expression of STA2 allows growth on media containing starch as the sole carbon source. L5366h and YHUM272 are 1278b derivative strains and were kindly provided by P. Sudberry and H.-U. Mösch, respectively.

The PCR-amplified P_FLO11-lacZ and P_STA2-lacZ integration cassettes were transformed into ISP15 and 1278b (YHUM272) to generate ISP15flo11::lacZ, ISP15sta2::lacZ, and 1278bflo11::lacZ. Integration into the native loci of FLO11 and STA2 was confirmed through Southern blot analysis and subsequent sequencing. All additional gene disruptions were obtained through the one-step gene replacement method (AUSUBEL et al. 1994 ) in wild-type ISP15 and YHUM272 and in the newly constructed lacZ reporter strains. The knockout cassettes for NRG1, NRG2, RME1, SFL1, and SOK2 were obtained through PCR amplification of the corresponding disrupted genes of the mutants from the BY4742 (BRACHMANN et al. 1998 ) mutant collection supplied by European Saccharomyces cerevisiae Archive for Functional Analysis (EUROSCARF).

The diploid strain 2N1278flo11::lacZ is derived from a cross between the two 1278b derivatives 1278bflo11::lacZ (YHUM272) and YHUM271 (kindly provided by H.-U. Mösch). The strain carries one functional FLO11 allele, while the second allele is replaced with the lacZ gene under control of the native FLO11 promoter. The RME1 alleles of 2N1278flo11::lacZ were deleted with the two cassettes, rme1::URA3 and rme1::kanMX4, to generate the recombinant diploid 2N1278flo11::lacZrme1/rme1.

Site-directed mutagenesis:
The genomic DNA of ISP15flo11::lacZ and 1278bflo11::lacZ served as templates for the site-directed mutagenesis of the putative RRE. Primer Fp-PFLO11-RREmut (Table 3) was used to convert the GTACCACAAAA nucleotide sequence to ATATTATAAAA. The subsequent PCR amplification of the RRE mutagenized P_FLO11-lacZ-HIS3 cassettes was performed with primers Fp-PFLO11-RREmut and Rp-FLO11 (+4.0 kb). The mutated lacZ reporter cassettes were reintroduced into wild-type ISP15 and 1278b (YHUM272) to generate ISP15flo11::lacZRREmut and 1278bflo11::lacZRREmut. The desired nucleotide changes were confirmed through sequence analysis.

Invasive growth, starch utilization, and ß-galactosidase assays:
The invasive growth and starch utilization plate assays were performed as described previously by GAGIANO et al. 1999A , GAGIANO et al. 1999B . Transformed strains for the ß-galactosidase assays were allowed to grow for 5 days when 5 ml of SCD liquid medium was inoculated to serve as starter cultures. The precultures were grown overnight and 5 ml SCD medium was freshly inoculated to an optical density at 600 nm (OD₆₀₀) of 0.05, while 5 ml SCGE medium was inoculated to an OD₆₀₀ of 0.15. To ensure that the cells were in the logarithmic growth phase, the SCD cultures were assayed at an OD₆₀₀ of between 1.0 and 1.5. Due to the slow generation time observed for 1278b strains grown in SCGE medium, the cultures were incubated for 24 hr to ensure that an OD₆₀₀ of at least 0.8 was reached before the cells were harvested and assayed. Three independent transformants were assayed and the differences in ß-galactosidase values never exceeded 15%. ß-Galactosidase activity is expressed as Miller units (AUSUBEL et al. 1994 ) and the data represent the average of three independent experiments.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

RME1 affects invasive growth and starch degradation:
RME1 was isolated from a 2µ-based S. cerevisiae genomic library, which was transformed into the starch-degrading ISP15 strain. Transformants were screened for enhanced ability to grow on starch as sole carbon source, a phenotype that suggests increased expression of the STA2 glucoamylase-encoding gene (LAMBRECHTS et al. 1996 ). As can be seen in Fig 1A, multiple copies of RME1 resulted in more efficient starch degradation on starch-containing media. Inversely, the deletion of RME1 led to a decrease in starch utilization. Since starch degradation and invasive growth are coregulated phenotypes, we assessed whether the absence or the increased copy number of RME1 would have a similar effect on invasive growth. Compared to the wild-type strain transformed with the 2µ-control plasmid, the 2µ-RME1-transformed strain invaded the agar more effectively, whereas the rme1 mutant exhibited a reduced invasiveness (Fig 1B). To assess whether these effects may be an indirect consequence of changes in growth rate, we assessed growth rate in various growth media and under growth conditions. No differences could be observed between the different strains (data not shown).

View larger version (40K): In this window In a new window Download PPT slide   Figure 1. RME1 regulates starch degradation and invasive growth. (A) Starch degradation phenotypes of ISP15 wild-type strains transformed with YEplac112 (2µ), YEplac112-RME1 (2µ-RME1), and rme1 on starch-containing SCS medium. Multiple copies of RME1 increase starch degradation, while the rme1 strain shows a reduction in phenotype. The halos surrounding the colonies reflect Sta2p glucoamylase activity. (B) The same strains as in A on SCD medium. (C) Invasive growth phenotypes of strains 1278b (YHUM272) transformed with YEplac112 and YEplac112-RME1 and 1278brme1 carrying YEplac112 on SCD medium. As for ISP15, the strain overexpressing RME1 shows increased invasiveness, while the rme1 strain shows a significant reduction. (D) Induction of invasive growth by YEplac112-RME1 is blocked in flo11 strain (1278bflo11::lacZ) on SCD medium. (E and F) RME1 regulates STA2 (ISP15) and FLO11 (ISP15 and 1278b) expression in SCD (E) and SCGE (F) liquid cultures. The genomic ORFs of STA2 and FLO11 were replaced with lacZ in the wild-type strains and the RME1 deletions were created in the newly constructed reporter strains. ß-Galactosidase activity is expressed in Miller units (AUSUBEL et al. 1994 ).

To verify that multiple RME1 copies and deletion of RME1 led to similar phenotypes in nonstarch-degrading strains, the effect of RME1 on invasive growth was also assessed in the 1278b genetic background. This strain was chosen because it is most commonly used for the genetic analysis of pseudohyphal differentiation and invasive growth. The data confirm the observations made in the ISP15 genetic background: Multiple copies of RME1 led to increased invasiveness, whereas the deletion resulted in a significant decrease in invasive growth (Fig 1C). Again, no growth defects could be observed for any of the strains (data not shown).

Since FLO11 has a well-documented role in cell-substrate adhesion and invasion (LAMBRECHTS et al. 1996 ; LO and DRANGINIS 1998 ; PAN and HEITMAN 2000 ), we decided to assess whether Rme1p requires Flo11p to enhance invasive growth. Fig 1D shows that 2µ-RME1 was no longer able to induce invasive growth in a strain deleted for FLO11, even after a prolonged incubation period of 12 days.

RME1 regulates the transcription of FLO11 and STA2:
Since Rme1p acts as a transcriptional regulator, we assessed whether RME1 copy number directly affects the transcription of STA2 and FLO11. For this purpose, we replaced the chromosomal ORFs of these genes with the ß-galactosidase-encoding lacZ gene. Fig 1E and Fig F, shows that the presence of 2µ-RME1 leads to increased lacZ activity in the three strains, ISP15sta2::lacZ, ISP15flo11::lacZ, and 1278bflo11::lacZ. We also compared the effect of RME1 in fermentable and nonfermentable carbon sources, since both FLO11 and STA2 are subjected to glucose repression. The expression levels conferred by the FLO11 promoter in the ISP15 strain were always 7- to 10-fold lower than those conferred by the STA2 promoter, and both genes showed lower expression in glucose (SCD) than in glycerol-ethanol (SCGE) medium (Fig 1E and Fig F), confirming previously published information (GAGIANO et al. 1999B ). The presence of 2µ-RME1 induced both promoters, P_FLO11 and P_STA2, 5- to 10-fold under both conditions. The deletion of RME1, on the other hand, decreased the expression levels of all reporter genes by 30% in both strains. These data correlate well with the phenotypes observed on plates (Fig 1, A–C), as well as with the reported reduction of CLN2 expression levels in an RME1 deletion strain (TOONE et al. 1995 ; FRENZ et al. 2001 ). It should also be noted that low expression levels of FLO11 in the ISP15 strain make the interpretation of the effects of RME1 deletion on FLO11 expression rather difficult, although repeated experiments always yielded similar data.

The data clearly show that multiple copies and deletion of RME1 result in similar phenotypes and transcriptional changes for both STA2 and FLO11, independently of the genetic background of the strain and of the nature of the carbon source.

Rme1p acts independently of signaling modules that regulate invasive growth:
We assessed whether the regulation of FLO11 by RME1 would be affected by the hyperactive allele of RAS2 or the deletion of signaling modules that regulate invasive growth. For this purpose, the 2µ-RME1 plasmid was transformed into strains with deletions or mutations in genes that affect cAMP-dependent signaling (RAS2^val19, gpa2, ras2) or the nutrient-regulated MAP kinase cascade (ras2, ste7, ste11, ste12, ste20). The experiments were conducted in the haploid 1278b genetic background. The data presented in Table 4 show that the deletion of either RAS2 or GPA2 did not affect the ability of 2µ-RME1 to induce P_FLO11-lacZ transcription. Both deletions (gpa2 and ras2) resulted in a decrease in basal reporter gene-encoded activity in SCD and in SCGE, but the level of induction conferred by the 2µ-RME1 plasmid was always comparable to, or slightly higher than, that observed in the wild type. The same was true in the reverse situation, when the effects of the hyperactive RAS2^val19 mutation were assessed in both wild-type and rme1 genetic backgrounds. The increase in transcription was almost identical in both strains, i.e., 7.8- and 8.3-fold in SCD (Fig 2) and 3-fold in SCGE (results not shown).

View larger version (40K): In this window In a new window Download PPT slide   Figure 2. RME1 deletion does not affect the ability of other transcriptional activators to induce FLO11. Histogram representing the induction ratios obtained for 1278bflo11::lacZ and 1278bflo11::lacZrme1 transformed with YCplac22-RAS2val19 and YEplac112 without insert or with FLO8, MSN1, MSS11, PHD1, and TEC1. The absolute SCD ß-galactosidase values were normalized to the YEplac112 control to obtain the induction ratios for every construct in each strain.

View this table: In this window In a new window   Table 4. Expression of PFLO11-lacZ in 1278b mutant strains

Similarly, multiple copies of RME1 were able to activate invasive growth in the absence of elements of the invasive growth-regulating MAP kinase cascade (Fig 3). As reported previously (MOSCH et al. 1996 ; GAGIANO et al. 1999B ), deletion of the different STE genes resulted in reduced invasive growth, with the strains ste20 and ste11 showing the severest phenotypes. Multiple copies of RME1 were able to restore the invasive growth phenotype in all mutants tested.

View larger version (111K): In this window In a new window Download PPT slide   Figure 3. Assessment of the effect of Rme1p in strains with MAPK gene deletions. L5366h (1278b), L5624h (ste20), L5625h (ste11), L5626h (ste7), and L5627h (ste12) were transformed with YEplac195 and YEplac195-RME1 and grown on SLAD for 5 days at 30° before washing. RME1 overexpression results in increased invasiveness in all cases.

Rme1p induces invasive growth and starch degradation independently of Cln1p and Cln2p:
We investigated whether the effect of Rme1p on FLO11 was dependent on the presence of Cln1p or Cln2p, since

1. Rme1p is known to control CLN2 expression (TOONE et al. 1995 ; FRENZ et al. 2001 ).

2. G₁ cyclins regulate invasive growth (LOEB et al. 1999 ; OEHLEN and CROSS 1998 ).

3. Deletion of RME1 causes a 30% reduction in FLO11 promoter activity (Fig 1E), similar to the reduction observed by TOONE et al. 1995 for CLN2 mRNA in an rme1 strain.

For this purpose, we generated strains deleted for CLN1, CLN2, or both in the ISP15 genetic background. The cln1cln2 double mutant showed clear growth defects and was excluded from the analysis. In accordance with the results of LOEB et al. 1999 , the cln1 strain showed the severest defect in invasive growth, while the cln2 strain also displayed a clear reduction (Fig 4). The presence of 2µ-RME1 in both the cln1 and the cln2 strains strongly enhanced the level of invasion. When tested on starch-containing SCS plates, the deletion of the cyclin genes did not lead to changes in starch degradation, and the presence of 2µ-RME1 resulted in similar increases in the wild-type and the two cyclin-mutated strains.

View larger version (79K): In this window In a new window Download PPT slide   Figure 4. Rme1p induces invasive growth and starch degradation independently of G1 cyclins. Wild-type ISP15, cln1, cln2, and rme1 were transformed with YEplac112 and YEplac112-RME1 and grown on SLAD and SCS (Sta2p activity).

Rme1p does not require other transcriptional regulators:
Several transcription factors have been shown to activate FLO11 expression (RUPP et al. 1999 ; GAGIANO et al. 1999A ). To assess whether Rme1p would require the presence of these factors, we transformed the 1278bflo11::lacZ strain and the isogenic mutants flo8, msn1, mss11, phd1, ste12, and tec1 with 2µ-RME1. The effects on transcription of FLO11 were assessed in both fermentable and nonfermentable carbon sources. Basal levels of lacZ-encoded ß-galactosidase activity in the wild-type strain grown on glucose-containing medium (SCD) were severely affected by deletions of FLO8, MSN1, MSS11, STE12, and TEC1 (Table 4), with expression levels being reduced at least 6-fold. As reported previously, the deletion of PHD1 did not affect FLO11 expression to the same extent, but resulted in a still significant reduction of 65% of reporter gene-encoded activity. On nonfermentable carbon sources, however, only deletions of FLO8, MSN1, and MSS11 resulted in a similarly severe decrease in lacZ expression, suggesting that the presence of STE12 and TEC1 may not be required to the same extent under glucose-derepressed conditions. This corroborates data published by RUPP et al. 1999 that showed that the FLO11 expression levels of ste12 and tec1 strains were close to wild-type levels in postdiauxic shift cultures, but significantly reduced during exponential growth on glucose. However, under both glucose-repressed and -derepressed conditions and in all the mutants, 2µ-RME1 was able to increase ß-galactosidase activity significantly. Interestingly, the deletion of the two genes that affect basal transcription levels most severely, MSS11 and FLO8, also resulted in the lowest 2µ-RME1-dependent induction in SCD. However, in SCGE the induction ratios are the highest for mss11 (18.7-fold) and flo8 (13.5-fold), which is probably due to the very low basal lacZ transcription levels.

In the reverse situation, all 2µ plasmids carrying the genes of the different factors were able to activate transcription by the same induction factor in the wild-type and rme1 strains (Fig 2). In all cases, the expression data also correlated well with the invasive growth phenotype of each strain (data not shown).

We also assessed the effect of the deletions on the STA2 reporter system in the ISP15 strain. An excellent correlation between starch degradation phenotypes and P_STA2-lacZ expression could be observed (data not shown). Furthermore, the ISP15 STA2 data also correlate well with the FLO11 data obtained in the 1278bflo11::lacZ strain, again demonstrating the coregulated nature of the two genes and the validity of the data for different genetic backgrounds.

The effect of the deletion of genes that have been shown to negatively affect FLO11 and/or STA2 expression is presented in Table 5. Deletions of NRG1, NRG2, and SOK2 result in a slight (nrg1) to a 2- and 3-fold increase (nrg2 and sok2, respectively) in P_FLO11-lacZ expression in SCD medium. The most significant effect is observed with the sfl1 strain, which shows a 25-fold increase in basal reporter gene activity. As observed for the transcriptional activators described above, none of the deletions appeared to affect the ability of 2µ-RME1 to induce lacZ expression, although the level of induction in the sfl1 is reduced to 1.2- and 1.1-fold in SCD and SCGE, respectively. However, this may be due to the very high basal level of expression in this strain, which may not allow for further increases in expression levels.

The hypothesis that Rme1p acts independently of the repressor Sfl1p is supported by the data obtained for STA2 expression in the ISP15 strain. In this case, the deletion of SFL1 did not derepress the STA2 gene to the same extent, and 2µ-RME1 was able to induce transcription significantly by a factor of 4.4. Another important difference between the two strains can be observed in the response to the deletion of NRG2. Indeed, the deletion appears not to affect STA2 expression significantly in ISP15, contrarily to the effect on FLO11 expression observed in the 1278bflo11::lacZ strain. Nrg2p also appears to mediate glucose repression, since the deletion of NRG2 leads to a twofold increase in lacZ expression in SCD, but no induction can be observed in SCGE medium.

Rme1p induces FLO11 expression via an Rme1p response element:
Sequence analysis of P_FLO11 and P_STA2 revealed the presence of a putative RRE, GTACCACAAAA, at positions -1427 and -1314, respectively (Fig 5). The only difference between this sequence and the previously identified RREs in the promoters of IME1 and CLN2 is a T to A substitution in position 6 of the consensus sequence in P_FLO11 and P_STA2. To assess the role of this putative RRE, we mutated the GTACCACAAAA nucleotide stretch to ATATTATAAAA in the FLO11 promoters of ISP15flo11::lacZ and 1278bflo11::lacZ, since the guanine and cytosine nucleotides had been shown to be required for Rme1p-DNA interaction (SHIMIZU et al. 1998 , SHIMIZU et al. 2001 ). Fig 6A shows that 2µ-RME1 was no longer able to properly activate the P_FLO11-lacZ with the mutated RRE. In strain ISP15flo11::lacZ, the 2µ-RME1 plasmid resulted in the production of ß-galactosidase, as indicated by the dark color of the colony, whereas the strain with the RRE mutation exhibited very little activity.

View larger version (64K): In this window In a new window Download PPT slide   Figure 5. Rme1p response elements in the promoters of FLO11, STA2, IME1, and CLN2. W, A or T; R, A or G; D, A or G or T.

View larger version (37K): In this window In a new window Download PPT slide   Figure 6. Rme1p requires the PFLO11 RRE to induce lacZ expression. (A) ISP15flo11::lacZ and ISP15flo11::lacZRREmut were transformed with YEplac195 and YEplac195-RME1 and grown on SCLD (0.1% glucose) supplemented with X-gal, for 12 days. The dark color of the colony formed by strain ISP15flo11::lacZ transformed with YEplac195-RME1 is indicative of lacZ expression. The strain carrying the mutation in the putative RRE does not show a similar induction when transformed with the same plasmid. (B) ß-Galactosidase activity of the ISP15 strains used in A measured after growth in SCD (see MATERIALS AND METHODS). (C) ß-Galactosidase activity of strains 1278bflo11::lacZ, 1278bflo11::lacZrme1, and 1278bflo11::lacZRREmut transformed with YEplac181, YEplac181-FLO8, and YEplac181-RME1 and grown in SCD as described in MATERIALS AND METHODS.

The values of ß-galactosidase activity indicated a 30% reduction in activity of the FLO11 promoter when the RRE sequence was mutated in both the ISP15 and the 1278b reporter strains (Fig 6B and Fig C). This reduction is similar to the reduction observed in the RME1-deleted 1278b strain (Fig 6C). The RRE mutations also significantly reduced the ability of 2µ-RME1 to induce the reporter gene. However, transcriptional activation by multiple copies of RME1 was not entirely abolished, since the 2µ-RME1 plasmid still resulted in a twofold increase in ß-galactosidase activity, compared to the eightfold increase observed in the wild-type ISP15 strain.

To further verify whether RREmut specifically affected RME1-dependent activation, 1278b reporter strains were transformed with 2µ-FLO8 and 2µ-RME1 plasmids. Fig 6C shows that the mutated promoter was fully activated by Flo8p, in terms of both absolute ß-galactosidase units and induction ratio. Reporter gene-encoded activity increased 3.1- and 3.4-fold in the presence of 2µ-FLO8 in wild-type and RRE-mutated strains, respectively, while the corresponding values for 2µ-RME1 are 3 and 1.2. Similar data were obtained when multiple copies of MSN1, MSS11, PHD1, and TEC1 were assessed in the RRE mutant strain (results not shown).

The very slight residual induction of lacZ activity by the 2µ-RME1 plasmid in both the ISP15 and 1278b RREmut reporter strains (Fig 6B and Fig C) may suggest that the promoter of FLO11 contains a second RRE. Both the promoters of IME1 and CLN2 contain two Rme1p response elements each (SHIMIZU et al. 1997 , SHIMIZU et al. 2001 ; FRENZ et al. 2001 ). However, no other sequence with significant homology to the identified RREs could be identified in the FLO11 and STA2 promoters.

Effects of Rme1p in diploid strains:
RME1 expression is strongly repressed in diploid cells (MITCHELL and HERSKOWITZ 1986 ). We nevertheless assessed whether RME1 affected invasive growth similarly in diploid and haploid cells. The isogenic diploid strains 2N1278flo11::lacZ (2N) and 2N1278flo11::lacZrme1/rme1 (2Nrme1) transformed with the 2µ-control and 2µ-RME1 plasmids were tested for their ability to invade different growth substrates. No difference in invasive growth could be observed between the wild-type strain and the RME1-deleted strain (Fig 7A). In the presence of 2µ-RME1, the wild-type and RME1-mutant strains presented no observable phenotypes when grown on SCD medium. However, a significant increased invasiveness is exhibited when both strains were grown on nonfermentable carbon sources, with the strongest increase being observed on nitrogen-limited SLAGE medium. We also assessed whether RME1 affected the formation of pseudohyphae in the diploid strains. The only significant difference was that elongated cells and pseudohyphae formation could be observed 48 hr after spotting on the SLAD medium in the 2µ-RME1 transformed strain, whereas both the wild type and the disrupted strain required an additional 24 hr before elongated cells could be observed. However, total cell elongation as well as the final length of individual filaments appeared unaffected. The rme1/rme1 strain formed pseudohyphae with an efficiency similar to that of wild type.

View larger version (48K): In this window In a new window Download PPT slide   Figure 7. Effect of RME1 in diploid strains. (A) Strains 2N (MATa/ URA3/ura3-52 flo11::lacZ-HIS3/FLO11) and 2Nrme1 (MATa/ flo11::lacZ-HIS3/FLO11 rme1::URA3/rme1::kanMX4) bearing either YEplac195 or YEplac195-RME1 were spotted onto SCD, SCGE, and SLAGE and allowed to grow for 5 days at 30° before washing. 2µ-RME1 is able to induce invasion on nonfermentable carbon sources, but not on glucose-containing SCD. Deletion of RME1 does not lead to a visible difference in invasive growth when compared to the wild type. (B) ß-Galactosidase activity on SCD and SCGE media of the diploid strains 2N1278flo11::lacZ (2N) and 2N1278flo11::lacZrme1/rme1 (2Nrme1) transformed with YEplac181 and YEplac181-RME1 (see Table 1 for relevant genotypes).

To quantify the effect of multiple copies of RME1 on FLO11 transcription in the diploid background, strains 2N1278flo11::lacZ and 2N1278flo11::lacZrme1/rme1, which both still contain one functional copy of FLO11, were also tested for ß-galactosidase activity. The lacZ expression levels were the same for the diploid wild-type reporter and the rme1/rme1 strains in both SCD and SCGE (Fig 7B). In the strains transformed with multiple copies of RME1, on the other hand, induction was dependent on the growth substrate, contrary to the situation in the haploid 1278b strain (Table 4). Indeed, the 2µ-RME1-transformed diploids showed virtually no lacZ induction when grown in SCD, while a twofold induction above wild-type level was observed when 2µ-RME1-transformed diploids were grown in SCGE medium. Rme1p therefore is able to induce FLO11 expression in diploid strains in the presence of nonfermentable carbon sources such as glycerol and ethanol (Fig 7B) and to increase invasive growth under conditions of nitrogen limitation (Fig 7A).

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Rme1p controls nutrient-dependent cellular differentiation:
Our data provide evidence that Rme1p acts as a genetic switch between nutrient-controlled growth forms of S. cerevisiae and, in particular, induces invasive growth while repressing meiosis in haploid cells:

1. Multiple copies of RME1 significantly enhance FLO11 and STA2 transcription as well as the associated phenotypes invasive growth and starch degradation.

2. Deletion of RME1 leads to a 30% reduction in the transcription of FLO11 and STA2, which is also reflected in the associated phenotypes.

3. A specific sequence within the promoters of FLO11 and STA2 confers Rme1p responsiveness.

4. The mutation of this promoter element leads to a reduction in basal transcription levels similar to that resulting from the deletion of RME1.

Previously, the ability of Rme1p to activate CLN2 expression, coupled with the cell cycle-dependent expression of RME1, has been taken as evidence for the involvement of this protein in the regulation of mitosis (TOONE et al. 1995 ). Taken together, these and our data suggest that Rme1p plays a general role as a transcriptional regulator of genes that are central to the control of nutrient-dependent cellular growth forms, i.e., unicellular mitotic growth, invasive and pseudohyphal growth, and spore formation.

Haploid vs. diploid strains:
Our data clearly indicate that Rme1p enhances invasive growth in haploid strains by activating the expression of FLO11. In diploid strains, however, deletion of RME1 did not reduce invasion or FLO11 transcription under any of the conditions tested here. These data suggest that Rme1p may not be relevant for the regulation of invasion and pseudohyphal differentiation in diploids. However, multiple copies of RME1 activated invasion and FLO11 expression in diploids in a nutrient-dependent manner, requiring the absence of glucose and being enhanced by low levels of available nitrogen. These observations may indicate that Rme1p does play a role in the regulation of invasion in diploid cells, but that the specific conditions required to monitor these phenotypes may not have been tested here. Alternatively, the data may be explained by the fact that the a12 repressor in heterozygous MATa/MAT diploid strains strongly represses RME1 transcription (MITCHELL and HERSKOWITZ 1986 ). FLO11 itself is also repressed by the same repressor, and it is therefore possible that the induction of FLO11 observed in the diploid 2µ-RME-transformed strain may be due to a12 titration. However, this scenario would not explain the fact that induction in diploids appears to be dependent on specific growth conditions.

Conditions promoting sporulation in diploid strains and invasive growth in haploid strains are very similar, but for one essential difference: sporulation is favored by the complete depletion of nitrogen sources, whereas invasion requires that nitrogen sources be present, at least in limited amounts. Rme1p could therefore be specifically required to favor invasion and inhibit sporulation in haploids and diploids under conditions when the risk of wrongly activating the sporulation pathway is highest. This hypothesis is strengthened by data of GASCH et al. 2000 , indicating that RME1 expression is induced in response to nitrogen limitation.

Recent evidence shows that RME1 is expressed in a cell cycle-dependent manner in both haploid and diploid cells, peaking in late M/G₁ of the mitotic cell cycle (FRENZ et al. 2001 ). On the basis of these results, the authors suggested that Rme1p is linked to the promotion of cell cycle progression in both cell types and may be required for the proper regulation of alternative developmental pathways.

Rme1p regulates FLO11 transcription via an RRE:
Rme1p acts directly via an RRE sequence in the promoter of the FLO11 gene. As in the case of the RREs in P_IME1 (COVITZ and MITCHELL 1993 ) and P_CLN2 (TOONE et al. 1995 ), the FLO11 and STA2 RREs are situated far upstream of the ATG translation start codons, in positions -1427 and -1314, respectively. Mutations within the FLO11 RRE significantly reduced, but did not completely eliminate, the ability of multiple copies of RME1 to activate transcription. This might suggest the presence of a second RRE in the promoter of FLO11, resembling the situation in P_CLN2 and P_IME1. However, careful scanning did not reveal the presence of a second consensus sequence in the 3.5-kb sequence of P_FLO11 and P_STA2.

The RRE is situated in an area that was pinpointed as being essential for the regulation of FLO11 by several groups (RUPP et al. 1999 ; GAGIANO et al. 1999A ; PAN and HEITMAN 2002 ). In particular, PAN and HEITMAN 2002 showed that Flo8p acts in close proximity to the identified RRE. Furthermore, KOBAYASHI et al. 1999 proposed that Flo8p might act via a sequence that contains the RRE in the promoters of STA1 (STA2 homolog) and FLO11. It therefore is highly significant that the mutations in the RRE did not affect the ability of Flo8p to activate FLO11, indicating that the presence or absence of Rme1p on the FLO11 promoter does not affect Flo8p activity.

Rme1p acts independently of known signaling mechanisms and of transcriptional regulators of invasive growth:
Rme1p acts independently of the invasive growth-regulating signaling pathways, the cAMP/PKA pathway, and the invasive growth-modulating MAPK pathway. It also does not require the G₁ cyclins. In fact, the deletion of CLN1 or CLN2 has no effect on the ability of Rme1p to induce invasive growth.

The data also show that other transcriptional regulators of FLO11 and STA2 were not affected by Rme1p. Indeed, all factors investigated were still able to confer similar levels of induction or repression in an rme1 and in a wild-type strain when present on a multiple copy plasmid. Similarly, 2µ-RME1 has led to increased FLO11 expression in strains deleted for any of these factors.

Possible mechanism of Rme1p-dependent regulation of FLO11:
It is unclear how Rme1p interacts with other elements that regulate invasive and pseudohyphal growth and which signal is responsible for this regulation. A possible link between RME1 and invasive growth may be established through the further investigation of factors that regulate RME1 transcription. For example, Swi5p has been shown to regulate RME1 expression (TOONE et al. 1995 ) and has recently also been implicated in the regulation of FLO11 (PAN and HEITMAN 2000 ).

It has been suggested that Rme1p acts by excluding other factors from promoters (COVITZ et al. 1994 ; SHIMIZU et al. 1997 ). Since this exclusion may occur at sites that are situated at significant distances from the RRE, it has been hypothesized that this effect may be chromatin dependent (COVITZ et al. 1994 ). The activation of FLO11 transcription by Rme1p therefore may be due to the exclusion of one or several transcriptional repressors. We investigated whether the effect of RME1 is dependent on the exclusion of specific or general repressor proteins that regulate FLO11 transcription, including Sok2p (PAN and HEITMAN 2000 ), Sfl1p (ROBERTSON and FINK 1998 ; CONLAN and TZAMARIAS 2001 ; PAN and HEITMAN 2002 ), Nrg1p, or Nrg2p (KUCHIN et al. 2002 ). Our results show that transformants carrying 2µ-RME1 resulted in elevated P_FLO11-lacZ and P_STA2-lacZ expression in strains deleted for any of these repressor genes. Similarly, strains lacking the functional activators Flo8p, Msn1p, Mss11p, Phd1p, Ste12p, and Tec1p also exhibited higher levels of reporter gene activity in the presence of 2µ-RME1.

A role for Rme1p in lifting general repression appears the most likely hypothesis and would also best fit other, previously described regulatory roles of the protein. In this regard the Tup1p-Ssn6p general corepressor complex (CONLAN and TZAMARIAS 2001 ) is a possible candidate for Rme1p-related function. Although it was shown that Rme1p and the Tup1p-Ssn6p repressor complex act independently to repress IME1 transcription (MIZUNO et al. 1998 ), the possibility remains that these proteins interact functionally to regulate FLO11 transcription, since Rme1p seems to play an activating rather than a repressive role in this context. Other potential proteins involved in Rme1p activity include components of RNA polymerase II holoenzyme, since Rgr1p and Sin4p have been shown to be required for RME1-dependent repression of IME1 (COVITZ et al. 1994 ; BLUMENTAL-PERRY et al. 2002 ), and Sin4p has also been implicated in FLO11 repression (CONLAN and TZAMARIAS 2001 ).

Taken together, our data suggest that Rme1p controls cellular adaptation to the nutritional status of the environment and may act as the central regulatory element of a new, previously unidentified pathway. Other proteins, in particular Sok2p, have also been implicated in similar multiple regulatory roles, including repression of meiosis, activation of mitosis, and control of invasive and pseudohyphal differentiation (SHENHAR and KASSIR 2001 ). However, Sok2p acts as a repressor of invasive and pseudohyphal growth and, according to our data, does not appear to interact with Rme1p. Considering the previously published evidence regarding Rme1p, we suggest that an Rme1p-dependent pathway may act as a general cellular coordinator, rather than as a specific input/specific output mechanism, and may tilt the cellular machinery toward one or another differentiation status, according to cell type and environmental conditions.

	FOOTNOTES

¹ Present address: The Australian Wine Research Institute, Waite Road, Urrbrae, SA 5064 Adelaide, Australia.

	ACKNOWLEDGMENTS

The authors thank B. Futcher, J. Winderickx, P. Sudberry, and H.-U. Mösch for strains; J. Arensburg for critical reading of the manuscript; and M. Steiner and W. Schwarzer for technical assistance. This work was supported by grants from the South African Wine Industry (Winetech) and the National Research Foundation (NRF) of South Africa.

Manuscript received May 13, 2003; Accepted for publication July 8, 2003.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

AUSUBEL, F. M., R. BRENT, R. E. KINGSTON, D. D. MOORE, J. G. SEIDMAN et al., 1994 Current Protocols in Molecular Biology. John Wiley & Sons, New York.

BAUER, F. F., and I. S. PRETORIUS, 2001 Pseudohyphal and invasive growth in Saccharomyces cerevisiae, pp. 109–133 in Focus on Biotechnology—Applied Microbiology, edited by A. DURIEUX and J.-P. SIMON. Kluwer Academic Publishers, Dordrecht, The Netherlands.

BERBEN, G., J. DUMONT, V. GILLIQUET, P. A. BOLLE, and F. HILGER, 1991  The YDp plasmids: a uniform set of vectors bearing versatile gene disruption cassettes for Saccharomyces cerevisiae.. Yeast 7:475-477.[Medline]

BLUMENTAL-PERRY, A., W. LI, G. SIMCHEN, and A. P. MITCHELL, 2002  Repression and activation domains of Rme1p structurally overlap, but differ in genetic requirements. Mol. Biol. Cell 13:1709-1721.[Abstract/Free Full Text]

BOTSTEIN, D., F. C. FALCO, S. E. STEWART, M. BRENNAN, and S. SCHERER et al., 1979  Sterile host yeast (SHY): a eukaryotic system of biological containment for recombinant DNA experiments. Gene 8:17-24.[Medline]

BRACHMANN, C. B., A. DAVIES, G. J. COST, E. CAPUTO, and J. LI et al., 1998  Designer deletion strains derived from Saccharomyces cerevisiae S288C: a useful set of strains and plasmids for PCR-mediated gene disruption and other applications. Yeast 14:115-132.[Medline]

CONLAN, R. S. and D. TZAMARIAS, 2001  Sfl1 functions via the co-repressor Ssn6-Tup1 and the cAMP-dependent protein kinase Tpk2. J. Mol. Biol. 309:1007-1015.[Medline]

COVITZ, P. A. and A. P. MITCHELL, 1993  Repression by the yeast meiotic inhibitor RME1.. Genes Dev. 7:1598-1608.[Abstract/Free Full Text]

COVITZ, P. A., W. SONG, and A. P. MITCHELL, 1994  Requirement for RGR1 and SIN4 in RME1-dependent repression in Saccharomyces cerevisiae.. Genetics 138:557-586.

CROSS, F. R., 1995  Starting the cell cycle: What's the point? Curr. Opin. Cell Biol. 7:790-797.[Medline]

FRENZ, L. M., A. L. JOHNSON, and L. H. JOHNSTON, 2001  Rme1, which controls CLN2 expression in Saccharomyces cerevisiae, is a nuclear protein that is cell cycle regulated. Mol. Genet. Genomics 266:374-384.[Medline]

GAGIANO, M., D. VAN DYK, F. F. BAUER, M. G. LAMBRECHTS, and I. S. PRETORIUS, 1999a  Divergent regulation of the evolutionarily closely related promoters of the Saccharomyces cerevisiae STA2 and MUC1 genes. J. Bacteriol. 181:6497-6508.[Abstract/Free Full Text]

GAGIANO, M., D. VAN DYK, F. F. BAUER, M. G. LAMBRECHTS, and I. S. PRETORIUS, 1999b  Msn1p/Mss10p, Mss11p and Muc1p/Flo11p are part of a signal transduction pathway downstream of Mep2p regulating invasive growth and pseudohyphal differentiation in Saccharomyces cerevisiae.. Mol. Microbiol. 31:103-106.[Medline]

GAGIANO, M., F. F. BAUER, and I. S. PRETORIUS, 2002  The sensing of nutritional status and the relationship to filamentous growth in Saccharomyces cerevisiae.. FEMS Yeast Res. 2:433-470.[Medline]

GAGIANO, M., M. BESTER, D. VAN DYK, J. FRANKEN, and F. F. BAUER et al., 2003  Mss11p is a tran