SGM Journals
Research Article

Ambient pH signalling in the yeast Yarrowia lipolytica involves YlRim23p/PalC, which interacts with Snf7p/Vps32p, but does not require the long C terminus of YlRim9p/PalI

Sylvie Blanchin-Roland, Grégory Da Costa, Claude Gaillardin

  • Laboratoire de Microbiologie et Génétique Moléculaire, AgroParisTech, INRA, UMR1238, CNRS, UMR2585, F-78850 Thiverval-Grignon, France
  • Correspondence
    Sylvie Blanchin-Roland
    Sylvie.Blanchin-Roland{at}grignon.inra.fr

    • Microbiology 2008; 154(6):1668–1676 · https://doi.org/10.1099/mic.0.2008/017046-0

    View at publisher PubMed

    Abstract

    A conserved ambient pH signal transduction pathway has been evidenced in both ascomycetous yeasts and filamentous fungi, called the Rim or Pal pathway, respectively. However, closely related PalC orthologues are found only in Yarrowia lipolytica and in filamentous fungi, where the Rim9p/PalI factor has a much longer C-terminal tail than in other yeasts. We show here that, like Aspergillus nidulans palI mutants, a Ylrim9Δ mutant has a less extreme phenotype than other mutants of the pathway, whereas rim9 mutants in Saccharomyces cerevisiae and Candida albicans reportedly exhibit a tight Rim phenotype. Deletion of the long C-terminal tail of YlRim9p/PalI had no phenotypic effect on ambient pH signalling. We also show that the Y. lipolytica PalC orthologue, named YlRim23p, is absolutely required for the alkaline pH response. Its only interactant identified in a genome-wide two-hybrid screen is YlSnf7/Vps32p, confirming the link between the Rim and the Vps pathways. YlRim13p and YlRim20p both interact with YlSnf7/Vps32p but not with YlRim23p. The long C-terminal tail of YlRim9p/PalI interacts neither with YlRim23p nor with YlSnf7/Vps32p. These results show that YlRim23p is a bona fide component of the Rim pathway in Y. lipolytica and that it participates in the complexes linking pH signalling and endocytosis.

    Edited by: M. Molina

    INTRODUCTION

    Extracellular pH regulation plays major roles in the control of many biological processes, including the survival of fungi and bacteria and the virulence of some of these micro-organisms. Environmental pH profoundly affects gene expression in many organisms, especially in those which are able to grow over a wide pH range. Ambient pH-responsive genes include those encoding permeases, plasma-membrane transporters, secreted proteins and enzymes involved in the synthesis of exported metabolites. In the yeast Yarrowia lipolytica, transcription of the XPR2 and AXP1 genes, encoding alkaline and acidic extracellular proteases, respectively, is tightly controlled by a combination of environmental stimuli, including ambient pH and nutrient availability (Ogrydziak et al., 1977; Ogrydziak, 1993). In Candida albicans, the genes encoding the three proteases Sap4p, Sap5p and Sap6p are under the control of extracellular pH, Sap5p being required for E-cadherin degradation during mucosal tissue invasion (Villar et al., 2007).

    These genes are regulated by a conserved pH signalling pathway called Pal in filamentous fungi and Rim in yeasts. This pathway has been intensively investigated in Aspergillus nidulans (Penalva & Arst, 2002; Herranz et al., 2005; Hervas-Aguilar et al., 2007; Galindo et al., 2007; Penas et al., 2007; Calcagno-Pizarelli et al., 2007), in Y. lipolytica (Lambert et al., 1997; Tréton et al., 2000; Gonzalez-Lopez et al., 2002; Blanchin-Roland et al., 2005), in several other ascomycetes, including Saccharomyces cerevisiae (Li & Mitchell, 1997; Xu & Mitchell, 2001; Vincent et al., 2003; Xu et al., 2004; Boysen & Mitchell, 2006), in C. albicans (Ramon & Fonzi, 2003; Li et al., 2004; Kullas et al., 2004; Cornet et al., 2005; Barwell et al., 2005; Baek et al., 2006), in the basidiomycete Ustilago maydis (Arechiga-Carvajal & Ruiz-Herrera, 2005) and in other fungi (Penalva & Arst, 2002, 2004). At alkaline pH, a cascade of six pal or RIM genes activates the zinc-finger transcriptional factor PacC/Rim101p through a complex C-terminal proteolytic processing event (see below). Previous reports have established that components of the endocytic trafficking cooperate with the Rim pathway in the recruitment of Rim20p and its targeting to Rim101p. Endosomal sorting complex required for transport (ESCRT)-I, -II and -III are three protein complexes that act sequentially in the sorting of both ubiquitinated membrane protein and biosynthetic cargos to the multivesicular body (MVB) vesicles, which are ultimately delivered to the lumen of the vacuole for degradation (Hurley & Emr, 2006). ESCRT-III proteins Snf7/Vps32p and Vps20p and all ESCRT-I and -II components are required for Rim101p activation in S. cerevisiae (Xu et al., 2004; Rothfels et al., 2005; Hayashi et al., 2005). This functional link between the Rim and Vps pathways is conserved in C. albicans (Xu et al., 2004; Kullas et al., 2004; Cornet et al., 2005) and Y. lipolytica (Blanchin-Roland et al., 2005). The PacC/Rim101p truncated form is able to activate alkaline pH-responsive genes and to repress acid-induced genes (for reviews see Penalva & Arst, 2002, 2004). At acidic pH, the Rim/Pal pathway is admittedly inactive, although a specific processing event of Rim101p has been reported in C. albicans (Li et al., 2004). In Y. lipolytica, the Rim pathway seemed to be indirectly required at acidic pH for the activation of the acid-responsive AXP1 gene (Gonzalez-Lopez et al., 2002).

    Besides its major role in the ambient alkaline pH response and growth, the Rim pathway appears essential for mating and sporulation in Y. lipolytica (Lambert et al., 1997) and it controls sporulation in S. cerevisiae by activating IME1 expression (Li & Mitchell, 1997) and corepressing, with Nrg1, DIT1 expression in mitotic cells (Rothfels et al., 2005). It also regulates cell differentiation during haploid invasive growth and diploid pseudohyphal differentiation in S. cerevisiae (Lamb & Mitchell, 2003). It is required for the hyphal formation at alkaline pH in C. albicans, in particular by repressing NRG1 expression in a pH-dependent way (El Barkani et al., 2000; Lotz et al., 2004). Moreover, the resistance of cells to high levels of Ca2+ and Li+ requires activity of the Rim pathway, and the subset of ESCRT components needed for Rim101p processing in S. cerevisiae (Lamb & Mitchell, 2003; Platara et al., 2006), in C. albicans (Kullas et al., 2007) and in Fusarium oxysporum (Caracuel et al., 2003). In addition, the Rim pathway participates directly in cell-wall assembly, and acts in parallel with the protein kinase C (PKC) pathway in S. cerevisiae (Castrejon et al., 2006). In C. albicans, this pathway regulates PHR1 and PHR2, encoding two isoforms of the cell-wall glycosidases important for adaptation to ambient pH (Muhlschlegel & Fonzi, 1997), and represses the acid-expressed cell-wall gene RBR1, encoding a glycosylphosphatidylisositol (GPI) protein required for filamentation (Lotz et al., 2004). Finally, defects in ambient pH response lead to reduced virulence in C. albicans, A. nidulans and other pathogenic fungi (Penalva & Arst, 2002; Davis, 2003; Bignell et al., 2005; Cornet et al., 2005; Mitchell et al., 2007; Villar et al., 2007).

    According to the current model, largely elaborated through studies in A. nidulans, the pH signalling pathway is organized into two complexes: a plasma-membrane complex, including Rim21p/PalH, Rim9p/PalI and Rim8p/PalF, and an endosomal-membrane complex, including Rim20p/PalA, Snf7/Vps32p, Rim13p/PalB, to which Rim101p/PacC is recruited for its proteolytic activation. The arrestin-like protein PalF (Rim8p) binds strongly to two regions within the C-terminal cytosolic tail of the 7TM PalH protein, which is the likely pH sensor (Herranz et al., 2005). At alkaline pH, PalF is phosphorylated and ubiquitinated in a pH- and PalH-dependent manner. These PalF modifications are partially dependent on the 3TM protein PalI, but independent of the three other Pal proteins: PalA, PalB and PalC (Herranz et al., 2005). Recently, PalI (Rim9p) was shown to localize to the plasma membrane where it may assist localization of PalH (Calcagno-Pizarelli et al., 2007). PalF phosphorylation and ubiquitination may lead to the endocytosis of the PalF–PalH complex, transducing the signal to the downstream complex (Herranz et al., 2005). Several reports in A. nidulans, S. cerevisiae and C. albicans clearly show that PalA/Rim20p interacts with both Snf7/Vps32p, a subunit of ESCRT-III (see above), and PacC/Rim101p (Vincent et al., 2003; Xu & Mitchell, 2001). Both Rim20p–Rim101p and Rim20p–Snf7/Vps32p interactions are required for Rim101p processing. An interaction between the calpain-like protease Rim13p/PalB and Snf7/Vps32p was detected during one whole-genome screen of two-hybrid interactants in S. cerevisiae (Ito et al., 2001), but no direct interaction between Rim13p/PalB and Rim101p/PacC has been reported. In A. nidulans, PalB was recently shown to be the ‘signalling’ protease that catalyses the first, pH-regulated cleavage of PacC (Penas et al., 2007), whereas the second processing step involves the proteasome (Hervas-Aguilar et al., 2007).

    PalC was the only Pal protein with no homologue in hemiascomycetous yeasts, except Y. lipolytica. However, recently, YGR122w was identified as a likely S. cerevisiae orthologue (Galindo et al., 2007). Ygr122wp was previously shown to bind Snf7/Vps32p (Ito et al., 2001; Uetz et al., 2000) and to be required for activation of Rim101p (Rothfels et al., 2005; Barwell et al., 2005), and ygr122wΔ results in Li+ hypersensitivity (Galindo et al., 2007). A mutational analysis of PalC revealed conserved regions, one potentially functional as a BRO1 domain (Tilburn et al., 2005). While this manuscript was in preparation, PalC was shown to localize to cortical structures at alkaline pH and to bind Snf7/Vps32p (Galindo et al., 2007).

    In this paper, we show that YlRIM23, the orthologue of PalC, is required for the pH response as well as for mating in Y. lipolytica. We also report that YlRim23p binds to YlSnf7/Vps32p and confirm that YlSnf7/Vps32p binds to both YlRim13p and YlRim20p. We further show that the long C-terminal tail of YlRim9p does not seem to be involved in the pH response.

    METHODS

    Strains and sequence data.

    The Escherichia coli strain used was Mach1-T1 (Invitrogen). Yeast strains are described in Table 1⇓. Unless otherwise stated, sequence data were obtained from the complete genome assembly of Y. lipolytica (; Dujon et al., 2004).

    Table 1.

    Strains used in this study

    Culture media and phenotypic tests.

    Complete YPD medium and derepressing medium (Y) have been described previously (Lambert et al., 1997). Solid and liquid media were buffered at pH 4.0 (with 0.2 M citrate buffer), at pH 7.0 (with 0.2 M sodium phosphate buffer) or at pH 8.0 (with 25 mM Tris/HCl) and supplemented with 0.1 g uracil l−1. The pH of liquid media was stable within 0.3 pH units during growth. Growth at various pHs was assayed as previously described (Blanchin-Roland et al., 2005). Estimation of efficiency of mating for compatible Y. lipolytica auxotrophs was done as described by Lambert et al. (1997).

    DNA and RNA techniques.

    Standard recombinant DNA techniques were performed essentially as previously described (Gonzalez-Lopez et al., 2002). PCR products were routinely checked by sequencing; sequences were obtained from the DNA sequencing department of Genome Express. They were assembled and annotated using the GCG package (University of Wisconsin, Madison, WI, USA). All transformation events were checked by colony PCR and confirmed by Southern analysis. Gene expression was determined by real-time quantitative reverse transcriptase-PCR as previously described (Blanchin-Roland et al., 2005).

    Deletion of YlRIM23 and YlRIM9.

    The rim23Δ : : URA3 and the rim9Δ : : URA3 deletion cassettes were constructed according to Fickers et al. (2003). The promoter and terminator regions of each gene were amplified using primers RIM23FP1/RIM23P2+RIM23T1/RIM23HT2 or RIM9CP1/RIM9P2+RIM9T1/RIM9ET2, respectively (see Table 2⇓ for primer sequences). For each gene, the resulting fragments were annealed and amplified with the flanking primers. These cassettes were cloned into pBSKS+ to generate plasmids BSPTRIM23 and BSPTRIM9, respectively. The loxR-URA3-loxP module (Fickers et al., 2003) was inserted between promoter and terminator sequences at the I-SceI site, generating plasmids pINA1375 and pINA1359, respectively. The cassettes were excised from these plasmids and purified with the QIAquick gel extraction kit (Qiagen). The RIM9 deletion cassette was integrated by double crossover at the YlRIM9 locus of the SY12 strain to give four independent clones of the SY1375 strain (Ylrim9Δ-13 allele). The RIM23 deletion cassette was similarly used to generate two independent clones of the SY1359 strain (Ylrim23Δ-10 allele). YlURA3 was excised from the SY1375 strain by using the plasmid pUB4-CRE (Fickers et al., 2003) to give the SY1375u strain.

    Table 2.

    Primers used in this study

    To delete YlRIM23 in the Ylrim9Δ-13 background and in the SY1372u strain carrying the YlRIM9SL allele (see below), a deletion cassette was constructed from the loxR-Hph-loxP module (Fickers et al., 2003) and cloned into BSPTRIM23 at the I-SceI site, to generate plasmid pINA1387. The deletion cassette was excised from pINA1387, purified as above, then integrated by double crossover at the YlRIM23 locus into the SY1372u strain to generate five independent clones of the SY1389 strain (YlRIM9SL, Ylrim23Δ-31) and into the SY1375u strain to generate four independent clones of the SY1390 strain (Ylrim9Δ-13, Ylrim23Δ-31).

    Subcloning of the YlRIM9 gene.

    The complete YlRIM9 ORF flanked by 660 bp of upstream and 544 bp of downstream sequences was rescued from plasmid AW0AA029A08 (AL414126). To increase the size of the upstream sequence, a PCR product was amplified from SY12 genomic DNA using primers RIM91 and RIM92r (Table 2⇑). Both this PCR product and the fragment carrying RIM9 were inserted into the integrative URA3 plasmid pINA300′ to give pINA1371.

    Construction of the mutant encoding a C-terminally truncated YlRim9p.

    The YlRIM9SL-1372 allele was obtained by overlapping PCR and contains two mutations, A889T and C894G, which replace the Arg297 and Tyr298 codons by nonsense codons within the RIM9 coding sequence. PCR products were obtained from genomic DNA with primer pairs RIM9CP1-RIM9end3r and RIM9end5-RIM9ET2 (Table 2⇑), mixed and amplified with flanking primers to generate a full-length product containing the internal mutations. This PCR product was inserted into pINA1371 to generate plasmid pINA1372, which was checked by sequencing. The BamHI-digested pINA1372 was targeted to the promoter of RIM9 in the SY1375u strain (Ylrim9Δ-13 allele) to give three independent clones of the SY1372 strain (YlRIM9SL allele). BamHI-digested pINA1371 was similarly targeted in the SY1375u strain to give three independent clones of the SY1371 strain (YlRIM9 allele). YlURA3 was excised from the SY1372 strain, using pUB4-CRE as above, to give the SY1372u strain.

    Yeast two-hybrid assay.

    S. cerevisiae PJ69-4A (Table 1⇑) and plasmids pAS2ΔΔ (TRP1, bla) and pACT2 (LEU2, bla) were used for the two-hybrid analysis (James et al., 1996). Primers were designed to amplify the complete RIM8 nucleotide sequence (RIM8DHSf and RIM8DHBa), the complete RIM13 nucleotide sequence (RIM13DHNc and RIM13DHBa), the complete RIM23 nucleotide sequence (RIM23DHNc and RIM23DHBa) and the C-terminal RIM9 coding sequence from codon 254 to the stop codon 724 (RIM9CDHNc and RIM9CDHBa) (Table 2⇑). The PCR products were digested by NcoI (or SfiI in the case of RIM8) and BamHI to allow in-frame cloning in the multisite of each two-hybrid vector. The recombinant vectors carrying the complete RIM20 nucleotide sequence were constructed by B. Tréton (unpublished). Empty and recombinant two-hybrid vectors were co-transformed into yeast cells and transformants were selected on minimal medium plus histidine, methionine, uracil and adenine for selection of the plasmids. Purified transformants were streaked on YNB plus histidine, methionine and uracil, on YNB plus adenine, methionine, uracil and 5 mM 3-aminotriazole or on YNB plus methionine and uracil for direct selection of the interactants (James et al., 1996).

    Two-hybrid library screening.

    The PJ69-4α strain was transformed with pAS2ΔΔ-RIM23 (see above), and a mid-exponential culture (120 ml) of the resulting transformant was mixed with an aliquot of each of the three Y. lipolytica two-hybrid libraries constructed in pGAD-C1 to C3 plasmids in PJ69-4A (Kabani et al., 2000) at a ratio of 3 : 1 (bait : prey). After mating at 30 °C for 16 h, cells were harvested and plated onto YNB plus adenine, methionine, uracil and 5 mM 3-aminotriazole for selection of His+ colonies. The number of diploid and parental cells was determined as previously described (Kabani et al., 2000). The number of diploids obtained (BC1, 9.3×107; BC2, 4.7×109; BC3, 7.5×105) was sufficient to ensure a representative sampling of the two first pools of the library (107 clones per pool). After 7–8 days at 30 °C, His+ diploids were replica-plated onto YNB plus histidine, methionine and uracil plates. Ade+ colonies were recovered from the third to the eighth day and streaked twice on YNB plus methionine and uracil plates to retain Ade+ His+ diploids.

    RESULTS AND DISCUSSION

    Role of YlRIM23 and YLRIM9 in ambient pH response

    Previous approaches led us to identify orthologues of all the A. nidulans Pal pathway members except one, PalC (Lambert et al., 1997; Gonzalez-Lopez et al., 2002). A putative orthologue (YALI0F17028 g) of palC was identified from the Y. lipolytica genome (Dujon et al., 2004; Tilburn et al., 2005). The predicted 441 aa protein was named YlRim23p. A likely PalC orthologue, encoded by YGR122w, was recently identified in S. cerevisiae (Galindo et al., 2007). Moreover, PalH has two homologues in S. cerevisiae and C. albicans, Rim21p and Dfg16p (Barwell et al., 2005), but only one in Y. lipolytica, Rim21p. We found two putative paralogues of YlRim21p (YALI0B22814g and YALI0B06710g) in the Y. lipolytica genome. Deletion of each of these ORFs had no effect on pH response (not shown). This suggests that the Y. lipolytica genome carries only one functional equivalent of PalH/YlRim21 and none of ScDfg16p.

    To check if YlRIM23 was involved in ambient pH signalling, the SY1359 strain carrying the null mutation (Ylrim23Δ-10) was created (see Methods). We previously observed that the Ylrim9-D7 insertional mutant had a milder effect than other rim mutants on growth or PHR1 expression at alkaline pH (Gonzalez-Lopez et al., 2002). To check whether the same leaky phenotype was displayed by a null mutant, the SY1375 strain carrying the Ylrim9Δ-13 allele was constructed (see Methods). The effects of YlRIM23 or YlRIM9 deletion on the transcription of different pH-regulated genes were assessed by real-time quantitative PCR on total RNAs extracted from cultures grown in Y medium at pH 4.0 and pH 8.0, using actin transcript as a reference. As ambient alkaline pH-responding genes, we chose XPR2 and YlPHR1 and as acidic pH-responding gene, YlPHR2 (Blanchin-Roland et al., 2005).

    At pH 8.0, both the Ylrim23Δ-10 and Ylrim9Δ-13 mutants nearly abolished XPR2 transcription, and the Ylrim23Δ-10 mutant reduced YlPHR1 expression about 15-fold, whereas the Ylrim9Δ-13 mutant reduced it about 6-fold (Fig. 1⇓). These observations were confirmed by measuring the expression of the XPR2′ : lacZ fusion integrated at the XPR2 locus of these strains: the residual β-galactosidase activity was about 1.4 % for Ylrim23Δ-10 and 14.6 % for Ylrim9Δ-13. At pH 4.0, no or very low transcription of these alkaline reporter genes could be detected (Fig. 1⇓). At alkaline pH, in the Ylrim23Δ-10 mutant as in the Ylrim9Δ-13 mutant, expression of the acid-induced gene YlPHR2 was derepressed more than 38-fold compared to the wild-type level at the same pH, reaching its level at acidic pH in the reference strain in the case of Ylrim9Δ-13 or exceeding it in the case of Ylrim23Δ-10 (Fig. 1⇓). Taken together, these results strongly suggest that, like other YlRIM genes (Gonzalez-Lopez et al., 2002; Lambert et al., 1997), YlRIM23 is absolutely required for transcriptional activation of alkaline-induced genes and for repression of acid-induced genes at pH 8.0. The Ylrim9Δ-13 mutant, by contrast, led to a somewhat less extreme phenotype, as previously stated (Gonzalez-Lopez et al., 2002).

    Figure image not available in archive
    Fig. 1.

    Expression patterns of alkaline-induced genes (XPR2 and PHR1) and an acid-induced gene (PHR2) in Y. lipolytica at pH 4.0 and pH 8.0. The strains were: SY12 (WT), SY1359u (rim23), SY1375u (rim9), SY1371 (rim9+RIM9), SY1372 (rim9+RIM9SL), SY1389 (rim23+RIM9SL), SY1390 (rim9+rim23). RIM9SL stands for the YlRIM9 C-terminally truncated allele (see Methods). Levels of mRNA were measured by quantitative real-time PCR and expressed as a percentage of actin gene expression levels. Results are means±sd from four quantifications.

    Ylrim23Δ-10 induced the characteristic growth defect of Rim mutants, affecting growth at alkaline pH but not at neutral or acidic pH (Fig. 2⇓) (Gonzalez-Lopez et al., 2002). As previously observed for the insertional rim9-D7 mutant (Gonzalez-Lopez et al., 2002), the Ylrim9Δ-13 mutation had a less pronounced effect on growth at alkaline pH.

    Figure image not available in archive
    Fig. 2.

    Effects of the different mutations on sensitivity of Y. lipolytica to ambient pH. Droplets of serial dilutions of a late-exponential-phase culture in YPD medium were deposited on solid YPD medium buffered at pH 3.5 with 0.2 M citrate buffer, and at pH 7.0–8.5 with 0.2 M MOPS buffer, and incubated at 28 °C for 40 h. The strains were SY1390 (rim9+rim23), SY1389 (rim23+RIM9SL), SY1375u (rim9), SY1371 (rim9+RIM9), SY1372 (rim9+RIM9SL), SY1359u (rim23), LAM26-03 (rim13) and SY12 (WT).

    All these results establish that the YlRIM23 gene, like other YlRIM genes, is required for cellular responses to ambient alkaline pH. The null mutation Ylrim9Δ leads to a less extreme phenotype than other Ylrim mutations, as measured by activation of alkaline-induced genes, repression of an acid-induced gene (Fig. 1⇑), growth at alkaline pH (Fig. 2⇑) and efficiency of mating (see below), confirming the data obtained with the insertional mutant described by Gonzalez-Lopez et al. (2002). In a similar way, a palI null mutation in A. nidulans allows partial growth at pH 8.0, whereas all other pal mutations completely prevent growth at this pH (Arst et al., 1994; Denison et al., 1998). The phenotype of rim9 mutants in yeasts has been poorly investigated up to now. In S. cerevisiae, rim9 mutants exhibit a tight defect in sporulation (Su & Mitchell, 1993); in C. albicans, the rim9 −/− mutant displays a tight defect in ambient pH response (M. Cornet, personal communication).

    The long YlRim9p C terminus does not seem to be involved in the pH response

    Alignment of PalI/Rim9p sequences showed that YlRim9p (742 aa) (Gonzalez-Lopez et al., 2006) and its orthologues in filamentous fungi are much longer than their orthologues in other yeasts (239 aa for ScRim9p). Strong sequence conservation is restricted to the regions including the hydrophobic, putative transmembrane segments. The highly hydrophilic C-terminal tails that lie outside the region of homology between all the PalI/Rim9p sequences represent about 75 % of YlRim9p, 66 % of PalI, 82 % of MgPalI and 72 % of NcPalI. We noticed that deletion of a long Rim9p/PalI in Y. lipolytica and filamentous fungi results in a leaky phenotype, whereas the short Rim9p appears absolutely required in S. cerevisiae and C. albicans. Although we failed to detect any structural or functional patterns within these poorly conserved sequences, we wondered whether this large C-terminal tail in YlRim9p played a role in the pH response. To address this question, we established the phenotype of a mutant expressing a C-terminally truncated YlRim9 polypeptide. Point mutations were introduced to replace the adjacent residues R297 and Y298 by nonsense codons (YlRIM9SL allele, see Methods). This construct replaced the wild-type gene at its normal location (SY1372 strain). As a control, the wild-type allele was used to replace the null mutation Ylrim9Δ-13 (SY1371 strain). The effects of the C-terminally truncated form of YlRim9p on the transcription of different pH-regulated genes were assessed by real-time quantitative PCR at pH 4.0 and pH 8.0. At pH 8.0, identical levels of XPR2 or YlPHR1 transcripts were measured in SY1372 (YlRIM9SL mutation) and in the isogenic control SY1371. They were reduced by 30 % and 6 % for XPR2 and YlPHR1, respectively, compared to the wild-type strain, reflecting unidentified background differences (Fig. 1⇑). These observations were fully confirmed by measuring the expression of the XPR2′ : lacZ fusion integrated at the XPR2 locus of SY1372 and SY1371 (72 % of residual β-galactosidase activity for both strains). At alkaline pH, in SY1372 as in SY1371, expression of the acid-induced gene YlPHR2 was also comparable to wild-type levels at the same pH (Fig. 1⇑). At pH 4.0, the YlRIM9SL mutation had no effect on the transcription of these alkaline- or acid-responsive genes (Fig. 1⇑). Taken together, these results strongly suggest that the C-terminal tail of YlRim9p is not required for transcriptional activation of alkaline-induced genes and for repression of acid-induced genes at pH 8.0.

    Both Y. lipolytica and filamentous fungi carry a conserved YlRIM23/PalC gene, besides a long version of YlRIM9/PalI. We wondered whether YlRim23p/PalC may counteract a putative negative effect of the YlRim9p/PalI C-terminal tail on pH signal transduction. According to this hypothesis, the null rim23 mutant in the RIM9SL context should behave like the wild-type strain. RIM23 was also deleted in the rim9Δ-13 mutant. All clones of SY1389 (rim23Δ-1, RIM9SL) and of SY1390 (rim23Δ-31, rim9Δ-13) exhibited low levels of XPR2-driven β-galactosidase activity at pH 8.0 (about 0.03 % of wild-type levels). Transcription of alkaline- or acidic-responsive genes was affected in a similar way and to the same extent by the double mutations or by the single mutation Ylrim23Δ-10 (see Fig. 1⇑). The double mutations induced the characteristic growth defect of Rim mutants, drastically crippling growth at alkaline but not at neutral pH (Fig. 2⇑). Taken together, these results show that YlRIM23 is absolutely required for the RIM101-dependent pH response, but does not counteract a putative negative effect of the YlRim9p C-terminal tail on pH signal transduction. This indicates that in Y. lipolytica at least, and possibly in filamentous fungi, the long C-terminal tail of YlRim9p/PalI does not play a major role in the pH response and that its presence is not related to that of YlRim23p/PalC.

    Mating in Y. lipolytica requires YlRim23p and, to a lesser extent, YlRim9p

    We previously reported that mating in Y. lipolytica requires activation of the Rim pathway in both parents. Mutants affecting RIM8, RIM13, RIM20, or RIM21 mated less efficiently than the control strains, yielding few pal/+ and even fewer pal/pal diploids (Lambert et al., 1997). As shown in Table 3⇓, efficiency of mating was drastically reduced compared to wild-type contexts in +×rim23Δ-10 crosses, reduced in +×rim8-21 and slightly reduced in +×rim9Δ-13. The RIM9SL C-terminal truncation alone did not affect efficiency of mating whereas no diploids could be obtained from +×rim23Δ-1, RIM9SL and +×rim23Δ-31, rim9Δ-13 confrontations. This indicates that mating requires Rim23 like other Rim factors indicated above, and to a lesser extent Rim9, but not the C-terminal tail of Rim9p.

    Table 3.

    Effect of RIM status on mating

    Screening of the two-hybrid library for partners of YlRim23p

    To investigate the function of YlRim23p, we looked for partners of this factor using the two-hybrid screen (Fields & Song, 1989). Exhaustive matings against three genomic libraries (Kabani et al., 2000), yielded 19, 175 and 68 His+ Ade+ diploids, respectively (see Methods). The inserted genomic fragments were directly amplified from yeast colonies by PCR and sequenced. From 132 sequences out of a total of 149 analysed, redundant and overlapping fragments were found for only one ORF, which is a strong argument for a true interaction. This ORF matches the YlSNF7/VPS32 gene (accession number YALI0C16027g), hereafter designated SNF7 for simplicity. The fusion points were generated by either TaqI or HpaII, which are the two ‘CG’ enzymes cutting in the promoter sequence in-frame with the coding sequence, at positions –75 bp and – 69 bp relative to the A of ATG, respectively. No fusion point within the coding sequence was generated by the ‘CG’ enzymes. This suggests that at least the N-terminal region of Snf7 is required for the interaction with Rim23p. No other candidate protein was repeatedly obtained in this screening. Plasmids were purified from one representative clone for each fusion point and co-transformed together with either pAS2ΔΔ or pAS2ΔΔ-RIM23 in the PJ69-4A strain. After checking the expression of the HIS3 and ADE2 reporter genes, the clones turned out to be true positives while all the controls were negative (Fig. 3a⇓, A and B). These results indicated no auto-activation of the expression of the reporter genes and reproduced the interaction of Snf7p with Rim23p. This result is in line with recently published data from A. nidulans demonstrating by two-hybrid and recombinant protein pull-down assays that PalC binds Snf7p (Galindo et al., 2007). PalC was recently shown to be recruited to plasma-membrane-associated structures in an alkaline pH- and PalH-dependent but PalA-independent manner, which led to the suggestion that PalC can link the putative pH-sensing complex and the endosomal complex (Galindo et al., 2007).

    Figure image not available in archive
    Fig. 3.

    Two-hybrid interactions within the Rim/Pal pathway. (a) YlSnf7p interacts with YlRim23p (A and B) and with YlRim13p (C and D). Co-transformants were grown on medium lacking Trp and Leu (A and C) and on the quadruple selection medium –Leu –Trp – His –Ade (B and D) using strain PJ69-4A. Parts A and B: 1, pAS2ΔΔ+pACT2; 2, pGADSNF7-C2/6+pAS2ΔΔ; 3, pGADSNF7-C2/47+ pAS2ΔΔ; 4, pAS2ΔΔ-RIM23+pACT2; 5, pGADSNF7-C2/6+pAS2ΔΔ-RIM23; 6, pGADSNF7-C2/47+pAS2ΔΔ-RIM23. Parts C and D: 1, pGADSNF7-C2/6+pAS2ΔΔ; 2, pAS2ΔΔ-RIM13+pACT2; 3, pAS2ΔΔ+pACT2; 4, 5 and 6, three independent clones carrying pGADSNF7-C2/6+pAS2ΔΔ-RIM13. (b) Network of YlRim23p interactions tested in this work. Data were obtained with the –Leu –Trp –His –Ade system with the S. cerevisiae PJ69-4A strain. Arrows and T symbols stand for positive or negative two-hybrid interactions for a given prey-to-bait combination, respectively.

    YlSnf7p binds to YlRim13p or YlRim20p in two-hybrid interactions, in addition to YlRim23p

    The screening of the two-hybrid library indicated only YlSnf7p as a possible in vivo direct partner of YlRim23p. No two-hybrid interaction could be detected between YlRim23p and YlRim8p, or YlRim13p, or YlRim20p, or the C-terminal tail of YlRim9p. In addition, this region was unable to bind YlSnf7p. In contrast, two-hybrid assays confirmed that YlSnf7p binds to YlRim20p (data not shown) and also to YlRim13p (Fig. 3a⇑, C and D). Fig. 3(b)⇑ summarizes the network of two-hybrid interactions tested. These assays confirm the interactions between Snf7p and Rim20p/PalA/AIP1/Alix already described in S. cerevisiae, A. nidulans and C. albicans (Ito et al., 2001; Vincent et al., 2003) and between Snf7p and Rim13p, revealed only by one whole-genome screen of two-hybrid interactants in S. cerevisiae (Ito et al., 2001). They also suggest the absence of interaction between YlRim23p and any of the other Rim factors tested.

    Acknowledgments

    We thank Mehdi Kabani, Anita Boisramé, Marie-Thérèse Le Dall and Brigitte Tréton for providing the two-hybrid vectors and helpful advice. This work was supported by the Institut National de la Recherche Agronomique (INRA) and the Centre National de la Recherche Scientifique (CNRS).

    References

    1. Arechiga-Carvajal, E. T. & Ruiz-Herrera, J. (2005). The RIM101/pacC homologue from the basidiomycete Ustilago maydis is functional in multiple pH-sensitive phenomena. Eukaryot Cell 4, 999–1008.
    2. Arst, H. N., Jr, Bignell, E. & Tilburn, J. (1994). Two new genes involved in signalling ambient pH in Aspergillus nidulans. Mol Gen Genet 245, 787–790.
    3. Baek, Y. U., Martin, S. J. & Davis, D. A. (2006). Evidence for novel pH-dependent regulation of Candida albicans Rim101, a direct transcriptional repressor of the cell wall beta-glycosidase Phr2. Eukaryot Cell 5, 1550–1559.
    4. Barwell, K. J., Boysen, J. H., Xu, W. & Mitchell, A. P. (2005). Relationship of DFG16 to the Rim101p pH response pathway in Saccharomyces cerevisiae and Candida albicans. Eukaryot Cell 4, 890–899.
    5. Bignell, E., Negrete-Urtasun, S., Calcagno, A. M., Haynes, K., Arst, H. N., Jr & Rogers, T. (2005). The Aspergillus pH-responsive transcription factor PacC regulates virulence. Mol Microbiol 55, 1072–1084.
    6. Blanchin-Roland, S., Da Costa, G. & Gaillardin, C. (2005). ESCRT-I components of the endocytic machinery are required for Rim101-dependent ambient pH regulation in the yeast Yarrowia lipolytica. Microbiology 151, 3627–3637.
    7. Boysen, J. H. & Mitchell, A. P. (2006). Control of Bro1-domain protein Rim20 localization by external pH, ESCRT machinery, and the S. cerevisiae Rim101 pathway. Mol Biol Cell 17, 1344–1353.
    8. Calcagno-Pizarelli, A. M., Negrete-Urtasun, S., Denison, S. H., Rudnicka, J. D., Bussink, H. J., Múnera-Huertas, T., Stanton, L., Hervás-Aguilar, A., Espeso, E. A. & other authors (2007). Establishment of the ambient pH signaling complex in Aspergillus nidulans: PalI assists plasma membrane localization of PalH. Eukaryot Cell 6, 2365–2375.
    9. Caracuel, Z., Casanova, C., Roncero, M. I., Di Pietro, A. & Ramos, J. (2003). pH response transcription factor PacC controls salt stress tolerance and expression of the P-Type Na+-ATPase Ena1 in Fusarium oxysporum. Eukaryot Cell 2, 1246–1252.
    10. Castrejon, F., Gomez, A., Sanz, M., Duran, A. & Roncero, C. (2006). The RIM101 pathway contributes to yeast cell wall assembly and its function becomes essential in the absence of mitogen-activated protein kinase Slt2p. Eukaryot Cell 5, 507–517.
    11. Cornet, M., Bidard, F., Schwarz, P., Da Costa, G., Blanchin-Roland, S., Dromer, F. & Gaillardin, C. (2005). Deletions of endocytic components VPS28 and VPS32 affect growth at alkaline pH and virulence through both RIM101-dependent and RIM101-independent pathways in Candida albicans. Infect Immun 73, 7977–7987.
    12. Davis, D. (2003). Adaptation to environmental pH in Candida albicans and its relation to pathogenesis. Curr Genet 44, 1–7.
    13. Denison, S. H., Negrete-Urtasun, S., Mingot, J. M., Tilburn, J., Mayer, W. A., Goel, A., Espeso, E. A., Penalva, M. A. & Arst, H. N., Jr (1998). Putative membrane components of signal transduction pathways for ambient pH regulation in Aspergillus and meiosis in Saccharomyces are homologous. Mol Microbiol 30, 259–264.
    14. Dujon, B., Sherman, D., Fischer, G., Durrens, P., Casaregola, S., Lafontaine, I., de Montigny, J., Marck, C., Neuvéglise, C. & other authors (2004). Genome evolution in yeasts. Nature 430, 35–44.
    15. El Barkani, A., Kurzai, O., Fonzi, W. A., Ramon, A., Porta, A., Frosch, M. & Muhlschlegel, F. A. (2000). Dominant active alleles of RIM101 (PRR2) bypass the pH restriction on filamentation of Candida albicans. Mol Cell Biol 20, 4635–4647.
    16. Fickers, P., Le Dall, M. T., Gaillardin, C., Thonart, P. & Nicaud, J. M. (2003). New disruption cassettes for rapid gene disruption and marker rescue in the yeast Yarrowia lipolytica. J Microbiol Methods 55, 727–737.
    17. Fields, S. & Song, O. (1989). A novel genetic system to detect protein-protein interactions. Nature 340, 245–246.
    18. Galindo, A., Hervas-Aguilar, A., Rodriguez-Galan, O., Vincent, O., Arst, H. N., Jr, Tilburn, J. & Penalva, M. A. (2007). PalC, one of two Bro1 domain proteins in the fungal pH signalling pathway, localizes to cortical structures and binds Vps32. Traffic 8, 1346–1364.
    19. Gonzalez-Lopez, C. I., Szabo, R., Blanchin-Roland, S. & Gaillardin, C. (2002). Genetic control of extracellular protease synthesis in the yeast Yarrowia lipolytica. Genetics 160, 417–427.
    20. Gonzalez-Lopez, C. I., Ortiz-Castellanos, L. & Ruiz-Herrera, J. (2006). The ambient pH response Rim pathway in Yarrowia lipolytica: identification of YlRIM9 and characterization of its role in dimorphism. Curr Microbiol 53, 8–12.
    21. Hayashi, M., Fukuzawa, T., Sorimachi, H. & Maeda, T. (2005). Constitutive activation of the pH-responsive Rim101 pathway in yeast mutants defective in late steps of the MVB/ESCRT pathway. Mol Cell Biol 25, 9478–9490.
    22. Herranz, S., Rodriguez, J. M., Bussink, H. J., Sanchez-Ferrero, J. C., Arst, H. N., Jr, Penalva, M. A. & Vincent, O. (2005). Arrestin-related proteins mediate pH signaling in fungi. Proc Natl Acad Sci U S A 102, 12141–12146.
    23. Hervas-Aguilar, A., Rodriguez, J. M., Tilburn, J., Arst, H. N., Jr & Penalva, M. A. (2007). Evidence for the direct involvement of the proteasome in the proteolytic processing of the Aspergillus nidulans zinc-finger transcription factor PACC. J Biol Chem 282, 34735–34747.
    24. Hurley, J. H. & Emr, S. D. (2006). The ESCRT complexes: structure and mechanism of a membrane-trafficking network. Annu Rev Biophys Biomol Struct 35, 277–298.
    25. Ito, T., Chiba, T., Ozawa, R., Yoshida, M., Hattori, M. & Sakaki, Y. (2001). A comprehensive two-hybrid analysis to explore the yeast protein interactome. Proc Natl Acad Sci U S A 98, 4569–4574.
    26. James, P., Halladay, J. & Craig, E. A. (1996). Genomic libraries and a host strain designed for highly efficient two-hybrid selection in yeast. Genetics 144, 1425–1436.
    27. Kabani, M., Boisrame, A., Beckerich, J. M. & Gaillardin, C. (2000). A highly representative two-hybrid genomic library for the yeast Yarrowia lipolytica. Gene 241, 309–315.
    28. Kullas, A. L., Li, M. & Davis, D. A. (2004). Snf7p, a component of the ESCRT-III protein complex, is an upstream member of the RIM101 pathway in Candida albicans. Eukaryot Cell 3, 1609–1618.
    29. Kullas, A. L., Martin, S. J. & Davis, D. (2007). Adaptation to environmental pH: integrating the Rim101 and calcineurin signal transduction pathways. Mol Microbiol 66, 858–871.
    30. Lamb, T. M. & Mitchell, A. P. (2003). The transcription factor Rim101p governs ion tolerance and cell differentiation by direct repression of the regulatory genes NRG1 and SMP1 in Saccharomyces cerevisiae. Mol Cell Biol 23, 677–686.
    31. Lambert, M., Blanchin-Roland, S., Le Louedec, F., Lepingle, A. & Gaillardin, C. (1997). Genetic analysis of regulatory mutants affecting synthesis of extracellular proteinases in the yeast Yarrowia lipolytica: identification of a RIM101/pacC homolog. Mol Cell Biol 17, 3966–3976.
    32. Li, W. & Mitchell, A. P. (1997). Proteolytic activation of Rim1p, a positive regulator of yeast sporulation and invasive growth. Genetics 145, 63–73.
    33. Li, M., Martin, S. J., Bruno, V. M., Mitchell, A. P. & Davis, D. A. (2004). Candida albicans Rim13p, a protease required for Rim101p processing at acidic and alkaline pHs. Eukaryot Cell 3, 741–751.
    34. Lotz, H., Sohn, K., Brunner, H., Muhlschlegel, F. A. & Rupp, S. (2004). RBR1, a novel pH-regulated cell wall gene of Candida albicans, is repressed by RIM101 and activated by NRG1. Eukaryot Cell 3, 776–784.
    35. Mitchell, B. M., Wu, T. G., Jackson, B. E. & Wilhelmus, K. R. (2007). Candida albicans strain-dependent virulence and Rim13p-mediated filamentation in experimental keratomycosis. Invest Ophthalmol Vis Sci 48, 774–780.
    36. Muhlschlegel, F. A. & Fonzi, W. A. (1997). PHR2 of Candida albicans encodes a functional homolog of the pH-regulated gene PHR1 with an inverted pattern of pH-dependent expression. Mol Cell Biol 17, 5960–5967.
    37. Ogrydziak, D. M. (1993). Yeast extracellular proteases. Crit Rev Biotechnol 13, 1–55.
    38. Ogrydziak, D. M., Demain, A. L. & Tannenbaum, S. R. (1977). Regulation of extracellular protease production in Candida lipolytica. Biochim Biophys Acta 497, 525–538.
    39. Penalva, M. A. & Arst, H. N., Jr (2002). Regulation of gene expression by ambient pH in filamentous fungi and yeasts. Microbiol Mol Biol Rev 66, 426–446.
    40. Penalva, M. A. & Arst, H. N., Jr (2004). Recent advances in the characterization of ambient pH regulation of gene expression in filamentous fungi and yeasts. Annu Rev Microbiol 58, 425–451.
    41. Penas, M. M., Hervas-Aguilar, A., Munera-Huertas, T., Reoyo, E., Penalva, M. A., Arst, H. N., Jr & Tilburn, J. (2007). Further characterization of the signaling proteolysis step in the Aspergillus nidulans pH signal transduction pathway. Eukaryot Cell 6, 960–970.
    42. Platara, M., Ruiz, A., Serrano, R., Palomino, A., Moreno, F. & Arino, J. (2006). The transcriptional response of the yeast Na+-ATPase ENA1 gene to alkaline stress involves three main signaling pathways. J Biol Chem 281, 36632–36642.
    43. Ramon, A. M. & Fonzi, W. A. (2003). Diverged binding specificity of Rim101p, the Candida albicans ortholog of PacC. Eukaryot Cell 2, 718–728.
    44. Rothfels, K., Tanny, J. C., Molnar, E., Friesen, H., Commisso, C. & Segall, J. (2005). Components of the ESCRT pathway, DFG16, and YGR122w are required for Rim101 to act as a corepressor with Nrg1 at the negative regulatory element of the DIT1 gene of Saccharomyces cerevisiae. Mol Cell Biol 25, 6772–6788.
    45. Su, S. S. & Mitchell, A. P. (1993). Identification of functionally related genes that stimulate early meiotic gene expression in yeast. Genetics 133, 67–77.
    46. Tilburn, J., Sanchez-Ferrero, J. C., Reoyo, E., Arst, H. N., Jr & Penalva, M. A. (2005). Mutational analysis of the pH signal transduction component PalC of Aspergillus nidulans supports distant similarity to BRO1 domain family members. Genetics 171, 393–401.
    47. Tréton, B., Blanchin-Roland, S., Lambert, M., Lepingle, A. & Gaillardin, C. (2000). Ambient pH signalling in ascomycetous yeasts involves homologues of the Aspergillus nidulans genes palF and paIH. Mol Gen Genet 263, 505–513.
    48. Uetz, P., Giot, L., Cagney, G., Mansfield, T. A., Judson, R. S., Knight, J. R., Lockshon, D., Narayan, V., Srinivasen, M. & other authors (2000). A comprehensive analysis of protein-protein interactions in Saccharomyces cerevisiae. Nature 403, 623–627.
    49. Villar, C. C., Kashleva, H., Nobile, C. J., Mitchell, A. P. & Dongari-Bagtzoglou, A. (2007). Mucosal tissue invasion by Candida albicans is associated with E-cadherin degradation, mediated by transcription factor Rim101p and protease Sap5p. Infect Immun 75, 2126–2135.
    50. Vincent, O., Rainbow, L., Tilburn, J., Arst, H. N., Jr & Penalva, M. A. (2003). YPXL/I is a protein interaction motif recognized by aspergillus PalA and its human homologue, AIP1/Alix. Mol Cell Biol 23, 1647–1655.
    51. Xu, W. & Mitchell, A. P. (2001). Yeast PalA/AIP1/Alix homolog Rim20p associates with a PEST-like region and is required for its proteolytic cleavage. J Bacteriol 183, 6917–6923.
    52. Xu, W., Smith, F. J., Jr, Subaran, R. & Mitchell, A. P. (2004). Multivesicular body-ESCRT components function in pH response regulation in Saccharomyces cerevisiae and Candida albicans. Mol Biol Cell 15, 5528–5537.