OUP user menu

The RIM101 pathway has a role in Saccharomyces cerevisiae adaptive response and resistance to propionic acid and other weak acids

Nuno P. Mira, Artur B. Lourenço, Alexandra R. Fernandes, Jorg D. Becker, Isabel Sá-Correia
DOI: http://dx.doi.org/10.1111/j.1567-1364.2008.00473.x 202-216 First published online: 1 March 2009


The physiological function of the Saccharomyces cerevisiae RIM101 signaling pathway is extended in this study beyond alkaline pH-induced responses. The transcription factor Rim101p is demonstrated to be required for maximal tolerance to weak acid-induced stress, at pH 4.0, but does not exert protection against low pH itself (range 4.5–2.5), when a strong acid is used as the acidulant. The Rim101p-dependent alterations of the yeast transcriptome following exposure to propionic acid stress (at pH 4.0) include genes of the previously described Rim101p regulon but also new target genes, in particular KNH1, involved in cell wall β-1,6-glucan synthesis and the uncharacterized ORF YIL029c, both required for maximal propionic acid resistance. Clustering of the genes that provide resistance to propionic acid reveals the enrichment of those involved in protein catabolism through the multivesicular body pathway and in the homeostasis of internal pH and vacuolar function. The analysis of the network of interactions established among all the identified propionic acid resistance determinants shows an enrichment of interactions around the RIM101 gene and highlights the role of proteins involved in Rim101p proteolytic processing. RIM101 expression is shown to be required to counteract propionic acid-induced cytosolic acidification and for proper vacuolar acidification and cell wall structure, these having positive implications for a robust adaptive response and resistance to stress promoted by this food preservative.

  • Rim101p
  • weak acids
  • response to stress
  • internal pH homeostasis
  • cell wall
  • yeast


The ability of yeasts and other fungi to respond to the alkalinization of the surrounding medium is mediated by a complex interplay of signalling mechanisms involving the RIM101 pathway and the calcineurine and SNF1 pathways (Lamb, 2001; Penalva & Arst Jr, 2002; Serrano, 2002; Lamb & Mitchell, 2003; Viladevall, 2004; Platara, 2006; Ruiz, 2008). The C2H2 zinc finger transcription factor Rim101p was first identified as a positive regulator of the meiotic gene program, sporulation and invasive growth (Su & Mitchell, 1993; Li & Mitchell, 1997; Lamb, 2001). The ability of Rim101p to promote changes in gene expression is controlled through the proteolytic removal of the C-terminal domain and requires the activity of several upstream products. These include the calpain-like protease Rim13p and the protease scaffold Rim20p, the arrestin-related Rim8p and the transmembrane proteins Rim21p and Rim9p (Li & Mitchell, 1997; Penalva & Arst Jr, 2002; Herranz, 2005; Calcagno-Pizarelli, 2007). More recently, DFG16 and YGR122w genes, encoding poorly characterized plasma membrane proteins, were also implicated in Rim101p proteolytic processing (Barwell, 2005; Rothfels, 2005). Thirty-five genes whose transcription is Rim101p dependent in Saccharomyces cerevisiae (17 upregulated genes and 18 downregulated genes) were identified (Lamb & Mitchell, 2003). These genes have known or predicted roles in cell wall maintenance and organization, in iron uptake or are uncharacterized putative membrane proteins (Lamb, 2001; Lamb & Mitchell, 2003). It was also demonstrated that the known binding site for the Rim101p homologue PacC of Aspergillus nidulans, TGCCAAG, occurs in the promoter region of most of the upregulated genes in Δrim101 but not in the downregulated gene promoters (Lamb & Mitchell, 2003). Rim101p acts directly to cause repression at the promoters of seven of the upregulated genes in a Δrim101 mutant background (Lamb & Mitchell, 2003), in particular of NRG1 and SMP1, encoding two transcription repressors. The positive effect of Rim101p in S. cerevisiae transcription is largely due to the direct repression of Nrg1p and Smp1p (Lamb & Mitchell, 2003). The biological function of S. cerevisiae Rim101p may differ from the function of its homologues in other fungi because most of the Rim101p/PacC homologues act as direct activators of alkaline pH-induced genes (Penalva & Arst Jr, 2002; Lamb & Mitchell, 2003). The deletion of RIM101 in S. cerevisiae also leads to susceptibility phenotypes toward sodium and lithium ions, to the antifungal agent fluconazole and to poor growth at low temperatures (Su & Mitchell, 1993; Lamb & Mitchell, 2003; Parsons, 2003), reinforcing the idea that, in this species, RIM101 may have other roles beyond alkaline pH-induced responses.

In the present work, it is demonstrated that the RIM101 pathway is required for S. cerevisiae adaptation and resistance to weak acids (at constant pH 4.0). The molecular mechanisms underlying Rim101p-mediated adaptation to propionic acid were further examined using genome-wide approaches. Propionic acid is largely used in the preservation of bakery and fresh dairy products (Suhr & Nielsen, 2004); however, despite the presence of high concentrations of this or other weak acid food preservatives, some spoilage yeasts are able to resume growth after an initial extended lag phase period (Piper, 2001; Suhr & Nielsen, 2004). The experimental model S. cerevisiae has been widely used to elucidate the molecular mechanisms behind resistance to weak acids in fungi, a knowledge essential for the development of suitable preservation strategies. The antimicrobial activity of weak acids at low pH relies on the effects of the undissociated acid form that may reach values close to the total concentration, depending on medium pH and the weak acid pKa (Piper, 2001; Suhr & Nielsen, 2004). Undissociated forms of many weak acids are liposoluble and may stimulate the passive influx of H+ into the cell by increasing plasma membrane permeability and affecting its biological function, in particular solute transport (Piper, 2001; Teixeira, 2007). The dissociation in the near-neutral cytoplasm of the permeant acid form leads to an additional acidification of the cell interior and to the accumulation of the counter ion. The acidification of the cytosol may lead to the inhibition of critical metabolic pathways (Krebs, 1983; Holyoak, 1997). The membrane effects of weak acids are particularly deleterious in the case of the more lipophilic compounds (Piper, 2001; Teixeira, 2007). Because, in general, weak acids are largely ineffective at pH values above 5.5, at which they are present as the nonliposoluble anionic form, it is frequently difficult to separate the direct effect of the presence of the antimicrobial form and the stress caused by the low pH value itself, when the external medium is acidified below the pKa value.

Saccharomyces cerevisiae adaptation and resistance to weak acids involve at least three regulatory systems mediated by: (1) the transcription factors Msn2p/Msn4p (Schüller, 2004), involved in the general stress response, (2) War1p, required for weak acid-induced transcriptional activation of PDR12, encoding a plasma membrane ABC efflux pump presumably involved in the efflux of benzoate and sorbate (Piper, 2001; Schüller, 2004), and (3) Haa1p, required for adaptation and resistance especially to the more hydrophilic acetic and propionic acids (Fernandes, 2005). In the present work, we show evidences indicating that the RIM101 pathway is also required for the effective adaptation and resistance of S. cerevisiae to weak acids. Based on the screen of the yeast disruptome for propionic acid susceptibility phenotypes, a number of putative biological targets and processes behind the Rim101p-mediated adaptive response to propionic acid were hypothesized and tested.

Materials and methods

Strains and growth media

The parental strain S. cerevisiae BY4741 (MATa, his3Δ1, leu2Δ0, met15Δ0 and ura3Δ0) and the Euroscarf collection of derived mutant strains, with all nonessential ORFs individually deleted using the KanMX cassette (http://web.uni-frankfurt.de/fb15/mikro/euroscarf/), were used. Cells were batch cultured at 30 °C, with orbital agitation (250 r.p.m.), in MM4 liquid medium that contains, per liter: 1.7 g yeast nitrogen base without amino acids or NH4+ (Difco Laboratories, Detroit, MI), 20 g glucose, 2.65 g (NH4)2SO4, 20 mg methionine, 20 mg histidine, 60 mg leucine, 20 mg uracil, 40 mg tryptophan and 30 mg lysine, all from Sigma (Spain). Yeast peptone dextrose (YPD) medium [per liter, 20 g glucose, 20 g Bacto peptone (Difco) and 10 g yeast extract (Difco)], agarized with 20 g L−1 agar (Iberagar S.A., Barreiro, Portugal), was used to assess the viable cell concentration. Iron-depleted growth medium was obtained by supplementing YPD medium with 100 μM of the iron chelator bathophenanthroline sodium disulfonate (BPS) (Sigma).

Genome-wide screening for propionic acid susceptibility and data analysis

To screen the entire Euroscarf deletion mutant collection for sensitivity to propionic acid, the different strains were grown in 96-well plates in MM4 medium, at pH 4.0, for 24 h. Using a 96-pin replica platter, these cells were spotted onto the surface of MM4 (pH 4.0) agarized medium either or not supplemented with propionic acid (final concentration of 18 or 20 mM). Susceptibility phenotypes were registered after incubation at 30 °C for 2–3 days, depending on the severity of growth inhibition. The eventual over- or under-representation of specific terms related to the physiological function of genes found to be required for propionic acid resistance among our data set was determined using GOToolBox (Martin, 2004). The Fischer exact test was used to correct the data, and enrichment was considered for P-values below 0.01. The interaction networks were obtained using string software (http://string.embl.de/), which takes into account the physical and functional associations between yeast proteins.

Other susceptibility assays

The susceptibility of the parental strain BY4741 and the deletion mutant Δrim101 to weak acids was assessed by comparing the growth curves of both strains in MM4 medium (acidified to pH 4.0 with HCl) or in this same medium supplemented with 60 mM acetic acid, 20 mM propionic acid, 8 mM butyric acid or 0.9 mM benzoic acid (K+ salt) at a constant final pH of 4.0. For this, the acids were dissolved in water and the pH of the final stock solution was adjusted to 4.0 with HCl or with NaOH, in the case of benzoic K+ salt. The acidification of MM4 medium, from pH 4.5 to 2.5, was carried out using HCl as the acidulant. Cell growth was followed by measuring culture OD600 nm during batch cultivation. Cell viability was followed by determining the number of CFU onto YPD solid medium. Colonies were counted after 2 days of incubation at 30 °C.

Spot assays of parental strain or Δrim101 cells harboring the recombinant plasmid pSS179RIM101 (Su & Mitchell, 1993), or the cloning vector alone, were performed by growing the cells in MM4 solid medium at pH 4.5 (lacking uracil to assure plasmid maintenance) supplemented with propionic acid in inhibitory concentrations (ranging from 8 to 9 mM) or in this same growth medium alkalinized to pH 7.3. The cell suspensions used to inoculate the agar plates for complementation experiments were harvested in the mid-exponential phase (OD600 nm=0.2±0.02) and then diluted to an initial suspension of standardized OD600 nm=0.05±0.005.

The effect of RIM101 expression in yeast growth in the iron-depleted medium was assessed in YPD medium or in this same standard medium supplemented with 100 μM of the iron chelator BPS. Cell suspensions used to prepare the spot assays were grown in YPD until the mid-exponential phase (OD600 nm=0.4±0.05) and diluted to a standardized OD600 nm=0.05±0.005. These cell suspensions and two subsequent dilutions (1 : 5 and 1 : 10) were applied as spots (4 μL) onto the surface of iron-depleted growth media and incubated at 30 °C for 3–5 days, depending on the severity of growth inhibition.

Transcriptomic analysis of the effect of RIM101 expression in yeast response to propionic acid

The transcriptomic analysis of the yeast response to propionic acid was carried out using cells of S. cerevisiae BY4741 and BY4741_Δrim101 previously grown in MM4 (pH 4.0) until the mid-exponential phase (standard OD600 nm=0.8) and then reinoculated into fresh medium either or not supplemented with 20 mM of propionic acid (at pH 4.0). After 45 min of incubation in the absence (samples A) or the presence of the weak acid (samples B), cells were harvested, immediately frozen in liquid nitrogen and kept at −80 °C until RNA extraction. Total RNA extraction was performed according to the hot phenol method. RNA was processed for use on Affymetrix (Santa Clara, CA) Yeast Genome S98 GeneChip arrays using the manufacturer's standard protocols as described before (Teixeira, 2006a). To ensure the reliability of the analysis, each GeneChip experiment was performed with biological replicates and the replicate data for the same sample were weighted gene-wise using inverse squared SE as weights. Moreover, normalized CEL intensities of a total of 10 arrays were used to obtain model-based gene expression indices based on a Perfect Match220 only model (Li & Hung Wong, 2001). Eight arrays, corresponding to two replicates of each condition under study, were analyzed. The remaining two arrays were only used to build the referred model. Only genes called Present in at least one of the compared arrays were kept for downstream analysis. Finally, all genes compared were considered to be differentially expressed if they were called Present in at least one of the arrays and if the 90% lower confidence bound of the fold change between the experiment and the baseline was >1.2. Based on these criteria, and in general, only median transcription changes >2.0 or <2.0 were considered as alterations to the transcriptome in response to propionic acid. The description of gene function is based on the information available in SGD (http://www.yeastgenome.org).

Cytosolic and vacuolar pH values

Cytosolic (pHc) and vacuolar (pHv) pH values were assessed by fluorescence microscopy using the pH-sensing fluorescent probes 5-(and 6-)carboxyfluorescein diacetate, succinimidyl-ester [5(6)-CFDA, SE] and 5-(and 6-)carboxy-2,7-dichlorofluorescein diacetate, succinimidyl-ester [5(6)-CDCFDA, SE], respectively. Both techniques make use of membrane-permeant fluorescent probes, that accumulate in the yeast cytosol or vacuole and that undergo pH-dependent changes, in both its absorption and its fluorescence emission spectra in a wide range of pH values (Roberts, 1991). This property has been exploited to successfully determine cytosolic and vacuolar pH in yeast cells (Roberts, 1991; Preston, 1997; Viegas, 1998; Vindelov & Arneborg, 2002; Fernandes, 2003) and also to screen for mutants defective in maintenance of a low vacuolar pH (Roberts, 1991; Preston, 1997). Briefly, to assess cytosolic pH, yeast cells were harvested by filtration, washed twice with CF buffer (50 mM glycine, 10 mM NaCl, 5 mM KCl and 1 mM MgCl2 in 40 mM Tris-100 mM MES, pH 4.0) and resuspended in 2 mL of CF buffer to an OD600 nm) of 10. The probe 5(6)-CFDA, SE was added to the cell suspension (final concentration of 40 μM) and the mixture was vortexed in one burst of 10 s and incubated for 20 min at 30 °C with orbital agitation (250 r.p.m.). After fluorescent labelling, cells were centrifuged at 5500 g for 5 min (at 4 °C), washed twice with CF buffer and finally resuspended in 2 mL of the same buffer. Cells were immediately examined with a Zeiss Axioplan microscope equipped with epifluorescence interference filters (Zeiss BP450-490 and Zeiss LP520). For pHv, the same protocol was used, with the exception that the harvested cells were resuspended in 2 mL of CF buffer supplemented with 2% glucose to assure proper vacuolar staining by the 5(6)-CDCFDA, SE probe (Preston, 1997; Viegas, 1998; Vindelov & Arneborg, 2002) (final concentration of 45 μM). In both cases, fluorescent emission was collected with a cooled CCD camera (Cool SNAPFX, Roper Scientific Photometrics). Bright-field images for pHc determination were obtained concurrently and recorded at 1-min intervals, each experiment being completed within 15 min. The images were analyzed using metamorph 3.5, and fluorescence images were background corrected using dark-current images. In each experiment, at least 500 individual cells were examined and the value of fluorescence intensity emitted by each cell was obtained pixel –by pixel in the region of interest to estimate pHc or pHv. The conversion of fluorescence intensity signals of 5(6)-CFDA, SE into pHc values was performed based on an in vivo calibration curve using cell suspensions grown in the absence of acid. Cells were loaded with the fluorescent probe as described above and incubated for 30 min with 0.5 mM carbonyl cyanide m-chlorophenylhydrazone to dissipate the plasma membrane pH gradient. Then, the external pH was adjusted (in the range 3.0–7.0) by the addition of HCl or NaOH. The average pHc values presented are means of, at least, three independent experiments. Fluorescence values registered in the vacuole were not converted into values of vacuolar pH.

β-1,3-glucanase sensitivity assay

To monitor the eventual structural changes occurring in yeast cell wall, a lyticase susceptibility assay was conducted essentially as described by Simões (2006). Cells of the parental strain and the deletion mutants tested were cultivated for 3 h in MM4 (pH 4.0) either or not supplemented with propionic acid (20 mM; pH 4.0). These cells were washed with distilled water and resuspended in 0.1 mM sodium phosphate buffer (pH 7.5). After addition of 20 μg mL−1 of lyticase (a β-1,3-glucanase from Arthrobacter luteus, Sigma), cell lysis was followed by the decrease of the initial OD600 nm (%) of the cell suspension.

Propionic acid-induced cell death experiments

To compare the kinetics of propionic acid-induced cell death of unadapted or adapted cell populations of the parental strain BY4741 and the deletion mutant Δrim101, cells were cultivated in MM4 (pH 4.0) medium or in this basal medium supplemented with 20 mM propionic acid (at pH 4.0), respectively, and harvested after 45 min of incubation. These cells were reinoculated (OD600 nm=0.01) in the same medium supplemented with a lethal concentration of 250 mM of propionic acid. The decrease of cell viability under propionic acid challenge was accompanied during 60 min. The viability values obtained are means of, at least, three independent death experiments that gave rise to identical patterns.


RIM101 expression is required for adaptation and resistance to weak acids

The comparison of the growth curves of S. cerevisiae BY4741 and its mutant devoid of the RIM101 gene in MM4 medium, either or not supplemented with the different weak acids tested (at pH 4.0), indicates that Rim101p is required to reduce the initial period of adaptation to acetic, propionic, butyric and benzoic acids and to decrease acid-induced inhibition of exponential growth (Fig. 1a, and results not shown). For the equivalent weak acid concentrations tested, leading to a similar lag phase duration of 18 h in the parental strain, the level of protection provided by RIM101 expression decreased with weak acid lipophilicity (Fig. 1a). Specifically, the adaptation period for the Δrim101 mutant increased to 35, 30 and 22 h for acetic, propionic and benzoic acids, respectively (Fig. 1a). Keeping in mind previous observations from other authors indicating that Rim101p confers protection to S. cerevisiae growth in alkaline environments, but not at low pH (Lamb, 2001), we tested the effect of RIM101 deletion when the growth medium was acidified down to pH 2.5, using the strong acid HCl as the acidulant, and found no detectable effect (Fig. 1b). Altogether, these results support the concept that RIM101 expression is required for protection against weak acid-induced stress but not against the acidification of the surrounding medium down to growth-inhibitory values, considering that a strong acid is used as an acidulant. Given that the higher protective role of Rim101p was registered under stress imposed by the more hydrophilic short-chain acetic and propionic acids (Fig. 1), our attention was focused on the more poorly characterized molecular mechanisms underlying yeast adaptation and resistance to propionic acid. The role of Rim101p in alleviating propionic acid susceptibility was confirmed by complementation of the Δrim101 susceptibility phenotype toward the weak acid by RIM101 expression from plasmid pSS179RIM101 (results not shown).


The expression of the RIM101 gene is required for yeast adaptation and resistance to weak acids, but not to low pH resulting from the acidification of the growth medium with a strong acid. (a) Growth curves of the parental strain BY4741 (◻) and the deletion mutant Δrim101 (○) in MM4 medium at pH 4.0 (control) or in the same basal medium supplemented with 60 mM acetic acid (log P=0.24), 20 mM propionic acid (log P=0.32) or 0.9 mM benzoic acid (log P=1.71). Log P is the logarithm of the partition coefficient octanol–water (P) of the acids tested. (b) Growth curves of the parental strain BY4741 (◻) and the Δrim101 deletion mutant (○) in MM4 growth medium acidified with HCl to the indicated pH.

Yeast disruptome screening: importance of vacuolar function and of the RIM101 pathway in propionic acid resistance

The yeast deletion mutant collection was screened to search for key players in providing resistance to propionic acid. The individual elimination of 254 genes/ORFs was found to lead to a susceptibility phenotype to inhibitory concentrations of propionic acid (Table 1). These genes were clustered according to their biological function, and the results indicate the enrichment in this dataset of genes involved in intracellular trafficking (particularly of those involved in vesicle-mediated transport), in protein catabolism through the multivesicular body (MVB) system, in generation of ATP and energy, in ergosterol biosynthesis, in post-translational protein modifications and in the regulation of intracellular pH (pHi) (Fig. 2a). Genes involved in the regulation of pHi are those essentially related to the assembly and function of vacuolar H+-ATPase (V-ATPase): VMA1, VMA2, VMA4, VMA5 and TFP3, encoding several subunits of the peripheral and membrane domain of this protein pump, and VPH2 and VMA22, encoding two membrane proteins involved in V-ATPase assembly. Vesicle-mediated transport class is essentially composed by vacuolar sorting proteins (VPS) that comprise multimeric complexes involved in the formation of vesicles responsible for intracellular trafficking between Golgi compartments, the endosome and the vacuole (VPS16, VAM16, PEP5 and PEP3, among others). A large number of other VPS genes form another enriched class. These genes encode several subunits of the endosomal sorting complexes required for intracellular transport (ESCRT) –VPS28, SRN2, STP22 (ESCRT-I), SNF8 and VPS36 (ESCRT-II) and VPS20, DID4 and VPS24 (ESCRT-III), which are involved in the catabolism of proteins through the MVB pathway (Katzmann, 2002). MVBs are part of the endosomal system of eukaryotic cells that are formed by invagination of the endosomal membrane to receive the transmembrane proteins that will be later sorted and degraded in the lumen of the lysosome or the vacuole. The enrichment of genes encoding proteins involved in post-translational modification of proteins was also registered. This class includes several members of the RIM signalling pathway involved in Rim101p proteolytic processing (RIM101, RIM8 and RIM13) and protein kinases, in particular, the HOG1 kinase that mediates the HOG signalling pathway also involved in resistance to acetic (Mollapour & Piper, 2006) and citric acids (Lawrence, 2004), and the SLT2 kinase involved in the regulation of the maintenance of cell wall integrity (Klis, 2006). The analysis of the complex network of interactions established among all the genes or ORFs required for propionic acid resistance reveals a main interaction network centered in the members of the RIM101 signalling pathway and several components of the ESCRT protein complexes (Fig. 2b). The connection between RIM101 signalling pathway and the MVB pathway is probably related to the fact that most ESCRT-defective mutants are also defective in Rim101p processing (Xu, 2004; Hayashi, 2005). The disruptome analysis showed an increased sensitivity to propionic acid of strains in which Rim101p processing is impaired, either in mutants deleted for genes of the RIM101 pathway –DFG16, RIM13, RIM8 and RIM101– or the MVB pathway –VPS20, SNF8, VPS36, STP22, SNF8, VPS28, VPS24 and DID4. The susceptibility phenotypes registered for members of the RIM101 signalling pathway through the test of the yeast disruptome were confirmed based on the comparison of their growth curves in liquid medium (results not shown). Moreover, the Δrim13, Δrim9, Δrim20 and Δrim21 mutants, devoid of genes encoding the other members of the RIM101 signalling pathway that could not be identified based on the disruptome analysis, are also susceptible to propionic acid, based on the comparison of the growth curves (results not shown). BRO1 and DOA4 encode proteins that also make part of this main interaction network and are involved in the coordination and progress of the MVB pathway. Although their elimination does not affect Rim101p processing (Xu & Mitchell, 2001; Rothfels, 2005), Δbro1 and Δdoa4 mutant strains are unable to mediate efficient Rim101p repression (Rothfels, 2005). Altogether, these results support the idea that propionic acid resistance requires the entire RIM101 pathway and a fully functional association between RIM101 and MVB pathways, involved in the regulation of Rim101p activity. The disruptome screenings reported in the literature for tolerance to sorbic acid (Mollapour, 2004) and citric acid (Lawrence, 2004) did not describe the involvement of Rim101p. However, according to our results, the effect of Rim101p expression in resistance to the more lipophilic acids is smaller or undetectable (Fig. 1, and results not shown) and may not be detected in those disruptome screenings.

View this table:

List of genes whose elimination leads to an increase in yeast susceptibility to propionic acid

functionPropionic acid susceptible mutants
Amino acids, vitamins and nucleotide metabolismADE1, ADE3, DHH1, MED2, NPP1, PDX3, RIB4, RNR4, SIN4, SPT8, THI6, THR1, TRP5, URA8
Carbohydrate metabolismALD6, GPH1, GUT1, KGD2, MLS1, OAR1, PDC1, PFK1, RPE1, SFA1, YIG1, YVH1, ZWF1
Lipid metabolismBTS1, CDC50, CRD1, ERG2, ERG24, ERG3, ERG4, ERG5, ERG6, IPK1, PSD1, RHR2, SUR1
DNA and RNA metabolismARP5, CTK2, DBP3, EAF7, HOP2, HPR1, IST3, LEA1, MRM1, MSS18, MSH3, NPL6, POP2, RAD23, REF2, SAP30, SDS3, SEH1, SHU2, TFB5
Metal homeostasisFRE8
Cell cycleBUB1, DOC1, EMI1, KTI11, PCL9, SPO22, SPO7
Cell morphogenesisBUD25, MDY2, NEM1, RVS161, SAC6
Cell wallBAG7, ECM1, END3, KNH1, SLT2, TIR2
Mitochondrial functionATP15, ATP4, ATP5, ATP7, COQ1, COQ4, COR1, COX19, COX20, COX9, FIS1, INH1, MDM38, MHR1, MIP1, MNE1, MRPL11, MRPL20, MRPL35, MRPL36, PET100, POR1, RSM27
Protein translationPET309, CBP6, EUG1, FMT1, GCN1, GCN2, MEF1, MRP21, MRPL8, MRP16, MSW1, PET112, RPL31A, RPL33B, RPL36B, RPL40A, RPL41B, RPP2A, RPS16a, RPS27B, RSM18, SLM5, TEF4, TIF3, ZUO1
Protein modification and signal transductionAKR1, ANP1, APE3, ATG15, BRE5, BUD32, CBP3, CTK3, DFG16, DOA1, DOA4, GEM1, HDA2, KEX2, KSS1, MSS2, NAT3, NPR1, PBS2, PKH2, PKP1, PMP1,PRB1, RDI1, RIM13, RIM8, RMD5, SEM1, SET7, SNF1, STE20, TPK2, UBP6, UFD2, VAM6, VID28, YAK1
Transcription factorsGZF3, HAP5, OPI1, RIM101, RPI1, RSF2, SFP1, WAR1, ZAP1
TransportAGP2, BAP2, MAL31, MUP1, NHA1, NPR2, PDR12, PET8, QDR1, RCY1, TPO1, WHI2, YIA6
Intracellular protein traffickingAPQ12, APL5, APM1, ATG20, ATG22, BRO1, DID4, GCS1, GET1, GET3, GOS1, PEP3, PEX21, RIC1, SNF8, SRN2, SSO2, STP22, TOM5, URE2, VPS16, VPS20, VPS24, VPS28, VPS33, VPS36, VPS4, VPS52
Vacuolar acidificationTFP1, TFP3, VMA5, VMA9, VPH2, VMA22, VPH1, VMA4
UnknownYBL012c, YBL071c-B, YBR056w, YCL007c, YCR025c, YCR045c, YCR050c, YCR090c, YCR100c, YDL063c, YDL173w, YDL180w, YDL183c, YDL241w, YDR008c, YDR065w, YDR114c, YDR115w, YDR230w, YDR537c, YGL024w, YGL041c-B, YGL188w-a, YGL188c-A, YGR050c, YGR150c, YHL026c, YPR175w-A, YIL028w, YIL029c, YIL064w, YIL077c, YIL092w, YIR016w, YJL120w, YJR018w, YJR079w, YKL096c-B, YKL118w, YLR414c, YMR098c, YMR242w-A, YNL296w, YOL014w, YOL050c, YOR200w, YOR331c, YPR099c
  • Genes are clustered according with their physiological function based on the information available in SGD (http://www.yeastgenome.org). The magnitude of the susceptibility phenotype of the corresponding deletion mutants is also indicated: nonunderlined genes indicate a strong susceptibility phenotype of the corresponding deletion mutant whereas underlined genes indicate a moderate susceptibility phenotype.


(a) Categorization, based on the biological process taxonomy of gene ontology, of the genes required for propionic acid resistance. Genes were clustered using GoToolBox and only the classes found to be statistically over-represented in our dataset (P-value below 0.01; the gene frequency within each class is indicated by black bars), compared with the frequency registered in the genome (white bars), were selected. (b) Main interaction network map of the determinants of resistance to propionic acid, focused on the RIM101 gene and based on the results of the screening of the yeast disruptome. The map was performed using the string software and shows the protein and/or the genetic interactions of Rim101p with other proteins that are also determinants of resistance to propionic acid. The existence of an interaction between two genes at a gene or at a protein level is represented by a connection between two nodes, and the major functional groups of genes are also indicated. Genes whose products affect Rim101p processing are highlighted in gray and genes whose expression was found to be regulated by Rim101p, based on our microarray analysis, are underlined.

Rim101p-dependent alterations of the yeast transcriptome in the presence of propionic acid

To identify the Rim101p-dependent alterations in genome-wide expression occurring in propionic acid-stressed cells, the transcriptomes of the parental strain and the Δrim101 mutant were compared during the early response to propionic acid stress and in cells cultivated in the absence of the acid (Fig. 3). Weak acid-stressed cells used for transcriptomic analysis were harvested after 45 min of sudden exposure to propionic acid (Fig. 3). Under the experimental conditions used in this study, this corresponds to the early cell response to the acid. In fact, in the presence of the propionic acid, a latency period of 4.5 or 6 h was observed for the parental and the Δrim101 strains, respectively, while the reinoculation of the unsupplemented medium with the same inoculum led to no detectable growth delay (Fig. 3). During the propionic acid adaptation period, the concentration of the viable cell population was maintained or slightly reduced in the case of the Δrim101 population (Fig. 3).


Growth curves of Saccharomyces cerevisiae BY4741 (▪, ◻) and the deletion mutant Δrim101 (●, ○) in MM4 (pH 4.0) (dark symbols) or in this same basal medium supplemented with 20 mM propionic acid (pH 4.0) (open symbols). The arrows indicate the times of cultivation at which cell samples were harvested to compare their transcriptomes. Growth was followed by measuring culture OD600 nm and the concentration of viable cells, assessed as the number of CFU per milliliter of cell culture (CFU mL−1).

In the absence of propionic acid stress, the deletion of RIM101 was found to affect the transcription levels of 32 genes by a ratio above or below twofold (Table 2). Several of these genes had already been described as being dependent on RIM101 expression in cells grown in YPD growth medium at a pH close to neutrality (Lamb & Mitchell, 2003). Nevertheless, under the experimental conditions used in our study, a set of 22 new Rim101p-regulated genes has emerged: ZPS1, YOL014w, YLL053c, YKR075c, YIL029c, YER067w, SUL1, SRL3, SMF1, SGE1, MSN4, MNN4, MDH2, KNH1, INO1, HPF1, DUR3, HSP26, AQY2, AFR1, JID1 and YHL044w (Table 2). The Rim101p-dependent genes encode proteins involved in cell wall metabolism and iron homeostasis and transcription factors or proteins of unknown or uncharacterized function (Table 2). The putative Rim101p-binding site TGCCAAG (Lamb & Mitchell, 2003) occurs in the promoter region of 75% of the upregulated genes in the Δrim101 background, but in only one of the downregulated genes (results not shown), suggesting that Rim101p may directly repress the genes whose promoter region has this binding site. Moreover, consistent with the idea that in S. cerevisiae Rim101p functions essentially as a repressor, the positive effect of Rim101p over gene transcription being due to the direct repression of NRG1 and SMP1 expression (Lamb & Mitchell, 2003), 65% of the upregulated genes in Δrim101 cells are documented targets of Nrg1p or Smp1p (CWP1, FET4, SMF1, UTR2, YDR133c, YKR075c, ZPS1 and SGE1), as indicated by the transcription regulatory associations gathered in the YEASTRACT database (Teixeira, 2006b) (results not shown).

View this table:

Results of the comparative analysis of the transcriptome of the parental strain Saccharomyces cerevisiae BY4741 (wt) and the mutant BY4741_Δrim101rim), cultivated in MM4 medium (pH 4.0) (c) or in this medium supplemented with 20 mM propionic acid, at pH 4.0 (ac)

ORF or genefunctionEmbedded ImageEmbedded ImageAcid susceptibility
Downregulated genes in Δrim101Not induced by propionic acidFET4Plasma membrane Fe(II) low-affinity transporter5.75.3
ZPS1Putative GPI-anchored protein; transcription induced under zinc limitation and at alkaline pH4.56.3
YDR133cUnknown function4.55.2
YIL029cUnknown function3.53.9+++
ARN4Siderophore–iron chelate transporter3.43.4
CWP1Mannoprotein involved in cell wall organization2.83.9+
UTR2Putative chitin transglycosidase involved in cell wall maintenance2.82.8
COS8Subtelomeric protein; potential role in the unfolded protein response2.12.9
KNH1Involved in cell wall β-1,6 glucan synthesis2.12.7+++
SGE1Plasma membrane transporter MFS transporter conferring resistance to cationic dyes2.32.2
SRL3Unknown function2.12.6
MDH2Malate dehydrogenase involved in gluconeogenesis1.42.2
AFR1α factor pheromone receptor regulator1.42.2
INO1Inositol-1-phosphate synthase1.42.1
SMF1Divalent metal ion transporter with broad specificity1.42.1
YKR075cUnknown function; expression regulated by glucose1.42.0
Propionic acid-inducedYDL241wUnknown function5.72.9+
YOL014wUnknown function5.13.5+
JID1Probable Hsp40p cochaperone3.52.0
BAG7GTPase-activating protein involved in the stimulation of Rho1p2.51.5++
CIN5Transcription factor of the yAP family involved in drug resistance and salt tolerance2.31.5
YHL044wUnknown function2.12.0
ICS2Unknown function2.11.3
MSN4Transcription factor involved in stress response1.52.3
YPL014wUnknown function1.02.1
Upregulated genes in Δrim101Not induced by propionic acidDIT1Enzyme required for spore wall maturation0.50.5
MNN4Putative positive regulator of mannosylphosphate transferase0.50.5
DUR3Polyamine and urea plasma membrane transporter0.50.4
RIM8Uncharacterized protein required for Rim101p processing0.40.5+
YJR061wSimilar to Mnn4p0.30.4
YOR389wUnknown function0.30.3
HPF1Mannoprotein involved in cell wall organization0.30.3
YLL053cUnknown function0.20.2
AQY2Water channel0.20.2
SMP1Putative transcription factor similar to Rlm1p0.10.4
YPL277cUnknown function0.20.2
Propionic acid-inducedHSP26Small heat shock protein, expressed under stress0.40.6
NRG1Transcriptional repressor involved in response to alkaline pH and glucose catabolic repression0.40.5
  • Genes whose transcript levels in Δrim101 cells were altered over twofold and below 0.5-fold, compared with the parental strain, are listed as two different groups: the up- or the downregulated genes in Δrim101 cells. The susceptibility of the different deletion mutants toward propionic acid compared with parental strain susceptibility, assessed by screening the yeast disruptome, is also shown.

  • * Indicates that at least in one of the conditions tested the alteration of the transcript levels from the different genes is below the threshold values of twofold or 0.5-fold used.

  • –, susceptibility identical to the parental strain; +, ++, +++, increasing levels of susceptibility.

Remarkably, only a very limited number (6%) of the 191 genes whose transcript levels were altered during the early response to propionic acid in the parental strain (Supporting Information, Table S1) exhibit different transcript levels in the Δrim101 mutant (Table 2). Moreover, even in the absence of the weak acid, most of these genes showed different transcription levels in the parental strain and in the Δrim101 mutant (Table 2).

Rim101p-mediated increased resistance to propionic acid involves its role in cell wall assembly

Although Rim101p was found to affect the transcription level of several genes in yeast cells under propionic acid aggression (Table 2), only a small subset of these genes was confirmed to be required for yeast resistance to propionic acid, as shown by the chemical genomic profiling carried out complemented by the comparison of the growth curves in liquid media supplemented with the acid (Table 2 and results not shown). In particular, the expression of KNH1 and CWP1 genes, whose function is related to cell wall assembly and structure, was found to be required for propionic acid resistance. KNH1, encoding a protein involved in cell wall β-1,6-glucan synthesis, exerted the highest protective effect (Table 2 and Fig. 4a) while CWP1, encoding a cell wall mannoprotein linked to the β-1,3- and β-1,6-glucan heteropolymer, exerted a more moderate protection (Table 2 and Fig. 4a). Other genes whose expression affects the cell wall structure and that are more actively transcribed in the presence of Rim101p, specifically ZPS1 and UTR2, showed no detectable effect in propionic acid resistance (Table 2).


Effect of the expression of RIM101 and RIM101-target genes CWP1 and KNH1 on yeast resistance to propionic acid (a) and in conferring protection against lyticase activity (b). (a) The susceptibility of cells of the parental strain Saccharomyces cerevisiae BY4741 (◻, ▪) and of the deletion mutants Δrim101 (●, ○), Δcwp1 (▲, △) and Δknh1 (◆, ◊) to propionic acid was compared in MM4 (pH 4.0) (dark symbols) or in this same basal medium supplemented with 20 mM propionic acid (pH 4.0) (open symbols). (b) Susceptibility to lyticase activity was compared in the same strains using cultures grown in MM4 (at pH 4.0) until the mid exponential phase, which were then used to reinoculate this same medium (closed symbols) or this medium supplemented with 20 mM of propionic acid (after pH 4.0) (open symbols). After 3 h of incubation, the different cell populations were harvested, washed with water and resuspended in 0.1 M sodium phosphate buffer at pH 7.5. Following addition of 20 μg of lyticase (Sigma) per millilitre of this cell suspension, the decrease of the OD600 nm was measured periodically. Data are the means±SD of at least three independent experiments.

The role of Rim101p and of the above-mentioned Rim101p-target genes in yeast resistance to lyticase was assessed using a very simple but valuable assay in monitoring cell wall alterations in which the accessibility of this β-1,3-glucanase to the internal layer of β-1,3-glucan of the yeast cell wall is estimated (Shimoi, 1998; Simões, 2006). The results indicate that the expression of RIM101 and, less significantly, of CWP1, increases the resistance of unstressed cells against the action of lyticase (Fig. 4b). In agreement with previous observations that suggest a remodelling of the cell wall structure in response to acetic and benzoic acids (Simões, 2006), a remarkable increase of resistance to the lyticase activity of parental cells was registered following 3 h of incubation in the presence of propionic acid (Fig. 4b). The elimination of RIM101 moderately affected the resistance of the adapted cells to lyticase (Fig. 4b). Among the Rim101p-target genes identified in this study whose expression may affect cell wall assembly, only CWP1 expression exhibited a role similar to RIM101 expression in mediating propionic acid-increased resistance to lyticase (Fig. 4b). This observation is consistent with the protective role of CWP1 against the weak acid aggression and, with the exception of KNH1, with the undetectable effect of the other genes in providing resistance to the acid (Fig. 4a). Despite the very important role of KNH1 as a propionic acid resistance determinant, its elimination did not lead to higher susceptibility to lyticase. It is known that mutants affected in β-1,6-glucan biosynthesis compensate this loss by increasing β-1,3-glucan levels, thus increasing resistance to β-1,3-glucanases (Shahinian, 1998), a phenotype that was also observed in our study (Fig. 4b).

The absence of Rim101p affects the homeostasis of internal pH specially in the presence of propionic acid stress

Keeping in mind the strong susceptibility phenotypes toward propionic acid of mutants devoid of genes involved in V-ATPase function, the hypothesized role of Rim101p in the homeostasis of pHc and pHv values was tested in cells incubated for 45 min in the absence or the presence of propionic acid under conditions close to those used for microarray experiments. In the absence of propionic acid stress, the elimination of the RIM101 gene was found to slightly affect the pHc of cells incubated in MM4 medium, at pH 4.0; the average pHc was reduced from 6.5 in the parental strain to 6.3 in the Δrim101 mutant (Table 3). The normal acidification of the vacuole was also affected in the absence of Rim101p, as indicated by the increased fluorescence of 5(6)-CDCFDA, the SE probe used for specific vacuolar staining (Table 3). In the presence of propionic acid stress, the elimination of RIM101 affected the control of cytosolic pH more significantly leading to higher internal acidification; pHc was reduced to c. 5.7 in the parental strain, compared with 5.4 in the Δrim101 mutant (Table 3). The reduction of vacuolar pH registered in response to propionic acid in the parental strain was not observed in Δrim101 (Table 3). This reduction of vacuolar pH under weak acid stress was hypothesized before to be an adaptive response that helps the cell to control cytosolic acidification to more physiological values (Carmelo, 1997). This adaptive response was attributed to the putative activation of V-ATPase activity (Carmelo, 1997), which was more recently demonstrated to occur under stress imposed by the acid herbicide 2,4-D (Fernandes, 2003). The deficient acidification of the vacuole in cells lacking Rim101p was much more evident under propionic acid stress (Table 3). The assessment of pHv of cells of Δyil029c, Δyol014w, Δydl241w and Δbag7 deletion mutants, grown either in the presence or the absence of propionic acid, did not show any significant difference, compared with the parental strain vacuolar pH (results not shown). These mutants for genes of unknown or poorly characterized function were chosen because the expression of these genes was found to lead to higher tolerance to propionic acid and their transcription levels were higher in the parental strain compared with the Δrim101 mutant (Table 2). However, their postulated role in mediating Rim101p-dependent control of vacuolar acidification was not confirmed.

View this table:

Effect of RIM101 expression in the homeostasis of pHc and pHv

StrainPropionic acidpHcpHv (fluorescence) (arbitrary units)
Parental strain6.5 ± 0.06251.4 ± 16.0
+5.7 ± 0.09187.3 ± 15.0
Δrim1016.3 ± 0.1288.9 ± 5.0
+5.4 ± 0.1315.2 ± 19.5
  • Cells of the parental strain and the mutant Δrim101 were harvested after 45 min of growth in the absence (−) or the presence (+) of 20 mM propionic acid, and values of pHc and pHv were assessed through fluorescence microscopy. Values of fluorescence were converted into pHc values using an in vivo calibration curve prepared as described in Materials and methods. Results are means of at least three independent experiments.

Because yeast mutants with perturbed vacuolar acidification were found to exhibit a growth defect under iron limitation as a result of being unable to load copper into the catalytic center of the high-affinity iron transporter Fet3p (Davis-Kaplan, 2004), we have tested and confirmed that the elimination of RIM101 affects yeast growth in YPD medium supplemented with the iron chelator BPS (results not shown). However, iron supplementation (10 and 100 μM) of the growth medium did not improve yeast resistance to propionic acid or rescued the susceptibility phenotype of the Δrim101 mutant toward this weak acid (results not shown).

RIM101 expression reduces propionic acid-induced death

The results shown until now indicate that the elimination of RIM101 leads to a defect in cytosolic pH homeostasis and in proper vacuolar acidification and cell wall structure (Fig. 4 and Table 3), either in the presence or the absence of propionic acid-induced stress. This suggests that RIM101 expression may provide protection against propionic acid in cells that have not been previously adapted to the acid. Indeed, the expression of RIM101 in unadapted cells delayed the initiation of propionic acid-mediated exponential death induced by a lethal concentration of propionic acid (250 mM) (Fig. 5) and decreased the specific rate of exponential death induced by the acid. Previous exposure of the parental strain and the Δrim101 populations to a lower but inhibitory concentration of propionic acid (20 mM) drastically reduced acid-induced death (Fig. 5), as expected as a result of yeast adaptation to the acid. The deletion of RIM101 from these previously adapted cells slightly increased the initial loss of viability induced by propionic acid, presumably delaying the eventual appearance of the phase of rapid exponential death (Fig. 5). This putative protective effect may result from the above-described Rim101p-dependent adaptive mechanisms that are induced in response to propionic acid.


Rim101p provides protection against propionic acid-induced death. Death experiments were carried out using cells of the parental strain BY4741 (◻, ▪) and the deletion mutant Δrim101 (●, ○) grown for 45 min in MM4 (pH 4.0) (unadapted populations; closed symbols) or in this same basal medium supplemented with 20 mM of propionic acid (adapted populations; open symbols). These cells were harvested by filtration and reinoculated into fresh MM4 (pH 4.0) medium supplemented with a lethal concentration of propionic acid (250 mM), and the acid-induced death was followed based on the quantification of CFU.


In the present work, it was proved that the transcription factor Rim101p is required for the maximal tolerance of S. cerevisiae to weak acid-induced stress, at a constant pH (pH 4.0) below their pKa, but does not exert protection against growth inhibition by low pH itself (in the range of pH 4.5–2.5). Because it is difficult to separate the effect of the antimicrobial undissociated form of weak acids and the effect of low pH itself, these two independent effects are frequently confused in the literature. In this study, the antimicrobial effect of low pH was separated from the weak acid effect using the strong HCl acid as the acidulant.

The comparison of the transcriptomes of the parental strain and the mutant with the RIM101 gene deleted indicates that the transcription factor Rim101p is active not only at neutral–alkaline pH but also at acidic pH, as suggested before for its Candida albicans homologue (Li, 2004). Indeed, the transcription of genes such as NRG1, SMP1, RIM8, YJR061w, YOR389w and YPL277c, whose Rim101p-mediated repression is dependent on a processed active form of the transcription factor (Lamb & Mitchell, 2003), was found to be repressed by Rim101p in cells cultivated at pH 4.0 either in the presence or the absence of propionic acid stress. Moreover, the genes that are up or downregulated twofold or more in the Δrim101 mutant, at pH 4.0, largely overlap with those identified before as being Rim101-target genes in cells grown in YPD medium (presumably at a pH close to neutrality) (Lamb & Mitchell, 2003). The activation of Rim101p in S. cerevisiae occurs by proteolytic removal of an inhibitory C-terminal region (Li & Mitchell, 1997; Penalva & Arst Jr, 2002) and there is the idea that a processed form of Rim101p predominates at alkaline pHs while in acidic environments the full-length unprocessed form predominates. However, in C. albicans, Rim101p was recently found to undergo a novel processing event at acidic pHs (Li, 2004), even though Rim101p had no apparent role in gene expression at acidic pH (Bensen, 2004). However, Rim101p also regulates the expression of several genes that are not regulated by pH (Bensen, 2004) and it was suggested that the processed form of Rim101p under acidic conditions could be responsible for the regulation of these genes (Li, 2004). Although Rim101p processing has not been reported in S. cerevisiae or A. nidulans at acidic pH, the possibility that this event also occurs in these fungi has been considered (Li, 2004). This hypothesis is consistent with the results of the present work pointing out to the fact that Rim101p is active at acidic pH, thus promoting alkaline pH-independent processes. Susceptibility assays indicate that the genes involved in Rim101p processing (e.g. Rim9p, Rim13p, Rim20p, Rim21p and Rim8p) are required for maximal propionic acid resistance, involving for the first time Rim101p processing in increased tolerance to weak acid aggression.

Weak acid stress leads to internal acidification of yeast cells (Holyoak, 1997; Piper, 2001; Fernandes, 2003), and their ability to grow or maintain viability at a high weak acid concentration may also reflect their capacity to maintain control over their internal pH. Indeed, the proper functioning of yeast cells relies on maintenance of their pHi within relatively narrow limits, as large deviations from the normal pH severely inhibit metabolism as a result of suboptimal activities for cytosolic enzymes (Holyoak, 1997; Piper, 2001). Additionally, the maintenance of the electrochemical potential of plasma membrane is crucial to the secondary active transport of nutrients across the plasma membrane and for detoxification. The post-translational activation and the essential role of plasma membrane H+-ATPase (PM-ATPase) in the control of pHi homeostasis under weak acid stress in yeast are known (Carmelo, 1997; Holyoak, 1997; Viegas, 1998; Makrantoni, 2007). The deletion of RIM101 was found to have a negative effect over the control of the propionic acid-induced internal acidification. Among the other reasons discussed below, the decreased transcription in Δrim101 of SIA1 (1.9-fold, below the threshold level and thus not included in Table 2) may underlie this deficiency because it encodes a protein involved in glucose-induced activation of PM-ATPase activity (de la Fuente, 1997), whose function is crucial to counteract weak acid-induced intracellular acidification (Holyoak, 1997; Viegas, 1998). Our results also reinforce previous observations (Carmelo, 1997; Fernandes, 2003; Makrantoni, 2007), suggesting that the function of V-ATPase is also required for resistance to weak acid stress, presumably by partially controlling cytosolic pH values under weak acid stress by actively pumping H+s into the vacuole. The yeast vacuole was proposed before to play a role in the homeostasis of pHi (Carmelo, 1997). Vacuolar pH is maintained at mildly acidic pH around 6.0 when the cytoplasm is close to neutral pH, in unstressed cells grown at pH 4.5; when cells are incubated with weak acids, vacuolar pH decreases (Carmelo, 1997). The maintenance of the proton gradient across the vacuolar membrane may be due to the modulation of the vacuolar membrane potential resulting from the activation of V-ATPase or by alteration of its ion conductivity. The first hypothesis is in agreement with recent studies where the crucial role of yeast V-ATPase activity in mediating resistance to the food preservative sorbic acid (Makrantoni, 2007) and the weak acid herbicide 2,4-D (Fernandes, 2005) is highlighted.

In yeast, the V-ATPase acidifies the vacuole by driving the translocation of protons into the lumen. This acidification is crucial for many cellular processes besides cytosolic pH homeostasis, including endocytosis, targeting of newly synthesized lysosomal enzymes, protein processing and the coupled transport of small molecules (Forgac, 2007). The V-ATPase is also responsible for acidification of post-Golgi vesicular compartments (Forgac, 2007). MVB sorting requires a pH gradient across the endosomal membrane, and disruption of vacuolar acidification causes defective vacuolar sorting of proteins as well as the accumulation and missorting of precursor forms of various proteins (Supek, 1994; Perzov, 2000; Forgac, 2007). In particular, the amount of the major form of PM-ATPase, Pma1p, in the plasma membrane of mutants affected in V-ATPase function is markedly reduced and a large amount of a nonactive form of the protein is accumulated in the endoplasmic reticulum (ER), suggesting that the passage from ER to the Golgi is impaired in these mutants (Perzov, 2000). Consequently, V-ATPase function may also be crucial to maintain suitable levels of PM-ATPase activity to efficiently counteract the dissipation of plasma membrane proton gradient and propionic acid-induced acidification of the cytosol. This adaptive process is affected in the Δrim101 mutant and is consistent with the susceptibility phenotypes against propionic acid exhibited by mutants deleted for genes encoding subunits that compose the multimeric structure of V-ATPase (TFP1, TFP3, VMA5, VMA9, VPH2, VMA22 and VMA4). A failure to maintain the pH gradient across the endosomal membrane, coupled with the activation of the Rim101p pathway, was suggested before to represent the physiological significance of the existing link between the Rim101p pathway and MVB sorting (Xu, 2004; Hayashi, 2005). Another consequence of defective acidification is the failure to load the multicopper oxidase Fet3p with copper within a post-Golgi endosomal compartment, thus resulting in growth sensitivity under iron limitation (Radisky, 1997; Davis-Kaplan, 2004). This phenotype was also observed in S. cerevisiaeΔrim101 cells, as reported before for the C. albicansΔrim101 mutant (Bensen, 2004), suggesting that in this pathogenic yeast, vacuolar acidification may also be defective in the absence of RIM101 expression. Despite our efforts, the mechanism underlying the defective acidification of vacuolar pH in the Δrim101 mutant could not be attributed to any of the poorly characterized Rim101p targets identified in our study, YIL029c, YDL241w, YOL014w and BAG7, also implicated in propionic acid resistance. Moreover, according to our microarray data, Rim101p could only be implicated, directly or indirectly, in the transcriptional regulation of 1 (specifically of the RIB4 gene, whose transcription level was 1.8-fold higher in the parental strain compared with the Δrim101 mutant and therefore not listed in Table 2) of the 65 genes whose deletion leads to growth defects typical of vacuolar membrane ATPase mutants (Vma) (Sambade, 2005). These genes encode the structural subunits of V-ATPase, VPS and putative regulators of V-ATPase (Sambade, 2005). RIB4 encodes a 6,7-dimethyl-8-ribityllumazine synthase involved in riboflavin and flavin adenine dinucleotide (FAD) cofactor biosynthesis (García-Ramírez, 1995). Because of experimental difficulties in growing the auxotrophic Δrib4 mutant, we could not confirm the hypothesized involvement of this Rim101p target gene in the lack of proper vacuolar acidification in Δrim101 cells. However, the screening of the yeast disruptome indicated that RIB4 expression confers protection toward propionic acid. Remarkably, RIB4 interacts with the protein kinase Dbf2p, which was recently implicated in the modulation of V-ATPase activity (Graumann, 2004; Makrantoni, 2007). Moreover, RIB4 deletion implicates a reduction in FAD cofactor biosynthesis, which may affect ATP levels limiting the in vivo V-ATPase activity.

Direct experimental evidence assigning a role to the RIM101 pathway in the assembly of S. cerevisiae cell wall was obtained before (Castrejon, 2006). The present study implicates RIM101 expression in cell wall resistance to lyticase and in cell wall remodelling during adaptation to propionic acid. Two genes involved in cell wall assembly and organization, CWP1 and KNH1, exhibited reduced transcript levels in the absence of Rim101p being determinants of resistance to propionic acid, with KNH1 providing the highest protective effect. KNH1 is involved in β-1,6-glucan synthesis (Dijkgraaf, 1998), which is essential by interconnecting all the cell wall components into a strong lattice (Klis, 2006) and CWP1 encodes a cell wall mannoprotein, linked to β-1,3 and β-1,6-glucan heteropolymer through a phosphodiester bond (Kapteyn, 1997; Klis, 2006). The ORF YIL029c, another Rim101p target gene and a major determinant of resistance to propionic acid, belongs to a group of genes that were proposed to represent the signature of the cell wall compensatory mechanism after damage (Lagorce, 2003). Global expression analyses carried out in recent years have indicated that yeast cells respond to environmental stress by modifying cell wall organization. This adaptive response will limit the size and type of molecules that may penetrate the cell wall and come in contact with the plasma membrane, supporting the notion that cell wall porosity is rapidly and markedly reduced in response to stress. Cell wall remodelling may compensate for the damage produced in the cell wall and increase protection against diverse stresses (Kapteyn, 1997; Shimoi, 1998; Lagorce, 2003; Klis, 2006; Simões, 2006). It should be emphasized that none of the Rim101p dependent cell wall biosynthetic genes required for propionic acid resistance that were identified in this study is transcriptionally activated (equal or above twofold) in response to the acid. However, either in the presence or the absence of propionic acid stress, their transcript levels in the Δrim101 mutant were below the levels registered in the parental strain.

In summary, in the present study, we demonstrate that RIM101 expression is required for proper internal pH homeostasis, vacuolar acidification and cell wall structure, with positive implications for a robust adaptive response to weak acid stress in yeast. These uncovered physiological roles highlight the importance of this signalling pathway in S. cerevisiae response to weak acids. Remarkably, the fungicidal action of fluconazole in C. albicans is elicited by short-chain weak acids, in particular by acetic acid, under conditions where the fungicide alone is not effective (Moosa, 2004). Considering that the RIM101 pathway is required for C. albicans virulence in vivo (Davis, 2000), the indications emerging from our study may be extensive to this pathogenic yeast and may provide guidance for the development of more suitable treatment strategies.


Table S1. Up-regulated genes after 45 minutes of exposure of Saccharomyces cerevisiae BY4741 to 20 mM propionic acid obtained in the microarrays analysis carried out. Genes whose transcript levels were activated by more than 2-fold, compared with mRNA values registered in control cells, were selected and are presented. Gene function is also shown based on the information available at SGD ().


We thank Dr A.P. Mitchell (Department of Microbiology, University of Columbia,) for providing the pSS179RIM101 plasmid. This research was supported by FEDER, POCTI and PDCT Programmes and Fundação para a Ciência e a Tecnologia (FCT) [contracts: POCTI/AGR/45347/2002 and PDCT/BIO/56838/2004 and PhD fellowships to N.P.M. (SFRH/BD/17456/2004) and A.B.L. (SFRH/23437/2005)].


  • Editor: André Goffeau


View Abstract