The role of nutrient regulation and the Gpa2 protein in the mating pheromone response of C. albicans


*E-mail; Tel. (+1) 415 476 8783; Fax (+1) 415 502 4315.


Although traditionally classified as asexual, the human fungal pathogen Candida albicans can undergo highly efficient mating. A key component of this mating is the response to pheromone, which is mediated by a conserved kinase cascade that transduces the signal from the pheromone receptor to a transcriptional response in the nucleus. In this paper we show (i) that the detailed response of C. albicans to the alpha pheromone differs among clinical isolates, (ii) that the response depends critically on nutritional conditions, (iii) that the entire response is mediated by the Ste2 receptor, and (iv) that, in terms of genes induced, the response to alpha pheromone in C. albicans shows only marginal overlap with the response in Saccharomyces cerevisiae. We further investigated the nutritional control of pheromone induction and identify the GPA2 gene as a critical component. We found that Δgpa2/Δgpa2 mutants are hypersensitive to pheromone and, unlike wild-type strains, show efficient cell cycle arrest (including the formation of characteristic halos on solid medium) in response to mating pheromone. These results indicate that C. albicans, like several other fungal species but unlike S. cerevisiae, integrates signals from a nutrient-sensing pathway with those of the pheromone response MAP kinase pathway to generate the final transcriptional response.


Candida albicans and Saccharomyces cerevisiae are thought to have diverged from a common ancestor between 100 and 900 million years ago (Gargas et al., 1995; Hedges et al., 2004). C. albicans can grow as both a commensal organism and as a pathogen in humans, and is the most common fungal pathogen isolated from humans (Edmond et al., 1999). In contrast, most strains of S. cerevisiae do not cause disease in mammals and are found in soil, on the surface of fruit, and in the wine and bread industries.

Although originally identified as an asexual organism, C. albicans has now been shown to mate, both in the laboratory and in a mammalian host (Hull et al., 2000; Magee and Magee, 2000). Mating in C. albicans differs from that in other fungi in that it requires that a and α cells undergo an epigenetic switch from the white phase, the predominant form in which cells are round and smooth, to the opaque phase, in which cells are elliptical and exhibit cell surface irregularities (Miller and Johnson, 2002). White-opaque phase switching is linked with the differential expression of approximately 400 genes; although several mating-specific genes are upregulated in the opaque phase, most of the genes differentially regulated in the white-to-opaque transition do not seem to be involved in mating (Lan et al., 2002; Tsong et al., 2003). It seems likely that white–opaque switching links mating to important features of pathogenesis in a mammalian host (Gow, 2002; Johnson, 2003; Magee and Magee, 2004; Soll, 2004).

Several laboratories have previously identified the MFα gene, which encodes the alpha mating pheromone in C. albicans (Lan et al., 2002; Bennett et al., 2003; Lockhart et al., 2003a; Panwar et al., 2003). Alpha pheromone is secreted by α opaque cells and a opaque cells respond to it in the early stages of mating (Miller and Johnson, 2002; Lockhart et al., 2003b). Processing of the precursor alpha pheromone is predicted to generate both 13-mer and 14-mer peptides (MF13 and MF14), both of which have been shown to induce morphological changes in opaque a-type cells (Panwar et al., 2003). Transcriptional profiling identified 65 genes that were upregulated in response to MF13 in opaque a cells (Bennett et al., 2003), and northern analysis also identified several genes that were up- or downregulated by pheromone (Lockhart et al., 2003b). More recently, transcriptional profiling of cells in mating mixes of opaque a and α cells was used to identify additional genes regulated during mating (Zhao et al., 2005). These experiments, together with genetic studies (Chen et al., 2002; Magee et al., 2002), demonstrated that C. albicans utilizes a signalling pathway that is conserved in S. cerevisiae, the MAP kinase cascade, to mediatethe response to pheromone. Although the signalling pathways are highly conserved between S. cerevisiae and C. albicans, the transcriptional outputs are markedly different (Bennett et al., 2003). For example, most genes induced by pheromone in C. albicans are not induced in S. cerevisiae; these include a number of genes previously implicated in filamentous growth and virulence, suggesting a close link between mating, filamentous growth, and virulence.

In this paper, we have analysed the response of C. albicansa cells to pheromone (both MF13 and MF14 peptides) in several different laboratory media, in both white and opaque phases, and we also examined the pheromone response in different clinical isolates of C. albicans. As judged both by the morphological and transcriptional response to pheromone, all of these variations influenced the pheromone response.

We further investigated the effects of growth media on the pheromone response. The strongest response – as judged by both the number of genes significantly induced and repressed and by the magnitude of these effects – was observed when nutrients were limited. Previous work in C. albicans has identified the G protein-coupled receptor, Gpr1, and the G protein, Gpa2, as nutritional sensors regulating filamentous growth in a/α cells (Sanchez-Martinez and Perez-Martin, 2002; Miwa et al., 2004; Maidan et al., 2005). To test if the Gpr1/Gpa2 complex contributes to the nutrient dependence of pheromone signalling in C. albicans, we constructed a gpa2/gpa2 mutant and analysed its response to pheromone. The Δgpa2 mutant produced an enhanced pheromone response in C. albicans in nutrient-rich medium, particularly in YPD (yeast extract/peptone + dextrose). This enhanced response was observed in several ways; the most dramatic being the formation of distinct ‘halos’, indicative of tight cell cycle arrest, observed when mating pheromone was spotted onto plates containing the Δgpa2 strain. These results show that, like several other fungi but unlike S. cerevisiae, nutrients regulate the pheromone response in C. albicans, and that this regulation occurs through integration of signals from a nutrient-sensing pathway and the pheromone signalling pathway.


Pheromone response in C. albicans is dependent on growth medium and strain background

Previously, we showed that 65 genes are induced more than threefold in C. albicans opaque a cells in response to the 13mer alpha pheromone, MF13 (Bennett et al., 2003). These results were based on experiments carried out with one isolate (derived from SC5314) in Synthetic Complete medium supplemented with glucose (SCD). We subsequently found, however, that C. albicans mating occurs at much higher frequency under other conditions, and that mating efficiency can be strongly influenced by the genetic backgrounds of the mating pairs (Bennett et al., 2005). We therefore set out to test whether the transcriptional response to alpha pheromone is influenced by culture medium, strain background, or both. As both 13mer and 14mer alpha pheromones were shown to have biological activity, we also compared their effects on the pheromone response pathway. Although only opaque cells of C. albicans are capable of high-frequency mating, we also examined whether white cells could respond to alpha pheromone. Finally, to test if the STE2 gene encodes the receptor for the entire alpha pheromone response, we examined the transcriptional response of C. albicansΔste2/Δste2 cells to pheromone.

The global transcriptional response to alpha pheromone was determined using microarrays that contained 11 325 spots representing 6550 nuclear genes, at least 95% of the estimated number of genes in the C. albicans genome. In each experiment, cells were treated with alpha pheromone dissolved in DMSO (final concentration of alpha factor was 10 μg ml−1) or were mock treated with DMSO. Unless stated otherwise, cells were collected 4 h after induction and the genes induced by alpha pheromone determined by microarray analysis, as described in Experimental procedures. The 4 h time point was chosen as this gave the maximal transcriptional response to pheromone, as indicated by previous studies (Bennett et al., 2003). The transcriptional response of a opaque cells to pheromone matches their morphological response, as mating projections begin to appear at 2 h and are prominent by 4 h (Bennett et al., 2003; Lockhart et al., 2003a; Panwar et al., 2003). Figure 1A shows genes that were induced in response to alpha pheromone under a number of experimental conditions, while Fig. 1B shows genes that were repressed by pheromone under the same conditions. Due to the large number of experimental variables tested (media, strain background, pheromone, white/opaque phase, etc.), not every microarray experiment was repeated multiple times; however, the key conclusions discussed in this paper were based on multiple experimental repeats. In addition, SAM (Significance Analysis of Microarrays) was used to provide statistical analysis for many of the experiments.

Figure 1.

Microarray analysis of the response of SC5314-derived opaque and white phase a/a cells to alpha pheromone. Opaque and white phase cells were grown in logarithmic phase at 23°C and treated with pheromone (10 μg ml−1) in DMSO or mock treated (see Methods and Materials). Cells were collected after 4 h incubation with pheromone and mRNA prepared as previously described (Bennett et al., 2003). Labelled cDNA was prepared and hybridized against DNA microarrays representing 6500 ORFs.
A. Genes induced by pheromone.
B. Genes repressed by pheromone.
Only those genes whose change in expression was significant by SAM are shown (SAM analysis performed for opaque cells in SpiderM medium conditions). Genes regulated by pheromone are summarized in Tables 1 and 2. All experiments are pheromone-treated cultures with the exception of lane 13, which is a direct hybridization of cells grown in SpiderM medium (cDNA labelled with red Cy5) versus SCD medium (cDNA labelled with green Cy3). SC5314 (opaque) refers to strain RBY731, SC5314 (white) refers to strain RBY717, and SC5314 (Δste2ste2) refers to strain RBY941 (see Table 3).

Growth medium affects the response of opaque cells to alpha pheromone

Although many experiments are represented in Fig. 1, nutrient regulation of the pheromone response pathway is most evident in the comparison between SpiderM medium and other types of growth media. SpiderM medium (1% nutrient broth, 0.4% KH2PO4, 1% mannitol, pH 7.2) has previously been shown to induce filamentous growth in white-phase cells of C. albicans (Liu et al., 1994) and also to promote efficient mating between opaque a and alpha cells (Bennett et al., 2005). In contrast to SCD medium (synthetic complete + dextrose) and YPD medium (yeast extract/peptone + dextrose), C. albicans cells grown in SpiderM are under nutrient-limiting conditions, and this is presumed to be the signal for induction of hyphae in white-phase cells (Liu et al., 1994). For our microarray experiments, cells were initially grown in SCD medium, washed and resuspended in SpiderM medium immediately before pheromone addition (see Experimental procedures). Figure 1A, lanes 8–11 show that the transcriptional response to MF13 in SpiderM medium was significantly different from that in either SCD or YPD media (Fig. 1A, lanes 1–6). In particular, a large number of genes were induced in response to alpha pheromone in SpiderM medium that were not induced in either SCD or YPD media. SAM was used as a statistical method of identifying the genes whose expression change was significant, based on the amplitude of the expression change and the reproducibility of the change in multiple experiments (Tusher et al., 2001). Using SAM (with a median false discovery rate of zero), we identified 95 genes that were upregulated by pheromone in SCD medium, and 15 genes that were downregulated by pheromone. In comparison, the response in SpiderM medium revealed significant induction of 330 genes and repression of 194 genes (see Tables S1 and S2; median false discovery rate zero). Many of these 524 genes show small, but statistically significant, changes in gene expression. To concentrate our further analysis on those genes showing the largest effects, we have arbitrarily narrowed the gene list to those whose expression level changed by more than threefold, giving a set of 144 genes induced, and 66 genes repressed by pheromone in SpiderM medium (Tables 1 and 2).

Table 1.  Genes induced in response to alpha pheromone in SpiderM medium.
Nameorf19orf6InductionFunctionS. cerevisiae
  1. Table of significantly induced genes identified by SAM analysis and whose expression changed by more than threefold. The full set of genes that passed SAM analysis are available in Table S1. Analysis was performed on four repeat experiments examining the response of opaque a cells to MF13 pheromone in SpiderM medium.

orf6.3434orf19.9386orf6.3434294Putative lectin binding domain 
MFALPHAorf19.11961orf6.4306212Alpha pheromoneNo
FUS1orf19.8748orf6.3131195Required for cell fusion in matingInduced
orf6.7473orf19.9629orf6.7473140Putative aspartyl proteaseNo
PRM1orf19.8286orf6.656282Plasma membrane fusionInduced
FIG1orf19.7782orf6.347671Involved in cell polarization and cell fusionInduced
DG1106orf19.11282orf6.291962Similarity to dictyostelium protein required for development 
SST2orf19.11698orf6.276758Adaptation to pheromoneInduced
HWP2orf19.10888orf6.293353Hwp1-related cell wall protein 
POLorf19.9710orf6.131242Reverse transcriptase 
SAP4orf19.13139orf6.380340Secreted aspartyl proteaseNo
HOL1orf19.9156orf6.533739Member of multidrug resistance protein familyNo
SAP7orf19.8376orf6.363538Secreted aspartyl proteaseNo
FGR23orf19.9183orf6.55538Extracellular alpha 1, 4 glucan 
HST6orf19.7440orf6.860029ATP dependent permeaseInduced
HWP1orf19.8899orf6.488328Hyphal wall protein 
orf6.333orf19.380orf6.33328Leucine rich repeat gene family protein 
STE2orf19.8315orf6.401223Alpha factor pheromone receptorInduced
BUD16orf19.9387orf6.343522Bud-site selectionNo
RAM1orf19.5046orf6.332320Farnesyltransferase that prenylates a-factorNo
AMAN6orf19.9470orf6.80220Similar to alpha-1,6-mannanase 
RHR2orf19.12892orf6.867318Glycerol-1-phosphatase activity, response to osmotic stressNo
RBT1orf19.1327orf6.488917Hwp1-related cell wall protein 
CEK2orf19.8091orf6.285416MAP kinase in mating cascadeInduced
ECE1orf19.10882orf6.288616Filament specific cell wall protein 
orf6.7686orf19.12721orf6.768615Weak homology with protein involved in transport from ER to GolgiNo
SAP6orf19.5542orf6.362413Secreted aspartyl proteaseNo
RBT4orf19.13583orf6.53712Pathogen-related protein, repressed by Tup1 
SAP5orf19.13032orf6.442712Secreted aspartyl proteaseNo
KRE1orf19.11855orf6.454212Cell wall glycoproteinInduced
RIM2orf19.4499orf6.419111Mitochondrial carrier family, required for respirationNo
ABP2orf19.11013orf6.150111Fatty acid synthesisNo
orf6.7100orf19.9048orf6.710010Unknown, homology to dehydrogenases 
YHR048Worf19.10898orf6.43369Member of multidrug resistance protein familyNo
ABPDorf19.7938orf6.44449Similar to Myo1, actin-binding protein 
orf6.4778orf19.13064orf6.47789Unknown, leucine rich repeat family 
PIRIN-1orf19.2467orf6.49688Novel, conserved nuclear protein from humans 
YGR287Corf19.11465orf6.56588Alpha glucosidaseNo
SLU7orf19.14119orf6.880483′ Splice site selectionNo
MNN4orf19.10474orf6.66598Transfer of mannosylphosphate to oligosaccharidesNo
RSN1orf19.12268orf6.52117Membrane protein of unknown functionNo
YNL134Corf19.9932orf6.36777Alcohol dehydrogenaseInduced
SHA3orf19.11153orf6.64807Putative serine/threonine kinase; adaptation to glucoseRepressed
YNL247Worf19.12397orf6.59117Cysteine metabolismNo
GPX3orf19.7730orf6.15587Thiol peroxidaseInduced
RAM2orf19.12280orf6.52236Farnesyltransferase that prenylates a-factorNo
YEA4orf19.8962orf6.41686Transporter required for cell wall chitin synthesisInduced
YJR054Worf19.13916orf6.75516Vacuole proteinNo
PLB3orf19.8309orf6.19856Phospholipase BInduced
EXG1orf19.2990orf6.19826Glucan 1,3-beta-glucosidase activityNo
MIH1orf19.3069orf6.15905Tyrosine phosphatase involved in cell cycle controlNo
YBR159Worf19.11340orf6.52495Microsomal beta-keto-reductaseNo
DAP1orf19.6867orf6.71315Haem binding protein; damage responseNo
NAB6orf19.12950orf6.46625Putative RNA binding proteinNo
SPO7orf19.14174orf6.71495Membrane protein required for sporulationNo
PCL1orf19.10172orf6.12584G1/S-specific cyclinPCL2 induced
CPH1orf19.4433orf6.6954Homologue of Ste12, transcription factorInduced
orf6.2403orf19.12042orf6.24034Putative transcription factor in S. pombe 
ORC4orf19.11697orf6.27664Subunit of origin replication complexNo
LEA1orf19.8845orf6.57274Splicing factorNo
CHS7orf19.9980orf6.36224Chitin biosynthesisNo
BNI1orf19.12393orf6.59074Nucleates formation of actin filaments, smoo tip formationNo
YBR042Corf19.137orf6.34774Phospholipid biosynthesisNo
CPP1orf19.12330orf6.73454Adaptation to pheromoneInduced
EMP46orf19.8350orf6.60124Coated vesicleInduced
YIL088Corf19.8524orf6.72724Amino acid transportNo
RAD9orf19.11751orf6.31114Checkpoint arrestRepressed
TCP1orf19.8031orf6.62634Chaperone component, maintenance of actin cytoskeletonNo
GFA1orf19.1618orf6.64544First step of chitin synthesisInduced
MSM1orf19.11432orf6.24174Methionine biosynthesisNo
CEK1/ERK1orf19.10404orf6.18194MAP kinaseNo
DAL5orf19.10720orf6.43914Damage responseInduced
RAX2orf19.11249orf6.39544Maintenance of bud-site selection in bipolar buddingRepressed
MOH1orf19.10876orf6.25794Unknown, homology to kinase Snf7No
KAP122orf19.9641orf6.74614Regulatory protein for drug resistance proteinNo
DIP5orf19.2445orf6.36234Amino acid transportNo
MYO2orf19.12482orf6.79944Type V myosin, establishment of cell polarityNo
QRI1orf19.11741orf6.34924UDP-N-GlcNAc pyrophosphorylase, important in cell wall biosynthesisNo
EHT1orf19.10558orf6.77774Serine hydrolase, lipid metabolismNo
URE2orf19.8339orf6.10924Regulation of nitrogen utilizationNo
TOS1orf19.1690orf6.41494Cell wall protein of unknown functionNo
MTLa1orf19.3201orf6.18843Transcription factor, forms complex with a2 
TAO3n.a.orf6.41463Protein involved in cell morphogenesis and smoo tip formationNo
MUM2orf19.11526orf6.44103Cytoplasmic protein required for meiosis in S. cerevisiaeNo
RHO1orf19.10362orf6.49373GTPase involved in establishment of cell polarityInduced
YNL101Worf19.8735orf6.21743Amino acid transporter from the vacuoleNo
CHS2orf19.7298orf6.83793Chitin synthaseRepressed
RBT2orf19.8991orf6.63843Ferric reductase, Tup1-regulated 
UBC8orf19.12016orf6.66873Ubiquitin-conjugating enzyme, negatively regulates gluconeogenesisInduced
SCW10orf19.9345orf6.7083Cell wall protein, conjugation during matingInduced
CUP9orf19.13867orf6.76463Repressor of a peptide transporterInduced
CSP37orf19.10066orf6.23883Putative membrane protein 
orf6.7571orf19.13936orf6.75713Weak homology with USO1, required for ER to Golgi transportNo
ASK10orf19.11525orf6.44093Component of RNA polymerase II holoenzymeNo
CHS3orf19.12403orf6.60533Chitin synthaseInduced
AXL1orf19.7342orf6.84233Bud-site selection, pheromone maturationInduced
Table 2.  Genes repressed in response to alpha pheromone in SpiderM medium.
Nameorf19orf6RepressionFunctionS. cerevisiae
  1. Table of significantly repressed genes identified by SAM analysis and whose expression changed by more than threefold. The full set of genes that passed SAM analysis are available in Table S2. Analysis was performed on four repeat experiments examining the response of opaque a cells to MF13 pheromone in SpiderM medium.

YNL234Worf19.1541orf6.8217Weak similarity to AGA1 (a-agglutinin anchor)No
orf6.2493orf19.4075orf6.24937Similar to Hyr1, hyphally regulated protein 
SAP3orf19.6001orf6.90366Secreted aspartyl protease 
YJR110Worf19.4745orf6.36576Phosphatidylinositol 3-phosphate phosphataseNo
orf6.1902orf19.853orf6.19026Homology with secreted aspartyl protease 
SAP1orf19.5714orf6.46446Secreted aspartyl protease 
OPS4orf19.4934orf6.60566Opaque specific protein 
OPT2orf19.2292orf6.13645Oligopeptide transporterNo
JEN1orf19.7447orf6.86075Lactate-proton symporterNo
DAN4orf19.6334orf6.55275Possible cell wall mannoproteinNo
orf6.2482orf19.1978orf6.24825Inositol metabolism, transporter 
DYN1orf19.5999orf6.90384Dynein heavy chain, microtubule motorRepressed
MCM21orf19.3494orf6.19384Component of centromere/kinetochore networkNo
DSE4orf19.10584orf6.78034Daughter cell specific protein; glucanaseRepressed
POLRTorf19.6078orf6.2414Reverse transcriptase 
INT1orf19.11733orf6.17324Adhesion and virulence geneRepressed (BUD4)
SWE1orf19.4867orf6.7064Protein kinase regulates G2/M transitionRepressed
GIN1n.a.orf6.21884Weak homology to Zn finger transcription factor 
LTE1orf19.2238orf6.23944GDP/GTP exchange factor, required for mitotic exitNo
CHS1orf19.5188orf6.34994Chitin synthaseInduced
YDL222Corf19.6489orf6.10744Possibly affects actin distribution and budding patternInduced
NAG1orf19.9703orf6.24704Glucosamine-6-phosphate isomerase 
SAP2orf19.3708orf6.53064Secreted aspartyl protease 
MCM6orf19.2611orf6.42743Involved in DNA replicationNo
KIP2orf19.9315orf6.20203Kinesin-related motor; mitotic spindle positioningRepressed
NAG2orf19.2157orf6.6843N-acetylglucosamine-6-phosphate deacetylase 
LIP1orf19.12284orf6.23983Secretory lipase 1No
PRY1orf19.10303orf6.36043Unknown, related to plant pathogenesis proteinsRepressed
YAT2orf19.2809orf6.45843Putative carnitine acetyltransferaseNo
PRP8orf19.6442orf6.78123U5 snRNA-associated splicing factorNo
FLO1orf19.3620orf6.32903Cell wall protein involved in flocculationNo
MCD1orf19.7634orf6.89823Sister chromatid cohesion; peaks in S-phaseRepressed
DAP1orf19.6867orf6.71313Damage response protein; haem binding proteinNo
MCM5orf19.5487orf6.41813Involved in DNA replicationNo
MRC1orf19.659orf6.65523Mediator of replication checkpointRepressed
RNR1orf19.5779orf6.48513Ribonucleotide reductase large subunitRepressed
POL30orf19.4616orf6.79153Proliferating cell nuclear antigen (PCNA)Repressed
RNR2orf19.5801orf6.66353Ribonucleotide reductase small subunitNo
YVC1orf19.2209orf6.35753Vacuolar cation channelNo
USO1orf19.3100orf6.1493Integrin homologue; ER to Golgi transportNo
MCM4orf19.202orf6.59343Involved in DNA replicationRepressed
CLOST1orf19.716orf6.26373Homology with protein from Clostridium perfringens 
LAP3orf19.539orf6.67513Aminopeptidase of cysteine protease familyInduced

We considered two possibilities for the increased number of genes induced by pheromone in SpiderM medium compared with SCD medium: (i) the pheromone response, averaged over the whole population of cells, has a higher amplitude in SpiderM medium, so genes that are only marginally induced in SCD medium are now induced above the threefold cut-off in SpiderM medium. (ii) Pheromone induces a distinct second set of genes in SpiderM medium that remain unchanged during the pheromone response in SCD medium. To distinguish between these possibilities we examined more closely the induction ratios of a number of genes in both SCD medium and SpiderM medium (Fig. 2). From this analysis we conclude that growth of cells in SpiderM medium leads to a general increase in the response of all affected genes to pheromone, rather than induction of a distinct set of genes in SpiderM medium. In principle, the enhanced pheromone response in SpiderM medium could be due to either an enhanced response in each individual cell or to an increase in the fraction of cells that response to pheromone. We believe, based on the morphological response of cells to pheromone, that the latter idea at least partially explains the increase: 80% of opaque a/a cells grown on SpiderM medium form mating projections in response to α-factor, while only 40% of cells in SCD show this morphological response (see below).

Figure 2.

Comparison of pheromone induction in SCD and SpiderM media. The fold induction of several genes is compared in SCD and SpiderM media, taking data from the microarray analysis. For each gene, the fold induction is an average from at least four independent microarray experiments (using MF13 pheromone, as shown in Fig. 1A). The error bars represent the SE (standard error) calculation for the averaged data. The first seven genes are genes induced more than threefold in SCD medium and the second seven genes are induced less than threefold in SCD medium. Both sets of genes were further upregulated by pheromone in SpiderM medium (see text for details).

We have previously discussed the genes induced by pheromone in SCD medium (Bennett et al., 2003) and here we note in passing that the larger gene set observed for SpiderM medium include genes involved in cell wall formation (GFA1, CHS7, YEA4, KRE1 and MNN4) and several more that may play an active role in the mating process, including BNI1, MYO2 and BUD16. There is also a large class of genes whose functions and possible roles in mating are unknown.

We also observed a number of genes that were significantly downregulated in response to pheromone, and this also occurred in a media-dependent manner (see Fig. 1B). Some genes were repressed in most media conditions, while other genes were only repressed when cells were grown in SpiderM medium (Fig. 1B, lanes 8–11). SAM was used to identify those genes whose repression by pheromone was significant (see Table S2), and the subset whose expression changed by more than threefold is listed in Table 2. Many of these genes are upregulated in SpiderM medium compared with SCD medium in the absence of pheromone (Fig. 1B, lane 13), but are downregulated upon addition of alpha pheromone (lanes 8–11). This cluster of pheromone-repressed genes included SAP1, SAP3 and OPS4, which have previously been shown to be repressed in the strain P37005 in response to alpha pheromone (Lockhart et al., 2003a). In addition, we noted the repression of many genes whose homologues function in DNA replication and cell growth in S. cerevisiae. For example, we noted that five of the six subunits of the MCM2-7 hexameric complex that acts in DNA replication initiation were downregulated by pheromone (see Table S2). Several additional components of the DNA replication machinery were downregulated, including DNA polymerases (POL12, DPB2), DNA primase (PRI1), and the DNA polymerase clamp, POL30. Furthermore, the G1 cyclin, CLN1, was also repressed by pheromone treatment, and this gene has been shown to exhibit cell cycle-specific regulation in C. albicans (Loeb et al., 1999). Indeed, heterologous expression of CaCLN1 in S. cerevisiae caused dominant-negative inhibition of pheromone response (Whiteway et al., 1992). The downregulation of so many genes required for progression through the cell cycle is consistent with pheromone-mediated cell cycle arrest, as occurs in S. cerevisiae. In a subsequent section we confirm that pheromone induces cell cycle arrest in C. albicans, and that growth medium affects this process.

Modulation of the pheromone response as detected by microarray analysis

In addition to the effect of media on the pheromone response, a number of other factors were found to influence the transcriptional response, as outlined below.

(i) The 13mer alpha pheromone induces a stronger transcriptional response in opaque a cells than the 14mer alpha pheromone.  As mentioned in the introduction, the alpha pheromone precursor protein is predicted to contain three repeats of the mature pheromone sequence (Bennett et al., 2003; Lockhart et al., 2003a; Panwar et al., 2003). Release of the mature alpha peptides by proteolytic processing is predicted to generate two peptides 13 amino acids in length (MF13) and one peptide 14 amino acids in length (MF14). We compared the transcriptional response elicited by MF13 and MF14 peptides in opaque a cells, in both SCD and SpiderM media (lanes 1–4, 7, 11 and 12 of Fig. 1A). In both SCD and SpiderM media the response to MF13 (10 μg ml−1) was considerably stronger than that to MF14 at the same concentration. For example, RBT1 was induced 34-fold by MF13 in SpiderM medium but only sixfold by MF14 in the same medium. These results showing an enhanced transcriptional response to MF13 are consistent with earlier studies showing that MF13 caused a stronger morphological response than MF14 when added in comparable concentrations to opaque a/a cells (Panwar et al., 2003).

(ii) Deletion of STE2 blocks the entire transcriptional response to alpha pheromone.  The STE2 gene in S. cerevisiae encodes the cell surface receptor for alpha pheromone and deletion of STE2 completely blocks the response to pheromone (Roberts et al., 2000). Previously, we showed that deletion of the STE2 orthologue in C. albicansa cells blocked both the morphological response of a cells to alpha pheromone and the mating of a cells. In contrast, deletion of STE2 from α cells had no effect on their mating efficiency (Bennett et al., 2003). To determine if Ste2 was required for all aspects of the transcriptional response to alpha pheromone in C. albicans, we analysed the response of Δste2/Δste2a cells to MF13 in SCD medium (Fig. 1A, lanes 16 and 17). No genes showed significant induction in the Δste2 mutant upon exposure to alpha pheromone, indicating that the STE2 gene is absolutely required for all aspects of the pheromone-induced transcriptional programme.

(iii) Growth medium influences the pheromone response of white cells.  Although opaque cells are the mating-competent form of C. albicans, we also examined the transcriptional response of white cells to pheromone. Unlike opaque cells, white cells show no morphological response to pheromone (Bennett et al., 2003; Lockhart et al., 2003a; Panwar et al., 2003). However, white cells can exhibit a limited transcriptional response, and it was found that this response is dependent upon the media conditions used. The clearest effects were seen when strains grown in Lee's medium were compared with strains grown in SCD medium. Whereas white cells grown in Lee's medium showed induction of 30 genes (> threefold) in response to pheromone (Fig. 1A, lanes 21 and 23–25), no significant gene induction was observed when the experiment was performed in SCD medium (lanes 18, 19 and 22). In contrast to opaque cells where the strongest pheromone response was seen in SpiderM medium, white cells showed the strongest response in Lee's medium (see Fig. 1A, lane 21).

(iv) C. albicans strain background influences the response to alpha factor.  The experiments described above utilized an a/a derivative of the standard laboratory strain SC5314. We compared the response in SC5314 cells to that of a clinical isolate, P37005, which is a natural a/a isolate (Lockhart et al., 2002). Using opaque cells, the transcriptional profile of the P37005 strain in Lee's medium (Fig. 1A, lanes 14 and 15), was very similar to that of the SC5314 strain in SpiderM medium (lane 12), when both strains were challenged with MF14 pheromone.

In contrast to opaque cells, white a/a cells from SC5314 and P37005 strains responded very differently to pheromone. In particular, in experiments carried out in Lee's medium, the P37005 white strain showed a strong response to pheromone whereas the SC5314 strain (Fig. 1A, compare lane 21 with lane 24) showed a markedly different response that was obviously much weaker. These results demonstrate that strain background has a significant effect on the transcriptional response induced by pheromone, and this is particularly evident in white cells.

(v) Identification of a conserved Ste12/Cph1 binding motif in pheromone-induced genes.  To identify potential regulatory sites in the promoters of pheromone-induced genes, we submitted the promoter sequences of 31 strongly induced genes to the motif-identification program, MEME ( (Bailey and Elkan, 1994). Most of the induced genes contained a common motif TTGTTTC/GA that closely resembled the binding site of the S. cerevisiae Ste12 transcription factor, TGTTTCA (also known as PRE or pheromone response element) (Dolan et al., 1989; Errede and Ammerer, 1989). The C. albicans motif was present in 20 out of the 31 upstream regions queried, often in multiple repeats; no other significantly conserved sequences of this length were found. In total, we found 37 matches to the C. albicans PRE motif in the 31 promoters (an average ratio of 1.19 motif per promoter) and this was found to be significant when compared with the promoters of all the genes in the C. albicans genome (P-value < 1.5 × 10−10). The consensus S. cerevisiae Ste12 motif TGTTTCA was also previously identified in the promoters of C. albicans HWP1, RBT4, SAP5 and SAP6 genes by promoter searching (Lane et al., 2001), and overexpressed C. albicans Ste12/Cph1 protein was shown to bind to the S. cerevisiae PRE (Malathi et al., 1994).

In summary, using transcriptional profiling we have shown that multiple factors influence the response to pheromone in C. albicans. These include the white and opaque phase transition, strain background, the specific mating pheromone (MF13 or MF14), and, in particular, the growth medium. We also show that all instances of gene induction are dependent on the Ste2 protein, the putative receptor for alpha factor.

Role of the Gpa2 protein in regulating the response to pheromone in C. albicans

Of all the variables we examined, growth media have the largest effect on the pheromone response; we therefore investigated this effect in greater depth. Several yeast species, including Schizosaccharomyces pombe and Cryptococcus neoformans, require signalling inputs from both a pheromone-inducible pathway and a nutrient-sensing pathway for efficient mating. In these fungi, nutrient sensing occurs via a conserved cAMP signalling pathway including Gpr1 (G protein-coupled receptor) and Gpa2 (a Gα protein), which lie at the beginning of the pathway (for review see Lengeler et al., 2000).

In C. albicans, the Gpr1/Gpa2 proteins have been shown to interact and to function in the nutrient regulation of filamentous growth (Miwa et al., 2004; Maidan et al., 2005). However, there is some controversy as to whether these C. albicans proteins signal directly to the MAP kinase pathway (Sanchez-Martinez and Perez-Martin, 2002) or signal through the cAMP pathway (Miwa et al., 2004; Maidan et al., 2005). To test whether C. albicans Gpa2 is involved in nutrient sensing during pheromone response we made a homozygous knockout of the GPA2 gene in a and α cells (see Experimental procedures). Deletion of the GPA2 gene in a/α cells was previously shown to lead to a defect in filamentation on several hyphae-inducing media (Sanchez-Martinez and Perez-Martin, 2002; Miwa et al., 2004; Maidan et al., 2005). We found that Δgpa2/Δgpa2a and α cells were defective for filamentation when tested on solid SpiderM medium plates or medium containing serum (see Fig. 3), showing that the behaviour of a and α cells is similar to that of a/α cells in this regard. Reintroduction of a single copy of the GPA2 gene (see Experimental procedures) restored filamentation on Spider or serum-containing media (data not shown), confirming that Gpa2 is required for proper filamentation in a/α, a/a and α/α cells.

Figure 3.

Defects in filamentation due to deletion of the GPA2 gene. Hypha formation was compared in white cells of wild-type (WT) and Δgpa2 mutant a and α strains. Strains were grown on solid YPD medium or solid SpiderM medium at 30°C, or on solid YPD medium containing 10% FBS at 37°C. Strains RBY1173, RBY1144 and RBY1145 are a-type strains, and strains RBY1174, RBY1143 and RBY1146 are α-type strains.

To test the role of Gpa2 in nutrient sensing during mating, we first examined the morphological response of cells to pheromone under different media conditions. Figure 4A shows the percentage of opaque a/a cells that formed mating projections in response to alpha pheromone (after 4 h) in both wild-type and Δgpa2 mutants. In contrast to the results with a/a opaque cells, Δgpa2a/a opaque cells showed efficient formation of mating projections irrespective of the growth medium (Fig. 4A). Thus, deletion of the GPA2 gene removed the negative effects of growth media on the morphological response to pheromone. This conclusion was confirmed by demonstrating that reintroduction of the GPA2 gene (strain RBY1205) restored the nutrient control of mating projections.

Figure 4.

Efficiency of mating projection formation in cells exposed to pheromone.
A. The fraction of cells making mating projections in wild-type (RBY1179) and Δgpa2 mutant (RBY1167) strains was compared in YPM, YPD, SpiderM and SpiderD media. In opaque cells from the wild-type strain, the fraction of the population that make mating projections in Spider medium was significantly higher than in YP medium. In contrast, deletion of the GPA2 gene led to a high percentage response to pheromone under all four media conditions tested.
B. Addition of cAMP to the medium did not significantly affect the efficiency of the response to mating pheromone. The error bars represent the SE calculation for the averaged data.

In S. pombe and C. neoformans, addition of exogenous cAMP to the media can override nutritional signals from the Gpr1/Gpa2 sensor (Isshiki et al., 1992; Mochizuki and Yamamoto, 1992; Alspaugh et al., 1997). We examined whether the addition of 10 mM N6, 2′-O-dibutyryl-AMP (dbcAMP) would alter the morphological response to pheromone in wild-type a/a cells. dbcAMP is a non-metabolic derivative of cAMP and has previously been used to model high cAMP levels in C. albicans (Miwa et al., 2004; Maidan et al., 2005). Addition of dbcAMP was found to have little or no effect on the efficiency of the response to pheromone in either YPD or SpiderM medium (see Fig. 4B). We also found that the addition of dbcAMP did not affect the efficiency of the response of the Δgpa2/Δgpa2 mutant to pheromone (data not shown). These results, taken with the published literature, suggest that high cAMP levels alone are not sufficient to influence the pheromone response.

Deletion of GPA2 increases the transcriptional response to pheromone

The experiments described above indicate that wild-type cells show a heightened response to pheromone in nutrient-poor media compared with rich media, as evident by the increased percentage of cells showing a morphological response (Fig. 4A) and by the increased transcriptional response to pheromone averaged over the population of cells (Fig. 1A). To examine the role of the GPA2 gene on the transcriptional response to pheromone, we analysed the expression of several genes by quantitative PCR (QPCR), in the presence and absence of the GPA2 gene and in the presence and absence of pheromone. Figure 5 shows the level of expression of four pheromone-inducible genes (FUS1, DG1106, RBT1 and HST6) in wild-type and Δgpa2 strains grown under different media conditions. These genes were chosen to represent a range of biological functions and pheromone induction ratios.

Figure 5.

Quantitative PCR was used to compare the transcriptional response to pheromone in wild-type and Δgpa2 mutant strains. The level of expression of four genes (FUS1, DG1106, RBT1 and HST6) was compared in the presence and absence of pheromone in both a wild-type strain (RBY1179) and in Δgpa2 strains (RBY1166 and RBY1167). Expression levels were normalized to the PAT1 gene (see Experimental procedures). The error bars represent the SE calculation for the averaged data. Units of expression are arbitrary.

Pheromone induction of three of the four genes (FUS1, DG1106 and RBT1) was greater in the Δgpa2 mutant than in the wild-type strain under all media conditions tested (see Fig. 5). The increased response of the Δgpa2 mutant was particularly evident in the comparison of YPM and YPD media. However, as seen in Fig. 5, the transcriptional response of Δgpa2 strains was not completely independent of the media. Thus, deletion of the GPA2 gene leads to an increased transcriptional response to pheromone under each of the tested media conditions, but does not render the response to pheromone completely independent of the media used.

Δgpa2 strains exhibit distinct ‘halos’ upon treatment with mating pheromone

In S. cerevisiae, the spotting of alpha mating pheromone onto a nascent lawn of MATa cells results, when the lawn has grown to maturity, in the formation of a ‘halo’ around the pheromone spot due to efficient growth arrest of the neighbouring cells. This assay has been used as a semi-quantitative measure of pheromone activity, as the size of the halo is proportionate to pheromone activity (Manney, 1983). In C. albicans, the addition of alpha pheromone to a wild-type lawn of opaque a/a cells failed to produce a clear halo in the laboratory strain SC5314 (A. Uhl, R.J. Bennett and A.D. Johnson, unpubl. obs., and see below) and induced only an indistinct halo in the clinical isolate P3745 (Panwar et al., 2003). As Δgpa2 strains showed an increased morphological and transcriptional response to pheromone we tested the behaviour of these strains in halo assays. Figure 6 compares halo formation in a lawn of opaque wild-type (Fig. 6A) and opaque Δgpa2 (Fig. 6B) cells, upon which alpha pheromone was directly spotted. In wild-type cells, the MF13 pheromone induced a weak response, in which neighbouring cells did not undergo complete growth arrest but grew more slowly than unaffected cells, resulting in an indistinct zone of reduced growth. The MF14 pheromone produced a very localized effect, preventing cell growth within the spotted area, but having no effect on neighbouring cells (data not shown). In opaque a/aΔgpa2 cells, the MF13 pheromone induced a much clearer halo, as seen in Fig. 6B. Some microcolonies were still evident within the halo, indicating that growth arrest was not absolute, but the zone of growth arrest was pronounced and extended well beyond the boundaries of the original pheromone spot. Colonies were picked from the zones of inhibition and were examined microscopically; they appeared to be white cells that had spontaneously switched from the opaque phase, and hence would escape pheromone arrest. Spontaneous opaque to white switching occurs at frequencies of approximately 10−3 (Rikkerink et al., 1988), a value consistent with the number of colonies appearing in the zones of growth arrest.

Figure 6.

Halo formation in wild-type and Δgpa2 strains. A lawn of opaque a cells from (A) a wild-type strain (RBY1179) (B) a Δgpa2 mutant strain (RBY1167), or (C) a GPA2 addback strain (RBY1205), was plated on solid SpiderM medium plates. In addition, pheromone (MF13 dissolved in 10% DMSO) or controls (10% DMSO) were spotted onto the plates (black circles). Plates were allowed to grow for 3 days at room temperature. Whereas a very faint ‘halo’ was apparent around the MF13 spot on the wild-type cells (A) or the GPA2 addback cells (C), a distinct zone of growth arrested cells was evident around the MF13 spot on the Δgpa2 cells (B).

To confirm that halo formation was enhanced in the absence of the GPA2 gene, we also compared halo formation following reintroduction of a single copy of the wild-type GPA2 gene into the Δgpa2/Δgpa2 mutant. Figure 6C shows that addback of the GPA2 gene restored the wild-type phenotype, indicating that the enhanced halo formation is dependent on the absence of the GPA2 gene. Halo formation by the Δgpa2 strain did not extend to all media: for example, this strain did not form halos on YPD medium. These findings are consistent with our observations (discussed above) that deletion of GPA2 removed some but not all of the nutritional controls of the pheromone response.

The formation of clear halos using Δgpa2 cells suggested that these cells were undergoing efficient cell cycle arrest in response to pheromone and FACS analysis confirmed this idea (Fig. 7). Whereas almost all the cells from the Δgpa2 strain arrested in G1 in response to pheromone, a significant fraction of wild-type cells continued to divide, as evident from the large fraction of cells in G2 (Fig. 7B). Following the removal of pheromone, cells re-entered the cell cycle, and by 120 min the majority of Δgpa2 cells had replicated their DNA and were in G2 of the cell cycle (Fig. 7A). Wild-type cells also re-entered the cell cycle following the removal of alpha pheromone (Fig. 7B), although the population was less synchronous due to the incomplete arrest of these cells by pheromone.

Figure 7.

FACS analysis of alpha pheromone-induced cell cycle arrest and release. Opaque cells from Δgpa2gpa2 (A) and wild-type (B) strains were grown in YPD at 23°C in logarithmic phase (Log phase) prior to addition of alpha pheromone (MF13) or a mock control. Cells were collected after 5 h incubation with α-factor and analysed by FACS. In addition, following pheromone treatment, some cells were washed to remove pheromone and allowed to recover from cell cycle arrest. Cells were resuspended in fresh YPD medium at 23°C and collected at 30 min intervals for FACS analysis. YPD medium was used for this experiment as cells re-entered the cell cycle more efficiently in this medium than in SpiderM medium (data not shown). The wild-type strain was RBY1175 and the Δgpa2 mutant strain was RBY1167.

FACS analysis of opaque cells in different media

Recent studies by Zhao et al. revealed that cells grown to saturation phase responded more strongly to pheromone than cells grown in exponential phase (Zhao et al., 2005). The increased response of cells that had entered saturation phase correlated with their being in the G1 phase of the cell cycle, and hence ready to respond to pheromone. One possibility for the varying response to pheromone observed in this work is that media composition influences the fraction of cells in the G1 phase of the cell cycle. To test this idea, we analysed cells grown on different media by FACS. Figure 8 shows opaque cells grown in logarithmic phase in YPM, YPD, SpiderM or SpiderD media. Cells grown in YP or Spider medium showed a very similar FACS profile (compare Fig. 8A and C, or B and D), indicating that the increased pheromone response observed in Spider medium over YP medium is not due to an increased fraction of the population being in G1. Interestingly, an increased fraction of cells in G1 was observed when cells were grown in mannitol-containing medium compared with glucose-containing medium (compare Fig. 8A and B, or C and D). The increased fraction of cells in G1 may help explain the slight increase in pheromone response observed with media containing mannitol rather than glucose (compare percentage of cells responding to pheromone in Fig. 4A); however, the predominant effect of growth media on pheromone induction is direct and cannot be attributed simply to the fraction of cells in G1.

Figure 8.

FACS analysis of cells grown in different media conditions. Opaque cells from wild-type strain RBY1179 were grown in logarithmic phase, collected, and analysed by FACS, as described in Experimental procedures. Cells were grown in YPM (A), YPD (B), SpiderM (C) or SpiderD (D).

Mating efficiency of Δgpa2 strains in different media conditions

As described above, the response to pheromone is enhanced in Δgpa2 cells, as evidenced by an increased fraction of cells responding, by an enhanced transcriptional response of the population, and by efficient growth arrest. We next tested whether deletion of the GPA2 gene affected mating, quantified under a standard set of growth conditions.

In Fig. 9 are shown results from quantitative mating assays, in which auxotrophic opaque a and α strains were mated and prototrophic mating products selected and counted (see Experimental procedures). As previously observed, mating efficiency of wild-type strains is affected by the media conditions; in particular, mating was repressed in glucose-containing media relative to mannitol-containing media (Bennett et al., 2005). Under several different media conditions, deletion of the GPA2 gene led to a significant increase in mating efficiency. For example, mating in YPD and SpiderD media was increased threefold in the Δgpa2 cells relative to wild-type cells (see Fig. 9).

Figure 9.

Mating efficiency of wild-type (WT) and Δgpa2 strains. Mating between wild-type a and α strains (RBY1177 mated to RBY1180, and RBY1178 mated to RBY1179) was compared with mating between Δgpa2a and α strains (RBY1166 mated to RBY1169, and RBY1167 mated to RBY1168). The efficiency of mating was determined by mating the auxotrophic strains and plating on selective media. Matings were carried out in liquid media for 24 h, as described in Experimental procedures. The error bars represent the SE calculation for the averaged data.


The response to pheromone in S. cerevisiae is one of the most highly studied signalling pathways, and has become a model system for understanding many aspects of signal transduction (for recent review see Bardwell, 2005). A receptor-linked MAP kinase cascade transduces the pheromone signal from outside the cell to the nucleus where a transcriptional response compromising approximately 485 genes occurs (Roberts et al., 2000). In C. albicans, the principle components of this MAP kinase cascade are highly conserved (Chen et al., 2002; Magee et al., 2002; Bennett et al., 2003; Lockhart et al., 2003a; Panwar et al., 2003), but, as we show here, both the output of the signalling pathway and the way it is controlled differ significantly.

We show that strain background, the structure of the alpha pheromone peptide, the phenomenon of white-opaque switching, and media composition all have a significant effect on the nature of the pheromone response in C. albicans. Perhaps the most important of these observations is the dependence of the pheromone response on the nutrients available in the medium. For example, we found that the pheromone induction in SpiderM medium, a low-nutrient medium containing mannitol as the carbon source, had a much higher amplitude (as measured by number of genes significantly induced, their induction ratios, and the fraction of cells exhibiting morphological responses to mating pheromone) than that in nutrient-rich media. Under these optimal conditions approximately 140 genes were induced more than threefold by pheromone. Relatively few of these genes overlap with the 156 genes induced more than twofold in S. cerevisiae (see Table 1) (Roberts et al., 2000). Thus, while the signal transduction pathway is virtually identical between S. cerevisiae and C. albicans, the set of genes it regulates has diverged considerably.

We further investigated the nutritional control over the pheromone response pathway in C. albicans and found that it is mediated in part by the Gpa2 protein. Gpa2 is highly conserved in fungi where it functions with the Gpr1 protein to mediate nutrient signalling (Lengeler et al., 2000; Thevelein et al., 2005). Deletion of C. albicans GPA2 increased (i) the morphological response of C. albicans to pheromone, (ii) the magnitude of the transcriptional response to pheromone, particularly in rich, glucose-containing media, (iii) the degree of cell cycle arrest in response to pheromone, and (iv) the efficiency of mating. This integration of nutritional and pheromone signalling pathways is apparently absent in S. cerevisiae, but it has been documented in S. pombe and C. neoformans (for review see Lengeler et al., 2000).

The heightened pheromone response of C. albicansΔgpa2 mutants causes tight cell cycle arrest and provides a useful tool for future studies of C. albicans. FACS analysis confirmed that pheromone-treated Δgpa2 cells were efficiently arrested in the G1 phase of the cell cycle. Unlike a/a derivatives of the laboratory strain SC5314, the Δgpa2a/a strain produces pronounced ‘halos’ of growth inhibition around spots of alpha pheromone. The ability to induce cell cycle arrest using pheromone has been a powerful tool for synchronization of the cell cycle in S. cerevisiae, and the present results suggest that the efficient pheromone arrest observed in C. albicansΔgpa2 mutants will be a similarly powerful tool for cell cycle studies in this organism. Recently, another report of heightened sensitivity to pheromone was demonstrated by deletion of the SST2 gene in C. albicans (Dignard and Whiteway, 2006). SST2 acts to downregulate the MAP kinase pheromone response in S. cerevisiae (Dietzel and Kurjan, 1987), and its removal therefore increases the sensitivity to pheromone exposure. Again, halo formation was observed in sst2 mutants but not in wild-type strains. These results, taken together with our observations, indicate that the response to pheromone in C. albicans is graded, with moderate responses causing morphological changes and only stronger responses (so far, achievable only with mutant strains) eliciting cell cycle arrest.

In summary, we have shown that pheromone response in C. albicans is regulated by multiple factors including nutrient composition, strain background and the chemical nature of the pheromone. In exploring the nutrient dependence, we found C. albicans, like several other fungi but unlike S. cerevisiae, regulates its mating in response to external nutritional cues, and this regulation depends on Gpa2-mediated signalling. These studies show that wherever C. albicans normally undergoes mating, be it in the host or in some as yet undiscovered external niche, nutrient sensing is a critical input. These studies also highlight the difference in the transcriptional response of S. cerevisiae and C. albicans to pheromone. Although these yeasts utilize the same proteins to detect pheromone (the Ste2 receptor for α-pheromone), to transmit the signal (Gpa1/Cag1, Ste4, Ste18, Ste20/Cst20, Ste11, Ste7, Fus3/Cek2), and to activate gene expression (Ste12/Cph1) (Chen et al., 2002; Magee et al., 2002; Bennett et al., 2003), the many genes whose expression is regulated are distinctive for each yeast, with the responses encompassing relatively few genes in common. This observation is testament to the vast amount of transcriptional rewiring that has occurred since C. albicans and S. cerevisiae last shared a common ancestor some 300 million years ago (Gargas et al., 1995; Hedges et al., 2004).

Experimental procedures

Media and reagents

Standard laboratory media were prepared as described previously (Guthrie and Fink, 1991). Lee's medium has been described previously (Bedell and Soll, 1979). Lee's + D and Lee's + M refer to Lee's medium supplemented with either 2% glucose or 2% mannitol. Spider medium contained 1.35% agar, 1% nutrient broth, 0.4% potassium phosphate (pH 7.2) supplemented with either 2% glucose (SpiderD) or 2% mannitol (SpiderM) (Liu et al., 1994). SCD medium refers to synthetic complete medium supplemented with 2% glucose. YPD and YPM media refer to 1% yeast extract, 2% peptone media supplemented with either 2% glucose or 2% mannitol respectively.

Alpha pheromone peptides MF13 (GFRLTNFGYFEPG) and MF14 (GFRLTNFGYFEPG) were synthesized by Genemed Synthesis. N6, 2′-O-dibutyryl-AMP (dbcAMP) was purchased from Sigma.


The starting strain SNY152 (leu2/leu2 his1/his1 arg4/arg4) (Noble and Johnson, 2005) was used for construction of Δgpa2/Δgpa2 deletion strains. First, a and alpha derivatives of SNY152 were generated by growth of this strain on sorbose medium, as previously described (Janbon et al., 1998; Bennett et al., 2003), to create RBY1132 (a/a strain) and RBY1133 (α/α strain) (Table 3). PCR products for targeting the GPA2 ORF were generated using oligonucleotides 1, 5′-GTCTTATTATTATTGAATCATCG, and 3, 5′-cacggcgcgcctagcagcggCGAAGCACAAGAACCCATGG, to first amplify the 5′ flank of GPA2, and oligonucleotides 4, 5′-gtcagcggccgcatccctgcGAATTGGTGATGGCACTGT, and 6, 5′-CACACATCGCGTTTGTCGTG, to amplify the 3′ flank of GPA2. Selectable marker sequences (Candida dubliniensis HIS1, Candida maltosa LEU2 and C. albicans ARG4) were also amplified by PCR from the plasmids pSN52, pSN40 and pSN44 respectively, as described (Noble and Johnson, 2005). Fusion PCR products were then generated by using oligonucleotides 1 and 6 to amplify DNA in reactions containing the flank PCR products together with a marker PCR product (Noble and Johnson, 2005). The first allele of GPA2 was replaced using the HIS1 marker in both RBY1132 and RBY1133 to create RBY1140/1141 and RBY1142/1143 respectively. The second allele of GPA2 was then replaced by the LEU2 gene in RBY1140/1141 to create RBY1144/1145 and by the ARG4 gene in RBY1142/1143 to create RBY1146/1147. Correct integration of the PCR products was verified by PCR across the 5′ and 3′ disruption junctions, and loss of the GPA2 gene was also confirmed using PCR primers internal to the GPA2 ORF. The resulting strains were switched to the opaque phase by passaging on SCD medium to yield RBY1166/1167 and RBY1168/1169.

Table 3.  Strains used in this study.
RBY717a/a, ura3::imm434/URA3 iro1::imm434/IRO1WhiteBennett et al. (2003)
RBY731a/a, ura3::imm434/URA3 iro1::imm434/IRO1OpaqueBennett et al. (2003)
RBY722α/α, ura3::imm434/URA3 iro1::imm434/IRO1WhiteBennett et al. (2003)
RBY734α/α, ura3::imm434/URA3 iro1::imm434/IRO1OpaqueBennett et al. (2003)
RBY941a/a, ste2::hisG/ste2::URA3 ura3::imm434/ura3::imm434OpaqueThis study
P37005a/aWhite/opaqueLockhart et al. (2002)
RBY1143α/α, leu2::hisG/leu2::hisG his1::hisG/his1::hisG arg4::hisG/arg4::hisG gpa2::HIS1/GPA2*WhiteThis study
RBY1144a/a, leu2::hisG/leu2::hisG his1::hisG/his1::hisG arg4::hisG/arg4::hisG gpa2::HIS1/gpa2::LEU2*WhiteThis study
RBY1145a/a, leu2::hisG/leu2::hisG his1::hisG/his1::hisG arg4::hisG/arg4::hisG gpa2::HIS1/gpa2::LEU2*WhiteThis study
RBY1146α/α, leu2::hisG/leu2::hisG his1::hisG/his1::hisG arg4::hisG/arg4::hisG gpa2::HIS1/gpa2::ARG4*WhiteThis study
RBY1147α/α, leu2::hisG/leu2::hisG his1::hisG/his1::hisG arg4::hisG/arg4::hisG gpa2::HIS1/gpa2::ARG4*WhiteThis study
RBY1166a/a, leu2::hisG/leu2::hisG his1::hisG/his1::hisG arg4::hisG/arg4::hisG gpa2::HIS1/gpa2::LEU2*OpaqueThis study
RBY1167a/a, leu2::hisG/leu2::hisG his1::hisG/his1::hisG*arg4::hisG/arg4::hisG gpa2::HIS1/gpa2::LEU2OpaqueThis study
RBY1168α/α, leu2::hisG/leu2::hisG his1::hisG/his1::hisG arg4::hisG/arg4::hisG gpa2::HIS1/gpa2::ARG4*OpaqueThis study
RBY1169α/α, leu2::hisG/leu2::hisG his1::hisG/his1::hisG arg4::hisG/arg4::hisG gpa2::HIS1/gpa2::ARG4*OpaqueThis study
RBY1173a/a, leu2::hisG/leu2::hisG his1::hisG/HIS1*WhiteThis study
RBY1174α/α, leu2::hisG/leu2::hisG his1::hisG/HIS1*WhiteThis study
RBY1177a/a, leu2::hisG/leu2::hisG his1::hisG/HIS1*OpaqueThis study
RBY1178α/α, leu2::hisG/leu2::hisG, his1::hisG/HIS1*OpaqueThis study
RBY1179a/a, arg4::hisG/arg4::hisG, his1::hisG/HIS1*OpaqueThis study
RBY1180α/α, arg4::hisG/arg4::hisG, his1::hisG/HIS1*OpaqueThis study
RBY1205a/a, leu2::hisG/leu2::hisG his1::hisG/his1::hisG arg4::hisG/arg4::hisG gpa2::HIS1/gpa2::LEU2/GPA2::SAT1-FLIP (addback)*URA3/Δura3::imm434, IRO1/Δiro1::imm434OpaqueThis study

The HIS1 gene was added back to several control strains to give them the same auxotrophies as the Δgpa2/Δgpa2 strains used in the quantitative mating experiments (see strain table). The C. dubliniensis HIS1 gene was amplified from pSN52 with primers 5′-GAGGAGACAGAAGTTAGTAG-3′ and 5′-TATGGTGCTCATGGCTACGC-3′. The PCR product was transformed into sorbose-derived a and α derivatives of SNY87 (Δleu2/Δleu2, Δhis1/Δhis1) (Noble and Johnson, 2005) to create RBY1177 and 1178. Similarly, the PCR product was transformed into sorbose-derived a and α derivatives of SNY95 (Δarg4/Δarg4, Δhis1/Δhis1) to create RBY1179 and RBY1180. These strains were switched to opaque as described above.

A wild-type copy of the GPA2 gene was reintroduced into Δgpa2/Δgpa2 strains to confirm the Δgpa2 phenotype. First, the wild-type GPA2 sequence, together with 1 kb of upstream sequence and 700 bp of the 3′ UTR, was amplified by PCR from SC5314 using primers 5′-GCGGCCATGGGCCCCACGAGAATGAATGAAAAAACCG-3′ and 5′-GGCGGTACTCGAGATCACCTGCCAAAACTCCTC-3′. Using the restriction sites ApaI and XhoI (underlined), the PCR product was introduced into pSFS2A (Reuss et al., 2004), which contains a dominant nourseothricin resistance marker (caSAT1). The resulting plasmid was linearized with AflII and integrated into the 3′ UTR of the GPA2 gene in strain RBY1166 (Δgpa2/Δgpa2) to generate RBY1205 (Δgpa2/Δgpa2/GPA2). Colonies were selected on YPD medium containing 200 μg ml−1 nourseothricin grown at 30°C, as previously described (Reuss et al., 2004), and the integration of the GPA2 gene checked by PCR using primers 5′-TCTCCTTCTCCTTCTTGATC-3′ and 5′-GGCGGCATCCGCGGGAATTGGTGATGGCAGCTGT-3′.

The STE2 gene was knocked out in CAI4 strains of C. albicans (Fonzi and Irwin, 1993) using the method of Wilson et al. (1999). Both alleles of STE2 were deleted using a PCR product based on pDDB57 as previously described (Bennett et al., 2003). Following deletion of the second allele, a derivatives were obtained by growth on sorbose medium and switched to the opaque phase, to form RBY941.

Halo assay

A lawn of C. albicans cells was formed by plating approximately 105 opaque cells (from an overnight culture grown in SCD medium) onto solid medium. MF13 or MF14 pheromone (10 μg ml−1 in 10% DMSO) or a 10% DMSO control was spotted (2 μl) onto the lawn of cells. Plates were incubated at room temperature for 2–3 days and photographed.

Quantitative mating assay

Quantitative mating assays were performed by modification of a previous protocol (Miller and Johnson, 2002). Leu- and Arg- mating strains were grown in the opaque phase at 23°C overnight in liquid SCD medium, and approximately 3 × 107 cells of each strain were mixed together. For mating on solid media, the mixed strains were deposited onto 0.8 μm filters using a Millipore Vacuum Sampling Manifold, and then placed on the surface of agar plates containing different media for 24 h at 23°C. For mating in liquid media, the mixed strains were washed in 5 ml of H2O, and then resuspended in 1 ml of the test liquid media, and grown at 23°C overnight. In both cases, cells were collected and various dilutions plated onto Leu- Arg- media to select for mating conjugants, and onto Leu- and Arg- media to monitor the parent + conjugation population.

Microarray analysis

Typically, for white and opaque cultures, 10–50 ml cultures were grown overnight at 23°C in SC + 100 μg ml−1 uridine and 55 μg ml−1 adenine. These cultures were used to inoculate 500–4000 ml cultures of the same medium, YPD medium, or Lee's medium, and grown overnight. When the OD of the culture reached 1.0, a sample of the culture was removed and frozen (zero time point). To the remainder of the culture, alpha pheromone dissolved in 10% DMSO was added (final concentration 10 μg ml−1), or 10% DMSO alone (control). The final concentration of DMSO in these cultures was 0.01%. Samples were taken from the cultures at 2 h or 4 h, collected by filtration and frozen down. In the case of alpha pheromone treatment of cells in SpiderM medium, cells were grown to OD 1.0 in SCD medium, washed with H2O, and resuspended in the same volume of SpiderM medium, prior to addition of alpha factor. Preparation of mRNA, cDNA and microarray analysis was carried out as previously described (Bennett et al., 2003).

Microarrays were analysed by SAM for the pheromone response in SCD medium (six microarray repeats) and Spider medium (four microarray repeats). The set of repeats for each experiment was analysed using the one-class response, the K-nearest neighbours setting, and the default random number seed (1234567). The delta values for both data sets were selected as the lowest value at which the median false discovery rate was zero. The delta value for the pheromone response in SpiderM medium data was 1.52, and the delta value for the pheromone response in SCD medium data was 1.28. The genes that passed the cut-off in SpiderM medium are included in Tables S1 and S2. For presenting the SpiderM medium data in the manuscript (Tables 1 and 2), an additional cut-off was implemented, choosing only those genes whose expression changed more than threefold. The microarray data will be made available on the Johnson lab web site at

Promoter analysis of the genes induced by pheromone was performed using MEME (Bailey and Elkan, 1994). Identification of the Cph1/Ste12 binding motif was confirmed by comparing the frequency of the motif in the pheromone-induced gene set (1.19) with the frequency in the promoters of all the genes in the C. albicans genome (0.35). This was found to be significant with a Poisson P-value < 1.55 × 10−10.

Quantitative PCR assays

Cell cultures were prepared as for microarrays except small-scale (3 ml) cultures were harvested for each experiment. Total RNA was prepared using a hot phenol procedure (Bennett et al., 2003). To eliminate DNA contamination, RNA was re-extracted twice with hot phenol (pH 4.3; Fisher Scientific) followed by chloroform extraction and precipitation with ethanol. RNA was reverse transcribed with Superscript (Stratagene) and cDNA amplified by QPCR in a DNA Engine 2 Opticon 2 (MJ Research). Signals from each experimental sample were normalized to signals from expression of the PAT1 gene, whose expression was not regulated by pheromone. Wild-type strain values are averaged from four independent experiments and Δgpa2/Δgpa2 strain values are averaged from three independent experiments.

FACS analyses

For analysis of cells grown in different media, opaque cells from strain RBY1175 were grown in SCD medium at 23°C overnight, washed in H2O, and resuspended in fresh medium (YPM, YPD, SpiderM or SpiderD) at a concentration of 107 cells ml−1. After 4 h incubation at 23°C, cells were collected by centrifugation and prepared for FACS analysis as previously described (Hull et al., 2000). To analyse cells during pheromone cell cycle arrest and release from arrest, opaque cells from RBY1175 or RBY1167 were grown to exponential phase in YPD at 23°C (2 × 107 cells ml−1) and then incubated with pheromone (MF13; 10 μg ml−1 in DMSO) or a control (DMSO alone) for 5 h. Cells were collected for FACS analysis, or washed in H2O and resuspended in fresh YPD medium. The resuspended cells were incubated at 23°C for another 150 min, with cells collected at 30 min intervals for FACS analysis.


We would like to thank Virgil Rhodius for assistance with the analysis of the microarray data using SAM, Brian Tuch for promoter analysis of the pheromone-induced genes, Andrew Uhl for unpublished observations, and other members of the Johnson lab for their help and support. This work was supported in part by grants from the Burroughs Wellcome Fund (993218) and NIH (RO1 AI49187) to A.D.J.