Ste18p Is a Positive Control Element in the Mating Process ofCandida albicans

Hui Lu,a* Yuan Sun,b Yuan-Ying Jiang,a Malcolm Whitewayb

Center for New Drug Research, Department of Pharmacology, School of Pharmacy, Second Military Medical University, Shanghai, Chinaa; Biology Department, ConcordiaUniversity, Montreal, Quebec, Canadab

Heterotrimeric G proteins are an important class of eukaryotic signaling molecules that have been identified as central elementsin the pheromone response pathways of many fungi. In the fungal pathogen Candida albicans, the STE18 gene (ORF19.6551.1)encodes a potential � subunit of a heterotrimeric G protein; this protein contains the C-terminal CAAX box characteristic of �subunits and has sequence similarity to � subunits implicated in the mating pathways of a variety of fungi. Disruption of thisgene was shown to cause sterility of MTLa mating cells and to block pheromone-induced gene expression and shmoo formation;deletion of just the CAAX box residues is sufficient to inactivate Ste18 function in the mating process. Intriguingly, ectopic ex-pression behind the strong ACT1 promoter of either the G� or the G� subunit of the heterotrimeric G protein is able to suppressthe mating defect caused by deletion of the G� subunit and restore both pheromone-induced gene expression and morphologychanges.

Heterotrimeric G protein-mediated signal transduction is ubiqui-tous and regulates many critical processes in eukaryotic cells (for

a review, see reference 1). The general paradigm for the function ofthese heterotrimeric G proteins is that the G� and G�� elements areactivated through dissociation or conformational changes triggeredby GTP binding to the G� subunit in response to ligand binding to aG protein-linked 7-transmembrane-spanning receptor protein (2).The signal is ultimately inactivated by reassociation after GTP hydro-lysis to GDP (3), often through the aid of a regulator of G proteinsignaling (RGS) protein that serves as a GTPase-activating protein(GAP) for the process (4).

The involvement of a heterotrimeric G protein in a fungal mat-ing signaling pathway was first noted for the baker’s yeast Saccha-romyces cerevisiae (5–7); for a review, see reference 8. Detailedgenetic analysis in S. cerevisiae suggested that the functioning ofthis module fit well to the paradigm established through biochem-ical analysis of mammalian G protein systems, in that the � and ��subunits played distinct (and opposing) roles in the signaling pro-cess. Although many of the well-studied mammalian pathwaysused the � subunit as the effector activator, the yeast system was aninitially identified member of the now-extensive class of pathwaysin which the �� subunit served to activate the downstream com-ponents of the pathway, in this case by directing the membraneassociation and activation of members of an Ste5p-scaffolded mi-togen-activated protein (MAP) kinase module (9, 10). The S.cerevisiae G� (ScG�) subunit serves a role primarily in downregu-lating the pheromone signaling pathway (11), while the activatedMAP kinase triggers the activity of the transcription factor Ste12p,which induces the transcription of pheromone-responsive genes(12), and also phosphorylates and stabilizes Far1p to initiate cellcycle arrest (13).

Recently the number of fungal mating pathways known toinvolve a heterotrimeric G protein has increased. The fungalpathogen Candida albicans has been demonstrated to show mat-ing between MTLa and MTL� strains in vivo (14) and in vitro (15)and to have a pheromone response pathway similar to that of S.cerevisiae (16). As is the case in S. cerevisiae, C. albicans also hasa unique G� subunit implicated in regulation of cyclic AMP

(cAMP) signaling, designated Gpa2 (17), while the classic hetero-trimeric G protein is implicated in the mating process. However,in the C. albicans system, loss of either the CaG� or CaG� of themating pathway heterotrimeric G protein (encoded by CAG1 andSTE4, respectively) results in full sterility (18), unlike in S. cerevi-siae, where loss of the G� subunit leads to constitutive signaling.In Schizosaccharomyces pombe, the pheromone signaling pathwayis regulated only by the SpG�1 subunit; it appears that neither theSpG� subunit nor the SpG� subunit is involved in this process(19). In Cryptococcus neoformans, there are three G� subunits, oneG� subunit, and two G� subunits; CnG�2 upregulates the pher-omone response pathway, while CnG�3 inhibits the signalingpathway (20). In Kluyveromyces lactis, the signal transductionpathway that mediates mating is positively regulated by both theKlG� (21) and KlG� (22) subunits of the heterotrimeric G pro-tein, while loss of the KlG� subunit produces only a minor matingdefect (23). Thus, while a heterotrimeric G protein is a commonelement in many fungal mating response pathways, the specificroles of the subunits appear to differ significantly from one systemto another.

In the present study, we have focused on the function of theheterotrimeric G protein G� subunit of the fungal pathogen C.albicans. Mating in C. albicans is a somewhat more complex pro-cess than that in S. cerevisiae or K. lactis, in that the typical C.albicans strain is a nonmating a/� diploid. Mating type homozy-

Received 6 December 2013 Accepted 28 January 2014

Published ahead of print 31 January 2014

Address correspondence to Yuan-Ying Jiang, 13761571578@163.com, or MalcolmWhiteway, malcolm.whiteway@concordia.ca.

* Present address: Hui Lu, Department of Pharmacy, Lanzhou General Hospital ofPLA, Lanzhou, Gansu, China.

Supplemental material for this article may be found at http://dx.doi.org/10.1128/EC.00320-13.

Copyright © 2014, American Society for Microbiology. All Rights Reserved.

doi:10.1128/EC.00320-13

April 2014 Volume 13 Number 4 Eukaryotic Cell p. 461– 469 ec.asm.org 461

on June 15, 2020 by guesthttp://ec.asm

.org/D

ownloaded from

gosis must occur prior to mating: a low percentage of clinicalisolates have become homozygous at the MTL locus, forming ei-ther a/a or �/� cells (24), and this homozygosis can be selected forin the lab by growth of a/� strains on sorbose medium (15). How-ever, homozygosis of the mating type locus MTL is not sufficientfor mating in C. albicans; it simply eliminates the a1-�2 repressorthat blocks potential activation of the mating-competent opaquestate (25). In the absence of the a1-�2 repressor, cells can establishthis mating-competent epigenetic state, which is characterized byhigh-level expression of the WOR1 gene and by unique cell andcolony morphologies (26). However, once the mating-competentopaque state is achieved, C. albicans mating proceeds in a mannersimilar to that of S. cerevisiae and K. lactis.

In S. cerevisiae, G� is critical for normal mating and loss ofG� blocks mating in strains with pathways activated either nor-mally, by overproduction of Ste4p (G�) (27), or by deletion ofGPA1(G�) (7). In contrast, in K. lactis cells, the G� subunit is

essentially dispensable for mating, as loss of G� compromisesmating only moderately and then only when both partners lackthe subunit (23). As the K. lactis and S. cerevisiae lineages divergedfrom the C. albicans lineage before the split between K. lactis and S.cerevisiae, it is of interest to establish what role the G� of thepathogen plays in mating signal transduction. We found that, as inS. cerevisiae, the C. albicans G� is required for mating. However, incontrast to the situation in S. cerevisiae, this requirement can beeliminated by ectopic expression of either the G� or G� subunit ofthe G protein.

MATERIALS AND METHODSStrains and culture conditions. The C. albicans strains and oligonucleo-tides used in this work are listed in Tables 1 and 2. For general growth andmaintenance of the strains in the white phase, the cells were cultured infresh YPD medium (1% yeast extract, 2% Bacto peptone, 2% dextrose, 2%agar for solid medium, pH 6.5) at 30°C. Strains were switched from the

TABLE 1 C. albicans strains used in this study

Strain Parent Mating type Description Source

3294 CNC43 a/a his1/his1 ura3/ura3 arg5,6/arg5,6 P. T. Magee3315 CNC43 �/� trp1/trp1 lys2/lys2 P. T. MageeLH001 3294 a/a STE18/ste18::HIS1 ura3/ura3 arg5,6/arg5,6 This studyLH002 LH001 a/a ste18::HIS1/ste18::URA3 arg5,6/arg5,6 This studyLH003 LH002 a/a ste18::HIS1/ste18::HIS1 ura3/ura3 arg5,6/arg5,6 This studyLH006 LH003 a/a ste18::HIS1/ste18::HIS1 ura3/ura3 arg5,6/arg5,6 (CIP10) This studyLH011 3294 a/a STE18-carboxy-terminal MTD/ste18-carboxy-terminal MTD::HIS1 ura3/ura3 arg5,6/arg5,6 This studyLH012 LH011 a/a ste18-carboxy-terminal MTD::HIS1/ste18-carboxy-terminal MTD::URA3 arg5,6/arg5,6 This studyLH004 LH003 a/a ste18::HIS1/ste18::HIS1 RPS1/rps1::STE18-URA3 arg5,6/arg5,6 This studyLH005 LH003 a/a ste18::HIS1/ste18::HIS1 RPS1/rps1::STE18�C-URA3 arg5,6/arg5,6 This studyLH021 LH003 a/a ste18::HIS1/ste18::HIS1 RPS1/rps1::act-STE4-URA3 arg5,6/arg5,6 This studyLH022 LH003 a/a ste18::HIS1/ste18::HIS1 RPS1/rps1::act-CAG1-URA3 arg5,6/arg5,6 This study

TABLE 2 Oligonucleotides used in this work

Name Description Sequence (5= to 3=)a

STE18-TB-F STE18 deletion PCR cassette forwardprimer

TTTTGATGTAAAAATTAACATGAAAGATTGTGTTTCAGAATTTTCTCCACCTACAACAACAACGACGACAATAACTAGATtatagggcgaattggagctc

STE18-TB-R STE18 and STE18�C deletion PCRcassette reverse primer

ATGTATATATATATATATAAATACATATGTGTGTGATTTCATTCTTGTGGGTTGATTAATTGGAGAACTATTTCTGTCGTgacggtatcgataagcttga

STE18�C-TB-F STE18�C deletion PCR cassette forwardprimer

TGGAGTTTACCTCCAGATCAGAATAGATTTGCCAAATATAAACAGTTGAGAAATGCACGCAATTCATCTCAAGCTACAGTTtatagggcgaattggagctc

STE18-F-ex STE18 forward external primer GATTATTACAAGGTGCATTTGCSTE18-R-ex STE18 reverse external primer AACATTGAAAGCTCAATTAGGCSTE18-F-in STE18 forward internal primer GAATTCAAGAGTTGACTAATCGHIS1-F HIS1 forward primer TTTAGTCAATCATTTACCAGACCGHIS1-R HIS1 reverse primer TCTATGGCCTTTAACCCAGCTGURA3-F URA3 forward primer TTGAAGGATTAAAACAGGGAGCURA3-R URA3 forward primer ATACCTTTTACCTTCAATATCTGGSTE18-SalI STE18 and STE18�C forward primer

for reintegrationCTGATGGTCGACGATTATTACAAGGTGCATTTGC

STE18-HindIII STE18 reverse primer for reintegration CCCAAGCTTGGGAACATTGAAAGCTCAATTAGGCSTE18�C-HindIII STE18�C reverse primer for

reintegrationCCCAAGCTTGGGTTAAACTGTAGCTTGAGATGAATTGCG

RPS1-R-in RPS1 reverse internal primer TTTCTGGTGAATGGGTCAACGACACT-F Actin promoter internal primer TTTTCTAATTTTCACTCCTGGCAG1-F-in CAG1 forward internal primer ATTGAACAAAGTTTACAATTGCGTCCAG1-R-in CAG1 reverse internal primer TCATTAGTATCGTCTGGTTTGCCSTE4-F-in STE4 forward internal primer ACTATACAACACCTTGCGAGGASTE4-R-in STE4 reverse internal primer CAGTTGCCAAAGCTACACCATCa Lowercase letters represent sequences used to prime the synthesis of the selectable marker, and boldface letters indicate introduced restriction enzyme cleavage sites.

Lu et al.

462 ec.asm.org Eukaryotic Cell

on June 15, 2020 by guesthttp://ec.asm

.org/D

ownloaded from

white phase to the opaque phase in two rounds of screening on plates withsynthetic N-acetylglucosamine (GlcNAc) (0.67% yeast nitrogen base,0.15% amino acid mix with uridine at 100 �g/ml, 2% GlcNAc, and 2%agar for solid medium) (28) and synthetic dextrose (SD) medium (0.67%yeast nitrogen base, 0.15% amino acid mix with uridine at 100 �g/ml, 2%dextrose, and 2% agar for solid medium). Phloxine B was added to nutri-ent agar for opaque colony staining (18). Cultures in SD medium at roomtemperature were used to maintain the cells in the opaque phase, and thetypical oblong cell morphology phenotype of the cells in the opaque phasewas confirmed by microscopy.

Disruption of the STE18 gene and deletion of the C terminus of theSTE18 gene. The C. albicans sequence (assembly 19) from the CandidaGenome Database (http://www.candidagenome.org/) was used as the ref-erence for the genomic sequence. The two alleles of the STE18 gene(ORF19.6551.1) were deleted from the MTLa strain 3294, obtained ini-tially from P. T. Magee (Table 1). All the disruptions were done using atwo-step PCR method as described previously (29), with the replacementof the first allele with HIS1 and of the second allele with URA3. Oligonu-cleotides STE18-TB-F and STE18-TB-R were used to prepare the cassettesfor the deletion of the STE18 gene. The strain produced by replacing thefirst copy of the STE18 gene by HIS1 in the MTLa strain 3294 was namedLH001. The correct insertion of the HIS1 cassette at the STE18 locus wasconfirmed by PCR analysis of genomic DNA from strain LH001 witholigonucleotides STE18-F-ex plus HIS1-R, STE18-R-ex plus HIS1-F, andSTE18-F-ex plus STE18-R-ex. Oligonucleotides STE18-F-ex and STE18-R-ex flank and are external to the recombination sites of the PCR cassettes.Oligonucleotides HIS1-F and HIS1-R are internal relative to the HIS1gene of the PCR cassettes. The second copy of the STE18 gene was deletedfrom strain LH001 by replacement with the URA3 cassette to generate theste18 null strain LH002. The correct insertion of the URA3 cassette at theSTE18 locus was confirmed by PCR with oligonucleotides STE18-F-explus URA3-R and STE18-R-ex plus URA3-F. The carboxyl terminus ofthe STE18 gene was deleted using a similar strategy. OligonucleotidesSTE18�C-TB-F and STE18-TB-R were used to prepare the PCR cassettes.The strain produced by deleting one allele from the parent strain 3294 wasnamed LH011. The correct insertion of the HIS1 cassette at the carboxylterminus of the STE18 gene was confirmed by PCR with oligonucleotidesSTE18-F-in plus HIS1-R and STE18-R-ex plus HIS1-F. Strain LH011 wasthen transformed with the URA3 cassette to remove the CAAX box of thesecond allele to STE18 gene to generate the carboxyl-terminal CAAX box-deleted strain LH012. The correct insertion site of the URA3 cassette wasconfirmed by PCR with oligonucleotides STE18-F-in plus STE18-R-ex.

Reintegration. A copy of the wild-type gene for complementationexperiments was reintegrated at the RPS1 locus in the ste18� strain asdescribed previously (18). The recipient strain LH002 was treated with5-fluoroorotic acid (5-FOA) to recover the URA3 marker. A resultinguridine-negative strain was named LH003. For the STE18 gene, a 1,320-bpDNA fragment from genomic DNA was amplified by PCR using oligonu-cleotides STE18-SalI and STE18-HindIII. Oligonucleotide STE18-SalIcontains an exogenous SalI restriction site, absent in the STE18 gene se-quence, near its 5= end, and STE18-HindIII contains an exogenous Hin-dIII restriction site in the 3=-end noncoding sequence of the STE18 gene.The PCR fragment was digested with SalI and HindIII, the resulting1.3-kb fragment was ligated with vector CIp10 (30) also cut with SalI andHindIII, and Escherichia coli strain DH5� was transformed with the con-struct. The integrity of the clone with respect to the STE18 wild-typesequence was confirmed by DNA sequencing. The selected clone for thewild-type STE18 gene was named plasmid pCIP-STE18 and was digestedwith the enzyme StuI for transformation of strain LH003. The new STE18strain was named LH004. A similar strategy and protocol were used for thereintegration of the STE18�C gene. An 820-bp fragment was amplifiedwith oligonucleotides STE18-SalI and STE18�C-HindIII. Oligonucleo-tide STE18�C-HindIII was designed with an exogenous HindIII restric-tion site, absent in the STE18 gene sequence, and is positioned in the 3=end of the STE18 gene open reading frame (ORF), which did not include

the last 21 bp (GGTTGTTGTACAATTGTTTAA). This PCR fragmentwas digested with SalI and HindIII, and the 820-bp fragment was ligatedto the CIp10 vector cleaved with the same two enzymes. The integrity ofthe clone was confirmed by DNA sequencing, and the selected clone withthe STE18�C gene sequence was named plasmid pCIP-STE18�C. Thisplasmid was digested with StuI, and strain LH003 was transformed withthe construct. This new STE18 strain was named LH005. The integrationof pCIP-STE18 and pCIP-STE18�C at the correct site in the RPS1 locuswas confirmed by PCR. A strain carrying an insertion of the originalvector CIP10 with no insert was also constructed and designated LH006.

Ectopic expression of either the STE4 gene or the CAG1 gene in ste18null mutant strains. To overexpress the STE4 gene or the CAG1 gene inste18� strains, we used plasmids pl390 (wild-type STE4 gene under con-trol of the ACT1 promoter) or pl391 (wild-type CAG1 gene under controlof the ACT1 promoter) (18) for transformation. Plasmids pl390 and pl391were digested with StuI and then transformed into ste18� strains. Theste18� strain transformed with plasmid pl390 was named LH021, whilethat with plasmid pl391 was named LH022. The correct integration of theplasmids was confirmed by PCR.

Microarray analysis. The transcriptional response to pheromonetreatment was measured in the wild-type strain, in a strain (LH003) de-leted for STE18, and in STE18-deleted strains overexpressing either theSTE4 gene (LH021) or the CAG1 gene (LH022). Standard protocols wereperformed as follows: we collected cells from SD-complete cultures in logphase with and without pheromone induction, and we used the hot-phe-nol method for RNA extraction (18), the polyA Spin mRNA isolation kitfor mRNA isolation (18), and reverse transcription for cDNA production,followed by indirect chemical labeling for dye addition. Arrays obtainedfrom NRC-BRI (18) were hybridized in a hybridization chamber at 42°Cfor overnight incubation and then washed and scanned. The scanning wasdone using a GenePix4000B microarray scanner, and the images wereanalyzed in GenePix Pro 4.1; the output data were processed with Mi-crosoft Excel 2013 and MultiExperiment Viewer (MeV).

Mating assays. Patch mating experiments, using auxotrophic markercomplementation with strain 3315 as the MTL� tester strain, were done asdescribed previously (31). Briefly, all the assayed strains were maintainedin the opaque phase at room temperature. Experimental and tester strainswere streaked as straight lines on separate YPD plates. After 24 h of incu-bation at room temperature, the two sets of streaks were crossed onto asingle fresh YPD plate and incubated for 24 h at room temperature. After24 h of incubation, cells were replicated to dropout plates containing SDmedium minus five amino acids (uridine, histidine, arginine, tryptophan,and lysine) for selection of mating products and to YPD plates as a control.All the plates were incubated at room temperature for 5 days prior toscoring.

Quantitative assays were done as follows. Opaque cells of the testerstrain 3315 and experimental strains were cultured overnight in SD-com-plete liquid medium in a shaker at room temperature. Each of the exper-imental strains was mixed with the tester strain in fresh SD-completeliquid medium (107 cells/ml for each strain) and then incubated in theshaker for 48 h at room temperature. Cells were quantified with a micro-scopic counting chamber, collected by centrifugation, washed, and resus-pended with water before plating onto SD-Trp� Lys�, SD-Arg�, or SD-Trp� Lys� Arg� medium for prototrophic selection and colony counting.

Pheromone response assay. Strains were examined microscopicallyfor shmoo formation after treatment with synthetic �-factor. Strains inopaque phase were incubated for 24 h at room temperature, diluted for afinal optical density at 600 nm (OD600) of 0.05, and treated with phero-mone (1 mg/ml) for 12 h before being photographed. Strains were visu-alized and photographed using a Nikon Eclipse TS100 microscope with amagnification of �400 using differential interference contrast (DIC) op-tics.

Alignment of G protein � subunit sequences in ascomycete yeasts.Multiple-protein-sequence alignments were performed with the MAFFTweb application (http://mafft.cbrc.jp/alignment/software/) and visual-

Ste18p in Candida albicans

April 2014 Volume 13 Number 4 ec.asm.org 463

on June 15, 2020 by guesthttp://ec.asm

.org/D

ownloaded from

ized with Jalviewer (version 2). The protein sequences of the differentascomycete yeast species were downloaded from the Fungal OrthogroupsRepository (http://www.broadinstitute.org/regev/orthogroups/) hostedby the Broad Institute, MIT. Pairwise protein alignments and identitypercentages were established with BLASTP, which is from BLAST (http://blast.ncbi.nlm.nih.gov/Blast.cgi).

Microarray data accession number. Microarray data have been de-posited in the GEO database under accession number GSE54031.

RESULTSORF19.6551.1 (CaSTE18) encodes a typical G� subunit of het-erotrimeric G protein in C. albicans. The C. albicans genomecontains a single copy of a gene (STE18 [ORF19.6551.1]) that hasthe structural and expression characteristics of a typical � subunitof heterotrimeric G proteins involved in mating. Expression ofthis gene is repressed by the a1/�2 repressor and is thus limited toMTL homozygous cells, in either the white or opaque state (32).This expression pattern is found for other genes such as FAR1 andSTE4, which have been shown to be part of the mating pheromoneresponse pathway in this organism (18, 33). Analysis of the de-duced primary structure of the protein showed a high degree ofidentity with the ScG� subunit of S. cerevisiae (40% identity and58% similarity) and the KlG� subunit of K. lactis (38% identityand 56% similarity), two other functionally characterized � sub-units of fungal mating response G proteins. CaG� is 90 aminoacids long and contains the conserved C-terminal CCAAX motif(CCTIV) that is a potential target for farnesylation at Cys87 andfor palmitoylation at the preceding Cys86 (34) (Fig. 1).

The CaG� subunit of the heterotrimeric G protein is re-quired for pheromone-induced gene induction and for matingin C. albicans. In S. cerevisiae, the ScG� subunit is required formating and loss of ScG� results in full sterility (7), but in K. lactisloss of the KlG� subunit leads to only a slight mating defect (23).Mating in C. albicans requires both the CaG� and CaG� subunitsof the heterotrimeric G proteins (18); when the genes encodingthe CaG�1 subunit or the CaG� subunit were deleted, the cellsbecame fully sterile. The effects of CaG� inactivation have notbeen investigated. In order to ascertain the role of the G protein �subunit in the pheromone response pathway in C. albicans, a nullmutation was created in strain 3294 (MTLa). This ste18 null mu-tant was generated by a complete deletion of the whole ORF withHIS1 and URA3 cassettes by homologous recombination accord-ing to the strategy described in Materials and Methods. The cor-rect insertion of the HIS1 and URA3 cassettes at the STE18 locus

was confirmed by PCR analysis of genomic DNA (see the supple-mental material).

We then identified opaque derivatives of the ste18 null mutantstrain on phloxine B plates and tested them for mating capacity.When the opaque ste18 null mutant strain was tested in a cross-patch mating assay, no prototrophic products derived from mat-ing were detected, showing that the ste18 null strain was totallysterile (Fig. 2). In addition, the strain was totally defective in pher-omone-induced gene expression, as treatment of opaque ste18strains with synthetic �-factor failed to induce expression of any ofa set of classic pheromone-responsive genes (Fig. 3). The cells werealso defective in shmoo formation in the presence of �-factor (seethe supplemental material). This sterile, nonresponsive pheno-type was a result of the loss of Ste18p function in the mutant strain,

FIG 1 Alignment of candidate G protein � subunit sequences encoded by 14 species of ascomycete yeasts. The deduced amino acid sequences of G� subunitsencoded by the C. albicans STE18 gene and its homologs in 13 related ascomycete yeasts are shown, and the conserved C-terminal tail is highlighted.

FIG 2 Ste18 null mutant strains are sterile. Mating was assayed by auxotrophicmarker complementation between strains of opposite mating types as de-scribed in Materials and Methods. The mating assay for MTLa ste18� strains,strains with STE18 reintegrated (ste18� STE18), and strains with STE18�Creintegrated (ste18� STE18�C) is shown. No colonies were formed by com-plementation of the ste18� strains (LH002 and LH006) and the strain withSTE18�C reintegrated (LH005), while the strain with STE18 reintegrated(LH004) reverted the sterile phenotype. WT, wild type. The original matingcross is shown at the bottom corner.

Lu et al.

464 ec.asm.org Eukaryotic Cell

on June 15, 2020 by guesthttp://ec.asm

.org/D

ownloaded from

as the reintroduction of a single copy of the STE18 gene at theRPS1 locus reestablished mating competence (Fig. 2). Thus,Ste18p, along with both the G� subunit Ste4p and the G� subunitCag1p, is a positive component in the pheromone response path-way of C. albicans.

The CAAX box of the CaG� subunit is critical for mating inC. albicans. The C terminus of the G� subunit is subjected to acomplex posttranslational processing that generates a hydropho-bic region of the protein through the addition of a methyl groupand lipid moieties (for a review, see reference 35). As an initialmeans of investigating whether this modified domain is requiredfor mating in C. albicans, we deleted the CAAX (cysteine, ali-phatic, aliphatic, X) box (CCTIV) of both copies of the CaSTE18gene (see the supplemental material). Opaque versions of doublymodified strains were identified on phloxine B plates to assess therole of the CaG� subunit carboxy terminus in the mating process.Like the ste18 null mutant strain, the ste18�C mutant strain wasunable to mate (Fig. 4). We also reintroduced a single copy of theSTE18�C gene (without the last 21 bp) at the RPS1 locus in theste18 null mutant strain, and this STE18�C strain was also unableto mate (Fig. 2). These results demonstrated that the carboxylterminus of CaSte18p is critical for mating in C. albicans; reversetranscription-PCR (RT-PCR) analysis showed that the CAAX-de-leted gene was expressed at normal levels (data not shown), so themating defect could not be attributed to reduced expression of themutant allele.

Ectopic expression of either the G� subunit or the G� sub-unit of the heterotrimeric G protein permits mating in the ab-sence of the G� subunit in C. albicans. In S. cerevisiae, overex-pression of the STE4 gene product led to cell cycle arrest of haploidcells and suppressed the sterility of cells defective in the matingpheromone receptors encoded by the STE2 and STE3 genes (27).The G� subunit of the heterotrimeric G protein triggered the

pheromone response pathway in the absence of the G� subunit inK. lactis. Overexpression of the CAG1 gene and the STE4 geneproduct in wild-type strains of C. albicans did not result in anyconstitutive expression of pheromone-responsive genes or lead toincreased shmoo formation or cell cycle arrest in the presence ofthe pheromone. In addition to this, overproduction of the STE4gene did not suppress the sterility caused by deletion of the CAG1gene, while the overexpression of the CAG1 gene was similarlyunable to suppress the sterility caused by deletion of the STE4 gene(18). We assessed the role of ectopic expression of the CAG1 andSTE4 genes in the absence of the STE18 gene in C. albicans byexpressing either the CAG1 gene or the STE4 gene using the strongACT1 promoter in the ste18 null mutant strain. This ectopic ex-pression had been shown to increase gene expression from 7- to15-fold in wild-type cells (18). Expression of either the CAG1 geneor the STE4 gene is able to suppress the mating defect caused bydeletion of the G� subunit (Fig. 5 and 6).

We further assessed the consequences of the ectopic expressionof the G� or G� subunit on pheromone-induced gene expressionin cells lacking STE18. The strains containing G� or G� expressedunder ACT1 control showed a higher level of gene induction in thepresence of �-factor than did the wild-type cells (Fig. 3). Thisenhancement in response was associated with an increase in quan-titative mating over wild-type levels as well (Table 3). Intriguingly,although all the responsive strains showed shmoo formation (seethe supplemental material), they showed no evidence for a clearcell cycle arrest leading to the formation of halos in a classic haloassay (data not shown).

DISCUSSION

The � subunits of heterotrimeric G proteins play key roles in thefunction of these important signaling molecules. Structural stud-ies have shown that � subunits form a very stable coiled-coil in-teraction with the N terminus of a G� subunit; this association isalmost as stable as that which would be generated by a covalent

FIG 3 Transcriptional response to pheromone treatment. The values shownrepresent the averages from 3 independent biological samples for 17 phero-mone-inducible genes from C. albicans. The stronger signals have the deeperred color. The final column presents a graphical summary of the data. Deletionof the STE18 gene eliminates the pheromone induction of the genes, whileectopic expression of either STE4 or CAG1 enhances the responsiveness toinduction by pheromone treatment. The complete data files are accessible atGEO through accession no. GSE54031.

FIG 4 Strains with C termini of CaG� subunits (CCTIV) deleted are sterile.The mating assay was done as described in Materials and Methods. No pro-totrophic colonies from the ste18�C strain LH012 were detected after 5 days ofincubation at room temperature. WT, wild type.

Ste18p in Candida albicans

April 2014 Volume 13 Number 4 ec.asm.org 465

on June 15, 2020 by guesthttp://ec.asm

.org/D

ownloaded from

linkage between the proteins (36). In addition, G� subunits typi-cally undergo a complex posttranslational modification process oftheir carboxyl terminus. These proteins have a CAAX (cysteine,aliphatic, aliphatic, X) box motif at their C termini, and the cys-teine residue of this motif is a target for the addition of both aprenyl (typically a farnesyl or geranylgeranyl) residue and amethyl residue (for a review, see reference 35). These modifica-tions, and a further addition of a palmitoyl residue to a secondcysteine often found adjacent to the CAAX box, generate a highlyhydrophobic G� C terminus that helps to anchor the protein,

together with the attached G� subunit, to cellular membranes (fora review, see reference 37).

Deletion of the G� subunit of the C. albicans mating responseG protein generates sterility. In the present study, this test wasdone only in the MTLa background, but there is no reason toanticipate that the function is in any way mating type dependent.The positive functioning of the G� subunit is similar to the situa-tion in S. cerevisiae, where the loss of the G protein � subunitSte18p results in the loss of mating competence (7). However, therole of the G� subunit in C. albicans contrasts with the role of G�subunit in K. lactis, where the loss of the G� subunit produces onlya slight mating defect (23). Because S. cerevisiae and K. lactis di-verged after the split from the common ancestor with C. albicans,it would appear that the ancestral G� subunit was essential forpheromone response and mating and that this role became re-duced in the lineage leading to K. lactis. However, the observationthat ectopic expression of either the G� or G� subunit of the C.albicans heterotrimeric G protein can suppress the mating defectcaused by deletion of the G� subunit suggests that the molecularrole of the ancestral G� may actually be closer to that of the K.lactis protein than to that of the S. cerevisiae one. In S. cerevisiae,changes in the stoichiometry of the � or � subunit did not influ-ence the need for the G� protein. Furthermore, engineering theG� subunit with the addition of a C-terminal membrane anchorwas not sufficient to allow signal transduction, implying that theG� subunit played a role beyond simply serving as a membraneattachment (10). Because directly anchoring the Ste5p scaffold tothe membrane through the addition of a membrane attachmentmotif activated the response pathway (10), a likely nonanchoringrole of the G� subunit in yeast was to serve as part of the Ste5pbinding interface. Considering that the overproduction of eitherthe C. albicans G� or G� subunit can bypass the need forCaSte18p, it is likely that G� is not an essential component in anySte5p binding surface for this organism. The C. albicans Ste5p isconsiderably smaller than the yeast Ste5 protein, and the bindinginterface with the G protein, if any, is currently undefined (38, 39).Similarly, the K. lactis system does not have an essential role for theG� subunit, perhaps because the normal level of the G� or the G�protein in this organism was sufficient to permit mating in itsabsence. K. lactis has a candidate Ste5 protein, but its molecularfunction has not yet been assessed, and its link, if any, to the Gprotein has not been established.

The observation that deletion of the CAAX box of CaSte18p

FIG 5 Ectopic expression the STE4 gene in a ste18� strain reverts the sterilephenotype. The mating assay was done as described in Materials and Methods.Prototrophic colonies from the ste18� strain with the STE4 gene overex-pressed were detected after 5 days of incubation at room temperature. WT,wild type.

FIG 6 Ectopic expression of the CAG1 gene in a ste18� strain reverts the sterilephenotype. The mating assay was done as described in Materials and Methods.Prototrophic colonies from the ste18� strain with the CAG1 gene overex-pressed (LH022) were detected after 5 days of incubation at room temperature.WT, wild type.

TABLE 3 Quantitative mating of wild-type strain 3294 and mutantstrains

Relevant genotype (strain)Mating frequency(104, mean SD)a

Wild type (3294) 2.63 0.21�ste18 (LH003) 0�ste18 STE18 (LH004) 6.84 0.45�ste18 STE18�C (LH005) 0�ste18 CIP10 (LH006) 0STE18�C (LH012) 0�ste18 STE4 (LH021) 12.0 0.6�ste18 CAG1 (LH022) 4.83 0.87a The STE18 replacement allele and the ectopically expressed CAG1 and STE4 genesenhance mating over the wild-type level. The mating frequency represents the numberof observed prototrophic colonies divided by the number of potentially mating-competent input wild-type 3294 cells or input derivative 3294 cells.

Lu et al.

466 ec.asm.org Eukaryotic Cell

on June 15, 2020 by guesthttp://ec.asm

.org/D

ownloaded from

causes sterility while overproduction of the G� or G� subunit cansuppress even complete deletion of the G� gene further suggeststhat membrane association of the G protein is a key component offunction in the C. albicans mating pathway. It is possible that acritical role of this membrane association is to bring the G� sub-unit into proximity with a membrane-linked effector. This couldexplain the ability of overproduction of either G� itself or G� tobypass the need for G�; subunit overproduction would generatemore �� dimers, and with G� having its own membrane attach-ment capability due to myristoylation, this increases the overall

amount of membrane-associated G� (Fig. 7). The C. albicans andK. lactis pathways may thus represent relatively unspecialized het-erotrimeric G protein modules, with all subunits playing a positiverole in the mating process and the G� subunit being relativelydispensable, while S. cerevisiae is more specialized, with the � and�� subunits having distinct functions and the G� subunit beingcritical for effector activation. It is also intriguing that the ectopicexpression of G� or G� in the ste18 null mutant actually enhancedmating responsiveness measured by pheromone-induced gene ex-pression, while similar ectopic expression in wild-type cells had

FIG 7 Roles of G protein subunits in the mating process in C. albicans. All three G protein subunits are required for the mating (A). Loss of any one of them leadsto full sterility (B, C, and D). The CAAX box of the Ste18p is critical for mating of C. albicans (E). Ectopic expression behind the strong ACT1 promoter of theSTE4 gene did not rescue the sterility caused by absence of the G� subunit, while the ectopic expression of the CAG1 gene was similarly unable to suppressthe sterility caused by the absence of the G� subunit (H and I). Intriguingly, the ectopic expression of either the CAG1 gene or the STE4 gene is able to suppressthe mating defect caused by deletion of the G� subunit (F and G).

Ste18p in Candida albicans

April 2014 Volume 13 Number 4 ec.asm.org 467

on June 15, 2020 by guesthttp://ec.asm

.org/D

ownloaded from

little effect (18). This suggests that while the signaling from ectop-ically expressed G� or G� in the absence of Ste18 is capable ofallowing mating, the altered subunit stoichiometry affects the pro-cess quantitatively.

Overall the functions of only a limited number of fungalmating pathway G� subunits have been assessed, but their roleshave been found to be surprisingly variable. Less functionalvariability has been noted for higher eukaryotes. However, themammalian G�5 subunit has been found to associate with reg-ulator of G protein signaling (RGS) proteins with a G� homol-ogy region, as well as with classical G� subunits (40), showingthat even in higher eukaryotes, G protein function can acceptvariation in the involvement of the G� subunit. It is likely thatcontinued genetic and biochemical analysis of the functions ofG protein systems in fungal mating pathways will provide im-portant insights into the functional plasticity of this importantclass of signaling molecules.

ACKNOWLEDGMENTS

We thank Daniel Dignard (Biotechnology Research Institute) for plas-mids pl390 and pl391. We thank Cunle Wu (Biotechnology ResearchInstitute) for helpful discussion about this study. We thank HannahRegan (McGill University) and Pierre Cote (Concordia University) fortheir comments on the manuscript.

Hui Lu was supported by a scholarship from the China ScholarshipCouncil on the MOE-NRC Research and Postdoctoral Fellowship Pro-gram. Work in the Whiteway lab was supported in part by CIHR grantMOP 42516.

REFERENCES1. Oldham WM, Hamm HE. 2008. Heterotrimeric G protein activation by

G-protein-coupled receptors. Nat. Rev. Mol. Cell Biol. 9:60 –71. http://dx.doi.org/10.1038/nrm2299.

2. Tesmer JJ. 2010. The quest to understand heterotrimeric G protein sig-naling. Nat. Struct. Mol. Biol. 17:650 – 652. http://dx.doi.org/10.1038/nsmb0610-650.

3. Sprang SR. 1997. G protein mechanisms: insights from structural analy-sis. Annu. Rev. Biochem. 66:639 – 678. http://dx.doi.org/10.1146/annurev.biochem.66.1.639.

4. Hollinger S, Hepler JR. 2002. Cellular regulation of RGS proteins: mod-ulators and integrators of G protein signaling. Pharmacol. Rev. 54:527–559. http://dx.doi.org/10.1124/pr.54.3.527.

5. Dietzel C, Kurjan J. 1987. The yeast SCG1 gene: a G alpha-like proteinimplicated in the a- and alpha-factor response pathway. Cell 50:1001–1010. http://dx.doi.org/10.1016/0092-8674(87)90166-8.

6. Miyajima I, Nakafuku M, Nakayama N, Brenner C, Miyajima A, Kai-buchi K, Arai K, Kaziro Y, Matsumoto K. 1987. GPA1, a haploid-specificessential gene, encodes a yeast homolog of mammalian G protein whichmay be involved in mating factor signal transduction. Cell 50:1011–1019.http://dx.doi.org/10.1016/0092-8674(87)90167-X.

7. Whiteway M, Hougan L, Dignard D, Thomas DY, Bell L, Saari GC,Grant FJ, O’Hara P, MacKay VL. 1989. The STE4 and STE18 genes ofyeast encode potential beta and gamma subunits of the mating factorreceptor-coupled G protein. Cell 56:467– 477. http://dx.doi.org/10.1016/0092-8674(89)90249-3.

8. Bardwell L. 2005. A walk-through of the yeast mating pheromone responsepathway. Peptides 26:339–350. http://dx.doi.org/10.1016/j.peptides.2004.10.002.

9. Whiteway MS, Wu C, Leeuw T, Clark K, Fourest-Lieuvin A, ThomasDY, Leberer E. 1995. Association of the yeast pheromone response Gprotein beta gamma subunits with the MAP kinase scaffold Ste5p. Science269:1572–1575. http://dx.doi.org/10.1126/science.7667635.

10. Pryciak PM, Huntress FA. 1998. Membrane recruitment of the kinasecascade scaffold protein Ste5 by the Gbetagamma complex underlies acti-vation of the yeast pheromone response pathway. Genes Dev. 12:2684 –2697. http://dx.doi.org/10.1101/gad.12.17.2684.

11. Metodiev MV, Matheos D, Rose MD, Stone DE. 2002. Regulation of

MAPK function by direct interaction with the mating-specific Galpha inyeast. Science 296:1483–1486. http://dx.doi.org/10.1126/science.1070540.

12. Breitkreutz A, Boucher L, Tyers M. 2001. MAPK specificity in the yeastpheromone response independent of transcriptional activation. Curr.Biol. 11:1266 –1271. http://dx.doi.org/10.1016/S0960-9822(01)00370-0.

13. Butty AC, Pryciak PM, Huang LS, Herskowitz I, Peter M. 1998. The roleof Far1p in linking the heterotrimeric G protein to polarity establishmentproteins during yeast mating. Science 282:1511–1516. http://dx.doi.org/10.1126/science.282.5393.1511.

14. Hull CM, Raisner RM, Johnson AD. 2000. Evidence for mating of the“asexual” yeast Candida albicans in a mammalian host. Science 289:307–310. http://dx.doi.org/10.1126/science.289.5477.307.

15. Magee BB, Magee PT. 2000. Induction of mating in Candida albicans byconstruction of MTLa and MTLalpha strains. Science 289:310 –313. http://dx.doi.org/10.1126/science.289.5477.310.

16. Chen J, Lane S, Liu H. 2002. A conserved mitogen-activated proteinkinase pathway is required for mating in Candida albicans. Mol. Micro-biol. 46:1335–1344. http://dx.doi.org/10.1046/j.1365-2958.2002.03249.x.

17. Bennett RJ, Johnson AD. 2006. The role of nutrient regulation and the Gpa2protein in the mating pheromone response of C. albicans. Mol. Microbiol. 62:100–119. http://dx.doi.org/10.1111/j.1365-2958.2006.05367.x.

18. Dignard D, Andre D, Whiteway M. 2008. Heterotrimeric G-proteinsubunit function in Candida albicans: both the alpha and beta subunits ofthe pheromone response G protein are required for mating. Eukaryot. Cell7:1591–1599. http://dx.doi.org/10.1128/EC.00077-08.

19. Shpakov AO, Pertseva MN. 2008. Signaling systems of lower eukaryotesand their evolution. Int. Rev. Cell Mol. Biol. 269:151–282. http://dx.doi.org/10.1016/S1937-6448(08)01004-6.

20. Hsueh YP, Xue C, Heitman J. 2007. G protein signaling governing cellfate decisions involves opposing Galpha subunits in Cryptococcus neofor-mans. Mol. Biol. Cell 18:3237–3249. http://dx.doi.org/10.1091/mbc.E07-02-0133.

21. Savinon-Tejeda AL, Ongay-Larios L, Valdes-Rodriguez J, Coria R. 2001.The KlGpa1 gene encodes a G-protein alpha subunit that is a positivecontrol element in the mating pathway of the budding yeast Kluyveromy-ces lactis. J. Bacteriol. 183:229 –234. http://dx.doi.org/10.1128/JB.183.1.229-234.2001.

22. Kawasaki L, Savinon-Tejeda AL, Ongay-Larios L, Ramirez J, Coria R.2005. The Gbeta(KlSte4p) subunit of the heterotrimeric G protein has apositive and essential role in the induction of mating in the yeastKluyveromyces lactis. Yeast 22:947–956. http://dx.doi.org/10.1002/yea.1278.

23. Navarro-Olmos R, Kawasaki L, Dominguez-Ramirez L, Ongay-LariosL, Perez-Molina R, Coria R. 2010. The beta subunit of the heterotrimericG protein triggers the Kluyveromyces lactis pheromone response pathwayin the absence of the gamma subunit. Mol. Biol. Cell 21:489 – 498. http://dx.doi.org/10.1091/mbc.E09-06-0472.

24. Soll DR. 2004. Mating-type locus homozygosis, phenotypic switching andmating: a unique sequence of dependencies in Candida albicans. Bioessays26:10 –20. http://dx.doi.org/10.1002/bies.10379.

25. Zordan RE, Galgoczy DJ, Johnson AD. 2006. Epigenetic properties ofwhite-opaque switching in Candida albicans are based on a self-sustainingtranscriptional feedback loop. Proc. Natl. Acad. Sci. U. S. A. 103:12807–12812. http://dx.doi.org/10.1073/pnas.0605138103.

26. Zordan RE, Miller MG, Galgoczy DJ, Tuch BB, Johnson AD. 2007.Interlocking transcriptional feedback loops control white-opaque switch-ing in Candida albicans. PLoS Biol. 5:e256. http://dx.doi.org/10.1371/journal.pbio.0050256.

27. Whiteway M, Hougan L, Thomas DY. 1990. Overexpression of the STE4gene leads to mating response in haploid Saccharomyces cerevisiae. Mol.Cell. Biol. 10:217–222.

28. Huang G, Yi S, Sahni N, Daniels KJ, Srikantha T, Soll DR. 2010.N-Acetylglucosamine induces white to opaque switching, a mating pre-requisite in Candida albicans. PLoS Pathog. 6:e1000806. http://dx.doi.org/10.1371/journal.ppat.1000806.

29. Dignard D, El-Naggar AL, Logue ME, Butler G, Whiteway M. 2007.Identification and characterization of MFA1, the gene encoding Candidaalbicans a-factor pheromone. Eukaryot. Cell 6:487– 494. http://dx.doi.org/10.1128/EC.00387-06.

30. Murad AM, Lee PR, Broadbent ID, Barelle CJ, Brown AJ. 2000. CIp10,an efficient and convenient integrating vector for Candida albicans. Yeast16:325–327. http://dx.doi.org/10.1002/1097-0061(20000315)16:4�325::AID-YEA538�3.0.CO;2-#.

Lu et al.

468 ec.asm.org Eukaryotic Cell

on June 15, 2020 by guesthttp://ec.asm

.org/D

ownloaded from

31. Dignard D, Whiteway M. 2006. SST2, a regulator of G-protein signalingfor the Candida albicans mating response pathway. Eukaryot. Cell 5:192–202. http://dx.doi.org/10.1128/EC.5.1.192-202.2006.

32. Srikantha T, Borneman AR, Daniels KJ, Pujol C, Wu W, SeringhausMR, Gerstein M, Yi S, Snyder M, Soll DR. 2006. TOS9 regulates white-opaque switching in Candida albicans. Eukaryot. Cell 5:1674 –1687. http://dx.doi.org/10.1128/EC.00252-06.

33. Cote P, Whiteway M. 2008. The role of Candida albicans FAR1 in regulationof pheromone-mediated mating, gene expression and cell cycle arrest. Mol.Microbiol. 68:392–404. http://dx.doi.org/10.1111/j.1365-2958.2008.06158.x.

34. Hirschman JE, Jenness DD. 1999. Dual lipid modification of the yeastggamma subunit Ste18p determines membrane localization ofGbetagamma. Mol. Cell. Biol. 19:7705–7711.

35. McIntire WE. 2009. Structural determinants involved in the formationand activation of G protein betagamma dimers. Neurosignals 17:82–99.http://dx.doi.org/10.1159/000186692.

36. Sondek J, Bohm A, Lambright DG, Hamm HE, Sigler PB. 1996. Crystalstructure of a G-protein beta gamma dimer at 2.1A resolution. Nature379:369 –374. http://dx.doi.org/10.1038/379369a0.

37. Gautam N, Downes GB, Yan K, Kisselev O. 1998. The G-protein be-tagamma complex. Cell Signal. 10:447– 455. http://dx.doi.org/10.1016/S0898-6568(98)00006-0.

38. Cote P, Sulea T, Dignard D, Wu C, Whiteway M. 2011. Evolutionaryreshaping of fungal mating pathway scaffold proteins. mBio 2(1):e00230–10.

39. Yi S, Sahni N, Daniels KJ, Lu KL, Huang G, Garnaas AM, Pujol C,Srikantha T, Soll DR. 2011. Utilization of the mating scaffold protein inthe evolution of a new signal transduction pathway for biofilm develop-ment. mBio 2(1):e00237–10.

40. Xie K, Allen KL, Kourrich S, Colon-Saez J, Thomas MJ, Wickman K,Martemyanov KA. 2010. G�5 recruits R7 RGS proteins to GIRK channelsto regulate the timing of neuronal inhibitory signaling. Nat. Neurosci.13:661– 663. http://dx.doi.org/10.1038/nn.2549.

Ste18p in Candida albicans

April 2014 Volume 13 Number 4 ec.asm.org 469

on June 15, 2020 by guesthttp://ec.asm

.org/D

ownloaded from