|Home | About | Journals | Submit | Contact Us | Français|
Members of the dual-specificity tyrosine-phosphorylated and regulated kinase (DYRK) family perform a variety of functions in eukaryotes. We used gene disruption, targeted pharmacologic inhibition, and genome-wide transcriptional profiling to dissect the function of the Yak1 DYRK in the human fungal pathogen Candida albicans. C. albicans strains with mutant yak1 alleles showed defects in the yeast-to-hypha transition and in maintaining hyphal growth. They also could not form biofilms. Despite their in vitro filamentation defect, C. albicans yak1Δ/yak1Δ mutants remained virulent in animal models of systemic and oropharyngeal candidiasis. Transcriptional profiling showed that Yak1 was necessary for the up-regulation of only a subset of hypha-induced genes. Although downstream targets of the Tec1 and Bcr1 transcription factors were down-regulated in the yak1Δ/yak1Δ mutant, TEC1 and BCR1 were not. Furthermore, 63% of Yak1-dependent, hypha-specific genes have been reported to be negatively regulated by the transcriptional repressor Tup1 and inactivation of TUP1 in the yak1Δ/yak1Δ mutant restored filamentation, suggesting that Yak1 may function upstream of Tup1 in governing hyphal emergence and maintenance.
Candida albicans is an opportunistic fungal pathogen in humans, causing various forms of candidiasis ranging from superficial mucosal infections to life-threatening systemic diseases, predominantly in patients with a compromised immune system. In addition, C. albicans, like many pathogens, adheres to and subsequently form biofilms on implanted devices, especially on indwelling intravascular catheters. Organisms growing in biofilms have decreased susceptibility to antifungals (Donlan and Costerton, 2002 ; Douglas, 2003 ; Nobile and Mitchell, 2006 ). C. albicans is a polymorphic organism that undergoes reversible morphogenetic transitions between budding yeast, pseudohyphal and hyphal growth forms. The ability to transition from the yeast form to the hyphal form has been considered a major requirement for virulence (Saville et al., 2003 ), but it is also critical for biofilm formation (Ramage et al., 2002 ; Richard et al., 2005 ).
A variety of factors, including temperature, amino acids, pH changes, and serum can induce the yeast-to-hypha transition in C. albicans (Ernst, 2000 ; Eckert et al., 2007 ). The contribution of a range of signaling pathways to morphogenesis has been investigated, suggesting that the yeast-to-hypha transition is the subject of both positive and negative regulation (Eckert et al., 2007 ; Whiteway and Bachewich, 2007 ). Central to the positive regulation of filamentation is the transcription factor Efg1 that acts downstream of the cAMP signaling pathway. Indeed, a C. albicans efg1Δ/efg1Δ mutant is unable to form hyphae under a variety of hypha-inducing conditions (Lo et al., 1997 ; Stoldt et al., 1997 ), indicating that Efg1 plays a critical role at the onset of filamentation. Nevertheless, it has also been reported that efg1Δ/efg1Δ mutants display increased filamentation during growth under anaerobic or microaerophilic (embedded) conditions, suggesting that Efg1 acts as a possible repressor of filamentation under certain conditions (Giusani et al., 2002 ; Doedt et al., 2004 ). Negative regulation of filamentation is achieved by the general repressor Tup1 in association with the DNA-binding proteins Nrg1 and Rfg1. These negative regulators of filamentation act independently of Efg1 and other positive regulators of filamentation (Braun and Johnson, 1997 ; Braun et al., 2001 ; Kadosh and Johnson, 2001 ; Murad et al., 2001b ). The interplay between positive and negative regulation of the yeast-to-hypha switch is illustrated by the complex structure of the promoter regions of hypha-specific genes such as ALS3 and HWP1, which integrate inputs from multiple activators and repressors and are under the control of chromatin remodeling proteins (Argimon et al., 2007 ; Kim et al., 2007 ). To date, the exact relationships, interconnections, and feedback regulations of the signaling pathways and transcription factors that positively and negatively regulate filamentation remain largely uncharacterized.
Candida glabrata is another pathogenic yeast that represents the second leading cause of candidiasis (Pfaller and Diekema, 2007 ). C. glabrata is relatively distantly related to C. albicans, and it has only been reported to form pseudohyphae (Csank and Haynes, 2000 ). However, it is able to form biofilms, and we have previously taken advantage of this property to identify C. glabrata genes that are involved in biofilm formation (Iraqui et al., 2005 ). A genetic screen for C. glabrata insertional mutants altered in their ability to form biofilms enabled us to identify four genes involved in biofilm formation: YAK1, encoding a protein kinase; RIF1 and SIR4, encoding two components of the subtelomeric silencing machinery; and EPA6, encoding a member of the C. glabrata family of Epa adhesins (Iraqui et al., 2005 ). Experimental evidence suggested a model whereby the C. glabrata Yak1 kinase is required to repress the subtelomeric silencing machinery and in turn allow expression of the EPA6 and EPA7 adhesin genes that contribute to cell-to-cell adhesion within biofilms (Iraqui et al., 2005 ). In C. albicans, adhesins such as Als3 and Als1 are structurally related to the C. glabrata Epa adhesins, and they have been implicated in biofilm formation (Chandra et al., 2001 ; Garcia-Sanchez et al., 2004 ; Nobile et al., 2006a ; Zhao et al., 2006 ). However, C. albicans ALS genes are not located at subtelomeres (van het Hoog et al., 2007 ), suggesting that mechanisms other than silencing are involved in the regulation of these genes during biofilm formation.
In the present study, we investigated the C. albicans Yak1 kinase, a member of the family of dual-specificity tyrosine-phosphorylated and regulated kinases (DYRKs; Becker and Joost, 1999 ). The founding member of the DYRK family, the Saccharomyces cerevisiae Yak1 kinase, was originally identified as a suppressor of the lethality associated with the loss of either RAS function or the Tpk1, Tpk2, and Tpk3 catalytic subunits of the cAMP-dependent protein kinase A (PKA) (Garrett and Broach, 1989 ). The different phenotypes associated with deletion or overexpression of the S. cerevisiae YAK1 gene suggest that Yak1 acts as a negative regulator of growth and modulates PKA-regulated processes such as heat sensitivity and pseudohyphal growth (Garrett and Broach, 1989 ; Hartley et al., 1994 ; Ward and Garrett, 1994 ; Zhang et al., 2001 ). The function of Yak1 may be directly regulated by PKA, because it is phosphorylated in vitro and in vivo by PKA (Garrett et al., 1991 ; Zappacosta et al., 2002 ; Ptacek et al., 2005 ). S. cerevisiae Yak1 and other DYRKs are Ser/Thr protein kinases whose activity is modulated by autophosphorylation at a conserved tyrosine residue (Garrett et al., 1991 ; Kassis et al., 2000 ). Known targets of S. cerevisiae Yak1 include Bcy1, the regulatory subunit of PKA; Crf1, a corepressor of the forkhead-like transcription factor Fhl1 involved in the repression of ribosomal protein genes; and Pop2, which is required for expression of glucose-repressed genes (Griffioen et al., 2001 ; Moriya et al., 2001 ; Martin et al., 2004 ). Phosphorylation by Yak1 modulates the nucleocytoplasmic distribution of these proteins and their contribution to growth-related processes. Recent results indicate that S. cerevisiae Yak1 may influence the nucleocytoplasmic distribution of the Msi1 (Cac3) protein and its association in the chromatin assembly factor-I (CAF-I), suggesting a potential role of Yak1 in regulating chromatin structure (Pratt et al., 2007 ).
In this study, we investigated the role of C. albicans Yak1 in biofilm formation, the yeast-to-hypha transition, and hyphal maintenance by using null and conditional alleles of the C. albicans YAK1 gene. We identified genes whose up-regulation at the yeast-to-hypha switch is dependent upon a functional Yak1, and we showed that Yak1 is not required for the proper activation of the cAMP/PKA/Efg1 pathway and transcription factors necessary for the up-regulation of Yak1-dependent hypha-specific genes. Moreover, we demonstrated that deletion of TUP1 could bypass the Yak1 requirement for filamentation. Together, our results suggest that Yak1 may function in the Tup1 pathway to govern hyphal emergence and maintenance.
C. albicans strains used in this study are listed in Table 1. These strains were routinely grown at 30°C in YPD (1% yeast extract, 2% peptone, and 2% glucose) or in YNB (0.67% yeast nitrogen base without amino acids; Difco, Detroit, MI) with 0.4 or 2% glucose, supplemented with 20 mg/l histidine, 20 mg/l arginine, and 40 mg/l uridine when necessary. Different protocols were used to induce hyphal formation on solid medium or in liquid culture. Filamentation on plate was assessed on YPD FCS 20% (YPD, 20% fetal calf serum [Sigma-Aldrich, St. Louis, MO]), and 1.5% Bacto-agar) and on RPMI medium (1× RPMI [Invitrogen, Carlsbad, CA] buffered with 50 mM HEPES, pH 7.5, and 1.5% Bacto-agar). Cells (5 × 105) in 5 μl of water were spotted on plates that were incubated at 37°C, under atmospheric or 5% CO2 concentration for 2–10 d. Agar invasion was tested by washing the cells at the surface of YPD FCS 20% or RPMI plates and observing whether cells remained attached to the agar. Filamentation in embedded condition was done by embedding ≈100 yeast cells between two layers of YPS medium (1% yeast extract, 2% peptone, 2% saccharose, and 1% agar), and then incubating plates at 22°C for 7 d (Brown et al., 1999 ). Colonies on solid agar were observed with an SZX9 stereomicroscope at 16–57× and photographed using a DP-12 digital system (Olympus, Melville, NY). Filamentation in liquid media was assessed by inoculating 2–5 × 106 cells ml−1 of an overnight culture grown in YPD at 30°C in Lee's medium (Lee et al., 1975 ) or YPD FCS 10% (Sigma-Aldrich) and incubation with shaking at 37°C. Filamentation was assessed at various time points. 4-Amino-1-(tert-butyl)-3-(1′-naphthylmethyl)pyrazolo-(3,4-d)-pyrimidine (1-NMPP1; Toronto Research Chemicals, North York, ON, Canada) was dissolved in dimethyl sulfoxide at 30 mM and used at the final concentrations of 1.5–5 μM. cAMP was added to solid or liquid media at the final concentration of 5 mM. Cells were stained with 5 μg/ml Calcofluor white for 5 min, and then they were examined on a Nikon eclipse E 600 microscope with a 40× objective under fluorescence conditions, and photographed.
All strains constructed for this study are derivatives of BWP17 (Wilson et al., 1999 ). The lithium-acetate procedure was used to transform C. albicans as described previously (Walther and Wendland, 2003 ). Replacement of the complete open reading frames (ORFs) of both alleles of C. albicans YAK1 by polymerase chain reaction (PCR)-generated ARG4 and HIS1 disruption cassettes flanked by 100 base pairs of target homology region was performed through sequential transformation as described by Gola et al. (2003) . Primers used for amplification of the ARG4 and HIS1 cassettes and confirmation of the deletions are listed in Table 2. Independent transformants were produced and the disruptions were verified by PCR on whole yeast cells as described previously (Gola et al., 2003 ) and by Southern analysis using a 834-base pair PCR product (primers YakF1 and YakR1) encompassing 109 base pairs of the 5′-noncoding region and the first 725 coding nucleotides (nt) of the YAK1 gene as a probe on EcoRI-digested genomic DNA prepared with the Yeast DNA Purification kit (Epicenter Technologies, Madison, WI). Before phenotypic analysis, if necessary, the strains were sequentially converted to histidine prototrophy or arginine prototrophy by using the HIS1 or ARG4 genes that had been amplified from the pFA-HIS1 or pFA-ARG4 plasmids, respectively (Gola et al., 2003 ), and to uracil prototrophy using the StuI-linearized CIp10 plasmid that carries the URA3 gene and integrates at the RPS10 locus (Murad et al., 2000 ).
The C. albicans YAK1 gene along with 885 base pairs of the 5′-noncoding and 629 base pairs of the 3′-noncoding regions was amplified with primers YakSalI and YakEcoRV (Table 2). After digestion with SalI and EcoRV, the PCR product was inserted into XhoI, EcoRV-digested CIp10 (Murad et al., 2000 ), yielding pCIP-YAK1. Site-directed mutagenesis of the YAK1 gene was performed using pCIP-YAK1 as a template for amplification and oligonucleotides YakA1 and YakA2 (Table 2) carrying mutations that direct the change of Phe507 into Ala in the Yak1 protein and introduce an AluI restriction site in the YAK1 ORF. After amplification, DpnI digestion was used to counterselect parental DNA. The occurrence of the mutation in derivatives of pCIP-YAK1 obtained upon Escherichia coli K-12 transformation was verified first by testing the presence of the AluI site and then by sequence analysis. This resulted in the identification of plasmid pCIP-YAK1F507A that differs from pCIP-YAK1 by the expected nucleotide changes only. StuI-linearized pCIP-YAK1 and pCIP-YAK1F507A were used to transform C. albicans strain CEC359, yielding strains CEC366 and CEC382, respectively (Table 1).
Plasmids pSKM28 with TPK1 under the control of the PCK1 promoter (Bockmuhl et al., 2001 ) and pRC2312P-H with EFG1 under the control of the PCK1 promoter (Stoldt et al., 1997 ) were kindly provided by J. Ernst (Düsseldorf Universität, Germany), and they were used to transform C. albicans strain CEC359 yielding strains CEC399 and CEC396, respectively. SphI-digested plasmid p383c was used to disrupt TUP1 in the C. albicans yak1Δ/yak1Δ strain, CEC359, as described previously (Braun and Johnson, 1997 ). Candidate heterozygous TUP1/tup1Δ::(hisG-URA3-hisG) and TUP1/tup1Δ::hisG and homozygous tup1Δ::hisG/tup1Δ::(hisG-URA3-hisG) mutants were verified by PCR on whole yeast cells by using oligonucleotides hisG_up and Tup_V5 or hisG_down and Tup_V3 (Table 2) and by Southern analysis as described previously (Braun and Johnson, 1997 ).
Biofilms were produced in microfermentors as described previously (Garcia-Sanchez et al., 2004 ). Biofilms were formed on plastic slides (Thermanox; Nalge Nunc, Rochester, NY). After contact with the plastic slide, cells adherent to the surface were incubated at 37°C with a continuous flow of YNB medium containing 0.4% glucose, 0.1 g/l arginine, 0.1 g/l histidine, 0.2 g/l methionine, and 0.01 g/l uridine and air. Under these conditions, growth of the planktonic cells is minimized. After 41 h of growth, the biofilm in each microfermentor was resuspended in water, filtered, and dried 3 d at 70°C. Quantification of the biofilm dry mass was done using a precision scale, and the data presented are representative of three independent experiments that used two microfermentors per strain. Adherence of yeast cells to Thermanox was tested as previously described (Peltroche-Llacsahuanga et al., 2006 ).
Biofilms formed on Thermanox by strains SC5314 (YAK1/YAK1; Table 1) and CEC362 (yak1Δ/yak1Δ; Table 1) were analyzed with a JEOL JSM-6700F apparatus, which is an ultrahigh-resolution field emission scanning electron microscope (FESEM) equipped with a cold-field-emission gun and a strongly excited conical lens. The secondary-electron-image resolution was 1 nm at 15 kV and 2.2 nm at 1 kV. Biofilms were frozen using a Gatan Alto 2500 cryo-stage and cryo-preparation chamber (Gatan UK, Oxford, United Kingdom) dedicated for use in FESEM. First, samples were frozen in slush nitrogen and cryo-transferred under vacuum to the preparation chamber for ice sublimation and metal shadowing with argon. Material was then transferred (under secondary vacuum) from the Gatan preparation chamber to the scanning electron microscope (SEM) stage for sample observation. The working temperature with the SEM stage module was −110°C. SEM working conditions were as follows: accelerating voltage, 1 or 2 kV; probe current, 30 pA; semi-in-lens secondary electron image; and working distance 21 nm.
Total RNA was prepared from 70-ml cultures of strain SC5314 (Table 1, WT) and CEC362 (Table 1, yak1Δ/yak1Δ). Saturation cultures grown in YPD were diluted to 2 × 106 cells ml−1 and grown for 1 h in YPD at 30°C (yeast phase) or grown for 1 h in Lee's medium at 37°C (hyphal phase). Total RNA was isolated using RNAeasy according to the manufacturer's instructions (QIAGEN, Valencia, CA). Total RNA samples (5 μg) were indirectly labeled using Atlas PowerScript Fluorescent Labeling kit (Clontech, Mountain View, CA) with 1 μg of oligo(dT) (Invitrogen), according to the conditions recommended by the manufacturer. cDNAs were coupled with cyanines by using Cy3 Mono-Reactive Dye or Cy5 Mono-Reactive Dye (GE Healthcare, Chalfont St. Giles, United Kingdom). Fluorescent cDNAs were then purified with the Fluortrap matrix (Clontech). Equal quantities of Cy3- and Cy5-labeled cDNAs were mixed and concentrated by Microcon YM-30 (Millipore, Billerica, CA). Purified cDNAs were hybridized with microarrays containing ≈6000 C. albicans probes representing nearly the whole genome according to the conditions recommended by the manufacturer (Eurogentec, Seraing, Belgium). Arrays were scanned with an Axon 4000a scanner with fixed photomultiplier tube (550 for Cy3 and 650 for Cy5). Four comparisons were performed: hyphal phase wild-type cells versus yeast phase wild-type cells; hyphal phase mutant cells versus yeast phase mutant cells; hyphal phase wild-type cells versus hyphal phase mutant cells; and yeast phase wild-type cells versus yeast phase mutant cells. For each comparison, two biological replicates were used and each biological replicate was subjected to technical replicates with dye-swaps. Data were acquired and analyzed by Genepix Pro5.0 (Axon Instruments, Foster City, CA). Data normalization using Lowess and analysis were performed using GeneSpring (Silicon Genetics, Redwood City, CA). Raw data are available at http://www.galarfungail.org/data/Goyard/data.htm. Normalized data are available in Supplemental Table 1S. The p values for differentially expressed genes between the compared strains or conditions were adjusted with Benjamini–Hochberg's transformation. A list of genes that show statistically significant >3-fold differences in the hyphal phase wild-type cells versus hyphal phase mutant cells or yeast phase wild-type cells versus yeast phase mutant cells comparisons (p < 0.01) is available in Supplemental Table 2S. A list of genes that show statistically significant >3-fold differences in the hyphal phase wild-type cells versus yeast phase wild-type cells comparison (p < 0.01) along with corresponding ratio and p values for the three other comparisons is available in Supplemental Table 3S.
Total RNA was isolated from cultures of strain CEC362 (yak1Δ/yak1Δ), CEC366 (yak1Δ/yak1Δ/YAK1), and BWP17AHU (YAK1/YAK1) that had been grown to saturation in YPD, diluted to 2 × 106 cells ml−1, and grown for 0, 15, 30, 60, or 180 min in Lee's medium at 37°C. Total RNA was isolated using RNAeasy (QIAGEN) including a first step of DNase treatment according to the manufacturer's instructions. Purified RNA was DNase digested a second time using the DNA-free kit (Ambion, Austin, TX) to ensure the absence of genomic DNA contamination.
All RNA samples were verified as DNA free by using them directly as template in a PCR assay. Lack of a PCR product in the DNase-treated sample indicated that products amplified in the subsequent assay were not derived from trace amounts of genomic DNA. RNA concentration was measured spectrophotometrically and 500 ng from each sample was used for cDNA synthesis with the iScript kit according to the manufacturer's instructions (Bio-Rad, Hercules, CA). Quantitative (q)RT-PCR was carried out on an iCycler (Bio-Rad). Oligonucleotide primers were designed with OligoPerfect Designer (Invitrogen), and they were 22–25 bases (Table 2). Primer pairs had melting temperatures of 60°C, and they resulted in amplicons of 120–150 base pairs. All primer pairs produced products of the expected sizes. For qRT-PCR, ABsolute* QRT-PCR (ABgene, Epsom, Surrey, United Kingdom) was used according to the manufacturer's recommendations. Before transcript quantification, efficiency values were determined for all primer pairs, efficiencies of 90% or greater were observed. Each reaction mixture contained a 200 nM concentration of each primer and 10 μl of cDNA product diluted 1:50. Each sample was run in duplicates and between duplicates the variation of the two independent CT was <0.5. Reactions were run in sealed 96-well plates (ABgene) by using recommended cycle times and temperatures. Relative expression levels were calculated using the delta-delta Ct method, with C. albicans translation elongation factor EFT3 transcripts as standard. The relative expression was calculated as 2(Ct target yeast − Ct EFT3 yeast) − (Ct target hyphae − Ct EFT3 hyphae). All the data were normalized to the expression level of cells of strain CEC366 (yak1Δ/yak1Δ/YAK1) at time zero.
The virulence of C. albicans strains SC5314 (YAK1/YAK1), CEC362 (yak1Δ/yak1Δ), and CEC366 (yak1Δ/yak1Δ/YAK1) in mouse models of hematogenously disseminated and oropharyngeal candidiasis were determined as described previously (Cornet et al., 2005 ; Park et al., 2005 ). In the disseminated candidiasis models, the mice were infected via the tail vein. Two different inocula were tested (2 × 105 and 2 × 106 organisms per mouse), and the mice were monitored for survival for 14 d. In the oropharyngeal candidiasis model, the mice were immunosuppressed with cortisone acetate, and then they were inoculated orally with each of the strains. After 5 d of infection, the oral fungal burden of the mice was determined by quantitative culture. In all experiments, each strain of C. albicans was tested in seven mice.
A search of assembly 19 of the C. albicans genome sequence (Jones et al., 2004 ) was performed to identify ORFs encoding proteins homologous to S. cerevisiae and C. glabrata Yak1. Two contiguous open reading frames, namely, orf19.148 and orf19.147, were identified (Figure 1). In the course of assembly 19 annotation (Braun et al., 2005 ), these ORFs have been merged through an artificial frame-shift to generate a unique ORF, referred to as orf19.147 (Figure 1). The same frame-shift was observed for the two allelic ORFs, namely, orf19.7788 and orf19.7787 (Figure 1). The region encompassing this frame-shift was amplified from genomic DNA of C. albicans strain SC5314 and the amplification product directly sequenced. The resulting sequence (GenBank accession no. EU443248) differed from that in assembly 19 at six positions, yielding a unique 2427-nt ORF (Figure 1). This ORF was identical to that of YAK1 in C. albicans WO-1 (http://www.broad.mit.edu/annotation/genome/candida_albicans/Home.html), except for several microsatellite-like regions. When the deduced amino acid sequence was used to search the S. cerevisiae or C. glabrata proteome, it identified the two Yak1 proteins, suggesting that it is orthologuous to S. cerevisiae and C. glabrata Yak1.
Figure 1 shows a schematized organization of the C. albicans Yak1 protein, similar to that observed for S. cerevisiae and C. glabrata Yak1 and other Yak1 proteins identified in different yeast genomes (data not shown). Yak1 proteins share a poorly conserved low-complexity amino-terminal domain of ≈400 amino acids. In species of the CTG clade of Saccharomycotina (Fitzpatrick et al., 2006 ), this domain includes two well-conserved regions (Figure 1). The carboxy-terminal domain of Yak1 proteins harbors a conserved (>60% identity) catalytic domain typical of Ser/Thr protein kinases. This probable catalytic domain is separated by a low-complexity region from a second but partial Ser/Thr protein kinase catalytic domain. Four conserved putative PKA phosphorylation sites were identified in the C. albicans Yak1 protein, three sites (K136RLS, R209RMS, and R258RSS) in the amino-terminal low-complexity region, and one site (R351RCS) in a domain that is conserved among all yeast Yak1 proteins and located upstream of the catalytic domain (Figure 1). The corresponding four PKA phosphorylation sites in the S. cerevisiae protein have been shown to be phosphorylated (Zappacosta et al., 2002 ). The conserved tyrosine residue that is subject to autophosphorylation in other kinases of the DYRK family is located at position 588. Low-complexity regions in the C. albicans Yak1 proteins show intraspecific variations at the level of amino acid repeats (data not shown).
C. albicans is an obligate diploid and auxotrophies have been associated with altered virulence (Lay et al., 1998 ; Brand et al., 2004 ). To evaluate gene function in this species, it is necessary to perform two transformations to successively inactivate both alleles by gene replacement. It is also important to correct all auxotrophies. Furthermore, this organism has significant genomic instability, which can be induced by transformation (Selmecki et al., 2005 ). Therefore, it is necessary to reintegrate a functional wild-type allele of the gene of interest back into the homozygous null mutant to ensure that the wild-type phenotype is restored and that the observed phenotype is not the result of a secondary mutation that was inadvertently introduced during transformation (Noble and Johnson, 2007 ).
To investigate the function of C. albicans YAK1, strains deleted for one allele (CEC364, yak1Δ::ARG4/YAK1 and CEC365, yak1Δ::HIS1/YAK1; Table 1) or two alleles of YAK1 (CEC362, yak1Δ::HIS1/yak1Δ::ARG4 and CEC363, yak1Δ::ARG4/ yak1Δ::HIS1; Table 1) were constructed through targeted replacement of the YAK1 ORF by the HIS1 or ARG4 genes (Gola et al., 2003 ). Replacement of the YAK1 ORF and its absence in homozygous null mutants was confirmed by PCR and Southern blot analysis (data not shown). All strains were rendered prototroph through reconstitution of the HIS1 or ARG4 loci or integration at the RPS10 locus of the URA3 gene carried by the CIp10 plasmid (Murad et al., 2000 ). In addition, a derivative of strain CEC359 carrying a wild-type copy of YAK1 at the RPS10 locus was generated and is referred to as CEC366 (yak1Δ::HIS1/yak1Δ::ARG4 RPS10/RPS10::CIp10-URA3-YAK1; Table 1). No difference in growth rates were observed when the CEC362 (yak1Δ./yak1Δ) and CEC366 (yak1Δ/yak1Δ/YAK1) were grown in rich medium at 30°C or in minimal medium at 30°C or 37°C, indicating that YAK1 is not necessary for normal planktonic growth of C. albicans yeast cells (data not shown).
C. glabrata yak1 null mutants showed a fivefold reduction in their ability to form biofilms in a microtiter plate assay for biofilm formation (Iraqui et al., 2005 ). Whereas wild-type C. glabrata formed biofilms consisting of colonies of up to 100 cells adhering to plastic, biofilms of the C. glabrata yak1 null mutants consisted of single cells or colonies of <10 cells, with no alteration of initial adherence. Therefore, we tested the ability of the different C. albicans yak1Δ mutants to form biofilms. For this purpose, a continuous-flow microfermenter model (Garcia-Sanchez et al., 2004 ) was used and the biofilm biomass formed after 41 h of biofilm growth at 37°C was measured. Data presented in Figure 2A show that strain CEC362 that lacked the two YAK1 alleles was impaired for biofilm formation, with a biomass representing ≈20% of that of a biofilm formed by YAK1 wild-type C. albicans strains. Similar results were obtained with an independent yak1Δ/yak1Δ mutant strain (CEC363; data not shown). The heterozygous YAK1/yak1Δ strain CEC364 as well as the complemented yak1Δ/yak1Δ/YAK1 strain CEC366 exhibited a biofilm biomass close to that observed for the SC5314 or BWP17AHU control strains (Figure 2A), indicating that the biofilm formation defect of the CEC362 strain was a consequence of YAK1 inactivation.
This defect in biofilm formation might be due to a defect in adherence of the mutant yeast cells to the plastic surface used to grow C. albicans biofilms. Adherence to Thermanox plastic of the SC5314 wild-type and CEC362 yak1Δ/yak1Δ strains were compared. No significant difference could be observed indicating that the defect in biofilm formation was not due to a lack of initial adherence to the plastic surface (data not shown). Cryo-scanning electron microscopy was used to assess the architecture of the biofilm formed by the yak1Δ/yak1Δ null mutant. Although the biofilm formed by C. albicans wild-type strain SC5314 consisted of a very dense and organized structure of intertwined long hyphal elements that seemed embedded in a matrix, such a structure was not observed for the CEC362 yak1Δ/yak1Δ mutant strain and only microcolonies consisting mostly of yeast or pseudohyphal cells and apparently devoid of matrix material could be seen (Figure 2B).
C. albicans mutants with defects in filamentation are unable to form a dense and mature biofilm (Ramage et al., 2002 ; Richard et al., 2005 ). Because the thin biofilm formed by the CEC362 yak1Δ/yak1Δ mutant strain consisted mostly of pseudohyphal cells despite the presence of some hyphae (Figure 2B), we hypothesized that the defect in biofilm formation of the mutant strain could reflect a general defect in morphogenesis. The ability of the different C. albicans yak1 mutants to produce hyphae was first examined on solid agar media under Efg1-dependent—YPD FCS 20% and RPMI—and Efg1-independent—embedded YPS—hypha-inducing conditions. Results presented in Figure 3A show that the CEC362 yak1Δ/yak1Δ strain exhibited a strong filamentation defect in all three media. Whereas the wild-type strain formed wrinkled colonies (pseudohyphal and mycelial cells) on YPD FCS 20% and developed peripheral long hyphae, strain CEC362 formed smooth colonies and peripheral hyphae could not be detected. Reintegration of a wild-type copy of YAK1 resulted in a wrinkled colony but with fewer peripheral hyphae than wild type (Figure 3A). On RPMI medium, the yak1Δ/yak1Δ strain showed some filamentation, but uniform filamentation around the colony was never seen, and filaments seemed shorter than those observed for the wild-type or reconstituted strains (Figure 3A). Moreover, agar invasion was abolished on YPD FCS 20% and RPMI solid medium when the YAK1 gene was inactivated (data not shown). In embedded conditions, the wild-type strain formed homogeneous fluffy, furry colonies, whereas no filamentation was observed with strain CEC362 (Figure 3A). These results indicated that the C. albicans YAK1 gene was necessary for efficient filamentation on different solid media.
The ability of the yak1Δ/yak1Δ strain to form filaments was also assessed in liquid media under Efg1-dependent hypha-inducing conditions—Lee's medium and YPD FCS 10%. Results presented in Figure 3A show that the YAK1 gene was also necessary for efficient filamentation in Lee's medium, because strain CEC362 was unable to form true hyphae with nearly 90% cells forming pseudohyphae. In contrast, when yeast cells of the yak1Δ/yak1Δ mutant were exposed to 10% FCS, they originally underwent hyphal differentiation in a manner indistinguishable of the wild-type strain, but they eventually exhibited abnormal patterns of filamentation, including lateral and apical budding (Figure 3B). These results indicated that the Yak1 protein is dispensable for the initiation of hyphal morphogenesis in response to serum but necessary for hyphal maintenance in this condition.
Shokat and colleagues have shown that the replacement in the ATP binding pocket of protein kinases of a conserved bulky residue—the so-called gate-keeper residue—by a short chain amino acid such as glycine or alanine results in a kinase that is sensitive to inhibition by derivatives of the tyrosine kinase inhibitor PP1 (Bishop et al., 2000 ; Bishop et al., 2001 ). We used this chemical genetics strategy to generate a C. albicans strain with conditional Yak1 function. On the basis of an alignment of the Yak1 amino acid sequence with sequences in the kinase database available at http://sequoia.ucsf.edu/ksd, a mutation that increased the size of the ATP binding pocket of Yak1 was introduced by substituting a phenylalanine at position 507 with an alanine (Figure 1A).
A copy of the yak1F507A allele was introduced at the RPS10 locus in the CEC359 yak1Δ/yak1Δ strain. As shown in Figure 4A, the resulting CEC382 yak1Δ/yak1Δ/yak1F507A strain showed wild-type filamentation when incubated in Lee's medium, suggesting that the F507A mutation does not alter the activity of the Yak1 kinase in the absence of PP1 derivatives. In contrast, when 1.5 μM 1-NMPP1 was added to Lee's medium, hyphal differentiation of the yak1Δ/yak1Δ/yak1F507A strain was abolished, whereas this drug had no impact on the filamentation of a C. albicans wild-type strain (Figure 4A). Addition of 5 μM 1-NMPP1 to RPMI 1640 agar plates inhibited filamentation and agar invasion of the yak1Δ/yak1Δ/yak1F507A strain, but it had no effect on the filamentation and agar invasion of a yak1Δ/yak1Δ/YAK1 strain (data not shown).
We took further advantage of the conditional yak1 allele to probe the role of Yak1 during hyphal elongation. Addition of 1-NMPP1 at different time points after transfer of yeast cells of the CEC382 yak1Δ/yak1Δ/yak1F507A strain into Lee's medium resulted in an arrest of hyphal elongation and promoted lateral and apical bud formation from hyphae as illustrated in Figure 4B. Together, these results indicated that the Yak1 protein kinase activity is necessary for the initiation of hyphal formation as well as for hyphal elongation and maintenance.
The ability to make the transition from the yeast to hyphal forms has been considered a major requirement for C. albicans virulence (Saville et al., 2003 ). Our observation that the C. albicans YAK1 gene is required for hyphal emergence and/or elongation in vitro led us to examine whether it is also required for C. albicans virulence. The virulence of C. albicans strains SC5314 (YAK1/YAK1), CEC362 (yak1Δ/yak1Δ), and CEC366 (yak1Δ/yak1Δ RPS10/RPS10::YAK1) was examined in murine models of systemic candidiasis (Cornet et al., 2005 ) and oropharyngeal candidiasis (Park et al., 2005 ). Similar mortality rates were observed for the three strains upon systemic candidiasis (data not shown). In the murine model of oropharyngeal candidiasis no significant difference in oral fungal burden and histopathology of oral tissues were observed between the three strains (Figure 3C; data not shown). Together, these data indicated that YAK1 was dispensable for virulence during systemic and oropharyngeal candidiasis mediated by C. albicans.
For further insight into the function of Yak1 in hyphal morphogenesis, transcript profiling of C. albicans strains SC5314 (YAK1/YAK1) and CEC362 (yak1Δ/yak1Δ) cultivated 1 h in YPD at 30°C, i.e., in the yeast phase, or incubated 1 h in Lee's medium at 37°C, i.e., in the hyphal phase, was performed. Comparison of the transcript profiles obtained for each strain in the two culture conditions or for both strains cultivated in the same condition was achieved.
Data presented in Figure 5A and Supplemental Table 2S show that Yak1 functionality was associated with greater or equal to threefold up-regulation of 42 genes and lesser or equal to threefold down-regulation of 18 genes in the hyphal phase compared with wild-type cells. In the yeast phase, Yak1 functionality was associated with ≥3-fold up-regulation of nine genes and ≤3-fold down-regulation of six genes, the expression of most of which was similarly altered in the hyphal phase (Figure 5A).
That inactivation of YAK1 resulted in alterations of gene expression mostly in the hyphal phase suggested that Yak1 might be involved in regulating the differential expression of genes during the yeast-to-hypha transition. Therefore, we investigated whether the 42 genes that show Yak1-dependent increased expression in the hyphal phase were also up-regulated upon the yeast-to-hypha transition in wild-type C. albicans. Data presented in Supplementary Table 3S show that 178 genes were up-regulated by more than threefold when C. albicans strain SC5314 was induced to undergo the yeast-to-hypha transition in Lee's medium. Genes that have been previously described as being up-regulated upon hyphal growth were found among these 178 genes, including ALS3, HWP1, HYR1, ECE1, IHD1, HGC1, RBT1, RBT5, RBT4, SOD5, SNZ1, PHR1, SEC24, CHA1, and RDI1 (Birse et al., 1993 ; Hoyer et al., 1998 ; Staab and Sundstrom, 1998 ; Sharkey et al., 1999 ; Braun et al., 2000 ; Murad et al., 2001a ; Nantel et al., 2002 ; Martchenko et al., 2004 ; Zheng et al., 2004 ). As shown in Figure 5B, of the 42 genes that show Yak1-dependent increased expression in the hyphal phase, 32 were also found among the 178 yeast-to-hypha up-regulated genes. A list of these 32 genes is shown in Table 3, and it includes the hypha-specific genes ALS3, HYR1, HWP1, ECE1, IHD1, HGC1, and RBT1. The remaining 146 yeast-to-hypha up-regulated genes showed less than threefold differences between wild-type and mutant hyphae (see Table 4 for examples and Supplemental Table 3S). Consistently, most of these 146 genes were up-regulated when the C. albicans strain CEC362 (yak1Δ/yak1Δ) was induced to undergo the yeast-to-hypha transition in Lee's medium (see Table 4 for examples and Supplemental Table 3S). Together, these data indicated that although up-regulation of the group of 32 genes in response to hypha-inducing conditions is dependent on the function of Yak1, this is not the case for the group of 146 genes.
Further examination of the group of 32 genes showed that although several genes did not show increased expression when the yak1Δ/yak1Δ strain was induced to undergo the yeast-to-hypha transition (see for examples HYR1, HGC1, CFL11, and CAS4; Table 3), others had expression levels significantly higher in hyphal than yeast cells of the mutant strain (see for examples ECE1, ALS3, HWP1, and HGT1; Table 3). For these latter genes, we observed that 1) the activation ratio was always lower in the mutant strain than in the wild-type strain and 2) the level of expression in wild-type yeast cells was higher than that in yeast cells of the yak1Δ/yak1Δ mutant strain (Table 3). Consequently, despite up-regulation of these genes in response to hypha-inducing conditions, their expression was always significantly diminished in hypha of the mutant strain relative to the wild type. These data suggested that Yak1 contributes to the transcriptional regulation of a subset of hypha-induced genes in the yeast and hyphal forms.
C. albicans genes that are differentially regulated at the yeast-to-hypha transition are the subject of complex positive and negative regulation by an array of transcription factors, including Efg1, Tec1, Bcr1, Rim101, Tup1, Nrg1, and Rfg1 (Eckert et al., 2007 ; Whiteway and Bachewich, 2007 ). Interestingly, we noted a significant enrichment for Tup1-regulated genes in the group of 32 Yak1-dependent, hypha up-regulated genes in contrast to the group of 146 Yak1-independent, hypha up-regulated genes. Indeed, 62.5% of the genes in the former group have been reported to be under Tup1-negative regulation, whereas this is the case for 24.7% of the genes in the latter group. Tup1-negatively regulated genes account for 6.5% of the whole C. albicans gene set (Garcia-Sanchez et al., 2005 ; Kadosh and Johnson, 2005 ).
To confirm the role of Yak1 in the hypha-induced transcriptional activation of hypha-specific genes, real-time PCR was used to examine the expression of HWP1, HYR1, ALS3, and RBT5 during a time course experiment for hyphal differentiation. These genes were chosen because three (HWP1, HYR1, and ALS3) showed Yak1-dependent hypha up-regulation in our transcript profiling experiment, whereas the latter (RBT5) showed Yak1-independent hypha up-regulation (Tables 3 and and4).4). Data presented in Figure 6A show that mRNA levels of the four genes increased during hyphal differentiation of the CEC366 yak1Δ/yak1Δ/YAK1 strain. The same observation was made when mRNA levels of RBT5 were monitored during hyphal differentiation of the CEC362 yak1Δ/yak1Δ strain, indicating that RBT5 up-regulation is independent of the function of Yak1. In contrast, mRNA levels of HYR1, HWP1, and ALS3 did not increase in the CEC362 strain to the same extent as that seen in the CEC366 strain (Figure 6), consistent with the microarray data presented above and the Yak1-dependent hypha up-regulation of HWP1, HYR1, and ALS3.
The cAMP–PKA pathway plays a key role in the control of morphogenesis through regulation of the activity of the transcription factor Efg1 and subsequent activation of genes encoding transcriptional factors and hypha-specific genes (Eckert et al., 2007 ; Whiteway and Bachewich, 2007 ). Results presented above showed that efficient hyphal morphogenesis and hypha-induced expression of several hypha-specific genes, such as HWP1, HYR1, and ALS3, are dependent upon a functional YAK1 gene. Thus, we investigated whether Yak1 might regulate the cAMP–PKA–Efg1 regulatory cascade.
Addition of cAMP or overexpression of the PKA catalytic subunit Tpk1 can restore filamentation to C. albicans strains that have defects in components that act upstream of adenylyl cyclase (Maidan et al., 2005 ). To test whether the Yak1 kinase is involved in activating hyphal morphogenesis upstream of adenylyl cyclase, the effect of cAMP on the filamentation and agar invasion defects of the yak1Δ/yak1Δ CEC362 strain was tested. Addition of 5 mM cAMP to solid YPD FCS 20% medium or liquid Lee's medium did not restore filamentation and agar invasion of strain CEC362 (data not shown). Moreover, transformation of a C. albicans yak1Δ/yak1Δ strain with plasmids allowing overexpression of TPK1 (Bockmuhl et al., 2001 ) or EFG1 (Stoldt et al., 1997 ) did not alleviate the filamentation and agar invasion defects associated to the deletion of YAK1 (data not shown). Thus, artificial activation of the cAMP–PKA–Efg1 pathway is not sufficient to suppress the filamentation defect of a yak1Δ mutant.
The TEC1 and BCR1 genes are targets of the cAMP–PKA–Efg1 pathway, and they show up-regulation upon induction of hyphal growth. Efg1 is responsible for the activation of TEC1, whereas Tec1 activates BCR1 whose product in turn activates HWP1, HYR1, ALS3, and RBT5 (Nobile and Mitchell, 2005 ; Cao et al., 2006 ). Interestingly, TEC1 was found within the group of 146 genes whose transcriptional activation in response to a hypha-inducing signal did not require a functional YAK1 gene (Table 4). BCR1 was not found among the genes that showed more than threefold induction upon hyphal morphogenesis in the wild-type strain. However, as shown in Table 4, BCR1 was up-regulated 2.8-fold (p = 2.5 × 10−4) in the wild-type strain and 1.9-fold (p = 0.04) in the mutant strain undergoing hyphal differentiation, suggesting that transcriptional activation of BCR1 in response to a hypha-inducing signal did not require a functional YAK1. To confirm this observation, the expression of TEC1 and BCR1 was monitored using real-time PCR. Quantification of TEC1 and BCR1 mRNA levels was associated with important variability between biological replicates, possibly due to the low level of expression of these genes and variations in the kinetics of germ tube emergence between biological replicates. However, we consistently observed up-regulation of both genes upon hyphal induction in the CEC366 (yak1Δ/yak1Δ/YAK1) and CEC362 (yak1Δ/yak1Δ) strains. These results suggested that the defect in the transcriptional activation of hypha-specific genes such as HWP1, HYR1, and ALS3 observed in the yak1Δ/yak1Δ mutant is probably not due to defects in the up-regulation of TEC1 and BCR1, and more generally in the activation of the cAMP–PKA–Efg1 pathway.
C. albicans yeast-to-hypha transition is the subject of negative regulation by the general repressor Tup1 in association with the DNA-binding proteins Nrg1 and Rfg1 (Braun and Johnson, 1997 ; Braun et al., 2001 ; Kadosh and Johnson, 2001 ; Murad et al., 2001b ). Because 63% of the Yak1-dependent, hypha-specific genes have been reported to be negatively regulated by Tup1 (see above), we hypothesized that Yak1 may function in the Tup1 pathway to govern hyphal emergence and maintenance. Inactivation of Tup1 is associated with constitutive filament formation in media that are not normally conducive to filamentation (Braun and Johnson, 1997 ). Thus, a C. albicans yak1Δ/yak1Δ tup1Δ/tup1Δ strain was generated through successive disruption of the two TUP1 alleles in the yak1Δ/yak1Δ CEC359 strain. As shown in Figure 7, the resulting yak1Δ/yak1Δ tup1Δ/tup1Δ CEC991 strain showed constitutive filamentation when grown under noninducing conditions in a manner similar to the YAK1/YAK1 tup1Δ/tup1Δ BCA2-10 strain (Braun and Johnson, 1997 ). Thus, deletion of the TUP1 gene was able to bypass the requirement for Yak1 in C. albicans filamentation.
Results presented in this study have identified the Yak1 protein kinase as a novel regulator of hyphal growth in C. albicans. Indeed, inactivation of Yak1 through gene deletion or chemical inhibition is associated with defects in 1) the emergence of germ tubes, 2) the maintenance of hyphal growth, and 3) the expression of a number of genes that are normally up-regulated upon hyphal morphogenesis.
We originally investigated the function of C. albicans Yak1 because of the role of the orthologous Yak1 kinase in the formation of biofilms by C. glabrata (Iraqui et al., 2005 ). Our results showed that the C. albicans yak1Δ/yak1Δ mutants were severely impaired in their ability to form biofilms. This is likely to reflect the contribution of C. albicans Yak1 to filamentation, this process being central to the establishment of mature biofilms by C. albicans (Ramage et al., 2002 ; Richard et al., 2005 ), and to the expression of the Als3 and Hwp1 adhesins that have been proposed to mediate cell-to-cell interactions within biofilms (Nobile and Mitchell, 2005 ; Nobile et al., 2006a ,b ). Thus, both C. glabrata Yak1 and C. albicans Yak1 play important roles in regulating the expression of genes encoding adhesins mediating cell-to-cell interactions and in biofilm formation. C. glabrata Yak1 is thought to regulate the expression of the EPA genes through negative regulation of the subtelomeric silencing machinery (Iraqui et al., 2005 ), a process that is unlikely to account for the role of C. albicans Yak1 in the regulation of the ALS3 and HWP1 genes because these genes are not located at subtelomeric regions (van het Hoog et al., 2007 ). Yet, a role of C. albicans Yak1 in controlling gene expression through regulation of chromatin remodeling cannot be excluded because S. cerevisiae Yak1 was recently shown to influence the nucleocytoplasmic distribution of the Msi1 (Cac3) protein and its association in the CAF-I (Pratt et al., 2007 ). Whether this is also the case for C. albicans Yak1 and influences the observed binding of chromatin remodeling proteins to the promoter regions of Yak1-regulated genes will remain to be investigated. In this regard, the HWP1 control region critical for hypha-induced activation has recently been shown to be bound by proteins involved in chromatin remodeling (Kim et al., 2007 ) and the Swi/Snf chromatin remodeling complex is essential for hyphal development and expression of hypha-specific genes (Mao et al., 2006 ).
Our data showed that Yak1 is required at the onset of filamentation under some hypha-inducing conditions. Strikingly, however, the yak1Δ/yak1Δ mutants did not show defects in germ tube emergence when serum was used to induce the yeast-to-hypha switch. There, germ tubes formed efficiently but, after a certain time, hypha started to produce lateral and apical buds (Figure 3B), suggesting that Yak1 is also necessary for the maintenance of hyphal growth. This conclusion was reinforced through analysis of a C. albicans mutant strain expressing a 1-NMPP1-sensitive Yak1 kinase. Addition of 1-NMPP1 to hypha of this strain promoted the production of lateral branches with pseudohyphal morphology (Figure 4B). Thus, there is a requirement of the Yak1 kinase activity for both the initiation and the maintenance of hyphal growth. Although much is known about the molecular requirements for the initiation of hyphal morphogenesis (Eckert et al., 2007 ; Whiteway and Bachewich, 2007 ), the mechanisms that underlie the maintenance of hyphal growth and return to the yeast growth phase remain poorly understood. This in part due to the fact that most C. albicans morphogenetic mutants are defective at early stages of hyphal morphogenesis and that promoter-replacement methods that could be used for conditional expression of these genes are not optimal to survey their contribution to later stages of morphogenesis. In this regard, our chemical genetic approach to the study of Yak1 provides for rapid inactivation of the kinase activity without changes in the growth medium and should be applicable to other kinases that have been implicated in hyphal morphogenesis such as Tpk1, Tpk2 and components of the MAP kinase pathways (Bockmuhl et al., 2001 ; Alonso-Monge et al., 2006 ). However, examples of C. albicans mutants that show defects in the maintenance of hyphal elongation and production of lateral buds have been reported. For example, overexpression of the transcriptional regulator Efg1 in hypha triggers the production of lateral yeast cells, suggesting that down-regulation of EFG1 is important for the maintenance of hyphal growth (Tebarth et al., 2003 ). Inactivation of the G1 cyclin Cln1 is also associated with a defect in hyphal maintenance (Loeb et al., 1999 ), suggesting that proper control of the cell cycle is an important aspect of hyphal maintenance as it is of other aspects of morphogenesis (Berman, 2006 ). In S. cerevisiae, Yak1 interacts with the Hrt1 component of the Skp1–Cdc53–F-box complex (Uetz et al., 2000 ) that is involved in ubiquitination of G1 cyclins and their subsequent degradation. It has been proposed that the contribution to pseudohyphal growth of S. cerevisiae Yak1 might result from its inhibition of G1 cyclin degradation (Zhang et al., 2001 ). A similar mechanism might explain the Yak1-dependent regulation of hyphal maintenance in C. albicans.
That yak1Δ/yak1Δ mutants were not defective for germ tube emergence in the presence of serum, whereas they were in other hypha-inducing conditions is not entirely unexpected. Other C. albicans mutants with defects in filamentation and agar invasion under a variety of inducing conditions are still able to form germ tubes when exposed to serum (Loeb et al., 1999 ; Zakikhany et al., 2007 ). This is possibly because serum triggers filamentation through multiple signaling pathways, whereas other inducers might be more specific (Harcus et al., 2004 ). The ability to switch between yeast and hyphal forms is a central component of C. albicans virulence as indicated by the reduced virulence of numerous C. albicans mutants that are defective for hyphal morphogenesis or that can be locked into one of the two morphotypes (Lo et al., 1997 ; Gow et al., 2002 ; Saville et al., 2003 ). Yet, the yak1Δ/yak1Δ mutants did not show any virulence alteration in animal models of systemic and oropharyngeal candidiasis. This suggests that the conditions that are encountered by C. albicans in these animal models are—such as serum—sufficient to trigger hyphal morphogenesis in a Yak1-independent manner. Further study of the hyphal differentiation program of yak1Δ/yak1Δ mutants exposed to serum or other conditions mimicking the animal environment will be needed to identify the signaling pathways that allow to bypass the Yak1 requirement for filamentation.
The activation of adenylyl cyclase, PKA and the transcriptional regulator Efg1 is at the center of the yeast-to-hypha transition, mediating the response to several environmental stimuli (Lo et al., 1997 ; Stoldt et al., 1997 ; Eckert et al., 2007 ), and we have investigated whether a role of Yak1 in regulating filamentation could be through the cAMP/PKA/Efg1 pathway. Several observations suggest that it is not the case: 1) Yak1 is required for hyphal differentiation in conditions that require Efg1 but also in microaerophilic conditions where Efg1 is dispensable; 2) addition of cAMP or overexpression of TPK1, encoding a catalytic subunit of PKA, or EFG1 does not suppress filamentation defects associated to a yak1Δ/yak1Δ mutation suggesting that Yak1 does not act upstream of the cAMP/PKA/Efg1 pathway; and 3) transcriptional targets of Efg1 such as TEC1 (Schweizer et al., 2000 ) are efficiently activated in response to hyphal induction in cells that lack Yak1. Moreover, we have observed that up-regulation of the BCR1 gene is not altered in a yak1Δ/yak1Δ mutant exposed to hypha-inducing conditions. Similarly, RBT5, a target of the Efg1–Tec1–Bcr1 transcriptional activation cascade (Nobile and Mitchell, 2005 ), showed Yak1-independent hypha up-regulation (Figure 6). Strikingly however, hypha-induced up-regulation of Bcr1 transcriptional targets such as HWP1, HYR1, and ALS3 (Nobile and Mitchell, 2005 ) was altered when Yak1 is lacking. Together, these data suggest that the filamentation defect associated to the absence of Yak1 is due to defective up-regulation of hypha-induced genes, several of which are necessary for efficient hyphal morphogenesis, independently of the activation of the cAMP/PKA/Efg1 pathway and the Tec1 and Bcr1 transcriptional regulators located downstream of this pathway. Although one cannot exclude that Yak1 may positively regulate morphogenesis through phosphorylation of transcriptional activators such as Tec1 or Bcr1, this seems unlikely because RBT5, a downstream target of these activators, is still strongly up-regulated in cells that lack Yak1, and C. albicans bcr1Δ mutants are not defective for filamentation (Nobile and Mitchell, 2005 ).
Hypha-induced genes are also negatively regulated by Tup1 in conjunction with the DNA-binding proteins Nrg1 and Rfg1 (Braun and Johnson, 1997 ; Braun et al., 2001 ; Murad et al., 2001b ; Kadosh and Johnson, 2005 ), and there is evidence that chromatin remodeling contributes to their expression (Kim et al., 2007 ; Mao et al., 2006 ). Therefore, it is tempting to speculate that Yak1 may modify the activity of the Tup1-Nrg1, Tup1-Rfg1, or chromatin remodeling complexes in response to hypha-inducing signals and consequently allow derepression or activation to take place. A large fraction of the genes that are hypha-induced in a Yak1-dependent manner have been shown to be repressed by Tup1 (Kadosh and Johnson, 2001 ; Garcia-Sanchez et al., 2005 ), suggesting that Yak1 may interfere with Tup1-mediated repression. This hypothesis is supported by our observation that inactivation of the TUP1 gene can bypass the Yak1 requirement for filamentation (Figure 7), suggesting that Yak1 might function as a repressor of Tup1. However, Tup1-repression is also observed for a number of genes that are not dependent upon Yak1 for hyphal-dependent induction such as RBT5, and we have not identified features that could distinguish these two groups of genes. Therefore, constitutive filamentation of the tup1Δ yak1Δ mutant could also indicate that the derepression of Tup1-repressed genes allows progression of the morphogenetic program independently of any positive regulation that may involve Yak1.
In S. cerevisiae, Yak1 has been proposed to act downstream of PKA, regulating different processes related to growth. PKA consensus phosphorylation sites in the S. cerevisiae Yak1 protein have been shown to be genuinely phosphorylated (Zappacosta et al., 2002 ). These phosphorylation sites are conserved in C. albicans Yak1, and it is conceivable that PKA-mediated phosphorylation of Yak1 at the onset of filamentation may contribute to its function in filamentation, linking the positive and negative regulatory branches of hypha-induced gene regulation. Yet, there are unexplored PKA-independent regulatory aspects to the function of Yak1 in S. cerevisiae (Garrett et al., 1991 ; Ward and Garrett, 1994 ; Zappacosta et al., 2002 ). PKA-independent regulation of Yak1 in C. albicans may in particular explain its contribution to filamentation under microaerophilic conditions whose relation to the cAMP/PKA pathway is unclear.
As discussed above, C. albicans Yak1 might contribute to the regulation of hyphal emergence and maintenance through different mechanisms that are not mutually exclusive: 1) inhibition of the repression exerted by Tup1, Nrg1, and Rfg1 on hypha-inducible genes; 2) regulation of chromatin remodeling needed for the expression of hypha-induced genes; and 3) regulation of ubiquitination and in turn G1 cyclin activity. The identification of the C. albicans proteins that are phosphorylated by Yak1 is likely to shed light on the actual mechanisms that are involved in regulating filamentation in a Yak1-dependent manner. In this regard, the availability of a 1-NMPP1–sensitive form of Yak1 should facilitate the identification of such targets. Indeed, protein kinases whose “gate-keeper” residue has been modified also use radiolabeled analogues of ATP as phosphate donors in in vitro phosphorylation reactions, and this property has been used advantageously to identify their cellular targets (Shah et al., 1997 ; Bishop et al., 2001 ; Habelhah et al., 2001 ; Dephoure et al., 2005 ).
We thank J. Wendland, A. Johnson, A. Brown, and J. Ernst for plasmids. We are grateful to members of our laboratories and to A. Brown for stimulating discussions and comments on the manuscript. P.K. was supported by Swiss National Science Foundation grant 81BS-69445, the Novartis Stiftung, the Freiwillige Akademische Gesellschaft, the European Commission QLK2-2000-00795, and EGIDE. This work was supported by European Commission grant MRTN-CT-2003-504148 and Institut Pasteur Programme Transversal de Recherche 173. S.G.F. was supported by grant 5R0 1DE017088 from the National Institutes of Health.
This article was published online ahead of print in MBC in Press (http://www.molbiolcell.org/cgi/doi/10.1091/mbc.E07-09-0960) on March 5, 2008.