INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Candida albicans is the fungus most frequently identified from clinical isolates. It can cause a variety of opportunistic infections, including deadly systemic candidiasis in immunocompromised patients (for a review, see reference 52). C. albicans is capable of dramatic morphological switching between budding yeast growth and filamentous hyphal growth. Both growth forms coexist in infected tissues. Because mutant strains defective in morphological switching are much less virulent than wild-type strains (14, 22, 35, 39), competence to perform the switch has been linked with pathogenicity in C. albicans. Hyphal cells have been suggested to aid in adhesion and penetration of epithelial or endothelial cell layers to facilitate the infection (24).

C. albicans cells are able to respond to and integrate a large variety of environmental signals during their morphological development. Serum, nitrogen starvation, high temperature, and neutral pH, for example, promote hyphal development (51). Hyphal development is also accompanied by transcriptional induction of many hypha-specific genes, such as ECE1, HWP1, and HYR1 (5, 7, 59). A conserved mitogen-activated protein (MAP) kinase pathway has been shown to regulate hyphal development; mutations in Cst20 (PAK), Hst7 (MEK), Cek1 (MAP kinase), and Cph1 (a transcription factor) partially block hyphal colony formation on certain hypha-inducing media (15, 31, 34, 36). The Cek1 MAP kinase pathway functions in parallel with Efg1, a member of a family of basic-helix-loop-helix proteins important for developmental processes in several fungi (39, 60). Efg1 may function downstream of Tpk2, the catalytic subunit of protein kinase A (PKA), in hyphal development (57). Furthermore, a C. albicans Ras protein has been shown to be required for serum-induced hyphal differentiation (20). The complicated nature of dimorphic regulation is underscored by the discovery of more signaling pathways necessary for proper hyphal development. A two-component histidine kinase, Cos1/Nik1, and a Hog1 MAP kinase are involved in hyphal morphogenesis (2, 15, 48, 49, 58). More recently, we have found a G₁ cyclin-dependent kinase (Cdk) to be important for hyphal development under specific hypha-inducing conditions and for transcription of hypha-specific genes (42). Negative regulators of hyphal development have also been identified. The deletion of TUP1, which encodes a global transcriptional corepressor, causes hyperfilamentation under yeast growth conditions (9). Considering that C. albicans cells can respond to a large number of extracellular signals and growth conditions in monitoring hyphal development, they are likely to utilize many parallel signal transduction pathways to integrate these signals.

Many of the regulatory components for dimorphic switching are conserved in filamentous fungi despite their enormous diversity in size and shape and their genetic distance. For example, elements of the same conserved MAP kinase pathway involved in hyphal development in C. albicans are also required for filamentous growth in other fungi (45). In Saccharomyces cerevisiae, the switch from a unicellular yeast growth to a pseudohyphal growth upon nitrogen starvation depends on this MAP kinase pathway. Four protein kinases, Ste20, Ste11, Ste7, and Kss1, function in sequence to activate the transcriptional factors Ste12 and Tec1 (12, 37, 44, 46). Similarly, the same MAP kinase pathway is necessary for filamentous growth and virulence in the corn smut Ustilago maydis (6) and for appressorium formation and virulence in the rice fungus Magnaporthe grisea (65). Cyclic AMP (cAMP)/PKA is another conserved signal transduction pathway important for filamentous growth in several fungi (45). In S. cerevisiae, changing the level of intracellular cAMP either by an activated allele of RAS2 GTPase or by the G protein  subunit homologue Gpa2 affects the amount of filamentation (23, 32, 43). The cAMP-mediated signal transduction in filamentous growth is independent of the Kss1 MAP kinase pathway; instead it requires Flo8, a transcriptional regulatory necessary for pseudohyphal growth (38, 53, 55). The cAMP/PKA pathway is also important for filamentous growth, virulence, and mating in the human pathogen Cryptococcus neoformans (3). In addition to the MAP kinase and cAMP/PKA pathways, the Cdk Cdc28 has been shown to regulate filamentous growth in S. cerevisiae (1, 19, 41). Depending on its associated cyclins, Cdc28 plays different roles in filamentous growth (1, 19, 41).

Here we report the identification of another protein kinase in the same Cdc2 subfamily as MAP kinases and Cdks. We have designated it Crk1, for Cdc2-related kinase. Disruption of both copies of CRK1 in C. albicans leads to defective hyphal formation under all conditions examined. Furthermore, crk1/crk1 mutants fail to induce hypha-specific genes and are avirulent in mice. Consistent with mutant phenotypes, the ectopic expression of a CRK1 catalytic domain promotes the formation of hyphal colonies under conditions suited for yeast growth. Expression of the Crk1 catalytic domain in S. cerevisiae and C. albicans mutants defective in components of known filamentation signaling pathways suggested that Crk1 can activate filamentous growth via a route independent of the filamentation MAP kinase pathway and that of Phd1/Efg1. The relationship between Ras1 and Crk1 in hyphal development is also discussed.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Strains and culture conditions. The C. albicans and S. cerevisiae strains used in this study are listed in Tables 1 and Table 2, respectively. Yeast strains were routinely grown on YPD medium or on SD medium for selection of prototropic strains (56). Synthetic low-ammonia medium (SLAD) was used for observing pseudohyphal colony formation of S. cerevisiae (23). Invasive growth of S. cerevisiae was examined as described by Roberts and Fink (54) except that uracil-deficient synthetic complete medium (SCUra) was used instead of YPD medium. Transformation of S. cerevisiae was performed as described by Ito et al. (29). C. albicans strains were cultured as described previously (42). Ura C. albicans strains were selected on 5-fluoro-orotic acid (FOA)-containing medium (8). The protoplasting method of Kurtz et al. (33) was used for C. albicans transformation. Cell and colony morphologies were photographed as described by Loeb et al. (42).

Cloning and sequencing of CRK1. Two oligonucleotides, 5'AAAATTTGTGAC(or T)TTTGGTTTA and 5'TCTTGCTAAACCA, were synthesized according to a nucleotide sequence that encodes KICDFGLAR, a conserved region in the subdomain VII of Cdc2-related protein kinases. The two oligonucleotides were annealed to each other, and the two ends were labeled by blunt-end filling in with Klenow enzyme in the presence of [-³²P]dATP (Amersham). The oligonucleotide was used as a hybridization probe to screen a C. albicans genomic library inserted in GEM12 phage (Promega) (10). The hybridization was performed at 40°C in 6× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate)-1× Denhart solution-100 µg of yeast tRNA/ml-0.05% sodium pyrophosphate. The membrane was washed with 6× SSC at room temperature. Sixty-five positive  plaques were isolated. The recombinant  DNA was digested with restriction enzymes and analyzed by Southern hybridization using the same oligonucleotide probe. The 65  phage clones were classified into 13 groups. Two groups had restriction patterns of the known MAP kinase genes CEK1 and MKC1 (50, 63). The inserts from the other putative  phage clones were released by digestion with BamHI and cloned into the BamHI site of pBSK (Stratagene). Two clones, pBSZS1 and pBSZS2, had much stronger hybridization signals than the others and were analyzed in detail. pBSZS2 was found to contain a gene for a new MAP kinase (J. Chen et al., unpublished data). A 9-kb BamHI fragment in pBSZS1 was digested with EcoRI, PstI, and Bal31 and then subcloned into pBSK for sequence analysis. Nucleotide sequences of the DNA fragment that hybridized to the probe were determined by the dideoxy-chain termination method using Sequenase (U.S. Biochemical) and [-³⁵S]dATP (Amersham). Protein sequence comparisons were conducted by using the BLAST algorithm of Altschul et al. (4). Plasmid pBSZS1 contained a DNA sequence with a 2,241-bp open reading frame, corresponding to a protein of 746 amino acids, designated Crk1.

Plasmid and C. albicans strain construction. A 4-kb SacI fragment containing the entire coding region of CRK1 was subcloned from pBSZS1 into the SacI site of a pUC19 vector, generating plasmid pUC19CRK1 (Table 3). The internal 2-kb EcoRV-XhoI fragment in plasmid pUC19CRK1 was replaced with a 4.8-kb SalI-ScaI hisG-URA3-hisG fragment from plasmid pCUB6 (21) (see Fig. 3A), generating plasmid pUC19CRK1URA3. SacI-digested pUC19CRK1URA3 DNA (see Fig. 3A) was used to transform Candida ura3/ura3 strain CAI4 (21) to produce CRK1/crk1 and crk1/crk1 strains (Table 1).

Southern and Northern analyses. Methods for DNA isolation and Southern blot hybridization were as previously described (11). Total RNA extraction and Northern blot hybridization were performed as described in Current Protocols in Molecular Biology (26). DNA probes were labeled with the Bethesda Research Laboratories random-primer labeling kit and [-³²P]dATP (Amersham); 4-kb SacI and 2-kb EcoRV-XhoI CRK1 fragments from pUC19CRK1, a 1.4-kb XbaI-ScaI URA3 fragment from pUR3 (30), and a ClaI-SalI ACT1 fragment from plasmid p1595/3 (18) were used as probes. C. albicans ECE1 and HWP1 PCR products were used for probing Northern blots. The primers used were 5'GCCATCCACCATGCTCC and 5'GTGCTACTGAGCCGGCATCTC for ECE1 and 5'TGCTCCAGGTACTGAATCCGC and 5'GGCAGATGGTTGCATGAGTGG for HWP1. The 2-kb CRK1 fragment (see Fig. 3) was used as a probe in Northern hybridization. The sizes of mRNAs on Northern blots correlated with the expected lengths based on information from the C. albicans genome database.

Kinase assays. For kinase assays, plasmids pVTUHACRK1 and pVTUHACRK1N were introduced into YPH499. Extract preparation, immunoprecipitation, and kinase assays of immune complexes were performed as described elsewhere (61, 64). For each immunoprecipitation, 2.5 mg of protein total extract was used, with 1 µg of myelin basic protein (MBP; Sigma) or histone H1 protein (Sigma) as a substrate.

Bioassay for response to -factor. A halo bioassay was performed as described by Irie et al. (28). In short, 0.1 ml of overnight culture (10⁷ cells) was mixed with 5 ml of 0.7% soft agar and spread onto a YPD plate. Whatman paper disks (6 mm in diameter) were placed on the nascent lawn. Different quantities (0, 0.5, and 5 µg) of synthetic -factor (Sigma) were dotted onto each disk in 5-µl aliquots. Photographs were taken after 48 h.

Virulence studies. The virulence of C. albicans strains was tested as described by De Bernardis et al. (16). C. albicans strains were grown on SDUra plates for 48 h at 30°C. The cells were suspended in physiological saline solution and counted in a hemacytometer. Following quantitation, cells were adjusted to densities of 5 × 10⁷ and 5 × 10⁶ cells/ml. Each C. albicans strain was tested for virulence by injecting 0.1 ml of cells (5 × 10⁶ and 5 × 10⁵ cells) into the lateral tail veins of ICR male mice (18 to 21 g each; Shanghai Laboratory Animal Center, Chinese Academy of Sciences, Shanghai, China). Eight mice were injected for each strain. Surviving mice were observed daily after infection with C. albicans.

Nucleotide sequence accession number. The GenBank accession number for the CRK1 nucleotide sequence is U92261.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Cloning of protein kinase gene CRK1. We used an oligonucleotide probe-based screen to clone putative protein kinases in the Cdc28/Cdc2 subfamily from a C. albicans genomic library. The oligonucleotide sequence was designed from a region in the kinase subdomain VII, which is conserved among all MAP kinases and Cdks (Fig. 1B; Materials and Methods). Four putative kinase genes were cloned: one for a new MAP kinase (Chen et al., unpublished); two known MAP kinase genes, CEK1 and MKC1 (50, 63); and a novel gene that encodes a 746-amino-acid predicted protein (Fig. 1). The amino-terminal half of the coding sequence contains all 11 kinase catalytic domains that are highly conserved among members of the Cdc2 subfamily (Fig. 1B) (27). It shares the highest similarity with S. cerevisiae Sgv1 (28) (47% identical and 63% similar). A Schizosaccharomyces pombe Cdc2-like gene ranked second in our BLAST search, with 46% identity and 61% similarity. Two human kinases, a sequence of cDNA isolated from brain tissue and a PITALRE kinase (Pk11/Cdk9) (17, 25), also gave comparable scores in the search. In addition, the C. albicans kinases have an insertion common to all members of the Cdc2 branch of the kinase family (27) (Fig. 1B, underlined region between X and XI). Thus, we designated the C. albicans protein Crk1, for Cdc2-related kinase. However, Crk1 lacks some of the conserved residues that are known to be important for Cdk functions (47), including the highly conserved regulatory residue Tyr15 (Fig. 1B, Tyr19 for Cdc28), whose phosphorylation state modulates the kinase activity, and the conserved PSTAIRE sequence (Fig. 1B, domain III) important for interacting with the cell cycle-regulated cyclins. It also lacks the conserved MEK phosphorylation site TXY at the L12 region between subdomains VII and VIII of all MAP kinases (66). Very few similarities outside the kinase domains exist between Crk1 and the other four kinases, except for a short sequence immediately following subdomain XI which shows significant similarity between Crk1 and Sgv1 (Fig. 1B).

View larger version (94K): [in this window] [in a new window]   FIG. 1.   Comparison of the Crk1 sequence to sequences of other Cdc2-related kinases. (A) Diagram of predicted functional domains in Crk1. The shaded region contains the conserved kinase domain, as shown in panel B. Potential nuclear localization sequences (based on the PSORT program) near the carboxyl terminus are also indicated. (B) Sequence alignment of C. albicans Crk1 (CaCrk1) kinase domains with those of S. cerevisiae Sgv1 (ScSgv1), an S. pombe Cdc2 homologue (SpCdc2h; accession no. AB004534), a human Ser/Thr kinase (hukinase; accession no. AB020711), the human PITALRE kinase HuKp11/Cdk9, S. cerevisiae Cdc28, and S. cerevisiae Fus3. Subdomains are labeled according to Hanks et al. (27). Shaded residues represent identities among these kinases. Conserved phosphorylation sites in Cdc28 and Fus3 are indicated with asterisks and dots, respectively. Underlined sequences denote the insertion unique for the Cdc2 branch of the kinases. The arrow indicates the ending position of Crk1N and Sgv1N. (C) Kinase activity associated with Crk1 and Crk1N immunocomplexes. Yeast cell extracts were immunoprecipitated with anti-HA antibodies. Immunocomplexes were assayed for the ability to phosphorylate MBP (2-h exposure).

CRK1 suppresses the hypersensitive pheromone-induced growth arrest phenotype of S. cerevisiae sgv1 mutants. As shown in Fig. 1, Crk1 is most similar to S. cerevisiae Sgv1 in protein sequence. SGV1 was isolated in a mutant screen for suppressors that could repress hyperadaptation from pheromone-induced growth arrest in GPA1^Val50 cells (28). GPA1 encodes the  subunit of the G protein for pheromone receptors in S. cerevisiae. It also plays a positive role in promoting recovery from pheromone-induced growth arrest. The effect of sgv1 on recovery from pheromone treatment is not specific to the GPA1^Val50 mutation, as sgv1 mutants in otherwise wild-type strains are more sensitive to pheromone-induced growth arrest (28). The same sgv1 mutation also causes temperature-sensitive and cold-sensitive growth phenotypes, consistent with the fact that SGV1 is essential for vegetative growth (28).

View larger version (67K): [in this window] [in a new window]   FIG. 2.   Suppression of S. cerevisiae sgv1 mutants by C. albicans Crk1. (A) Pheromone-induced growth arrest assay. Haploid S. cerevisiae sgv1 mutants were transformed with vector (pVTU) (1), SGV1 (pVTUSGV1) (2), SGV1N (pVTUSGV1N) (3), CRK1 (pVTUCRK1) (4), and CRK1N (pVYUCRK1N) (5). Approximately 107 cells were plated on each YPD plate. Sterile filter disks were placed on the nascent cell lawns; -factor in the amounts of 0 ng (top), 50 ng (left), and 500 ng (right) was added to the disks. Plates were incubated for 2 days at 30°C. (B) Effect of temperature on growth. The strains used for panel B were used to test growth properties at 37 and 18°C. Cells were streaked onto SCUra plates, which were incubated for 5 days at 30°C, 7 days at 37°C, and 7 days at 18°C.

Chromosomal deletion of CRK1 in C. albicans. To elucidate cellular functions of CRK1, we deleted CRK1 in C. albicans. Part of the CRK1 coding region was replaced by URA3 with two flanking sequences of hisG for gene deletion by homologous recombination (Fig. 3A). A sequential gene disruption strategy was used to delete both copies of CRK1 in C. albicans as described by Fonzi and Irwin (21). Of 185 transformants from the first round of transformation, 90% had the hisG-URA3-hisG insertion at the CRK1 locus, based on Southern hybridizations (Fig. 3B, lane 2). The pattern of Southern hybridization with the 4-kb CRK1 probe is consistent with integration of the crk1::hisG-URA3-hisG construct at the CRK1 locus. After growth selection on FOA to remove the URA3 marker, the second copy of CRK1 was deleted by another round of transformation with the same crk1::hisG-URA3-hisG construct; 16 out of 78 transformants displayed the homologous recombination at the second copy of the CRK1. This was determined by the loss of the wild-type CRK1 gene shown by Southern hybridization (Fig. 3B, middle, lane 3).

View larger version (53K): [in this window] [in a new window]   FIG. 3.   Disruption of the C. albicans CRK1 gene. (A) Restriction map and disruption strategy for CRK1. (B) Southern analysis of transformants with the CRK1 disruption construct. Genomic DNA from the recipient strain (lane 1, CAI4), a heterozygote transformant (lane 2, CAW1), an FOAr/ura3 derivative of CAW1 (lane 3, CAW2), a homozygote transformant (lane 4, CAW3), and an FOAr/ura3 derivative of CAW3 (lane 5, CAW4) were digested with BamHI. The Southern blot on the left was probed with the 4-kb SacI fragment shown in panel A. The BamHI site in the hisG-URA3-hisG sequence generated two new hybridization fragments of 10.4 and 1.4 kb from the original 9-kb wild-type BamHI fragment. The 10.4-kb crk1::hisG-URA3-hisG fragment became an 8-kb crk1::hisG fragment after selection on an FOA plate to loop out the URA3 and one copy of hisG. This size difference between crk1::hisG and crk1::hisG-URA3-hisG is evident in lane 4, where the doublet represents fragments of 10.4 and 8 kb, respectively. The Southern blot in the middle was probed with the 2-kb EcoRV-XhoI fragment shown in panel A. The EcoRV-XhoI region was replaced with the hisG-URA3-hisG sequence in the deletion construct. Therefore, the 2-kb probe is expected to hybridize only to the 9-kb BamHI fragment from the wild-type CRK1 locus. Homozygous crk1/crk1 mutants do not contain the 9-kb BamHI fragment. The Southern blot on the right was probed with C. albicans URA3.

View larger version (82K): [in this window] [in a new window]   FIG. 4.   Effects of CRK1 disruption on cell growth. (A) crk1/crk1 strains grow slower than wild type. Wild-type (WT; SC5314), CRK1/crk1 (CAW1), crk1/crk1 (CAW3), crk1/crk1 carrying a vector (CAW5), and crk1/crk1 carrying CRK1 (CAW6) were grown on a YPD plate for 5 days at 22°C. (B) Comparison of cell morphologies. Wild-type (SC5314) and crk1/crk1 (CAW3) cells were grown in YPD medium at 22°C for 15 h and photographed.

Crk1 is necessary for hyphal development. Deletion of CRK1 caused a profound defect in hyphal development on all solid hypha-inducing media tested (Fig. 5A). On serum-containing agar medium, crk1/crk1 strains produced mostly round cells, with a very low percentage of stunted hyphal cells in the initial hours, whereas wild-type strains produced long hyphae. After 3 days of incubation, the wild-type strains generated florid hyphal colonies, whereas crk1/crk1 strains produced round colonies (Fig. 5A, top row). The colonies remained round even after a longer incubation time (Fig. 5A). The heterozygous strain CRK1/crk1 produced intermediate hyphal colonies (Fig. 5A). The defective hyphal growth was most likely caused by the CRK1 deletion, since reintroducing a wild-type CRK1 gene on an autonomous replicating plasmid rescued the mutant phenotype (Fig. 5A). On solid Lee's medium, crk1/crk1 strains also failed to develop hyphal colonies after 5 days (Fig. 5A, third row). On Lee's medium, the wild-type strain produced highly filamentous hyphal colonies and the heterozygous strain made intermediate hyphal colonies. We consistently observed that the defect in hyphal development persisted after 7 days. crk1/crk1 strains were also defective in developing hyphal colonies on all other solid media that we tried, including Spider medium and SLAD medium (data not shown).

View larger version (229K): [in this window] [in a new window]   FIG. 5.   crk1/crk1 strains are defective in hyphal development. (A) crk1/crk1 strains cannot develop hyphal colonies. Ura+ strains, including wild type (WT; SC5314), CRK1/crk1 (CAW1), crk1/crk1 (CAW3), and crk1/crk1 carrying ectopically expressed CRK1 (CAW6), were plated at a density of about 50 colonies per plate on solid serum-containing medium and solid Lee's medium. Colonies were photographed after incubation at 37°C for the days (d) indicated. (B) crk1/crk1 strains are impaired in hyphal filament formation in liquid media. Overnight cultures of the four Ura+ strains used for panel A were diluted in YPD medium containing 10% serum or modified Lee's medium for hyphal induction. Cells were photographed after incubation at 37°C for the time indicated. (C) crk1/crk1 mutants are defective in induction of hypha-specific genes. RNA from wild-type cells grown in YPD at 30°C for 6 h was used as a control for gene expression in the yeast growth form (left). RNA from cells of the same Ura+ strains induced for 3.5 h by serum (panel B, top row) and induced for 6 h in Lee's medium (panel B, middle row) were subjected to Northern analysis, as shown in panel C, middle and right gels, respectively. Northern probes are as indicated. The image for the CRK1 Northern blot was obtained after 3 days of exposure. Images for ECE1-, HWP1-, and ACT1-probed filters were obtained after 3 h of exposure. Transcript levels were quantified with a PhosphorImager.

crk1/crk1 is avirulent in mice. The dimorphic transition ability of C. albicans has been linked with its pathogenicity in mice (22, 35, 39). Here we assessed the virulence of crk1/crk1 mutant strains by the intravenous injection of mice (Materials and Methods). Injection of mice with wild-type C. albicans cells is fatal. Injection with an inoculum of 5 × 10⁶ cells caused all mice to die in 6 days, and a smaller inoculum of 5 × 10⁵ cells killed all mice in 13 days (Fig. 6A). The heterozygotic CRK1/crk1 mutant strain was less virulent than the wild-type strain despite being capable of hyphal development (Fig. 5A and B). All mice survived for over 20 days after injection with 5 × 10⁵ cells, and 60% survived after 10 days with an inoculum of 5 × 10⁶ cells (Fig. 6B). The crk1/crk1 cells were avirulent at both inoculum sizes: all mice survived for more than 20 days after injection with 5 × 10⁶ or 5 × 10⁵ crk1/crk1 cells (Fig. 6C). This is comparable to the virulence level of the cph1/cph1 efg1/efg1 double mutant (Fig. 6F), which has been previously shown to be avirulent with a similar inoculum size (39). We also used cph1/cph1 and hst7/hst7 single mutants as controls in our experiments. The two strains showed similar survival curves and were slightly less virulent than the wild type (Fig. 6E and D) (39). In comparison, the heterozygotic CRK1/crk1 mutant was less virulent than either cph1/cph1 or hst7/hst7 strains.

View larger version (22K): [in this window] [in a new window]   FIG. 6.   Virulence assay. ICR male mice were injected with wild-type (WT; SC5314; A), CRK1/crk1 (CAW1; B), crk1/crk1 (CAW3; C), hst7/hst7 (JKC129; D), cph1/cph1 (JKC19; E), and cph1/cph1 efg1/efg1 (HLC54; F) strains. The mice were injected with 5 × 105 () and 5 × 106 (×) C. albicans cells. Mice injected with either crk1/crk1 cells or cph1/cph1 efg1/efg1 cells all survived for more than 20 days.

CRK1N promotes invasive or filamentous growth in S. cerevisiae through Flo8 but not through the filamentation MAP kinase pathway or Phd1. Many of the regulatory components of dimorphic transition are conserved between C. albicans and S. cerevisiae. Furthermore, several C. albicans hyphal regulatory proteins were identified by their ability to promote pseudohyphal growth in S. cerevisiae (36, 60). Therefore, we decided to use S. cerevisiae as an initial step to investigate potential regulatory targets of Crk1.

View larger version (43K): [in this window] [in a new window]   FIG. 7.   CRK1N stimulated filamentous and invasive growth in S. cerevisiae. (A) Colony morphologies of isogenic wild-type (WT; CG31), ste7/ste7 (HLY351), ste12/ste12 (HLY352), tec1/tec1 (HLY2002), phd1/phd1 ste12/ste12 (L6235), and flo8/flo8 (HLY852) strains carrying vector (left) or CRK1N (right) grown on SLAD at 30°C for 4 days. (B) Total and invasive growth of wild-type (L5528), ste7 (HLY367), ste12 (HLY362), tec1 (HLY2000), and flo8 (HLY850) strains carrying a vector (left) or CRK1N (right) after 5 days of growth on SCUra.

Ectopic expression of CRK1N promotes hyphal growth under conditions favorable for yeast growth. While complementing the hypha-defective phenotype in crk1/crk1 mutants, we observed that the ectopic expression of CRK1 enhanced hyphal growth under conditions that are otherwise favorable for the yeast form of growth. The ectopic expression of CRK1 allowed formation of visible hyphal colonies after 3 days of growth on solid YPD medium (Fig. 8A). No filaments were observed in the wild-type control strain grown under the same conditions until day 5 (Fig. 8A). We also observed that the expression of the catalytic domain of Crk1 promoted more filamentous growth than Crk1 (Fig. 8A), indicating that Crk1N might be more active than Crk1.

View larger version (55K): [in this window] [in a new window]   FIG. 8.   Ectopic expression of CRK1- or CRK1N-promoted filamentation under yeast growth conditions. (A) Colony phenotypes of wild-type (SC5314), crk1/crk1 with vector (CAW5), crk1/crk1 with ectopic expression of CRK1 (CAW6), and CRK1N (CAW7) strains grown on YPD medium at 30°C for 3 days. (B) Induction of ECE1 transcription by ectopic expression of CRK1 and CRK1N. Wild-type (SC5314), CRK1/crk1 (CAW1), crk1/crk1 carrying a vector (V; CAW5), crk1/crk1 carrying CRK1 (CAW6), and crk1/crk1 carrying CRK1N (CAW7) strains were grown in YPD at 30°C for 6 h, and total RNA was extracted for Northern analysis. The Northern blot was probed with CRK1, ECE1, and ACT1 and exposed for 3 days (for the CRK1 and ECE1 transcripts) and 3 h (for ACT1).

CRK1 can promote hyphal development through a pathway that is independent of Cph1 and Efg1 in C. albicans. To address the function of Crk1 in the context of known C. albicans signaling components, we expressed CRK1N in various C. albicans strains defective in the filamentation MAP kinase pathway and the parallel pathway Efg1. CRK1N suppressed the hyphal formation defect in cst20/cst20 strains (Fig. 9). The effect of CRK1N was evident after 3 days but more obvious after a longer incubation. Although Cst20 is supposedly in the same pathway as Hst7 and Cph1, the filament-promoting activity of Crk1N was much weaker in hst7/hst7 and cph1/cph1 strains than in the cst20/cst20 strain. As shown in Fig. 9, hst7/hst7 and cph1/cph1 colonies were more filamentous than cst20/cst20 colonies, but with the expression of CRK1N, they were less filamentous than the cst20/cst20 strain. This indicated that the signaling pathway from Cst20 to Hst7 might not be linear. We also showed that CRK1N could partially suppress the defect of hyphal development in the efg1/efg1 mutants (Fig. 9). Surprisingly, when transformed with CRK1N, efg1/efg1 cph1/cph1 double mutants generated more hyphal filaments than either efg1/efg1 or cph1/cph1 single mutants with CRK1N (Fig. 9). Comparable to this observation, RAS1^V13 promoted dramatic hyphal formation in efg1/efg1 cph1/cph1 double mutant. Subtle differences existed between CRK1N- and RAS1^V13-activated filamentation. CRK1N-promoted hyphal filamentation was most evident around the initial streaks and well-separated single colonies, whereas RAS1^V13-activated filamentation was seen throughout the streak regardless of the colony density. Interestingly, the expression of RAS1^V13 in either MAP kinase pathway-defective strains or efg1/efg1 strains led to the formation of large wrinkled sheet-like colonies (Fig. 9). The fact that both CRK1N and RAS1^V13 generated more filaments in the double mutant than in each of the single mutants indicated that there might be complicated negative regulation by Cph1 and Efg1 on filamentation. The phenotypes of CRK1N and RAS1^V13 in the efg1/efg1 cph1/cph1 strain suggested that they both could activate hyphal development through pathways independent of Efg1 and Cph1.

View larger version (120K): [in this window] [in a new window]   FIG. 9.   Functional relationship of Crk1 with the filamentation MAP kinase pathway, Efg1, and Ras1 in C. albicans hyphal development. The C. albicans mutant strains indicated on the left (described in Table 1) were transformed with CRK1N (BES119CRK1N) and RAS1V13 (pQF145.2). Both genes are under the control of the MAL2 promoter. The C. albicans transformants were grown on an SCUra+sucrose (2%) plate containing 50 mM succinate at pH 5 for 4 days at 30°C (20). Well-separated colonies near the edge of each steak were photographed.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

A new member of the Cdc2-related protein kinase family with a regulatory carboxyl terminus. We have cloned a novel gene, CRK1, which encodes a Ser/Thr kinase with a catalytic domain highly conserved among kinases of the Cdk subfamily. The kinase domain of Crk1 is most similar to those of the S. cerevisiae protein Sgv1 and three other Cdc2-related protein kinases. In addition to being similar in sequence, Crk1 may also share overlapping functions with Sgv1, because CRK1 partially suppresses the hypersensitive pheromone-induced growth arrest phenotype in sgv1 mutants. This suppression is specific to sgv1, as neither Crk1 nor Crk1N promotes adaptation to pheromone induction in wild-type strains. However, Crk1 and Sgv1 also have nonoverlapping functions. First, Crk1 cannot complement the conditional growth defect of sgv1. Second, overexpression of full-length Sgv1 or its catalytic domain does not promote invasive or filamentous growth in S. cerevisiae (Chen, unpublished observation). Third, Sgv1 is essential for viability in S. cerevisiae whereas Crk1 is dispensable in C. albicans. Since the deletion of SGV1 leads to lethality in S. cerevisiae, it is impossible to address whether sgv1/sgv1 mutants will block pseudohyphal growth. All of these findings indicate that functional differences exist between Crk1 and Sgv1. Similar to Crk1, the Ras proteins from S. cerevisiae and C. albicans also have functional differences; S. cerevisiae Ras proteins are essential, whereas the C. albicans Ras1 protein is not. The functional differences between Crk1 and Sgv1 may reflect variation in substrate specificity. The substrate specificity is likely to be defined by a region in the catalytic domain of Crk1 and Sgv1 since the activity to promote invasive/filamentous growth and the ability to complement the conditional growth in sgv1 are supported by the catalytic domains of Crk1 and Sgv1, respectively.

Role of Crk1 in hyphal development in C. albicans. Crk1 is required for hyphal development under all hypha-inducing conditions investigated. Deleting CRK1 dramatically impaired hyphal formation under various hypha-inducing conditions, whereas the ectopic expression of its catalytic domain promoted hyphal colony formation even under conditions favorable for yeast form growth. Crk1 is probably not directly responsible for changes in cytoskeleton that are necessary for the polarized growth during hyphal development. Rather, several lines of evidence support its role in regulating the transcriptional program of hypha-specific genes. First, crk1/crk1 mutants are severely impaired in the induction of two hypha-specific genes. Second, the catalytic domain of Crk1 can induce the expression of hyphal genes under yeast growth conditions when hyphal genes are normally undetectable. Third, the ectopic expression of the Crk1 catalytic domain in S. cerevisiae promoted invasive growth, a phenomenon caused by the expression of a cell wall protein, Flo11 (40). Flo11 is necessary for invasive growth, and its expression is regulated by transcription factors required for invasive/pseudohyphal growth. Finally, the Crk1 sequence predicts two conserved basic bipartite nuclear localization sequences at the carboxyl terminus (Fig. 1), suggesting a nuclear function. Taken together, these findings indicate that Crk1 plays a role in regulating the hyphal transcriptional program. This could be achieved by its phosphorylation of some transcription factor(s) or regulator(s) important for hyphal development.

Integrative regulation of multiple pathways for hyphal development. Our data suggest that Ras1 and Crk1 regulate additional pathways independent of Cph1 and Efg1, and all of these pathways converge to regulate a common set of hypha-specific genes. This is reminiscent of the pseudohyphal development program in S. cerevisiae, where the Kss1 MAP kinase pathway and the cAMP-regulated pathway converge on the promoter of FLO11 (53, 55). Since both crk1/crk1 and efg1/efg1 mutants are unable to induce hyphal development under similar in vitro hyphal growth conditions, it is unlikely that each pathway responds to a particular hypha-inducing condition; rather, each pathway may respond to a specific signal in a hypha-inducing condition. The strong activation for any one pathway may be enough to reach a threshold required for hyphal development, but in many cases, integrated inputs from more than one pathway may be required to reach the threshold required for hyphal development. A defect in any one of the signaling pathways will reduce the total integrated inputs and hamper hyphal development. This integrative regulation may be necessary for fine-tuning of the signaling system, such as in rapid response versus sustained activation. Alternatively, multiple pathways could be used by C. albicans cells to sense subtle differences in the growth conditions of its native host environment. Our defined laboratory media, which have been chosen for its all or no hyphal growth property, may fail to mimic the subtle differences in growth conditions in the host.

Virulence. The crk1/crk1 is avirulent under the conditions used in our investigation. The reduced growth rate of crk1/crk1 may contribute to the reduced virulence. The reduced virulence could also be due to the impaired ability of crk1/crk1 strains to undergo hyphal formation, as many mutants defective in hyphal formation have been shown to be less virulent or avirulent compared to the wild type (14, 22, 35, 39). Alternatively, there may be other targets of Crk1 that are not involved in hyphal growth but are required for virulence. Given its role in virulence, uncovering the proteins regulated by Crk1 should reveal new targets for antifungal drug development.