Properties of Various RHO1 Mutant Alleles of Cryptococcus neoformans

Abstract

The RHO1 homologue of Cryptococcus neoformans complemented Saccharomyces cerevisiae rho1 mutations. The results of overexpression and site-specific mutagenesis of CnRHO1 in C. neoformans and S. cerevisiae indicated that although CnRHO1 could functionally substitute for the RHO1 gene of S. cerevisiae, mutants of cnrho1 manifested unique features in certain aspects.

The Rho GTPases belong to the Ras-related superfamily of small G proteins, which function as molecular switches between the active, GTP-bound form and the inactive, GDP-bound form (for reviews, see references 14, 33, and 34). In Saccharomyces cerevisiae, RHO1 is essential (26) and Rho1p is localized at the cell periphery, at the bud growing site (37). Rho1p interacts with various proteins which are important for cell wall synthesis and actin organization. The genetic and biochemical evidence demonstrates that Rho1p interacts with and activates protein kinase C (Pkc1p), which maintains cell wall integrity through the activation of the mitogen-activated protein kinase cascade (19, 31). Rom7p/Bem4p, which is related to bud emergence, has been identified as a downstream target of Rho1p (15). Rho1p may control the actin cytoskeleton by binding Bni1p (12, 21). Bni1p is the yeast homologue of the mammalian proteins called formins, which participate in morphogenesis (12). Rho1p was also identified as the putative regulatory subunit of β-1,3-glucan synthase, which produces a major structural component of the yeast cell wall (10, 32). The regulation of β-1,3-glucan synthesis by Rho1p has also been demonstrated in Schizosaccharomyces pombe (1) and Candida albicans (22).

The fungal cell wall confers cells with rigidity and protects them from osmotic pressure. It consists mainly of β-glucans, mannoproteins, and small amounts of chitin (4, 7, 13). Among these components, β-1,3-glucan is the major component of many fungal cell walls. From studies with S. cerevisiae, it has been demonstrated that β-1,3-glucan synthase is composed of at least one large integral membrane catalytic subunit and a small loosely membrane-associated regulatory subunit (16, 20, 29). The catalytic subunit has been suggested to be encoded by FKS1 and FKS2 (9, 16, 28). The regulatory subunit is encoded by RHO1 (10, 27, 32). Recently, sequence analysis of the Cryptococcus neoformans RHO1 cDNA (CnRHO1) indicated that the putative protein encoded by CnRHO1 contains 197 amino acids and shares a high degree of sequence identity with Rho1p in other fungal species (35). The function of CnRHO1 in C. neoformans, however, was not elucidated. C. neoformans is an important fungal pathogen that causes meningoencephalitis, primarily in immunocompromised patients (25). Since the regulation of glucan synthase by CnRHO1 has not yet been established, it is of interest to study the roles of CnRHO1 in C. neoformans.

Cloning of the CnRHO1 gene.

A temperature-sensitive strain of S. cerevisiae (HNY21, MATa ura3 leu2 trp1 his3 ade2 rho1-104^ts; a gift from E. Cabib) was transformed with a cDNA library of C. neoformans constructed in an S. cerevisiae vector, pGAD10. Three transformants, which were able to grow at 37°C, were obtained (data not shown). This result suggested that a cloned cDNA of C. neoformans could functionally suppress the temperature-sensitive mutation rho1-104 in S. cerevisiae. The sequence of the cloned C. neoformans cDNA insert showed great similarity to the RHO1 gene of S. cerevisiae and was almost identical to the published sequence of CnRHO1 cDNA of C. neoformans, which was obtained by PCR amplification (35). The differences between the cloned cDNA and the published CnRHO1 cDNA sequence were at positions 266 (T→C) and 500 (G→C) (35). This discrepancy may be attributed to the different strains which were used to construct the cDNAs. The cloned cDNA, however, encoded the same 21.7-kDa putative protein as the published cDNA sequence and hybridized to a single 7.0-kb BamHI fragment of genomic DNA from C. neoformans (data not shown). By comparing the cDNA and genomic sequences, we found that the CnRHO1 gene contains seven introns. Most of the introns exhibit the consensus splice donor (GTNNGY) and acceptor (YAG), except that the splice donor of intron 6 is GTCCCT. The details for construction of all the plasmids in this study are available upon request.

Effects of CnRHO1 mutations in S. cerevisiae.

Since the cDNA of CnRHO1 could function in S. cerevisiae, we constructed a variety of mutations in the cDNA of CnRHO1 to further study its function in S. cerevisiae. The first type of mutations was nucleotide substitution mutations in the phosphate and magnesium binding sites, which are similar to those in other deregulated GTPases (3). We generated two site-specific mutations, cnrho1-Q64L and cnrho1-G15V, which produce constitutively active rho1 mutants in several different organisms. Three alleles, cDNA of CnRHO1, cnrho1-Q64L, and cnrho1-G15V, were placed under the control of a GAL1 promoter and expressed in S. cerevisiae strain HNY21. When transformants containing these plasmids were grown on glucose, they exhibited a temperature-sensitive phenotype, similar to the vector control (Fig. 1A, columns 1, 2, and 3 versus column 4). In the presence of galactose, transformation with the wild-type CnRHO1 did not affect the growth of yeast cells at 30°C (Fig. 1A, column 1). This observation was different from the previously reported result in which overexpression of S. cerevisiae RHO1 under the control of a GAL1 promoter caused growth arrest in yeast (11). Cells transformed with cnrho1-Q64L were not able to grow on galactose medium (Fig. 1A, column 2). This growth arrest phenotype of cnrho1-Q64L was similar to that of the rho1-Q68L phenotype of S. cerevisiae (31). These results suggested that the effect of overexpression of CnRHO1 alleles in S. cerevisiae is similar but not identical to that of RHO1 alleles in S. cerevisiae. We noted that the growth of yeast cells was slightly reduced when cnrho1-G15V was overexpressed in S. cerevisiae at 30°C (Fig. 1A, column 3). This observation differed from the effect of a similar mutation in rho1 of S. pombe, in which overexpression of rho1-G15V caused growth arrest (1). Despite the slight reduction in growth at 30°C, cnrho1-G15V-containing transformants were the only strains which could grow at 37°C on galactose medium (Fig. 1A, column 3), while overexpression of CnRHO1 and cnrho1-Q64L failed to suppress the temperature-sensitive phenotype of HNY21 at 37°C.

FIG. 1.

(A) Overexpression of CnRHO1 alleles in S. cerevisiae. Different alleles of CnRHO1 were placed under the control of the GAL1 promoter and expressed in the HNY21 strain background. Transformants were grown as indicated. Column 1, GAL1(p)::CnRHO1; column 2, GAL1(p)::cnrho1-Q64L; column 3, GAL1(p)::cnrho1-G15V; column 4, vector. (B) Expression of CnRHO1 effector mutants in S. cerevisiae. Different alleles of CnRHO1 were placed under the control of the native S. cerevisiae RHO1 promoter and terminator. The resulting plasmids were transformed into a Δrho1 strain of S. cerevisiae, as described in the text. Transformants were incubated at either room temperature (RT) or 37°C. Column 1, RHO1(p)::CnRHO1; column 2, RHO1(p)::cnrho1-V39T; column 3, RHO1(p)::cnrho1-E41I. (C) Overexpression of CnRHO1 in C. neoformans. Different alleles of CnRHO1 were placed under the control of the C. neoformans GAL7 promoter and transformed into the LP1 strain. Transformants were grown as indicated. Column 1, GAL1(p)::CnRHO1; column 2, GAL7(p)::cnrho1-Q64L; column 3, GAL1(p)::cnrho1-G15V; column 4, vector. Gal + Pne, galactose medium with 128 μg of pneumocandin B₀/ml. (D) Phenotypes of TYCC314 and TYCC295. Replacing the CnRHO1 allele with the cnrho1-E41I allele in C. neoformans generated a temperature-sensitive phenotype (TYCC314). B-4476-FO5 is a congenic strain containing wild-type CnRHO1. The temperature-sensitive phenotype of TYCC314 was suppressed by 1 M sorbitol. Reconstitution of cnrho1-E41I back to the wild-type allele restored growth at 37°C (TYCC295).

Another type of mutation, the effector mutation, is known to inhibit the interaction of GTPases with target molecules (31, 35). To determine the consequence of effector mutations of CnRHO1, three alleles, cDNA of CnRHO1, cnrho1-V39T, and cnrho1-E41I, were placed under the control of the S. cerevisiae RHO1 promoter as well as the terminator. These constructs were transformed into S. cerevisiae strain JDY6-7A[pRS316(RHO1)] (MATa ade2 ura3 leu2 trp1 his3 rho1::HIS3 pRS316. RHO1[URA3]; a gift from E. Cabib), which contains a rho1 deletion in the genome while harboring a URA3 plasmid expressing the RHO1 gene. Transformants were subsequently transferred to 5-fluoroorotic acid medium to isolate uracil auxotrophs. If these constructs could functionally substitute for the RHO1 gene of S. cerevisiae, viable cells could be recovered after removing the RHO1-containing URA3 plasmid in strain JDY6-7A[pRS316(RHO1)]. We found that CnRHO1 cDNA could functionally complement the rho1 deletion in JDY6-7A[pRS316(RHO1)] (Fig. 1B, column 1). We also found that mutations in the effector region of CnRHO1 cDNA caused a growth reduction at 30°C in S. cerevisiae, and the effect on growth was more severe in cnrho1-V39T than in cnrho1-E41I (Fig. 1B, columns 2 and 3). Both the rho1-E41I and rho1-V39T mutations of S. cerevisiae have been shown to behave as temperature-sensitive alleles in yeast (31), but mutations in the effector region that cause reduction in growth at 30°C have not been described. Lastly, transformants of the wild-type CnRHO1 were viable at 37°C, while strains derived from transformants containing effector mutations of CnRHO1 cDNA failed to grow at 37°C. These results suggested that although CnRHO1 can functionally substitute for the RHO1 gene of S. cerevisiae, CnRHO1 does not behave the same way as the RHO1 homologue in some yeast strains.

Overexpression of CnRHO1 in C. neoformans.

Since overexpression of the deregulated mutations of cnrho1 cDNA in S. cerevisiae resulted in interesting phenotypes, we attempted to determine the effects of constitutively active mutations of CnRHO1 on the growth of C. neoformans. Three cDNA clones, CnRHO1, cnrho1-Q64L, and cnrho1-G15V, were placed under the control of the C. neoformans GAL7 promoter. The resulting constructs were transformed into an ade2 ura5 recipient strain (LP1). PCRs were performed to confirm the existence of unaltered fusion constructs by using specific primers which were designed to detect the GAL7-CnRHO1 fusion construct (data not shown). PCR-positive transformants were transferred onto galactose medium to overexpress each construct of CnRHO1. Unlike the results of overexpression of CnRHO1 in S. cerevisiae described above, growth of C. neoformans at either 30 or 37°C was not affected when CnRHO1 was overexpressed (Fig. 1C, column 1). The growth of transformants containing cnrho1-Q64L was greatly reduced on galactose medium, but these transformants were still viable at either 30 or 37°C (Fig. 1C, column 2). In contrast, overexpression of cnrho1-G15V did not affect growth of C. neoformans at either temperature (Fig. 1C, column 3). These results showed that the effects of overexpression of CnRHO1, as well as overexpression of deregulated cnrho1 alleles, in C. neoformans are different from those in S. cerevisiae.

S. pombe is hypersensitive to the glucan synthase inhibitor papulacandin B when RHO1 is overexpressed (1). Pneumocandin lipopeptides are a class of compounds also known to inhibit β-1,3-glucan synthesis (8, 24). C. neoformans, however, is notably less susceptible to these inhibitors (2, 17, 23). We tested whether overexpression of CnRHO1 could cause hypersensitivity to pneumocandin B₀. The presence of up to 128 μg of pneumocandin B₀ per ml had no obvious effect on growth in a strain overexpressing CnRHO1 or cnrho1-G15V, while pneumocandin B₀ slightly exacerbated the retarded growth in cells overexpressing the cnrho1-Q64L allele (Fig. 1C, Gal + Pne). Thus, C. neoformans was resistant to pneumocandin B₀ even when CnRHO1 was overexpressed. Another interesting observation was that when CnRHO1 cDNA was expressed in a rho1 deletion mutant of S. cerevisiae, the resulting strain was as sensitive as control strains to pneumocandin B₀ (data not shown).

Glucan synthase activity in strains overexpressing CnRHO1.

Since the β-1,3-glucan-specific fluorochrome aniline blue poorly stains Cryptococcus cells, it has been suggested that β-1,3-glucan is not prevalent in Cryptococcus cell walls (31). This observation is further supported by chemical analysis of cell wall components (18). On the other hand, a different study suggests that CnFKS1, which may encode a β-1,3-glucan synthase catalytic subunit, may be required for the viability of C. neoformans (37). As glucan synthase activity of many fungi is regulated by RHO1p, it was of interest to study the relationship between RHO1p and glucan synthase in C. neoformans. To determine whether CnRHO1 could modulate glucan synthase activity, glucan synthase activity was determined using the membrane fraction of C. neoformans cells in which various alleles of CnRHO1 had been overexpressed. Enzyme activity was quantitated by measuring the incorporation of UDP-glucose into a trichloroacetic acid-insoluble fraction, as described previously (36) but with modifications. The reaction mixtures contained crude membrane fraction, 50 mM Tris (pH 7.5), 10% glycerol, 0.5 mM EDTA, 25 mM KF, 0.25% (wt/vol) bovine serum albumin, 20 μM [γ-S]GTP, UDP-[¹⁴C]glucose (specific activity, 319 mCi/mmol; Amersham) and 10 mM unlabeled UDP-glucose (specific activity, 250,000 cpm/μmol). After 2 h of incubation with gentle agitation, reactions were terminated with ice-cold trichloroacetic acid. Glucan synthase activity was expressed as nanomoles per milligram per hour. Glucan synthase activity (mean ± standard deviation) decreased 39.7% in transformants containing only the vector when GTP was excluded from the reaction mixture (55.7 ± 2.1 versus 33.6 ± 6.2). Thus, as in other fungi, the glucan synthase activity of C. neoformans was affected by the presence of GTP. In cryptococcal transformants overexpressing the wild-type CnRHO1 allele, glucan synthase activity in the presence or absence of GTP (61.6 ± 9.3 versus 36.8 ± 5.1) was not significantly different from that of the corresponding vector control. In contrast, the effect of the presence of GTP on glucan synthase activity was much lower in strains overexpressing the activated mutant alleles cnrho1-Q64L (61.4 ± 3.4 versus 51.8 ± 6.1; 15.6% reduction) and cnrho1-G15V (57.9 ± 1.9 versus 52.3 ± 0.4; 9.7% reduction). These results suggested that, as observed with S. cerevisiae (32), when cells overexpressed the deregulated cnrho1 alleles, addition of exogenous GTP to the reaction had little influence on the activity of glucan synthase.

We noted that in the presence of GTP, only a small difference in glucan synthase activity was observed between the strains overexpressing wild-type CnRHO1 and strains containing vector alone. Similarly, a small increase in glucan synthase activity was observed in strains overexpressing the constitutively active mutants, including cnrho1-Q64L and cnrho1-G15V. This minor increase in glucan synthase activity in response to the overexpression of CnRHO1 alleles differed from observations with S. pombe, in which overexpression of S. pombe RHO1 resulted in more than fourfold increase of glucan synthase activity, and overexpression of the constitutively active mutants rho1-G15V and rho1-Q64L caused a >20-fold increase (1). A similar significant influence on glucan synthase activity was observed when the RHO1 gene was overexpressed in S. cerevisiae (32).

Construction of a temperature-sensitive RHO1 allele in C. neoformans.

We attempted to delete the CnRHO1 gene in C. neoformans, with no success (data not shown). It is likely that CnRHO1 is an essential gene, as is the case in other organisms. Since cnrho1-E41I cDNA behaved as a temperature-sensitive allele in S. cerevisiae, we anticipated that a replacement of the wild-type CnRHO1 gene with a cnrho1-E41I allele in the genome of C. neoformans might generate a temperature-sensitive phenotype. The cnrho1-E41I mutation was generated in vitro and was used to construct a replacement plasmid, pYCC314 (Fig. 2B). Plasmid pYCC314 was transformed into C. neoformans, and transformants were selected by a positive-negative selection method to obtain cells containing a gene replacement (5). A PCR screening method was applied by using insertion-specific primers to identify putative transformants, which presumably carry the replacement plasmid at the CnRHO1 locus (Fig. 2B). Southern blot analysis was then used to confirm the integration of cnrho1-E41I at the CnRHO1 locus by a gene replacement event. Figure 2F shows that a 7.0-kb wild-type CnRHO1 fragment was replaced with a 10-kb fragment in TYCC314 due to homologous integration, as depicted in Fig. 2C.

FIG. 2.

(A) Genomic map of the CnRHO1 gene. (B) Construct for generating the temperature-sensitive cnrho1 allele pYCC314. For simplicity, the URA5 gene is not drawn to scale. (C) Genomic map of the temperature-sensitive cnrho1 allele in TYCC314. (D) Map of pYCC295 used in cotransformation (6). (E) Map of reconstituted CnRHO1. (F) Southern blot for the temperature-sensitive cnrho1 strain TYCC314. (G) Southern blot for CnRHO1-reconstituted strain TYCC295. DNA was isolated, digested with restriction enzyme, fractionated on a 0.8% agarose gel, and analyzed by Southern blotting. The blots were hybridized with a probe of the 7.0-kb BamHI fragment of CnRHO1. A, AatII; B, BamHI; E, EcoRI; S, SmaI; Sp, SphI. Arrowheads indicate primer locations; gray boxes indicate CnRHO1 coding region; crosses indicate crossing over; dashed lines indicate the chromosomal region flanking CnRHO1; and stars indicate E41I. B-4476, wild type; TYCC314, temperature-sensitive cnrho1 strain; TYCC295, CnRHO1-reconstituted strain. Numbers to the right of the gels in panels F and G indicate the sizes of DNA markers in kilobases.

As predicted, TYCC314 showed a temperature-sensitive phenotype at 37°C, while the congenic strain, B-4476FO5, grew well at both room temperature and 37°C (Fig. 1D). Since the temperature-sensitive phenotype caused by a similar mutation of rho1 in S. cerevisiae could be suppressed by 1 M sorbitol (31), we grew TYCC314 on 1 M sorbitol at a restrictive temperature and found that the temperature tolerance was restored in TYCC314 (Fig. 1D). The growth rate of TYCC314 was also compared to the growth rate of B-4476FO5 in liquid culture. The growth rate of TYCC314 was slightly lower than that of B-4476FO5, but TYCC314, like B-4476FO5, continued to grow at 25°C after prolonged incubation. At 37°C, however, TYCC314 showed a much longer doubling time, and the optical density at 600 nm of the culture ceased to increase by 10 h after the culture was shifted to 37°C. Viability of the cells after 12 h of incubation at 37°C was compared between TYCC314 and B-4476FO5. The proportion of viable cells of TYCC314 was only 11.5% of that of B-4476FO5. In addition, some TYCC314 cells grown at 37°C were undergoing lysis, while no lysis was observed for the control strain, B-4476FO5 (data not shown).

Glucan synthase activity was also measured at 37°C in the membrane fractions isolated from TYCC314 and B-4476FO5. No significant difference in glucan synthase activity was found between TYCC314 and B-4476FO5 (80.8 ± 2.6 versus 76.2 ± 3.4). To confirm that the temperature sensitivity of TYCC314 was caused by the cnrho1-E41I mutation and not by other hidden mutations, the cnrho1-E41I allele of TYCC314 was restored to the wild type by a cotransformation method (Fig. 2C, D, and E). One of the putative adenine auxotrophic transformants, TYCC295, was isolated and analyzed by Southern blot analysis. TYCC295 showed the same hybridization pattern as the control wild-type strain, which indicated that the CnRHO1 allele was reconstituted (Fig. 2G). Lastly, the temperature-sensitive phenotype of TYCC314 was also corrected in TYCC295 (Fig. 1D). These results suggested that the temperature-sensitive phenotype of TYCC314 was caused by a mutation in the cnrho1 gene, and replacement of the mutant allele with the wild-type allele restored the ability of the C. neoformans strain to grow at 37°C.

PKC1 is a downstream target of RHO1. It has been shown that PKC1-R398P suppresses the temperature-sensitive phenotype of rho1-E45I and of a RhoA mutant strain of S. cerevisiae (31). The temperature-sensitive phenotype of a RhoA mutant strain can also be suppressed by the presence of 1 M sorbitol. Because TYCC314 behaved as a temperature-sensitive strain and the temperature sensitivity could be rescued by 1 M sorbitol, it would be interesting to see whether PKC-R398P of C. neoformans, when available, could suppress the temperature sensitive phenotype. Since RHO1 of S. cerevisiae or S. pombe interacts with several different genes in different pathways, similar properties may also exist in CnRHO1 of C. neoformans. Future studies isolating the suppressors of TYCC314 may further elucidate the functions of CnRHO1.

Nucleotide sequence accession number.

The GenBank accession number for the genomic sequence of CnRHO1 is AF242351.