Insight into the Role of HOG Pathway Components Ssk2p, Pbs2p, and Hog1p in the Opportunistic Yeast Candida lusitaniae

Abstract

In the present study, we have investigated the role of SSK2, PBS2, and HOG1, encoding modules of the high-osmolarity-glycerol mitogen-activated protein kinase pathway in Candida lusitaniae. Functional analysis of mutants indicated that Ssk2p, Pbs2p, and Hog1p are involved in osmotolerance, drug sensitivity, and heavy metal tolerance but not in oxidant stress adaptation.

One important component of the Saccharomyces cerevisiae signaling network is the high-osmolarity-glycerol (HOG) mitogen-activated protein kinase (MAPK) pathway, which responds to osmotic, oxidative, and heavy metal stress (for reviews, see references 14 and 30). This transduction module also plays a broad role in regulating morphology, growth, and adaptation to various conditions in a number of other fungal species (1, 3, 7, 8, 10, 12, 13, 15-18, 20, 21, 25, 27, 32, 33, 35, 36). The HOG MAPK signaling pathway consists of MAPK Hog1p, MAPK kinase (MAPKK) Pbs2p, and MAPKK kinase (MAPKKK) Ssk2p, which communicate by a phosphorylation cascade. It is now accepted that the HOG pathway can be regulated by two upstream branches. The first branch is referred to as the two-component system and includes the histidine kinase receptor Sln1p, the histidine-containing phosphotransferase Ypd1p, and the response regulator Ssk1p. Ssk1p modulates the Ssk2p-Pbs2p-Hog1p activity by regulating the Ssk2p MAPKKK. The second branch is initiated by the Sho1p adaptor protein, which makes use of several proteins (Ste20p, Ste50p, and Ste11p) to link the HOG cascade through the Pbs2p MAPKK (31).

In two previous studies, we reported the characterization of genes (SLN1, NIK1, CHK1, YPD1, SSK1, and SKN7) encoding members of the two-component system in the opportunistic yeast Candida lusitaniae (5, 29). In the present study, we investigated the contribution of the downstream HOG MAPK pathway elements Ssk2p, Pbs2p, and Hog1p in the stress adaptation, drug sensitivity, mating ability, and yeast-to-pseudohyphae morphological transition of this fungal species.

The sequences of the C. lusitaniae SSK2 (CLUG_02562.1), PBS2 (CLUG_03729.1), and HOG1 (CLUG_01883.1) genes were retrieved from the annotated Candida database of the Broad Institute (http://www.broad.mit.edu/). To determine the role of these genes in C. lusitaniae, null mutants were constructed for each one by using the URA3 blaster strategy (5, 24) (see Fig. S1 in the supplemental material). All disruptant and reintegrant strains are listed in Table 1.

TABLE 1.

Candida lusitaniae strains

Strain	Genotype	Mating type	Parent	Generation (h)c
6936a	Wild type	MATa		1.23 ± 0.02
CL38b	Clinical isolate	MATα		ND
6936 ura3[Δ360]b	ura3[Δ360]	MATa	6936	ND
PC1 (α)b	ura3[Δ360]	MATα	CL38	ND
ssk2Δ	ura3[Δ360]ssk2Δ::REP-URA3-REP	MATa	6936 ura3[Δ360]	1.31 ± 0.02
pbs2Δ	ura3[Δ360]pbs2Δ::REP-URA3-REP	MATa	6936 ura3[Δ360]	1.37 ± 0.03
hog1Δ	ura3[Δ360]hog1Δ::REP-URA3-REP	MATa	6936 ura3[Δ360]	1.26 ± 0.05
ssk2Δα	ura3[Δ360]ssk2Δ::REP-URA3-REP	MATα	PC1 (α)	ND
pbs2Δα	ura3[Δ360]pbs2Δ::REP-URA3-REP	MATα	PC1 (α)	ND
hog1Δα	ura3[Δ360]hog1Δ::REP-URA3-REP	MATα	PC1 (α)	ND
ssk2ΔREP	ura3[Δ360]ssk2Δ::REP	MATa	ssk2Δ	ND
pbs2ΔREP	ura3[Δ360]pbs2Δ::REP	MATa	pbs2Δ	ND
hog1ΔREP	ura3[Δ360]hog1Δ::REP	MATa	hog1Δ	ND
ssk2Δ + SSK2	ura3[Δ360]ssk2Δ::REP pVAX-URA3-SSK2	MATa	ssk2ΔREP	ND
pbs2Δ + PBS2	ura3[Δ360]pbs2Δ::REP pVAX-URA3-PBS2	MATa	pbs2ΔREP	ND
hog1Δ + HOG1	ura3[Δ360]hog1Δ::REP pVAX-URA3-HOG1	MATa	hog1ΔREP	ND

^aReference strain from Centraalbureau voor Schimmelcultures (Utrecht, The Netherlands).

^bDescribed in reference 5.

^cThe values are means ± standard deviations based on three individual replicates. ND, not determined.

We first investigated the putative involvement of Ssk2p, Pbs2p, and Hog1p in the integration and transduction of various environmental stress signals. For this purpose, we analyzed the growth, osmotolerance, oxidative stress response, drug sensitivity, and metalloid tolerance of the ssk2Δ, pbs2Δ, and hog1Δ mutants. All strains exhibited similar doubling times in liquid yeast extract-peptone-dextrose (YPD) medium (Table 1). Furthermore, no difference was observed in the growths (colony lengths and aspects) of the wild-type and mutant strains when the strains were cultured in solid YPD or yeast nitrogen base medium (data not shown). The growths of the ssk2Δ, pbs2Δ, and hog1Δ mutants are affected on high-osmolarity medium (1.5 M sorbitol, 1 M NaCl or KCl) (Fig. 1). Interestingly, we found that the ssk2Δ mutant is slightly less sensitive to these osmolytes than the pbs2Δ and hog1Δ mutants. However, these results indicate that C. lusitaniae Ssk2p, Pbs2p, and Hog1p could be involved in the capacity of adaptation of yeast cells to hyperosmotic conditions. All the strains tested (wild-type, mutant, and reintegrant strains) displayed similar phenotypes when grown under oxidant conditions (H₂O₂ or menadione). The three mutants exhibited hypersensitivity to UV irradiation (2,880 J m⁻²) and high temperature (43°C), implying that Ssk2p, Pbs2p, and Hog1p are also essential for responses to these signals. Moreover, the ssk2Δ, pbs2Δ, and hog1Δ mutants were sensitive to methylglyoxal (30 mM), suggesting that HOG pathway components are required to protect cells from this metabolic by-product (2, 19). C. lusitaniae ssk2Δ, pbs2Δ, and hog1Δ cells displayed similarly increased resistance to the cell wall biogenesis inhibitor Congo red relative to wild-type cells. However, we found these mutants slightly sensitive to calcofluor white (a compound that interferes with cell wall formation) and to the purine analogue caffeine. Finally, the ssk2Δ, pbs2Δ, and hog1Δ mutants showed impaired growth on cadmium- or arsenite-containing media, suggesting a role for Ssk2p, Pbs2p, and Hog1p in the regulation of heavy metal detoxification (Fig. 1).

FIG. 1.

Sensitivity of the wild-type strain; representative single mutants ssk2Δ, pbs2Δ, and hog1Δ; and the reintegrant ssk2Δ + SSK2, pbs2Δ + PBS2, and hog1Δ + HOG1 strains to different stresses. All strains were grown overnight in YPD liquid medium, the cells were counted, and then dilutions (10⁵ to 10² cells) were spotted onto a YPD plate supplemented or not supplemented (control) with 1.5 M sorbitol, 1 M NaCl or KCl, 8 mM H₂O₂, 100 μM menadione, 80 μg ml⁻¹ calcofluor white, 200 μg ml⁻¹ Congo red, 10 mM caffeine, 30 mM methylglyoxal, 4 μg ml⁻¹ fenpiclonil, 12 μg ml⁻¹ iprodione, 50 μM Cd(NO₃)₂, or 800 μM AsNaO₂. To test UV sensitivity, cells on YPD plates were exposed to UV for 12 s (2,880 J m⁻²). Each plate was further incubated for 1 day at 35°C and photographed. For high-temperature sensitivity, spotted cells were incubated for 1 day at 43°C and photographed. This experiment was done in triplicate, and pictures show representative plates for each condition. WT, wild type.

In C. lusitaniae, mutation of NIK1 or SSK1, encoding histidine kinase receptor or response regulator proteins, respectively, leads to severe dicarboximide and phenylpyrrole resistance (5, 29). We thus tested if the deletion of genes encoding downstream elements could be responsible for a comparable antifungal sensitivity pattern. C. lusitaniae ssk2Δ, pbs2Δ, and hog1Δ cells displayed strong fenpiclonil and iprodione resistance (Fig. 1). Nevertheless, neither hypersensitivity nor resistance toward the clinical antifungals amphotericin B, 5-fuorocytosine, and fluconazole was observed (not shown). Therefore, we conclude that the HOG pathway modulates filamentous-fungus-specific antifungals (dicarboximides and phenylpyrroles) but not clinical antifungal susceptibility in C. lusitaniae.

We then investigated if the deletion of the SSK2, PBS2, or HOG1 genes could have an impact upon the in vitro mating ability of C. lusitaniae. Indeed, morphological changes occur during the sexual reproduction of C. lusitaniae, notably during conjugation (9). Genetic crosses were performed under the same conditions as those described previously, using the reference mating type tester strains 6936 (MATa) and CL38 (MATα) (9, 23). No variation in the number of conjugation tubes and produced tetrads (echinulate ascospores) was detected for any unilateral (ssk2Δ with CL38, pbs2Δ with CL38, or hog1Δ with CL38) or bilateral (ssk2Δ with ssk2Δα, pbs2Δ with pbs2Δα, or hog1Δ with hog1Δα) genetic cross. Both groups of data indicate that Ssk2p, Pbs2p, and Hog1p do not participate in the mating process of C. lusitaniae.

C. lusitaniae can efficiently switch to pseudohyphal growth when cultured on solid medium depleted of a nitrogen source, such as yeast carbon base (YCB) (5, 11). Therefore, the capacity of mutants to differentiate pseudohyphae was investigated. For this purpose, approximately 10⁶ cells (in drops of 4 μl) of wild-type, ssk2Δ, pbs2Δ, and hog1Δ strains were spotted on various YCB solid media. Figure 2 shows representative pictures of pseudohyphae formation after 48 h of growth on YCB medium supplemented or not supplemented with a discriminatory concentration of each compound used in the drop plate assays described above. The concentrations indicated in Fig. 2 were used because higher and lower concentrations either were found too toxic or had no effect on the pseudohyphal development of the wild-type strain. The lengths of the pseudohyphae emerging from the edges of the colonies are reported in Table S2 in the supplemental material.

FIG. 2.

Pseudohyphal-growth ability of mutants. Amounts of 10⁶ cells (contained in drops of 4 μl) of the wild-type (WT) strain and representative mutants ssk2Δ, pbs2Δ, and hog1Δ were spotted onto YCB solid medium supplemented or not supplemented with a discriminatory concentration of sorbitol (0.8 M), NaCl (0.4 M), KCl (0.4 M), H₂O₂ (1 mM), menadione (60 μM), calcofluor white (50 μg ml⁻¹), Congo red (100 μg ml⁻¹), caffeine (1 mM), methylglyoxal (15 mM), iprodione (2 μg ml⁻¹), fenpiclonil (0.5 μg ml⁻¹), Cd(NO₃)₂ (25 μM), or AsNaO₂ (500 μM). Pictures were taken 48 h after spotting. This experiment was done in triplicate, and pictures show representative structures observed for each sample. The lengths of the pseudohyphae emerging from the edges of the colonies are reported in Table S2 in the supplemental material. Bar, 300 μm.

Homogeneously distributed pseudohyphae of equivalent lengths were obtained for all four strains on unsupplemented YCB medium (Fig. 2). More precisely, morphogenesis per se was not altered, since the numbers of lateral cells branching along the pseudohyphae do not significantly differ between the mutants and the wild-type strain. Addition of 0.8 M sorbitol produced a homogeneous reduction of pseudohyphae growth of the wild-type strain and partially inhibited pseudohyphal development of ssk2Δ. Similar results were observed in the presence of 0.4 M NaCl or KCl, in which ssk2Δ formed some sporadic bunch-like pseudohyphae (Fig. 2). The most important result of these experiments was that both mutants exhibited even stronger sensitivities to osmostress since the pseudohyphal growths of the pbs2Δ and hog1Δ strains were completely abolished on this high-osmolarity medium. These observations imply that Ssk2p, Pbs2p, and Hog1p likely play a role in osmotolerance during the pseudohyphal development of C. lusitaniae. A significant effect of methylglyoxal was also observed since the growths of the ssk2Δ, pbs2Δ, and hog1Δ mutants were restricted to bunch-like pseudohyphae, compared to the homogenous pseudohyphal formation of the wild type at the same concentration. As expected, although the switch of the wild-type strain was altered on YCB medium containing a discriminatory concentration of fenpiclonil or iprodione, the pseudohyphal developments of the ssk2Δ, pbs2Δ, and hog1Δ mutants were similar to those observed on drug-free YCB medium. The dicarboximide and phenylpyrrole resistance mediated by inactivation of HOG pathway components also occurred during the morphogenetic transition in C. lusitaniae. Finally, the effects of an oxidant (H₂O₂ or menadione), cell wall assembly inhibitors (Congo red and calcofluor white), caffeine, and metalloids on yeast cells, as detailed above, were also investigated on pseudohyphal formation. Strikingly, the mutants showed no significant reduction of pseudohyphal development in the presence of discriminatory concentrations of these compounds relative to the growth of the wild type (Fig. 2).

To summarize, the results obtained in this work strengthen the putative architecture of the C. lusitaniae two-component signaling pathway recently proposed (see reference 29; an update is provided in Fig. 3). Indeed, Ssk2p, Pbs2p, and Hog1p are clearly implicated in osmoadaptation, as reported for a large pallet of yeast and for filamentous-fungus models (3, 6, 12, 16, 26). In addition, the HOG MAPK components appear to be essential for UV and heat shock tolerance, as recently described for Cryptococcus neoformans (3). However, as suggested in a previous work, the Ssk1p-Ssk2p-Pb2p-Hog1p branch seems not to be required for oxidant adaptation in C. lusitaniae. In this way, the Sln1p-Ypd1p-Skn7p pathway likely takes charge of this function in C. lusitaniae (5, 29). Our data also provide evidence that the HOG pathway is involved in response to the heavy metals cadmium and arsenite, as recently reported for C. albicans and S. cerevisiae (6, 34). The present study (in addition to previous studies [5, 29]) supports the hypothesis that a Nik1p-Ypd1p-Ssk1p-Ssk2p-Pbs2p-Hog1p circuitry is probably crucial for dicarboximide and phenylpyrrole sensitivity (4, 16, 22).

FIG. 3.

Putative model for the action of the two-component system in C. lusitaniae. A possible mechanism that incorporates our experimental results from this study and our previous studies (5, 29) is proposed. The HOG MAPK cascade could operate downstream of the Sln1p HKR, the HPt protein Ypd1p, and the Ssk1p response regulator to regulate the osmotic stress response, pseudohyphal growth, UV adaptation, the heat shock response, and the heavy metal tolerance. In addition, the Nik1p-Ypd1p-Ssk1p-Ssk2p-Pbs2p-Hog1p-mediated transduction pathway probably regulates dicarboximide (dicarb.) and phenylpyrrole (phenylp.) sensitivity. Genes encoding Sho1p, Cdc42p, Ste11p, Ste50p, and Ste20p (dotted boxes) are located in the C. lusitaniae genome (unpublished data), but it is likely that the Sho1p branch does not regulate the HOG MAPK cascade (see Fig. S2 in the supplemental material). It is possible that the HOG pathway could integrate signal from a novel osmotic-stress signaling branch, as recently proposed for C. albicans (6, 28). Dotted arrows indicate mechanisms or interactions not yet fully elucidated.

Preliminary experiments showed that the C. lusitaniae sho1Δ mutant does not display any strong overlapping stress-sensitive phenotypes exhibited by pbs2Δ and hog1Δ cells (see Fig. S2 in the supplemental material). These results thus suggested that in C. lusitaniae, Sho1p branch could not interact with the Pbs2p-Hog1p cascade. In this way, it is possible that the HOG pathway could integrate a signal from a novel osmotic-stress signaling module, as recently proposed for C. albicans (6, 28). Future investigations will aim to characterize the putative ability of the Sln1p, Nik1p, and Chk1p histidine kinase receptors to regulate the HOG cascade in this emerging pathogen.