Glucanase Induces Filamentation of the Fungal Pathogen Candida albicans

Candida albicans is the most common human fungal pathogen. Many organisms, including C. albicans, secrete glucanases under different environmental conditions. Here, we report a novel role for beta-1, 3- glucanase in inducing Candida albicans to form filaments at 22°C and enhancing filamentation at 37°C in nutrient-rich medium. Quorum sensing, the efg1-signaling and cek1 MAP kinase pathways are involved in this process. Our data suggest that the natural antifungal agent beta–glucanase may support morphologic transformation of Candida albicans at a wide range of ambient temperatures.


Introduction
Candida albicans is a fungus that can undergo multiple morphological transitions. C. albicans is known to infect a wide variety of hosts from warm-blooded animals to insects [1]. Hyphal formation is essential for virulence and invasion in a variety of hosts [2]. Many environmental stimuli can stimulate Candida albicans to form hyphae, including, but not limited to, serum, CO 2 , quorum sensing molecules, and pH [3][4][5][6]. However, most of these stimuli require incubation physiological temperature (37uC) for efficient hyphal growth. Although hyphal formation has been reported at lower temperatures, the conditions used were complex, requiring a matrix [7] or amino acid-containing liquid synthetic medium [8].
Many organisms have beta-1, 3-glucanase genes and secrete glucanases into the environment, including C. albicans [9,10]. C. albicans possesses three cell wall related exo-beta-1, 3-glucanases, Xog1, Exg2, and Spr1 [11]. CAMP65 was also identified as a putative beta-glucanase and is required for hyphal morphogenesis [12]. Although these hydrolytic enzymes are believed to play cell wall remodeling roles during growth and morphogenesis in filamentous fungi, they may play multiple physiological roles, which have not been fully elucidated in C. albicans. In this work, we show that beta-1, 3-glucanase, an enzyme produced by many bacteria, fungi and plants [13][14][15], permits C. albicans to overcome the temperature requirement for hyphal transformation in yeast peptone dextrose (YPD) medium.

Candida albicans strains and growth conditions
The wild type strain SC5314, and strains SN425 (reference stain), SGH284 (a biosensor reporter strain), HLC54 (egf1/cph1 double mutant), cph1 mutant, cek1 mutant, CJN2302 (efg1 mutant), cht2 mutant and CJN2318 (efg1 revertant) were grown in YPD medium and maintained on YPD agar plates (Table 1). These strains were described in detail elsewhere [2,[16][17][18]. To induce filamentation, the cells were grown to stationary phase overnight in YPD broth, and then inoculated in 5 ml glucanase-supplemented media (YPD, DMEM or RPMI) at 10 6 cells per ml and incubated at 22uC, 30uC or 37uC, for 18 h without shaking. For glucanase dose response experiments beta-1, 3-glucanase/lyticase, purified from two microbial sources (from Arthrobacter luteus or Trichoderma harzianum, Sigma), was added in YPD at concentrations ranging between 0.1-100 mg/ml and organisms were inoculated and incubated as described above. Glucanase inactivated by heating at 95uC for 10 minutes was used as control. To induce true hyphae, 10% FBS was used as positive control and organisms were grown at 37uC in an aerobic incubator with 5% CO 2 . Hyphal units were enumerated under phase contrast microscopy.

Fluorescence Microscopy
For staining the cell wall or nuclear material, the cells were collected by centrifugation, washed with PBS and stained with Calcofluor white (Sigma) or Hoechst 33258 (BIO-RAD), respectively, for 10 minutes. After washing in PBS the cells were mounted on slides and observed under a fluorescence microscope. To examine the effect of glucanase treatment on the cell wall glucan, cells were stained with a monoclonal antibody highly specific for (1R6) branched, (1R3)-b-D-glucans (BFDiv, Biothera), as we previously described [19]. To examine cell viability, a fluorescence LIVE-DEAD viability stain kit (Molecular Probes) was used to stain the cells, according to manufacturer's instructions. In some experiments farnesol or tyrosol were added to the media at 10-200 mM (Sigma) [20,21].

Quantitative Real-time PCR Assay
The assay was described in detail previously [22]. Briefly, 5 ml cell cultures were grown as described above and total RNA was extracted using the RiboPure yeast kit (Ambion, Inc.). RNA concentrations and quality were determined by measuring the absorbance at 260 nm and 280 nm (ND-1000 spectrophotometer, NanoDrop Technologies). Equal amounts of RNA (3 mg in 20 ml reactions) were reverse transcribed with oligo(dT) primers using Superscript reverse transcriptase II (Invitrogen). Primers used were as follows: EFB1, Forward: 59-CAT TGA TGG TAC TAC TGC CAC -39; Reverse: 59-TTT ACC GGC TGG CAA GTC TT -39.  The cells were grown in 5 ml YPD medium in BD 14 ml plastic tubes at 22uC without shaking for 18 h. Heat-inactivated glucanase was used as control at the same concentration. B: Mannosidase (100 mg/ml) did not induce hyphae under the same conditions as glucanase. Heat-inactivated mannosidase was used as control at the same concentration. C, D: Glucanase (100 mg/ml) induced all planktonic cells to grow as filaments in DMEM or RPMI media. Cells were incubated in respective media at 22uC without shaking for 18 h. Heat inactivated glucanase was used as control at the same concentration. E: Hypha formation was dependent on glucanase concentration in YPD medium. Cells were inoculated in YPD media containing 0.1-100 mg/ml of glucanase and incubated at 22uC without shaking for 18 h. F, G: Glucanase (100 mg/ml) enhanced filamentation at 30uC and 37uC in YPD medium. Cells were incubated in a water bath without shaking for 18 h. Filaments were longer than those at 22uC. doi:10.1371/journal.pone.0063736.g001 Table 1.

Strains
Genotype Reference

Results and Discussion
C. albicans standard laboratory strain SC5314 grows as a yeast form in YPD at 22uC [6,23]. In order to test the effect of glucanase under these conditions, we inoculated 10 6 cells per ml in 5 ml YPD medium containing 50 mg/ml b-1, 3-glucanase. After overnight growth at 22uC without shaking, we found that glucanase induces C. albicans to form long filaments (Fig. 1.A). C. albicans SC5314 does not form filaments in the presence of heatinactivated glucanase under these conditions (Fig. 1.A). Results with b-1, 3-glucanase from two different microbial sources were similar (data not shown). To confirm that this effect was specific to glucanase, we tested mannosidase under the same conditions and found that it did not induce filamentation (Fig. 1.B). In glucanasesupplemented DMEM or RPMI media without serum, filamentation was even more pronounced, with almost 100% of cells forming filaments (Fig. 1.C and D). Glucanase showed a dosedependent effect on filamentation with an active concentration as low as 0.5 mg/ml, whereas concentrations above 50 mg/ml did not increase the number of hyphal cells further (Fig. 1.E). Finally, to test whether filamentation in the presence of glucanase was temperature-dependent, we examined the effect of glucanase at 30uC and 37uC. Filamentation was increased by glucanase significantly at these temperatures, but in contrast to 22uC where C. albicans cannot form filaments in YPD without glucanase, a lower degree of filamentation is possible at these higher temperatures in YPD medium in the absence of glucanase ( Fig. 1.F and G).
We found that glucanase treatment does not reduce the immunoreactivity of the filaments to an anti-b-glucan antibody, arguing against the possibility of major hydrolytic dissolution of the b-glucan layer of the cell wall under these conditions (not shown). However, we noticed that glucanase treatment reduces binding of Calcofluor White (CFW), a chitin-specific dye, to the cell wall (Fig2. A). This suggested that glucanase treatment reduced the cell wall chitin content. We hypothesized that glucanase may upregulate chitinase gene expression, which may in turn lead to chitin degradation, and analyzed the expression of the CHT2 and CHT3 genes. These chitinase genes are repressed during hyphal morphogenesis as they play roles in cell separation [24,25]. In agreement with this, CHT3 gene transcription was downregulated during exposure to glucanase (Fig. 2.B). However, CHT2 gene transcription was significantly increased, suggesting that this enzyme may be responsible for the reduction in the chitin signal in CFW-stained filaments (Fig. 2.B). To confirm the role of CHT2 in this process we stained a cht2 mutant with CFW and found that the intensity of CFW binding on the glucanase-induced filaments was restored to levels similar to serum-induced hyphae in the mutant or the wild type background (Fig. 2.A). Collectively these data support the notion that glucanase reduces the chitin content in the cell wall of C. albicans by upregulating CHT2 gene expression.
CFW staining also showed that the sides of the cell walls in glucanase-induced filaments were parallel, resembling true hyphae. However, we noticed that the first septum was located at the mother-bud neck, whereas in 10% serum-induced hyphae, the first septum was located within hyphae (Fig. 2.A). The septal localization pattern was medium-independent (not shown) and suggests that in glucanase-induced hyphae, the nuclear division occurred across the mother-bud neck, as seen in yeast or pseudohyphae [4]. To examine nuclear division, we stained the DNA with Hoechst stain, a DNA-specific dye. As shown in Fig. 2.C, in 10% serum-induced hyphae, nuclear division took place throughout the filament, but in glucanase-induced hyphae, nuclear division was noted across the mother-bud neck, as is typical of yeast cells.
In order to test whether hyphal specific genes were expressed in glucanase-treated cultures, we used a dual approach. First the biosensor reporter strain SGH284 was used which has a hyphalspecific HWP1-RFP reporter, and a constitutive TDH3-GFP reporter that expresses GFP in both yeast and hyphal forms [26]. As shown in Fig. 2.D, the RFP signal could be observed on the surface of the glucanase-induced filaments but not on yeast cells, showing that the hypha-specific HWP1 gene was strongly expressed. To confirm this finding we analyzed HWP1 gene expression by qRT-PCR and included other genes that are strongly upregulated in hyphae-inducing conditions such as HYR1, ALS3 and UME6 [27][28][29]. Indeed, a strong upregulation of all four hyphal-associated genes was observed, with HWP1 gene expression level increasing about 2,000 fold, HYR1 about 150 fold, ALS3 about 19 fold, and UME6 about 25 fold, compared to inactivated glucanase controls (Fig. 2.E).
We then investigated the signaling pathway involved in glucanase-induced filament formation. Hyphal development in C. albicans is under transcriptional control of two principal signaling pathways, the MAP kinase pathway, predominantly represented by the transcription factor CPH1, and the cAMP pathway, represented by EFG1. The efg1/cph1 double mutant is unable to respond to most hyphae-inducing environmental stimuli; we thus first examined the effects of glucanase on this double mutant strain [2]. As expected the double mutant did not form hyphae in response to glucanase (Fig. 3.A). We next tested efg1 and cph1 single mutants for their response to glucanase in order to further dissect which of the two signaling pathways is involved. As shown in Fig. 3.C, the efg1 mutant did not form hyphae in contrast to the reference (Fig. 3.A) and complemented strains (not shown), suggesting that the effects of glucanase on filamentation are under the control of the cAMP signaling pathway. In contrast, the cph1 mutant formed hyphae in the presence of glucanase (Fig. 3.D), confirming that this transcription factor is not involved in the filamentation response. Cek1 is a MAP kinase that acts in concert with Efg1 in the adaptation response to defective cell wall glycosylation [30,31]. To examine the involvement of this MAP kinase in the glucanase-induced filamentation response, we tested the ability of a cek1 mutant to form filaments under these conditions. As seen in Fig. 3.E, the cek1 mutant did not form filaments in response to glucanase, indicating that Cek1is involved in glucanase-induced filamentation. These results also suggest that Cek1 is involved in phosphorylation of Efg1, or that there is crosstalk between the EFG1 and CEK1/MAP kinase signaling pathways during adaptation to glucanase exposure. In C. albicans, cell-density dependent quorum sensing molecules can affect the transition of yeast to hyphal forms and vice versa [20,21]. In order to test whether glucanase-induced filamentation is dependent on cell density, we inoculated different yeast cell concentrations and treated them with 100 mg/ml glucanase. We found that the effect of glucanase was dependent on the initial inoculum size, since filamentation was not observed when the inoculum exceeded 10 6 cells/ml (data not shown). We thus hypothesized that quorum sensing molecules can limit the effect of glucanase on filamentation, and tested the effect of farnesol, a known quorum sensing molecule that inhibits hyphal formation [21]. Farnesol inhibited glucanase-induced filamentation (Fig. 4A), although a few germ tubes were still present that appeared to be dead as seen with the Live-Dead dye (Fig. 4B). We also tested another quorum sensing molecule, tyrosol [20] that supports hyphal formation, but found no enhancement on the hyphaeinducing effect of glucanase (data not shown).
C. albicans is often found along with other microorganisms in a microbial community called a biofilm [19,32]. C. albicans and many other organisms possess glucanase genes and secrete glucanases into the environment that can digest glucan. Cell wall associated glucanases in C. albicans remodel cell wall during cell division and morphogenesis [12,33]. In the plant kingdom, numerous plants also secrete glucanases, which may act as antifungal agents that lyse the fungal cell wall and protect them from fungal invasion [34][35][36]. Here, we showed evidence that beta-1, 3-glucanases can induce filamentation of C. albicans, even at low temperatures and that this process is controlled by the efg1 and cek1 pathways. To our knowledge, this is the first report that beta-1, 3-glucanases from bacteria or fungi induce filamentation or any physiological response in C. albicans, and our findings suggest that this may be a protective adaptive response to these cell wall glucan-damaging enzymes. The glucanase-exposed yeast cells grow pseudohyphae first, but then continue to develop into true hyphae, as suggested by expression of hyphae-enriched genes that are associated with virulence in mammalian hosts. These observations also suggest that the presence of glucanases in polymicrobial biofilms may influence the virulence potential of C. albicans by increasing filamentation in a variety of host environments.