A novel co-regulation exists between the first step of GPI (glycosylphosphatidylinositol) anchor biosynthesis and the rate-determining step of ergosterol biosynthesis in Candida albicans. Depleting CaGpi19p, an accessory subunit of the enzyme complex that initiates GPI biosynthesis, down-regulates ERG11, altering ergosterol levels and drug response. This effect is specific to CaGpi19p depletion and is not due to cell wall defects or GPI deficiency. Additionally, down-regulation of ERG11 down-regulates CaGPI19 and GPI biosynthesis.
- Candida albicans
- glycosylphosphatidylinositol anchor biosynthesis (GPI anchor biosynthesis)
- mutual co-regulation
GPI (glycosylphosphatidylinositol) anchors are complex glycolipids which anchor proteins to the outer surface of cell membranes in eukaryotes and, in organisms which possess them, to the cell wall as well. They are important for pathogenicity and virulence of many pathogens, including Candida albicans. Approximately 35% of the putative GPI-anchored proteins of C. albicans appear to be important for maintenance of physiology and pathogenicity of this fungus .
GPI biosynthesis begins with transfer of GlcNAc (N-acetylglucosamine) from UDP-GlcNAc to phosphatidylinositol, catalysed by a multi-subunit GPI-GnT (GPI-N-acetylglucosaminyltransferase) complex with one catalytic and five or six accessory subunits . No function has been directly ascribed to the accessory subunits. However, GPI-GnT of mammals and Saccharomyces interact with other signalling and biosynthetic pathways [3–6], and the accessory proteins may participate in species-specific regulatory interactions.
We previously characterized CaGPI19, which encodes an accessory subunit of the GPI-GnT and reported a link between GPI biosynthesis and drug response in C. albicans . Heterozygous and conditional-null mutants of CaGPI19 were resistant to Amp B (amphotericin B); the conditional null had higher azole sensitivity also. This was coupled with gene dosage-dependent hyperfilamentation, suggesting that interactions of GPI-GnT with other pathways also exist in C. albicans.
The drug response phenotypes suggest specific interactions with ergosterol biosynthesis . We explore this further in the present study. We show that depletion of CaGpi19p down-regulates ERG11, lowering ergosterol levels. We also notice higher Ras signalling. These effects are specific to disruption of the first step of GPI biosynthesis and are not due to general GPI deficiency or cell wall defects. We also show that CaGPI19 and ERG11 are mutually co-regulated.
Strains and media
The heterozygote, Cagpi19/CaGPI19, and conditional-null, Cagpi19/MET3-GFP-CaGPI19, strains were made in BWP17 and grown in minimal medium containing methionine/cysteine to repress CaGPI19 expression . The chs2 mutant in SGY243 background was a gift from Professor N. Gow . We used CAI4, a Δura3 strain closely comparable with SGY243 , as a control for chs2. The CaSMP3 mutants in the CAI4 background were a gift from Professor P. Orlean . The nomenclature used for the strains in the present paper is as used by the respective groups.
Generation of erg11/ERG11 strain
One allele of ERG11 was knocked out in BWP17 by the method described in . The deletion cassette was constructed by amplifying the HIS1 marker flanked by 59-bp sequences from the upstream and downstream regions of ERG11 using primers ERG11-HD-FP and ERG11-HD-RP (Supplementary Table S1 at http://www.BiochemJ.org/bj/443/bj4430619add.htm) and used for transforming BWP17 cells.
Monitoring GPI anchor levels in erg11/ERG11 strain
The level of GPI-anchored proteins in the heterozygous ERG11 mutant was assessed using the pEcm331.GFP.Ecm331c construct, containing a GFP (green fluorescent protein)-tagged fragment of ECM33p with a GPI-anchor attachment signal sequence (a gift from Professor B. Wong ). The construct was used to transform BWP17 and erg11/ERG11, and fluorescence in cells was quantified using an Olympus IX71 microscope.
RNA preparation and real-time PCR detection
RNA was extracted as described in . RNA (3 μg) was used for generating cDNA. Transcripts were amplified by real-time PCR (Applied Biosystems) with primers ERG11-FP and ERG11-RP for ERG11, ERG1-FP and ERG1-RP for ERG1, and GAPDH-FP and GAPDH-RP for GAPDH (glyceraldehyde-3-phosphate dehydrogenase) (as an internal control) (Supplementary Table S1). Primers for CaGPI19 were as described in . Quantification of transcripts was achieved using SYBR Green PCR Master Mix (Applied Biosystems) and calculated relative to the control by the comparative CT method.
GC–MS analysis of total membrane sterols
Total membrane sterols were extracted using the method described in  with modifications (see the Supplementary Online Data at http://www.BiochemJ.org/bj/443/bj4430619add.htm). GC–MS was performed on a Shimadzu GC QP2010 Plus instrument using a 60 m wax column with an electrospray ionization source and helium carrier gas. A 2 μl sample was injected using a splitless injection at 280°C and a column flow of 1.21 ml/min. The oven temperature was set at 100°C, raised to 240°C at 20°C/min and a final temperature of 300°C at 3°C/min with hold for 10 min. Cholesterol (Sigma–Aldrich) was used as an internal standard; 50 μl of a 1 mg/ml stock was added to the sterol extract in petroleum ether before drying. Ergosterol and lanosterol (Sigma–Aldrich) (1 mg/ml in chloroform) were the external standards. Quantification was made after normalizing the area of the ergosterol peaks to that of the internal standard. Sterol content was normalized to cell mass and presented as relative to the wild-type which was set at 1.
R6G (rhodamine 6G) influx assay
The R6G influx assay was performed as described in  (see the Supplementary Online Data).
PKA (protein kinase A) signalling (cAMP/PKA activity)
To study Ras signalling, the activity of the cAMP-dependent downstream effector PKA was tested using the non-radioactive PKA activity kit (Promega) according to the manufacturer's instructions.
Cell wall integrity pathway activation by PKC (protein kinase C)
Cell lysates from 50 ml of mid-exponential-phase cultures were used. BWP17 cells treated with 100 μg/ml CFW (Calcofluor White) was used as a positive control. A 200 μg sample of total protein (50 μl) was used for dot blot assays (see the Supplementary Online Data). The blots were probed first with 1:1000 diluted phospho-p44/p42 MAPK (mitogen-activated protein kinase) [ERK1/2 (extracellular-signal-regulated kinase 1/2)] (Thr202/Tyr204) antibody (Cell Signaling Technology) and next with 1:15000 LI-COR IR-dye anti-rabbit secondary antibody-800, then scanned using the LI-COR Odyssey Infrared Imaging System. The fluorescence intensity, corresponding to the combined presence of phosphorylated Mkc1 and Cek1, was quantified by Alpha Imager Software.
Deletion of CaGPI19 leads to down-regulation of ergosterol biosynthesis
Assessment of the drug response of the heterozygote and conditional null for CaGPI19, revealed resistance to Amp B in both the heterozygote and conditional null and, in the conditional-null strain, increased sensitivity to azoles as well . Both drug classes target ergosterol directly or indirectly; Amp B binds ergosterol in the cell membrane and creates cytotoxic pores, whereas azoles target Erg11p, an enzyme in the sterol biosynthetic pathway . Hence we suspected alteration in membrane ergosterol levels in the mutants.
GC–MS analysis revealed ~14% less ergosterol in the conditional null compared with the wild-type, suggesting impaired ergosterol biosynthesis (Figure 1A). Since this mutant was azole-sensitive, we also investigated the expression status of ERG11. Real-time PCR revealed a ~50% decrease in ERG11 transcripts compared with BWP17 (Figure 1B). The heterozygote showed no significant variation in ergosterol content (Figure 1A), but had a ~23% decrease in ERG11 transcripts (Figure 1B). Our protocol for sterol extraction perhaps lacked the sensitivity to detect small differences in sterols between this mutant and the wild-type, but, since the mutant was not azole-sensitive , the differences in ergosterol levels, if any, are not expected to be large. GC–MS analysis also showed a 3-fold increase in lanosterol, the substrate of Erg11p, in the conditional null (Figure 1C), thus there appears to be a block in sterol biosynthesis at the level of ERG11.
To test whether down-regulation of sterol biosynthesis started at ERG11 or further upstream in the pathway, we looked at ERG1 transcript levels. If the down-regulation was at ERG1, then ERG11 would also be affected, as components of the pathway are transcriptionally co-regulated . Real-time PCR showed a ~60% increase in ERG1 transcripts in the conditional null relative to BWP17 and a ~25% increase in the heterozygote (Figure 1B). We also tested the response of the mutants to terbinafine, a drug specific for Erg1p . Higher terbinafine-resistance was observed (Figure 1D), suggesting higher Erg1p expression. A decrease in ergosterol levels may prompt a feedback up-regulation of upstream components of the pathway, in this case ERG1. Similar effects have been reported previously [8,14,16].
Down-regulation of ergosterol biosynthesis in CaGPI19 mutants therefore appears to be at the level of ERG11.
ERG11 down-regulation is not due to general cell wall defects or GPI deficiency
To test whether ERG11 down-regulation was a general consequence of cell wall defects, we studied the chs2 mutant, which is defective in chitin biosynthesis .
The chs2 strain was sensitive to azoles (Figure 2A). But ERG11 transcripts were significantly higher in the chs2 strain (Figure 2B) (as were ERG1 levels; L. Hauhnar and S. S. Komath, unpublished work), perhaps ensuring higher ergosterol levels to compensate for a defective cell wall . Since chs2 has cell wall defects, we probed whether the drug-sensitivity could be due to greater accumulation of cytotoxic drugs. Indeed, chs2 showed increased influx of the drug R6G (Figure 2C). Also, unlike CaGPI19 mutants , chs2 showed fewer hyphae (Table 1). Thus the phenotypes of CaGPI19 mutants do not arise from generic cell wall defects.
To test whether the CaGPI19 phenotypes are specifically linked to the first step of GPI biosynthesis, we studied the CaSMP3 mutants, which are affected in transfer of the fourth mannose to the GPI precursor, a late but essential step of the pathway .
CaSMP3 mutants were also sensitive to azoles (Figure 2D). But ERG11 transcripts were higher in the Casmp3/CaSMP3 heterozygote (Figure 2E). The conditional null under repressive conditions (in glucose) also exhibited a 2-fold increase in ERG11 when compared with growth under permissive conditions (in maltose). However, under permissive conditions, it showed an anomalous decrease in transcript levels as compared with CAI4 and Casmp3/CaSMP3 (Supplementary Figure S1 at http://www.BiochemJ.org/bj/443/bj4430619add.htm). Since ERG11 transcription is greater in glucose than in galactose , the anomaly may be due to different carbon sources in the growth medium. Given the significant increase in ERG11 transcripts for Casmp3/CaSMP3 compared with the wild-type, however, we conclude that the ERG11 is up-regulated in CaSMP3 mutants.
CaSMP3 mutants also have cell wall defects . Expectably, a greater amount of R6G enters the mutant cells compared with CAI4 (Figure 2F). Thus azole-sensitivity and ERG11 up-regulation appear to be a response to cell wall damage in CaSMP3 mutants. This was in contrast with the CaGPI19 mutants which did not show higher R6G influx compared with control (Supplementary Figure S2 at http://www.BiochemJ.org/bj/443/bj4430619add.htm). Thus ERG11 down-regulation in CaGPI19 mutants is not a response to GPI deficiency itself.
CaGPI19 mutants show increased cAMP/PKA signalling
We previously reported that CaGPI19 mutants are hyperfilamentous in serum and liquid Spider medium . Serum-dependent hyphal induction in C. albicans depends on Ras1p . In yeast, Ras is a core component of the cAMP/PKA signalling pathway  and in C. albicans, this pathway is involved in hyphal morphogenesis . To determine whether Ras1p signalling was affected in the mutants, we looked at the downstream cAMP/PKA signalling. The conditional null had higher cAMP-dependent PKA activity (77.5±7%) compared with the wild-type (30.67±2.7%) (Figures 3A and 3B), suggesting increased Ras1p signalling. No significant increase in PKA activity was seen in Cagpi19/CaGPI19 (35±7.1%), within the sensitivity of this experiment.
In both Saccharomyces cerevisiae and C. albicans, a reliable indicator of PKA activity is hyphal growth on fermentable sugars, such as glucose, which are substrates for the cAMP/PKA pathway, compared with growth on non-fermentable sugars (glycerol) [19,20]. Consistent with this, reversion of the hyperfilamentous phenotype of CaGPI19 mutants was seen on glycerol (Figure 3C). Furthermore, CaGPI19 mutants showed higher heat-shock-sensitivity (Figure 3D), typical of cells with increased Ras activity .
Since CaGPI19 mutants exhibit clumping, indicative of possible cell wall damage , we explored whether the MAPK-dependent PKC cell wall integrity signalling pathway may be up-regulated [20,21]. As shown in Figure 3(E), the specific antibody consistently picked up lower levels of the phosphorylated forms of MAPK in the cell lysates of CaGPI19 mutants. On the other hand, when cell wall damage was induced by CFW, it activated cell wall integrity signalling and higher levels of phosphorylated MAPK were observed. Thus hyperfilamentation in the CaGPI19 mutants is not due to cell wall integrity signalling.
Next, we tested whether lower ergosterol was a factor in the hyperfilamentation of CaGPI19 mutants . If so, we hypothesized that sequestering ergosterol from the membrane of wild-type cells would induce hyperfilamentation. BWP17 cells were spotted on to hyphae-inducing medium in the presence and absence of β-CD (β-cyclodextrin), a well-established sterol-sequestering agent [22,23]. Cells grown in β-CD showed significantly more filamentation (Figure 3F), and had lower ergosterol levels (0.66±0.01 relative to the membranes of untreated cells), suggesting an association between ergosterol depletion and filamentation.
CaGPI19 appears to be mutually regulated with ERG11
To confirm the link between ergosterol and filamentation, we generated a heterozygous mutant of ERG11, erg11/ERG11. We hypothesized that erg11/ERG11 would mimic the reduced ergosterol levels of the CaGPI19 mutants. Thus, if reduced ergosterol levels could trigger hyperfilamentation, we would see it here.
The erg11/ERG11 mutant had lower ERG11 transcripts (Figure 4A) and higher azole-sensitivity (Figure 4B), confirming the decrease in Erg11p levels. The tendency for hyphal growth of erg11/ERG11 was higher than for BWP17 (Figure 4C and Table 1). Most interestingly, however, the CaGPI19 transcript levels were lower in the mutant (~40% decrease) compared with BWP17 (Figure 4D). This was also reflected in lower levels of GPI-anchored GFP–Ecm33p in this mutant (~20% decrease; Figure 4E). Thus ERG11 and CaGPI19 appear to be mutually co-regulated.
Our results show clearly that CaGPI19 and ERG11 are mutually down-regulated in C. albicans, suggesting mutual co-regulation of the two pathways. That the co-regulation happens at the first step of GPI biosynthesis on the one hand and the rate-limiting step of sterol biosynthesis  on the other points to the fine balance that must exist between the two pathways in the cell. Both GPI and ergosterol biosynthesis occur on the endoplasmic reticulum. Direct interactions between components of the pathways may provide an opportunity for regulation. In Saccharomyces, a physical interaction between Erg11p and the GPI-GnT subunit, Gpi2p, was shown by yeast two-hybrid assays . It is possible that a similar physical interaction exists in C. albicans, allowing mutual regulation.
GPI anchoring and ergosterol have an intimate relationship; GPI-anchored proteins rely on association with lipid rafts to be transported to the plasma membrane  and are found to be localized in sterol- and sphingolipid-rich lipid rafts in many organisms, including mammals and yeast [26–28]. Association with GPI-anchored proteins is also necessary for raft association of some membrane proteins , thus there is a role for GPI-anchored proteins in the recruitment to microdomains and trafficking of other proteins, in addition to their own trafficking. In C. albicans, the polarization of lipid rafts in the membrane has been shown to contribute to germ tube formation and hyphal morphogenesis . This is significant as many GPI-anchored proteins are involved in virulence, notably the adhesins . These GPI-anchored proteins may be moved to sites of adhesion via lipid raft polarization along the hyphae. Other GPI-anchored proteins such as Ecm33p are known to be necessary for cell wall biogenesis and remodelling . Since hyphal invasion involves remodelling of the cell wall in order for the polarization of the germ tube, it is possible that regulation of ergosterol levels in the membrane and the resultant changes in membrane fluidity allow signalling to and transport of these proteins to sites of action.
Is there also a link between ergosterol depletion and Ras hyperactivation? There are two possible models. One is a linear connection between Ras activity and ergosterol levels; depletion of CaGpi19p down-regulates ERG11, which in turn affects membrane packing. As Ras is membrane-localized, changes in membrane dynamics due to ergosterol depletion, or loss of recognizable structural features of ergosterol, may contribute to higher Ras signalling. Complete loss of ergosterol and accumulation of methyl sterols results in abnormal or no hyphae . Thus, whereas complete loss of ergosterol may impair hyphal growth, smaller alterations in its levels may be a mechanism for regulating the morphological switch. An alternative model is that CaGpi19p independently affects both Erg11p and Ras1p. In S. cerevisiae, the homologous Ras2p interacts with GPI-GnT via the Eri1p subunit [5,6]. A similar interaction may exist in C. albicans as well.
These results add a new dimension to the complex biochemistry of sterol and GPI anchor biosynthesis in eukaryotes. They also suggest a new set of probable drug targets involving the accessory proteins of the GPI-GnT complex in the GPI biosynthetic pathway of C. albicans. Given that the accessory proteins are far less conserved than the catalytic subunits, they could provide a window of opportunity to combat infections by a pathogen notorious for developing multi-drug resistance, often via ERG11 mutations and/or up-regulation in the case of azoleresistance.
The study was conceived and designed by Sneha Komath, Guiliana Victoria and Bhawna Yadav. Experiments were carried out by Guiliana Victoria, Bhawna Yadav, Lalremruata Hauhnar, Priyanka Jain and Shilpi Bhatnagar. Data analysis and writing of the paper was by Sneha Komath, Guiliana Victoria and Bhawna Yadav with specific input from the others.
This work was supported by the University Grants Commission (UGC) of India [grant number 34-282\2008 (SR)] and the Department of Biotechnology (DBT) of India [grant number BT/PR10689/BRB/10/616/2008] (to S.S.K.). For fellowships, G.S.V. thanks the UGC, and L.H., B.Y. and P.J. thank the Council of Scientific and Industrial Research (CSIR) of India.
We thank Dr Aaron Mitchell for the BWP17 strain. GC–MS was carried out at Advanced Instrumentation Research Facility (AIRF), Jawaharlal Nehru University.
Abbreviations: Amp B, amphotericin B; β-CD, β-cyclodextrin; CFW, Calcofluor White; GFP, green fluorescent protein; GlcNAc, N-acetylglucosamine; GPI, glycosylphosphatidylinositol; GPI-GnT, GPI-N-acetylglucosaminyltransferase; MAPK, mitogen-activated protein kinase; PKA, protein kinase A; PKC, protein kinase C; R6G, rhodamine 6G
- © The Authors Journal compilation © 2012 Biochemical Society