|Home | About | Journals | Submit | Contact Us | Français|
This study shows the importance of PDK1, TOR and PKC signaling pathways to the basal tolerance of Cryptococcus neoformans toward fluconazole, the widely used drug for treatment of cryptococcosis. Mutations in genes integral to these pathway resulted in hypersensitivity to the drug. Upon fluconazole treatment, Mpk1, the downstream target of PKC was phosphorylated and its phosphorylation required Pdk1. We show genetically that the PDK1 and TOR phosphorylation sites in Ypk1 as well as the kinase activity of Ypk1 are required for the fluconazole basal tolerance. The involvement of these pathways in fluconazole basal tolerance was associated with sphingolipid homeostasis. Deletion of PDK1, SIN1, or YPK1 but not MPK1 affected cell viability in the presence of sphingolipid biosynthesis inhibitors. Concurrently, pdk1Δ, sinΔ1, ypk1Δ, and mpk1Δ exhibited altered sphingolipid content and elevated fluconazole accumulation compared with the wild-type. The fluconazole hypersensitivity phenotype of these mutants, therefore, appears to be the result of malfunction of the influx/efflux systems due to modifications of membrane sphingolipid content. Interestingly, the reduced virulence of these strains in mice suggests that the cryptococcal PDK1, PKC, and likely the TOR pathways play an important role in managing stress exerted either by fluconazole or by the host environment.
Cryptococcus neoformans is the most common cause of fungal meningoencephalitis. The primary predisposing factor for cryptococcosis is a compromised immune system such as the case in HIV infected patients or with other underlying conditions. Cryptococcal meningoencephalitis is fatal unless treated and its mortality rate is high even with the most advanced treatment (Kwon-Chung & Bennett, 1992, Perfect & Casadevall, 2002). Fluconazole (FLC), a triazole antifungal drug, has been the agent most widely used for prophylactic therapy as well as for the long-term management of common mycoses such as candidiasis and cryptococcosis owing to its efficacy and safety (Zonios & Bennett, 2008). Triazoles target the P450 enzyme lanosterol 14α-demethylase, Erg11. The generally accepted mode of antifungal action of triazoles, based on the Saccharomyces cerevisiae model, is inhibition of ergosterol biosynthesis. It is a multi-mechanistic process that is initiated by the inhibition of two cytochrome P450 enzymes involved in the catalysis of lanosterol 14α-demethylation (Erg11) and Δ22 desaturation (Erg3) (Kelly et al., 1997). The inhibition of these enzymes results in the blockage of ergosterol biosynthesis, accumulation of toxic sterol intermediates, and the disruption of membrane integrity (Cowen & Steinbach, 2008, Anderson et al., 2003, Cowen, 2008).
Long-term maintenance therapy with azoles creates a condition favorable for the emergence of resistance to the drug, and increased azole resistance in vitro has been shown to be predictive of treatment failures and infection relapses in C. neoformans (Perfect & Cox, 1999). The molecular basis of azoles resistance has been extensively characterized in S. cerevisiae and pathogenic Candida species such as C. albicans and C. glabrata (Kontoyiannis et al., 1999, Lamping et al., 2007, White et al., 2002, Tsai et al., 2006, Sanglard & Odds, 2002, Helmerhorst et al., 2006, Brun et al., 2004, Bennett et al., 2004, Akins, 2005). Mechanisms known for the emergence of azole resistance in these fungi include: increased production of multidrug transporters (Lupetti et al., 2002, Cowen & Steinbach, 2008, Cowen et al., 2002), mutations in ergosterol biosynthetic pathway genes (Sanglard et al., 1998, Marichal et al., 1999), amplification of genomic regions that contain ergosterol biosynthetic pathway genes and transcription factors that positively regulate a subset of efflux pump genes (Selmecki et al., 2006, Selmecki et al., 2008), as well as activation of Hsp90 that may facilitate the cell-response to drug stress (Cowen et al., 2006, Cowen & Lindquist, 2005).
C. neoformans is phylogenetically distant from these well studied fungi and the mechanism of azole resistance in this organism is poorly understood. Unlike in Candida species, isolation of FLC resistant mutants have rarely been reported in C. neoformans and the emergence of resistance has most often been documented with clinical outcomes of AIDS patients receiving azole maintenance therapy (Armengou et al., 1996, Venkateswarlu et al., 1997, Paugam et al., 1994, Birley et al., 1995, Berg et al., 1998). An intriguing pattern of azole resistance called heteroresistance was reported in C. neoformans isolated from patients with recurrent episodes of infection (Sionov et al., 2009, Mondon et al., 1999). Further studies have shown that heteroresistance to azoles is intrinsic to all C. neoformans strains tested and heteroresistant subpopulations in each clone adapt to high concentrations of FLC by forming disomies of multiple chromosomes (Sionov et al., 2010, Sionov et al., 2009, Mondon et al., 1999). However, the molecular mechanism(s) involved in disomy formation remains an enigma. In order to expand our knowledge on the mechanism of azole tolerance which may provide insight into our understanding of clinically observed resistance mechanisms in C. neoformans as well as to find ways to improve therapeutic effect of azoles for cryptococcosis, we screened a mutant library and identified strains exhibiting FLC hypersensitivity. We found homologs representing components of the signaling cascade controlled by the mammalian phosphoinositide-dependent kinase (PDK1) to be crucial for responses to FLC. PDK1 is a serine/threonine kinase that controls a complex network of signaling cascades including responses to insulin and several growth factors, glucose uptake, regulation of apoptosis, translation initiation and others (for review see (Vanhaesebroeck & Alessi, 2000, Mora et al., 2004)). In addition, we found that the components of the protein kinase C (PKC) cell wall integrity signaling pathway as well as a component of the target of rapamycin (TOR) pathway were also involved in FLC response. The PKC signaling pathway is responsible for cell wall remodeling through the cell cycle and also functions in response to various stresses (Levin, 2005, Lesage & Bussey, 2006, Gerik et al., 2005). TOR was first identified in S. cerevisiae (Heitman et al., 1991) and is a highly conserved serine/threonine kinase that regulates cell growth in response to environmental changes such as nutrient availability or cellular energy status (for review see (Otsubo & Yamamato, 2008)). In the present study, we describe a regulatory module important for basal tolerance to FLC in C. neoformans that is also associated with sphingolipid homeostasis under the stress imposed by FLC.
We screened a library containing 1,201 deletion mutants of C. neoformans (Liu et al., 2008) for strains that were sensitive to fluconazole (FLC-s) and obtained a total of 32 FLC-s mutants. We chose 9 FLC-s mutants for further studies based on their degree of drug sensitivity and the availability of annotation for each mutated gene (Fig. S1 and Table S2). Several mechanisms may contribute to alterations in FLC sensitivity as in other organisms. For instance, mutations in the FLC target, Erg11, or other enzymes involved in the ergosterol biosynthesis pathway may alter the efficacy of FLC, change the cellular sterol content, or cause accumulation of toxic sterol intermediates upon FLC treatment. Alternatively, FLC-s mutants may alter influx or efflux systems resulting in increased cellular accumulation of FLC that maximizes the impact of the drug. Therefore, we examined whether these mutants had any defect in sterol biosynthesis and/or transport of the drug. There was no noticeable reduction in ergosterol content between wild-type strain H99 and the mutants (Fig. S2). In contrast, some of the FLC-s mutants, such as sin1, mpk1 and pdk1, accumulated greater amounts of 3H-FLC than the wild-type strain (Fig. 1A). Accumulation of FLC in the cells could have been due to a malfunction in the influx/efflux system. Since AFR1 is the only ATP-binding cassette transporter thus far known to be involved in the efflux of FLC in C. neoformans and its expression reportedly increases upon FLC treatment (Sionov et al., 2009, Posteraro et al., 2003), we measured the expression levels of AFR1 in these mutants. In the presence or absence of FLC, the expression levels of AFR1 were comparable between wild-type and the FLC-s strains (Fig. S3). Although it is not known whether the protein levels or genomic location of AFR1 has been altered in these mutants, we hypothesized that FLC influx/efflux system is likely altered in some of these FLC-s mutants by a mechanism(s) previously uncharacterized.
Although none of the nine FLC-s mutants showed significant defects in the mechanisms known to be involved in FLC hypersensitivity, three of the strains harbored a mutation in either the SIN1, MPK1, or PDK1 gene. These genes are homologs of components in the signaling cascades, TOR, MAPK, and PDK1 in mammalian and other eukaryotic systems. These genes control various cellular responses and their regulatory functions are known to be interrelated but are less known to be involved in azole susceptibility in pathogenic fungi. We, therefore, focused our attention on their role in the basal tolerance of FLC. Since MPK1 has been extensively studied in the H99 strain (Kojima et al., 2006, Kraus et al., 2003), we independently deleted and complemented the PDK1 and SIN1 genes in H99 to obtain the strains of interest in the same genetic background. Our sin1 and pdk1 deletion mutants in H99 displayed a similar phenotype as the original mutants obtained from the library and complementation of their mutations restored FLC accumulation and hypersensitivity to FLC to wild-type levels (Figs. 1B and and2A2A and data not shown). We noted that growth of our pdk1Δ strain on YPD medium was considerably weaker than the original pdk1Δ strain (Fig. 2A vs. Fig. S1). We have isolated several independent pdk1 deletion mutants and all of them displayed a similar phenotype (data not shown). Although poor growth of the pdk1Δ strain on YPD agar confounded the interpretation of FLC susceptibility test results, it was clear that the pdk1Δ strain was hypersensitive to FLC.
The C. neoformans PDK1 gene shares similarity with the KSG1 (kinase responsible for sporulation and growth 1) gene of Schizosaccharomyces pombe which is a homolog of the mammalian PDK1 gene (Graub et al., 2003, Niederberger & Schweingruber, 1999, Inagaki et al., 1999). A typical PDK1 kinase consists of a catalytic domain containing an ATP-binding site, an active site, and a substrate binding site. C. neoformans PDK1 homolog encodes a putative 1,206-aa protein whose catalytic domains (281-777aa, designated by SMART program, http://smart.embl-heidelberg.de/) are 35% identical to the human PDK1 kinase catalytic domains. However, the C. neoformans Pdk1 protein has an extra 228 amino acid (347-575aa) between the ATP binding site and the active site (data not shown). S. cerevisiae contains two functionally redundant PDK1 homologs, PKH1 and PKH2, and these two genes share an essential role in cell growth (Inagaki et al., 1999). The cryptococcal deletion mutant library contains a mutant annotated as PKH201 which shares some similarity to PKH2 of S. cerevisiae. However, Pkh201 does not contain the PDK1 catalytic domain and the pkh201 deletion mutant was not hypersensitive to FLC (Fig. 2C).
C. neoformans SIN1 shares 38% and 19.8% identity in their protein sequences with S. pombe Sin1 and S. cerevisiae Avo1, respectively. Sin1 of S. pombe was originally isolated as a Sty1 (or SAPK) interacting protein and identified as a component of the TORC2 along with Ste20 and Wat1 (Wilkinson et al., 1999, Hartmuth & Petersen, 2009, Matsuo et al., 2007). In C. neoformans, interaction of the TOR homolog with the Tor1 inhibitor rapamycin via FKBP12, a TOR binding protein, has been reported in the serotype D strain B-3501A (Cruz et al., 1999). Interestingly, the C. neoformans sin1Δ strain was not only sensitive to FLC but also was hypersensitive to rapamycin (Fig. 2A and D). In S. cerevisiae, the rapamycin/FKBP12 complex does not bind TORC2. It is thought that components of TORC2 complex of S. cerevisiae, such as Avo1 (the Sin1 homolog), bound to Tor2 may hinder the binding of rapamycin/FKBP12 rendering the TORC2 complex insensitive to rapamycin (Loewith et al., 2002). It is possible that C. neoformans Sin1 may interact with TOR complex and deletion of SIN1 may render the strain hypersensitive to rapamycin treatment.
Mpk1 is a homolog of Slt2/MAPK kinase and its role in maintenance of the cell wall has been characterized in C. neoformans (Kraus et al., 2003, Kojima et al., 2006). Although the sensitivity to FLC of C. neoformans mpk1Δ has not been demonstrated previously, the C. albicans Mkc1 (a homolog of Mpk1) is phosphorylated and activated upon exposure to FLC (LaFayette et al.). It has been shown that protein kinase C1 (Pkc1) signaling regulates the mitogen-activated protein kinase (MAPK) cascade comprised of a linear series of protein kinases including MAPKKK (Bck1/2), MAPKK (Mkk1/2), and MAPK (Slt2) that relay signals to the terminal transcription factors in S. cerevisiae (Zhao et al., 2007). C. neoformans mutants lacking components of the PKC pathway, Pkc1, Mkk2 and Bck1, also showed hypersensitivity to FLC (Fig. 2C) suggesting that PKC-MAPK cascade is required for FLC resistance.
It is noteworthy that C. neoformans PKC1 contains a serine/threonine protein kinase domain with high similarity to the known PDK1 target proteins (Vanhaesebroeck & Alessi, 2000). In addition to PKC1, two more genes, SCH9 and YPK1 in the H99 genome encode proteins with high similarity to the PDK1 target proteins. These putative PDK1 targets belong to the AGC (protein kinase A/ protein kinase G/ protein kinase C) kinase family which is highly conserved and known to play myriad roles in cellular growth, proliferation, and survival in other organisms (Pearce et al.). The alignment of putative PDK1 target protein sequences reveals that both PDK1 and PDK2 phosphorylation sites and sequences surrounding the sites are well conserved in all three proteins (data not shown). This raised the possibility that C. neoformans Pdk1 may also phosphorylate and activate these putative downstream target proteins (Casamayor et al., 1999). However, sch9Δ did not show any discernible sensitivity to FLC, suggesting that Sch9 is not a downstream target of the Pdk1 regulatory pathway associated with FLC response in C. neoformans (Fig. 2C). In contrast, deletion of YPK1 caused a substantial reduction in FLC resistance (Fig. 2C). In S. cerevisiae, strains lacking YPK1 were shown to be sensitive to fluconazole (Li et al., 2010). These results suggest that C. neoformans may respond to FLC by utilizing the PDK1 signaling pathway through the well conserved downstream targets, Pkc1 and Ypk1 (Fig. 8).
It has been shown that PDK1 homologs of S. cerevisiae, Pkh1 and Pkh2, function upstream of Pkc1 and activate the Pkc1 kinase and the Pkc1-effector MAPK pathway in response to various stresses (Inagaki et al., 1999). To determine whether C. neoformansresponds to FLC treatment through the PDK1-PKC-MAPK pathway by phosphorylating the terminal MAPK, we monitored the status of C. neoformans Mpk1 phosphorylation using the commercially available MAPK-phosphorylation-specific antibody. Mpk1 phosphorylation was detected in the wild-type strain exposed to FLC (Fig. 3, lane 2). However, Mpk1 phosphorylation was also detected in the wild-type strain treated with DMSO, a commonly used solvent for FLC (Fig. 3, lane 1). When the wild-type strain was treated with FLC solubilized in water, a strong Mpk1 phosphorylation signal was detected while only a faint signal was observed in the water-treated control sample (Fig. 3 lane 3 and 4). Similarly, a faint signal was observed in the water-treated samples derived from pdk1Δ, sin1Δ, and ypk1Δ (data not shown). Interestingly, Mpk1phosphorylation was diminished in pdk1Δ strain treated with FLC but unaffected in sin1Δ or ypk1Δ (Fig. 3 lane 5 and data not shown). These data suggest that Pdk1 is necessary for induction of Mpk1 phosphorylation and raises the possibility that Pdk1 may control the activation of PKC cell integrity pathway in C. neoformans. Unlike pdk1Δ and sin1Δ, however, mpk1Δ was not sensitive to 1M KCl and 1M NaCl (Fig. 2B), implying that Pdk1 uses downstream effectors other than the MAPK pathway in response to the stresses exerted by these compounds.
S. cerevisiae Ypk1 and its homolog in S. pombe, Gad8, can both be phosphorylated by upstream kinases. Since C. neoformans Ypk1 shares 55% and 69% identity with the Ypk1 and Gad8, respectively, it is possible that C. neoformans Ypk1 is also phosphorylated in response to FLC treatment and relays the signal to downstream targets/effectors. To test whether C. neoformans Ypk1 is phosphorylated upon FLC treatment, we tagged Ypk1 with the FLAG epitope at its C-terminus. The resulting construct was transformed into a H99 ypk1 deletion mutant to replace the deleted allele. Such transformants restored the FLC sensitivity of ypk1Δ to wild-type levels indicating that the Ypk1-FLAG fusion protein was functional (Fig. 4B). The expression of the Ypk1-FLAG protein was confirmed by western blot analysis with anti-FLAG antibody as the protein migrated with the expected molecular mass (64 Kda) (Fig. 4A). To test the phosphorylation status of Ypk1, the strain expressing Ypk1-FLAG was treated with 16 μg/ml FLC for various time periods. The total protein lysate was immunoprecipitated with anti-FLAG M2 magnetic beads and resolved in SDS-PAGE. In contrast to a series of slow migrating bands diagnostic of Ypk1 phosphorylation status reported in other organisms, C. neoformans Ypk1 migrated as a single species without any mobility shift regardless of FLC treatment (Fig. 4A). Although we used several different extraction buffers and methods, we failed to detect the slow migrating bands of Ypk1 even when the cells were treated with FLC for up to 6 hr (data not shown). In addition, since antibody specific for C. neoformans Ypk1 was not available, we used anti-phospho-serine and anti-phospho-threonine antibodies for immunoblotting. However, both antibodies failed to detect any phosphorylation of the Ypk1-FLAG fusion protein (data not shown).
To explore the inability to detect Ypk1 phosphorylation in C. neoformans, we attempted to show the functional importance of the presumptive phosphorylation sites in Ypk1. C. neoformans Ypk1 contains putative consensus sites, T402 and S562, for Pdk1 and Pdk2 phosphorylation, respectively. We constructed Ypk1-FLAG fusion constructs containinga T402A or S562A mutation and used the constructs to replace the ypk1Δ allele. In transformants in which the ypk1Δ allele was replaced by a Ypk1(S562A)-FLAG construct, FLC resistance was recovered to levels comparable to the wild-type (Fig. 4B). This indicates that the conserved putative Pdk2 phosphorylation site, S562, is dispensable for Ypk1 function associated with FLC resistance. In contrast, transformants of Ypk1(T402A)-FLAG remained as sensitive to FLC as the ypk1Δ deletion mutant (Fig. 4B). The levels of Ypk1 protein, however, were lower in the T402A derivative compared to the control but were comparable to the S562A derivative (Fig. S4). These data indicate that the putative Pdk1 phosphorylation site, T402 residue, play a role in the stability and function of Ypk1.
To test whether C. neoformans Ypk1 functions as a protein kinase and the ATP-binding domain is essential for its function in response to FLC, we mutated two residues of Ypk1, D386 and K274, which reside equivalently to D488 and K376 in S. cerevisiae Ypk1. These residues are predicted to be directly involved in anchoring of the γ-phosphate group of ATP (Knighton et al., 1992). Substitution of these residues with any other amino acid abolishes the ability of Ypk1 to transfer phosphate to its target amino acids, resulting in a kinase-dead version of Ypk1 (Roelants et al., 2010, Roelants et al., 2002). Interestingly, both Ypk1 (D386A)-FLAG and Ypk1(K274A)-FLAG constructs failed to complement the FLC-s phenotype of the ypk1Δ strain (Fig. 4B), suggesting that Ypk1 kinase activity is necessary for its function in C. neoformans.
In S. cerevisiae, ypk1Δ is hypersensitive to the TORC1 inhibitor rapamycin (Gelperin et al., 2002). We found that C. neoformans ypk1Δ and pdk1Δ are sensitive to rapamycin (Fig 2D), which implies a regulatory link between TORC1 and Ypk1.We examined, therefore, whether the conserved TOR phosphorylation site in Ypk1 is important for response to FLC treatment in C. neoformans. The S543 residue of C. neoformans Ypk1 (corresponding to S527 of S. pombe Gad8) in the presumptive turn motif was substituted with alanine in the Ypk1(S543A)-FLAG construct and was expressed in the ypk1Δ deletion mutant. Interestingly, Ypk1(S543A)-FLAG failed to confer FLC resistance (Fig. 4B). These results imply that turn motif phosphorylation at S543 is important for Ypk1 function and support the possibility that C. neoformans TORC1 might be involved in Ypk1 regulation in response to FLC treatment.
To study the importance of TOR pathway in fluconazole resistance, we attempted to delete the TOR homolog in C. neoformans. We failed repeatedly to delete TOR1 (CNAG_06642), suggesting deletion of TOR1 might be lethal (data not shown). The essentiality of TOR1 has been suggested in a serotype D strain of C. neoformans TOR-like (Davidson et al., 2002). There is another gene, CNAG_05220 designated as TLK1 for kinase, which shares 57% identity with Tor1 of S. cerevisiae but lacks the conserved FKBP-rapamycin binding domain. However, deletion of TLK1, did not affect the sensitivity of the strain to FLC (Fig. 2C).
In fission yeast, Tor1 forms a complex with conserved proteins, Ste20, Wat1 and Sin1. It is possible that C. neoformans Tor1 may form a complex with the Sin1 homolog along with other TOR complex components and relay the incoming signal to Ypk1. To examine whether Sin1 and Ypk1 function in the same pathway, we transformed the Ypk1-FLAG construct in sin1Δ. When the Ypk1-FLAG protein levels in the sin1Δ transformant (sin1Δ+YPK1A) were close to the control strain (ypk1Δ+YPK1), the sin1Δ transformant remained sensitive to FLC and rapamycin (Fig. 5A and B). However, the sensitivity to FLC and rapamycin was complemented in the sin1Δ transformant that expressed higher amounts of Ypk1-FLAG (sin1Δ+YPK1B). This data suggests that C. neoformans SIN1 is upstream of YPK1 in FLC response. In contrast, over-expression of Ypk1 in a pdk1Δ deletion mutant was not sufficient to confer FLC resistance (Fig. 5), suggesting that Ypk1 is not the only downstream effector of Pdk1 or alternatively, phosphorylation at the Pdk1 site of Ypk1 by Pdk1 is absolutely essential for Ypk1 activity in the cryptococcal response to FLC.
The importance of the interaction between membrane ergosterol and sphingolipids in determining drug susceptibility has been reported in C. albicans (Mukhopadhyay et al., 2004). Interestingly, homologs of Pdk1 and its downstream kinases are involved in sphingolipid-mediated signaling in other organisms (Sun et al., 2000, Roelants et al., 2010, Luo et al., 2008). Furthermore, Ypk2 has been shown to regulate ceramide biosynthesis in S. cerevisiae (Aronova et al., 2008). Since the sterol profile of the mutants did not show noticeable changes in ergosterol content, we investigated whetherhypersensitivity to FLC is related to sphingolipid homeostasis and/or sphingolipid-mediated signaling. We used myriocin (MYR), aureobasidin A (ABA) and phytosphingosine (PHS), to examine the importance of the aforementioned genes in sphingolipid biosynthesis. It is known that myriocin blocks the production of endogenous phytosphingosine, aureobasidin A blocks production of the complex inositol-containing sphingolipids, and addition of exogenous phytosphingosine elevates intracellular phytosphingosine causing toxicity. Interestingly, growth of the three FLC-s mutants, pdk1Δ, sin1Δ, and ypk1Δ, but not mpk1Δ, was hampered by the inhibitors at the concentrations which do not affect growth of the wild-type strain (Fig. 6A). These observations suggest that deletion of either PDK1, SIN1 or YPK1 but not MPK1, is deleterious to cell viability when sphingolipid biosynthesis is compromised by these inhibitors.
To examine the possibility whether PDK1, SIN1, and YPK1 exert FLC basal tolerance in C. neoformans through modulation of sphingolipid biosynthesis, we used 3H-myoinositol and thin layer chromatography to analyze the complex sphingolipids. Our experiments were carried out in the presence or absence of FLC to see the effect of each mutation as well as FLC exposure on sphingolipid biosynthesis. When the relative amounts of complex sphingolipids were compared within each strain between the FLC treated and non-treated cells, no obvious difference was observed (Fig. 6B). These data suggest that FLC treatment does not affect the accumulation of complex sphingolipids in all three deletion mutant strains as well as in the wild-type strain. However, the relative amounts of the complex sphingolipids IPC, MIPC, and MIP2C+ (a combination of M(IP)2C and other complex sphingolipids; see materials and methods) were significantly reduced in pdk1Δ and ypk1Δ strains compared with those in the wild-type strain (Fig. 6B and C). In sin1Δ, MIPC was the only complex sphingolipid that showed a slight reduction compared to the wild-type strain. In mpk1Δ, there was an increase of IPC compared to the wild-type (p = 0.06 and p = 0.05 in the absence and presence of FLC, respectively). These results demonstrate that PDK1, SIN1, YPK1, and MPK1 play a role in maintaining complex sphingolipid homeostasis.
Since the deletion mutants showed increased sensitivity to various stresses, we investigated the impact of the strains’ diminished ability to handle environmental stresses on their virulence. The growth rate at 37°C for ypk1Δ and sin1Δ was slightly slower than the wild-type (3.4 ± 0.2h and 3.9 ± 0.3h vs. 2.5 ± 0.3h, respectively) and pdk1Δ grew significantly slower compared with the wild-type (7.8 ± 2.1h vs. 2.5 ± 0.3h). Expression of other major virulence factors such as capsule and melanin in the deletion mutants was comparable to wild-type (data not shown). Groups of 10 mice were challenged with different yeast strains via tail vein injection. pdk1Δ, sin1Δ, and ypk1Δ strains all showed reduced virulence (p<0.001 compared to wild-type H99) (Fig. 7). Furthermore, complementation of these mutations increased the virulence compared to the original deletion mutant (p<0.001). It was not surprising that virulence of pdk1Δ is reduced since it grows much more slowly at 37°C than the wild-type strain. Significantly, ypk1Δ and sin1Δ strains showed only a slight reduction in the growth rate at 37°C and yet clearly demonstrated a reduction in virulence. Apparently, SIN1 and YPK1 are required by C. neoformans to handle the stress encountered in the host environment.
Mechanisms of azole resistance and the role of cellular signaling in fungal drug resistance have been extensively studied in Candida species (for review see (Cowen, 2008, Shapiro et al., 2011)). However, the role of well conserved kinases in FLC basal tolerance has not been well documented. In the course of dissecting the mechanisms involved in the ability of C. neoformans to manage FLC stress, we found that PDK1 and PKC-MAPK signaling pathways contribute to FLC basal tolerance. Both pathways, however, showed differential involvement in response to salt and rapamycin treatment.
Azole drugs are known to block the biosynthesis of ergosterol and result in the accumulation of a toxic sterol intermediate that disrupts membrane integrity, culminating in severe membrane stress. It has also been shown that fungal azole susceptibility is related to sphingolipid metabolism. For example, deletion of genes involved in sphingolipid biosynthetic pathways such as SLD1 encoding sphingolipid Δ8-desaturase, HSX11 encoding glucosylceramide synthase, and IPT1 encoding inositol phosphoryl transferase, results in hypersensitivity to FLC in several species (Pasrija et al., 2005, Prasad et al., 2005, Oura & Kajiwara, 2008, Oura & Kajiwara, 2010). In addition, an antimicrobial peptide inhibitor of the fungal plasma membrane ATPase (Pma1), which is associated with lipid rafts is known to block the azole resistance in C. albicans (Monk et al., 2005). Moreover, acquisition of the multidrug resistance phenotype in C. albicans is accompanied by upregulation of genes required for normal lipid metabolism that constitute membrane rafts (Pasrija et al., 2005, Mukhopadhyay et al., 2004, Mukhopadhyay et al., 2002). Consistent with the notion that ergosterol and sphingolipids are tightly coupled for membrane integrity and drug susceptibility, we show that FLC-s mutants, pdk1Δ, sin1Δ, and ypk1Δ are hypersensitive to chemicals that inhibit sphingolipid synthesis and exhibit reduced content of complex sphingolipids. We speculate that the FLC-s phenotype of our mutants can be partially attributed to impaired membrane rafts due to modified sphingolipid content.
Membrane rafts modulate signaling events via compartmentalization of signaling molecules through specific affinity between sphingolipids, sterols, and membrane proteins and/or signaling molecules (Moffett et al., 2000, Wu et al., 2004, Martin & Konopka, 2004). A stable interaction between ergosterol and sphingolipid is essential to maintain the stability of sphingolipid-rich microdomains (Mukhopadhyay et al., 2004). Since the sphingolipid contents are altered in our mutants, integrity of the membrane rafts may have been disturbed. As a consequence, FLC causing a blockage in ergosterol biosynthesis may become more deleterious in these mutants. Little is known about how azole drugs mechanically influence membrane properties and fungal viability. Abe et al. showed that treatment with fluconazole decreased the rigidity of the plasma membrane in S. cerevisiae (Abe et al., 2009). It has been postulated that lipid and sterol compositions in the plasma membrane play a role in azole accumulation (Luo et al., 2008, Kohli et al., 2002). Similarly, we have shown that our mutants accumulate more FLC than the wild-type strain. In light of the recent study showing that the import of FLC proceeds via facilitated diffusion (Mansfield et al.), it is possible that modifications in membrane composition and/or flexibility could affect facilitated diffusion and maximize the impact of the drug simply by increased accumulation of FLC inside the cells.
The components of the PKC-MAPK signaling pathway are known to be vital for maintaining integrity of the cell, allowing full resistance to both oxidative and nitrosative stresses in C. neoformans (Kraus et al., 2003, Gerik et al., 2008). Recently, PKC signaling was shown to regulate FLC resistance in C. albicans via a circuitry comprised of the Mpk1 homolog, Mkc1 (LaFayette et al., 2010). Deletion of SLT2/MPK1 in S. cerevisiae also leads to fluconazole hypersensitivity (Parsons et al., 2004). We show that deletion mutants of each component in the pathways, PKC1, MKK2, BCK1 and MPK1, are hypersensitive to FLC suggesting that the PKC-MAPK cascade is involved in FLC susceptibility in C. neoformans. Although the sphingolipid biosynthetic pathway has been linked to Pkc1 in C. neoformans (for review see (Shea & Del Poeta, 2006)), it is also possible that a general cell wall deficiency rather than changes in sphingolipid composition allows the drug to enter the cell more readily and cause FLC hypersensitivity.
While Mpk1 activation is thought to be initiated by Pkc1 and carried out in a linear mode by a series of upstream kinases in the PKC pathway, the S. cerevisiae PDK1 homologs, Pkh1/Pkh2, function upstream of Pkc1 activating the kinase and the Pkc1 effector, the MAPK pathway (Inagaki et al., 1999). In fact, deletion of PDK1 diminishes the Mpk1 phosphorylation suggesting the possibility that Pdk1 may function upstream of the PKC-MAPK signaling pathway in response to FLC treatment. However, we cannot rule out the possibility that Pdk1 directly phosphorylates Mpk1.
Mammalian PDK1 and its functional homologs have been shown to phosphorylate and activate a number of AGC protein kinases in vitro, which include p70S6k, PKB, SGK (serum/glucocorticoid-inducible kinase) and PKC isoforms (Vanhaesebroeck & Alessi, 2000). All these targets contain the conserved phosphorylation sequence in their activation loop, which is termed PDK1 site (Thr-Phe-Cys-Gly-Thr-X-Glu-Tyr, where the bold Thr represents the phosphorylated residue and X represents any amino acid). In addition, full activation of some targets also requires the phosphorylation of a second site, PDK2 (phe-X-X-Ar-Ser/Thr-Ar, where the residue in bold is the phosphorylated amino acid and Ar represents an aromatic residue). Although phosphorylation of Ypk1 is yet to be confirmed biochemically in C. neoformans, our results are clear that mutation of the putative Pdk1 phosphorylation site in Ypk1 and kinase-dead version of Ypk1 renders the FLC-s phenotype. This strongly suggests that Ypk1 is phosphorylated by Pdk1 and Ypk1 functions as a kinase to phosphorylate the target(s) yet to be identified in C. neoformans.
Although we failed to delete the putative TOR homolog and directly test its functional relationship with FLC tolerance, several compelling observations led us to hypothesize that the TOR pathway might also be involved in C. neoformans basal tolerance toward FLC. First, deletion of the SIN1 homolog, a known component of the TOR complex in other organisms, is hypersensitive to FLC as well as to rapamycin. Second, over-expression of Ypk1 rescues the FLC-s phenotype of sin1Δ. Third, the conserved TOR phosphorylation site in Ypk1 is important in C. neoformans’ response to FLC treatment. In S. pombe, the Ypk1 homolog, Gad8, has been isolated as a multicopy suppressor of tor1-g2 sterile mutant and is regulated by Tor1 as well as by Pdk1 (Matsuo et al., 2003). In addition, S. cerevisiae TORC2 is proposed to activate Slm1 and Slm2, phosphoinositide effectors, which then downregulate the turnover of complex sphingolipids (Tabuchi et al., 2006, Beeler et al., 1998). Given that deletion of SIN1 or YPK1 reduces the accumulation of complex sphingolipids, it is plausible that Tor1, along with Sin1 and Ypk1, is involved in maintaining homeostasis of complex sphingolipids in C. neoformans. It is known that TORC2 also regulates the Rho1/PKC cell wall integrity pathway in S. cerevisiae (Torres et al., 2002). However, since the presence of rapamycin did not alter the FLC basal tolerance in the wild-type strain H99 (data not shown) and mpk1Δ was not hypersensitive to rapamycin (Fig. 2D), it is not clear if C. neoformans also channels the signal through the TORC to PKC pathway in response to FLC treatment. Figure 8 summarizes our findings on the possible involvement of PDK1, TOR and PKC-MAPK signaling pathway in C. neoformans FLC basal tolerance.
It is interesting to observe that deletion of PDK1, SIN1, and YPK1 results in a significant reduction of virulence. In addition, C. neoformans mutants lacking Mpk1 also exhibit attenuated virulence in the mouse model of cryptococcosis (Kraus et al., 2003). Since these genes are conserved in many fungi, the importance of these genes in virulence along with FLC basal tolerance may be exploited for future drug designs to treat fungal infections. Furthermore, understanding of the FLC basal tolerance and in depth characterization of these pathways will further shed light on how C. neoformans handles stress. Our study provides a basis to further our understanding on clinically observed drug resistance in C. neoformans and may contribute towards improved designs for a new generation antifungal azoles.
This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institute of Allergy and Infectious Diseases/National Institutes of Health. The protocol was approved by the Committee on the Ethics of Animal Experiments of National Institute of Allergy and Infectious Diseases (Permit Number: LCID50E).
C. neoformans serotype A genome sequenced strain H99 was used as the wild-type. List of strains relevant to this study is in the Table S1 and S2. Yeast extract-peptone-dextrose (YPD) and Yeast Nitrogen Base (YNB) were as described previously (Chang & Kwon-Chung, 1994). Rapamycin (Sigma R0395) is dissolved in ethanol to make 1mg/ml stock solution and added to YPD medium to 1 ng/ml. Phytosphingosine hydrochloride (PHS) (Sigma P2795) was dissolved in 95% ethanol to make 20 mM stock solution. Aureobasidin A (Clonetech. Cat#630466) and Myriocin (Sigma M1177) were dissolved in methanol to make 1 mM stock solution. FLC was either dissolved in DMSO to prepare 50 mg/ml stock or in water to make 10 mg/ml stock.
We screened the signature tag mutagenesis (STM) deletion collection of the C. neoformans CM018 (a serotype A strain derived from H99)(www.atcc.org) (Liu et al., 2008) to identify mutants sensitive to fluconazole. The STM strains were first replica-spotted to YPD agar media containing 16 μg/ml FLC which is the maximum concentration of FLC that allows the growth of H99 cells. Mutants that failed to grow were selected and re-tested on media containing different concentration of FLC (8, 16, and 32 μg/ml) to verify their levels of FLC sensitivity. A total of 32 FLC-sensitive (FLC-s) mutants were isolated from initial screening of 1,201 deletion mutant strains. The afr1Δ, sre1Δ, and scp1Δ strains were found among the 32 FLC-s mutants and previous reports showed that deletion of each of these genes results in FLC-sensitivity (Sionov et al., 2009, Chang et al., 2007). Isolation of known FLC-s mutants verified this screening approach to be suitable for identification of the genes important for the azole resistance in C. neoformans.
Identification of C. neoformans homologs was performed by Blast search of the H99 genomes at Broad Institute (http://www.broad.mit.edu/annotation/fungi/cryptococcus_neoformans/index.html). In order to confirm the phenotypes of the STM deletion mutants, we reconstructed the selected deletion strains in H99 and retested the phenotype. All genes were deleted by biolistic transformation with PCR fusion using a strategy similar to that described for Clostridium difficile (Kuwayama et al., 2002). Briefly, the left end and the right end of the locus were amplified with the primers listed in Table S3. The linear disruption cassette was then used to homologously integrate into the strains by biolistic transformation (Toffaletti et al., 1993). Transformants were screened by colony PCR and confirmed by Southern blot hybridization. The wild-type genes YPK1, SIN1, and PDK1 were PCR-amplified from H99 using the primers listed in Table S3. The PCR clones were sequenced, cloned into a vector containing the norseothricin (NAT) resistance marker and transformed into deletion mutants by the biolistic method. PCR was used to identify integrative transformants containing the intact wild-type gene and Southern blot analysis was used to confirm the results.
Sterol analysis was carried out as described before (Sionov et al., 2009). Briefly, the log phase YPD grown cells were harvested by centrifugation and washed once with sterile distilled water. The pellets were resuspended in 9 ml methanol; 4.5 ml 60% (wt/vol) KOH was added together with 5 μg cholesterol (used as an internal recovery standard). Cell suspensions were heated to 75°C in a water bath for 2 h to complete the saponification; the sterols were then extracted with hexane and analyzed by gas chromatography (GC) with an Agilent 6850 gas chromatograph with an HP-1 fused silica column and a flame ionization detector at a temperature programmed from 200°C (1 min) to 300°C at a rate of 10°C/min. For identification of each sterol, a GC-MS (gas chromatograph-mass spectrometer) instrument (Polaris Q, from Thermo Electron, coupled to a Focus GC) was used to obtain mass spectral data in the EI mode for each of the GC peaks. This GC-MS instrument used a Restek 5MS fused silica column (30 m length, 0.25 mm i.d., 25 μm film thickness) at a program temperature from 200°C (1 min) to 300°C at a rate of 10°C/min. Cholesterol and ergosterol were identified by comparisons with standards. The relative amount of ergosterol was determined by comparing the area under the ergosterol peak in the GC using cholesterol as internal standard.
Isolation and analysis of genomic DNA was carried out as described previously (Chang & Kwon-Chung, 1994). For gene expression analysis, overnight cultures of wild-type and each deletion mutant strain were refreshed to OD600= 0.2, grown in fresh YPD media for 3 hrs, treated with FLC at the final concentration of 16 μg/ml for 30 min or 2 hr. RNA was extracted from yeast cells using Trizol (Invitrogen, Carlsbad, CA), treated with RNAse-free Dnase (Ambion, Austin, TX) for the removal of genomic DNA, and purified with Rneasy MinElute cleanup kit (Qiagen, Valencia, CA). cDNA was synthesized using high-capacity cDNA archive kit (Applied Biosystems, Foster City, CA) and used in real time reverse transcription PCR(RT-PCR) with TaqMan universal PCR master mix and the ABI PRISM 7700 sequence detection system (Applied Biosystems, Foster City, CA). The primers used in RT-PCR are listed in Table S3. Data were normalized with GPD1 level and expressed as the amount in each deletion mutant strain relative to that in H99.
Cells were grown in YPD medium under the indicated conditions. Cultures were grown overnight in YPD at 30 °C with shaking and then diluted to an optical density (OD600) of 0.4 in 20 ml of pre-warmed, fresh YPD medium. The cells were grown with shaking for an additional 2h and then treated as indicated. The cells were harvested, resuspended in ice-cold stop buffer (0.9% NaCl, 1 mM NaN3, 10 mM Na-EDTA, 50 mM NaF) and then washed once in ice-cold stop buffer. The cells were resuspended in 0.2 ml lysis buffer (50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 05 mM Na-EDTA, 5 mM Na-EGTA, 0.2 mM Na3VO4, 50 mM KF, 30 m M Na4O7P2.H2O, 50 mM NaF, 1X protease inhibitors (Roche, no. 11836170001), and 10 μl/ml each phosphatase inhibitor cocktail 1 and cocktail 2 (Sigma, P-2850 and P-5726, respectively). Lysis was achieved with five times of 45 sec 0.5-mm zirconia/silica bead beating with intermittent cooling on ice using FastPrep-24 (Scientific Industries). Lysates were spun down at 4°C and the supernatants were collected. Protein concentrations were determined using Quick Start Bradford dye reagent (Bio-Rad, no. 500-0205). For each sample, 30 μg of total protein was loaded on 8% sodium dodecyl sulfate (SDS)-polyacrylamide gels, and proteins were transferred to a PVDF membrane.
For phosphorylated Mpk1 determination, a 1:1000 dilution of phospho-p44/42 MAPK (Thr202/Tyr204) rabbit polyclonal antibody (Cell Signaling Technology, no. 9101) was used. Total Mpk1 protein was detected using a 1:2000 dilution of DYKDDDDK (FLAG) tag rabbit polyclonal antibody (Cell Signaling Technology, no. 2368). All primary antibodies were diluted in TBST (25 mM Tris-HCl (pH 8.1), 145 mM NaCl, 0.1% Tween 20) plus 5% bovine serum albumin and allowed to bind overnight at 4°C. Secondary antibodies used were goat anti-rabbit immunoglobulin G peroxidase conjugated (Sigma, A-6154) and were diluted 1:1000 in TBST. Proteins were detected using the SuperSignal West Dura kit (Thermo Scientific, no 34076).
Protein extract was prepared and quantitated as described above. A total of 1 mg of protein extract was incubated with Anti-FLAG M2 beads (M8823, Sigma Aldrich) for 3 hr at 4 °C. The beads were collected, washed 5 times with wash buffer (50 mM Tris–HCl, pH 7.5, 150 mM NaCl) and eluted with 25 μl 2X SDS-PAGE sample by boiling for 3 min. Supernatant was collected for immunoblotting. Protein samples were size fractionated on Novex 4–12% Bis-TRIS gradient gels using 4-morpholinepropanesulfonic acid buffer (Invitrogen) or on regular SDS-PAGE gels with a tris-glycine buffer. Note that these different methods cause slight differences in the running behavior of proteins. Size fractionated proteins were subsequently transferred onto membranes. The membranes were incubated with specific antibodies as indicated. Bound antibodies were detected as described above.
Cultures were grown at 30°C overnight in YPD liquid media and diluted with fresh YPD broth to 108/ml, as determined by optical density at 600nm. Fluconazole accumulation was measured by using the drug that had been tritiated by gas exchange to a specific activity of 629 GBq/mM (Amersham Biosciences). [3H] fluconazole was added to 6 ml cell suspension at a final concentration of 7.4 kBq/ml (0.2 μCi/ml, 3.6 ng/ml). After 60 min of incubation with rotation at 225 rpm (30°C), samples were washed three times with 6 ml PBS (pH 7.0) containing 100 μM unlabeled fluconazole and triplicates of 2 ml samples were filtered on a vacuum manifold (Millipore, Bedford, Mass.) with 24-mm-diameter GF/C glass fiber filters (Whatman, Maidstone, Kent, United Kingdom), which had been presoaked in 100 μM unlabeled fluconazole in PBS. Zero-time samples were chilled on ice for 30 min before the addition of tritiated fluconazole and kept on ice until they were washed and filtered. The filters were dried overnight at 37°C, placed in Hydrofluor scintillation fluid (National Diagnostic, Atlanta, GA) and allowed to stand overnight, and the cells were counted with a 200CA Tricarb liquid scintillation counter (Packard, Downer’s Grove, IL). Cell number was estimated based on optical density at 600 nm measured after a washing step for both 0 and 60 min samples, which allows consideration of the growth difference. Data were normalized to the level of 3H-FLC in H99 and presented as relative % uptake.
Overnight cultures were transferred to YNB without inositol for 2 hrs at 30°C. 3H-Myo-inositol (NET1177005 at 1 mCi/ml) was added at a concentration of 5 μCi/2×107 cells/ml for 1 hr before adding FLC (16 μg/ml). Cultures with or without FLC treatment were harvested 2hrs later. Cells suspensions at same OD600 were stopped with 0.1 mg/ml BSA and 5% TCA and put on ice for 10min. Cells were centrifuged and resuspended in 1 ml of ethanol/H2O/diethylether/pyridine (30:30:10:2). Samples were incubated at 60°C in a water-bath for 1 hr and sonicated two times during the incubation. The extracted lipids were dried and dissolved in chloroform/methanol (5:4) and 100 μl of the extract was spotted on silica TLC plate (EMD Chemical Inc. 60W HX733723) using a spotter. The lipids were resolved in chloroform/methanol/4.5N ammonium hydroxide (90:45:10) for 75 min. Radioactive bands were visualized by X-ray film after spraying with En3Hance (NEN Life Science Products 6NE970C). Sphingolipids and phosphatidylinositols (PI) were distinguished by treating the extracted lipids with monomethylamine/methanol/H2O/butanol (5:4:3:1) at 53°C for 45 min. The same volume of water and butanol/petroleum ether (b.p. 40-60°C)/ethyl formate (20:4:1) were added to the sample after cooling. The sample was mixed vigorously and spun. The lipid layer was dried and analyzed on TLC alongside samples without deacylation. Inositol phosphorylceramide (IPC), mannosyl-inositol phosphorylceramide (MIPC), mannosyl-di(inositol phosphoryl) ceramide (M(IP)2C were identified using a strain of S. cerevisiae as a reference based on previous publication since complex sphingolipids of C. neoformans have not been analyzed under the same experimental conditions. We noted that at least two hitherto unidentified complex sphingolipids were detected on TLC in C. neoformans but not in S. cerevisiae (Fig. 6B). For quantitative analysis of sphingolipids, X-ray films obtained from various exposure times of TLC plates were scanned with GS-800 Imaging Densitometer using Quantity One 1-D analysis software (BioRad). The relative amounts of lipid in each strain are expressed as a ratio between the signal density of each lipid and the corresponding lipid signal in the FLC untreated H99 sample. Because of the poor resolution of the signals for M(IP)2C and the slower migrating bands of unknown complex sphingolipids, the signals of M(IP)2C and all the slower migrating bands were combined and designated as MIP2C+ in our comparison.
Female BALB/c mice (6-8 weeks old) were injected via the lateral tail vein with 0.2 ml of a suspension of each yeast strain (2.5×104/ml) as described previously (Chang & Kwon-Chung, 1994) and the mortality was monitored. Kaplan-Meier analysis of survival was performed with JMP software for Macintosh (SAS Institute, Cary, NC).
Exponentially growing cultures (OD600 = 0.5-1.0) were washed, resuspended in 0.9% NaCl and adjusted to OD600 = 0.1. The cell suspensions were serially diluted, spotted onto the indicated media, and incubated for 3-4 days at 30°C
The FLAG epitope was inserted in frame at the carboxy terminus of YPK1 by PCR amplification and the norseothricin (NAT) resistance marker from pAI3 was cloned into the 3′ flanking region of YPK1. The resulting plasmid (pHL164) was sequenced to confirm that no errors had been introduced during amplification and used to transform pdk1Δ, sin1Δ as well as ypk1Δ. Transformants were first screened for loss of G418 resistance and for gain of norseothricin resistance to select integrative transformants containing the fusion construct at the YPK1 locus. PCR and Southern blot analysis were employed to confirm the integration event.
To generate a catalytically inactive (kinase-dead) version of Ypk1(K376A) and Ypk1(D386A), Lys376 (AAA) was replaced with Ala (GCT) and Asp386 (GAT) was replaced with Ala (GCT) respectively by PCR. These positions correspond to conserved residues critical for recognition of Mg2+-ATP substrate in all protein kinases (Hanks & Hunter, 1995). Likewise, to attempt to generate phosphorylation defective Ypk1 derivatives – T402A, S562A, and S543A, three presumptive phosphorylation sites, Thr402 (ACC), Ser562 (TCC), and Ser543 (TCG) that match the consensus sequence for phosphorylation by PDK1, PDK2 and TOR1, respectively, were replaced with alanine (GCT). Primers were designed to incorporate each mutation/substitution as well as restriction enzymes to swap the resulting fragment into pHL164. Each fragment containing the corresponding mutation was amplified from pHL164 by fusion PCR, cloned into pCR2.1, and sequenced to verify the sequences. The fragment of each construct was cloned into pHL164 to replace the wild-type corresponding fragments using internal restriction enzyme sites. 1.3kb NdeI-SalI fragment of Ypk1 from K274A mutant, 1.8kb NdeI-XhoI fragments of D386A and T402A, 1kb SalI-SnaB1 fragments of S543A and S562A were used to replace corresponding segments in pHL164 to create the final constructs, pHL177-181, respectively. Each construct was transformed into the ypk1Δ deletion mutant and the transformants were first screened for the loss of G418 resistance and gain of norseothricin resistance to select integrative transformants containing the fusion construct at the YPK1 locus. PCR and Southern blot analysis were used to confirm the integration event.
Figure S1. Plate assay for FLC sensitivity. Cells of exponentially growing cultures of original clones from the library screen were serially diluted and spotted onto media containing 8 μg/ml of FLC. The hypA mutant was not selected for further studies because of its unstable phenotype. The name of each strain was indicated on the left.
Figure S2. Ergosterol content of original clones from the library screen. Sterols of the log phase YPD grown cells of each indicated strain were extracted and analyzed by gas chromatography/mass spectrometry. The amount of ergosterol is expressed as the relative ratio to cholesterol, the internal recovery standard.
Figure S3 Relative expression levels of AFR1. Strains of original clones from the library screen were used for the study. RNAs were isolated from strains grown in vitro without FLC (A) and with 16μg/ml FLC (B). Quantitation of the relative transcript levels of AFR1 was performed by real-time RT-PCR analysis. Data were normalized with GPD1 level and expressed as the amount in each deletion mutant strain relative to that in H99.
Figure S4. Western blot of Ypk1-FLAG constructs. The Ypk1 expression levels were analyzed from the indicated strains expressing Ypk1-FLAG derivatives using anti-FLAG antibody.
Table S1. List of strains relevant to this study.
Table S2. List of STM strains chosen for this study.
Table S3. List of primers used in this study
We thank Y. J. Kim, T. Balla, M. A. Singh, and M. Del Poeta for suggestion and technical help in analysis of complex sphingolipids, Andy Han for technical help in initial screening of the mutants and A. Varma for critical discussions and reading of the manuscript. This study was supported by funds from the intramural program of the National Institute of Allergy and Infectious Diseases, NIH. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.