Sulfiredoxin plays peroxiredoxin-dependent and –independent roles via the HOG signaling pathway in Cryptococcus neoformans and contributes to fungal virulence

Summary

Mechanisms of oxidative stress resistance are crucial virulence factors for survival and proliferation of fungal pathogens within the human host. In this study we have identified and functionally characterized the role of sulfiredoxin, Srx1, in oxidative stress resistance of Cryptococcus neoformans causing fungal meningoencephalitis and regulation of peroxiredoxins, Tsa1 and Tsa3, and thioredoxins, Trx1 and Trx2. The C. neoformans HOG (High Osmolarity Glycerol) pathway was essential for the transcriptional regulation of SRX1 under peroxide stress conditions. A gene deletion study revealed that Srx1 was required for cells to counteract peroxide stress, but not other oxidative damaging agents. HOG1 was found to be essential for the induction of adaptive response to peroxide stress with concurrent repression of ergosterol biosynthesis in an SRX1-independent manner. Consistent with this, phosphorylation of C. neoformans Hog1 was modulated by both low and high doses of exogenous hydrogen peroxide treatment. Immunoblot analysis using the C. neoformans Tsa1 specific antibody revealed that both Srx1 and Trx1 were essential for recycling of oxidized Tsa1. In addition to its role in peroxide sensing and response C. neoformans Srx1 was also found to be required for a peroxiredoxin-independent function in promoting fungicide-dependent cell swelling and growth arrest. Finally we showed the importance of C. neoformans Srx1 in fungal pathogenesis by demonstrating its requirement for full virulence using a mouse infection model.

Introduction

Cryptococcus neoformans, a basidiomycete fungus, is ubiquitously found in diverse environmental niches, including trees, soils, bird guano, and even aquatic conditions. Under certain environmental conditions, cells having opposite mating types (α and a) may initiate sexual differentiation, which involves the processes of pheromone production, cell-to-cell fusion, filamentous growth with clamp connection, basidium formation, and production of four chains of basidiospores. The basidiospores or dried yeasts are considered to be infectious propagules, which can be inhaled through the upper respiratory tract and establish the initial infection at the lung alveoli of the lower respiratory tract, causing fungal pneumonia. Upon subsequent dissemination into the bloodstream, C. neoformans can infect the brain by passing through the blood-brain brain via either trancytosis or “Trojan horse” mechanism, causing meningoencephalitis. The systemic cryptococcosis is fatal to, albeit not limited to, immunocompromised individuals such as AIDS patients if left untreated (see reviews (Hull & Heitman, 2002; Lin & Heitman, 2006; Kronstad et al., 2011; Idnurm et al., 2005)).

To defend against C. neoformans infection, the host employs several types of innate immune cells. The alveolar macrophage located at the lung alveoli is one of such phagocytic cells (McQuiston & Williamson, 2012; Brummer, 1998; Garcia-Rodas & Zaragoza, 2012). During dissemination into other tissues, other phagocytic cells, such as neutrophils and monocytes, are also known to play a key role in limiting C. neoformans infection (Seider et al., 2010; Urban et al., 2006; Serbina et al., 2008). These cells phagocytize the invading pathogen by forming phagosomes, which are subsequently fused to lysosomes generating phagolysosomes. Activated phagocytes increase uptake of oxygen through the respiratory burst and within the acidic phagolysosome generate various toxic reactive oxygen species (ROS) or nitrosative species (RNS), including superoxide anions (O₂^•−), hydrogen peroxide (H₂O₂), hydroxyl radicals (^•OH), hypochlorous acid (HOCl, from Cl and H₂O₂ via myeoloperoxidase), singlet oxygen (¹O₂), and nitric oxide (NO). The ROS and RNS can kill ingested pathogens by oxidizing critical cellular components. In response to the innate immune response by the host, C. neoformans employs a variety of cellular defense mechanisms. During infection, C. neoformans produces two major virulence factors, polysaccharide capsule and melanin pigment. Both capsule and melanin allow C. neoformans to resist the phagocytosis by the phagocytes and thereby allow the pathogen to avoid clearance during the progression of cryptococcosis (see reviews (Bose et al., 2003; McFadden & Casadevall, 2001; Zaragoza et al., 2009; Casadevall et al., 2000; Jacobson, 2000; Langfelder et al., 2003)). Once engulfed by the phagocytes regardless of the antiphagocytic molecules, C. neoformans activates a series of oxidative stress response signaling cascades not only to detoxify the ROS or RNS but also to repair the damages caused by the oxidative insult. Impairment of the oxidative defense mechanisms results in significant reduction in virulence of the pathogen (Brown et al., 2007).

The general oxidative stress response and redox regulation systems have been well characterized in the model yeasts, such as Saccharomyces cerevisiae (see reviews, (Herrero et al., 2008; Pereira et al. 2001; Antelmann & Helmann, 2011)). During the oxidative burst as well as normal respiration, O₂^•−, the precursor of most ROS, is generated and subsequently converted to H₂O₂ spontaneously or through catalysis by superoxide dismutases (SODs). S. cerevisiae contains a Cu,Zn-dependent Sod1, which is located at the cytoplasm and mitochondrial intermembrane space, and a Mn-dependent Sod2 localized at the mitochondrial matrix. H₂O₂ is further detoxified by its full reduction to water (H₂O) through catalase and peroxidase systems. The budding yeast contains two heme-associated catalases, Cta1 and Ctt1, which are localized to peroxisomes and cytoplasm, respectively. Unlike SOD and catalases that use the redox status of associated metals, the glutathione peroxidase (Gpx) and peroxiredoxins (Prxs; also known as thioredoxin peroxidases) reduce inorganic and organic peroxides utilizing electrons donated by reduced glutathione (GSH) and thioredoxin (Trx), respectively. In the presence of reduced transition metals such as iron, H₂O₂ is partly reduced, which generates even stronger oxidants, ^•OH, by the Fenton reaction.

Similar antioxidant defense systems have been identified and partially characterized in C. neoformans. The cytosolic Cu,Zn-SOD, Sod1, has been shown to have a pivotal role in defense against ROS and its deletion severely attenuates virulence of the pathogen (Cox et al., 2003). C. neoformans contains four catalases (Cat1-4), among which CAT2 and CAT4 are closest orthologs of yeast CTA1 and CTT1, respectively. Unexpectedly, however, deletion of all four CAT genes did not affect sensitivity to ROS or virulence of the pathogen (Giles et al., 2006), indicating that other peroxide-detoxifying systems may have a full complementary role in C. neoformans. Indeed, two peroxidase systems, Gpx and Prx, have been discovered. C. neoformans contains two glutathione peroxidases, Gpx1 and Gpx2, both of which are involved in defense against organic peroxides, such as tert-butyl hydroperoxide (an alkyl peroxide; tBOOH) and cumen hydroperoxide (an aromatic peroxide), but not against H₂O₂ or O₂^•− (Missall et al., 2005). Although both Gpx proteins are required for survival in macrophages, deletion of GPX1 and GPX2 does not affect virulence of C. neoformans in mice (Missall et al., 2005), suggesting that other peroxidase or antioxidant systems can compensate for the loss.

In C. neoformans, the Prx system appears to be important for both oxidative stress response and virulence. Generally, Prxs are classified into 2-Cys or 1-Cys Prxs. During reaction with peroxide, typical 2-Cys Prxs attack the substrate peroxide with the peroxidatic cysteine, which is oxidized to cysteine sulfenic acid (-SOH). In the second step of the catalysis, the peroxidatic cysteine sulfenic acid from one Prx subunit reacts with the resolving cysteine from the other Prx subunit to form homodimers through an intersubunit disulfide bond. In contrast, atypical 2-Cys Prxs form an intramolecular disulfide bond. On the other hand, 1-Cys Prxs have a monomeric form with a single active cysteine site. The sulfenic acid is recycled to the reduced form (R-SH) with the help of disulfide-specific oxidoreductases and particularly by thioredoxins (Wood et al., 2003; Roussel et al., 2009). Upon hyperoxidation, however, Prxs are further oxidized and inactivated by forming sulfinic acids (R-SO₂H). Although this had been considered an irreversible process, a sulfiredoxin (Srx) was discovered to recycle the sulfinic acid form its sulfenic acid derivative in an ATP-dependent reaction (Biteau et al., 2003). The sulfinic acid is often oxidized to the sulfonic acid (R-SO₃H), which is an irreversible process that inactivates peroxidase activity but enhances molecular chaperone activity of Prxs (Lim et al., 2008). In yeasts, the oxidized Prxs can activate an AP-1 like transcription factor, such as Yap1 in the budding yeast or Pap1 in the fission yeast, by inducing conformational changes through the formation of disulfide bonds. Activated AP-1 like transcription factors tend to accumulate within the nucleus and induce a plethora of stress defense genes to counteract the incoming oxidative stress (Herrero et al., 2008). In C. neoformans, two peroxiredoxins, Tsa1 and Tsa3, have been discovered and Tsa1, but not Tsa3, counteracts both organic and inorganic peroxides (Missall et al., 2004). Furthermore, deletion of TSA1, but not TSA3, abolishes virulence of the pathogen, indicating that the Tsa1-mediated Prx system is a key cryptococcal thiol peroxidase system. In addition, two thioredoxin proteins, Trx1 and Trx2, have been discovered in C. neoformans with Trx1 playing a major role in fungal growth and pathogenesis (Missall and Lodge, 2005). However, it is unknown how the hyperoxidized form (R-SO₂H) of peroxiredoxins is recycled in C. neoformans and which signaling cascades control the Prx- and Trx-systems.

Our recent independent transcriptome analysis studies, one in response to peroxide induced oxidative stress and the other in a stress-activated HOG signaling pathway, revealed that expression of a sulfiredoxin-like gene, SRX1, is highly upregulated in response to oxidative stress (H₂O₂) and this upregulation is dependent on functional Hog1 (Ko et al., 2009; Upadhya et al., 2013). Nevertheless, the function of Srx1 remains unknown in C. neoformans. In this study we have functionally characterized the role of Srx1 in relation to its capacity to repair and recycle oxidized Tsa1. Remarkably, here we found that in addition to its role in Prx-dependent oxidative stress response, C. neoformans Srx1 had Prx-independent function in promoting fungicide-mediated cell growth arrest. Furthermore we discovered that Srx1 is also required for full virulence of C. neoformans. Therefore, this is a first report of the role of Srx1 in fungal pathogenesis.

Results

C. neoformans genome has a single sulfiredoxin-like gene, SRX1

Two independent global transcriptome analysis studies of C. neoformans experiencing oxidative stress induced by the exposure to exogenous H₂O₂ under two different growth conditions have revealed that expression of a peroxiredoxin (Prx, thiol specific antioxidase) gene, TSA1 (CNAG_03482.2), and a sulfiredoxin-like gene, CNAG_00654.2 (named SRX1), are strongly induced in a HOG-dependent manner (Ko et al., 2009; Upadhya et al., 2013)(Supplemental Fig. S1A). Although TSA1 has been previously identified and characterized in C. neoformans (Missall et al., 2004), SRX1 has not been studied before in this pathogen. Therefore we aimed to functionally characterize Srx1 in relation to the peroxiredoxin system and the HOG pathway.

First we analyzed the genomic DNA structure and protein coding sequence of SRX1. For this purpose, we performed 5’/3’ rapid amplification of cDNA ends (RACE) and the whole cDNA analysis. The SRX1 gene is composed of one intron and two exons and has 83 bp of 5’-untranslated region (UTR) and 56-bp of 3’-UTR terminator region. The coding sequence analysis revealed that SRX is expected to encode a 142 amino acid protein, which is identical to that predicted by the Broad Institute. A key cysteine residue (Cys84 of S. cerevisiae Srx1), which is conserved in most of sulfiredoxins in eukaryotes and forms a transient disulfide with peroxiredoxins, is also present in the Cryptococcus Srx1 (Cys106) and even its surrounding amino acid residues (GGCHR) are completely conserved (Supplemental Fig. S1B). The BLAST analysis showed that SRX1 (CNAG_00654.2) is the only sulfiredoxin-like gene in C. neoformans, similar to other fungi that contain only a single Srx1-like sulfiredoxin.

C. neoformans Hog1 mediates transcriptional regulation of SRX1 under oxidative stress inducing conditions

To further analyze expression patterns of genes involved in the peroxiredoxin, thioredoxin, and sulfiredoxin systems, we performed Northern blot analysis for SRX1, TSA1, TSA3, TRX1, and TRX2 with or without exposure to H₂O₂ (Fig. 1 and Supplemental Fig. S2). Expression of SRX1 was induced as early as 10 min post peroxide treatment and peaked at 30 and 60 minutes before gradually decreasing to a level similar to that of untreated control at 3 hr post peroxide treatment (Fig. 1A and Supplemental Fig. S2A). Transcript levels of TSA1 also increased and peaked at 1 and 2 hr post peroxide treatment and subsequently showed a decreased transcription at 3 hr time point (Fig. 1B and Supplemental Fig. S2B). Expression of TSA3, which encodes a protein homologous to mitochondrial thiol peroxidase of S. cerevisiae, was negatively regulated by peroxide. TSA3 transcript levels decreased until 60 min post peroxide treatment before starting to show transcriptional upregulation with a peak transcript levels at 3 hr post peroxide treatment (Fig. 1C and Supplemental Fig. S2C). Unlike peroxiredoxins and sulfiredoxins, transcription of C. neoformans thioredoxin genes was not significantly influenced by exogenous H₂O₂ treatment (Fig. 1D and 1E, Supplemental Fig. S2D and S2E).

Fig. 1. The role of C. neoformans Hog1 in controlling the expression of sulfiredoxin, peroxiredoxin and thioredoxin genes under H₂O₂-induced oxidative stress condition.

Northern blot analysis using the total RNA isolated from the WT (H99) strain and hog1Δ mutant grown to mid-logarithmic phase at 30ΔC in YPD medium and exposed to 2.5 mM of H₂O₂ (A-E and G) or 0.2 mM (F). Each membrane was hybridized with gene specific probes washed and developed. The same membrane was reprobed with the C. neoformans ACT1 gene specific probe. The relative expression level of each gene was quantitatively measured with phosphorimager analysis after normalization with ACT1 expression levels. Each value at indicated time points represents expression levels relative to the levels at zero time point. For (F), ethidium bromide staining results of rRNA were used as loading controls.

To determine whether transcriptional expression of above genes induced by peroxide treatment is mediated by C. neoformans Hog1, we performed Northern expression analysis after treating wild-type and the hog1Δ mutant with exogenous H₂O₂. Interestingly, peroxide induced upregulation of SRX1 gene expression seen in wild-type cells was highly reduced in the hog1Δ mutant, indicating that the HOG signaling pathway is an essential mediator of peroxide stress signaling that regulates SRX1 gene expression (Fig. 1A). Basal expression levels of TSA1 were slightly higher in the hog1Δ mutant than the wild-type whereas its extent of induction in response to peroxide treatment was not significantly affected in the hog1Δ strain compared to wild-type cells (Fig. 1B). Expression of TSA3 was reduced in response to 2.5 mM H₂O₂ in a HOG1-independent manner, although its basal expression levels also increased in the hog1Δ mutant similar to TSA1 (Fig. 1C). While expression of TRX1 was not influenced by the presence or absence of peroxide either in wild-type or hog1Δ cells (Fig. 1D), TRX2 gene expression was elevated in hog1Δ cells compared to wild-type and its expression was not influenced further by peroxide treatment (Fig. 1E).

In Schizosaccharomyces pombe, SRX1 expression is induced by both low levels (0.2 mM) and high levels (1 mM) of H₂O₂ and this transcriptional regulation is mediated by two different signaling pathways. At low levels of H₂O₂, Pap1, a S. cerevisiae Yap1 ortholog, mediates the transcriptional regulation of SRX1. However, at higher concentration of H₂O₂, Sty1, a Hog1 ortholog, plays a critical role in regulating SRX1 expression (Vivancos et al., 2005). In C. neoformans, however, the Pap1/Yap1 ortholog has not been identified. Therefore we also measured SRX1 expression in response to low levels of H₂O₂ (0.2 mM) in wild-type and hog1Δ mutant strains. SRX1 expression was only transiently induced (peaked at 10 min and then declined) in response to 0.2 mM H₂O₂ (Fig. 1F), strongly suggesting that induction levels of SRX1 depend on peroxide concentration. Interestingly, this increase was observed only in the wild-type cells and not in the hog1Δ mutant, indicating that unlike in S. pombe, C. neoformans Hog1 may regulate the expression of SRX1 even at lower concentrations of H₂O₂ (Fig. 1F).

To further confirm the role of Hog1 in regulation of SRX1 expression, we examined SRX1 expression in the hog1Δ::HOG1^T171A+Y173A strain (YSB253), in which dual phosphorylation sites are mutated, and the hog1Δ::HOG1^K49S+K50N strain (YSB308), in which residues essential for Hog1 kinase activity are mutated (Fig. 1G). The peroxide-mediated induction of SRX1 was significantly reduced to the extent of that in the hog1Δ mutant in dual phosphorylation-defective or kinase-dead hog1 mutants. These data firmly demonstrate that both dual phosphorylation and kinase activity of Hog1 is required for regulation of SRX1 expression. However, the weak induction of SRX1 in the hog1Δ mutant indicates that other signaling pathway(s) may exert minor control the peroxide-mediated SRX1 induction. Taken together, sulfiredoxin expression is under dynamic control through the HOG pathway in response to peroxide stress in C. neoformans.

C. neoformans Srx1 is critical for sensing and responding to exogenous peroxide treatment

Peroxide-dependent modulation of SRX1 expression implied that Srx1 could be functionally involved in oxidative stress response. To elucidate the role of Srx1 in C. neoformans, we generated the srx1Δ deletion mutants in H99, KN99α, and KN99a strain backgrounds as described in Experimental procedures. To verify the specificity of the observed phenotypes to the Srx1 protein we constructed the srx1Δ::SRX1 complemented strain. As members of C. neoformans peroxiredoxin and thioredoxin families have been earlier shown to be critical to counteract oxidative stress (Missall et al., 2004; Missall and Lodge, 2005), we wanted to compare their phenotypes to that of the srx1Δ mutant. Therefore we generated the tsa1Δ, tsa3Δ, tsa1Δ tsa3Δ, trx1Δ, trx2Δ, and trx1Δ trx2Δ double mutants in the H99 strain background (data not shown). Deletion of SRX1 did not cause any defects in cellular growth at temperature ranging from 25°C to 39°C, indicating that Srx1 is dispensable for normal growth of C. neoformans. As expected, the srx1Δ mutant was highly susceptible to H₂O₂ compared to wild-type. The magnitude of the peroxide sensitivity in the srx1Δ mutant was dependent on the concentration of H₂O₂ used for the treatment. The srx1Δ mutant was slightly sensitive compared to wild-type at a concentration of 1 mM of H₂O₂ whereas it exhibited greater sensitivity to higher concentrations (>2 mM) of H₂O₂ (Fig. 2A). In contrast, tsa1Δ, tsa1Δ tsa3Δ, trx1Δ and trx1Δ trx2Δ mutant strains were found to be very sensitive even to 1 mM H₂O₂ (Fig. 2A). These data indicate that Srx1 is required for withstanding a higher concentration of H₂O₂ whereas Trx and Prx systems are required for survival at a lower concentration of H₂O₂. In response to organic peroxide, such as tert-butyl hydroperoxide (tBOOH), the srx1Δ, tsa1Δ, tsa1Δ tsa3Δ, trx1Δ, and trx1Δ trx2Δ mutants exhibited similar levels of high sensitivity (Fig. 2A). Deletion of either TSA3 or TRX2 alone did not cause any changes in peroxide sensitivity (Fig. 2A), which is consistent with their limited role in peroxide resistance as previously published (Missall et al., 2004; Missall and Lodge, 2005).

Fig. 2. Srx1 is required for peroxide sensing and response in C. neoformans.

C. neoformans stains [Wild-type (H99), hog1Δ (YSB64), srx1Δ (YSB596), srx1Δ::SRX1 (YSB1214), tsa1Δ (YSB1273), tsa3Δ (YSB1204), tsa1Δ tsa3Δ (YSB2735), trx1Δ (YSB1667), trx2Δ (YSB1791), and trx1Δ trx2Δ (YSB1796)] were grown overnight at 30°C in liquid YPD medium, 10-fold serially diluted, and spotted on YPD agar containing the indicated concentration of H₂O₂, tert-butyl hydroperoxide (tBOOH) (A), two superoxide anion generators (menadione and paraquat) (B), or a thiol-specific oxidant diamide (C). Cells were incubated at 30°C for 3 days and photographed.

Superoxide anion (O₂^•−) is normally converted to H₂O₂ via the action of superoxide dismutases (SODs). Therefore we thought that Srx1 could be also involved in counteracting superoxide anions. However, Srx1 appeared to be dispensable for this function. In response to two superoxide generators, menadione and paraquat, the srx1Δ mutant exhibited wild-type levels of sensitivity (Fig. 2B). Indeed, the srx1Δ and hog1Δ mutants showed slightly increased resistance to paraquat than wild-type for unknown reasons. It is probable that superoxide generated by menadione and paraquat may not produce sufficiently high levels of H₂O₂, which requires Srx1 action. Similar to the srx1Δ mutant, the tsa1Δ mutant also exhibited wild-type levels of sensitivity to both menadione and paraquat (Fig. 2B). Interestingly, however, the tsa3Δ mutant exhibited slightly higher sensitivity to paraquat and the tsa1Δ tsa3Δ double mutant exhibited even higher sensitivity to both menadione and paraquat than tsa1Δ and tsa3Δ single mutants (Fig. 2B), suggesting that Tsa3 may be required for defense against superoxide anion. Both trx1Δ and trx1Δ trx2Δ mutants, but not the trx2Δ mutant, displayed increased sensitivity to menadione and paraquat. Therefore it is in agreement with the aforementioned hypothesis that Prx and Trx1 are required for counteracting low levels of H₂O₂. When the cells were exposed to diamide, which is a membrane-permeable, thiol-specific oxidizing agent but does not produce any ROS, the srx1Δ mutant did not show any increased sensitivity (Fig. 2C). In contrast, both Tsa1 and Tsa3 play redundant roles in diamide response. The tsa1Δ and tsa3Δ mutants exhibited increased diamide-sensitivity and the tsa1Δ tsa3Δ double mutants showed even higher sensitivity than each single mutant (Fig. 2C). Similarly TRX1 also seems to play a role in diamide response (Fig. 2C). These results indicate that unlike peroxiredoxins and thioredoxins, C. neoformans Srx1 appears to be primarily involved in sensing and responding to high levels of peroxide.

In addition to the components of oxidative stress resistance, survival of the fungal pathogens within the host macrophages also requires mechanisms to neutralize host induced nitrosative stress. Therefore we tested the contribution of Srx1 to the survival of C. neoformans upon nitrosative stress. When tested on a medium containing sodium nitrite, a nitric oxide donor at pH 4, the srx1Δ mutant was as resistant to NaNO₂ as wild-type, indicating that C. neoformans sulfiredoxin is dispensable for nitrosative stress sensing and response (data not shown).

C. neoformans HOG1 is required for peroxide adaptive response in an Srx1-independent manner

In S. cerevisiae and other yeasts, pre-exposure to low levels of H₂O₂ has been shown to induce resistance to subsequent challenge with a higher dose of exogenous H₂O₂, which is a phenomenon defined as adaptive response. Because the C. neoformans HOG pathway plays a major role in mediating transcriptional response to peroxide stress, we addressed whether C. neoformans has the ability to adapt to oxidative stress, and if such mechanism exists, determine whether it depends on functional Hog1. When wild-type C. neoformans cells pre-exposed to a range of H₂O₂ (0.2-1.0 mM) for different time periods were challenged with a higher dose of H₂O₂ (2.5 mM), there was a increase in the basal levels of oxidative stress resistance (Fig. 3A). One hr pre-exposure to 0.2 mM H₂O₂ was sufficient to induce adaptive response to peroxide stress (Fig. 3A). Importantly, the C. neoformans HOG1 gene was found to be essential for mediating adaptive response to oxidative stress (Fig 3A). Regardless of the amount of H₂O₂ or time of pre-exposure, the hog1Δ mutant exhibited almost identical levels of sensitivity to 2.5 mM H₂O₂, strongly indicating that the peroxide-adaptation is mediated through the HOG pathway. Moreover, in spite of its role in counteracting peroxide stress, Srx1 appeared to be dispensable for inducing adaptation to peroxide stress. Pre-exposure to 0.2-0.5 mM H₂O₂ for one to two hr significantly increased the resistance of the srx1Δ mutant to subsequent challenge with 1.5 mM H₂O₂ (Fig. 3B). Pre-exposure to 1 mM H₂O₂ appeared to be fairly toxic to the srx1Δ mutant despite the short incubation time (Fig. 3B). In conclusion, the C. neoformans Hog1 not only controls oxidative stress sensing and response partly via Srx1, but also is required for oxidative stress adaptation in an Srx1-independent manner.

Fig. 3. C. neoformans HOG1 is essential for inducing an adaptive response to oxidative stress in an Srx1-independent manner.

C. neoformans strain [Wild-type (H99), hog1Δ (YSB64), and srx1Δ (YSB596)] were grown overnight at 30°C in liquid YPD medium and sub-cultured for about 4 hr to 1.0 of OD₆₀₀. Cells were pre-exposed to the indicated concentration of H₂O₂ for 30 min, 1 hr, and 2 hr and spotted on YPD agar medium containing the indicated concentration of H₂O₂. Cells were incubated at 30°C for 2 days and photographed.

Ergosterol biosynthesis is repressed during peroxide induced oxidative stress in a Hog1-dependent and Srx1-independent manner in C. neoformans

In S. cerevisiae, ergosterol biosynthesis is repressed during osmotic and oxidative stress responses by the Hog1-dependent and -independent transcriptional downregulation of ERG11 and ERG2 (Montañés et al, 2011). This finding led us to address whether similar type of repression of ergosterol biosynthesis occurs during oxidative stress in C. neoformans and if so whether Hog1 and/or Srx1 are involved in this process. First, we examined whether ERG11 and ERG2 expressions are downregulated during oxidative stress response. In agreement with the finding in S. cerevisiae, expression levels of ERG11 and ERG2 were significantly downregulated within 30 min in response to 2.5 mM H₂O₂ in the wild-type strain (Fig. 4A). In contrast, the repression of ERG11 and ERG2 was either delayed or did not occur in the hog1Δ mutant (Fig. 4A). Downregulation of ERG11 was delayed and only occurred after 60 min exposure to H₂O₂ in the hog1Δ mutant, suggesting that the HOG and other signaling pathways are involved in the process. More strikingly, H₂O₂-mediated downregulation of ERG2 did not occur at all in the hog1Δ mutant (Fig. 4B). These data strongly indicate that the HOG pathway downregulates ERG11 and ERG2 expression to repress ergosterol biosynthesis during oxidative stress response in C. neoformans. This is in stark contrast to the case in S. cerevisiae, where the HOG pathway is largely dispensable for ERG gene repression during oxidative stress response (Montañés et al, 2011). In the srx1Δ mutant, however, ERG11 and ERG2 expressions were reduced like those in the wild-type strain (Fig. 4A and 4B), suggesting that Srx1 is dispensable for repression of ergosterol biosynthesis during oxidative stress in C. neoformans.

Fig. 4. Repression of ergosterol biosynthesis by the HOG pathway in C. neoformans.

Northern blot analysis using the total RNA isolated from the WT (H99) strain and hog1Δ and srx1Δ mutants grown to mid-logarithmic phase at 30°C in YPD medium and exposed to 2.5 mM of H₂O₂ (A and B). Each membrane was hybridized with gene specific probes (ERG11 for A and ERG2 for B), washed and developed. Ethidium bromide staining results of rRNA were used as loading controls. (C) C. neoformans strain [Wild-type (H99), hog1Δ, and srx1Δ] were grown overnight at 30°C in liquid YPD medium and sub-cultured for about 4 hr to 1.0 of OD₆₀₀. Cells were treated with (indicated as ‘+’) or without (indicated as ‘-‘) the 10 μg/ml of fluconazole for 1 hr and spotted on YPD agar medium containing the indicated concentration of H₂O₂. Cells were incubated at 30°C for 2 days and photographed.

To further support this finding, we tested whether artificial repression of cellular ergosterol content by treatment with azole drugs can increase resistance to oxidative stress. In fact, fluconazole-treated wild-type cells exhibited a greater resistance to peroxide stress than non-treated wild-type cells whereas fluconazole treatment did not affect the oxidative stress sensitivity in the hog1Δ mutant (Fig. 4C). All these data further corroborate that Hog1-mediated ergosterol repression is required for oxidative stress response and adaptation. Interestingly, however, fluconazole treatment did not enhance peroxide resistance in the srx1Δ mutant (Fig. 4C), which shows normal ergosterol repression during oxidative stress. Taken together, ergosterol biosynthesis is repressed for resistance to peroxide stress in a Hog1-dependent manner in C. neoformans.

Phosphorylation status of C. neoformans Hog1 is modulated by exogenous peroxide treatment

A previous report has demonstrated the unique phosphorylation pattern of C. neoformans Hog1 in a serotype A strain upon subjecting the cells to high osmolarity stress (Bahn et al., 2005). To gain further evidence for the role of C. neoformans Hog1 during exposure to both low (0.2 mM) and high (2.5 mM) doses of H₂O₂ we tested the phosphorylation status of Hog1 during peroxide exposure. When we treated C. neoformans H99 strains to 1 M NaCl, high osmolarity stress caused the dephosphorylation of Hog1 as shown in Fig. 5, which is consistent with the previously reported data (Bahn et al., 2005). The majority of Hog1 was found to be dephosphorylated at 20 min after the treatment with 1 M NaCl (Fig. 5). Similar to high osmolarity stress, exposure of the cells to oxidative stress induced by treating wild-type cells with 2.5 mM H₂O₂ caused dephosphorylation of Hog1 (Fig. 5). After 20 min of 2.5 mM peroxide treatment the majority of Hog1 was found to be in its dephosphorylated form. However, 2.5 mM of peroxide treatment resulted in a slower rate of Hog1 dephosphorylation than that observed with salt stress. Interestingly, a modest level of dephosphorylation of Hog1 was also observed at the 20 min time point when cells were treated with a low dose of 0.2 mM of H₂O₂ (Fig. 5). These immunoblot studies provide further evidence that C. neoformans Hog1 plays an important role during oxidative stress induced by both low and high doses of exogenous H₂O₂.

Fig. 5. C. neoformans Hog1 undergoes dephosphorylation when subject to oxidative stress under both low (0.2 mM) and high (2.5 mM) concentrations of exogenous H₂O₂.

C. neoformans KN99 strains were grown in YPD. At an OD₆₅₀ of 1.0, cells were treated with 0.2 mM and 2.5 mM of H₂O₂ to induce mild and strong oxidative stress respectively. As a positive control cells were exposed to 1M NaCl to induce high osmolarity stress. At various time points cells were collected, washed, and used for the preparation of total cell lysate. Equal amount of protein was subject to immunoblot analysis. Phosphorylation of the Hog1 was monitored using phospho-p38 MAPK antibody. The same blot was probed with Hog1-specific polyclonal antibody to measure the level of total Hog1 protein.

Functional Srx1 is essential for the reductive recycling of oxidized Tsa1

Due to the critical requirement of Tsa1 during oxidative stress and the sensitivity of C. neoformans trx1Δ and srx1Δ mutants to peroxide, we wanted to characterize biochemically the potential role of C. neoformans Trx1 and Srx1 in the recycling of oxidized and hyperoxidized cysteine residues of Tsa1, respectively. To this end, we treated logarithmically growing cells of wild-type, trx1Δ and srx1Δ strains with 6 mM H₂O₂. Protein extracts from whole cell lysates were separated on non-reducing and reducing polyacrylamide gels in the presence of 0.1% SDS and were subjected to immunoblot analysis using a polyclonal antibody raised against C. neoformans Tsa1. Tsa1 protein in wild-type cells was found to be present as monomers and dimers (Fig. 6). In the absence of Trx1, a majority of Tsa1 was found to be locked in their oxidized dimers even in the absence of exogenous peroxide. Addition of peroxide did not influence significantly the composition of Tsa1 dimers in the absence of Trx1 (Fig. 6, non-reducing condition). However, treatment of srx1Δ cells with peroxide resulted in the migration of Tsa1 protein at a position corresponding to its monomeric state potentially reflecting its hyperoxidized form, since these monomers cannot associate to form protein dimers (Fig. 6, non reducing condition). Separation of proteins under reducing conditions resulted in a single Tsa1 specific band for all the samples (Fig. 6 reducing condition). These immunoblot studies clearly indicate that in the absence of functional Srx1, the majority of Tsa1 loses its potential to form dimers, suggesting the importance of Srx1 in the repair and recycling of catalytic cysteine residues of Tsa1.

Fig. 6. C. neoformans Srx1 is essential for Tsa1 recycling.

Immunoblot analysis of C. neoformans Tsa1 from wild-type, trx1Δ and srx1Δ cells before and after treating with 6 mM H₂O₂ at various time points. Cells were lysed in 20% TCA to prevent the formation of non specific disulfide bonds during the preparation of total cell lysate and active –SH groups were protected from oxidation by reacting with iodoacetamide. A total of 50 μg of protein was separated on a 10% Bis-Tris gel and samples were run under non-reducing and reducing conditions using MOPS buffer containing 0.1% SDS. Actin specific antibodies were used as a loading control.

Peroxiredoxin-independent role of SRX1 in fungicide-dependent cell swelling and growth arrest

Based on the finding that Hog1 is required for peroxide-mediated SRX1 induction (Fig. 1), we addressed whether the hyperactivation of the HOG pathway can trigger induction of SRX1 even in the absence of peroxide. For this purpose, we used a phenylpyrrole class of fungicide, fludioxonil, which is known to hyperactivate the HOG pathway and thereby trigger over-accumulation of intracellular glycerol, which subsequently causes cytokinesis defects and cell swelling in C. neoformans (Kojima et al., 2006). Therefore, mutation of any signaling component in the HOG pathway, including SSK1, SSK2, PBS2, or HOG1, confers almost complete resistance to fludioxonil (Kojima et al., 2006; Bahn et al., 2007). We found that SRX1 expression was transiently induced (after 10 min exposure) upon fludioxonil treatment in the wild-type strain, but not in the hog1Δ mutant (Fig. 7A).

Fig. 7. The Prx-independent role of Srx1 in mediating cellular response to fludioxonil via the HOG pathway without involvement in intracellular glycerol synthesis in C. neoformans.

(A) Northern blot analysis of total RNA isolated from the WT (H99) strain and hog1Δ mutant at various time points after treating the cells grown to middle logarithmic phase at 30°C in YPD medium with 40 μg/ml fludioxonil. (B) C. neoformans stains [Wild-type (H99), hog1Δ (YSB64), srx1Δ (YSB596), srx1Δ::SRX1 (YSB1214), tsa1Δ (YSB1273), tsa3Δ (YSB1204), trx1Δ (YSB1667), trx2Δ (YSB1791), and trx1Δ trx2Δ (YSB1796)] were grown overnight at 30°C in liquid YPD medium, 10-fold serially diluted, and spotted on YPD agar containing 10 μg/ml of fludioxonil . Cells were incubated at 30°C for 3 days and photographed. (C and D) Each strain [Wild-type (H99), hog1Δ (YSB64), srx1Δ (YSB596), srx1Δ::SRX1 (YSB1214)] was grown overnight at 30°C in YPD medium and subcultured in fresh YPD medium to an OD₆₀₀=1.0. Then the cells were re-incubated in YPD medium with or without fludioxonil (10 μg/ml) at 30°C for 48 hr with shaking, and then photographed (C). Bars in the picture indicate 10 μm. (D) Cell diameter of each strain was quantitatively measured by using the NIS-Elements AR 4.00.00 image analysis software (Nikon). The Y-axis indicates average diameter of the cell (μm, micron). Total 100 cells per strain were measured. (E) To quantitatively measure intracellular glycerol accumulation, each strain was grown to the middle logarithm phase in liquid medium and re-incubated in YPD medium containing fludioxonil (10 μg/ml) at 30°C for 3 hr with shaking. The glycerol content in each strain was measured with the UV-glycerol assay kit by following the manufacturer's instruction. Four independent experiments with duplicate technical replicates were performed. Error bars indicate standard deviation. Statistic differences in relative cell diameter (D) and glycerol content (E) were determined by Bonferroni's multiple comparison test. Each symbol indicates the following: *, P < 0.05; NS, not significant (P > 0.05).

Next we tested whether Srx1 is involved in resistance to fludioxonil. Surprisingly, the absence of Srx1 rendered C. neoformans cells more resistant to fludioxonil treatment (Fig. 7B). The srx1Δ mutant was significantly more resistant to fludioxonil than wild-type and its complemented strain, albeit to a less extent than the hog1Δ mutant. Although the tsa1Δ mutant exhibited slightly increased resistance to a low concentration of fludioxonil, deletion of the two Prx genes, TSA1 and TSA3, rendered cells to be even more susceptible to fludioxonil than the wild-type strain (Fig. 7B), suggesting that Srx1 may promote fludioxonil sensitivity in a manner independent of the Prx system. Interestingly, Trx1 and Trx2 play opposing roles in fludioxonil susceptibility. The trx1Δ mutant was more sensitive to fludioxonil than the wild-type whereas the trx1Δ trx2Δ mutant was slightly more resistant to a low concentration of fludioxonil than the trx1Δ mutant (Fig. 7B). In summary, the srx1Δ mutant showed higher fludioxonil resistance than any other Prx or Trx mutant strains.

Next we monitored whether Srx1 promotes cytokinesis defects and cell swelling upon exposure to fludioxonil. As reported before, the wild-type cells became significantly swollen upon fludioxonil treatment due to over-accumulation of intracellular glycerols whereas the hog1Δ mutant cells were unaffected (Fig. 7C and 7D). Notably, similar to the hog1Δ mutant, the srx1Δ mutant was not significantly swollen by fludioxonil (Fig. 7C and 7D), indicating that Srx1 could be involved in controlling intracellular glycerol content. Unexpectedly, however, induction of intracellular glycerol in response to fludioxonil was normal in the srx1Δ mutant whereas it was defective in the hog1Δ mutant as reported previously (Fig. 7E), suggesting that Srx1 does not regulate intracellular glycerol synthesis per se. Taken together, Srx1 has a Prx-independent role in promoting cellular response to fludioxonil without any involvement in intracellular glycerol synthesis in C. neoformans.

Srx1 is dispensable for azole and polyene drug susceptibility

The capability of Srx1 to promote cellular response to the fungicide fludioxonil led us to test the role of Srx1 in susceptibility to other antifungal drugs, such as azoles and polyenes, which are clinically used. In C. neoformans, the HOG pathway represses expression of some ergosterol biosynthesis genes, such as ERG11. Therefore deletion of HOG1 increases basal expression levels of ERG11, which confers resistance to azole drugs (fluconazole or ketoconazoles), and cellular ergosterol levels, which subsequently increases the membrane binding of amphotericin B and the drug-susceptibility (Ko et al., 2009)(Fig. 8). Unlike the case in fludioxonil resistance, the srx1Δ mutant exhibited almost wild-type levels of resistance to azole drugs (ketoconazole, itraconazole and fluconazole) and amphotericin B (Fig. 8A). Supporting this, unlike the hog1Δ mutant having increased basal expression of ERG11, the srx1Δ mutant has almost wild-type levels of ERG11 expression with or without fluconazole (Fig. 8B). Similarly, both Trx1 and Trx2 seemed to be dispensable for azole drug susceptibility, although deletion of TRX1 slightly increased polyene susceptibility (Fig. 8A). Interestingly, the tsa1Δ mutant showed slightly increased resistance only to fluconazole whereas the tsa3Δ mutant was more susceptible to all azole drugs than wild-type strain (Fig. 8A). However, both tsa1Δ and tsa3Δ mutants exhibited wild-type levels of basal and fluconazole-induced ERG11 (Fig. 8C). Instead, the basal levels of ERG3 appeared to be lower in both tsa3Δ and tsa1Δ tsa3Δ mutants than wild-type strain, implying that Tsa3 may be partly involved in ERG3 expression. Therefore, it is possible that decreased ERG3 expression by tsa3 mutation may further decrease cellular ergosterol content and exacerbate cell viability in addition to the effect by Erg11 inhibition with fluconazole. Taken together, Srx1 is not required for ergosterol biosynthesis and azole/polyene drug susceptibility.

Fig. 8. The role of Prx, Trx, Srx systems in azole and polyene drug susceptibility.

(A) C. neoformans stains [Wild-type (H99), hog1Δ (YSB64), srx1Δ (YSB596), srx1Δ::SRX1 (YSB1214), tsa1Δ (YSB1273), tsa3Δ (YSB1204), tsa1Δ tsa3Δ (YSB2735), trx1Δ (YSB1667), trx2Δ

(YSB1791), and trx1Δ trx2Δ (YSB1796)] were grown overnight at 30°C in liquid YPD medium, 10-fold serially diluted, and spotted on YPD agar containing 0.8 μg/ml amphotericin B (AMB), 14 μg/ml fluconazole (FCZ), 0.04 μg/ml itraconazole (ICZ), and 0.2 μg/ml ketoconazole (KCZ). (B) Northern blot analysis for measuring ERG11 and ERG3 induction by fluconazole treatment. C. neoformans strain [WT (H99) strain and hog1Δ and srx1Δ mutants] grown to mid-logarithmic phase at 30°C in YPD medium were treated with or without 10 μg/ml fluconazole for 90 min and then total RNA was isolated. Each membrane was hybridized with gene specific probes (ERG11 and ERG3), washed and developed. Ethidium bromide staining results of rRNA were used as loading controls.

A strain deleted for SRX1 has wild type levels of capsule, melanin and mating efficiency, but is required for full virulence in a mouse inhalation model of infection

We further examined other Prx-independent roles of Srx1 in C. neoformans, such as production of two major virulence factors, melanin and capsule, and the mating process for generation of infectious propagules. Melanin is an antioxidant, polyphenol complex pigment, which is synthesized by laccases, Lac1 and Lac2, upon carbon starvation. The srx1Δ mutant produced normal levels of melanin as compared to wild-type and other mutants (Supplemental Fig. S3A). Melanin production defects observed in the trx1Δ or trx1Δ trx2Δ mutant appeared to result from their growth defects. Polysaccharide capsule is one of the antiphagocytic molecules deposited on the surface of C. neoformans. As reported before (Missall and Lodge, 2005; Missall et al., 2004), Tsa1, Tsa3, Trx1 and Trx2 were not involved in capsule production (Supplemental Fig. S3B). Similarly, the srx1Δ mutant was as efficient in capsule production as wild-type (Supplemental Fig. S3B). Host infection by C. neoformans appears to be initiated by inhalation of infectious propagules (spores or dried yeasts) through the respiratory tract. Srx1 was dispensable for the mating process of C. neoformans. Upon co-incubation with an opposite mating partner (either wild-type KN99a strain or the a srx1Δ mutant), the srx1Δ mutant showed filamentous growth and formed basidia and basidiospores as efficiently as wild-type (Supplemental Fig. S4). Taken together, Srx1 is dispensable for production of two major virulence factors, capsule and melanin, as well as for generation of spores through mating in C. neoformans.

Regardless of a lack of its role in capsule and melanin synthesis, the finding that Srx1 is required for counteracting peroxide stresses under control by the HOG pathway led us to examine its role in virulence of C. neoformans. When we tested the virulence of the srx1Δ mutant by using nasal inhalation murine model of cryptococcosis, the srx1Δ mutant showed significantly attenuated virulence compared to wild-type strain (H99)(Fig. 9A). In a supportive evidence for the specific role of Srx1 in virulence, the srx1Δ::SRX1 complemented strain exhibited wild-type levels of virulence (Fig. 9A). The srx1Δ mutant constructed in the KN99α strain background also exhibited similar results as the mutants generated in the H99 strain background (Fig. 9B).

Fig. 9. Srx1 is required for full virulence of C. neoformans by promoting its survival and proliferation in the host.

(A and B) Groups of 10 mice (female CBA/J) were infected with each C. neoformans strain through inhalation. Virulence was recorded as mortality of mice, which were monitored for 75 days after infection. Mice that lost 25% of their starting bodyweight were considered to be ill and were sacrificed. The percentage of mice survived was plotted against days post infection. (C) Fungal burden in the mice pulmonary tissue that were infected with 10⁵ of wild type (KN99), tsa1Δ (LCCN 717), srx1Δ (RUCN 1102) and srx1Δ (RUCN 1103) C. neoformans strains as above. At various time points post infection, lungs were harvested and homogenized and serial dilutions were plated for CFU enumeration. Error bars indicate standard error of the mean for three mice per treatment group.

C. neoformans tsa1Δ strains were previously shown to be avirulent in mouse model of cryptococcal infection with reduced fungal burden in the lungs (Missall et al., 2004), suggesting that reduced proliferation or survival of peroxide sensitive mutants resulted in reduced virulence. Therefore we tested whether the srx1Δ mutants also had reduced fungal burden. Figure 9C shows that at 1 and 3 days post infection, fungal burden of tsa1Δ and two independent srx1Δ strains were comparable to that of wild-type, but at 10 days post infection, the fungal burden of tsa1Δ and srx1Δ strains were significantly reduced with at least ten-fold less cells than that of wild-type. This suggests that Srx1 is required for full virulence of C. neoformans by promoting its survival and proliferation in the host.

Discussion

In this study we functionally characterized a sulfiredoxin-like protein, Srx1, in relation to its role in Tsa1 recycling and the HOG-signaling pathway in C. neoformans. We found that Srx1 plays a major role in sensing and responding to hydrogen peroxide and organic peroxide, but not to other types of oxidative or nitrosative damaging agents. We believe that the role of Srx1 in counteracting peroxide stress enables C. neoformans to survive within the host phagocytic cells. Supporting this idea, C. neoformans strains deleted for the SRX1 gene were severely attenuated in virulence. Remarkably we have discovered that Srx1 also has a Prx-independent role in the regulation of fludioxonil-mediated growth arrest without directly controlling intracellular glycerol content. Therefore it appears that Srx1 acts as a defensive molecule during oxidative stress response but as an offensive molecule driving cell death upon treatment with a fungicide (summarized in Fig. 10). In fact the Prx-independent role for Srx1 has been previously documented. The human Srx1 serves as a regulator of the glutathionylation/degluthathionylation (or thiol switch) process, which is one of critical post-translational modification implicated in a number of human diseases, including Parkinson's disease (Findlay et al., 2005; Findlay et al., 2006). To the best of our knowledge, this is the first functional study of the fungal Srx-like protein and its implication in virulence of fungal pathogens.

Fig. 10. The proposed model for the role of sulfiredoxin, thioredoxins, and peroxiredoxins in C. neoformans.

A sulfiredoxin (Srx1) in C. neoformans has two-independent roles in oxidative stress response and fludioxonil-mediated cell death, both of which are mediated by the HOG pathway, consisting of the two-component phosphorelay system containing histidine kinases (Tco proteins), a phosphotransfer protein (Ypd1), and a response regulator (Ssk1), and the MAPK module containing the Ssk2 MAPKKK, the Pbs2 MAPKK (may act as a scaffolding protein), and Hog1 MAPK. In response to peroxide, Tco2 and other unknown sensors appear to activate the HOG pathway and induce SRX1 expression in a dose-dependent manner (Bahn et al., 2006). Srx1 is required for recycling of a peroxiredoxin, Tsa1, which plays a key role in detoxifying incoming peroxide. The oxidized, sulfenic acid (SOH) of Tsa1 is reduced back to the cysteine residue (SH) by the thioredoxin system (Trx1). In the absence of Srx1, Tsa1 can be inactivated by hyperoxidation upon exposure to high levels of peroxide and eventually degraded, which make cells more vulnerable to peroxide stress. Another peroxiredoxin, Tsa3, plays a minor role in defending cells against certain oxidants, including superoxide anions and diamide, and antifungal drugs. The HOG pathway is involved in peroxide adaptation without involvement of Srx1. In response to fludioxonil, which is a phenylpyrrole class of fungicide, Tco1 and Tco2 activate the HOG pathway (Bahn et al., 2006), which activates expression of its target genes, including GPD1 (Ko et al., 2009) and HRK1 (Kim et al., 2011), and triggers over-accumulation of intracellular glycerol. Upon the increased internal turgor pressure, cells become swollen, which causes cytokinesis defect and growth arrest (Kojima et al., 2006). In addition, the HOG pathway activates expression of SRX1, which promotes cell swelling without controlling intracellular glycerol synthesis.

The identity and function of Srx1 was first reported in S. cerevisiae in pioneering work by Biteau et al. (Biteau et al., 2003). In response to peroxide, the N-terminal cysteine thiol of Prxs is oxidized to a unstable sulfenic acid, which either forms a disulfide with another Prx and can be subsequently recycled back to a cysteine thiol by Trx or is further oxidized to more stable sulfinic acid (oxidative inactivation). The latter form of Prx inactivation and its reversion by Srxs are known to be conserved only in eukaryotes, but not in prokaryotes. Cys84 of Srx1 forms a disulfide bond with sulfenic acid of Prx in the ATP-dependent manner (Biteau et al., 2003). We found that the Cys84 residue of S. cerevisiae Srx1 is also highly conserved in C. neoformans Srx1 (Cys106). The S. cerevisiae Srx1 has two additional cysteine residues at position 48 and 106 in addition to the conserved catalytic Cys84 residue (Supplemental Fig. S1). Formation of an intramolecular disulfide bond between Cys48 and active site Cys84 of S. cerevisiae Srx1 has been shown to be an essential step during the reduction of the hyperoxidized sulfinic form of Prx1 cysteine (Roussel et al., 2009). In the subsequent step of catalysis, the disulfide bond between Cys48and Cys84 has been shown to be targeted for reduction and recycling by thioredoxin. Thus in addition to recycling the mildly oxidized active site cysteine of Prx1, S. cerevisiae Trx1 is also required for reducing oxidized Srx1. However, such Trx1-dependent recycling of Srx1 is absent in human and mouse Srx1 catalysis due to the lack of cysteine residues corresponding to Cys48 of S. cerevisiae Srx1. Interestingly the cysteine residue corresponding to Cys48 of S. cerevisiae Srx1 is also absent in the C. neoformans Srx1, suggesting that similarly to mouse and human, C. neoformans Srx1 may depend on reductants other than Trx1 for its catalysis. Moreover, unlike human and mouse Srxs, the C. neoformans Srx1 contains one additional cysteine residue at position 16 (Supplemental Fig. S1), whose functional role during oxidative stress remains to be investigated.

One notable finding in this study is the regulation of Cryptococcus SRX1 by Hog1 in both low and high concentration of H₂O₂. In the fission yeast S. pombe, SRX1 is also transcriptionally induced in response to both low and high H₂O₂ stress by two separate signaling pathways, the Pap1 and the Sty1-Atf1 signaling pathways, respectively (Vivancos et al., 2005). Interestingly, it is not known whether Srx1 is controlled by the HOG pathway in S. cerevisiae. In C. neoformans, SRX1 expression was transiently and weakly induced in response to low levels of H₂O₂ (0.2 mM) whereas it was strongly induced in response to high levels of H₂O₂ (2.5 mM, Fig. 2). The H₂O_2-induced SRX1 expression was abolished in the hog1Δ mutant indicating that the HOG pathway is a major controller of SRX1 expression during both high and low levels of H₂O₂ treatment. The observed dephosphorylation of Hog1 during both mild and strong oxidative stress further confirms its role in oxidative stress resistance. Furthermore the HOG pathway also appears to be critical for maintenance of basal intracellular ROS levels in C. neoformans. We found that the hog1Δ mutant displayed a very high levels of intracellular ROS compared to wild-type (Y.-S. Bahn, unpublished).

An additional novel finding of this study was that C. neoformans Hog1 is involved in inducing the adaptive response to oxidative stress. Even though the SRX1 gene is induced in a Hog1-dependent manner during oxidative stress, its function is not required for triggering an adaptation response. Similar to acute oxidative stress, exposure to mild oxidative stress is also accompanied by dephosphorylation of Hog1 suggesting that Hog1-specific phosphatases may be involved in triggering the adaptation response to oxidative stress. Analysis of the 286 H₂O₂-sensitive deletion strains in S. cerevisiae identified eight genes to be very important for adaptation to oxidative stress. These include Yap1 and Skn7 transcription factors (Ng et al., 2008). The role of Yap1 and Skn7 in adaptive and acute oxidative stress response was also discovered by a adaptation stress screen using the S. cerevisiae gene deletion collection (Kelley and Ideker, 2009). The role of Yap1 and its target genes, such as catalase genes, in inducing adaptive oxidative response was also documented in S. cerevisiae by independent studies (Ouyang et al., 2011; Izawa et al., 1996). Furthermore, additional mechanisms of oxidative stress adaptation were discovered to be dependent on NADPH generating enzyme systems (Izawa et al., 1998). Conversely, in the human fungal pathogen Candida albicans, adaptive oxidative stress response is independent of either Cap1 (a S. cerevisiae Yap1 ortholog) or Hog1, suggesting the presence of additional mechanisms of adaptive response in the pathogen (Gonzalez-Parraga et al., 2010).

Based on data presented by this study and others (Montañés et al, 2011), the repression of ergosterol biosynthesis appears to be a common stress adaptation process among fungi. In S. cerevisiae, ergosterol biosynthesis is highly repressed by downregulation of ERG2 and ERG11 in response to osmotic and oxidative stresses (Montañés et al, 2011). Here our study showed that the similar type of ergosterol biosynthesis repression occurs in C. neoformans in response to oxidative stress. In fact, our previous transcriptome analysis showed that expression of ERG2 and ERG11 genes appear to be downregulated by osmotic stress and fludioxonil treatment in C. neoformans (Ko et al., 2009). Furthermore, we recently found that ERG2 and ERG11 genes are also downregulated during temperature upshift in C. neoformans (unpublished data by YS Bahn). In S. cerevisiae, two main activators, Ecm22 (under normal conditions) and Upc2 (under sterol depletion), control ERG genes (Vik and Rine, 2001; Davies et al., 2005). In response to osmotic stress, the HOG pathway recruits two transcriptional repressors, Mot3 and Rox1, to repress ECM22 and UPC2 and downregulate expression of ERG2 and ERG11. In response to oxidative stress, however, Hog1, Mot3, and Rox1 appear to be only partially involved in ergosterol repression (Montañés et al, 2011). Our present study showed that the HOG pathway plays a significant role in ergosterol repression for oxidative stress adaptation in C. neoformans. However, orthologs to Mot3, Rox1, Ecm22, and Upc2 are not apparent in the Cryptococcus genome. Therefore, it remains to be elucidated how the HOG pathway represses ERG genes in C. neoformans during stress adaptation.

The physiological meaning of ergosterol repression during stress adaptation is not clear at this point. The most likely explanation is that changes in cellular ergosterol content will affect cell membrane fluidity and permeability, which influences the influx and efflux of toxic chemicals or ions. Montañés et al. showed that overexpression of ERG2 and ERG11 increased stress sensitivity and inhibition of ERG11 by azole drugs renders cells resistant to stresses (Montañés et al, 2011). In contrast, some S. cerevisiae ergosterol pathway mutants (erg3Δ and erg6Δ) exhibit hypersensitivity to H₂O₂ due to a decreased permeability barrier to H₂O₂ (Branco et al., 2004). Furthermore, it has been shown that transcriptional induction of ergosterol biosynthesis genes by Mga2 increases the membrane ergosterol content, which in turn limits the diffusion of H₂O₂ across the membrane and renders cells more resistant to subsequent challenge by a higher dose of H₂O₂ (Kelley and Ideker, 2009). In these cases, the accumulation of some ergosterol precursors may affect the plasma membrane permeability. Taken together, it is likely that dynamic regulation of ergosterol biosynthesis appears to be critical for maintaining cellular homeostasis.

In S. pombe a majority of Tpx1 was found to be present in reduced monomer form in actively growing cells and were shifted to oxidized dimer forms upon exposure to exogenous H₂O₂ (Day et al., 2012). In actively growing C. neoformans cells, however, Tsa1 is mainly present both as monomer and dimer. Our immunoblot analysis further confirms that the function of Trx1 in reducing mildly oxidized Tsa1 disulfide dimers is conserved in C. neoformans. The significance of the function of Trx1 in reducing mildly oxidized Tsa1 cysteine residues is evident in the increased sensitivity of trx1Δ strains to exogenous peroxide treatment (Fig.2). The observation that in the absence of Srx1, C. neoformans Tsa1 monomers loses its ability to form homodimers clearly reflects the importance of Srx1 in Tsa1 repair and recycling. Srx1 catalyzes the reduction of sulfinic acid derivative at the expense of ATP, suggesting the importance of ATP during oxidative stress. Supporting this we have recently shown that specific inhibition of mitochondrial ATP synthesis results in the increased sensitivity of C. neoformans cells to exogenous peroxide stress (Upadhya et al., 2013).

The role of Srx1 in resistance to fludioxonil, a phenylpyrrole class of fungicide, is rather unexpected. Originally we thought that artificial induction of SRX1 by fludioxonil treatment, which is known to hyperactivate the HOG pathway, is not physiologically meaningful for actual fludioxonil resistance because it has been suggested that the toxic effects of fludioxonil mainly result from increased intracellular turgor pressure imposed by over-accumulated intracellular glycerols and do not appear to be correlated to the oxidative stress response (Kojima et al., 2006). It could be speculated that fludioxonil-mediated cell cycle arrest may perturb normal cellular redox-balance, in which Srx1 is indirectly involved. Supporting this hypothesis, the tsa1Δ mutant exhibited resistance, albeit to a lesser extent than the srx1Δ mutant, to fludioxonil. Surprisingly, however, the tsa3Δ mutant showed increased susceptibility and the tsa1Δ tsa3Δ double mutant, which lacks two Prxs, was even more susceptible to fludioxonil than the wild-type strain and tsa3Δ mutant, which contrasts with the fludioxonil-resistant srx1Δ mutant. Therefore, Srx1 appears to have some oxidative stress-independent (Prx-independent) roles in mediating the fludioxonil-dependent HOG activation. The findings in this study revealed additional fungicidal mechanism of fludioxonil. It seems that cell swelling and growth defects caused by fludioxonil treatment is not only caused by over-accumulation of intracellular glycerol by hyperactivation of the HOG pathway through its downstream effectors, such as Gpd1 (glycerol-3-phosphate dehydrogenase) and Hrk1 (Hog1-regulated kinase (Kim et al., 2011)), but also by the induction of Srx1 by the HOG pathway (Fig. 10). Deletion of SRX1 conferred a significant resistance to fludioxonil without causing cell swelling, but did not affect intracellular glycerol induction levels, suggesting that Srx1 promotes fludioxonil sensitivity without directly controlling intracellular glycerol synthesis. As an alternative explanation, deletion of SRX1 may counteract increased cellular turgor pressure conferred by enhanced intracellular glycerol content upon fludioxonil treatment. Its detailed mechanism of Srx1-mediated fludioxonil sensitivity needs to be further addressed in future studies.

The role of Srx1 in promoting antifungal drug susceptibility appears to be limited to the case for fludioxonil. Deletion of SRX1 does not affect expression of ERG genes (ERG11 and ERG2) and cellular susceptibility to other clinical drugs, such as azole and polyene drugs (Fig. 8). Instead, the role of Tsa3 in resistance to fludioxonil and azole drugs is notable (Fig. 7B and 8A), because any physiological roles of the second Prx have not been suggested before. At this point, it remains elusive how Tsa3 confers resistance to fludioxonil and azole drugs. However, the finding that basal expression levels of ERG3 are altered by TSA3 deletion suggests that Tsa3 may be directly or indirectly involved in ergosterol biosynthesis. This possibility should be addressed in future studies.

The role of Srx1 in virulence of C. neoformans further supports the importance of this protein for the pathogen to survive within the host during infection. Srx1 plays a significant role in virulence of C. neoformans, albeit to lesser extent than the peroxiredoxin and thioredoxins as per the previously reported virulence data. The tsa1Δ mutant is virtually avirulent whereas the tsa3Δ mutant is as virulent as wild-type (Missall et al., 2004), suggesting that Tsa1, but not Tsa3, plays a major role in virulence of C. neoformans. In contrast, Trx1 and Trx2 play redundant roles in the virulence of C. neoformans. One of the potential reasons for the attenuated virulence of the srx1Δ strains may be due to their decreased ability to survive and proliferate under the host lung conditions as revealed by the fungal burden assay (Fig. 9C).

In conclusion, our present study extends the role of HOG signaling pathway in oxidative stress resistance by establishing the requirement of functional Hog1 for inducing an adaptive response to oxidative stress and controlling the Srx1-mediated Prx recycling system. Our genetic and biochemical data emphasizes the critical importance of Srx1 in Prx-mediated oxidative stress resistance and virulence in C. neoformans. Finally the unique role of C. neoformans Srx1 in contributing to fludioxonil induced growth arrest in a Prx-independent manner suggests a novel role for this antioxidant molecule in fungal biology.

Experimental procedures

Strain and media

The strains used in this study are listed in supplemental Table S1. All C. neoformans strains were cultured in YPD (yeast extract-peptone-dextrose) medium or YNB (yeast nitrogen base without amino acids, 2% glucose) medium. Agar-based DME (Dulbecco modified Eagle) medium for capsule production was prepared by combining filter-sterilized 2× DME liquid medium (pH 7.2, Invitrogen Corp.) with autoclaved 2% agar solution. Mating assays were performed on the V8 mating medium containing 5% V8 juice (Campbell's Soup Co.), 0.5 g/L of KH₂PO₄ and 4% Bacto-agar (Difco) with the pH adjusted to 7.2 before autoclaving. Melanin production was assessed on Niger seed medium containing indicated concentration.

Rapid amplification of cDNA ends (RACE) and coding sequence analysis of SRX1

To make cDNA used in RACE, the wild-type C. neoformans H99 strain was grown in 50 ml YPD medium for 16 hours at 30°C and cell were pelleted at 4°C, frozen in liquid nitrogen, and lyophilized. The total RNA was isolated using Ribo-EX (Geneall). According to the manufacturer's instruction (GeneRacer kit, Invitrogen), cDNAs for RACE were synthesized and used for touchdown PCR and nested PCR with appropriate 5’ and 3’ RACE primers listed in Table S2 and provided by the kit. All PCR products were cloned into pTOP vector (Enzynomics) and sequenced. The sequence for the SRX1 gene from wild-type strain H99 was deposited in Genbank (accession number: JN15977).

Deletion of peroxiredoxins ( TSA1/3), sulfiredoxin (SRX1), and thioredoxin (TRX1/TRX2) genes

All primers necessary for amplification of disruption cassettes, diagnostic PCR screening, and Southern blot probe for each gene are listed in supplemental Table S2. Each disruption cassette was generated by overlap PCR and introduced into C. neoformans serotype A strain H99 (MATα) or KN99a (MATa) by biolistic transformation, as previously described (Kim et al., 2009, Davidson, 2002; Bahn, 2005). The tsa1Δ tsa3Δ double mutants were constructed by introducing the TSA1 disruption cassette marked with nourseothricin resistant marker (NAT) into the tsa3Δ mutant (YSB1204). The trx1Δ trx2Δ double mutants were generated by introducing the NEO-marked TRX2-deletion-cassette into the trx1Δ mutant (YSB1667). All PCR amplifications were performed by using the Ex-Taq polymerase (Takara) and transformants were selected on YPD containing nourseothricin. Each mutant strain was confirmed by diagnostic PCR and Southern blot analysis. The Southern hybridization and analysis were performed as described before (Jung et al., 2011).

Construction of the srx1Δ::SRX1 complemented strain

To verify the phenotypes of the srx1Δ mutants, the srx1Δ::SRX1 complemented strain was constructed as follows. The 2 kb-fragment containing the full-length SRX1 gene was amplified by PCR using primers B2245 and B2246 listed in Table S2 and cloned into pTOP vector (Enzynomics). After sequencing, the SRX1 gene insert was subcloned into pJAF12, which contains NEO (neomycin resistance gene) selectable marker, generating a plasmid pJAF12-SRX1. The pJAF12-SRX1 plasmid was linearized by SnaBl digestion and biolistically transformed into the srx1Δ mutant strain (YSB596). To confirm the targeted re-integration of SRX1 gene through a single cross-over event, diagnostic PCR was performed.

Total RNA isolation and Northern blot analysis

For Northern blot analysis, each strain was grown in YPD medium at 30°C for 16 hr. Then the overnight culture was inoculated at 1:20 dilution into fresh YPD medium and further incubated for about 4 hr at 30°C to 1.0 of an optical density at 600 nm (OD₆₀₀). For zero-time samples, a portion of the culture was used. For time-point stress samples, the remaining medium were added with hydrogen peroxide (H₂O₂) at a final concentration of 2.5 mM or fluconazole at a final concentration of 10 μg/ml and further incubated for the indicated amount of time. Each sampled portion of the culture was frozen in liquid nitrogen for 30 min and lyophilized overnight. The total RNAs were isolated by Ribo-Ex (Geneall) as described before (Jung et al., 2011). The concentration of total RNA samples was measured at OD₂₆₀ and ten μg of each RNA sample was applied to gel electrophoresis. Agarose gel electrophoresis, membrane transfer, hybridization, washing and film development for Northern blot analysis were performed as described before (Jung et al., 2011).

Assay for capsule and melanin production

For capsule induction, each C. neoformans strain was incubated for 16 hr at 30°C in YPD medium, spotted onto the agar-based DME medium, and further incubated for 2 days at 37°C. After incubation, capsule production levels were stained by India ink (Remel BACTIDROP INDIA INK) and observed microscopically. Cell and capsule images were captured by DIC microscopy. To quantitate measurement of capsule, cryptocrit analysis was performed (Alspaugh et al., 2002; Jung et al., 2011). For melanin production, cells were spotted onto Niger seed medium containing 0.1%, 0.5% and 1% glucose. Then, the plates were incubated for up to 7 days at 30°C or 37°C. Melanin production was monitored and photographed daily.

Mating assay

All strains for the mating assay were initially grown in YPD medium for 16 hr at 30°C and resuspended in water. Equal concentrations of MATα and MATa cells (10⁷ cells/ml) were mixed, spotted (5 μl) onto V8 mating medium and incubated in the dark at room temperature for 1-2 weeks. These spots were monitored weekly by observing filamentation morphology and photographed using an Olympus BX51 microscope equipped with STOP insight digital camera.

Sensitivity test for oxidative and nitrosative stress response and antifungal drugs

For oxidative stress sensitivity and antifungal drug susceptibility tests, cells were incubated in YPD medium overnight at 30°C, serially diluted in water, and spotted (3.5 μl) onto solid YPD medium containing indicated concentration of each oxidant, including diamide, menadione, tert-butyl hydroperoxide (tBOOH), paraquat and H₂O₂, or antifungal drugs, including amphotericin B, fluconazole, itraconazole, and ketoconazole. For nitrosative stress sensitivity test, cells grown in YNB (pH 4) liquid medium overnight at 30°C were spotted onto solid YNB medium (pH 4) containing the indicated concentration of NaNO₂ (sodium nitrite). Each plate was incubated for 2-5 days, and photographed during the incubation period.

Oxidative stress adaptation test

For oxidative stress adaptation each strain was grown in 50 ml YPD medium at 30°C for 16 hr. Then 5 ml of the overnight culture was inoculated into 100 ml of fresh YPD medium and further incubated for about 4 hr at 30°C to OD₆₀₀=1.0. For inducing the adaptive response to oxidative stress, the culture medium were pre-treated with 0, 0.2, 0.5, and 1 mM H₂O₂ and incubated for 30 min, 1 hr, and 2 hr. Cells were then washed with 1 ml cold PBS buffer and 3.5 μl of the cell suspension was spotted onto solid YPD medium containing 1, 1.5, 2, and 2.5 mM H₂O₂. Each plate was incubated for 2-4 days and photographed during the incubation period.

Immunoblot analysis of Hog1 phosphorylation

C. neoformans cells were grown to an OD₆₀₀ of 1.0-1.5 in YPD medium at 30°C and were subject to oxidative and osmolarity stress. Yeast cultures were treated with either 0.2 mM or 2.5 mM H₂O₂ to induce mild and acute oxidative stress, respectively. For exposing the cells to salt stress, 1 M NaCl was used. At various time points after the addition of the reagent, 25 ml of cell suspension was withdrawn and diluted with equal volume of ice-cold stop mix (0.9% NaCl, 1 mM NaN₃, 10 mM EDTA, and 50 mM NaF (Kamada et al., 1995; Surana et al., 1991). The cells were harvested at 652 g at 4°C for 10 min and then washed once in ice-cold stop buffer. The cell pellet was resuspended in 1 ml of 1x lysis buffer (50 mM Tris-HCl pH 7.5, 150 mM NaCl, 5 mM EDTA pH 8.2, 5 mM EGTA, 0.2 mM Na₃VO₄, 50 mM KF, 30 mM sodium pyrophosphate, 15 mM p-nitrophenylphosphate, 1x protease inhibitor cocktail (Roche, #11836170001), and 10 μl/ml each phosphatase inhibitor cocktail II and III (Sigma, P-5726 and P0044). Approximately 1 ml of the resuspended cells was beaten for 10 min at 4°C with 0.75 ml of 0.5 mm zirconia/silica beads (BioSpec Products, Inc., #11079105z) on a Disruptor Genie (Scientific Industries, Inc. NY, USA). Lysates were spun down at 20,800 g at 4°C for 20 min and the supernatants collected. Protein concentrations were determined using Quick Start Bradford dye reagent (Bio-Rad, #500-0205). For each sample, 50 μg of total protein were loaded on 10% Mini-PROTEAN TGX precast gels (Bio-Rad, #456-1033) and transferred to nitrocellulose. For phosphorylated Hog1 determination, a 1:10,000 dilution of phospho-p38 MAPK (Thr180/Tyr182) rabbit monoclonal antibody (Cell Signaling Technology, #4511S) was used. Total Hog1 protein independent of phosphorylation was detected using a 1:4,000 dilution of Hog1 specific rabbit polyclonal antibody (Santa cruz Biotechnology, Inc, cat # 25757). Membranes were blocked for 1 hr at 25°C using blocking buffer (25 mM Tris-HCl (pH 8.1), 145 mM NaCl, 0.1% Tween 20 (TBST) containing 10% non-fat dry milk powder). All primary antibodies were diluted in TBST and allowed to bind overnight at 4°C. Secondary antibodies used were goat anti-rabbit immunoglobulin G peroxidase conjugated (Sigma, A-6154) in a dilution of 1:4,000 in TBST and were allowed to bind for 1 hr at 25^oC. Proteins were detected using the ECL Plus Western blotting detection system (GE Healthcare, #RPN2132).

Cloning, expression and purification of recombinant C. neoformans Tsa1 for antibody generation

Full-length cDNA corresponding to C. neoformans TSA1 (CNAG_03482) was amplified using primers (pET28Tsa1For and pET28Tsa1Rev, Table S2) containing Nde1 and Xho1 sites, respectively. Purified PCR product was cloned into pET28b (Novagen, USA) expression vector at Nde1 and Xho1 restriction sites so as to have a N-terminal Histidine tag. After sequence verification a positive clone was transformed into BL21DE3 host cells and protein induction was carried out using Overnight Express Auto-induction system (Novagen cat no 71300) following the procedure supplied with the reagent. One gram (wet wt) of a bacterial cell pellet that was induced overnight was resuspended in 5 ml of Bugbuster protein extraction buffer (Novagen 70584). The cell suspension was treated with 5 μl of benzonase (Novagen cat 70746) and 2 mg of lysozyme for 30 min with shaking. Cell suspension was centrifuged at 14,000 g for 15 min to separate cell debris. The supernatant was collected and used for purification of His-tagged Tsa1 protein using HIS bond purification kit (Novagen cat no 70239) by following the protocol supplied with the regent. Purified protein was confirmed by SDS-polyacrylamide gel and was used for immunization to rabbit host by GenScript, USA, Inc (Piscataway, NJ) according to the manufacturer's protocols. Specificity of the antibodies was confirmed using the lysates from wild-type and tsa1Δ cells.

C. neoformans Tsa1 immunoblot analysis

Wild-type and mutant strains were grown in YPD. At an optical density of 1.0, cells were treated with 6 mM of H₂O₂. At various time points after the addition of H₂O₂, 25 ml cell suspension was withdrawn and centrifuged at 4°C to pellet the cells. Pelleted cells were washed with 20% TCA (trichloro acetic acid) and pelleted kept frozen in dry ice until use. For assaying the role of Srx1 in Tsa1 recycling, cells were resuspended in one ml of 20% TCA and were broken by vortexing them (5 min for a total of 5 times with cooling in-between) in the presence of 0.5 mm zirconia beads (BioSpec products Inc, OK, USA) using Disruptor Genie (Scientific Industries, Inc. NY, USA) at 4°C. Disrupted cells were centrifuged at 15,000 g at 4°C. The pellet was washed with one ml of cold acetone. After the removal of traces of acetone by evaporation, protein pellet was resuspended in 100 mM Tris-HCl (pH 8.8) containing 1% SDS, 20 mM EDTA and 100 mM of iodoacetamide. Tubes were incubated in dark for 20 min. Samples were centrifuged at 12000 g for 10 min at room temperature. Supernatant was collected and used as the whole cell lysate after estimating the protein content by Bio-Rad DC protein assay (Bio-Rad laboratories, USA). Proteins were separated using non-reduced polyacrylamide gels and transferred to nitrocellulose membrane. C. neoformans Tsa1 specific antibodies were used at 1:4000 dilution in TBS-T buffer containing 5% non-fat dry milk powder and 0.2% bovine serum albumin and incubated at room temperature for 1 hr. Secondary antibodies used were goat anti-rabbit immunoglobulin G peroxidase conjugated (Sigma, A-6154) at a dilution of 1:5,000 employing the same conditions used for the primary antibody. A monoclonal antibody specific to human β-actin (ab8224, Abcam, Inc, Cambridge, MA, USA) was used as a loading control at 1:1000 dilution in 5% BSA in TBS-T. Proteins were detected using the ECL Plus Western blotting detection system (GE Healthcare, #RPN2132).

Virulence assay

The wild-type strain (H99 and KN99Δ), srx1Δ mutant (YSB596 and RUCN1103), and its complemented strain (YSB1214 and JLCN872) were grown overnight in YPD broth. The cells were centrifuged and washed with PBS buffer. Virulence studies were performed using a murine inhalation model of infection. CBA/J female mice were infected with 1×10⁵ cells in 0.05 ml PBS buffer. Mice were monitored daily and those showing the signs of being morbidity (weight loss, extension of the cerebral portion of the cranium) were sacrificed by CO₂ asphyxiation. Experiments involving murine models were conducted following the protocol approved by the Institutional Animal Care and Use Committee (IACUC).

Fungal burden assay

Three mice per group (CBA/J female mice ) were infected intranasally with 1×10⁵ of wild type (KN99α), tsa1Δ (LCCN 717) and two independent srx1Δ (RUCN 1102 and RUCN 1103) strains as described above. Lung tissues were harvested at 1 day (d), 3 d, and 10 d post infection (PI). Mice were sacrificed by CO₂ asphyxiation. For the determination of colony forming units (CFU), lungs from each mouse were placed in 2.0 mL of 1x PBS (pH 7), homogenized, serially diluted, plated onto YPD supplemented with 100 mg/mL streptomycin and ampicillin, incubated 2 d at 30°C. CFUs were calculated per mg of lung tissue and expressed per mouse lung.

Statistical analysis

All statistical analyses were performed using GraphPad Prism (GraphPad Software, Inc., San Diego, CA). One-way analysis of variance (ANOVA) was used to analyze differences in lung CFUs using Dunnett's multiple comparison test.