Sterol Regulatory Element Binding Protein (Srb1) Is Required for Hypoxic Adaptation and Virulence in the Dimorphic Fungus Histoplasma capsulatum

The Histoplasma capsulatum sterol regulatory element binding protein (SREBP), Srb1 is a member of the basic helix-loop-helix (bHLH), leucine zipper DNA binding protein family of transcription factors that possess a unique tyrosine (Y) residue instead of an arginine (R) residue in the bHLH region. We have determined that Srb1 message levels increase in a time dependent manner during growth under oxygen deprivation (hypoxia). To further understand the role of Srb1 during infection and hypoxia, we silenced the gene encoding Srb1 using RNA interference (RNAi); characterized the resulting phenotype, determined its response to hypoxia, and its ability to cause disease within an infected host. Silencing of Srb1 resulted in a strain of H. capsulatum that is incapable of surviving in vitro hypoxia. We found that without complete Srb1 expression, H. capsulatum is killed by murine macrophages and avirulent in mice given a lethal dose of yeasts. Additionally, silencing Srb1 inhibited the hypoxic upregulation of other known H. capsulatum hypoxia-responsive genes (HRG), and genes that encode ergosterol biosynthetic enzymes. Consistent with these regulatory functions, Srb1 silenced H. capsulatum cells were hypersensitive to the antifungal azole drug itraconazole. These data support the theory that the H. capsulatum SREBP is critical for hypoxic adaptation and is required for H. capsulatum virulence.


Introduction
Histoplasmosis is a disease that occurs following inhalation of air-borne H. capsulatum infectious spores or conidia [1]. As the fungus infects the host, it frequently encounters a diverse range microenvironmental conditions that can influence fungal morphology and genetic profile. The internal temperature change within the host from 25°C to 37°C induces a physical change from the mycelial form to the pathogenic budding yeast form [2]. Macrophages and dendritic cells are among the first set of cells to defend the host from attacks by the fungus. Within macrophages, yeasts cells are nutrient starved and are exposed to an unfavorable acidic environment. However, they are able to replicate and survive [3]. Therefore, because surviving varying microenvironmental stress conditions is critical for H. capsulatum pathogenesis, it must be able to adapt quickly. The H. capsulatum expression profile in response to temperature [4] nitrosative stress [5] and iron deficiency [6] has confirmed that the ability to survive environmental stressors is tightly regulated at the level of transcription. Thus, this manuscript is focused on understanding the mechanism that H. capsulatum utilizes to survive relativeoxygen deprivation (hypoxia).
In the human pathogenic fungi Aspergillus fumigatus [7] and Cryptococcus neoformans [8,9], and in the fission yeast Schizosaccharomyces pombe [10,11], a sterol regulatory element binding protein (SREBP) has been identified as a key regulator of hypoxia. Fungal SREBPs control the expression of genes involved in the biosynthesis of ergosterol, lipids and heme (reviewed in [12][13][14][15]). Although the manner in which SREBP functions is similar in these fungal species, they are not identical, and much remains to be studied. Under normal oxygen and sterol conditions, the C. neoformans and S. pombe SREBP, Sre1 are membrane bound in a complex with Scp1 (SREBP cleavage activating protein) [8]. However, A. fumigatus contains no functional characterized Scp1 [15]. Periods of low oxygen or decreased sterols trigger the cleavage of fungal SREBPs by Stp1 in C. neoformans, or by via mechanisms involving an ubiquitin ligase and other proteases A. fumigatus and S. pombe [8,11,16]. The N-terminal region is then released and binds to sterol regulatory elements (SRE) in the nucleus, inducing the expression of hypoxic-genes [7,8,11]. Unlike the other fungi, S. pombe contains another regulatory mechanism in the nucleus. When O 2 levels are high Ofd1 binds to Sre1 targeting it for proteasomal degradation; and when O 2 levels are low Ofd1 transcription and protein levels increase, but is inhibited in the nucleus by the nuclear transporter, Nro1 [15,[17][18][19][20][21][22]. Although putative Ofd1 homologs are identified in A. fumigatus and C. neoformans, they do not serve the same function [13,15].
Mammalian SREBPs share common features with fungal SREBPs, and function to regulate the expression of cholesterol uptake, cholesterol synthesis and lipid metabolism genes [23]. Like fungi, mammalian SREBPs require cleavage from the cell membrane for activation and subsequent induction of downstream gene targets [23]. However, activation is sterol dependent, and cleavage from the cell membrane is inhibited when sterols levels are adequate. Mammalian SREBPs directly activate the expression of more than 30 genes required for lipid (LDL receptor and lipoprotein lipase genes) and sterol (Erg family of genes) metabolism, but have not been shown to be influenced by changes in oxygen levels [23][24][25][26][27].
The response to hypoxia in H. capsulatum is the transcriptional induction of its SREBP, Srb1 together with the hypoxia-responsive genes (HRGs), NADP/FAD, ABC, NIT50 and RSP/ GEF. In addition, the genes that encode C-8 sterol isomerase (Erg2) and C-5 sterol desaturase (Erg3) increase under hypoxia in a time-dependent manner [28]. Unlike the other fungi studied, H. capsulatum is a dimorphic fungus that is internalized by the hosts' immune cells and infection results in the formation of a hypoxic granuloma. Despite the biological importance of oxygen to H. capsulatum, little information exists regarding the processes underlying hypoxia tolerance in this dimorphic fungus.
This study is focused on characterizing the function of Srb1 in H. capsulatum in order to advance our understanding of how H. capsulatum adapts to hypoxia. We postulated that H. capsulatum Srb1 may similarly control adaptation to hypoxia and the hypoxic response may be essential for fungal growth and pathogenesis. We report here that Srb1 is necessary for H. capsulatum survival under hypoxia, itraconazole susceptibility and is required for fungal virulence.

Srb1 is necessary for H. capsulatum survival under hypoxia
To determine whether H. capsulatum requires Srb1 for its survival under hypoxia, we silenced the gene encoding Srb1 using RNA interference (RNAi) in the wild type strain, G217B. The telomeric plasmid shuttle vector targeting Srb1 for degradation, was constructed based on the predicted encoding Srb1 gene sequence, and vectors were transformed into H. capsulatum as previously described [29], S1 Fig. This method resulted in a reduction in Srb1 expression ranging from 14% to 63%, rigorously confirmed by qRT-PCR, Fig 1A. The Srb1:RNAi strains were named UC69 (18.8% knockdown), UC71 (63.5% knockdown), UC73 (39.9% knockdown) and UC75 (14.5% knockdown), and two wild-type-vector strains transformed with the vector only vector:RNAi, designated UC70 and UC72. Wild-type-vector strain UC70 and UC71 (silenced by 63.5%) were selected for extensive analysis. Both strains displayed normal growth rates and morphology over 3 days of culture under normoxic conditions, compared to wild type strain G217B in Histoplasma Macrophage Media (HMM) , Fig 2A. Srb1 is required for the expression of other hypoxia-responsive genes H. capsulatum induces the expression of 4 key gene transcripts: NIT50 (DQ350702), an ABC transporter (XM_001544189), a NADP/FAD oxidoreductase (XM_001544225) and an RSP/ GEF (XM_001539669), after 24 hours of in vitro hypoxia within a hypoxia chamber [28]. Therefore, we sought to determine whether Srb1 influenced the upregulation of these hypoxiaresponsive genes (HRG). We first evaluated the levels of Srb1 mRNA expression under hypoxia when Srb1 is silenced by comparing UC71 to wild-type and wild-type-vector strain, UC70. When cultured under hypoxia for 8 hours, the period when Srb1 expression is highest in H. capsulatum, there is an insignificant increase in Srb1 in UC71 compared to wild-type (G217B) and UC70, Fig 1B. This defect in Srb1 upregulation under hypoxia is detected as early as 4 hours after hypoxia treatment and maintained after 12 hours, Fig 1C. This data confirms that RNAi silencing technology was not only able to suppress Srb1 expression under normoxia, but also maintain gene silencing under hypoxia.
To determine the influence of Srb1 silencing on hypoxia-responsive genes, the basal HRG expression of UC71 was compared to wild-type and UC70 under normoxia, and also after 8 hours of hypoxia compared to normoxia. Quantitative RT-PCR (qRT-PCR) analysis of HRGs showed that under normal atmospheric conditions of~21% O 2 , only ABC mRNA expression was significantly lower in Srb1 silenced strain UC71, compared to UC70, Fig 1D. Suggesting that unlike the other HRGs, ABC mRNA is strongly regulated by Srb1 and can influence its transcription even under normoxia. As previously reported, NIT50, ABC, NADP/FAD and RSP/GEF were significantly upregulated during hypoxia in wild-type and wild-type-vector strains [28]. However, when Srb1 is silenced, ABC, NADP/FAD and RSP/ GEF expression was significantly lower than wild-type-vector UC70, after 8 hours of hypoxia, Fig 1E. Though an increase in ABC mRNA was observed in UC70 compared to wildtype under conditions of both normoxia and hypoxia, this difference was not found to be statistically significant, Fig 1D and 1E. This strongly suggests that Srb1 is involved in the regulation of HRG mRNA expression and these genes may be potential downstream targets of the transcription factor.
While we have demonstrated that H. capsulatum remains viable under O 2 concentrations of less than 1% [28], we sought to determine the effect of Srb1 silencing on H. capsulatum during periods of extended hypoxia. Optical density, used to quantify culture turbidity, was used as a measurement of yeast growth. Compared to H. capsulatum growth before hypoxia (time-point 0), after 3 days of hypoxia (1% O 2 , 5% CO 2 , 94% N2), there was no change in wild-type-vector strain UC70, while there was a significant increase in yeast concentration in wild-type-vector strain UC72. In contrast, hypoxia resulted in the killing of Srb1 silenced yeasts UC71 and UC73, silenced by 63.5% and 39.3%, respectively, Fig 2B. When transferred back to oxygen replete conditions after hypoxia incubation, UC70 and UC72 replenished growth, while no growth of UC71 and UC73 was observed, Fig 2B. Our lab has previously demonstrated that  wild type (G217B) H. capsulatum growth is inhibited under hypoxia and displays visual differences in colony size, but not colony number when compared to normoxia [28].
As a secondary method to determine the phenotype of H. capsulatum when Srb1 is silenced, yeast strains were streaked onto solid agar plates to compare colony growth under normoxia to hypoxia. After 11 days of incubation under hypoxia, wild-type-vector strains UC70 and UC72 showed a visible increase in yeast density, while UC71 continued to show limited or no growth and changed to a brownish color, Fig 2C. Not surprisingly, strain UC69 silenced by only 18.8%, was able to survive 11 days of hypoxia, as demonstrated by the noticeable increase in yeast density, Fig 2C. In a third experiment, yeast strains cultured on solid agar under hypoxia for 11 days, were placed in normoxia for an additional 11 days. UC71 growth remained impaired compared to wild-type-vectors, UC70 and UC72, Fig 2D (ii). Strain UC75 (14.5% knockdown) however, continued to replicate and was unaffected by hypoxia much like UC69 (18.8% knockdown), Fig 2D (ii) and 2C.
In parallel, each strain from Fig 2D (i) (11 days of hypoxia), was re-streaked and placed under normoxia or hypoxia for an additional 11 days, Fig 2D (iii) and (iv), respectively. In congruence with growth in liquid HMM, wild-type-vector strains, UC70 and UC72 were able to replenish growth after replacement under normoxia, however only minimal growth was observed in Srb1 silenced strain UC71, Fig 2D (iii). In contrast, after 11 days of hypoxia, wildtype and wild-type-vector strains continued to replicate, while there was no visible growth in UC71, Fig 2D (iv). UC75 maintained a phenotype similar to that of wild-type-vector strains under both conditions. Taken together these data demonstrate that Srb1 silencing has an irreversible effect on H. capsulatum's ability to recover from hypoxia. H. capsulatum can quickly adapt to hypoxia, and Srb1 is involved in mediating this response by an undefined mechanism.

Srb1 silencing inhibits the expression of the ergosterol biosynthetic pathway genes during hypoxic adaptation
To determine if Srb1 is a regulator of genes involved in the ergosterol biosynthetic pathway, we examined the effect of Srb1 silencing on the mRNA expression of genes encoding the enzymes Erg2, Erg3, Erg11 and Erg25 in yeast grown in normoxia compared to 8 hours of hypoxia. These enzymes needed for ergosterol biosynthesis, all require oxygen for their function [30]. There are 2 paralogs of the Erg11 gene of which have been designated Erg11a and Erg11b. Fig 3 shows that upon switching to hypoxia for 8 hours, Erg2, Erg3, Erg11a, and Erg25 was significantly upregulated in wild-type and UC70, but not UC71 cells. Expression of these genes under normal atmospheric conditions was only slightly reduced in UC71 compared to UC70, therefore basal expression was not significantly affected, Fig 3. Interestingly, Srb1 had no effect on Erg11b expression, and its expression remained comparable between all strains under both conditions. This suggests that Srb1 is required for the maximum induction of Erg2, Erg3, Erg11a and Erg25, and that as in mammals, may also have an essential role in regulating the genes necessary for ergosterol biosynthesis.

Loss of Srb1 results in impaired growth H. capsulatum within macrophages
To examine the role of Srb1 during H. capsulatum infection, we tested the ability of Srb1 depleted yeast cells to replicate within macrophages. Bone marrow derived murine macrophages (BMDMs) were infected with wild-type, control vector (UC70) and Srb1 silenced H. capsulatum strains (UC71, UC75) at a multiplicity of infection (MOI) of 0.1 and 1. Two (2) hours following co-incubation, BMDMs were washed with fresh media and either lysed or allowed to incubate for a total of 24 hours. At both 2 and 24 hours post infection, macrophages were lysed and plated to quantify the remaining viable yeast by determining the number of colony-forming units (CFUs). Fig 4A shows that the rate of phagocytosis between strains UC70, UC71, UC75 was similar to wild-type, as equal number of yeasts cells were phagocytosed by the macrophages, observed

Srb1 is required for H. capsulatum virulence
Infectivity studies were undertaken to examine the virulence of Srb1 silenced H. capsulatum, and to determine its importance during the course animal infections. To assay virulence, we inoculated groups of 6 C57BL/6 mice with a lethal dose (2×10 7 ) of either wild-type, wild-typevector (UC70, UC72), or Srb1 silenced strain (UC71, UC73, UC75), intranasally. Wild-type and UC70 challenged mice succumbed to infection by day 13 and 14 post infection, respectively, Fig 5A. In contrast, all mice challenged with either UC71 or UC73 (39.9% knockdown) survived through three weeks, showed no adverse symptoms and surpassed their initial body weight, Fig 5A and 5B. Four (4) out of six (6) mice inoculated with wild-type vector UC72, succumbed to infection by day 10; while two (2) out of six (6) mice infected with UC75 (14.5% knockdown) also died by 12 d.p.i. Macroscopically the lungs of mice infected with the UC71 Srb1 silenced strain, showed almost no signs of infection and appeared pink and spongy. While lungs of mice challenged with either UC70 or wild-type strains were dark red and firm with the appearance of lesions, Fig 5C. The corresponding CFU cultured from four (4) mouse lungs inoculated with UC71 was remarkably lower than those challenged with wild-type H. capsulatum, Fig 5D. In a separate single experiment we observed a decrease in lung fungal burden in mice infected with a sublethal dose (2x10 6 ) of UC71 compared to UC70 at 7 d.p.i., Fig 5E. This data strongly supports the theory that Srb1 is required for H. capsulatum virulence in mice.

Srb1 silencing confers H. capsulatum sensitivity to itraconazole
We have found that without Srb1 expression, induction of 4 oxygen-requiring ergosterol biosynthesis genes (Erg 2, Erg3, Erg11a and Erg25) was impaired during hypoxia. Therefore, we sought to determine whether Srb1 silencing affected the ability of the ergosterol synthesis inhibitor, itraconazole to exert its fungistatic effects on H. capsulatum. Itraconazole targets the ergosterol pathway in fungi particularly the product of the Erg11 gene. Erg11 encodes lanosterol 14-alpha-demethylase, the enzyme that converts lanosterol to ergosterol [31]. We cultured Srb1 silenced strain UC71, WT and wild-type-vector strain UC70 with varying concentrations of itraconazole and measured their growth by culture optical density after 6 days. A concentration of 1.25 ng/ml of intraconazole inhibited wild-type and wild-type-vector growth by 55 and 57% respectively. Surprisingly, this concentration was enough to inhibit UC71 growth by 76%, Fig 6. With 2.5 ng/ml of itraconazole, both wild-type and UC70 growth was inhibited by 61%, while UC71 growth was inhibited by 88%. At 5 ng/ml of itraconazole, a concentration known to completely inhibit H. capsulatum, there was an 81% (wild-type), 83% (UC70) and 91% (UC71) decrease in growth compared to control not treated with itraconazole. Silencing Srb1 dramatically increases the susceptibility of H. capsulatum to itraconazole. The concentration of drug needed to inhibit growth is much lower when Srb1 is silenced. This susceptibility phenotype is similar to that observed with SrbA mutants in A. fumigatus and Sre1 mutants in C. neoformans [32,33].
In Fig 3, we examined the mRNA abundance of Erg11a and Erg11b transcripts when Srb1 is silenced, and a statistically significant decrease in mRNA abundance was observed in Erg11a mRNA UC71 compared to wild type. Erg11b mRNA abundance was also consistently higher in the Srb1 silenced strain than Erg11a. Thus, these results suggest that a decrease in the Erg11a transcript in the UC71 strain may confer the increased susceptibility of this strain to itraconazole and that Srb1 may directly regulate Erg11a transcript abundance.

Discussion
The mechanisms that underlie H. capsulatum gene regulation under hypoxia are complex and intricate. While we have identified a role for Srb1 in this pathway, much remains to me elucidated and studied. Our studies provide evidence that the H. capsulatum sterol regulatory element binding protein, Srb1 is crucial for adaptation to hypoxia in H. capsulatum. When Srb1 expression is silenced by both 63.5% and 39.9%, H. capsulatum cannot survive under oxygen conditions of less than 1%, and is incapable of replenishing growth under normoxia. However, silencing by 18.8% or 14.5% did not result in measurable changes in H. capsulatum growth, Fig  2. Thus, the ability of H. capsulatum to survive hypoxia was dose dependent on the degree of  silencing of Srb1; and effective silencing of Srb1, of more than 39.9%, resulted in a strain that was defective in growth under conditions of hypoxia, but unaffected when under normal atmospheric oxygen conditions (~21% O 2 ). In eukaryotes, control of metabolic processes becomes critical for survival during microenvironmental stress conditions; and under low-oxygen conditions sterol, fatty acids, iron, and heme homeostasis can be severely dysregulated [11,16,30,[34][35][36][37]. This data shows that in order to exert its function, more than 60% of Srb1 expression is necessary for H. capsulatum to survive hypoxia.
Other key players in the hypoxia response network include an ABC transporter, an NADP/ FAD oxidoreductase, an RSP/GEF and NIT50. These genes were previously reported to be significantly upregulated under hypoxia in H. capsulatum, but are yet to be functionally characterized [28]. In this study, upregulation was once again examined under hypoxia and data presented at an 8 hour time point, the timing of maximal upregulation, although smaller amplitude up regulation was noted at 2 and 4 hour timepoints. The 8 hour timepoint may reflect maximal upregulation but may be influenced by factors other than hypoxia such as nutritional deprivation. When Srb1 is silenced hypoxic induction of ABC, NADP/FAD and RSP/GEF, but not NIT50, was inhibited, Fig 1D and 1E. In other eukaryotic organisms, these putative H. capsulatum proteins are known to be involved in the transport of metabolic products, lipids and sterols, and drugs (ABC); ergosterol biosynthesis associated and coordinately regulated with Erg11p (NADP/FAD), responsive to nitrosative stress (NIT50) and guanine nucleotide exchange (RSP/GEF) [5,31,[38][39][40][41][42][43]; all processes essential for basic metabolism. Many of those molecules whose biosynthesis is increased under hypoxia, are likely indirect sensors of oxygen that orchestrate changes in overall fungal metabolism to allow H. capsulatum to adapt to hypoxia [13]. This data supports the hypothesis that Srb1 is involved in the regulation of basic H. capsulatum metabolic processes during hypoxic stress. Nonetheless, how H. capsulatum senses and responds to changing oxygen levels remains an area of investigation.
We provide evidence to support Srb1 as the functional homolog of fungal SREBP, and SREBPs are known to regulate the expression of many genes in different eukaryotes. Mammalian SREBP and C. neoformans Sre1 are transcriptional activators that regulate expression of every sterol biosynthetic enzyme [8,44]. Whereas S. pombe Srel and A. fumigatus SrbA do not control the entire sterol biosynthetic pathway, but predominantly enzymes involved in oxygenrequiring reactions [7,10]. Here, our initial analysis of Srbl-dependent gene expression indicates that H. capsulatum Srbl controls expression of genes required for the oxygen-dependent reactions (ERG2, ERG3, ERG11 and ERG25). Srb1 as a transcriptional regulator may regulate genes under conditions other than hypoxia but the current studies examined only this experimental condition.
Ergosterol is a critical part of the fungal cell membrane and inhibition of its synthesis results increases cellular permeability causing leakage of its cellular contents [45]; and as the name suggests, Srb1 may also be essential for ergosterol homeostasis in H. capsulatum. We found that Srb1 silencing resulted in a strain of H. capsulatum that was more sensitive to the antifungal drug, itraconazole. Yet the mechanism for the increased susceptibility and whether this fungal SREBP mediates ergosterol biosynthesis remains unknown. Itraconazole interacts with 14-α demethylase (ERG11), a cytochrome P-450 enzyme that converts lanosterol to ergosterol [8,13]. The putative NADP/FAD, cytochrome P-450 reductase is also associated and coordinately regulated with Erg11p [31]. This increased sensitivity is likely to be due to the decreased expression of Erg11a and NADP/FAD in the absence of Srb1.
Silencing Srb1 expression in H. capsulatum allowed us to study whether it is necessary for virulence and progression of murine histoplasmosis. Disrupting Srb1 expression resulted in an avirulent H. capsulatum strain that cannot effectively replicate within the macrophage and is not pathogenic in mice. The H. capsulatum strain silenced by 63.5% (UC71), exhibited enhanced susceptibility to killing by macrophages compared to wild-type. It is well established that once internalized H. capsulatum is able to replicate within unactivated macrophages [46][47][48], but we were not able to detect a significant increase in wild-type yeast replication within BMDMs, Fig 4B and 4C. The lack of replication if wild-type yeast may be related to the derivation of macrophages by using GM-CSF to differentiate bone-marrow monocytes, as GM-CSF, if incubated with monocytes for extended periods can stimulate macrophage fungistatic activity and inhibit H. capsulatum replication [49].
Mice infected with UC71 showed no signs of illness and survived until sacrifice at 21 d.p.i. while wild-type infected mice died by day 14. In addition, lungs extracted from UC71 infected mice appeared healthy compared to wild-type and wild-type-vector infected mice. It is possible that this virulence defect in Srb1 silenced strains is due to the presence of hypoxic microenvironments in host tissue during infection, in which H. capsulatum is not able to adapt. Though not tested, we postulate that without Srb1 there is a decrease in ergosterol production, necessary for cell wall and membrane maintenance as a result of differences in Srb1-dependent gene expression between wild-type and UC71.
While Srb1 appears to be a master regulator of hypoxia in H. capsulatum, we also hypothesize that a complex network of proteins and enzymes may also be necessary for adaptation to hypoxia. Ofd1 regulates both DNA binding and the degradation of Sre1 in S. pombe in an oxygen-dependent manner. In the presence of oxygen, Ofd1 binds to Sre1 and keeps Sre1 from binding to DNA by accelerating Sre1 degradation [10,50]. During low oxygen tension, Ofd1-Sre1 binding is inhibited by Nro1, slowing down Sre1 degradation and allowing it to bind to DNA and upregulate hypoxic gene transcription [17,19]. An Ofd1 homolog identified in H. capsulatum may be one such interactive proteins and its expression is upregulated under hypoxia (data not shown). The transcriptional regulation and role of Ofd1 in the hypoxic response differs between S. pombe and A. fumigatus and C. neoformans [51].These studies indicate that the function of Srb1 as a hypoxic transcription factor is conserved across fungi in both ascomycetes (S. pombe, A. fumigatus) and a basidiomycete (C. neoformans). Our data provides evidence that H. capsulatum, a dimorphic intracellular pathogen, has evolved a complex pathway to survive not only nutrient and mineral deficiencies but also changes in oxygen levels. The mechanisms that underlie H. capsulatum gene regulation under hypoxia are complex and intricate. While we have identified a role for Srb1 in this pathway, much remains to be elucidated and studied.

Strains and Culture Conditions
All Histoplasma capsulatum strains utilized in this study were derived from the ATCC Histoplasma capsulatum wild-type strain G217B (ATCC # 26032) and are listed in Table 1. H. capsulatum strains were maintained on Histoplasma-Macrophage Medium (HMM) [52], at 37°C with 5% CO 2 . For solid medium HMM was solidified with 0.8% agarose (Lonza SeaKem ME, Pittsburgh, PA, USA) and supplemented with 25 M FeSO4. Liquid cultures were grown until log phase in an orbital shaker with continuous shaking for 72 hours at 200 rpm. For growth of blasticidin resistant strains, plates and media were supplemented with 100 μg/mL blasticidin S (Invivogen, San Diego, CA, USA) where appropriate. Normoxic conditions were deemed general atmospheric levels within the incubator (~21% O 2 ). Hypoxic conditions (<1% O 2 ), were achieved using a controlled atmospheric chamber (BioSpherix C-Chamber, Winston-Salem, NC, USA) with O 2 and CO 2 levels controlled by an external controller. The chamber was maintained at 37°C and kept at~1% O 2 utilizing a gas mixture of 94% N 2 and 5% CO 2 . Growth rates of RNA interference (RNAi) yeast strains were determined in liquid media, by measurement of culture turbidity at an absorbance of 590 or 600 nm. Precise enumeration of yeasts was done by hemocytometer counts.
For itraconazole experiments, all yeast strains were cultured under normoxia within a shaking incubator at 37°C for 6 days. Yeasts were grown in media containing HMM supplemented with 1.25, 2.5, and 5 ng/ml of intraconazole dissolved in DMSO; and HMM with DMSO served as the no treatment control. Optical density was used to as a measure of yeast growth and experiments were conducted in triplicate. UC69, UC71, UC73 and UC75 (Srb1:RNAi). H. capsulatum Srb1:RNAi strains were constructed in G217B wild-type strain using a hairpin-loop telomeric silencing plasmid to knockdown the expression of Srb1 (S1 Fig). An H. capsulatum shuttle vector pSK-Tel-Kan-Blast-186 was previously constructed by fusion of the pSKII+ backbone containing the origin of replication and multiple cloning sites, with a fragment from pCR83 containing H. capsulatum telomere sequence repeats flanking the kanamycin resistance cassette, a fragment containing the Aspergillus terreus blasticidin deaminase gene BSD under control of the gpd promoter flanked by the Aspergillus nidulans trpC terminator, and a fragment of pCR186 (a kind gift from Chad Rappleye) containing a gfp silencing cassette under the control of the H. capsulatum pH2AB promoter [53,54]. Restriction endonuclease digestion and ligation was used to replace the gfp gene with Srb1. A 470 bp fragment of the Srb1 coding region with compatible restriction site end sequences, was generated by PCR and cloned into the plasmid constructs in forward and reverse directions using relevant primers Table 2. The resulting plasmid containing the Srb1 cDNA under control of the H2AB promoter was linearized through PacI restriction digestion and transformed into H. capsulatum G217B by electroporation under standard conditions [55]. Transformants were subsequently cultured in liquid HMM media for 24 hours then passaged in HMM supplemented with 100 μg/mL blasticidin S and plated on HMM-blasticidin solid media. Individual colony transformants were replica plated onto HMM-blasticidin. Plasmid transformed yeast were selected based on resistance to the antibiotic blasticidin S due to the presence of the plasmid encoding blasticidin resistance gene, BSD, which confers resistance in H. capsulatum. The loss of the srb1 gene was verified by quantitative Real-Time PCR (qRT-PCR) using Srb1 specific primers in clones only able to grow in the presence of blasticidin S.

Strain Construction
UC70 and UC72 (Wild-type-vectors). Using the same protocol described above the H. capsulatum shuttle vector pSK-Tel-Kan-Blast-186 was constructed without the Srb1 silencing cassette. The empty vector plasmid was linearized with PacI restriction endonuclease, electroporated into H. capsulatum G217B and cultured in HMM for 24 hours [54]. Transformants with plasmid vectors were selected based on their ability to grow on HMM plates containing 100 μg/mL blasticidin S.

Yeast RNA isolation and Quantitative Real-Time PCR
H. capsulatum yeast cells were collected from 5 mL HMM liquid cultures by centrifugation for 10 min at 1,600 rpm. Total RNA was extracted from fresh cell pellets using a MasterPure Yeast RNA Purification Kit (Epicentre Biotechnologies, Madison, WI, USA) following the manufacturer's instructions, and treated with DNase I for removal of contaminating genomic DNA. RNA concentration and purity was determined using a NanoDrop ND-1000 spectrophotometer (ThermoScientific, Florence, KY, USA), at an absorbance ratio of 260 and 280 nm. cDNA was synthesized using 1 μg of RNA reverse transcribed using SuperScript II reverse transcriptase (Life Technologies, Grand Island, NY, USA), random primer mix and deoxynucleotide triphosphate (dNTPs). For qRT-PCR reactions templates were diluted 1:80 in a PCR mix containing each gene-specific primer pair and SYBR Green dye (Life Technologies, Grand Island, NY, USA). To ensure accuracy of cDNA amplification, at least one primer of each pair was generated to span an intron, and their inability to amplify genomic DNA was confirmed before use. Standard curves were generated for each primer set using 5-fold serial dilutions of pooled cDNA. Gene expression was measured on Applied Biosystems 7500 FAST PCR machine (Applied Biosystems, Foster City, CA, USA). For each sample, values obtained were normalized to the levels of glyceraldehyde 3-phosphate dehydrogenase (GAPDH), and RNA levels were calculated using the standard curve ddCT method. Oligonucleotide primers used for qRT-PCR study are listed in Table 2.

Intranasal Mouse Infection
Six (6) week old C57BL/6 male mice purchased from Charles River Laboratories, Inc., (Wilmington, MA, USA) were used for infection. Mice were housed and maintained in micro-isolator cages in a barrier facility, under pathogen-free conditions, and supplied with sterilized food and water. All animal procedures were approved by the Institutional Animal Care and Use Committee of the Cincinnati VA Medical Center under protocol # 14-10-03-01. As described in the animal protocol, mice with histoplasmosis develop some level of distress manifest by ruffled fur, reduced activity and tachypnea. Animals infected with sub-lethal inoculum spontaneously recover from the infection over a 2-3 week period. Animals administered a larger lethal inoculum or animals with progressive infection manifest worsening tachypnea, severe hunching, and immobility. Animals were monitored at least twice each day to identify animals meeting a humane endpoint allowing sacrifice prior to spontaneous death. Mice with severe tachypnea, severe hunching or immobility in whom it was thought death may be imminent were euthanized. Euthanasia was performed using CO 2 exposure in a gradual fill enclosure from a compressed CO 2 cylinder followed by cervical dislocation as a secondary method. No analgesia was provided as it has not been deemed capable of relieving the distress seen with active infection. In the course of the experiments 25 of 185 animals died as a result of the Histoplasma challenge without sacrifice despite twice daily close monitoring. The death rate was higher than anticipated and that recorded in prior studies. All of these animals were determined to have died as a result of severe Histoplasma pneumonia. Mice were infected intranasally (i.n.) with a lethal dose (2 x 10 7 ) mid-exponential phase H. capsulatum yeast cells resuspended in 20 μL PBS. Groups of six (6) mice were infected with either G217B, UC70, UC71, UC72, UC73, UC75 strain. To avoid yeast clumping, all yeast cultures were vortexed prior to infection. The accuracy of infection was measured by determination of colony forming units of infection cultures (CFUs). Mice were weighed daily until 21 days post infection or until they succumbed to infection and were sacrificed.
In a separate experiment, lungs from four (4) lethally infected animals were removed aseptically before sacrifice 14 d.p.i., photographed and then homogenized for enumeration of fungal burden by CFU/lung, Fig 5D. For sublethal experiments, groups of 4 mice were infected intranasally (i.n.) with (2 x 10 6 ) mid-exponential phase H. capsulatum yeast cells UC70 and UC71 resuspended in 20 μL PBS, Fig 5E. Lungs were removed, homogenized and plated on HMM at 7 d.p.i for the enumeration of surviving CFU.

Macrophage Infection
Bone marrow derived macrophages (BMDM) isolated from C57BL/6 mice were infected with 5 x 10 5 or 5 x 10 4 mid-log phase H. capsulatum yeast cells in DMEM (Dulbecco's Modified Eagle Medium, Gibco, NY). BMDMs were harvested and cultured following standard protocol using GM-CSF [56]. Macrophages (5× 10 5 ) were placed in 48-well plates and kept at 37°C in a 5% CO 2 incubator overnight. Non-adherent cells were removed by aspiration and washing 3× with DMEM, and the plates kept at 37 0 C in a 5% CO 2 incubator. For infections, H. capsulatum yeast cells were grown to log phase in HMM or HMM with 100 μg/mL blasticidin S where appropriate. BMDMs were pelleted and washed 3x in sterile PBS, resuspended in DMEM supplemented with 10% fetal bovine serum, then vortexed twice for 5 seconds, and counted by hemocytometer. Macrophages were infected with H. capsulatum at a multiplicity of infection (MOI) of 1 yeasts per macrophage (1:1) or 0.1 yeasts per macrophage (0.1:1), yeast to macrophage ratio. Approximately 5 ×10 5 (in 100 uL) H. capsulatum cells were used to infect 5 ×10 5 BMDMs in 48-well cell culture dishes for 2 hours in DMEM supplemented with 10% fetal bovine serum, and gentamicin. After 2 hours non-adherent yeasts were removed by aspiration and the BMDM monolayer washed with DMEM 3x. Two groups of macrophages were tested as follows: a) to determine the number of yeasts phagocytosed after 2 hours the infected cells were harvested and lysed in sterile water by repetitive pipetting, b) fresh medium was added to each well for an additional 24 hours. After which, the culture medium was removed and the BMDM washed 3x with medium, then lysed in sterile water by repetitive pipetting. The lysates were centrifuged and yeasts were suspended in 1 mL of PBS or sterile water. Dilutions were plated onto HMM plates and incubated at 37°C for 1 week to determine CFUs.

Statistical Analysis
All experiments were performed in duplicate or triplicate. All statistical analyses were performed using GraphPad Prism for Windows. The results are expressed as means ± S.E.M. The statistical significance between the control and experimental data was calculated using one-way or two way analysis of variance (ANOVA) with paired comparisons. A p-value of less than 0.05 was considered to be statistically significant. For mouse survival, log rank (Mantel Cox) test was performed on survival curves.

Ethics Statement
All animal studies were carried out in accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health and with the prior approval of the Cincinnati VA Medical Center Institutional Animal Care and Use Committee under protocol #14-10-03-01.