A link between virulence and homeostatic responses to hypoxia during infection by the human fungal pathogen Cryptococcus neoformans

Abstract

Fungal pathogens of humans require molecular oxygen for several essential biochemical reactions, yet virtually nothing is known about how they adapt to the relatively hypoxic environment of infected tissues. We isolated mutants defective in growth under hypoxic conditions, but normal for growth in normoxic conditions, in Cryptococcus neoformans, the most common cause of fungal meningitis. Two regulatory pathways were identified: one homologous to the mammalian sterol-response element binding protein (SREBP) cholesterol biosynthesis regulatory pathway, and the other a two-component-like pathway involving a fungal-specific hybrid histidine kinase family member, Tco1. We show that cleavage of the SREBP precursor homolog Sre1-which is predicted to release its DNA-binding domain from the membrane-occurs in response to hypoxia, and that Sre1 is required for hypoxic induction of genes encoding for oxygen-dependent enzymes involved in ergosterol synthesis. Importantly, mutants in either the SREBP pathway or the Tco1 pathway display defects in their ability to proliferate in host tissues and to cause disease in infected mice, linking for the first time to our knowledge hypoxic adaptation and pathogenesis by a eukaryotic aerobe. SREBP pathway mutants were found to be a hundred times more sensitive than wild-type to fluconazole, a widely used antifungal agent that inhibits ergosterol synthesis, suggesting that inhibitors of SREBP processing could substantially enhance the potency of current therapies.

Figure 1. Mutants in SREBP Pathway and Tco1 Are Sensitive to Hypoxia

Growth in normoxic and hypoxic conditions. Cultures diluted to OD_600nm = 0.6 were diluted serially in 10-fold increments prior to being spotted onto YPD plates. Plates were incubated in normoxic or hypoxic (controlled atmosphere chamber; less than 0.2% oxygen) conditions in the dark at 37 °C. Top and bottom panels are from the same plates, middle panels are from plates grown under the same conditions. WT, wild-type.

Figure 2. Predicted Protein Domains for SREBP Pathway Component Orthologs and Tco1 Histidine Kinase Family Members

(A) Protein domains as predicted by SMART (http://smart.embl-heidelberg.de) for C. neoformans Sre1, Scp1, and Stp1. Transmembrane segments for Sre1 were predicted using MEMSAT3 (http://bioinf.cs.ucl.ac.uk). Arrowhead indicates site of 3×FLAG insertion into Sre1. bHLH, bHLH DNA-binding domain; peptidase, peptidase M50 family domain. (B) ClustalW sequence alignment between C. neoformans Sre1 and its orthologs in C. albicans, Ustilago maydis, S. pombe, and human (SREBP-1A), depicting a portion of the predicted DNA-binding domain. All five contain a conserved tyrosine residue (indicated by asterisk) specific to the SREBP family of bHLH transcription factors. (C) Domains as predicted by SMART for Tco1 homologs. C. neoformans Tco1, and orthologs to Tco1 from B. dermatiditis, H. capsulatum, C. albicans, and Neurospora crassa. HATPase, histidine kinase ATPase domain; HisKA, histidine kinase A domain; REC, cheY-like homologous receiver domain.

Figure 3. Phenotypic Characterization of Mutants in the SREBP Pathway and in TCO1

(A) Growth in YNB medium at 37 °C. Growth was monitored via OD₆₀₀ measurements. Data shown are an average of three independent cultures for each strain, and error bars represent standard deviations (SD). (B) Capsule assays. Above, capsule synthesis was induced in wild-type (WT), sre1Δ-1, and tco1Δ-1 strains, and subsequently visualized using India ink staining. Images were taken at 160× magnification. Below, quantification of the ratio of capsule diameter to cell diameter for capsule-induced cells from the indicated strains. Data shown represent the mean ± SD, n ≥ 30. p-Values (asterisks) were derived using Student's t-test. (C) Melanin assays. The indicated strains were grown to saturation and spotted onto L-DOPA–containing medium. The plates were then cultured at 37 °C in the dark.

Figure 4. FLAG-Sre1 Is Processed in Low Oxygen Conditions

Proteins extracts from wild-type and two independent cultures of FLAG-SRE1 grown in normoxic and hypoxic (bubbling mixed gas into flasks; 0.2% oxygen) conditions were fractionated by SDS-PAGE and immunoblotted with antibodies against the FLAG epitope. Immunoblotting with antibodies against the cyclin-dependent protein kinase PSTAIRE was used to control for variation in loading.

Figure 5. Microarray Hybridization and RT-QPCR Studies Demonstrate That a Subset of Genes Involved in the Hypoxia Response Require Sre1 for Their Regulation

(A) Summary of profiling transcriptional response to hypoxia in sre1Δ and wild-type (WT) cells. cDNA from wild-type cells cultivated in normoxic conditions was hybridized against cDNA from wild-type cells exposed to hypoxic conditions on arrays containing 6,846 features from the genome of C. neoformans. Similarly, cDNA from sre1Δ-1 grown in normoxia was hybridized against cDNA from sre1Δ-1 grown in hypoxia. Statistical analysis of the resulting arrays was conducted using the software SAM. The transcriptional response to hypoxia in wild-type was identified as the set of genes with statistically significant changes in gene expression as determined by SAM. A cutoff of 2-fold or greater changes in gene expression was imposed upon this gene set to determine the number of genes up- and down-regulated in wild-type in response to hypoxia. The arrays hybridized with cDNA from sre1Δ-1 were compared to those hybridized with wild-type cDNA to determine which hypoxia-induced transcriptional changes were significantly different between sre1Δ-1 and wild-type. The resulting gene set is summarized above, where relative degree of induction/repression in sre1Δ-1 and wild-type were compared by dividing fold-change in expression in sre1Δ-1 with fold-change in expression in wild-type. (B) RT-QPCR. cDNA from sre1Δ-1 and wild-type cells grown in normoxia and hypoxia were amplified using primers against ERG1, ERG3, ERG5, and ACT1. Values obtained for ERG1, ERG3, and ERG5 were normalized against ACT1 for each sample to give relative expression, and then expression for the three genes were normalized to their expression in wild-type in normoxic conditions. Error bars represent standard deviation across four samples.

Figure 6. Hypoxia-Sensitive Mutants Display Proliferation and Virulence Defects

(A) Tail vein inoculation competition experiments. Approximately 1:1 mixtures of wild-type (WT) and the indicated mutant (marked with a NAT resistance gene) were injected into mice (A/J) intravenously via the tail vein (2 × 10⁶ total cells/mouse). The actual proportion of mutant cells in each inoculum were determined by plating a dilution of the inoculum on nonselective medium and then assaying 100–200 individual colonies for NAT resistance. At 10 d post-infection, animals were sacrificed and the lungs, brains, and spleen from each animal were homogenized and serial dilutions were plated. Then, 100–200 colonies per organ were assayed for NAT resistance to determine the percentage of mutant cells. Error bars represent the standard deviation from four mice per inoculum. (B) Intranasal inoculation competition experiments. Approximately 1:1 mixtures of wild-type and the indicated mutant were inoculated intranasally into mice (A/J) (5 × 10⁵ total cells/mouse). The actual proportions of mutant cells were determined as in (A). At 21 d post-infection, animals were sacrificed and the lungs from each animal were treated as in (A). Error bars represent the standard deviation from four mice per inoculum. (C) Virulence assays. Eight to ten mice (A/J) were injected intravenously via the tail vein with 2 × 10⁵ cells of the indicated strain and progression to severe morbidity was monitored.

Figure 7. Mutants in the SREBP Pathway, but Not the tco1Δ Mutant, Display Hypersensitivity to Fluconazole

Cells (10⁴ cells/ml) of wild-type and each mutant strain were cultured in YNB for three d in 96-well format in the presence of 0.0045–30 μg/ml fluconazole, after which relative growth was assessed using OD₆₀₀ measurements. Data shown are an average of at least three independent cultures for each strain. Error bars represent standard deviations.