Activation of the heat shock transcription factor Hsf1 is essential for the full virulence of the fungal pathogen Candida albicans

Abstract

The evolutionarily conserved heat shock transcription factor Hsf1 plays a central role in thermal adaptation in the major fungal pathogen of humans, Candida albicans. Hsf1 becomes hyperphosphorylated in response to heat shock and activates the transcription of genes with heat shock elements (HSEs) in their promoters, these genes contributing to thermal adaptation. However, the relevance of Hsf1 activation to C. albicans virulence is not clear as this pathogen is thought to be obligately associated with warm blooded animals, and this issue has not been tested because HSF1 is essential for viability in C. albicans. In this study, we demonstrate that the HSE regulon is active in C. albicans cells infecting the kidney. We also show the CE2 region of Hsf1 is required for activation and that the phosphorylation of specific residues in this domain contributes to Hsf1 activation. C. albicans HSF1 mutants that lack this CE2 region are viable. However, they are unable to activate HSE-containing genes in response to heat shock, and they are thermosensitive. Using this HSF1 CE2 deletion mutant we demonstrate that Hsf1 activation, and hence thermal adaptation, contributes significantly to the virulence of C. albicans.

Fig. 1

Construction of C. albicans HSF1 mutants. The top line illustrates the domain structure of C. albicans Hsf1, roughly to scale, by analogy to S. cerevisiae Hsf1. Hsf1 domains include two transcriptional activation regions (AR1 and AR2), a DNA-binding domain (DBD), a trimerisation region (TRI), a control element (CE2) and a carboxy-terminal region (CTM). Schematic representations of C. albicans mutations created in this study lie below the wild type Hsf1 structure (Hsf1). Truncations of the CTM region (CTMt) and all of the carboxy-terminus up to and including the CE2 region (CE2t) were created at the HSF1 locus in an HSF1/hsf1 heterozygote, these constructs being expressed from the endogenous HSF1 promoter. Hsf1 phosphorylation sites in the CE2 region are highlighted in yellow. BamHI sites were introduced on each site of the CE2 region to create the wild type control (SDM-WT) for the subsequent site directed mutagenesis of these sites to alanine, in blue (SDM-A) or glutamate, in red (SDM-E). The CE2 region was also deleted from SDM-WT to create the mutation ΔCE2. These constructs were expressed from the ACT1 promoter (Section 2).

Fig. 2

Removal of the CE2 and AR2 regions attenuates the ability of Hsf1 to activate the HSE regulon in response to heat shock and confers thermosensitivity. (A) Northern analysis of the HSP70, HSP90 and HSP140 mRNAs using ACT1 as an internal loading control. RNA was isolated from C. albicans strains hsf1/HSF1 (CLM61-1), CE2t (SN127) and CTMt (SN128) (Table 1) grown at 30 °C in YPD to exponential phase and then subjected to a 30 to 45 °C heat shock for 30 min, or maintained at 30 °C (control). (B) Activation of the HSE-lacZ reporter was measured by assaying β-galactosidase activities in heat shocked C. albicans hsf1/HSF1, CE2t and CTMt cells expressing either the basal-lacZ or HSE-lacZ reporter (SN65, SN66, SN138, SN141, SN148, and SN151; Table 1). Fold-induction in HSE-lacZ expressing cells is shown relative to equivalent cells expressing the basal-lacZ reporter. (C) Impact of Hsf1 truncations upon thermotolerance. C. albicans cells were serially diluted, spotted onto YPD, and incubated overnight at 30 °C, 37 °C or 42 °C: HSF1, hsf1/HSF1 (CLM61-1); CE2t (SN127); CTMt (SN128).

Fig. 3

Influence of Hsf1 phosphorylation sites upon Hsf1 phosphorylation and thermotolerance in C. albicans. (A) Mutagenesis of putative phosphorylation sites at S571, T575, S577 and S578 does not prevent Hsf1 phosphorylation. C. albicans cells expressing different FLAG-tagged Hsf1 mutants were subjected to a 30 to 45 °C heat shock or maintained at 30 °C, and protein extracts analysed by Western blotting with an anti-FLAG antibody: SDM-WT, SN250; SDM-A, SN251; SDM-E, SN252 (Table 1). Some extracts were also treated with λ phosphatase to confirm that the observed band shifts represented Hsf1 phosphorylation. Bands corresponding to phosphorylated (Hsf1-P) and unphosphorylated Hsf1 are highlighted by arrows on the right. (B) Mutations in the Hsf1 CE2 domain affect thermotolerance. C. albicans hsf1/tet-HSF1 cells (CLM62-1) transformed with various plasmids were grown for 6 h in the presence of 20 μg/ml doxycycline to down-regulate Hsf1, subjected to a 30 to 45 °C heat shock for 30 min and then viability assayed by plating on YPD and determining colony forming units (CFU): SDM-WT (SN254); SDM-A (SN255); SDM-E (SN256); ΔCE2 (SN257) (Table 1). Viability was measured relative to heat stressed control cells (SDM-WT, SN254). Means and standard deviations from triplicate experiments are shown: ^*p < 0.05; ^**p < 0.01.

Fig. 4

Deletion of the CE2 domain blocks detectable Hsf1 phosphorylation and renders C. albicans thermosensitive. (A) CE2 deletion blocks Hsf1 phosphorylation. C. albicans cells expressing FLAG-tagged versions of Hsf1 were heat shocked or maintained at 30 °C, and protein extracts examined by western blotting: SDM-WT, SN250; ΔCE2, SN253 (Table 1). (B) The CE2 domain is required for thermotolerance. Serially diluted C. albicans hsf1/tet-HSF1 cells (CLM62-1) expressing different HSF1 alleles, or the empty pACT1 vector as a control, were spotted onto YPD containing doxycycline and incubated overnight at 30 °C or 37 °C.

Fig. 5

Expression of HSE-containing genes in C. albicans cells from infected mouse kidneys. HSP90, HSP104, lacZ and ENO1 mRNA levels were measured relative to the internal ACT1 mRNA control by qRT-PCR. (A) Control experiments assaying transcript levels in C. albicans SN1 cells (expressing HSE-lacZ) grown in vitro in YPD at different temperatures. Error bars are from triplicate assays from two independent qRT-PCR analyses on each sample. Similar results were obtained for two independent experiments. (B) C. albicans mRNA levels in cells from infected kidneys of mice with systemic candidiasis: W1–W3, three different mice infected with SC5314; B1–B3, three mice infected with SN1 (expressing basal-lacZ); H1–H6, six mice infected with SN2 (expressing HSE-lacZ). The time post-infection (in days) that each mouse was sacrificed is indicated.

Fig. 6

C. albicans mutants that are unable to activate Hsf1 display attenuated virulence in the mouse model of systemic candidiasis. (A) Kidney burdens measured at day 3: WT, SC5314; HSE-lacZ (SN2); CTMt (SN128); CE2t (SN127) (Table 1). (B) Infection Outcome Scores calculated after three days (means and standard deviation for data from six animals): ^∗p < 0.05; ^∗∗p < 0.01. Higher outcome scores reflect more severe infection.