Hsp90 orchestrates transcriptional regulation by Hsf1 and cell wall remodelling by MAPK signalling during thermal adaptation in a pathogenic yeast

Abstract

Thermal adaptation is essential in all organisms. In yeasts, the heat shock response is commanded by the heat shock transcription factor Hsf1. Here we have integrated unbiased genetic screens with directed molecular dissection to demonstrate that multiple signalling cascades contribute to thermal adaptation in the pathogenic yeast Candida albicans. We show that the molecular chaperone heat shock protein 90 (Hsp90) interacts with and down-regulates Hsf1 thereby modulating short term thermal adaptation. In the longer term, thermal adaptation depends on key MAP kinase signalling pathways that are associated with cell wall remodelling: the Hog1, Mkc1 and Cek1 pathways. We demonstrate that these pathways are differentially activated and display cross talk during heat shock. As a result ambient temperature significantly affects the resistance of C. albicans cells to cell wall stresses (Calcofluor White and Congo Red), but not osmotic stress (NaCl). We also show that the inactivation of MAP kinase signalling disrupts this cross talk between thermal and cell wall adaptation. Critically, Hsp90 coordinates this cross talk. Genetic and pharmacological inhibition of Hsp90 disrupts the Hsf1-Hsp90 regulatory circuit thereby disturbing HSP gene regulation and reducing the resistance of C. albicans to proteotoxic stresses. Hsp90 depletion also affects cell wall biogenesis by impairing the activation of its client proteins Mkc1 and Hog1, as well as Cek1, which we implicate as a new Hsp90 client in this study. Therefore Hsp90 modulates the short term Hsf1-mediated activation of the classic heat shock response, coordinating this response with long term thermal adaptation via Mkc1- Hog1- and Cek1-mediated cell wall remodelling.

Figure 1. Hsf1 down-regulates Hsp90.

(A) Inhibiting Hsp90 with geldanamycin leads to Hsf1 activation. C. albicans cells were incubated with 20 µM geldanamycin for one hour, and Hsf1 phosphorylation examined by western blotting. (B) Depleting Hsp90 leads to Hsf1 activation. Doxycycline treatment of C. albicans tetO-HSP90 cells was used to ectopically down-regulate Hsp90 levels, and Hsf1 phosphorylation was then examined by western blotting: black arrow, phosphorylated Hsf1; white arrow, non-phosphorylated Hsf1. Hsf1 phosphorylation was confirmed using lambda phosphatase controls. (C) Hsp90 depletion causes protracted Hsf1 phosphorylation during a 30°C–42°C heat shock. Exponentially growing C. albicans tetO-HSP90 cells were transferred from 30°C to 42°C, and 20 µg/ml doxycycline was added. Hsf1 phosphorylation was examined every hour up to 4 hours. Data reflect the outcomes for at least two independent replicate experiments. (D) HSP104 transcript levels increase following Hsp90 depletion in tetO-HSP90 cells. WT (wild type, SN95: Table 1) and tetO-HSP90 cells (CaLC1411) were treated with 0 or 20 µg/ml doxycycline for 7 hours, and HSP104 transcript levels were measured and normalised to the ACT1 loading control. Wild-type cells were also subjected to a 30 minute 30°C–42°C heat shock for HSP104 comparison.

Figure 2. Hsf1 is an Hsp90 client in C. albicans.

Hsf1 and Hsp90 physically interact as revealed by co-immunoprecipitation. (A) Hsp90 co-purifies with FLAG-Hsf1 after immunoprecipitation of extracts from unstressed cells with anti-FLAG M2 affinity agarose. Hsp90 was not co-immunoprecipitated from control cells lacking tagged Hsf1. (B) Reciprocal co-immunoprecipitation was performed using Hsp90-TAP showing that FLAG-Hsf1 copurifies with Hsp90-TAP upon immunoprecipitation with IgG agarose. FLAG-Hsf1 was not co-immunoprecipiated with IgG agarose in control cells lacking FLAG-Hsf1. Data reflect the outcomes for three independent replicate experiments. (C) Hsp90 co-immunoprecipitates with FLAG-Hsf1 0, 10 and 60 minutes after a 30°C to 42°C heat shock, but does not co-immunoprecipitate in control cells lacking FLAG-Hsf1. The membrane was re-probed for Hsp70, which co-purifies with FLAG-Hsf1 at 60 minutes post-heat shock. The reciprocal co-immunopreciptation validates these results: FLAG-Hsf1 co-immunoprecipitates with Hsp90-TAP at 0, 10 and 60 minutes post-heat shock. Re-probing the membranes for Hsp70, reveals that Hsp70 interacts with Hsp90-TAP at 60 minutes post-heat shock. (D) Localisation of Hsp90-GFP in response to elevated temperature. C. albicans CaLC1855 cells were treated with a 42°C heat shock and fixed at 0, 10 and 60 minutes post-heat shock. Hsp90 is localised in the cytosol at 0 and 10 minutes post-heat shock. Significant accumulation of Hsp90 in the nucleus is observed 60 minutes post-heat shock, as determined by co-staining with DAPI. Scale bars, 5 µm.

Figure 3. Key MAP kinase signalling pathways contribute to thermal adaptation in C. albicans.

Figure depicting protein kinase mutants screened for temperature phenotypes. Temperature sensitivity was assayed by monitoring the growth of C. albicans mutants following a 30°C–42°C heat shock. Triplicate primary screens examined transposon mutants , and triplicate secondary screens tested C. albicans null mutants (Table 1). Temperature resistant kinases are displayed in the blue column, kinases with no apparent role in thermal adaptation are displayed in the pale columns, and temperature sensitive kinases are present in the pink (10–19%) to red (>20%) columns.

Figure 4. Differential activation profiles of MAP kinases in response to heat shock.

(A) Phosphorylation of C. albicans Mkc1, Cek1 and Hog1 during a 30°C–42°C heat shock revealed by western analysis using phospho-specific antibodies. Total kinase levels were monitored using antibodies against total Hog1, FLAG-tagged Mkc1 and TAP-tagged Cek1. Actin served as an internal loading control. (B) Expression of target genes for Mkc1 (PGA13), Cek1 (PMT4) and Hog1 (RHR2) during a 30°C–42°C heat shock, as determined by qRT-PCR of the corresponding transcripts relative to the internal ACT1 mRNA control.

Figure 5. Cross-talk between MAP kinase pathways during heat shock.

Differential MAP kinase activation in response to a 30°C–42°C heat shock in MAP kinase mutants. (A) Activation of Hog1, Mkc1 and Cek1 was determined in an mkc1Δ mutant relative to the internal Act1 control. (B) Activation of Hog1, Mkc1 and Cek1 was assayed in a cek1Δ mutant relative to the internal Act1 control. (C) Activation of Hog1, Mkc1 and Cek1 was determined in a hog1Δ mutant relative to the internal Act1 control.

Figure 6. Impact of MAP kinase cross-talk upon stress resistance during thermal adaptation.

Cek1 signalling is required for the Calcofluor White resistance of hog1 cells at high temperatures. MIC assays were performed in YPD medium supplemented with different concentrations of NaCl, Calcofluor White or Congo Red. Plates were incubated statically at 25°C, 30°C, 37°C and 42°C for 48 hours. For each strain, optical densities were averaged for duplicate measurements and growth is quantitatively displayed with colour as indicated with the colour bar. Data are representative of three biological replicates. WT, wild type (NGY152: Table 1).

Figure 7. The cross-talk between thermal adaptation and cell wall stress resistance is not mediated via stress cross-protection.

(A) Stress cross-protection in C. albicans wild-type cells (NGY152: Table 1) was not observed for cells pre-treated with a heat stress and then subjected to cell wall stress, but was observed for cells exposed to a secondary oxidative stress. The data represent cell survival after exposure to a 30°C–42°C heat shock followed by a subsequent cell wall (CFW, CR), osmotic (NaCl) or peroxide (H₂O₂) stress (see Materials and Methods) and the data are expressed relative to unstressed cells (dark bars). Control cells (grey bars), were not exposed to the prior 30°C–42°C heat shock. (B) The reciprocal assay was performed, whereby C. albicans wild type cells were exposed to a prior stress (cell wall: CFW, CR; osmotic: NaCl; peroxide: H₂O₂) followed by a 30°C–42°C heat shock (dark bars). These data are expressed relative to unstressed cells. Control cells (grey bars) correspond to cells exposed only to the 30°C–42°C heat shock. (C) Wild type, hog1Δ (JC50) or cap1Δ cells (JC128: Table 1) were pre-treated with a 30°C–42°C heat shock, followed by a H₂O₂ stress (dark bars). Control cells (grey bars) were not exposed to the 30°C–42°C heat shock. The data represent the level of survival compared to unstressed cells. All data are the means from three independent assays: ** paired, two-tailed t-test, p<0.01.

Figure 8. Hsp90 depletion decreases stress resistance.

(A) Impact of Hsp90 depletion upon stress resistance (CFUs) was measured following doxycycline treatment of C. albicans wild-type and tetO-HSP90 cells (SN95 and CaLC1411: Table 1): white bars, wild type, no doxycycline; light grey bars, wild type plus 20 µg/ml doxycycline; dark grey bars, tetO-HSP90 no doxycycline; black bars, tetO-HSP90 plus 20 µg/ml doxycycline; CFW, 100 µg/ml Calcofluor White; CR, 100 µg/ml Congo Red; HS, 30°C–42°C heat shock; 5 mM H₂O₂; 1 M NaCl. (B) Impact of Hsp90 inhibition with the pharmacological inhibitor geldanamycin. Cells were grown for 7 hours in the absence or presence of 10 µM geldanamycin and subjected to the same stresses as in (A). CFUs determined from untreated cells. All data are the means from three independent assays: ** paired, two-tailed t-test, p<0.01.

Figure 9. Hsp90 depletion affects MAP kinase signalling in the presence and absence of stress.

C. albicans wild-type and tetO-HSP90 cells (SN95 and CaLC1411: Table 1) were treated with 0 or 20 µg/ml doxycycline for seven hours. Mkc1, Cek1 and Hog1 phosphorylation levels were then assayed by western analysis in unstressed cells or cells treated as follows: (A) 30 minute 30°C–42°C heat shock; (B) 30 minutes with 100 µg/ml Calcofluor White (CFW); (C) 10 minutes with 5 mM H₂O₂; or (D) 12 minutes with 1 M NaCl. Total Mkc1 levels were assayed using Mkc1-6xHis-FLAG tagged cells in SN95 (CaLC681) and tetO-HSP90 (caLC648) cells (Table 1). Hsp90 levels and total kinase levels for Hog1 were also examined by western blotting relative to the internal Act1 loading control.

Figure 10. Hsp90 depletion destabilises Cek1.

C. albicans CEK1-TAP (CaLC2287) and tetO-HSP90 CEK1-TAP (CaLC2288: Table 1) cells were treated with 0 or 20 µg/ml doxycycline for seven hours and subjected to western analysis. Decreased Cek1-TAP levels were observed following Hsp90 depletion in unstressed cells and in cells treated as follows: (A) 30 minute 30°C–42°C heat shock; or (B) 30 minutes with 100 µg/ml Calcofluor White (CFW). Hsp90 protein levels were examined confirming significant depletion following doxycycline treatment of tetO-HSP90 cells. Actin served as the internal loading control.

Figure 11. Hsp90 depletion affects cell wall architecture.

C. albicans wild-type cells (SN95), and tetO-HSP90 (CaLC1411: Table 1) were treated with 0 or 20 µg/ml doxycycline for seven hours. Chitin levels were assayed by Calcofluor White staining, scale bars are 10 µm (A) and quantification of fluorescence levels (B). Data represent means from fifty cells: ** paired, two-tailed t-test, p<0.01. The architecture of the cell wall was examined by transmission electron microscopy where scale bars are 200 µm (C), and the thickness of the cell wall quantified in n = 30 cells (D): ** paired, two-tailed t-test, p<0.01.

Figure 12. Hsp90 coordinates the activities of multiple signalling pathways that contribute to thermotolerance in C. albicans - a model.

Hsf1 activation is required for thermotolerance . Hog1, Mkc1 and Cek1 signalling are also required for thermotolerance (Figure 3), but these MAP kinases are not essential for Hsf1 phosphorylation (Figure 4). Instead, these pathways promote thermotolerance in part via cell wall remodelling , . Hsp90 coordinates much of this activity. Hsf1 (Figures 1 and 2), Hog1, Mkc1 and Cek1 (Figure 10) are all Hsp90 client proteins , . Changes in ambient temperature affect interactions between Hsp90 and Hsf1 (Figure 2), and probably affect Hsp90 interactions with Hog1, Mkc1 and Cek1 thereby modulating the activities of these signalling pathways and their inputs to thermal adaptation. Increases in ambient temperature activate Hsf1, thereby inducing the expression of protein chaperones (HSPs) including Hsp90, which promotes thermal adaptation in the shorter term. Hsp90 then down-regulates Hsf1 and modulates Mkc1, Hog1 and Cek1 signalling, which in the longer term influences cell wall architecture (Figure 11), leading to the thermotolerance of C. albicans.