HSP90 inhibitors disrupt a transient HSP90-HSF1 interaction and identify a noncanonical model of HSP90-mediated HSF1 regulation

Abstract

Heat shock factor 1 (HSF1) initiates a broad transcriptional response to proteotoxic stress while also mediating a cancer-specific transcriptional program. HSF1 is thought to be regulated by molecular chaperones, including Heat Shock Protein 90 (HSP90). HSP90 is proposed to sequester HSF1 in unstressed cells, but visualization of this interaction in vivo requires protein crosslinking. In this report, we show that HSP90 binding to HSF1 depends on HSP90 conformation and is only readily visualized for the ATP-dependent, N-domain dimerized chaperone, a conformation only rarely sampled by mammalian HSP90. We have used this mutationally fixed conformation to map HSP90 binding sites on HSF1. Further, we show that ATP-competitive, N-domain targeted HSP90 inhibitors disrupt this interaction, resulting in the increased duration of HSF1 occupancy of the hsp70 promoter and significant prolongation of both the constitutive and heat-induced HSF1 transcriptional activity. While our data do not support a role for HSP90 in sequestering HSF1 monomers to suppress HSF1 transcriptional activity, our findings do identify a noncanonical role for HSP90 in providing dynamic modulation of HSF1 activity by participating in removal of HSF1 trimers from heat shock elements in DNA, thus terminating the heat shock response.

Figure 1

HSF1 is bound by HSP90 in the “closed” conformation. (A) HSP90 ATP-driven conformational cycle. HSP90α E47A/HSP90β E42A mutants bind ATP but cannot hydrolyze it and remain trapped in the “closed” N-domain dimerized conformation. Conversely, HSP90α D93A/HSP90β D88A mutants do not bind ATP and remain trapped in an “open” N-domain undimerized conformation. (B) HSF1 interaction with HSP90 “closed” and “open” mutants. HEK293 cells were co-transfected with HA-HSF1 and with each FLAG-HSP90 construct, harvested and analyzed for HSF1/HSP90 association by anti-FLAG IP - HSF1 WB. Pull-down experiments were repeated at least twice. (C) Reproducible and reciprocal HSF1 interaction with “closed” HSP90α E47A and HSP90β E42A mutants in the presence and absence of MoO4. Interactions were assessed by anti-FLAG (HSP90) IP - HSF1 WB (top panel) or anti-FLAG (HSF1) IP–HA (HSP90) WB (bottom panel). Pull-down experiments were repeated at least twice. (D) Effect of HSP90 inhibitors on HSF1-HSP90α E47A interaction. HEK293 cells were transfected with HA-HSF1 along with FLAG-HSP90α E47A, treated with N-terminal or C-terminal HSP90 inhibitors for 4 hours and analyzed by anti-FLAG (HSP90) IP- HSF1 WB. Pull down experiments were repeated at least twice. Full blots used to create cropped figure panels are shown in Supplementary Figure 6.

Figure 2

Mapping interaction sites of “closed” HSP90 on HSF1. (A) Schema of FLAG-HSF1 C-terminal truncation constructs and their expression in HEK293 cells. The associated blot shows expression of each construct in HEK293 cells. Equal amounts of DNA were used for transfection. For construct C477, the upper band represents the intact truncation (designated by an asterisk), while the lower band is likely a proteolytic fragment. (B) Schema of FLAG-HSF1 N-terminal truncation and internal deletion constructs and their expression in HEK293 cells. The associated blot shows expression of each construct in HEK293 cells. Equal amounts of DNA were used for each transfection. Note the poor expression of N195. For N195-C383, the lower band represents the intact construct (designated by an asterisk); the upper band may reflect a post translationally modified peptide. (C) Interaction of selected FLAG-HSF1 C-terminal truncation constructs with HA-tagged HSP90α E47A. HEK293 cells were transfected with either FLAG-HSF1 constructs or FLAG-pcDNA control and HA-HSP90α E47A, and then analyzed by anti-FLAG (HSF1) IP followed by HA (HSP90α E47A) WB. Pull downs were performed in triplicate and western blots analyzed by densitometry. All HA IP bands were normalized to their respective FLAG IP signals (see Supplemental Figure 6, compilation of full blots) and then normalized to the WT signal. A one-way ANOVA utilizing multiple comparisons was performed to determine significant differences in HSP90 interaction between truncated HSF1 constructs and F-HSF1 WT. A representative blot is shown above the bar graph. Significance is indicated by asterisks (**Indicates p < 0.01). The p-value for the difference between C339 and WT interaction was 0.0534. (D) Interaction of FLAG-HSF1 N-terminal truncation and internal deletion constructs with HA- tagged HSP90α E47A: HEK293 cells were transfected and analyzed as in panel C. Significance is indicated by asterisks (**Indicates p < 0.01 and ****Indicates p < 0.0001). Full blots used to create cropped figure panels and for densitometry analysis for the bar graphs are shown in Supplementary Figure 6.

Figure 3

Effects of HSF1 truncation and internal deletion mutants on hsp70 promoter-reporter activity and HSF1 dimerization. (A) Expression of TAD-deleted HSF1 constructs inhibits endogenous HSF1 driven hsp70 promoter-reporter activity, while removal of the HR-A/B domain abrogates dominant negative activity. HEK293 cells were transfected with hsp70 promoter-reporter construct and HSF1 C-terminal truncation constructs, treated with DMSO or 17AAG for 1 hour, heat shocked at 42 °C for 30 minutes and allowed to recover for 4 hours before harvesting. (B) Disruption of HA-tagged HSP90α E47A interaction with HSF1 C-terminal RD truncations by 17AAG. HEK293 cells were transfected with HA-HSP90α E47A and FLAG-HSF1 RD truncation mutants, then analyzed by anti-FLAG (HSF1) IP followed by HA (HSP90α E47A) WB. Pull-down experiments were repeated at least twice. (C) Deletion of the DBD and retention of the HR-A/B domain inhibits hsp70-reporter activity driven by endogenous HSF1. HEK293 cells were transfected, treated and processed as described in panel A. (D) Deletion of the HR-A/B domain disrupts HSF1 dimerization. HEK293 cells were transfected with WT HA-HSF1 and FLAG-HSF1 deletion constructs, then analyzed by anti-FLAG IP and HA WB. Pull-down experiments were repeated at least twice. (E) Summary of the effects of HSF1 domain deletions on hsp70-reporter activity. Full blots used to create cropped figure panels are shown in Supplementary Figure 6.

Figure 4

Impact of RD phosphorylation on HSP90 inhibitor effect on HSF1 activity and mobility shift in SDS-PAGE, and on HSF1 interaction with HSP90 (A) Schema of WT and dPRD FLAG-HSF1 constructs depicting RD phosphorylation sites mutated to alanine in dPRD-HSF1. (B) The N-terminal HSP90 inhibitor 17AAG enhances transcriptional activity of dPRD HSF1. HEK293 cells were transfected with hsp70-reporter construct and HSF1 C-terminal truncation constructs, treated with DMSO or 17AAG for 1 hour, heat shocked at 42 °C for 30 minutes and allowed to recover for 4 hours before harvesting. This experiment was repeated three times and values represent mean +/− S.D. (C) N-terminal HSP90 inhibitors induce a mobility shift in dPRD HSF1 similarly to that seen with WT-HSF1. HEK293 cells were transfected with WT or dPRD HSF1, treated with the three N-terminal HSP90 inhibitors as shown, and harvested and HSF1 was analyzed by WB. Full blots used to create cropped figure panels are shown in Supplementary Figure 6.

Figure 5

HSP90 inhibition increases duration, but not extent, of HSF1 binding to the hsp70 promoter and enhances HSF1 transcriptional activity in response to heat shock. (A) Chromatin immunoprecipitation using HSF1 antibody followed by PCR amplification of the hsp70 promoter sequence (top, representative DNA gel; bottom, graphical display of band optical density data from 3 independent experiments +/− S.D.). HEK293 cells were either subjected to heat shock (42 °C/30 minutes), treated with 17AAG, or both. The 17AAG groups were pre-exposed to the HSP90 inhibitor (10 μM, 1 hour) prior to heat shock. Samples were then put back in a 37 °C incubator to recover for the indicated times. Values for the ‘no drug, no HS’ control (first lane) were set to 1, and remaining values are expressed as fold above or below this baseline value. This experiment was repeated three times and values represent mean +/− S.D. A two-way ANOVA was used to analyze the samples for statistically significant differences compared to other samples at a given time. Significant difference compared to ‘no drug, no HS’ is indicated by asterisks (***Indicates p < 0.001, **Indicates p < 0.01, and *Indicates p < 0.05); ‘17AAG + HS’ is also significantly lower than ‘17AAG’ alone at 6 hr (^##p < 0.01) and significantly higher than ‘HS’ alone at 2 and 4 hr (⁺⁺⁺Indicates p < 0.001, ⁺Indicates p < 0.05). (B) Endogenous HSP70 mRNA was quantified by RT-qPCR in HEK293 cells subjected to heat shock alone, 17AAG alone, or heat shock after pre-exposure to 17AAG, using identical experimental conditions as in panel A. This experiment was performed in triplicate and values represent mean +/− S.D. A two-way ANOVA was used to analyze for statistically significant differences. Notably, there was a statistically significant difference between ‘17AAG alone’ and ‘17AAG + HS’ samples at 4 and 6 hours (***Indicates p < 0.001 and ****Indicates p < 0.0001).