Hsf1 activation inhibits rapamycin resistance and TOR signaling in yeast revealed by combined proteomic and genetic analysis

Abstract

TOR kinases integrate environmental and nutritional signals to regulate cell growth in eukaryotic organisms. Here, we describe results from a study combining quantitative proteomics and comparative expression analysis in the budding yeast, S. cerevisiae, to gain insights into TOR function and regulation. We profiled protein abundance changes under conditions of TOR inhibition by rapamycin treatment, and compared this data to existing expression information for corresponding gene products measured under a variety of conditions in yeast. Among proteins showing abundance changes upon rapamycin treatment, almost 90% of them demonstrated homodirectional (i.e., in similar direction) transcriptomic changes under conditions of heat/oxidative stress. Because the known downstream responses regulated by Tor1/2 did not fully explain the extent of overlap between these two conditions, we tested for novel connections between the major regulators of heat/oxidative stress response and the TOR pathway. Specifically, we hypothesized that activation of regulator(s) of heat/oxidative stress responses phenocopied TOR inhibition and sought to identify these putative TOR inhibitor(s). Among the stress regulators tested, we found that cells (hsf1-R206S, F256S and ssa1-3 ssa2-2) constitutively activated for heat shock transcription factor 1, Hsf1, inhibited rapamycin resistance. Further analysis of the hsf1-R206S, F256S allele revealed that these cells also displayed multiple phenotypes consistent with reduced TOR signaling. Among the multiple Hsf1 targets elevated in hsf1-R206S, F256S cells, deletion of PIR3 and YRO2 suppressed the TOR-regulated phenotypes. In contrast to our observations in cells activated for Hsf1, constitutive activation of other regulators of heat/oxidative stress responses, such as Msn2/4 and Hyr1, did not inhibit TOR signaling. Thus, we propose that activated Hsf1 inhibits rapamycin resistance and TOR signaling via elevated expression of specific target genes in S. cerevisiae. Additionally, these results highlight the value of comparative expression analyses between large-scale proteomic and transcriptomic datasets to reveal new regulatory connections.

Figure 1. Proteomic analysis strategy and results.

(A) Sample preparation workflow for quantitative proteomic analysis of rapamycin treatment in BJ5465 yeast cells. (B) Functional categorization of 127 proteins showing abundance changes of 1.5 fold or greater due to rapamycin treatment. The number of proteins from each category, and their relative percentages are also indicated on the pie chart. (C) Correlation or anticorrelation (described as similar or opposite changes between proteins and RNA, respectively) for rapamycin affected proteins (obtained via proteomic analysis in this study) and gene transcripts (obtained by microarray analysis of rapamycin treated yeast cells; *, , and heatshock/oxidative stress; **[13]).

Figure 2. Cells with increased Hsf1 transcriptional activity are hypersensitive to rapamycin treatment.

(A) Rapamycin sensitivity of HSF1, hsf1-R206S, F256S, and hsf1-R256S cells (upper panel). FPR1-dependent rapamycin sensitivity of hsf1-R206S, F256S cells (lower panel). (B) Rapamycin sensitivity of SSA1 SSA2, ssa1-3 ssa2-2, and ssa1-3 ssa2-2 hsf1P215Q cells. Cells were grown to saturation at 25°C and serial dilutions (50,000, 5000, and 500 cells per spot) were spotted on YPD plates supplemented with 25 nM rapamycin or drug carrier solvent (methanol) and assayed for growth at 25°C for the indicated durations of time. ssa1-3 ssa2-2 cells and derivatives were grown identically but spotted at a density of 5000 and 500 cells/spot.

Figure 3. Effect of hsf1-R206S, F256S mutation on expression of HSE4Ptt-CYC1-LacZ reporter and Hsf1 target genes.

(A) hsf1-R206S, F256S and isogenic HSF1 cells transformed with HSE4Ptt-CYC1-lacZ plasmid were grown overnight in minimal selective media at 23°C to an OD₆₀₀ of 0.5 units, and then shifted to 25°C, 29°C, or 33°C, for 90 minutes prior to determination of β-galactosidase activity. (B) mRNA levels of diverse classes of Hsf1 targets in hsf1-R206S, F256S cells relative to HSF1 cells. The promoter region of HSP12 is known to have ‘step’ heat shock elements (HSEs), while that of SSA3/4, HSP78, and HSP42 have perfect HSEs . Although canonical HSEs have not been found in promoter regions of PIR3 and YRO2, these were identified in global CHIP-on-CHIP experiments as Hsf1 targets . CUP1-1 has a variant HSE . Cells were grown at 25°C, and processed for RNA isolation, real-time PCR analysis, and analyzed as described in materials and methods section. Relative expression of each gene was normalized to actin and expressed as an average fold induction in hsf1-R206S, F256S cells versus unperturbed wild type cells.

Figure 4. Reduced TOR signaling in hsf1-R206S, F256S cells.

(A) Expression level of genes representing five different pathways repressed by TOR function, upon rapamycin treatment in HSF1 cells (left panel), and in hsf1-R206S, F256S cells (right panel, in absence of rapamycin treatment). (B) Expression level of ribosomal protein (RP) genes and RAP1, a positive regulator of RP genes, upon rapamycin treatment in HSF1 cells (left panel) and in hsf1-R206S, F256S cells (right panel, in absence of rapamycin treatment) (C) Mobility of Gln3-myc ₁₃ in HSF1 cells treated with or without rapamycin and hsf1-R206S, F256S cells with or without rapamycin treatment as indicated above. Cells were grown to log-phase at 25°C and treated with 200nM rapamycin or methanol alone and processed for RNA isolation or total protein extraction as described in materials and methods section.

Figure 5. Role of Msn2/4 and Gln3/Gat1 in TOR-regulated phenotypes seen in hsf1-R206S, F256S cells.

(A) Effect of deleting MSN2, 4 on elevated expression of Msn2/4 targets in hsf1-R206S, F256S cells (B) Effect of deleting MSN2, 4 on rapamycin induced expression of Msn2/4 targets in HSF1 cells (C) Effect of deleting GLN3 alone or both GLN3 and GAT1 on elevated expression of NCR genes in hsf1-R206S, F256S cells (D) Effect of MSN2/4, GLN3/GAT1, or HYR1 deletions on rapamycin sensitivity of hsf1-R206S, F256S cells. Relative expression of each gene was normalized to actin and expressed as an average fold induction relative to wild type cells.

Figure 6. Deletion of Hsf1 target genes, PIR3 and YRO2 partially suppress TOR-regulated phenotypes of hsf1-R206S, F256S cells.

A) Suppression of rapamycin sensitivity of hsf1-R206S, F256S cells by deletion of PIR3 and YRO2. HSF1 and hsf1-R206S, F256S cells bearing the indicated gene deletions were grown to saturation at 25°C and 5000 cells each were spotted on YPD plates supplemented with methanol alone (rapamycin solvent), 10 nM, and 25 nM rapamycin, respectively. B) PIR3 deletion reduced expression of multiple TOR-repressed genes in hsf1-R206S, F256S cells. Expression level of genes was monitored by RT-PCR as explained in materials and methods section. C) Effect of PIR3 and YRO2 deletion on temperature sensitivity of hsf1-R206S, F256S cells. Indicated strains were streaked out on YPD plates and allowed to grow 3 days at 34°C.

Figure 7. Over expression of MSN2, MSN4 or HYR1 does not inhibit TOR signaling (A) Effect of over expression of MSN2, MSN4 or HYR1 on rapamycin resistance of wild type cells.

Wild type HS170T cells (HSF1 cells isogenic to hsf1-R206S, F256S cells used in this study) were transformed with 2μ plasmids for over expression of the relevant genes, and spotted on selective media supplemented with 25 nM Rapamycin (or methanol) at 50,000, 5000, and 500 cells per spot and assayed for growth at 25°C (B) Effect of MSN2 over expression on TOR signaling ‘readouts’ assayed by real-time PCR (C) Effect of MSN2 over expression versus rapamycin treatment, on expression level of CTT1, a classical Msn2 target gene. RNA isolation, cDNA synthesis, real-time PCR conditions, and analysis of data are described in materials and methods section.