Hsp90 mutants with distinct defects provide novel insights into cochaperone regulation of the folding cycle

Abstract

Molecular chaperones play a key role in maintaining proteostasis and cellular health. The abundant, essential, cytosolic Hsp90 (Heat shock protein, 90 kDa) facilitates the folding and activation of hundreds of newly synthesized or misfolded client proteins in an ATP-dependent folding pathway. In a simplified model, Hsp70 first helps load client onto Hsp90, ATP binding results in conformational changes in Hsp90 that result in the closed complex, and then less defined events result in nucleotide hydrolysis, client release and return to the open state. Cochaperones bind and assist Hsp90 during this process. We previously identified a series of yeast Hsp90 mutants that appear to disrupt either the 'loading', 'closing' or 'reopening' events, and showed that the mutants had differing effects on activity of some clients. Here we used those mutants to dissect Hsp90 and cochaperone interactions. Overexpression or deletion of HCH1 had dramatically opposing effects on the growth of cells expressing different mutants, with a phenotypic shift coinciding with formation of the closed conformation. Hch1 appears to destabilize Hsp90-nucleotide interaction, hindering formation of the closed conformation, whereas Cpr6 counters the effects of Hch1 by stabilizing the closed conformation. Hch1 and the homologous Aha1 share some functions, but the role of Hch1 in inhibiting progression through the early stages of the folding cycle is unique. Sensitivity to the Hsp90 inhibitor NVP-AUY922 also correlates with the conformational cycle, with mutants defective in the loading phase being most sensitive and those defective in the reopening phase being most resistant to the drug. Overall, our results indicate that the timing of transition into and out of the closed conformation is tightly regulated by cochaperones. Further analysis will help elucidate additional steps required for progression through the Hsp90 folding cycle and may lead to new strategies for modulating Hsp90 function.

Fig 1. Effect of HCH1 overexpression varies between mutant groups.

Hsc82 mutants were expressed in strain JJ816 (hsc82hsp82). Cells expressing the indicated mutant were transformed with empty vector (p41KanTEF, -) or a plasmid overexpressing HCH1 (p41KanTEF-Hch1-myc, + HCH1). Cells were grown overnight at 30°C, serially diluted 10-fold, plated on selective media (YPD + 200 μg/ml G418) and grown for two days at the indicated temperature. A: Mutants that alter residues required for direct interaction between Hsp70 and Hsp90 (R46G and G309S: loading mutants), or mutants that disrupt the ability of Hsc82 to interact with Sba1 and Cpr6 in the presence of AMP-PNP (S481Y, A583T: closing mutants). B. Mutants that alter residues associated with ATP hydrolysis or associated conformational changes (S25P, K102E, Q380K: reopening mutants). C. Additional strains expressing reduced levels of Hsp82, Hsp70 or containing a deletion in STI1 (strains GRS4, JJ1480 and JJ623) were transformed with empty vector (p41KanTEF, -), a plasmid overexpressing HCH1 (p41KanTEF-Hch1-myc, + HCH1), and grown as above. All strains were growth at 30°C except the strain lacking STI1. D. Effects of HCH1 overexpression were quantified in three independent growth assays. Representative images of each sample set at shown. Left, mutations that affect loading and closing steps. Right, mutations that affect reopening steps.

Fig 2. Effect of AHA1 overexpression varies between mutant groups.

As in Fig 1, except that cells were independently transformed with a plasmid overexpressing AHA1 (p41KanTEF-Aha1-myc, +AHA1). Cells were then grown overnight at 30°C, serially diluted 10-fold, and plated on selective media (YPD + G418) and grown for two days at the indicated temperature. A cartoon shows the approximate location of the mutations within Hsp90. Dark gray, Hsp90 dimer. Circle, client, X, approximate site of mutations. 70, Hsp70. Below. Effects of HCH1 of AHA1 overexpression were quantified using at least two independent growth assays.

Fig 3. Deletion of the NxNNWHW sequence or the D53K alteration of Hch1 disrupts function.

As in Fig 1, except that cells were also transformed with a plasmid overexpressing HCH1 containing a deletion of the NxNNWHW sequence at the amino terminus (p41KanTEF-Hch1Δ11-myc, +hch1Δ11, or p41KanTEF-Hch1-D53K-myc, +hch1-D53K). Cells were grown overnight at 30°C, serially diluted 10-fold, and plated on selective media (YPD + G418) and grown for two days at the indicated temperature. Below. Effects of overexpression of wild-type or mutant Hch1 were quantified using at least two independent growth assays.

Fig 4. Identification of hsc82-M116I, which is specifically dependent on HCH1 and alters a residue in the lid that closes over bound ATP.

A. hsc82-M116I was expressed in isogenic hsc82hsp82 strains that do not contain deletion of any cochaperones (WT) or contain individual deletion of the cochaperone listed. Strains are listed in S2 Table. Cells were serially diluted 10-fold and grown at the indicated temperature for two days. B. WT HSC82 or hsc82-M116I were expressed in an hsc82hsp82 strain (WT) or an hch1hsc82hsp82 strain (hch1). Cells were transformed with empty vector (pRS426, -) or a plasmid overexpressing HCH1 (pRS426-HCH1, + HCH1), grown overnight at 30°C, serially diluted 10-fold, plated on selective media and grown for two days at the indicated temperature. C. Location of M116 and K102 (shown in pick) mapped onto the closed AMP-PNP bound structure (PBD 2CG9). Figure generated with EZMOL [78], nucleotide shown in orange. Below. Growth assays were quantified using at least two independent growth assays.

Fig 5. The effect of deletion of individual non-essential cochaperones varies according to mutant groupings.

Plasmids expressing indicated WT or mutant forms of Hsc82 were expressed in isogenic hsc82hsp82 strains that do not contain deletion of any cochaperones (WT) or contain individual deletion of the cochaperone listed. Cells were grown overnight at 30°C, serially diluted 10-fold, plated on rich media (YPD) and grown for two days at the indicated temperature. If no growth is shown on the 30° plate (such as with hsc82-R46G in the sti1 strain), the cells were inviable. Below. Growth assays were quantified using at least two independent growth assays.

Fig 6. Overexpression of CPR6 promotes Cpr6-Hsc82 interaction or stabilization in the absence of exogenous nucleotide.

A. Strain JJ110 (cpr6hsc82hs82) expressed untagged Cpr6, His-CPR6 under the ADH promoter or His-CPR6 under the GPD promoter [42]. Cell lysates were incubated in the absence or presence of 5 mM AMP-PNP for 5 min. at 30°C prior to incubation with nickel resin. The stained gel is shown, as well as immunoblots showing Hsc82/Hsp82 bound to nickel resin or in whole cell lysates. B. Quantification of the relative levels of Hsc82/Hsp82 bound to His-Cpr6. C. Model for the opposing in vivo effects of Hch1 and Cpr6. Overexpression of HCH1 exacerbates defects of the loading and closing mutants. Overexpression of HCH1 or AHA1 overcomes defects of the reopening mutants. Overexpression of CPR6 promotes or stabilizes the closed conformation and release of Cpr6 is required for cycle progression.

Fig 7. Sensitivity to Hsp90 inhibitor also correlates with mutant grouping.

Strains were grown in YPD media and 10-fold serial dilutions were prepared and placed on agar plates with or without the indicated concentration of NVP-AUY922 where indicated and grown for 48 hours at the indicated temperatures. Growth assays were quantified using at least two independent growth assays (S9 Fig).