Nucleotide exchange is sufficient for Hsp90 functions in vivo

Abstract

Hsp90 is an essential eukaryotic chaperone that regulates the activity of many client proteins. Current models of Hsp90 function, which include many conformational rearrangements, specify a requirement of ATP hydrolysis. Here we confirm earlier findings that the Hsp82-E33A mutant, which binds ATP but does not hydrolyze it, supports viability of S. cerevisiae, although it displays conditional phenotypes. We find binding of ATP to Hsp82-E33A induces the conformational dynamics needed for Hsp90 function. Hsp90 orthologs with the analogous EA mutation from several eukaryotic species, including humans and disease organisms, support viability of both S. cerevisiae and Sz. pombe. We identify second-site suppressors of EA that rescue its conditional defects and allow EA versions of all Hsp90 orthologs tested to support nearly normal growth of both organisms, without restoring ATP hydrolysis. Thus, the requirement of ATP for Hsp90 to maintain viability of evolutionarily distant eukaryotic organisms does not appear to depend on energy from ATP hydrolysis. Our findings support earlier suggestions that exchange of ATP for ADP is critical for Hsp90 function. ATP hydrolysis is not necessary for this exchange but provides an important control point in the cycle responsive to regulation by co-chaperones.

Cochaperones convey the energy of ATP hydrolysis for directional action of Hsp90

“…Thus, conformational changes are thermally driven and not fuelled by ATP turnover. A recent study suggests that it is not the hydrolysis of ATP but the repositioning of its gamma phosphate which is sufficient for Hsp90 function 19 . They emphasize that returning to the open, ADP-bound state is essential, which can occur either by nucleotide exchange (non-directional) or by hydrolysis (directional).…”

“…This would allow first for the formation of a functional complex and then, e.g. after ATP hydrolysis or nucleotide exchange 19 , for recycling of all components. We speculate that this is the long searched-for function of Hsp90’s ATPase function, namely to disassemble a functional complex after the job is done.…”

“…Upon ATP-hydrolysis, the complex opens again and releases the maturated kinase as well as the cochaperone, leaving Hsp90 ready for yet another cycle 13 , 15 – 17 . Alternatively, it was shown that Hsp90 does not close upon ATP binding itself 18 , but instead, Reidy et al 19 proposed recently that its conformational changes are triggered by the proper positioning of ATP’s gamma phosphate. Furthermore, it was proposed that the cycle might also be linked to dephosphorylation of a kinase 20 .…”

Insights into Hsp90 mechanism and in vivo functions learned from studies in the yeast, Saccharomyces cerevisiae

“…As shown in Figure 1B , most of the differences between the two isoforms are located in the amino-terminal domain, which is likely the basis for differences in ATPase activity and sensitivity to Hsp90 inhibitors that bind the nucleotide-binding pocket ( Girstmair et al, 2019 ). Early studies showed that ATP hydrolysis was also essential in yeast, but subsequent studies suggest that nucleotide exchange, rather than hydrolysis, is sufficient for viability ( Zierer et al, 2016 ; Reidy et al, 2023 ). Additional structural studies, coupled with functional analysis in yeast, identified residues within a flexible loop of the middle domain that play an important role regulating ATP hydrolysis ( Meyer et al, 2003 ).…”

Cochaperones convey the energy of ATP hydrolysis for directional action of Hsp90

“…Upon ATPhydrolysis, the complex opens again and releases the maturated kinase as well as the cochaperone, leaving Hsp90 ready for yet another cycle (Verba and Agard 2017;Eckl et al 2015;Keramisanou et al 2022;Schopf et al 2017). Alternatively, it was shown that Hsp90 does not close upon ATP binding itself (Lopez et al 2021), but instead (Reidy et al 2023) proposed recently that its conformational changes are triggered by the proper positioning of ATP's gamma phosphate. Furthermore, it was proposed that the cycle might also be linked to dephosphorylation of a kinase (Jaime-Garza et al 2023).…”