Jump to ContentJump to Main Navigation
Show Summary Details
More options …

Biological Chemistry

Editor-in-Chief: Brüne, Bernhard

Editorial Board: Buchner, Johannes / Lei, Ming / Ludwig, Stephan / Sies, Helmut / Thomas, Douglas D. / Turk, Boris / Wittinghofer, Alfred

12 Issues per year


IMPACT FACTOR 2017: 3.022

CiteScore 2017: 2.81

SCImago Journal Rank (SJR) 2017: 1.562
Source Normalized Impact per Paper (SNIP) 2017: 0.705

Online
ISSN
1437-4315
See all formats and pricing

Access brought to you by:

provisional account

More options …
Volume 399, Issue 4

Issues

Interaction of the middle domains stabilizes Hsp90α dimer in a closed conformation with high affinity for p23

Kamil Synoradzki
  • Department of Experimental Pharmacology, Mossakowski Medical Research Centre, Polish Academy of Sciences, 5 Pawinskiego St., 02-106 Warsaw, Poland
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
/ Przemyslaw Miszta
  • Faculty of Chemistry, Biological and Chemical Research Centre, University of Warsaw, 1 Pasteura St., 02-093 Warsaw, Poland
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
/ Egidijus Kazlauskas
  • Department of Biothermodynamics and Drug Design, Institute of Biotechnology, Vilnius University, Saulėtekio Al. 7, 10257 Vilnius, Lithuania
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
/ Aurelija Mickevičiūtė
  • Department of Biothermodynamics and Drug Design, Institute of Biotechnology, Vilnius University, Saulėtekio Al. 7, 10257 Vilnius, Lithuania
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
/ Vilma Michailovienė
  • Department of Biothermodynamics and Drug Design, Institute of Biotechnology, Vilnius University, Saulėtekio Al. 7, 10257 Vilnius, Lithuania
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
/ Daumantas Matulis
  • Department of Biothermodynamics and Drug Design, Institute of Biotechnology, Vilnius University, Saulėtekio Al. 7, 10257 Vilnius, Lithuania
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
/ Slawomir Filipek
  • Faculty of Chemistry, Biological and Chemical Research Centre, University of Warsaw, 1 Pasteura St., 02-093 Warsaw, Poland
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
/ Pawel Bieganowski
  • Corresponding author
  • Department of Experimental Pharmacology, Mossakowski Medical Research Centre, Polish Academy of Sciences, 5 Pawinskiego St., 02-106 Warsaw, Poland
  • Email
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
Published Online: 2018-01-16 | DOI: https://doi.org/10.1515/hsz-2017-0172

Abstract

The human genome encodes two highly similar cytosolic Hsp90 proteins called isoforms Hsp90α and Hsp90β. Of the 300 client proteins for Hsp90 identified so far only a handful interact specifically with one Hsp90 isoform. Here we report for the first time that Hsp90 cochaperone p23 binds preferentially to Hsp90α and that this interaction is mediated by the middle domain of Hsp90α. Based on the homology modeling, we infer that the middle domains in the Hsp90α dimer bind stronger with each other than in the Hsp90β dimer. Therefore, compared to Hsp90β, Hsp90α may adopt closed conformation more easily. Hsp90 interacts with p23 in the closed conformation. Hsp90α binds human recombinant p23 about three times stronger than Hsp90β but with significantly smaller exothermic enthalpy as determined by isothermal titration calorimetry of direct binding between the purified proteins. As p23 binds to Hsp90 in a closed conformation, stabilization of the Hsp90α dimer in the closed conformation by its middle domains explains preference of p23 to this Hsp90 isoform.

This article offers supplementary material which is provided at the end of the article.

Keywords: Hsp90α; Hsp90β; human; isoform; middle domain; p23

Introduction

Hsp90 (90 kDa heat shock protein) is a indispensable chaperone in eukaryotic cells, crucial for maintaining proper folding and stability of 10% of the proteome (Zhao et al., 2005).

Hsp90 acts as a homodimer with the monomers composed of three domains with distinct functions. The C-terminal domain (C-domain) holds together monomers of Hsp90. The middle domain (M-domain) contains a major site of interaction with the client proteins. The N-terminal domain (N-domain) possesses both ATP-binding and ATP-hydrolytic activity, essential for Hsp90 function. In addition, each domain can bind numerous cochaperones that modulate the function of Hsp90 in two ways. Some of them, such as Cdc37, Sgt1, Hop or Gcunc45, facilitate the interaction of Hsp90 with clients. Others, such as Aha1 and p23, influence Hsp90 activity mostly by changes in ATP hydrolysis rate. Aha1 stimulates Hsp90 ATPase activity, whereas p23 decreases ATP hydrolysis (Panaretou et al., 2002; Richter et al., 2004; Siligardi et al., 2004). The Hsp90 dimer may adopt open, closed or intermediate conformations depending on bound ATP or cochaperones. All three domains are involved in the binding of Hsp90 monomers in the closed conformation. The monomers in the open conformation are held together only by the dimerization of the C-domains (Csermely et al., 1993; Nemoto et al., 1995; Sullivan et al., 1997; Ali et al., 2006).

Most eukaryotic genomes contain only one gene encoding cytosolic form of Hsp90, with the exception of vertebrates that have two genes that express two proteins called isoforms Hsp90α (Hsp90AA1) and Hsp90β (Hsp90AB1) (Chen et al., 2006). Both isoforms interact with a similar set of clients and cochaperones. However, studies of the mouse knockouts lacking one Hsp90 isoform demonstrated that Hsp90α and Hsp90β perform specific functions and that their chaperoning activity does not completely overlap. Hsp90α deletion results in male infertility, due to the failure in sperm maturation. Other than that phenotype, development of both sexes seems unaffected (Grad et al., 2010; Kajiwara et al., 2012). In contrast, deletion of Hsp90β results in the failure of placenta development by both male and female embryos which leads to death at an early stage of development (Voss et al., 2000).

Together, based on the available data, as many as 135 proteins may interact exclusively with Hsp90α, whereas Hsp90β may interact with more than 400 clients (Echeverria et al., 2011; Zuehlke et al., 2015). Some proteins can be chaperoned effectively by one, but not the other isoform. For example, endothelial nitric oxide synthase (eNOS) activity decreases is cells that are treated with anti Hsp90α siRNA, but silencing of Hsp90β results in increased eNOS activity (Cortes-Gonzalez et al., 2010). The Hsp90 isoforms exert an opposite effect on the cellular level of the potassium channel KCNQ4. The KCNQ4 level decreases when Hsp90β is silenced using specific siRNA, whereas silencing Hsp90α has the opposite effect and leads to an increase in KCNQ4 levels (Gao et al., 2013).

The above studies document the isoform-specific activity of Hsp90 but do not address the question of what underlies these differences at the level of the Hsp90 structure and its interaction with other proteins. Three isoform-specific cochaperones were identified thus far. Gcunc45 and Aarsd1 preferentially bind to Hsp90β, whereas Aha1 interacts stronger with Hsp90α (Chadli et al., 2008; Synoradzki and Bieganowski, 2015; Echeverria et al., 2016).

Here we report that cochaperone p23 interacts preferentially with Hsp90α. Our calculations indicate that the M-domain of Hsp90α forms dimer more stable than the M-domain of Hsp90β. We propose that the M-domains stabilize Hsp90α in a closed conformation that binds Aha1 and p23 with high affinity.

Results

Hsp90α interacts strongly with p23

We determined by co-immunoprecipitation that more p23 cochaperone binds to Hsp90α than to Hsp90β (Figure 2A). We have determined previously, that Aha1 binds preferentially to Hsp90α, and that this binding is mediated by the M-domain of Hsp90α (Synoradzki and Bieganowski, 2015). Therefore we decided to compare p23 binding of the wild type Hsp90 isoforms and the Hsp90 hybrids in which the middle domains were replaced by the homologous fragments from the other isoform (H7 and H8) (Figure 1). We used human HEK-293 cells that expressed alleles of Hsp90 with a point mutation (Hsp90α I128T and Hsp90β I123T) that rendered them insensitive to inhibition by the known specific inhibitor of the Hsp90 ATPase, 17-AAG (Zurawska et al., 2010). HEK-293 cells transfected with these mutants acquired resistance to 17-AAG because the growth of the cells was supported by the transfected Hsp90 despite inactivation of the endogenous Hsp90. The selective pressure of the inhibitor assured high expression of the transfected Hsp90s and allowed for the analysis of the transfected Hsp90 activity without interference from the wild-type Hsp90 (Synoradzki and Bieganowski, 2015). Results of immunoprecipitation revealed that indeed more p23 bound to Hsp90α and Hsp90 hybrid H8 that contained the M-domain of Hsp90α (Figure 2B).

Diagram of Hsp90 hybrid genes used in this work. Sequences of Hsp90α and Hsp90β labeled in dark and light gray, respectively. Numbers indicate amino acid residues at the junctions of the fragments used to create hybrid sequences.
Figure 1:

Diagram of Hsp90 hybrid genes used in this work.

Sequences of Hsp90α and Hsp90β labeled in dark and light gray, respectively. Numbers indicate amino acid residues at the junctions of the fragments used to create hybrid sequences.

p23 binds preferentially to Hsp90α M-domain. (A) Hsp90α binds more p23 than Hsp90β. (B) Hsp90 hybrid that contains α M-domain binds more p23 than Hsp90β. Anti-Flag antibody was used to isolate protein complexes form the cells expressing Flag-tagged Hsp90α, Hsp90β and hybrids H7 and H8. Control represents IP made from the HEK-293 cells transfected with plasmid pcDNA3.1.
Figure 2:

p23 binds preferentially to Hsp90α M-domain.

(A) Hsp90α binds more p23 than Hsp90β. (B) Hsp90 hybrid that contains α M-domain binds more p23 than Hsp90β. Anti-Flag antibody was used to isolate protein complexes form the cells expressing Flag-tagged Hsp90α, Hsp90β and hybrids H7 and H8. Control represents IP made from the HEK-293 cells transfected with plasmid pcDNA3.1.

We have further analyzed differences in p23 binding to Hsp90α and Hsp90β isoforms using isothermal titration calorimetry (ITC). Indeed, p23 bound to Hsp90α about 3 times tighter than to Hsp90β (Figure 3 and Table 1). Surprisingly, the Hsp90β-p23 interaction was much more enthalpy-driven than Hsp90α-p23 interaction, indicating significant differences in the nature of binding.

Isothermal titration calorimetry of p23 interaction with Hsp90α and Hsp90β showing that human p23 binds tighter to Hsp90α, and with a significantly smaller exothermic enthalpy than to Hsp90β. (A) Fitted integrated ITC data curves; (B) raw data curves (upper panel: Hsp90α; lower panel: Hsp90β).
Figure 3:

Isothermal titration calorimetry of p23 interaction with Hsp90α and Hsp90β showing that human p23 binds tighter to Hsp90α, and with a significantly smaller exothermic enthalpy than to Hsp90β.

(A) Fitted integrated ITC data curves; (B) raw data curves (upper panel: Hsp90α; lower panel: Hsp90β).

Table 1:

Affinity constant of p23 binding to Hsp90α and Hsp90β, obtained by isothermal titration calorimetry.

Intact M-domain of Hsp90β is essential for its effective function in yeasts

Yeast Hsp90 genes can be replaced with either human Hsp90 isoform, but yeasts that express Hsp90β grow faster than yeasts that express Hsp90α (Scroggins et al., 2007). We compared growth of yeast strains that expressed only wild type Hsp90α, Hsp90β or one of the hybrid human Hsp90 genes. The results presented in Figure 4A showed that exchange of the whole M-domain in either isoform results in slower growth. Yeast strains carrying hybrids H12, H15, H16, H17 with parts of the M-domain of Hsp90β replaced with the homologous sequences from Hsp90α grew markedly worse than the strain with the complete Hsp90β (t-test, p<0.05). We did not observe such differences in growth rate when the same hybrids were expressed in HEK-293 cells (Figure 4B).

Expression of the hybrid Hsp90s affects growth of yeast but not of human cells. (A) Growth rate of yeasts that express Hsp90s with the hybrid M-domains. Wild type yeasts (WT) and yeasts that express indicated Hsp90 proteins as a sole source of this chaperone were cultured in YPD medium at 30°C. Doubling times were calculated from the growth curves measured by optical density at 600 nm. (B) Human cells that express 17-AAG-resistant alleles of Hsp90 hybrids and non-transfected cells grow at similar rate. HEK-293 cells transfected with plasmids expressing indicated hybrid Hsp90s (H12, 15, 16, 17), wild-type Hsp90α (α), and Hsp90β (β) and selected for 17-AAG resistance were cultured in medium supplemented with 1 μm 17-AAG. The doubling time of the cells was calculated from the growth curves measured using the MTT assay. Graphs represents mean and SD value from three experiments.
Figure 4:

Expression of the hybrid Hsp90s affects growth of yeast but not of human cells.

(A) Growth rate of yeasts that express Hsp90s with the hybrid M-domains. Wild type yeasts (WT) and yeasts that express indicated Hsp90 proteins as a sole source of this chaperone were cultured in YPD medium at 30°C. Doubling times were calculated from the growth curves measured by optical density at 600 nm. (B) Human cells that express 17-AAG-resistant alleles of Hsp90 hybrids and non-transfected cells grow at similar rate. HEK-293 cells transfected with plasmids expressing indicated hybrid Hsp90s (H12, 15, 16, 17), wild-type Hsp90α (α), and Hsp90β (β) and selected for 17-AAG resistance were cultured in medium supplemented with 1 μm 17-AAG. The doubling time of the cells was calculated from the growth curves measured using the MTT assay. Graphs represents mean and SD value from three experiments.

Hsp90α activity in yeasts is limited by its ATPase activity

An amino acid substitution I128T in Hsp90α (and corresponding I123T in Hsp90β) increases an intrinsic ATPase activity of this protein (Zurawska et al., 2010). To test how this mutation affected the yeasts’ growth rate, we compared strains that expressed wild type and mutant of Hsp90α and Hsp90β. We also tested the influence of the I128T (I123T) mutation in the strains with AHA1 gene deletion. Substitution I128T in Hsp90α increased the growth rate of the yeasts that expressed this mutant protein. The I128T mutation reversed also the slow growth phenotype caused by AHA1 gene deletion (Figure 5). The effects of the I123T mutation on the growth of the strains that expressed Hsp90β was not statistically significant.

Mutation that increases Hsp90 ATPase activity compensates AHA1 deletion in yeasts. (A) Growth of the yeast strains that express wild type human Hsp90α, or its I128T mutant as a sole source of this protein was compared to the same strains with an additional deletion of the AHA1 gene. Hsp90α I128T allele rescued slow growth phenotype caused by the lack of Aha1 protein and increased growth rate of the AHA1 wild type strain. (B) Growth of the yeast strains that express Hsp90β or Hsp90β I123T mutant was compared to the strains with an additional deletion of AHA1 gene. No effect on growth of the AHA1 deletion or the I123T mutation in Hsp90β was observed. Growth rates were compared using the t-test.
Figure 5:

Mutation that increases Hsp90 ATPase activity compensates AHA1 deletion in yeasts.

(A) Growth of the yeast strains that express wild type human Hsp90α, or its I128T mutant as a sole source of this protein was compared to the same strains with an additional deletion of the AHA1 gene. Hsp90α I128T allele rescued slow growth phenotype caused by the lack of Aha1 protein and increased growth rate of the AHA1 wild type strain. (B) Growth of the yeast strains that express Hsp90β or Hsp90β I123T mutant was compared to the strains with an additional deletion of AHA1 gene. No effect on growth of the AHA1 deletion or the I123T mutation in Hsp90β was observed. Growth rates were compared using the t-test.

M-domains have higher energy of interactions and large contact surface in Hsp90α than in Hsp90β

The modeled structure of Hsp90α and Hsp90β dimers were subjected to 200 ns molecular dynamics (MD) in implicit solvent. Both structures stabilized after about 100 ns measured with root-mean-square displacement (RMSD) from the initial structure and Hsp90α displacement was slightly larger than this of Hsp90β. For individual domains Hsp90α displacement was larger for N and C domains while for M domain it was the same in both isoforms (Supplementary Figure 1). The RMSD plots demonstrate that models of the Hsp90α and Hsp90β dimers are stable but cannot be used to compare stabilities of the isoforms in the closed conformation as they are calculated as a deviation from the initial structures of Hsp90α and Hsp90β which are different although prepared from the same source, yeast Hsp90 (PDB id:2CG9), using homology modeling. Because of different residues and different length of Hsp90α and Hsp90β their initial and final structures are different what is exemplified in RMSD levels. The RMSF profiles (root-mean-square fluctuations for individual residues) for Hsp90α and Hsp90β were similar to one another and it was not possible to assign groups of residues with markedly different behavior (data not shown). For both simulations we calculated the energy of interactions (sum of van der Waals and electrostatic interactions) and the contact surface [solvent accessible solvent area (SASA)] between domains in a dimer. The results of the simulations are summarized in Table 2. There is a large difference in the energy of interaction for the C and M domains over three-fold for Hsp90α and over two-fold for Hsp90β. There are some differences in domain-domain interactions between Hsp90α and Hsp90β for N-N (15%) and C-C (40%) domains, however, the M-M interactions in Hsp90α is over five-fold larger than in Hsp90β form. Similar dependences are in SASA values: small differences between Hsp90α and Hsp90β for N-N and C-C domain contact, but for M-M domains the SASA value for Hsp90α is over two-fold larger than for Hsp90β. Summing up the interactions of N+M domains between monomers of Hsp90 there is over two-fold increase in energy of interactions for Hsp90α compared to Hsp90β form, and over 30% increase in SASA values. Nearly the same dependence exist after adding the cross-inter-domain interactions due to their small values. Figure 6 presents surfaces of residues participating in the same- and cross-domain interactions on Hsp90α and Hsp90β monomers.

Table 2:

Energy and SASA values for the same-inter-domain interactions and for cross-inter-domain interactions.

The interface of Hsp90 monomer-monomer interactions is larger in Hsp90α. The residues participating in M-M interactions are shown as dark surfaces, the residues participating in M-N and M-C interactions as bright surfaces.
Figure 6:

The interface of Hsp90 monomer-monomer interactions is larger in Hsp90α.

The residues participating in M-M interactions are shown as dark surfaces, the residues participating in M-N and M-C interactions as bright surfaces.

Discussion

The presented results demonstrate that p23 binding to Hsp90α is stronger than its binding to Hsp90β. While this binding is several times weaker than previously reported for corresponding yeast proteins (Panaretou et al., 2002; Richter et al., 2004; Siligardi et al., 2004), it could be explained by the lack of posttranslational modifications in the bacterial expression systems. The analysis of the hybrid Hsp90s composed of fragments of Hsp90α and Hsp90β revealed that strong binding of p23 depends on the presence of the α M-domain in the studied Hsp90. We reported earlier that Hsp90α and Hsp90β differ in their ability to bind to the cochaperone Aha1, and that this difference can be traced to the middle domain of Hsp90 (Synoradzki and Bieganowski, 2015). The preferential binding of Aha1 and p23 to the same Hsp90s might suggest the formation of the triple Aha1-p23-Hsp90 complex. Hsp90 concentration in the cell is much higher than the concentration of p23 and Aha1 and each cochaperone may bind Hsp90 dimer that is unoccupied by the other cochaperone (Ghaemmaghami et al., 2003; Finka and Goloubinoff, 2013). Therefore a noticeable amount of the triple complex could form only if binding of one of the cochaperones stimulates binding of the other cochaperone. The existing data demonstrate that the opposite is true and p23 and Aha1 compete for the binding to Hsp90 (Harst et al., 2005). Therefore we concluded that both cochaperones bind independently Hsp90 in a conformation that is stabilized by the α M-domain.

The M-domain of Hsp90 is composed of two distinct subdomains (Meyer et al., 2003). Our results proved that, despite high homology, the M-domains of Hsp90α and Hsp90β formed functional units in which parts cannot be interchanged without adverse effect on their function in yeasts. Whatever the reasons for the decreased activity of the hybrids with the swapped M-domain or its fragments, they are visible only in the heterologous environment of the yeast cell. The same hybrids expressed in the human cells support their growth similarly to the wild type Hsp90s. This difference may result from the more effective interaction of the hybrids with human cochaperones than with their yeast homologs.

Yeast Hsp90 ATPase is 10-fold more active compared to the human Hsp90 (Richter et al., 2008). Growth of the yeast strain with deleted AHA1 or SBA1 genes is not compromised unless cells are subjected to stressful conditions, indicating that yeast Hsp90 is relatively independent from the cochaperones (Fang et al., 1998; Panaretou et al., 2002). Yeast Hsp90 can be substituted by its human homologs, but the resulting strains grow slower (Scroggins et al., 2007). This indicates that unlike yeast Hsp90, human Hsp90, especially Hsp90α, needs help from cochaperones to function effectively. In yeasts that express only human Hsp90, Hsp90α activity depends on Aha1, whereas Hsp90β is not affected by the absence of this protein (Synoradzki and Bieganowski, 2015). The low intrinsic ATPase activity is clearly a limiting factor for the human Hsp90α in yeasts because strains expressing Hsp90α mutant with increased ATPase activity grows faster than strains that express wild type protein. Moreover, this mutation compensates for the lack of Aha1, resulting in the growth rate restoration.

The recent reports on the Hsp90β-specific inhibitor, and Aha1-Hsp90α interaction, as well as results of this investigation define the M-domain as a region of the major differences between Hsp90 isoforms (Synoradzki and Bieganowski, 2015; Yim et al., 2016). Our calculations of the energy of interaction and the surface area (SASA) of the interacting domains in both Hsp90 isoforms suggest that Hsp90α M-domain forms dimer more stable than its Hsp90β homolog. Therefore Hsp90α may adopt closed conformation more readily than Hsp90β. Both Aha1 and p23 bind to Hsp90 in its closed conformation, with ATP occupying the ATPase catalytic center (Ali et al., 2006; Retzlaff et al., 2010). As we demonstrated, Hsp90α is more dependent on the interaction with Aha1 and p23 and has higher affinity for these cochaperones. The M domains dimerization may help Hsp90α to adopt the closed conformation suitable for the Aha1 and p23 binding. SASA can account for an increase of entropy of binding (diminishing of solvent accessible surface mostly due to hydrophobic interactions) which is lacking in energy of interaction. Interestingly, the SASA value for M-M interactions in Hsp90α is over two times larger than for Hsp90β, and the energy of domain-domain interaction is over five-fold larger in Hsp90α. The RMSD values are nearly the same for M domains of Hsp90α and Hsp90β so such large difference of energy of interactions does not result from a simple relaxation of the structure but is a consequence of intrinsic interactions. For larger stability of the Hsp90 closed conformation the total N+M interactions are important as they are breaking during opening of Hsp90 and they are binding more strongly (over two-fold) for Hsp90α (SASA larger over 30%) implicating higher stability of the whole dimer.

If Hsp90α is predominantly in a closed conformation, while Hsp90β – in a predominantly open, then the change in enthalpy may indicate that Hsp90β must undergo a large conformational change prior to the p23 binding reaction. This large conformational change may be reflected by the large exothermic enthalpy change, which is not observed in the Hsp90α case.

Hsp90α binds stronger than Hsp90β to several client proteins and in the cells subjected to heat shock an affinity of Hsp90α for several protein kinases increases, whereas an affinity of Hsp90β remains unchanged (Taherian et al., 2008; Prince et al., 2015). It is therefore possible that the Hsp90α stronger affinity for cochaperones and clients evolved as an adaptation of Hsp90α to function effectively under stressful conditions.

Materials and methods

Plasmids

Plasmids with Hsp90 hybrid genes H7 and H8 were described earlier (Synoradzki and Bieganowski, 2015). Plasmids with the Hsp90 hybrid genes for this study were constructed as described earlier using PCR-generated fragments of the previously constructed plasmids (Synoradzki and Bieganowski, 2015). Details on the Hsp90 hybrids used in this study are provided in Figure 1. The names of the hybrids are consistent with previous work (Synoradzki and Bieganowski, 2015). For yeast expression hybrid Hsp90 genes were cloned under the control of a strong constitutive promoter of TDH3 gene in p423TDH3 plasmid with HIS3 marker, and 2μ origin of replication (Zurawska et al., 2010). Plasmids that expressed in human cells wild type or 17-AAG resistant alleles of Hsp90 with single amino acid substitution, Hsp90α Ile128Thr and Hsp90β Ile123Thr were described earlier (Zurawska et al., 2010).

Sequences of primers used to generate Hsp90α/Hsp90β junctions during Hsp90 hybrids construction are listed in the Supplementary Table 1.

Human Hsp90 expression plasmids for Escherichia coli were constructed as described earlier (Cikotiene et al., 2009; Zubriene et al., 2010). Human p23 was cloned into pET28a-6×His-p23 vector, adding a His-tag at the N-terminus.

HEK-293 cells culture and immunoprecipitation

Human embryonic kidney cells HEK-293 were cultured in Iscove’s modified Dulbecco’s medium, supplemented with 17-allylamino-demethoxygeldamycin (17-AAG) when necessary. For the growth rate assay 2000 cells/well were seeded in a 96-well plate in 200 μl of IMDM medium without phenol red. Growth of the cells was measured every 24 h using a tetrazolium salt MTT-based assay (Cell Proliferation Kit I, Roche Poland, Warsaw, Poland). One hundred microliters of medium were removed from the assayed wells and 10 μl of MTT reagent was added. The labeling reaction was conducted for 3 h at 37°C in a humidified chamber with 5% CO2. The reaction was terminated by the overnight incubation with 100 μl of the solubilization solution. Optical density at 600 nm was measured using an Epoch Microplate Spectrophotometer (BioTek, Winooski, VT, USA). Doubling time was calculated using the equation: Doubling time=(time1−time2)×log(2)/log(OD1/OD2).

For immunoprecipitation HEK-293 cells were transfected with plasmids carried 17-AAG resistant mutants of wild type and hybrid Flag-tagged Hsp90 and selected in medium containing 1 μm 17-AAG (Sigma, St. Louis, MO, USA). Immunoprecipitation was performed as described earlier (Synoradzki and Bieganowski, 2015). Immunoprecipitation of the wild type Hsp90α and Hsp90β was carried out 48 h after transfection of HEK-293 cells with plasmids that expressed Flag-tagged alleles of these proteins. The buffer used for this immunoprecipitation was supplemented with 10 mm sodium molybdate to stabilize Hsp90-p23 complexes (Hartson et al., 1999; Sun et al., 2012).

Protein purification and isothermal titration calorimetry

N-His-tagged full-length Hsp90 proteins were expressed in Escherichia coli BL21(DE3) and purified as described previously (Zubriene et al., 2010). N-His-tagged p23 was also expressed in E. coli BL21(DE3) and purified using a Ni-IDA affinity column in a buffer containing 20 mm Tris (pH 7.4), 0.2 m NaCl, 10 mm glycerol, 20 mm 2-mercaptoethanol, and eluted using imidazole gradient (from 50 mm to 0.6 m). The proteins were dialyzed against buffer containing 50 mm Tris (pH 7.5, 25°C), 100 mm NaCl, 5 mm MgCl2. AMP-PNP (Jena Bioscience, Jena, Germany) was added to all protein samples at final concentration of 2 mm.

ITC experiments were carried out on MicroCal iTC200 isothermal titration calorimeter (Malvern, Westborough, USA) with the active cell volume of 0.2048 ml. The concentration of Hsp90 protein in the cell was 40 μm, while the syringe contained 400 μm of p23. Titrations were carried out at 25°C, using 19 injections of 2 μl each, injected at 150 s intervals. Data was processed and analysed using Origin software, provided by the calorimeter manufacturer. All experiments were repeated at least three times.

Yeast strains construction and growth assays

Saccharomyces cerevisiae strain used in this study had genotype ade2 ura3 trp1 leu2 his3 lys2 MATa hsc82::G418 hsp82::G418 and carried a Hsp90α gene cloned in a plasmid with URA3 marker (Zurawska et al., 2010). Plasmids with HIS3 marker that carried different forms of the human Hsp90 were introduced by transformation into this strain. The HIS3+ transformants were selected for the URA3 marker loss in the media containing 1% 5-fluoroorotic acid (Boeke et al., 1987). The resulting transformants were cultured at 30°C in YPD medium supplemented with adenosine.

Yeasts growth assays were performed at 30°C in liquid YPD media supplemented with 40 μg/ml of adenosine sulfate. Overnight cultures were diluted to initial concentration of OD600=0.05 and dispensed in aliquots of 200 μl/well in 96-well plate. Plate was incubated in an Epoch Microplate Spectrophotometer (BioTek) in a room termostated at 30°C and growth of the cultures was measured at 30 min intervals at 600 nm. Data were used to fit the exponential growth equation. Doubling time was calculated from the part of data collected during an exponential growth phase.

Molecular modeling

The structure of human dimeric Hsp90α and Hsp90β with ATP bound was prepared on a basis of the crystal structure of yeast Hsp90-Sba1 closed chaperone complex from the Protein Data Bank (PDB id:2CG9) (Ali et al., 2006). The homology modeling service Phyre2 (Kelley et al., 2015) was used to create human Hsp90α and Hsp90β monomers. There are lacking fragments (compared to the template crystal structure) in Hsp90α: 1–15 (N-domain), 230–281 (M-domain), 619–631 (C-domain), 690–724 (C-terminus); in Hsp90β: 1–10 (N-domain), 225–273 (M-domain), 611–623 (C-domain). The above lacking fragments were modeled in Phyre2 using a library of fragments of known protein structures. The dimeric forms of Hsp90 were obtained by structural alignment with crystal structure of yeast Hsp82 in PYMOL program (DeLano). All energy minimizations and MD simulations were performed in NAMD program version 2.10 using all-atom force field CHARMM27 (Phillips et al., 2005) in implicit solvent. ATP was parametrized in the same force field. All figures of molecular structures were created using the VMD program (v.1.9.2) (Humphrey et al., 1996). The simulated systems were initially subjected to 10 000 steps of energy minimization and then 100 ns MD of thermal equilibration with increasing temperature from 20 K to 298 K. The MD simulations were conducted using Langevin (stochastic) dynamics (Kubo et al., 1991). It introduces viscosity which is lacking in simulations in implicit solvents. In this approach the molecules in the system interact with a stochastic heat bath via random forces and dissipative forces. The friction coefficient for Langevin dynamics of 50 ps−1 was used and temperature was set to 298 K. Non-bonded interactions were damped employing a switching function for van der Waals and electrostatic interactions using cutoff of 14 Å. For each investigated system (Hsp90α and Hsp90β dimers) 200 ns MD simulation was performed with a time step of 2 fs. All bond lengths were constrained using SHAKE algorithm (Ryckaert et al., 1977).

Acknowledgments

This work was supported by the Polish National Science Centre grant no. NN303818640 and 2016/23/B/NZ6/02536. The sponsor played no role in study design, in the collection, analysis and interpretation of data, in the writing of the report or in the decision to submit the article for publication. The authors declare no conflicts of interest.

References

  • Ali, M.M., Roe, S.M., Vaughan, C.K., Meyer, P., Panaretou, B., Piper, P.W., Prodromou, C., and Pearl, L.H. (2006). Crystal structure of an Hsp90-nucleotide-p23/Sba1 closed chaperone complex. Nature 440, 1013–1017. CrossrefPubMedGoogle Scholar

  • Boeke, J.D., Trueheart, J., Natsoulis, G., and Fink, G.R. (1987). 5-Fluoroorotic acid as a selective agent in yeast molecular genetics. Methods Enzymol. 154, 164–175. PubMedCrossrefGoogle Scholar

  • Chadli, A., Felts, S.J., and Toft, D.O. (2008). GCUNC45 is the first Hsp90 co-chaperone to show α/β isoform specificity. J. Biol. Chem. 283, 9509–9512. Web of SciencePubMedCrossrefGoogle Scholar

  • Chen, B., Zhong, D., and Monteiro, A. (2006). Comparative genomics and evolution of the HSP90 family of genes across all kingdoms of organisms. BMC Genomics 7, 156. PubMedCrossrefGoogle Scholar

  • Cikotiene, I., Kazlauskas, E., Matuliene, J., Michailoviene, V., Torresan, J., Jachno, J., and Matulis, D. (2009). 5-Aryl-4-(5-substituted-2,4-dihydroxyphenyl)-1,2,3-thiadiazoles as inhibitors of Hsp90 chaperone. Bioorg. Med. Chem. Lett. 19, 1089–1092. CrossrefWeb of SciencePubMedGoogle Scholar

  • Cortes-Gonzalez, C., Barrera-Chimal, J., Ibarra-Sanchez, M., Gilbert, M., Gamba, G., Zentella, A., Flores, M.E., and Bobadilla, N.A. (2010). Opposite effect of Hsp90α and Hsp90β on eNOS ability to produce nitric oxide or superoxide anion in human embryonic kidney cells. Cell Physiol. Biochem. 26, 657–668. PubMedCrossrefWeb of ScienceGoogle Scholar

  • Csermely, P., Kajtar, J., Hollosi, M., Jalsovszky, G., Holly, S., Kahn, C.R., Gergely, P., Jr., Soti, C., Mihaly, K., and Somogyi, J. (1993). ATP induces a conformational change of the 90-kDa heat shock protein (hsp90). J. Biol. Chem. 268, 1901–1907. PubMedGoogle Scholar

  • DeLano, W.L. The PyMOL Molecular Graphics System (San Carlos, CA, USA: DeLano Scientific LLC). Google Scholar

  • Echeverria, P.C., Bernthaler, A., Dupuis, P., Mayer, B., and Picard, D. (2011). An interaction network predicted from public data as a discovery tool: application to the Hsp90 molecular chaperone machine. PLoS One 6, e26044. CrossrefPubMedWeb of ScienceGoogle Scholar

  • Echeverria, P.C., Briand, P.A., and Picard, D. (2016). A remodeled Hsp90 molecular chaperone ensemble with the novel cochaperone aarsd1 is required for muscle differentiation. Mol. Cell Biol. 36, 1310–1321. Web of ScienceCrossrefPubMedGoogle Scholar

  • Fang, Y., Fliss, A.E., Rao, J., and Caplan, A.J. (1998). SBA1 encodes a yeast hsp90 cochaperone that is homologous to vertebrate p23 proteins. Mol. Cell Biol. 18, 3727–3734. PubMedCrossrefGoogle Scholar

  • Finka, A. and Goloubinoff, P. (2013). Proteomic data from human cell cultures refine mechanisms of chaperone-mediated protein homeostasis. Cell Stress Chaperones 18, 591–605. Web of ScienceCrossrefPubMedGoogle Scholar

  • Gao, Y., Yechikov, S., Vazquez, A.E., Chen, D., and Nie, L. (2013). Distinct roles of molecular chaperones HSP90α and HSP90β in the biogenesis of KCNQ4 channels. PLoS One 8, e57282. Web of ScienceCrossrefPubMedGoogle Scholar

  • Ghaemmaghami, S., Huh, W.K., Bower, K., Howson, R.W., Belle, A., Dephoure, N., O’Shea, E.K., and Weissman, J.S. (2003). Global analysis of protein expression in yeast. Nature 425, 737–741. CrossrefPubMedGoogle Scholar

  • Grad, I., Cederroth, C.R., Walicki, J., Grey, C., Barluenga, S., Winssinger, N., De Massy, B., Nef, S., and Picard, D. (2010). The molecular chaperone Hsp90β is required for meiotic progression of spermatocytes beyond pachytene in the mouse. PLoS One 5, e15770. CrossrefPubMedWeb of ScienceGoogle Scholar

  • Harst, A., Lin, H., and Obermann, W.M. (2005). Aha1 competes with Hop, p50 and p23 for binding to the molecular chaperone Hsp90 and contributes to kinase and hormone receptor activation. Biochem. J. 387, 789–796. PubMedCrossrefGoogle Scholar

  • Hartson, S.D., Thulasiraman, V., Huang, W., Whitesell, L., and Matts, R.L. (1999). Molybdate inhibits hsp90, induces structural changes in its C-terminal domain, and alters its interactions with substrates. Biochemistry 38, 3837–3849. PubMedCrossrefGoogle Scholar

  • Humphrey, W., Dalke, A., and Schulten, K. (1996). VMD: visual molecular dynamics. J. Mol. Graphics Model. 14, 33–38. CrossrefGoogle Scholar

  • Kajiwara, C., Kondo, S., Uda, S., Dai, L., Ichiyanagi, T., Chiba, T., Ishido, S., Koji, T., and Udono, H. (2012). Spermatogenesis arrest caused by conditional deletion of Hsp90α in adult mice. Biol. Open. 1, 977–982. Web of SciencePubMedCrossrefGoogle Scholar

  • Kelley, L.A., Mezulis, S., Yates, C.M., Wass, M.N., and Sternberg, M.J. (2015). The Phyre2 web portal for protein modeling, prediction and analysis. Nat. Protoc. 10, 845–858. PubMedWeb of ScienceCrossrefGoogle Scholar

  • Kubo, R., Toda, M., and Hashitsume, N. (1991). Statistical Physics II: Nonequilibrium Statistical Mechanics (Berlin, Germany: Springer-Verlag). Google Scholar

  • Meyer, P., Prodromou, C., Hu, B., Vaughan, C., Roe, S.M., Panaretou, B., Piper, P.W., and Pearl, L.H. (2003). Structural and functional analysis of the middle segment of hsp90: implications for ATP hydrolysis and client protein and cochaperone interactions. Mol. Cell 11, 647–658. PubMedCrossrefGoogle Scholar

  • Nemoto, T., Ohara-Nemoto, Y., Ota, M., Takagi, T., and Yokoyama, K. (1995). Mechanism of dimer formation of the 90-kDa heat-shock protein. Eur. J. Biochem. 233, 1–8. PubMedCrossrefGoogle Scholar

  • Panaretou, B., Siligardi, G., Meyer, P., Maloney, A., Sullivan, J.K., Singh, S., Millson, S.H., Clarke, P.A., Naaby-Hansen, S., Stein, R., et al. (2002). Activation of the ATPase activity of hsp90 by the stress-regulated cochaperone aha1. Mol. Cell 10, 1307–1318. PubMedCrossrefGoogle Scholar

  • Phillips, J.C., Braun, R., Wang, W., Gumbart, J., Tajkhorshid, E., Villa, E., Chipot, C., Skeel, R.D., Kale, L., and Schulten, K. (2005). Scalable molecular dynamics with NAMD. J. Comput. Chem. 26, 1781–1802. CrossrefPubMedGoogle Scholar

  • Prince, T.L., Kijima, T., Tatokoro, M., Lee, S., Tsutsumi, S., Yim, K., Rivas, C., Alarcon, S., Schwartz, H., Khamit-Kush, K., et al. (2015). Client proteins and small molecule inhibitors display distinct binding preferences for constitutive and stress-induced Hsp90 isoforms and their conformationally restricted mutants. PLoS One 10, e0141786. PubMedCrossrefWeb of ScienceGoogle Scholar

  • Retzlaff, M., Hagn, F., Mitschke, L., Hessling, M., Gugel, F., Kessler, H., Richter, K., and Buchner, J. (2010). Asymmetric activation of the hsp90 dimer by its cochaperone aha1. Mol. Cell 37, 344–354. PubMedCrossrefWeb of ScienceGoogle Scholar

  • Richter, K., Walter, S., and Buchner, J. (2004). The Co-chaperone Sba1 connects the ATPase reaction of Hsp90 to the progression of the chaperone cycle. J. Mol. Biol. 342, 1403–1413. PubMedCrossrefGoogle Scholar

  • Richter, K., Soroka, J., Skalniak, L., Leskovar, A., Hessling, M., Reinstein, J., and Buchner, J. (2008). Conserved conformational changes in the ATPase cycle of human Hsp90. J. Biol. Chem. 283, 17757–17765. CrossrefWeb of SciencePubMedGoogle Scholar

  • Ryckaert, J., Ciccotti, G., and Berendsen, H. (1977). Numerical integration of the Cartesian equations of motion of a system with constraints. J. Comput. Phys. 23, 327–341. CrossrefGoogle Scholar

  • Scroggins, B.T., Robzyk, K., Wang, D., Marcu, M.G., Tsutsumi, S., Beebe, K., Cotter, R.J., Felts, S., Toft, D., Karnitz, L., et al. (2007). An acetylation site in the middle domain of Hsp90 regulates chaperone function. Mol. Cell 25, 151–159. PubMedWeb of ScienceCrossrefGoogle Scholar

  • Siligardi, G., Hu, B., Panaretou, B., Piper, P.W., Pearl, L.H., and Prodromou, C. (2004). Co-chaperone regulation of conformational switching in the Hsp90 ATPase cycle. J. Biol. Chem. 279, 51989–51998. CrossrefPubMedGoogle Scholar

  • Sullivan, W., Stensgard, B., Caucutt, G., Bartha, B., McMahon, N., Alnemri, E.S., Litwack, G., and Toft, D. (1997). Nucleotides and two functional states of hsp90. J. Biol. Chem. 272, 8007–8012. PubMedCrossrefGoogle Scholar

  • Sun, L., Prince, T., Manjarrez, J.R., Scroggins, B.T., and Matts, R.L. (2012). Characterization of the interaction of Aha1 with components of the Hsp90 chaperone machine and client proteins. Biochim. Biophys. Acta 1823, 1092–1101. Web of ScienceCrossrefPubMedGoogle Scholar

  • Synoradzki, K. and Bieganowski, P. (2015). Middle domain of human Hsp90 isoforms differentially binds Aha1 in human cells and alters Hsp90 activity in yeast. Biochim. Biophys. Acta 1853, 445–452. CrossrefPubMedWeb of ScienceGoogle Scholar

  • Taherian, A., Krone, P.H., and Ovsenek, N. (2008). A comparison of Hsp90α and Hsp90β interactions with cochaperones and substrates. Biochem. Cell Biol. 86, 37–45. PubMedWeb of ScienceCrossrefGoogle Scholar

  • Voss, A.K., Thomas, T., and Gruss, P. (2000). Mice lacking HSP90β fail to develop a placental labyrinth. Development 127, 1–11. PubMedGoogle Scholar

  • Yim, K.H., Prince, T.L., Qu, S., Bai, F., Jennings, P.A., Onuchic, J.N., Theodorakis, E.A., and Neckers, L. (2016). Gambogic acid identifies an isoform-specific druggable pocket in the middle domain of Hsp90β. Proc. Natl. Acad. Sci. USA 113, E4801–E4809. Google Scholar

  • Zhao, R., Davey, M., Hsu, Y.C., Kaplanek, P., Tong, A., Parsons, A.B., Krogan, N., Cagney, G., Mai, D., Greenblatt, J., et al. (2005). Navigating the chaperone network: an integrative map of physical and genetic interactions mediated by the hsp90 chaperone. Cell 120, 715–727. CrossrefPubMedGoogle Scholar

  • Zubriene, A., Gutkowska, M., Matuliene, J., Chaleckis, R., Michailoviene, V., Voroncova, A., Venclovas, C., Zylicz, A., Zylicz, M., and Matulis, D. (2010). Thermodynamics of radicicol binding to human Hsp90 α and β isoforms. Biophys. Chem. 152, 153–163. Web of SciencePubMedCrossrefGoogle Scholar

  • Zuehlke, A.D., Beebe, K., Neckers, L., and Prince, T. (2015). Regulation and function of the human HSP90AA1 gene. Gene 570, 8–16. Web of SciencePubMedCrossrefGoogle Scholar

  • Zurawska, A., Urbanski, J., Matuliene, J., Baraniak, J., Klejman, M.P., Filipek, S., Matulis, D., and Bieganowski, P. (2010). Mutations that increase both Hsp90 ATPase activity in vitro and Hsp90 drug resistance in vivo. Biochim. Biophys. Acta 1803, 575–583. Web of ScienceCrossrefPubMedGoogle Scholar

Supplemental Material:

The online version of this article offers supplementary material (https://doi.org/10.1515/hsz-2017-0172).

About the article

Received: 2017-05-31

Accepted: 2017-12-14

Published Online: 2018-01-16

Published in Print: 2018-03-28


Citation Information: Biological Chemistry, Volume 399, Issue 4, Pages 337–345, ISSN (Online) 1437-4315, ISSN (Print) 1431-6730, DOI: https://doi.org/10.1515/hsz-2017-0172.

Export Citation

©2018 Walter de Gruyter GmbH, Berlin/Boston.Get Permission

Supplementary Article Materials

Comments (0)

Please log in or register to comment.
Log in