Hsp90 is an ATP-dependent molecular chaperone that assists folding and conformational maturation/maintenance of many proteins. It is a potential cancer drug target because it chaperones oncoproteins. A prokaryotic homolog of Hsp90 (HtpG) is essential for thermo-tolerance in some bacteria and virulence of zoonotic pathogens. To identify a new class of small molecules which target prokaryotic and eukaryotic Hsp90s, we studied the effects of a naturally occurring cyclic sesquiterpene, zerumbone, which inhibits proliferation of a wide variety of tumor cells, on the activity of Hsp90. Zerumbone enhanced the ATPase activity of cyanobacterial Hsp90 (Hsp90SE), yeast Hsp90, and human Hsp90α. It also enhanced the catalytic efficiency of Hsp90SE by greatly increasing kcat. Mass analysis showed that zerumbone binds to cysteine side chains of Hsp90SE covalently. Mutational studies identified 3 cysteine residues (one per each domain of Hsp90SE) that are involved in the enhancement, suggesting the presence of allosteric sites in the middle and C-terminal domains of Hsp90SE. Treatment of cyanobacterial cells with zerumbone caused them to become very temperature-sensitive, a phenotype reminiscent of cyanobacterial Hsp90 mutants, and also decreased the cellular level of linker polypeptides that are clients for Hsp90SE. Zerumbone showed cellular toxicity on cancer-derived mammalian cells by inducing apoptosis. In addition, zerumbone inhibited the binding of Hsp90/Cdc37 to client kinases. Altogether, we conclude that modification of cysteine residues of Hsp90 by zerumbone enhances its ATPase activity and inhibits physiological Hsp90 function. The activation of Hsp90 may provide new strategies to inhibit its chaperone function in cells.

Introduction

The 90-kDa heat shock proteins (Hsp90s) are a family of molecular chaperones that are evolutionarily conserved. Hsp90 in co-operation with other chaperones and co-chaperones plays essential roles in stabilization, regulation, and/or assembly of a wide range of protein clients that include the eukaryotic protein kinases, transcription factors, and mutated/overexpressed oncoproteins [1,2]. Hsp90 consists of three domains: an N-terminal domain, a middle domain, and a C-terminal domain. It functions as a homodimer which is formed via the C-terminal domain interactions [3,4]. Hsp90 shows a very weak ATPase activity. An ATP-binding site is present in the N-terminal domain of Hsp90 [5]. ATP binding promotes a transient N-terminal dimerization of the two protomers, which is accompanied by docking the middle domain onto the N-terminal domain of the same protomer. These structural rearrangements lead to the hydrolysis of ATP to ADP [1,2]. It is assumed that ATP binding and hydrolysis are essential to the in vivo function of the cytosolic Hsp90 in eukaryotes [6,7]. However, recent evidence indicates that ATP hydrolysis is not required for viability at least in yeast [8].

Hsp90s are present in cytosol, mitochondria, chloroplast, and endoplasmic reticulum in eukaryotic cells [9]. Prokaryotes also possess a homolog of Hsp90 called HtpG [10]. Compared with eukaryotic Hsp90s, much less work has been done to elucidate the chaperone function/mechanism of prokaryotic Hsp90s in vitro and in vivo. We have been working on Hsp90 in the cyanobacterium Synechococcus elongatus PCC 7942, an oxygenic photosynthetic prokaryote. Hsp90 in cyanobacteria is not essential under normal conditions, but becomes essential/important under stresses [1113]. We have identified cellular clients for the cyanobacterial Hsp90 [1416], and showed that Hsp90 physically interacts with Hsp70 (DnaK) to assist unfolding/folding of denatured proteins [17]. E. coli Hsp90 also associates with DnaK directly for collaboration in protein folding [18], suggesting that a co-chaperone may not be necessary for the prokaryotic Hsp90/Hsp70 chaperone machine to function. This is in contrast with the fact that cytosolic Hsp90s in eukaryotes require the assistance of co-chaperones [1].

Inhibitors and activators are useful to elucidate the mechanisms of enzyme-catalyzed reactions. In the same way, small molecules that control the Hsp90 ATPase activity may facilitate elucidation of the chaperone mechanism of Hsp90. These molecules may also become drug candidates since Hsp90 is being studied as a potential drug target for the treatment of cancer [1922]. Hsp90 is required for stability and function of mutated and overexpressed oncoproteins, and cancer cells are more sensitive to Hsp90 inhibitors than non-transformed cells [23]. Most of the Hsp90-targetting compounds including those that are clinically relevant are inhibitors that bind to the ATP-binding pocket, located in the N-terminal domain of Hsp90, to block the ATP-biding/hydrolysis [1922]. There are several activators for Hsp90 although not many are known. Tamoxifen and its metabolite 4-hydroxytamoxifen [24], Hsp90Mod3/4 [25], and goniothalamin [26] are among the few known activators for Hsp90. In silico, biochemical, or biophysical analyses indicate that all of them bind to the N-terminal domain of Hsp90. Tamoxifen and goniothalamin show anticancer activity. Recently, derivatives of 2-phenylbenzofurans that were rationally designed to bind to the middle-domain/C-terminal domain interface have been found to stimulate the Hsp90 ATPase activity allosterically [27].

Zerumbone, a cyclic sesquiterpene, was first isolated from the rhizome oil of Zingiber zerumbet and was determined to possess anti-inflammatory effect [28]. The most investigated biological activity of zerumbone is its ability to inhibit proliferation of a wide variety of tumor cells including cells derived from colon and breast cancers [29], leukemia, and myeloid and liver cancers [30]. Zerumbone induces suppression of skin tumor promoter phorbol ester-induced Epstein–Barr virus activation in Raji cells [28], inhibition of superoxide and nitric oxide generation, cyclooxygenase-2 expression, tumor necrosis factor-α release in activated leukocytes, and apoptosis in human colonic adenocarcinoma cell lines [31]. In vivo studies have shown that zerumbone suppresses skin tumors in mice [32], colon and lung carcinogenesis in mice [33], azoxymethane-induced aberrant crypt foci formation in rats [29], and human breast cancer-induced bone loss in nude mice [34]. More recently, the key cell signaling pathways modulated by zerumbone have been described [35]. However, the molecular mechanisms that underlie these activities are yet to be characterized [36].

In the present study, we found that the ATPase activity of Hsp90 is enhanced by its modification with zerumbone via cysteine residues present in any of the three domains of Hsp90. As far as we know, the present study is the first one to show enhancement of the Hsp90 ATPase activity by cysteine modification. Our study suggests that although zerumbone activates the ATPase activity of Hsp90, it inhibits the cellular function of Hsp90 in both prokaryotes and eukaryotes. Zerumbone provides a new chemical tool to study a cysteine residue of Hsp90 for controlling the ATPase activity of Hsp90 and to study the effect of unregulated Hsp90 activation on biological function of Hsp90 in prokaryotes and eukaryotes.

Experimental

Reagents

Zerumbone (MW = 218.3) whose structure is shown in Figure 1 was purchased from AdipoGen. Radicicol and geldanamycin were purchased from Sigma–Aldrich and Life Technologies, respectively.

Zerumbone (ZER) activates the ATPase activity of the cyanobacterial Hsp90SE.

Figure 1.
Zerumbone (ZER) activates the ATPase activity of the cyanobacterial Hsp90SE.

Chemical structure of ZER whose molecular mass is 218.33 (A). Effect of varying concentrations of ZER on the ATPase activities of Hsp90SE (B) and effect of ZER (50 µg/ml of a reaction mixture) on the Michaelis constant (Km) for ATP and kcat of Hsp90SE (C and D). The ATPase activity in the absence of ZER in B (100%) was 1.20 ± 0.19 (min−1). The activities were measured in the presence of 2 µM Hsp90SE. In all the points, data from three replicates are presented as mean ± SEM. Some error bars are covered by plot symbols. The Km and Vmax values were estimated using nonlinear fitting operated by the GraphPad Prism software.

Figure 1.
Zerumbone (ZER) activates the ATPase activity of the cyanobacterial Hsp90SE.

Chemical structure of ZER whose molecular mass is 218.33 (A). Effect of varying concentrations of ZER on the ATPase activities of Hsp90SE (B) and effect of ZER (50 µg/ml of a reaction mixture) on the Michaelis constant (Km) for ATP and kcat of Hsp90SE (C and D). The ATPase activity in the absence of ZER in B (100%) was 1.20 ± 0.19 (min−1). The activities were measured in the presence of 2 µM Hsp90SE. In all the points, data from three replicates are presented as mean ± SEM. Some error bars are covered by plot symbols. The Km and Vmax values were estimated using nonlinear fitting operated by the GraphPad Prism software.

Antibodies and expression plasmids

Anti-30 kDa rod linker polypeptide antibody and anti-Hsp90 antibody were described previously [16,37,38]. Anti-PARP and an antibody specific for caspase-cleaved form of PARP (Asp214) were from Cell Signaling Technology. Anti-tubulin antibody (clone DM1A), HRP-conjugated anti-FLAG antibody (clone M2), and anti-FLAG affinity resin (M2) were from Sigma–Aldrich. Anti-Hsp70 antibody (clone K20) and HRP-conjugated anti-Cdc37 antibody (clone E4) were from Santa Cruz Biotechnology.

Mammalian expression plasmid for FLAG-tagged Cdk4 was described previously [39]. The catalytic domain (amino acid 343–648) of human Raf1 was produced by PCR amplification with primers 5- GGATCCACCATGGAAATAGAAGCCAGTGAAGTG-3′ and 5′-GCTAGGATCCCTAGAAGACAGGCAGCCTCGG-3′, and the resulting fragment was isolated, digested with BamHI, and ligated into the BamHI site of pFLAG-CMV2 (Sigma) to construct the mammalian expression plasmid for FLAG-tagged Raf1CD. The whole coding region was confirmed by direct sequencing.

Purification of various molecular chaperones from the cyanobacterium Synechococcus elongatus PCC 7942, E. coli, the fission yeast Schizosaccharomyces pombe, and Homo sapiens

Construction of a strain which overexpresses C-terminally His-tagged HtpG (Hsp90SE) and C-terminally His-tagged DnaK2 from S. elongatus PCC 7942, and their purification method, were described previously [16,17]. Purification of C-terminally His-tagged HtpG from E. coli (Hsp90EC) was described previously [16]. Construction, expression, and purification of C-terminally His-tagged ClpB1 from S. elongatus PCC 7942 and N-terminally GST-tagged human Hsp90α will be reported elsewhere. The GST tag was removed by PreScission Protease (GE Healthcare Life Sciences). Construction of a strain which overexpresses non-tagged Hsp90 from S. pombe, and its purification method, were described previously [40]. E. coli GroELwas purchased from Takara Bio.

Construction of cysteine substitution mutants of the cyanobacterial Hsp90SE

Site-directed mutagenesis to change a cysteine(s) to an alanine(s) was performed as described in the QuikChange II site-directed mutagenesis kit manual (Agilent Technologies). PCR reactions were performed with the pET-21a expression vector (Invitrogen) containing the full-length htpG gene from S. elongatus PCC 7942 (AB010001; see references [11] and [16]) and mutagenic primers given in Table 1 which were synthesized by Eurofins Genomics. All the site-directed mutations were confirmed by sequencing the constructed plasmid DNAs. To construct a triple cysteine mutant Hsp90SE (C152A/C367A/C601A), pET21a carrying the double mutant (C367A/C601A) htpG gene was first constructed with pET21a containing the C367A htpG mutant as a template, and CHG_C601A Fw and CHG_C601A Rv as a pair of primers. Then, using the plasmid as a template, PCR was performed with CHG_C152A Fw and CHG_C152A Rv as a pair of primers. All the mutant Hsp90s were purified as described above for the wild-type Hsp90SE.

Table 1
Primers used for construction of Hsp90SE mutants

A pair of forward (Fw) and reverse (Rv) primers are given in the table.

Mutation Primer names Primer sequences 
C152A CHG_C152A Fw
CHG_C152A Rv 
GCGACACCTGTTCATTGGACTGCTGATGGATCTCC
GGAGATCCATCAGCAGTCCAATGAACAGGTGTCGC 
C197A CHG_C197A Fw
CHG_C197A Rv 
GCTTGTTAAGACCTATGCTGACTTCATGCCCGTCC
GAACGGGCATGAAGTCAGCATAGGTCTTAACAAGC 
C282A CHG_C282A Fw
CHG_C282A Rv 
CTATTCGCCAATCAAGTCTTTGTCAGCGATCACTG
CAGTGATCGCTGACAAAGACTTGATTGGCGAATAG 
C291A CHG_C291A Fw
CHG_C291A Rv 
GTCAGCGATCACGCTGAAGAAGTTGTGCCGCG
CGCGGCACAACTTCTTCAGCGTGATCGCTGAC 
C367A CHG_C367A Fw
CHG_C367A Rv 
CAGCACCTTTGTGAAGTTTGGCGCTCTCAATGATC
GATCATTGAGAGCGCCAAACTTCACAAAGGTGCTG 
C427A CHG_C427A Fw
CHG_C427A Rv 
CGTGTCTATTACGCTACGGATGCTGCTAGCCAAGC
GCTTGGCTAGCAGCATCCGTAGCGTAATAGACAGC 
C601A CHG_C601A Fw
CHG_C601A Rv 
GGTCCTGGCAGAGCAGCTCGCTCGACACATCTATG
CATAGATGTGTCGAGCGAGCTGCTCTGCCAGGACC 
Mutation Primer names Primer sequences 
C152A CHG_C152A Fw
CHG_C152A Rv 
GCGACACCTGTTCATTGGACTGCTGATGGATCTCC
GGAGATCCATCAGCAGTCCAATGAACAGGTGTCGC 
C197A CHG_C197A Fw
CHG_C197A Rv 
GCTTGTTAAGACCTATGCTGACTTCATGCCCGTCC
GAACGGGCATGAAGTCAGCATAGGTCTTAACAAGC 
C282A CHG_C282A Fw
CHG_C282A Rv 
CTATTCGCCAATCAAGTCTTTGTCAGCGATCACTG
CAGTGATCGCTGACAAAGACTTGATTGGCGAATAG 
C291A CHG_C291A Fw
CHG_C291A Rv 
GTCAGCGATCACGCTGAAGAAGTTGTGCCGCG
CGCGGCACAACTTCTTCAGCGTGATCGCTGAC 
C367A CHG_C367A Fw
CHG_C367A Rv 
CAGCACCTTTGTGAAGTTTGGCGCTCTCAATGATC
GATCATTGAGAGCGCCAAACTTCACAAAGGTGCTG 
C427A CHG_C427A Fw
CHG_C427A Rv 
CGTGTCTATTACGCTACGGATGCTGCTAGCCAAGC
GCTTGGCTAGCAGCATCCGTAGCGTAATAGACAGC 
C601A CHG_C601A Fw
CHG_C601A Rv 
GGTCCTGGCAGAGCAGCTCGCTCGACACATCTATG
CATAGATGTGTCGAGCGAGCTGCTCTGCCAGGACC 

ATPase assay for Hsp90s, GroEL, DnaK2, and ClpB1

In the present study, we used two different methods to measure the ATPase activity. Results obtained with the two methods were similar or the same. Method 1 was used to generate all the results except some of those presented in Table 2.

Table 2
ATPase activities of various molecular chaperones and activities of pyruvate kinase and lactate dehydrogenase in the absence (−ZER) or presence (+ZER) of ZER

Except human Hsp90α, the concentration of ZER in a reaction mixture was 50 µg/ml. In the case of human Hsp90α, it was 100 µg/ml. The concentration of Hsp90SE, Hsp90SP, Hsp90α, or DnaK2 in a reaction mixture was 2 µM. The concentrations for GroEL, ClpB1, pyruvate kinase, and lactate dehydrogenase were 1 µM, 0.1 µM, 1 nM, and 0.5 nM, respectively. Data from three replicates are presented as mean ± SEM. The number of cysteine residues per protein monomer is indicated on the right-hand side. The ATPase activities of Hsp90s, DnaK2, and GroEL were assayed by the method 2 described in ‘Experimental’. The activities of Hsp90SE shown in this table were lower than those in other tables and figures since the ATP concentration in a reaction mixture used in the method 2 was 0.2 mM, which was lower than that (6 mM) used in the method 1.

 −ZER (mol/min/mol) +ZER (mol/min/mol) Activity change (%) # of cysteines 
Hsp90SE 0.41 ± 0.02 1.09 ± 0.02 265 
Hsp90SP 0.35 ± 0.03 0.88 ± 0.01 251 
Hsp90α 0.09 ± 0.06 0.42 ± 0.04 467 
DnaK2 0.44 ± 0.02 0.38 ± 0.01 86 
GroEL 3.75 ± 0.01 3.29 ± 0.11 88 
ClpB1 46.9 ± 1.05 40.9 ± 1.89 87 
Pyruvate kinase 2.88 × 104 ± 706 2.86 × 104 ± 765 99 
Lactate dehydrogenase 2.21 × 104 ± 848 2.05 × 104 ± 244 93 
 −ZER (mol/min/mol) +ZER (mol/min/mol) Activity change (%) # of cysteines 
Hsp90SE 0.41 ± 0.02 1.09 ± 0.02 265 
Hsp90SP 0.35 ± 0.03 0.88 ± 0.01 251 
Hsp90α 0.09 ± 0.06 0.42 ± 0.04 467 
DnaK2 0.44 ± 0.02 0.38 ± 0.01 86 
GroEL 3.75 ± 0.01 3.29 ± 0.11 88 
ClpB1 46.9 ± 1.05 40.9 ± 1.89 87 
Pyruvate kinase 2.88 × 104 ± 706 2.86 × 104 ± 765 99 
Lactate dehydrogenase 2.21 × 104 ± 848 2.05 × 104 ± 244 93 

Method 1

The total volume of the reaction mixture was 1 ml. Two nmol of the cyanobacterial Hsp90SE, E. coli Hsp90EC or 0.1 nmol of the cyanobacterial ClpB1 was incubated in the presence of zerumbone in 50 mM Hepes-KOH (pH 8.0) containing 20 mM KCl for 5 min at 37°C. Then, 2 mM phosphoenolpyruvate, 0.2 mM NADH, 50 μg of rabbit muscle lactate dehydrogenase (Oriental Yeast), and 30 μg of rabbit muscle pyruvate kinase (Oriental Yeast) were added to the medium. ATP hydrolysis was initiated by the addition of 6 mM ATP and 6 mM MgCl2 (MgCl2 was added to a reaction mixture in equimolar amounts to ATP). The absorbance at 340 nm was monitored continuously for 5 min at 37°C in a Shimadzu UV-1600PC spectrophotometer equipped with a thermostated cell holder. ATP hydrolysis by Hsp90 was terminated by the addition of 4 or 8 nmol of radicicol. The remaining activity in the presence of radicicol was subtracted as background, and thus all the Hsp90 activities reported in the present study were the ‘net’ activities that were inhibited by radicicol. Dimethyl sulfoxide (DMSO) was used to dissolve zerumbone. The effect of DMSO on the activity of Hsp90 or ClpB1 was also evaluated in a control reaction containing the same volume of DMSO as the corresponding zerumbone solution.

Method 2

Results given in Table 2 except for those for ClpB1, pyruvate kinase, and lactate dehydrogenase were obtained by the following method. The assay mixture (total volume, 1 ml) consisted of 50 mM Hepes-KOH (pH 8.0), 1 mM MgCl2, 20 mM KCl, 2 mM phosphoenolpyruvate, 30 µg of pyruvate kinase, 100 µM NADH, 50 µg of lactate dehydrogenase, and 0.2 mM ATP. ATP hydrolysis was initiated by the addition of 2 nmol of Hsp90SE, the yeast Hsp90SP, or Hsp90α to the reaction mixture pre-incubated at 37°C. After 5 min at 37°C, zerumbone (50 µg) was added to the reaction mixture, and the ATP hydrolysis was re-monitored for 5 min at 37°C. Finally, the reaction catalyzed by Hsp90 was terminated by the addition of 4 or 8 nmol of radicicol. The remaining activity in the presence of radicicol was subtracted as background. ATPase activities for E. coli GroEL and the cyanobacterial DnaK2 were measured as described above, using 1 nmol of GroEL or 2 nmol of DnaK2.

Mass spectrometry

A 250 µl aliquot of 50 mM Hepes-KOH (pH 8.0) containing 10 µM Hsp90SE was incubated in the presence or absence of zerumbone (10 or 50 µg/ml) at 37°C for 10 min. Then the solution was dialyzed with the Thermo Scientific Slide-A-Lyzer® MINI Dialysis Unit in 250 ml of 50 mM Hepes-KOH (pH 8.0) at 4°C for 2 h. After repeating this dialysis procedure twice, the dialysate was centrifuged and the supernatant solution was used for nano-flow liquid chromatography (LC) - tandem mass spectrometry (MS/MS) analysis [41].

Hsp90SE treated in the presence or absence of zerumbone was subjected to TPCK-treated trypsin digest in the digestion buffer (10 mM Tris–HCl, pH 8.0) for 12 h at 37°C. The digest mixture was separated using a nano-flow LC (Easy nLC, Thermo Fisher Scientific) on an NTCC analytical column (C-18 reversed-phase column, ID 0.075 × 100 mm, 3 µm bead size, Nikkyo Technos). Buffer A and B were composed of aqueous 0.1% formic acid and acetonitrile, respectively. Peptides were eluted from the C-18 column using a linear gradient of 1−35% buffer B in buffer A over 10 min at a flow rate of 300 nL/min and then analyzed using a Q-Exactive mass spectrometer (Thermo Fisher Scientific) with a nanospray ion source using the data-dependent TOP10 MS/MS method in the mass range of m/z = 300–600, 600–900, and m/z = 900–2000. Peptide identifications were made using MS/MS Ions Search in the MASCOT program v2.3 (Matrix Science Inc.). Peptides modified or unmodified by zerumbone were subjected to the LC-MS/MS using the target MS/MS mode. MS/MS spectra and chromatograms were processed using Qual Browser (Thermo Xcalibur 2.2).

Measurement of fluorescence spectra of tryptophan

Fluorescence spectra of tryptophan from Hsp90SE were measured at 37°C on a Shimadzu RF-5300 spectrofluorophotometer equipped with a thermostated cell holder. The emission fluorescence was scanned from 300 to 400 nm with the excitation wavelength set at 290 nm. The reaction mixture contained 2 µM Hsp90SE in 50 mM Hepes-KOH (pH 8.0), 20 mM KCl, 6 mM MgCl2 and zerumbone (10 or 50 µg/ml). The total volume of the mixture was 800 µl. The reaction mixture was pre-incubated at 37°C for 5 min before all the measurements. The effect of DMSO was also examined in a control medium containing the same volume of DMSO as the corresponding zerumbone sample.

Cyanobacterial and E. coli cell culture, cell extraction, and detection of the Hsp90SE clients by immunoblot analysis

S. elongatus PCC 7942 was grown in BG-11 liquid medium at 30 or 42°C with or without 20 µg/ml zerumbone as described previously [11,42]. The apparent absorbance at 730 nm was measured as an indication of the cyanobacterial cell number. E. coli K-12 W3110 was grown in Luria-Bertani liquid medium with or without 20 µg/ml zerumbone at 37 or 42°C. The apparent absorbance at 540 nm was measured to construct a growth curve.

To detect the 30 kDa rod linker polypeptide in cell extracts of S. elongatus PCC 7942, immunoblot analysis was performed with rabbit polyclonal antibodies raised against the polypeptide from S. elongatus PCC 7942 as described previously [16].

Mammalian cell culture, transfection, cell extraction, and immunoprecipitation

COS7 and KB cells were cultured in DMEM supplemented with 10% FCS in humidified air containing 5% CO2. COS7 cells were transfected with plasmids by electroporation, and cell extracts were prepared as described previously [4345]. Extracts with equal amounts of proteins were incubated with anti-FLAG affinity resin for 12 h at 4°C and the co-immunoprecipitation of Hsp90 and Cdc37 with client kinases was examined as described previously [39,46].

Cell imaging and counting

Phase contrast images of cells were obtained using a microscope (Zeiss Axiovert 200M with Plan-NEOFLUAR 20× lens). Cells were detached from dishes by trypsinization, collected by centrifugation, and resuspended in culture medium for cell number counting.

Results

Zerumbone activates the ATPase activity of Hsp90s

We have been searching for small molecules that affect the activity of Hsp90 [26,42] because they are useful tools for basic research and therapeutic purposes. A small, natural compound zerumbone (Figure 1A) has anticancer activity. The fact that Hsp90 is essential for cancer cell survival led us to speculate that it may interact with Hsp90. Thus, we examined the effect of zerumbone on the ATPase activity of Hsp90s from prokaryotic and eukaryotic cells (Table 2). All the Hsp90 ATPase activities measured for the present study were the net activities that were completely inhibited by a specific inhibitor of Hsp90, radicicol. Unexpectedly, Hsp90s from the cyanobacterium S. elongatus (Hsp90SE), the fission yeast Schizosaccharomyces pombe (Hsp90SP), and Homo sapience (Hsp90α) were activated by the molecule. The ATPase enhancement of both Hsp90SE and Hsp90SP by 50 µg/ml zerumbone was two- to three-fold. That of the human Hsp90α by 100 µg/ml zerumbone was more than four-fold. Zerumbone had no significant effect on the ATPase activity of other major molecular chaperones such as DnaK2, GroEL, and ClpB1 (Table 2). Zerumbone did not have significant effects on the activities of pyruvate kinase and lactate dehydrogenase, which were used as coupling enzymes for the measurement of the ATPase activities in this study (Table 2).

As described in the Introduction, the chaperone function/mechanism of prokaryotic Hsp90s is still poorly understood. We focused on Hsp90SE and examined the effect of varying concentrations of zerumbone on its ATPase activity. Zerumbone enhanced the activity of Hsp90SE in a concentration-dependent manner. The ATPase activity at 100 µg/ml zerumbone was three times higher than that in the absence of the small molecule (Figure 1B). When the kinetic parameters of the Hsp90 were investigated, the Km for ATP was found to decrease in the presence of zerumbone. On the other hand, the kcat value increased greatly (Figure 1C,D). The kcat/Km value in the presence of zerumbone was more than three times higher than that in the absence of the compound. Thus, zerumbone enhances the catalytic efficiency of Hsp90.

Zerumbone causes conformational changes in Hsp90

It has been shown that conformational rearrangements of Hsp90 in the N-terminal and middle domains are induced by ATP binding to the N-terminal domain [1,2]. Tryptophan fluorescence is sensitive to the local environment of the fluorophore. Thus, it is a sensitive marker of the conformational changes of proteins [47]. The cyanobacterial Hsp90 contains one and 4 tryptophan residues in the N-terminal and middle domains, respectively. No tryptophan is present in the C-terminal domain. Previously, the tryptophan fluorescence of Hsp90 was shown to be reduced in the presence of ATPγS, which was explained as being due to the conformational changes caused by nucleotide binding [48]. We also observed a decrease in the intrinsic tryptophan fluorescence of Hsp90SE in the presence of the non-hydrolyzable ATP analog adenylyl-imidodiphosphate (AMP-PNP). Similarly, zerumbone induced reduction in the fluorescence of Hsp90SE (Supplementary Figure S1).

Zerumbone modifies cysteine residues on Hsp90SE

Zerumbone provides electrophilic Michael addition centers for irreversible covalent binding of target proteins via nucleophilic residues [31,36]. Proteins such as Keap1 and HuR that are denatured in the presence of SDS and 2-mercaptoethanol have been shown to bind to zerumbone and this binding is suppressed when they are pre-incubated with a thiol-modifier N-ethylmaleimide, suggesting that these denatured proteins bind to zerumbone via a thiol group [49].

To examine the hypothesis that the enhancement of the ATPase activity by zerumbone is due to the modification of cysteine residues in Hsp90SE, we first counted the number of cysteines present in the Hsp90s used in the present study. As presented in Table 2, all the three Hsp90s that were activated by zerumbone contained multiple cysteine residues. To confirm that cysteine residues of Hsp90SE are directly modified by zerumbone, we performed MALDI-mass spectrometry with cyanobacterial Hsp90SE that had been treated by zerumbone. We treated Hsp90SE by zerumbone under non-denaturing conditions in order to correlate the modification with the enhancement of the ATPase activity. The Hsp90SE treated with zerumbone was trypsin-digested, and the mixture was subjected to nLC-MS/MS. The cysteine-containing peptides modified with zerumbone were confirmed by MS/MS spectra (Table 3 and Figure 2). The results showed that all the 7 cysteines of Hsp90SE were modified by zerumbone to some extent under the present experimental conditions although no cysteine was found to be fully modified, up to the concentration of 50 µg/ml. We were unable to measure the exact extent of modification in each cysteine, and to discern whether or not all the 7 cysteines are modified at the same time by the present methods.

MS/MS spectra of ZER-modified peptides of the cyanobacterial Hsp90SE.

Figure 2.
MS/MS spectra of ZER-modified peptides of the cyanobacterial Hsp90SE.

C* of an amino acid sequence indicate a cysteine residue modified by ZER. The y- and b-series fragment ions were annotated in spectra and summarized in the peptide sequences. The upper, middle, and lower spectra show peptides containing C152, C367, and C601, respectively.

Figure 2.
MS/MS spectra of ZER-modified peptides of the cyanobacterial Hsp90SE.

C* of an amino acid sequence indicate a cysteine residue modified by ZER. The y- and b-series fragment ions were annotated in spectra and summarized in the peptide sequences. The upper, middle, and lower spectra show peptides containing C152, C367, and C601, respectively.

Table 3
Amino acid sequences and masses of peptides of the cyanobacterial Hsp90SE containing cysteine residues
Amino acid sequences of isolated peptide fragments containing cysteine residues or ZER-modified cysteine residues Detected mass of each peptide fragment Mascot ion score Expect value 
GATPVHWTCDGSPSFELSEGSR 1160.5201 (2+) 46 2.80 × 10−5 
GATPVHWTC*DGSPSFELSEGSR 1269.6072 (2+) 75 3.30 × 10−8 
TYCDFMPVPIALEGEVLNK 713.6883 (3+) 50 1.00 × 10−5 
TYC*DFMPVPIALEGEVLNK 786.4114 (3+) 67 2.20 × 10−7 
LFCNQVFVSDHCEEVVPR 707.6680 (3+) 54 4 × 10−6 
LFC*NQVFVSDHCEEVVPR 780.3923 (3+) 59 1.30 × 10−6 
LFC*NQVFVSDHC*EEVVPR 853.1130 (3+) 86 2.8 × 10−9 
FGCLNDQK 462.7158 (2+) 54 4.20 × 10−6 
FGC*LNDQK 571.7990 (2+) 16 0.022 
VYYCTDAASQATYIELFR 1057.4996 (2+) 111 7.70 × 10−12 
VYYC*TDAASQATYIELFR 1166.5873 (2+) 83 4.70 × 10−9 
DGHSPSQVLAEQLCRHIY 1026.9993 (2+) 32 0.00056 
DGHSPSQVLAEQLC*RHIY 1136.0832 (2+) 25 0.0035 
Amino acid sequences of isolated peptide fragments containing cysteine residues or ZER-modified cysteine residues Detected mass of each peptide fragment Mascot ion score Expect value 
GATPVHWTCDGSPSFELSEGSR 1160.5201 (2+) 46 2.80 × 10−5 
GATPVHWTC*DGSPSFELSEGSR 1269.6072 (2+) 75 3.30 × 10−8 
TYCDFMPVPIALEGEVLNK 713.6883 (3+) 50 1.00 × 10−5 
TYC*DFMPVPIALEGEVLNK 786.4114 (3+) 67 2.20 × 10−7 
LFCNQVFVSDHCEEVVPR 707.6680 (3+) 54 4 × 10−6 
LFC*NQVFVSDHCEEVVPR 780.3923 (3+) 59 1.30 × 10−6 
LFC*NQVFVSDHC*EEVVPR 853.1130 (3+) 86 2.8 × 10−9 
FGCLNDQK 462.7158 (2+) 54 4.20 × 10−6 
FGC*LNDQK 571.7990 (2+) 16 0.022 
VYYCTDAASQATYIELFR 1057.4996 (2+) 111 7.70 × 10−12 
VYYC*TDAASQATYIELFR 1166.5873 (2+) 83 4.70 × 10−9 
DGHSPSQVLAEQLCRHIY 1026.9993 (2+) 32 0.00056 
DGHSPSQVLAEQLC*RHIY 1136.0832 (2+) 25 0.0035 

C* indicates a ZER-modified cysteine residue.

Modifications of one of cysteine residues in each domain are potentially involved in the enhancement of the ATPase activity of the cyanobacterial Hsp90 by zerumbone

Zerumbone modified all the cysteine residues of Hsp90SE rather promiscuously. Furthermore, it was not possible to quantify the extent of modification in each cysteine residue of Hsp90SE accurately by mass analysis. Thus, we could not discern the significance of the modification in any specific cysteine residue for enhancement of the Hsp90 ATPase activity. We therefore employed mutational studies to identify which cysteine residues are responsible for the enhancement.

We mutated all 7 cysteines in Hsp90SE to alanine individually to find which mutation abolishes enhancement of the ATPase activity by zerumbone. As shown in Figure 3A, mutations in C197 of the N-terminal domain, and three cysteines (C282, C291, and C427) of the middle domain did not have a significant effect on the enhancement, while mutations in C152 of the N-terminal domain, C367 of the middle domain, and C601 of the C-terminal domain suppressed it (Figure 3B), suggesting that these 3 cysteines are potentially involved in the enhancement of the ATPase activity via the modification by zerumbone. The C152A mutation caused the highest suppressive effect on the enhancement. Only a small increase in the ATPase activity was observed in this mutant protein. The activity enhancement in the presence of zerumbone was further decreased by mutation of all the three cysteine residues. Surprisingly, the C152A mutation and the triple mutation resulted in great enhancement of the ATPase activity in the absence of zerumbone (Table 4). None of the three cysteines C152, C367, and C601 are conserved in yeast and human Hsp90α (Supplementary Figure S2).

Mutation of one of cysteine residues in each domain abolishes the enhancement of the ATPase activity of the cyanobacterial Hsp90SE by ZER.

Figure 3.
Mutation of one of cysteine residues in each domain abolishes the enhancement of the ATPase activity of the cyanobacterial Hsp90SE by ZER.

Effect of 10 or 50 µg/ml ZER on the ATPase activity of the wild-type (WT) or cysteine mutants of Hsp90SE whose concentration in a reaction mixture was 2 µM. In all the points, data from three replicates are presented as mean ± SEM. Some error bars are covered by plot symbols. The ATPase activities of WT and the mutants obtained in the absence of ZER are presented in Table 4. (A) Effect of ZER on the wild-type Hsp90SE (WT) and cysteine mutants of Hsp90SE (C197A, C291A, C282A, and C427A). (B) Effect of ZER on the wild-type Hsp90SE (WT), and cysteine mutants of Hsp90SE (C152A, C367A, C601A, and the triple mutant C152A/C367A/C601A). A Student's t-test was performed to assess statistical significance (P-value < 0.05) between the WT activity and each of the four mutants in B. (C) Location of cysteine residues on Hsp90SE. NTD, N-terminal domain; MD, middle domain; CTD, C-terminal domain.

Figure 3.
Mutation of one of cysteine residues in each domain abolishes the enhancement of the ATPase activity of the cyanobacterial Hsp90SE by ZER.

Effect of 10 or 50 µg/ml ZER on the ATPase activity of the wild-type (WT) or cysteine mutants of Hsp90SE whose concentration in a reaction mixture was 2 µM. In all the points, data from three replicates are presented as mean ± SEM. Some error bars are covered by plot symbols. The ATPase activities of WT and the mutants obtained in the absence of ZER are presented in Table 4. (A) Effect of ZER on the wild-type Hsp90SE (WT) and cysteine mutants of Hsp90SE (C197A, C291A, C282A, and C427A). (B) Effect of ZER on the wild-type Hsp90SE (WT), and cysteine mutants of Hsp90SE (C152A, C367A, C601A, and the triple mutant C152A/C367A/C601A). A Student's t-test was performed to assess statistical significance (P-value < 0.05) between the WT activity and each of the four mutants in B. (C) Location of cysteine residues on Hsp90SE. NTD, N-terminal domain; MD, middle domain; CTD, C-terminal domain.

Table 4
The ATPase activities of the wild-type and various cysteine mutants of the cyanobacterial Hsp90SE

The concentration of Hsp90SE was 2 µM. Data from three replicates are presented as mean ± SEM.

Hsp90SE and its mutants ATPase activity (mol/min/mol) 
WT 1.19 ± 0.02 
C152A 7.28 ± 0.06 
C197A 1.48 ± 0.01 
C282A 0.98 ± 0.01 
C291A 0.95 ± 0.02 
C367A 1.16 ± 0.04 
C427A 1.38 ± 0.03 
C601A 1.81 ± 0.09 
C152A/C367A/C601A 9.27 ± 0.41 
Hsp90SE and its mutants ATPase activity (mol/min/mol) 
WT 1.19 ± 0.02 
C152A 7.28 ± 0.06 
C197A 1.48 ± 0.01 
C282A 0.98 ± 0.01 
C291A 0.95 ± 0.02 
C367A 1.16 ± 0.04 
C427A 1.38 ± 0.03 
C601A 1.81 ± 0.09 
C152A/C367A/C601A 9.27 ± 0.41 

As described above, it was impossible to discern whether or not all the 7 cysteines are modified with zerumbone at the same time by the present mass analysis. Thus, we tried to determine how many of the seven cysteine side chains of Hsp90SE are modified by zerumbone. The decreased amount of sulfhydryls accessible to the Ellman reagent was ∼2 residues in the presence of 50 µg/ml zerumbone (data not shown), suggesting that two of the above cysteines are involved in the enhancement of the ATPase activity at the zerumbone concentration.

Studies with the E. coli Hsp90, and humulene, a structural analog of zerumbone

The results shown above provide evidence that chemical modification of cysteine residues is involved in enhancement of the Hsp90 ATPase activity. To study this further, we tested whether zerumbone exerts any effects on the ATPase activity of E. coli Hsp90EC, which has no cysteine. As illustrated in Table 5, the activity enhancement was much smaller than that observed with the cyanobacterial Hsp90SE. Humulene is a structural analog lacking only the reactive carbonyl group in zerumbone, and has been used as a reference compound for zerumbone. It did not enhance the ATPase activity of Hsp90SE, but rather inhibited it to a small extent. These results support the hypothesis that activation of the Hsp90SE ATPase activity is largely due to the covalent modification of cysteine residues of Hsp90SE.

Table 5
Effect of ZER or humulene at the concentration of 50 µg/ml on the ATPase activity of the cyanobacterial Hsp90 (Hsp90SE) and the E. coli Hsp90 (Hsp90EC)

The number of cysteines of each Hsp90 is also indicated in the Table. The concentration of the Hsp90s was 2 µM. Data from three replicates are presented as mean ± SEM.

Hsp90 # of cysteines Small molecule Changes in activity (%) when compared with the activity in the absence of ZER or humulene 
Hsp90SE ZER 209 
Hsp90EC ZER 121 
Hsp90SE Humulene 86 
Hsp90 # of cysteines Small molecule Changes in activity (%) when compared with the activity in the absence of ZER or humulene 
Hsp90SE ZER 209 
Hsp90EC ZER 121 
Hsp90SE Humulene 86 

Zerumbone causes high temperature-sensitive phenotype and reduction in the cellular level of Hsp90 clients in cyanobacteria

In contrast with eukaryotes, E. coli and B. subtilis can grow and survive in the absence of Hsp90 under normal and heat stress conditions [10,50]. However, Hsp90SE mutants exhibit severe defects at high temperatures, although they behave like the wild-type under normal conditions [11]. They do not grow at a moderately high temperature, and greatly reduce high-temperature survival. Among different mutations of heat shock genes in the cyanobacterium Synechocystis sp. PCC 6803, only mutation of the hsp90 (htpG) gene resulted in a clearly distinguishable growth defect, that is, no growth at a moderately high temperature [51]. Thus, no growth at a moderately high temperature may be used as one of the signature phenotypes which indicate functional defects in the cyanobacterial Hsp90. If zerumbone affects the in vivo function of Hsp90, we can predict that it will increase the high temperature sensitivity of cyanobacteria. On the other hand, it would be predicted not to have any significant effects on the growth of E. coli regardless of growth temperature since an Hsp90 mutant of E. coli has only a minor phenotype, showing a slight growth disadvantage at high temperatures [10]. Thus, we examined the effect of zerumbone on growth of S. elongatus and E. coli. As far as we know, the effect of zerumbone on prokaryotic cells has not previously been elucidated. As shown in Figure 4A,B, the cyanobacterium could not grow at 42°C in the presence of zerumbone, while only a small growth inhibition was observed at 30°C, the normal growth temperature used in the present study. In contrast, zerumbone did not show any effect on the growth of E. coli at either 37 or 42°C (Figure 4C,D). These results also suggest that zerumbone does not have any damaging effects on cellular proteins/components of E. coli.

ZER causes a high temperature-sensitive phenotype and reduction in the cellular level of an Hsp90 client in cyanobacterium S. elongatus, whereas it does not affect the growth of E. coli.

Figure 4.
ZER causes a high temperature-sensitive phenotype and reduction in the cellular level of an Hsp90 client in cyanobacterium S. elongatus, whereas it does not affect the growth of E. coli.

Growth of S. elongatus (A and B) and E. coli (C and D) at a normal growth temperature and a mild high temperature in the presence (square symbols) or absence (triangle symbols) of 20 µg/ml ZER. When ZER was not added to the culture medium, the same volume of DMSO, the solvent used to dissolve ZER, was added. The apparent absorbance at 730 and 540 nm was used to measure the relative number of cells for S. elongatus and E. coli, respectively. In all the points, data from three replicates are presented as mean ± SEM. Some error bars are covered by plot symbols. The figure E shows the cellular level of the 30 kDa rod linker polypeptide of the phycobilisome in S. elongatus cells grown at 30 and 42°C in the presence (ZER) and absence (DMSO) of 20 µg/ml ZER. Total soluble proteins (20 µg) was loaded to each lane, separated by SDS–PAGE, and the 30 kDa rod linker polypeptide of the phycobilisome was specifically detected by immunoblot analysis using polyclonal antibodies raised against the polypeptide. Relative band intensity (%) that is indicated below each protein band was obtained by ImageJ.

Figure 4.
ZER causes a high temperature-sensitive phenotype and reduction in the cellular level of an Hsp90 client in cyanobacterium S. elongatus, whereas it does not affect the growth of E. coli.

Growth of S. elongatus (A and B) and E. coli (C and D) at a normal growth temperature and a mild high temperature in the presence (square symbols) or absence (triangle symbols) of 20 µg/ml ZER. When ZER was not added to the culture medium, the same volume of DMSO, the solvent used to dissolve ZER, was added. The apparent absorbance at 730 and 540 nm was used to measure the relative number of cells for S. elongatus and E. coli, respectively. In all the points, data from three replicates are presented as mean ± SEM. Some error bars are covered by plot symbols. The figure E shows the cellular level of the 30 kDa rod linker polypeptide of the phycobilisome in S. elongatus cells grown at 30 and 42°C in the presence (ZER) and absence (DMSO) of 20 µg/ml ZER. Total soluble proteins (20 µg) was loaded to each lane, separated by SDS–PAGE, and the 30 kDa rod linker polypeptide of the phycobilisome was specifically detected by immunoblot analysis using polyclonal antibodies raised against the polypeptide. Relative band intensity (%) that is indicated below each protein band was obtained by ImageJ.

Previously, we identified linker polypeptides of the phycobilisome, a light-harvesting complex, as clients of Hsp90SE in cyanobacterial cells [16]. Hsp90SE can stabilize linker polypeptides to suppress its aggregation. Linker polypeptides including the 30 kDa rod linker polypeptide are thought to play an important role in the assembly of the phycobilisome protein complex which accounts for ∼50% of the total cyanobacterial cellular proteins in mass [52]. Thus, misassembly or incomplete assembly of the phycobilisome may lead to the disruption of protein homeostasis and cellular dysfunction. Thus, we decided to analyze the cellular level of the 30 kDa rod linker polypeptide. As shown in Figure 4E, the level of the linker polypeptide was reduced significantly in the presence of zerumbone at both 30 and 42°C. The reduction was more evident at 42°C than at 30°C. Levels of other linker polypeptides as well as other unidentified proteins were also reduced in zerumbone-treated cells (Supplementary Figure S3).

Zerumbone induced apoptosis and suppressed proliferation in cultured mammalian cells

Zerumbone activated the ATPase activity of the yeast and human Hsp90s as well as the cyanobacterial Hsp90 as illustrated in Table 2. Thus, we examined the effect of zerumbone treatment on cultured mammalian cells. The human cervix carcinoma epidermoid adherent KB cell line (now known to be a subline of HeLa) has been used extensively in studies of metabolism, signaling, and apoptosis of cells as well as cancer chemotherapy screening, thus we chose KB cells as a model for the analysis of drug-induced cell death in the present study. First, nearly confluent cells were treated with increasing concentrations of zerumbone and a fixed concentration of solvent DMSO (0.55%) for 25 h. Zerumbone showed dose-dependent cell-killing activity as shown by phase-contrast cell images (Figure 5A) and cell number counting (Figure 5B). Cell-killing activity was prominent at 4.4 µg/ml of zerumbone, and cells were completely dead and floating by the treatment with 22 µg/ml zerumbone. Cleavage of PARP by caspases has been widely accepted as a specific marker for the cellular apoptotic process. We found that 16 h treatment with zerumbone induced accumulation of caspase-cleaved PARP at concentrations higher than 11 µg/ml, as shown in Figure 5C, indicating that cells underwent the apoptotic process and the cell-killing activity of zerumbone was not due to the general non-specific toxicity. No changes of the levels of Hsp90 or a loading control protein tubulin were observed. In addition, cells could not proliferate in the presence of 4.4 µg/ml of zerumbone, and cells could not survive in the presence of 11 µg/ml of zerumbone (data not shown). These results indicate that zerumbone shows cellular toxicity on cancer-derived mammalian cells by inducing apoptosis, and possesses anti-proliferative activity.

ZER induced apoptosis in cultured mammalian cells.

Figure 5.
ZER induced apoptosis in cultured mammalian cells.

Nearly confluent KB cells were treated with increasing concentrations of ZER in a fixed concentration (0.55%) of DMSO as the solvent for 25 h, and the phase contrast cell images were taken (A). The numbers of living attached cells were counted and indicated as a percentage of the initial cell number (B). Averages of duplicate wells were shown with standard deviation bars. The effect of increasing concentrations of ZER (16 h) on the cleavage of an endogenous caspase substrate PARP was examined by western blotting analysis (C). Equal amounts of total proteins were loaded. Left, total PARP (upper panel) and caspase-cleaved PARP (lower panel). Right, the amounts of Hsp90 (upper panel) and tubulin (lower panel) are shown.

Figure 5.
ZER induced apoptosis in cultured mammalian cells.

Nearly confluent KB cells were treated with increasing concentrations of ZER in a fixed concentration (0.55%) of DMSO as the solvent for 25 h, and the phase contrast cell images were taken (A). The numbers of living attached cells were counted and indicated as a percentage of the initial cell number (B). Averages of duplicate wells were shown with standard deviation bars. The effect of increasing concentrations of ZER (16 h) on the cleavage of an endogenous caspase substrate PARP was examined by western blotting analysis (C). Equal amounts of total proteins were loaded. Left, total PARP (upper panel) and caspase-cleaved PARP (lower panel). Right, the amounts of Hsp90 (upper panel) and tubulin (lower panel) are shown.

We then examined the effect of zerumbone on the association of Hsp90 and its kinase-targeting co-chaperone Cdc37 with client protein kinases. Cdk4 and Raf1CD (the catalytic domain of Raf1) were expressed as FLAG-tagged proteins in COS7 cells and the binding of endogenous Hsp90 and Cdc37 with Cdk4 and Raf1CD was examined by co-immunoprecipitation experiments. Cells were treated with DMSO (vehicle control), geldanamycin (positive control), or zerumbone before extraction. Figure 6A clearly indicates that treatment of cells with zerumbone (4.8 µg/ml 20 h) suppressed the binding of both Hsp90 and Cdc37 to Cdk4 and Raf1CD. The effect of zerumbone (lanes 3 and 6) was weaker than that of geldanamycin (lanes 2 and 5). The disruption of the association of Hsp90 and Cdc37 with client kinases often results in destabilization of the client kinases. In fact, we observed modest but reproducible reduction of Cdk4 and Raf1CD by zerumbone treatment (Figure 6B, top and second panels, lane 3) as in the case of geldanamycin treatment (Figure 6B, lane 2). The cellular levels of Hsp90 and Cdc37 remained unchanged by the zerumbone treatment (Figure 6B, third and fourth panels). Geldanamycin treatment has been shown to robustly induce Hsp70 up-regulation in cells, and we also observed a higher level of Hsp70 in geldanamycin-treated cells (Figure 6B, bottom, lane 2). Interestingly, zerumbone treatment did not cause significant induction of Hsp70 (Figure 6B, lane 3). We have also examined the binding of Hsp90 and Cdc37 to other client kinases v-Src and MOK, and we observed significant reduction in the association of both Hsp90 and Cdc37 with these client kinases in the presence of zerumbone (Figure 6C, lanes 3 and 6), whereas geldanamycin treatment almost completely abolished the association (Figure 6C, lanes 2 and 5). Taken altogether, these results indicated that zerumbone inhibits the physiological function of Hsp90/Cdc37 chaperones to bind to and support the stability of client kinases.

ZER inhibited the association of Hsp90 and Cdc37 with client protein kinases without significant induction of Hsp70.

Figure 6.
ZER inhibited the association of Hsp90 and Cdc37 with client protein kinases without significant induction of Hsp70.

(A) Binding of endogenous Hsp90 and Cdc37 to exogenously expressed FLAG-tagged Cdk4 (lanes 1–3) and Raf1CD (lanes 4–6) was examined by co-immunoprecipitation experiments. Cells were treated with vehicle (DMSO), geldanamycin (5 µM 4 h), or ZER (4.8 µg/ml 20 h). The amounts of Hsp90 (top), Cdc37 (middle), and Cdk4 or Raf1CD (bottom) were shown by western blotting with anti-Hsp90, anti-Cdc37, or anti-FLAG antibody. (B) The cellular levels of proteins (Cdk4, Raf1CD, Hsp90, Cdc37, and Hsp70) were shown by western blotting with corresponding antibodies. Equal amounts of proteins of total cell lysates were loaded in each lane. (C) The binding of Hsp90 and Cdc37 to FLAG-tagged v-Src or MOK was examined by co-immunoprecipitation experiments. Cells were treated as in (A), and the amounts of Hsp90 (top), Cdc37 (middle), and v-Src (bottom, lanes 1–3) or MOK (bottom, lanes 4–6) in the immunoprecipitates were shown by western blotting. All of the experiments in (AC) were independently performed more than two times, resulting in the identical conclusion and the representative western blotting results of the same gel were shown with splicing lines indicating the removal of non-relevant lanes. The positions of molecular mass markers (MagicMark XP, Invitrogen) are shown on the left. The signal intensities of the bands obtained by Image Reader LAS-4000 were quantitated with a software Multi Gauge V3.11 (Fuji Film) and shown underneath the gel images as percentages to the DMSO control lanes.

Figure 6.
ZER inhibited the association of Hsp90 and Cdc37 with client protein kinases without significant induction of Hsp70.

(A) Binding of endogenous Hsp90 and Cdc37 to exogenously expressed FLAG-tagged Cdk4 (lanes 1–3) and Raf1CD (lanes 4–6) was examined by co-immunoprecipitation experiments. Cells were treated with vehicle (DMSO), geldanamycin (5 µM 4 h), or ZER (4.8 µg/ml 20 h). The amounts of Hsp90 (top), Cdc37 (middle), and Cdk4 or Raf1CD (bottom) were shown by western blotting with anti-Hsp90, anti-Cdc37, or anti-FLAG antibody. (B) The cellular levels of proteins (Cdk4, Raf1CD, Hsp90, Cdc37, and Hsp70) were shown by western blotting with corresponding antibodies. Equal amounts of proteins of total cell lysates were loaded in each lane. (C) The binding of Hsp90 and Cdc37 to FLAG-tagged v-Src or MOK was examined by co-immunoprecipitation experiments. Cells were treated as in (A), and the amounts of Hsp90 (top), Cdc37 (middle), and v-Src (bottom, lanes 1–3) or MOK (bottom, lanes 4–6) in the immunoprecipitates were shown by western blotting. All of the experiments in (AC) were independently performed more than two times, resulting in the identical conclusion and the representative western blotting results of the same gel were shown with splicing lines indicating the removal of non-relevant lanes. The positions of molecular mass markers (MagicMark XP, Invitrogen) are shown on the left. The signal intensities of the bands obtained by Image Reader LAS-4000 were quantitated with a software Multi Gauge V3.11 (Fuji Film) and shown underneath the gel images as percentages to the DMSO control lanes.

In total, our study indicates that zerumbone inhibits the cellular function of both prokaryotic and eukaryotic Hsp90s although it activates the ATPase activity of the chaperone.

Discussion

In the present study, we investigated the effects of zerumbone on the Hsp90 ATPase activity. Zerumbone is a natural dietary plant metabolite that shows anticancer activity in several tumor models. However, the molecular mechanisms underlying the antitumor activity remain to be elucidated. We have shown here that zerumbone stimulates the Hsp90 ATPase activity of both prokaryotic and eukaryotic Hsp90s. Zerumbone increases the catalytic efficiency (the kcat/Km value) of the cyanobacterial Hsp90SE more than three times by decreasing Km and increasing kcat (Figure 1). This enhancement is specific and is not seen for other chaperones with intrinsic ATPase activity (Table 2). Activities of other chaperones and enzymes that contain cysteines were not significantly affected and, if anything, were slightly inhibited in the presence of zerumbone.

All the Hsp90 activators reported so far interact with Hsp90 non-covalently (see Introduction), whereas zerumbone modifies cysteine residues covalently [31,36]. Our present study supports the hypothesis that this covalent modification leads to enhancement of the ATPase activity of various Hsp90s. Zerumbone increased the ATPase activity of all examined Hsp90s except the E. coli Hsp90, which does not contain any cysteine residues, by several-fold compared with controls (Tables 2 and 5). α-humulene, a structural analog of zerumbone lacking the reactive carbonyl group did not enhance the ATPase activity of the cyanobacterial Hsp90 (Table 5), supporting the hypothesis that covalent modification by zerumbone is involved in the enhancement. Modification of cysteine residues of Hsp90SE by zerumbone was confirmed by mass analysis (Table 3 and Figure 2). Mutational studies identified the cysteine residues that are responsible for the enhancement. Mutations in 3 cysteine residues (Cys-152, Cys-367, Cys-601) suppressed the enhancement, indicating that their modification by zerumbone causes the activation of the Hsp90SE ATPase activity (Figure 3). Interestingly, these residues are spread over all three domains, indicating that it is possible to regulate the ATPase activity by modifying a cysteine residue in any one of the domains. Cys-152 is located at a distance from the ATP-binding site of the N-terminal domain. Cys-367 is located at a site of the middle domain that corresponds to Val-429 of the yeast Hsp90. Glu-431 as well as Val-429 are located in one of the α-helices that are sandwiched between two αβα subdomains. The mutation E431K is shown to decrease ATPase activity [53], indicating that modifications of amino acid residues in the α-helix can influence ATPase activity.

It is evident that the activation due to the Cys-601 modification is taking place allosterically, as the C-terminal domain is not directly involved in either the ATP binding or the ATP hydrolysis. Thus, the cysteine residue in the C-terminal domain is a potential target for controlling the Km and kcat of Hsp90 allosterically. Cys-601 is located in helix 4 of the C-terminal domain. Helices 4 and 5 (shown in the dotted square in Figure 7) create the majority of the dimer interface through a four-helix bundle [4]. Thus, modification of this cysteine may influence the stability of the Hsp90 dimer. A yeast Hsp90 mutant whose alanine at 577 in the C-terminal domain (that corresponds to Ser-522 in Hsp90SE) was substituted with cysteine or isoleucine strengthened the association of the two C-terminal domains of the Hsp90 dimer to increase the ATPase activity four-fold [54]. Modification of Cys-601 of Hsp90SE by zerumbone may stabilize the dimerization to enhance ATPase activity.

Sites of cysteine (Cys) modification on cyanobacterial Hsp90SE, human Hsp90α, and human Hsp90β with various small compounds shown in the table.

Figure 7.
Sites of cysteine (Cys) modification on cyanobacterial Hsp90SE, human Hsp90α, and human Hsp90β with various small compounds shown in the table.

Each number (1 to 7) with an arrow points to the modified site whose information is given in the table below. NTD, N-terminal domain; MD, middle domain; CTD, C-terminal domain. A dotted square present within CTD indicates the helices 4 and 5 that create the majority of the dimer interface. Some Cys modifications have been shown to enhance or inhibit the Hsp90 ATPase activity as indicated in the table.

Figure 7.
Sites of cysteine (Cys) modification on cyanobacterial Hsp90SE, human Hsp90α, and human Hsp90β with various small compounds shown in the table.

Each number (1 to 7) with an arrow points to the modified site whose information is given in the table below. NTD, N-terminal domain; MD, middle domain; CTD, C-terminal domain. A dotted square present within CTD indicates the helices 4 and 5 that create the majority of the dimer interface. Some Cys modifications have been shown to enhance or inhibit the Hsp90 ATPase activity as indicated in the table.

Modifications of cysteine residues of Hsp90 by small molecules have been shown to take place both in vitro and in vivo (Figure 7). Among them, S-nitrosylation that is mediated by nitric oxide modifies Cys-597 of the C-terminal domain of Hsp90α [55], resulting in the inhibition of its ATPase activity. 4-hydroxy-2-nonenal, an α, β-unsaturated aldehyde generated through lipid peroxidation during oxidative stress targets Cys-572, another cysteine residue of the C-terminal domain of Hsp90α to inhibit its chaperone activity [56]. Isothiocyanates are naturally occurring small molecules that have medical benefits. 6-methylsulfinylhexyl isothiocyanate modifies Cys-521 of Hsp90β [57]. Interestingly, all the above cysteines are located at the middle/C-terminal domain interface where an allosteric site is present [27]. Sulphoxythiocarbamate targets various cysteine residues of Hsp90β (Figure 7), forming stable thiocarbamate adducts. It causes destabilization of client oncoproteins of Hsp90 and inhibits cell proliferation [58]. None of the above cysteines are conserved in Hsp90SE (Figure 7). Thus, our present study revealed novel sites to control Hsp90 function through covalent modification of cysteine residues.

There are many reports showing that zerumbone has anti-proliferative activity against cancers as described in the Introduction. Thus, the observation that zerumbone activates the Hsp90 ATPase activity (Table 2) is unexpected, because most of antitumor drugs targeting Hsp90 inhibit its ATPase activity [1922]. It is possible that the activation of the Hsp90 ATPase activity may not be related to the anti-proliferative activity. Thus, we examined whether zerumbone affects the cellular function of Hsp90. First, we studied the effects of the compound on the thermo-tolerance of the cyanobacterium S. elongatus and the cellular level of an Hsp90SE client. Zerumbone inhibited cyanobacterial growth completely at a moderately high temperature (Figure 4). The characteristics displayed in the presence of zerumbone are reminiscent of those of an Hsp90 mutant of this cyanobacterium [11]. Further evidence supporting the hypothesis that zerumbone suppresses Hsp90SE function in vivo is shown by the reduction in the cellular level of its client, the 30 kDa rod linker polypeptide of the phycobilisome. Similar reduction is also observed in the Hsp90 mutant [16]. We should point out that zerumbone did not have negative effects on the growth of E. coli (Figure 4). Zerumbone appears to modify cysteine residues of various proteins non-specifically [36,59]. Thus, it is reasonable to expect that zerumbone modified proteins non-specifically in the E. coli cell without having a negative impact on its growth. This suggests that many cellular proteins modified by zerumbone may not lose their function (see also Table 2), which may be one of the reasons why zerumbone is safe as a dietary compound. It is unlikely that differential effects on the growth of the two prokaryotes exerted by zerumbone are due to cell permeability, as cell membranes have been shown to be permeable to zerumbone [36,59].

We further studied the effects of zerumbone on the chaperone function of Hsp90 in mammalian cells. Zerumbone inhibited the interaction of Hsp90 with its clients such as Cdk4, Raf1, v-Src, and MOK in cells (Figure 6). We also detected a decrease in the cellular level of Cdk4 and Raf1 in the presence of zerumbone as with geldanamycin, a well-described destabilizer of the Hsp90 clients. The decrease may be caused by the inhibition of the interaction of Hsp90 with the clients. There are reports that Hsp90 clients such as MMP9, Survivin, Bcl-XL, Bcl-2, and IκBα (https://www.picard.ch/downloads/Hsp90interactors.pdf) decrease their cellular levels in the presence of zerumbone [35]. The decrease may be also due to the inhibition of their interaction with Hsp90 in the presence of zerumbone. The reduction of Hsp90 clients in a cell may be one of the reasons for the cytotoxicity caused by zerumbone (Figure 5). In addition to zerumbone, Hsp90Mod-4 and derivatives of 2-phenylbenzofurans inhibited the folding/activation of glucocorticoid receptor in a yeast cell and decreased the levels of various clients in cancer cells [25,27], respectively, indicating that these activators decease the in vivo chaperone function of Hsp90.

There is evidence that shows the enhancement of the ATPase activity causes the inhibition of the chaperone function of Hsp90 in vivo and in vitro. Nathan and Lindquist [60] identified mutations in Hsp90 that cause high temperature sensitivity. Later, it was found that one of the mutations, T22I, resulted in the enhancement of the ATPase activity in vitro [53]. The mutant strain shows a growth defect even under normal conditions. In the mutant yeast cells which express the T22I mutant of Hsp90, the Hsp90 clients, glucocorticoid receptor and v-Src, do not fold efficiently and their cellular levels are lowered [60]. Furthermore, we showed recently that an Hsp90 ATPase activator, goniothalamin inhibits the in vitro chaperone function of Hsp90SE to assist refolding of a denatured protein [26].

Our results as well as those reported previously suggest that excessive Hsp90 ATPase activity exerts a negative effect on Hsp90 chaperone function. However, regulation of Hsp90 chaperone function may be far from straightforward. A recent study showed no correlation between the ATPase activity and in vivo chaperone function of Hsp90 [8]. They found that the time spent by Hsp90 in a specific conformation state is important rather than the overall time for the ATP hydrolysis. For example, cells do not survive when their mutated Hsp90s do not spend sufficient time in the open conformation. This suggests that it may be the case that the zerumbone modification restricts how long Hsp90 spends in an open conformation. This hypothesis will be tested in our future work.

Abbreviations

     
  • AMP-PNP

    adenylyl-imidodiphosphate

  •  
  • CBB

    Coomassie brilliant blue

  •  
  • DMSO

    Dimethyl sulfoxide

  •  
  • Hsp90

    the 90-kDa heat shock protein

  •  
  • Hsp90EC

    HtpG of Escherichia coli

  •  
  • Hsp90SE

    HtpG of Synechococcus elongatus

  •  
  • Hsp90SP

    Hsp90 of Schizosaccharomyces pombe

  •  
  • LC

    liquid chromatography

  •  
  • MDH

    malate dehydrogenase

  •  
  • MS/MS

    tandem mass spectrometry

  •  
  • WT

    wild-type

  •  
  • ZER

    zerumbone

Author Contribution

H.N. conceived, designed, and co-ordinated the study, and wrote the whole paper. I.J. provided the initial sample of zerumbone, and wrote part of the Introduction concerning zerumbone. Y.A. and T.K. designed, performed, and analyzed the experiments shown in Figures 1 and 3, and Tables 2, 4, and 5. Y.N. and T.K. designed, performed, and analyzed the experiments shown in Figure 4. T.S. and N.D. designed, performed and analyzed the experiments shown in Figure 2 and Table 3. Y.M. designed, performed, and analyzed the experiments shown in Figures 5 and 6. Y.M. also wrote part of the Results and Discussion concerning results shown in Figures 5 and 6. All the authors reviewed the results and approved the final version of the manuscript.

Funding

This work was supported in part by Grant-in-aids for Scientific Research (C) [No. 24580102 and 15K07349] to H.N. from the Ministry of Education, Science, Sports and Culture of Japan. It was also supported by the Bilateral Joint Research Projects (2014–2016) from the Ministry of Education, Science, Sports and Culture of Japan.

Acknowledgments

We are grateful to Dr. Peter A. Lund of University of Birmingham for carefully proofreading the manuscript.

Competing Interests

The Authors declare that there are no competing interests associated with the manuscript.

References

References
1
Li
,
J.
,
Soroka
,
J.
and
Buchner
,
J.
(
2012
)
The Hsp90 chaperone machinery: conformational dynamics and regulation by co-chaperones
.
Biochim. Biophys. Acta, Mol. Cell. Res.
1823
,
624
635
2
Pearl
,
L.H.
(
2016
)
Review: the HSP90 molecular chaperone-an enigmatic ATPase
.
Biopolymers
105
,
594
607
3
Minami
,
Y.
,
Kimura
,
Y.
,
Kawasaki
,
H.
,
Suzuki
,
K.
and
Yahara
,
I.
(
1994
)
The carboxy-terminal region of mammalian HSP90 is required for its dimerization and function in vivo
.
Mol. Cell. Biol.
14
,
1459
1464
4
Harris
,
S.F.
,
Shiau
,
A.K.
and
Agard
,
D.A.
(
2004
)
The crystal structure of the carboxy-terminal dimerization domain of htpG, the Escherichia coli Hsp90, reveals a potential substrate binding site
.
Structure
12
,
1087
1097
5
Prodromou
,
C.
,
Roe
,
S.M.
,
O'Brien
,
R.
,
Ladbury
,
J.E.
,
Piper
,
P.W.
and
Pearl
,
L.H.
(
1997
)
Identification and structural characterization of the ATP/ADP-binding site in the Hsp90 molecular chaperone
.
Cell
90
,
65
75
6
Panaretou
,
B.
,
Prodromou
,
C.
,
Roe
,
S.M.
,
O'Brien
,
R.
,
Ladbury
,
J.E.
,
Piper
,
P.W.
et al. 
(
1998
)
ATP binding and hydrolysis are essential to the function of the Hsp90 molecular chaperone in vivo
.
EMBO J.
17
,
4829
4836
7
Obermann
,
W.M.J.
,
Sondermann
,
H.
,
Russo
,
A.A.
,
Pavletich
,
N.P.
and
Hartl
,
F.U.
(
1998
)
In vivo function of Hsp90 is dependent on ATP binding and ATP hydrolysis
.
J. Cell Biol.
143
,
901
910
8
Zierer
,
B.K.
,
Rübbelke
,
M.
,
Tippel
,
F.
,
Madl
,
T.
,
Schopf
,
F.H.
,
Rutz
,
D.A.
et al. 
(
2016
)
Importance of cycle timing for the function of the molecular chaperone Hsp90
.
Nat. Struct. Mol. Biol.
23
,
1020
1028
9
Johnson
,
J.L.
(
2012
)
Evolution and function of diverse Hsp90 homologs and cochaperone proteins
.
Biochim. Biophys. Acta, Mol. Cell Res.
1823
,
607
613
10
Bardwell
,
J.C.
and
Craig
,
E.A.
(
1988
)
Ancient heat shock gene is dispensable
.
J. Bacteriol.
170
,
2977
2983
11
Tanaka
,
N.
and
Nakamoto
,
H.
(
1999
)
Htpg is essential for the thermal stress management in cyanobacteria
.
FEBS Lett.
458
,
117
123
12
Hossain
,
M.M.
and
Nakamoto
,
H.
(
2002
)
Htpg plays a role in cold acclimation in cyanobacteria
.
Curr. Microbiol.
44
,
291
296
13
Hossain
,
M.M.
and
Nakamoto
,
H.
(
2003
)
Role for the cyanobacterial HtpG in protection from oxidative stress
.
Curr. Microbiol.
46
,
70
76
14
Watanabe
,
S.
,
Kobayashi
,
T.
,
Saito
,
M.
,
Sato
,
M.
,
Nimura-Matsune
,
K.
,
Chibazakura
,
T.
et al. 
(
2007
)
Studies on the role of HtpG in the tetrapyrrole biosynthesis pathway of the cyanobacterium Synechococcus elongatus PCC 7942
.
Biochem. Biophys. Res. Commun.
352
,
36
41
15
Saito
,
M.
,
Watanabe
,
S.
,
Yoshikawa
,
H.
and
Nakamoto
,
H.
(
2008
)
Interaction of the molecular chaperone HtpG with uroporphyrinogen decarboxylase in the cyanobacterium Synechococcus elongatus PCC 7942
.
Biosci. Biotechnol. Biochem.
72
,
1394
1397
16
Sato
,
T.
,
Minagawa
,
S.
,
Kojima
,
E.
,
Okamoto
,
N.
and
Nakamoto
,
H.
(
2010
)
Htpg, the prokaryotic homologue of Hsp90, stabilizes a phycobilisome protein in the cyanobacterium Synechococcus elongatus PCC 7942
.
Mol. Microbiol.
76
,
576
589
17
Nakamoto
,
H.
,
Fujita
,
K.
,
Ohtaki
,
A.
,
Watanabe
,
S.
,
Narumi
,
S.
,
Maruyama
,
T.
et al. 
(
2014
)
Physical interaction between bacterial heat shock protein (Hsp) 90 and Hsp70 chaperones mediates their cooperative action to refold denatured proteins
.
J. Biol. Chem.
289
,
6110
6119
18
Genest
,
O.
,
Hoskins
,
J.R.
,
Camberg
,
J.L.
,
Doyle
,
S.M.
and
Wickner
,
S.
(
2011
)
Heat shock protein 90 from Escherichia coli collaborates with the DnaK chaperone system in client protein remodeling
.
Proc. Natl Acad. Sci. U.S.A.
108
,
8206
8211
19
Whitesell
,
L.
and
Lindquist
,
S.L.
(
2005
)
HSP90 and the chaperoning of cancer
.
Nat. Rev. Cancer
5
,
761
772
20
Neckers
,
L.
, and
Workman
,
P.
(
2012
)
Hsp90 molecular chaperone inhibitors: are we there yet?
Clin. Cancer Res.
18
,
64
76
21
Miyata
,
Y.
,
Nakamoto
,
H.
and
Neckers
,
L.
(
2013
)
The therapeutic target Hsp90 and cancer hallmarks
.
Curr. Pharm. Des.
19
,
347
365
22
Jhaveri
,
K.
,
Taldone
,
T.
,
Modi
,
S.
and
Chiosis
,
G.
(
2012
)
Advances in the clinical development of heat shock protein 90 (Hsp90) inhibitors in cancers
.
Biochim. Biophys. Acta, Mol. Cell Res.
1823
,
742
755
23
Kamal
,
A.
,
Thao
,
L.
,
Sensintaffar
,
J.
,
Zhang
,
L.
,
Boehm
,
M.F.
,
Fritz
,
L.C.
et al. 
(
2003
)
A high-affinity conformation of Hsp90 confers tumour selectivity on Hsp90 inhibitors
.
Nature
425
,
407
410
24
Zhao
,
R.
,
Leung
,
E.
,
Grüner
,
S.
,
Schapira
,
M.
and
Houry
,
W.A.
(
2010
)
Tamoxifen enhances the Hsp90 molecular chaperone ATPase activity
.
PLoS ONE
5
,
e9934
25
Zierer
,
B.K.
,
Weiwad
,
M.
,
Rübbelke
,
M.
,
Freiburger
,
L.
,
Fischer
,
G.
,
Lorenz
,
O.R.
et al. 
(
2014
)
Artificial accelerators of the molecular chaperone Hsp90 facilitate rate-limiting conformational transitions
.
Angew. Chem. Int. Ed. Engl.
53
,
12257
12262
26
Yokoyama
,
Y.
,
Ohtaki
,
A.
,
Jantan
,
I.
,
Yohda
,
M.
and
Nakamoto
,
H.
(
2015
)
Goniothalamin enhances the ATPase activity of the molecular chaperone Hsp90 but inhibits its chaperone activity
.
J. Biochem.
157
,
161
168
27
Sattin
,
S.
,
Tao
,
J.
,
Vettoretti
,
G.
,
Moroni
,
E.
,
Pennati
,
M.
,
Lopergolo
,
A.
et al. 
(
2015
)
Activation of Hsp90 enzymatic activity and conformational dynamics through rationally designed allosteric ligands
.
Chemistry
21
,
13598
13608
28
Murakami
,
A.
,
Takahashi
,
M.
,
Jiwajinda
,
S.
,
Koshimizu
,
K.
and
Ohigashi
,
H.
(
1999
)
Identification of zerumbone in Zingiber zerumbet Smith as a potent inhibitor of 12-O-tetradecanoylphorbol-13-acetate-induced Epstein-Barr virus activation
.
Biosci. Biotechnol. Biochem.
63
,
1811
1812
29
Kirana
,
C.
,
McIntosh
,
G.H.
,
Record
,
I.R.
and
Jones
,
G.P.
(
2003
)
Antitumor activity of extract of Zingiber aromaticum and its bioactive sesquiterpenoid zerumbone
.
Nutr. Cancer
45
,
218
225
30
Yodkeeree
,
S.
,
Sung
,
B.
,
Limtrakul
,
P.
and
Aggarwal
,
B.B.
(
2009
)
Zerumbone enhances TRAIL-induced apoptosis through the induction of death receptors in human colon cancer cells: Evidence for an essential role of reactive oxygen species
.
Cancer Res.
69
,
6581
6589
31
Murakami
,
A.
,
Takahashi
,
D.
,
Kinoshita
,
T.
,
Koshimizu
,
K.
,
Kim
,
H.W.
,
Yoshihiro
,
A.
et al. 
(
2002
)
Zerumbone, a Southeast Asian ginger sesquiterpene, markedly suppresses free radical generation, proinflammatory protein production, and cancer cell proliferation accompanied by apoptosis: the alpha,beta-unsaturated carbonyl group is a prerequisite
.
Carcinogenesis
23
,
795
802
32
Murakami
,
A.
,
Tanaka
,
T.
,
Lee
,
J.-Y.
,
Surh
,
Y.-J.
,
Kim
,
H.W.
,
Kawabata
,
K.
et al. 
(
2004
)
Zerumbone, a sesquiterpene in subtropical ginger, suppresses skin tumor initiation and promotion stages in ICR mice
.
Int. J. Cancer
110
,
481
490
33
Kim
,
M.
,
Miyamoto
,
S.
,
Yasui
,
Y.
,
Oyama
,
T.
,
Murakami
,
A.
and
Tanaka
,
T.
(
2009
)
Zerumbone, a tropical ginger sesquiterpene, inhibits colon and lung carcinogenesis in mice
.
Int. J. Cancer
124
,
264
271
34
Sung
,
B.
,
Murakami
,
A.
,
Oyajobi
,
B.O.
and
Aggarwal
,
B.B.
(
2009
)
Zerumbone abolishes RANKL-induced NF-kB activation, inhibits osteoclastogenesis, and suppresses human breast cancer-induced bone loss in athymic nude mice
.
Cancer Res.
69
,
1477
1484
35
Prasannan
,
R.
,
Kalesh
,
K.A.
,
Shanmugam
,
M.K.
,
Nachiyappan
,
A.
,
Ramachandran
,
L.
,
Nguyen
,
A.H.
et al. 
(
2012
)
Key cell signaling pathways modulated by zerumbone: role in the prevention and treatment of cancer
.
Biochem. Pharmacol.
84
,
1268
1276
36
Kalesh
,
K.A.
,
Clulow
,
J.A.
and
Tate
,
E.W.
(
2015
)
Target profiling of zerumbone using a novel cell-permeable clickable probe and quantitative chemical proteomics
.
Chem. Commun.
51
,
5497
5500
37
Koyasu
,
S.
,
Nishida
,
E.
,
Kadowaki
,
T.
,
Matsuzaki
,
F.
,
Iida
,
K.
,
Harada
,
F.
et al. 
(
1986
)
Two mammalian heat shock proteins, HSP90 and HSP100, are actin-binding proteins
.
Proc. Natl Acad. Sci. U.S.A.
83
,
8054
8058
38
Miyata
,
Y.
and
Yahara
,
I.
(
2000
)
p53-independent association between SV40 large T antigen and the major cytosolic heat shock protein, HSP90
.
Oncogene
19
,
1477
1484
39
Miyata
,
Y.
and
Nishida
,
E.
(
2004
)
CK2 controls multiple protein kinases by phosphorylating a kinase-targeting molecular chaperone Cdc37
.
Mol. Cell. Biol.
24
,
4065
4074
40
Ishida
,
M.
,
Tomomari
,
T.
,
Kanzaki
,
T.
,
Abe
,
T.
,
Oka
,
T.
and
Yohda
,
M.
(
2013
)
Biochemical characterization and cooperation with co-chaperones of heat shock protein 90 from Schizosaccharomyces pombe
.
J. Biosci. Bioeng.
116
,
444
448
41
Michalski
,
A.
,
Damoc
,
E.
,
Hauschild
,
J.-P.
,
Lange
,
O.
,
Wieghaus
,
A.
,
Makarov
,
A.
et al. 
(
2011
)
Mass spectrometry-based proteomics using Q Exactive, a high-performance benchtop quadrupole Orbitrap mass spectrometer
.
Mol. Cell. Proteomics
10
,
M111.011015
42
Minagawa
,
S.
,
Kondoh
,
Y.
,
Sueoka
,
K.
,
Osada
,
H.
and
Nakamoto
,
H.
(
2011
)
Cyclic lipopeptide antibiotics bind to the N-terminal domain of the prokaryotic Hsp90 to inhibit the chaperone activity
.
Biochem. J.
435
,
237
246
43
Miyata
,
Y.
,
Akashi
,
M.
and
Nishida
,
E.
(
1999
)
Molecular cloning and characterization of a novel member of the MAP kinase superfamily
.
Genes Cells
4
,
299
309
44
Miyata
,
Y.
,
Chambraud
,
B.
,
Radanyi
,
C.
,
Leclerc
,
J.
,
Lebeau
,
M.-C.
,
Renoir
,
J.-M.
et al. 
(
1997
)
Phosphorylation of the immunosuppressant FK506-binding protein FKBP52 by casein kinase II (CK2): regulation of HSP90-binding activity of FKBP52
.
Proc. Natl Acad. Sci. U.S.A.
94
,
14500
14505
45
Miyata
,
Y.
and
Nishida
,
E.
(
2011
)
DYRK1A binds to an evolutionarily conserved WD40-repeat protein WDR68 and induces its nuclear translocation
.
Biochim. Biophys. Acta, Mol. Cell Res.
1813
,
1728
1739
46
Miyata
,
Y.
,
Ikawa
,
Y.
,
Shibuya
,
M.
and
Nishida
,
E.
(
2001
)
Specific association of a set of molecular chaperones including HSP90 and Cdc37 with MOK, a member of the mitogen-activated protein kinase superfamily
.
J. Biol. Chem.
276
,
21841
21848
47
Lakowicz
,
J.R.
(
2006
)
Principles of Fluorescence Spectroscopy
,
3rd edn
.
Springer
,
New York, NY
48
Csermely
,
P.
,
Kajtár
,
J.
,
Hollósi
,
M.
,
Jalsovszky
,
G.
,
Holly
,
S.
,
Kahn
,
C.R.
et al. 
(
1993
)
ATP induces a conformational change of the 90-kDa heat shock protein (hsp90)
.
J. Biol. Chem.
268
,
1901
1907
PMID:
[PubMed]
49
Ohnishi
,
K.
,
Irie
,
K.
and
Murakami
,
A.
(
2009
)
In vitro covalent binding proteins of zerumbone, a chemopreventive food factor
.
Biosci. Biotechnol. Biochem.
73
,
1905
1907
50
Versteeg
,
S.
,
Mogk
,
A.
and
Schumann
,
W.
(
1999
)
The Bacillus subtilis htpG gene is not involved in thermal stress management
.
Mol. Gen. Genet.
261
,
582
588
51
Rowland
,
J.G.
,
Pang
,
X.
,
Suzuki
,
I.
,
Murata
,
N.
,
Simon
,
W.J.
and
Slabas
,
A.R.
(
2010
)
Identification of components associated with thermal acclimation of photosystem II in Synechocystis sp. PCC6803
.
PLoS ONE
5
,
e10511
52
Grossman
,
A.R.
,
Schaefer
,
M.R.
,
Chiang
,
G.G.
and
Collier
,
J.L.
(
1993
)
The phycobilisome, a light-harvesting complex responsive to environmental conditions
.
Microbiol. Rev.
57
,
725
749
PMID:
[PubMed]
53
Hawle
,
P.
,
Siepmann
,
M.
,
Harst
,
A.
,
Siderius
,
M.
,
Reusch
,
H.P.
and
Obermann
,
W.M.
(
2006
)
The middle domain of Hsp90 acts as a discriminator between different types of client proteins
.
Mol. Cell Biol.
26
,
8385
8395
54
Retzlaff
,
M.
,
Stahl
,
M.
,
Eberl
,
H.C.
,
Lagleder
,
S.
,
Beck
,
J.
,
Kessler
,
H.
et al. 
(
2009
)
Hsp90 is regulated by a switch point in the C-terminal domain
.
EMBO Rep.
10
,
1147
1153
55
Martinez-Ruiz
,
A.
,
Villanueva
,
L.
,
Gonzalez de Orduna
,
C.
,
Lopez-Ferrer
,
D.
,
Higueras
,
M.A.
,
Tarin
,
C.
et al. 
(
2005
)
S-nitrosylation of Hsp90 promotes the inhibition of its ATPase and endothelial nitric oxide synthase regulatory activities
.
Proc. Natl Acad. Sci. U.S.A.
102
,
8525
8530
56
Carbone
,
D.L.
,
Doorn
,
J.A.
,
Kiebler
,
Z.
,
Ickes
,
B.R.
and
Petersen
,
D.R.
(
2005
)
Modification of heat shock protein 90 by 4-hydroxynonenal in a rat model of chronic alcoholic liver disease
.
J. Pharmacol. Exp. Ther.
315
,
8
15
57
Shibata
,
T.
,
Kimura
,
Y.
,
Mukai
,
A.
,
Mori
,
H.
,
Ito
,
S.
,
Asaka
,
Y.
et al. 
(
2011
)
Transthiocarbamoylation of proteins by thiolated isothiocyanates
.
J. Biol. Chem.
286
,
42150
42161
58
Zhang
,
Y.
,
Dayalan
,
N.S.
,
Samarasinghe
,
K.
,
Van Hecke
,
G.C.
,
Pheely
,
A.
,
Boronina
,
T.N.
et al. 
(
2014
)
Sulphoxythiocarbamates modify cysteine residues in HSP90 causing degradation of client proteins and inhibition of cancer cell proliferation
.
Br. J. Cancer
110
,
71
82
59
Ohnishi
,
K.
,
Ohkura
,
S.
,
Nakahata
,
E.
,
Ishisaka
,
A.
,
Kawai
,
Y.
,
Terao
,
J.
et al. 
(
2013
)
Non-specific protein modifications by a phytochemical induce heat shock response for self-defense
.
PLoS ONE
8
,
e58641
60
Nathan
,
D.F.
and
Lindquist
,
S.
(
1995
)
Mutational analysis of Hsp90 function: interactions with a steroid receptor and a protein kinase
.
Mol. Cell. Biol.
15
,
3917
3925

Supplementary data