ITZ-1, a Client-Selective Hsp90 Inhibitor, Efficiently Induces Heat Shock Factor 1 Activation

Article abstract of DOI:10.1016/j.chembiol.2009.12.012

ITZ-1 is a chondroprotective agent that inhibits interleukin-1β-induced matrix metalloproteinase-13 (MMP-13) production and suppresses nitric oxide-induced chondrocyte death. Here we describe its mechanisms of action. Heat shock protein 90 (Hsp90) was identified as a specific ITZ-1-binding protein. Almost all known Hsp90 inhibitors have been reported to bind to the Hsp90 N-terminal ATP-binding site and to simultaneously induce degradation and activation of its multiple client proteins. However, within the Hsp90 client proteins, ITZ-1 strongly induces heat shock factor-1 (HSF1) activation and causes mild Raf-1 degradation, but scarcely induces degradation of a broad range of Hsp90 client proteins by binding to the Hsp90 C terminus. These results may explain ITZ-1's inhibition of MMP-13 production, its cytoprotective effect, and its lower cytotoxicity. These results suggest that ITZ-1 is a client-selective Hsp90 inhibitor.

Full text of DOI:10.1016/j.chembiol.2009.12.012

Chemistry & Biology
Article
ITZ-1, a Client-Selective Hsp90 Inhibitor,
Efficiently Induces Heat Shock Factor 1 Activation
1
1
1
1
1
Haruhide Kimura,1,2, Hiroshi Yukitake, Yasukazu Tajima, Hirobumi Suzuki, Tomoko Chikatsu, Shinji Morimoto,
Yasunori Funabashi,1 Hiroaki Omae,1 Takashi Ito,1 Yukio Yoneda,2 and Masayuki Takizawa1
1Pharmaceutical Research Division, Takeda Pharmaceutical, Osaka 532-8686, Japan
*
2Laboratory of Molecular Pharmacology, Division of Pharmaceutical Sciences, Kanazawa University Graduate School of Natural Science
and Technology, Kanazawa 920-1192, Japan
*Correspondence: Kimura_Haruhide@takeda.co.jp
DOI 10.1016/j.chembiol.2009.12.012
SUMMARY
normal articular cartilage (Presle et al., 1999; Ku¨ hn et al., 2004).
Because there are no known therapeutic agents proven to
ITZ-1 is a chondroprotective agent that inhibits inter- prevent erosion of articular cartilage or slow disease progres-
sion, agents that interfere with cartilage destruction would be
of considerable therapeutic value (Ku¨ hn et al., 2004; Aigner
et al., 2006; Ge et al., 2006).
leukin-1b–induced matrix metalloproteinase–13
(MMP-13) production and suppresses nitric oxide–
induced chondrocyte death. Here we describe its
mechanisms of action. Heat shock protein 90
(Hsp90) was identified as a specific ITZ-1–binding
protein. Almost all known Hsp90 inhibitors have
been reported to bind to the Hsp90 N-terminal ATP-
binding site and to simultaneously induce degrada-
To develop therapeutic drugs for OA, we have searched for
chondroprotective agents and found N-{(1Z)-5,6-Dimethyl-
3-oxo-8-[(4,4,5,5,5,-pentafluoropentyl)thio]-2,3-dihydro-1H-imi-
dazo[5,1-c][1,4]thiazin-1-ylidene}-4methylbenzenesulfonamide,
ITZ-1 (Kimura et al., 2009), which suppresses IL-1b–induced
cartilage degradation in vitro and in vivo (Pettipher et al., 1986;
tion and activation of its multiple client proteins.
However, within the Hsp90 client proteins, ITZ-1
Cawston et al., 1995; Bottomley et al., 1997). ITZ-1 selectively
inhibits IL-1b–induced MMP-13 production with a half-maximal
strongly induces heat shock factor–1 (HSF1) activa- inhibitory concentration (IC50) of 0.51 mM via inhibition of extra-
cellular signal–regulated protein kinase (ERK) activation (Vincenti
and Brinckerhoff, 2002; Ahmed et al., 2004, 2005), and
tion and causes mild Raf-1 degradation, but scarcely
induces degradation of a broad range of Hsp90 client
proteins by binding to the Hsp90 C terminus. These
results may explain ITZ-1’s inhibition of MMP-13
suppresses NO-induced death in human articular chondrocytes
(HACs) (Terauchi et al., 2003). Thus, ITZ-1 could be a promising
lead compound for a disease-modifying anti-OA drug program.
production, its cytoprotective effect, and its lower
However, the precise mechanisms of action of ITZ-1 have been
cytotoxicity. These results suggest that ITZ-1 is a
poorly understood.
client-selective Hsp90 inhibitor.
Here, we describe the molecular mechanisms of ITZ-1. We
identified heat shock protein 90 (Hsp90) (Whitesell and Lindquist,
2005; Sharp and Workman, 2006; Neckers, 2007) as a target
protein of ITZ-1 by use of drug-affinity chromatography (Harding
et al., 1989; Taunton et al., 1996; Shimizu et al., 2000; Sato et al.,
2007) and have further elucidated its downstream molecular
mechanisms. Hsp90 is involved in the folding, activation, and
assembly of several proteins, known as Hsp90 client proteins,
and the inhibition of Hsp90 function has been considered to
simultaneously induce degradation and activation of these
multiple client proteins (Whitesell and Lindquist, 2005; Sharp
and Workman, 2006; Neckers, 2007). However, several recent
reports have suggested the existence of small molecule inhibi-
tors of Hsp90 with client selective properties (Yu et al., 2005;
Salehi et al., 2006; Gallo, 2006; Ansar et al., 2007; Lu et al.,
2009). Interestingly, ITZ-1 strongly induced heat shock factor–1
(HSF1) activation and caused mild Raf-1 degradation without
affecting other Hsp90 client proteins by binding to the Hsp90 C
terminus (Whitesell and Lindquist, 2005). Considering the
definitive evidence that Hsp90 plays pivotal roles in multiorgan
physiology and pathology, the discovery we present here may
provide an approach to the development of therapeutic drugs
targeting Hsp90.
INTRODUCTION
Osteoarthritis (OA) is a degenerative joint disease characterized
by slow but irreversible progressive destruction of articular
cartilage. The sustained production of interleukin (IL)–1 in the
affected joint may play a key role in OA pathogenesis (Aigner
et al., 2006; Ge et al., 2006). IL-1 induces up-regulation of matrix
metalloproteinases (MMPs) that degrade principal matrix macro-
molecules, such as proteoglycans and collagens, in cartilage.
Proteoglycans have a high turnover rate, whereas type II collagen
has a low turnover rate; thus, type II collagen degradation is a
committed step in OA progression. Within the approximately
20-member MMP-family, MMP-1, MMP-8, and MMP-13 are
interstitial collagenases that degrade type II collagen; in partic-
ular, MMP-13 may play a pivotal role in cartilage degradation
(Ge et al., 2006; Takaishi et al., 2008).
IL-1 has also been reported to induce death in chondrocytes,
which regulate cartilage metabolism, probably via nitric oxide
(NO) production (Yasuhara et al., 2005). Because the chondro-
cytes constitute the only cell type of the articular cartilage,
survival of the chondrocytes is essential for the maintenance of
18 Chemistry & Biology 17, 18–27, January 29, 2010 ª2010 Elsevier Ltd All rights reserved

Chemistry & Biology
ITZ-1 a Client-Selective Hsp90 Inhibitor
F
F
F
HN
A
D
B
C
O
F
F
F
F
O
O
F
S
S
O
N
O
HO
S
HO
S
H
2N
S
H
2N
S
S
S
S
S
O
O
O
O
N
N
N
N
S
NH
O
O
O
O
HCl
HCl
N
S
S
S
S
NH
NH
NH
NH
N
N
N
N
ITZ-1
O
O
O
O
O
ITZ-1
ITZ-6
ITZ-7
ITZ-8
ITZ-9
ITZ-2
ITZ-3
ITZ-4
ITZ-5
120
100
80
60
40
20
0
F
F
F
HN
O
O
O
O
O
N
S
N
S
N
H
S
S
N
S
S
S
S
O
O
O
O
N
N
N
N
H
H
H
O
O
O
O
S
S
S
S
NH
NH
NH
NH
N
N
N
N
O
O
O
O
ITZ-6
ITZ-7
ITZ-8
ITZ-9
-20
0.3
1
3
10
mM
E
F
ITZ-4-column
89 - 13
ITZ-2-column
1 2 3 4 5 6 7 8 91011
ITZ-3-column
1 2 3 4 5 6 7 8 91011
ITZ-4-column
1 2 3 4 5 6 7 8 91011
control-column
1 2 3 4 5 6 7 8 91011
ITZ-5-column
1 23 4 89
MW [kDa] 1 234
-
-
-
13
MW [kDa]
175.0
175.0
83.0
62.0
47.5
83.0
62.0
32.5
25.0
47.5
32.5
25.0
16.5
6.5
Figure 1. Identification of Hsp90 as an ITZ-1-Binding Protein
(A) Chemical structure of ITZ-1.
(B) Chemical structures of ITZ-1 derivatives used for drug-affinity chromatography.
(C) Chemical structures of ITZ-1 derivatives used to estimate the potencies of derivatives shown in (B) on the matrix.
(D) Inhibitory activities of the ITZ-1 derivatives against IL-1b–induced MMP-13 production in HACs. Results shown are means SD (n = 3).
(E) Identification of Hsp90 from protein extracts as an ITZ-1–binding protein by drug-affinity chromatography. ITZ-2-, ITZ-3-, and ITZ-4-columns were used as
positive columns. As a control, a matrix on which no compound was immobilized was used (control-column). Lane 1, molecular weight (MW) marker; lane 2,
protein extract; lane 3, flow-through fraction; lanes 4 and 5, binding buffer wash fractions; lanes 6–10, 1 M NaCl-eluted fractions; and lane 11, glycine-HCl
buffer-eluted fraction. Arrows indicate the Hsp90 protein.
(F) Direct binding of His-hHsp90a to the ITZ-4-column. The ITZ-5-column was used as a negative control. Lane 1, MW marker; lane 2, His-hHsp90a; lane 3,
flow-through fraction; lanes 4–8, binding buffer wash fractions; and lanes 9–13, 6 M urea buffer-eluted fractions. Arrows indicate the His-hHsp90a protein.
See also Table S1.
RESULTS
with ITZ-6, ITZ-7, ITZ-8, and ITZ-9, respectively (Figure 1C;
Table S1). ITZ-6, ITZ-7, and ITZ-8 inhibited IL-1b–induced
MMP-13 production and exhibited an IC50 of 3.2 mM, 2.0 mM,
and 4.0 mM, respectively, whereas ITZ-9, even at 10 mM, did not
Identification of Hsp90 as a Possible ITZ-1–Binding
Protein
The initial goal of the study was to identify the possible target show any detectable activity (Figure 1D). We hypothesized that
proteins of ITZ-1 (Figure 1A; see Table S1 available online) by use common binding proteins of matrix coupled with ITZ-2, ITZ-3,
of drug-affinity chromatography. A terminal carboxyl or amino and ITZ-4, but not ITZ-5, would also bind to ITZ-1. Thus, columns
group was needed to couple small molecules to AF-amino or of ITZ-2–, ITZ-3–, and ITZ-4–coupled matrixes (ITZ-2-column,
AF-carboxy TOYOPEARL 650 matrix; however, ITZ-1 does not ITZ-3-column, and ITZ-4-column) were used as positive
possess these groups. Thus, two carboxyl derivatives, ITZ-2 columns, and an ITZ-5–coupled matrix column (ITZ-5-column)
and ITZ-3, and two amino derivatives, ITZ-4 and ITZ-5, of ITZ-1 was used as a negative column. We also utilized a column of
were prepared for affinity chromatography (Figure 1B; Table S1). AF-amino TOYOPEARL 650 matrix on which no compound
ITZ-2 and ITZ-4 are structurally similar to ITZ-1. ITZ-3 and ITZ-5 was immobilized as another negative control column (control-
were selected because previous studies have shown that intro- column).
duction of a trifluoromethyl group to the 4-position of benzene
There was a possibility that the binding affinity of ITZ-5 to the
ring in the (phenylsulfonyl)imino group retained activity, whereas target proteins was only 10 times less than that of ITZ-2, ITZ-3,
introduction of an acetylamino group to the same position re- and ITZ-4 (Figure 1D), which could result in the binding of target
sulted in a decrease in activity. Because each compound was proteins to both positive and negative columns in some condi-
immobilized to the matrix by coupling with a spacer arm at the tion. Therefore, we used control-column instead of ITZ-5-column
terminal carboxyl or amino group, we estimated the potency of for initial search. Chondrocyte protein extracts were applied to
each compound on the matrix by using compounds coupled the ITZ-2-, ITZ-3-, ITZ-4-, and control-column. After washing
with a long alkyl chain instead of a spacer arm. Potencies of the columns with binding buffer, bound proteins were eluted
ITZ-2, ITZ-3, ITZ-4, and ITZ-5 on the matrix were estimated with 1 M NaCl-containing buffer followed by glycine-HCl buffer.
Chemistry & Biology 17, 18–27, January 29, 2010 ª2010 Elsevier Ltd All rights reserved 19

Chemistry & Biology
ITZ-1 a Client-Selective Hsp90 Inhibitor
Br
By analyzing each fraction, a common binding protein with a
molecular mass of ꢀ90 kDa (indicated by arrows) was identified
in the glycine-HCl buffer-eluted fractions of the three positive
columns (Figure 1E). This 90-kDa protein was not obtained
from the control-column, indicating that it was retained in these
positive columns by binding to the compound, but not to the
solid support. Amino acid sequence analysis revealed that the
90-kDa protein was Hsp90. The two closely related cytosolic iso-
forms of Hsp90 (Hsp90a and Hsp90b) were identified (Whitesell
and Lindquist, 2005).
A
B
S
S
O
N
S
S
F
S
S
O
O
N
N
O
S
O
O
NH
S
S
N
O
NH
NH
HO
N
N
O
O
O
HO
O
ITZ-10
ITZ-11
ITZ-12
C
120
100
80
60
40
20
0
ITZ-1
ITZ-10
ITZ-11
ITZ-12
His-hHsp90a
ITZ-10
To exclude the possibility that the interactions between the
ITZ-1 derivatives and Hsp90 are indirect or nonspecific, we
analyzed the binding of ITZ-1 derivatives to purified recombinant
Hsp90a (His-hHsp90a) using the drug-coupled columns. The
ITZ-4-column was used as a positive column, and the ITZ-5-
column was used as a negative column. His-hHsp90a was
diluted with binding buffer and then applied to the drug-coupled
columns. After washing the columns with binding buffer, bound
protein was eluted with the elution buffer containing 6 M urea.
Analysis of each fraction by SDS-PAGE followed by silver stain-
ing revealed that His-hHsp90a was retained in the positive ITZ-4-
column, but not in the negative ITZ-5-column, and was eluted
with elution buffer (Figure 1F). These results suggest that ITZ-1
derivatives interact directly with Hsp90a, and the rank order of
the derivatives in the inhibition of MMP-13 production might
agree well with that of the binding affinity for Hsp90.
ITZ-1
D
0 25 50 100 (mM) 0 25 50 100 (mM)
ITZ-11
ITZ-12
0 25 50 100 (mM) 0 25 50 100 (mM)
0.1 0.3
1
3
10
mM
Figure 2. Direct Interaction between ITZ-1 and Hsp90
(A) Chemical structures of ITZ-1 derivatives.
(B) MMP-13 production inhibitory activities of ITZ-1 derivatives. Results shown
are means SD (n = 3).
(C) Specific binding of His-hHsp90a to ITZ-3–immobilized matrixes. ITZ-3–
immobilized matrix or control matrix was incubated with His-hHsp90a with
agitation, and then bound protein was analyzed by SDS-PAGE followed by
CBB staining.
(D) Displacement studies with ITZ-1 derivatives by the binding assay using
ITZ-3-immobilized matrixes. The binding reactions were performed as in (C)
in the presence of free compounds. See also Table S1and Figure S1.
Direct Interaction Between ITZ-1 and Hsp90a
Next, we investigated the direct interaction between ITZ-1 deriv-
atives and Hsp90a by two distinct binding assays. To obtain
further insights into the structure-activity relationship for ITZ-1
derivatives, we selected ITZ-1 and three other derivatives without
the terminal carboxyl/amino group or the long alkyl chain (ITZ-
10, ITZ-11, and ITZ-12; Figure 2A; Table S1). ITZ-1 was most
potent, followed by ITZ-10, ITZ-11, and ITZ-12, in the inhibition
of IL-1b–induced MMP-13 production in HACs (Figure 2B).
First, we analyzed the ITZ-1-Hsp90a interaction with the ITZ-
3–immobilized matrix. To test the functionality of the ITZ-3–im-
mobilized matrix, ITZ-3–immobilized matrix or control matrix
was incubated with His-hHsp90a with agitation, and then bound
protein was analyzed by SDS-PAGE followed by Coomassie bril-
liant blue R250 (CBB) staining. His-hHsp90a was recovered from
the ITZ-3–immobilized matrix, but not from the control matrix
(Figure 2C). This result shows that ITZ-3–immobilized matrix can
be used to measure the affinity of ligands that bind to the ITZ-3–
binding site of Hsp90a. Under these conditions, the binding
reactions were performed in the presence of free compounds.
Analysis of the bound protein to the ITZ-3–immobilized matrix re-
and ITZ-1 were detected; however, when other 6His-tagged
proteins were used instead of His2-hHsp90a, [3H]-ITZ-1 binding
was undetectable (Figure S1A). These data indicate that ITZ-1
itself interacts specifically with Hsp90a.
In the SPA assay measuring the competitive inhibition of
[3H]-ITZ-1 binding to His2-hHsp90a, ITZ-1 was the most potent,
followed by ITZ-10, ITZ-11, and ITZ-12, and exhibited IC50 of
1.9 mM, 3.9 mM, 86 mM, and 100 mM, respectively (Figure S1B).
Note that the rank order of the derivatives in the inhibition of
MMP-13 production agrees well with the binding affinities for
Hsp90a (Figures 2B and D and Figure S1B). ITZ-1 might inhibit
MMP-13 production via binding to Hsp90. Interestingly, gelda-
namycin (GA), which binds to the Hsp90 N-terminal ATP-binding
site (Whitesell and Lindquist, 2005; Sharp and Workman, 2006;
Neckers, 2007), hardly inhibited the ITZ-1-Hsp90a interaction
at concentrations less than 100 mM (Figure S1B). ITZ-1 might
not bind to the N-terminal ATP-binding site of Hsp90.
vealed that ITZ-1 was most potent, followed by ITZ-10, ITZ-11, ITZ-1 Inhibits IL-1b–Induced ERK Activation
and ITZ-12, in the competitive inhibition of Hsp90a-ITZ-3 binding via Raf-1 Degradation
(Figure 2D).
We have reported that ITZ-1 selectively inhibited IL-1b–induced
Next, we developed a scintillation proximity assay (SPA assay) extracellular signal–regulated protein kinase (ERK) activation
based on the detection of the radioligand [3H]-ITZ-1 binding without affecting p38 kinase or c-Jun N-terminal kinase (JNK)
to a 6His-tagged Hsp90a protein (His2-hHsp90a), which was activation (Kimura et al., 2009). As an Hsp90 client protein,
captured by Ysi (2–5 mm) copper His-tag SPA beads (Graziani Raf-1, which triggers ERK activation, is one IL-1b–signaling
et al., 1999). In this assay, [3H]-ITZ-1 binding to Hsp90a led to intermediate (Vincenti and Brinckerhoff, 2002), and Raf-1 protein
the excitation of the scintillant contained in the SPA beads, levels are known to be decreased by Hsp90 inhibition (Stancato
and the concomitant light emission was measured. Using 40 et al., 1997; Sharp and Workman, 2006; Neckers, 2007). We
nM of [3H]-ITZ-1, specific interactions between His2-hHsp90a examined whether ITZ-1 could decrease Raf-1 protein levels
20 Chemistry & Biology 17, 18–27, January 29, 2010 ª2010 Elsevier Ltd All rights reserved

Chemistry & Biology
ITZ-1 a Client-Selective Hsp90 Inhibitor
ITZ-1
ITZ-12
To obtain insights into the ITZ-1–binding site, we assessed
whether ITZ-1 could inhibit ATPase activity. A colorimetric assay
for inorganic phosphate, based on the formation of a phospho-
molybdate complex and subsequent reaction with malachite
green, was used to measure the ATPase activity of human
Hsp90a (Rowlands et al., 2004). The N-terminal ligands exhibited
potent inhibitory activities (IC50 of GA, 17-AAG, RAD, and CCT
were 3.1 mM, 4.6 mM, 4.1 mM, and 3.7 mM, respectively), whereas
neither ITZ-1 nor NOV showed detectable activities at concen-
trations <100 mM (Figure 4A). These data suggest that ITZ-1
does not bind to the N-terminal ATP-binding site.
-
+
+
+
+
+
-
+
+
+
+
+
IL-1b (10 ng/ml)
(mM)
(mM)
Raf-1
b-Actin
100 76 63 42 32
100 110 85 42 20
100 110 91 96 95
100 120 99 84 99
Phospho-ERKs
Total-ERKs
Figure 3. Effects of ITZ-1 and ITZ-12 on the Raf-1 Protein Levels and
IL-1b–Induced ERK Activation in HACs
HACs were treated with ITZ-1 or ITZ-12 for 24 hr followed by stimulation with
IL-1b for 30 min. Raf-1 protein levels and total or phosphorylated protein levels
of ERK in equal amounts of whole-cell lysates were detected by western blot
analysis and quantitated by densitometry. b-actin was used as a control.
In the competitive inhibition of Hsp90a-ITZ-3 binding, GA,
17-AAG, and RAD at 100 mM and NOV at 1000 mM did not exhibit
any activity, whereas CCT at 100 mM showed partial inhibition
(Figure 4B). These results suggest that ITZ-1 does not bind
to Hsp90 at the same site as GA, 17-AAG, RAD, and NOV do
in solution, or the binding affinity of ITZ-1 for Hsp90 is much
stronger compared to those compounds.
using ITZ-12 with a lower Hsp90 binding affinity as a control
(Figure 2D; Figure S1B). HACs were pretreated with compounds
for 24 hr and then stimulated with 10 ng/ml of IL-1b for 30 min to
induce ERK activation. ITZ-1, but not ITZ-12, decreased Raf-1
protein levels and inhibited the IL-1b–induced ERK activation
at similar concentrations (Figure 3). ITZ-1, by binding to Hsp90,
might inhibit the MMP-13 production through an inhibition of
ERK activation via decreased Raf-1 protein levels.
The ITZ-1–binding site on Hsp90 was further examined by
analyzing several deletion fragments for binding to the ITZ-3–im-
mobilized matrix (Whitesell and Lindquist, 2005). Interestingly,
removal of 77% of the N-terminal domain and CR between resi-
dues 48 and 290 (designated D1) or removal of CR and 70% of
the middle domain between residues 197 and 528 (designated
D2) from His-hHsp90a retained ITZ-3 binding, whereas removal
of the C-terminal domain between 645 and 732 (designated D3)
markedly impaired ITZ-3 binding (Figure 4C). The binding of D1
and D2 to the ITZ-3–immobilized matrix was also competitively
inhibited by ITZ-1, suggesting that the ITZ-1–binding domain
on Hsp90 is within its C-terminal domain.
ITZ-1 Binds to the Hsp90 C Terminus Without Inhibiting
Hsp90 ATPase Activity
Hsp90 consists of three domains that have important functional
interactions: the N-terminal domain contains the ATP-binding
site and has ATPase activity, the middle domain interacts with
client proteins and cochaperons and activates the ATPase
activity, and the C-terminal domain is responsible for its inherent ITZ-1 Is Comparable to GA in the Induction of Hsp
dimerization. In eukaryotes, a flexible, highly charged region (CR) Expression
connects the N-terminal domain to the middle domain (Whitesell Because the Hsp90-binding domain of ITZ-1 seemed to be
and Lindquist, 2005). GA, 17-allylaminogeldanamycin (17-AAG), different from those of other Hsp90 inhibitors, including GA,
radicicol (RAD), and CCT018159 (CCT) bind to the N-terminal 17-AAG, and RAD, we determined to investigate the Hsp90
ATP-binding site, whereas novobiocin (NOV) has been reported inhibitory properties of ITZ-1. Hsp90 inhibitors have been
to bind with poor affinity to a second ATP-binding site on the reported to induce Hsp expression by dissociating the HSF1/
Hsp90 C-terminal domain (Marcu et al., 2000a, 2000b; Whitesell Hsp90 complex (Zou et al., 1998; Sittler et al., 2001; Mimnaugh
and Lindquist, 2005; Sharp and Workman, 2006; Neckers, 2007). et al., 2004; Shen et al., 2005). To test whether ITZ-1 could
A
B
C
ITZ-1
100
Binding to
ITZ-1
GA
80
60
40
20
0
ITZ-3 matrix
0 12.5 25 50 (mM)
1
210 272
CR
629 732
C
17-AAG
RAD
CCT
NOV
Wt
D1
D2
D3
+
+
+
-
N
M
48
290
197
528
645
0.1
1
10
100
-20
mM
Figure 4. ITZ-1 Binds to the C-terminal Domain of Hsp90 Without Inhibiting Hsp90 ATPase Activity
(A) Effects of ITZ-1 and various Hsp90 inhibitors on Hsp90a ATPase activity. A colorimetric assay for inorganic phosphate, based on the formation of a phospho-
molybdate complex and subsequent reaction with malachite green, was used to measure the ATPase activity of human Hsp90a. Results shown are means SD
(n = 3).
(B) Displacement studies with ITZ-1 and various Hsp90 inhibitors by the binding assay using ITZ-3–immobilized matrixes.
(C) Schematic representation of the human Hsp90a (His-hHsp90a; Wt) and its deletion mutants (D1 ꢁ D3). N, N-terminal domain; CR, flexible, highly charged
region; M, middle domain; C, C-terminal domain. Specific binding of Wt, D1, D2, and D3 to ITZ-3–immobilized matrixes was evaluated as in Figure 2C. Effect
of ITZ-1 on the interaction between ITZ-3–immobilized matrix and these Hsp90 proteins were assessed as in Figure 2D.
Chemistry & Biology 17, 18–27, January 29, 2010 ª2010 Elsevier Ltd All rights reserved 21

Chemistry & Biology
ITZ-1 a Client-Selective Hsp90 Inhibitor
IL-1b (-)
ITZ-1 (mM) 0 0.01 0.1
IL-1b (+)
7
6
5
4
3
2
1
0
7
6
5
4
3
2
1
0
Hsp70,IL-1(-)
Hsp70,IL-1(+)
Hsp90,IL-1(-)
Hsp90,IL-1(+)
A
B
Hsp70,IL-1(-)
Hsp70,IL-1(+)
Hsp90,IL-1(-)
Hsp90,IL-1(+)
1
0 0.01 0.1
1
1
Hsp70
Hsp90
b-Actin
IL-1b (-)
0 0.01 0.1
IL-1b (+)
GA (mM)
1
0 0.01 0.1
0
0.01 0.1
ITZ-1 (mM)
1
10
0
0.01 0.1
1
10
Hsp70
Hsp90
b-Actin
GA (mM)
C
8
7
6
5
4
3
2
1
0
8
GA
ITZ-1
7
6
5
4
3
2
1
0
17-AAG
RAD
CCT
NOV
ITZ-10
ITZ-11
ITZ-12
IP Ab : cIgG anti-Hsp90
cIgG
0
anti-Hsp90
ITZ-1
D
GA
1
(mM)
10
0
0.1
1
10 0.1
HSF1
Hsp90
1
2
3
4
5
6
7
8
9 10 11 12 13 14 15
0
0.01 0.1
mM
1
10
0
0.01 0.1
1
10 100 1,000
mM
Figure 5. HSF1 Activation and Hsp70 and Hsp90 Induction by ITZ-1
(A) Up-regulation of Hsp70 and Hsp90 mRNA by ITZ-1 or GA. HACs were treated with ITZ-1 or GA in the presence or absence of IL-1b (10 ng/ml) for 24 hr.
Messenger RNA levels of Hsp70 and Hsp90 were measured by real-time quantitative PCR. Results shown are means SD (n = 3).
(B) Up-regulation of Hsp70 and Hsp90 protein by 48 hr treatment with ITZ-1 or GA. Protein levels of Hsp70 and Hsp90 were detected by western blot analysis.
b-actin was used as a control.
(C) Induction of Hsp70 mRNA expression by ITZ-1 derivatives or various Hsp90 inhibitors in the presence of IL-1b (10 ng/ml) costimulation. Results shown are
means SD (n = 3).
(D) ITZ-1 and GA, but not ITZ-12, disrupted the HSF1-Hsp90 complex in the cell lysates obtained from IL-1b–stimulated HAC. Coimmunoprecipitation using an
anti-Hsp90 antibody followed by western blot analysis of HSF1 (top) and Hsp90 (bottom) was performed. Coimmunoprecipitation was performed in the presence
or absence of 1 mM ITZ-1, ITZ-12, or GA (lanes 1–6) and various concentrations of ITZ-1 or GA (lanes 7–15). See also Figure S2.
induce Hsp expression, HACs were stimulated with ITZ-1 or GA good agreement with their order of binding affinity for Hsp90a.
in the presence or absence of IL-1b costimulation. Both ITZ-1 Moreover, the potency of ITZ-1 in the dissociation of HSF1/
and GA increased mRNA and protein levels of Hsp70 and Hsp90 complexes was equal to that of GA (Figure 5D). This result
Hsp90 in HACs in the presence of IL-1b costimulation (Figures agreed well with the results shown in Figures 5A and 5B. These
5A and 5B). It is well known that there are various approaches results suggest that ITZ-1 induced Hsp expression via binding
to induce Hsp expression using pharmacological interventions. to Hsp90 and that the potency of ITZ-1 was at least equivalent
Therefore, we investigated in detail whether ITZ-1 induced Hsp to that of GA.
expression by binding to Hsp90.
We first examined the structure-activity relationship in the ITZ-1 Is Over Three Orders of Magnitude Less Potent
induction of Hsp70 expression for the ITZ-1 derivatives. With in the Degradation of Several Hsp90 Client Proteins
IL-1b costimulation, ITZ-1 and ITZ-10 dramatically up-regulated than GA
Hsp70 mRNA, whereas ITZ-11 and ITZ-12 showed little or no Hsp90 inhibitors have been reported to induce degradation of
Hsp70 up-regulation at concentrations <10 mM (Figure 5C). its multiple client proteins (Whitesell and Lindquist, 2005; Sharp
This rank order was in good agreement with their order in the and Workman, 2006; Neckers, 2007). To further address the
binding affinity for Hsp90a (Figure 2D; Figure S1B). Thus, ITZ-1 Hsp90 inhibitory property of ITZ-1, HACs were treated with
may induce Hsp70 expression via binding to Hsp90. ITZ-1 was ITZ-1 or known Hsp90 inhibitors for 24 hr, and then degrada-
more potent than known Hsp90 inhibitors in the induction of tion of Hsp90 client proteins, such as Raf-1, glucocorticoid
Hsp70 expression; the concentrations of ITZ-1, GA, 17-AAG, receptor (GR), Akt, epidermal growth factor receptor (EGF-R),
RAD, CCT, and NOV required to double the Hsp70 mRNA level and receptor interacting protein (RIP), was analyzed (Stancato
(the minimum effective concentration, MEC2.0) were 0.042 mM, et al., 1997; Lewis et al., 2000; Bagatell et al., 2005; Zhang et al.,
0.19 mM, 0.55 mM, >10 mM, 0.97 mM, and 240 mM, respectively 2007a, 2007b). Interestingly, ITZ-1 was less potent by over three
(Figure 5C). On the other hand, ITZ-1 was the least toxic to HACs orders of magnitude than was GA in the degradation of a majority
(Figure S2); the effective concentration 50 (EC50) of ITZ-1, GA, of these client proteins (Figure 6A). ITZ-1 did not induce R50%
17-AAG, RAD, CCT, and NOV were >10 mM, 0.32 mM, 1.2 mM, degradation of client proteins except Raf-1 at 10 mM, whereas
13 mM, >10 mM, and 200 mM, respectively; thus, this potent GA simultaneously reduced these client proteins at concentra-
Hsp induction was not a byproduct of cytotoxicity.
tions >0.01 mM. Moreover, ITZ-1 at 0.1 mM induced more than
The coimmunoprecipitation (IP) study conducted in the pres- a 40% decrease of Raf-1, whereas there were no other client
ence of free compounds in the lysates revealed that ITZ-1 at proteins that were decreased by more than 40% with ITZ-1 at
1 mM, but not ITZ-12, drastically decreased the amount of HSF1 10 mM. Both ITZ-1 and GA required IL-1b costimulation to induce
coprecipitated with Hsp90 (Figure 5D). The rank order was in Hsp expression (Figures 5A and 5B); thus, we investigated the
22 Chemistry & Biology 17, 18–27, January 29, 2010 ª2010 Elsevier Ltd All rights reserved

Chemistry & Biology
ITZ-1 a Client-Selective Hsp90 Inhibitor
A
B
ITZ-1
GA
ITZ-1
GA
(mM)
(mM)
(mM)
(mM)
Raf-1
Raf-1
b-Actin
GR
100 71 57 41 13
100 72 53 44 38
100 77 73 20 13
100 110 90 10
9
b-Actin
GR
100 87 110 16 17
100 83 91 15 35
100 95 98 120 120
100 96 87 92 77
100 110 130 130 120
100 85 100 110 100
100 110 120 17 21
100 88 68 33 20
Akt
RIP
b-Actin
b
-Actin
Akt
EGF-R
100 94 91 94 76
100 94 99 86 64
100 100 110 34 24
100 130 140 15 16
100 94 79 21 28
100 98 110 100 110
b-Actin
RIP
b-Actin
Figure 6. ITZ-1 Selectively Induces Raf-1 Protein Degradation
HACs were treated with ITZ-1 or GA for 24 hr in the absence (A) or presence (B) of IL-1b (10 ng/ml) costimulation. Protein levels of various Hsp90 client proteins in
equal amounts of whole-cell lysates were assessed by western blot analysis and quantitated by densitometry. b-actin was used as a control. Dotted lines indicate
the concentration of compound required to induce a R50% decrease of each client protein level. See also Figure S3.
client protein degradation by ITZ-1 or GA in the presence of IL-1b role in the identification of drug-binding proteins (Harding et al.,
costimulation. However, IL-1b costimulation did not affect their 1989; Taunton et al., 1996). However, in many cases, nonspecific
client protein degradation properties (Figure 6B). Other Hsp90 binding of proteins to solid support or to the compound itself
inhibitors also simultaneously induced degradation of these has complicated identification of target proteins, and examina-
client proteins (Figure S3A). The R50% degradation of these tion of the precise connection between the function of the
five client proteins by ITZ-1, GA, 17-AAG, RAD, CCT, and NOV selected binding proteins and activity of the drugs was not done
was seen at concentrations of >10 mM, 0.01 mM, 0.1 mM, 0.1 mM, sufficiently (Shimizu et al., 2000; Sato et al., 2007). To facilitate
10 mM, and 1000 mM, respectively.
drug target identification, we estimated the potency of each
To determine whether ITZ-1 down-regulates Raf-1 at the compound on the matrix after coupling with a long alkyl chain,
protein level, HACs were treated with ITZ-1 together with cyclo- instead of a spacer arm, at their terminal carboxyl or amino
heximide (CHX), a protein synthesis inhibitor (Nomura et al., group. We also adopted a system that can select activity-asso-
2005). Evaluation of cellular Raf-1 protein level by western blot ciated structure-specific binding proteins of compounds using
analysis revealed that preexisting Raf-1 protein was markedly three columns of active compounds immobilized matrixes.
decreased by ITZ-1 and CHX. In contrast, when cells were These devices might confer our higher purification efficiency of
treated with CHX alone, Raf-1 protein decreased to only 44% Hsp90 as the target protein of ITZ-1.
of the initial amount after 24 hr, indicating a protein half-life of
about 24 hr. These data demonstrate that down-regulation of pathway inhibitor (U0126), but not a p38 kinase inhibitor
Raf-1 protein occurred at the protein level (Figure S3B). (SB203580) or a JNK inhibitor (SP600125), selectively inhibited
We have previously shown that ITZ-1 as well as an ERK-MAPK
Previous research has suggested a link between GA and IL-1b–induced MMP-13 production (Kimura et al., 2009). Here
degradation of Hsp90 client proteins via the proteasome we showed that the rank order of ITZ-1 derivatives in the inhibi-
pathway (Mimnaugh et al., 2004; Nomura et al., 2005). We exam- tion of MMP-13 production agreed well with their binding affinity
ined whether proteasomal degradation mediates the loss of for Hsp90a. Furthermore, ITZ-1, but not ITZ-12, decreased Raf-1
Raf-1 in ITZ-1–treated HACs using the proteasome inhibitor protein levels and inhibited the IL-1b–induced ERK activation.
MG132. Pretreatment with MG132 resulted in a remarkable Thus ITZ-1, by binding to Hsp90, might inhibit MMP-13 produc-
suppression of ITZ-1–induced Raf-1 degradation in HACs (Fig- tion by suppressing IL-1b–induced ERK activation via decreased
ure S3C). This result suggests that Raf-1 was degraded by the Raf-1 protein levels.
proteasome.
The rank order of ITZ-1 derivatives affecting the dissociation of
HSF1/Hsp90 complex and up-regulation of Hsp70 expression
also agreed well with their binding affinity for Hsp90a. These
results suggest that ITZ-1 induced Hsp expression by binding
DISCUSSION
ITZ-1 is a chondroprotective agent that suppresses IL-1b– to Hsp90. The Hsp up-regulation might confer the chondropro-
induced cartilage degradation in vitro and in vivo. ITZ-1 also tective effects of ITZ-1 in vitro and in vivo (Kimura et al., 2009)
inhibits IL-1b–induced MMP-13 production and suppresses because Hsp70 induction has been shown to prevent NO-
NO-induced chondrocyte death (Kimura et al., 2009). Here, we induced apoptosis in chondrocytes and to efficiently help in the
show the successful identification of Hsp90 as a target protein recovery from osteoarthritis (Terauchi et al., 2003; Ku¨ hn et al.,
of ITZ-1. Drug-affinity chromatography has played an important 2004; Ge et al., 2006; Grossin et al., 2006).
Chemistry & Biology 17, 18–27, January 29, 2010 ª2010 Elsevier Ltd All rights reserved 23

Chemistry & Biology
ITZ-1 a Client-Selective Hsp90 Inhibitor
inhibitors. Our results, as well as results with A4 and AEG3482,
suggest that the C-terminal domain is suitable for the develop-
ment of HSF-1–selective Hsp90 inhibitors.
Table 1. Hsp90 Inhibitory Properties of ITZ-1 and Several Known
Hsp90 Inhibitors
ITZ-1 GA 17-AAG RAD CCT NOV
Up-regulation of Hsps, especially that of Hsp70, has been
shown to be of great therapeutic potential for various diseases,
such as ischemic heart disease, diabetes, stroke, trauma, and
various neurodegenerative diseases (Auluck and Bonini, 2002;
Hsp70 mRNA induction,
0.042 0.19 0.55
>10 0.97 240
MEC2.0 (mM)
Raf-1 degradation, EC50 (mM)
1
0.01 0.1
0.01 0.1
0.1 10
0.1 10
1,000
1,000
Client protein
>10
¨
Sittler et al., 2001; Shen et al., 2005; Soti et al., 2005). Several
degradation, EC50 (mM)
reports have shown that GA, via Hsp70 induction, is effective
as a cytoprotective agent (Sittler et al., 2001; Auluck and Bonini,
2002; Shen et al., 2005). However, it is well known that GA is
toxic to cells probably as a result of nonspecific degradation of
a broad range of client proteins. ITZ-1 was at least 300 times
less toxic than GA for growing HAC, probably because of its
client selectivity (Table 1). ITZ-1 type Hsp90 inhibitors may be
beneficial as Hsp inducers with lower cytotoxicity.
Cytotoxicity, EC50 (mM)
>10
0.32 1.2
13
>10 200
ATPase inhibition, IC50 (mM) >100 3.1 4.6
4.1 3.7 >100
MEC2.0 for Hsp70 mRNA induction was determined as the concentration
of each compound to double the Hsp70 mRNA level. EC50 for Raf-1
degradation was determined as the concentration of each compound
required to induce a R50% decrease in Raf-1 protein levels. EC50 for
client protein degradation was determined as the concentration of each
compound required to induce a R50% decrease of the five client protein
levels.
Both ITZ-1 and GA required IL-1b costimulation to induce
Hsp expression in chondrocytes. To become transcriptionally
competent, HSF1 requires homotrimerization, translocation from
So far, Hsp90 inhibitors, especially those that bind at the N- the cytoplasm to the nucleus, hyperphosphorylation, and acqui-
terminal domain, have been considered to simultaneously induce sition of DNA-binding activity (Zou et al., 1998). Both ITZ-1 and
degradation or activation of multiple Hsp90 client proteins GA dissociated HSF1/Hsp90 complexes in the cell lysate; there-
(Whitesell and Lindquist, 2005; Sharp and Workman, 2006; fore, IL-1b signal may affect the activation step of free HSF1.
Neckers, 2007). In fact, all known Hsp90 inhibitors used in this Actually, IL-1b stimulation has been reported to involve trimer
study simultaneously reduced cellular protein levels of several formation of HSF1 in human embryonic lung fibroblast cells
Hsp90 client proteins in HACs. However, the Hsp90 inhibitory (Sasaki et al., 2002).
property of ITZ-1 was client protein selective, compared with
In summary, our results suggest that ITZ-1 is a client selective
these other Hsp90 inhibitors. ITZ-1 strongly induced Hsps Hsp90 inhibitor that selectively induces robust HSF1 activation
expression and mildly caused Raf-1 degradation but scarcely and mild Raf-1 degradation. ITZ-1 type Hsp90 inhibitors are
induced degradation of client proteins other than Raf-1. ITZ-1, promising as Hsp70 inducers, as well as disease-modifying
GA, 17-AAG, RAD, CCT, and NOV required >240, 0.053, 0.18, anti-OA drugs. It seems difficult to discover Hsp90 inhibitors
>0.01, 10, and 4.2 times lower concentrations, respectively, for with such mechanisms of action by classical genetics because
Hsp70 expression, compared with client protein degradation Hsp90 null mutants are lethal in eukaryotes (Young et al.,
(Table 1). Because HACs were used for these studies, different 2001). Furthermore, ITZ-1 type Hsp90 inhibitors could not be
uptake and metabolism rates of ITZ-1 by cells in each assay obtained by conventional assays such as the ATPase assay.
had a negligible effect on the experimental results.
Thus, the two binding assays we used here are useful to identify
This unique property of ITZ-1 might be due to its unique additional ITZ-1 type Hsp90 inhibitors. Our study demonstrates
Hsp90-binding site; almost all known Hsp90 inhibitors bind to that forward chemical genetics can result in the more efficient
the N-terminal ATP-binding domain and inhibit the ATPase identification of new drugs and new molecular targets (Lokey,
activity (Rowlands et al., 2004; Whitesell and Lindquist, 2005; 2003; Saghatelian and Cravatt, 2005; Spring, 2005; Kawasumi
Sharp and Workman, 2006; Neckers, 2007), whereas ITZ-1 binds and Nghiem, 2007).
to Hsp90 at a site in the C-terminal domain without affecting
ATPase activity. Although NOV has been reported to bind to SIGNIFICANCE
the Hsp90 C-terminal domain, its binding affinity was too weak
to understand the contribution of this site to the regulation of ITZ-1, a lead compound for an antiosteoarthritis (OA) drug
Hsp90 function (Marcu et al., 2000a, 2000b; Whitesell and Lind- program, was previously found to have protective effects
quist, 2005). ITZ-1 may be useful for better understanding the against interleukin-1b (IL-1b)–induced cartilage degradation
function of the Hsp90 C-terminal domain. Recently, a NOV in vitro and in vivo by both selective inhibition of IL-1b–
analog A4, which could induce client protein degradation at induced matrix metalloproteinase–13 (MMP-13) production
ꢀ70-fold lower concentrations than NOV, has been reported to via inhibition of extracellular signal–regulated protein kinase
up-regulate Hsp90 levels at concentrations 1,000–10,000 fold (ERK) activation and suppression of nitric oxide (NO)–
lower than that required for client protein degradation (Yu et al., induced chondrocyte death. Here, we report the elucidation
2005; Ansar et al., 2007; Lu et al., 2009). AEG3482, an antiapop- of its precise mechanisms of action. Heat shock protein 90
totic compound that has been shown to bind Hsp90 at a site (Hsp90) was identified as a specific ITZ-1–binding protein.
different from GA, has also been reported to induce HSF-1– ITZ-1 induced degradation of an Hsp90 client protein, Raf-
dependent Hsp70 production without causing client protein 1, which triggers IL-1b–induced ERK activation and up-regu-
degradation (Salehi et al., 2006; Gallo, 2006). Therefore, it is con- lation of a cytoprotective heat shock protein 70 (Hsp70) by
ceivable that ITZ-1, a C-terminal inhibitor, possesses a mecha- dissociating heat shock factor 1 (HSF1)/Hsp90 complexes,
nism of action distinct from that of GA and other N-terminal which may account for its cytoprotective effect as well
24 Chemistry & Biology 17, 18–27, January 29, 2010 ª2010 Elsevier Ltd All rights reserved

Chemistry & Biology
ITZ-1 a Client-Selective Hsp90 Inhibitor
to the columns. After a wash with five volumes of binding buffer, binding
proteins were eluted with five volumes of elution buffer (10 mM Tris-HCl
[pH 7.5], 20% glycerol, 1 mM EDTA, 500 mM NaCl, and 6 M Urea). The eluted
proteins were analyzed by SDS-PAGE followed by silver staining.
as a selective inhibition of MMP-13 production. Almost all
known Hsp90 inhibitors, including geldanamycin (GA), have
been reported to bind to the N-terminal ATP-binding site
of Hsp90 and simultaneously induce degradation and activa-
tion of its multiple client proteins. However, ITZ-1 specifi-
cally bound to the C-terminal domain of Hsp90 without
affecting Hsp90s ATPase activity. Interestingly, ITZ-1
strongly induced HSF1 activation and caused mild Raf-1
degradation but scarcely induced degradation of a broad
range of Hsp90 client proteins besides Raf-1. These results
suggest that ITZ-1 is a client-selective Hsp90 inhibitor.
ITZ-1 type Hsp90 inhibitors are promising as Hsp70 inducers
for various diseases, as well as disease-modifying anti-OA
drugs. It seems difficult to discover Hsp90 inhibitors with
such a mechanism of action by classical genetics because
Hsp90 null mutants are lethal in eukaryotes. Furthermore,
an ITZ-1 type Hsp90 inhibitor could not be obtained by
conventional assays such as the ATPase assay. Our study
demonstrates that forward chemical genetics can result in
the efficient identification of new drugs and new molecular
targets.
Protein Identification by LC-MS/MS
The eluted proteins were subjected to SDS-PAGE, and then the gels were
stained with CBB. Protein spots were excised from the gel and subjected to
in-gel digestion followed by LC-MS/MS sequence analysis at APRO Life
Science Institute.
Expression and Purification of Recombinant Proteins
Expression and purification of recombinant proteins were accomplished by
the pET expression system (Merck) with His Bind Resin according to the
manufacturer’s instructions. Detailed methods are described in the Supple-
mental Experimental Procedures.
In Vitro Binding Assay with ITZ-3–Immobilized Matrix
One microgram of His-hHsp90a or deletion mutants was incubated with
sample in 200 ml of binding buffer (10 mM Tris-HCl [pH 7.5], 20% glycerol,
1 mM EDTA, 500 mM NaCl, and 1 mM DTT) at room temperature for 30 min.
After addition of 5 ml of ITZ-3–immobilized matrix or control matrix, the
mixtures were further incubated for 40 min with agitation. The matrixes were
washed five times with 200 ml of binding buffer and then heated for 5 min at
95ꢂC in SDS-PAGE sample buffer. The samples were subjected to SDS-
PAGE followed by CBB staining.
EXPERIMENTAL PROCEDURES
Evaluation of MAPK Activation and Hsp90 Client Protein
Degradation
Materials
GA, malachite green, polyvinyl alcohol, ammonium molybdate, CHX, and anti-
b-actin antibody (A-5441) were from Sigma. Anti-Raf-1 antibody (sc-133), anti-
GR antibody (sc-8992), anti-EGF-R antibody (sc-03), and protein A/G-agarose
(sc-2003) were from Santa Cruz Biotechnology. Phosphorylated and nonphos-
phorylated state specific antibodies for ERK p44/p42 (9100), anti-Akt antibody
(9270), and anti-rabbit IgG, HRP-linked antibody (7074) were from Cell
Signaling Technology. Anti-Hsp70 antibody (H53220), anti-Hsp90 antibody
(H38220), and anti-RIP antibody (551042) were from BD Bioscience. Anti-
HSF1 antibody (SPA-901) was from Assay Design. Peroxidase-labeled affinity
purified antibody to mouse IgG (g) (074-1802) was from KPL; 17-AAG, RAD,
and NOV were from Wako. CCT was from Tocris Bioscience. ATP was from
Roche. IL-1b was from R & D Systems. HACs and Chondrocyte Growth
Medium (CGM) were from Lonza Walkersville. A 1:1 (v/v) mixture of Dulbecco’s
modified Eagle’s MEM (DMEM) and Ham’s F-12 medium (DMEM/F-12) was
from Invitrogen. Detailed methods for the syntheses of ITZ-1 derivatives are
described in Supplemental Information.
ERK activation was evaluated as described elsewhere (Kimura et al., 2009). To
evaluate client protein degradation, HACs pr-cultured in CGM at a density of
4 3 104 cells/ml in 6-well plates for 1 day were further cultured in DMEM/F-12
supplemented with 100 U/ml penicillin, 100 mg/ml streptomycin, and 20 mg/ml
gentamicin at 37ꢂC in 5% CO2/95% humidified air for 1 day, and then treated
with samples for 24 hr. HACs were lysed with lysis buffer (Kimura et al., 2009)
and then subjected to western blot analysis.
Western Blot Analysis
Samples were subjected to SDS-PAGE, transferred onto nitrocellulose filters
(Hybond-ECL; GE Healthcare), and immunoblotted with the primary antibodies
followed by horseradish peroxidase–conjugated secondary antibodies. The
immune complexes were visualized with an enhanced chemiluminescence
system (ECL Detection Kit; GE Healthcare) and Hyperfilm ECL (GE Healthcare),
according to the manufacturer’s instructions. Protein levels were calculated
from densitometrical analysis by GS-800 Calibrated Densitometer (Bio-Rad).
Assay for MMP Production
ATPase Activity Assay
IL-1b–induced production of MMPs by HACs was measured as described
elsewhere (Kimura et al., 2009).
The ATPase activity assay was performed as described elsewhere, using
human Hsp90a (ATGen) (Rowlands et al., 2004).
Evaluation of the Induction of Hsp Expression
Drug-Affinity Chromatography
HACs maintained in CGM were suspended in DMEM/F-12 supplemented with
50 U/ml penicillin and 50 mg/ml streptomycin at a density of 5.5 3 104 cells/ml
and were cultured in 96-well or 12-well plates at 37ꢂC in 5% CO2/95% humid-
ified air for 1 day. After medium exchange, HACs were treated with samples
and 10 ng/ml of IL-1b for 24 or 48 hr. Real-time quantitative PCR was per-
formed as described elsewhere (Kimura et al., 2009). Primers used for Hsp70
analysis were Hsp70-F (50-CGAGAGGGTGTCAGCCAAGA-30) and Hsp70-R
(50-AGCCACGAGATGACCTCTTGAC-30). Primers used for Hsp90 analysis
were Hsp90-F (50-TTGGTCCTGTGCGGTCACT-30) and Hsp90-R (50-CCATC
GGTTGGTCTTGGG-30). For western blot analysis, HACs were washed with
PBS and then lysed with lysis buffer (Kimura et al., 2009).
Two milliliters of 100 mM imidazo[5,1-c][1,4]thiazine derivatives dissolved in
75–95% DMF and 20 mg of 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide
hydrochloride (EDC) were coupled to 1 ml of AF-Amino TOYOPEARL 650M
(Tosoh) or AF-Carboxy TOYOPEARL 650M (Tosoh), according to the manufac-
turer’s instructions.
HACs were homogenized in four volumes of binding buffer (10 mM Tris-HCl
[pH 7.5], 20% glycerol, 1 mM EDTA, 100 mM NaCl, 0.5 mM p-APMSF, and
1 mM DTT) with
a teflon homogenizer. The homogenate was spun at
100,000 3 g for 60 min at 4ꢂC, and then the supernatant was recovered and
applied to the columns of drug-coupled matrixes. Columns were washed twice
with five volumes of binding buffer, and then binding proteins were eluted with
five volumes of elution buffer (10 mM Tris-HCl [pH 7.5], 20% glycerol, 1 mM
EDTA, 1 M NaCl, 0.5 mM p-APMSF, and 1 mM DTT) by collecting 5 fractions
equal to column volume, and with three volumes of glycine-HCl buffer (0.1 M
glycine-HCl [pH 2.7]).
Immunoprecipitation
HACs were cultured in the presence 10 ng/ml of IL-1b for 24 hr, were washed
with PBS, were lysed with RIPA buffer (50 mM Tris-HCl [pH 8.0], 150 mM NaCl,
1% NP-40, 0.1% SDS, 0.5% deoxycholic acid, 1 mM sodium orthovanadate V,
and 1 mM microcystin-LR) with proteinase inhibitors (Complete; Boeringer
His-hHsp90a was diluted to 50 mg/ml with binding buffer (10 mM Tris-HCl
[pH 7.5], 20% glycerol, 1 mM EDTA, 500 mM NaCl, and 1 mM DTT) and applied
Chemistry & Biology 17, 18–27, January 29, 2010 ª2010 Elsevier Ltd All rights reserved 25

Chemistry & Biology
ITZ-1 a Client-Selective Hsp90 Inhibitor
Mannheim), and were centrifuged at 16,0003 g for 10 min, and then superna-
tant (lysates) was recovered. Next, 600 ml of lysates (2 mg/ml) were preincu-
bated with samples for 30 min and then were incubated with 30 mg of anti-
Hsp90 antibody (SPA-840; Assay Design) or control IgG (I4131; Sigma) for
30 min. Subsequently, the lysates were incubated with a 50-ml slurry of protein
A/G-agarose with rotation for 3 hr at 4ꢂC. The immunoprecipitates were
collected by centrifugation, washed seven times with RIPA buffer, and then
subjected to SDS-PAGE followed by western blot analysis using anti-HSF1
or anti-Hsp90 antibodies.
Grossin, L., Cournil-Henrionnet, C., Pinzano, A., Gaborit, N., Dumas, D.,
Etienne, S., Stoltz, J.F., Terlain, B., Netter, P., Mir, L.M., and Gillet, P. (2006).
Gene transfer with Hsp70 in rat chondrocytes confers cytoprotection in vitro
and during experimental osteoarthritis. FASEB J. 20, 65–75.
Harding, M.W., Galat, A., Uehling, D.E., and Schreiber, S.L. (1989). A receptor
for the immunosuppressant FK506 is a cis-trans peptidyl-prolyl isomerase.
Nature 341, 758–760.
Kawasumi, M., and Nghiem, P. (2007). Chemical genetics: elucidating bio-
logical systems with small-molecule compounds. J. Invest. Dermatol. 127,
1577–1584.
SUPPLEMENTAL INFORMATION
Kimura, H., Yukitake, H., Suzuki, H., Tajima, Y., Gomaibashi, K., Morimoto, S.,
Funabashi, Y., Yamada, K., and Takizawa, M. (2009). The chondroprotective
agent ITZ-1 inhibits interleukin-1b-induced matrix metalloproteinase-13
production and suppresses nitric oxide-induced chondrocyte death. J. Phar-
macol. Sci. 110, 201–211.
Supplemental Information includes three figures, one table, and Supplemental
Experimental Procedures and can be found with this article online at doi:10.
1016/j.chembiol.2009.12.012.
ACKNOWLEDGMENTS
Ku¨ hn, K., D’Lima, D.D., Hashimoto, S., and Lotz, M. (2004). Cell death in carti-
lage. Osteoarthritis Cartilage 12, 1–16.
We thank Kiyofumi Yamada, Seiichi Tanida, Hidekazu Sawada, Masaomi
Miyamoto, Hideaki Nagaya, and Takeo Wada for their helpful discussions.
We also thank Yoshimi Sato for help with the drug-affinity chromatography.
The authors declare they have no competing financial interests.
Lewis, J., Devin, A., Miller, A., Lin, Y., Rodriguez, Y., Neckers, L., and Liu, Z.G.
(2000). Disruption of Hsp90 function results in degradation of the death domain
kinase, receptor-interacting protein (RIP), and blockage of tumor necrosis
factor-induced nuclear factor-kB activation. J. Biol. Chem. 275, 10519–10526.
Lokey, R.S. (2003). Forward chemical genetics: progress and obstacles on the
Received: September 4, 2009
Revised: December 19, 2009
Accepted: December 21, 2009
Published online: January 21, 2010
path to a new pharmacopoeia. Curr. Opin. Chem. Biol. 7, 91–96.
Lu, Y., Ansar, S., Michaelis, M.L., and Blagg, B.S.J. (2009). Neuroprotective
activity and evaluation of Hsp90 inhibitors in an immortalized neuronal cell
line. Bioorg. Med. Chem. 17, 1709–1715.
REFERENCES
Marcu, M.G., Chadli, A., Bouhouche, I., Catelli, M., and Neckers, L.M. (2000a).
The heat shock protein 90 antagonist novobiocin interacts with a previously
unrecognized ATP-binding domain in the carboxyl terminus of the chaperone.
J. Biol. Chem. 275, 37181–37186.
Ahmed, S., Wang, N., Lalonde, M., Goldberg, V.M., and Haqqi, T.M. (2004).
Green tea polyphenol epigallocatechin-3-gallate (EGCG) differentially inhibits
interleukin-1b-induced expression of matrix metalloproteinase-1 and ꢁ13 in
human chondrocytes. J. Pharmacol. Exp. Ther. 308, 767–773.
Marcu, M.G., Schulte, T.W., and Neckers, L.M. (2000b). Novobiocin and
related coumarins and depletion of heat shock protein 90-dependent signaling
proteins. J. Natl. Cancer Inst. 92, 242–248.
Ahmed, S., Wang, N., Hafeez, B.B., Cheruvu, V.K., and Haqqi, T.M. (2005).
Punica granatum L. extract inhibits IL-1b-induced expression of matrix metal-
loproteinases by inhibiting the activation of MAP kinases and NF-kB in human
chondrocytes in vitro. J. Nutr. 135, 2096–2102.
Mimnaugh, E.G., Xu, W., Vos, M., Yuan, X., Isaacs, J.S., Bisht, K.S., Gius, D.,
and Neckers, L. (2004). Simultaneous inhibition of hsp90 and the proteasome
promotes protein ubiquitination, causes endoplasmic reticulum-derived
cytosolic vacuolization, and enhances antitumor activity. Mol. Cancer Ther.
3, 551–566.
Aigner, T., Sachse, A., Gebhard, P.M., and Roach, H.I. (2006). Osteoarthritis:
pathobiology targets and ways for therapeutic intervention. Adv. Drug Deliv.
Rev. 58, 128–149.
Neckers, L. (2007). Heat shock protein 90: the cancer chaperone. J. Biosci. 32,
Ansar, S., Burlison, J.A., Hadden, M.K., Yu, X.M., Desino, K.E., Bean, J., Neck-
ers, L., Audus, K.L., Michaelis, M.L., and Blagg, B.S.J. (2007). A non-toxic
Hsp90 inhibitor protects neurons from Ab-induced toxicity. Bioorg. Med.
Chem. Lett. 17, 1984–1990.
517–530.
Nomura, M., Nomura, N., and Yamashita, J. (2005). Geldanamycin-induced
degradation of Chk1 is mediated by proteasome. Biochem. Biophys. Res.
Commun. 335, 900–905.
Auluck, P.K., and Bonini, N.M. (2002). Pharmacological prevention of Parkin-
Pettipher, E.R., Higgs, G.A., and Henderson, B. (1986). Interleukin 1 induces
leukocyte infiltration and cartilage proteoglycan degradation in the synovial
joint. Proc. Natl. Acad. Sci. USA 83, 8749–8753.
son disease in drosophila. Nat. Med. 8, 1185–1186.
Bagatell, R., Beliakoff, J., David, C.L., Marron, M.T., and Whitesell, L. (2005).
Hsp90 inhibitors deplete key anti-apoptotic proteins in pediatric solid tumor
cells and demonstrate synergistic anticancer activity with cisplatin. Int. J.
Cancer 113, 179–188.
Presle, N., Cipolletta, C., Jouzeau, J.Y., Abid, A., Netter, P., and Terlain, B.
(1999). Cartilage protection by nitric oxide synthase inhibitors after intraartic-
ular injection of interleukin-1b in rats. Arthritis Rheum. 42, 2094–2102.
Bottomley, K.M., Borkakoti, N., Bradshaw, D., Brown, P.A., Broadhurst, M.J.,
Budd, J.M., Elliott, L., Eyers, P., Hallam, T.J., Handa, B.K., et al. (1997).
Inhibition of bovine nasal cartilage degradation by selective matrix metallopro-
teinase inhibitors. Biochem. J. 323, 483–488.
Rowlands, M.G., Newbatt, Y.M., Prodromou, C., Pearl, L.H., Workman, P., and
Aherne, W. (2004). High-throughput screening assay for inhibitors of heat-
shock protein 90 ATPase activity. Anal. Biochem. 327, 176–183.
Cawston, T.E., Ellis, A.J., Humm, G., Lean, E., Ward, D., and Curry, V. (1995).
Interleukin-1 and oncostatin M in combination promote the release of collagen
fragments from bovine nasal cartilage in culture. Biochem. Biophys. Res.
Commun. 215, 377–385.
Saghatelian, A., and Cravatt, B.F. (2005). Assignment of protein function in the
postgenomic era. Nat. Chem. Biol. 1, 130–142.
Salehi, A.H., Morris, S.J., Ho, W.C., Dickson, K.M., Doucet, G., Milutinovic, S.,
Durkin, J., Gillard, J.W., and Barker, P.A. (2006). AEG3482 is an antiapoptotic
compound that inhibits Jun kinase activity and cell death through induced
expression of heat shock protein 70. Chem. Biol. 13, 213–223.
Gallo, K.A. (2006). Targeting HSP90 to halt neurodegeneration. Chem. Biol. 13,
115–116.
Ge, Z., Hu, Y., Heng, B.C., Yang, Z., Ouyang, H., Lee, E.H., and Cao, T. (2006).
Sasaki, H., Sato, T., Yamaguchi, N., Okamoto, T., Kobayashi, D., Iyama, S.,
Kato, J., Matsunaga, T., Takimoto, R., Takayama, T., et al. (2002). Induction
of heat shock protein 47 synthesis by TGF-b and IL-1b via enhancement of
the heat shock element binding activity of heat shock transcription factor 1.
J. Immunol. 168, 5178–5183.
Osteoarthritis and therapy. Arthritis Rheum. 55, 493–500.
Graziani, F., Aldegheri, L., and Terstappen, G.C. (1999). High throughput scin-
tillation proximity assay for the identification of FKBP-12 ligands. J. Biomol.
Screen. 4, 3–7.
26 Chemistry & Biology 17, 18–27, January 29, 2010 ª2010 Elsevier Ltd All rights reserved

Chemistry & Biology
ITZ-1 a Client-Selective Hsp90 Inhibitor
Sato, S., Kwon, Y., Kamisuki, S., Srivastava, N., Mao, Q., Kawazoe, Y., and
Uesugi, M. (2007). Polyproline-rod approach to isolating protein targets of
bioactive small molecules: isolation of a new target of indomethacin. J. Am.
Chem. Soc. 129, 873–880.
Terauchi, R., Takahashi, K.A., Arai, Y., Ikeda, T., Ohashi, S., Imanishi, J.,
Mazda, O., and Kubo, T. (2003). Hsp70 prevents nitric oxide-induced
apoptosis in articular chondrocytes. Arthritis Rheum. 48, 1562–1568.
Vincenti, M.P., and Brinckerhoff, C.E. (2002). Transcriptional regulation of
collagenase (MMP-1, MMP-13) genes in arthritis: integration of complex
signaling pathways for the recruitment of gene-specific transcription factors.
Arthritis Res. 4, 157–164.
Sharp, S., and Workman, P. (2006). Inhibitors of the Hsp90 molecular chap-
erone: current status. Adv. Cancer Res. 95, 323–348.
Shen, H.Y., He, J.C., Wang, Y., Huang, Q.Y., and Chen, J.F. (2005). Geldana-
mycin induces heat shock protein 70 and protects against MTPT-induced
dopaminergic neurotoxicity in mice. J. Biol. Chem. 280, 39962–39969.
Whitesell, L., and Lindquist, S.L. (2005). Hsp90 and the chaperoning of cancer.
Nat. Rev. Cancer 5, 761–772.
Shimizu, N., Sugimoto, K., Tang, J., Nishi, T., Sato, I., Hiramoto, M., Aizawa, S.,
Hatakeyama, M., Ohba, R., Hatori, H., et al. (2000). High-performance affinity
beads for identifying drug receptors. Nat. Biotechnol. 18, 877–881.
Yasuhara, R., Miyamoto, Y., Akaike, T., Akuta, T., Nakamura, M., Takami, M.,
Morimura, N., Yasu, K., and Kamijo, R. (2005). Interleukin-1b induces death in
chondrocyte-like ATDC5 cells through mitochondrial dysfunction and energy
depletion in a reactive nitrogen and oxygen species-dependent manner.
Biochem. J. 389, 315–323.
Sittler, A., Lurz, R., Lueder, G., Priller, J., Hayer-Hartl, M.K., Hartl, F.U.,
Lehrach, H., and Wanker, E.E. (2001). Geldanamycin activates a heat shock
response and inhibits huntingtin aggregation in a cell culture model of Hunting-
ton’s disease. Hum. Mol. Genet. 10, 1307–1315.
Young, J.C., Moarefi, I., and Hartl, F.U. (2001). Hsp90: a specialized but essen-
¨
Soti, C., Nagy, E., Giricz, Z., Vigh, L., Csermely, P., and Ferdinandy, P. (2005).
tial protein-folding tool. J. Cell Biol. 154, 267–273.
Heat shock proteins as emerging therapeutic targets. Br. J. Pharmacol. 146,
769–780.
Yu, X.M., Shen, G., Neckers, L., Blake, H., Holzbeierlein, J., Cronk, B., and
Blagg, B.S.J. (2005). Hsp90 inhibitors identified from a library of novobiocin
analogues. J. Am. Chem. Soc. 127, 12778–12779.
Spring, D.R. (2005). Chemical genetics to chemical genomics: small molecules
offer big insights. Chem. Soc. Rev. 34, 472–482.
Stancato, L.F., Silverstein, A.M., Owens-Grillo, J.K., Chow, Y.H., Jove, R., and
Pratt, W.B. (1997). The hsp90-binding antibiotic geldanamycin decreases Raf
levels and epidermal growth factor signaling without disrupting formation of
signaling complexes or reducing the specific enzymatic activity of Raf kinase.
J. Biol. Chem. 272, 4013–4020.
Zhang, H., Yang, Y.C., Zhang, L., Fan, J., Chung, D., Choi, D., Grecko, R.,
Timony, G., Karjian, P., Boehm, M., and Burrows, F. (2007a). Dimeric ansamy-
cins—a new class of antitumor Hsp90 modulators with prolonged inhibitory
activity. Int. J. Cancer 120, 918–926.
Zhang, L., Nephew, K.P., and Gallagher, P.J. (2007b). Regulation of death-
associated protein kinase. Stabilization by Hsp90 heterocomplexes. J. Biol.
Chem. 282, 11795–11804.
Takaishi, H., Kimura, T., Dalal, S., Okada, Y., and D’Armiento, J. (2008). Joint
diseases and matrix metalloproteinases: a role for MMP-13. Curr. Pharm. Bio-
technol. 9, 47–54.
Taunton, J., Hassig, C.A., and Schreiber, S.L. (1996). A mammalian histone de-
acetylase related to the yeast transcriptional regulator rpd3p. Science 272,
408–411.
Zou, J., Guo, Y., Guettouche, T., Smith, D.F., and Voellmy, R. (1998). Repres-
sion of heat shock transcription factor HSF1 activation by Hsp90 (Hsp90
complex) that forms a stress-sensitive complex with HSF1. Cell 94, 471–480.
Chemistry & Biology 17, 18–27, January 29, 2010 ª2010 Elsevier Ltd All rights reserved 27