ANTICANCER RESEARCH 29: 797-808 (2009)

Review Inhibition of Cancer Invasion and Metastasis by Targeting the Molecular Heat-shock Protein 90

FUMITAKA KOGA1, KAZUNORI KIHARA1 and LEN NECKERS2

1Department of Urology, Tokyo Medical and Dental University Graduate School, 1-5-45 Yushima, Bunkyo-ku, Tokyo 113-8519, Japan; 2Urologic Oncology Branch, Center for Cancer Research, National Cancer Institute, Bethesda, MD 20892, U.S.A.

Abstract. Heat-shock protein 90 () is a molecular strategies have made their way from the bench to the patient, chaperone required for the stability and function of numerous and are now used as part of the standard clinical modalities oncogenic signaling proteins that determine the hallmarks of available against a variety of malignancies. cancer. Receptor tyrosine kinases (RTKs) and hypoxia-inducible Heat-shock protein 90 (Hsp90) is a molecular chaperone factor-1 (HIF-1)-mediated pathways, commonly activated in required for the stability and function of conditionally aggressive cancer, potentiate each other and thus efficiently activated signaling proteins as well as multiple mutated, promote cancer invasion and metastasis. Hsp90 inhibitors, by chimeric and/or overexpressed signaling proteins, which interacting specifically with a single molecule, Hsp90, cause the promote cancer cell growth and survival (1, 2). Its role in destabilization and eventual degradation of multiple Hsp90 client cancer has received much attention in the past decade since proteins. These agents impede the cellular processes involved in Hsp90 functions as an integral part of the machinery that cancer invasion and metastasis by simultaneously impairing allows cancer cells to escape normal regulation. Several multiple Hsp90-dependent signaling proteins including HIF-1α, selective Hsp90 inhibitors are currently undergoing clinical most RTKs and their hub mediators Src, Raf-1 and Akt. Recently, evaluation. This review focuses on the role of Hsp90 in a fraction of Hsp90 identified on the cell surface has been found cancer invasion and angiogenesis, the crucial steps for the to play a crucial role in cancer invasion and metastasis. The development of cancer metastasis, and discusses the potential first-in-class , 17-allylamino-17-demethoxy- utility and problems of Hsp90 inhibitors in clinical settings. , is currently in phase II clinical trials. The potential utility and problems of Hsp90 inhibitors in clinical Physiological Role of Hsp90 settings are discussed. A fuller understanding of the roles of as a Molecular Chaperone Hsp90 in cancer biology and accumulating clinical data on Hsp90 inhibitors will guide us toward the goal of optimizing the Hsp90 is an extremely abundant chaperone protein, use of these agents in the clinic. comprising up to 1 to 2% of total cellular protein. It increases by 2- to 10-fold during environmental stress; this Accumulating evidence on molecular pathways in cancer up-regulation could be explained as a part of the protective cells has allowed the development of novel compounds that mechanisms that enhance cell survival (3). Hsp90 exerts target specific oncogenic pathways. Molecular therapeutic essential housekeeping functions as a molecular chaperone, such as facilitating protein re-folding, translocation of proteins between cellular compartments, suppression of protein aggregation, supporting functional maturation of Correspondence to: Fumitaka Koga, MD, Ph.D., Department of signaling proteins and facilitating normal protein turnover Urology, Tokyo Medical and Dental University Graduate School, (4). A folding pathway can either lead to a functional, 1-5-45 Yushima, Bunkyo-ku, Tokyo 113-8519, Japan. Tel/Fax: +81 properly folded client protein, or to degradation of the same 358035295, e-mail: [email protected] protein (5). Under stress, this dual function helps repair or Key Words: Cancer, invasion, metastasis, molecular chaperones, heat- degrade damaged proteins to re-establish protein homeostasis shock protein 90, geldanamycin, molecularly targeted therapeutics, (6). To date, more than 100 proteins have been reported as review. Hsp90 clients (see the website maintained by D Picard,

0250-7005/2009 $2.00+.40 797 ANTICANCER RESEARCH 29: 797-808 (2009) http://www.picard.ch/downloads/Hsp90interactors.pdf). limitless replicative potential, v) self-sufficiency in growth These include many kinases such as ErbB2 (also known as signaling, and vi) ability to evade . As shown in HER2/neu), Akt, Bcr-Abl (a fusion protein of BCR and Abl Figure 2, Hsp90 plays a pivotal role in the acquisition and associated with chronic myelogenous leukemia), Raf-1, and maintenance of each of these capabilities. Given the potential cyclin dependent kinase (CDK) 4, as well as transcription of Hsp90 inhibition to simultaneously abrogate the six factors such as hypoxia-inducible factor-1α (HIF-1α), , hallmarks of a cancer cell, targeting Hsp90 would appear to and steroid receptors. be a reasonable anticancer strategy. Figure 1 shows nucleotide-dependent cycling of the Hsp90-based super-chaperone machine. Newly synthesized Hsp90 Inhibitors Hsp90 client protein first associates with an Hsp70/Hsp40 chaperone complex (7). This assembly is then linked to The benzoquinoid ansamycin antibiotics, first isolated from Hsp90 via p60Hop, an Hsp90/Hsp70-interacting protein (8). the actinomycete Streptomyces hygroscopicus var. geldanus At this point, Hsp90 is in its ADP-bound conformation. var. nova, include geldanamycin (GA) and its semi-synthetic Replacement of ADP by ATP in an amino-terminal derivatives, 17-allylamino-17-demethoxygeldanamycin nucleotide binding pocket of Hsp90 alters the Hsp90 (17-AAG) and water-soluble 17-demethylaminoethylamino- conformation, releasing the p60Hop and the Hsp70/Hsp40 17-demethoxygeldanamycin (17-DMAG) (16) (Figure 3). complex, and recruiting another set of co-chaperones, GA has shown antitumor activity against v-Src-transformed including p50Cdc37 and p23. This ATP-bound chaperone cancer cells in vivo as well as in vitro (17). GA was initially complex folds and stabilizes client proteins, and temporarily considered to exhibit its antitumor activity by directly holds them in a state that is now ready to bind to ligand or inhibiting tyrosine kinases. However, an in vitro kinase assay respond to a stimulus. If a client protein fails to receive its revealed that GA neither directly interacts with Src nor stimulus or bind to its ligand, it will be recycled through new inhibits its phosphorylating activity. Finally, Hsp90 was chaperone interactions. If the ATP-dependent step does not identified as the direct target of GA (18). take place, the client will eventually undergo ubiquitination Ansamycin Hsp90 inhibitors, radicicol derivatives (19) and proteasomal degradation (9, 10). Thus, ATP/ADP and other small molecule N-terminal inhibitors, as well as nucleotide occupancy of the amino-terminal binding pocket the peptide shepherdin (20, 21) replace the nucleotide in the in Hsp90 drives the Hsp90 super-chaperone machine to bind Hsp90-binding pocket with an affinity greater than ATP or to, chaperone and release client proteins. ADP, thus short-circuiting the Hsp90 chaperone cycle. From the point of view of evolution, Hsp90 functions as a Cycling of the chaperone machine is critical to its function. capacitor of genetic and phenotypic variation (11, 12). The The Hsp90 inhibitors prevent the nucleotide-dependent chaperoning function of Hsp90 can buffer genetic variation cycling and thus interfere with the Hsp90 chaperone function at the protein level, allowing polymorphic variants to as if one stops the rotation of a bicycle wheel by inserting a accumulate silently until they are released in response to stick between the spokes. GA binding promotes stable stress, which creates increased demand for Hsp90 to assembly of the super-chaperone machine that resembles the facilitate refolding of its usual client proteins as well as new, chaperone’s ADP-bound conformation, resulting in targeting stress-destabilized clients. The stock of genetic variation in of the client proteins to the ubiquitin-proteasomal system, certain individuals can thereby exceed the buffering capacity where they are degraded (22) (Figure 4). Thus, Hsp90 of Hsp90 and produce diverse, genotype-specific inhibition uniformly shortens the half-lives of Hsp90 client phonotypes. In this way, previously hidden genetic variations proteins. may become available for natural selection to enhance the Since Hsp90 is essential for the function of normal cells as survival of distinct genotypes within a population. well as tumor cells, one might be concerned that inhibition of its functions might not be selective for malignancy. Both The Cancer Chaperone Hsp90 preclinical and clinical observations, however, have shown that Hsp90 inhibitors can be given in vivo at doses and Cancer is a disease of genetic instability. Although only a schedules that exert antitumor activity without causing host few specific mutations seem to be sufficient for the toxicity (23). This is the case for several Hsp90 inhibitors, development of the malignant phenotype, over 10,000 including a synthetic purine mimetic PU24FCl (24) and the mutations have been found at the time of diagnosis of colon novel peptidomimetic inhibitor of the amino-terminal Hsp90 cancer (13, 14). Such genetic instability allows a cell to nucleotide-binding site, shepherdin (20, 21), as well as eventually acquire six capabilities that characterize 17-AAG and 17-DMAG (25-27). Indeed, Hsp90 inhibitors malignancies, as suggested by Hanahan and Weinberg (15). have been shown to concentrate in tumor tissue, while being These are i) sustained angiogenesis, ii) invasion and rapidly cleared from normal tissue with a half-life similar to metastasis, iii) insensitivity to anti-growth signaling, iv) that of the drug in plasma (24-27). Recently, a possible

798 Koga et al: Inhibition of Invasion and Metastasis by Targeting Hsp90 (Review)

Figure 1. Nucleotide-dependent cycling of the Hsp90-based super-chaperone machine. Ub, ubiquitin.

explanation for this phenomenon has been proposed. Kamal lymphatic vessel formation (angiogenesis), iii) intravasation and colleagues suggested that enhanced drug binding to into the vessels, iv) transit in the circulation and tumor cell Hsp90 reflects the activity of the Hsp90 extravasation, and v) establishment of metastases (30). chaperone machine in tumor vs. normal cells (28). Tumor Acquisition of invasive and angiogenic potential is thus Hsp90 is present in an activated super-chaperone complex critical for metastasis. The RTK and HIF-1 pathways play a with high affinity for ATP and high ATPase activity, whereas pivotal role in these processes (31-33). Hsp90 in normal cells predominantly exists in nucleotide- RTK signaling induces diverse biological effects including unbound, uncomplex form. Thus, assembly of an Hsp90 cell proliferation, cell survival, invasion and angiogenesis. multi-chaperone complex is a prerequisite for the activation Highly invasive and metastatic carcinomas commonly of Hsp90. The difference in Hsp90 conformation most likely overexpress RTKs such as the hepatocyte growth factor contributes to the tumor selectivity of the Hsp90 inhibitors. (HGF) receptor Met and epidermal growth factor receptor (EGFR) (31, 32). Clinically, overexpression of RTKs is Reciprocal Relationship between Receptor associated with tumor progression and poor prognosis of Tyrosine Kinase (RTK) and HIF-1 to Promote patients with diverse carcinomas (31, 34-36). Moreover, the Cancer Invasion and Angiogenesis expression of these RTK ligands in the cancer tissues and serum is associated with a poor prognosis of the cancer- Invasion into adjacent tissues and metastasis to distant sites bearing patients (37, 38). are major features of cancer cells and the cause of 90% of Hypoxia exists in proliferative tumors and the extent of human cancer death (29). The metastatic process consists of tumor hypoxia correlates with tumor progression and i) invasion to adjacent stroma, ii) new blood and/or metastasis (39). Under hypoxia, HIF-1α protein is stabilized,

799 ANTICANCER RESEARCH 29: 797-808 (2009)

Figure 2. The cancer chaperone Hsp90. Hsp90 plays a central role in supporting all of the six hallmark traits of cancer, suggested by Hanahan and Weinberg (15), by chaperoning a subset of client proteins as shown. HIF-1α, hypoxia-inducible factor-1α; VEGF, vascular endothelial growth factor; RTK, receptor tyrosine kinase; MMP2, matrix metalloproteinase 2; CDK4, 6, cyclin-dependent kinase 4, 6; IGF-1R, insulin-like growth factor-1 receptor; IKK, inhibitor of kappa B kinase.

is dimerized with HIF-1β, and thus can bind to the hypoxia response element to transactivate a battery of genes involved in the promotion of angiogenesis and glucose metabolism to adapt to a stressful environment (40). The induction of these genes not only gives tumors a survival advantage, but also promotes invasion and angiogenesis (33, 40). Indeed, HIF- 1α expression is associated with unfavorable clinical outcomes in various malignancies (41, 42). As mentioned above, both the RTK and HIF-1 pathways are commonly activated in aggressive cancer cells. Recent studies have suggested that these two pathways potentiate each other by forming a positive feedback loop (43, 44) (Figure 5). RTK signaling transcriptionally induces HIF-1α (45, 46) while hypoxia enhances RTK signaling via HIF-1- mediated up-regulation of RTKs (43, 44, 47). This reciprocal Figure 3. Chemical structures of geldanamycin (GA) and its derivatives, relationship between the two distinct oncogenic pathways 17-allylamino-17-demethoxygeldanamycin (17-AAG) and water-soluble sustains activation of both pathways, thus efficiently 17-demethylaminoethylamino-17-demethoxygeldanamycin (17-DMAG). promoting cancer invasion and angiogenesis (43, 44, 47).

Targeting Hsp90 to Simultaneously Inhibit simultaneous impairment of multiple, crucial proinvasive and Multiple Proinvasive and Proangiogenic Pathways proangiogenic pathways.

Since Hsp90 clients, including RTKs and HIF-1α, play key Multiple inhibition of RTK signaling pathways. The majority roles in cancer invasion and angiogenesis, targeting Hsp90 of wild-type RTKs, including EGFR (59), Met (43, 52) and is a reasonable anticancer treatment strategy to prevent insulin-like growth factor-1 receptor (IGF-1R) (60), are cancer metastases. Indeed, many preclinical studies have subject to regulation only during their maturation process. In shown that Hsp90 inhibitors efficiently impede cancer contrast, ErbB2 (61) and some activating mutant forms of invasion, angiogenesis and/or metastasis (43, 48-58). The RTK, such as mutant EGFR (62), Kit (the stem cell factor anti-metastatic effects of Hsp90 inhibitors are ascribed to receptor) (63) and fms-like tyrosine kinase receptor-3 (Flt-3)

800 Koga et al: Inhibition of Invasion and Metastasis by Targeting Hsp90 (Review)

Figure 4. Geldanamycin (GA) short-circuits the cycling of the Hsp90-based super-chaperone machine, eventually leading Hsp90 client proteins to degradation via the ubiquitin-proteasomal system. Ub, ubiquitin.

Figure 6. Receptor tyrosine kinase (RTK) signaling and its biological effects promoting cancer invasion and angiogenesis. *Indicates Hsp90 Figure 5. Positive feedback loop between receptor tyrosine kinase (RTK) client. Erk, extracellular signal-regulated kinase; PI3K, phosphoinositide signaling and hypoxia-inducible factor-1 (HIF-1) pathway to promote 3-kinase; VEGF, vascular endothelial growth factor; uPA, urokinase-type cancer invasion and angiogenesis. plasminogen activator; MMP, matrix metalloproteinase.

801 ANTICANCER RESEARCH 29: 797-808 (2009)

(64), are dependent on Hsp90 also for the maintenance of the the stability of VEGF receptors depend on Hsp90 function and mature form and are subject to regulation of their kinase thus Hsp90 inhibitors destabilize VEGF receptors (55, 73). activity by the chaperone. Inhibition of Hsp90 induces slow Therefore, Hsp90 inhibitors efficiently block the proliferation degradation of the former “weak” clients, while it causes a and differentiation of endothelial cells, consequently inhibiting dramatic decrease in kinase activity and protein level of the the neovascularization of proliferating tumors. latter “strong” clients. The degree to which clients depend on Hsp90 is defined by the accessibility of Hsp90 to its specific Inhibition of activation of focal adhesions. Upon extracellular binding loop, which is normally buried in the quaternary stimuli such as growth factors and the engagement of structure of the kinase (65-67). Subtle mutations of a weak extracellular matrix to integrins, focal adhesions are formed client can transform it into a strong client (65). and activated. Focal adhesion kinase (FAK) is a crucial The Ras/Raf-1/extracellular-signal regulated kinase (Erk) component of focal adhesions and functions as an integrator and phosphoinositide 3-kinase (PI3K)/Akt pathways mediate to control cell motility (80). In response to extracellular diverse biological effects of the RTK signaling relevant to stimuli, FAK is activated by autophosphorylation at Tyr397 invasion and angiogenesis (Figure 6). Hsp90 inhibition and recruits Src to the phosphorylated Tyr397 residue via the potently abates RTK signaling pathways, even at a low dose Src homology-2 domain of Src, leading to conformational (43), not only by destabilizing and inactivating RTKs activation of Src and formation of a dual-activated FAK/Src themselves but also by destabilizing their downstream nodal complex as a part of activated focal adhesions (80, 81). effector kinases including Src (68), Raf-1 (69) and Akt (70). Within this complex, Src further phosphorylates tyrosine Furthermore, proinvasive and proangiogenic cellular residues of the FAK and fully activated FAK recruits p130Cas processes induced by activation of these hub kinases, such and growth factor receptor-bound protein 2 (Grb2) to as increased cell motility (71), proteolytic activation of modulate focal adhesion dynamics (80). Extracellular stimuli matrix metalloproteinase (MMP) 2 (72) and vascular also activate a small GTPase Rho, inducing organized endothelial growth factor (VEGF) signaling (73), are assembly of the contractile actin-myosin filament (stress sensitive to Hsp90 inhibition. The mechanisms by which fiber), which anchors to the focal adhesions, eventually Hsp90 inhibition interferes with these processes are promoting cell motility (82, 83). described later. Collected data suggest that Hsp90 plays an essential role in the activation of focal adhesions. Hsp90 participates in Inhibition of HIF-1 activity. HIF-1α is normally labile and focal adhesion assembly induced by VEGF (73, 84), HGF expressed at low levels in normoxic cells. The molecular and integrin ligands (43). GA impairs such extrinsic stimuli- basis for the instability of HIF-1α in normoxia depends on induced FAK activation, focal adhesion assembly, stress fiber von Hippel Lindau protein (VHL), an E3 ubiquitin ligase formation, and consequently cell motility (43, 73, 84). targeting HIF-1α for -dependent degradation (40). Hypoxia normally impairs VHL function, thus allowing Roles of Extracellular Hsp90 in HIF-1α to accumulate. Cancer Invasion and Metastasis HIF-1α interacts with Hsp90 (74) and is destabilized by Hsp90 inhibitors under hypoxic conditions (51, 75). In the Although Hsp90 is an intracellular chaperone, it has recently presence of Hsp90 inhibitors, HIF-1α undergoes VHL- been found expressed on the cell surface of highly metastatic independent proteasomal degradation (75). In addition, malignant cells such as melanoma cells (85) and Hsp90 inhibitors hamper HIF-1 transcriptional activity (75- fibrosarcoma cells (72). Cell surface Hsp90 is also involved 77). HIF-1 derived from Hsp90 inhibitor-treated cells fails in migration of neuronal cells during development of the to bind DNA (76), suggesting that Hsp90 critically mediates nervous system (86). In melanoma cells, cell surface Hsp90 the proper conformation of HIF-1 heterodimers and the expression positively correlates with metastatic potential recruitment of cofactors. (85), and cell migration is experimentally inhibited by HIF-1 transactivates a variety of genes including RTKs (33, Hsp90 antibody (86), or by GA coupled with cell- 40, 43, 44, 47), urokinase-type plasminogen activator (uPA) impermeable agarose beads (72). These data suggest that cell (78), uPA receptor (40, 79), MMP2 (40) and VEGF (33, 40), surface Hsp90 is involved in cancer cell invasion and eventually promoting cancer progression. VEGF, the most metastasis and is a potential target for the development of potent angiogenic factor, stimulates angiogenesis by binding novel invasion and metastasis inhibitors. to and activating its cell surface receptor, promoting As for the role of cell surface Hsp90 in cell invasion, endothelial cell proliferation and migration. Hsp90 inhibition Eustace et al. have shown the involvement of cell surface attenuates VEGF signaling in two ways. First, Hsp90 Hsp90 in the activation process of MMP2 (72), which can inhibition remarkably reduces VEGF secretion by interdicting degrade extracellular matrices to provide cancer cells with the RTK and HIF-1 pathways as described above. In addition, access to the vasculature, allowing tumor dissemination.

802 Koga et al: Inhibition of Invasion and Metastasis by Targeting Hsp90 (Review)

MMP2 forms a complex with cell surface Hsp90 and Hsp90 clinical trials of nonansamycin Hsp90 inhibitors. Although no antibody or cell-impermeable GA beads impair MMP2 episode has been reported in the clinical trials, preclinical maturation, leading to diminished MMP2 activity (72). data suggest that 17-AAG may promote bone metastasis by Recently, Tsutsumi et al. have shown that cell surface Hsp90 potentiating osteoclast activity (95). However, this can be is involved in actin cytoskeletal dynamics (lamellipodia prevented by concurrent treatment with the bisphosphonate formation), which is necessary for cell motility (87). A cell- alendronate (Yano et al., unpublished data). In addition, impermeable GA derivative, DMAG-N-oxide, impairs the Hsp90 inhibition can transiently activate oncogenic kinases integrin-mediated formation of focal adhesion complex, such as Src (66) and ErbB2 (96) although the potential which incorporates Src, attenuating cancer cell invasion in a clinical significance of these effects remains to be determined. dose-dependent manner (87). Furthermore, DMAG-N-oxide A role for Hsp90 in the maturation of human ether-a-go-go- dramatically suppresses lung colonization of melanoma cells related gene (hERG) channels (97) has suggested the (87). Similar anti-metastatic effects in a murine melanoma potential for cardiac toxicity with chronic Hsp90 inhibition, model were also observed for Hsp90 antibody (56, 87). Very and particular care must be taken when using Hsp90 recently, cell surface Hsp90 has been reported to directly inhibitors in combination with doxorubicin (98). interact with the extracellular domain of ErbB2 (88). ErbB3 Given the unique properties of Hsp90 as a therapeutic ligand stimulation induces ErbB2/ErbB3 heterodimer target and tolerable toxicities in clinical trials, considerable formation, ErbB2 phosphorylation and activation of its efforts have been made to develop novel Hsp90 inhibitors downstream signaling, eventually promoting cell migration. with better pharmacological and toxicity profiles. Distinct Disruption of the cell surface Hsp90/ErbB2 interaction using from small molecules acting at the amino-terminal ATP- Hsp90 antibody inhibits the ErbB2/ErbB3 heterodimer binding site, other approaches such as affecting co-chaperone formation, reduces ErbB2 phosphorylation and its interactions would potentially interfere more specifically downstream signaling, and impairs cytoskeletal reorganization with the super-chaperone complex. and cell motility (88). Thus, a growing body of data suggests that cell surface Hsp90 plays a crucial role in integrin- and Conclusion RTK-mediated cell invasion, and that targeting cell surface Hsp90 represents a potential therapeutic approach to inhibit The era of molecularly potentially holds cancer invasion and metastasis. great promise in the treatment of cancer. Hsp90 inhibitors are an attractive anticancer strategy in that although they are Potential Utility and Problems of Hsp90 Inhibitors directed toward a specific molecular target, they in Clinical Settings simultaneously inhibit multiple oncogenic signaling pathways on which cancer cells depend for growth, survival To date, several phase I clinical trials of 17-AAG have been and metastasis. The complexity of Hsp90 function is only completed (25, 89-94). Encouragingly, the drug is well now becoming clear. A more complete understanding of the tolerated in both adult and pediatric patients. Hepatic toxicity is biological activity of Hsp90 in cancer and the continuing the major dose-limiting toxicity, which requires an intermittent accumulation of clinical data on Hsp90 inhibitors, with close dosing schedule. The current requirement for intermittent coordination between bench and bed, will help us to resolve dosing of GA derivatives appears to allow serum drug the above mentioned issues and thereby to define the optimal concentrations that would impair “the six hallmarks of cancer” use of these unique and interesting compounds in the clinic. but would not induce apoptosis in cancer cells. For instance, 2 infusion of 17-AAG at doses of 320-450 mg/m yielded Cmax Acknowledgements of 16 μM, maintaining its biologically active concentration of >100 nM for 24 hours or longer (25). By interfering with the This work was supported, in part, by grants-in-aids for Scientific functions of Hsp90 clients involved in “evading apoptosis”, GA Research (19791102) from the Ministry of Education, Culture, derivatives might be most useful in combination with other Sports, Science and Technology of Japan. therapeutics such as , radiotherapy and other molecular targeting therapies. Anti-invasive and metastatic References effects are also expected and indeed, numerous instances of stable disease have been reported in the 17-AAG clinical trials. 1 Neckers L: Hsp90 inhibitors as novel cancer chemotherapeutic agents. Trends Mol Med 8: S55-61, 2002. Despite the tolerability of Hsp90 inhibitors in the clinic, 2 Neckers L and Neckers K: Heat-shock protein 90 inhibitors as there are still many problems to be resolved. Determining novel cancer chemotherapeutic agents. Expert Opin Emerg whether hepatotoxicity of 17-AAG is caused directly by Drugs 7: 277-288, 2002. inhibition of Hsp90 function or is secondary to drug 3 Buchner J: Hsp90 & Co. – a holding for folding. Trends metabolism awaits results of currently ongoing phase I Biochem Sci 24: 136-141, 1999.

803 ANTICANCER RESEARCH 29: 797-808 (2009)

4 Whitesell L and Lindquist SL: HSP90 and the chaperoning of 22 Schneider C, Sepp-Lorenzino L, Nimmesgern E, Ouerfelli O, cancer. Nat Rev Cancer 5: 761-772, 2005. Danishefsky S, Rosen N and Hartl FU: Pharmacologic shifting 5 Wegele H, Muller L and Buchner J: Hsp70 and Hsp90 – a relay of a balance between protein refolding and degradation mediated team for protein folding. Rev Physiol Biochem Pharmacol 151: by Hsp90. Proc Natl Acad Sci USA 93: 14536-14541, 1996. 1-44, 2004. 23 Kelland LR, Sharp SY, Rogers PM, Myers TG and Workman P: 6 Jolly C and Morimoto RI: Role of the heat shock response and DT-Diaphorase expression and tumor cell sensitivity to molecular chaperones in oncogenesis and cell death. J Natl 17-allylamino, 17-demethoxygeldanamycin, an inhibitor of heat- Cancer Inst 92: 1564-1572, 2000. shock protein 90. J Natl Cancer Inst 91: 1940-1949, 1999. 7 Hernandez MP, Sullivan WP and Toft DO: The assembly and 24 Vilenchik M, Solit D, Basso A, Huezo H, Lucas B, He H, Rosen intermolecular properties of the hsp70-Hop-hsp90 molecular N, Spampinato C, Modrich P and Chiosis G: Targeting wide- chaperone complex. J Biol Chem 277: 38294-38304, 2002. range oncogenic transformation via PU24FCl, a specific 8 Smith DF, Sullivan WP, Marion TN, Zaitsu K, Madden B, inhibitor of tumor Hsp90. Chem Biol 11: 787-797, 2004. McCormick DJ and Toft DO: Identification of a 60-kiloDalton 25 Banerji U, O’Donnell A, Scurr M, Pacey S, Stapleton S, Asad Y, stress-related protein, p60, which interacts with hsp90 and Simmons L, Maloney A, Raynaud F, Campbell M, Walton M, hsp70. Mol Cell Biol 13: 869-876, 1993. Lakhani S, Kaye S, Workman P and Judson I: Phase I 9 Mimnaugh EG, Chavany C and Neckers L: Polyubiquitination pharmacokinetic and pharmacodynamic study of 17-allylamino, and proteasomal degradation of the p185c-erbB-2 receptor 17-demethoxygeldanamycin in patients with advanced protein-tyrosine kinase induced by geldanamycin. J Biol Chem malignancies. J Clin Oncol 23: 4152-4161, 2005. 271: 22796-22801, 1996. 26 Eiseman JL, Lan J, Lagattuta TF, Hamburger DR, Joseph E, Covey 10 Xu W, Marcu M, Yuan X, Mimnaugh E, Patterson C and JM and Egorin MJ: Pharmacokinetics and pharmacodynamics of Neckers L: Chaperone-dependent E3 ubiquitin ligase CHIP 17-demethoxy 17-[[(2-dimethylamino)ethyl]amino]geldanamycin mediates a degradative pathway for c-ErbB2/Neu. Proc Natl (17DMAG, NSC 707545) in C.B-17 SCID mice bearing MDA- Acad Sci USA 99: 12847-12852, 2002. MB-231 human breast cancer xenografts. Cancer Chemother 11 Queitsch C, Sangster TA and Lindquist S: Hsp90 as a capacitor Pharmacol 55: 21-32, 2005. of phenotypic variation. Nature 417: 618-624, 2002. 27 Xu L, Eiseman JL, Egorin MJ and D’Argenio DZ: Physiologically- 12 Rutherford SL and Lindquist S: Hsp90 as a capacitor for based pharmacokinetics and molecular pharmacodynamics of 17- morphological evolution. Nature 396: 336-342, 1998. (allylamino)-17-demethoxygeldanamycin and its active metabolite 13 Hahn WC and Weinberg RA: Modelling the molecular circuitry in tumor-bearing mice. J Pharmacokinet Pharmacodyn 30: 185- of cancer. Nat Rev Cancer 2: 331-341, 2002. 219, 2003. 14 Stoler DL, Chen N, Basik M, Kahlenberg MS, Rodriguez-Bigas 28 Kamal A, Thao L, Sensintaffar J, Zhang L, Boehm MF, Fritz LC MA, Petrelli NJ and Anderson GR: The onset and extent of and Burrows FJ: A high-affinity conformation of Hsp90 confers genomic instability in sporadic colorectal tumor progression. tumour selectivity on Hsp90 inhibitors. Nature 425: 407-410, 2003. Proc Natl Acad Sci USA 96: 15121-15126, 1999. 29 Sporn MB: The war on cancer. Lancet 347: 1377-1381, 1996. 15 Hanahan D and Weinberg RA: The hallmarks of cancer. Cell 30 Sahai E: Illuminating the metastatic process. Nat Rev Cancer 7: 100: 57-70, 2000. 737-749, 2007. 16 Isaacs JS: Heat-shock protein 90 inhibitors in antineoplastic 31 Birchmeier C, Birchmeier W, Gherardi E and Vande Woude GF: therapy: is it all wrapped up? Expert Opin Investig Drugs 14: Met, metastasis, motility and more. Nat Rev Mol Cell Biol 4: 569-589, 2005. 915-925, 2003. 17 Uehara Y, Hori M, Takeuchi T and Umezawa H: Phenotypic 32 Holbro T, Civenni G and Hynes NE: The ErbB receptors and change from transformed to normal induced by benzoquinonoid their role in cancer progression. Exp Cell Res 284: 99-110, 2003. ansamycins accompanies inactivation of p60src in rat kidney 33 Semenza GL: HIF-1 and tumor progression: pathophysiology cells infected with Rous sarcoma virus. Mol Cell Biol 6: 2198- and therapeutics. Trends Mol Med 8: S62-67, 2002. 2206, 1986. 34 Dei Tos AP: The biology of epidermal growth factor receptor and 18 Whitesell L, Mimnaugh EG, De Costa B, Myers CE and Neckers its value as a prognostic/predictive factor. Int J Biol Markers 22: LM: Inhibition of heat-shock protein HSP90-pp60v-src S3-9, 2007. heteroprotein complex formation by benzoquinone ansamycins: 35 Lengyel E, Prechtel D, Resau JH, Gauger K, Welk A, essential role for stress proteins in oncogenic transformation. Lindemann K, Salanti G, Richter T, Knudsen B, Vande Woude Proc Natl Acad Sci USA 91: 8324-8328, 1994. GF and Harbeck N: C-Met overexpression in node-positive 19 Soga S, Shiotsu Y, Akinaga S and Sharma SV: Development of breast cancer identifies patients with poor clinical outcome radicicol analogues. Curr Cancer Drug Targets 3: 359-369, independent of Her2/neu. Int J Cancer 113: 678-682, 2005. 2003. 36 Mita K, Nakahara M and Usui T: Expression of the insulin-like 20 Gyurkocza B, Plescia J, Raskett CM, Garlick DS, Lowry PA, growth factor system and cancer progression in hormone-treated Carter BZ, Andreeff M, Meli M, Colombo G and Altieri DC: prostate cancer patients. Int J Urol 7: 321-329, 2000. Antileukemic activity of shepherdin and molecular diversity of 37 Gohji K, Nomi M, Niitani Y, Kitazawa S, Fujii A, Katsuoka Y and hsp90 inhibitors. J Natl Cancer Inst 98: 1068-1077, 2006. Nakajima M: Independent prognostic value of serum hepatocyte 21 Plescia J, Salz W, Xia F, Pennati M, Zaffaroni N, Daidone growth factor in bladder cancer. J Clin Oncol 18: 2963-2971, 2000. MG, Meli M, Dohi T, Fortugno P, Nefedova Y, Gabrilovich 38 Thogersen VB, Sorensen BS, Poulsen SS, Orntoft TF, Wolf H DI, Colombo G and Altieri DC: Rational design of and Nexo E: A subclass of HER1 ligands are prognostic markers shepherdin, a novel anticancer agent. Cancer Cell 7: 457-468, for survival in bladder cancer patients. Cancer Res 61: 6227- 2005. 6233, 2001.

804 Koga et al: Inhibition of Invasion and Metastasis by Targeting Hsp90 (Review)

39 Brizel DM, Scully SP, Harrelson JM, Layfield LJ, Bean JM, 54 Park JW, Yeh MW, Wong MG, Lobo M, Hyun WC, Duh QY and Prosnitz LR and Dewhirst MW: Tumor oxygenation predicts for Clark OH: The heat-shock protein 90-binding geldanamycin the likelihood of distant metastases in human soft tissue inhibits cancer cell proliferation, down-regulates oncoproteins, sarcoma. Cancer Res 56: 941-943, 1996. and inhibits epidermal growth factor-induced invasion in thyroid 40 Semenza GL: Targeting HIF-1 for cancer therapy. Nat Rev cancer cell lines. J Clin Endocrinol Metab 88: 3346-3353, 2003. Cancer 3: 721-732, 2003. 55 Sanderson S, Valenti M, Gowan S, Patterson L, Ahmad Z, 41 Bos R, van der Groep P, Greijer AE, Shvarts A, Meijer S, Pinedo Workman P and Eccles SA: Benzoquinone ansamycin heat- HM, Semenza GL, van Diest PJ and van der Wall E: Levels of shock protein 90 inhibitors modulate multiple functions required hypoxia-inducible factor-1alpha independently predict prognosis for tumor angiogenesis. Mol Cancer Ther 5: 522-532, 2006. in patients with lymph node negative breast carcinoma. Cancer 56 Stellas D, Karameris A and Patsavoudi E: Monoclonal antibody 97: 1573-1581, 2003. 4C5 immunostains human melanomas and inhibits melanoma 42 Theodoropoulos VE, Lazaris A, Sofras F, Gerzelis I, Tsoukala cell invasion and metastasis. Clin Cancer Res 13: 1831-1838, V, Ghikonti I, Manikas K and Kastriotis I: Hypoxia-inducible 2007. factor 1 alpha expression correlates with angiogenesis and 57 Xie Q, Gao CF, Shinomiya N, Sausville E, Hay R, Gustafson M, unfavorable prognosis in bladder cancer. Eur Urol 46: 200-208, Shen Y, Wenkert D and Vande Woude GF: Geldanamycins 2004. exquisitely inhibit HGF/SF-mediated tumor cell invasion. 43 Koga F, Tsutsumi S and Neckers LM: Low dose geldanamycin Oncogene 24: 3697-3707, 2005. inhibits hepatocyte growth factor and hypoxia-stimulated 58 Zagzag D, Nomura M, Friedlander DR, Blanco CY, Gagner JP, invasion of cancer cells. Cell Cycle 6: 1393-1402, 2007. Nomura N and Newcomb EW: Geldanamycin inhibits migration 44 Swinson DE and O’Byrne KJ: Interactions between hypoxia and of glioma cells in vitro: a potential role for hypoxia-inducible epidermal growth factor receptor in non-small-cell lung cancer. factor (HIF-1alpha) in glioma cell invasion. J Cell Physiol 196: Clin Lung Cancer 7: 250-256, 2006. 394-402, 2003. 45 Nakamura K, Martin KC, Jackson JK, Beppu K, Woo CW and 59 Sakagami M, Morrison P and Welch WJ: Benzoquinoid Thiele CJ: Brain-derived neurotrophic factor activation of TrkB ansamycins (herbimycin A and geldanamycin) interfere with the induces vascular endothelial growth factor expression via maturation of growth factor receptor tyrosine kinases. Cell Stress hypoxia-inducible factor-1alpha in neuroblastoma cells. Cancer Chaperones 4: 19-28, 1999. Res 66: 4249-4255, 2006. 60 Lang SA, Moser C, Gaumann A, Klein D, Glockzin G, Popp FC, 46 Tacchini L, Dansi P, Matteucci E and Desiderio MA: Hepatocyte Dahlke MH, Piso P, Schlitt HJ, Geissler EK and Stoeltzing O: growth factor signalling stimulates hypoxia inducible factor-1 Targeting heat-shock protein 90 in pancreatic cancer impairs (HIF-1) activity in HepG2 hepatoma cells. 22: insulin-like growth factor-I receptor signaling, disrupts an 1363-1371, 2001. interleukin-6/signal-transducer and activator of transcription 47 Pennacchietti S, Michieli P, Galluzzo M, Mazzone M, Giordano 3/hypoxia-inducible factor-1alpha autocrine loop, and reduces S and Comoglio PM: Hypoxia promotes invasive growth by orthotopic tumor growth. Clin Cancer Res 13: 6459-6468, 2007. transcriptional activation of the met protooncogene. Cancer Cell 61 Xu W, Mimnaugh E, Rosser MF, Nicchitta C, Marcu M, Yarden 3: 347-361, 2003. Y and Neckers L: Sensitivity of mature Erbb2 to geldanamycin 48 Alqawi O, Moghaddas M and Singh G: Effects of geldanamycin is conferred by its kinase domain and is mediated by the on HIF-1alpha-mediated angiogenesis and invasion in prostate chaperone protein Hsp90. J Biol Chem 276: 3702-3708, 2001. cancer cells. Prostate Cancer Prostatic Dis 9: 126-135, 2006. 62 Shimamura T, Lowell AM, Engelman JA and Shapiro GI: 49 Benezra R, Henke E, Ciarrocchi A, Ruzinova M, Solit D, Rosen Epidermal growth factor receptors harboring kinase domain N, Nolan D, Mittal V and de Candia P: Induction of complete mutations associate with the heat-shock protein 90 chaperone regressions of oncogene-induced breast tumors in mice. Cold and are destabilized following exposure to geldanamycins. Spring Harb Symp Quant Biol 70: 375-381, 2005. Cancer Res 65: 6401-6408, 2005. 50 de Candia P, Solit DB, Giri D, Brogi E, Siegel PM, Olshen AB, 63 Fumo G, Akin C, Metcalfe DD and Neckers L: 17-Allylamino- Muller WJ, Rosen N and Benezra R: Angiogenesis impairment 17-demethoxygeldanamycin (17-AAG) is effective in down- in Id-deficient mice cooperates with an Hsp90 inhibitor to regulating mutated, constitutively activated KIT protein in completely suppress HER2/neu-dependent breast tumors. Proc human mast cells. Blood 103: 1078-1084, 2004. Natl Acad Sci USA 100: 12337-12342, 2003. 64 George P, Bali P, Cohen P, Tao J, Guo F, Sigua C, Vishvanath A, 51 Mabjeesh NJ, Post DE, Willard MT, Kaur B, Van Meir EG, Fiskus W, Scuto A, Annavarapu S, Moscinski L and Bhalla K: Simons JW and Zhong H: Geldanamycin induces degradation of Cotreatment with 17-allylamino-demethoxygeldanamycin and hypoxia-inducible factor 1alpha protein via the proteosome FLT-3 kinase inhibitor PKC412 is highly effective against human pathway in prostate cancer cells. Cancer Res 62: 2478-2482, acute myelogenous leukemia cells with mutant FLT-3. Cancer 2002. Res 64: 3645-3652, 2004. 52 Maulik G, Kijima T, Ma PC, Ghosh SK, Lin J, Shapiro GI, 65 Citri A, Harari D, Shohat G, Ramakrishnan P, Gan J, Lavi S, Schaefer E, Tibaldi E, Johnson BE and Salgia R: Modulation of Eisenstein M, Kimchi A, Wallach D, Pietrokovski S and Yarden the c-Met/hepatocyte growth factor pathway in small cell lung Y: Hsp90 recognizes a common surface on client kinases. J Biol cancer. Clin Cancer Res 8: 620-627, 2002. Chem 281: 14361-14369, 2006. 53 Nguyen DM, Desai S, Chen A, Weiser TS and Schrump DS: 66 Koga F, Xu W, Karpova TS, McNally JG, Baron R and Neckers Modulation of metastasis phenotypes of non-small cell lung L: Hsp90 inhibition transiently activates Src kinase and promotes cancer cells by 17-allylamino 17-demethoxy geldanamycin. Ann Src-dependent Akt and Erk activation. Proc Natl Acad Sci USA Thorac Surg 70: 1853-1860, 2000. 103: 11318-11322, 2006.

805 ANTICANCER RESEARCH 29: 797-808 (2009)

67 Xu W, Yuan X, Xiang Z, Mimnaugh E, Marcu M and Neckers 81 Lai JF, Kao SC, Jiang ST, Tang MJ, Chan PC and Chen HC: L: Surface charge and hydrophobicity determine ErbB2 binding Involvement of focal adhesion kinase in hepatocyte growth to the Hsp90 chaperone complex. Nat Struct Mol Biol 12: 120- factor-induced scatter of Madin-Darby canine kidney cells. J 126, 2005. Biol Chem 275: 7474-7480, 2000. 68 Xu Y, Singer MA and Lindquist S: Maturation of the tyrosine 82 Burridge K and Wennerberg K: Rho and Rac take center stage. kinase c-src as a kinase and as a substrate depends on the Cell 116: 167-179, 2004. molecular chaperone Hsp90. Proc Natl Acad Sci USA 96: 109- 83 Nobes CD and Hall A: Rho, rac, and cdc42 GTPases regulate the 114, 1999. assembly of multimolecular focal complexes associated with actin 69 Schulte TW, Blagosklonny MV, Ingui C and Neckers L: stress fibers, lamellipodia, and filopodia. Cell 81: 53-62, 1995. Disruption of the Raf-1-Hsp90 molecular complex results in 84 Rousseau S, Houle F, Kotanides H, Witte L, Waltenberger J, destabilization of Raf-1 and loss of Raf-1-Ras association. J Biol Landry J and Huot J: Vascular endothelial growth factor (VEGF)- Chem 270: 24585-24588, 1995. driven actin-based motility is mediated by VEGFR2 and requires 70 Sato S, Fujita N and Tsuruo T: Modulation of Akt kinase activity concerted activation of stress-activated protein kinase 2 by binding to Hsp90. Proc Natl Acad Sci USA 97: 10832-10837, (SAPK2/p38) and geldanamycin-sensitive phosphorylation of 2000. focal adhesion kinase. J Biol Chem 275: 10661-10672, 2000. 71 Lang SA, Klein D, Moser C, Gaumann A, Glockzin G, Dahlke 85 Becker B, Multhoff G, Farkas B, Wild PJ, Landthaler M, Stolz MH, Dietmaier W, Bolder U, Schlitt HJ, Geissler EK and W and Vogt T: Induction of Hsp90 protein expression in Stoeltzing O: Inhibition of heat-shock protein 90 impairs malignant melanomas and melanoma metastases. Exp Dermatol epidermal growth factor-mediated signaling in gastric cancer 13: 27-32, 2004. cells and reduces tumor growth and vascularization in vivo. Mol 86 Sidera K, Samiotaki M, Yfanti E, Panayotou G and Patsavoudi Cancer Ther 6: 1123-1132, 2007. E: Involvement of cell surface HSP90 in cell migration reveals a 72 Eustace BK, Sakurai T, Stewart JK, Yimlamai D, Unger C, novel role in the developing nervous system. J Biol Chem 279: Zehetmeier C, Lain B, Torella C, Henning SW, Beste G, 45379-45388, 2004. Scroggins BT, Neckers L, Ilag LL and Jay DG: Functional 87 Tsutsumi S, Scroggins B, Koga F, Lee MJ, Trepel J, Felts S, proteomic screens reveal an essential extracellular role for hsp90 Carreras C and Neckers L: A small molecule cell-impermeant alpha in cancer cell invasiveness. Nat Cell Biol 6: 507-514, Hsp90 antagonist inhibits tumor cell motility and invasion. 2004. Oncogene 27: 2478-2487, 2008. 73 Le Boeuf F, Houle F and Huot J: Regulation of vascular 88 Sidera K, Gaitanou M, Stellas D, Matsas R and Patsavoudi E: A endothelial growth factor receptor 2-mediated phosphorylation Critical role for HSP90 in cancer cell invasion involves of focal adhesion kinase by heat-shock protein 90 and Src kinase interaction with the extracellular domain of HER-2. J Biol Chem activities. J Biol Chem 279: 39175-39185, 2004. 283: 2031-2041, 2008. 74 Gradin K, McGuire J, Wenger RH, Kvietikova I, Whitelaw ML, 89 Bagatell R, Gore L, Egorin MJ, Ho R, Heller G, Boucher N, Toftgard R, Tora L, Gassmann M and Poellinger L: Functional Zuhowski EG, Whitlock JA, Hunger SP, Narendran A, Katzenstein interference between hypoxia and dioxin signal transduction HM, Arceci RJ, Boklan J, Herzog CE, Whitesell L, Ivy SP and pathways: competition for recruitment of the Arnt transcription Trippett TM: Phase I pharmacokinetic and pharmacodynamic factor. Mol Cell Biol 16: 5221-5231, 1996. study of 17-N-allylamino-17-demethoxygeldanamycin in pediatric 75 Isaacs JS, Jung YJ, Mimnaugh EG, Martinez A, Cuttitta F and patients with recurrent or refractory solid tumors: a pediatric Neckers LM: Hsp90 regulates a von Hippel Lindau-independent oncology experimental therapeutics investigators consortium hypoxia-inducible factor-1 alpha-degradative pathway. J Biol study. Clin Cancer Res 13: 1783-1788, 2007. Chem 277: 29936-29944, 2002. 90 Goetz MP, Toft D, Reid J, Ames M, Stensgard B, Safgren S, 76 Hur E, Kim HH, Choi SM, Kim JH, Yim S, Kwon HJ, Choi Y, Adjei AA, Sloan J, Atherton P, Vasile V, Salazaar S, Adjei A, Kim DK, Lee MO and Park H: Reduction of hypoxia-induced Croghan G and Erlichman C: Phase I trial of 17-allylamino-17- transcription through the repression of hypoxia-inducible factor-1 demethoxygeldanamycin in patients with advanced cancer. J Clin alpha/aryl hydrocarbon receptor nuclear translocator DNA binding Oncol 23: 1078-1087, 2005. by the 90-kDa heat-shock protein inhibitor radicicol. Mol 91 Grem JL, Morrison G, Guo XD, Agnew E, Takimoto CH, Pharmacol 62: 975-982, 2002. Thomas R, Szabo E, Grochow L, Grollman F, Hamilton JM, 77 Minet E, Mottet D, Michel G, Roland I, Raes M, Remacle J and Neckers L and Wilson RH: Phase I and pharmacologic study of Michiels C: Hypoxia-induced activation of HIF-1: role of HIF-1 17-(allylamino)-17-demethoxygeldanamycin in adult patients alpha-Hsp90 interaction. FEBS Lett 460: 251-256, 1999. with solid tumors. J Clin Oncol 23: 1885-1893, 2005. 78 Tacchini L, Matteucci E, De Ponti C and Desiderio MA: 92 Ramanathan RK, Egorin MJ, Eiseman JL, Ramalingam S, Hepatocyte growth factor signaling regulates transactivation of Friedland D, Agarwala SS, Ivy SP, Potter DM, Chatta G, genes belonging to the plasminogen activation system via Zuhowski EG, Stoller RG, Naret C, Guo J and Belani CP: Phase hypoxia inducible factor-1. Exp Cell Res 290: 391-401, 2003. I and pharmacodynamic study of 17-(allylamino)-17- 79 Krishnamachary B, Berg-Dixon S, Kelly B, Agani F, Feldser D, demethoxygeldanamycin in adult patients with refractory Ferreira G, Iyer N, LaRusch J, Pak B, Taghavi P and Semenza advanced cancers. Clin Cancer Res 13: 1769-1774, 2007. GL: Regulation of colon carcinoma cell invasion by hypoxia- 93 Solit DB, Ivy SP, Kopil C, Sikorski R, Morris MJ, Slovin SF, inducible factor 1. Cancer Res 63: 1138-1143, 2003. Kelly WK, DeLaCruz A, Curley T, Heller G, Larson S, Schwartz 80 Mitra SK, Hanson DA and Schlaepfer DD: Focal adhesion L, Egorin MJ, Rosen N and Scher HI: Phase I trial of kinase: in command and control of cell motility. Nat Rev Mol 17-allylamino-17-demethoxygeldanamycin in patients with Cell Biol 6: 56-68, 2005. advanced cancer. Clin Cancer Res 13: 1775-1782, 2007.

806 Koga et al: Inhibition of Invasion and Metastasis by Targeting Hsp90 (Review)

94 Weigel BJ, Blaney SM, Reid JM, Safgren SL, Bagatell R, 97 Ficker E, Dennis AT, Wang L and Brown AM: Role of the Kersey J, Neglia JP, Ivy SP, Ingle AM, Whitesell L, Gilbertson cytosolic chaperones Hsp70 and Hsp90 in maturation of the RJ, Krailo M, Ames M and Adamson PC: A phase I study of cardiac potassium channel HERG. Circ Res 92: e87-100, 2003. 17-allylaminogeldanamycin in relapsed/refractory pediatric 98 Gabrielson K, Bedja D, Pin S, Tsao A, Gama L, Yuan B and patients with solid tumors: a Children’s Oncology Group study. Muratore N: Heat-shock protein 90 and ErbB2 in the cardiac Clin Cancer Res 13: 1789-1793, 2007. response to doxorubicin injury. Cancer Res 67: 1436-1441, 95 Price JT, Quinn JM, Sims NA, Vieusseux J, Waldeck K, 2007. Docherty SE, Myers D, Nakamura A, Waltham MC, Gillespie MT and Thompson EW: The 90 inhibitor, 17-allylamino-17-demethoxygeldanamycin, enhances osteoclast formation and potentiates bone metastasis of a human breast cancer cell line. Cancer Res 65: 4929-4938, 2005. 96 Xu W, Yuan X, Beebe K, Xiang Z and Neckers L: Loss of Hsp90 Received May 28, 2008 association up-regulates Src-dependent ErbB2 activity. Mol Cell Revised August 6, 2008 Biol 27: 220-228, 2007. Accepted September 16, 2008

807