Geldanamycin and Its 17-Allylamino-17-Demethoxy

[CANCER RESEARCH 63, 3241–3246, June 15, 2003] Geldanamycin and its 17-Allylamino-17-Demethoxy Analogue Antagonize the Action of Cisplatin in Human Colon Adenocarcinoma Cells: Differential Caspase Activation as a Basis for Interaction1

Irina A. Vasilevskaya,2 Tatiana V. Rakitina, and Peter J. O’Dwyer University of Pennsylvania Cancer Center, Philadelphia, Pennsylvania 19104

ABSTRACT mal degradation (10, 11). The ability of GA to deplete Raf1, ErbB2, and mutant p53 in breast cancer cells was found to correlate with its Several chaperone-binding drugs based on geldanamycin (GA) have antiproliferative activity (12), making it a plausible candidate for use been synthesized, and one of them, 17-allylamino-17-demethoxygeldanamycin (17-AAG), is being developed in the clinic. Interest in the use of in cancer treatment. However, GA caused liver toxicity in preclinical 17-AAG in combination with cytotoxic drugs led us to study both GA and studies (13). A search for more tolerable derivatives has yielded 17-AAG with cisplatin (DDP) in the human colon adenocarcinoma cell 17-AAG, a less toxic analogue that retains the tumoricidal features of lines HT29 and HCT116. We performed isobologram analysis of combi- GA. Like its parent compound, 17-AAG inhibits several signaling nations of DDP with GA or 17-AAG in these cell lines using the standard pathways through binding to Hsp 90, which results in destabilization 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay to of signaling complexes and degradation of its client proteins by the evaluate cell survival. In HCT116, the effects of GA and 17-AAG with proteasome in a variety of cell lines (14–16). Treatment with 17-AAG DDP were additive and schedule dependent. In HT29 both GA and has been shown to inhibit tumor growth and induce apoptosis in colon 17-AAG antagonized DDP effects resulting in cytotoxicity less than ex- pected. We hypothesized that the antagonism in HT29 cells might be a cancer, glioblastoma, and breast cancer cell lines (16–18). These consequence of altered p53 function in this cell line. Accordingly, we tested studies led to the active clinical development of 17-AAG as a poten- GA/17-AAG and DDP in combination in the HCTp5.2 cell line, which tial anticancer drug (19, 20). expresses a dominant-negative form of p53. In these cells too, the GA Recent work suggests that 17-AAG enhances the cytotoxic ef- analogues antagonized DDP, suggesting a role for p53 in the observed fects of paclitaxel in non-small cell lung cancer cell lines and effects. Investigation of the DDP-induced signaling pathways revealed that xenografts (21) and of Taxol and doxorubicin in breast cancer cell ansamycins block the activation of mitogen-activated protein kinase and lines (22), making it a promising candidate for combination treat- c-Jun NH2-terminal kinase pathways and c-Jun expression in HT29 cells while exerting incomplete inhibitory effects in HCT116 and HCTp5.2 cell ment of solid tumors. Doxorubicin exerts its cytotoxicity by caus- lines. Therefore, effects on signaling are thought not to underlay the ing DNA strand breaks as a consequence of topoisomerase II antagonism in the latter model. The ansamycins inhibited DDP-induced inhibition (23). Enhancement of its action by ansamycins prompted activation of caspases 8 and 3 in HT29 and HCTp5.2 but not in HCT116 us to evaluate the effects of GA and 17-AAG in colon tumor cell cells, which we postulate to be the basis for higher survival of p53-deficient lines when used in combination with another DNA-damaging drug, cells when treated with combinations of the two drugs. DDP [cis-diamminedichloroplatinum (II)], which exhibits a broad spectrum of activity against several tumor types and is routinely INTRODUCTION used for the treatment of solid tumors (24, 25). DDP treatment leads to formation of DNA platinum adducts, DNA damage, cell Specific inhibitors targeted to a variety of signal transduction cycle arrest, and cell death, which occurs primarily by apoptosis, pathways are in clinical development as potential anticancer agents (1, although under certain circumstances DDP treatment can cause 2). Several have activity as single agents, but studies in animal models death by necrosis (26, 27). DDP treatment activates caspase cas- suggest that most will be considerably more active in combination cades regulated both by cellular receptors such as Fas/CD95 (28) with chemotherapy (3). Among the inhibitors in clinical development and mitochondrial release of cytochrome c (29). Cell lines express- 3 are agents targeted to the molecular chaperone Hsp 90 (4, 5). ing either wild-type or mutated p53 protein were shown to dem- The benzoquinone ansamycins, herbimycin A and GA, were first onstrate growth arrest and were able to undergo apoptosis after described as inhibitors of tyrosine kinases and were shown to reverse DDP exposure (30, 31). DDP also activates signaling pathways cell transformation by oncogenic kinases such as Src, Abl, and ErbB resulting in the induction of MAPKs. The roles of MAPK, JNK, (6). Later, it was shown that ansamycins do not affect kinases directly and p38 kinase in mediating DDP effects are under active inves- but instead serve as inhibitors of Hsp 90 (7), a chaperoning protein tigation, and there are numerous conflicting reports concerning responsible for proper protein folding and an important participant in their input in both enhancement and attenuation of DDP cytotox- a variety of cellular processes (8). GA acts by blocking the binding of icity (32–36). These mechanisms may also be critical to an inter- ATP to Hsp 90 (9), which leads to destabilization of Hsp 90 com- action with ansamycins because we have previously shown that plexes with its client proteins rendering them available for proteaso- GA was able to inhibit signaling pathways using Hsp 90 and its analogues as chaperones (37). Received 11/11/02; accepted 4/15/03. The costs of publication of this article were defrayed in part by the payment of page Our results show that when added simultaneously, DDP and GA/ charges. This article must therefore be hereby marked advertisement in accordance with 17-AAG have an additive effect in HCT116 colon cancer cell line but 18 U.S.C. Section 1734 solely to indicate this fact. are antagonistic in HT29 cells. Our initial hypothesis that these 1 Supported in part by NIH Grant CA 49820. 2 To whom requests for reprints should be addressed, at University of Pennsylvania, phenotypes reflect altered p53 status in these cells was supported by 1020 BRB II/III, 421 Curie Boulevard, Philadelphia, PA 19104. Phone: (215) 573-7300; antagonism of DDP and ansamycins in the HCT-116-derived cell line, Fax: (215) 573-7049; E-mail: [email protected]. 3 HCTp5.2, expressing dominant-negative form of p53. Additional The abbreviations used are: hsp90, Mr 90,000 heat shock protein; GA, geldanamycin; 17-AAG, 17-allylamino-17-demethoxygeldanamycin; DDP, cisplatin [cis-diamminedi- dissection on the basis of observed antagonism indicated that it results chloroplatinum (II)]; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bro- from inhibition of DDP-induced Fas-mediated apoptosis in this mide; MAPK, mitogen-activated protein kinase; JNK, c-Jun NH2-terminal kinase; ERK, extracellular signal-regulated kinase (same as MAPK); CI, combinatorial index. cell line. 3241

MATERIALS AND METHODS DNA Ladder Analysis. Cells were treated as described above. After 20 h, DNA was isolated using Apoptotic DNA-Ladder Kit (Roche Diagnostics Cell Lines and Reagents. Colon cancer cell lines HT29 (functionally Corp.) according to manufacturer recommendations. Agarose gel electrophore- p53-deficient, mismatch repair proficient) and HCT116 (opposite characteris- sis of 10 ␮g of DNA/lane was performed followed by visualization of DNA tics) were purchased from American Type Culture Collection (Manassas, VA). with ethidium bromide. Cells were cultivated in Eagle’s MEM supplemented with 10% fetal bovine serum, penicillin (100 units/ml), streptomycin (100 units/ml), and fungizone RESULTS (Life Technologies, Inc., Grand Island, NY). The HCTp5.2 cell line was generated by transfection with pcDNA3 plasmid expressing a dominant- DDP and GA/17-AAG Exhibit Additive Effects in HCT116 and negative form of p53 (gift from Dr. Wafik El-Deiry, University of Pennsyl- Antagonize Each other in HT29 Cells. Treatment of colon cancer vania, Philadelphia, PA) followed by isolation of stable transfectant. Cultures cell lines with DDP, GA, and its analogue, 17-AAG, results in were maintained in a humidified incubator at 37°C in 5%CO2-95% air. GA was purchased from Life Technologies, Inc., and 17-AAG was kindly provided dose-dependent cytotoxicity. Fig. 1 demonstrates survival curves of by Dr. Edward Sausville (DTP; National Cancer Institute, Bethesda, MD). HT29 and HCT116 cell lines treated with DDP, GA, and 17-AAG ␮ Stock solutions of ansamycins (1 mM) were prepared in DMSO, aliquoted, and with IC50s ranging from 8.8 to 4.5 M for DDP, 6.8 to 12 nM for GA, stored at Ϫ20°C. DDP (Sigma Chemical Co., St. Louis, MO) was dissolved in and from 18 to 50 nM for 17-AAG, respectively. To evaluate the sterile PBS at a final concentration of 1 mg/ml (3.33 mM). impact of the combined treatment with DDP and GA/17-AAG on MTT Assay and Isobologram Analysis. HT29, HCT116, and HCTp5.2 human colon adenocarcinoma cell lines, HT29 and HCT116 cells cells were plated in 96-well plates at a density of 2000 cells/well in 140 ␮lof were treated with various concentrations of drugs alone and in com- medium 20–24 h before addition of drugs. DDP was then added in concen- bination followed by survival assessment (MTT assay) and isobolo- ␮ trations of 0, 1, 3, 6, 8, and 10 M from left to the right side of the plate (over gram analysis. We used the method described by Tsai et al. (39) to two plates, each concentration of DDP in quadruplicate). GA or 17-AAG was create isobolograms at the IC50 level: IC50 unit values of GA or added cross-wise in concentrations of 0–50 and 0–75 nM, respectively. Cyto- 17-AAG Ͻ 1 were plotted against corresponding IC unit values of toxicity was measured using the standard MTT assay after a 72-h exposure to 50 drugs (38). In sequential experiments, cells were treated totally for 96 h, with DDP. The distribution of dots along the line connecting values of 1 the second compound in combination introduced with a 24-h delay. Results constitutes an additive effect of two drugs while scattering below or

were quantified using ELx800 Universal Microplate Reader (Bio-Tek Instru- above represents synergism and antagonism, respectively. As Fig. 2 ments, Inc.) at 595 nm wavelength, and control absorbance was designated as demonstrates, the effect of GA/17-AAG and DDP on cell survival is 100%, with percentage of it as a cell survival. MTT results were analyzed as described by Tsai et al. (39). Briefly, after each experiment, survival curves were generated, two for DDP and GA or 17-AAG alone and 13 for the 42 pairs

of drug combinations. The IC50s for each drug in combination were deter-

mined, and IC50 units were derived as ratio of IC50 for DDP and GA/17-AAG

in this particular drug combination relative to IC50 of drug alone (this value

was designated as 1) for each cell line. Isobolograms were generated at IC50 levels. Each plot represents values generated in at least three independent experiments for simultaneous treatment of cells and two experiments for sequential drug addition. The method of Chou et al. (40) was also used to determine CIs for drugs. Cell Treatment, Protein Extracts Preparation, and Western Blotting. Cells were treated for 24 h with 0.01% DMSO (control), 500 nM GA or 17-AAG alone, 30 ␮M DDP alone, and with combinations of DDP with GA or 17-AAG. Protein extracts were prepared as described previously (37). Extracts were diluted with lysis buffer to obtain equal protein concentration, aliquoted, and stored at Ϫ70°C. Samples for electrophoresis were prepared, separated in discontinuous SDS gels, and transferred onto Immobilon-P membranes (Mil- lipore, Bedford, MA). Western blotting was carried out according to the manufacturer recommendations, using horseradish peroxidase-conjugated sec- ondary antibodies (Santa Cruz Biotechnology, Santa Cruz, CA) and the enhanced chemiluminescence plus detection system (Amersham, Arlington Heights, IL). The following antibodies were used: mouse monoclonal antibodies against p21, p53, phospho-ERK, phospho-c-Jun, and phospho-activating transcription factor-2 and rabbit polyclonal antibodies raised against ERK, JNK, and p38 from Santa Cruz Biotechnology; rabbit polyclonal antibodies against phosphorylated forms of p53, JNK, and p38 from Cell Signaling Technology (Beverly, MA); and goat polyclonal antibodies raised against ␤-actin (Santa Cruz Biotechnology). Mouse monoclonal antibodies against caspase 8 were purchased from Oncogene Research Products (Boston, MA). Caspase Activity Assay. ApoTarget caspase colorimetric protease assay kits were purchased from BioSource International, Inc. (Camarillo, CA). Cells were plated in 10-cm Petri dishes and subjected to drug treatment as described above. After 20 h, cellular extracts were isolated, and assays were carried out according to manufacturer recommendations. Synthetic peptides conjugated with chromophore (p-nitroaniline) served as substrates for measuring the activity of caspases, which subsequently was quantified by measuring the absorbance of light by free chromophore using microtiter plate reader at 405 Fig. 1. Cytotoxic effects of ansamycins and DDP in colon cancer cell lines. Survival ᭜ f Œ nm. Comparison of the absorbance of p-nitroaniline from apoptotic sample curves for HT29 ( ), HCT116 ( ), and HCTp5.2 ( ) cell lines after treatment with DDP (A), GA (B), and 17-AAG (C) are based on the results of MTT assays after continuous with an untreated control allowed determination of the fold increase in caspase exposure to drugs for 72 h. Average values from at least three independent experiments activity. All experiments were performed at least twice. are shown, bars represent SD. 3242

treatment causing a significant increase in phospho-p53 levels, but the functional p53 deficiency resulted in blockade or strong inhibition of DDP-induced p21 expression. The cytotoxic effects of ansamycins in HCTp5.2 are dose depend-

ent, with IC50s for GA and 17-AAG (20 and 48 nM) close to those in HCT116 (Fig. 1), suggesting that p53 function is not a key determinant of the cellular responses to ansamycins in HCT116. DDP sensitivity in these cells, however, was 2-fold lower than that of the ␮ parental line (with IC50 of 8 M), reaffirming the importance of intact p53 for DDP cytotoxicity in this model. Isobologram analysis revealed partial recapitulation of antagonism between ansamycins and DDP in HCTp5.2 cell line (Fig. 3B), with CIs higher than these in HCT116 cells (Table 1). Ansamycins Exert Distinct Effects on DDP Signaling through MAPK Pathways in HT29 and HCT116 Cells. DDP has been shown to induce MAPK-signaling pathways in various cell lines, whereas ansamycins, on the other hand, are known inhibitors of signaling. We investigated the effect of GA and 17-AAG on DDP- induced signaling through MAPKs in HT29, HCT116, and HCTp5.2 Fig. 2. Combinations of ansamycins and DDP are additive in HCT116 and antagonistic cell lines. Fig. 4 demonstrates that in HT29 cells, ansamycins block in HT29 cell lines. Shown are the isobolograms at IC50 based on the results of MTT assays for HT29 and HCT116 cell lines treated with combinations of GA/17-AAG and DDP DDP-induced activation of ERK and JNK, causing profound inhibi- added simultaneously. Each diagram combines the results of at least three independent tion of phosphorylation of activating transcription factor-2 and c-Jun experiments. and blockade of c-Jun synthesis. In HCT116 and HCTp5.2, the effects

additive in HCT116 cells, whereas in the HT29 cell line, these drugs

exhibit antagonism. In addition, CIs at the IC50 of DDP and of GA/17-AAG were calculated by the method of Chou et al. (40) for all experiments performed. The CI values Ͻ 0.9 or Ͼ 1.1 reflect syner- gistic or antagonistic interaction of drugs, respectively. Results of all experiments performed are summarized in Table 1. Because the importance of the schedule of drug administration was demonstrated for the combination of Taxol and 17-AAG (22) in breast cancer cell lines, we examined if sequential addition of DDP and ansamycins to HCT116 and HT29 cells would change the outcome observed with simultaneous treatment. Experiments with the second drug in combination added with a 24-h delay were carried out, and the results analyzed as described above. The effects of GA and 17-AAG antagonized those of DDP in both cell lines irrespective of the order in which drugs were added (Table 1). Stable Expression of Dominant-negative p53 in HCT116 Recapitulates the Antagonistic Phenotype of Combinations of DDP with Ansamycins. The importance of p53 in cellular responses to DNA damage is well established (41). Because HT29 and HCT116 differ in p53 status, we hypothesized that loss of p53 function in HT29 could be the basis of the observed antagonism between DDP and ansamycins. Using HCT116 as the parental line, we generated HCTp5.2, a cell line stably expressing a dominant-negative form of 143 p53 in which Val was substituted to Ala, resulting in elimination of Fig. 3. Expression of mutant p53 recapitulates the antagonistic phenotype in HCTp5.2 sequence-specific transcriptional activity. Fig. 3A shows differences line after combined treatment with ansamycins and DDP. A, status of p53 pathways in colon cancer cell lines studied. Cells were treated for 24 h with 30 ␮M DDP and 500 nM in induction of the p53 pathway in HT29, HCT116, and HCTp5.2 as GA or 17-AAG alone and in combination, and protein extracts were isolated as described a response to treatment with DDP, GA, and 17-AAG alone or in in “Materials and Methods” and subjected to electrophoresis followed by Western blotting combination. In the presence of DDP, HCT116 cells demonstrated analysis. Western blotting with ␤-actin antibodies was performed to monitor protein content of each sample. Picture shown is representative for at least two independent induction of p53 with corresponding elevation of p21 protein. In experiments. B, isobolograms at IC50 for HCTp5.2 cell line were generated similarly to HT29 and HCTp5.2 cells, p53 protein was up-regulated with DDP those for HT29 and HCT116 cells.

Table 1 Effect of combination of GA/17-AAG and cisplatin in colon cancer cell linesa DDP ϩ DDP ϩ GA3 17-AAG3 DDP3 DDP3 Cell line GA 17-AAG DDP DDP GA 17-AAG HT29 1.382 Ϯ 0.22 1.19 Ϯ 0.09 1.24 Ϯ 0.22 1.25 Ϯ 0.1 1.14 Ϯ 0.1 1.24 Ϯ 0.09 HCT116 0.984 Ϯ 0.02 1.02 Ϯ 0.07 1.18 Ϯ 0.14 1.26 Ϯ 0.19 1.12 Ϯ 0.06 1.14 Ϯ 0.09 HCTp5.2 1.217 Ϯ 0.06 1.09 Ϯ 0.06 N/Ab N/A N/A N/A a Combinatorial indices at IC50 for colon cancer cell lines treated with DDP and GA/17-AAG are shown; values represent results from at least three independent experiments. b N/A, not available. 3243

same dynamic was observed for caspase 3 (Fig. 6B). These results confirmed our conclusion that the loss of functional p53 protein in HCTp5.2 cell line was associated with inhibition of DDP-induced apoptosis in the presence of ansamycins.

DISCUSSION The aim of this study was to investigate the effects of combined treatment of colon cancer cell lines with GA/17-AAG and DDP. Because among the known effects of ansamycins are inhibition of signaling through MAPK pathways, we also wished to determine its role in the interaction of these Hsp 90 inhibitors and DDP, the cytotoxicity of which depends upon DNA damage-induced signaling (42, 43). We observed additivity between ansamycins and DDP in HCT116 colon cancer cells but were surprised that this combination was antagonistic in HT29 cells. The demonstration of antagonism led to an investigation of its cause. We show that in HT29 cell line activation of JNK associated with DDP-induced apoptosis was blocked by ansamycins, leading to com- plete inhibition of c-Jun expression. The depletion of this major component of the activator protein 1 transcription factor makes the effect of ansamycins on DDP signaling in HT29 cells even more profound, expanding it to a multitude of not only JNK targets but also to activator protein 1-regulated genes, known to participate in cell growth, proliferation, and death pathways (44). Because p53 function is absent in this cell line, there is a possibility that signaling inhibition Fig. 4. Effects of ansamycins on DDP-induced signaling through MAPKs in HT29, HCT116, and HCTp5.2 cell lines are cell specific. Cells were treated as described in a may contribute to the antagonism but only through p53-independent legend for Fig. 3A; protein extracts were isolated and subjected to electrophoresis mechanisms, which we are currently investigating. In the HCT116 followed by Western blotting analysis. Picture shown is representative of at least two line, MAPK signaling was less affected by ansamycins, which to- independent experiments. gether with the presence of functional p53 could contribute to additive effects of drugs in this cell line; this assumption was confirmed by of ansamycins on signaling pathways are less pronounced, resulting in occurrence of the antagonistic phenotype in the HCT116-derived partial inhibition, the degree of which seems to be independent of p53 HCTp5.2 cells, where it was accompanied by the attenuation of function. Thus, the antagonism of DDP and ansamycins in HCTp5.2 cannot be attributed to altered DDP signaling through MAPK pathways in this cell line. Higher Survival of HCTp5.2 Cells Treated with DDP Is because of Abrogated Apoptosis. One of the reasons for the greater survival of HCTp5.2 treated with DDP and ansamycins could be the abrogation of DDP-induced p53-dependent apoptosis in this cell line. The apoptotic DNA ladder analysis (Fig. 5A) confirmed that assumption. In addition, it was supported by the observations of cell morphology after treatment with high concentrations of DDP and 17-AAG alone and in combination for 24 h, whereas HCTp5.2 cells did not exhibit significant visible changes, HCT116 cells demonstrated such charac- teristic features of cell death as cell shrinkage, membrane blebbing, and formation of apoptotic bodies. We were unable, however, to detect any significant difference in contents neither of several Bcl-2 family members, nor AKT in HCT116 and HCTp5.2 cells (data not shown). However, in accordance with the loss of p53 transcriptional activation function in HCTp5.2, induction of Fas by DDP was greatly reduced (Fig. 5B) in this cell line. The increase of Fas expression in DDP-treated HCT116 was not affected by the presence of ansamycins and was accompanied by higher degree of procaspase 8 processing than that in HCTp5.2 cells (Fig. 5B). Taken together, these data initially localized the effect of the drug combination in HCTp5.2 cells to the Fas-mediated apoptotic pathway. It was confirmed by measuring the activation of caspases in HT29, HCT116, and HCTp5.2 cells lines after combined treatment with DDP and 17-AAG. The results Fig. 5. Higher survival of HCTp5.2 cells is because of abrogated DDP-induced apoptosis. HCT116 and HCTp5.2 cells were treated as described in Fig. 2A and subjected revealed marked differences in the activation of caspases: the pres- to DNA fragmentation assay (A) or Western blotting analysis of the induction of Fas and ence of 17-AAG inhibited DDP-induced activation of caspase 8 in processing of procaspase 8 (B) in response to drug treatment. For DNA ladders, DNA was isolated as described in “Materials and Methods.” Western blotting analysis was per- HT29 cells and, to a lesser degree, in the HCTp5.2 cell line, whereas formed as described above; picture shown is representative of at least two independent the HCT116 cell line demonstrated enhanced activation (Fig. 6A). The experiments. 3244

of both GA and 17-AAG (Յ2–5nM, respectively) as compared with control samples in MTT assays. The p53 protein mediates cell cycle arrest and/or apoptosis in response to genotoxic stress based mostly on its function as a transcription factor (52) and loss of this function can dramatically change the outcome of treatment with DNA-damaging agents, depending on the nature of the p53 mutation (53). The HCTp5.2 cells express a dominant-negative form of p53 in which the substitution Val1433Ala renders it transcriptionally inactive. This particular mutant belongs to the group of so-called discriminatory p53 mutants, which have been shown to cause induction of reporter genes under the control of the p21 promoter and partial cell cycle arrest in some cellular models but are dominant negative for induction of apoptosis (54, 55), although diminished elevation of Fas expression has been reported with these constructs. There are conflicting reports concerning involvement of Fas/Fas ligand system in DDP-induced apoptosis (56, 57). Recent studies of Micheau et al. (58) suggest that drug-induced death pro- ceeds through induction of Fas, its clustering and interaction with Fas-associated death domain followed by activation of caspase-8 and caspase-3. We observed the absence or strong inhibition of DDP- induced Fas expression in HT29 and HCTp5.2 cells, respectively. Accordingly, DDP-induced activation of both caspase-8 and caspase-3 were impaired in these lines but not in HCT116, where addition of ansamycins caused additional increase in caspase activation. This data confirmed that loss of transcriptional p53 activity in HCTp5.2 cells contributed to inhibition of apoptotic pathways induced by DDP in this cell line and localize the effect of the ansamycins to the Fas-mediated pathway in particular. In summary, the activity and cellular responses to ansamycins Fig. 6. Treatment with DDP and 17-AAG causes differential activation of caspases in HT29, HCT116, and HCTp5.2 cell lines. HT29 cells (Ⅺ), HCT116 cells (f), and combined with DDP are dependent upon the genetic background of HCTp5.2 cells (o) were treated with 30 ␮M DDP and 500 nM GA/17-AAG alone or in the tumor cell model. In our HCT116 model, we varied the activity of combination for 20 h; protein extracts were isolated and used in colorimetric caspase 8 (A) and caspase 3 (B) assays as described in “Materials and Methods.” Graphs represent the a single gene, p53, and demonstrated that this reconstituted an antag- average values of three independent experiments; bars represent SD. onistic interaction that was observed in a p53-deficient cell line HT29. It is estimated that Ͼ80% of human tumors have loss or mutation of p53: thus, combinations of this nature need to be approached with apoptosis. It was critical to explore this interaction further if one is to great caution in clinical trials. It is not established that p53 is the use such combinations in clinical trials. only, nor less the most important, determinant of the interaction. The significance of scheduling in treatment of cells with Taxol and However, the data support a careful analysis of drug interactions 17-AAG was shown previously: pretreatment with 17-AAG abrogated with ansamycins. Taxol-induced apoptosis (22, 45). In our experiments, alternative scheduling not only did not affect antagonism of ansamycins and DDP in HT29 cells but led to the antagonism of these drugs in the HCT116 ACKNOWLEDGMENTS cell line as well. Because the effects of ansamycins (when added first) on cell cycle distribution and multiple signaling pathways occurred We thank Dr. Edward Sausville (National Cancer Institute) for kindly providing 17-AAG and Dr. Wafik El-Deiry (University of Pennsylvania) for before DDP addition, it could contribute to the observed antagonism. the gift of pcDNA3 plasmid expressing a dominant-negative form of p53 and Recently the binding of DDP to Hsp 90 leading to conformational critical review of the manuscript. We also thank Drs. Steven Johnson and changes was described (46–48). The consequences of this phenom- Kang-Shen Yao for constructive discussions and reviewing of the manuscript. enon are not completely understood, but it is possible to speculate that they could have a part in antagonism of DDP and ansamycins when DDP was introduced first in scheduling experiments. REFERENCES The cytotoxicity of DDP is the result of its binding to DNA, 1. Hanahan, D., and Weinberg, R. A. The hallmarks of cancer. Cell, 100: 57–70, 2000. followed by formation of intra- and interstrand DNA adducts, which 2. Gibbs, J. B. Anticancer drug targets: growth factors and growth factor signaling. are poorly repaired in human cells. The formation of DNA adducts J. Clin. Investig., 105: 9–13, 2000. 3. Sirotnak, F. M., Zakowski, M. F., Miller, V. A., Scher, H. I., and Kris, M. G. Efficacy consequently leads to inhibition of transcription, cell cycle arrest, and of cytotoxic agents against human tumor xenografts is markedly enhanced by coad- apoptosis (49). The tumoricidal effects of ansamycins used alone have ministration of ZD1839 (Iressa), an inhibitor of EGFR tyrosine kinase. Clin. Cancer been primarily associated with their ability to disrupt signaling Res., 6: 4885–4892, 2000. 4. Neckers, L. Hsp90 inhibitors as novel chemotherapeutic agents. Trends Mol. Med., 8: through survival pathways (12, 14, 16, 50). Used in combination, S55–S61, 2002. however, the ability of ansamycins to affect signaling through stress 5. Workman, P., and Maloney, A. HSP90 as a new therapeutic target for cancer therapy: pathways leading to cell death could interfere with cytotoxic effects of the story unfolds. Expert Opin. Biol. Ther., 2: 3–24, 2002. 6. Uehara, Y., Myrakawi, Y., Mizuno, S., and Kawai, S. Inhibition of transforming chemotherapeutic agents. In fact, GA at low doses was shown to activity of tyrosin kinase oncogenes by herbimycin A. Virology, 164: 294–298, 1988. prevent neurotoxic effects of DDP, vincristine, and Taxol on cultured 7. Whitesell, L., Mimnaugh, E. G., De Costa, B., Myers, C. E., and Neckers, L. M. Inhibition of heat-shock protein Hsp90-pp6␤v-src heteroprotein complex formation embryonic neurons (51). We consistently observed slight enhance- by benzoquinone ansamycins: essential role for stress proteins in oncogenic transfor- ment of cell survival occurring in the presence of low concentrations mation. Proc. Natl. Acad. Sci. USA, 91: 8324–8328, 1994. 3245

8. Richter, K., and Buchner, J. Hsp90: Chaperoning signal transduction. J. Cell. Physiol., of extracellular signal-regulated kinase activity increases sensitivity to cisplatin. Clin. 188: 281–290, 2001. Cancer Res., 5: 1007–1014, 1999. 9. Grenert, J. P., Sullivan, W. P., Fadden, P., Haystead, T. A. J., Clark, J., Mimnaugh, 33. Wang, X., Martindale, J. L., and Holbrook, N. J. Requirement for ERK activation in E., Krutzsch, H., Ochel, H. J., Schulte, T. W., Sausville, E., Neckers, L. M., and Toft, cisplatin-induced apoptosis. J. Biol. Chem., 275: 39435–39443, 2000. D. O. The amino-terminal domain of heat shock protein 90 (hsp90) that binds 34. Sanchez-Perez, I., Murguia, J. R., and Perona, R. Cisplatin induces a persistent geldanamycin is an ATP/ADP switch domain that regulates hsp90 conformation. activation of JNK that is related to cell death. Oncogene, 16: 533–540, 1998. J. Biol. Chem., 272: 23843–23850, 1997. 35. Hayakawa, J., Ohmich, M., Kurachi, H., Ikegami, H., Kimura, A., Matsuoka, T., 10. Whitesell, L., Sutphin, P., Wen, G. A., Schulte, T., Blagosklonny, M. V., and Jikihara, H., Mercola, D., and Murata, Y. Inhibition of extracellular signal-regulated Neckers, L. Geldanamycin-stimulated destabilization of mutated p53 is mediated by protein kinase or c-Jun N-terminal protein kinase cascade, differentially activated by proteasome in vivo. Oncogene, 14: 2809–2816, 1997. cisplatin, sensitized human ovarian cancer cell line. J. Biol. Chem., 274: 31648– 11. Schulte, T. W., Won, G. A., and Neckers, L. M. Geldanamycin-induced destabiliza- 31654, 1999. tion of Raf-1 involves the proteasome. Biochem. Biophys. Res. Commun., 239: 36. Sanches-Prieto, R., Rojas, J. M., Taya, Y., and Gutkind, S. A role for p38 mitogen- 655–659, 1997. activated protein kinase pathway in the transcriptional activation of p53 on genotoxic 12. Won, G. A., Schnur, R. C., Neckers, L., and Blagosklonny, M. V. Depletion of stress by chemotherapeutic agents. Cancer Res., 60: 2464–2472, 2000. p185erbB2, Raf- 1 and mutant p53 proteins by geldanamycin derivatives correlates 37. Vasilevskaya, I. A., and O’Dwyer, P. J. Effects of geldanamycin on signaling through with antiproliferative activity. Cancer Chemother. Pharmacol., 40: 60–64, 1997. activator-protein 1 in hypoxic HT29 human colon adenocarcinoma cells. Cancer Res., 13. Supko, J. G., Hickman, R. L., Grever, M. R., and Malspeis, L. Preclinical pharma- 59: 3935–3940, 1999. cological evaluation of geldanamycin as an antitumor agent. Cancer Chemother. 38. Mosmann, T. Rapid colorimetric assay for cellular growth and survival: application Pharmacol., 36: 305–315, 1995. to proliferation and cytotoxicity assays. J. Immunol. Methods, 65: 55–63, 1983. 14. Schulte, T., and Neckers, L. The benzoquinone ansamycin 17-allylamino-17-deme- 39. Tsai, C-M., Gazdar, A. F., Venzon, D. J., Steinberg, S. M., Dedrick, R. L., Mulshine, thoxy-geldanamycin binds to HSP90 and shares important biologic activities with J. L., and Kramer, B. S. Lack of in vitro synergy between etoposide and cis- geldanamycin. Cancer Chemother. Pharmacol., 42: 273–279, 1998. diammine-dichloroplatinum (II). Cancer Res., 49: 2390–2397, 1989. 15. Kelland, L. R., Sharp, S. Y., Rogers, P. M., Myers, T. G., and Workman, P. 40. Chou, T-C., Motzer, R. J., Tong, Y., and Bosl, G. J. Computerized quantification of DT-Diaphorase expression and tumor cells sensitivity to 17-allylamino,17-deme- synergism and antagonism of taxol, topotecan, and cisplatin against human terato- thoxygeldanamycin, an inhibitor of heat shock protein 90. J. Natl. Cancer Inst. carcinoma cell growth: a rational approach to clinical protocol design. J. Natl. Cancer. (Bethesda), 91: 1940–1949, 1999. Inst. (Bethesda), 86: 1517–1524, 1994. 16. Hostein, I., Robertson, D., DiStefano, F., Workman, P., and Clarke, P. A. Inhibition 41. Liu, Y., and Kulesz-Martin, M. p53 protein at the hub of cellular DNA damage of signal transduction by the Hsp90 inhibitor 17-allylamino-17-demethoxygeldana- response pathways through sequence-specific and non-sequence-specific DNA bind- mycin results in cytostasis and apoptosis. Cancer Res., 61: 4003–4009, 2001. ing. Carcinogenesis (Lond.), 22: 851–860, 2001. 17. Yang, J., Yang, J-M., Iannone, M., Shih, J., Lin, Y., and Hait, W. N. Disruption of the 42. Gjerset, R. A., Lebedeva, S., Haghigni, A., Turia, S. T., and Mercola, D. Inhibition EF-2 kinase/Hsp90 protein complex: a possible mechanism to inhibit glioblastoma by of Jun kinase pathway blocks DNA repair, enhances p53-mediated apoptosis and Geldanamycin. Cancer Res., 61: 4010–4016, 2001. promotes gene amplification. Cell Growth Differ., 10: 545–554, 1999. 18. Mu¨nster, P. N., Srethapakdi, M., Moasser, M. M., and Rosen, N. Inhibition of heat 43. Pan, B., Yao, K-S., Monia, B., Dean, N. M., McKay, R. A., Hamilton, T. C., and shock protein 90 function by ansamycins causes the morphological and functional O’Dwyer, P. J. Reversal of cisplatin resistance in human ovarian cancer cell lines by differentiation of breast cancer cells. Cancer Res., 61: 2945–2952, 2001. a c-jun oligodeoxynucleotide (ISIS 10582): evidence for the role of transcription 19. Wilson, R. H., Takimoto, C. H., Agnew, E. B., Morrison, G., Grollman, F., Thomas, factor overexpression in determining resistant phenotype. Biochem. Pharmacol., 63: R. R., Saif, M. W., Hopkins, J., Allegra, C., Grochow, L., Szabo, E., Hamilton, J. M., 1699–1707, 2002. Monhan, B. P., Neckers, L., and Grem, J. L. Phase I pharmacological study of 44. Shaulian, E., and Karin, M. AP-1 as a regulator of cell life and death. Nat. Cell Biol., 17-(allylamino)-17-demethoxygeldanamycin (AAG) in adult patients with advanced 4: E131–E136, 2002. solid tumors. Proc. Am. Soc. Clin. Oncol. Annu. Meet., 20: 82a, 2001. 45. Nguen, D. M., Chen, A., Mixon, A., and Schrump, D. S. Sequence-dependent 20. Mu¨nster, P. N., Tong, W., Schwartz, L., Larson, S., Keneson, K., De La Cruz, A., enhancement of paclitaxel toxicity in non-small cell lung cancer by 17-allylamino Rosen, N., and Scher, H. Phase I trial of 17-(allylamino)-17-demethoxygeldanamycin 17-demethoxy-geldanamycin. J. Thorac. Cardiovasc. Surg., 118: 908–915, 1999. (17AAG) in patients (pts) with advanced solid malignancies. Proc. Am. Soc. Clin. 46. Itoh, H., Ogura, M., Komatsuda, A., Wakui, H., Miura, A. B., and Tashima, Y. A Oncol. Annu. Meet., 20: 83a, 2001. novel chaperone-activity-reducing mechanism of the 90-kDa molecular chaperone 21. Nguyen, D. M., Logang, D., Chen, A., Stewart, J. H., IV, Tabibi, E., and Schrump, D. S. Enhancement of paclitaxel-mediated cytotoxicity in lung cancer cells by HSP90. Biochem. J., 343: 697–703, 1999. 17-allylamino geldanamycin: in vitro and in vivo analysis. Ann. Thorac. Surg., 72: 47. Marcu, M. G., Chadli, A., Bouhouche, I., Gatelli, M., and Neckers, L. M. The heat 371–379, 2001. shock protein 90 antagonist novobiocin interacts with a previously unrecognized 22. Mu¨nster, P. N., Basso, A., Solit, D., Norton, L., and Rosen, N. Modulation of Hsp90 ATP-binding domain in the carboxyl-terminus of the chaperone. J. Biol. Chem., 275: function by ansamycins sensitizes breast cancer cells to chemotherapy-induced ap- 37181–37186, 2000. optosis in RB- and schedule-dependent manner. Clin. Cancer Res., 7: 2228–2236, 48. Soti, C., Racz, A., and Csermely, P. A nucleotide-dependent molecular switch 2001. controls ATP binding at C-terminal domain of Hsp90: N-terminal nucleotide binding 23. Reinert, K. E. Anthracycline-binding induced DNA stiffening, bending and elonga- unmasks a C-terminal binding pocket. J. Biol. Chem., 277: 7066–7075, 2002. tion: stereochemical implications from viscometric investigations. Nucleic Acid Res., 49. Jordan, P., and Carmo-Fonseca, M. Molecular mechanisms involved in cisplatin 11: 3411–3440, 1983. cytotoxicity. Cell. Mol. Life Sci., 57: 1229–1235, 2000. 24. Rosenberg, B. Platinim complexes for the treatment of cancer: why the search goes 50. Mu¨nster, P. N., Marchion, D. C., Basso, A., and Rosen, N. Degradation of HER2 by on. In: B. Lippert (ed.), Cisplatin: Chemistry and Biochemistry of a Leading Anti- ansamycins induces growth arrest and apoptosis in cells with HER2 overexpression Ј cancer Drug, p. 3. Zurich, Switzerland: Verlag Helvetica Chimica Acta, 1999. via a HER3, phosphatidylinositol 3 -kinase-AKT-dependent pathway. Cancer Res., 25. O’Dwyer, P., Stevenson, J., and Johnson, S. Clinical status of cisplatin, carboplatin 62: 3132–3137, 2002. and other platinum-based antitumor drugs. In: B. Lippert (ed.), Cisplatin: Chemistry 51. Sano, M. Radicicol and geldanamycin prevent neurotoxic effects of anti-cancer drugs and Biochemistry of a Leading Anticancer Drug, p. 31. Zurich, Switzerland: Verlag on cultured embryonic sensory neurons. Neuropharmacology, 40: 947–953, 2001. Helvetica Chimica Acta, 1999. 52. Chao, C., Saito, S., Kang, J., Anderson, C. W., Appella, E., and Xu, Y. p53 26. Johnson, S. W., Stevenson, P. J., and O’Dwyer, P. J. Cisplatin and its analogues. In: transcriptional activity is essential for p-53 dependent apoptosis following DNA V. T. DeVita, S. Hellman, and S. A. Rosenberg (eds.), Cancer: Principles and Practice damage. EMBO J., 19: 4967–4975, 2000. of Oncology, Ed. 6, pp. 377–388. Philadelphia: Lippincott, Williams & Wilkins, 53. Bunz, F., Hwang, P. M., Torrance, C., Waldman, T., Zhang, Y., Dillehay, L., 2001. Williams, J., Lengauer, C., Kinzler, K. W., and Vogelstein, B. Disruption of p53 in 27. Gonzales, V. M., Fuertes, M. A., Alonso, C., and Perez, J. M. Is cisplatin-induced human cancer cells alters the responses to therapeutic agents. J. Clin. Investig., 104: death always produced by apoptosis? Mol. Pharmacol., 59: 657–663, 2001. 263–269, 1999. 28. Fulda, S., Los, M., Friesen, C., and Debatin, K-M. Chemosensitivity of solid tumor 54. Ryan, K. M., and Voudsen, K. H. Characterization of structural p53 mutants which cells in vitro is related to activation of the CD95 system. Int. J. Cancer, 76: 105–114, show selective defects in apoptosis but not cell cycle arrest. Mol. Cell. Biol., 18: 1998. 3692–3698, 1998. 29. Kojima, H., Endo, K., Moriyama, H., Tanaka, Y., Alnemri, E. S., Slapak, C. A., 55. Aurelio, O. N., Kong, X-T., Gupta, S., and Stanbridge, E. J. p53 mutants have Teicher, B., Kufe, D., and Datta, R. Abrogation of mitochondrial cytochrome C selective dominant-negative effects on apoptosis but not growth arrest in human release and caspase-3 activation in acquired multidrug resistance. J. Biol. Chem., 273: cancer cell lines. Mol. Cell. Biol., 20: 770–778, 2000. 16647–16650, 1998. 56. Ferreira, C. G., Tolis, C., Span, S. W., Peters, G. J., van Lopik, T., Kummer, A. J., 30. Allday, Inman, G. J., Crawford, D. H., and Farrell, P. J. DNA damage in human B Pinedo, H. M., and Giaccone, G. Drug-induced apoptosis in lung cancer cells is not cells can induce apoptosis, proceeding from G1-S when p53 is transactivation com- mediated by the Fas/FasL (CD95/APO1) signaling pathway. Clin. Cancer Res., 6: petent and G2-M when it is transactivation defective. EMBO J., 14: 4994–5005, 203–212, 2000. 1995. 57. Micheau, O., Hammann, A., Solary, E., and Dimanche-Boitrel, M-T. STAT-1- 31. Seki, T., Ohba, N., Makino, R., Funatomi, H., and Mitamura, K. Mechanism of independent up-regulation of procaspase -3 and -8 in cancer cells treated with growth-inhibitory effect of cisplatin on human pancreatic cancer cells and status of cytotoxic drugs. Biochem. Biophys. Res. Commun., 256: 603–607, 1999. p53 gene. Anticancer Res., 21: 1919–1924, 2001. 58. Micheau, O., Solary, E., Hammann, A., and Dimanche-Boitrel, M-T. Fas ligand- 32. Persons, D. L., Yazlovitskaya, E. M., Cui, W., and Pelling, J. C. Cisplatin-induced independent, FADD-mediated activation of the Fas death pathway by anticancer activation of mitogen-activated protein kinases in ovarian carcinoma cells: inhibition drugs. J. Biol. Chem., 274: 7987–7992, 1999.

3246

Downloaded from cancerres.aacrjournals.org on September 27, 2021. © 2003 American Association for Cancer Research. Geldanamycin and its 17-Allylamino-17-Demethoxy Analogue Antagonize the Action of Cisplatin in Human Colon Adenocarcinoma Cells: Differential Caspase Activation as a Basis for Interaction

Irina A. Vasilevskaya, Tatiana V. Rakitina and Peter J. O'Dwyer

Cancer Res 2003;63:3241-3246.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/63/12/3241

Cited articles This article cites 55 articles, 25 of which you can access for free at: http://cancerres.aacrjournals.org/content/63/12/3241.full#ref-list-1

Citing articles This article has been cited by 8 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/63/12/3241.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/63/12/3241. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.