- Email: [email protected]
Multiple Kinase Pathways Involved in the Different De Novo Sensitivity of Pancreatic Cancer Cell Lines to 17-AAG
Journal of Surgical Research 176, 147–153 (2012) doi:10.1016/j.jss.2011.09.017
Multiple Kinase Pathways Involved in the Different De Novo Sensitivity of Pancreatic Cancer Cell Lines to 17-AAG Heping Liu, Ph.D.,* Ti Zhang, M.D., Ph.D.,§ Rong Chen, Ph.D.,† David J. McConkey, Ph.D.,‡ John F. Ward, M.D.,‡ and Steven A. Curley, M.D.*,1 *Department of Surgical Oncology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas; †Department of Experimental Therapeutics, The University of Texas M. D. Anderson Cancer Center, Houston, Texas; ‡Department of Urology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas; and §Department of Hepatobiliary Surgery, Tianjin Cancer Hospital, Tianjin Medical University, Tianjin, China Originally submitted April 22, 2011; accepted for publication September 9, 2011
Background. 17-Allylamino-17-demethoxygeldanamycin (17-AAG) specifically targets heat shock protein (HSP)90 and inhibits its chaperoning functions for multiple kinases involved in cancer cell growth and survival. To select responsive patients, the molecular mechanisms underlying the sensitivity of cancer cells to 17-AAG must be elucidated. Materials and Methods. We used cytotoxicity assays and Western blotting to explore the effects of 17-AAG and sorafenib on cell survival and expression of multiple kinases in the pancreatic cancer cell lines AsPC-1 and Panc-1. Gene cloning and transfection, siRNA silencing, and immunohistochemistry were used to evaluate the effects of mutant p53 protein on 17-AAG sensitivity. Results. AsPC-1 and Panc-1 responded differently to 17-AAG, with half maximal inhibitory concentration (IC50) values of 0.12 and 3.18 mM, respectively. Comparable expression of HSP90, HSP70, and HSP27 was induced by 17-AAG in AsPC-1 and Panc-1 cells. P-glycoprotein and mutant p53 did not affect 17-AAG sensitivity in these cell lines. Multiple kinases are more sensitive to HSP90 inhibition in AsPC-1 than in Panc1 cells. After 17-AAG treatment, p-Bad (S112) decreased in AsPC-1 cells and increased in Panc-1 cells. Sorafenib markedly increased p-Akt, p-ERK1/2, p-GSK-3b, and p-S6 in both cell lines. Accordingly, 17-AAG and sorafenib acted antagonistically in AsPC-1 and Panc-1 cells, except at high concentrations in AsPC-1 cells. Conclusions. Differential inhibition of multiple kinases is responsible for the different de novo sensitiv1 To whom correspondence and reprint requests should be addressed at Department of Surgical Oncology, Unit 1484, The University of Texas M. D. Anderson Cancer Center, 1400 Pressler Street, Houston, TX 77030. E-mail: [email protected].
ity of AsPC-1 and Panc-1 cells to HSP90 inhibition. P-glycoprotein and mutant p53 protein did not play a role in the sensitivity of pancreatic cancer cells to 17-AAG. Ó 2012 Elsevier Inc. All rights reserved. Key Words: 17-AAG; pancreatic cancer; chemosensitivity; kinase.
INTRODUCTION
Heat shock proteins (HSPs) constitute a large family of proteins that have been classified according to their molecular weight. For example, HSP70 is the 70-kDa HSP; other family members are HSP27, HSP60, and HSP90. HSPs, especially HSP70, can be induced by stress such as heat or cold shock, cytotoxic drugs, and radiation. The 90-kDa chaperone protein, HSP90, is highly conserved and ubiquitously expressed in all living organisms. HSP90 requires several co-chaperones, such as Cdc37, to form a highly complicated complex that regulates the folding/unfolding, assembly, and maturation of its client proteins [1]. HSP90 regulates more than 200 cellular client proteins. Among them are serine/threonine or tyrosine kinases, such as c-Raf-1, PKB/Akt, Bcr-Abl, CDK4, and v-Src, which are crucial to proliferation, differentiation, and survival of cancer cells [1]. The ansamycin antibiotic geldanamycin is a novel anticancer drug that can specifically bind to an ATP/ADP pocket in the N-terminal domain of HSP90 and inhibit HSP90 function by preventing full chaperone cycling [2]. Treatment with geldanamycin results in the accumulation of unprocessed chaperone–client protein complexes within cells, and the premature release and
147
0022-4804/$36.00 Ó 2012 Elsevier Inc. All rights reserved.
148
JOURNAL OF SURGICAL RESEARCH: VOL. 176, NO. 1, JULY 2012
degradation of vital cellular proteins in the proteasome leads to cell death [3, 4]. Preclinical experiments led to the surprising observation that cancer cells, because of their so-called ‘‘kinase addiction,’’ are significantly more sensitive to HSP90 inhibition than nontransformed cells are [5, 6]. This finding raised the possibility that HSP90 inhibitors could be used in cancer patients as new chemotherapeutic agents. A geldanamycin derivative with a more favorable toxicity profile, 17-allylamino-17-demethoxy-geldanamycin (17-AAG), has entered clinical phase I/II trials for breast cancer, myeloma, and metastatic melanoma after being tested in xenograft and transgenic mouse models [7–9]. The membrane transporter P-glycoprotein (also called ABCB1 or MDR1) functions as a drug efflux pump that actively transports drugs from the inside to the outside of cells and leads to a decreased intracellular accumulation of drugs needed to kill cancer cells. P-glycoprotein is known to be involved in cancer cell resistance to many chemotherapeutic agents [10]. Several studies have also shown that P-glycoprotein was involved in the resistance of tumor cells to geldanamycin or its derivatives [11, 12]. Inactivating p53 mutations are the most common genetic alterations found in human cancers. Unlike mutations in other tumor suppressor genes, most p53 mutations are missense mutations, of which about 97% occur within the sequence-specific DNA binding domain. Accumulation of mutant p53 protein may lead to novel biological effects in cancer cells, consistent with a gain-of-function phenotype [13, 14]. Moreover, mutant p53 protein is a HSP90 client protein [15] and may modulate the responsiveness of cancer cells to chemotherapeutic drugs [16]. To accurately predict clinical response to 17-AAG and to select patients who will benefit from 17-AAG in future clinical use, it is important to elucidate the molecular mechanisms underlying the de novo sensitivity of cancer cells to 17-AAG or similar HSP90 inhibitors. Pancreatic cancer is notorious for lack of response to conventional chemotherapy. In the current study, we sought to determine the genetic determinants that regulate the sensitivity of AsPC-1 and Panc-1 cells, two human pancreatic cancer cell lines, to 17-AAG.
MATERIALS AND METHODS Cell Lines and Reagents The pancreatic cancer cell lines AsPC-1 and Panc-1 were obtained from the American Type Culture Collection (Manassas, VA) and grown in Dulbecco’s modified Eagle’s medium containing 10% fetal bovine serum, 2 mM glutamine, and 1 mM sodium pyruvate in a 37 C humidified incubator with 5% CO2. 17-AAG was purchased from Sigma (St. Louis, MO) and dissolved in methanol to make 5-mM stock. Nexavar (sorafenib) pills (Bayer HealthCare Pharmaceuticals) were
obtained from the M. D. Anderson Cancer Center pharmacy and dissolved in DMSO to make 10-mM stock.
Western Blotting Whole-cell lysates were separated in either 7.5% or 10% denaturing polyacrylamide gel , depending on the molecular sizes of the proteins, as described previously [17], and transferred by electroblotting to nitrocellulose membranes (Bio-Rad Laboratories, Hercules, CA). Primary antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA), CalBiochem (San Diego, CA), and Cell Signaling Technology (Beverly, MA). The membranes were incubated with the appropriate primary antibodies, and specific immunoreactivity was detected with fluorescently labeled secondary antibodies (Alexa Fluor 680 [Invitrogen, Carlsbad, CA] and IRDye 800CW [LI-COR Biosciences, Lincoln, NE]) using the Odyssey imager and software ver. 3 (LI-COR). To confirm approximately equal loading, the membranes were also incubated with anti-b-actin antibody.
Cytotoxicity Assay Cytotoxicity was assessed with the 3-(4,5 dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide (MTT) (Trevigen, Gaithersburg, MD) colorimetric assay [18]. Cells were plated at 2–3 3 103 per well in 96-well plates in triplicate and allowed to adhere overnight. Cells were then treated with appropriate drugs at varying concentrations for 72 h at 37 C. MTT was added to each well, and incubation was continued at 37 C for 4 h. Cell viability was measured by dissolving the cellular reduced product of MTT in 200 mL of DMSO and reading at 570 nm with the FLUOstar Omega microplate reader (BMG Labtech, Chicago, IL). Each test was performed at least three times.
Statistical Analysis Half maximal inhibitory concentration (IC50) values were calculated with SigmaPlot ver. 10.0.1 (Systat Software, San Jose, CA). The two-tailed Student’s t-test was used for unpaired comparisons. To characterize synergistic or antagonistic interactions between agents, we performed median dose–effect analysis with CalcuSyn software (Biosoft, Ferguson, MO) [19]. Data are presented as mean 6 SD of at least three replicate experiments.
RESULTS Effects of HSP90 Inhibition on Tumor Cell Survival and Multiple Kinase Pathways
To determine the responses of two human pancreatic cancer cell lines, AsPC-1 and Panc-1, to 17-AAG, we used MTT assays. The sensitivity of Panc-1 cells was substantially dependent on the confluence of the cell cultures, so Panc-1 cells exhibited confluencedependent resistance (CDR) (Fig. 1A). AsPC-1 cells, however, did not exhibit substantial CDR. To compare the sensitivity of the two cell lines to 17-AAG, we grew the cells to 60%–80% confluence before plating them for cytotoxicity assays. We found that Panc-1 cells were much more resistant to 17-AAG than AsPC-1 cells were, consistent with the results of a previous study [20]; Panc-1 and AsPC-1 cells had IC50 values of 3.18 and 0.12 mM, respectively (Fig. 1B). The different responses of AsPC-1 and Panc-1 cells to 17-AAG may be due to the different sensitivities of
LIU ET AL.: MULTIPLE KINASES REGULATE CANCER SENSITIVITY TO 17-AAG
149
FIG. 1. Multiple kinase signaling pathways are more sensitive to 17-AAG in AsPC-1 cells than in Panc-1 cells. (A), (B) We used MTT cell survival assays to determine the cytotoxic effect of 17-AAG on AsPC-1 and Panc-1 cells after 72 h of treatment. Each experiment was performed in triplicate. Bars, SD. (A) Panc-1 cells exhibited CDR. (B) Sensitivity was determined in the logarithmic growth phase (60%–80% confluence), and IC50 values were 0.12 mM for AsPC-1 cells and 3.18 mM for Panc-1 cells. (C) Cell lysates were prepared after 24 h of treatment with 17-AAG at different concentrations and separated on polyacrylamide gel. The membrane was probed with the indicated antibodies. Lanes 1–6, AsPC-1 cells treated with 17-AAG at 0, 0.05, 0.1, 0.2, 0.4, and 0.8 mM; lanes 7–12, Panc-1 cells treated with 17-AAG at 0, 0.4, 0.8, 1.6, 3.2, and 6.4 mM.
multiple kinases to HSP90 inhibition. We determined the sensitivities of these kinases to HSP90 inhibition by Western blotting after 24 h of treatment with varying concentrations of 17-AAG. Although the total amounts of the kinases were unaffected, the levels of phosphorylated ERK1/2 (p-ERK1/2) and p-MEK1/2, in the MAP kinase pathway, were much more sensitive to 17-AAG in AsPC-1 cells than in Panc-1 cells. Similarly, levels of p-S6 ribosomal protein and phosphorylated glycogen synthase kinase-3b (p-GSK-3b) were much more sensitive in AsPC-1 cells than in Panc-1 cells. In contrast, levels of p-Bad (S112) changed in opposite directions in the two cell lines with increasing 17-AAG concentration, decreasing in AsPC-1 cells and increasing in Panc-1 cells, whereas p-Bad (S136) was almost undetectable in both AsPC-1 and Panc-1 cells (Fig. 1C). p-Bad is sequestered in the cytosol by binding to 14-3-3 to promote cell survival [21]; the differential changes of p-Bad (S112) levels after 17-AAG treatments might be responsible for the different cytotoxic effects of HSP90 inhibition in AsPC-1 and Panc-1 cells. In addition, CDK4 levels were more sensitive to 17-AAG in AsPC-1 cells than in Panc-1 cells (data not shown). Expression of Heat Shock Proteins, not P-Glycoprotein, Is Induced by 17-AAG Treatment
Because 17-AAG inhibits HSP90 chaperoning function [2], we determined the expression levels of various HSPs in AsPC-1 and Panc-1 cells. We found that the
expression levels of the three major HSPs, HSP27, HSP70, and HSP90, were comparable in AsPC-1 and Panc-1 cells without 17-AAG treatment. After 72 h of 17-AAG treatment, expression of HSP70 was strongly induced, whereas expression of HSP27 and HSP90 was only moderately induced (Fig. 2A). Induction of HSP70 expression was more pronounced in Panc-1 cells (6.63 the level without 17-AAG treatment) than in AsPC-1 cells (2.43 the untreated level) (Fig. 2C and D). P-glycoprotein has been reported to mediate the resistance of cancer cells to 17-AAG [11, 12, 22]. However, after 72 h of treatment with 17-AAG at indicated concentrations, P-glycoprotein was not detectable in AsPC-1 or Panc-1 cells with Western blotting (Fig. 2B), whereas the positive control showed strong P-glycoprotein expression. The positive control H460/ TaxR was prepared from a lung cancer cell line, H460, with acquired paclitaxel resistance established by repeated treatment with Taxol [23]. On the basis of our microarray data for pancreatic cancer cell lines [24], we used a similar pharmacogenomic approach to determine whether there was a correlation between 17-AAG sensitivity and mRNA expression levels of P-glycoprotein in nine pancreatic cancer cell lines (MIAPaCa-2, BxPC-3, AsPC-1, L3.6p1, Hs766T, MPanc96, SU86.86, CFPAC, and Panc-1); we did not identify any association of P-glycoprotein mRNA levels with 17-AAG resistance (data not shown). Panc-1 cells harbor a missense mutation in p53 DNA binding domain and express high levels of mutant p53
150
JOURNAL OF SURGICAL RESEARCH: VOL. 176, NO. 1, JULY 2012
FIG. 2. Expression of HSP27, HSP70, and HSP90, but not of P-glycoprotein, is induced by the HSP90 inhibitor 17-AAG. (A), (C), (D) Cellular extracts were prepared and treated with the indicated concentrations of 17-AAG for 72 h. Expression levels of the three major HSPs (HSP27, HSP70, and HSP90) were determined by probing with their corresponding antibodies. (A) Compared with b-actin, we found that expression of HSP70 was strongly induced, whereas expression of HSP27 and HSP90 was only moderately induced. Signal intensities were calculated for (C) AsPC-1 and (D) Panc-1 cells. (B) Expression of P-glycoprotein (MDR1) was not detectable in AsPC-1 or Panc-1 cells with or without 17-AAG treatment. A positive control, TaxR, was included in this Western blot.
protein, whereas AsPC-1 cells harbor a frameshift (null) mutation in p53 with no detectable expression of mutant p53 protein (data not shown). We used p53 siRNA and expression vectors and found that mutant p53 protein did not have any impact on the sensitivity of Panc-1 and AsPC-1 cells to 17-AAG (Figs. S1 and S2, Supplemental Data). Sorafenib (Nexavar), a Multiple Kinase Inhibitor, Regulates 17-AAG Sensitivity
Sorafenib, originally developed as a Raf kinase inhibitor, is a multiple kinase inhibitor [25]. We found that sorafenib could paradoxically up-regulate the levels of multiple phosphorylated kinase substrates in AsPC-1 and Panc-1 cells as determined by Western blotting after 24 h of 17-AAG treatment (Fig. 3). Thus, combined treatment with 17-AAG and sorafenib had an antagonistic effect on multiple kinase pathways. We found that levels of p-ERK1/2 (T202/Y204), p-Akt (S473), p-S6 (S235/236), and p-GSK-3b were elevated after combined treatment with 17-AAG and sorafenib for 24 h. The total levels of most of these proteins were unaffected, but total Akt decreased markedly at increasing concentrations of 17-AAG and sorafenib in AsPC-1 cells. Sorafenib treatment also shifted multiple lowmolecular-weight isoforms of B-Raf to high-molecularweight ones in AsPC-1 cells, with no apparent change in Panc-1 cells (Fig. 3).
To further confirm our observation that the inherent differences in multiple kinases regulated the sensitivity of AsPC-1 and Panc-1 cells to 17-AAG, we extended our Western blot determinations to cytotoxicity assays. AsPC-1 and Panc-1 cells had very similar responses to sorafenib, with IC50 values of about 11.82 mM. At 5 mM, sorafenib seemed to decrease the sensitivity of Panc-1 cells to 17-AAG, increasing IC50 by almost a factor of 5, while not significantly altering the sensitivity of AsPC-1 cells to 17-AAG (Fig. 4A and B). To clarify the antagonistic interaction of 17-AAG and sorafenib, we determined the combination index (CI) in AsPC-1 and Panc-1 cells according to Chou and Talalay [19]. Consistent with our results from Western blotting, we found that 17-AAG and sorafenib mostly had an antagonistic effect (CI > 1), except that they acted synergistically (CI < 1) at high concentrations in AsPC-1 cells (Fig. 4C and D). The change in p-HSP90 (T4/5) as a percentage of total HSP90 in AsPC-1 cells and Panc-1 cells after 24 h of treatment with 17-AAG and sorafenib showed a similar pattern with our Western blotting and CI determinations: the percentage decreased with increasing concentration of 17-AAG and sorafenib, except at the highest concentration of 17-AAG in AsPC-1 cells (Fig. 4E and F). DISCUSSION
Chemosensitivity is likely to be an inherent, complex phenotype, with genetic polymorphisms, protein
LIU ET AL.: MULTIPLE KINASES REGULATE CANCER SENSITIVITY TO 17-AAG
151
FIG. 3. Sorafenib antagonizes the effects of 17-AAG on multiple kinase pathways. Cell lysates were prepared after 24 h of treatment with 17-AAG and sorafenib at the indicated concentrations and separated on polyacrylamide gel. The membrane was probed with the indicated antibodies. The treatments had no effect on the level of the co-chaperone molecule CDC37 or its activated form, p-CDC37. However, 17-AAG and sorafenib acted antagonistically on the Akt(p-Akt), Wnt (p-GSK3b), and MAP kinase (p-ERK1/2) signaling pathways, more markedly so in AsPC-1 cells than in Panc-1 cells. 17-AAG and sorafenib had no effect on the total amounts of the indicated proteins, except that they markedly decreased the level of total Akt at the highest concentrations (2 3 IC50) in AsPC-1 cells. A ¼ 17-AAG; S ¼ sorafenib. The numbers indicate concentrations in mM.
expression alterations, and post-translational modifications playing significant roles. In this study, two pancreatic cancer cell lines, AsPC-1 and Panc-1, provided us with an excellent model to explore the molecular determinants of sensitivity to 17-AAG, a novel anticancer agent that inhibits the chaperoning function of HSP90 [6]. In the logarithmic growth phase (60%–80% confluence), AsPC-1 cells were much more sensitive to 17AAG than Panc-1 cells. Many kinases responsible for cell proliferation and survival are client proteins of HSP90. Targeting HSP90 with 17-AAG leads to degradation of these kinases and cell death [9]. Thus, the difference in the responsiveness of the multiple kinases to HSP90 inhibition might be the basis for the different cytotoxic effects of 17-AAG on AsPC-1 and Panc-1 cells. Our Western blotting results indeed showed that after 24 h of 17-AAG treatment, several key phosphorylated kinases (ERK1/2, MEK1/2, and GSK-3b) decreased to much lower levels in AsPC-1 cells than in Panc-1 cells. Growth factors can stimulate Bad phosphorylation, which suppresses cell apoptosis and promotes cell survival [21]. We found that HSP90 inhibition resulted in decreased p-Bad (S112) levels in AsPC-1 cells but increased levels in Panc-1 cells. p-Bad (S112) levels may be a predictive factor for the sensitivity of pancreatic cancer cells to 17-AAG, although the underlying mechanism for this effect remains to be elucidated.
As cytotoxic drugs, HSP90 inhibitors can activate the heat shock response [11, 26]. Our results also revealed that treatment with 17-AAG induced expression of the three major HSPs, HSP90, HSP70, and HSP27, especially HSP70 (Fig. 2A), which increased to 6.63 and 2.43 the level without 17-AAG treatment in Panc-1 and AsPC-1 cells, respectively. The stronger heat shock response in Panc-1 cells might be attributable to 17AAG resistance because induction of stress response proteins, such as HSP27 and HSP70, by HSP90 inhibition could offset the cytotoxic effects of 17-AAG [11]. In the microarray profiling of 60 human tumor cell lines (NCI-60), their sensitivity to geldanamycin and its analogs displayed negative correlation with mRNA expression levels of P-glycoprotein, suggesting a role of P-glycoprotein in chemoresistance [22]. However, our microarray data of nine pancreatic cancer cell lines did not show any association between 17-AAG sensitivity and P-glycoprotein mRNA expression levels. Moreover we did not detect any protein expression of P-glycoprotein in AsPC-1 or Panc-1 cells, indicating that P-glycoprotein was not involved in the responsiveness of AsPC-1 and Panc-1 cells to 17-AAG. Mutant p53 protein in AsPC-1 and Panc-1 cells had no impact on responsiveness to HSP90 inhibition. We found that sorafenib, a multiple kinase inhibitor, could paradoxically up-regulate the phosphorylated
152
JOURNAL OF SURGICAL RESEARCH: VOL. 176, NO. 1, JULY 2012
FIG. 4. Sorafenib and 17-AAG have antagonistic effects in AsPC-1 and Panc-1 cells. (A), (B) AsPC-1 and Panc-1 cells had similar responses to sorafenib, with IC50 values of about 11.82 mM. Sorafenib at 5 mM had a negligible impact on the sensitivity of AsPC-1 cells but greatly increased the resistance of Panc-1 cells to 17-AAG, according to the IC50 values. Bars, SD. The CI curves of sorafenib and 17-AAG for (C) AsPC-1 cells and (D) Panc-1 cells, simulated with CalcuSyn software, showed antagonism, except at high concentrations in AsPC-1 cells. Red squares indicate actual measurements. (E), (F) p-HSP90 (T4/5) as a percentage of total HSP90 changed in a similar way in AsPC-1 and Panc-1 cells after 17-AAG and sorafenib treatments, except at the highest concentration of 17-AAG in AsPC-1 cells. (F) Signal intensities calculated from (E). 1–3, 17-AAG treatments; 4–6, sorafenib treatments.
components in multiple kinase pathways. To confirm our conclusion that the sensitivity of AsPC-1 and Panc-1 cells to HSP90 inhibitors was related with their intrinsic multiple kinase pathways, we determined the combinatorial effects of 17-AAG and sorafenib on both AsPC-1 and Panc-1 cells. As expected, sorafenib, at a low concentration (5 mM), greatly increased the resistance of Panc-1 cells to 17-AAG, according to the IC50 values, but had a negligible impact on the sensitivity of AsPC-1 cells to 17-AAG. The CIs of 17-AAG and sorafenib also indicated that they acted antagonistically in AsPC-1 and Panc-1 cells, except at high concentrations in AsPC-1 cells. At these concentrations, combined 17AAG and sorafenib treatments markedly reduced total Akt levels in AsPC-1 cells, a likely explanation for the synergism of 17-AAG and sorafenib in AsPC-1 cells. Thus, these data further demonstrate that HSP90 regulates the kinase cascades differently in AsPC-1 cells than in Panc-1 cells and that this difference is responsible for the different cytotoxicity of 17-AAG in these two cell lines. Although the molecular mechanism underlying the different HSP90 functions in AsPC-1 and Panc-1 cells remains elusive at present, our research
represents an important and significant attempt in this area. Our data may be of clinical application for predicting the patients’ responses to HSP90 inhibition. ACKNOWLEDGMENTS The authors thank Yue Lu for help in preparing the manuscript. Yafang Li helped with some statistical analyses and the CI calculation. Dr. Bingliang Fang kindly provided the positive control (H460/ TaxR) for the P-glycoprotein. The authors also thank Karen R. Muller, Kristine K. Ash, and Yolanda Brittain for editing and submitting this manuscript. This work was supported by the Kanzius Cancer Research Foundation.
SUPPLEMENTARY DATA
Supplementary data related to this article can be found online at doi: 10.1016/j.jss.2011.09.017. REFERENCES 1. Blume-Jensen P, Hunter T. Oncogenic kinase signaling. Nature 2001;411:355. 2. Whitesell L, Mimnaugh EG, De Costa B, et al. Inhibition of heat shock protein HSP90-pp60v-src heteroprotein complex formation by benzoquinone ansamycins: Essential role for stress proteins in oncogenic transformation. Proc Natl Acad Sci U S A 1994;91:8324.
LIU ET AL.: MULTIPLE KINASES REGULATE CANCER SENSITIVITY TO 17-AAG 3. Sepp-Lorenzino L, Ma Z, Lebwohl DE, et al. Herbimycin A induces the 20 S proteasome- and ubiquitin-dependent degradation of receptor tyrosine kinases. J Biol Chem 1995;270:16580. 4. Mimnaugh EG, Chavany C, Neckers L. Polyubiquitination and proteasomal degradation of the p185c-erbB-2 receptor proteintyrosine kinase induced by geldanamycin. J Biol Chem 1996; 271:22796. 5. Kamal A, Boehm MF, Burrows FJ. Therapeutic and diagnostic implications of Hsp90 activation. Trends Mol Med 2004; 10:283. 6. Neckers L. Using natural product inhibitors to validate Hsp90 as a molecular target in cancer. Curr Top Med Chem 2006;6:1163. 7. Modi S, Stopeck AT, Gordon MS, et al. Combination of trastuzumab and tanespimycin (17-AAG, KOS-953) is safe and active in trastuzumab-refractory HER-2 overexpressing breast cancer: A phase I dose-escalation study. J Clin Oncol 2007;25:5410. 8. Anderson KC, Hannah AL, Pazdur R, et al. A strategic framework for novel drug development in multiple myeloma. Br J Haematol 2007;138:153. 9. Solit DB, Osman I, Polsky D, et al. Phase II trial of 17allylamino-17-demethoxygeldanamycin in patients with metastatic melanoma. Clin Cancer Res 2008;14:8302. 10. Endicott JA, Ling V. The biochemistry of P-glycoprotein-mediated multidrug resistance. Annu Rev Biochem 1989;58:137. 11. McCollum AK, TenEyck CJ, Stensgard B, et al. P-Glycoproteinmediated resistance to Hsp90-directed therapy is eclipsed by the heat shock response. Cancer Res 2008;68:7419. 12. Benchekroun MN, Schneider E, Safa AR, et al. Mechanisms of resistance to ansamycin antibiotics in human breast cancer cell lines. Mol Pharmacol 1994;46:677. 13. Lee JM, Bernstein A. Apoptosis, cancer and the p53 tumor suppressor gene. Cancer Metastasis Rev 1995;14:149. 14. Lang GA, Iwakuma T, Suh YA, et al. Gain of function of a p53 hot spot mutation in a mouse model of Li-Fraumeni syndrome. Cell 2004;119:861.
153
15. Blagosklonny MV, Toretsky J, Bohen S, et al. Mutant conformation of p53 translated in vitro or in vivo requires functional HSP90. Proc Natl Acad Sci USA 1996;93:8379. 16. Lowe SW, Ruley HE, Jacks T, et al. p53-dependent apoptosis modulates the cytotoxicity of anticancer agents. Cell 1996;74:957. 17. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970;227:680. 18. Mosmann T. Rapid colorimetric assay for cellular growth and survival: Application to proliferation and cytotoxicity assays. J Immunol Methods 1983;65:55. 19. Chou TC, Talalay P. Quantitative analysis of dose-effect relationships: The combined effects of multiple drugs or enzyme inhibitors. Adv Enzyme Regul 1984;22:27. 20. Weinstein JN, Myers TG, O’Connor PM, et al. An informationintensive approach to the molecular pharmacology of cancer. Science 1997;275:343. 21. Zha J, Harada H, Yang E, et al. Serine phosphorylation of death agonist BAD in response to survival factor results in binding to 14-3-3 not BCL-X(L). Cell 1996;87:619. 22. Huang Y, Blower PE, Liu R, et al. Chemogenomic analysis identifies geldanamycins as substrates and inhibitors of ABCB1. Pharm Res 2007;24:1702. 23. Teraishi F, Wu S, Sasaki J, et al. P-glycoprotein-independent apoptosis induction by a novel synthetic compound, MMPT [5-[(4-methylphenyl)methylene]-2-(phenylamino)-4(5H)-thiazolone]. J Pharmacol Exp Ther 2005;314:355. 24. Arumugam T, Ramachandran V, Fournier KF, et al. Epithelial to mesenchymal transition contributes to drug resistance in pancreatic cancer. Cancer Res 2009;69:5820. 25. Wilhelm SM, Carter C, Tang L, et al. BAY 43-9006 exhibits broad spectrum oral antitumor activity and targets the RAF/ MEK/ERK pathway and receptor tyrosine kinases involved in tumor progression and angiogenesis. Cancer Res 2004;64:7099. 26. Powers MV, Workman P. Inhibitors of the heat shock response: Biology and pharmacology. FEBS Lett 2007;581:3758.
Multiple Kinase Pathways Involved in the Different De Novo Sensitivity of Pancreatic Cancer Cell Lines to 17-AAG Heping Liu, Ph.D.,* Ti Zhang, M.D., Ph.D.,§ Rong Chen, Ph.D.,† David J. McConkey, Ph.D.,‡ John F. Ward, M.D.,‡ and Steven A. Curley, M.D.*,1 *Department of Surgical Oncology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas; †Department of Experimental Therapeutics, The University of Texas M. D. Anderson Cancer Center, Houston, Texas; ‡Department of Urology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas; and §Department of Hepatobiliary Surgery, Tianjin Cancer Hospital, Tianjin Medical University, Tianjin, China Originally submitted April 22, 2011; accepted for publication September 9, 2011
Background. 17-Allylamino-17-demethoxygeldanamycin (17-AAG) specifically targets heat shock protein (HSP)90 and inhibits its chaperoning functions for multiple kinases involved in cancer cell growth and survival. To select responsive patients, the molecular mechanisms underlying the sensitivity of cancer cells to 17-AAG must be elucidated. Materials and Methods. We used cytotoxicity assays and Western blotting to explore the effects of 17-AAG and sorafenib on cell survival and expression of multiple kinases in the pancreatic cancer cell lines AsPC-1 and Panc-1. Gene cloning and transfection, siRNA silencing, and immunohistochemistry were used to evaluate the effects of mutant p53 protein on 17-AAG sensitivity. Results. AsPC-1 and Panc-1 responded differently to 17-AAG, with half maximal inhibitory concentration (IC50) values of 0.12 and 3.18 mM, respectively. Comparable expression of HSP90, HSP70, and HSP27 was induced by 17-AAG in AsPC-1 and Panc-1 cells. P-glycoprotein and mutant p53 did not affect 17-AAG sensitivity in these cell lines. Multiple kinases are more sensitive to HSP90 inhibition in AsPC-1 than in Panc1 cells. After 17-AAG treatment, p-Bad (S112) decreased in AsPC-1 cells and increased in Panc-1 cells. Sorafenib markedly increased p-Akt, p-ERK1/2, p-GSK-3b, and p-S6 in both cell lines. Accordingly, 17-AAG and sorafenib acted antagonistically in AsPC-1 and Panc-1 cells, except at high concentrations in AsPC-1 cells. Conclusions. Differential inhibition of multiple kinases is responsible for the different de novo sensitiv1 To whom correspondence and reprint requests should be addressed at Department of Surgical Oncology, Unit 1484, The University of Texas M. D. Anderson Cancer Center, 1400 Pressler Street, Houston, TX 77030. E-mail: [email protected].
ity of AsPC-1 and Panc-1 cells to HSP90 inhibition. P-glycoprotein and mutant p53 protein did not play a role in the sensitivity of pancreatic cancer cells to 17-AAG. Ó 2012 Elsevier Inc. All rights reserved. Key Words: 17-AAG; pancreatic cancer; chemosensitivity; kinase.
INTRODUCTION
Heat shock proteins (HSPs) constitute a large family of proteins that have been classified according to their molecular weight. For example, HSP70 is the 70-kDa HSP; other family members are HSP27, HSP60, and HSP90. HSPs, especially HSP70, can be induced by stress such as heat or cold shock, cytotoxic drugs, and radiation. The 90-kDa chaperone protein, HSP90, is highly conserved and ubiquitously expressed in all living organisms. HSP90 requires several co-chaperones, such as Cdc37, to form a highly complicated complex that regulates the folding/unfolding, assembly, and maturation of its client proteins [1]. HSP90 regulates more than 200 cellular client proteins. Among them are serine/threonine or tyrosine kinases, such as c-Raf-1, PKB/Akt, Bcr-Abl, CDK4, and v-Src, which are crucial to proliferation, differentiation, and survival of cancer cells [1]. The ansamycin antibiotic geldanamycin is a novel anticancer drug that can specifically bind to an ATP/ADP pocket in the N-terminal domain of HSP90 and inhibit HSP90 function by preventing full chaperone cycling [2]. Treatment with geldanamycin results in the accumulation of unprocessed chaperone–client protein complexes within cells, and the premature release and
147
0022-4804/$36.00 Ó 2012 Elsevier Inc. All rights reserved.
148
JOURNAL OF SURGICAL RESEARCH: VOL. 176, NO. 1, JULY 2012
degradation of vital cellular proteins in the proteasome leads to cell death [3, 4]. Preclinical experiments led to the surprising observation that cancer cells, because of their so-called ‘‘kinase addiction,’’ are significantly more sensitive to HSP90 inhibition than nontransformed cells are [5, 6]. This finding raised the possibility that HSP90 inhibitors could be used in cancer patients as new chemotherapeutic agents. A geldanamycin derivative with a more favorable toxicity profile, 17-allylamino-17-demethoxy-geldanamycin (17-AAG), has entered clinical phase I/II trials for breast cancer, myeloma, and metastatic melanoma after being tested in xenograft and transgenic mouse models [7–9]. The membrane transporter P-glycoprotein (also called ABCB1 or MDR1) functions as a drug efflux pump that actively transports drugs from the inside to the outside of cells and leads to a decreased intracellular accumulation of drugs needed to kill cancer cells. P-glycoprotein is known to be involved in cancer cell resistance to many chemotherapeutic agents [10]. Several studies have also shown that P-glycoprotein was involved in the resistance of tumor cells to geldanamycin or its derivatives [11, 12]. Inactivating p53 mutations are the most common genetic alterations found in human cancers. Unlike mutations in other tumor suppressor genes, most p53 mutations are missense mutations, of which about 97% occur within the sequence-specific DNA binding domain. Accumulation of mutant p53 protein may lead to novel biological effects in cancer cells, consistent with a gain-of-function phenotype [13, 14]. Moreover, mutant p53 protein is a HSP90 client protein [15] and may modulate the responsiveness of cancer cells to chemotherapeutic drugs [16]. To accurately predict clinical response to 17-AAG and to select patients who will benefit from 17-AAG in future clinical use, it is important to elucidate the molecular mechanisms underlying the de novo sensitivity of cancer cells to 17-AAG or similar HSP90 inhibitors. Pancreatic cancer is notorious for lack of response to conventional chemotherapy. In the current study, we sought to determine the genetic determinants that regulate the sensitivity of AsPC-1 and Panc-1 cells, two human pancreatic cancer cell lines, to 17-AAG.
MATERIALS AND METHODS Cell Lines and Reagents The pancreatic cancer cell lines AsPC-1 and Panc-1 were obtained from the American Type Culture Collection (Manassas, VA) and grown in Dulbecco’s modified Eagle’s medium containing 10% fetal bovine serum, 2 mM glutamine, and 1 mM sodium pyruvate in a 37 C humidified incubator with 5% CO2. 17-AAG was purchased from Sigma (St. Louis, MO) and dissolved in methanol to make 5-mM stock. Nexavar (sorafenib) pills (Bayer HealthCare Pharmaceuticals) were
obtained from the M. D. Anderson Cancer Center pharmacy and dissolved in DMSO to make 10-mM stock.
Western Blotting Whole-cell lysates were separated in either 7.5% or 10% denaturing polyacrylamide gel , depending on the molecular sizes of the proteins, as described previously [17], and transferred by electroblotting to nitrocellulose membranes (Bio-Rad Laboratories, Hercules, CA). Primary antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA), CalBiochem (San Diego, CA), and Cell Signaling Technology (Beverly, MA). The membranes were incubated with the appropriate primary antibodies, and specific immunoreactivity was detected with fluorescently labeled secondary antibodies (Alexa Fluor 680 [Invitrogen, Carlsbad, CA] and IRDye 800CW [LI-COR Biosciences, Lincoln, NE]) using the Odyssey imager and software ver. 3 (LI-COR). To confirm approximately equal loading, the membranes were also incubated with anti-b-actin antibody.
Cytotoxicity Assay Cytotoxicity was assessed with the 3-(4,5 dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide (MTT) (Trevigen, Gaithersburg, MD) colorimetric assay [18]. Cells were plated at 2–3 3 103 per well in 96-well plates in triplicate and allowed to adhere overnight. Cells were then treated with appropriate drugs at varying concentrations for 72 h at 37 C. MTT was added to each well, and incubation was continued at 37 C for 4 h. Cell viability was measured by dissolving the cellular reduced product of MTT in 200 mL of DMSO and reading at 570 nm with the FLUOstar Omega microplate reader (BMG Labtech, Chicago, IL). Each test was performed at least three times.
Statistical Analysis Half maximal inhibitory concentration (IC50) values were calculated with SigmaPlot ver. 10.0.1 (Systat Software, San Jose, CA). The two-tailed Student’s t-test was used for unpaired comparisons. To characterize synergistic or antagonistic interactions between agents, we performed median dose–effect analysis with CalcuSyn software (Biosoft, Ferguson, MO) [19]. Data are presented as mean 6 SD of at least three replicate experiments.
RESULTS Effects of HSP90 Inhibition on Tumor Cell Survival and Multiple Kinase Pathways
To determine the responses of two human pancreatic cancer cell lines, AsPC-1 and Panc-1, to 17-AAG, we used MTT assays. The sensitivity of Panc-1 cells was substantially dependent on the confluence of the cell cultures, so Panc-1 cells exhibited confluencedependent resistance (CDR) (Fig. 1A). AsPC-1 cells, however, did not exhibit substantial CDR. To compare the sensitivity of the two cell lines to 17-AAG, we grew the cells to 60%–80% confluence before plating them for cytotoxicity assays. We found that Panc-1 cells were much more resistant to 17-AAG than AsPC-1 cells were, consistent with the results of a previous study [20]; Panc-1 and AsPC-1 cells had IC50 values of 3.18 and 0.12 mM, respectively (Fig. 1B). The different responses of AsPC-1 and Panc-1 cells to 17-AAG may be due to the different sensitivities of
LIU ET AL.: MULTIPLE KINASES REGULATE CANCER SENSITIVITY TO 17-AAG
149
FIG. 1. Multiple kinase signaling pathways are more sensitive to 17-AAG in AsPC-1 cells than in Panc-1 cells. (A), (B) We used MTT cell survival assays to determine the cytotoxic effect of 17-AAG on AsPC-1 and Panc-1 cells after 72 h of treatment. Each experiment was performed in triplicate. Bars, SD. (A) Panc-1 cells exhibited CDR. (B) Sensitivity was determined in the logarithmic growth phase (60%–80% confluence), and IC50 values were 0.12 mM for AsPC-1 cells and 3.18 mM for Panc-1 cells. (C) Cell lysates were prepared after 24 h of treatment with 17-AAG at different concentrations and separated on polyacrylamide gel. The membrane was probed with the indicated antibodies. Lanes 1–6, AsPC-1 cells treated with 17-AAG at 0, 0.05, 0.1, 0.2, 0.4, and 0.8 mM; lanes 7–12, Panc-1 cells treated with 17-AAG at 0, 0.4, 0.8, 1.6, 3.2, and 6.4 mM.
multiple kinases to HSP90 inhibition. We determined the sensitivities of these kinases to HSP90 inhibition by Western blotting after 24 h of treatment with varying concentrations of 17-AAG. Although the total amounts of the kinases were unaffected, the levels of phosphorylated ERK1/2 (p-ERK1/2) and p-MEK1/2, in the MAP kinase pathway, were much more sensitive to 17-AAG in AsPC-1 cells than in Panc-1 cells. Similarly, levels of p-S6 ribosomal protein and phosphorylated glycogen synthase kinase-3b (p-GSK-3b) were much more sensitive in AsPC-1 cells than in Panc-1 cells. In contrast, levels of p-Bad (S112) changed in opposite directions in the two cell lines with increasing 17-AAG concentration, decreasing in AsPC-1 cells and increasing in Panc-1 cells, whereas p-Bad (S136) was almost undetectable in both AsPC-1 and Panc-1 cells (Fig. 1C). p-Bad is sequestered in the cytosol by binding to 14-3-3 to promote cell survival [21]; the differential changes of p-Bad (S112) levels after 17-AAG treatments might be responsible for the different cytotoxic effects of HSP90 inhibition in AsPC-1 and Panc-1 cells. In addition, CDK4 levels were more sensitive to 17-AAG in AsPC-1 cells than in Panc-1 cells (data not shown). Expression of Heat Shock Proteins, not P-Glycoprotein, Is Induced by 17-AAG Treatment
Because 17-AAG inhibits HSP90 chaperoning function [2], we determined the expression levels of various HSPs in AsPC-1 and Panc-1 cells. We found that the
expression levels of the three major HSPs, HSP27, HSP70, and HSP90, were comparable in AsPC-1 and Panc-1 cells without 17-AAG treatment. After 72 h of 17-AAG treatment, expression of HSP70 was strongly induced, whereas expression of HSP27 and HSP90 was only moderately induced (Fig. 2A). Induction of HSP70 expression was more pronounced in Panc-1 cells (6.63 the level without 17-AAG treatment) than in AsPC-1 cells (2.43 the untreated level) (Fig. 2C and D). P-glycoprotein has been reported to mediate the resistance of cancer cells to 17-AAG [11, 12, 22]. However, after 72 h of treatment with 17-AAG at indicated concentrations, P-glycoprotein was not detectable in AsPC-1 or Panc-1 cells with Western blotting (Fig. 2B), whereas the positive control showed strong P-glycoprotein expression. The positive control H460/ TaxR was prepared from a lung cancer cell line, H460, with acquired paclitaxel resistance established by repeated treatment with Taxol [23]. On the basis of our microarray data for pancreatic cancer cell lines [24], we used a similar pharmacogenomic approach to determine whether there was a correlation between 17-AAG sensitivity and mRNA expression levels of P-glycoprotein in nine pancreatic cancer cell lines (MIAPaCa-2, BxPC-3, AsPC-1, L3.6p1, Hs766T, MPanc96, SU86.86, CFPAC, and Panc-1); we did not identify any association of P-glycoprotein mRNA levels with 17-AAG resistance (data not shown). Panc-1 cells harbor a missense mutation in p53 DNA binding domain and express high levels of mutant p53
150
JOURNAL OF SURGICAL RESEARCH: VOL. 176, NO. 1, JULY 2012
FIG. 2. Expression of HSP27, HSP70, and HSP90, but not of P-glycoprotein, is induced by the HSP90 inhibitor 17-AAG. (A), (C), (D) Cellular extracts were prepared and treated with the indicated concentrations of 17-AAG for 72 h. Expression levels of the three major HSPs (HSP27, HSP70, and HSP90) were determined by probing with their corresponding antibodies. (A) Compared with b-actin, we found that expression of HSP70 was strongly induced, whereas expression of HSP27 and HSP90 was only moderately induced. Signal intensities were calculated for (C) AsPC-1 and (D) Panc-1 cells. (B) Expression of P-glycoprotein (MDR1) was not detectable in AsPC-1 or Panc-1 cells with or without 17-AAG treatment. A positive control, TaxR, was included in this Western blot.
protein, whereas AsPC-1 cells harbor a frameshift (null) mutation in p53 with no detectable expression of mutant p53 protein (data not shown). We used p53 siRNA and expression vectors and found that mutant p53 protein did not have any impact on the sensitivity of Panc-1 and AsPC-1 cells to 17-AAG (Figs. S1 and S2, Supplemental Data). Sorafenib (Nexavar), a Multiple Kinase Inhibitor, Regulates 17-AAG Sensitivity
Sorafenib, originally developed as a Raf kinase inhibitor, is a multiple kinase inhibitor [25]. We found that sorafenib could paradoxically up-regulate the levels of multiple phosphorylated kinase substrates in AsPC-1 and Panc-1 cells as determined by Western blotting after 24 h of 17-AAG treatment (Fig. 3). Thus, combined treatment with 17-AAG and sorafenib had an antagonistic effect on multiple kinase pathways. We found that levels of p-ERK1/2 (T202/Y204), p-Akt (S473), p-S6 (S235/236), and p-GSK-3b were elevated after combined treatment with 17-AAG and sorafenib for 24 h. The total levels of most of these proteins were unaffected, but total Akt decreased markedly at increasing concentrations of 17-AAG and sorafenib in AsPC-1 cells. Sorafenib treatment also shifted multiple lowmolecular-weight isoforms of B-Raf to high-molecularweight ones in AsPC-1 cells, with no apparent change in Panc-1 cells (Fig. 3).
To further confirm our observation that the inherent differences in multiple kinases regulated the sensitivity of AsPC-1 and Panc-1 cells to 17-AAG, we extended our Western blot determinations to cytotoxicity assays. AsPC-1 and Panc-1 cells had very similar responses to sorafenib, with IC50 values of about 11.82 mM. At 5 mM, sorafenib seemed to decrease the sensitivity of Panc-1 cells to 17-AAG, increasing IC50 by almost a factor of 5, while not significantly altering the sensitivity of AsPC-1 cells to 17-AAG (Fig. 4A and B). To clarify the antagonistic interaction of 17-AAG and sorafenib, we determined the combination index (CI) in AsPC-1 and Panc-1 cells according to Chou and Talalay [19]. Consistent with our results from Western blotting, we found that 17-AAG and sorafenib mostly had an antagonistic effect (CI > 1), except that they acted synergistically (CI < 1) at high concentrations in AsPC-1 cells (Fig. 4C and D). The change in p-HSP90 (T4/5) as a percentage of total HSP90 in AsPC-1 cells and Panc-1 cells after 24 h of treatment with 17-AAG and sorafenib showed a similar pattern with our Western blotting and CI determinations: the percentage decreased with increasing concentration of 17-AAG and sorafenib, except at the highest concentration of 17-AAG in AsPC-1 cells (Fig. 4E and F). DISCUSSION
Chemosensitivity is likely to be an inherent, complex phenotype, with genetic polymorphisms, protein
LIU ET AL.: MULTIPLE KINASES REGULATE CANCER SENSITIVITY TO 17-AAG
151
FIG. 3. Sorafenib antagonizes the effects of 17-AAG on multiple kinase pathways. Cell lysates were prepared after 24 h of treatment with 17-AAG and sorafenib at the indicated concentrations and separated on polyacrylamide gel. The membrane was probed with the indicated antibodies. The treatments had no effect on the level of the co-chaperone molecule CDC37 or its activated form, p-CDC37. However, 17-AAG and sorafenib acted antagonistically on the Akt(p-Akt), Wnt (p-GSK3b), and MAP kinase (p-ERK1/2) signaling pathways, more markedly so in AsPC-1 cells than in Panc-1 cells. 17-AAG and sorafenib had no effect on the total amounts of the indicated proteins, except that they markedly decreased the level of total Akt at the highest concentrations (2 3 IC50) in AsPC-1 cells. A ¼ 17-AAG; S ¼ sorafenib. The numbers indicate concentrations in mM.
expression alterations, and post-translational modifications playing significant roles. In this study, two pancreatic cancer cell lines, AsPC-1 and Panc-1, provided us with an excellent model to explore the molecular determinants of sensitivity to 17-AAG, a novel anticancer agent that inhibits the chaperoning function of HSP90 [6]. In the logarithmic growth phase (60%–80% confluence), AsPC-1 cells were much more sensitive to 17AAG than Panc-1 cells. Many kinases responsible for cell proliferation and survival are client proteins of HSP90. Targeting HSP90 with 17-AAG leads to degradation of these kinases and cell death [9]. Thus, the difference in the responsiveness of the multiple kinases to HSP90 inhibition might be the basis for the different cytotoxic effects of 17-AAG on AsPC-1 and Panc-1 cells. Our Western blotting results indeed showed that after 24 h of 17-AAG treatment, several key phosphorylated kinases (ERK1/2, MEK1/2, and GSK-3b) decreased to much lower levels in AsPC-1 cells than in Panc-1 cells. Growth factors can stimulate Bad phosphorylation, which suppresses cell apoptosis and promotes cell survival [21]. We found that HSP90 inhibition resulted in decreased p-Bad (S112) levels in AsPC-1 cells but increased levels in Panc-1 cells. p-Bad (S112) levels may be a predictive factor for the sensitivity of pancreatic cancer cells to 17-AAG, although the underlying mechanism for this effect remains to be elucidated.
As cytotoxic drugs, HSP90 inhibitors can activate the heat shock response [11, 26]. Our results also revealed that treatment with 17-AAG induced expression of the three major HSPs, HSP90, HSP70, and HSP27, especially HSP70 (Fig. 2A), which increased to 6.63 and 2.43 the level without 17-AAG treatment in Panc-1 and AsPC-1 cells, respectively. The stronger heat shock response in Panc-1 cells might be attributable to 17AAG resistance because induction of stress response proteins, such as HSP27 and HSP70, by HSP90 inhibition could offset the cytotoxic effects of 17-AAG [11]. In the microarray profiling of 60 human tumor cell lines (NCI-60), their sensitivity to geldanamycin and its analogs displayed negative correlation with mRNA expression levels of P-glycoprotein, suggesting a role of P-glycoprotein in chemoresistance [22]. However, our microarray data of nine pancreatic cancer cell lines did not show any association between 17-AAG sensitivity and P-glycoprotein mRNA expression levels. Moreover we did not detect any protein expression of P-glycoprotein in AsPC-1 or Panc-1 cells, indicating that P-glycoprotein was not involved in the responsiveness of AsPC-1 and Panc-1 cells to 17-AAG. Mutant p53 protein in AsPC-1 and Panc-1 cells had no impact on responsiveness to HSP90 inhibition. We found that sorafenib, a multiple kinase inhibitor, could paradoxically up-regulate the phosphorylated
152
JOURNAL OF SURGICAL RESEARCH: VOL. 176, NO. 1, JULY 2012
FIG. 4. Sorafenib and 17-AAG have antagonistic effects in AsPC-1 and Panc-1 cells. (A), (B) AsPC-1 and Panc-1 cells had similar responses to sorafenib, with IC50 values of about 11.82 mM. Sorafenib at 5 mM had a negligible impact on the sensitivity of AsPC-1 cells but greatly increased the resistance of Panc-1 cells to 17-AAG, according to the IC50 values. Bars, SD. The CI curves of sorafenib and 17-AAG for (C) AsPC-1 cells and (D) Panc-1 cells, simulated with CalcuSyn software, showed antagonism, except at high concentrations in AsPC-1 cells. Red squares indicate actual measurements. (E), (F) p-HSP90 (T4/5) as a percentage of total HSP90 changed in a similar way in AsPC-1 and Panc-1 cells after 17-AAG and sorafenib treatments, except at the highest concentration of 17-AAG in AsPC-1 cells. (F) Signal intensities calculated from (E). 1–3, 17-AAG treatments; 4–6, sorafenib treatments.
components in multiple kinase pathways. To confirm our conclusion that the sensitivity of AsPC-1 and Panc-1 cells to HSP90 inhibitors was related with their intrinsic multiple kinase pathways, we determined the combinatorial effects of 17-AAG and sorafenib on both AsPC-1 and Panc-1 cells. As expected, sorafenib, at a low concentration (5 mM), greatly increased the resistance of Panc-1 cells to 17-AAG, according to the IC50 values, but had a negligible impact on the sensitivity of AsPC-1 cells to 17-AAG. The CIs of 17-AAG and sorafenib also indicated that they acted antagonistically in AsPC-1 and Panc-1 cells, except at high concentrations in AsPC-1 cells. At these concentrations, combined 17AAG and sorafenib treatments markedly reduced total Akt levels in AsPC-1 cells, a likely explanation for the synergism of 17-AAG and sorafenib in AsPC-1 cells. Thus, these data further demonstrate that HSP90 regulates the kinase cascades differently in AsPC-1 cells than in Panc-1 cells and that this difference is responsible for the different cytotoxicity of 17-AAG in these two cell lines. Although the molecular mechanism underlying the different HSP90 functions in AsPC-1 and Panc-1 cells remains elusive at present, our research
represents an important and significant attempt in this area. Our data may be of clinical application for predicting the patients’ responses to HSP90 inhibition. ACKNOWLEDGMENTS The authors thank Yue Lu for help in preparing the manuscript. Yafang Li helped with some statistical analyses and the CI calculation. Dr. Bingliang Fang kindly provided the positive control (H460/ TaxR) for the P-glycoprotein. The authors also thank Karen R. Muller, Kristine K. Ash, and Yolanda Brittain for editing and submitting this manuscript. This work was supported by the Kanzius Cancer Research Foundation.
SUPPLEMENTARY DATA
Supplementary data related to this article can be found online at doi: 10.1016/j.jss.2011.09.017. REFERENCES 1. Blume-Jensen P, Hunter T. Oncogenic kinase signaling. Nature 2001;411:355. 2. Whitesell L, Mimnaugh EG, De Costa B, et al. Inhibition of heat shock protein HSP90-pp60v-src heteroprotein complex formation by benzoquinone ansamycins: Essential role for stress proteins in oncogenic transformation. Proc Natl Acad Sci U S A 1994;91:8324.
LIU ET AL.: MULTIPLE KINASES REGULATE CANCER SENSITIVITY TO 17-AAG 3. Sepp-Lorenzino L, Ma Z, Lebwohl DE, et al. Herbimycin A induces the 20 S proteasome- and ubiquitin-dependent degradation of receptor tyrosine kinases. J Biol Chem 1995;270:16580. 4. Mimnaugh EG, Chavany C, Neckers L. Polyubiquitination and proteasomal degradation of the p185c-erbB-2 receptor proteintyrosine kinase induced by geldanamycin. J Biol Chem 1996; 271:22796. 5. Kamal A, Boehm MF, Burrows FJ. Therapeutic and diagnostic implications of Hsp90 activation. Trends Mol Med 2004; 10:283. 6. Neckers L. Using natural product inhibitors to validate Hsp90 as a molecular target in cancer. Curr Top Med Chem 2006;6:1163. 7. Modi S, Stopeck AT, Gordon MS, et al. Combination of trastuzumab and tanespimycin (17-AAG, KOS-953) is safe and active in trastuzumab-refractory HER-2 overexpressing breast cancer: A phase I dose-escalation study. J Clin Oncol 2007;25:5410. 8. Anderson KC, Hannah AL, Pazdur R, et al. A strategic framework for novel drug development in multiple myeloma. Br J Haematol 2007;138:153. 9. Solit DB, Osman I, Polsky D, et al. Phase II trial of 17allylamino-17-demethoxygeldanamycin in patients with metastatic melanoma. Clin Cancer Res 2008;14:8302. 10. Endicott JA, Ling V. The biochemistry of P-glycoprotein-mediated multidrug resistance. Annu Rev Biochem 1989;58:137. 11. McCollum AK, TenEyck CJ, Stensgard B, et al. P-Glycoproteinmediated resistance to Hsp90-directed therapy is eclipsed by the heat shock response. Cancer Res 2008;68:7419. 12. Benchekroun MN, Schneider E, Safa AR, et al. Mechanisms of resistance to ansamycin antibiotics in human breast cancer cell lines. Mol Pharmacol 1994;46:677. 13. Lee JM, Bernstein A. Apoptosis, cancer and the p53 tumor suppressor gene. Cancer Metastasis Rev 1995;14:149. 14. Lang GA, Iwakuma T, Suh YA, et al. Gain of function of a p53 hot spot mutation in a mouse model of Li-Fraumeni syndrome. Cell 2004;119:861.
153
15. Blagosklonny MV, Toretsky J, Bohen S, et al. Mutant conformation of p53 translated in vitro or in vivo requires functional HSP90. Proc Natl Acad Sci USA 1996;93:8379. 16. Lowe SW, Ruley HE, Jacks T, et al. p53-dependent apoptosis modulates the cytotoxicity of anticancer agents. Cell 1996;74:957. 17. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970;227:680. 18. Mosmann T. Rapid colorimetric assay for cellular growth and survival: Application to proliferation and cytotoxicity assays. J Immunol Methods 1983;65:55. 19. Chou TC, Talalay P. Quantitative analysis of dose-effect relationships: The combined effects of multiple drugs or enzyme inhibitors. Adv Enzyme Regul 1984;22:27. 20. Weinstein JN, Myers TG, O’Connor PM, et al. An informationintensive approach to the molecular pharmacology of cancer. Science 1997;275:343. 21. Zha J, Harada H, Yang E, et al. Serine phosphorylation of death agonist BAD in response to survival factor results in binding to 14-3-3 not BCL-X(L). Cell 1996;87:619. 22. Huang Y, Blower PE, Liu R, et al. Chemogenomic analysis identifies geldanamycins as substrates and inhibitors of ABCB1. Pharm Res 2007;24:1702. 23. Teraishi F, Wu S, Sasaki J, et al. P-glycoprotein-independent apoptosis induction by a novel synthetic compound, MMPT [5-[(4-methylphenyl)methylene]-2-(phenylamino)-4(5H)-thiazolone]. J Pharmacol Exp Ther 2005;314:355. 24. Arumugam T, Ramachandran V, Fournier KF, et al. Epithelial to mesenchymal transition contributes to drug resistance in pancreatic cancer. Cancer Res 2009;69:5820. 25. Wilhelm SM, Carter C, Tang L, et al. BAY 43-9006 exhibits broad spectrum oral antitumor activity and targets the RAF/ MEK/ERK pathway and receptor tyrosine kinases involved in tumor progression and angiogenesis. Cancer Res 2004;64:7099. 26. Powers MV, Workman P. Inhibitors of the heat shock response: Biology and pharmacology. FEBS Lett 2007;581:3758.