Correlation of expression of P-glycoprotein and inhibitor of apoptosis proteins to chemosensitivity in gastrointestinal carcinoma tissues

Transcription

1 [Chinese Journal of Cancer 27:11, Construction ; of a November eukaryotic 2008]; vector 2008 expressing Sun Yat-sen MASPIN University ene and its Cancer effect Center on cell apoptosis in gastric cancer cell line SGC7901 Basic Research Paper Correlation of expression of P-glycoprotein and inhibitor of apoptosis proteins to chemosensitivity in gastrointestinal carcinoma tissues Jie Han,* Bi-Bo Tan, Wei Geng, Bing-Rong Lv, Jian-Hui Zhao Department of Gastrointestinal Surgery, Hebei Provincial People s Hospital, Shijiazhuang, Hebei P.R. China Key words: gastrointestinal neoplasm, multidrug resistance, inhibitor of apoptosis proteins, P-glycoprotein, chemosensitivity test Background and Objective: P-glycoprotein (P-gp) mediated classical drug resistance and inhibition of the apoptotic pathway are the two mostly investigated mechanisms of multidrug resistance (MDR). Coexpression and interaction of MDR-related factors result in pleiotropic drug resistance in cancer cells. This study was to investigate the correlation of expressions of multiple MDR-related factors, such as P-gp, p53, Survivin or bcl-2, to chemosensitivitity in gastrointestinal carcinomas. Methods: Eighty-four tissue specimens of gastrointestinal carcinomas were analyzed. Expressions of P-gp, p53, Survivin and bcl-2 were determined by immunohistochemistry (IHC). Drug chemosensitivity of nine drugs to cancer cells were measured by MTT assay. Results: The positive staining of P-gp, p53, Survivin and bcl-2 were detected in 96.4%, 64.3%, 89.3% and 60.7% of all specimens, respectively. The expression of P-gp and bcl-2 (r = , p < 0.05), and the expression of survivin and bcl-2 (r = , p < 0.05) were positively correlated. The inhibition rates of paclitaxel (PTX), oxaliplatin (OXA) or cisplatin (DDP) on cancer cells were significantly lower in the group with strong P-gp expression than that with weak P-gp expression (all p < 0.05). The strong expression of p53 was correlated with decreased inhibition rates of PTX and DDP on cancer cells (p < 0.05, p < 0.01). When the expression of Survivin was increased, the inhibition rates of vincristine (VCR) or DDP on cancer cells were reduced significantly (p < 0.05, p < 0.01), but the inhibitory effect of OXA was remarkably increased (P < 0.01). The inhibition rates of 5-fluorouracil (5-FU), VCR, epirubicin (EADM) and OXA on cancer cells were lower in the group with strong *Correspondence to: Jie Han; Department of Gastrointestinal Surgery, Hebei Provincial People s Hospital, Shijiazhuang, Hebei, , P.R. China; Tel.: ; hanjie01@yahoo.com.cn Submitted: 04/08/08; Revised: 07/21/08; Accepted: 08/11/08 This paper was translated into English from its original publication in Chinese. Translated by: Beijing Xinglin Meditrans Center ( and Hua He on 10/20/08. The original Chinese version of this paper is published in: Ai Zheng(Chinese Journal of Cancer), 27(11); Previously published online as a Chinese Journal of Cancer E-publication: expression of bcl-2 than in that with weak expression of bcl-2 (p < 0.05). Conclusions: The expression of MDR-related factors in gastrointestinal carcinomas is associated with drug resistance of only certain chemotherapy drugs. Multiple factors and mechanisms should be considered when assessing the influence of MDR related factors on drug resistance. It is proven that multidrug resistance (MDR) in tumor involves different mechanisms and different checkpoints of cytotoxicity, including reduction in the accumulation of the drug within the cell until resistance against apoptosis is developed. Commonly, several mechanisms are intertwined and simultaneous expression of multiple genes or proteins in relation to MDR is observed in tumor cells. Currently, studies on tumor MDR are mostly performed in MDR cells. There are reports that the expression levels of MDR factors are used to monitor individualized chemotherapy clinically. However, there are few reports on the relationship between the expression of MDR-related factors in gastrointestinal tumors and tumor sensitivity towards chemotherapeutic drugs. 1-3 This paper investigated the correlation of expressions of P-glycoprotein (P-gp), apoptotic inhibitory proteins, such as p53, Survivin, and bcl-2, to chemosensitivity in gastrointestinal carcinoma. Materials and Methods Samples. Fresh tumor samples were taken from 84 patients with gastrointestinal carcinoma, who were diagnosed surgically and pathologically in Hebei General Hospital from October 2006 to December Among them, 50 cases were gastric carcinoma and 34 cases were colorectal carcinoma. There were 51 males and 33 females, aged between 22 to 80 years, with a medium age of 62 years. None of them received any treatment prior to surgery. Fresh tumor tissues with a size of 1.0 cm x 0.5 cm x 0.5 cm were taken after resection, and were kept in culture media containing double antibodies (100 µ/ml penicillin and 200 µg/ml streptomycin), to conduct MTT assay later. In addition, each tumor sample was fixed in neutral formaldehyde and routinely embedded into the paraffin block before being sliced for immunohistochemistry. Immunohistochemistry. The rat anti-human P-gp, p53, survivin, and bcl-2 monoclonal antibodies and the immunohistochemistry kit was purchased from Beijing Zhongshan Chinese Journal of Cancer 423

2 Biotechnology Company (ZYMED Company, USA). The tissues were stained by the SP method to measure expressions of P-gp, p53, survivin, and bcl-2. The staining procedure was strictly performed according to the instructions of the manual. PBS was used in replace of primary antibodies as negative control, while known positive slides were used as positive control. Results were determined by two pathologists double-blindedly. Five highmagnification fields of view (400 ) were randomly selected from each slide, in which 100 cells were counted. The results were judged based on the staining intensity and the number of positive cells. Staining intensity was scored as 0 (no detectable stain), 1 (light yellow), 2 (yellow), and 3 (brownish yellow). The percentage of positive cells was scored as 0 ( 5%), 1 (6 25%), 2 (26 50%), 3 (51 75%), and 4 ( 76%). The product of the two scores was regarded as the final score to classify the results into 1, negative (-), 2 3, weak positive (+), 4 7, (++), positive, and 8, strong positive (+++). Scores 3 was regarded as weak expression, and scores 4 was strong expression. Measurement of drug sensitivity in vitro. DMEM and pepsin were the products of GIBCO Company, USA. MTT kit was purchased from Sigma Company, USA. The cell culture plate was the licensed product of Tailun Biotech Company (ZL X) and the enzyme-labeling instrument DNM-9602A was purchased from Beijing Perlong Company. Tumor tissues were ground and the cell concentration was adjusted to /ml using a cell isolator. The plate and DMEM were pre-heated to 37. Nine test drugs were 5-Fluorouracil (5-FU), Vincristine (VCR), Etoposide (VP-16), Pentoxifylline (PTX), Epirubicin (EADM), OPT, Oxaliplatin (OXA), Cisplatin (DDP), and Methotrexate (MTX), and the final drug concentration was the peak drug concentration in plasma. 4 Each treatment were performed in triplicate, while cells without treatment were set as control. Cell suspension (200 µl) was seeded into the well, gently shaken, and incubated at 37, with 5% CO 2 and saturated humidity for 48 h. A volume of 20 µl MTT (5 mg/ml) was added into the well and was cultured for another 3 24 h. Following centrifugation at 177 g for 10 min, the supernatant was discarded and the residue liquid was removed. For each well, 100 µl DMSO was added and gently shaken again. The absorption value (A) of each well was measured at the wavelength of 570 nm using the enzymelabeling instrument. The computer software, which was provided by the Tailun Biotech Company, was used to calculate The average inhibitory rate (IR) on tumor growth was calculated based on the following formula: (1 average A value of the each treatment group/ average A value of the control group) 100% Statistical analysis. Experimental data were analyzed by the SPSS 10.0 software. The IR of drugs is presented as average ± standard deviation. The inter-group difference was compared using the t-test. The double variable relational analysis of drug-resistant factors was using the CMH Chi-square test and the related coefficients of Spearman grading were calculated. p < 0.05 indicated statistical significance in the difference. Figure 1. Expressions of P-gp, Survivin and bcl-2 in gastrointestinal carcinomas (SP 200). (A) No brown-yellow granules of positive P-gp staining appear in the negative control. (B) A large amount of brown-yellow granules of positive P-gp staining appear in the cells. (C) No brown-yellow granules of positive Survivin staining appear in the negative control. (D) A large amount of brown-yellow granules of positive Survivin staining appear in the cells. (E) No brown-yellow granules of positive bcl-2 staining appear in the negative control. (F) A large amount of brown-yellow granules of positive bcl-2 staining appear in the cells. Results The expressions of P-gp, p53, Survivin, and bcl-2 proteins in digestive tumor. Positive staining of P-gp, Survivin, and bcl-2 proteins appeared as yellow particles on the cell membrane and/ or in the cytoplasm (Fig. 1), while the p53 protein would show its nucleus in yellow color. In 84 cases of gastric carcinoma and colorectal carcinoma, the positive rates of P-gp, p53, Survivin, and bcl-2 were higher and they were 96.4%, 64.3%, 89.3%, and 60.7%, respectively (Table 1). In digestive tumor tissue, there were 4 types of MDR-related factors, which their expressional relationships were analyzed by CMH Chi-square test and Spearman grading analysis. There was no relationship between p53 and P-gp, Survivin, and bcl-2 in tumor tissue (r = , , , p > 0.05). No significant relationship was seen in the expression strengths of P-gp and Survivin (r = , p > 0.05). In tumor tissue, the expression of P-gp and bcl-2 and the expression of Survivin and bcl-2 showed positive relationship in their expression strengths (r = , r = , p < 0.05) (Table 2). The relationship between the expression of multi-drug resistant factors in digestive tumor and the sensitivity to 424 Chinese Journal of Cancer 2008; Vol. 27 Issue 11

3 Table 1 chemotherapeutic drugs. In digestive tumor tissue, the four MDR-related factors expression levels varied throughout the uses of nine different chemotherapeutic drugs, as shown by different inhibitory rates on tumor cells (Table 1). In the group with strong expression of P-gp, PTX, OXA, and DDP s inhibitory rates were significantly lower than those of the group with weak expression (they were respectively: t = , p = ; t = , p = ; t = , p = ). In the group with strong expression of p53, PTX and DDP s inhibitory rates on tumor cells were dramatically reduced (t = , p = and t = , p = 0.002). When Survivin was strongly expressed, the inhibitory rates on tumor cells by VCR and DDP were significantly lowered (t = , p = and t = , p = ), but the inhibitory rate of OXA was rather increased (t = , p = ). In the group of strong expression of bcl-2, the inhibitory rates on tumor cells by four drugs (5-FU, VCR, EADM, OXA) were much lower than those of the group with weak expression (they were respectively: t = , p = ; t = , p = ; t = , p = ; t = , p = ). Discussion Correlations among the expression of MDR-related factors in 84 gastrointestinal carcinomas The classical drug-resistant pathway, which is mediated by P-gp, and the inhibition on cell apoptosis, are the two mostly investigated mechanisms in research for MDR of tumor. Because the tolerance of tumor cell to a chemotherapy was determined by expression and function of genes/proteins related to drug-resistance, and the sensitization test results of in vitro chemotherapeutic experiment of tumor was consistent to clinical chemotherapeutic drug tolerance, 5 this paper, therefore, used the MTT method on in vitro sensitization test to study the relationship between the expressions of MDR-related factors of gastrointestinal tumor and the sensitivity to chemotherapeutic drugs. It was discovered that the expressions of MDR-related factors were much higher in gastrointestinal tumor and the inhibitory rates in eight out of the nine common clinical chemotherapeutic drugs were lower. It suggested that digestive tumors were much resistant to chemotherapy. Some researchers suggested the use of the expression levels of some MDR-related factors to tumors as predicative index for the effect of a chemotherapeutic drug. 1,2 However, no report of whether the results of in vitro and other studies on solid tumors could reflect upon higher resistance in gastrointestinal tumor was available. P-gp, through energy-dependent efflux channels, could pump natural alkaloids, anti-tumor antibiotics, alkylating agents, steroids, and others, out of cell, and these drugs were called MDR-related drugs. This study showed that the tumor tissues with high expression of P-gp developed tolerance to only PTX, while they were not tolerant to EADM, VCR, and VP-16. It suggested that the high expression of P-gp in tumor tissue did not induce tolerance to all MDR-related drugs, and this result was contradictory to some studies in vitro. 7 Also, although 5-FU was a non-mdr-related drug and the high expression of P-gp in tumor tissue in this study did not induce tolerance to it, there were some researches that proved the high expression of P-gp in tumor tissue could induce more sensitivity to anti-metabolic drug such as Gemcitabine. 2 It suggested that the expression and function of P-gp in tumor cell were still influenced by other factors. There were researches that discovered the high expression of P-gp was not altered in MCF7/Adr cell strains after transfection of bcl-2, but the reduction in ATP and activity of PKC also lowered the effect of P-gp, which further convert them to be more sensitive to ADR and VCR. 7 The intracellular metabolism of glucose could also regulate the cascade reactions of expression and function of P-gp. 8 The gastric cancer cell line, which was tolerant to Adriamycin, showed no high expressions of MDR1 gene and P-gp, because the expression of MDR1 mrna at cellular level was heterogeneous. 9 Therefore, we believed that the expression level of P-gp in digestive tumor tissue could only partially reflect the condition of multidrug resistance. If only based on this as a guiding index for clinical usage of chemotherapeutic drug, it would not be complete and accurate. Many chemotherapeutic drugs, such as VP-16, OPT, DDP, VCR, MMC, and etc., achieve their functions through DNA injury and apoptotic induction and the inhibitor of apoptosis proteins (IAPs) such as Survivin and bcl-2 can result in the MDR-phenotype in tumor cells after the inhibition of induced apoptosis by chemotherapeutic drug as it acts on target site. This study demonstrated that the phenomenon of developed drug tolerance to gastrointestinal tumor in five out of the nine chemotherapeutic drugs was related to the hyper-expression of survivin and bcl-2. The high expression of survivin in tumor is related to the drug resistance to VCR and DDP, while the high expression of bcl-2 causes drug resistance to 5-FU, VCR, OXA, and EADM, suggesting that Survivin and bcl-2 act on different points of the apoptotic pathway to inhibit the drug-induced apoptosis. The bcl-2 is the terminal regulatory point in the process of apoptosis and the activation of cascade reaction for proteinase is terminated when the release of cytochrome from mitochondria is interfered. Survivin, which is located downstream of bcl-2, directly inhibits the core of apoptosis, Caspase, to prevent the process. In this research, the expressions of Survivin and bcl-2 in tumor tissues had positive correlation to the apoptotic process and this result was in accordance to most researches. 10 Also, when tumor became resistant to VCR, the expressions of Survivin and bcl-2 were higher suggesting both had synergic effect when inhibiting drug-induced apoptosis. Chinese Journal of Cancer 425

4 Table 2 Correlation of the expression of four MDR-related factors to inhibition rates of nine chemotherapy drugs in 84 gastrointestinal carcinoma patients MDR, multidrug resistance; P-gp, P-glycoprotein; bcl-2, B-cell leukemia/lymphoma-2; 5-FU, 5-fluorouracil; VCR, vincristine; VP-16, etoposide; PTX, paclitaxel; EADM, epirubicin; OPT, hydroxycamptothecin; OXA, oxaliplatin; DDP, cisplatin; MTX, mrthotrexate. All values are presented as mean ± SD of 84 patients in each group. a p < 0.05, b p < 0.01 vs. Weak group. Mutant protein of p53, which participates in development of MDR in tumor, is due to the uncontrolled regulation on cell apoptosis or at the transcriptional level, the expressions of drug-resistant genes/proteins such as P-gp, Survivin, and bcl-2 are up-regulated, even though the correlation of p53 to the expression of various MDR-related factors remained controversial. This study did not find such correlation between the p53 protein and the expressions of P-gp, Survivin, and bcl-2, suggesting the complexity of MDR in tumor. This varying finding was probably also related to the method, measuring technique, and different cell strains by the investigator. Solid tumor with mutant p53 can be tolerant to many MDR-related drugs or non-mdr-related drugs. In this study, it was found that the high expression of p53 in gastrointestinal tumor tissue was also related to the drug resistance to DDP, proving that p53 was a closely related gene to Cisplatin drug and cell apoptosis. At the same time, it was found that the tumor tissue was tolerant to PTX, which was not consistent with the finding of mutant p53 in turning tumor cell to become sensitive to PTX in other report. 2 Some scholars researches on colon cancer samples had discovered that the mutation in p53 did not induce change in the expression of P-gp in tumor tissue, 11 but rather acted on micro-tubular drug, PTX, to increase expression of P-gp and efflux of PTX to produce the effect of drug resistance. 12 Many different forms of drug tolerance in tumor cells are the synergic results of multiple MDR mechanisms and many MDR-related factors. This paper further analyzed the relationship of MDR in gastrointestinal tumor to the simultaneous expression of many MDR-related factors, to discover that the drug resistance to VCR, PTX, OXA, and DDP in tumor involved two or more MRD-related factors. There was positive correlation between P-gp of tumor and the expression of bcl-2, as well as the expression strengths of Survivin and bcl-2. It suggested that MDR of tumor, while possibly based on one mechanism, was more controlled by the regulatory network of multidrug resistant genes/proteins. This study also discovered that the high expression of P-gp also developed drug resistance to non-mdr-related drugs such as OXA and DDP, and these Cisplatin drugs were not substrates of P-gp. Possible explanation was that many factors such as GST-π, p53, and others were also involved in drug tolerance to Cisplatin. The abnormal transduction of apoptotic signals such as NF-κB, which was induced by the change in expression of apoptotic-related proteins, was related to the drug resistance to Cisplatin. 1,9 The complex network of regulation, which consists of many drug resistant mechanisms and multiple MDR-related genes/proteins, in tumor cell determines the properties of drug resistance to chemotherapeutic drugs. The in vitro MDR research under controlled conditions cannot reflect the complex process of multiple factors and multiple steps and the assessment of the relationship between the expression of some MDR-related factors and the sensitivity to chemotherapeutic drugs must also take into account of the influence by other factors. This study initially proved that the use of expression levels of MDR-related factors to predict sensitivity of the gastrointestinal tumor to a chemotherapeutic drug could not reflect completely the complex process of multiple factors and steps in MDR of tumor. Currently, the assessment system for multidrug resistance in gastro- 426 Chinese Journal of Cancer 2008; Vol. 27 Issue 11

5 intestinal tumor is still not completed. Therefore, the relationship of many MDR-related factors in gastrointestinal tumor and their influences on the sensitivity to chemotherapy must be further investigated. Acknowledgements Grant: Sci-Tech Project Foundation of Hebei Province (No D-73) References [1] Ahmed F E. Molecular markers that predict response to colon cancer therapy[j]. Expert Rev Mol Diagn, 2005,5(3): [2] Zhao X D, Zhang Y. Routine chemotherapeutic drug treatment effectiveness predictive molecules and chemotherapeutic drug selsction. Ai Zheng, 2006, 25 (12): [in Chinese] [3] Han J, Lü B R, Tan B B, et al. The changes of multidrug resistance in metastatic lymph nodes of gastrointestinal tumors. Zhonghua Shi Yan Wai Ke Za Zhi, 2006,23 (12): [in Chinese] [4] Zhang Y, Shen W M. Choice of culture media in MTT tumor chemosensitivity test. Yao Xue Fu Wu Yu Yan Jiu, 2004, 4(3): [in Chinese] [5] Han J, Tan B B, Lü B R, et al. Differences between gastric and colorectal cancer in the chemosensitivity test of primary culture on tumor cells. Zhonghua Pu Tong Wai Ke Za Zhi, 2008, 23 (6): [in Chinese] [6] Xiao Y, Li J D, Shi H L, et al. Predictive value of In Vitro MTT assay chemosensitivity test of cytotoxic drug activity in cervical cancer. Ai Zheng, 2007, 26(4): [in Chinese] [7] Del Bufalo D, Biroccio A, Trisciuoglio D, et al. Bcl-2 has differing effects on the sensitivity of breast cancer cells depending on the antineoplastic drug used. Eur J Cancer, 2002, 38(18): [8] Ledoux S, Yang R, Friedlander G, et al. Glucose depletion enchances P-glycopretein expression in hepatoma cells: role of endoplasmic reticulum stress response. Cancer Res, 2003,63(21): [9] Satta T, Isobe K I, Yamuchi M, et al. Expression of MDR1 and Glutatione S Transferase-π genes and chemosensitivities in human gastrointestinal cancer. Cancer, 1992,69(4): [10] Lu C D, Altieri DC, Tanigawa N, et al. Expression of a novel antiapoptosis gene, survivin, correlated with tumor cell apoptosis and p53 accumulation in gastric carcinoma. Cancer Res, 1998,58(9): [11] De Angelis P, Stokke T, Smedshammer L, et al. P-glycoprotein is not expressed in a majority of colorectal carcinomas and is not regulated by mutant p53 in vivo. Br J Cancer, 1995,72(2): [12] Martello L A, Verdier-Pinard P, Shen H J, et al. Elevated levels of microtubule destablizing factors in a Taxol-resistant/dependent A549 cell line with an alpha-tubulin mutation. Cancer Res, 2003,63(6): Chinese Journal of Cancer 427