TRAP1, a novel mitochondrial chaperone responsible for multi-drug resistance and protection from apoptotis in human colorectal carcinoma cells

Cancer Letters 279 (2009) 39–46

Contents lists available at ScienceDirect

Cancer Letters journal homepage: www.elsevier.com/locate/canlet

TRAP1, a novel mitochondrial chaperone responsible for multi-drug resistance and protection from apoptotis in human colorectal carcinoma cells Eleonora Costantino a,1, Francesca Maddalena a,1, Serena Calise b, Annamaria Piscazzi a, Virginia Tirino d, Alberto Fersini e, Antonio Ambrosi e, Vincenzo Neri e, Franca Esposito b,c, Matteo Landriscina a,* a

Clinical Oncology Unit, Department of Medical Sciences, University of Foggia, Viale Pinto 1, 71100 Foggia, Italy Department of Biochemistry and Medical Biotechnology, University of Naples Federico II, Via S. Pansini 5, 80131 Naples, Italy CEINGE Biotecnologie Avanzate, 80145 Naples, Italy d Department of Experimental Oncology, Istituto Nazionale Tumori, Naples, Italy e General Surgery Unit, Department of Surgical Sciences, University of Foggia, Viale Pinto 1, 71100 Foggia, Italy b c

a r t i c l e

i n f o

Article history: Received 8 December 2008 Received in revised form 10 January 2009 Accepted 13 January 2009

Keywords: TRAP1 Colorectal carcinoma Chemoresistance Oxidative stress Apoptosis Chemotherapy

a b s t r a c t TRAP1 is a component of a pro-survival mitochondrial pathway up-regulated in tumor cells. The evaluation of TRAP1 expression in 26 human colorectal carcinomas showed up-regulation in 17/26 tumors. Accordingly, TRAP1 levels were increased in HT-29 colorectal carcinoma cells resistant to 5-fluorouracil, oxaliplatin and irinotecan. Thus, we investigated the role of TRAP1 in multi-drug resistance in human colorectal cancer. Interestingly, TRAP1 overexpression leads to 5-fluorouracil-, oxaliplatin- and irinotecan-resistant phenotypes in different neoplastic cells. Conversely, the inhibition of TRAP1 activity by TRAP1 ATPase antagonist, shepherdin, increased the sensitivity to oxaliplatin and irinotecan in colorectal carcinoma cells resistant to the single agents. These results suggest that the increased expression of TRAP1 could be part of a pro-survival pathway responsible for multi-drug resistance. Ó 2009 Elsevier Ireland Ltd. All rights reserved.

1. Introduction Drug resistance is the major cause of anticancer treatment failure. Indeed, drug resistance is a multifactorial phenomenon involving multiple interrelated or independent pathways, including changes in cellular responses, such as an increased ability to repair DNA damage or tolerate stress conditions, as well as acquired mechanisms for escaping apoptosis [1]. A wide variety of pathways have been identified as responsible for drug resistance; how-

* Corresponding author. Tel.: +39 0881 736241; fax: +39 0881 733614. E-mail address: [email protected] (M. Landriscina). 1 These authors contributed equally to this work. 0304-3835/$ - see front matter Ó 2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.canlet.2009.01.018

ever, these insights have resulted only in limited improvements in patient survival. Adaptive responses to stress conditions, such as increases in DNA repair activities or antioxidant defenses, may contribute to drug resistance and escape from apoptosis in tumor cells [2]. Cancer cells produce increased amounts of reactive oxygen species (ROS). A consequent antioxidant adaptive response favors a redox imbalance and an altered redox regulation of cellular signaling, thus the activation of pro-survival mechanisms [2]. Adaptive responses to oxidative stress have been the object of intense investigation [3,4]. Recently, an experimental system of Saos-2 osteosarcoma cells chronically adapted to grow in mild oxidizing conditions, induced by the GSH-depleting agent diethylmaleate (DEM), allowed us to isolate several

40

E. Costantino et al. / Cancer Letters 279 (2009) 39–46

mRNAs differentially expressed between DEM-adapted and non-adapted cells [5]. Among others, we identified TNF receptor-associated protein 1 (TRAP1), encoding for a mitochondrial heat-shock protein homologous to Hsp90 family members [6]. Interestingly, Saos-2 cells overexpressing TRAP1 exhibited a phenotype resistant to H2O2- or cisplatin-induced DNA damage and apoptosis, suggesting that this protein protects against oxidative stress [7]. It has been recently proposed that TRAP1 and Hsp90 are components of a mitochondrial survival pathway which is selectively activated in tumor cells and is responsible for antagonizing the proapoptotic activity of cyclophilin D and thus favoring mitochondria integrity and cell survival [8]. Interestingly, strategies aimed at inhibiting TRAP1 function, based on novel TRAP1 ATPase antagonists, induce sudden collapse of mitochondrial function and apoptosis, thus improving the efficacy of anticancer treatments [9,10]. In such a perspective, TRAP1/Hsp90 chaperone may represent a novel molecular target for overcoming drug resistance. We evaluated the role of TRAP1 up-regulation in inducing a multi-drug resistant phenotype in human colorectal cancer cells, and studied the capacity of TRAP1 to antagonize the proapoptotic activity of 5-fluorouracil (FU), oxaliplatin (l-OHP) and irinotecan (IRI), three chemotherapeutic agents which represent a standard of care in human colorectal carcinoma [11]. 2. Patients and methods 2.1. Patients Between January 2006 and May 2008, specimens from both tumor and normal, non-infiltrated peritumoral mucosa were obtained from 26 patients with colorectal carcinoma during surgical removal of the neoplasm. Samples were divided into 125 mm3 pieces, one specimen was fixed in formalin to confirm the histopathological diagnosis, while the others were frozen in liquid nitrogen for immunoblot and Real Time RT-PCR analysis. Expressed written informed consent to use biological specimens for investigational procedures was obtained from all patients. 2.2. Chemicals, cell cultures and constructs FU, l-OHP and IRI and all reagents were purchased from Sigma–Aldrich, Italy unless otherwise specified. Shepherdin and control scrambled peptides were kindly provided by Dr. Altieri (University of Massachusetts Medical School, USA). Human HT-29 colon carcinoma and Saos-2 osteosarcoma cells were cultured in DMEM containing 10% fetal bovine serum (FBS), 1.5 mM glutamine, and 100 units/ml penicillin and streptomycin. Saos-2 osteosarcoma cells stably transfected with TRAP1 (Saos-2 TRAP1 cells) were previously generated [7]. Full-length TRAP1 cDNA and a deletion mutant of TRAP1 (dnTRAP!), lacking the aminoterminal domain, were cloned in pRc-CMV vector (Invitrogen, Italy). The transfection vector pEGFP-F (Clontech, USA), which encodes for a farnesylated enhanced GFP (EGFP-F), was used as a reporter vector both to monitor transfection efficiency and as a cotransfection marker.

HT-29 cells were transiently cotransfected with pRc-CMV (mock) or TRAP1 constructs, and pEGFP-F (ratio 3:1, respectively) using Lipofectamine 2000 Reagent (Invitrogen, Italy), and 24 h later culture medium was supplemented twice (every 24 h) with 100 lM FU or only once with 10 lM l-OHP. Drug-induced apoptosis was evaluated by a cytofluorimetric analysis by sorting only for pEGFPpositive cells after 48 h of drugs treatment. HT-29 colon carcinoma cells resistant to FU (HT-29 FUR), l-OHP (HT-29 l-OHP-R) or IRI (HT-29 IRI-R) were selected as previously reported [12]. Cells were stabilized in presence of 20 lM FU, 1 lM IRI and 3 lM l-OHP. The resistance to each drug was assessed by MTT dye assay (see below) and by measuring apoptosis in presence of increasing concentrations of FU, l-OHP or IRI. 2.3. Cytotoxicity assays Growth inhibition was measured using the MTT dye assay as previously described [12]. Each experiment was performed three times using four replicates for each drug concentration. Apoptosis was evaluated by cytofluorimetric analysis of annexin V and 7-amino-actinomycin D (7AAD) positive cells by using the FITC-Annexin V/7-AAD Kit (Beckman Coulter, Italy). Stained cells were analyzed by ‘‘EPICS XL” Flow Cytometer (Beckman Coulter, Italy). Ten thousand events were collected per sample. Positive staining for annexin V as well as double staining for annexin V and 7-AAD was interpreted as signs of, respectively, early and late phases of apoptosis [13]. Apoptosis was evaluated in HT-29 cells transiently cotransfected with pRcCMV (mock), TRAP1 or dnTRAP1 constructs and pEGFP-F vector by propidium iodide (PI) labeling. Cells were resuspended in 100 ll of binding buffer (0.1 mM Hepes/NaOH, 1.4 M NaCl, 25 mM CaCl2, pH 7.4) and incubated for 15 min in the dark with 2 ll PI. Cells were analyzed with a Becton Dickinson FACS-Vantage flow cytometer (Becton Dickinson, USA). Experiments were performed four times using three replicates for each drug concentration. Data are referred only to pEGFP-F positive cells. 2.4. Immunoblot analysis Immunoblot analysis was performed as previously reported [14]. Specific proteins were detected by using a mouse monoclonal anti-TRAP1 antibody (Santa Cruz Biotechnology, Italy), a mouse monoclonal anti-thymidylate synthase antibody (Histoline Laboratories, Italy) or a mouse monoclonal anti-GAPDH antibody (Santa Cruz Biotechnology, Italy). In order to compare expression in different tumor specimens, TRAP1 protein levels were quantified by densitometric analysis using the Quantity One 4.5.0 software (BioRad Laboratories GmbH, USA) and expressed as time increase in tumors compared to the levels in the respective peritumoral non-infiltrated mucosa. 2.5. RNA extraction and Real Time RT-PCR analysis Total RNA from cell pellets was extracted using the TRIzol Reagent (Invitrogen, Italy). For the first strand synthesis of cDNA, 5 lg of RNA were used in a 20 ll reaction mixture

41

E. Costantino et al. / Cancer Letters 279 (2009) 39–46

utilizing a cDNA Superscript II (Invitrogen, Italy). One ll of cDNA sample was amplified using the Platinum SYBR Green qPCR Supermix UDG (Invitrogen, USA) in an iCycler iQ Real Time Detection System (BioRad Laboratories GmbH, USA). The following primers were utilized: TRAP1, forward 50 -GACGCACCGCTCAACAT-30 , reverse 50 -CACATCA AACATGGACGGTTT-30 ; TS, forward 50 -GGAGTTGACCAACT GCAAAGAGTG-30 , reverse 50 -GATGTGCGCAATCATGTACGT GAG-30 ; DPD, forward 50 -GGTCTTGCTAGCGCAACTCC-30 , reverse 50 -CCTTTAGTTCAGTGACACTTTGA-30 ; TP, forward 50 -TCATCCAGAGCCCAGAGCAGA-30 , reverse 50 -TCTGCTCTG GGCTCTGCATCA-30 ; GAPDH, forward 50 -CAAGGCTGAGAA CGGGAA-30 , reverse 50 -GCATCGCCCCACTTGATTTT-30 . Reaction conditions were 50 °C for 2 min, 95 °C (90 °C for GAPDH and TRAP1) for 2 min, followed by 45 cycles of 15 s at 95 °C (30 s at 90 °C for GAPDH and TRAP1), 30 s at 62 °C (60 °C for GAPDH and TRAP1), 30 s at 72 °C.

Tumor stage

N

%

T1 T2 T3 T4

1 6 11 8

3.8 23.1 42.3 30.8

Nx N0 N1 N2

2 10 7 7

7.7 38.5 26.9 26.9

2.6. Statistical analysis

M0 M1

17 9

65.4 34.6

Metastatic site

N

%

Liver Lung Nodes Peritoneum

9 6 16 3

34.6 23.1 61.5 11.5

The paired Student’s T test was used to establish the statistical significance between different levels of apoptotic cell death in mock and TRAP1 or dnTRAP1 transfected cells and between the differential rates of apoptosis induced by chemotherapeutic agents in scrambled- and shepherdin-treated tumor cells. The Chi-square test was used to establish the association between TRAP1 protein and mRNA levels. Statistically significant values (p < 0.05) are reported in the Section 3 and the Tables. 3. Results 3.1. TRAP1 is up-regulated in human colorectal carcinomas and in colorectal carcinoma cells resistant to FU, l-OHP and IRI TRAP1 protein levels were evaluated in a cohort of 26 human colorectal carcinomas. Patient characteristics are reported in Table 1. In addition, TRAP1 mRNA expression was assessed in a subgroup of 16 colorectal tumors. TRAP1 protein and mRNA expression was interpreted as up-regulated if found to have increased at least three times in comparison with the corresponding non-infiltrated peritumoral mucosa. Fig. 1A shows an immunoblot analysis of TRAP1 in five tumors, chosen as representative analytical results of all samples, while TRAP1 protein and mRNA expression levels in the whole series are reported in Fig. 1B. TRAP1 protein expression was up-regulated in 17/26 (65.4%) colorectal tumors. Consistently, a significant correlation between TRAP1 expression levels by immunoblot and Real Time PCR analysis (p = 0.012) was observed. Starting from these preliminary observations, we evaluated the role of TRAP1 in favoring a chemoresistant phenotype in vitro. To this aim, HT-29 colorectal carcinoma cells adapted to toxic concentrations of FU, l-OHP and IRI were selected. Interestingly, immunoblot analysis showed increased TRAP1 protein levels in HT-29 cells adapted to l-OHP, FU and IRI (Fig. 2A). Of note, HT-29 FU-R cells exhibited a significant up-regulation of thymidylate synthase (TS) gene and protein expression with no major modifications in the mRNA expression of dihydropyrimidine dehydrogenase (DPD) and thymidine phosphorylase (TP), three genes known to be responsible for resistance to FU [15] (Fig. 2B). 3.2. The up-regulation of TRAP1 expression results in a multi-drug resistant phenotype in vitro The increased expression of TRAP1 in the majority of human colorectal carcinomas and in human colorectal carcinoma drug-resistant cells, prompted us to study the role of the TRAP1 gene in favoring resistance to these antiblastic agents. This issue was evaluated by using HT-29 colorectal carcinoma cells transiently transfected with TRAP1 expression vector. In addition, Saos-2 osteosarcoma cells were also chosen as a control

Table 1 Baseline characteristics of the patients. Patients

26

Age Range Median

59–87 73

Sex Female Male

8 18

for these studies since previously had exhibited the up-regulation of TRAP1 gene expression upon adaptation to chronic mild oxidizing conditions [5], and after transfection with TRAP1 gene revealed a phenotype resistant to H2O2- or cisplatin-induced DNA damage and apoptosis [7]. In preliminary experiments, Saos-2 cells stably transfected with an empty vector or with TRAP1 expression vector were exposed to increasing concentrations of FU (Fig. 3A), l-OHP (Fig. 3B) and IRI (Fig. 3C) for 48 h and evaluated for cell viability by MTT incorporation analysis. Interestingly, TRAP1 stable clones displayed a reduced sensitivity to all agents, as compared to controls (Fig. 3). In addition, the rate of apoptotic cell death was assessed in the same cell lines upon exposure to FU, l-OHP and IRI for 48 h. As reported in Table 2, all treatments induced a significant increase in the rate of apoptosis in Saos-2 control cells. Conversely, high levels of TRAP1 protein present in TRAP1 stable transfectants (Fig. 3, insert) provided protection against FU-, l-OHP- and IRI-induced apoptosis. In parallel experiments, TRAP1 or a deletion mutant of TRAP1 (dnTRAP1), which is unable to traslocate into mitochondria, and behaves as a dominant negative over the endogenous TRAP1 activity, were transiently transfected in HT-29 colorectal carcinoma cells. pEGFP-F vector was used as a cotransfection marker in order to monitor transfection efficiency. HT-29 cells transfected with TRAP1 or dnTRAP1 expression vectors or mock-transfected HT-29 cells were exposed to 100 lM FU and 10 lM l-OHP for 48 h and apoptotic cell death was assessed only in pEGFP-positive cells sorted by cytofluorimetry. Interestingly, consistently with the observations obtained in osteosarcoma TRAP1 stable transfectants, the up-regulation of TRAP1 expression protected cells from FU- and l-OHP-induced cell death which, otherwise, induced high levels of apoptosis in mock-transfected cells (Table 2). Even more importantly, TRAP1 dominant negative mutant significantly rescued the sensitivity of HT-29 cells to the proapoptoic activity of FU and l-OHP (Table 2). Similarly, interference of TRAP1 levels by siRNAs rescued the sensitivity to the proapoptotic drugs as well (data not shown). These results suggest that TRAP1 may be involved in a multi-drug resistant phenotype in vitro. 3.3. The TRAP1 ATPase antagonist, shepherdin, restores sensitivity to antiblastic agents in vitro We further evaluated whether the inhibition of TRAP1 ATPase activity by shepherdin is able to restore sensitivity to chemotherapeutic agents. Indeed, shepherdin was previously described as a tumor-selective agent which, upon binding to mitochondrial TRAP1/Hsp90

42

E. Costantino et al. / Cancer Letters 279 (2009) 39–46

Fig. 1. TRAP1 protein levels in human colorectal carcinomas: (A) total cell lysates from five colorectal tumors (T) and the corresponding non-infiltrated peritumoral mucosa (M) were separated by SDS–PAGE and immunoblotted with a mouse monoclonal anti-TRAP1 and a mouse monoclonal anti-GAPDH antibodies and (B) TRAP1 protein and mRNA levels in human colorectal carcinomas expressed as time increase in tumors compared to the corresponding peritumoral non-infiltrated mucosa.

complexes, inhibits their chaperone activity via an ATP competition mechanism [8]. To address this issue, we used wild type HT-29 cells and HT-29 cells resistant to FU, l-OHP and IRI, which had previously been characterized by increased expression of TRAP1 (Fig. 2A). Cells were exposed to shepherdin or cell-permeable scrambled peptidomimetic for 5 h either alone or in combination with FU, l-OHP or IRI. Cells previously exposed to the combined treatment with the chemotherapeutic drug and the cell-permeable peptide were further incubated in the presence of the same antiblastic agent for an additional 19 h. The ability of shepherdin to restore the sensitivity to the specific chemotherapeutic agent is expressed as a ratio between the apoptotic rates induced by the specific agent in the presence of shepherdin or of scrambled peptides. As reported in Table 3, wild type HT-29 cells treated with 50 lM shepherdin did not exhibit any increase in the rate of apoptotic cell death (shepherdin/scrambled ratio of 1.2:1). By contrast, the simultaneous exposure of HT-29 cells to shepherdin and FU, l-OHP or IRI-induced a 2–3 times increase in the rate of apoptosis compared to the treatment with the same chemotherapeutic agent combined with scrambled control peptide (shepherdin/scrambled ratio of 2.3:1, 3.5:1 and 2.1:1, respectively) (Table 3).

In parallel experiments, HT-29 cells resistant to toxic concentrations of FU, l-OHP and IRI were exposed to 50 lM shepherdin in combination with the respective resistance antiblastic agent, and this treatment did not result in increased rates of apoptotic cell death (data not shown), possibly due to the high levels of TRAP1 present in these cells. Thus, experiments were performed by using 100 lM shepherdin which, as previously demonstrated in other cancer cells [8], is responsible per se for a significant increase in cell apoptosis with a shepherdin/scrambled ratio of 3:1, 2.3:1 and 2.7:1, respectively in HT-29 FU-R, HT-29 l-OHP-R and HT-29 IRI-R cells (Table 3). Interestingly, the simultaneous exposure of HT-29 l-OHP-R and HT-29 IRI-R cells to shepherdin and, respectively, l-OHP or IRI resulted in the up-regulation of drug-induced apoptosis (shepherdin/scrambled ratio of 6.4:1 and 4:1, respectively), whereas the incubation of HT-29 FU-R cells with FU and shepherdin did not restore sensitivity to FU (Table 3). The high levels of apoptotic cell death in drug-resistant cells exposed to scrambled are independent from the peptide, since a comparable cell death is observed in control resistant cells without scrambled peptide. In addition, 100 lM scrambled did not induce apoptosis in wild type cells, as well. This feature is specific and shared by all three cell lines and is likely due to the presence of the resistance drug

E. Costantino et al. / Cancer Letters 279 (2009) 39–46

Fig. 2. (A) TRAP1 protein levels in colorectal carcinoma cells resistant to single chemotherapeutic agents. Total cell lysates from wild type HT-29 (WT) and HT-29 cells resistant to 5-fluorouracil (FU), oxaliplatin (l-OHP) or irinotecan (IRI) were separated by SDS–PAGE and immunoblotted with a mouse monoclonal anti-TRAP1 and a mouse monoclonal anti-GAPDH antibody. (B) Expression of thymidylate synthase (TS), dihydropyrimidine dehydrogenase (DPD), and thymidine phosphorylase (TP) genes in wild type HT-29 and HT-29 cells resistant to FU (HT-29 FU-R cells). Total RNA was extracted from HT-29 and HT-29 FU-R cells and evaluated for expression of TS, DPD and TP by Real Time RT-PCR. Insert: TS protein expression in wild type HT-29 (line 1) and HT-29 FU-R (line 2) cells. Cell lysates were separated by SDS–PAGE and immunoblotted with a mouse monoclonal anti-TS antibody.

in the culture media, which may induce cell death in non-adapted subclones. These results suggest that targeting TRAP1/Hsp90 in colorectal carcinoma cells may overcome drug resistance.

4. Discussion Apoptosis is the essential response induced in tumor cells by most chemotherapeutic agents, and an overwhelming number of results support the association between activation of survival mechanisms and/or downregulation of proapoptotic pathways and resistance to chemotherapy [16]. We recently studied the role of TRAP1 in the adaptation to mild conditions of oxidative stress and demonstrated that this gene may be responsible for protecting from ROS-induced DNA damage and cisplatin-triggered apoptosis [7]. Since several mechanisms involved in ROS-adaptive responses have also been described as mechanisms responsible for resistance to chemotherapeutic agents in tumor cells [17], we further explored, at a preclinical level, the role of TRAP1 in inducing a chemoresistant phenotype in human colorectal carcinoma. The results reported here is, to our knowledge, the first evidence that TRAP1 expression is up-regulated in about 60% of human colorectal cancers and that TRAP1 up-regulation induces a multi-drug resistant phenotype in colon carcinoma cells. Of note, we observed that HT-29 cells resistant to FU, l-OHP and IRI express increased TRAP1 protein levels and that the inhibition of TRAP1 activity by shepherdin or a dnTRAP1 deletion mutant sensitizes

43

drug-resistant and wild type colorectal carcinoma cells to anticancer agents These lines of evidence may be relevant for either basic scientists or clinical oncologists. Indeed, our results support the hypothesis that oxidant- and drug-resistant cells share common molecular mechanisms of adaptation to stress. It has been proposed that increased antioxidant networks of cancer cells may play a critical role in preventing apoptotic cell death in response to anticancer treatments [17]. In such a perspective, a molecular feature common to several tumor cell types is the overexpression of nonenzymatic and enzymatic antioxidant defenses, and this up-regulation has received growing attention as a negative modulator of programmed cell death. We have previously observed that overexpression of the mitochondrial enzyme MnSOD significantly correlates with high tumor grade and a worst prognosis in human brain gliomas [18,19] and it is clear that even small amounts of this enzyme seem to be crucial for cell resistance to inflammatory stimuli and anticancer drugs and to prevent oncogene-induced apoptosis [20]. Several lines of evidence support a role for molecular chaperones in driving cell transformation and resistance to apoptosis. It has been proposed that, because of its restricted repertoire of client proteins, mainly kinases and signaling molecules, Hsp90 chaperones occupy a critical role in cellular homeostasis [21]. Hsp90 chaperones are, indeed, required for the activity of several key regulators of apoptosis and through these associations may confer survival advantages to tumor cells [22]. Furthermore, the role of TRAP1/Hsp90 chaperones has been reinforced by recent evidence reported by Kang et al. suggesting that this pathway may represent an organelle-specific network responsible for maintaining homeostasis in tumor cell mitochondria [8]. Indeed, tumor cells organize a mitochondrial chaperone network which involves Hsp90, TRAP1 and cyclophilin D and is responsible for antagonizing the proapoptotic function of cyclophilin D, thus favoring cell viability [8]. Conversely, inhibition of mitochondrial Hsp90 chaperones using mitochondrial-directed ATPase antagonists causes the sudden collapse of mitochondrial membrane potential, release of cytochrome c, and massive death of tumor but not of normal cells [8]. Thus, it is likely that the high levels of TRAP1 observed not only in the majority of human colorectal carcinomas, but also in other human malignancies [23], may be considered an additional mitochondrial antiapoptotic mechanism, which may account for a multi-drug resistant phenotype as observed in vitro. Indeed, the evaluation of molecular markers of resistance to specific chemotherapeutic agents is a field of intense investigation in human colorectal cancer. The up-regulation of TS, DPD and TP genes, the increased protein expression of topoisomerase-1, as well as specific polymorphisms in nucleotide excision repair genes have already been shown to predict resistance to FU, IRI and l-OHP [15,24,25], but, so far, these efforts have not provided any useful tools for tailoring patient treatment in daily clinical practice. Certainly, further studies are required to evaluate whether the up-regulation of TRAP1 expression may be predictive of resistance to currently used chemotherapy regimens in human colorectal tumors and whether

44

E. Costantino et al. / Cancer Letters 279 (2009) 39–46

Fig. 3. Reduced sensitivity of TRAP1 stable transfectants to FU, l-OHP and IRI: Saos-2 stably transfected with an empty vector (Control) or with TRAP1 expression vector (TRAP1) were incubated in the presence of increasing concentrations of FU (A), l-OHP (B) and IRI (C) and evaluated for drug-induced cytotoxicity by MTT incorporation. Results are reported as percentage of MTT incorporation compared with the corresponding untreated control. Insert: TRAP1 protein expression in mock Saos-2 cells (line 1) and Saos-2 TRAP1 stable transfectants (line 2). Cell lysates were separated by SDS–PAGE and immunoblotted with a mouse monoclonal anti-TRAP1 antibody.

the expression of TRAP1 may represent a decision tool for selecting the appropriate chemotherapy schedule in patients. In such a perspective, it is noteworthy that the TRAP1 antagonist, shepherdin, is unable to revert resistance to FU in vitro, suggesting that multiple mechanisms are involved in drug resistance in colorectal tumor cells and that the up-regulation of TS may be dominant over TRAP1 antiapoptotic activity in favoring resistance to FU.

Intriguingly, shepherdin restored sensitivity to IRI and l-OHP in colorectal tumor cells resistant to the single agents. Indeed, shepherdin has already been characterized as a novel anticancer agent able specifically to bind mitochondrial TRAP1/Hsp90 complexes and to inhibit their ATPase activity, thus selectively favoring tumor cell death [8,10]. In such a perspective, the TRAP1/Hsp90 pathway may be regarded as a novel molecular target in human

E. Costantino et al. / Cancer Letters 279 (2009) 39–46 Table 2 Rates of apoptotic cell death in Saos-2 osteosarcoma cells stably transfected with pRc-CMV (mock) or TRAP1 constructs and in HT-29 colorectal carcinoma cells transiently transfected with pRc-CMV (mock), TRAP1 or dnTRAP1 constructs and pEGFP-F vector treated with FU, l-OHP or IRI. Apoptotic rates in HT-29 cells are calculated based only on sorted pEGFP-F positive cells. Mock

TRAP1

dnTRAP1

p-Value

Saos-2 cells Vehicle FU 1 lM FU 5 lM l-OHP 1 lM l-OHP 5 lM IRI 1 lM IRI 3 lM

0.7 ± 0.2 2.3 ± 0.4 11.5 ± 0.7 17.0 ± 0.5 18.5 ± 0.7 9.0 ± 0.6 10.8 ± 0.5

4.5 ± 0.3 4.9 ± 0.5 6.7 ± 0.2 5.1 ± 0.3 5.9 ± 0.4 4.0 ± 0.4 3.3 ± 0.6

– – – – – – –

n.s. 0.011 0.001 0.002 0.01 0.005

HT-29 cells Vehicle FU 100 lM l-OHP 10 lM

3.4 ± 0.7 32.6 ± 3.5 32.1 ± 1.8

4.1 ± 0.9 4.8 ± 1.7* 5.9 ± 1.3*

3.8 ± 1.1 52.3 ± 3.9** 39.9 ± 1.6**

0.01*, 0.034** 0.004*, 0.045**

45

We suggest that the identification of TRAP1 in these networks may open promising therapeutic approaches for human cancer. Taken together, these results suggest evaluating shepherdin effects in IRI and/or l-OHP chemoresistant colorectal carcinomas overexpressing TRAP1, as a novel molecular-targeted therapy to overcome drug resistance, rescuing sensitivity to l-OHP- and/or IRI-based chemotherapy. 5. Conflict of interest statement All authors declare that there are no possible financial or personal conflicts of interest.

* and **; statistical significance between different levels of apoptotic cell death in mock and TRAP1 or dnTRAP1 transfected cells.

Acknowledgement This work was supported by Grants from MIUR-PS 35126/Ind. References

Table 3 Rates of apoptotic cell death in wild type colorectal carcinoma HT-29 cells and HT-29 cells resistant to FU (FU-R), l-OHP (l-OHP-R) or IRI (IRI-R) treated with FU, l-OHP or IRI in the presence of shepherdin or control scrambled peptides. Apoptosis (% ± SD)

Ratio (±SD)

p-Value

Scrambled

Shepherdin

1.7 ± 0.1 2.3 ± 0.3 1.6 ± 0.2 1.8 ± 0.2

2.0 ± 0.2 5.1 ± 0.2 5.3 ± 0.2 3.8 ± 0.2

1.2 ± 0.1 2.3 ± 0.4 3.5 ± 0.5 2.1 ± 0.3

0.005 <0.0001 0.004

HT-29 FU-R cells Control 12.4 ± 1.1 13.8 ± 1.2 FU 100 lM

36.9 ± 1.8 35.9 ± 2.1

3.0 ± 0.2 2.6 ± 0.4

n.s.

HT-29 OHP-R cells Control 4.3 ± 0.7 3.9 ± 0.9 l-OHP 15 lM

9.8 ± 2.3 23.8 ± 4.1

2.3 ± 0.3 6.4 ± 1.5

0.016

HT-29 IRI-R cells Control IRI 10 lM

20.8 ± 2.1 36.1 ± 1.4

2.7 ± 0.7 4.0 ± 0.7

0.001

HT-29 cells Control FU 10 lM l-OHP 5 lM IRI 10 lM

8.0 ± 1.7 9.3 ± 1.5

colorectal cancer. A derivative of geldanamycin, 17AAG, has been proposed for clinical use in cancer patients for its ability to block the TRAP1/Hsp90 pathway. 17AAG has shown significant antitumor activity in vitro and in animal models and thus has entered clinical testing in cancer patients [26]. Down-regulation of client proteins has been reported in lymphocytes at well-tolerated doses of 17AAG and early evidence of therapeutic activity has been described in human solid tumors by using 17AAG alone or in combination with paclitaxel [26]. However, recent reports suggest that 17AAG is unable to accumulate in mitochondria [8]: hence, Hsp90 inhibition as well as cytochrome c release cannot occur, nor apoptosis can be activated. By contrast, shepherdin induces mitochondrial permeability transition pores to open and triggers selective tumor cell death due to the Antennapedia sequence which drives the peptide inside purified tumor mitochondria [8].

[1] T.R. Wilson, D.B. Longley, P.G. Johnston, Chemoresistance in solid tumours, Ann. Oncol. 17 (Suppl. 10) (2006) 315–324. [2] S. Pervaiz, Pro-oxidant milieu blunts scissors: insight into tumor progression, drug resistance, and novel druggable targets, Curr. Pharm. Des. 12 (34) (2006) 4469–4477. [3] I. Manoli, S. Alesci, M.R. Blakman, Y.A. Su, O.M. Rennert, G.P. Chrousos, Mitochondria as key components of the stress response, Trends Endocrinol. Metab. 18 (5) (2007) 190–198. [4] C.T. Taylor, J. Pouyssegur, Oxygen, hypoxia, and stress, Ann. NY Acad. Sci. 1113 (2007) 87–94. [5] N. Montesano Gesualdi, G. Chirico, M.T. Catanese, G. Pirozzi, F. Esposito, AROS-29 is involved in adaptive response to oxidative stress, Free Rad. Res. 40 (5) (2006) 467–476. [6] S.J. Felts, B.A. Owen, P. Nguyen, J. Trepel, D.B. Donner, D.O. Toft, The hsp90-related protein TRAP1 is a mitochondrial protein with distinct functional properties, J. Biol. Chem. 275 (5) (2000) 3305– 3312. [7] N. Montesano Gesualdi, G. Chirico, G. Pirozzi, E. Costantino, M. Landriscina, F. Esposito, Tumor necrosis factor-associated protein 1 (TRAP1) protects cells from oxidative stress and apoptosis, Stress 10 (4) (2007) 342–350. [8] B.H. Kang, J. Plescia, T. Dohi, J. Rosa, S.J. Doxsey, D.C. Altieri, Regulation of tumor cell mitochondrial homeostasis by an organelle-specific Hsp90 chaperone network, Cell 131 (2) (2007) 257–270. [9] F. Guo, K. Rocha, P. Bali, M. Pranpat, W. Fiskus, S. Boyapalle, S. Kumaraswamy, M. Balasis, B. Greedy, E.S. Armitage, N. Lawrence, K. Bhalla, Abrogation of heat shock protein 70 induction as a strategy to increase antileukemia activity of heat shock protein 90 inhibitor 17allylamino-demethoxy geldanamycin, Cancer Res. 65 (22) (2005) 10536–10544. [10] J. Plescia, W. Salz, F. Xia, M. Pennati, N. Zaffaroni, M.G. Daidone, M. Meli, T. Dohi, P. Fortugno, Y. Nefedova, D.I. Gabrilovich, G. Colombo, D.C. Altieri, Rational design of shepherdin, a novel anticancer agent, Cancer Cell 7 (5) (2005) 457–468. [11] R.M. Goldberg, Advances in the treatment of metastatic colorectal cancer, Oncologist 10 (Suppl. 3) (2005) 40–48. [12] C. Barone, M. Landriscina, M. Quirino, M. Basso, C. Pozzo, G. Schinzari, E. D’Argento, N. Trigila, A. Cassano, Schedule-dependent activity of the 5-fluorouracil and irinotecan combination in the treatment of human colorectal cancer: in vitro evidences and a phase I dose-escalating clinical trial, Br. J. Cancer 96 (1) (2007) 21–28. [13] T.C. George, D.A. Basiji, B.E. Hall, D.H. Lynch, W.E. Ortyn, D.J. Perry, M.J. Seo, C.A. Zimmerman, P.J. Morrissey, Distinguishing modes of cell death using the imagestream multispectral imaging flow cytometer, Cytometry A. 59 (2) (2004) 237–245. [14] M. Landriscina, S.A. Altamura, L. Roca, M. Gigante, A. Piscazzi, E. Cavalcanti, E. Costantino, C. Barone, M. Cignarelli, L. Gesualdo, E. Ranieri, Reverse transcriptase inhibitors induce cell differentiation and enhance the immunogenic phenotype in human renal clear-cell carcinoma, Int. J. Cancer 122 (12) (2008) 2842–2850.

46

E. Costantino et al. / Cancer Letters 279 (2009) 39–46

[15] R. Soong, N. Shah, M. Salto-Tellez, B.C. Tai, R.A. Soo, H.C. Han, S.S. Ng, W.L. Tan, N. Zeps, D. Joseph, R.B. Diasio, B. Iacopetta, Prognostic significance of thymidilate synthase, dihydropyrimidine dehydrogenase and thymidine phosphorylase protein expression in colorectal cancer patients treated with or without 5-fluorouracilbased chemotherapy, Ann. Oncol. 19 (5) (2008) 915–919. [16] T.T. Tan, E. White, Therapeutic targeting of death pathways in cancer: mechanisms for activating cell death in cancer cells, Adv. Exp. Med. Biol. 615 (2008) 81–104. [17] J.D. Pennington, T.J. Wang, P. Nguyen, L. Sun, K. Bisht, D. Smart, D. Gius, Redox-sensitive signaling factors as a novel molecular targets for cancer therapy, Drug Resist. Updat. 8 (5) (2005) 322–330. [18] M. Landriscina, F. Remiddi, F. Ria, B. Palazzotti, M.E. De Leo, M. Iacoangeli, R. Rosselli, M. Scerrati, T. Galeotti, The level of MnSOD is directly correlated with grade of brain tumours of neuroepithelial origin, Br. J. Cancer 74 (12) (1996) 1877–1885. [19] F. Ria, M. Landriscina, F. Remiddi, R. Rosselli, M. Iacoangeli, M. Scerrati, G. Pani, S. Borrello, T. Galeotti, The level of manganese superoxide dismutase content is an independent prognostic factor for glioblastoma. Biological mechanisms and clinical implications, Br. J. Cancer 84 (4) (2001) 529–534. [20] G. Pani, R. Colavitti, B. Bedogni, S. Fusco, D. Ferraro, S. Borrello, T. Galeotti, Mitochondrial superoxide dismutase: a promising target

[21] [22] [23]

[24]

[25]

[26]

for new anticancer therapies, Curr. Med. Chem. 11 (10) (2004) 1299– 1308. J.C. Young, I. Moarefi, F.U. Hartl, Hsp90: a specialized but essential protein-folding tool, J. Cell Biol. 154 (2) (2001) 267–273. L. Whitesell, S.L. Lindquist, HSP90 and the chaperoning of cancer, Nat. Rev. Cancer 5 (10) (2005) 761–772. W. Fang, X. Li, Q. Jiang, Z. Liu, H. Yang, S. Wang, S. Xie, Q. Liu, T. Liu, J. Huang, W. Xie, Z. Li, Y. Zhao, E. Wang, F.M. Marincola, K. Yao, Transcriptional patterns, biomarkers and pathways characterizing nasopharyngeal carcinoma of Southern China, J. Transl. Med. 6 (2008) 32. M. Monzo, I. Moreno, A. Navarro, R. Ibeas, R. Artells, B. Gel, f. Martinez, J. Moreno, R. Hernandez, M. Navarro-Vigo, Single nucleotide polymorphisms in nucleotide excision repair genes XPA, XPD, XPG and ERCC1 in advanced colorectal cancer patients treated with firstline oxaliplatin/fluoropyrimidine, Oncology 72 (5–6) (2007) 364–370. M.S. Braun, S.D. Richman, P. Quirke, C. Daly, J.W. Adlard, F. Elliot, J.H. Barret, P. Selby, A.M. Meade, R.J. Stephens, M.K. Parmar, M.T. Seymour, Predictive biomarkers of chemotherapy efficacy in colorectal cancer: results from the UK MRC FOCUS trial, J. Clin. Oncol. 26 (16) (2008) 2690–2698. G. Chiosis, M. Vilenchik, J. Kim, D. Solit, Hsp90: the vulnerable chaperone, Drug Discov. Today 9 (20) (2004) 881–888.