2-Methoxy-5-Amino-N-Hydroxybenzamide Sensitizes Colon Cancer Cells to TRAIL-Induced Apoptosis by Regulating Death Receptor 5 and Survivin Expression Free

TNF-related apoptosis-inducing ligand (TRAIL), a member of the TNF superfamily (1), is capable of inducing apoptosis after interaction with death receptors (DR; ref. 2). Five homologous human receptors for TRAIL have been identified. Two of these receptors, DR4 (TRAIL-R1) and DR5 (TRAIL-R2), have cytoplasmic death domains and activate apoptotic pathways following TRAIL binding (3). TRAIL-R3, also termed decoy receptor (DcR) 1, and TRAIL-R4 (DcR2) are expressed on the cell surface but lack a functional intracellular death domain, thus conferring protection against TRAIL-induced apoptosis (4). The last TRAIL receptor is osteoprotegerin, a secreted, low-affinity receptor, whose physiologic relevance remains unclear (5). Interaction of TRAIL with DR4 and DR5 results in caspase-8 activation through Fas-associated death domain (FADD) in the death-inducing signaling complex. Activated caspase-8 can induce apoptosis by activating either directly downstream effector caspases (e.g., caspase-3, caspase-6, and caspase-7) or the intrinsic mitochondria-mediated apoptotic pathway (6).

TRAIL induces apoptosis in a variety of cancer cells but not in nontransformed cells (2, 7), thus emphasizing the role of this molecule in the immune surveillance against tumors. Indeed, preclinical studies have shown that soluble forms of TRAIL suppress the growth of human tumor xenografts, with no apparent systemic toxicity (8, 9). Similarly, agonistic antibodies targeting DR4 or DR5 have been reported to exert antineoplastic effects in various models of carcinogenesis (10, 11). Therefore, given the selectivity of its proapoptotic actions, TRAIL is considered as an attractive target for anticancer therapies. However, common cancers do not easily undergo apoptosis and are resistant to TRAIL-based therapies (12–14). This has been, for example, documented in patients with colon cancer (15, 16). Although the exact mechanism underlying the diminished susceptibility of colon cancer to TRAIL-induced apoptosis is not fully understood, previous studies have shown that colon cancer cells express reduced levels of DR4/DR5 (17) and high survivin (18), an antiapoptotic factor that suppresses TRAIL-induced apoptosis (19). These observations together with the demonstration that the above defects in colon cancer cells are reversible suggest the need of drug combinations to overcome the colon cancer cell resistance to TRAIL.

2-Methoxy-5-amino-N-hydroxybenzamide (herein termed 2-14) is a derivative of mesalamine, an anti-inflammatory drug commonly used in the management of patients with inflammatory bowel diseases. In recent years, both in vitro and in vivo studies have shown that mesalamine inhibits multiple biological pathways that sustain colon cancer cell growth, thereby reducing the risk of developing inflammatory bowel disease-related colon cancer (20–22). However, the antineoplastic effects of mesalamine are seen only with doses that are not always reached in the gut following oral administration of the drug (23, 24). Therefore, we developed several derivatives of mesalamine and selected 2-14 as it was the more potent inhibitor of colon cancer cell growth (25). Because our previous study showed that treatment of colon cancer cell lines with 2-14 triggers endoplasmic reticulum stress (25), a phenomenon that has been involved in the regulation of the expression of various components of the TRAIL-driven apoptotic pathway (26, 27), we hypothesized that 2-14 could restore the sensitivity of colon cancer cells to TRAIL. In this study, we assessed the in vitro and in vivo effect of 2-14 on TRAIL-induced colon cancer cell apoptosis.

All reagents were from Sigma-Aldrich unless specified. Mesalamine and 2-14 (both from Giuliani SpA) were dissolved in culture medium as a 100 or 25 mmol/L stock solution, respectively. The chemical structure of 2-14 and mesalamine are illustrated in Fig. 1A. The pH of mesalamine solution was adjusted to 7.4 with NaOH. Colon cancer cell lines were cultured in McCoy's 5A (HCT-116 and HT-29) or Dulbecco's Modified Eagle's Media (HT-115) or RPMI 1640 (DLD-1 and the murine colon cancer cell line CT26) medium, all supplemented with 10% FBS and 1% penicillin/streptomycin (all from Lonza), in a 37°C, 5% CO₂, fully humidified incubator. HCT-116, HT-29, DLD-1, and CT26 cells were obtained from the American Type Culture Collection. HT-115 cells were obtained from the European Collection of Cell Cultures. The above-mentioned cell lines were procured more than 6 months ago and have not been tested recently for authentication in our laboratory.

To test the susceptibility of colon cancer cells to TRAIL, cells were treated with increasing doses of recombinant TRAIL (12.5–100 ng/mL; human: catalogue no. 310-04, mouse: catalogue no. 315-19; Peprotech) for 24 to 48 hours. To assess whether 2-14 sensitizes colon cancer cells to TRAIL-induced apoptosis, cells were preincubated with 2-14 (0.75–3 mmol/L) for 8 hours and then stimulated with TRAIL (50 ng/mL) for further 16 to 36 hours. In some experiments, cells were preincubated with the extracellular signal–regulated kinase (ERK) inhibitor PD98059 (20 μmol/L) or dimethyl sulfoxide (DMSO; vehicle) for 1 hour and then stimulated with 2-14. In parallel, cells preincubated with PD98059 or DMSO were stimulated or nor with 2-14 for 8 hours and then either left untreated or treated with TRAIL (50 ng/mL) for further 24 hours. To evaluate the role of the proteasome pathway in the 2-14–mediated regulation of survivin, cells were preincubated with the proteasome inhibitor lactacystin (10 μmol/L) or DMSO for 30 minutes before adding 2-14 for further 16 hours.

To assess cell death and apoptosis, cells were stained with fluorescein isothiocyanate–conjugated Annexin V according to the manufacturer's instructions (Immunotech) and 5 μg/mL propidium iodide (PI) for 30 minutes at 4°C, and their fluorescence was measured using FL-1 and FL-2 channels of FACSCalibur using CellQuest Pro software. Analysis of apoptotic cells was conducted as previously described (28).

Caspase-8 activation was quantified by flow cytometry using the CaspGLOW Staining Kit (Biovision).

Cells were stained with monoclonal anti-human phycoerythrin-conjugated DR4 or DR5 (Biolegend) or phycoerythrin-conjugated mouse control IgG1 (BD Biosciences) and examined by flow cytometry. Positivity was defined as percentage of DR4- or DR5-positive cells and as mean fluorescence intensity (MFI; ref. 29).

Total RNA was extracted from cells by using TRIzol reagent according to the manufacturer's instructions (Invitrogen). A constant amount of RNA (1 μg/sample) was reverse transcribed into cDNA, and 1 μL of cDNA per sample was then amplified by real-time PCR using SYBR Green Supermix (Bio-Rad). Human primers were as follows: DR4: FWD: 5′-CACAGCAATGGGAACATAGC-3′, REV: 5′-CAGGGACTTCTCTCTTCTTC-3′; DR5: FWD: 5′-GCCCCACAACAAAAGAGGTC-3′, REV: 5′-GGAGGTCATTCCAGTGAGTG-3′; DcR1: FWD: 5′-AAAGTTCGTCGTCGTCATCG-3′, REV: 5′ACAGGCTCCAGTATGTTCTG-3′; DcR2: FWD: 5′-AAGTTCCCCAGCAGACAGTG-3′, REV: 5′-GTACATAGCAGGCAAGAAGGC-3′; and survivin: FWD: 5′ACGACCCCATAGAGGAACATA-3′, REV: 5′-CGCACTTTCTCCGCAGTTTC-3′. Mouse primers were as follows: DR5: FWD: 5′-AGTAGTGCTGCTGATTGGAG-3′, REV: 5′-CCTGTTTTCTGAGTCTTGCC-3′; DcTRAIL-R1: FWD: 5′-AATCCCCCATACTCAAGGAC-3′, REV: 5′-ATTTGGCACTCGCATTTCCG-3′; DcTRAIL-R2: FWD: 5′-TGCTGCTGCTGCTGAATCTG-3′, REV: 5′GTTCCTGGGTGACACTTCTC-3′; and β-actin: FWD: 5′-AAGATGACCCAGATCATGTTTGAGACC-3′, REV: 5′-AGCCAGTCCAGACGCAGGAT-3′. RNA expression was calculated relative to the housekeeping β-actin gene on the base of the ddC_t algorithm.

Total proteins were extracted using the following lysis buffer: 10 mmol/L HEPES, 1 mmol/L EDTA, 60 mmol/L KCl, 0.2% IGEPAL CA-630, 1 mmol/L sodium fluoride, 10 μg/mL aprotinin, 10 μg/mL leupeptin, 1 mmol/L DTT, and 1 mmol/L phenylmethylsulfonylfluoride, separated on an SDS-PAGE gel, and blots were then incubated with the following antibodies: PPAR-γ, p-ERK, ERK, p-38, p38, p-JNK, JNK, CHOP, Bak, Bax, Bcl-2, XIAP, Bcl-xL, and survivin (all from Santa Cruz Biotechnology); c-FLIP_S/L (Alexis); and β-actin, followed by a secondary antibody conjugated to horseradish peroxidase. For immunoprecipitation studies, the protein lysates were prepared from cells either left untreated or treated with 1.5 mmol/L 2-14 for 1 to 3 hours. Proteins were immunoprecipitated with 2 μg of anti-ubiquitin (Santa Cruz Biotechnology) or control isotype antiserum for 2 hours followed by incubation with protein A/G agarose beads (Upstate) overnight. The resulting immunoprecipitates were washed thoroughly 4 times with cold lysis buffer, separated by SDS/PAGE, and immunoblotted with an antibody against survivin. In another set of experiments, the same proteins were immunoprecipitated with 2 μg anti-survivin or control isotype antiserum as described above and immunoblotted with an antibody against ubiquitin. Blots were stripped and incubated with anti-Smad3 antibody (Santa Cruz Biotechnology) to ascertain equivalent loading of the lanes. Computer-assisted scanning densitometry (Total Lab, AB.EL Science-Ware Srl) was used to analyze the intensity of the immunoreactive bands.

HT-29 and DLD-1 cells were transfected with DR4, DR5, CHOP, survivin, or control short interfering RNA (siRNA; Santa Cruz Biotechnology) using Lipofectamine 2000 reagent (Invitrogen). The efficiency of the siRNA transfection was assessed using fluorescein-conjugated control siRNA (Santa Cruz Biotechnology).

CT26-derived grafts were generated in BALB/c mice as described (25). After 2 weeks, mice with similar tumor volume, determined by caliper measurements, were divided into 4 groups of 8 mice each. The control group received daily intraperitoneal injections of PBS, the second group received intraperitoneal injection of 1 mg/kg/mouse recombinant mouse TRAIL (catalogue no. 1121-TL/CF; R&D Systems) dissolved in PBS every second day starting from day 15, the third group received intraperitoneal injection of 16 mg/kg/mouse 2-14 dissolved in PBS every second day starting from day 14, and the last group received intraperitoneal injection of both TRAIL and 2-14 at dose and time points indicated above. Mice were sacrificed at day 28. Tumors were excised, photographed, their volume calculated as previously described (25), and used for immunofluorescence and terminal deoxynucleotidyl transferase–mediated dUTP nick end labeling (TUNEL) assay. Studies were approved by the Local Ethics Committee.

Immunofluorescence was conducted by using Tyramide Cy3 and fluorescein systems according to the manufacturer's instructions (Perkin Elmer). DR5 and survivin antibodies were from Santa Cruz Biotechnology. Sections were evaluated using a fluorescence microscope (Olympus). In addition, DR5 and survivin were evaluated in colonic sections of mice with colitis-associated colon cancer, induced by azoxymethane and dextran sulfate sodium, treated or not with 2-14, as previously described (25). Apoptotic cells were evaluated by using a commercially available TUNEL assay (Apoptag Plus; Millipore).

Data were analyzed using the Student's t test for comparison between groups or ANOVA followed by Scheffè's test for multiple comparisons. Significance was defined as values of P < 0.05.

In preliminary studies, we confirmed that colon cancer cells respond differently to TRAIL in terms of apoptosis (30). TRAIL dose dependently enhanced HCT-116 and HT-115 cell apoptosis whereas DLD-1 and, particularly, HT-29 cells were largely resistant against TRAIL-induced cell death (Supplementary Fig. S1A and B). Therefore, DLD-1 and HT-29 cells were selected to examine whether 2-14 sensitizes colon cancer cells to TRAIL. On the basis of these preliminary data, we selected the dose of 50 ng/mL TRAIL for the next studies. Initially, we evaluated whether 2-14 sensitizes colon cancer cells to TRAIL-induced apoptosis. Preincubation of DLD-1 and HT-29 cells with 2-14 significantly increased the percentage of apoptosis following TRAIL treatment (Fig. 1B). Similarly, 2-14 and TRAIL synergized in inducing HCT-116 and HT-115 cell apoptosis (not shown).

To dissect the mechanism by which 2-14 enhances the sensitivity of colon cancer cells to TRAIL, we next looked at the expression of DR4 and DR5. Among the 4 cell lines tested, HT-29 cells expressed the lowest level of DR5 whereas DR4 was expressed at low level in both HCT-116 and HT-29 (Supplementary Fig. S2). The expression of DR4 and DR5 in DLD-1 cells and the percentage of DR4/DR5–positive DLD-1 cells were not affected by mesalamine, whereas the drug slightly increased the number of DR4- and DR5-positive HT-29 cells only when used at 30 mmol/L (Supplementary Fig. S3A and B). Treatment of HT-29 cells with 2-14 significantly enhanced the RNA transcripts for DR4 and DR5 (Supplementary Fig. S4A). In contrast, 2-14 slightly enhanced DcR1 RNA expression only when used at 3 mmol/L but did not alter DcR2 RNA (Supplementary Fig. S4B). Flow cytometric data showed that treatment of HT-29 cells with 2-14 increased the fraction of DR5-positive and, to a lesser extent, DR4-positive cells (Fig. 2A) and this was paralleled by enhanced DR5 but not DR4 MFI expression (Supplementary Fig. S5A). Virtually, all DLD-1 cells expressed DR4 and DR5 (Supplementary Fig. S2), even though they were largely resistant to TRAIL. Therefore, we hypothesized that the ability of 2-14 to promote TRAIL-induced DLD-1 apoptosis could reflect changes in the amount of DR4/DR5. To address this issue, we evaluated the MFI for DR4 and DR5 in 2-14–treated cells. 2-14 enhanced DR5, but not DR4, MFI expression (Supplementary Fig. S5B). Consistently, 2-14 significantly increased DR5 but not DR4 RNA expression (Supplementary Fig. S4A). No change in DcR1 and DcR2 transcripts was seen in DLD-1 following 2-14 exposure (Supplementary Fig. S4B).

DR4/DR5-driven signals lead to activation of caspase-8, an enzyme involved in the TRAIL-dependent apoptotic pathway (6). So, we next monitored caspase-8 activation by flow cytometry. The percentage of HT-29 and DLD-1 cells expressing active caspase-8 was not changed by TRAIL (Fig. 2B). IN contrast, a dose-dependent increase in the fraction of active caspase-8–expressing cells was seen following 2-14 treatment. Combination of 2-14 and TRAIL significantly enhanced the fraction of caspase-8–expressing cells as compared with cells treated with single compounds (Fig. 2B).

To determine whether upregulation of DR4/DR5 by 2-14 plays a role in the sensitization of HT-29 and DLD-1 cells to TRAIL, we tested the effect of 2-14 + TRAIL on the apoptosis of cells transfected with DR4 and/or DR5 siRNA. By using fluorescently labeled siRNA, we initially observed that nearly 75% of the cells were transfected and that less than 5% of transfected cells were PI positive (data not shown). The fraction of cells positive for DR4/DR5 was markedly reduced by specific siRNA in both HT-29 (Fig. 2C) and DLD-1 cells (Supplementary Fig. S6). Knockdown of DR5, but not DR4, abrogated the proapoptotic effect of 2-14 and TRAIL on both cell lines (Fig. 2D), as well as significantly reduced the fraction of apoptotic DLD-1 cells after exposure with TRAIL alone (Fig. 2D, right). Cells with silencing of both DR4 and DR5 exhibited a response to 2-14 and TRAIL not different from that seen in DR5-deficient cells (Fig. 2D). Altogether these data indicate that DR5, and not DR4, is involved in the 2-14–mediated TRAIL-induced apoptosis.

An upstream event in the induction of TRAIL receptors is activation of the mitogen-activated protein kinase pathway (31). In colon cancer cells, phosphorylation of ERK, which correlates with its activation, is followed by the induction of CHOP (32), a transcription factor that directly regulates DR5 expression through a specific binding site in the 5′ flanking region of the DR5 gene (33). ERK was strongly and rapidly phosphorylated following 2-14 treatment in HT-29 (Fig. 3A, top) and DLD-1 cells (Supplementary Fig. S7A) and this activation persisted up to 48 hours (data not shown). 2-14 did not affect p38 phosphorylation but slightly enhanced c-jun-NH,-kinase (JNK) phosphorylation in HT-29 (Fig. 3A, top) but not in DLD-1 cells (Supplementary Fig. S7A). Consistently, 2-14 enhanced CHOP expression in both cell lines (Fig. 3A, bottom and Supplementary Fig. S7B). Pharmacologic inhibition of ERK activation with PD98059 reduced the 2-14–driven CHOP induction (Fig. 3B, top) and DR5 expression (Fig. 3B, bottom) and completely reverted the 2-14–mediated TRAIL-induced apoptosis (Fig. 3C). Silencing of CHOP reverted the 2-14–mediated DR5 upregulation, confirming the involvement of this transcription factor in the upregulation of DR5 following 2-14 treatment (Fig. 3D). Similar results were seen in DLD-1 cells (Supplementary Fig. S7C–F).

The demonstration that DLD-1 cells are resistant against TRAIL-driven apoptosis despite they express DR4 and DR5 suggests that further molecules are involved in the response of colon cancer cells to TRAIL. To this end, we compared the expression of various pro- and antiapoptotic intracellular proteins between TRAIL-resistant (i.e., HT-29 and DLD-1) and TRAIL-sensitive (i.e., HCT-116 and HT-115) cell lines (Supplementary Fig. S8A). Bak and Bax expression was more pronounced in HT-115, DLD-1, and HT-29 cells as compared with HCT-116. Because these are proapoptotic molecules (34), it is unlikely they are involved in the resistance of DLD-1 and HT-29 to TRAIL. Similarly, the reduced expression of the antiapoptotic protein Bcl-2 we detected in DLD-1 and HT-29 is unlikely to be involved in the defective TRAIL-induced apoptosis. DLD-1 cells exhibited enhanced expression of the long and short isoforms of FLIP, an inhibitor of DR signaling (35), whereas this protein was expressed at low level in HT-29 cells (Supplementary Fig. S8A). 2-14 slightly reduced c-FLIP in DLD-1 cells only when used at 3 mmol/L whereas no relevant changes in c-FLIP expression were seen in 2-14–treated HT-29 (Supplementary Fig. S8B), arguing against a major role of this molecule in the 2-14–mediated TRAIL apoptosis. Expression of X-linked inhibitor of apoptosis protein (XIAP) did not differ among the 4 cell lines and was not affected by 2-14 treatment in HT-29 and DLD-1 cells (Supplementary Fig. S8A and B) whereas both Bcl-xL and survivin, 2 antiapoptotic proteins, were upregulated in HT-29 and DLD-1 cells as compared with HCT-116 cells (Supplementary Fig. S8A). 2-14 treatment did not alter Bcl-xL protein expression (Supplementary Fig. S8B). Because survivin is known to interfere with TRAIL-induced cell death in various cancer cells (28, 36, 37) and HT-29 and DLD-1 cells transfected with survivin siRNA became susceptible to TRAIL-induced apoptosis (Fig. 4A), we hypothesized that survivin could be involved in the 2-14–mediated TRAIL-induced apoptosis. 2-14 significantly reduced survivin protein expression in a dose-dependent fashion both in HT-29 and DLD-1 cells (Fig. 4B). Time course studies showed that, in 2-14–treated cells, reduction of survivin protein expression (Fig. 4C) occurred earlier (i.e., 8 hours) than inhibition of RNA transcripts (i.e., 32 hours; Fig. 5A), thus suggesting a posttranscriptional control of survivin expression by 2-14. Indeed, in line with previous studies (38), we next showed that 2-14 enhanced the ubiquitination of survivin (Fig. 5B) and treatment of cells with the proteasome inhibitor, lactacystin, prevented the 2-14–induced survivin downregulation (Fig. 5C).

To translate these observations into mice, we assessed whether 2-14 and TRAIL synergized in an in vivo syngeneic colon cancer model in which tumors are generated by injecting the murine colon cancer cell line, CT26, into BALB/c mice. Initially, we showed that TRAIL did not induce CT26 cell apoptosis (Supplementary Fig. S9A). 2-14 restored the susceptibility of CT26 cells to TRAIL (Supplementary Fig. S9B). Like DLD-1 and HT-29 cells, CT26 responded to 2-14 by enhancing the level of phosporylated ERK, CHOP, and DR5 (Supplementary Fig. S10A–C). 2-14 slightly enhanced the expression of DcTRAIL-R1, the murine homologous of human DcR1, without altering the expression of DcTRAIL-R2, the murine homologous of human DcR2 (Supplementary Fig. S10C). Moreover, 2-14 reduced survivin protein level (Supplementary Fig. S10D), without affecting survivin RNA expression (data not shown). In this context, it is noteworthy that CT26 do not express DR4 (39).

Subsequently, mice with CT26-derived grafts were injected with TRAIL and/or 2-14. The graft tumor volume of mice receiving TRAIL alone did not significantly differ from that seen in mice receiving PBS. In contrast, mice treated with 2-14 exhibited a significant decrease in the tumor volume in comparison with controls (Fig. 6A). The tumor growth of mice receiving both 2-14 and TRAIL was significantly reduced as compared with mice treated with 2-14 alone (Fig. 6A). Mice treated with 2-14 and TRAIL exhibited enhanced cell apoptosis (Fig. 6B). In line with the in vitro observations, grafts of mice treated with 2-14, either alone or in combination with TRAIL, showed DR5 upregulation and survivin reduction (Fig. 6C). To confirm this later finding in another model of colon cancer, we evaluated whether 2-14 modulates DR5 and survivin in the azoxymethane/dextran sulfate sodium-induced colitis-associated colon cancer. Mice treated with 2-14 exhibited a marked DR5 upregulation in the tumoral areas, which was associated with diminished expression of survivin (Supplementary Fig. S11).

In recent years, a considerable amount of work has been done to elucidate mechanisms that sustain cancer cell growth and survival. This progress has contributed to show that neoplastic cells are largely resistant against apoptosis induced by physiologic stimuli. One such defect involves the TRAIL-mediated cell death program. Indeed, many cancer cells, including colon cancer cells, exhibit an altered expression of molecules that interfere with TRAIL-induced apoptosis. Therefore, compounds that overcome colon cancer cell resistance to TRAIL could improve the way we manage colon cancer patients.

In this study, we showed that 2-14, a derivative of mesalamine, restored the susceptibility of DLD-1 and HT-29 cells to TRAIL-induced apoptosis, and this phenomenon was associated with enhanced expression of DR5 and, to a lesser extent of, DR4. However, silencing of DR5, but not DR4, abrogated the effect of 2-14 on TRAIL-induced apoptosis in HT-29 and DLD-1. These data are not surprising because it is known that DR5 binds TRAIL with greater affinity than DR4 (5) and that, even in the absence of DR4, DR5 is sufficient to deliver TRAIL-induced apoptotic signals (27, 40, 41). DR5 expression can be positively regulated by the transcription factor PPAR-γ (42), and by p53 and CHOP, 2 proteins that bind to and enhance DR5 promoter activity (43). High PPAR-γ protein expression in colon cancer cells associates with enhanced activity of this transcription factor (44). We were not able to see any change in PPAR-γ protein expression in HT-29 and DLD-1 cells following treatment with 2-14 (personal unpublished observations), arguing against a role of PPAR-γ in the 2-14–driven DR5 increase. It is also unlikely that p53 is involved in the 2-14–induced DR5 upregulation in DLD-1 and HT-29 cells, as these 2 cell lines lack a functional p53. In contrast, our data suggest that activation of the ERK/CHOP pathway can play a key role in upregulating DR5 in DLD-1 and HT-29 cells. Indeed, 2-14 enhanced ERK phosphorylation in these cells and this event was associated with CHOP induction. Notably, pharmacologic inhibition of ERK pathway prevented the 2-14–induced DR5/CHOP expression and TRAIL-driven apoptosis. The involvement of CHOP in our system was confirmed by knockdown experiments showing that silencing of CHOP reverted the 2-14–induced DR5 upregulation.

We also showed that 2-14–treated cells exhibited enhanced expression of active JNK, which has been associated with DR4 induction in other systems (31). However, as pointed out above, knockdown of DR4 did not interfere with the ability of 2-14 to make colon cancer cells sensitive to TRAIL. Moreover, pharmacologic inhibition of JNK did not affect TRAIL-induced apoptosis in our system (personal unpublished observation), arguing against a role of JNK pathway in the 2-14–mediated TRAIL-induced apoptosis.

The demonstration that DLD-1 cells were resistant against TRAIL-induced apoptosis despite their elevated expression of DR5 suggests that, at least in this cell line, further molecules contribute to make DLD-1 resistant to TRAIL. To address this issue, we analyzed the expression of several pro- and antiapoptotic molecules in colon cancer cells, which are either resistant or sensitive to TRAIL. The most consistent finding we observed was the increased expression of Bcl-xL and survivin, 2 antiapoptotic molecules, in DLD-1 and HT-29 cells. However, 2-14 reduced survivin but not Bcl-xL expression. These observations together with the demonstration that silencing of survivin restored the sensitivity of DLD-1 and HT-29 cells to TRAIL strongly suggest that 2-14–mediated survivin downregulation is a crucial event in the induction of colon cancer cell apoptosis following treatment with 2-14 and TRAIL. This hypothesis is consistent with previously published studies showing that high survivin is seen in the majority of human cancers (45, 46), where this protein is supposed to act as an opposition factor to various anticancer therapies (36, 37). Time course studies showed that reduction of survivin protein was seen as early as 8 hours following 2-14 treatment whereas inhibition of survivin RNA expression occurred at later time points (i.e., 32 hours). It is thus likely that regulation of survivin by 2-14 occurs at posttranscriptional level. Indeed, by immunoprecipitation and Western blotting, we next showed that 2-14 increased survivin ubiquitination and that lactacystin, a proteasome inhibitor, prevented the 2-14–induced survivin downregulation. Our data are consistent with previous studies showing that survivin can be polyubiquitinated in vivo at lysine residues and degraded by the ubiquitin/proteasome pathway (38) and that antitumoral drugs can promote proteasome-mediated survivin degradation (28). We were not, however, able to show that proteasome inhibitors reverted the effect of 2-14 on the TRAIL-induced apoptosis (personal unpublished observation). This probably might rely on the fact that inhibition of proteasome pathway can alter the expression/activity of additional proteins involved in the control of cell death.

Next, we translated our data in vivo in mice using 2 distinct models of colon cancer. Initially, we confirmed that 2-14 enhanced DR5 expression and downregulated survivin in CT26-derived xenografts. In this model, the combined therapy with 2-14 and TRAIL enhanced the colon cancer cell apoptosis and caused a more pronounced tumor suppression than that seen with single compounds. Similar results were seen in a murine model of colitis-associated colon cancer in which the antineoplastic effect of 2-14 associated with the induction of DR5 and downregulation of survivin.

In conclusion, our data show that 2-14 sensitizes colon cancer cells to TRAIL-induced apoptosis in vitro and in vivo and this may be a novel therapeutic strategy for colon cancer that are resistant against TRAIL-based therapies. Further studies are, however, needed to define the pharmacokinetic properties of 2-14 and the effects of the combined treatment with 2-14 and TRAIL on vital functions of the host.

No potential conflicts of interest were disclosed.

This work received support from the “Fondazione Umberto di Mario,” Rome, Italy, Giuliani SpA, and Associazione Italiana per la Ricerca sul Cancro (A.I.R.C. no. 9148) to G. Monteleone.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.