The HSP70 and Autophagy Inhibitor Pifithrin-μ Enhances the Antitumor Effects of TRAIL on Human Pancreatic Cancer

Members of the TNF cytokine family, such as TNF-α and the Fas ligand, play important roles in inflammation and immunity (1). However, their use in anticancer therapy is limited because they cause severe cytotoxicity to normal cells. In contrast, TRAIL, a member of the TNF superfamily, can induce apoptosis in cancer cells while causing almost no cytotoxicity to normal cells (2). Therefore, TRAIL has been applied for several clinical trials (3, 4). TRAIL binds to death receptors (DR) and activates caspases, leading to apoptosis. However, some cancer cells acquire TRAIL resistance (5). The binding of TRAIL to death receptors activates NF-κB, PKC, mitogen-activated protein kinase (MAPK), and AKT signaling (6, 7), thereby promoting cell proliferation and apoptosis resistance. Notably, activation of NF-κB signaling is an important TRAIL resistance mechanism (6, 8). Indeed, combination therapy with TRAIL plus NF-κB inhibitors has been reported to be a promising treatment strategy (9–11).

Cancer cells acquire therapy resistance through a variety of mechanisms. One is increased expression of HSP70 (12). In contrast to the very low level in unstressed normal cells, HSP70 expression is rapidly increased in response to various stresses (13, 14). Importantly, increased expression of HSP70 in cancer cells has been reported to be associated with malignant features and poor prognosis in patients with cancer (15). Autophagy, meanwhile, has received much attention as a mechanism of therapy resistance in cancer cells. Its fundamental role in cells is cytoprotection under starvation and stress conditions (16), but this protective function can render cancer cells therapy-resistant (17). Indeed, many reports suggest that inhibition of autophagy in cancer cells can restore their susceptibility to anticancer therapy (18–20). Recently, we reported that autophagy inhibits apoptosis of human prostate and breast cancer cells treated with an innate adjuvant receptor ligand (21, 22). These lines of evidence suggest that HSP70 and autophagy are promising targets for therapies intended to enhance TRAIL-induced antitumor effects, something that has not yet been fully investigated.

Pifithrin (PFT)-μ (2-phenylethynesulfonamide) was initially identified as a small-molecule inhibitor of p53 binding to mitochondria (23). Recently, it was revealed to interact selectively with HSP70 and to inhibit its functions (24). In addition, it induces altered autophagy, leading to inhibition of the later autophagic pathway. Furthermore, PFT-μ inhibits the degradation of IκBα, resulting in inhibition of the NF-κB pathway. This information led us to test the possibility that PFT-μ enhances TRAIL-induced antitumor effects in human pancreatic cancer. In this study, after confirming that HSP70 and autophagy play protective roles in TRAIL-induced antitumor effects, we showed that the combination of TRAIL and PFT-μ decreased the viability and colony-forming ability of pancreatic cancer cells and increased their cell death. Using a xenograft mouse model, we also showed that the combination therapy significantly decreased pancreatic tumor growth as compared with treatment with either agent alone. These results suggest that PFT-μ is a promising enhancer of TRAIL-induced antitumor effects on human pancreatic cancer.

Four human pancreatic cancer cell lines (MiaPaca-2, Panc-1, AsPC-1, and BxPC-3), which were kindly provided by Dr. K. Takenaga (Shimane University Faculty of Medicine, Shimane, Japan), were maintained in Dulbecco's Modified Eagle's Medium (Sigma-Aldrich) supplemented with 10% fetal calf serum (FCS; Invitrogen) and 20 μg/mL gentamicin (Sigma-Aldrich) at 37°C in a humidified atmosphere containing 5% CO₂. No authentication was done by the authors.

Cell viability was evaluated using the 2-(2-methoxy-4-nitrophenyl)-3-(4-nitrophenyl)-5-(2, 4-disulfophenyl)-2H-tetrazolium monosodium salt (WST-8) assay (Nacalai Tesque). Briefly, cells were seeded in flat-bottomed 96-well plates. The next day, TRAIL (PeproTech Inc.) and/or PFT-μ (Santa Cruz Biotechnology, or Cayman Chemical) were added. Two days later, WST-8 was added to each well and the plates were read at a wavelength of 450 nm after 3 hours. For inhibition assays, z-VAD-fmk (R&D Systems) was added and dimethyl sulfoxide (DMSO) was used as a vehicle control.

Cell death was measured using the Annexin V–FITC Apoptosis Detection Kit (BioVision) and propidium iodide (PI). To examine the cell cycle and proliferation of cancer cells, a bromodeoxyuridine (BrdUrd)/7-amino-actinomycin D (7AAD) Proliferation Kit (Becton Dickinson) was used according to the manufacturer's instructions. Analysis was conducted using a FACSCalibur flow cytometer (Becton Dickinson).

Cells were lysed with a mammalian protein extraction reagent (M-PER; Thermo Scientific) containing a protease inhibitor cocktail (Nacalai Tesque). Equal amounts of protein were resolved on 4% to 12% gradient or 12% SDS-PAGE gels and then transferred to polyvinylidene fluoride membranes. The membranes were blocked and the blots then incubated with the following primary antibodies: anti-LC3 (MBL), anti-cyclin D1 (Cell Signaling Technology), anti-c-Myc (Epitomics), anti-beclin-1 (Cell Signaling Technology), anti-cathepsin L (Santa Cruz Biotechnology), anti-IκBα (Cell Signaling Technology), anti-β-actin (BioLegend), and anti-α-tubulin (Santa Cruz Biotechnology). Goat anti-rabbit and goat anti-mouse alkaline phosphatase–conjugated secondary antibodies (Invitrogen) were used to detect the primary antibodies.

Transfection of siRNA was conducted using Lipofectamine RNAiMAX (Invitrogen), according to the manufacturer's instructions. HSP70 siRNA (sc-29357) and beclin-1 siRNA (sc-29797) were purchased from Santa Cruz Biotechnology. Control siRNA (#6568) was purchased from Cell Signaling Technology. Three days after siRNA transfection, cancer cells were used for subsequent experiments.

LC3B (NM_022818) was amplified by PCR and inserted into the pcDNA3.1/NT-GFP-TOPO vector (Invitrogen) in frame with the GFP sequence. Transfection of plasmids was conducted using Lipofectamine 2000 (Invitrogen), according to the manufacturer's instructions. Cells were cultured on round microscope cover glasses in 24-well plates with the indicated reagents for 2 days. After incubation with Hoechst 33342 (5 μg/mL) for 30 minutes, cells were fixed with 3% formalin and placed on slide glasses with 4 μL of mounting medium for fluorescence (Vectashield; Vector Laboratories, Inc.). To examine NF-κB translocation to the nucleus, cells were cultured on round microscope cover glasses in 24-well plates for 1 day. After incubation with PFT-μ (20 μmol/L) for 5 hours, cells were additionally cultured in the presence of TRAIL and Hoechst 33342 (5 μg/mL) for 30 minutes. After fixation and permeabilization with 3% formalin and 1% Triton X, respectively, cells were stainined with anti-NF-κB p65 antibody (Cell Signaling Technology), followed by Alexa Fluor 488–conjugated anti-rabbit immunoglobulin G (IgG) F(ab′)2 fragment (Cell Signaling Technology). Confocal imaging was conducted using an Olympus FV1000-D laser scanning microscope (Olympus).

Cancer cells were seeded in 6-well plates in the presence or absence of TRAIL and/or PFT-μ. Two days later, the medium was replaced with medium that contained no reagent and the culture was continued for an additional 10 days. Thereafter, colonies were counted after fixation with methanol and staining with 0.05% crystal violet.

BALB nu/nu female mice, purchased from CLEA Japan Inc., were maintained under specific pathogen-free conditions. Experiments were carried out according to the ethical guidelines for animal experimentation of the Shimane University Faculty of Medicine (approval number: IZ24-5). Mice were inoculated in the right flank with 4 × 10⁶ MiaPaca-2 cells with Matrigel (Japan BD Biosciences) at a volume ratio of 1:1. On day 20, the mice were pooled and divided into 4 groups. On days 1, 2, and 3 after grouping, the mice were injected intraperitoneally with PFT-μ (100 μL). As a vehicle control, 100 μL of DMSO was injected. On day 2 after grouping, the mice were injected once intratumorally with TRAIL (50 μL). As a vehicle control, 50 μL of culture medium was injected. Thereafter, tumor size was measured twice weekly.

Data were evaluated statistically using an unpaired two-tailed Student t test or ANOVA with Scheffe post hoc test. A P value of less than 0.05 was considered to indicate statistical significance.

First, we examined the expression of DR4 and DR5 on 4 pancreatic cancer cell lines. Although the expression of DR4 was relatively low on 3 cell lines and was negative on Panc-1, all cell lines were strongly positive for DR5 (Supplementary Fig. S1). We next determined the TRAIL susceptibility of these cell lines (Fig. 1A). MiaPaca-2, BxPC-3, and Panc-1 cells were highly, moderately, and lowly sensitive to TRAIL, respectively, and their susceptibility was dose-dependent. In contrast, AsPC-1 cells were entirely resistant to TRAIL. In subsequent experiments, we focused on MiaPaca-2 and Panc-1 cells.

To elucidate the roles of HSP70 in TRAIL-induced antitumor effects on human pancreatic cancer cells, we compared the sensitivities of MiaPaca-2 and Panc-1 cells in which HSP70 expression was blocked by RNA interference (Fig. 1B). Knockdown of HSP70 decreased the viability of MiaPaca-2 and Panc-1 cells treated with TRAIL (Fig. 1C). Knockdown of HSP70 markedly increased the percentage of Annexin V⁺ Panc-1, but not MiaPaca-2, cells upon TRAIL treatment (Fig. 1D). As shown in Fig. 1E, knockdown of HSP70 increased Annexin V⁺/PI⁻ early apoptotic and Annexin V⁺/PI⁺ late apoptotic and/or necrotic Panc-1 cells upon TRAIL treatment. A notable finding is that Panc-1 cells, which were less sensitive to TRAIL, became sensitive to low-dose TRAIL after knockdown of HSP70. Overall, these results indicate that HSP70 plays a protective role in TRAIL-induced antitumor effects on human pancreatic cells.

We next tested the possibility that autophagy plays a protective role in the TRAIL resistance of pancreatic cancer cells. LC3 exists in 2 forms: LC3-type I, which is cytosolic, and its proteolytic derivative, LC3-type II, which localizes to the autophagosomal membrane (25). As shown in Fig. 2A, TRAIL treatment increased the expression of LC3-type II in MiaPaca-2 cells, whereas LC3-type II was expressed in Panc-1 cells without TRAIL treatment. We also assessed autophagy by confocal imaging of LC3 foci in GFP-LC3 fusion protein-expressing cancer cells (Fig. 2B). These 2 cell lines were transiently transfected with a plasmid-encoding GFP-LC3, and GFP-LC3 foci were subsequently examined. No GFP-LC3 foci were detected in control GFP/NT–transfected MiaPaca-2 and Panc-1 cells. Although GFP-LC3 foci were detected in GFP-LC3–transfected MiaPaca-2 and Panc-1 cells, even without TRAIL treatment, GFP-LC3 foci seemed to increase in both number and size after TRAIL treatment. To elucidate the role of autophagy in TRAIL-induced antitumor effects, the expression of beclin-1, a molecule required for autophagy, was knocked down by transfection of beclin-1 siRNA (Fig. 2C). Knockdown of beclin-1 decreased the viabilities and increased the percentages of Annexin V⁺ MiaPaca-2 and Panc-1 cells upon TRAIL treatment (Fig. 2D and E). Collectively, these results indicate that, similar to HSP70, autophagy plays a protective role in TRAIL-induced antitumor effects on pancreatic cancer cells.

Because both HSP70 and autophagy were suggested to function protectively upon TRAIL treatment of pancreatic cancer cells, we next determined whether TRAIL-induced antitumor effects were enhanced by PFT-μ, which inhibits both the functions of HSP70 and autophagy (24), as described in the Introduction. Figure 3A shows the structure of PFT-μ. As shown in Fig. 3B, when administered alone, PFT-μ dose-dependently decreased the viability of all 4 pancreatic cancer cell lines. When suboptimal doses of TRAIL and PFT-μ were combined, additive effects were evident in 3 cell lines other than AsPC-1 (Fig. 3C). Combined treatment with suboptimal doses of TRAIL and PFT-μ significantly decreased the colony-forming ability of MiaPaca-2 and Panc-1 cells compared with those treated with either agent alone (Fig. 3D). Because the continuous presence of PFT-μ for 12 days drastically inhibited colony formation, we cultured cancer cells for 2 days with TRAIL and/or PFT-μ and subsequently for 10 days without reagents.

In the studies described earlier, we examined the antitumor effects on pancreatic cancer cells by measuring viability 2 days after the administration of TRAIL and/or PFT-μ. However, such effects on viability may reflect alterations in cell death and/or growth. Therefore, we next examined the underlying mechanism of action of PFT-μ in detail. Although suboptimal doses of TRAIL slightly increased Annexin V⁺ MiaPaca-2 and Panc-1 cells, PFT-μ moderately increased the percentages of Annexin V⁺ cells (Fig. 4A). Combining PFT-μ with TRAIL further increased the number of Annexin V⁺ cells. Because PFT-μ was reported to induce caspase-independent apoptosis (24), we attempted to confirm the result in our experimental systems. TRAIL-induced increase of Annexin V⁺ MiaPaca-2 and Panc-1 cells was completely inhibited by the pan-caspase inhibitor z-VAD, whereas z-VAD rescued in part PFT-μ–induced cell death (Fig. 4B). This result indicates that PFT-μ increased Annexin V⁺ cells in both caspase-dependent and -independent manners. We next determined whether growth arrest was involved in the antitumor effects of PFT-μ by evaluating BrdUrd uptake and 7AAD staining. PFT-μ significantly decreased the percentages of BrdUrd⁺ S-phase MiaPaca-2 and Panc-1 cells (Fig. 4C and D) and increased the apoptotic sub-G₁ fraction in MiaPaca-2 cells. We also examined the expression of proliferation-related proteins in MiaPca-2 and Panc-1 cells and found that PFT-μ decreased cyclin D1 expression in MiaPaca-2 cells but not Panc-1 cells. Collectively, these results indicate that PFT-μ induces 2 types of antitumor responses in pancreatic cancer cells: cell death, both caspase-dependent and -independent, and cell-growth arrest.

We next attempted to elucidate the mechanisms by which PFT-μ enhanced TRAIL-induced antitumor effects. To evaluate the influence of PFT-μ on the autophagy system, we determined its effect on the degradation of pro-cathepsin L (Fig. 5A). PFT-μ increased the expression of LC3 type II. However, this increase was not the result of enhanced autophagy but rather inhibition of the autophagy-lysosomal system, because degradation of pro-cathepsin L to cathepsin L was clearly inhibited, as reported previously (24). We also examined the effect of PFT-μ on NF-κB signaling in TRAIL-treated cancer cells, because activation of NF-κB is a major mechanism of TRAIL resistance of cancer cells (6, 8). As expected, TRAIL significantly decreased the level of IκBα, indicating activation of the NF-κB pathway. Although the IκBα level was decreased in cells treated with PFT-μ alone, PFT-μ restored IκBα expression in cells treated with TRAIL (Fig. 5B). TRAIL and PFT-μ had no marked effect on HSP70 expression in cancer cells. We next directly confirmed that the combination treatment inhibited the NF-κB pathway using confocal imaging. Although the treatment with either of TRAIL or PFT-μ induced NF-κB translocation to the nucleus, the combination treatment decreased the expression of NF-κB in the nucleus (Fig. 5C). These results suggest that PFT-μ can inhibit the autophagy–lysosomal system and antagonize TRAIL-associated NF-κB activation in pancreatic cancer cells.

Finally, we evaluated whether combination therapy with TRAIL plus PFT-μ exerted an antitumor effect against established human pancreatic cancer in a xenograft mouse model. Nude mice were inoculated with MiaPaca-2 cells and were grouped when tumor diameters reached 8 to 9 mm. After grouping, PFT-μ (25 mg/kg) was injected intraperitoneally on days 1, 2, and 3 and TRAIL was injected locally on day 2. No change in body weight resulted, suggesting no severe adverse event (data not shown). Although the systemic administration of PFT-μ had no antitumor effect and local injections of TRAIL decreased tumor growth moderately, but not significantly, combination therapy with TRAIL plus PFT-μ significantly suppressed tumor growth (Fig. 6A and B).

Because pancreatic cancer is highly resistant to conventional anticancer therapies and is associated with a very poor prognosis (26), new treatment modalities to enhance the efficacy of current treatments are required. In this study, we investigated the possibility that PFT-μ, a small-molecule HSP70 inhibitor that has the ability to alter autophagy (24), could enhance TRAIL-induced antitumor effects on human pancreatic cancer cells. We found that both HSP70 and autophagy are, at least in part, responsible for the TRAIL resistance of cancer cells and that PFT-μ enhances TRAIL-induced antitumor effects on human pancreatic cancer cells.

HSP70 is a potent heat-inducible survival protein that confers cytoprotection against various death-inducing stimuli and increases tumorigenicity (27–29). It has been suggested to be a promising target in cancer treatment (30). In this study, knockdown of HSP70 significantly decreased cell viability and increased the percentage of Annexin V⁺ pancreatic cancer cells after TRAIL treatment (Fig. 1). The increase of Annexin V⁺ cells in response to TRAIL was stronger in the less TRAIL-sensitive Panc-1 cells than in highly TRAIL-sensitive MiaPaca-2 cells. Although only one report suggests that TRAIL-induced apoptosis in cancer cells is enhanced by HSP70 (31), many reports have shown that HSP70 inhibits TNF-α- and Fas-mediated apoptosis in cancer cells (32–34). In this study, we have shown that siRNA knockdown of HSP70 increased Annexin V⁺ Panc-1 cells after TRAIL treatment (Fig. 1). To our knowledge, this is the first report that HSP70 contributes to the TRAIL resistance of human pancreatic cancer cells. Another intriguing finding is that autophagy is protective in TRAIL-treated pancreatic cancer cells. This result is compatible with other reports (20). Autophagy has received much attention in various cell biology fields (16, 17). Despite the reports of autophagic cell death (18, 35, 36) and autophagy-dependent antitumor immune response (37), the fundamental role of autophagy is thought to be cytoprotection under starvation and stress conditions (16). Many reports suggest that autophagy functions cytoprotectively in cancer cells (19, 20, 36, 38). In line with this, we previously reported that autophagy protects against apoptosis in human prostate and breast cancer cells after treatment with an innate adjuvant receptor ligand (21, 22). Therefore, in this study, we investigated the participation of autophagy in the TRAIL resistance of pancreatic cancer cells. Autophagy was constitutively induced in MiaPaca-2 and Panc-1 cells, and TRAIL treatment apparently augmented autophagy in MiaPaca-2 cells. We also showed that autophagy inhibition by knockdown of beclin-1 increased the susceptibility of pancreatic cancer cells to TRAIL (Fig. 2). Our findings about the protective roles of HSP70 and autophagy in TRAIL treatment encouraged us to use the HSP70 and autophagy inhibitor PFT-μ to enhance TRAIL-induced antitumor effects on pancreatic cancer cells.

Death receptor–mediated TRAIL signaling is involved in activation of the NF-κB pathway, leading to TRAIL resistance (6, 8). Indeed, the proteasome inhibitor bortezomib enhances TRAIL-induced antitumor effects by inhibiting NF-κB (9–11). Interestingly, PFT-μ decreases IκBα degradation by inhibiting the autophagy–lysosomal system (24). PFT-μ also inhibits the proteasome system (39). We then determined whether PFT-μ inhibited TRAIL-induced activation of the NF-κB pathway. As expected, TRAIL decreased IκBα expression in cancer cells and cotreatment with PFT-μ inhibited IκBα degradation. In addition, confocal imaging revealed that the combination treatment inhibited NF-κB translocation to the nucleus. This may be the third mechanism by which PFT-μ augmented TRAIL-induced antitumor effects on pancreatic cancer cells.

How does HSP70 inhibit TRAIL-induced cell death? HSP70 interacts with Apaf-1 to prevent caspase activation (40, 41). Interestingly, HSP70 has been reported to localize to the membranes of lysosomes, promote cancer cell viability, and inhibit TNF-induced cell death by inhibiting lysosomal membrane permeabilization (42). Thus, HSP70 enhances survival by stabilizing the lysosomes in cancer cells. Because PFT-μ binds HSP70, inhibiting the autophagy–lysosomal system, it is possible that PFT-μ inhibits HSP70-induced stabilization of lysosomal membrane permeabilization, resulting in increased cell death. We plan to investigate this possibility.

HSP70 can bind to apoptosis-inducing factor (AIF), which induces caspase-independent apoptosis by translocation into the nucleus (43). As PFT-μ induces cell death in both caspase-dependent and -independent manners in cancer cells (Fig. 4B), we evaluated the role of AIF in PFT-μ–induced caspase-independent cell death. However, RNA interference of AIF had no effect on PFT-μ–induced cell death of pancreatic cancer cells (data not shown); thus, AIF likely does not participate in PFT-μ–induced cell death.

There are several possible explanations about how autophagy protects cancer cells from TRAIL-induced cell death. First, autophagy degrades caspases. Indeed, autophagy can degrade active caspase-8 (44), which is required for extrinsic signal-mediated caspase-dependent apoptosis. Second, although we cannot exclude the possibility of participation of the proteasome system, the autophagy–lysosomal system degrades IκBα, resulting in activation of the NF-κB pathway. Several reports have indicated that activation of this pathway can cause cancer cells to become resistant to TRAIL (6, 8). Interestingly, a recent report suggests that the mitochondrial tumor suppressor ARF interacts with HSP70 and that PFT-μ can reduce the ability of ARF to induce autophagy by selectively interacting with HSP70 and blocking ARF trafficking to mitochondria (45). ARF thus seems to play a crucial role in the altered autophagy induced by PFT-μ. We plan to elucidate this mechanism in more detail.

Besides cell death, PFT-μ induced growth arrest of pancreatic cancer cells. PFT-μ treatment significantly decreased the S-phase fraction in 2 cancer cell lines (Fig. 4C). The decreased cyclin D1 expression might, at least in part, account for PFT-μ–induced cell growth arrest. We conducted a long-term (12-day) colony-formation assay. A suboptimal dose of PFT-μ significantly decreased the colony-forming ability of MiaPaca-2 cells. These data support the hypothesis that PFT-μ arrests cell growth.

Among the 4 pancreatic cancer cell lines, only AsPC-1 was resistant to TRAIL and PFT-μ showed no sensitizing effects on AsPC-1. Thus, some TRAIL susceptibility seems to be necessary for the sensitizing effects of PFT-μ. In addition, only AsPC-1 was entirely TRAIL-resistant even though there was no difference in DR4 and DR5 expression among the 4 cell lines (Supplementary Fig. S1). Although we have not elucidated the underlying mechanisms, AsPC-1 may have been preferentially positive for nonfunctional decoy receptors or antiapoptotic molecules including bcl-2, bcl-xL, and c-FLIP, as reported previously (46, 47).

In conclusion, we show here that PFT-μ has the potential to induce 2 types of antitumor response; that is, cell death and cell growth arrest, in human pancreatic cancer cells and that it enhances TRAIL-induced antitumor effects both in vitro and in vivo. These results suggest that PFT-μ shows promise for use as a therapy intended to enhance the antitumor effects of TRAIL or agonistic antibodies against death receptors.

No potential conflicts of interest were disclosed.

Conception and design: H. Monma, M. Harada

Development of methodology: S. Okano

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): H. Monma, N. Harashima, M. Harada

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): H. Monma, M. Harada

Writing, review, and/or revision of the manuscript: H. Monma, Y. Tajima, M. Harada

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): N. Harashima, T. Inao

Study supervision: M. Harada

The authors thank Ms. Tamami Moritani for her technical assistance.

This study was supported in part by Grants-in aid for Scientific Research from the Ministry of Education, Culture, Sports, Science, and Technology, Japan (no. 18591449 to M. Harada and no. 23701074 to N. Harashima), and from Shimane University “S-TAKUMI Medical Nanotechnology” Project (M. Harada).

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.