MUC1-Mediated Metabolic Alterations Regulate Response to Radiotherapy in Pancreatic Cancer

Pancreatic cancer is the fourth leading cause of cancer-related deaths in the United States (1). It has the lowest survival rate of 6% and is projected to be the second leading cause of cancer-related deaths by 2030 (1, 2). Current therapeutic options for advanced pancreatic cancer include chemotherapy and radiation therapy (RT). However, these options provide only marginal increases in the survival rate due to the therapeutic resistance of pancreatic tumors (3). Clinical trials indicate a significant response to radiation in only 20% of primary pancreatic tumors (4, 5). Although multiple factors cause resistance to RT, the biological mechanisms mediating such innate resistance are currently being explored.

Pancreatic tumors present elevated DNA damage responses (6). Also, pancreatic cancer cell models can be sensitized to radiation through inhibition of the nonhomologous end joining (NHEJ) mode of DNA repair (7). Efficient DNA damage responses would depend on anabolic alterations that could provide cancer cells with nucleotide pools upon radiation-induced DNA damage (8). Pancreatic tumors are highly glycolytic, as indicated by high [¹⁸F]-fluorodeoxyglucose uptake in positron emission tomography (9–11). The success of RT in pancreatic cancer patients inversely correlates to the overall glycolytic index of the tumors (12, 13). Although the exact role of glycolysis in imparting radiation resistance remains poorly defined, the increased glycolytic index may provide cells with increased nucleotide pools by routing carbon flux through the pentose phosphate pathway (PPP). Factors that could mediate the radiation-induced DNA repair mechanisms through upregulation of nucleotide metabolism are least studied in pancreatic cancer.

Mucin1 (MUC1) is a transmembrane glycoprotein that is overexpressed and aberrantly localized in pancreatic ductal adenocarcinoma (14, 15). MUC1 serves as a sensor of extracellular signaling cues and transmits these signals to the nucleus, modulating the transcriptional profile of cancer cells (15). These functions allow MUC1 to facilitate tumor cell growth, invasion, motility, and cell survival under harsh conditions. Furthermore, MUC1 mediates DNA damage response by interacting with ATM in breast cancer cells (16). MUC1 also interacts with c-Abl and prevents its nuclear targeting to block DNA damage–induced apoptosis (17). Besides its role in mediating DNA damage repair, MUC1 also facilitates metabolic reprogramming in pancreatic cancer cells (18, 19). We previously reported that the MUC1 cytoplasmic domain directly interacts with and stabilizes hypoxia-inducible factor-1 alpha (HIF-1α), which can, in turn, alter the metabolic gene expression in pancreatic cancer cells (18, 20). We also identified that MUC1-mediated metabolic alterations induce increased flux of carbon through the PPP, which primarily generates the precursor metabolites for de novo nucleotide synthesis in cancer cells (18, 21). Based on our previous studies, we hypothesized that MUC1-mediated metabolic alterations play a role in radiation resistance in pancreatic cancer cells. Our present study demonstrates that radiation resistance in pancreatic cancer is dependent on MUC1-mediated metabolic flux. We utilized mass spectrometry–based metabolomics technology and pharmacologic inhibition of MUC1-induced aberrant metabolism to decipher the role of MUC1-mediated metabolic alterations in imparting radiation resistance to pancreatic cancer.

Pancreatic cancer cell lines S2-013 and Capan2 with MUC1 overexpression (S2-013.MUC1) and MUC1 knockdown (Capan2-sh.MUC1) have been described previously (18). Additional MUC1 knockdowns of FG and HPAF2 were developed using established shMUC1 constructs. Osteosarcoma cell line U2OS (U2OS SA-GFP) was provided by Dr. Jeremy Stark (Beckman Research Institute of the City of Hope, Duarte, CA; ref. 22). After transfection of pcDNA3.MUC1F construct (23) into these cells, the transfected cells were cultured in pyruvate-free DMEM, supplemented with 10% FBS and 1 mg/mL G418, for 7 to 14 days. G418-resistant cells were isolated and utilized for the experiments. An empty vector was stably transfected to establish the control cell lines. The cell lines were validated by STR profiling and tested for mycoplasma every 6 months.

Cells maintained in DMEM with 10% FBS were irradiated by utilizing a linear accelerator available in the Department of Radiation Oncology at UNMC. Briefly, cells were seeded in culture dishes 16 hours before irradiation. Irradiation performed by placing culture plates on 10 cm of solid water (phantom material used for radiation beam calibration) by positioning plates in the center of the 40 cm × 40 cm radiation field with dose verifications was evaluated using Metal Oxide Semiconductor Field Effect Transistor detectors which enables the dose verification through the radiation-induced threshold shifts. Irradiation was conducted with 6 MV X-rays at a rate of 2.73 Gy/min from the posterior direction, with the cells on the flask base being 100 cm from the X-ray source.

Cells irradiated with fractionated or single-dose radiation were trypsinized and subsequently cultured for cell survival assays. The medium was replaced with fresh DMEM with or without inhibitors before and immediately after irradiation. Cells cultured in the respective media for 72 hours were evaluated for cell survival using either MTT or trypan blue exclusion method using BIO-RAD TC20 automated cell counter.

Clonogenic assays were performed to determine cellular response to radiation (24). Experimental (MUC1-knockdown and -overexpressed cell line models) and control cells with similar seeding densities were chosen for clonogenic assays, and colonies formed at the end of experiments were washed, fixed in methanol, and stained with 0.4% crystal violet in 25% methanol. Colonies containing >50 cells for each well were counted (in triplicate). Surviving fraction at each dose was determined by using the formula: [(number of surviving colonies in dose X)/(number of cells seeded for dose X (average colonies arising from the nonirradiated cells (0 Gy)/number of nonirradiated cells seeded); 25].

Cells were washed twice with cold PBS and lysed in radioimmunoprecipitation assay lysis buffer at 4°C for 30 minutes. Pellets were separated by centrifugation at 13,000 rpm for 10 minutes, and supernatants were collected for protein estimation. Equal quantities of denatured proteins were separated by electrophoresis using SDS-PAGE gels and transferred to activated, PVDF membranes. Western blotting was performed using primary antibodies against MUC1-CT (Armenian Hamster monoclonal antibody), actin, and beta-tubulin (Clones J5 and E7, respectively, from Developmental Studies Hybridoma Bank).

Cells seeded at 40% density on sterile glass coverslips in 24-well plates were used for evaluating DNA damage response. Fresh media with or without inhibitors were added immediately before and after irradiation. Cells were rinsed with PBS to remove the media at specific time points after treatment and fixed in 4% paraformaldehyde. Cells were permeabilized with 0.2% tween-20 for 5 minutes at room temperature and washed thrice with PBS, followed by blocking at room temperature with 1% nonfat milk in PBS containing 0.05% Tween. Fixed cells were subsequently subjected to incubation at 4°C overnight with the primary antibody [Phospho-Histone H2A.X (Ser139; 20E3) Rabbit mAb #9718 from Cell Signaling Technology, 1:1,000 in blocking solution], followed by three washes with PBS and then incubated with secondary antibody conjugated to Alexa fluor-644 for 30 minutes, in dark at room temperature. After washing 3 times with PBS, coverslips were mounted on DAPI-fluoro mount-G. Immunofluorescence imaging capturing was performed at 20X magnification by using a DMI6000 Leica microscope.

Animal studies were conducted according to an approved protocol by the University of Nebraska Medical Center Institutional Animal Care and Use Committee. Female athymic nude mice were injected subcutaneously with one tumor each with 2 million S2-013.Neo or S2-013.MUC1 cells at the left flank. Tumor volume was calculated using the formula: volume = [length *(width^2)]/2. Mice were grouped into four treatment groups and treated with saline control, bromopyruvate (BrPA) alone, radiation alone, and BrPA + radiation, once all tumors reached 100 mm³ or greater. BrPA treatment started on day 0 and given i.p. daily at 7.5 mg/kg. Radiation treatments were started on day 1 and given at a dose of 4 Gy for 5 consecutive days (20 Gy total). Mice were anesthetized using ketamine/xylazine mix and irradiated with the body shielded using a lead shield while tumor remained exposed to radiation. Tumor irradiation was performed with a 160 kV RAD Source RS 2000 X-Ray irradiator at a dose rate of 1.2 Gy/min. Tumor sizes were monitored for every 3 days until 27 days, after day 0.

Cells (0.75 × 10⁷) were cultured for 24 hours in normal DMEM, and culture medium was exchanged with fresh media before irradiation. Cells were exposed to ionizing radiation (IR; 4 Gy) followed by rinsing with PBS and frozen on dry ice for metabolomic studies. Polar metabolomics and data analyses were performed as described previously (26–28).

Radiation-induced DNA damage would resultantly inhibit tumor cell proliferation (29). To determine the role of MUC1 in regulating radiation resistance in pancreatic cancer, we evaluated cell survival using MUC1-knockdown and -overexpressed cell line models. Reducing MUC1 expression using MUC1 knockdown resulted in sensitization of FG, HPAF2, and Capan2 pancreatic cancer cell lines to radiation. The survival of FG.shMUC1 cells upon 2 to 6 Gy radiation exposure was lesser than the control FG.shScr cells (Fig. 1A and B). Similarly, HPAF2 and Capan2 cells exposed to radiation (2–8 Gy) showed a significant reduction in survival upon MUC1 knockdown. Reciprocally, we observed increased cell survival in S2-013 cells under conditions of MUC1 expression when irradiated with 2 to 6 Gy doses of irradiation as shown in Figure 1A and B. MUC1-expressing HPAF2 and S2-013 cells also exhibited higher clonogenic potential at lower doses of irradiation (2–6 Gy), compared with controls (Fig. 1C). Furthermore, to evaluate if the effect of MUC1 expression on radiation resistance is only specific for pancreatic cancer cell lines, we utilized an osteosarcoma cell line model, U2OS, which expresses low levels of MUC1. Exogenous expression of MUC1 in U2OS (U2OS.MUC1) cells enhanced their survival upon irradiation compared with the cells stably expressing vector control (U2OS.Neo; Fig. 1D).

Radiation-induced cellular damage includes the modification of DNA bases resulting in DNA strand breaks (28). Phosphorylation of histone 2A (γH2AX) is an indicator of DNA damage. Also, low levels of γH2AX are indicative of effective DNA damage repair occurring in irradiated cells. In our irradiation study, we observed that MUC1-expressing cells stained less positively for γH2AX 12 hours after irradiation in comparison with either control or MUC1-knockdown cells (Fig. 2A). Furthermore, we evaluated the role of NHEJ repair and homologous repair (HR) in DNA damage repair of MUC1-expressing cells using NHEJ inhibitor Nu-7026 and HR inhibitor RI-1, respectively (30, 31). Both the treatments abrogated MUC1-induced differences in γH2AX staining as the level of γH2AX staining increased by combination treatments with DNA damage repair inhibitors and radiation (Fig. 2A). Inhibition of DNA damage repair by RI-1 significantly reduced the survival of irradiated MUC1-expressing cells. However, Nu-7026 diminished the survival of irradiated pancreatic cancer cells irrespective of their MUC1 expression status (Fig. 2B).

Our previous studies demonstrated that MUC1 expression promotes glucose uptake and glycolytic metabolism in pancreatic cancer cells (18). To determine if the MUC1 expression also induces differential metabolic changes upon irradiation, we performed LC-MS/MS–based metabolomics analysis of polar metabolite extracts from control and irradiated cell lines. Partial least square discriminant analysis (PLS-DA) of the polar metabolite components indicated an overall metabolic shift in pancreatic cancer cells with MUC1 knockdown or overexpression, under control and irradiated conditions (Fig. 3A and B, PLS-DA plots). Further comparison of metabolite levels among the control and irradiated cells indicated higher levels of glycolytic, PPP, and nucleotide biosynthetic pathway intermediates in MUC1-expressing cells, which increase within 10 minutes of irradiation (4 Gy; Fig. 3A and B, bar graphs). Based on the increased nucleotide levels observed in irradiated MUC1-expressing cells, we evaluated if nucleoside supplementation to the MUC1-knockdown or S2-013.Neo cells would decrease their responsiveness to radiation, to the level observed in MUC1-expressing cells. As expected, nucleoside supplementation prevented the radiation-induced DNA damage in pancreatic cancer cells irrespective of MUC1 expression (Fig. 4A). Nucleoside supplementation along with radiation significantly enhanced survival of the Capan2.shMUC1 and S2-013.Neo cells in comparison with the radiation treatments (Fig. 4B). Furthermore, comparison of relative survival between Capan2.shMUC1 cells and combination-treated Capan2.shMUC1 cells and S2-013.Neo cells with combination-treated S2-013.Neo cells showed that nucleosides recovered the survival in combination-treated cells.

Our findings indicated a significant increase in glycolysis, PPP, and nucleotide metabolism in irradiated MUC1-expressing pancreatic cancer cells. Hence, we next determined if metabolic inhibition of glycolysis with 3-bromopyruvate (BrPA) and PPP with 6-amino nicotinamide (6AN) would sensitize MUC1-expressing cells to radiation. Cell survival analyses showed a reduction in cell survival with BrPA (2.5 μmol/L) and 6AN (50 μmol/L) treatments in both control and experimental cell lines, similar to the radiation treatment alone. Treatment with BrPA reduced cell survival in four different pancreatic cancer cells, and the combination of BrPA along with radiation was more effective in reducing cell survival irrespective of MUC1 expression (Fig. 5A). Though PPP inhibitor, 6AN enhanced the sensitivity of MUC1-expressing cells to radiation; its combination with radiation was less effective in comparison with the combinations of radiation and BrPA (Fig. 5A). Glycolytic flux into PPP provides intermediates for de novo nucleotide biosynthesis required for proliferating cancer cells. We observed that the pretreatment with glycolytic inhibitor, BrPA, before irradiation diminished the levels of metabolites in glycolysis and PPP, in FG and S2-013 cell types (Supplementary Fig. S1). Relative levels of glycolysis pathway metabolites including G6P, F6P, FBP, and G3P decreased significantly in control and experimental cell line models upon treatment with either BrPA alone or in combination with radiation (Supplementary Fig. S1). Dual treatment with BrPA and radiation reduced most of the PPP-metabolite levels and de novo nucleotide biosynthesis pathway, as some of the nucleotides were relatively decreased in BrPA and combination-treated MUC1-expressing FG and S2-013 cells (Supplementary Fig. S1). Based on our BrPA-treated metabolomic studies, we anticipated that the inhibitory effect of BrPA on MUC1-mediated glycolysis and PPP would suppress MUC1-regulated DNA damage upon irradiation. Therefore, we evaluated the effects of BrPA on DNA damage in combination with radiation. We observed that BrPA treatment enhanced DNA damage in both control and experimental cell lines (Fig. 5B).

We further evaluated the effects of BrPA and radiation on clonogenicity. The combination of BrPA and radiation significantly reduced clonogenicity of MUC1-expressing cells in both HPAF2 and S2-013 cell line models (Fig. 6A). Based on our in vitro findings, we evaluated the effects of BrPA on MUC1-induced radiation resistance using subcutaneous xenograft tumor models. Tumor growth studies showed that irradiation reduced tumor growth in S2-013.Neo group tumors in comparison with the saline-treated S2-013.Neo group (Fig. 6B, left). In contrast to the irradiated S2-013.Neo group mice, tumors in mice implanted with S2-013.MUC1 cells did not exhibit significant differences in tumor growth among the saline, radiation, and BrPA-treated groups (Fig. 6B, right). Interestingly, both S2-013.Neo and S2-013.MUC1 groups treated with the combination of radiation and BrPA showed a significant decrease in tumor growth in comparison with their respective saline-treated groups from 18 days of combination treatments. The tumors were validated for MUC1 expression by IHC upon necropsy (Supplementary Fig. S2).

Exposure of tumor cells to stress conditions such as RT can induce oncogene-regulated stress responses, which in turn lead to radiation resistance in tumor cells (32, 33). Oncogene-regulated mechanisms allow tumor cells in overcoming radiation-induced DNA damage, thus imparting radiation resistance to tumor cells (34). However, the exact nature of signaling pathways and mechanisms that confer radiation resistance remains poorly defined. Our present findings demonstrate that MUC1 oncogene abrogates cytotoxic effects of RT. Cell survival and clonogenicity upon irradiation were enhanced by the MUC1 expression, whereas reduction was observed by the MUC1 knockdown in pancreatic cancer cells (Fig. 1). Furthermore, MUC1 expression increased survival of osteosarcoma cells upon irradiation ascertaining the role of MUC1 in mediating radiation resistance in different cancer cell types, which is in agreement with another study wherein MUC1 knockdown radiosensitized breast cancer cell lines (16).

Cancer cells overcome DNA damage induced by irradiation through efficient DNA damage repair mechanisms (35). Such repair facilitates radiation resistance as DNA strand breaks induced by radiation are fixed, allowing the cells to survive and proliferate (36). Our present findings indicate that the MUC1 expression enhances DNA damage repair in pancreatic cancer cells as evident from relatively low γH2AX staining within the same interval (12 hours) of irradiation in MUC1-expressing cells compared with the control cells (Fig. 2). A previous study demonstrated that pancreatic cancer cells innately possess high DNA damage (7). NHEJ is the primary mechanism that facilitates DNA damage repair in pancreatic cancer cells upon irradiation followed by HR mechanism (7). Hence, we investigated the relative contribution of NHEJ and HR in MUC1-mediated DNA damage repair. Sustenance of radiation-induced γH2AX staining with inhibition of DNA-PK (involved in NHEJ repair) or RAD51 (involved in HR repair) in both control and MUC1-expressing cells indicated that both NHEJ and HR play roles in DNA damage response in pancreatic cancer cells independent of MUC1 expression. However, we noticed a significant decrease in the survival of irradiated MUC1-expressing cells exposed to RI-1, in comparison with the MUC1-low cells. Such difference was not observed in the case of Nu-7026 exposure, indicating that MUC1-expressing cells are dependent on HR-mediated DNA damage repair; but the NHEJ-mediated repair is independent of MUC1 expression in irradiated pancreatic cancer cells (Fig. 2). Thus, we further investigated possible mechanisms leading to the decrease in radiation sensitivity upon MUC1 expression.

Pancreatic cancer cells exhibit strong glycolytic phenotype (11). We previously reported that MUC1 expression alters the metabolism in pancreatic cancer cells (18, 20). MUC1-expressing pancreatic cancer cells exhibit enhanced glycolysis, TCA cycle, and PPP under steady-state conditions (18). Such metabolic alterations in tumor cells can facilitate DNA damage repair induced by radiation (29). Metabolomic analysis performed in the present study demonstrates that pancreatic cancer cells alter their metabolism upon irradiation independent of MUC1 expression (Fig. 3). Furthermore, our current findings show that radiation-induced metabolic alterations were distinct under conditions of MUC1 expression. These findings ascertain that MUC1 expression coordinates upregulation of the interconnected pathways and mediates the metabolic response of cancer cells in response to radiation. Our results demonstrate that the levels of nucleotide precursors in PPP (including the ribose-phosphate/ribulose-5-phosphate, RP) and that of nucleotides in both purine (GDP and dAMP) and pyrimidine biosynthesis pathways are enhanced in both control and MUC1-expressing cell lines. However, such changes are more pronounced in MUC1-expressing cells, suggesting that de novo nucleotide biosynthesis pathways are upregulated upon irradiation in a MUC1-dependent manner. We anticipated that the MUC1-mediated radiation resistance could be due to the enhancement of nucleotide metabolism upon irradiation. This could facilitate efficient DNA damage repair by supplementing the nucleotide pools essential for the replacement of damaged nucleotides. As expected, nucleoside supplementation reduced the DNA damage in control cells similar to that of the MUC1-expressing cells (Fig. 4A). Also, nucleoside supplementation mostly enhanced the cell survival in both MUC1-knockdown/MUC1-low and MUC1-expressing cells upon irradiation, to a similar extent (Fig. 4B). These results indicate that the MUC1-mediated resistance to radiation response is at least in part due to the increase in nucleotide metabolism. Irradiation leads to MUC1-mediated increase of glycolytic flux into PPP and downstream nucleotide metabolism pathways which are essential for the DNA damage repair and survival of the irradiated MUC1-expressing cells.

Oncogene-mediated aberrant metabolism provides a survival advantage for tumor cells (37). In addition, it also causes the susceptibility of tumor cells to inhibitors of metabolic pathways (38). Our present findings demonstrate radiosensitization of MUC1-expressing cells with BrPA, a metabolic inhibitor characterized as an efficient chemotherapeutic agent against pancreatic cancer (39, 40). Our results also show that BrPA treatment reduced the metabolite levels in glycolysis, PPP, and nucleotide biosynthetic pathways 12 hours after treatment in MUC1-expressing cells (Supplementary Fig. S1). Furthermore, inhibition of glycolysis resulted in radiosensitization of MUC1-expressing cells by preventing DNA damage repair (Fig. 5). Thus, inhibition of nucleotide metabolism by BrPA treatment resultantly inhibited survival and clonogenicity of MUC1-expressing cells by enhancing DNA damage (Figs. 5 and 6A).

MUC1 is expressed in most of the pancreatic tumors and promotes pathologic progression of the disease (41–45). MUC1 expression also confers drug resistance in pancreatic cancer (46). However, the role of MUC1 in regulating radiation responsiveness in pancreatic cancer is least studied, compared with other cancer models (16, 47, 48). Our present findings establish the role of MUC1 in imparting poor radiation responsiveness to pancreatic cancer cells. We identified that enhancement of glycolysis and PPP by MUC1 expression diminishes radiation responsiveness in pancreatic cancer cells. Our in vivo studies substantiate the role of MUC1 in imparting radiation resistance to pancreatic cancer. Notably, we also identified that MUC1 expression diminishes radiation responsiveness in pancreatic tumors (Fig. 6B). We determined that combination of BrPA along with radiation sensitizes pancreatic tumors to radiation. Thus, our findings establish the efficacy of targeting MUC1-mediated nucleotide metabolism for radiosensitization of pancreatic tumor cells. A potential molecular mechanism that could facilitate the MUC1-induced metabolic reprogramming could be through HIF-1α. Low-dose irradiation enhances HIF-mediated metabolic responses as reported in a recent study (49), and our previous study showed that MUC1-mediated HIF-1α stabilization leads to upregulation of glycolytic and PPP metabolic pathway genes in MUC1-expressing pancreatic cancer cells (18). Our results establish that metabolic responses elicited in MUC1-expressing cells can be effectively targeted using BrPA, which reduces radiation resistance in pancreatic cancer cells. This modality needs further evaluation in clinical studies wherein glycolysis-dependent pancreatic tumors can be pretreated with glycolytic inhibitors before irradiation, for enhancing the efficacy of fractionated radiation therapies.

No potential conflicts of interest were disclosed.

Conception and design: V. Gunda, J. Souchek, N.V. Chaika, C. Lin, P.K. Singh

Development of methodology: V. Gunda, J. Souchek, C. Lin

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): V. Gunda, J. Souchek, J. Abrego, S.K. Shukla, G.D. Goode, E. Vernucci, S. Li, S. Wang, T. Bessho, C. Lin

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): V. Gunda, J. Souchek, S.K. Shukla, R.J. King, F. Yu, T. Bessho, C. Lin, P.K. Singh

Writing, review, and/or revision of the manuscript: V. Gunda, J. Souchek, S.K. Shukla, S. Li, S. Wang, F. Yu, T. Bessho, C. Lin, P.K. Singh

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): V. Gunda, R.J. King, S. Wang, T. Bessho, C. Lin

Study supervision: C. Lin, P.K. Singh

Other (performing metabolite experiments): J. Abrego

Other (helped in survival and metabolite experiments): A. Dasgupta

The authors acknowledge the Fred & Pamela Buffett Cancer Center Support Grant (P30CA036727, NCI) for supporting shared resources.

This work was supported in part by funding from an NIH grant (R01 CA163649, R01 CA210439, and R01 CA216853, NCI) to P.K. Singh, the Pancreatic Cancer Action Network-AACR Career Development Award (13-20-25-SING) to P.K. Singh, and a Specialized Programs of Research Excellence (award SPORE, 2P50 CA127297, NCI) to P.K. Singh.

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.