HNRNPC regulates RhoA to induce DNA damage repair and cancer‐associated fibroblast activation causing radiation resistance in pancreatic cancer

Abstract Pancreatic cancer (PC) is one of the most lethal types of cancer due to its asymptomatic nature in the early stages and consequent late diagnosis. Its mortality rate remains high despite advances in treatment strategies, which include a combination of surgical resection and adjuvant therapy. Although these approaches may have a positive effect on prognosis, the development of chemo‐ and radioresistance still poses a significant challenge for successful PC treatment. Heterogeneous nuclear ribonucleoprotein C1/C2 (HNRNPC) and RhoA have been implicated in the regulation of tumour cell proliferation and chemo‐ and radioresistance. Our study aims to investigate the mechanism for HNRNPC regulation of PC radiation resistance via the RhoA pathway. We found that HNRNPC and RhoA mRNA and protein expression levels were significantly higher in PC tissues compared to adjacent non‐tumour tissue. Furthermore, high HNRNPC expression was associated with poor patient prognosis. Using HNRNPC overexpression and siRNA interference, we demonstrated that HNRNPC overexpression promoted radiation resistance in PC cells, while HNRNPC knockdown increased radiosensitivity. However, silencing of RhoA expression was shown to attenuate radiation resistance caused by HNRNPC overexpression. Next, we identified RhoA as a downstream target of HNRNPC and showed that inhibition of the RhoA/ROCK2‐YAP/TAZ pathway led to a reduction in DNA damage repair and radiation resistance. Finally, using both in vitro assays and an in vivo subcutaneous tumour xenograft model, we demonstrated that RhoA inhibition can hinder the activity of cancer‐related fibroblasts and weaken PC radiation resistance. Our study describes a role for HNRNPC and the RhoA/ROCK2‐YAP/TAZ signalling pathways in mediating radiation resistance and provides a potential therapeutic target for improving the treatment of PC.


| INTRODUC TI ON
Pancreatic cancer (PC) is the eleventh most common cancer worldwide and seventh leading cause of cancer mortality in both men and women due to its poor prognosis and late-stage diagnosis. 1 Earlystage PC is typically asymptomatic, while late-stage symptoms, such as abdominal pain, jaundice and pruritus, are non-specific, 2 making correct diagnosis especially challenging. Current diagnostic methods rely on a combination of computed tomography scans, magnetic resonance imaging, ultrasound imaging, cholangiopancreatography, positron emission tomography, angiography, blood tests and biopsy results. 3 PC risk factors include smoking, alcohol, obesity, certain dietary habits, occupational exposure to metal and pesticides, advanced age, male gender, ethnicity, diabetes mellitus, family history, genetic factors, ABO blood group, pancreatitis and some chronic infections. 4 While some advances in treatment strategies have been made, the five-year survival rate for PC remains low at 9%. 5 Although surgical resection and adjuvant therapy are common treatment strategies for PC depending on the stage of disease, 6 recent studies have demonstrated the benefits of applying a neoadjuvant approach, which include controlling potential micrometastases (usually present at the time of diagnosis), and determining which patients would benefit from surgery while sparing unsuitable patients from major surgical intervention. 7,8 Current first-line chemotherapeutic regimens for metastatic pancreatic cancer patients involve treatment with either FOLFIRINOX (folinic acid, 5-fluorouracil, irinotecan and oxaliplatin) or gemcitabine plus nab-paclitaxel, 9 with recent studies indicating that FOLFIRINOX may be the better option. 10 However, the eventual and inevitable development of chemo-and radioresistance significantly limits their treatment efficacy. 9,11 PC cells are particularly prone to developing endogenous and exogenous resistance to gemcitabine. 5 Therefore, it is important to investigate additional agents that might enhance chemotherapy efficacy in drug-sensitive cells and reduce resistance in drug-resistant cancer cells.
Previous studies have reported elevated levels of heterogeneous nuclear ribonucleoproteins C1/C2 (HNRNPC) in some cancer cells, while HNRNPC knockdown resulted in significant arrest of cell proliferation and tumour growth in breast cancer. 12 Conversely, elevated levels of HNRNPC may contribute to a metabolic environment that is beneficial for proliferation. 13 A more in-depth investigation has established an association between mRNA-binding protein hnRNP-Q1 and small GTPase RhoA. In particular, hnRNP-Q1 is involved at the cellular level in RhoA-dependent cellular morphogenesis and the molecular level as a RhoA mRNA translation repressor. 14 These findings also suggested that the 3′ untranslated region (UTR) of RhoA mRNA is essential for hnRNP-Q1-mediated regulation of RhoA expression. Furthermore, previous studies have also shown that RhoA silencing by small-interfering RNA (siRNA) reduces proliferation and migration of tumour cells and may improve the cytotoxic effect of chemotherapy in human colon cancer cells, thereby reversing chemoresistance. 15 Modulation of RhoA activity has been shown to sensitize cells to γ-radiation by attenuating the DNA damage response and repair pathways. 16 Rho-associated protein kinases such as ROCK2 are considered to be key Rho downstream effectors, while ROCK inhibitors have anti-tumour properties. 17 In fact, RhoA/ROCK signalling may be important for radiation resistance, with ROCK2 in particular having a role in the regulation of cell division. 18 ROCK2 has also been associated with chemotherapy resistance, with inhibition of ROCK2 signalling sensitizing gemcitabine-resistant PC cells to gemcitabine-mediated DNA damage. 19 In addition, YES-associated protein (YAP), which is a downstream target of RhoA signalling, 20 is upregulated in PC and participates in tumorigenesis via epithelial-mesenchymal transition-related factors. 21 Recently, the RhoA/ROCK pathway has been directly linked to the regulation of the YAP/TAZ pathway in fibroblasts. 22,23 RhoA/ROCK signalling is critically required for the activation of YAP in multiple biological processes 24 including stem cell development 25 and atherosclerosis. 26 However, whether there is a link between HNRNPC and the RhoA/ROCK2 and YAP/TAZ signalling pathways in PC remains unknown.
Cancer-associated fibroblasts (CAFs) or activated fibroblasts secrete cytokines, growth factors and matrix-degradation proteins that promote cancer cell proliferation and progression and participate in carcinogenesis, angiogenesis and metastasis. 27 CAFs are identified by the expression of activated fibroblast markers, such as fibroblast activation protein (FAP), α-smooth muscle actin (α-SMA) and fibroblast-specific protein 1. 28 Recent studies have suggested that the Hippo pathway and its transcriptional effectors YAP and TAZ are required for fibroblast activation. Nuclear YAP localization is a universal feature of CAFs, and YAP is critical for many aspects of the tumour-promoting CAF function. 29 However, although the RhoA/ROCK2-YAP/TAZ axis has been linked to fibrotic activity and fibroblast differentiation, 23,30 its role in PC remains unknown.
To determine whether the RhoA/ROCK2-YAP/TAZ signalling pathways have a role in PC, this study aimed to explore the mechanism for HNRNPC regulation of PC radiation resistance. We hypothesized that (1) HNRNPC positively regulates RhoA posttranscriptionally, enhances DNA damage repair and promotes PC radiation resistance through the ROCK2/γH2AX pathway and that (2) RhoA activates the YAP pathway, enhances the activation of CAFs and further leads to radiation resistance.
To test our hypothesis, we studied the role of HNRNPC, RhoA, ROCK2, YAP and TAZ in PC tumour and paired adjacent normal tissues and the human PC cell lines BxPC-3 and Panc-1 using cell viability, qRT-PCR, Western blotting, DNA damage repair and fragmentation, RNA immunoprecipitation and immunohistochemistry assays. A xenograft tumour model in nude mice was established to verify our in vitro findings. In summary, this study uncovered HNRNPC as a potential therapeutic target for improving PC radiation resistance.

| Ethics statement
Written informed consent was obtained before collecting PC samples from patients. The Ethics and Institutional Review Committee of the Ruijin Hospital of Shanghai Jiao Tong University of China approved the study, which was conducted in accordance with the ethical standards conveyed in the Declaration of Helsinki.

| Survival analysis using the Kaplan-Meier Plotter web tool
The online database Kaplan-Meier plotter (http://kmplot.com/analy sis/index.php?p=service) was used to determine the five-year survival rate of pancreatic cancer patients. Kaplan-Meier plotter performed survival analyses based on gene expression levels.

| Cell lines and culture
Human PC cell lines (BxPC-3 and Panc-1) were purchased from the Chinese Academy of Sciences Cell Bank of Type Culture Collection (Shanghai, China). All cell lines were cultured in Dulbecco's modified Eagle's medium (DMEM, HyClone, Logan, USA) supplemented with 10% foetal bovine serum (FBS), 100 U/mL penicillin and 100 μg/mL streptomycin. The cultures were incubated at 37°C in a humidified atmosphere containing 5% CO 2 . Pancreatic cancer cells were pretreated with clofibrate for 24 h and then were exposed to different dosages of ionizing radiation using an X-ray linear accelerator (Rad Source, Suwanee, GA, USA) at a fixed dose rate of 1.15 Gy/min. 31

| Colony formation assay
Cells were seeded in six-well plates (500 cells/well) and incubated for 24 h. After treatment, cells were cultured for an additional two weeks in drug-free medium. Cells were then fixed with 4% formaldehyde, stained with 0.1% crystal violet and the number of colonies was counted.

| Cell viability assay
Cells were seeded in 96-well plates (4 × 10 3 cells/well). Forty-eight hours after treatment (0-, 2-, 4-, 6-or 8-Gy radiation), MTT solution was added and the plates were placed in a CO 2 incubator for 4 h at 37°C. The optical density was measured on a microplate reader at 490 nm. Each group was tested in triplicate.

| qRT-PCR assay
Total RNA was extracted using Fastgen1000 (Fastgen, Shanghai, China) according to the manufacturer's instructions and reverse transcribed into cDNA with Prime Script RT (Takara, Dalian, China).

| RNA immunoprecipitation assay
The RNA immunoprecipitation assay (RIP) assay was conducted with a Magna RIP RNA-binding protein immunoprecipitation kit (Millipore, Billerica, MA, USA). Cells were first lysed with RIP buffer (Millipore), then lysates were incubated with Sepharose beads (Bio-Rad, Hercules, CA, USA) pre-coated with HNRNPC antibody.
Immunoglobulin G antibody was used as the control. Finally, the quantity of RhoA in the immunoprecipitated complexes was determined by qRT-PCR.

| Luciferase reporter assay
Wild-type and mutant RhoA 3′UTRs were cloned into a luciferase vector (Promega, Madison, WI, USA) and co-transfected with HNRNPC-OE vector into PC cells using Lipofectamine 3000 transfection reagent. Cells were harvested for luciferase activity analysis after 48 h.

| PC-normal fibroblast (NF) co-culture model
To obtain PC-conditioned media, PC cells were cultured to 50% confluence in DMEM supplemented with 10% FBS, then the media was changed to contain 1% FBS, 100 U/mL penicillin and 100 μg/mL streptomycin. After 48 h, the PC-conditioned media was centrifuged and filtered before incubation with isolated NFs as described previously 21 for 48 h. To establish the PC-NF co-culture model, PC cells and NFs were combined at a ratio of 2:1 and seeded into six-well plates.

| Immunohistochemical analysis
Sections (3µm thick) were deparaffinized, rehydrated and endogenous peroxidase activity was blocked via incubation with 3% hydrogen peroxide for 15 min at room temperature. Antigens were retrieved in a sodium citrate buffer (pH 6.0) for 3 min using a pressure cooker. Sections were blocked with 5% bovine serum albumin

| Xenograft studies in nude mice
Female BALB/c nude mice (16,  The tumour volume was calculated as (length/2) × (width 2 ). 21 The animal experimental protocol was approved by the Animal Experiment Ethics Committee of Shanghai Jiao Tong University.

| Statistical analysis
All values are presented as mean ± SEM. Statistical analysis was performed using SPSS 18.0 software (IBM, Chicago, IL, USA). Twotailed Student's t-test or one-way ANOVA were used for data analysis. A p-value of < 0.05 was considered statistically significant.

| HNRNPC and RhoA are highly expressed in PC patient tissues
To establish the relationship between HNRNPC, RhoA and radiation resistance in PC tissues, we examined HNRNPC and Both HNRNPC and RhoA mRNA expression levels were significantly higher in PC patient tumour samples than the corresponding adjacent non-tumour tissue ( Figure 1A, B). Moreover, RhoA expression was positively correlated with HNRNPC expression ( Figure 1C). IHC staining revealed that HNRNPC and RhoA protein expression levels were also significantly increased in the PC patient tumour samples compared to the adjacent normal tissue ( Figure 1D, E). Kaplan-Meier analysis (http://kmplot.com/analy sis/index.php?p=service) indicated that the overall survival rate of PC patients was significantly lower in patients expressing high HNRNPC levels ( Figure 1F). Finally, we demonstrated that HNRNPC mRNA expression was also significantly higher in multiple PC cell lines (AsPC-1, BxPC-3, Capan-1, CFPAC-1 and PanC-1) compared to control human pancreatic ductal epithelial cells (HPDECs) ( Figure 1G). Taken together, these data demonstrate that HNRNPC and RhoA are highly expressed in PC tissues and that high HNRNPC expression is associated with poor patient prognosis.  Figure 2G, H). Thus, taken together, our findings suggest that HNRNPC overexpression promotes radiation resistance in PC cells, while HNRNPC knockdown increases radiosensitivity.

| Radiation resistance induced by HNRNPC overexpression is mediated by RhoA
Using Starbase (http://starb ase.sysu.edu.cn/index.php), we predicted that HNRNPC interacted with RhoA. To further examine the relationship between HNRNPC and RhoA, we measured RhoA mRNA and protein expression in HNRNPC-overexpressing and -silenced cells using qRT-PCR and Western blot analysis. We found that RhoA mRNA and protein expression levels were significantly increased in cells overexpressing HNRNPC, while knockdown of HNRNPC led to a significant reduction in RhoA ( Figure 3A

| RhoA is a downstream target of HNRNPC
As HNRNPC is an RNA binding protein, we next sought to determine whether RhoA was a direct target of HNRNPC binding. We confirmed that HNRNPC is bound to RhoA mRNA by RIP assay ( Figure 4A). Consistent with these findings, significantly increased luciferase activity was observed in HNRNPC-overexpressing  F) The cell viability of HNRNPC-OE or HNRNPC siRNA-treated cells exposed to different doses of irradiation was measured by MTT assay. (G, H) Colony formation in BxPC-3 and PanC-1 cells treated with control, HNRNPC-OE, siHNRNPC#1, siHNRNPC#2 or siNC and exposed to 4 Gy radiation. Data are presented as mean ± SEM, n = 3. *p < 0.05, ** p <0.01 and ***p < 0.001 using ANOVA BxPC-3 and PanC-1 cells co-transfected with wild-type RhoA 3′UTR WT, but not mutated RhoA 3′UTR mut ( Figure 4B). Next, we investigated the signalling pathways downstream of HNRNPCinduced RhoA. Since RhoA has recently been implicated in fibrosis through the activation of the ROCK/YAP/TAZ axis, 23 we examined the effects of HNRNPC overexpression and silencing on the mRNA and protein expression of these pathways. We found that HNRNPC overexpression led to increased ROCK2, YAP and TAZ protein ( Figure 4C) and mRNA (Supporting Information Figure S1 Figure S1). Taken together, our findings demonstrate that RhoA is a downstream target of HNRNPC, which acts through the RhoA/ROCK2-YAP/TAZ axis to promote fibrosis.

F I G U R E 4 RhoA is a downstream target of HNRNPC. (A)
The interaction between HNRNPC and RhoA was examined by RNA immunoprecipitation. (B) Luciferase reporter assay was used to detect HNRNPC binding to RhoA 3'UTR. (C) The protein levels of downstream target genes of RhoA after HNRNPC knockdown or overexpression in BxPC-3 and PanC-1 cells were determined by Western blotting. Data are given as mean ± SEM, n = 3. *p < 0.05, **p < 0.01 and ***p < 0.001, using ANOVA examined expression of phosphorylated γH2AX, a marker of DNA double-strand breaks. We found increased levels of p-γH2AX after HNRNPC knockdown, while HNRNPC-OE treatment led to decreased p-γH2AX expression ( Figure 4C, Supplementary Figure 1).
To further elucidate the role of HNRNPC and RhoA in mediating radiation resistance in PC cells, we examined the effects of silencing RhoA expression on proliferation and DNA damage repair in PC cells exposed to irradiation. EdU staining revealed that RhoA interference reduced DNA synthesis and cell proliferation in BxPC-3 and PanC-1 cells in both control (0 Gy) and irradiated (4 Gy) cells ( Figure 5A).
TUNEL staining demonstrated that RhoA knockdown increased DNA fragmentation and apoptosis in both control (0 Gy) and irradiated (4 Gy) cells ( Figure 5B). RhoA silencing also led to a significant decrease in ROCK2 protein expression levels, while p-γH2AX levels were significantly increased especially after irradiation ( Figure 5C, D). Taken together, our findings show that inhibition of the RhoA/ ROCK2 pathway leads to decreased proliferation, increased apoptosis and increased DNA damage after radiation treatment, suggesting that RhoA has a role in mediating radiation resistance in PC cells via the RhoA/ROCK2-YAP/TAZ signalling pathways.

| RhoA inhibition can hinder CAF activity and weaken radiation resistance in PC
Since we observed increased expression of fibrotic markers in PC cells over-expressing HNRNPC, we next examined the effects of tumour cell supernatants on CAFs. We established a cell co-culture system using PC cell-conditioned media and NFs to examine the role of RhoA in CAF activation. Immunofluorescence staining revealed that culturing NFs in conditioned media obtained from RhoA-siRNAtreated BxPC-3 and PanC-1 cells led to a decrease in the expression of the CAF-related proteins α-SMA and FAP compared to control si-NC-treated cells ( Figure 5E). Similar results were observed by Western blot analysis ( Figure 5F). Together, these findings suggest that RhoA has a role in regulating CAF activation.
Next, we established a xenograft tumour model in nude mice to determine the effects of HNRNPC silencing on the radiation resistance of PC cells in vivo. NFs and control or si-HNRNPC-treated PanC-1-cells were co-cultured, then subcutaneously injected into nude mice. We found that mice injected with siNC-treated PanC-1/ NF co-cultures developed large tumours ( Figure 6A-C), which were significantly decreased after exposure to 10 Gy irradiation. Tumours that developed from NFs co-cultured with HNRNPC-silenced PanC-1 cells were significantly smaller, and irradiation of these PanC1/NF co-cultures with 10 Gy led to the most significant decrease in tumour growth ( Figure 6A-C). These findings suggest that silencing HNRNPC increases the radiation sensitivity of PC cells.
Western blot analysis of the tumour samples from mice injected with siNC-treated PanC-1/NF co-cultures revealed that exposure to 10 Gy irradiation resulted in a decrease in HNRNPC and RhoA protein expression, as well as increased p-γH2AX levels ( Figure 6D

| DISCUSS ION
PC is a lethal cancer that is asymptomatic in its early stages and has non-specific symptoms in later stages, making it difficult to diagnose. 1  while HNRNPC knockdown has been shown to slow down cell proliferation and tumour growth in breast cancer. 12 HNRNPC has also been identified as a key regulator of metastatic potential in glioblastoma cells. 38 Since PC cells generally develop radiation resistance during treatment, 39 determining the mechanisms underlying radioresistance is critical for the development of future treatment strategies.
The present study examined radiation-resistant PC tissue samples from clinical patients for both HNRNPC and RhoA levels, which are involved in the proliferation and migration of tumour cells. 40 Previously, an association between the mRNA-binding protein F I G U R E 5 Inhibition of the RhoA/ROCK2 pathway reduces DNA damage repair and radiation resistance. (A) EdU staining was used to measure DNA synthesis in BxPC-3 and PanC-1 cells treated with siRhoA or siNC and exposed to 4 Gy irradiation, bar = 100 μm. (B) TUNEL staining was used to evaluate apoptosis in BxPC-3 and PanC-1 cells treated with siRhoA or siNC and exposed to 4 Gy irradiation, bar = 100 μm. (C, D) Western blot analysis of ROCK2 and p-γH2AX expression in BxPC-3 and PanC-1 cells treated with siRhoA or siNC and exposed to 4 Gy irradiation. (E, F) NFs were cultured with conditioned media obtained from si-NC and siRhoA-treated BxPC-3 and PanC-1 cells. Immunofluorescence staining of activated fibroblast markers, FAP and α-SMA, bar = 50 μm (E). Western blot analysis of FAP and α-SMA protein expression (F). Data are given as mean ± SEM, n = 3. *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001 using ANOVA | 2333 XIA et Al.

F I G U R E 6
PanC-1/NF co-cultures enhance radiation resistance, while HNRNPC knockdown in mice reduces radiation resistance. A xenograft tumour model was established in nude mice using PanC-1/NF co-cultured cells, in which the PanC-1 cells had been treated with either si-NC or si-HNRNPC. Tumours were exposed to 0 Gy or 10 Gy radiation. hnRNP-Q1 and RhoA has been reported. 14 Here, we found that both HNRNPC and RhoA are expressed in radiation-resistant PC tissues. In addition, we demonstrated that RhoA was a downstream target of HNRNPC. Previous studies have suggested that HNRNPC is associated with the coordination of DNA-damage responses and radiation-induced apoptosis pathways. 41 Here, we found that overexpression of HNRNPC was associated with increased radioresistance, while knockdown of HNRNPC led to increased PC sensitivity.
Furthermore, inhibition of RhoA attenuated radiation resistance caused by HNRNPC overexpression.
The RhoA/ROCK pathway has been implicated in dynamic crosstalk regulation between cancer cells and their microenvironment, which may be used to inhibit cancer metastatic processes. 42,43 Rho inhibition has been shown to decrease tumour cell survival after radiation treatment, while ROCK inhibitors, in particular, may enhance chemo-and radiotherapy efficacy. 44 Rho/ROCK inhibitors may also act as pro-vascular agents to improve tumour blood flow and increase cell exposure to chemotherapy or even sensitize cells to the radiation effects. 45 Although a potential role for RhoA, which has been shown to regulate glioblastoma radioresistance via survivin, 46 and ROCK2 in radiation resistance has been suggested, their collective mechanism remains poorly understood. 18 PC radiosensitivity has also been enhanced via the YAP/TAZ signalling pathway. 47 Growing evidence suggests that YAP promotes resistance to various anti-cancer therapies, including radiation therapy. 48 Our results show that HNRNPC mediates radiation resistance through the RhoA/ROCK-YAP/TAZ pathways and identifies HNRNPC as a potential target to sensitize PC to radiation.
The regulatory effects of RhoA/ROCK2 and RhoA/YAP on PC radiation resistance were also investigated in the present study.
Inhibition of the RhoA/ROCK2 pathway reduced DNA damage repair and radiation resistance. Previous studies have demonstrated that CAFs participate in cytoskeletal alteration via the RhoA/ROCK pathway and increased YAP localization. 28 YAP fibroblast activity remodelled matrix organization and increased matrix stiffness, promoting cancer cell migration and invasion. 18 Here, we found that HNRNPC or RhoA inhibition limited the activity of CAFs and weakened PC radiation resistance. Overall, our study identified the HNRNPC-RhoA/ROCK2-YAP/TAZ axis in PC radiation resistance, providing a potential therapeutic target for improving PC treatment.
There are some limitations associated with this study. First, only two regulatory pathways of HNRNPC in PC radiation resistance were explored. Other regulatory signaling pathways may be involved and should be investigated in the future. Second, this work did not conduct an in-depth research on the two identified signalling pathways. Thus, more detailed studies are required to make definitive conclusions about the role of these pathways in radiation resistance. Tumour-associated fibroblasts are an important part of cancer development and treatment, while tumour radiation resistance is a major problem in tumour radiotherapy. In order to clarify the relationship between CAFs and tumour radiation resistance in the future, the regulatory mechanism for CAFs in PC cell radiation resistance should be investigated in greater detail to provide more therapeutic targets and explore additional directions for cancer treatment.
In summary, we have shown that inhibition of the HNRNPC-RhoA/ROCK2-YAP/TAZ axis can potentially sensitize PC cells to radiation, providing a novel method to enhance the treatment of PC.

CO N FLI C T O F I NTE R E S T
The authors confirm that there are no conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.