Targeting the Radiation-Induced ARv7-Mediated circNHS/miR-512-5p/XRCC5 Signaling With Quercetin Signaling Increases Prostate Cancer Radiosensitivity


 BackgroundRadiation therapy (RT) with androgen deprivation therapy (ADT) is an effective therapy to suppress the locally advanced prostate cancer (PCa). However, we unexpected found that RT could also inducing the androgen receptor splice variant 7 (ARv7) expression to decrease the radiosensitivity. MethodsThe study was designed to target ARv7 expression with Quercetin or ARv7-shRNA led to enhancing increase the radiation sensitivity to better suppress the PCa that involved the modulating the circNHS/miR-512-5p/XRCC5 signaling.ResultsMechanism studies revealed that RT-induced ARv7 may function via altering the circNHS/miR-512-5p/XRCC5 signaling to decrease the radiosensitivity. Results from preclinical studies using multiple in vitro cell lines and in vivo mouse models concluded that combining RT with small molecule of Quercetin to target full-length AR and ARv7 could lead to better efficacy to suppress PCa progression. Conclusion﻿Together, these results suggest that ARv7 may play key roles to alter the PCa radiosensitivity, and targeting this newly identified ARv7 mediated circNHS/miR-512-5p/XRCC5 signaling with Quercetin may help us to develop a novel RT to better suppress the progression of PCa.


Abstract
Background Radiation therapy (RT) with androgen deprivation therapy (ADT) is an effective therapy to suppress the locally advanced prostate cancer (PCa). However, we unexpected found that RT could also inducing the androgen receptor splice variant 7 (ARv7) expression to decrease the radiosensitivity.

Methods
The study was designed to target ARv7 expression with Quercetin or ARv7-shRNA led to enhancing increase the radiation sensitivity to better suppress the PCa that involved the modulating the circNHS/miR-512-5p/XRCC5 signaling.

Results
Mechanism studies revealed that RT-induced ARv7 may function via altering the circNHS/miR-512-5p/XRCC5 signaling to decrease the radiosensitivity. Results from preclinical studies using multiple in vitro cell lines and in vivo mouse models concluded that combining RT with small molecule of Quercetin to target full-length AR and ARv7 could lead to better e cacy to suppress PCa progression.

Conclusion
Together, these results suggest that ARv7 may play key roles to alter the PCa radiosensitivity, and targeting this newly identi ed ARv7 mediated circNHS/miR-512-5p/XRCC5 signaling with Quercetin may help us to develop a novel RT to better suppress the progression of PCa.

Background
Prostate cancer (PCa) is the second leading cause of cancer deaths and the most common cancer in men in the United States (1) and its morbidity is also rapidly rising in Asian (2). Radiation therapy (RT) including external beam radiotherapy (EBRT) and brachytherapy is one of the standard treatment modalities for managing primary or locally advanced disease (3). However, there are still 20-25% of PCa patients with noninvasive disease (stage T1-T2) relapse after high doses of radiation in less than 5 years (1). Clinically, the combination of RT and androgen deprivation therapy (ADT) is superior to RT alone for treatment of patients with localized, intermediate-risk and high-risk PCa, since ADT may abrogate the androgens/androgen receptor (AR)-induced the DNA damage repair (DDR) system so that RT can be more e ciently to suppress PCa (4,5).
Despite use of combined ADT and RT, biochemical recurrence was up to 50% of high metastatic risk PCa patients (6); One possible mechanism proposed to explain this was that ADT treatment could induce the expression of AR variants (ARVs) especially AR variants 7 (ARv7) in PCa (7)(8)(9) and ARVs could mediate DDR following RT (10).Therefore, it is necessary to inhibit both full-length AR and ARv7 signal pathway to effectively improve PCa radiosensitivity. Moreover, there is no study about the effect of ionizing radiation ARVs especially ARv7 expression in PCa.
Quercetin (Que, 3,3',4',5,7-pentahydroxy avone) is a bioactive plant-derived avonoid, abundant in fruits and vegetables, that has been used as nutrition supplement in several countries (11). Daily human intake of quercetin ranges from 10 to 100 mg depending on different dietary habits, and it can reach 500-1,000 mg if selected highly puri ed extracts are used (12). Recent studies also indicated that Que might have anti-cancer property that could suppress tumor cell growth in multiple in vitro cell lines and in vivo mouse models in many cancers (13)(14)(15). Its linkage to the ARVs, especially to the IR-induced ARv7 during RT, however, remains unclear.
Circular RNAs (circRNAs) are highly conserved and stable covalently closed RNA transcripts generated by back-splicing of a single pre-mRNA with gene-regulatory potential (16,17). Emerging evidence shows that circRNAs are closely related to human diseases, especially cancers, and may act as better biomarkers due to their abundance and stability (18,19). circRNAs have also been linked to the radiosensitivity of several cancer types (20,21) and we previously found circZEB1 could decrease radiosensitivity by mediating miR-141-3p/ZEB1 signaling pathway in PCa cells (22).
Here we found that IR may function via inducing the ARv7 expression to decrease the subsequent RT e cacy, and adding Que may then suppress the IR-induced both full-length AR and ARv7 to better increase the PCa radiosensitivity Methods 2.1. Cell lines and cell culture. C4-2, 22Rv1 and VCaP cell lines were obtained from American Type Culture Collection (ATCC) and maintained in RPMI 1640 medium containing 10% fetal bovine serum (FBS), antibiotics (100 unit/ml penicillin, 100 µg/ml streptomycin), and 2 mM glutamine (Invitrogen) in a humidi ed 5% CO2 environment at 37°C. All cell lines were authenticated as mycoplasma and bacteria free following ATCC instructions during the three months prior to experiments.
2.3. Clonogenic survival assays. Cells were plated in 60-mm plates at various densities, depending on the IR dosage, and allowed to attach overnight. Cells were treated with media and metformin or vehicle for 48 hours. Then, the cells were irradiated, and drug treatments were removed as soon as possible after IR, after which the cells were incubated for 14 days. Survival curves for each experiment were constructed by plotting the mean surviving fractions semilogarithmically as a function of the IR dose. The data were analyzed, and survival curves were plotted following a multitarget/single-hit model (SF = 1-[1-e -D/D0 ] N ) to plot the surviving fraction curve using GraphPad Prism 7.0 software, where SF is the surviving fraction and D is the radiation dose.
2.4. Quantitative real-time PCR. Total RNA was isolated using TRIzol ® Reagent (#15596026 ThermoFisher Scienti c) according to the manufacturer's instructions. Total RNA was reverse transcribed into cDNA using the iScript™ cDNA Synthesis Kit (# 1708891, Bio-Rad). The primer sequences are listed in Table S1. qRT-PCR was performed using a Bio-Rad iQ5 real-time thermal cycler and iQ™ SYBR® Green Supermix (#1708880, Bio-Rad). Relative mRNA expression levels were normalized against GAPDH levels (as an internal control) and determined by the 2 −ΔΔCt method.
Protein-antibody complexes were detected by SuperSignal™ West Femto Maximum Sensitivity Substrate (#34095, ThermoFisher Scienti c) using the Bio-Rad imaging system. 2.6. Neutral comet assay. Double strand breaks (DSBs) were assessed by a single-cell gel electrophoretic comet assay kits (Trevigen) under neutral conditions according to the manufacturer's protocol. Brie y, cells were harvested at the indicated times after a single dose of 6 Gy irradiation, mixed with 0.7% low melting point agarose, and plated on CometSlide microscope slides, followed by cell lysis. Cells then subjected to electrophoresis under neutral conditions and were stained with SYBR Gold (S11494, Thermo Fisher Scienti c). Comets were visualized using a Zeiss uorescent microscope and were scored (100 cells per treatment) followed by analysis using CASP Comet Assay Software. 2.7. Chromatin immunoprecipitation assay. Cells were cross-linked with 4% formaldehyde for 10 min followed by cell collection and sonication with a predetermined power to yield genomic DNA fragments 300-1000 bp in length. Lysates were precleared sequentially with normal rabbit IgG (#sc-2027, Santa Cruz Biotechnology) and Protein A/G. Anti-ARv7 antibody (2.0 µg) was added to the cell lysates and incubated at 4°C overnight. As a negative control, IgG was used in the reaction. Speci c primer sets were designed to amplify a target sequence within the human QKI gene promoter. PCR products were analyzed by agarose gel electrophoresis. 2.8. Luciferase reporter assay. The human promoter region of the QKI 5' promoter was constructed into the pGL3-basic vector (Promega). Site-directed mutagenesis of the ARv7 binding site in the NHS 5' promoter was achieved with the Quick-Change mutagenesis kit. C4-2 -IRR cells were plated in 24-well plates, and the cDNAs were transfected with Lipofectamine 3000 transfection reagent (Invitrogen) according to the manufacturer's instructions. PRL-TK was used as an internal control that served as the baseline control response. Luciferase activity was measured 48 hours after transfection by the Dual-Luciferase Reporter Assay System (Promega) according to the manufacturer's manual.
2.9. Pull-down assay with biotinylated circNHS probe. In brief, 1 × 10 7 C4-2-IRR cells were harvested, lysed, and sonicated. The cell lysate mixture was rotated overnight at 4°C after adding 2 µL RNase inhibitor and 500 pmol/L biotin-labeled antisense oligo against circNHS. The lysate mixture was rotated for 2 hours at 4°C after adding 10 µL streptavidin agarose beads (No. 88817, Pierce Biotechnology). Total RNA was extracted by TRIzol (Invitrogen) according to the manufacturer's protocol, reverse transcribed, and subjected to qPCR analysis to detect the miRNAs. 2.10. RNA in situ hybridization. The biotin-labeled circNHS probe was designed and synthesized by IDT (Integrated DNA Technologies, USA), and the probe sequence is available upon request. The probe signal was detected by Alexa Fluor™ 594 Tyramide SuperBoost™ Kit (Thermo Fisher Scienti c, USA) according to the manufacturer's instructions.
2.11. RNA isolation from nuclear and cytoplasmic fractions. The subcellular localization of circNHS was detected using the PARIS™ Kit according to the manufacturer's protocol (Thermo Fisher Scienti c, USA).
2.12. Actinomycin D assay. C4-2-IRR and 22Rv1 cells were seeded at 1 × 10 5 cells per well in a 6-well plate overnight and then exposed to 2 mg/L actinomycin D (Sigma, USA) for 6, 12, 18 and 24 hours. The cells were harvested at the indicated time points, and the stability of circNHS was analyzed using qRT-PCR.
2.13. Immuno uorescence. For immuno uorescence, the transfected cells were seeded on glass coverslips for 24 h and then incubated with primary antibodies against γ-H2AX at 4°C overnight. Cell nuclei were counterstained with 2 µg/ml DAPI for 5 min. Images were captured with a microscope using 20x objectives.

3.1.IR-induced ARv7 expression led to decrease radiosensitivity
Recent studies indicated that ADT treatment would induce the expression of ARv7 (23,24),the potential impact of IR treatment on ARv7 expression, however, remains unclear. Here we rst developed the radioresistant C4-2-IRR (C4-2-Ionizing Radiation Resistance) cell line via treating parental C4-2 cells with repeated 2 Gy radiation. Radioresistance of cells were con rmed in clonogenic assays by showing higher survival of C4-2-IRR cells than respective parental cells at indicated radiation dose (Fig. S1A). Interestingly, we found that IR could increase signi cantly the ARv7 expression at the protein and mRNA levels after IR in ARv7-positive 22Rv1 cells and 22Rv1 xenografts ( Fig. 1A and Fig. 1B).
To further examine if IR may function via inducing the ARv7 to increase the radioresistance, we rst added 5um enzalutamide to inhibit AR activity, then knocked down ARv7 with shARv7 in C4-2-IRR cells or ectopic overexpression of ARv7 in C4-2 cells. The results revealed that suppressed ARv7 in C4-2-IRR cells led to increase the PCa radiosensitivity (Fig. 1C). In contrast, increased ARv7 in C4-2 cells can then increase the radioresistance of PCa cells (Fig. 1D). As 22Rv1 cells have both AR and ARv7, we then applied the shAR and/or shARv7 to suppress the expression of AR and/or ARv7. The results revealed such suppression to AR alone or ARv7 alone could not led to decrease clonogenic survival following IR (Fig. 1E), and only knockdowning of both AR and ARv7 led to decrease the clonogenic survival following IR (Fig. 1E). Together, these results suggest that suppression both AR and ARv7 could lead to better effect to improve PCa radiosensitivity.
To further strengthen above conclusion, we added 5um enzalutamide to inhibit AR activity, then knocked down ARv7 with shARv7 in C4-2-IRR cells or ectopic overexpression of ARv7 in C4-2 cells. Next, we examined the ARv7 potential impact on PCa cells DDR by comet assays and the results from these assays also revealed that decreasing ARv7 in C4-2-IRR cells signi cantly increased DNA fragmentation following IR (Fig. 1F). Similar results from WB (Fig. 1G, left) analysis of the γ-H2AX levels showed that decreasing ARv7 in C4-2-IRR cells increased the DNA damage repair time. In contrast, increasing ARv7 resulted in less time to complete DDR in the C4-2 cells (Fig. 1G, right).
Together, results from Fig. 1 suggest that IR-induced ARv7 may decrease the continue RT e cacy.
3.2. Mechanism dissection of how ARv7 can alter the radiosensitivity: via increasing circNHS expression.
To further dissect the mechanism of how ARv7 can alter the radiosensitivity, we focused on circRNAs, as recent studies indicated that the expression of selected circRNAs could be altered after IR exposure (20,21). From circRNA Array data in radioresistant cancer cells (25), we rst chose the 10 most upregulated/downregulated circRNAs and detected their expression in C4-2-IRR cells model. The results revealed that only 7 of those 20 circRNAs had changed signi cantly compared with C4-2 parental cells ( Fig. 2A), and PCR results revealed that decreasing ARv7 led to signi cantly decreased expression of circNHS (hsa_circ_0089974) in C4-2-IRR cells (Fig. 2B). In contrast, increasing ARv7 led to signi cantly increased expression of circNHS in C4-2 cells (Fig. 2C). Moreover, we found that IR could increase signi cantly circNHS levels in 22Rv1 xenografts (Fig. 2D).
Next, results from circBase database analysis indicated that circNHS is generated from back-splicing of two exons (exons 2 and 3) of the NHS gene (chrX:17705861-17710588) (Fig. 2E). To avoid trans-splicing or genomic rearrangements, including head-to-tail splicing, we then applied multiple approaches to rule out such possibilities. We rst designed convergent primers to amplify NHS mRNA and divergent primers to amplify circNHS. Using cDNA and genomic DNA from C4-2-IRR and 22Rv1 cell lines as templates, circNHS was ampli ed from cDNA only by the divergent primers, whereas no ampli cation product was observed from genomic DNA (Fig. 2F). Furthermore, the results from the RNase R assay revealed that circNHS is the circular form, with better resistance to RNase R digestion than linear NHS (Fig. 2G). Finally, the results from adding actinomycin D to inhibit transcription also indicated that the half-life of circNHS is more stable than NHS mRNA in C4-2-IRR and 22Rv1 cells (Fig. 2H). Finally, the circular RNA microarrays data from Yang et al (26) showed that circNHS were highly expressed in high-grade (Gleason > 8) PCa tissues when compared with low-grade (Gleason < 6) PCa tissues (Fig. S1B).
Together, the results from Fig. 2 suggest that ARv7 may decrease radiosensitivity by increasing circNHS expression.

ARv7 increases circNHS expression via altering the transcriptional regulation
To further dissect the molecular mechanism of how ARv7 can increase circNHS expression, we examined circNHS transcriptional regulation, and our data indicated that ARv7 could increase NHS expression at both the mRNA and protein levels ( Fig. S1C and Fig. S1D). We then searched for potential androgen response elements (AREs) on the NHS 5' promoter region using the JASPAR database. The results revealed that 2 putative AREs were located within 2 kb of the NHS 5' promoter region (Fig. S1E). We then performed chromatin immunoprecipitation (ChIP) in vivo binding assays in 22Rv1 cells, and the results revealed that ARv7 could bind to the rst ARE (Fig. 3A), suggesting that ARv7 might be able to increase circNHS expression via direct binding to a ARE to exert its transcriptional regulation.
We then performed the luciferase reporter assay by inserting the 1 kb 5' promoter region of NHS containing ARE1 into the pGL3 luciferase backbone and also generated a version with a mutated ARE1 (Fig. S1F). The luciferase assay results revealed that decreasing ARv7 by adding ShARv7 signi cantly decreased luciferase activity in 22Rv1 transfected with the wild-type NHS promoter construct but not in cells with the mutant NHS promoter construct (Fig. 3B). Importantly, the results from the LinkedOmics database via TCGA data analysis also showed that NHS expression was positively correlated with AR expression in PCa patients (Fig. S1G).
Together, results from Fig. 3A-B suggest that ARv7 increases circNHS expression via altering the transcriptional regulation Next, to prove that ARv7 may function by altering circNHS expression to decrease radiosensitivity, we then suppressed the circNHS with the shcircNHS, and results revealed that suppressed circNHS (Fig. S1H) led to increase the radiosensitivity in C4-2-IRR cells (Fig. 3C). In contrast, increasing circNHS resulted in decreasing the radiosensitivity in C4-2 cells (Fig. 3D).
The results from the γ-H2AX assay also revealed that decreasing circNHS levels could increase the DNA damage repair time in C4-2-IRR cells (Fig. 3E, left panel). In contrast, increased circNHS expression resulted in less time to complete DNA damage repair in the C4-2 cell lines (Fig. 3E, right panel). Similarly, the results from the comet assay also revealed that decreasing circNHS in C4-2-IRR cells signi cantly increased DNA fragmentation following IR (Fig. 3F).
Finally, the results from the silencing experiment revealed that suppressing circNHS led to a partial reversal of ectopic overexpression of ARv7-suppressed radio-sensitivity in the C4-2-IRR cells (Fig. 3G) and increased circNHS led to a partial reversal of the shARv7-increased radio-sensitivity in the C4-2 cells (Fig. 3H).
Together, results from Fig. 3 suggest that ARv7 can function via increasing the circNHS to alter the radiosensitivity.
3.4. Mechanistic dissection of how the circNHS could decrease radiosensitivity: by competing with miR-512-5p/XRCC5 axis Next, to determine how circNHS could decrease radiosensitivity, we rst determined the subcellular localization of circNHS in PCa cell lines using a nuclear mass separation assay (Fig. 4A) and FISH analysis (Fig. 4B). The results revealed that circNHS was expressed mainly in the cytoplasm of PCa cells.
As early studies indicated that cytoplasmic circRNAs might compete with miRNAs to exert their function (27), we then performed an RNA immunoprecipitation (RIP) assay with an antibody against argonaute 2 (AGO2) in C4-2-IRR and 22Rv1 cells, and the results revealed that circNHS was signi cantly enriched by the AGO2 antibody (Fig. 4C), suggesting that circNHS might act as a binding platform for AGO2 and miRNAs. Based on these ndings, we predicted that circNHS might serve as a binding platform for miRNAs. We then surveyed 3 databases (28, 29) and found that 16 potential miRNAs could bind to circNHS (Fig. 4D). Using probes speci cally against circNHS to analyze these 16 candidate miRNAs in the complex, we found only a speci c enrichment of miR-512-5p and none of the rest 15 miRNAs, suggesting that miR-512-5p is one of the circNHS-associated miRNAs in PCa cells.
To further prove that the circNHS can alter radiosensitivity via binding to the miR-512-5p, we then constructed a mutated binding site of miR-512-5p (circNHS-Mut, Fig. S1I). The results from colony formation assays revealed that mutation of the miR-512-5p binding site could completely abolish circNHS-induced radiosensitivity in C4-2 cells (Fig. 4E).
Recent studies identi ed 542 ARv7-regulated genes via RNA-seq and CHIP-seq analyses (30). Among these genes, we found 9 DDR genes that may involve in DNA damage repair, and results from targetscan database predicted that XRCC5 and SPRTN can be the potential target gene of miR-512-5p (Fig. 4F).
Results from western blot analysis demonstrated that circNHS could regulate the expression of XRCC5 protein, and not SPRTN protein levels, suggesting circNHS may function via sponging miR-512-5p to increase the protein expression of XRCC5 (Fig. 4G). As expectedly, adding miR-512-5p inhibitor led to increase XRCC5 protein expression and could partly reverse the shcircNHS-decreased XRCC5 expression in C4-2 cells (Fig. 4H, left) while adding miR-141-3p mimics led to decreased XRCC5 protein expression and could effectively reverse the oe-circNHS increased XRCC5 expression in C4-2-IRR cells (Fig. 4H, right).
Finally, the results from the LinkedOmics database via TCGA data analysis also showed that XRCC5 expression was positively correlated with AR expression in PCa patients (Fig. S1J).
Together, the results from Fig. 4 suggest that circNHS might function by sponging miR-512-5p to alter XRCC5 protein expression and such regulation may then lead to alter the DDR pathway.
3.5. Preclinical studies in in vitro cell lines and in in vivo mouse model to target ARv7-mediated circNHS/miR-512-5p/XRCC5 signaling with small molecule of Que to increase radiosensitivity All above results suggest that IR may have unwanted side effects of inducing the ARv7 expression to increase the radioresistance. We are interested to see if adding small molecules can have similar effect to suppress the ARv7 expression. We are interested to test the Que, as our previous studies indicated that Que could suppress AR mRNA and protein expression (31). Here we found that adding Que could also signi cantly decrease the ARv7 expression at both mRNA and protein levels in 22Rv1 and VCaP cells (Fig. 5A-C).
Furthermore, adding 40µM Que to reduce ARv7 expression can then lead to increase the radiosensitivity to better suppress the C4-2-IRR and 22Rv1 cell growth using clonogenic assay (Fig. 5D-E). Mechanism dissection revealed that adding Que, and not Enz, could decrease DNA damage repair via detecting γ-H2AX levels (Fig. 5F).
To further validate these in vitro ndings in the in vivo mouse model, we established subcutaneous 22Rv1 xenografts in nude mice. Results showed that combined 2Gy IR treatment with Que treatment (75mg/kg) led to suppress the tumor growth approximately 70-75% on day 28, as compared to 2Gy IR alone or Que alone (Fig. 5G).
Together, results from Fig. 5 suggest that small molecule of Que can increase radiosensitivity to better suppress the PCa cells growth via targeting both AR and ARv7 levels.

Discussion
Studies reported that AR-FL mRNA is coded from eight exons and the protein contains four functional domains: an N-terminal domain (NTD), a central DNA-binding domain (DBD), a short hinge region and a C-terminal LBD (32). Constitutively active ARVs have now been discovered (33), most of which contain the NTD and the DBD, but lack the LBD owing to a truncation by a cryptic exon. As conventional ADT inhibits androgen-dependent activation of AR, the presence of C-terminal truncated ARVs provides a compelling mechanism for CRPC cells to circumvent ADT (34). ARVs are nuclear, constitutively active, bind to similar ARE, as that engaged by full-length AR (34). Among > 20 different ARVs that have been identi ed in clinical PC and PC cell lines and among these variants, AR-V7, originates from contiguous splicing of AR exons 1, 2, and 3 with the cryptic exon 3 (CE3) present within the canonical intron 3 of the AR gene, is one of the most well characterized AR-Vs and can be reliably measured in tissue and liquid biopsy specimens, and blood-based detection (35)(36)(37)(38). Owing to lack of LBD, AR-V7 can circumvent the pharmacological effects of second-generation antiandrogens such as enzalutamide, which targets LBD of AR directly. Besides,ARv7 could be induced by ADT treatment and also serves as a predictive biomarker for response to enzalutamide or abiraterone acetate (7,9,23) .
We and some groups have demonstrated that AR expression and activity are durably upregulated following IR in PCa models and AR can activate DDR pathways (4,5,10,39,40) in PCa, providing rationale for concurrent ADT + RT therapy (6, 41) is better than RT therapy alone. These preclinical studies also have led to testing combinations of RT with second-generation antiandrogens such as abiraterone, enzalutamide, apalutamide and AR degradation enhancer, ASC-J9 (40, 42-45). However, despite use of combined ADT and RT, the recurrence rate is up to 50% of high-risk PCa patients (6). To dissect this mechanism, we focus on the role of ARv7 expression and activity in PCa after RT and radioresistant PCa cells. Our in vitro and in vivo experiments results suggest that ARv7 is highly expressed in radioresistant PCa cells and IR could induce ARv7 expression in PCa which increase the clonogenic survival of PCa cells after irradiation. Moreover, these IR induced ARv7 could mediate PCa DDR and is not affected by enzalutamide treatment. Thus, our study may explain the clinical phenomenon that many patients rst respond to RT + ADT then eventually became biochemical disease relapse and that patients with ARv7 expression might not be eligible for such a combination treatment. Similarly, Yin et al. also found that ARVs alone can regulate DDR following PCa RT. To clarify these issues, further investigation on PCa RT clinical samples are necessary.
To further dissect the mechanism how ARv7 may alter the IR-sensitivity, recently studies found that circRNAs may play important roles in radiosensitivity. Several studies have revealed that circRNA levels might be substantially altered after IR, including esophageal cancer (3), cervical cancer (46), glioma (47), nasopharyngeal carcinoma (48), oral squamous cell carcinoma (49), colorectal cancer (50) and hepatocellular carcinoma (51). Previously, we found that circZEB1 was upregulated after IR in PCa cells and might function as a sponge for miR-141-3p (22). In this study, we analyzed our radioresistant C4-2 cell models and analyzed the twenty most altered circRNAs from circRNA-Array data (25). Studies showed that AR-V7 is constitutively active and capable of activating genes even in the absence of androgen (52). Our results showed that only circNHS could be regulated by ARv7 among these twenty circRNAs. By gain/loss of function assays of circNHS, we found that it could partially reverse the effects of ARv7 on PCa radiosensitivity. Moreover, there are currently no other published reports about circNHS functions now. To our knowledge, this is the rst study about circRNA function in radioresistant PCa cells to date, which might contribute to a better understanding of ARv7 function in RT.
Taken together, a requirement for inhibiting both full-length AR and ARv7 transcriptional activities is necessary to effectively improve radiosensitivity. To target the IR-induced both AR and ARv7 with small molecules, we screened and found that a small molecule Quercetin which is a avonoid widely found in vegetables and fruits particularly in onions, apples, and red wine (12). Moveover, quercetin is of very low toxicity and rarely produce any side effects even at high dose of 200 mg/kg given to rats and mice (53,54), and clinical trial showed that total consumption of 1000mg quercetin per day could be well tolerated in human not associated with any side effects (55, 56).Recent studies have found its antitumor role in many cancers and can function as radiosensitizer in many cancers (57-59). Our previous study also revealed that quercetin inhibited LNCaP cells growth through inhibiting androgen receptor expression and its inducible genes (31). In this study, we found that quercetin could inhibit the transcriptional activities of both full-length AR and ARv7 and enhance RT-mediated cell killing compared with second-generation antiandrogens such as enzalutamide. Thus, we provide proof-of-principle pre-clinical in vitro and in vivo evidence to rationalize the clinical use of quercetin to enhance the effect of IR as a potential strategy to improve the outcomes of PCa patients who undergo RT. Indeed, further randomized clinical trials are required to assess the effect of combining Que and RT for PCa patients.

Conclusions
In summary, our research here demonstrated that IR could induced ARv7 in PCa cells and IR-induced ARv7 could decrease the radiosensitivity through mediating circNHS/miR-512-5p/XRCC5 pathway. Targeting  Availability of data and materials The datasets used and/or analyzed during this study are available from the corresponding author on reasonable request.     and/or oeXRCC5 effects on C4-2-IRR cells survival after IR using clonogenic assay. (J) The oeARv7 and/or shXRCC5 effects on C4-2 cell survival after IR using clonogenic assay. **P < 0.01 compared with the controls. N.S., not signi cant compared with the controls. Data are presented as mean ± SEM. were drawn. Representative Images of tumors are presented after mice were sacri ced on day 28 and the tumor volume in each group was measured. Data are presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001 compared with the controls. N.S., not signi cant compared with the controls.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.