Berberine attenuates XRCC1‐mediated base excision repair and sensitizes breast cancer cells to the chemotherapeutic drugs

Abstract Berberine (BBR) is a natural isoquinoline alkaloid, which is used in traditional medicine for its anti‐microbial, anti‐protozoal, anti‐diarrhoeal activities. Berberine interacts with DNA and displays anti‐cancer activities, yet its effects on cellular DNA repair and on synthetic treatments with chemotherapeutic drugs remain unclear. In this study, we investigated the effects of BBR on DNA repair and on sensitization of breast cancer cells to different types of DNA damage anti‐tumoural drugs. We found BBR arrested cells in the cell cycle S phase and induced DNA breaks. Cell growth analysis showed BBR sensitized MDA‐MB‐231 cells to cisplatin, camptothecin and methyl methanesulfonate; however, BBR had no synergistic effects with hydroxurea and olaparib. These results suggest BBR only affects specific DNA repair pathways. Western blot showed BBR down‐regulated XRCC1 expressions, and the rescued XRCC1 recovered the resistance of cancer cells to BBR. Therefore, we conclude that BBR interferes with XRCC1‐mediated base excision repair to sensitize cancer cells to chemotherapeutic drugs. These finding can contribute to understanding the effects of BBR on cellular DNA repair and the clinical employment of BBR in treatment of breast cancer.

Berberine (BBR) is an isoquinoline alkaloid extracted from the rhizomes of Coptis, which displays multiple biochemical functions and anti-microbial and anti-inflammatory activities. 14,15 BBR also exhibits anti-cancer activities through suppressing cell proliferation and inducing tumour cell apoptosis. [16][17][18][19][20][21][22] BBR reportedly directly binds with DNA, radio-sensitizes lung and oesophageal cancer cells and sensitizes ovarian cancer cells to cisplatin and PARP inhibitors. 23,24 Consequently, BBR may interfere with cellular DNA repair and would be a promising adjuvant agent to sensitize breast cancer cells to chemotherapeutic drugs. However, the influence of BBR on DNA repair was not well defined.
In this study, we attempted to analyse the DNA repair pathways influenced by BBR. We analysed the effect of BBR on cell cycle and on cellular DNA breaks. We also tested the cell proliferation upon stimulating with BBR in combination with DNA damage agents including cisplatin, camptothecin, methyl methanesulfonate (MMS), hydroxurea (HU) and olaparib. Cisplatin is classified as an alkylating agent and is the most widely used anti-tumour drug, which reacts with DNA to form interstrand crosslinks to inhibit its replication. 25,26 MMS is also an alkylating agent to methylate DNA on N7-G and N3-A, which would cause single-strand breaks (SSB) and the repair is dependent on XRCC1-mediated base excision repair (BER). 27 Camptothecin is a topoisomerase inhibitor that prevents DNA ligations and generates DNA breaks at the sites of DNA replication. 28 HU is an anti-tumoural drug classified as an antimetabolite one, which inhibits ribonucleotide reductase to block the formation of deoxyribonucleotides. 29 Olaparib is a poly-ADP-ribose polymerase inhibitor, which prevents the parylation of DNA repair factors and is a promising anti-tumour drug in breast cancer treatment. 30 After finding the potential DNA repair pathway affected by BBR, we analysed the relevant DNA repair factors in BBR-treated cancer cells to find out the factors that were affected by BBR. These studies will contribute to understanding the acts of BBR in cellular DNA repair, also the clinical employments of BBR in chemotherapeutic and radiotherapeutic treatment of breast cancer.
Alexa fluor 488-conjugated secondary antibody was purchased from Proteintech. Hydroxyurea, MMS and olaparib were purchased from

| MTT assay
Cell viability was measured by MTT assay. 6 × 10 3 cells were seeded in a well in 96-well culture plates and cultured overnight. Cells were treated with BBR for 24 hours. MTT (Solabio) was added into each well; after 4 hours, the cells were measured at 492 nm using a microplate photometer (Thermo Fisher Scientific).

| Cell cycle analysis
Approximately 2 × 10 6 cells were exposed to BBR. 24 hours after treatment, the cells were harvested, washed and centrifuged at 400

| Comet assay
Cells were treated with different doses of BBR for one night. 3 × 10 4 cells were isolated for analysis. Dipped 1% normal melting agarose into a frosted microscope slide, 10 μL cells were mixed with 90 μL 0.7% low melting agarose. The microscope slide was treated with ice-cold lysis buffer (100 mmol/L EDTA, 2.5 mol/L NaCl, 10 mmol/L Tris Base, 100 mmol/L EDTA, 1% sodium sarcosinate, 10% DMSO and 1% Triton X-100) for 2 hours. Then the cells were denatured in alkaline buffer

| Drug combination index analysis
The combination index (CI) was calculated through the software CompuSyn (Biosoft). CI <0.7 was considered as synergism. CI <0.5 was considered as strong synergism. The IC50 values of BBR with different doses of drugs were calculated by Graphpad Prism 5 (Graphpad).

| Statistical analysis
Data analysis was completed by GraphPad Prism 6.0 software. t Test was used to determine the significance; *P < .05, **P < .01, ***P < .001, were considered statistically significant. All experiments were performed in triplicate. Data are expressed as the mean ± SD.

| Berberine induces DNA breaks
We next analysed the effects of BBR on cellular DNA damages.
Through comet assay, we found BBR increased the lengths of comet tails in a dose-dependent manner, which demonstrates BBR induces DNA breaks. At the treatment of 40 µmol/L dose of BBR, the tails of comet were increased, while 160 µmol/L BBR increased much more folds of tails of comet compared with the control (Figure 2A, B). These results demonstrate that BBR induces cellular DNA damage. As BBR also arrests cell cycle in S phase, it is possible that BBR interferes with the cellular DNA repair. CompuSyn. 20 μmol/L BBR has synergistic effects (CI < 0.7) with MMS and has strong synergistic effects (CI < 0.5) with cisplatin and camptothecin. However, BBR has no synergistic effects with HU and olaparib (Table 1, Figure S1).

| Berberine sensitizes MDA-MB-231 cells to cisplatin, camptothecin, MMS, but not to HU and olaparib
At the dose of 10 and 20 μmol/L, BBR effectively increased the sensitization of TNBC to cisplatin, camptothecin and MMS. Cisplatin is considered as a type of alkylating agent and one of the most effective drugs used to treat cancers. 11,31 BER and HR reportedly repair the DNA damages induced by cisplatin. 32 MMS is also an alkylating agent and induces single-strand breaks, which is repaired by XRCC1-dependent BER. 33,34 Camptothecin is a topoisomerase I inhibitor acting to induce DNA breaks. The cancer cells over-expressed XRCC1 resist to camptothecin and the XRCC1 deficient cells are sensitive to camptothecin. 35,36 Therefore, these data suggest XRCC1 would be an important target for BBR to function in DNA repair. We also tested the synergistic effects between BBR and HU, as well as between BBR and olaparib. Olaparib is a PARP1 inhibitor and HU is to inhibit ribonucleotide

| Berberine severely down-regulates the levels XRCC1 and ERCC1, slightly decreases RAD51 , but has no effects on PARP1
We next analysed the effects of low doses of BBR on expressions of DNA repair factors through western blot, which showed BBR slightly decreased the expressions of Rad51, but severely decreased XRCC1 and ERCC1 in a dose-dependent manner ( Figure 4A, 4). Rad51 plays vital roles in HR, thus results suggested BBR may slightly affect the HR efficiency through regulating Rad51 expression. In BER, XRCC1 is a scaffolding protein to recruit Ligase III, DNA polymerase and PARP1 to repair the single-strand breaks. Therefore, BBR may impair the BER efficacy through down-regulating XRCC1. ERCC1 is a key factor in nuclear excision repair (NER) and has critical roles in removing the DNA adducts caused by cisplatin. PARP1 is a member of poly-ADPribose polymerase (PARP) family, which is to add poly-ADP-ribose to DNA repair factors to affect their binding to chromatin, thus it is considered as a DNA break sensor and involved in different DNA repair pathways. We found low doses of BBR could not influence the levels of PARP1. We also detected the cellular free poly-ADP-ribose F I G U R E 2 Berberine induces DNA breaks. Comet assay showed 20, 40, 80 and 160 μmol/L BBR promoted the tails of comet in the comet assay (A). The folds of changes of comet tails upon different doses of BBR were quantified (control was set as 1) (B). 30 tails were analysed in this assay. t Test, ***P < .001 as compared with the control F I G U R E 3 Berberine sensitizes breast cancer cells to cisplatin, camptothecin and MMS, but has no synergistic effects with HU and olaparib. MTT assay was used to analyse the viability of MDA-MB-231 cell treated with BBR and DNA damage anti-tumoural drugs including cisplatin (Cis), camptothecin (CAMP), methyl methanesulfonate (MMS), hydroxyurea (HU) and olaparib (Ola; A-E). 10 μmol/L BBR sensitizes breast cancer cells to cisplatin, camptothecin and MMS, but has no synergistic effects with HU and olaparib. t Test, ns, no significant difference, **P < .01, ***P < .001 as compared with the control (PAR) in MDA-MB-231 cells treated with BBR. Unexpected, BBR increased the cellular free PAR ( Figure S2A), which suggest BBR may regulate the enzymatic activity of PARP. However, we find BBR didn't change the levels of PARP1 ( Figure 4A, 4). PARP family includes 16 family members, in which PARP1, PARP2, PARP4 and PARP5 have enzymatic activity to add PAR into the target proteins. 37 Thus, BBR might not directly act on PARP1, but probably on the other PARP family members to regulate the parylation process. We also tested the transcript levels of XRCC1 in cells treated with BBR, while the results showed BBR didn't affect the XRCC1 transcription ( Figure S2B). F I G U R E 4 Western analysis of cellular RAD51, XRCC1, PARP1 and ECRR1 expressions after treatment with BBR. Western results showed BBR obviously decreased the expressions of XRCC1 and ERCC1, slightly decreased RAD51, but had no effects on PARP1 (A). The western assays were repeated for three times, and the results were quantified (B). t Test, **P < .01, ***P < .001 as compared with the control

| Rescue of XRCC1 helps the breast cancer cells to resist BBR, while the over-expressed PARP1 cannot influence the effects of BBR on the growth of breast cancer cells
After finding BBR obviously decreases the levels of XRCC1, we next analyse whether XRCC1 is the vital factor for BBR to affect cellular DNA damage and to influence the cell growth. We artificially expressed XRCC1 in two TNBC, MDA-MB231 and BT549, and tested the growth of cancer cells treated with BBR ( Figure S3A). We found the rescued XRCC1 recovered the resistance of cancer cells to BBR ( Figure 5A, 5). These data demonstrate XRCC1 is an important target for BBR to affect cellular DNA repair and cell growth. As we found BBR didn't change PARP1 expressions but affect the cellular free PAR, we over-expressed PARP1 in cancer cells to detect the cell growth after treatment with BBR ( Figure 5C, 5; Figure S3B). The over-expressed PARP1 did not help the cells to resist to 80 μmol/L dose of BBR. These results demonstrate that PARP1 is not involved in the effects of BBR on DNA repair and cell growth.

CO N FLI C T O F I NTE R E S T
The authors have no ethical conflicts to disclose.