Enhanced expression of DNA polymerase eta contributes to cisplatin resistance of ovarian cancer stem cells

Reduced DNA Damage Formation and Enhanced DNA Repair Capacity Were Not Consistently Found in Ovarian CSCs upon Cisplatin Treatment.

Inefficient formation of DNA lesions and enhanced DNA repair have been implicated in cancer therapy resistance in CSCs (24). We sought to determine whether ovarian CSCs are also resistant to the formation of cisplatin-induced DNA lesions and exhibit enhanced DNA repair capacity. CD44⁺CD117⁺ cells were isolated from ovarian cancer cell lines 2008 and C13 (25, 26) by using fluorescence-activated cell sorting (FACS) and have been considered CSCs based on their characteristics (19, 27). The MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide] cell viability assay confirmed the cisplatin resistance of these CD44⁺CD117⁺ cells compared with their corresponding unsorted cells (SI Appendix, Fig. S1 A and B). However, we were unable to demonstrate a consistently reduced formation of 1,2-intrastrand cross-links (Pt-GG) in ovarian CSCs compared with the unsorted bulk cancer cells following the same amount of cisplatin treatment (SI Appendix, Fig. S2 A and C). No significant reduction of Pt-GG formation was found in another CSC population isolated from ovarian cancer cell line SKOV3 based on the spheroid formation (27, 28) (SI Appendix, Fig. S2 B and D). The removal rates of Pt-GG in these cells were further analyzed after equivalent amounts of Pt-GG were generated. Again, we failed to find an enhanced capacity of 2008-CSCs, C13-CSCs, and SKOV3-CSCs to remove cisplatin-induced DNA lesions (SI Appendix, Fig. S2 E–G). Given that cisplatin can induce other DNA lesions besides Pt-GG (29), we performed inductively coupled plasma mass spectrometry to determine the Pt content in DNA to account for all DNA-Pt adducts. Similarly, neither significant reduction of DNA-Pt adducts nor significant increase of the removal rate of DNA-Pt adducts was found in these CSCs compared with their corresponding bulk cancer cells (SI Appendix, Fig. S3 A–F). Taken together, these data suggest that inefficient DNA lesion formation and enhanced DNA repair capacity are not likely to be the cause of increased cisplatin resistance in ovarian CSCs. It is worth noting that even more DNA-Pt adducts were formed in C13-CSCs than their corresponding bulk cancer cells (SI Appendix, Figs. S2 A and C and S3B). The differences in the formation of DNA lesion and in DNA repair capacity among these ovarian CSCs suggest the complexity of cisplatin resistance, and there may likely be a distinct major mechanism for cisplatin resistance in ovarian CSCs with different backgrounds.

Expression of TLS Pol η Is Elevated in Ovarian CSCs. (V体育平台登录)

One of the important cell survival mechanisms following cisplatin treatment is TLS, which is mediated by specialized DNA Pols (11). The analyses of expression levels of various TLS Pols in ovarian CSCs demonstrated that POLH mRNA (encoding Pol η) is highly expressed in 2008-CSCs, C13-CSCs, and SKOV3-CSCs (Fig. 1 A–C and SI Appendix, Fig. S4). Interestingly, the POLH mRNA levels in 2008, 2008-CSCs, C13, and C13-CSCs correlate well with cisplatin sensitivity (SI Appendix, Figs. S1 vs. S5). We wished to extend these observations to the CSCs isolated from primary human ovarian tumors. Ovarian serous adenocarcinomas were disaggregated and subjected to different growth conditions, either for regular monolayer adherent growth of tumor cells or for the selection of self-renewing, nonadherent spheroids, which have been demonstrated to be CSCs (19). Disaggregated tumor cells from five patients were able to form spheroids under CSC-selective culture conditions (SI Appendix, Fig. S6A). The spheroids displayed significantly enhanced expression of Nanog, a marker of CSCs, compared with the bulk tumor cells (SI Appendix, Fig. S6B). Most importantly, the spheroid cells isolated from four out of five primary ovarian tumors also exhibited significantly increased expression level of POLH mRNA (Fig. 1D). In support of this finding, we also demonstrated an enhanced protein level of Pol η in 2008-CSCs, C13-CSCs, and SKOV3-CSCs compared with their corresponding bulk cancer cells (Fig. 1E). In addition, an elevated basal level of ub-PCNA was also revealed in these CSCs (Fig. 1F). These observations suggest an enhanced TLS activity in ovarian CSCs mainly mediated by elevated expression of Pol η.

VSports app下载 - Pol η Is Required for Cisplatin-Induced Enrichment of the CSC Population.

Cisplatin treatment efficacy is inversely correlated to the expression level of Pol η in various cancers (30–32). To determine whether Pol η down-regulation affects the efficacy of cisplatin treatment in ovarian cancers, we established a 2008 cell line with Pol η stable knockdown and generated xenografts by injecting cells s.c. into Athymic nude mice. Upon tumor presentation, mice were chronically treated with cisplatin six times during a period of 74 d. As shown in SI Appendix, Fig. S7 A and B, POLH-deficient transplants exhibited an enhanced response to cisplatin relative to POLH-proficient controls, indicating that down-regulation of Pol η is also able to sensitize ovarian cancer to cisplatin.

Cisplatin treatment is capable of inducing enrichment in a population of cells with CSC properties, probably due to killing of cisplatin-sensitive bulk cancer cells and survival of cisplatin-resistant CSCs (33, 34). We sought to assess the contribution of Pol η to cisplatin-induced enrichment of CSCs. In 2008, C13, and SKOV3 cells, treatment with cisplatin did increase the percentage of the CSC population defined as CD44⁺CD117⁺ cells. Down-regulation of Pol η expression by transient transfection of POLH siRNA into these cells reduced the cisplatin-enriched CSC population (Fig. 2 A–F and SI Appendix, Fig. S8 A–C). Consistently, treatment with cisplatin failed to enrich the CSC population in a 2008 cell line with POLH stable knockdown (SI Appendix, Fig. S9 A and B). We further validated this finding in a xenograft model with in vivo cisplatin treatment. Mice bearing POLH-proficient or POLH-deficient xenografts (Fig. 3A) were treated with cisplatin twice. Similar to the chronic treatment, this short-term cisplatin treatment also inhibited the growth of all xenografts, with POLH-deficient tumors exhibiting a less progressive growth dynamics and more significant tumor regression following treatments (Fig. 3B), in comparison with POLH-proficient transplants. The tumor cells were isolated 2 d after the second treatment and subjected to FACS to determine the percentage of CD44⁺CD117⁺ cells. As expected, in vivo cisplatin treatment enriched CSCs in the POLH-proficient, but not POLH-deficient xenografts (Fig. 3 C and D). Taken together, these data suggest Pol η plays a critical role in reducing the efficacy of cisplatin to shrink ovarian tumors and in the enrichment of CSC population upon cisplatin treatment.

Down-Regulation of TLS Pol η Sensitizes Ovarian CSCs to Cisplatin Treatment. (V体育安卓版)

To assess the contribution of Pol η to cisplatin resistance in CSCs, the expression of Pol η in 2008-CD44⁺ CD117⁺, C13-CD44⁺ CD117⁺, and SKOV3-spheroid cells was knocked down with either POLH siRNA or shRNA, and the cell sensitivity to cisplatin was determined. Down-regulation of Pol η promoted cisplatin-induced cell killing in all these ovarian CSC populations, with 2.1–8.3-fold reduction of IC50 (Fig. 4 A–C and SI Appendix, Fig. S10 A and B). Annexin V staining and FACS analysis of SKOV3-spheroid cells further revealed that the number of apoptotic cells increased in POLH–down-regulated cells upon cisplatin treatment (Fig. 4D). All these data suggest that Pol η-mediated TLS facilitates CSCs to survive cisplatin treatment, leading to an enrichment of the CSC population.

Decreased Expression of miR-93 Is Responsible for the Increased Pol η Expression in Ovarian CSCs.

Micro RNAs (miRNAs) can be differentially expressed in CSCs and regulate their characteristics (35). By using the web-based algorithms miRDB and miRanda, we identified two miRNAs that have high potential to bind to 3′-UTR of POLH (miR-93 and miR-20b) (Fig. 5A and SI Appendix, Fig. S11A). qRT-PCR analyses demonstrated that miR-93 level was significantly lower in all CSC populations derived from three ovarian cancer cell lines compared with their corresponding bulk cancer cells (Fig. 5 B and C and SI Appendix, Fig. S12A). The CSC populations derived from primary tumors exhibited significantly lower miR-93 level compared with their corresponding bulk tumor cells as well (Fig. 5D). However, a lower level of miR-20b was only found in C13 and SKOV3 CSCs (SI Appendix, Figs. S11 B and C and S12B), and no significant difference for miR-20b expression was found in the CSC population and bulk cancer cells derived from primary tumors (SI Appendix, Fig. S11D). Interestingly, those CSC populations (OV-7, OV-9, OV-10, and OV-11) showing lower levels of miR-93 also exhibited higher levels of POLH (Figs. 1D vs. 5D). These data suggest an inverse correlation between miR-93 and POLH expression levels in ovarian CSCs and indicate that miR-93 might regulate the expression of POLH.

To establish the regulatory role of miR-93 in POLH expression, 2008, C13, and SKOV3 cells were transfected with miR-93 inhibitors, whereas 2008-CD44⁺CD117⁺, C13-CD44⁺CD117⁺, and SKOV3-spheroid cells were transfected with miR-93 mimics. qRT-PCR analyses demonstrated that down-regulation of miR-93 in 2008 and C13 cells enhanced the POLH mRNA levels (SI Appendix, Fig. S13 A and B), whereas overexpression of miR-93 in 2008-CSCs and C13-CSCs reduced the POLH mRNA levels (SI Appendix, Fig. S13 C and D). Furthermore, the inhibitory effect of miR-93 on the protein level of Pol η was also revealed by immunoblotting analyses (Fig. 6 A and B). In contrast, transfection of miR-20b mimics did not have influence on the expression of Pol η in these CSCs (SI Appendix, Fig. S11E). Collectively, these data suggest that reduced expression of miR-93 contributes to the elevated expression of POLH in ovarian CSCs.

Given that miR-93 down-regulates Pol η expression, overexpression of miR-93 should be able to enhance the sensitivity of ovarian CSCs to cisplatin treatment by decreasing the Pol η level. Indeed, transfection of miR-93 mimics into 2008-CD44⁺CD117⁺, C13-CD44⁺CD117⁺, and SKOV3-spheroid cells increased cisplatin-induced cell death and cellular apoptosis (Fig. 6 C–H and SI Appendix, Fig. S14 A–C), indicating that enforced miR-93 expression sensitizes ovarian CSCs to cisplatin treatment.

Cell Culture and Cisplatin Treatment.

Human ovarian cancer cell line 2008 and its resistant cell line C13 (25) were kindly provided by Dr. Francois X. Claret (University of Texas, M. D. Anderson Cancer Center). The SKOV3 ovarian cancer cell line was provided by Dr. Thomas C. Hamilton (Fox Chase Cancer Center). The 2008 cells with stable knockdown of Pol η (2008-shPOLH-1) were successfully established in our laboratory. All cell lines were authenticated by DNA (short tandem repeat) profiling and maintained in Roswell Park Memorial Institute medium 1640 supplemented with 10% (vol/vol) fetal bovine serum. The 2008-CD44⁺CD117⁺, C13-CD44⁺CD117⁺, and SKOV3-spheroid cells were maintained in ultralow attachment plates in knockout Dulbecco’s modified Eagle medium/F12 medium supplemented with 20% knockout Serum Replacement (Life Technologies), 20 ng/mL epidermal growth factor, and 10 ng/mL basic fibroblast growth factor. All cells were grown at 37 °C in humidified atmosphere of 5% (vol/vol) CO₂ in air. Cisplatin (Sigma) stocking solution was prepared freshly with PBS and further diluted to the desired concentration with culture medium for cell treatment.

Flow Cytometry Analysis and Cell Sorting.

Anti–CD117-PE and anti–CD44-FITC (BD Pharmingen) were used for flow cytometric analysis and cell sorting. Briefly, cells were incubated with antibodies on ice for 40 min in the dark. After washing with cold PBS, cells were resuspended in 200 µL PBS and subjected to FACS analyses on a BD FACS Aria III at The Ohio State University Analytical Cytometry Shared Resource.

Isolation of Primary Tumor Cells from Freshly Removed Human Ovarian Tumors.

Freshly removed ovarian tumors were obtained from Department of Pathology at the Ohio State University within 4 h after surgery in accordance with a protocol approved by the Ohio State University’s IRB. The single tumor cells were obtained by enzyme digestion and cultured enrichment of epithelial tumor cells and CSCs under different culture conditions. See SI Appendix, Materials and Methods for detailed procedure.

qRT-PCR Analysis.

Total RNA was extracted using TRIzol reagent (Invitrogen), and the first-strand cDNA was generated by the High-Capacity cDNA Reverse Transcription kit (ABI) in a 20-μL reaction containing 1 μg of total RNA. A 2.5-μL aliquot of cDNA was amplified by Fast SYBR Green PCR Master Mix (Life Technologies) in each 20 μL reaction. PCR reactions were run on the ABI 7900 Fast Real-Time PCR system in the Ohio State University Comprehensive Cancer Center (OSUCCC) Nucleic Acid Core Facility. See SI Appendix, Materials and Methods for primer sequences.

"V体育官网入口" Immunoblotting.

Whole-cell lysates were prepared by boiling cell pellets for 10 min in SDS lysis buffer [2% (wt/vol) SDS, 10% (vol/vol) Glycerol, 62 mmol/L Tris⋅HCl, pH 6.8, and a complete miniprotease inhibitor mixture (Roche Applied Science)]. After protein quantification, equal amounts of proteins were loaded, separated on a polyacrylamide gel, and transferred to a nitrocellulose membrane. Protein bands were immunodetected with appropriate antibodies, e.g., goat anti-Pol η (Abcam), rabbit anti-Nanog (Cell Signaling Technology), mouse anti-Tubulin (Millipore), and mouse anti-Actin (Santa Cruz Technology).

miRNA Detection. (VSports最新版本)

For miRNA detection, a TaqMan MicroRNA Assay Kit (Applied Biosystems), including the following assays, was used: miR-20b (Assay ID: 00104) and miR-93 (Assay ID: 001090). All quantitative real-time PCR runs were carried out according to manufacturer’s instructions. RNU6B (Assay ID: 001093) and 18S rRNA (Applied Biosystems) were used for normalization. All PCR reactions were performed in triplicate.

Xenograft Tumor Growth.

Nonobese diabetic/severe combined immunodeficiency and Athymic nude (NCr-nu/nu) mice (6–8 wk, female, 20–25 g body weight) were obtained from National Cancer Institute (Frederick, MD). Animals were maintained in accordance with institutional policies, and all studies were performed with approval of the Institutional Animal Care and Use Committee at the Ohio State University. To generate xenografts, 5 × 10⁶ cells were mixed (1:1) with Matrigel (BD Biosciences) and injected s.c. into the flank of each mouse. Animals were treated with cisplatin i.p. twice (7 mg/kg; weekly) after xenografts reached 0.5 cm in diameter. Tumor growth was measured using calipers, and volumes were calculated based on the formula V = (a × b²)/2, in which a is the longest and b is the shortest diameter of the tumor. Xenograft cells were isolated after 2 d of the second treatment with the help of collagenase digestion and RBC lysis (eBioscience).

Detection of Cell Viability.

After 24 h of transfection with siRNA or shRNA, cells were reseeded and cultured for another 24 h in a 96-well plate at a density of 1,000 cells per well, then treated with cisplatin for 3 d. Cell viability was assessed by the MTT cell proliferation assay kit according to the manufacturer’s instruction (ATCC).

Detection of Apoptotic Cells.

The 2008-CSCs, C13-CSCs, and SKOV3-CSCs growing in ultralow attachment plates with CSC selective medium were transfected with either 100 nM siCtrl or siPOLH or miR-93 mimics with lipofectamine for 24 h and treated with cisplatin for 48 h. Cells were harvested and stained with Annexin V FITC assay kit (Cayman Chemical) and analyzed by FACS.

Statistical Analysis. (VSports在线直播)

Linear mixed effect models were used to take into account the observations from the same subject for the tumor growth studies. Tumor growth over time was compared among groups from those models. For the RT-PCR studies, the data were first normalized to the internal controls, and then ANOVA or linear mixed effects models were used for analysis for cell line studies and primary cell studies, respectively. All results were presented as mean ± SD, with a P value < 0.05 considered as statistically significant.

Additional laboratory reagents used are described in SI Appendix, Materials and Methods.