Abstract
Background/Aim: Clear cell carcinoma is a prevalent histological type of ovarian cancer in East Asia, particularly in Japan, known for its resistance to chemotherapeutic agents and poor prognosis. ARID1A gene mutations, commonly found in ovarian clear cell carcinoma (OCCC), contribute to its pathogenesis. Recent data revealed that the ARID1A mutation is related to better outcomes of cancer immunotherapy. Thus, this study aimed to investigate the immunotherapy treatment susceptibility of OCCC bearing ARID1A mutations. Materials and Methods: Expression of ARID1A was analyzed using western blotting in ovarian cancer cell lines. OCCC cell lines JHOC-9 and RMG-V were engineered to overexpress NY-ESO-1, HLA-A*02:01, and ARID1A. Sensitivity to chemotherapy and T cell receptor-transduced T (TCR-T) cells specific for NY-ESO-1 was assessed in ARID1A-restored cells compared to ARID1A-deficient wild-type cells. Results: JHOC-9 cells and RMG-V cells showed no expression of ARID1A protein. Overexpression of ARID1A in JHOC-9 and RMG-V cells did not impact sensitivity to gemcitabine. While ARID1A overexpression decreased sensitivity to cisplatin in RMG-V cells, it had no such effect in JHOC-9 cells. ARID1A overexpression reduced the reactivity of NY-ESO-1-specific TCR-T cells, as observed by the IFNγ ESLIPOT assay. Conclusion: Cancer immunotherapy is an effective approach to target ARID1A-deficient clear cell carcinoma of the ovary.
Epithelial ovarian cancer (EOC) is a major gynecological cancer, with 19,880 new cases diagnosed and 12,810 cases resulting in death annually in the USA (1). There are four major histological subtypes of EOC, including serous carcinoma, endometrioid carcinoma, clear cell carcinoma, and mucinous carcinoma. Although ovarian clear cell carcinoma (OCCC) is not common in Western countries, it is characteristically prevalent in East Asia, including Japan (2).
AT-rich interactive domain-containing protein 1A (ARID1A) is one of the proteins that form the SWItch/Sucrose Non-Fermentable (SWI/SNF) complex and is involved in the epigenetic regulation of gene expression (3). ARID1A mutations have been reported in many cancer types, including ovarian clear cell carcinomas (OCCC) (3-5). In OCCC, ARID1A gene mutations are present in more than 50% of cases and play a role in the re-activation of hTERT and metabolic changes that might be responsible for carcinogenesis in OCCC (3). ARID1A mutations frequently coexist with PIK3CA mutations in OCCC (6). Given the high rate of mutation and its functional role in carcinogenesis, ARID1A is considered a reasonable target for OCCC. Currently, several clinical trials targeting mutant ARID1A are underway; however, none of them has been approved at this moment (7).
Recently, cancer immunotherapy has been suggested as a candidate for ARID1A-mutant OCCC. ARID1A mutation, combined with high expression of CXCL13, serves as a biomarker for predicting immune checkpoint blockade (8). In OCCC, ARID1A mutation is related to high tumor mutation burden (TMB), suggesting the possible immunogenicity of ARID1A-mutant OCCC (9). These findings suggest that cancer immunotherapy might be a reasonable choice for ARID1A-mutant OCCC. However, the sensitivity of ARID1A-mutant OCCC to T cells is still elusive. In this study, we analyzed the sensitivity of ARID1A mutant OCCC to T cells using NY-ESO-1-specific TCR-T cells as a model effector T cell.
Materials and Methods
Cell lines, plasmids, and overexpression. Human ovary clear cell carcinoma cell lines JHOC-5, JHOC-9, RMG-V, and RMG-I were purchased from RIKEN CELL BANK (Tsukuba, Japan), and human ovary serous carcinoma cell lines AMOC-2, HUOA, OVCAR3, and HTBOA were purchased from ATCC (Manassas, VA, USA). JHOC-5, JHOC-9, RMG-V, and RMG-I were cultured in DMEM/F12 (Nacalai Tesque, Kyoto, Japan) supplemented with 10% fetal bovine serum (FBS) (Sigma-Aldrich, St. Louis, MO, USA). AMOC-2, HUOA, OVCAR3, and HTBOA were cultured in RPMI-1640 (Sigma-Aldrich) supplemented with 10% fetal bovine serum (FBS).
The NY-ESO-1 encoding plasmid pMXs-puro/NY-ESO-1 was used for NY-ESO-1 overexpression using PLAT-A Amphotropic packaging cells (Cosmo Bio inc., Tokyo, Japan), as described previously (10). NY-ESO-1 overexpressed cells were selected using DMEM/F12 supplemented with puromycin (1 μg/ml) (Sigma-Aldrich). The HLA-A02:01 encoding plasmid pMXs/HLA-A*02:01 was used for HLA-A*02:01 overexpression. After infection, HLA-A*02:01 overexpressed cells were analyzed by FACS Aria II (BD, San Jose, CA, USA), and HLA-A*02:01-positive cells were isolated. The pcDNA6-ARID1A plasmid was a gift from Dr. Ie-Ming Shih (Addgene plasmid #39311; http://n2t.net/addgene:39311; RRID:Addgene_39311) (11). ARID1A cDNA fragments were amplified by PCR using the following primers: 5′-CGGATCTAGCTAGTTAATTAAGCCACCATGGCCGCGCA GGTCG-3′ and 5′-CGACCGGCGCTCAGCTGGAATTCCGCT TCTGGAATGTGG-3′ for the 3655 bp 5′ ARID1A cDNA fragment, and 5′-CCACATTCCAGAAGCGGAATTCCA-3′ and 5′-TGGCGGCCGCTCGAGTCAATGGTGATGGTGATGAT-3′ for the 3380 bp 3′ ARID1A cDNA fragment. For PCR, PrimeSTAR GXL DNA polymerase (Takara Bio. Inc., Kusatsu, Japan) was used. The 5′ fragment was cloned into pMXs-neo by PacI and EcoRI sites, and the 3′ fragment was cloned into pMXs-neo by EcoRI and XhoI sites. After confirming the sequence, the pMXs-neo/ARID1A plasmid was transduced into PLAT-A cells by PEI MAX™ (Cosmo Bio inc.), and then the supernatant was transduced into JHOC-9 and RMG-V cells. The cells were selected using media containing 1mg/mL G418 (Thermo Fisher Scientific, Carlsbad, CA, USA).
Western blotting and flow cytometry. Western blotting was performed as described previously (12). Flow cytometry was performed as described previously (13). For the detection of HLA-A*02:01 by flow cytometry, a mouse monoclonal anti-HLA-A2 antibody (clone: BB7.2) was used. The hybridoma was purchased from ATCC, and the culture supernatant was used as a working solution. For the detection of NY-ESO-1 by western blotting, anti-NY-ESO-1 mouse monoclonal antibody (clone: E978) (Santa Cruz Biotechnology, Inc., Dallas, TX, USA) was used at a 200-fold dilution. For the detection of ARID1A by western blotting, anti-ARID1A rabbit monoclonal antibody (clone: EPR13501) (Abcam plc., Cambridge, UK) was used at a 1000-fold dilution.
Sensitivity to chemotherapeutic reagents and ELISPOT assay. Assessment of sensitivity to chemotherapeutic reagents using WST-8 (Dojindo, Kumamoto, Japan), was performed as described previously (14). Cisplatin (CDDP) and gemcitabine (GEM) were obtained from FUJIFILM Wako Chemicals (Osaka, Japan). Interferon-γ (IFNγ) ELISPOT assay was performed as described previously (14).
Statistical analysis. Student’s t-test was used to compare control group and experimental group in qRT-PCR analysis and IFNγ ELISPOT assay. p<0.05 was considered to indicate a significant difference.
Results
Expression of ARID1A protein in ovarian cancer cells. ARID1A protein deficiency, resulting from ARID1A gene mutation, is prevalent in clear cell carcinoma of the ovary. We analyzed the protein expression of ARID1A in ovarian cancer cell lines using western blotting with an anti-ARID1A antibody. Serous carcinoma cells, including AMOC-2, HUOA, OVCAR3, and HTBOA cells, expressed ARID1A. In clear cell carcinomas, JHOC-5 cells expressed ARID1A, while JHOC-9 and RMG-V cells did not express ARID1A, and RMG-I cells expressed ARID1A at a low level (Figure 1).
Expressions of ARID1A in ovarian cancer cells. The expression of ARID1A protein was analyzed by western blot. Human clear cell carcinoma cell lines JHOC-5, JHOC-9, RMG-I and RMG-V, human serous carcinoma cell lines AMOC-2, HUOA, OVCAR3 and HTBOA were used. β-actin was used as an internal control.
Overexpression of ARID1A, NY-ESO-1 and HLA-A*02:01. To address ARID1A deficiency in immune response, we transfected the ARID1A gene into JHOC-9 and RMG-V cells, establishing ARID1A-overexpressed cells. In a previous study, we generated HLA-A*02:01-restricted NY-ESO-1 peptide-specific TCR-T cells (13). We used these TCR-T cells as a model for cytotoxic T lymphocytes (CTLs). Since JHOC-9 and RMG-V were negative for both NY-ESO-1 and HLA-A*02:01, we overexpressed NY-ESO-1 and HLA-A*02:01, and their expression was assessed by western blotting and flow cytometry (Figure 2).
Overexpression of ARID1A, NY-ESO-1 and HLA-A*02:01. ARID1A, NY-ESO-1, and HLA-A*02:01 genes were overexpressed in JHOC-9 cells and RMG-V cells. NY-ESO-1 protein and ARID1A protein expression were evaluated by western blotting (A). HLA-A*02:01 expression was addressed by flow cytometry (B).
Restoration of ARID1A in ARID1A-deficient clear cell carcinoma did not alter the sensitivities to gemcitabine. In a previous study, ARID1A gene knockout increased the sensitivity of the clear cell carcinoma cell line RMG-I cells to gemcitabine (15). We addressed the sensitivities of ARID1A-restored clear cell carcinoma cells to chemotherapeutic reagents. Overexpression of ARID1A did not change sensitivity to cisplatin in JHOC-9 cells (Figure 3A). Additionally, overexpression of ARID1A did not alter the sensitivity to gemcitabine in JHOC-9 cells and RMG-V cells (Figure 3A and B). However, overexpression of ARID1A decreased the sensitivity to cisplatin in RMG-V cells (Figure 3A). The similar results could be obtained in more than 3 times independent experiments. Thus, restoration of ARID1A in ARID1A-deficient cells did not alter sensitivity to CDDP, and sensitivity to GEM depended on the cell lines.
Sensitivity to chemotherapeutic reagents. NY-ESO-1 and ARID1A overexpressing JHOC-9 cells and RMG-V cells were addressed for sensitivities to cisplatin (CDDP) (A) and gemcitabine (GEM) (B). NY-ESO-1 overexpressing cells were used as a negative control. The cells were incubated with CDDP or GEM at several concentrations for 2 days, then the viability was assessed by WST-8. Each value is the mean±standard deviation.
Overexpression of ARID1A in ARID1A-deficient clear cell carcinoma decreased the sensitivity to NY-ESO-1-specific TCR-T cells. To assess sensitivities to NY-ESO-1-specific TCR-T cells, an interferon-gamma (IFNγ) ELISpot assay was performed. The TCR-T cells exhibited reactivity to NY-ESO-1 peptide-pulsed T2 cells and did not react to peptide-unpulsed T2 cells, indicating specificity to NY-ESO-1 (Figure 4). TCR-T cells also showed reactivity to NY-ESO-1 and HLA-A*02:01-overexpressed JHOC-9 cells and RMG-V cells. Notably, ARID1A overexpression reduced the reactivity of TCR-T cells in both JHOC-9 cells and RMG-V cells, respectively. The similar results could be obtained in 3 different experiments. The results suggest that ARID1A overexpression in ARID1A-deficient cells reduces responsiveness to TCR-T cells.
IFNγ ELISpot assay. IFNγ ELISpot assay was performed using NY-ESO-1-specific TCR-T cells. In the assay, JHOC-9 cells and RMG-V cells were used as target cells. HLA-A*02:01 and NY-ESO-1 overexpressed cells and HLA-A*02:01, NY-ESO-1, and ARID1A overexpressed cells were also used as target cells. T2 cell was used as a negative control. NY-ESO-1 peptide pulsed T2 cell was used as a positive control. Each value is the mean±standard deviation.
Discussion
The prevalence of ARID1A mutation in ovarian clear cell carcinoma (OCCC) highlights its distinctive nature, with a detection rate of 46% in OCCC cases compared to the absence of ARID1A mutations in high-grade serous carcinoma, another major ovarian malignancy (4). A recent study in Japanese OCCC cases reported an even higher prevalence of ARID1A mutation at 69.4% (16). Notably, ARID1A mutations often coexist with PIK3CA mutations (6), emphasizing the essential role of PI3K/AKT signaling in the carcinogenesis of epithelial cells with ARID1A mutation. Additionally, frequent PTEN mutations, an inhibitor of PI3K, are observed in ARID1A-mutant OCCC cases (16). The activation of PI3K/AKT signaling has implications for the tumor microenvironment, promoting immune suppressive factors, such as PD-L1 and LAG-3 (17, 18).
While ARID1A mutations have been linked to interactions with mismatch repair (MMR) protein MSH2 and sensitivity to immune checkpoint blockade (19), direct evidence of cancer cells carrying ARID1A mutation being sensitive to cytotoxic T lymphocytes (CTLs) has been lacking. In this study, we investigated the sensitivity of ARID1A-deficient clear cell carcinoma cells in the ovary. The restoration of ARID1A in ARID1A-deficient OCCC cells resulted in reduced reactivity to NY-ESO-1-specific TCR-T cells (Figure 4). These data represent first evidence demonstrating a relationship between ARID1A expression and sensitivity to T cells. Currently, the exact mechanisms remain unknown; however, ARID1A deficiency may be associated with antigen processing or presentation, as the restoration of ARID1A did not alter the expression of both HLA-A*02:01 and NY-ESO-1.
Previous studies have highlighted the impact of ARID1A on the expression of tumor suppressor genes CDKN1A and SMAD3, as well as mitosis-related genes, suggesting potential avenues for further mechanistic analysis (11, 20). Nevertheless, our study emphasizes the need for additional research to elucidate the specific mechanisms underlying sensitivity to CTLs in ARID1A-deficient cells.
In the context of immunotherapy and ARID1A mutation, previous findings have indicated that ARID1A mutations can impair interferon (IFN) signaling and downstream gene expression. Low expression of the IFN-derived gene CXCL10 in ARID1A knockout cells results in reduced CTL infiltration, leading to less efficacy in immunotherapy models (21). ARID1A mutation has also been associated with better outcomes in immune checkpoint blockade (22). Our results align with these observations, suggesting that cancer cells carrying ARID1A mutations may be susceptible to CTLs. However, a more comprehensive collection of clinical data on immunotherapy in ARID1A mutant cancers is essential to draw definitive conclusions regarding the relationship between ARID1A mutation and immunotherapy outcomes.
In summary, our findings demonstrate that the restoration of ARID1A in ARID1A-deficient OCCC cells diminishes CTL reactivity. This highlights the complex interplay between ARID1A mutation, immune response, and potential implications for immunotherapeutic strategies in ovarian clear cell carcinoma.
Acknowledgements
The Authors would like to thank Ms. Junko Yanagawa for her technical assistance.
Footnotes
Conflicts of Interest
The Authors declare that they have no competing interests.
Authors’ Contributions
RT, YH, TS, and T. Torigoe contributed to the conception and design of the study. RT, AM, YM and T. Kubo acquired and analyzed the data. RT, YH, T. Kubo, T. Kanaseki, T. Tsukahara, and T. Torigoe interpreted the data. RT, YH, TS and T. Torigoe wrote the manuscript. TS and T. Torigoe approved the final version of the manuscript.
Funding
This study was supported by KAKENHI grants from the Japan Society for the Promotion of Science to Y. Hirohashi (23H02696). This work was supported by grants from the Japan Agency for Medical Research and Development Project for the Promotion of Cancer Research and Therapeutic Evolution to Y. Hirohashi (23ama221317h0002).
- Received February 15, 2024.
- Revision received March 31, 2024.
- Accepted April 10, 2024.
- Copyright © 2024 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
