Cisplatin resistance driver claspin is a target for immunotherapy in urothelial carcinoma

Bladder cancer is a major and fatal urological disease. Cisplatin is a key drug for the treatment of bladder cancer, especially in muscle-invasive cases. In most cases of bladder cancer, cisplatin is effective; however, resistance to cisplatin has a significant negative impact on prognosis. Thus, a treatment strategy for cisplatin-resistant bladder cancer is essential to improve the prognosis. In this study, we established a cisplatin-resistant (CR) bladder cancer cell line using an urothelial carcinoma cell lines (UM-UC-3 and J82). We screened for potential targets in CR cells and found that claspin (CLSPN) was overexpressed. CLSPN mRNA knockdown revealed that CLSPN had a role in cisplatin resistance in CR cells. In our previous study, we identified human leukocyte antigen (HLA)-A*02:01-restricted CLSPN peptide by HLA ligandome analysis. Thus, we generated a CLSPN peptide-specific cytotoxic T lymphocyte clone that recognized CR cells at a higher level than wild-type UM-UC-3 cells. These findings indicate that CLSPN is a driver of cisplatin resistance and CLSPN peptide-specific immunotherapy may be effective for cisplatin-resistant cases.


Introduction
Bladder cancer is the 10th most common cancer worldwide, with 550,000 new cases and more than 200,000 deaths per year worldwide [1,2]. Approximately 70% of new bladder cancer cases are non-invasive, and while most of these cases can be controlled surgically, the recurrence rate is high. The remaining 30% of new cases are muscle-invasive bladder cancer, and many of them have distant metastasis. Patients with distant metastasis have an extremely poor prognosis, with a 5 year survival rate of approximately 5%. For many years, the mainstay of treatment for bladder cancer with local invasion or distant metastasis has been chemotherapy, mainly cisplatin (CDDP), and although the response rate is not low, many patients eventually become resistant to treatment and the prognosis for advanced bladder cancer remains poor [1,3]. CDDP-based chemotherapy in neoadjuvant and adjuvant settings improves the overall survival of patients with muscle-invasive bladder cancer; therefore, the development of treatment resistance, especially to CDDP, has a direct impact on prognosis. Unlike many other cancers, the 1 3 survival rate of bladder cancer has not improved in 30 years [4]. Therefore, it is essential to analyze the mechanism underlying resistance to CDDP to improve the outcome of current bladder cancer therapy.
Bladder cancer is an immunogenic malignancy and cancer immunotherapy using BCG has been approved for decades [5]. Immune checkpoint blockade has also been approved for bladder cancer [6][7][8][9]. However, these immunotherapies are based on the non-specific augmentation of the immune response; thus, analysis of antigen-specificity in bladder cancer remains essential for further innovation in immunotherapy.
In this study, we established a novel CDDP-resistant bladder cancer subline and identified the novel immunotherapy target claspin (CLSPN) as a driver of CDDP resistance.

Establishment of a UM-UC3 CDDP-resistant (CR) cell line
The CR cell line was derived from the original parental UM-UC-3 cell line and J82 cell line by continuous exposure to CDDP at stepwise increasing concentrations. Initially, exponentially growing cells were exposed to 0.5 μmol/L CDDP. These cells were maintained in CDDP-containing DMEM supplemented with 10% FBS and 1% penicillin/streptomycin, and passaged upon reaching 70-80% confluency. At this point, the CDDP concentration was increased by approximately 1.5-fold and the above process was repeated continuously over a period of approximately 6 months. The final concentration of CDDP was 5 μmol/L. To address the transient response to CDDP, the cells were treated with CDDP at several concentrations for quantitative real-time (qRT)-PCR and 1 μmol/L for an interferon gamma (IFNγ) enzyme-linked immunospot (ELISPOT) assay for 2 days.

Small interfering RNA (siRNA)-mediated knockdown
The cells were transfected with siRNA targeting CLSPN (Thermo Fisher Scientific, Waltham, MA) using the Lipofectamine RNAi MAX reagent (Thermo Fisher Scientific). Non-targeting siRNA (Thermo Fisher Scientific) was used as a negative control. CLSPN knockdown was confirmed by qRT-PCR.

Western blotting and qRT-PCR
Western blotting was performed as previously described [10]. Briefly, 5.0 × 10 5 WT UM-UC-3 or CR cells were lysed in 100 μL radio-immunoprecipitation assay buffer. An anti-CLSPN rabbit polyclonal antibody (Cell Signaling Technology, Danvers, MA) was used at a dilution of 1:1,000. An anti-MDR1 rabbit monoclonal antibody (Cell Signaling Technology) was used at a dilution of 1:1,000. An anti-βactin mouse monoclonal antibody (Sigma-Aldrich) was used as a loading control at a dilution of 1:1,000. Anti-mouse IgG and anti-rabbit IgG secondary antibodies (KPL, Gaithersburg, MD) were used at a dilution of 1:2,000. The membrane was visualized with Chemiluminescent HRP Substrate (Millipore, Billerica, MA), and images were taken by an Odyssey Fc imaging System (LI-COR, Lincoln, NE). qRT-PCR was performed as previously described [13].

Cap analysis of gene expression (CAGE) analysis
CAGE library preparation, sequencing, mapping, gene expression, motif discovery analysis, and Gene Ontology enrichment analysis were performed by DNAFORM (Yokohama, Japan) as previously described [12].
CAGE tag count data were clustered by CAGEr [14] using the Paraclu algorithm [15] with default parameters. Differentially expressed genes (CPM > 1 and fold-change < 0.25, or CPM > 1 and fold-change > 4) were detected with the DESeq2 package (version 1.20.0) and used for Gene Ontology enrichment analysis by the clusterProfiler package [16]. Raw and processed CAGE data have been deposited in the NCBI GEO database (GSE217413).

Lactate dehydrogenase (LDH) cytotoxicity assay
To evaluate the function of the CTL clone, an LDH release assay was performed using a CyQUANT LDH Cytotoxicity Assay Kit (Thermo Fisher Scientific). WT UM-UC-3 and CR cells were plated in 96-well plates at a density of 2.0 × 10 4 cells/well in DMEM containing 10% FBS. After overnight culture, the medium was replaced with control medium (serum-free AIM-V) and the CLSPN-specific CTL clone yc3 was added at an effector-target ratio of 9:1, 3:1, or 1:1 and co-cultured for 8 h. The release of LDH was then measured according to the manufacturer's instructions.

Statistical analysis
Student's t test was used to compare two groups; p < 0.05 was considered to be statistically significant. Differences in the estimated frequency of CSCs/CICs were analyzed by the chi-square test.

Establishment of CR cells
In this study, we aimed to identify potential immunological targets that are expressed in CDDP-resistant subpopulations of urothelial carcinoma cells. We cultured the UM-UC-3 urothelial carcinoma cell line in CDDP-containing medium, and we increased the concentration of CDDP gradually over a period of approximately 6 months to 5 μM (Fig. 1A) to establish a CR cell line. CR cells showed resistance to CDDP compared with WT cells (Fig. 1B). A previous report indicated that chemoresistance is related to CSCs [20]; thus, we examined whether CSCs are enriched in CR cells. An ALDEFLUOR assay revealed that the aldehyde dehydrogenase-high (ALDH high ) population was increased, indicating that CSCs were enriched in CR cells (Fig. 1C). A sphere-formation assay revealed that the frequency of sphere-forming CSCs was increased in CR cells compared with WT cells (Fig. 1D).

CLSPN expression is upregulated in CR cells
To identify a potential immunological target, we screened the genes that were overexpressed in CR cells compared with WT cells. We performed CAGE analysis and found that 2284 genes were upregulated in CR cells (Supplemental Table S1). CDDP resistance responsive gene MDR1/ ABCB1 was 59.5-fold increased and the protein expression was confirmed by a Western blot (Fig. 2D). A previous study revealed that MDR1/ABCB1 is related to CDDP [21] and the results in this study was consistent with previous study. To narrow down the candidate genes, we referenced HLA ligandome analysis and a public database of gene expression (GTExPortal: https:// www. gtexp ortal. org). Previously, we performed HLA ligandome analysis using UM-UC-3 cells to identify HLA-A2-bound antigenic peptides [20]. Thus, we screened genes that were upregulated in CR cells, that encoded HLA-A2-bound antigenic peptides identified by HLA ligandome analysis, and whose expression in normal organs was less than 2 transcripts per million. Finally, we found that the CLSPN gene was overexpressed in CR cells and the candidate peptide sequence encoded by CLSPN was SLLNQPKAV ( Fig. 2A). CLSPN mRNA expression in normal organs was low or undetectable according to qRT-PCR (Fig. 2B). CLSPN mRNA expression was higher in CR cells than in WT cells (Fig. 2C). Western blot analysis revealed that CLSPN protein was expressed in CR and WT cells (Fig. 2D). To generalize the results, we established another CDDP-resistant line using an urothelial carcinoma cell J82. J82 CR cells also showed resistance to CDDP compared with that in J82 WT cells (Supplemental Figure S1A). qRT-PCR analysis and Western blot analysis revealed that CLSPN mRNA and protein were overexpressed in J82 CR cells compared with that in J82 WT cells.

CLSPN is a driver of CDDP resistance
CLSPN was expressed at higher levels in CR cells than in WT cells, and a previous study indicated that CLSPN is related to CSCs [22]. CLSPN is expressed in the G1/S phase of the cell cycle and has a role in the DNA checkpoint response [23]. Thus, we hypothesized that CLSPN might have a role in CDDP resistance. To assess the functions of CLSPN, we knocked down CLSPN mRNA using siRNA; CLSPN mRNA knockdown was confirmed by qRT-PCR (data not shown). siRNA knockdown of CLSPN mRNA increased the sensitivity of CR cells to CDDP (Fig. 3A). CDDP induced the expression of CLSPN mRNA in CR and WT cells (Fig. 3B).

Establishment and characterization of a CLSPN peptide-specific CTL clone
To generate CLSPN peptide-specific CTLs, we stimulated HLA-A*02:01-positive PBMCs with HLA-A*02:01-positive aAPCs pulsed with CLSPN peptide (Fig. 4A). After stimulation for 5 times, the cells were evaluated by an IFNγ Fig. 1 Establishment and characterization of a CR subline. A Schematic summary of the culture strategy for establishing a CR cell line. UM-UC-3 cells were cultured in medium containing CDDP at 0.5 μmol/L. The concentration of CDDP was increased over a period of approximately 6 months. B Sensitivity of WT and CR cells to CDDP. The cells were cultured in medium containing CDDP at several concentrations for 48 h. Cell viability was assessed by a WST-8 assay. Each value is the mean ± standard deviation (SD). Statistical analysis was performed using Student's t-test. C ALDEFLUOR assay. WT UM-UC-3 and CR cells were stained with BODIPY®-aminoacetaldehyde and analyzed by FACS. The ALDH high -positive gate was defined by a sample treated with the ALDH1 inhibitor DEAB. Numerical data indicate ALDH high rates. D Sphere-formation assay. The sphere-forming ability of WT UM-UC-3 and CR cells was assessed. The cells were seeded at 1.0 × 10 3 , 10 2 , 10 1 , and 10 0 cells/ well in an ultra-low attachment plate and cultured for 1 week. Sphereforming wells were counted. Stem cell frequency was analyzed using the ELDA web site. CI, confidence interval. Differences of the estimated frequencies of CSCs/CICs were analyzed by a chi-square test 1 3 ELISPOT assay. The number of IFNγ spots was increased in CLSPN peptide-pulsed T2 cells compared with peptide nonpulsed T2 cells (Fig. 4B). Flow cytometry revealed the presence of CLSPN tetramer-positive cells (Fig. 4B). To analyze the CTLs at the clonal level, we established the CLSPN-specific CTL clone yc3 by CLSPN tetramer-positive single cell sorting (Fig. 4B). Seven wells showed cell growth and we tested the cells by staining with CLSPN-HLA-A2 tetramer, finding that the yc3 clone was specific for CLSPN tetramer. Thus, we performed further analyses using the yc3 clone.
An IFNγ ELISPOT assay revealed that the yc3 clone recognized WT and CR cells (Fig. 4C). Although the yc3 clone recognized CR cells more strongly than WT cells, the difference did not reach statistical significance (p = 0.101) (Fig. 4C). Because CLSPN expression was induced by CDDP (Fig. 3B), we examined whether CDDP stimulation enhanced the reactivity of the yc3 clone. The addition of CDDP significantly increased the number of IFNγ spots compared with CDDP non-treated WT cells (Fig. 4D). To confirm the specificity of the yc3 clone, we knocked down CLSPN mRNA with siRNA and assessed the reactivity of the yc3 clone. Knockdown of CLSPN mRNA significantly decreased the number of IFNγ spots compared with WT cells (Fig. 4E). Finally, to confirm the cytotoxic activity of the yc3 clone, we examined cytotoxicity using an LDH release assay. The LDH release assay revealed that the yc3 clone showed higher cytotoxicity to CLSPN peptide-pulsed T2 cells than to control peptidepulsed T2 cells, peptide non-pulsed T2 cells, and negative control K562 cells (Fig. 5A). Furthermore, the yc3 clone showed higher cytotoxicity to CR cells than to WT cells (Fig. 5B). These results indicate that CLSPN peptide is endogenously expressed by CR cells, and the yc3 clone recognizes and kills CR cells in a CLSPN peptidedependent manner.

Discussion
In this study, we established a novel CDDP-resistant UM-UC-3 urothelial carcinoma cell line that showed a CSClike phenotype. CSCs are resistant to chemotherapy and radiotherapy by several mechanisms [24,25]. In contrast, CR cells showed a CSC-enriched phenotype with a high rate of ALDH high cells and high sphere-forming ability. In a previous study, a chemotherapy-resistant lung cancer subline showed a CSC-like phenotype with a high rate of ALDH high cells [20]. Thus, treatment resistance and CSCs might be an overlapping population in heterogenous cancer cells. Indeed, we found that CLSPN is overexpressed in UM-UC-3 CR cells, and our previous study revealed that CLSPN peptide is expressed in UM-UC-3 H10 cells, a CSC-like clone derived from UM-UC-3 cells [24]. CSCs are resistant to treatment by the expression of anti-apoptosis proteins or transporters [26]. CR cells also showed higher expression levels of the anti-apoptotic protein BIRC5 and transporter proteins including ABCB1, ABCB9, ABCB10, and ABCC1 than WT cells (Supplemental Table S1). A previous study revealed that CDDP-resistant or gemcitabine-resistant bladder cancer lines show high MDR1/ABCB1 expression and also crossresistance to other chemotherapeutic reagents [27]. Thus, CR cells were resistant to CDDP and might be multi-drug resistant due to high ABCB1 expression; however, they were sensitive to CTLs. Thus, CR cells should be sensitive to granzyme B and perforin, which are the molecules used by CTLs for cytotoxicity, suggesting that the molecular mechanisms related to resistance for CDDP and CTLs are different.
We did not address the relationship between antiapoptotic or transporter proteins and resistance to CDDP. However, CLSPN knockdown increased sensitivity to CDDP; thus, CLSPN is a driver of CDDP resistance. A previous study revealed that CLSPN is preferentially expressed in CSCs of gastric cancer [22]. Thus, CLSPN might be related to CDDP resistance and the CSC phenotype. However, CLSPN knockdown did not reduce the ALDH high population in our study (data not shown); thus, CLSPN might not have role in the maintenance of ALDH high cells.
CLSPN is reported to mediate the phosphorylation and activation of CHK1 by ATR during DNA replicative stress in the S phase of the cell cycle [23]. In addition, p38 stressactivated kinase phosphorylates CLSPN and protects cells from DNA damage during the S phase [29,30]. However, Fig. 4 Induction and characterization of a CLSPN peptide-specific CTL clone. A A schematic summary of CTL induction. PBMCs were obtained from an HLA-A*02:01-positive donor and CD8 + T cells were then isolated. CD8 + T cells were stimulated with HLA-A*02:01 + aAPCs (K562 cells transduced with CD80, CD83, and HLA-A*02:01 cDNAs) pulsed with CLSPN peptide (SLLNQP-KAV) at day 1, 8, 15, 22, and 29. At day 34 the CD8 + T cells were assessed by an ELISPOT assay and FACS. B CTL cloning strategy. Stimulated CD8 + T cells were analyzed by an IFNγ ELISPOT assay and FACS with CLSPN tetramer staining. For the IFNγ ELISPOT assay, T2 cells pulsed with CLSPN peptide and non-pulsed cells were used as targets. For FACS analysis, HIV-tetramer was used as a negative control. CLSPN tetramer-positive cells were single cell sorted and cultured. Cells grew in seven wells and one clone (yc3) showed reactivity to CLSPN peptide according to an IFNγ ELISPOT assay and FACS. C CLSPN peptide-specific CTL clone recognizes CR cells. The CLSPN peptide-specific CTL yc3 clone was evaluated for reactivity to WT UM-UC-3 and CR cells by an IFNγ ELISPOT assay. CLSPN peptide-pulsed T2 cells were used as a positive control and peptide non-pulsed T2 cells were used as a negative control. Each value is the mean ± SD. D CLSPN peptide-specific CTL clone recognizes CDDP-treated WT cells. The CLSPN peptide-specific CTL yc3 clone was evaluated for reactivity to WT UM-UC-3 cells and CDDP-treated WT cells by an IFNγ ELISPOT assay. WT cells were cultured in CDDP-containing medium (5 μM) for 48 h, and then used for assay. T2 cells were used as a negative control. Each value is the mean ± SD. E CLSPN knockdown by siRNA decreases the CTL response. WT UM-UC-3 cells were transfected with CLSPN siRNA, and 48 h later, the cells were evaluated by an IFNγ ELISPOT assay with the CLSPN peptide-specific CTL yc3 clone. T2 cells were used as a negative control. Each value is the mean ± SD ◂ Fig. 5 Cytotoxicity assay using the CLSPN peptide-specific CTL clone. A Cytotoxicity of the yc3 clone for CLSPN peptide-pulsed targets. The cytotoxicity of the yc3 clone was assessed by an LDH release assay. Target cells were co-cultured with the yc3 clone for 8 h at different effector/target ratios. HIV peptide-pulsed T2 cells, peptide non-pulsed T2 cells, and K562 cells were used as negative con-trols. Each value is the mean ± SD. B Cytotoxicity of the yc3 clone to WT UM-UC-3 and CR cells. The cytotoxicity of the yc3 clone was assessed by an LDH release assay. WT UM-UC-3 and CR cells were used as targets. Target cells were co-cultured with the yc3 clone for 8 h at different effector/target ratios. Each value is the mean ± SD CLSPN was shown to be related to DNA replication stress and to protect cancer cells in an ATR-CHK1 checkpointindependent manner [28]. Higher CLSPN mRNA expression is related to a poorer prognosis in lung cancer [28]. Other studies have revealed that CLSPN is overexpressed in gastric cancer, renal cell carcinoma, prostate cancer, bladder cancer, and low-grade glioma [22,[29][30][31][32]. CLSPN protein expression is highest in the S phase, and in the G2/M phase, CLSPN is phosphorylated by PLK1 to induce protein degradation by the ubiquitin-proteasome system [23,33]. Regarding this protein expression regulatory system, CLSPN protein expression is very low in normal tissues; however, it can be detected at higher levels in cancer tissues by immunohistochemical staining [22,[29][30][31]. These findings suggest that CLSPN protein might be relatively more stable in cancer cells compared with normal cells. Considering its common molecular function in cell biology, CLSPN might be expressed in a wide variety of malignancies and CLSPN-targeting immunotherapy might be applicable for several malignancies.
Antigenic peptides are produced by proteasomal antigen protein processing [34][35][36]. Fusion of the ubiquitin gene to an antigen gene enhances the CTL response [37]. Thus, CLSPN protein is a reasonable target for immunotherapy because it is degraded by the ubiquitin-proteasome system [23,33]. In this study, we demonstrated that CLSPN expression was induced by 2 day stimulation with CDDP. This quick response will enable us to design further potential combination immunotherapies of CDDP and CLSPN-targeting immunotherapy including peptide vaccination, mRNA vaccination, and engineered T cell therapy. Furthermore, CDDP stimulation induces the expression of the immune checkpoint molecule PD-L1 in oral squamous cell carcinoma cells [37]. Thus, combination immunotherapy with CDDP might require PD1/PD-L1 blockade.
Taken together, these findings suggest that CLSPN is a driver of CDDP resistance in UM-UC-3 bladder cancer cells, and CLSPN can be targeted by CLSPN peptide-specific CTLs. This study provides support for the use of immunotherapy targeting CDDP-resistant cases and combination therapy of CDDP and immunotherapy.