Clinical Correlation of Function and TCR vβ Diversity of MAGE-C2–Specific CD8⁺ T Cell Response in Esophageal Cancer Free

Elimination of cancer in humans depends on T cell–mediated immunotherapy that recognizes tumor Ags by utilizing their TCRs (1). CD8⁺ T cell–mediated adaptive cellular immunity plays important roles in antitumor immunity (2). In recent years, with the rapid development of medical research, genetically engineered T cell therapy, which mainly includes TCR T cells and chimeric Ag receptor T cells, was introduced that makes it possible to produce Ag-specific T cells. Increasing evidence has shed considerable light on the positive correlation between the number of tumor Ag-specific T cells and the survival of patients with distant melanoma metastasis (3). This has opened new possibilities, achieved dramatic tumor regression, and displayed clinical success in the treatment of viral infections (4) and tumors (5). Numerous preclinical studies showed that TCR-engineered effector T cells have the ability to recognize tumor-specific epitopes presented by MHC molecules on the tumor cell surface (6), which have been explored for more than two decades, and could mediate tumor lysis and eradication. TCR-based T cell therapy in which a patient’s own T cells are gene modified with a TCR has broadened the clinical applicability of Ag-specific T cells and demonstrated significant clinical responses in cancer patients, such as for metastatic melanoma (7) and multiple myeloma (8). Westdorp et al. (9) demonstrated that immunotherapy with blood-derived dendritic cell (DC) vaccination was feasible and induced functional Ag-specific T cells in most patients, and the presence of functional Ag-specific T cells was correlated with a beneficial clinical outcome.

It is well known that one of the most substantial impediments in the treatment of tumors is the identification of tumor Ags. Cancer germline (testis) Ags (CTAs), first discovered in 1991 (10), are selectively expressed in malignant cells, but not in normal tissues, except in human testicular germ cells (11). Due to their unique immunogenicity, CTAs are considered as ideal targets for cancer immunotherapy. Melanoma-associated Ags (MAGEs) such as MAGE-A3 (12) and MAGE-A4 (13) are a type of CTA. Rosenberg and colleagues (7) demonstrated that normal autologous T lymphocytes transduced with anti–TAA (tumor-associated Ag)-TCR genes ex vivo and reinfused into cancer patients can mediate the durable regression of tumors. The clinical application of TCRs directed against MART-1 (14) and NY-ESO-1 (15) Ags to treat various tumor types proved feasible in a substantial number of patients. Moreover, a MART-1–reactive, vβ22-expressing clone from one typical metastatic melanoma patient’s tumor-infiltrating lymphocytes (TILs) comprised from 3.9 to 25.5% of his circulating CD8⁺ lymphocytes after 403 d of treatment (5).

Despite improvements in treatment, esophageal cancer (ESCA) is still an important cause of cancer-related deaths, and the 5-y overall survival rate of patients is still very low (16). Therefore, search for immunotherapy targets is of great importance. Our previous studies have demonstrated that MAGE-C2 mRNA and protein expression levels were respectively identified in 53% of patients with non–small cell lung cancer (17) and 65% of patients with esophageal squamous cell carcinoma (ESCC) (18). MAGE-C2 can induce Ag-specific T cell responses in patients with melanoma (19). MAGE-C2 Ag is recognized by the MAGE-C2 TCR on the T lymphocyte surface, which is composed of TCR α- and β-chains (20, 21). Until now, the detailed and exclusive function of MAGE-C2 in ESCA has not been explored and remains unclear. In the current study, the relationship among MAGE-C2–specific CD8⁺ T cell frequency, their corresponding TCR β-chain V region (TCR vβ) subfamily distribution, and the overall survival rate of patients with ESCA is elucidated. Overall, this study provides an explanation for the frequency of functional MAGE-C2–specific CD8⁺ T cells as an independent prognostic factor in ESCA patients. Moreover, ESCA patients with high expression of TCR vβ16 had a good prognosis, hinting of the possibility of using primary T lymphocytes transduced with a TCR vβ16–expressing clone and reinfused into patients to mediate the durable regression of tumors.

Patients were recruited from the Department of Thoracic Surgery, The First Affiliated Hospital of Zhengzhou University. PBMCs were obtained from 73 HLA-A2⁺ ESCA patients expressing the MAGE-C2 Ag. The clinicopathological parameters of 73 ESCA patients are listed in Supplemental Table II. None of the patients had undergone chemotherapy, radiotherapy, or any other cancer treatment therapy prior to their participation in the study. The protocols used in this study were approved by the Zhengzhou University Medical Center Institutional Review Board. All participants provided written informed consent in accordance with the Declaration of Helsinki. Clinicopathological parameters were analyzed according to sex, stage, differentiation, and lymph node metastasis. Patients were staged according to the tumor–node–metastasis staging system of the American Joint Committee on Cancer (AJCC, 6th edition).

Solid tumor tissues were cut into small pieces and incubated with 300 μg/ml collagenase (Roche, Indianapolis, IN) and 50 μg/ml DNase I (Sigma-Aldrich, St Louis, MO) for 2 h at 37°C. Subsequently, the samples were mechanically disaggregated, and TILs were separated using centrifugation at 2500 rpm for 25 min on Ficoll-Paque Plus (Sigma-Aldrich). PBMCs were isolated from heparinized blood samples using Ficoll-Paque Plus density centrifugation, as previously described (22).

MAGE-C2–specific CD8⁺ T cells were selectively expanded from PBMCs and TILs by stimulation with MAGE-C2_336–344 (ALKDVEERV/HLA-A2) peptide-loaded DCs generated from CD14-selected monocytes (using CD14⁺ beads and manual MACS columns; Miltenyi Biotec, Bergisch Gladbach, Germany) pulsed with 10 μg/ml MAGE-C2 peptide for 2 h (23). DCs were then cocultured with autologous CD8⁺ T cells at an E:T ratio of 10:1 in RPMI 1640 media supplemented with 10% heat-inactivated human AB serum, IL-7 (10 ng/ml), and IL-2 (100 U/ml). At day 7, the T cells were restimulated with peptide-loaded DCs at an E:T ratio of 10:1. The T cells were analyzed after 2 wk and sorted using MAGE-C2/HLA-A2 tetramers (provided by the Ludwig Institute for Cancer Research, Brussels, Belgium). CMV_495–503 peptide (NLVPMVATV/HLA-A2) was used as negative control. The frequency and function of MAGE-C2–specific CD8⁺ T cells were analyzed using flow cytometry.

After two rounds of stimulation for 2 wk, the T cells were restimulated with 10 μg/ml MAGE-C2 peptide for 6 h, and then stained with surface markers (CD3, CD8, and MAGE-C2/HLA-A2 tetramer). Degranulation assays for the expression of CD107a were performed by staining the CD107a Ab in the dark for 30 min. Subsequently, the cells were fixed in Fix/Perm solution (BD Biosciences). After being washed with Perm/Wash buffer (BD Biosciences), the cells were labeled with Abs against IFN-γ (Abcam, Cambridge, U.K.) in the dark for 30 min. By labeling the CD107a and IFN-γ Abs, we analyzed the percentage of CD107a⁺IFN-γ⁺ and CD107a⁻IFN-γ⁻ cells in tetramer⁺ CD8⁺ T cells. These two groups of cells were then sorted using a flow sorter to further explore the differential distribution of the TCR vβ subfamily.

To further analyze the distribution of the TCR vβ subfamily, CD107a⁺IFN-γ⁺ and CD107a⁻IFN-γ⁻ cells were used to synthesize cDNA with one-step PCR. First, the lysis buffer was configured to mix the resuspension buffer and lysis enhancer at a ratio of 10:1. Thereafter, 22 μl of this mixture was added to each tube and incubated on ice for 10 min to lyse the cells, which were subsequently incubated at 75°C for 10 min using the one-step RT kit (Invitrogen, Thermo Fisher Scientific, Waltham, MA) according to the manufacturer’s instructions. Finally, DNA was digested with DNase to obtain RNA and subsequently subjected to reverse transcription to synthesize cDNA.

Quantitative RT-PCR (qRT-PCR) was performed using SYBR Premix Ex Taq II (Takara, Shiga, Japan) in an Agilent Mx3005P instrument. Primer sequences (Sangon Biotech, Shanghai, China) used for qRT-PCR are listed in Supplemental Table I. GAPDH was used for normalization. The data were analyzed using the 2^−ΔΔCt method.

The gene expression heatmaps of 30 CTAs in 184 patients with ESCA were obtained from The Cancer Genome Atlas (TCGA) RNA sequencing (RNA-seq) data (https://tcga.xenahubs.net). The gene expression of MAGEC2 in various kinds of tumors was obtained from the Tumor Immune Estimation Resource (TIMER) database. The correlation analysis between TBX21 and IFNAR2 in 184 patients with ESCA was obtained from TCGA RNA-seq data (https://tcga.xenahubs.net).

Data were compared using paired or unpaired two-tailed t tests, which were expressed as mean ± SEM by using GraphPad Prism v8 (GraphPad Software, La Jolla, CA). The Pearson χ² test was used to analyze the correlation between the two different groups of data. Overall survival curves were evaluated using the Kaplan–Meier method and log-rank tests. Results were considered statistically significant at p < 0.05.

We compiled a set of 30 CT-X genes and obtained RNA-seq data from 184 ESCA tumor samples from TCGA database. We used RSEM values to analyze the RNA-seq data. The frequency of CT genes expressed in ESCA is shown in Supplemental Fig. 1A. MAGE Ag, a type of CTA, is absent from healthy adult tissue except for male testicular germ cells, but it is overexpressed in various cancers (24). Based on our previous studies, we analyzed the expression of MAGE-C2 in tumor tissues and normal tissues for each cancer type (Supplemental Fig. 1B). We found that MAGE-C2 was expressed in tumor tissues of ESCA, but not in normal tissues. Although MAGE-C2 expression was significantly correlated with patient prognosis in our preliminary results (18), its detailed and unique functionality in ESCA has not yet been explored and remains unknown. MAGE-C2–specific T cells responses have been widely researched for its role in antitumor immunotherapy. Therefore, we wanted to further explore the relationship among spontaneously occurring T cell responses against MAGE-C2 peptide, clinicopathological parameters, and patient prognosis.

To investigate the significance of MAGE-C2–specific CD8⁺ T cells in patients with ESCA, the MAGE-C2/HLA-A2–specific tetramer was used to detect the frequency of MAGE-C2-tetramer⁺ T cells in PBMCs from HLA-A2⁺ MAGE-C2⁺ patients. The results demonstrated that MAGE-C2-tetramer⁺ T cells could be detected after stimulation with MAGE-C2 peptide-loaded DCs compared with CMV peptide in vitro (Fig. 1A). Furthermore, we simultaneously analyzed the percentage of MAGE-C2-tetramer⁺ T cells in TILs and PBMCs. Representative data of patients (n = 3) are shown in (Fig. 1B (p < 0.05). The percentage of MAGE-C2-tetramer⁺ T cells was higher in TILs than in corresponding PBMCs in paired ESCA samples (Fig. 1C; p < 0.05). Furthermore, in PBMCs, the percentage of MAGE-C2-tetramer⁺ T cells was relatively lower in lymph node metastasis and late-stage patients compared with non–lymph node metastasis and early-stage patients (Fig. 1D, 1E) and had no significant correlation with patient tumor differentiation and sex (Fig. 1F, 1G).

Considering that the cytotoxic capacity of CD8⁺ T cells had a major influence on disease progression, we analyzed the reactivity of MAGE-C2–specific CD8⁺ T cells in PBMCs after peptide stimulation by evaluating the production of IFN-γ by intracellular cytokine staining and the expression of CD107a by degranulation assays. MAGE-C2–specific CD8⁺ T cell responses were readily detectable after 2 wk of expansion during stimulation with MAGE-C2 peptide, and the expression of CD107a and IFN-γ was significantly higher in tetramer⁺ CD8⁺ T cells than in tetramer⁻ CD8⁺ T cells (Fig. 2A; p < 0.05). (Fig. 2B shows the statistics of 73 patients with ESCA. Our results demonstrated that the levels of the cytolytic indicator CD107a and/or the functional marker IFN-γ were significantly decreased in patients with poor outcome parameters, including lymphatic metastasis, lower differentiation, and advanced tumor stage in PBMCs (Fig. 2C–E; p < 0.05), whereas they were not correlated with the sex of the patients (Fig. 2F; p > 0.05).

Previous studies have shown that enforced expression of the T-box transcription factor (T-bet) has important and well-described roles in CD8⁺ T cell activation, differentiation, and memory formation (25). The higher the expression of T-bet, the better the prognosis of cancer patients (26). Additionally, it has been predicted to be important for the expression of known T-bet target genes, such as Ifng and Gzmb (27). T-bet–deficient CD4⁺ T cells are severely defective in Th1 cell differentiation and IFN-γ production (25). Using TCGA, we found that in patients with ESCA, the expression level of T-bet (TBX21) was positively related to IFN-γ (IFNAR2) (Fig. 3A; p < 0.05).

To further explore whether T-bet was correlated with MAGE-C2–specific CD8⁺ T cell responses, the relationship between the percentage of T-bet⁺ and tetramer⁺ in CD8⁺ T cells and between the percentage of T-bet⁺ and CD107a⁺, IFN-γ⁺ in tetramer⁺CD8⁺ T cells of PBMCs was analyzed. The frequencies of T-bet⁺ and tetramer⁺ in CD8⁺ T cells from two representative patients with ESCA are shown in (Fig. 3B. We analyzed the data and found that the frequency of T-bet⁺/CD8⁺ T cells was positively correlated with tetramer⁺/CD8⁺ T cells (Fig. 3C; p < 0.05), and T-bet⁺/tetramer⁺ was positively correlated with CD107a⁺ or IFN-γ⁺/tetramer⁺ (Fig. 3D, 3E; p < 0.05). Thereafter, we stratified patients according to the frequency of CD107a or/and IFN-γ of MAGE-C2–specific CD8⁺ T cells. Patients with higher frequencies of CD107a and/or IFN-γ–expressing MAGE-C2–specific CD8⁺ T cells in PBMCs had a significantly longer survival time and presented an obvious survival advantage over patients with lower frequencies of these measures (Fig. 3F–H; p < 0.05).

T lymphocytes recognize tumor-specific epitopes presented by the MHC molecules on the target cell surface by utilizing their TCRs composed of α- and β-chains. We have confirmed that the stronger the MAGE-C2–specific CD8⁺ T cell function, the better the prognosis of the patients. These results led us to further investigate whether the MAGE-C2 CD8⁺ TCR also influence the patient prognosis, which has not been elucidated thus far. Subsequently, to further analyze the correlation between TCR vβ expression and patient prognosis, we sorted the CD107a⁺IFN-γ⁺ and CD107a⁻IFN-γ⁻ of MAGE-C2 CD8⁺ T cells using a flow sorter to synthesize cDNA by a CellsDirect one-step qRT-PCR kit to detect the TCR vβ expression and distribution (Fig. 4A). The results showed that there were significant discrepancies in TCR vβ subfamily expression and distribution between CD107a⁺IFN-γ⁺ and CD107a⁻IFN-γ⁻ of MAGE-C2 CD8⁺ T cells (Fig. 4B and 4C, respectively) among different individuals. Differential expression levels of each TCR vβ of the same individual between CD107a⁺IFN-γ⁺ and CD107a⁻IFN-γ⁻ of MAGE-C2–specific CD8⁺ T cells are shown in (Fig. 4D. The top five TCR vβs in terms of frequency in 34 ESCA patients were vβ22, vβ10, vβ16, vβ5, and vβ14 (Tables I, II), and the percentage of these five TCR vβs was, respectively, 38, 35, 32, 26, and 24% (Table II). Subsequently, we further analyzed whether the top five highly expressed TCR vβs are correlated with patient prognosis. The results showed that a higher expression of TCR vβ16 was positively and significantly correlated with a better prognosis (Fig. 5C). Nevertheless, TCR vβ22, vβ10, vβ5, and vβ14 expression had no obvious impact on patient prognosis (Fig. 5A, 5B, 5D, 5E).

The role of tumor-infiltrating immune cells has been a subject of debate for many decades (28, 29). T cells are the most important determinants of the antitumor immune responses by detecting tumor Ags and killing tumor cells (30). Many researches have demonstrated a conspicuous survival benefit associated with the presence of TILs in various tumors (31). At present, tumor immunotherapy mainly focuses on the exploration and application of genetically engineered T cell therapy. TCR T cell therapy is one of the genetically engineered T cell therapies. Effector T cells use their TCRs to recognize the corresponding Ags on tumor cells and attack tumor cells. Despite the significant development of diagnostic techniques and therapeutic modalities, ESCA is still an important cause of cancer-related deaths (16). Therefore, the clinical challenges faced by ESCA patients, the paucity of effective treatment options, and dismal survival rates needed to be improved urgently. Until now, the exact mechanisms and causes leading to ESCA have been not systematically evaluated. Thus, there is urgency to find new treatments for ESCA.

MAGE Ags, a type of CTA, are selectively expressed in tumor tissues, but not in normal tissues, except in human testicular germ cells (11). Previous studies have demonstrated that MAGE Ags are correlated with an advanced stage and a worse prognosis of tumor patients, such as ESCC (18) and non–small cell lung cancer (17), suggesting that MAGE Ags can be used as ideal immunotherapeutic targets for cancer therapy (32). T-bet is a key transcriptional factor that regulates tumor-reactive CD8⁺ T cell effector differentiation. Sullivan et al. (27) demonstrated that the redundant roles of T-bet in T cells are sufficient for IFN-γ production, cytotoxicity, and antitumor responses by tumor-infiltrating T cells. More importantly, the higher expression of T-bet was significantly correlated with a better outcome in triple-negative breast cancer (33). Therefore, we want to further explore the relationship among Ag-specific T cell responses against MAGE-C2 peptide, T-bet expression, clinicopathological parameters, and prognosis of ESCA patients.

Our results demonstrated MAGE-C2–specific T cell responses in CD8⁺ T cells, which were detectable, derived from PBMCs were weaker than those in TILs after stimulation with MAGE-C2 peptide-loaded DCs for 2 wk ex vivo, suggesting that the immunosuppressive tumor microenvironment in vivo could be overcome in vitro. Moreover, the increased level of lineage-restricted transcription factor T-bet correlated with an improved Ag-specific T cell function, such as functional markers CD107a and IFN-γ. The expression of functional markers CD107a and IFN-γ were significantly higher in patients with non–lymph node metastasis and early-stage patients. These phenomena may be a consequence of the high expression of exhaustion markers on T cells in the advanced stage and lymphatic metastasis (23). Weide et al. (3) have shown that IFN-γ–releasing reactive T cells have a strong impact on the survival of patients with melanoma. Our results indicated that the higher the expression of CD107a and/or IFN-γ, the better the patient prognosis, suggesting that functional MAGE-C2–specific CD8⁺ T cell responses may indeed be an indicator of patient prognosis. Collectively, these studies underscore the importance of tumor Ag in malignant tissue in a highly restricted manner, which initiates a clinically effective Ag-specific T cell response.

In recent years, genetic engineering of autologous T cells to express TCRs that recognize the known tumor target Ags has provided a feasible and applicable treatment for patients with metastatic melanoma (7), colorectal carcinoma (34), and multiple myeloma (8). Based on their immunogenicity and frequencies, CTL or TCR gene-modified T cells against MAGE peptides may provide a potential alternative for cancer patients (35). A variety of different tumor-associated Ags can spontaneously induce CD8⁺ T cell responses, such as NY-ESO-1 and MAGE-A3 in patients with melanoma (3) and ESCC (23), respectively. Johnson et al. (36) showed that 3 of 16 (19%) patients treated with gp100 TCR-engineered T cells experienced an objective antitumor response. Rapoport et al. (8) demonstrated that >50% of patients with myeloma after using a TCR targeting NY-ESO-1 Ag mediated sustained Ag-specific antitumor responses. Morgan et al. (7) reported in 2006 that MART-1 TCR-modified lymphocytes could mediate tumor regression in humans. Most strikingly, a MART-1–reactive, vβ7-expressing clone from the TILs of one patient with metastatic melanoma comprised >66% of his circulating CD8⁺ lymphocytes 810 d after treatment (5). In our study, we have demonstrated that the expression and distribution of TCR vβ was distinctly different in CD107a⁺IFN-γ⁺ and CD107a⁻IFN-γ⁻ of MAGE-C2 CD8⁺ T cells. The expression of TCR vβ16 is closely correlated with the prognosis of ESCA patients.

In summary, circulating functional T cells targeting MAGE-C2 have a strong impact on the prognosis of ESCA patients. The higher the expression of TCR vβ16, the better the prognosis of the patients. However, for various reasons, transducing with TCR vβ16-expressing clone into T lymphocytes and reinfusion back into the patients to detect the curative effect has not been further researched in our present study and is required to be carried out in the future.

We thank Dr. Pierre van der Bruggen from the Ludwig Institute for Cancer Research–Brussels Branch for providing MAGE-C2/HLA-A2 tetramer.

The authors have no financial conflicts of interest.