Analysis of cytotoxic T lymphocytes from a patient with hepatocellular carcinoma who showed a clinical response to vaccination with a glypican-3-derived peptide

Glypican-3 (GPC3), which is a carcinoembryonic antigen, is overexpressed in human hepatocellular carcinoma (HCC). Previously, we performed a phase I clinical trial of GPC3-derived peptide vaccination in patients with advanced HCC, and reported that GPC3 peptide vaccination is safe and has clinical efficacy. Moreover, we proposed that a peptide-specific CTL response is a predictive marker of overall survival in patients with HCC who receive peptide vaccination. In this study, we established GPC3-derived peptide-specific CTL clones from the PBMCs of an HLA-A * 02:07-positive patient with HCC who was vaccinated with an HLA-A2-restricted GPC3 peptide vaccine and showed a clinical response in the phase I clinical trial. Established CTL clones were analyzed using the IFN-γ ELISPOT assay and a cytotoxicity assay. GPC3 peptide-specific CTL clones were established successfully from the PBMCs of the patient. One CTL clone showed cytotoxicity against cancer cell lines that expressed endogenously the GPC3 peptide. The results suggest that CTLs have high avidity, and that natural antigen-specific killing activity against tumor cells can be induced in a patient with HCC who shows a clinical response to vaccination with the GPC3 144–152 peptide.


Introduction
Primary liver cancer, which is frequently hepatocellular carcinoma (HCC), is the sixth most common cancer and third most frequent cause of cancer-related death worldwide, and it is becoming more prevalent not only in East Asia, South-East Asia, and Africa but also in Western countries (1)(2)(3). Recently, the multikinase inhibitor sorafenib was demonstrated to prolong overall survival (OS) in patients with advanced HCC, and it has become the standard drug for first-line systemic treatment (4)(5)(6). However, based on the Response Evaluation Criteria in Solid Tumors (RECIST), the response rate for sorafenib is rather low, and the incidence of adverse events is relatively high, especially in elderly patients (7). Therefore, the generation of a novel effective therapy for HCC is a priority.
On the basis of these results, we conducted a phase I clinical trial of a GPC3-derived peptide vaccine in 33 patients with advanced HCC. The HLA-A * 24:02-restricted GPC3 298-306 peptide was used for HLA-A * 24:02-positive patients and

Analysis of cytotoxic T lymphocytes from a patient with hepatocellular carcinoma who showed a clinical response to vaccination with a glypican-3-derived peptide
the HLA-A * 02:01-restricted GPC3 144-152 peptide was used for HLA-A * 02:01, HLA-A * 02:06 and HLA-A * 02:07-positive patients. We found that GPC3 vaccination was well-tolerated, and that the GPC3 peptide vaccine induced a GPC3-specific CTL response in almost all of the patients (27)(28)(29)(30). Moreover, the vaccination-induced GPC3-specific CTL response correlated with overall survival (OS); the OS was significantly longer in patients with high GPC3-specific CTL frequencies than in those with low GPC3-specific CTL frequencies (27).
In terms of clinical responses, one patient showed a partial response (PR) and 19 patients showed stable disease 2 months after initiation of treatment. One patient with HCC who showed a PR was HLA-A * 02:07-positive. In addition, several HLA-A * 02:01-restricted GPC3 peptide-specific CTL clones with cytotoxic activities against GPC3 were established from the peripheral blood mononuclear cells (PBMCs) of patients vaccinated in this trial (27).
The aims of the present study were: i) to establish GPC3-derived, peptide-specific CTL clones from the PBMCs of an HLA-A * 02:07-positive patient with HCC who showed a PR in the phase I clinical trial; and ii) to analyze the functions of these CTL clones.

Materials and methods
Ethics information. This study was approved by the Ethics Committee of the National Cancer Center and conformed to the ethical guidelines of the 1975 Declaration of Helsinki. All the patients gave written informed consent before entering the study at the National Cancer Center Hospital East (Chiba, Japan). The trial has been registered with the University Hospital Medical Information Network Clinical Trials Registry (UMIN-CTR no. 000001395).
CD107a staining and flow cytometry analysis. CD8 + T cells were isolated using human CD8 microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany) from PBMCs that were stimulated with the GPC3 144-152 peptide for 14 days. The CD8 + T cells were incubated with GPC3 144-152 -pulsed or HIV 19-27 -pulsed 1-87 cells at a ratio of 2:1 for 3.5 h at 37°C. CD107a-specific antibodies (BD Biosciences, San Jose, CA, USA) were included in the mixture during the incubation period.
Generation of CTL clones. CD8 + CD107a + cells were sorted using a FACSAria cell sorter (BD Biosciences). Sorted CTLs were stimulated and the CTL clones were established as previously described (28).
Cytotoxicity assay. Cytotoxic capacity was analyzed with the Terascan VPC system (Minerva Tech, Tokyo, Japan). The CTL clone was used as the effector cell type. Target cells were labeled in calcein-AM solution for 30 min at 37˚C. The labeled cells were then co-cultured with the effector cells for 4-6 h. Fluorescence intensity was measured before and after the culture period, and specific cytotoxic activity was calculated as previously described (28).
IFN-γ ELISPOT assay. Specific secretion of IFN-γ from human CTLs in response to stimulator cells was assayed using the IFN-γ ELISPOT kit (BD Biosciences), according to the manufacturer's instructions. Stimulator cells were pulsed with or without peptide for 1.5 h at room temperature and then washed three times. Responder cells were incubated with stimulator cells for 20 h. The resulting spots were counted using an ELIPHOTO counter (Minerva Tech).
Determination of recognition efficiency. Calcein-AM-labeled target cells were pulsed with various concentrations of peptide, starting at 10 -6 M and decreasing in log steps to 10 -14 M. The CTL clones were incubated with the target cells at an effector:target (E/T) ratio of 10:1 for 4 h. The recognition efficiencies of the CTL clones were defined as previously described (28).
Sequence analysis of TCR-β gene. Using the TRIzol reagent (Invitrogen), total cellular RNA was extracted from established CTL clones. The cDNA of the TCR-β gene was synthesized using the PrimeScript II 1st Strand cDNA Synthesis Kit (Takara) according to the manufacturer's instructions, with the modification that we used 200 nM of the primer specific for the TCR-β chain constant region. The cDNA products were subjected to 2-step PCR, as previously described by Yukie Tanaka-Harada (35,36), and the PCR products were purified and sequenced in the Applied Biosystems 3500 Genetic Analyzer (Applied Biosystems). The TCR-β variable (TRBV) gene, TCR-β joining (TRBJ) gene, TCR-β diversity (TRBD) alleles, and complementarity-determining region 3 (CDR3) sequences were identified using the IMGT databases (http://www.imgt.org/).

Results
GPC3 144-152 peptide-specific CTLs in the peripheral blood of the patient exert a clinical effect. We analyzed the immune responses of the patient who showed a PR following GPC3 144-152 peptide vaccination. In this patient, the supraclavicular lymph node metastases markedly regressed, two liver tumors disappeared, and the thoracic bone metastasis showed necrosis after the third vaccination (27). The levels of DCP decreased in the patients over the 2-month period. We evaluated the GPC3 144-152 -specific immune responses in the peripheral blood using the ex vivo IFN-γ ELISPOT assay. For the HLA-A * 02:07-positive patient with advanced HCC, the number and area of the spots increased after two rounds of vaccination, as compared with the pre-vaccination values, and the peak values were noted 10 weeks after the start of the treatment (Fig. 1A).
Establishment of GPC3 144-152 -specific CTL clones from the PBMCs of the patient. To investigate the ability of the GPC3 144-152 -specific CTLs induced by peptide vaccination to recognize antigen, we established CTL clones from the PBMCs of this patient 10 weeks after the start of treatment. The PBMCs were stimulated with the GPC3 144-152 peptide in vitro for 14 days. CD8 + T cells were isolated from the stimulated PBMCs, and then incubated with peptide-pulsed 1-87 cells. CD8 + CD107a + cells that reacted with the GPC3 144-152 -pulsed 1-87 cells were sorted to the single-cell level. Thus, we established GPC3 144-152 peptide-specific CTL clones.

Functional avidity of the GPC3 144-152 -specific CTL clones.
We evaluated the cytotoxicity profiles of the CTL clones for 1-87 cells pulsed with a decreasing concentration series (from 10 -6 to 10 -14 M) of the GPC3 144-152 peptide. The peptide concentration at which the curve reached 50% cytotoxicity was defined as the recognition efficiency of the clone. The recognition efficiencies of CTL clones 24-4-2, 24-4-7 and 24-2-10 were 10 -11 , 10 -9 and 10 -8 M, respectively (Fig. 2). This result suggests that CTL clone 24-4-2 has a higher avidity than the other two clones and, conversely, that CTL clone 24-2-10 has a lower avidity than the other two clones.

Established CTL clones have different sets of TCR-β alleles.
We analyzed the TCR-β gene sequences of the established CTL clones. The TRBV, TRBJ and TRBD alleles were identified using the IMGT databases. Thus, we identified the TRBV, TRBD and TRBJ alleles of the CTL clones (Table I). Each of the established CTL clones had different allele sets.  The established CTL clones were tested for avidity using 1-87 cells that were pulsed with various concentrations of the GPC3 144-152 peptide. The peptide concentration at which the curve crossed the 50% cytotoxicity mark was defined as the recognition efficiency of that clone. E/T is 10. A representative of three experiments is shown.

Discussion
Clinical trials of peptide-based vaccines are underway in several parts of the world. However, the monitoring of individual CTL post-vaccination has scarcely been reported in   immunotherapy trials. In the present study, we established HLA-A * 02:07 + GPC3 144-152 -specific CTL clones from the PBMCs of a patient who showed a PR following GPC3-derived peptide vaccination and we performed functional analyses against established CTL clones. This patient showed an increase in the number of CTLs specific for the GPC3-derived peptide in the peripheral blood after vaccination (Fig. 1A) (27,28). Ten weeks after the start of treatment, the GPC3 144-152 -specific CTL counts had increased approximately 18-fold, as compared with the pre-vaccination counts. In this case, analysis of the established CTL clones after vaccination could lend support to the notion that the vaccine-induced CTLs exert an antitumor effect, since few GPC3 144-152 -specific CTLs were detected before vaccination.
In the present study, we confirmed that GPC3 144-152 -specific CTL clones are cytotoxic for both GPC3 144-152 -pulsed 1-87 cells and JHH-7/HLA-A * 02:07 cells that express GPC3 peptide endogenously. Confirming that the GPC3 peptide-specific CTL clones kill cancer cells that express endogenously the antigen peptide is important because antigen-derived and CTL-inducible peptides are not necessarily presented by cancer cells that endogenously express the antigen (31)(32)(33). Three established CTL clones showed cytotoxic activities related to their avidity for GPC3 144-152 -pulsed 1-87 cells and JHH-7/ HLA-A * 02:07 cells that expressed the GPC3 peptide endogenously. These results show that although CTLs with different avidity can be isolated, only those CTLs with high avidity can kill cancer cells that express the antigen peptide endogenously. Several investigators have demonstrated a correlation between T-cell avidity and target recognition by T-cell populations that recognize murine tumor models and human cancers (34). Our results strongly support this observation.
The TCR usage of antigen-specific T cells is thought to be influenced by the affinity of the TCR for the antigen peptide-HLA class I complex. Several studies on the TCR usage of tumor-associated antigen (TAA)-specific T cells have used the TRBV gene family (35)(36)(37)(38)(39)(40)(41). These studies mainly analyzed the frequencies of TAA tetramer positive CD8 + T cells. Although it is important to examine quantitative aspects, such as the frequencies of TAA tetramer positive CD8 + T cells, the cytotoxicity of these T cells against cancer cells that express the TAA peptide endogenously cannot be confirmed. Moreover, GPC3 dextramer positive CD8 + T cells were not detected in the PBMCs of the patients with HCC before GPC3 peptide vaccination (27,28). To analyze biased usage of the TCR gene of GPC3 dextramer positive CD8 + T cells in the patients with HCC before and after GPC3 peptide vaccination, a new detection system with greater sensitivity ex vivo will be required. In the present study, we analyzed the TCR-β genes of the established GPC3 144-152 -specific CTL clones, to confirm that these CTL clones have different TCRs. Our experiments show that the established CTL clones have different TCR-β-chain allele sets, i.e., TRBV, TRBD and TRBJ alleles (Table I), and different CDR3 sequences (data not shown). These results suggest that various GPC3-specific CTLs are induced by GPC3 144-152 peptide vaccination.
We have reported previously on the detection via immunohistochemical staining of massive infiltration of CD8-positive T cells into the remaining liver tumor of this patient (27). It was difficult to confirm that these tumor-infiltrating CD8 + T cells have specificity for GPC3. Currently, we are conducting clinical testing of liver biopsies taken before and after GPC3 peptide vaccination of patients with advanced HCC. Our aim is to reveal the GPC3 peptide-specific immune responses induced by the GPC3-derived peptide vaccine in both the peripheral blood and the tumor. We are analyzing the TCR gene sequences of CD8 or GPC3 dextramer positive T cells in both the peripheral blood and tumor. Already in this trial, a remarkable clinical effect has been observed for an HLA-A * 02:07-positive patient with HCC who received GPC3 144-152 peptide vaccination (49).
HLA-A * 02:07 is present in the populations of East Asia, South-East Asia (7%), and northern India (11.5%) (26,(50)(51)(52). In southern China, the frequency of the HLA-A * 02:07 allele is reported to be even higher than the frequency of the HLA-A * 02:01 allele (53,54). In addition, about 75% of liver cancer cases occur in South-East Asia, including China, Hong Kong, Taiwan, Korea, India and Japan (55). Taking together these previous reports and our results, it appears that HLA-A * 02:07-positive patients with HCC are good candidates for GPC3 144-152 peptide vaccination. Further studies will be necessary to prove the clinical efficacy of GPC3 peptide vaccination for advanced HCC.
In conclusion, we present substantial evidence that GPC3 144-152 -specific CTLs with different TCR allele sets that are induced in patients with HCC who show a PR following GPC3 144-152 peptide vaccination indicate not only high avidity but also natural antigen-specific killing activity against tumor cells.