• Open Access

HLA-A2-restricted glypican-3 peptide-specific CTL clones induced by peptide vaccine show high avidity and antigen-specific killing activity against tumor cells

Authors

  • Toshiaki Yoshikawa,

    1. Section for Cancer Immunotherapy, Investigative Treatment Division, Research Center for Innovative Oncology, National Cancer Center Hospital East, Kashiwa
    2. Division of Immunology, Department of Molecular and Cellular Biology, School of Life Science, Faculty of Medicine, Tottori University, Yonago
    Search for more papers by this author
  • Munehide Nakatsugawa,

    1. Section for Cancer Immunotherapy, Investigative Treatment Division, Research Center for Innovative Oncology, National Cancer Center Hospital East, Kashiwa
    Search for more papers by this author
  • Shiro Suzuki,

    1. Section for Cancer Immunotherapy, Investigative Treatment Division, Research Center for Innovative Oncology, National Cancer Center Hospital East, Kashiwa
    Search for more papers by this author
  • Hirofumi Shirakawa,

    1. Section for Cancer Immunotherapy, Investigative Treatment Division, Research Center for Innovative Oncology, National Cancer Center Hospital East, Kashiwa
    Search for more papers by this author
  • Daisuke Nobuoka,

    1. Section for Cancer Immunotherapy, Investigative Treatment Division, Research Center for Innovative Oncology, National Cancer Center Hospital East, Kashiwa
    Search for more papers by this author
  • Noriko Sakemura,

    1. Section for Cancer Immunotherapy, Investigative Treatment Division, Research Center for Innovative Oncology, National Cancer Center Hospital East, Kashiwa
    Search for more papers by this author
  • Yutaka Motomura,

    1. Section for Cancer Immunotherapy, Investigative Treatment Division, Research Center for Innovative Oncology, National Cancer Center Hospital East, Kashiwa
    Search for more papers by this author
  • Yukie Tanaka,

    1. Division of Hematology, Saitama Medical Center, Jichi Medical University, Saitama, Japan
    Search for more papers by this author
  • Shin-Ichi Hayashi,

    1. Division of Immunology, Department of Molecular and Cellular Biology, School of Life Science, Faculty of Medicine, Tottori University, Yonago
    Search for more papers by this author
  • Tetsuya Nakatsura

    Corresponding author
    1. Section for Cancer Immunotherapy, Investigative Treatment Division, Research Center for Innovative Oncology, National Cancer Center Hospital East, Kashiwa
    Search for more papers by this author

To whom correspondence should be addressed.
E-mail: tnakatsu@east.ncc.go.jp

Abstract

Glypican-3 (GPC3) is an onco-fetal antigen that is overexpressed in human hepatocellular carcinoma (HCC), and is only expressed in the placenta and embryonic liver among normal tissues. Previously, we identified an HLA-A2-restricted GPC3144–152 (FVGEFFTDV) peptide that can induce GPC3-reactive CTLs without inducing autoimmunity in HLA-A2 transgenic mice. In this study, we carried out a phase I clinical trial of HLA-A2-restricted GPC3144–152 peptide vaccine in 14 patients with advanced HCC. Immunological responses were analyzed by ex vivoγ-interferon enzyme-linked immunospot assay. The frequency of GPC3144–152 peptide-specific CTLs after vaccination (mean, 96; range, 5–441) was significantly larger than that before vaccination (mean, 6.5; range, 0–43) (P < 0.01). An increase in the GPC3144–152 peptide-specific CTL frequency was observed in 12 (86%) of 14 patients after vaccination. Additionally, there was a significant correlation between the maximum value of GPC3144–152 peptide-specific CTLs after vaccination and the dose of the peptide injected (P = 0.0166, r = 0.665). Moreover, we established several GPC3144–152 peptide-specific CTL clones from PBMCs of patients vaccinated with GPC3144–152 peptide by single cell sorting using Dextramer and CD107a antibody. These CTL clones had high avidity (the recognition efficiency showing 50% cytotoxicity was 10−10 or 10−11 M) and could recognize HCC cell lines expressing GPC3 in an HLA-class I-restricted manner. These results suggest that GPC3144–152 peptide vaccine can induce high avidity CTLs capable of killing HCC cells expressing GPC3. This trial was registered with University Hospital Medical Information Network number 000001395. (Cancer Sci 2011; 102: 918–925)

In peptide-based vaccine trials, occasional marked clinical regressions of melanoma have been observed after peptide vaccination; however, tumor regressions have not correlated well with T cell responses measured in peripheral blood lymphocytes.(1–3) This may be because the clinical response to a vaccine was unrelated to the immune response to that vaccine or due to inadequate immune response monitoring. Moreover, vaccination with synthetic peptides has occasionally induced ineffective CTL responses due to various mechanisms.(4–9) When evaluating T cell response to peptide vaccines, it is important to confirm that the peptide is presented naturally on cancer cells and that responding CTLs lyse human cancer cells.

Glypican-3 (GPC3) is specifically overexpressed in human hepatocellular carcinoma (HCC).(10) The expression of GPC3 was correlated with a poor prognosis in HCC patients.(11) Moreover, GPC3 is useful not only as a novel tumor marker, but also as a target antigen for immunotherapy in several studies with mice.(12–14) We identified HLA-A*24:02-restricted GPC3298–306 (EYILSLEEL) and HLA-A*02:01-restricted GPC3144–152 (FVGEFFTDV) peptides, both of which can induce GPC3-reactive CTLs without inducing autoimmunity,(15) and reported a preclinical study using a mouse model with a view to designing an optimal schedule for the clinical trials of a GPC3-derived peptide vaccine and showed dose-dependency in the immunizing effect of the peptide vaccine.(16)

In this study, we completed the phase I clinical trial of a GPC3-derived peptide vaccine for 30 patients with advanced HCC (manuscript in preparation). Among them, 16 patients had the HLA-A24 gene and 14 had the HLA-A2 gene. Here, we describe the immunological evaluation of HLA-A2-restricted GPC3144–152 peptide vaccine in a phase I trial involving 14 patients. We highlight three important points: (i) HLA-A2-restricted GPC3144–152 peptide is immunogenic in advanced HCC patients; (ii) dose-dependent effects of GPC3144–152 peptide vaccine; and (iii) establishment of CTL clones showing not only high avidity but also natural antigen-specific killing activity against HCC cells.

Materials and Methods

Patients.  Fourteen patients with advanced HCC were injected with HLA-A2-restricted GPC3144–152 (FVGEFFTDV) peptide vaccine at the National Cancer Center Hospital East (Kashiwa, Japan). HLA-A2 gene-positive status was determined by genomic DNA typing tests (Mitsubishi Chemical Medience, Tokyo, Japan). All patients gave written informed consent before entering the study. The profiles of the 14 patients are summarized in Table 1. 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.

Table 1.   Summary of profiles of 14 patients with advanced human hepatocellular carcinoma who participated in this study, with their clinical and immunological responses before and after vaccination with HLA-A2-restricted GPC3144–152 peptide
Pt.HLAAge (years)SexStageDose of peptide (mg)Clinical response†GPC3-specific CTLs‡
PrePostChange
  1. †The clinical response was evaluated according to the Response Evaluation Criteria in Solid Tumors (RECIST) guidelines. ‡Peripheral blood was taken from each patient before and after vaccination, and glypican-3 (GPC3)-specific CTLs were measured by ex vivoγ-interferon enzyme-linked immunospot assay. F, female; M, male; PD, progressive disease; PR, partial response; Pt., patient; SD, stable disease; +, increase; −, decrease.

A2-1A*02:06/A*02:0767MIV0.3SD4340
A2-2A*02:0162MIIIA0.3PD018+
A2-3A*02:0155MIIIA0.3SD110+
A2-4A*02:0168FIIIC1.0SD1615
A2-5A*02:0172MIIIA1.0SD16101+
A2-6A*02:01/A*02:0662MII1.0PD023+
A2-7A*02:0167FIV3.0SD023+
A2-8A*02:0158MIIIA3.0SD0101+
A2-9A*02:0152MIV10.0SD1100+
A2-10A*02:0170MIV10.0PD05+
A2-11A*02:0168MII10.0PD1125+
A2-12A*02:0775FIV30.0PR11196+
A2-13A*02:0652MIV30.0PD2151+
A2-14A*02:0167MIV30.0PD0441+

Treatment protocol.  Vaccinations with GMP grade peptide, GPC3144–152 (FVGEFFTDV) (American Peptide Co., Sunnyvale, CA, USA) emulsified with incomplete Freund’s adjuvant (Montanide ISA-51 VG; Seppic, Paris, France) were carried out intradermally three times at 14-day intervals. Five incremental dose levels at 0.3, 1, 3, 10, and 30 mg/body were planned for the peptide administration.

Preparation of PBMCs.  Peripheral blood (30 mL) was obtained from each patient at times designated in the protocol (before the first vaccination and 2 weeks after each vaccination) and centrifuged using a Ficoll–Paque gradient.

Ex vivo interferon (IFN)-γ enzyme-linked immunospot (ELISPOT) analysis.  ELISPOT assay for the detection of antigen-specific IFN-γ producing T cells was carried out using the BD ELISPOT kit (BD Bioscience, San Jose, CA, USA) according to the manufacturer’s protocols. In brief, non-cultured PBMCs (5 × 105 cells/well) were added to plates in the presence of 10 μg/mL peptide antigens and incubated for 20 h at 37°C, 5% CO2. The GPC3 antigen was HLA-A2-restricted GPC3144–152 (FVGEFFTDV) peptide. The PBMCs with HLA-A2-restricted HIV19–27 (TLNAWVKVV) peptide were used as a negative control. The spots were automatically counted using the Eliphoto system (Minerva Tech, Tokyo, Japan).

Cell lines.  The human liver cancer cell line HepG2 (GPC3+, HLA-A*02:01/A*24:02), SK-Hep-1 (GPC3, HLA-A*02:01/A*24:02), the human melanoma cell line 526mel (GPC3+, HLA-A*02:01), and the human colon cancer cell line SW620 (GPC3, HLA-A*02:01/A*24:02) were used as target cells. T2 (HLA-A*02:01, TAP) was pulsed with GPC3144–152 peptide or HIV19–27 peptide at room temperature for 1 h. They were conserved in our laboratory.

Induction of GPC3144–152 peptide-specific CTLs from PBMCs.  The PBMCs were cultured (2 × 106 cells/well) with 10 μg/mL GPC3144–152 peptide in AIM-V medium supplemented with 10% human AB serum, recombinant human interleukin (IL)-2 for 14 days.

Dextramer staining and flow cytometry analysis.  The PBMCs were stained with HLA-A*02:01 Dextramer-RPE (GPC3144–152 [FVGEFFTDV], HIV19–27 [TLNAWVKVV]; Immudex, Copenhagen, Denmark) for 10 min at room temperature and anti-CD8-FITC (ProImmune, Oxford, UK) for 20 min at 4°C. Flow cytometry analysis was carried out using FACSAria cell sorter (BD Bioscience).

CD107a staining and flow cytometry analysis.  CD8+ T cells were isolated using human CD8 microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany) from PBMCs stimulated with GPC3144–152 peptide for 14 days. CD8+ T cells were incubated with T2 pulsed with GPC3144–152 or HIV19–27 peptide and HepG2 at a 2:1 ratio for 3.5 h at 37°C. CD107a-specific antibodies (BD Bioscience) were included during the incubation period.

Generation of CTL clones.  CD8+ GPC3 Dextramer+ or CD107a+ cells were sorted using a FACSAria cell sorter and seeded in a 96-well plate (1 cell/well) and stimulated by the addition of irradiated (100 Gy) allogeneic PBMCs (8 × 104 cells/well) as feeder cells, in AIM-V medium supplemented with 10% human AB serum, IL-2 (200 U/mL), and phytohemagglutinin-P (PHA) (5 μg/mL) for 14–21 days.

Response of CTL clones against cancer cell lines.  The CTL clones were cocultured with each cancer cell line as a target cell at the indicated effector/target (E/T) ratio, and cytotoxicity assay or IFN-γ ELISPOT assay was carried out. Blocking of HLA-class I or HLA-A2 was carried out as previously described.(15)

Cytotoxicity assay.  Cytotoxic activity against target cells was analyzed using the Terascan VPC system (Minerva Tech). Target cells were labeled with calcein AM (Dojindo, Kumamoto, Japan) solution for 30 min at 37°C. The labeled cells were then incubated with effector cells for 4–6 h. Fluorescence intensity was measured before and after the 4–6 h culture, and specific cytotoxic activity was calculated using the following formula: % cytotoxicity = {1− [(average fluorescence of the sample wells − average fluorescence of the maximal release control wells)/(average fluorescence of the minimal release control wells − average fluorescence of the maximal release control wells)]} × 100%.

Determination of recognition efficiency.  Calcein AM-labeled T2 target cells were pulsed with a range of peptide concentrations, starting at 10−6 M and decreasing by log steps to 10−14 M. The CTL clones were incubated with T2 target cells at a 10:1 E/T ratio for 4 h. For each CTL clone, % cytotoxicity was plotted against each peptide concentration. The peptide concentration at which the curve crossed 50% cytotoxicity was defined as the recognition efficiency of that clone.

Cold inhibition assay.  Calcein AM-labeled target cells were cultured with effector cells in a 96-well plate with cold target cells. T2 target cells, which were prepulsed with either HIV19–27 peptide or GPC3144–152 peptide, were used as cold target cells.

RNA interference.  Small interfering RNAs specific for human GPC3 were chemically synthesized double-strand RNAs (Invitrogen, Carlsbad, CA, USA). A non-silencing siRNA, AllStras Neg. Control siRNA, was obtained from Qiagen (Valencia, CA, USA). The GPC3-specific siRNA sequence used in this study was: 5′-GGAGGCUCUGGUGAUGGAAUGAUAA-3′. Synthetic siRNA duplexes were transfected using Lipofectamine RNAiMAX (Invitrogen) according to the manufacturer’s protocols.

Statistical analysis.  Student’s t-test was used to determine statistically significant differences between the two groups. Correlation between the frequency of GPC3-specific CTLs and the dose of the peptide injected was analyzed using Spearman’s rank correlation coefficient. Data from the ELISPOT assay using siRNA were statistically analyzed by one-way anova followed by Tukey’s multiple comparison test. Statistical significance was set as P < 0.05.

Results

Analysis of GPC3144–152 peptide-specific CTLs in PBMCs of vaccinated patients.  To analyze immune responses in the 14 patients vaccinated with GPC3144–152 peptide, we evaluated the GPC3144–152 peptide-specific immune responses by ex vivo IFN-γ ELISPOT assay. The representative data of patient A2-12 on changes in the frequency of GPC3144–152 peptide-specific CTLs before and after vaccination are shown in Figure 1(a). The frequencies of GPC3144–152 peptide-specific CTLs were 11 and 196 of 5 × 105 PBMCs at pre- and post-vaccination, respectively. The results of the comparison of the frequency of GPC3144–152 peptide-specific CTLs before vaccination and after vaccination in all patients are shown in Table 1 and Figure 1(b). GPC3144–152 peptide-specific CTLs were clearly detected in four and 14 of the 14 patients at pre- and post-vaccination, respectively. The frequency of GPC3144–152 peptide-specific CTLs after vaccination (mean, 96; range, 5–441) was significantly larger than that before vaccination (mean, 6.5; range, 0–43) (P < 0.01). An increase in GPC3144–152 peptide-specific CTLs was found in 12 (86%) of the 14 patients, except in two cases (patients A2-1 and A2-4). These results suggest that GPC3144–152 peptide vaccination can induce an increase in GPC3144–152 peptide-specific CTLs in HCC patients. Moreover, we compared the frequency of GPC3144–152 peptide-specific CTLs after vaccination for each dose of peptide injected. We found that the maximum value of GPC3144–152 peptide-specific CTLs after vaccination was significantly correlated with the dose of the peptide injected (P = 0.0166, r = 0.665) (Fig. 1c).

Figure 1.

 Changes in the frequency of GPC3144–152 peptide-specific CTLs before and after vaccination. Direct ex vivoγ-interferon enzyme-linked immunospot assay of PBMCs (5 × 105) was carried out. The Δ spot number indicates the number of GPC3144–152 peptide-specific CTLs calculated by subtracting the spot number in a well of HIV19–27 peptide. (a) Representative result showing the frequency of GPC3144–152 peptide-specific CTLs pre- and post-vaccination. (b) Changes in the frequency of GPC3144–152 peptide-specific CTLs before and after vaccination in all patients (A2-1–14). An increase in GPC3144–152 peptide-specific CTLs was observed in 12 (86%) of 14 patients. (c) The maximum number of GPC3144–152 peptide-specific CTLs after vaccination was significantly correlated with the dose of the peptide injected (P = 0.0166, r = 0.665).

Establishment of GPC3144–152 peptide-specific CTL clones by three different methods.  To further investigate the ability of GPC3144–152 peptide-specific CTLs induced by peptide vaccination to recognize an antigen, we established CTL clones from PBMCs of three vaccinated patients (patients A2-8, A2-9, and A2-14) by three different methods (Fig. 2). A representative clone from each patient is shown. In patient A2-9 (Fig. 2a), the frequency of GPC3144–152 peptide-specific CTLs was 50 of 5 × 105 PBMCs 1 month after the third vaccination, as determined by ex vivo ELISPOT assay, and 14 days after the in vitro stimulation with GPC3144–152 peptide, Dextramer assay was carried out. The population of CD8+ GPC3 Dextramer+ cells was 2.6% of all stimulated cells. These cells were sorted to a single cell in each well of a 96-well plate. Twenty-one days after cell sorting, peptide specificity was examined by Dextramer assay. The established CTL clone was CD8+ GPC3 Dextramer+ cells (99.7%) which did not react with HIV Dextramer as a negative control (Fig. 2a).

Figure 2.

 Establishment of GPC3144–152 peptide-specific CTL clones by three different methods. Left panels show the frequency of GPC3144–152 peptide-specific CTLs in the PBMCs used, as established by ex vivo interferon (IFN)-γ enzyme-linked immunospot (ELISPOT) assay. (a) The PBMCs of patient A2-9 were stimulated with GPC3144–152 peptide in vitro for 14 days. The population of CD8+ GPC3 Dextramer+ cells was sorted to a single cell. (b) CD8+ GPC3 Dextramer+ cells were directly sorted to a single cell from PBMCs of patient A2-14 without in vitro stimulation. (c) The PBMCs of patient A2-8 were stimulated with GPC3144–152 peptide in vitro for 14 days. CD8+ CD107a+ cells that reacted against HepG2 were sorted to a single cell. Right panels show Dextramer analysis of the established clones 21 days after cell sorting.

We next attempted to sort from small populations of GPC3144–152 peptide-specific CTLs without in vitro culture. In patient A2-14 (Fig. 2b), the frequency of GPC3144–152 peptide-specific CTLs was 329 of 5 × 105 PBMCs 2 weeks after the third vaccination, as determined by ex vivo ELISPOT assay; CD8+ GPC3 Dextramer+ cells could be clearly detected in 0.1% of PBMCs. CD8+ GPC3 Dextramer+ cells were directly sorted to a single cell from PBMCs without in vitro stimulation. The established CTL clone was CD8+ GPC3 Dextramer+ cells (99.9%) which did not react with HIV-Dextramer (Fig. 2b).

Finally, to establish high avidity and tumor-reactive CTLs from a heterogeneous population, we attempted to sort the population of CD8+ T cells which mobilized CD107a in response to naturally GPC3-expressing HepG2 cells. In the PBMCs from patient A2-8 (Fig. 2c), the frequency of GPC3144–152 peptide-specific CTLs was 39 of 5 × 105 PBMCs 1.5 months after the third vaccination, as determined by ex vivo ELISPOT assay, which were stimulated with GPC3144–152 peptide in vitro. After 14 days, the population of CD8+ GPC3 Dextramer+ cells was 1.4% of all stimulated cells. We incubated CD8+ T cells with T2 pulsed with GPC3144–152, HIV19–27 peptide, or HepG2. Approximately 2% and 18.7% of CD8+ T cells mobilized CD107a in response to HepG2 and T2 pulsed with GPC3144–152 peptide, respectively, but not in response to T2 pulsed with HIV19–27 peptide. CD107a+ CD8+ cells that reacted against HepG2 were sorted to a single cell. The established clone was CD8+ GPC3 Dextramer+ CTLs (99.9%) which did not react with HIV Dextramer (Fig. 2c). These results indicate that GPC3144–152 peptide-specific CTL clones were successfully established from PBMCs of patients injected with GPC3144–152 peptide vaccine by three different methods. Moreover, the result that patient A2-8 CTL clone that reacted to HepG2 had GPC3144–152 peptide specificity verified that GPC3144–152 peptide was present naturally on HepG2.

Analysis of GPC3144–152 peptide-specific avidity of three CTL clones.  To further characterize the GPC3144–152 peptide-specific avidity of the three CTL clones, we tested for the lysis of T2 cells pulsed with decreasing concentrations of GPC3144–152 or HIV19–27 peptide ranging from 10−6 to 10−14 M. The peptide concentration at which the curve crossed 50% cytotoxicity was defined as the recognition efficiency of that clone. The recognition efficiencies of patient A2-9, A2-14, and A2-8 clones were 10−10, 10−10, and 10−11 M, respectively (Fig. 3). These CTL clones did not react against T2 cells pulsed with HIV19–27 peptide. These results indicate that the established clones were GPC3144–152 peptide-specific and high avidity CTLs.

Figure 3.

 Analysis of the GPC3144–152 peptide specific avidity of the three CTL clones. The established CTL clones were tested for their avidities using various concentrations of GPC3144–152 (• ) or HIV19–27 (○) peptide-loaded T2 targets. The peptide concentration at which the curve crossed 50% cytotoxicity was defined as the recognition efficiency of that clone. Effector/target ratio = 10. The recognition efficiencies of patient A2-9 (a), A2-14 (b), and A2-8 (c) CTL clones were 10−10, 10−10, and 10−11 M, respectively.

Reactivity of three CTL clones against cancer cell lines.  We analyzed the IFN-γ production and cytotoxicity of the established CTL clones against cancer cell lines expressing HLA-A*02:01 and GPC3. We used SK-Hep-1 (GPC3, HLA-A*02:01+) and a human GPC3 gene transfectant, SK-Hep-1/hGPC3 (GPC3+, HLA-A*02:01+), as target cells. Production of IFN-γ in the three CTL clones was detected against SK-Hep-1/hGPC3, but not against SK-Hep-1 (Fig. 4a). Furthermore, these CTL clones showed specific cytotoxicity against SK-Hep-1/hGPC3 and HepG2 (GPC3+, HLA-A*02:01+), but not against SK-Hep-1 and SW620 (GPC3, HLA-A*02:01+) (Fig. 4b). These results indicate that all three CTL clones show cytotoxicity and the ability to produce IFN-γ against HLA-A*02:01+ GPC3+ HCC cell lines. Next, we examined whether these CTL clones respond to cancer cells weakly expressing GPC3. We used human melanoma cell line 526mel (GPC3+, HLA-A*02:01+) as a target cell that expresses GPC3 mRNA and protein at a lower level than the HCC cell lines (data not shown). Production of IFN-γ in patient A2-8 CTL clone (recognition efficiency: 10−11 M) were clearly detected against 526mel, whereas patient A2-9 CTL clone (recognition efficiency: 10−10 M) showed weak response to 526mel (Fig. 4c). Similarly, patient A2-8 CTL clone showed specific cytotoxicity against 526mel, whereas patient A2-9 CTL clone failed to lyse 526mel (Fig. 4d). These results suggest that higher avidity is essential to react to cancer cells weakly expressing GPC3.

Figure 4.

 Reactivity of three CTL clones against cancer cell lines. (a) γ-Interferon enzyme-linked immunospot assay of established CTL clones against SK-Hep-1/hGPC3 and SK-Hep-1/vec. Effector/target (E/T) ratio = 0.2. (b) Cytotoxic activities of the three CTL clones against SK-Hep-1/hGPC3 (bsl00001), SK-Hep-1/vec (□), HepG2 (• ), or SW620 (Δ) analyzed by cytotoxicity assay. (c) γ-Interferon enzyme-linked immunospot assay of established CTL clones against 526mel. E/T ratio = 0.2. (d) Cytotoxic activities of patient A2-8 (bsl00001) and A2-9 (□) CTL clone against 526mel analyzed by cytotoxicity assay.

Analysis of HLA-A2 and GPC3 restriction.  In a cold target inhibition assay, cytotoxicity against SK-Hep-1/hGPC3 of patient A2-9 clone was suppressed by the addition of GPC3144–152 peptide-pulsed T2 cells but not by the addition of HIV19–27 peptide-pulsed T2 cells (Fig. 5a). In an HLA blocking experiment, the IFN-γ production of patient A2-9 CTL clone was markedly inhibited by anti-HLA class I mAb and anti-HLA-A2 mAb as compared with that by IgG2a or IgG2b isotype control (P < 0.05) (Fig. 5b). Similarly, the cytotoxicity against SK-Hep-1/hGPC3 of patient A2-9 clone was markedly inhibited by anti-HLA class I mAb and anti-HLA-A2 mAb compared with that by IgG2a and IgG2b isotype control (P < 0.05) (Fig. 5c). These results clearly indicate that the CTL clone recognized SK-Hep-1/hGPC3 in an HLA-A2-restricted manner.

Figure 5.

 Analysis of HLA-A2 and glypican-3 (GPC3) restriction. (a) Cold target inhibition assay of patient A2-9 CTL clone against SK-Hep-1/hGPC3. Effector/target (E/T) ratio = 30. T2 was prepulsed with either HIV19–27 peptide or GPC3144–152 peptide, then used as cold target cells. Cold/hot target ratio = 10. The cytotoxicity of the CTL clone was inhibited by T2 pulsed with GPC3144–152 peptide but not by T2 pulsed with HIV19–27 peptide. (b) Inhibition of interferon (IFN)-γ production by anti-HLA class I mAb and anti-HLA A2 mAb. SK-Hep-1/hGPC3 used as target cells. E/T ratio = 0.02. The IFN-γ production of the CTL clone was markedly inhibited by anti-HLA class I mAb and anti-HLA-A2 mAb as compared with that by IgG2a and IgG2b isotype control (P < 0.05). Data are expressed as the mean ± SD. (c) Inhibition of cytotoxicity by anti-HLA class I mAb and anti-HLA A2 mAb. SK-Hep-1/hGPC3 used as target cells. E/T ratio = 30. The cytotoxicity of the CTL clone was markedly inhibited by anti-HLA class I mAb and anti-HLA-A2 mAb compared with that by IgG2a and IgG2b isotype control (P < 0.05). (d) The GPC3 expression on HepG2 treated with GPC3-siRNA or negative (neg)-siRNA for 24 h as determined by RT-PCR. (e) The GPC3 expression on HepG2 treated with GPC3-siRNA or neg-siRNA from 24 to 72 h as determined by Western blot analysis. The GPC3 expression of HepG2 was decreased from 24 to 72 h after treatment with GPC3 siRNA. (f) The IFN-γ production of the CTL clone against HepG2 treated with GPC3 siRNA. E/T ratio = 0.02. The IFN-γ production of the CTL clone was decreased by GPC3 siRNA (P < 0.05). Data are expressed as the mean ± SD.

Next, to ascertain the GPC3 antigen-specific response of a CTL clone, we examined GPC3 knockdown using siRNA on the GPC3+ HepG2 cell line. Representative data are shown in Figure 5(d–f). The GPC3 expression of HepG2 was clearly decreased by GPC3 siRNA on RT-PCR (Fig. 5d). Specifically, the GPC3 expression of HepG2 was decreased from 24 to 72 h following treatment with GPC3 siRNA on Western blot (Fig. 5e). We examined the IFN-γ production of patient A2-9 CTL clone against HepG2 treated with GPC3 siRNA. The IFN-γ production of the CTL clone was significantly decreased by GPC3 siRNA (P < 0.05) (Fig. 5f). These results indicate that HLA-A2-restricted GPC3144–152 peptide can be processed naturally by cancer cells, and the peptides in the context of HLA-A2 can be expressed on the cell surface of cancer cells in order to be recognized by a GPC3144–152 peptide-specific CTL clone.

Discussion

Salgaller et al.(17) failed to detect dose dependency between 1 and 10 mg in terms of the capacity of gp100 peptide to enhance immunogenicity in humans. Previously, we reported that the peptide emulsified with incomplete Freund’s adjuvant is stable, although the peptide is easily degraded in serum.(16) In this study, as with our previous report using a mouse model,(16) we found that the effect of GPC3144–152 peptide emulsified with incomplete Freund’s adjuvant between 0.3 and 30 mg, to induce specific CTLs, was dose-dependent.

GPC3144–152 (FVGEFFTDV) peptide was previously identified as an HLA-A*02:01-restricted peptide.(15) Moreover, we confirmed by binding assay that the peptide could also bind HLA-A*02:06 and HLA-A*02:07 molecules (data not shown). Therefore, we carried out a clinical trial for three types of HLA-A2 patient. Indeed, similar to HLA-A*02:01 patients, GPC3144–152 peptide-specific CTLs increased after vaccination in both HLA-A*02:06 and HLA-A*02:07 patients (Fig. 1b). These findings suggest that GPC3144–152 peptide is useful for not only HLA-A*02:01 patients but also HLA-A*02:06 and HLA-A*02:07 patients.

Notably, previous reports have shown that vaccination with synthetic peptides occassionally induced ineffective CTL responses due to various underlying mechanisms.(4–9) A possible mechanism is that responding T cells may have a very low affinity such that they recognize only target cells pulsed with high concentrations of the peptide and not tumor cells expressing the relevant epitopes at lower copy numbers. Alternatively, some antigen epitopes were not expressed on the surface of tumor cells.(18,19) When evaluating T-cell response to peptide vaccines, it is important to confirm that responding CTLs lyse human cancer cells. In the present study, although CTL clones established by Dextramer assay could react to HLA-A*02:01+ GPC3+ HCC cell lines, these clones failed to react to the HLA-A*02:01+ GPC3+ melanoma cell line 526mel expressing GPC3 mRNA and protein at a lower level than the HCC cell lines. Therefore, we attempted to establish CTL clones that are more tumor-reactive and with higher avidity than CTL clones established by Dextramer assay. Rubio et al.(20) showed that the surface mobilization of CD107a was useful for identifying and isolating functional tumor-reactive T cells with high recognition efficiency directly from PBMCs of cancer patients after vaccination. In the present study, the CTL clone showing the highest avidity (10−11 M) and tumor reactivity was established by CD107a mobilization assay. Moreover, this clone could also react to 526mel.

For patients with metastatic melanoma, adoptive cell therapy has emerged as the most effective treatment.(21,22) However, tumor-infiltrating lymphocytes with high avidity for tumor antigens can only be generated from some patients with melanoma.(21) Recent studies have shown that genes encoding T-cell receptors (TCRs) can be isolated from high avidity T cells that recognize cancer antigens, and retroviral or lentiviral vectors can be used to redirect lymphocyte specificity to these cancer antigens.(23–26) In the present study, we were able to successfully establish some high avidity CTL clones. We analyzed the TCR β-chain variable region gene families of these clones by RT-PCR and carried out gene sequencing (data not shown). These clones had different TCR genes. Our results raise the possibility that these clones might be applicable to adoptive cell therapy for a large number of HCC patients.

In conclusion, we proved in this study the dose-dependent effects of highly immunogenic GPC3144–152 peptide. Furthermore, we provided substantial evidence that CTLs showing not only high avidity but also natural antigen-specific killing activity against HCC cells could be induced in HCC patients by peptide vaccine.

Acknowledgments

This work was supported in part by Grants-in-Aid for Research on Hepatitis and for Clinical Research from the Ministry of Health, Labour and Welfare, Japan, and a Research Grant of the Princess Takamatsu Cancer Research Fund.

Disclosure Statement

The authors have no conflict of interest.

Ancillary