Clinical Cancer Research Bridging the Lab and the Clinic in Cancer Medicine Infection and Cancer: Biology, Therapeutics, and Prevention
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Komori, H.
Right arrow Articles by Nishimura, Y.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Komori, H.
Right arrow Articles by Nishimura, Y.
Clinical Cancer Research Vol. 12, 2689-2697, May 1, 2006
© 2006 American Association for Cancer Research

Human Cancer Biology

Identification of HLA-A2- or HLA-A24-Restricted CTL Epitopes Possibly Useful for Glypican-3-Specific Immunotherapy of Hepatocellular Carcinoma

Hiroyuki Komori1,2, Tetsuya Nakatsura1, Satoru Senju1, Yoshihiro Yoshitake1, Yutaka Motomura1,2, Yoshiaki Ikuta1,2, Daiki Fukuma1, Kazunori Yokomine1, Michiko Harao1,2, Toru Beppu2, Masanori Matsui3, Toshihiko Torigoe4, Noriyuki Sato4, Hideo Baba2 and Yasuharu Nishimura1

Authors' Affiliations: Departments of 1 Immunogenetics and 2 Gastroenterological Surgery, Graduate School of Medical Sciences, Kumamoto University; 3 Department of Microbiology, Saitama Medical School, Saitama; and 4 Department of Pathology, Sapporo Medical University School of Medicine, Sapporo, Japan

Requests for reprints: Yasuharu Nishimura, Department of Immunogenetics, Graduate School of Medical Sciences, Kumamoto University, Honjo 1-1-1, Kumamoto 860-8556, Japan. Phone: 81-96-373-5310, Fax: 81-96-373-5314; E-mail: mxnishim{at}gpo.kumamoto-u.ac.jp.


   Abstract
Top
 Abstract
Materials and Methods
 Results
 Discussion
 References
 
Purpose and Experimental Design: We previously reported that glypican-3 (GPC3) was overexpressed, specifically in hepatocellular carcinoma (HCC) and melanoma in humans, and it was useful as a novel tumor marker. We also reported that the preimmunization of BALB/c mice with dendritic cells pulsed with the H-2Kd-restricted mouse GPC3298-306 (EYILSLEEL) peptide prevented the growth of tumor-expressing mouse GPC3. Because of similarities in the peptide binding motifs between H-2Kd and HLA-A24 (A*2402), the GPC3298-306 peptide therefore seemed to be useful for the immunotherapy of HLA-A24+ patients with HCC and melanoma. In this report, we investigated whether the GPC3298-306 peptide could induce GPC3-reactive CTLs from the peripheral blood mononuclear cells (PBMC) of HLA-A24 (A*2402)+ HCC patients. In addition, we used HLA-A2.1 (HHD) transgenic mice to identify the HLA-A2 (A*0201)–restricted GPC3 epitopes to expand the applications of GPC3-based immunotherapy to the HLA-A2+ HCC patients.

Results: We found that the GPC3144-152 (FVGEFFTDV) peptide could induce peptide-reactive CTLs in HLA-A2.1 (HHD) transgenic mice without inducing autoimmunity. In five out of eight HLA-A2+ GPC3+ HCC patients, the GPC3144-152 peptide-reactive CTLs were generated from PBMCs by in vitro stimulation with the peptide and the GPC3298-306 peptide-reactive CTLs were also generated from PBMCs in four of six HLA-A24+ GPC3+ HCC patients. The inoculation of these CTLs reduced the human HCC tumor mass implanted into nonobese diabetic/severe combined immunodeficiency mice.

Conclusion: Our study raises the possibility that these GPC3 peptides may therefore be applicable to cancer immunotherapy for a large number of HCC patients.

Hepatocellular carcinoma (HCC) is now spreading rapidly, especially in Asian and Western countries. It is clear that patients with hepatitis B or C–based liver cirrhosis are at high risk for developing HCC (1), and patients with hepatitis treated surgically or by other therapies are also at high risk for recurrence (2). Furthermore, the liver function of these patients is often very poor, so further treatment for recurrence is often restricted. As a result, the prognosis of HCC remains poor, and new therapies for the prevention of cancer development and recurrence, i.e., adjuvant therapy, is urgently needed. As for melanoma, the age-adjusted incidence rates have been increasing in most fair-skinned populations in recent decades (3). In 2005, it is estimated that 59,580 Americans will be diagnosed to have melanoma, and 7,770 will die from the disease (4).

We and others previously reported that glypican-3 (GPC3) was overexpressed in most types of HCC (5–9) and melanoma in humans (8), and we also previously reported that an H-2Kd-restricted antigenic peptide, the mouse GPC3298-306 (EYILSLEEL) peptide, could be recognized by mouse CD8+ CTLs. In addition, these CTLs rejected tumor expressing mouse GPC3 both in vitro and in vivo (10). Because the structural motifs of peptides bound to HLA-A24 (A*2402) and mouse H-2Kd are similar, we investigated whether the GPC3 peptide was also useful as a cancer immunotherapy modality for HLA-A24+ HCC patients. The gene frequency of HLA-A24 (A*2402) is relatively high in Asian populations, especially in the Japanese, whereas it is low in Caucasians. On the other hand, The gene frequency of HLA-A2 (A*0201) is high among various ethic groups, including both Asians and Caucasians (11). Therefore, it is suggested that the HLA-A2-restricted and GPC3-derived CTL epitopes might be very useful for the immunotherapy of many patients with HCC and melanoma all over the world. In the present study, we identified human GPC3-derived CTL epitopes restricted by HLA-A2 using HLA-A2.1 (HHD) transgenic mice (Tgm) and examined whether these HLA-A2 or HLA-A24-restricted epitope peptides could induce GPC3-reactive CTLs from peripheral blood mononuclear cells (PBMC) of patients with HCC.


   Materials and Methods
Top
 Abstract
 Materials and Methods
Results
 Discussion
 References
 
Mouse. HLA-A2.1 (HHD) Tgm; H-2Db–/–ß2m–/– double knockout mice introduced with human ß2m-HLA-A2.1 ({alpha}1 {alpha}2)-H-2Db ({alpha}3 transmembrane cytoplasmic) (HHD) monochain construct gene were generated in the Department SIDA-Retrovirus, Unite d' Immunite Cellulaire Antivirale, Institut Pasteur, France (12, 13) and kindly provided by Dr. F.A. Lemonnier. Nonobese diabetic (NOD)/severe combined immunodeficiency (SCID) female mice at 6 weeks of age were purchased from CLEA Japan (Tokyo, Japan).

Patients, blood samples, and cell lines. Blood samples from patients with HCC were obtained during routine diagnostic procedures after obtaining a formal agreement signed by the patients in Kumamoto University Hospital from April to September 2005. Human liver cancer cell lines, SK-Hep-1 and T2-A0201 (a TAP-deficient and HLA-A*0201-positive cell line; refs. 14, 15), were provided by Kyogo Ito of Kurume University. Human liver cancer cell lines HepG2 and HuH-7 endogenously expressing GPC3, and GPC3– colon cancer cell line SW620, were kindly provided by the Cell Resource Center for Biomedical Research Institute of Development, Aging, and Cancer (Tohoku University, Sendai, Japan). C1R-A*2402 (an HLA-A*2402 transfectant of C1R cells expressing a trace amount of HLA class I molecule; ref. 15) were generous gifts from Dr. Masafumi Takiguchi. The expression of HLA-A2 and HLA-A24 in these cell lines were examined using flow cytometry with an anti-HLA-A2 monoclonal antibody (mAb), BB7.2 and anti-HLA-A24 mAb (One Lambda, Inc., Canoga Park, CA), respectively, in order to select target cell lines for CTL assays. The origins and HLA genotypes of these cell lines have been described elsewhere (16, 17). These cells were maintained in vitro in RPMI 1640 or DMEM supplemented with 10% FCS.

Induction of GPC3-reactive mouse CTLs and IFN-{gamma} enzyme-linked immunospot assay. Human GPC3-derived peptides (purity >90%) sharing the amino acid sequences with mouse GPC3 and carrying binding motifs for HLA-A*0201-encoded molecules, were identified using BIMAS software (BioInformatics and Molecular Analysis Section, Center for Information Technology, NIH, Bethesda, MD) and we purchased a total of nine peptides carrying HLA-A2 (A*0201) binding motifs (Table 1 ) from Biologica (Tokyo, Japan). The immunizations of mice with peptides were done as previously described (7). In brief, bone marrow (BM) cells (2 x 106) from HLA-A2.1 (HHD) Tgm were cultured in RPMI 1640 supplemented with 10% FCS, together with granulocyte macrophage colony-stimulating factor (5 ng/mL) and 2ME (0.8 ng/mL) for 7 days in 10-cm plastic dishes, and these BM-dendritic cells (DC) were pulsed with the mixture of GPC3 peptides carrying HLA-A2 binding motifs (1 µmol/L for each peptide) at 37°C for 2 hours. We primed the HLA-A2.1 (HHD) Tgm with this syngeneic BM-DC vaccine (5 x 105/mouse) into the peritoneal cavity once a week for two weeks. Seven days after the last immunization, the spleens were collected and CD4– spleen cells were isolated by negative selection with anti-CD4 microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany) to exclude any nonspecific IFN-{gamma} production by CD4+ spleen cells cocultured with the BM-DC. The CD4– spleen cells (2 x 106/well) were stimulated with syngeneic BM-DC (2 x 105/well) pulsed with each peptide in vitro. Then, 6 days later, the frequency of cells producing IFN-{gamma}/2 x 104 CD4– spleen cells upon stimulation with syngeneic BM-DC (1 x 104/well), pulsed with or without each peptide, was assayed in an enzyme-linked immunospot (ELISPOT) assay as previously described (18).


View this table:
[in this window]
[in a new window]
  		Table 1. GPC3-derived peptides conserved between human and mouse GPC3 and predicted to be bound to HLA-A2 (A*0201)

 
Induction of GPC3-reactive human CTLs. We isolated PBMCs from the heparinized blood of HLA-A24+ and/or HLA-A2+ Japanese patients with HCC or healthy donors by means of Ficoll-Conray density gradient centrifugation, and peripheral monocyte-derived DCs were generated as described previously (19, 20). CD8+ T cells were isolated using CD8 microbeads (Miltenyi Biotec) from the PBMC of the same donors, and thereafter, peptide-reactive CD8+ CTLs were generated (19, 20). Five days after the last stimulation, the cytotoxic activities of the CTLs were measured by a 51Cr release assay.

CTL responses against cancer cell lines. CTLs were cocultured with each cancer cell line as a target cell (5 x 103/well) at the indicated effector/target ratio and 51Cr release assay was done as described (21). The blocking of HLA-class I or HLA-DR, was done as follows. Before the coculture of CTLs with a cancer cell line in a 51Cr release assay or ELISPOT assay, target cancer cells were incubated for 1 hour with 10 µg/mL anti–class I mAb W6/32 or 10 µg/mL anti–HLA-DR mAb, H-DR-1, and then the effects of mAbs on either the cytotoxic activity or production of IFN-{gamma} by CTLs were examined as reported previously (22).

Histologic and immunohistochemical analysis. Immunohistochemical staining of CD8 or CD4 in tissue specimens of HLA-A2.1 (HHD) Tgm immunized with the GPC3144-152 peptides and the staining of apoptotic cells with terminal deoxynucleotidyl transferase–mediated nick end labeling methods (ApopTag fluorescein in situ apoptosis detection kits; Serologicals Corporation, Norcross, GA) in tumor specimens of patients with HCC were done as described previously (23, 24). In addition, immunohistochemical staining of HLA-class I in HCC tumor tissue specimens were done by using anti-HLA-class I mAb, EMR 8-5.5

Detection by ELISA of the serum-soluble GPC3 protein. Detection of the serum-soluble GPC3 protein was done by an indirect ELISA using the rabbit anti-GPC3 polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA) as described previously (7). We used recombinant human GPC3 protein (R&D Systems Inc., Minneapolis, MN) as a standard, and the presence of >106 ng/mL of serum GPC3 protein was considered to be positive.

Transfer of CTLs to the NOD/SCID mice implanted with a human HCC cell line. The transfer of GPC3-reactive CTLs to the immunodeficient mice implanted with a human HCC cell line was done as described previously (7). Briefly, we s.c. inoculated SK-Hep1/GPC3 cells (1 x 107) positive for both HLA-A2 and HLA-A24 at the right flank of NOD/SCID mice. When the diameter of these tumors reached 5 x 5 mm on day 9 after tumor inoculation into mice, we intravenously injected the mixture of GPC3 peptide-reactive CTL lines or irrelevant HIV peptides; HLA-A2-restricted SLYNTYATL peptide and HLA-A24-restricted RYLRDQQLL peptide, stimulated CD8+ T cells (3 x 106) established from four HLA-A24-positive or two HLA-A2-positive HCC patients, or saline alone. T cells were i.v. injected one more times on day 14. The CD8+ T cells stimulated with HLA-A24-restricted GPC3298-306 peptide or HIV (RYLRDQQLL) peptide and derived from two independent HLA-A24+ HCC patients were mixed, and injected into three NOD/SCID mice on day 9, and the mixture of peptide-stimulated CD8+ T cells from two other HLA-A24+ HCC patients distinct from the T cell donors at the first injection, were injected into the mice on day 14. The HLA-A2-restricted peptide-stimulated CD8+ T cells from one HLA-A2+ HCC patient were also injected into a NOD/SCID mouse on day 9, followed by the injection on day 14 with the peptide-stimulated CD8+ T cells derived from another HLA-A2+ HCC patient.

Statistical analysis. The two-tailed Student's t test was used to evaluate the statistical significance of differences in the data obtained by ELISPOT assay. The statistical significance of the differences in several factors between patients showing a successful CTL induction and other patients was assessed by a {chi}2 test. P < 0.05 was considered to be significant. Statistical analyses were made using the StatView 5.0 software package (Abacus Concepts, Calabasas, CA).


   Results
Top
 Abstract
 Materials and Methods
 Results
Discussion
 References
 
Identification of HLA-A2-restricted CTL epitopes by using HLA-A2.1 (HHD) Tgm. To identify HLA-A2-restricted epitopes by using HLA-A2.1 (HHD) Tgm, we selected nine kinds of peptides having amino acid sequences conserved between human and mouse GPC3 and having high predicted binding scores to HLA-A2 (A*0201; Table 1). CD4– spleen cells from HLA-A2.1 (HHD) Tgm immunized i.p. twice with BM-DCs pulsed with the mixture of these nine peptides were again stimulated in vitro with BM-DCs pulsed with each peptide, and we found that CD4– spleen cells stimulated in vitro with the GPC3144-152 peptide produced the largest amount of IFN-{gamma} in a peptide-specific manner in ELISPOT assays. These CD4– spleen cells (2 x 104/well), showed 36 ± 2.85 spot counts/well, in response to the BM-DCs pulsed with the GPC3144-152 peptide, whereas they showed 23 ± 1.84 spot counts/well in the presence of BM-DCs without peptide loading (P < 0.005) indicating that about (36-23) / 2 x 104 = 0.065% of CD4– spleen cells were reactive to the GPC3 peptide. When we used syngeneic BM-DCs pulsed with a HLA-A2-binding HIV-derived peptide; SLYNTYATL as a control, no significant response (8.84 ± 1.73) was observed. The summation of the diameter of the IFN-{gamma} ELISPOT observed in CD4– spleen cells stimulated with the GPC3144-152 peptide pulsed BM-DCs was 1,878 ± 131 µm, that stimulated with the HIV-derived SLYNTYATL peptide pulsed BM-DCs was 437 ± 77 µm, and that observed in the presence of BM-DC without peptide loading was 762 ± 131 µm (P < 0.001). These assays were done thrice with similar results. As shown in Fig. 1B , the differences in the spot counts (left) or spot diameters (right) between stimulations with peptide pulsed BM-DC and BM-DC without peptide loading clearly revealed the GPC3144-152 peptide-specific response of CD4– spleen cells. As for other peptides, no significant peptide-specific response was observed. These results suggest that the GPC3144-152 peptide could be a CTL epitope peptide in HLA-A2.1 (HHD) Tgm, and we also expected this GPC3144-152 peptide to be an epitope for human CTLs.


Figure 1
View larger version (18K):
[in this window]
[in a new window]
  		Fig. 1. Identification of HLA-A2-restricted CTL epitopes of GPC3 by using HLA-A2.1 (HHD) Tgm. A, protocol for identification of GPC3-derived and HLA-A2-restricted CTL epitopes. We primed the HLA-A2.1 (HHD) Tgm with BM-DCs (5 x 105) pulsed with the mixture of GPC3-derived peptides carrying HLA-A2 (A*0201) binding motif into the peritoneal cavity once a week for two weeks. Seven days after the last DC vaccination, spleens were collected and CD4– spleen cells (2 x 106/well) were stimulated with syngeneic BM-DCs (2 x 105/well) pulsed with each peptide in vitro for 6 days. We used these cultured CD4– spleen cells as responder cells in ELISPOT assay to evaluate GPC3-specific response of CTLs. B, bar graph, IFN-{gamma} ELISPOT counts/2 x 104 CD4– spleen cells cocultured with peptide pulsed BM-DCs subtracted with those cocultured with BM-DCs without peptide loading (left). Bar graph, summation of IFN-{gamma} ELISPOT diameters/2 x 104 CD4– spleen cells cocultured with peptide-pulsed BM-DCs subtracted with those cocultured with BM-DCs without peptide loading (right). Columns, mean of triplicate assays; bars, SE. All assays were done thrice with similar results.

 
The immunization of the HLA-A2-restricted peptide, GPC3144-152, did not induce autoimmunity in HLA-A2.1 (HHD) Tgm. It is well known that melanocyte-differentiation antigens such as MART-1 or gp100 are very useful for immunotherapy of melanoma patients, but they sometimes cause autoimmunity, such as vitiligo or uveitis, following vaccination. We previously reported that the immunization of the GPC3298-306 peptide did not cause autoimmunity in BALB/c mouse (9). To investigate whether the immunization of mice with HLA-A2-restricted GPC3-derived peptides causes autoimmunity, the immunohistochemical staining of several organs with anti-CD4 and anti-CD8 mAb was done in HLA-A2.1 (HHD) Tgm immunized with a mixture of nine GPC3 peptides 7 days before the analysis. As shown in Fig. 2 , we could not find any pathologic changes, such as lymphocyte infiltration or tissue destruction and repair in skin, lung, brain, heart, liver, and kidney of HLA-A2.1 (HHD) Tgm. The same result was also observed when mice were vaccinated with the GPC3144-152 peptide alone (n = 3; data not shown). These results indicate that the GPC3144-152 peptide-reactive CD8+ CTLs do not attack the normal tissue specimens that we investigated.


Figure 2
View larger version (32K):
[in this window]
[in a new window]
  		Fig. 2. Immunohistochemical staining with anti-CD4 or anti-CD8 mAb in tissue specimens of HLA-A2.1 (HHD) Tgm immunized with the GPC3 144-152 peptides. These tissue specimens were removed and analyzed 7 days after the second DC vaccination (original magnification, x200).

 
Induction of GPC3-reactive CTLs from PBMCs of HLA-A2- or HLA-A24-positive HCC patients. We evaluated the cytotoxic activity of CTLs that were induced with the GPC3298-306 or GPC3144-152 peptide from PBMCs isolated from HCC patients. PBMCs were isolated from HCC patients positive for HLA-A24 and/or HLA-A2, and CD8+ T cells sorted from the PBMCs were cocultured with autologous monocyte-derived DCs pulsed with each peptide as described in Materials and Methods. CTLs from PBMCs of HLA-A2+ HCC patients stimulated with the GPC3144-152 peptide or CTLs from PBMCs of HLA-A24+ HCC patients stimulated with the GPC3298-306 peptide exhibited cytotoxicity against peptide-pulsed target cells. The representative data of CTLs restricted by HLA-A2 or HLA-A24 were shown in Fig. 3A . The CTLs induced from PBMCs of patient A2-8 showed cytotoxic activity to T2-A0201 cells (HLA-A2+) pulsed with the GPC3144-152 peptide, but not to T2-A0201 cells without peptide loading by 51Cr release assay. The CTLs induced from PBMCs of patient A24-12 exhibited cytotoxic activity to the C1R-A*2402 cells (HLA-A24+) pulsed with the GPC3298-306 peptide, but not to C1R-A*2402 cells without peptide loading. These results indicate that these CTLs had peptide-specific cytotoxicity. Other CTLs induced from the nine patients A2-1, A2-2, A2-3, A2-4, A24-1, A24-3, A24-4, A24-6, and A24-7 similarly exhibited peptide-specific cytotoxicity against peptide-pulsed target cells (data not shown).


Figure 3
View larger version (23K):
[in this window]
[in a new window]
  		Fig. 3. CTL induction from PBMCs of HLA-A2- or HLA-A24-positive HCC patients. A and B, GPC3 peptide-reactive CTLs were generated from CD8+ T cells of HLA-A2+ and/or HLA-A24+ HCC patients. After three or four stimulations with autologous monocyte-derived DCs pulsed with the GPC3144-152 or GPC3298-306 peptide, the CTLs were subjected to a standard 51Cr release assay at the indicated effector/target ratio. Their cytotoxicity against the GPC3298-306 peptide pulsed C1R-A2402 cells or T2-A0201 cells, and each unpulsed cells (A), or GPC3– HLA-A2+, HLA-A24+ HCC cell line SK-Hep-1, GPC3– HLA-A2+, HLA-A24+colon cancer cell line SW620, and those cell lines transfected with the human GPC3 gene; SK-Hep-1/GPC3 or SW620/GPC3 (B) were examined by a 51Cr release assay. C and D, GPC3+ HLA-A2+, HLA-A24+ HCC cell line HepG2, GPC3+ HLA-A2–, HLA-A24– HCC cell line HuH-7, and GPC3– tumor cell lines SW620 and SK-Hep1 were used as target cells (left). Points, percentage of specific lysis calculated based on the mean values of a triplicate assay. D, inhibition of cytotoxicity by anti-HLA class I mAb (right). After the target HepG2 cells were incubated with anti-HLA class I mAb (W6/32, IgG2a) or anti-HLA DR mAb (H-DR-1, IgG2a), respectively, for 1 hour, the CTLs generated from PBMCs of patient A2-8 by stimulation with GPC3144-152 peptide (top) or CTLs generated from patient A24-6 using the GPC3298-306 peptide (bottom) were added. IFN-{gamma} production (top; IFN-{gamma} ELISPOT assay) and cytotoxicity (bottom; 51Cr release assay) were markedly inhibited by W6/32, but not by H-DR-1.

 
Furthermore, we used GPC3 transfectants, SK-Hep1/GPC3 (GPC3+, HLA-A2+, HLA-A24+) or SW620/GPC3 (GPC3+, HLA-A2+, HLA-A24+) as target cells and examined whether we could find GPC3-specific cytotoxic activity of CTLs. As shown in Fig. 3B, the CTLs induced from PBMCs of patient A2-3 by stimulation with the GPC3144-152 peptide showed specific cytotoxicity against SK-Hep1/GPC3, but not against GPC3-negative SK-Hep1. Similarly, the GPC3298-306 peptide–induced CTLs showed specific cytotoxicity against SW620/GPC3 in patient A24-7 or against SK-Hep1/GPC3 in patient A24-12, but not against SK-Hep1 or SW620, respectively, which did not endogenously express GPC3. These findings indicate that these peptides can be processed naturally in cancer cells, and the peptides in the context of HLA-A2 or HLA-A24 can be expressed on the cell surface of cancer cells to be recognized by the CTLs.

When we think about the application of GPC3 to cancer immunotherapy, the most important point is that these GPC3-reactive CTLs can exhibit specific cytotoxicity to the tumors endogenously expressing GPC3. We thus investigated whether these CTLs could kill human HCC cell lines expressing both endogenous GPC3 and the restriction HLA class I molecules. As shown in Fig. 3C, we could generate GPC3-reactive CTLs by stimulation with the GPC3144-152 peptide and these CTLs exhibited cytotoxic activity to HepG2 (GPC3+, HLA-A2+, and HLA-A24+), but not to HuH-7 (GPC3+, HLA-A2–, and HLA-A24–) or SW620 (GPC3–, HLA-A2+, and HLA-A24+) in patients A2-1, A2-3, and A2-4. Similarly, we could generate GPC3-reactive CTLs by stimulation of PBMCs with the GPC3298-306 peptide and these CTLs exhibited cytotoxic activity to HepG2, but not to HuH-7 or SK-Hep-1 (GPC3–, HLA-A2+, HLA-A24+) in patients A24-4, A24-7, and A24-12.

In an HLA-class I blocking experiment, anti-HLA class I mAb W6/32 markedly inhibited the IFN-{gamma} production stimulated with HepG2 cells in ELISPOT assay of CTLs generated from patient A2-8 by stimulation with the GPC3144-152 peptide (Fig. 3D, top), and inhibited cytotoxic activity against HepG2 cells in 51Cr release assay of CTLs generated from patient A24-6 by stimulation with the GPC3298-306 peptide (Fig. 3D, bottom), but anti-HLA-DR mAb, H-DR-1 did not inhibit the response of CTLs. These results clearly indicate that these CTLs recognized HepG2 in a HLA-class I–restricted manner.

As shown in Table 2 , we could induce GPC3-reactive CTLs from PBMCs in ~50% of either the HLA-A2– or HLA-A24-positive HCC patients. In patients A2-6, A24-5, A24-9, and A24-11 who did not express GPC3 in tumor tissues, GPC3-reactive CTLs could not be induced from their PBMCs. Among eight HLA-A2-positive HCC patients who expressed GPC3 in HCC tissue or produced soluble GPC3 in sera, patients A2-1, A2-2, A2-3, A2-4, A2-6, A2-7, A2-9, and A2-10, GPC3-reactive CTLs could be generated from the PBMCs of only four patients (50%). In patient A2-6, GPC3 was detected only in the serum but not in HCC tumor tissue. It was thought to be possible that the majority of GPC3 protein was secreted away in this type of HCC cell as described previously (7). Among six HLA-A24-positive patients who expressed GPC3 in tumor tissue, patients A24-1, A24-2, A24-3, A24-6, A24-10, and A24-12, GPC3-reactive CTLs could be generated from the PBMCs of only four patients (67%). We also examined whether it was possible to induce GPC3-specific CTLs from PBMCs isolated from healthy donors (each HLA type, n = 3), but we failed to generate GPC3-specific and HLA-A2- or HLA-A24-restricted CTLs even though PBMCs were stimulated with the peptides thrice in vitro (data not shown). These results suggest that GPC3-reactive CTLs could only be induced in patients who expressed GPC3 in tumor tissue, thus, indicating the existence of GPC3-reactive CTL precursors in patients with GPC3+ HCC. We also examined whether GPC3-reactive CTLs could be generated more frequently from PBMCs isolated from HCC patients positive for serum-soluble GPC3. As shown in Table 2, the presence of serum-soluble GPC3 did not correlate statistically with the successful induction of GPC3-reactive CTLs. As a result, we could not observe the enhancement of CTL induction efficiency via possible antigen presentation of soluble serum GPC3 through HLA-class II pathways to CD4+ T cells or cross-presentation through the HLA class I pathway to CD8+ T cells (25, 26) in patients positive for serum GPC3.


View this table:
[in this window]
[in a new window]
  		Table 2. Expression of GPC3 in HCC tissue, quantification of serum-soluble GPC3, and GPC3-specific CTL induction in HCC patients

 
Inoculation of the GPC3 peptide-induced CTLs reduced growth of a GPC3+ human HCC tumor cell line implanted into NOD/SCID mouse. To investigate the effects of GPC3 peptide-reactive CTL inoculation into the mice implanted with the GPC3+ human HCC cell line, we s.c. inoculated SK-Hep1/GPC3 cell lines positive for both HLA-A2 and HLA-A24 into NOD/SCID mice, and i.v. injected the mixture of CTLs generated from several HCC patients positive for HLA-A2 or HLA-A24 into mice implanted with SK-Hep1/GPC3 when the diameter of these tumors reached 5 x 5 mm in size as described in Materials and Methods. The CTLs injected into mice were prepared by stimulating peripheral blood CD8+ T cells with HLA-A2- or HLA-A24-restricted GPC3-epitope peptides or control-irrelevant HIV peptides as described in Materials and Methods. The tumor sizes of four individual mice in each group (Fig. 4A ) and mean ± SD of tumor sizes in each group (Fig. 4B) were evaluated. After 5 days from the second inoculation of GPC3 peptide-reactive CTLs, the tumor size of SK-Hep1/GPC3 was apparently reduced in comparison to the size of tumor mass implanted into NOD/SCID mice injected with control T cells or saline alone (P < 0.01). These results clearly indicate the efficacy of adoptive GPC3 peptide-reactive CTL transfer therapy for GPC3+ tumor in mice.


Figure 4
View larger version (18K):
[in this window]
[in a new window]
  		Fig. 4. Marked inhibition of growth of a GPC3-transfected human HCC cancer cell line, SK-Hep1/GPC3, engrafted into NOD/SCID mice after adoptive transfer of human CTLs induced by the GPC3 peptides. A, when tumor size reached 25 mm2 on day 9 after s.c. tumor implantation, human CTLs (3 x 106) reactive to HLA-A2-restricted ({blacksquare}) GPC3 peptide and generated from one HLA-A2+ donor, or those reactive to HLA-A24-restricted ({square}) GPC3 peptide and pooled from two HLA-A24+ donors were i.v. inoculated. On day 14, the inoculation of CTLs generated from the donors distinct from those at the first injection was repeated. The control CD8+ T cells stimulated with irrelevant HLA-A2-restricted ({blacktriangleup}) or HLA-A24-restricted ({triangleup}) HIV peptides were also injected into mice as a control. Tumor volumes in NOD/SCID mice given twice on days 9 and 14 with GPC3 epitope peptide–induced CTL lines (n = 4), control CD8+ T cells (n = 4), or saline alone (n = 4). Tumor size was expressed in square millimeters. B, points, mean tumor sizes in each group of mice; bars, ±SD (n = 4). Statistical significance was evaluated using t test.

 

   Discussion
Top
 Abstract
 Materials and Methods
 Results
 Discussion
References
 
In this article, we identified HLA-A24-restricted or HLA-A2-restricted GPC3 CTL epitope peptides, and found that GPC3-reactive CTLs could be generated from PBMCs stimulated with these peptides in ~50% of HCC patients. Vaccination based on these peptides did not induce autoimmunity in HLA-A2.1 (HHD) Tgm of a B57Bl/6 background. We previously identified the GPC3298-306 peptide to be a CTL epitope in BALB/c mouse, and we expected that this GPC3 peptide might also be present in human CTL in a HLA-A24-restricted manner. As expected, we could generate HLA-A24-restricted and the GPC3298-306 peptide-reactive human CTLs in this study. As a result, BALB/c mice may be useful for identifying HLA-A24-restricted CTL epitopes. HLA-A2.1 (HHD) Tgm was reported to be a versatile animal model for the preclinical evaluation of peptide-based immunotherapy (12, 13). We could also find its usefulness for the identification of HLA-A2-restricted antigenic epitope in this study.

In this study, we wanted to identify the most effective major CTL epitopes derived from GPC3. As a result, we used BM-DCs derived from HLA-A2.1 (HHD) Tgm and pulsed BM-DCs with the mixture of GPC3 peptides for the vaccination of mice. Some of the peptides tested stimulated the weak response of CTLs in an ELISPOT assay, and these peptides might also be useful for future analysis. It was recently reported that peptides having a weak affinity to MHC, which could not be predicted by a BIMAS system, could induce peptide-reactive CTLs with a cytotoxic activity (27). To search for more peptides that can be applicable for immunotherapy, it may be necessary to check these minor CTL epitopes in the future. In this study, the GPC3-derived peptides predicted to have high binding affinity to HLA-A2 molecules and having the amino acid sequences conserved between human and mouse GPCs were selected for the analysis. When we analyzed the amino acid sequence of human GPC3 protein, all of the top 28 human GPC3 peptides having high binding scores (>100) to HLA-A2 molecules shared the same amino acid sequences with mouse GPC3. Therefore, it is unlikely that we excluded many candidates of human GPC3-derived and HLA-A2-restricted CTL epitopes from the analysis by selecting the peptides having amino acid sequences shared between human and mouse GPC3. Furthermore, we have to consider the differences in the T cell repertoire in mice and humans. Thereby, we may miss GPC3 peptides recognized by human CTLs but not by mouse CTLs.

Considering ideal immunogenic target molecules for cancer immunotherapy, it is important to select a tumor antigen that could not be lost by tumor cells through immunoediting (28, 29). Recently, Capurro et al. reported that GPC3 is involved in the carcinogenesis and proliferation of HCC via regulation of noncanonical Wnt signals (30). Therefore, it may be possible that tumor cells cannot lose the GPC3 expression in order to continue to grow. Furthermore, according to an immunohistochemical analysis of the expression of HLA-class I molecules using newly developed specific mAb, EMR 8-5,5 we found that almost all HCC cells expressed HLA-class I as far as we could examine (Table 2). For these reasons, we think that GPC3 is a very useful candidate as a target tumor antigen for the immunotherapy of HCC. We and others previously reported that the expression of GPC3 in HCC was detected from an early stage and the quantification of the soluble GPC3 protein in sera was useful for a diagnosis of HCC at an early stage (5, 7). As a result, GPC3-based immunotherapy might be able to prevent the appearance of HCC in patients with hepatitis B or C–based liver cirrhosis.

In this study, we found that it is possible to induce GPC3-reactive CTLs by the stimulation of PBMCs with the two major GPC3 epitopes in vitro in 50% of the HCC patients having an appropriate HLA-class I allele. However, it is necessary to investigate more patients to estimate the probability of a successful induction of GPC3-reactive CTLs in HCC patients. We intended to know whether there was any correlation between successful induction of GPC3 peptide-reactive CTLs and prognosis or CTL infiltration into tumor tissue of these patients, therefore, we investigated the seven index cases; patients A2-10, A24-1, A24-2, A24-4, A24-9, A24-11, and A24-12, to see whether there was any correlation between successful induction of GPC3 peptide-reactive CTLs and prognosis or CTL infiltration into the tumor tissue of these patients. In three patients, A24-1, A24-4, and A24-12, who could generate GPC3 peptide-reactive CTLs, patient A24-12 recurred at 6 months after operation. In four patients, A2-10, A24-2, A24-9, and A24-11, who failed to induce GPC3-peptide-reactive CTLs, patient A24-9, whose HCC did not express GPC3, recurred at 6 months after operation, and patient A24-2 recurred at 3 months after operation and died 3 months after recurrence. These three recurred patients had extremely strong tumor invasion to the vasculature. Therefore, it was difficult to evaluate the correlation between the positive CTL response and clinical improvement at the present stage, and we have to increase the number of patients investigated and to do further statistical analyses on these relationships. In patients who could be examined for the infiltration of CD8-positive cells into their tumor specimens and for the existence of terminal deoxynucleotidyl transferase–mediated nick end labeling–positive cells in tumor tissue, patients A2-10, A24-1, A24-2, and A24-9, there was no strong correlation between the positive GPC3 peptide-reactive CTL response and for the existence of CD8-positive or terminal deoxynucleotidyl transferase–mediated nick end labeling–positive cells in the tumor tissues (data not shown). As shown in Fig. 4, we observed a regression of the tumor masses in NOD/SCID mice implanted with SK-Hep1/GPC3 and transferred i.v. with the GPC3 peptide-reactive CTLs in comparison to the mice injected with control CD8+ T cells or saline alone. Although the regression of tumor growth was observed for 2 weeks after the second transfer of CTLs, the tumors began to enlarge again after that period. We thought it was important to continue the transfer of CTLs again and again to obtain continuous regression of the GPC3-expressing tumor. These data suggest that the adoptive i.v. transfer of GPC3-reactive CTLs into mice bearing GPC3+ tumors was useful to inhibit tumor growth in the mouse tumor model.

In addition, it is most important to confirm the usefulness of GPC3-specific in vivo cancer immunotherapy in patients with HCC. Investigation of the presence of GPC3-specific CTLs in patients with melanoma are also eagerly awaited. We previously reported that DC differentiated in vitro from mouse embryonic stem cells transfected with the mouse GPC3 gene (24, 31) induced protective immunity against mouse melanoma cell line B16 F10 (32). We are now preparing a translational study of GPC3-based immunotherapy to reduce the risk of recurrence in HCC patients treated surgically. We will try to use the GPC3 epitope peptides identified in this study first, whereas in the second phase, we will make a trial of the peptide-pulsed DC vaccine. We expect that GPC3-based immunotherapy may be a novel treatment strategy that could potentially help to prevent the appearance, advance, and/or recurrence of HCC and melanoma.


   Acknowledgments
 
We thank Dr. Masafumi Takiguchi (Kumamoto University, Kumamoto, Japan), Kyogo Itoh (Kurume University, Kurume, Japan), and the Cell Resource Center for Biomedical Research Institute of Development, Aging and Cancer, Tohoku University for providing the cell lines; Dr. F.A. Lemonnier. (Dapartment SIDA-Retrovirus, Unite d' Immunite Cellulaire Antivirale, Institut Pasteur, France) for providing HLA-A2.1 (HHD) transgenic mice; Tatsuko Kubo (Department of Molecular Pathology, Kumamoto University) for technical assistance with immunohistochemical analyses, and Dr. Mikio Monji (Department of Immunogenetics, Graduate School of Medical Sciences, Kumamoto University) for helpful comments on technical procedures.


   Footnotes
 
Grant support: Grants-in-Aid 12213111 and 17015035 from the Ministry of Education, Science, Technology, Sports, and Culture, Japan, a Research Grant for Health Sciences from the Ministry of Health, Labor and Welfare, Japan, and by funding from Kirin Brewery Co., Oncotherapy Science Co., Eisai Pharmasential Co., and the Sagawa Foundation for Promotion of Cancer Research and Kumamoto Technology & Industry Foundation.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Note: T. Nakatsura is currently in the Immunotherapy Section, Investigative Treatment Division, Center for Innovative Medicine, National Cancer Center East, Kashiwa, Japan.

5 T. Torigoe, et al. Immunohistochemical analysis of HLA class I expression in tumor tissues revealed unusually high frequency of down-regulation in breast cancer tissues submitted. Back

Received 10/20/05; revised 2/28/06; accepted 3/ 2/06.


   References
Top
 Abstract
 Materials and Methods
 Results
 Discussion
 References
 

  1. Schafer DF, Sorrell MF. Hepatocellular carcinoma. Lancet 1999;353:1253–7.[CrossRef][Medline]
  2. Tung-Ping Poon R, Fan ST, Wong J. Risk factors, prevention, and management of postoperative recurrence after resection of hepatocellular carcinoma. Ann Surg 2000;232:10–24.[CrossRef][Medline]
  3. Tsao H, Atkins MB, Sober AJ. Management of cutaneous melanoma. N Engl J Med 2004;351:998–1012.[Free Full Text]
  4. Jemal A, Murray T, Ward E, et al. Cancer statistics. CA Cancer J Clin 2005;55:10–30.[Abstract/Free Full Text]
  5. Hippo Y, Watanabe K, Watanabe A, et al. Identification of soluble NH2-terminal fragment of glypican-3 as a serological marker for early-stage hepatocellular carcinoma. Cancer Res 2004;64:2418–23.[Abstract/Free Full Text]
  6. Capurro M, Wanless IR, Sherman M, et al. Glypican-3: a novel serum and histochemical marker for hepatocellular carcinoma. Gastroenterology 2003;125:89–97.[CrossRef][Medline]
  7. Nakatsura T, Yoshitake Y, Senju S, et al. Glypican-3, overexpressed specifically in human hepatocellular carcinoma, is a novel tumor marker. Biochem Biophys Res Commun 2003;306:16–25.[CrossRef][Medline]
  8. Nakatsura T, Kageshita T, Ito S, et al. Identification of glypican-3 as a novel tumor marker for melanoma. Clin Cancer Res 2004;10:6612–21.[Abstract/Free Full Text]
  9. Zhu ZW, Friess H, Wang L, et al. Enhanced glypican-3 expression differentiates the majority of hepatocellular carcinomas from benign hepatic disorders. Gut 2001;48:558–64.[Abstract/Free Full Text]
  10. Nakatsura T, Komori H, Kubo T, et al. Mouse homologue of a novel human oncofetal antigen, glypican-3, evokes T-cell-mediated tumor rejection without autoimmune reactions in mice. Clin Cancer Res 2004;10:8630–40.[Abstract/Free Full Text]
  11. Browning M, Krausa P. Genetic diversity of HLA-A2: evolutionary and functional significance. Immunol Today 1996;17:165–70.[CrossRef][Medline]
  12. Pascolo S, Bervas N, Ure JM, Smith AG, Lemonnier FA, Perarnau B. HLA-A2.1-restricted education and cytolytic activity of CD8(+) T lymphocytes from ß2 microglobulin (ß2m) HLA-A2.1 monochain transgenic H-2Db ß2m double knockout mice. J Exp Med 1997;185:2043–51.[Abstract/Free Full Text]
  13. Firat H, Garcia-Pons F, Tourdot S, et al. H-2 class I knockout, HLA-A2.1-transgenic mice: a versatile animal model for preclinical evaluation of antitumor immunotherapeutic strategies. Eur J Immunol 1999;29:3112–21.[CrossRef][Medline]
  14. Henderson RA, Michel H, Sakaguchi K, et al. HLA-A2.1-associated peptides from a mutant cell line: a second pathway of antigen presentation. Science 1992;255:1264–6.[Abstract/Free Full Text]
  15. Karaki S, Kariyone A, Kato N, Kano K, Iwakura Y, Takiguchi M. HLA-B51 transgenic mice as recipients for production of polymorphic HLA-A, B-specific antibodies. Immunogenetics 1993;37:139–42.[Medline]
  16. Wadee AA, Paterson A, Coplan KA, Reddy SG. HLA expression in hepatocellular carcinoma cell lines. Clin Exp Immunol 1994;97:328–33.[Medline]
  17. Matsui M, Machida S, Itani-Yohda T, Akatsuka T. Downregulation of the proteasome subunits, transporter, and antigen presentation in hepatocellular carcinoma, and their restoration by interferon-{gamma}. J Gastroenterol Hepatol 2002;17:897–907.[CrossRef][Medline]
  18. Bourgault Villada I, Moyal Barracco M, Ziol M, et al. Spontaneous regression of grade 3 vulvar intraepithelial neoplasia associated with human papillomavirus-16-specific CD4(+) and CD8(+) T-cell responses. Cancer Res 2004;64:8761–6.[Abstract/Free Full Text]
  19. Yoshitake Y, Nakatsura T, Monji M, et al. Proliferation potential-related protein, an ideal esophageal cancer antigen for immunotherapy, identified using complementary DNA microarray analysis. Clin Cancer Res 2004;10:6437–48.[Abstract/Free Full Text]
  20. Monji M, Nakatsura T, Senju S, et al. Identification of a novel human cancer/testis antigen, KM-HN-1, recognized by cellular and humoral immune responses. Clin Cancer Res 2004;10:6047–57.[Abstract/Free Full Text]
  21. Makita M, Hiraki A, Azuma T, et al. Antilung cancer effect of WT1-specific cytotoxic T lymphocytes. Clin Cancer Res 2002;8:2626–31.[Abstract/Free Full Text]
  22. Gomi S, Nakao M, Niiya F, et al. A cyclophilin B gene encodes antigenic epitopes recognized by HLA-A24-restricted and tumor-specific CTLs. J Immunol 1999;163:4994–5004.[Abstract/Free Full Text]
  23. Kai M, Nakatsura T, Egami H, Senju S, Nishimura Y, Ogawa M. Heat shock protein 105 is overexpressed in a variety of human tumors. Oncol Rep 2003;10:1777–82.[Medline]
  24. Matsuyoshi H, Senju S, Hirata S, Yoshitake Y, Uemura Y, Nishimura Y. Enhanced priming of antigen-specific CTLs in vivo by embryonic stem cell-derived dendritic cells expressing chemokine along with antigenic protein: application to antitumor vaccination. J Immunol 2004;172:776–86.[Abstract/Free Full Text]
  25. Guermonprez P, Valladeau J, Zitvogel L, Thery C, Amigorena S. Antigen presentation and T cell stimulation by dendritic cells. Annu Rev Immunol 2002;20:621–67.[CrossRef][Medline]
  26. Thomas AM, Santarsiero LM, Lutz ER, et al. Mesothelin-specific CD8(+) T cell responses provide evidence of in vivo cross-priming by antigen-presenting cells in vaccinated pancreatic cancer patients. J Exp Med 2004;200:297–306.[Abstract/Free Full Text]
  27. Bredenbeck A, Losch FO, Sharav T, et al. Identification of noncanonical melanoma-associated T cell epitopes for cancer immunotherapy. J Immunol 2005;174:6716–24.[Abstract/Free Full Text]
  28. Kawakami Y, Rosenberg SA. Human tumor antigens recognized by T-cells. Immunol Res 1997;16:313–39.[Medline]
  29. Tsuboi A, Oka Y, Udaka K, et al. Enhanced induction of human WT1-specific cytotoxic T lymphocytes with a 9-mer WT1 peptide modified at HLA-A*2402-binding residues. Cancer Immunol Immunother 2002;51:614–20.[CrossRef][Medline]
  30. Capurro MI, Xiang YY, Lobe C, Filmus J. Glypican-3 promotes the growth of hepatocellular carcinoma by stimulating canonical Wnt signaling. Cancer Res 2005;65:6245–54.[Abstract/Free Full Text]
  31. Senju S, Hirata S, Matsuyoshi H, et al. Generation and genetic modification of dendritic cells derived from mouse embryonic stem cells. Blood 2003;101:3501–8.[Abstract/Free Full Text]
  32. Motomura Y, Senju S, Nakatsura T, et al. Embryonic stem cell-derived dendritic cells expressing Glypican-3, a recently identified oncofetal antigen, induce protective immunity against highly metastatic mouse melanoma, B16–10. Cancer Res 2006;66:2414–22.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 J. Immunol.Home page
P. Blancou, R. Mallone, E. Martinuzzi, S. Severe, S. Pogu, G. Novelli, G. Bruno, B. Charbonnel, M. Dolz, L. Chaillous, et al.
Immunization of HLA Class I Transgenic Mice Identifies Autoantigenic Epitopes Eliciting Dominant Responses in Type 1 Diabetes Patients
J. Immunol., June 1, 2007; 178(11): 7458 - 7466.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Komori, H.
Right arrow Articles by Nishimura, Y.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Komori, H.
Right arrow Articles by Nishimura, Y.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online