Identification of a Gene Coding for a Protein Possessing Shared Tumor Epitopes Capable of Inducing HLA-A24-Restricted Cytotoxic T Lymphocytes in Cancer Patients1

[CANCER RESEARCH 59, 4056–4063, August 15, 1999] Identification of a Gene Coding for a Protein Possessing Shared Tumor Epitopes Capable of Inducing HLA-A24-restricted Cytotoxic T Lymphocytes in Cancer Patients1

Damu Yang, Masanobu Nakao, Shigeki Shichijo, Teruo Sasatomi, Hideo Takasu, Hajime Matsumoto, Kazunori Mori, Akihiro Hayashi, Hideaki Yamana, Kazuo Shirouzu, and Kyogo Itoh2 Cancer Vaccine Development Division, Kurume University Research Center for Innovative Cancer Therapy [D. Y., M. N., K. I.], and Departments of Surgery [A. H., H. Y., K. S.], Immunology [S. S., T. S., H. T., H. M., K. I.], and Otolaryngology [K. M.], Kurume University School of Medicine, Kurume, 830-0011, Japan

ABSTRACT we report a gene encoding epitopes that are capable of inducing CTLs in PBMCs of patients with SCCs and adenocarcinomas. Genes encoding tumor epitopes that are capable of inducing CTLs against adenocarcinomas and squamous cell carcinomas, two major human cancers histologically observed in various organs, have rarely been MATERIALS AND METHODS identified. Here, we report a new gene from cDNA of esophageal cancer cells that encodes a shared tumor antigen recognized by HLA-A2402- Generation of HLA-A2402-restricted CTLs. HLA-A2402-restricted and restricted and tumor-specific CTLs. The sequence of this gene is almost tumor-specific CTLs were established from the PBMCs of an esophageal identical to that of the KIAA0156 gene, which has been registered in cancer patient (HLA-A2402/A2601) by the standard method of mixed lymphocyte tumor cell culture, as reported previously (29). Briefly, the patient’s GenBank with an unknown function. This gene encodes a Mr 140,000 protein that is expressed in the nucleus of all of the malignant tumor cell PBMCs were repeatedly stimulated with the autologous tumor cell line (KE4) lines tested and the majority of cancer tissues with various histologies, in the culture medium [45% RPMI 1640, 45% AIM-V medium (Life Tech- including squamous cell carcinomas, adenocarcinomas, melanomas, and nologies, Inc., Grand Island, NY), and 10% FCS (Equitech Bio, Ingram, TX) leukemia cells. However, this protein was undetectable in the nucleus of with 100 units/ml interleukin 2 (Shiongi Pharmaceutical Co., Osaka, Japan)]. any cell lines of nonmalignant cells or normal tissues, except for the testis. The CTLs were tested for cytotoxicity to various cancer cells, EBV-trans- 51 Furthermore, this protein was expressed in the cytosol of all of the formed B cells, and normal cells in a 6-h Cr release assay, as reported proliferating cells, including normal cells and malignant cells, but not in previously (29), at various E:T ratios. normal tissues, except for the testis and fetal liver. Two peptides of this Identification of the Clone 13 Gene. Previously reported expression gene cloning methods were used in this study to identify a gene coding for tumor protein were recognized by HLA-A2402-restricted CTLs and were able to ϩ induce HLA-A24-restricted and tumor-specific CTLs from peripheral antigen recognized by CTLs (25). In brief, poly(A) RNA of the KE4 tumor ؉ cells was converted to cDNA, ligated to the SalI adapter, and inserted into the blood mononuclear cells of most of HLA-A24 cancer patients tested, but expression vector pSV-SPORT-1 (Life Technologies, Inc., Gaithersburg, MD). not from peripheral blood mononuclear cells of any healthy donors. These -؉ cDNA of HLA-A2402 or HLA-A0201 was obtained by reverse transcription peptides may be useful in specific immunotherapy for HLA-A24 cancer PCR and was cloned into the eukaryotic expression vector pCR3 (Invitrogen, patients with various histological types. San Diego, CA). Both 200 ng of plasmid DNA pools or clones of the KE4 cDNA library and 200 ng of the HLA-A2402 cDNA were mixed with 1 ␮lof Lipofectin in 70 ␮l of Opti-MEM (Life Technologies, Inc.) for 15 min. Thirty INTRODUCTION ␮l of the mixture were then added to the VA13 (2 ϫ 104) cells and incubated for 5 h. Next, 200 ␮l of RPMI 1640 containing 10% FCS were added and Many genes encoding tumor antigens and peptides that are recog- cultured for 2 days followed by the addition of CTLs (104 cells/well). After an nized by CTLs have been identified from melanoma cDNA (1–15). 18-h incubation, 100 ␮l of supernatant were collected to measure IFN-␥ by an Some of these peptides can induce CTLs that recognize tumor cells ELISA kit in a duplicate assay, as reported previously (25). This concentration from PBMCs3 of melanoma patients (16–20). Immunotherapy with (100 ng/ml) of HLA-A2402 cDNA was chosen based on the fact that expres- some of the peptides that are capable of inducing HLA class I- sion levels of HLA-A24 antigens on the surface of VA13 cells transfected with restricted CTLs has been shown to result in tumor regression in 50, 100, 200, and 400 ng/ml of HLA-A2402 cDNA were 35, 42, 40, and 20% ϩ by FACScan analysis with anti-HLA-A24 mAb, respectively. DNA sequenc- HLA-A0201 melanoma patients (21, 22). These results indicate that ing was performed with dideoxynucleotide sequencing method using a DNA identification of the peptides capable of inducing CTLs may provide Sequence kit (Perkin-Elmer Corp., Foster, CA), and the sequence was analyzed a new modality of cancer therapy. However, only a few tumor by the ABI PRISM 377 DNA Sequencer (Perkin-Elmer). rejection antigen genes have been identified from the adenocarcino- Northern Blot Analysis. Nylon membranes (Hybond-Nϩ; Amersham, mas and SCCs, which are, histologically, the most frequently ob- Buckinghamshire, United Kingdom) with UV-fixed total RNA (5 ␮g/lane) served cancers in various organs (23–28). In addition, peptides capa- from various tissues were prehybridized for 20 min and hybridized overnight ble of inducing CTLs against these cancers have not yet been fully at 65°C in the same solution [7% SDS, 1 mM EDTA, and 0.5 M Na2HPO4 (pH 32 identified. We have recently reported a SART1 gene coding for tumor 7.2)], containing the P-labeled clone 13 as a probe. The membranes were washed three times at 65°C in a washing buffer [1% SDS and 40 mM Na HPO antigens and peptides from the cDNA of SCCs (25). Here, we inves- 2 4 (pH 7.2)] and then autoradiographed. We tentatively named this full-length tigated new genes encoding CTL-directed antigens from SCCs, and gene encoding a tumor antigen recognized by HLA-A2402 restricted CTLs the SART3 gene. The relative expression level of the SART3 mRNA was calculated Received 3/22/99; accepted 6/18/99. by the following formula: index ϭ (SART3 density of a sample/␤-actin density The costs of publication of this article were defrayed in part by the payment of page of a sample) ϫ (␤-actin density of the KE4 tumor/SART3 density of the KE4 charges. This article must therefore be hereby marked advertisement in accordance with tumor). 18 U.S.C. Section 1734 solely to indicate this fact. Cloning of the SART3 Gene. The SART3 gene was obtained from the 1 This study was supported in part by Ministry of Education, Science, Sport and Culture (Japan) Grants-in-Aid 08266266, 09470271, 10153265, 09770985, 10671230, cDNA libraries of both KE4 and PBMCs of healthy donors by the standard 32 and 09671401 and Ministry of Health and Welfare (Japan) Grant H10-genome-003. colony hybridization method using the P-labeled clone 13 as a probe, as 2 To whom requests for reprints should be addressed, at Department of Immunology, reported previously (25). The differences in sequence at nucleotide positions Kurume University School of Medicine, 67 Asahi-machi Kurume, Fukuoka 830-0011, 88 and 547 of SART3 between the PBMCs and KE4 were further analyzed by Japan. Phone: 81-942-31-7551; Fax: 81-942-31-7699; E-mail: [email protected] treatment of the PCR products with restriction enzyme DdeI and AluI, respec- u.ac.jp. Ј 3 The abbreviations used are: PBMC, peripheral blood mononuclear cell; SCC, squa- tively. The primers used for the amplification were: 5 -CGAAACCTCGGCT- mous cell carcinoma; SART, SCC antigen recognized by T cells; ORF, open reading TCAGAAC-3Ј (forward)/5Ј-GGTCTTGTATGTCGCAGCGG-3Ј (reverse) frame; Ab, antibody; mAb, monoclonal Ab; PHA, phytohemagglutinin; aa, amino acid. and 5Ј-GATGGCCTGGACAGAGAGC-3Ј (forward) and 5Ј-CCACCAACT- 4056

GAGTACTGGCC-3Ј (reverse), which covered the regions around positions 88 the expanded cells were transferred to 24-well plate and incubated in the and 547, respectively. The PCR was carried out for 45 cycles of 0.5 min at absence of either a peptide or feeder cells for 14–20 days. Their CTL activity 94°C and 1 min at 60°C using the AmpliTaq Gold DNA polymerase (Perkin- was rechecked by an IFN-␥ production assay, and the cells were then tested for Elmer). their cytotoxicity by a 6-h 51Cr release assay at different E:T ratios. Preparation of the SART3-tag Fusion Protein in Expression Vector Constructs. For preparation of SART3/myc, SART3 cDNA was digested with RESULTS EcoRI and BstEII. The SART3 gene at positions 2390–2895 was amplified by reverse transcription-PCR using 5Ј-GGTGTTCAGGTACAGCACTTCC-3Ј Establishment of HLA-A24-restricted CTLs. A HLA-A2402- Ј Ј (forward) and 5 -GCTTGGCAAACTCGAGATTGGACATC-3 (reverse), and restricted and tumor-specific CTL line was established from PBMCs the amplified product was digested with BstEII and XhoI. The EcoRI-BstEII of an esophageal cancer patient (KE4, HLA-A2402/2601) by repeated and BstEII-XhoI fragments were then together ligated to the EcoRI and XhoI stimulation with KE4 autologous tumor cells, as reported previously sites of pcDNA3.1/Myc-His A vector (Invitrogen, San Diego, CA). The gene encoding a tag peptide was ligated to the 2880 position before the stop codon (29). This CTL line began to show HLA-A2402-restricted cytotoxicity at around day 33 of culture, after being stimulated five times with of the ORF, which was used as the SART3/myc. ϩ Western Blot Analysis. The SART3 obtained from the KE4 tumor cDNA the KE4 tumor cells. The CTLs lysed all of the HLA-A2402 SCC or library was digested with SalI and NotI at the multiple cloning site of pSV- adenocarcinoma cell lines tested. However, this CTL line did not lyse Ϫ SPORT and then ligated into the pGEX-4T-1 vector (Pharmacia Biotech AB, any HLA-A2402 SCC or adenocarcinoma cell lines or HLA- Uppsala, Sweden) for preparation of the SART3-glutathione S-transferase A2402ϩ PHA-activated T cells. Some of the representative results are fusion protein (SART3/GST) in expression vector constructs, as reported pre- shown in Fig. 1. The surface phenotype of this CTL line was ϩ Ϫ ϩ viously (25). Polyclonal anti-SART3/GST Ab was prepared by the immuniza- CD3 4 8 , and the CTL activity was inhibited by 0.1 mg/ml anti- tion of rabbits with the purified SART3/GST by methods reported previously CD8 (IgG2a; Nichirei, Tokyo, Japan) or anti-HLA-class I (IgG2a, (30). Anti-myc mAb (Invitrogen) was also used for the Western blot analysis. W6/32; American Type Culture Collection, Manassas, VA) but not by The samples were lysed with a buffer containing 10 mM Tris-HCl (pH 7.4), 0.1 mg/ml anti-CD4 (IgG1; Nichirei), anti-HLA-DR (IgG2a; SRL, 150 mM NaCl, 0.5% Triton X-100, 0.2 mM phenylmethylsulfonyl fluoride (Sigma Chemical Co., St. Louis, MO), and 0.03 trypsin inhibitor unit/ml Tokyo, Japan) mAb, or an irrelevant isotype-matched mAb [anti- aprotinin. They were then sonicated and centrifuged at 12,000 rpm for 20 min, CD13 (IgG1, MCS-2; developed in our laboratory) or anti-CD14 mAb and the supernatant was used as the cytosol fraction. The resulting pellet was (IgG2a; SRL); data not shown]. These results suggested that this CTL lysed with a buffer consisting of 7.2 mM urea, 1.6% Triton X-100, 0.8% DTT, line recognized tumor cells in an HLA-A2402-restricted manner, and and 2% lithium dodecyl sulfate and then centrifuged, and the supernatant was therefore, it was used in the following experiments. used as the nuclear fraction. The lysate was separated by SDS-PAGE. The Identification of the SART3 Gene. A total of 105 clones from the proteins in an acrylamide gel were blotted to a Hybond-polyvinylidene diflu- cDNA library of KE4 tumor cells were tested for their ability to oride membrane (Amersham) and then incubated with appropriate Abs for 4 h stimulate IFN-␥ production by HLA-A2402-restricted CTLs after at room temperature. The other methods used for the Western blot analysis cotransfection with HLA-A2402 into the VA13 cells. After repeated have been described previously (30). experiments with the several candidate clones, one clone (clone 13) Construction of Deletion Mutants. The SART3/pCMV-SPORT plasmid was digested with BamHI, HindIII, and SphI. The restriction sites of the three was confirmed to encode a tumor antigen recognized by the CTLs enzymes were restricted at multiple restriction sites of the pCMV-SPORT when cotransfected with HLA-A2402 (Fig. 2A). Namely, the CTLs ␥ vector and were unique at positions 1060, 1234, and 2165 of the SART3 gene, produced the significant amounts of IFN- by recognition of VA13

respectively. The linearized SART31–1060, SART31–1234, and SART31–2165/ cells cotransfected with HLA-A2402 cDNA (100 ng/well) and 12.5 to pCMV-SPORT were separated, purified, and ligated to prepare the three 200 ng/ml of clone 13 cDNA with a dose-dependent manner but not

deletion mutants (SART31–1060, SART31–1234, and SART31–2165). with HLA-A0201 and clone 13. Reduction of IFN-␥ production at the Peptides and CTL Assays. The SART3-derived peptides capable of bind- 400 ng/ml clone 13 cDNA could be due to the decreased levels of the ing to the HLA-A2402 molecules were searched for in the literature regarding SART3 protein expression, as measured by the Western blot analysis peptides for HLA-A24-binding motifs (31), and 21 different peptides were with the polyclonal anti-SART3 Ab (data not shown). synthesized. These peptides were kindly provided by Dr. M. Kanaoka (Re- The sequence of clone 13 proved to be 1290 bp. Expression of this search Division of Sumitomo Pharmaceutical Co., Osaka, Japan). The purity was Ͼ70% in most of the peptides, and Ͼ95% for the dose dependency assay and the induction of CTLs. For detection of antigen peptides, the HLA-A2402 or -A0201 cDNA (as a control)-transfected VA13 (2 ϫ 104) cells were pulsed with a peptide at a final concentration of 10 ␮M for 2 h, the CTLs (4 ϫ 104) were then added and incubated for 18 h, and 100 ␮l of supernatant were collected to measure IFN-␥ by ELISA in a triplicate assay, as reported previously (25). The KE4 CTL sublines that were used to test peptide speci- ficity were established from the parental HLA-A2402-restricted KE4 CTLs by the limiting dilution culture (10 cells/well), as reported previously (25). C1R- ␮ A*2402 cells (32) pulsed with 10 M of SART3109–118 or SART3109–118 were used as targets to test the peptide-specific reactivity of these CTLs. Induction of CTLs by the Peptides. The method for CTL induction by a peptide has been described previously (25). Briefly, PBMCs from healthy volunteers or cancer patients were incubated with 10 ␮M peptide in one well of a 24-well plate. After 7, 14, and 21 days of culture, the cells were incubated with irradiated (50 Gy) autologous PBMCs acting as antigen-presenting cells that had been preincubated with the same peptide at the same dose for 3 h. Effector cells were harvested at day 21 of culture, in the case of cancer patients, and at day 21 and 28 of culture, in the case of healthy donors, and most of the cells were immediately tested for their activity to produce IFN-␥ in response to various target cells by an ELISA (limit of sensitivity, 10 pg/ml). For a 51Cr release assay, the patients’ PBMCs (5000 cells/well), which had Fig. 1. The cytotoxicity of HLA-A24-restricted CTLs. The cytotoxicity of HLA-A24- 51 been stimulated three times with the peptide, were further cultured in a 96-well restricted CTLs against various target cells was tested by a 6-h Cr release assay at three E:T ratios. Target cells were the three SCCs (KE4, TE11, and QG56), a small cell U-bottomed microculture plate in the presence of feeder cells, consisting of carcinoma (LC65A), a large cell carcinoma (86-2), PHA blastoid cells (PHA-blast) of two ϩ 5 irradiated autologous or HLA-A24 allogenic PBMCs (2 ϫ 10 cells/well) healthy donors, and an EBV-transformed B cell line (EBV-BCL) of a healthy donor. Data that had been prepulsed with a corresponding peptide. Seven to 10 days later, points, means of triplicate determinants. 4057

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1999 American Association for Cancer Research. A GENE CODING FOR A SHARED TUMOR ANTIGEN gene at the mRNA level was investigated by Northern blot analysis using the radiolabeled clone 13 as a probe, and a band of ϳ3.8 kb was observed in all of the malignant and normal cell lines and also in all 15 normal tissues tested (heart, brain, placenta, lung, liver, skeletal muscle, kidney, pancreas, spleen, thymus, prostate, testis, ovary, small intestine, and colon). Some of the data are shown in Fig. 3. The relative level of mRNA expression was within the range of 1.2 Ϯ 0.3 in tumor cell lines and 0.7 Ϯ 0.2 in normal tissues, relative to the level of expression in the KE4 tumor cells, considered to be 1.0. In the tumor cell lines tested, relatively higher levels of expression were observed in Colo201 colon tumor (2.3), R27 breast tumor (1.9), and HSC4 oral tumor cells (1.8). In the normal tissues tested, relatively higher levels of expression were observed in the brain (1.1), placenta Fig. 4. Homology between SART3 and KIAA0156 cDNA. The indicated nucleotide (nt) (1.0), pancreas (1.1), and testis (1.1), and lower expression levels were and aa positions refer to the SART3 sequence. The sequences of SART3 cloned from KE4 seen in skeletal muscle (0.4), thymus (0.4), colon (0.4), and PBMCs and KIAA0156 are almost identical except at positions 137 and 3468, and at the ends of (0.2). mRNA expression of the SART3 was also observed in surgically 5Ј and 3Ј noncoding regions. The difference at position 137 causes a synonymous change. removed cancerous lesions, nonmalignant adjacent lesions, and be- The shaded areas represent the ORFs of clone 13, SART3, and KIAA0156 cDNAs. Arrows, positions of the two antigenic peptides. u, the segments of the SART-3 longer than nign uterine myomas (data not shown). These results suggest that this KIAA0156 cDNA at both the 5Ј and 3Ј ends. The sequence of segment in the 5Ј region (19 bp) is CGGACGCGTG GGGCGCAAG, whereas that in the 3Ј region (129 bp) is AAAACTTCTT GCTGAATGGT ACTCAGATGT GCATTCACAT ACAGATGTGT TTTGAAGTGG GTGTACCTTG CTTTACCTAA TAGATGTGTA AATAGAACTT TTGTAAGTCA AAAAAAAAAA AAAAAAAAA. The sequence of the SART3 cloned from the KE4 is available from European Molecular Biology Laboratory/GenBank/DNA Data Bank of Japan (accession no. AB020880).

gene is ubiquitously expressed at the mRNA level and that the 1290-bp cDNA was incomplete. A 3806-bp gene was then independ- ently cloned from the cDNA libraries of KE4 tumor and PBMCs of healthy donors using clone 13 as a probe with colony hybridization methods, as reported previously (25). The nucleotide sequences of these new clones were identical, with the exception of two nucleo- tides: the C in KE4 versus G in PBMCs at nucleotide position 88 and the C in the KE4 versus T in PBMCs at position 547 (Fig. 4). These differences could be due to polymorphism but not to point mutations because position 88 was C in the KE4, TE8, and TE9 esophageal SCC tumor cell lines, an EBV-transformed B cell line from the KE4 patient (BEC-1), and PBMCs from a healthy donor, whereas position 88 was G in fetal liver and VA13 fibroblast cells. PBMCs from another healthy donor and a sample of testis contained both C and G. Position Fig. 2. Recognition of the SART3 gene products by HLA-A24-restricted CTLs. Different amounts of a clone 13 (A)orSART3 cloned from the KE4 tumor cDNA (B) and 547 was C in the KE4, TE8, TE9, BEC-1, VA13, and fetal liver, 100 ng of HLA-A2402 or HLA-A0201 cDNA were cotransfected into VA13 cells, followed whereas a sample of PBMCs from a healthy donor contained both a C by testing their ability to stimulate IFN-␥ production by HLA-A24-restricted KE4 CTLs. and T. An aa translated from a codon containing nucleotide position The background of IFN-␥ production by the CTLs in response to VA13 cells (ϳ100 pg/ml) was subtracted from the values shown. Similar results were obtained in the SART3 88 was aspartic acid (GAC) or glutamic acid (GAG), whereas the aa gene cloned from the PBMCs (data not shown). was identical translated from a codon containing nucleotide position 547 (GCC, GCU ϭ alanine) in the ORF. This changed residue was not in the context of HLA-A24 antigen-binding motifs (31). The SART3 gene cloned from either the KE4 and PBMCs encoded the antigens recognized by the CTLs when cotransfected with HLA-A2402 but not with HLA-A0201 (Fig. 2B). The nucleotide sequence of this 3806-bp SART3 gene was almost identical to that of the KIAA0156 gene already registered in the GenBank (accession no. D63879). The KIAA0156 gene was isolated from a human myeloid cell line (KG-1) in a human gene cloning project (33). Except for its sequence and ubiquitous mRNA expression by Northern blot analysis, nothing is known about KIAA0156 according to a search at the literature level. KIAA0156 is 3660 bp, which is identical to the region of the SART3 located between nucleotide Fig. 3. Expression of the SART3 in human cell lines. mRNA expression of the SART3 positions 20 and 3679, except for positions 118 and 3449 (in which G and was investigated by the Northern blot analysis. Twenty-seven tumor cell lines were used A are replaced by A and C, respectively; Fig. 4). The SART3 gene cloned for the study: 8 SCCs (KE4, KE3, TE8, TE9, Kuma-1, HSC4, QG56, and Sq-1), 11 adenocarcinomas (A549, 1-87, PC-9, LK79, KMG-A, R27, MKN28, Colo201, Colo205, from KE4 encodes 963 aa in the second frame if its first ATG at and KM12LM), a large cell carcinoma (86-2), 3 hepatocellular carcinomas (KIM-1, nucleotide positions 20–22 is used as a start codon, with the aa sequence KYN1, and HAK3), 2 small cell lung carcinomas (LC-65A and LK79), 2 melanomas being completely identical to the translated aa sequence of KIAA0156 if (M36 and M73), and 1 B-cell leukemia (NALM-1). The origins of these tumor cells have been described previously (25, 29). PBMCs, BEC-1 EBV-transformed B-cell line, COS7 its ATG at positions 1–3 is used as a start codon (Fig. 4). These results kidney cells, and 16 normal tissues (heart, brain, placenta, lung, liver, skeletal muscle, suggest that 3660 bp of the reported KIAA0156 is a truncated form of the kidney, and pancreas on Human Multiple Tissue Northern Blot; and spleen, thymus, 3806 bp of the SART3 gene. The sequence of SART3 cloned from the prostate, testis, ovary, small intestine, colon, and peripheral blood leukocyte on Human Multiple Tissue Northern Blot II; Clontech, Palo Alto, CA) were also provided for KE4 cDNA is available from European Molecular Biology Laboratory/ Northern blot analysis using clone 13 as a probe. Some of the results are shown. GenBank/DDBJ (accession no. AB020880). 4058

Expression of the SART3 Protein. Expression of SART3 at the Table 1. Expression of SART3 protein in normal and cancer cells and tissuesa protein level in various cells and tissues was studied by Western blot Cell linesb Tissuesb analysis with the polyclonal anti-SART3/GST Ab. It recognized a Mr 140,000 band of a recombinant SART3 protein after cleavage of Cell/tissue type Cytosol Nucleus Cytosol Nucleus glutathione S-transferase with thrombin (data not shown). When Normal PBMCs 0/5b 0/5 SART3 was transfected to VA13 cells, an intensive Mr 140,000 band PHA blastoid cells 3/3 0/3 was observed in the cytosol fraction but not in the nuclear fraction Fibroblasts 2/2 0/2 (Fig. 5A). Furthermore, both this polyclonal Ab and anti-myc mAb Fetal liver 1/1 0/1 ϳ Newborn liver 0/1 0/1 recognized an Mr 143,000 band in the cytosol but not in the nucleus Liver 0/10 0/10 of VA13 cells transfected with the SART3 gene of positions 1–2880 Testis 2/2 2/2 (containing 953 aa) in conjunction with the pcDNA3.1/myc-His vector Placenta 0/1 0/1 Esophagus 0/3 0/3 (SART3/myc; Fig. 5A). The different migration of these bands (Mr 140,000 ϳ Cancer and 143,000) is probably due to a tag peptide (Mr 5000). In contrast to Head and neck SCC 2/2 2/2 14/20 (70) 10/20 (50) Esophageal SCC 8/8 8/8 3/5 (60) 4/5 (80) VA13 cells, a Mr 143,000 band was detected in both the cytosol and nuclear fractions of the KE4 or TE9 tumor cells when the SART3 was Lung cancer /myc Adenocarcinoma 4/4 4/4 7/10 (70) 6/10 (60) transfected to the KE4 (Fig. 5A) or TE9 tumor cells (data not shown). SCC 3/3 3/3 5/8 (63) 4/8 (50) A M 140,000 protein of SART3 was undetectable in either the Leukemia 12/12 12/12 4/4 (100) 4/4 (100) r c cytosol or nuclear fractions of any of the normal tissues tested except Melanoma 2/2 2/2 8/9 (89) ND a for testis and fetal liver under the used condition. In fetal liver, the M Expression of the SART3 protein in the cytosol and nucleus of various normal and r cancer cells and tissues was investigated by Western blot analysis with anti-SART3/GST 140,000 band was detected in only the cytosol fraction, and it was Ab. The tumor cell lines used for the study are described in the legend for Fig. 5. b Values represent no. of positive samples/total no. of samples (%) tested. c ND, not determined.

detectable in both the cytosol and nuclear fractions of testis. Part of

these results are shown in Fig. 5B. The Mr 140,000 of SART3 protein was not observed in unstimulated PBMCs, but it did become detectable in the cytosol fraction of activated PBMCs after stimulation with ␮ 10 g/ml PHA for 48 h (data not shown). The Mr 140,000 band was detected in both the cytosol and nucleus of all of the malignant tumor cell lines tested, including SCC or adenocarcinoma, leukemia, and melanoma cell lines as well as in more than half of fresh cancer tissues from various organs, including head and neck SCCs, esophageal SCCs, lung SCCs, lung adenocarcinomas, melanomas, and fresh leukemia cells. Some of these results are shown in Fig. 5B, and a summary is shown in Table 1. The SART3 protein was undetectable in surgically removed nonmalignant adjacent lesions or benign uterine myomas, regardless of the expression at the mRNA level.4 Identification of Antigenic Peptides. To identify the SART3- derived epitopes recognized by HLA-A2402-restricted CTLs, we ini- tially investigated the capability of the deletion mutants of the SART3 to stimulate IFN-␥ production by the CTLs. All three mutants, including the shortest one, retained the ability to stimulate IFN-␥ production by the CTLs (Fig. 6A), suggesting that the antigenic epitopes were located within the first 1060 bp. Meanwhile, each of the 21 different synthesized peptides with HLA-A2402 antigen-binding motifs among the 963 aa of the SART3 was loaded onto the HLA- A2402-transfected VA13 cells and tested for its ability to induce IFN-␥ production by the CTLs. High levels of IFN-␥ production were

induced by the SART3109–118 (VYDYNCHVDL) and SART3315–323 peptide (AYIDFEMKI; Fig. 6B), both of which are located in the region encoded by the shortest deletion mutant (Fig. 6A). A dose- dependent reaction was observed in these two peptides (Fig. 6C). Significant levels of IFN-␥ production were observed at a concentration of 0.1 ␮M each peptide. Low but significant levels of IFN-␥ Fig. 5. Expression of the SART3 protein. A, VA13 cells or KE4 tumor cells were production were also induced by the SART3141–150, SART3172–181, transfected with the SART3 or SART3myc gene. Two days later, these transfected cells and and SART3284–292 peptide (Fig. 6B), but consistent results were not untransfected cells were tested for their expression of the SART3 protein in both the obtained with these three peptides. cytosol and nuclear fractions by Western blot analysis with the polyclonal anti-SART3/ To confirm the presence of SART3 and SART3 pep- GST Ab or anti-myc mAb. B, expression of SART3 protein in both the cytosol and nuclear 109–118 315–323 fractions of various normal tissues, cell lines, and cancer tissues was investigated with the tide-specific CTLs, we tested 40 different KE4 CTL sublines for their anti-SART3/GST Ab. Some of the results are shown, and a summary is shown in Table 1. reactivity to either of the peptides. Four and 5 of the 40 different Normal tissues tested were testes, fetal liver, placenta, adult liver, lung, esophagus, and sublines showed the SART3 and SART3 peptide- colon. Normal cells used were PBMCs, PHA blastoid cells, and fibroblast cells (WI-38, 109–118 315–323 VA13). The tumor cell lines used were head and neck SCCs (Kuma-1 and Kuma-3), specific reactivity, respectively. The IFN-␥ production by these CTL esophageal SCCs (KE3, KE4, KE5, TE9, TE10, TE11, YES4, and YES5), lung adeno- sublines in response to a peptide was inhibited by 0.1 mg/ml anti-CD8 carcinomas (1-87, LK87, PC-9, and 11-18), lung SCCs (Sq-1, RERF-LC-AI, and QG56), leukemia cells (HPB-ALL, BALL-1, NALM16, ARH77, THP1, U937, HL60, ML-1, ML-2, KG-1, K562, and HEL), and melanomas (M36 and M73). 4 N. Tsuda, M. Sakamoto, T. Nishiad, et al., unpublished results. 4059

Fig. 7. Recognition of a peptide by the CTL sublines. Forty different KE4 CTL ␥ sublines were tested for their ability to produce IFN- by recognition of the SART3109–118 ␮ or SART3315–323 peptide that was loaded at 10 M on the C1R-A*2402 cells either alone or in the presence of 0.1 mg/ml anti-CD8, anti-HLA class I, anti-CD4, or anti-HLA-DR mAb. Among 40 CTL sublines, 4 and 5 sublines reacted to the SART3109–118 and SART3315–323, respectively. The representative results from the peptide-specific CTL sublines (10-81 and 10-47) are shown. One subline (10-31) was reactive to both of the peptides, and the other 30 sublines were not reactive to any of the peptides (data not shown).

or anti-HLA class I (W6/32) mAb but not by anti-CD4 or anti-HLA- class II mAb. The representative results are shown in Fig. 7. One subline (10-31) reacted to the two peptides, and the remaining 30 sublines did not react to either (data not shown).

Induction of CTLs by the SART3 Peptides. The SART3109–118 and SART3315–323 peptides were tested for their ability to induce CTLs from the PBMCs of seven HLA-A24ϩ patients with epithelial cancer (four patients with head and neck SCCs, two with esophageal SCCs, and one with lung adenocarcinoma) and from five healthy donors. None of the PBMCs from five healthy donors produced significant levels of IFN-␥ in an HLA-A24-restricted fashion when stimulated three times (data not shown) or even four times (Table 2)

in vitro with either SART3109–118 or SART3325–323. In contrast, PBMCs from five of seven patients produced significant levels of IFN-␥ in response to the HLA-A24ϩ tumor cells but not to the HLA-A24Ϫ tumor cells when stimulated three times in vitro with

SART3109–118 (Table 2). Similarly, PBMCs from six of seven cancer Fig. 6. Identification of the peptides recognized by CTLs. A, to identify the regions patients produced significant levels of IFN-␥ in response to the containing antigenic peptides for CTLs, we cotransfected deletion mutants of SART3 HLA-A24ϩ tumor cells but not to the HLA-A24Ϫ tumor cells when (SART31–1060, SART31–1234, and SART31–2165) and the full length of SART3 to VA13 cells (2 ϫ 104) with HLA-A2402 or -A0201 and, 2 days later, tested them for their stimulated with SART3325–323 (Table 2). Although the PBMCs from ability to stimulate IFN-␥ production by HLA-A24-restricted CTLs. The background the remaining cancer patients (PBMCs of patients 4 and 6 for of IFN-␥ production by the CTLs in response to VA13 cells transfected with both SART3 and those of patient 6 for SART3 ) failed to ϳ 109–118 325–323 HLA-A0201 ( 50 pg/ml) and each mutant was subtracted from the values shown in produce higher amounts of IFN-␥ (Ͼ100 pg/ml) in response to the the figure. B, to determine the peptide antigens, we loaded a series of 21 SART3 ϩ 4 ␥ oligopeptides (9–11-mers; 10 ␮M)for2htotheVA13 cells (2 ϫ 10 ) that were HLA-A24 tumor cells but still released significant levels of IFN- in transfected 2 days before with HLA-A2402 or -A0201. HLA-A24-restricted CTLs comparison to those in response to the HLA-A24Ϫ tumor cells. The ϫ 4 (4 10 ) were then added, incubated for 18 h, and the culture supernatant was PBMCs from all four of the cancer patients (patients 1, 2, 5, and 7) collected for measurement of IFN-␥ by the ELISA in triplicate assays. Columns, ϩ means of the triplicate assays. The background of IFN-␥ production by HLA-A24- tested lysed the HLA-A24 tumor cells but did not lyse the HLA- Ϫ ϩ restricted KE4 CTLs in response to each peptide loaded to the VA13 transfected with A24 tumor cells, the HLA-A24 EBV-transformed B cell line, or HLA-A0201 (about 50 pg/ml) was subtracted from the values shown. C, various doses the HLA-A24ϩ PHA-blastoid normal cells. Results of the three pa- of SART3109–118 and SART3315–323 were loaded for2honVA13 cells transfected with HLA-A2402 (Ⅺ)or-A0201 (F), followed by testing their ability to stimulate tients are shown in Fig. 8. The peptide-induced CTL activity tested in IFN-␥ production by HLA-A24-restricted KE4 CTLs. patients 2 and 5 was inhibited by 1 mg/ml anti-CD8 and anti-HLA- 4060

ϩ Table 2 Induction of HLA-A24-restricted CTL activity in HLA-A24 PBMCs by SART3-derived peptidesa IFN-␥ production (pg/ml) in response tob

KE3 KE4 PC-9 KE5 QG56 Donors (disease) Peptide (HLA-A2/24) (HLA-A24/26) (HLA-A2/24) (HLA-A11/11) (HLA-A26/46) Healthy donor 1 None 0 0 NDc 00 SART3109–118 00ND0 0 SART3315–323 00ND0 0 Healthy donor 2 None 0 0 ND 0 0 SART3109–118 00ND0 0 SART3315–323 00ND0 0 Healthy donor 3 None 0 0 ND 0 39 SART3109–118 00ND0 0 SART3315–323 00ND0 0 Healthy donor 4 None 0 0 0 0 0 SART3109–118 00000 SART3315–323 00000 Healthy donor 5 None 0 0 0 0 0 SART3109–118 00000 SART3315–323 00000 Patient 1 (head and neck SCC) None 0 0 0 0 59 SART3109–118 543 144 184 0 13 SART3315–323 516 379 651 0 0 Patient 2 (head and neck SCC) None 0 0 0 0 0 SART3109–118 405 420 502 0 86 SART3315–323 282 308 569 82 72 Patient 3 (head and neck SCC) None 135 15 209 0 146 SART3109–118 169 203 642 0 73 SART3315–323 1695 1690 1695 0 0 Patient 4 (head and neck SCC) None 45 45 0 31 54 SART3109–118 53 75 0 8 0 SART3315–323 103 117 41 35 50 Patient 5 (esophageal SCC) None 13 54 0 0 64 SART3109–118 838 824 299 7 497 SART3315–323 135 104 60 5 16 Patient 6 (esophageal SCC) None 0 0 41 0 0 SART3109–118 14 68 92 0 0 SART3315–323 32 35 94 0 0 Patient 7 (lung adenocarcinoma) None 0 10 0 0 0 SART3109–118 214 238 305 0 0 SART3315–323 243 265 413 0 0 ϩ ϩ a PBMCs from five HLA-A24 healthy donors and seven HLA-A24 cancer patients [head and neck SCCs (n ϭ 4), esophageal SCCs (n ϭ 2), and lung adenocarcinoma (n ϭ 1)] were stimulated in vitro with 10 ␮M peptide and tested for their ability to produce IFN-␥ by recognition of various target cells at an EϺT ratio of 2. b Values represent the means of triplicate determinations. The background IFN-␥ production in response to VA13 cells (50–100 pg/ml) has been subtracted from the experimental values shown in the table. c ND, not determined. class I (W6/32) but not anti-CD4 or anti-HLA-DR mAb in both the have also reported a similar expression pattern in the SART1 protein, IFN-␥ production assay and 51Cr release assay (data not shown). the other tumor rejection antigen isolated from the KE4 tumor cells (25), and the mechanisms involved in this discrepancy are presently DISCUSSION unclear. The SART3 protein was preferentially expressed in the cytosol of the majority of proliferating cells, including normal cells The nucleotide sequence of the SART3 gene is almost identical to and malignant cells. Furthermore, it was expressed in the nucleus of that of a gene with unknown function already registered in GenBank all of the malignant cell lines and the majority of cancer tissues, under the name KIAA0156 (33). The SART3 contains a segment that regardless of their histology and origin, but not in the nucleus of any is 19 bp longer than the 5Ј end of KIAA0156 cDNA. This segment of the normal or virus-transformed proliferating cell lines or normal contains the context of Kozak motif (AAGATGG; Refs. 34 and 35) tissues, except for testis. We have recently developed anti-SART3 for effective initiation of translation because the first ATG codon mAb useful for immunohistochemical staining of the SART3 protein resides at positions 20–22 of the SART3 gene or positions 1–3 of the in cells. The experiments with this mAb indicated that the SART3 KIAA0156 gene, respectively. The different lengths of 3Ј uncoding protein was expressed in both the cytosol and nuclear fractions of the regions between the SART3 and KIAA0156 are probably due to the KE4 tumor cells and spermatogonia of the testis. In contrast, it was processes by which the cDNA libraries are constructed. There are not expressed in the spermatids or Sertoli cells.5 Further studies are several As in the 3Ј unique segment of SART3 immediately flanking needed to clarify the location and distribution of the SART3 protein. the site corresponding to the 3Ј end of KIAA0156, and the oligo(dT) There are several motifs in the sequence of the SART3 protein, primer might anneal to this region during the construction of cDNA described as follows, that suggest its biological function: nuclear libraries. In addition, clone 13 likely originates from the same mech- localization signals around positions 612–615 and 641–647, RNA- anism because there are also multiple As located in identical regions binding motifs at positions 746–753 and 841–848, a tyrosine phos- of the SART3 and KIAA0156 genes that immediately flank 3Ј end of phorylation kinase site at positions 309–316, and an RGD cell attach- the clone 13 (data not shown). Furthermore, a band of 3.8 kb was ment sequence at positions 742–744. Indeed, we have recently shown observed in the Northern blot analysis. These results suggest that the that a tyrosine at position 316 of the SART3 was phosphorylated by 3660 bp of the reported KIAA0156 gene is a truncated form of the using the anti-SART3 mAb.6 Furthermore, it has recently been shown 3806 bp of the SART3 gene. that the SART3 protein is a nuclear RNA-binding protein (36). There- The SART3 protein was undetectable in normal tissues except for the testis and fetal liver under the used condition with the polyclonal 5 Y. Miyagi, T. Sasatomi, S. Shichijo, et al., unpublished results. Ab, regardless of its ubiquitous expression at the mRNA level. We 6 A. Yamada, K. Harada, S. Schichijo, et al., unpublished results. 4061

Fig. 8. Induction of CTLs by the SART3 peptides. The PBMCs from patients with head and neck SCCs (patient 2), esophageal SCCs (patient 5), and lung adenocarcinoma (patient 7) were stimulated with no peptide, SART3109–118, and SART3315–323, followed by a test of their cytotoxicity against HLA-A24ϩ tumor cells (KE4, KE3, and PC-9), EBV-transformed B cells from the KE4 patient (BEC-1), and PHA blastoid cells from a healthy donor or HLA-A24Ϫ target cells (KE5, QG56, and VA13) by a 6-h 51Cr release assay at three different E:T ratios. Detailed methods for the 51Cr release assay are described in the “Materials and Methods.” Data points, means of triplicate assays.

fore, the phosphorylated SART3 protein might play a role in the fected to these cells.7 In addition, there is a RGD motif in the SART3 metabolism of nuclear RNA. Further studies of the biological function at positions 742–744 that might be involved in the caspase-3-mediated of the SART3 protein are also needed. apoptosis (39). Dendritic cells may then acquire apoptotic tumor cells

The two SART3-derived peptides (SART3109–118 and SART3315– and effectively cross-present the SART3 antigens to T cells. 323) among the 21 peptides with HLA-A24 antigen-binding motifs It is presently unclear why the KE4 CTL line or the CTLs induced tested in this study were consistently recognized by HLA-A24- by the SART3 peptides were not cytotoxic to the nonmalignant restricted CTLs in repeated experiments. Because of the presence of proliferating cells, in which the SART3 is also expressed in their CTL sublines reacting to either of the peptides, the parental KE4 cytosol. One possible explanation is that these SART3-derived CTLs would consist of the mixtures of these peptide-specific clones. epitopes are expressed preferentially on the HLA class I groove of A dose dependency was observed in these two peptides. The malignant cells but not of nonmalignant cells. The nuclear SART3

SART3315–323 was a peptide capable of binding to HLA-A24 antigens might be an activated form and, thus, be ubiquitinized, processed, and located in the overlapping region of the clone 13 and the shortest loaded to the groove of HLA class I molecules. This assumption is ␥ SART31060 mutant, which retained the ability to stimulate IFN- partly based on the fact that a tyrosine at position 316 of the SART3 6 production by the CTLs. The SART3109–118 was also found in the is phosphrylated. Another possibility can be attributed to the different shortest SART31060 mutant. Both peptides were able to induce HLA- posttranslational modification of the epitopes between normal and A24-restricted CTLs from PBMCs of most of the HLA-A24ϩ cancer cancer cells. Some of the previously identified CTL epitopes have patients after being stimulated three times, but this induction did not posttranslational modifications, and the modifications have had a occur in response to either peptide in any of the HLA-A24ϩ healthy significant impact on the ability of the CTLs to recognize those donors tested, even after being stimulated four or five times. These peptides (40–42). Further studies are needed to clarify this issue. peptides were also able to induce the CTLs from the PBMCs of a The HLA-A24 allele is found in ϳ60% of Japanese (95% of them patient with gastric signet ring cell carcinoma (data not shown). These are genotypically A2402), 20% of Caucasians, and 12% of Africans results suggest that there is a preferential presence of the CTL pre- (43). The two SART3-derived peptides were able to induce HLA- cursors reacting to these SART3 epitopes in PBMCs of cancer pa- A24-restricted and tumor-specific CTLs in most of the cancer patients tients. In contrast, these precursors appear to be at less than detectable tested but none of the healthy donors. The SART3 protein and these levels in the PBMCs of healthy donors. Circulating T cells from peptides may be appropriate molecules for use in specific immuno- healthy donors might be immunologically tolerant of these peptides. therapy of HLA-A24ϩ cancer patients with various histological types. Alternatively, this difference may be due to the exclusive expression of SART3 in the nucleus of malignant cells. Recent results have shown that the dendritic cells-mediated cross-presentation of apo- ACKNOWLEDGMENTS ptotic tumor cells could, in part, account for the initial priming of antigen-specific CTLs in cancer patients (37, 38). Tumor cells may We thank Dr. M. Takiguchi at Kumamoto University (Kumamoto, Japan) for undergo apoptosis for many reasons in vivo, such as part of a host providing an C1R-A*2402 cell line and Dr. Kunzo Orita at Hayashibara Biochem- response to tumor cells, nutritional reasons, or the outcome of chemical Lab. Inc. (Okayama, Japan), for providing natural IFN-␥ for ELISA. otherapy or radiotherapy. Indeed, the SART-3 gene product itself can induce apoptosis of TE9 tumor cells but not VA13 cells when trans- 7 D. Yang, N. Tsuda, S. Shichijo, et al., unpublished results. 4062

REFERENCES Wunderlich, J. R., Parkhurst, M. A., Kawakami, Y., Seipp, C. A., Einhorn, J. H., and White, D. E. Immunologic and therapeutic evaluation of a synthetic peptide vaccine 1. van der Bruggen, P., Traversari, C., Chomez, P., Lurquin, C., De Plaen, E., Van den for the treatment of patients with metastatic melanoma. Nat. Med., 4: 321–327, 1998. Eynde, B., Knuth, A., and Boon, T. A gene encoding an antigen recognized by 22. Nestle, F. O., Alijagic, S., Gilliet, M., Sun, Y., Grabbe, S., Dummer, R., Burg, G., and cytolytic T lymphocytes on a human melanoma. Science (Washington DC), 254: Schadendorf, D. Vaccination of melanoma patients with peptide- or tumor lysate- 1643–1647, 1991. pulsed dendritic cells. Nat. Med., 4: 328–332, 1998. 2. Kawakami, Y., Eliyahu, S., Delgado, C. H., Robbins, P. F., Rivoltini, L., Topalian, 23. Gaugler, B., Brouwenstijn, N., Vantomme, V., Szikora, J. P., Van der Spek, C. W., S. L., Miki, T., and Rosenberg, S. A. Cloning of the gene coding for a shared human Patard, J. J., Boon, T., Schrier, P., and Van den Eynde, B. J. A new gene coding for melanoma antigen recognized by autologous T cells infiltrating into tumor. Proc. an antigen recognized by autologous cytolytic T lymphocytes on a human renal Natl. Acad. Sci. USA, 91: 3515–3519, 1994. carcinoma. Immunogenetics, 44: 323–330, 1996. 3. Kawakami, Y., Eliyahu, S., Delgado, C. H., Robbins, P. F., Sakaguchi, K., Appella, 24. Bra¨ndle, D., Brasseur, F., Weynants, P., Boon, T., and Van den Eynde, B. J. A E., Yannelli, J. R., Adema, G. J., Miki, T., and Rosenberg, S. A. Identification of a mutated HLA-A2 molecule recognized by autologous cytotoxic T lymphocytes on a human melanoma antigen recognized by tumor-infiltrating lymphocytes associated human renal cell carcinoma. J. Exp. Med., 183: 2501–2508, 1996. with in vivo tumor rejection. Proc. Natl. Acad. Sci. USA, 91: 6458–6452, 1994. 25. Mandruzzato, A., Brasseur, F., Andry, G., Boon, T., and van der Bruggen, P. A 4. Brichard, V., Van Pel, A., Wo¨lfel, T., Wo¨lfel, C., De Plaen, E., Lethe´, B., Coulie, P., CASP-8 mutation recognized by cytotoxic T lymphocytes on a human head and neck and Boon, T. The tyrosinase gene codes for an antigen recognized by autologous carcinoma. J. Exp. Med., 186: 785–793, 1997. cytolytic T lymphocytes on HLA-A2 melanomas. J. Exp. Med., 178: 489–495, 1993. 5. Gaugler, B., Van den Eynde, B., van der Bruggen, P., Romero, P., Gaforio, J. J., De 26. Shichijo, S., Nakao, M., Imai, Y., Takasu, H., Kawamoto, M., Niiya, F., Yang, D., Plaen, E., Lethe´, B., Brasseur, F., and Boon, T. Human gene MAGE-3 codes for an Toh, Y., Yamana, H., and Itoh, K. A gene encoding antigenic peptides of human antigen recognized on a melanoma by autologous cytolytic T lymphocytes. J. Exp. squamous cell carcinoma recognized by cytotoxic T lymphocytes. J. Exp. Med., 187: Med., 179: 921–930, 1994. 277–288, 1998. 6. Boe¨l, P., Wildmann, C., Sensi, M. L., Brasseur, R., Renauld, J. C., Coulie, P., Boon, 27. Gue´guen, M., Patard, J. J., Gaugler, B., Brasseur, F., Renauld, J. C., Van Cangh, P. J., T., and Van der Bruggen, P. BAGE: a new gene encoding an antigen recognized on Boon, T., and Van den Eynde, B. J. An antigen recognized by autologous CTLs on human melanomas by cytolytic T lymphocytes. Immunity, 2: 167–175, 1995. a human bladder carcinoma. J. Immunol., 160: 6166–6194, 1998. 7. Ven Den Eynde, B., Peeters, O., De Backer, O., Gaugler, B., Lucas, S., and Boon, T. 28. Hogan, K. T., Eisinger, D. P., Cupp, S. B., Lekstrom, K. J., Deacon, D. D., A new family of genes coding for an antigen recognized by autologous cytolytic T Shabanowitz, J., Hunt, D. F., Engelhard, V. H., Slingluff, C. L., and Ross, M. M. The lymphocytes on a human melanoma. J. Exp. Med., 182: 689–697, 1995. peptide recognized by HLA-A68.2-restricted, squamous cell carcinoma of the lung- 8. Coulie, P. G., Brichard, V., Van Pel, A., Wo¨lfel, T., Schneider, J., Traversari, C., specific cytotoxic T lymphocytes is derived from a mutated elongation factor 2 gene. Mattei, S., De Plaen, E., Lurquin, C., Szikora, J. P., Renauld, J. C., and Boon, T. A Cancer Res., 58: 5144–5150, 1998. new gene coding for a differentiation antigen recognized by autologous cytolytic T 29. Nakao, M., Yamana, H., Imai, Y., Toh, Y., Toh, U., Kimura, A., Yanoma, S., lymphocytes on HLA-A2 melanomas. J. Exp. Med., 180: 35–42, 1994. Kakegawa, T., and Itoh, K. HLA-A2601-restricted CTLs recognize a peptide antigen 9. Wang, R. F., Parkhurst, M. R., Kawakami, Y., Robbins, P. F., and Rosenberg, S. A. expressed on squamous cell carcinoma. Cancer Res., 55: 4248–4252, 1995. Utilization of an alternative open reading frame of a normal gene in generating a 30. Shichijo, S., Hayashi, A., Takaori, S., Tsunosue, R., Hoshino, T., Sakata, M., novel human cancer antigen. J. Exp. Med., 183: 1131–1140, 1996. Kuramoto, T., Oizumi, K., and Itoh, K. Detection of MAGE-4 protein in lung cancer. 10. Coulie, P. G., Lehmann, F., Lethe´, B., Herman, J., Lurquin, C., Andrawiss, M., and Int. J. Cancer., 54: 158–165, 1995. Boon, T. A mutated intron sequence codes for an antigenic peptide recognized by 31. Kubo, R. T., Sette, A., Grey, H. M., Appella, E., Sakaguchi, K., Zhu, N. Z., Arnott, cytolytic T lymphocytes on a human melanoma. Proc. Natl. Acad. Sci. USA, 92: D., Sherman, N., Shabanowitz, J., michel, H., Bodnar, W. M., Davis, T. A., and Hunt, 7976–7980, 1995. D. F. Definition of specific peptide motifs for four major HLA-A alleles. J. Immunol., 11. Robbins, P. F., El-Gamil, M., Li, Y. F., Kawakami, Y., Loftus, D., Appella, E., and 152: 3913–3924, 1994. Rosenberg, S. A. A mutated ␤-catenin gene encodes a melanoma-specific antigen 32. Ikeda-Morre, Y., Tomiyama, H., Miwa, K., Oka, S., Iwamoto, A., Kaneko, Y., and recognized by tumor infiltrating lymphocytes. J. Exp. Med., 183: 1185–1192, 1996. Takiguchi, M. Identification and characterization of multiple HLA-A24-restricted 12. Guilloux, Y., Lucas, S., Brichard, V. G., Van Pel, A., Viret, C., De Plaen, E., HIV-1 CTL epitopes: strong epitopes are derived from V regions of HIV-1. J. Im- Brasseur, F., Lethe´, B., Jotereau, F., and Boon, T. A peptide recognized by human munology, 159: 6242–6252, 1997. cytolytic T lymphocytes on HLA-A2 melanoma is encoded by an intron sequence of 33. Nagase, T., Seki, N., Tanaka, A., Ishikawa, K., and Nomura, N. Prediction of the the N-acetylglucosaminyltransferase V gene. J. Exp. Med., 183: 1173–1183, 1996. coding sequences of unidentified human genes. IV. The coding sequences of 40 new 13. Ikeda, H., Lethe´, B., Lehmann, F., Van Baren, N., Baurain, J. F., Smet, C. D., genes (KIAA0121-KIAA0160) deduced by analysis of clones from human cell line Chambost, H., Vitale, M., Moretta, A., Boon, T., and Coulie, P. G. Characterization KG-1. DNA Res., 2: 167–174, 1995. of an antigen that is recognized on a melanoma showing partial HLA loss by CTL 34. Kozak, M. Bifunctional messenger RNAs in eukaryotes. Cell, 47: 481–483, 1986. expressing an NK inhibitory receptor. Immunity, 6: 199–208, 1997. 35. Kozak, M. The scanning model for translation: an update. J. Cell Biol., 108: 229–241, 14. Lupetti, R., Pisarra, P., Verrecchia, A., Farina, C., Nicolini, G., Anichini, A., 1989. Bordignon, C., Sensi, M., Parmiani, G., and Traversari, C. Translation of a retained 36. Gu, J., Shimba, S., Nomura, N., and Reddy, R. Isolation and characterization of a new intron in tyrosinase-related protein (TRP) 2 mRNA generates a new cytotoxic T 110 kDa human nuclear RNA-binding protein (p110nrb). Biochim. Biophys. Acta, lymphocyte (CTL)-defined and shared human melanoma antigen not expressed in 1399: 1–9, 1998. normal cells of the melanocytic lineage. J. Exp. Med., 188: 1005–1016, 1998. 37. Albert, M. L., Darnell, J. C., Bender, A., Francisco, L. M., Bhardwaj, N., and Darnell, 15. Kawakami, T., Robbins, P. F., Wang, X., Tupesis, J. P., Parkhurst, M. R., Kang, X., R. B. Tumor-specific killer cells in paraneoplastic cerebellar degeneration. Nat. Med., Sakaguchi, K., Appella, E., and Rosenberg, S. A. Identification of new melanoma 4: 1321–1324, 1998. epitopes on melanosomal proteins recognized by tumor infiltrating T lymphocytes 38. Albert, M. L., Sauter, B., and Bhardwaj, N. Dendritic cells acquire antigen from restricted by HLA-A1, -A2, and -A3 alleles. J. Immunol., 161: 6985–6992, 1998. apoptotic cells and induce class I-restricted CTLs. Nature (Lond.), 392: 86–89, 1998. 16. Rivoltini, L., Kawakami, Y., Sakaguchi, K., Southwood, S., Sette, A., Robbins, P. F., 39. Buckley, C. D., Pilling, D., Henriquez, N. V., Parsonage, G., Threlfall, K., Scheel- Marincola, F. M., Salgaller, M. L., Yannelli, J. R., Appella, E., and Rosenberg, S. A. Toellner, D., Simmons, D. L., Akbar, A. N., Lord, J. M., and Salmon, M. RGD peptides Induction of tumor-reactive CTL from peripheral blood and tumor-infiltrating lympho- induce apoptosis by direct caspase-3 activation. Nature (Lond.), 397: 534–539, 1999. cytes of melanoma patients by in vitro stimulation with an immunodominant peptide of the human melanoma antigen MART-1. J. Immunol., 154: 2257–2265, 1995. 40. Skipper, J. C. A., Hendrickson, R. C., Gulden, P. H., Brichard, V., Van Pel, A., Chen, 17. Marincola, F. M., Rivoltini, L., Salgaller, M. L., Player, M., and Rosenberg, S. A. Y., Shabanowitz, J., Wo¨lfel, T., Slingluff, C. L., Boon, T., Hunt, D. F., and Engelhard, Differential anti-MART-1/MelanA CTL activity in peripheral blood of HLA-A2 V. H. An HLA-A2 restricted tyrosinase antigen on melanoma cells results from melanoma patients in comparison to healthy donors: evidence for in vivo priming by post-translational modification and suggests a novel processing pathway for mem- tumor cells. J. Immunother., 19: 266–277, 1996. brane proteins. J. Exp. Med., 183: 527–534, 1996. 18. Salgaller, M. L., Afsher, A., Marincola, F. M., Rivoltini, L., Kawakami, Y., and 41. Meadow, L. R., Wang, W., den Hann, J. M., Blokland, E., Reinhardus, C., Drijfhout, Rosenberg, S. A. Recognition of multiple epitopes in the human melanoma antigen J. W., Shabanowitz, J., Pierce, R., Agulnik, A., Bishop, C. E., Hunt, D. F., Goulmy, gp100 by peripheral blood lymphocytes stimulated in vitro with synthetic peptides. E., and Engelhard, V. H. The HLA*0201-restricted HY antigen contains a posttrans- Cancer Res., 55: 4972–4979, 1995. lationally modified cysteine that significantly affects T cell recognition. Immunity, 6: 19. Visseren, M. J. W., van Elsas, A., van der Voort, E. I. H., Ressing, M. E., Kast, W. M., 273–281, 1997. Schrier, P. I., and Melief, C. J. M. CTL specific for the tyrosinase autoantigen can be 42. Kittlesen, D. J., Thompson, L. W., Gulden, P. H., Skipper, J. C. A., Colella, T. A., induced from healthy donor blood to lyse melanoma cells. J. Immunol., 154: 3991– Shabanowitz, J. A., Hunt, D. F., Engelhard, V. H., and Slingluff, C. L. Human 3998, 1995. melanoma patients recognize an HLA-A1-restricted CTL epitope from tyrosinase 20. Tanaka, F., Fujie, T., Tahara, K., Mori, M., Takesako, K., Sette, A., Celis, E., and containing two cysteine residue: implications for tumor vaccine development. J. Im- Akiyoski, T. Induction of antitumor cytotoxic T lymphocytes with a MAGE-3- munol., 160: 2099–2106, 1998. encoded synthetic peptide presented by human leukocytes antigen-A24. Cancer Res., 43. Imanishi, T., Akaza, T., Kimura, A., Tokunaga, K., and Gojobori, T. Allele and 57: 4465–4468, 1997. haplotype frequencies for HLA and complement loci in various ethnic groups. In: K. 21. Rosenberg, S. A., Yang, J. C., Schwartzentruber, D. J., Hwu, P., Marincola, F. M., Tsuji, M. Aizawa, and T. Sasazuki (eds.), HLA 1991, Vol. 1, pp. 1065–1220. Oxford, Topalian, S. L., Restifo, N. P., Dudley, M. E., Schwarz, S. L., Spiess, P. J., United Kingdom: Oxford Scientific Publications, 1992.

4063

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1999 American Association for Cancer Research. Identification of a Gene Coding for a Protein Possessing Shared Tumor Epitopes Capable of Inducing HLA-A24-restricted Cytotoxic T Lymphocytes inCancer Patients

Damu Yang, Masanobu Nakao, Shigeki Shichijo, et al.

Cancer Res 1999;59:4056-4063.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/59/16/4056

Cited articles This article cites 42 articles, 28 of which you can access for free at: http://cancerres.aacrjournals.org/content/59/16/4056.full#ref-list-1

Citing articles This article has been cited by 32 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/59/16/4056.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/59/16/4056. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.