A gene encoding antigenic peptides of human squamous cell carcinoma recognized by cytotoxic T lymphocytes

Abstract

Except for melanomas, tumor antigens recognized by cytotoxic T lymphocytes (CTLs) are yet unidentified. We have identified a gene encoding antigenic peptides of human squamous cell carcinomas (SCCs) recognized by human histocompatibility leukocyte antigens (HLA)- A2601-restricted CTLs. This gene showed no similarity to known sequences, and encoded two (125- and 43-kilodalton [kD]) proteins. The 125-kD protein with the leucine zipper motif was expressed in the nucleus of the majority of proliferating cells tested, including normal and malignant cells. The 43-kD protein was expressed in the cytosol of most SCCs from various organs and half of lung adenocarcinomas, but was not expressed in other cancers nor in a panel of normal tissues. The three nonapeptides shared by the two proteins were recognized by the KE4 CTLs, and one of the peptides induced in vitro from peripheral blood mononuclear cells (PBMCs) the CTLs restricted to the autologous tumor cells. The 43-kD protein and this nonapeptide (KGSGKMKTE) may be useful for the specific immunotherapy of HLA-A2601(+) epithelial cancer patients.

Figure 1

Recognition of the SART-1 gene products by the KE4 CTLs. Different amounts of the 6A1-1D7 (Fig. 1 A) or the SART-1 cloned from the KE4 tumor (Fig. 1 B) and 100 ng of HLA-A2601 or HLA-A0201 cDNA were cotransfected into VA13 cells, followed by testing their ability to stimulate IFN-γ production by the KE4 CTLs. The background of IFN-γ production by the KE4 CTLs in response to VA13 cells (∼200 pg/ml) was subtracted in the figure. Similar results were obtained in the SART-1 cloned from the PBMCs (data not shown).

Figure 2

Expression of the SART-1. 21 tumor cell lines (KE4, KE3, TE8, Kuma-1, HSC4, QG56, Sq-1, A549, MKN28, Colo201, SW620, KMG-A, R-28, 86-2, LK79, LC65A, KIM-1, KYN-1, M36, M73, NALM-1; reference 18), PBMCs, Bec-1, COS, or 16 tissues (heart, brain, placenta, lung, liver, skeletal muscle, kidney, and pancreas on human multiple tissue Northern blot, and spleen, thymus, prostate, testis, uterus, small intestine, colon, and peripheral blood leukocyte on human multiple tissue Northern blot IV; Clontech Lab., Inc., Palo Alto, CA) were provided for Northern blot analysis with the 6A1-1D7 as a probe. Some of the results are shown in the figure.

Figure 3

Nucleotide sequence of the SART-1. The cloned cDNA (6A1-1D7) was initially provided for the nucleotide sequencing. The sequence of 6A1-1D7 is 990 bp long (positions 1,517–2,506, underlined by the solid line) that has an ORF of 201 bp long encoding 67 aa if the first AUG codon (1,663–1,665, underlined by the bold line) and stop codon (1,864–1,866, underlined by the dotted line) in the first frame are used for protein synthesis. One S-D and each of the two different S-D–like sequences are marked by the dot on the top. The 2,506-bp-long SART-1 was then cloned from cDNA libraries of the KE4 tumor and human PBMCs. The SART-1 has an ORF 2,400 bp long encoding 800 aa when the first AUG codon (39–41, underlined by the bold line) and stop codon (2,439–2,441, underlined by the dotted line) are used for protein synthesis in the third frame. There was only one nt difference at the position 812 (marked by ▾) between the SART-1 of KE4 tumor and PBMCs (cytosine in the KE4 tumor versus thymine in PBMCs). These sequence data are available from EMBL/GenBank/DDBJ under accession number AB006198.

Figure 4

Expression of SART-1₈₀₀ and SART-1₂₅₉ proteins. Tumor cell lines used for Western blot analysis were head and neck SCCs (Ca9-22, HSC3, HSC4, Kuma-1, and Kuma-3), esophageal SCCs (KE4, KE3, TE8, TE9, TE10, and TE11), lung adenocarcinomas (1-87, LK87, PC-9, A549, 11-18, and RERF-LC-MS), lung SCCs (Sq-1, RERF-LC-AI, and QG56), leukemia cells ( MOLT-4, HPB-ALL, HPB-MLT, HUT-102, BALL-1, NALM16, ARH77, THP1, U937, HL60, ML-1, ML-2, NALL-1, SPI-801, K562, and HEL), and melanomas (M36, M73) (18). PBMCs, PHA blasts, fibroblast cells (WI-38, VA13), and tumor tissues from various organs were also studied. (A) Expression of the SART-1₈₀₀ protein was investigated by Western blot analysis with anti–SART-1_800/GST Ab. Anti-myc monoclonal Ab (Invitrogen) was also used for analysis of VA13 transfected with the SART-1_800/myc. (B) Expression of the SART-1₂₅₉ was investigated with anti–SART-1_6A1-1D7/GST Ab. In the left gel, 70-, 43-, and 27-kD bands corresponded to the SART-1_6A1-1D7/GST, SART-1_6A1-1D7, and GST, respectively. The data of the cytosol fraction were shown. No bands were detected by this Ab in the nucleus of any samples tested (data not shown). (C) Expression of the SART-1₂₅₉ was investigated with anti– SART-1_67/GST and anti–SART-1_219/GST Abs. The data of the cytosol fraction were shown. (D) Anti-myc and anti–SART-1_219/GST Abs were used for analysis of COS cells transfected with the SART-1_6A1-1D7/myc or SART-1_219/myc. The total lysate was used for experiments.

Figure 4

Expression of SART-1₈₀₀ and SART-1₂₅₉ proteins. Tumor cell lines used for Western blot analysis were head and neck SCCs (Ca9-22, HSC3, HSC4, Kuma-1, and Kuma-3), esophageal SCCs (KE4, KE3, TE8, TE9, TE10, and TE11), lung adenocarcinomas (1-87, LK87, PC-9, A549, 11-18, and RERF-LC-MS), lung SCCs (Sq-1, RERF-LC-AI, and QG56), leukemia cells ( MOLT-4, HPB-ALL, HPB-MLT, HUT-102, BALL-1, NALM16, ARH77, THP1, U937, HL60, ML-1, ML-2, NALL-1, SPI-801, K562, and HEL), and melanomas (M36, M73) (18). PBMCs, PHA blasts, fibroblast cells (WI-38, VA13), and tumor tissues from various organs were also studied. (A) Expression of the SART-1₈₀₀ protein was investigated by Western blot analysis with anti–SART-1_800/GST Ab. Anti-myc monoclonal Ab (Invitrogen) was also used for analysis of VA13 transfected with the SART-1_800/myc. (B) Expression of the SART-1₂₅₉ was investigated with anti–SART-1_6A1-1D7/GST Ab. In the left gel, 70-, 43-, and 27-kD bands corresponded to the SART-1_6A1-1D7/GST, SART-1_6A1-1D7, and GST, respectively. The data of the cytosol fraction were shown. No bands were detected by this Ab in the nucleus of any samples tested (data not shown). (C) Expression of the SART-1₂₅₉ was investigated with anti– SART-1_67/GST and anti–SART-1_219/GST Abs. The data of the cytosol fraction were shown. (D) Anti-myc and anti–SART-1_219/GST Abs were used for analysis of COS cells transfected with the SART-1_6A1-1D7/myc or SART-1_219/myc. The total lysate was used for experiments.

Figure 4

Expression of SART-1₈₀₀ and SART-1₂₅₉ proteins. Tumor cell lines used for Western blot analysis were head and neck SCCs (Ca9-22, HSC3, HSC4, Kuma-1, and Kuma-3), esophageal SCCs (KE4, KE3, TE8, TE9, TE10, and TE11), lung adenocarcinomas (1-87, LK87, PC-9, A549, 11-18, and RERF-LC-MS), lung SCCs (Sq-1, RERF-LC-AI, and QG56), leukemia cells ( MOLT-4, HPB-ALL, HPB-MLT, HUT-102, BALL-1, NALM16, ARH77, THP1, U937, HL60, ML-1, ML-2, NALL-1, SPI-801, K562, and HEL), and melanomas (M36, M73) (18). PBMCs, PHA blasts, fibroblast cells (WI-38, VA13), and tumor tissues from various organs were also studied. (A) Expression of the SART-1₈₀₀ protein was investigated by Western blot analysis with anti–SART-1_800/GST Ab. Anti-myc monoclonal Ab (Invitrogen) was also used for analysis of VA13 transfected with the SART-1_800/myc. (B) Expression of the SART-1₂₅₉ was investigated with anti–SART-1_6A1-1D7/GST Ab. In the left gel, 70-, 43-, and 27-kD bands corresponded to the SART-1_6A1-1D7/GST, SART-1_6A1-1D7, and GST, respectively. The data of the cytosol fraction were shown. No bands were detected by this Ab in the nucleus of any samples tested (data not shown). (C) Expression of the SART-1₂₅₉ was investigated with anti– SART-1_67/GST and anti–SART-1_219/GST Abs. The data of the cytosol fraction were shown. (D) Anti-myc and anti–SART-1_219/GST Abs were used for analysis of COS cells transfected with the SART-1_6A1-1D7/myc or SART-1_219/myc. The total lysate was used for experiments.

Figure 5

Identification of regions containing antigenic peptides for CTL. Deletion mutants of 6A1-1D7 gene (6A1_1-492, 6A1_1-625, 6A1_1-736, 6A1_1-839, and 6A1_1-951) and the full length of 6A1-1D7 in A or mutants of SART-1 gene (SART-1_1-2,008, SART-1_1-2,141, SART-1_1-2,252, SART-1_1-2,355, and the others) and the full length of SART-1 in B were cotransfected to VA13 cells (2 × 10⁴) with HLA-A2601 or -A0201, and 2 d later these cells were tested for their ability to stimulate IFN-γ production by the KE4 CTLs. The background of IFN-γ production by the KE4 CTLs in response to VA13 cells transfected with both each mutant and HLA-A0201 (∼50 pg/ml) was subtracted in the figure.

Figure 6

Determination of peptide antigens. A series of 22 SART-1 oligopeptides (10 mer; 10 μM) in A or 6 different nonapeptides (10 μM) from three 10 mer (SART-1_736-745, SART-1_748-757, and SART-1_784-793) with deletion of one aa at position 1 or 10 in B were loaded for 2 h to the VA13 cells (2 × 10⁴) transfected with HLA-A2601 or -A0201. For IFN-γ production, the KE4 CTLs (10⁴) were added, incubated for 18 h, and the culture supernatant was collected for measurement of IFN-γ by the ELISA in duplicate assays. The background of IFN-γ production by the KE4 CTLs in response to each peptide loaded to the VA13 cells transfected with HLA-A0201 (∼50 pg/ml) was subtracted in the figure. In a ⁵¹Cr–release assay, these VA13 cells were labeled with Na₂ ⁵¹CrO₄ for 1 h followed by adding the KE4 CTLs (5 × 10⁴). 6 h later, the supernatant was harvested for measurement of the radioactivity in triplicate assays as reported (18). The background of percent lysis by the KE4 CTLs of the VA13 cells that were transfected with HLA-A0201 and loaded by each peptide was <5%.

Figure 6

Determination of peptide antigens. A series of 22 SART-1 oligopeptides (10 mer; 10 μM) in A or 6 different nonapeptides (10 μM) from three 10 mer (SART-1_736-745, SART-1_748-757, and SART-1_784-793) with deletion of one aa at position 1 or 10 in B were loaded for 2 h to the VA13 cells (2 × 10⁴) transfected with HLA-A2601 or -A0201. For IFN-γ production, the KE4 CTLs (10⁴) were added, incubated for 18 h, and the culture supernatant was collected for measurement of IFN-γ by the ELISA in duplicate assays. The background of IFN-γ production by the KE4 CTLs in response to each peptide loaded to the VA13 cells transfected with HLA-A0201 (∼50 pg/ml) was subtracted in the figure. In a ⁵¹Cr–release assay, these VA13 cells were labeled with Na₂ ⁵¹CrO₄ for 1 h followed by adding the KE4 CTLs (5 × 10⁴). 6 h later, the supernatant was harvested for measurement of the radioactivity in triplicate assays as reported (18). The background of percent lysis by the KE4 CTLs of the VA13 cells that were transfected with HLA-A0201 and loaded by each peptide was <5%.

Figure 7

CTL sublines recognizing each nonapeptide. 80 KE4 CTL sublines were tested for their ability to produce IFN-γ by recognition of each of the three 10 mer (SART-1_736-745, SART-1_748-757, and SART-1_784-793) that was loaded at 10 μM on the VA13 cells for 2 h transfected with HLA-A2601 or -A0201. Detailed methods are shown in the legend for Fig. 6. Four, five, or six of 80 of the KE4 CTL sublines reacted to the SART-1_736-745, SART-1_748-757, or SART-1_784-793, respectively. The representative results from the peptide-specific CTL sublines (No. 48 and 53: the SART-1_736-745–specific CTLs; No. 3 and 60: the SART-1_748-757–specific CTLs; and No. 13 and 36: the SART-1_784-793–specific CTLs) are shown in the figure. The other CTL sublines were mostly not reactive to any of the 10 mer, and only a few sublines were reactive to two of the three 10 mer (data not shown). None of the sublines were reactive to all the three peptides.

Figure 8

Dose dependency of nonapeptides. Various doses of each of the nonapeptides (SART-1_736-744, SART-1_749-757, and SART-1_785-793) were loaded for 2 h on VA13 cells transfected with HLA-A2601 or -A0201 followed by testing their ability to stimulate IFN-γ production by the parental KE4 CTLs. Detailed methods are shown in the legend for Fig. 6.

Figure 9

Determination of aa required for CTL-mediated recognition. Each aa of the three nonapeptides (SART-1_736-744, SART-1_749-757, and SART-1_785-793) was substituted by glycine when it was not glycine, or threonine when it was glycine. These substituents along with the parental nonapeptides (10 μM) were loaded for 2 h to the VA13 cells transfected with HLA-A2601 or -A0201 followed by testing their ability to stimulate IFN-γ production by the parental KE4 CTLs. Detailed methods are shown in the legend for Fig. 6.

Figure 9

Determination of aa required for CTL-mediated recognition. Each aa of the three nonapeptides (SART-1_736-744, SART-1_749-757, and SART-1_785-793) was substituted by glycine when it was not glycine, or threonine when it was glycine. These substituents along with the parental nonapeptides (10 μM) were loaded for 2 h to the VA13 cells transfected with HLA-A2601 or -A0201 followed by testing their ability to stimulate IFN-γ production by the parental KE4 CTLs. Detailed methods are shown in the legend for Fig. 6.

Figure 9

Determination of aa required for CTL-mediated recognition. Each aa of the three nonapeptides (SART-1_736-744, SART-1_749-757, and SART-1_785-793) was substituted by glycine when it was not glycine, or threonine when it was glycine. These substituents along with the parental nonapeptides (10 μM) were loaded for 2 h to the VA13 cells transfected with HLA-A2601 or -A0201 followed by testing their ability to stimulate IFN-γ production by the parental KE4 CTLs. Detailed methods are shown in the legend for Fig. 6.

Figure 10

Induction of CTLs by the nonapeptides. The cells from total of 144 microcultures (48 microcultures from PBMCs alone, 48 stimulated with the SART-1_736-744, and 48 with the SART-1_749-747) were independently tested for their activity to produce IFN-γ in response to the KE3, KE4 tumor, and VA13 cells. The cells from 10 of 48 of the microcultures from the PBMCs stimulated with the SART-1_736-744 produced IFN-γ by recognition of the KE4, but not the other cells. The representative result of one microculture (subline 11) showing positive IFN-γ production is shown in the figure. In contrast, the cells from none of 48 of the microcultures besides one from either PBMCs alone or PBMCs stimulated with the SART-1_749-757 produced IFN-γ by recognition of the KE4 cells. The representative result of one microculture (subline 6 from PBMCs alone or subline 14 from PBMCs with the SART-1_749-757) showing negative IFN-γ production is shown in the figure.