Identification of an epitope derived from CML66, a novel tumor‐associated antigen expressed broadly in human leukemia, recognized by human leukocyte antigen‐A*2402‐restricted cytotoxic T lymphocytes

Abstract

CML66 is a newly identified differentiation antigen that is expressed broadly in human leukemia and solid tumors, but its physiological function remains unknown. In the present study, to clarify the feasibility of CML66‐targeted cancer immunotherapy, we attempted to identify cytotoxic T lymphocyte (CTL) epitopes derived from CML66. An immunogenic CML66‐derived epitope (amino acid residues 76–84; YYIDTLGRI) capable of inducing human leukocyte antigen (HLA)‐A*2402‐restricted CTL specific for this peptide was identified. CML66‐derived peptide‐specific CTL efficiently lysed human leukemia cells, but not normal cells, in a HLA‐A*2402‐restricted fashion. Quantitative real‐time polymerase chain reaction revealed that CML66 mRNA is expressed abundantly in primary acute myeloid leukemia cells, acute lymphoid leukemia cells, and chronic myelogenous leukemia cells in advanced phase, and that the expression level of CML66 mRNA in normal cells is low compared with that in leukemia cells. CML66‐specific CTL precursors were detected in the peripheral blood of patients with acute leukemia. These data indicate that the CML66‐derived epitope identified in the present study is a new target antigen for cellular immunotherapy of human leukemia. (Cancer Sci 2008; 99: 1414–1419)

In order to develop efficacious and safe cancer immunotherapy, it is essential to identify tumor‐associated antigens (TAA) that are expressed exclusively in tumor cells and can be recognized by human cytotoxic T lymphocytes (CTL). Since 1991, when the gene encoding the melanoma‐associated antigen MAGE‐1 was first cloned,^{( 1 )} various types of human TAA that are recognized by CD4⁺ and CD8⁺ T lymphocytes have been identified.^{( 2} , ^{3 )} However, the number of leukemia‐associated antigens identified is still limited. To extend the spectrum of cellular immunotherapy for leukemia, identification of new leukemia‐associated antigens is essential.

CML66 is a novel TAA that was identified by the method of serological analysis of cDNA expression libraries (SEREX) in the serum of a patient with relapsed chronic myeloid leukemia (CML) who received allogeneic hematopoietic stem cell transplantation and donor lymphocyte infusion.^{( 4 )} CML66 was initially considered to be a leukemia‐associated antigen that might mediate the graft‐versus‐leukemia effect.^{( 5 )} CML66 is a 583‐amino acid protein with a molecular weight of 66 kDa. The CML66 gene is located on the long arm of chromosome 8 (8q23). Screening of mRNA expression in normal tissue libraries has shown that the physiological and constitutive expression of CML66 mRNA is exclusive to the testis, with slight expression detected in the heart. In contrast, CML66 mRNA appears to be overexpressed in most cases of acute myeloid leukemia (AML), CML, and solid tumors, including lung cancer and prostate cancer.^{( 4} , ^{5 )} Although CML66 is expressed broadly in tumor cells, the physiological function of this molecule remains obscure. Because CML66 was discovered initially by SEREX, it is possible that a T lymphocyte‐mediated immune response is induced in patients with leukemia. These previous findings suggested that CML66 might represent a candidate target for novel immunotherapy of leukemia, and therefore we attempted to identify the CML66‐derived epitopes recognized by CTL restricted by human leukocyte antigen (HLA)‐A*2402, which is the most prevalent HLA class I molecule in the Japanese population.

In the present study, we successfully identified an antigenic CTL epitope of CML66 using the strategy of reverse immunology. CML66 peptide‐specific CTL exerted cytotoxicity against leukemia cells in a HLA‐A*2402‐restricted manner. In addition, CML66‐specific CTL precursors were detected in the peripheral blood of patients with leukemia. On the basis of these data, we discuss the feasibility of cellular immunotherapy for leukemia and solid tumors targeting CML66.

Materials and Methods

Cell lines and primary cells.  Epstein–Barr virus‐immortalized B‐lymphoblastoid cell lines (LCL) were generated from peripheral blood B lymphocytes of healthy volunteers. LCL were cultured continuously in RPMI‐1640 medium supplemented with 10% heat‐inactivated fetal calf serum (FCS). The HLA‐A*2402 gene‐transduced T2 cell line (T2‐A24)^{( 6 )} was cultured in RPMI‐1640 medium supplemented with 10% FCS and 800 µg/mL geneticin (Life Technologies, Rockville, MD, USA). The HLA‐A*2402 gene‐transduced C1R cell line (C1R‐A24)^{( 7 )} was cultured in RPMI‐1640 medium supplemented with 10% FCS and 500 µg/mL hygromycin B (Sigma, St Louis, MO, USA). All human leukemia cell lines were cultured similarly to LCL. Written informed consent was obtained from each set of parents for the donation of cord blood, and also from each patient with leukemia for the donation of bone marrow aspirate or peripheral blood, according to the declaration of Helsinki. Mononuclear cells were isolated from these samples by Ficoll–Conrey density gradient centrifugation and stored in a liquid nitrogen freezer until use. In some experiments, CD34⁺ cells were isolated from leukemic bone marrow cells and cord blood mononuclear cells (CBMC) by using anti‐CD34 monoclonal antibody (moAb)‐coated magnetic beads (Miltenyi Biotec, Bergisch Gladbach, Germany) according to the manufacturer's instructions.

Synthetic peptides.  Considering the dominant prevalence of HLA‐A24 (HLA‐A*2402) in the Japanese population, four nine‐amino‐acid (9‐mer) peptides derived from the CML66 sequence, which were predicted to bind with high affinity to the HLA‐A*2402 molecule, were designed by computer algorithms available at the BioInformatics and Molecular Analysis Section website (http://bimas.dcrt.nih.gov/molbio/hla_bind),^{( 8 )} and the SYFPEITHI website (http://www.syfpeithi.de/home.htm).^{( 9 )} The sequences of these synthetic 9‐mer peptides were as follows: CML66‐22, GYKLSLEPL (residues 22–30); CML66‐70, WYQDSVYYI (residues 70–78); CML66‐76, YYIDTLGRI (residues 76–84); and CML66‐217, KYEIIKRDI (residues 217–225). The italic amino acids indicate the binding motifs for the HLA‐A*2402 molecule. Other peptides, namely WT1‐235Y (CYTWNQMNL; a heteroclitic peptide of the HLA‐A*2402‐binding WT1 peptide WT1_235–243 with M at position 236 replaced by Y)^{( 10 )} and WT1‐7 (DLNALLPAV; an HLA‐A*0201‐binding WT1 peptide WT1_7–15),^{( 11 )} were also prepared as positive and negative controls, respectively, to evaluate their binding affinity for the HLA‐A*2402 molecule. All peptides used in the present study were synthesized with greater than 80% purity (Greiner Bio‐One, Tokyo, Japan).

Generation of CML66 peptide‐specific CTL.  Peptide‐specific CTL were generated as reported previously.^{( 12} , ^{13 )} Briefly, monocytes were purified from peripheral blood mononuclear cells (PBMC) using anti‐CD14 moAb‐coated magnetic beads (Miltenyi Biotec) according to the manufacturer's instructions, and then cultured in RPMI‐1640 medium supplemented with 10% FCS, 500 U/mL recombinant human interleukin (IL)‐4 (Genzyme, Boston, MA, USA), and 800 U/mL recombinant human granulocyte‐macrophage colony‐stimulating factor (Kirin Pharma, Tokyo, Japan) for 3 days to generate immature dendritic cells (DC). Then, 100 U/mL recombinant human tumor necrosis factor‐α (Dainippon Pharmaceutical, Osaka, Japan) was added to induce maturation of the DC. On day 8 or 9, the matured DC were harvested for peptide loading to stimulate autologous CD8⁺ T lymphocytes. Autologous CD8⁺ T lymphocytes were isolated from the same volunteer donor using CD8 micromagnetic beads (Miltenyi Biotec). One million CD8⁺ T lymphocytes were cocultured with 1.0 × 10⁵ target peptide‐loaded (10 µmol/L) autologous DC in RPMI‐1640 medium supplemented with 10% human AB serum, 5 ng/mL human recombinant IL‐7 (Genzyme), and 100 pg/mL human recombinant IL‐12 (Genzyme) in 96‐well round‐bottomed plates. On the next day, recombinant human IL‐2 (Boehringer Mannheim, Mannheim, Germany) was added at a final concentration of 10 U/mL. Thereafter, CD8⁺ T lymphocytes were restimulated once a week with identical peptide‐loaded and mitomycin C (Kyowa Hakko, Tokyo, Japan)‐treated autologous PBMC. The antigen specificity of growing CD8⁺ T lymphocytes was screened by cytotoxic assays.

Assay for binding of peptides to the HLA‐A*2402 molecule.  The binding affinities of the synthesized peptides for the HLA‐A*2402 molecule were evaluated by a major histocompatibility complex (MHC) stabilization assay using T2‐A24 cells, as reported previously.^{( 6 )} Briefly, 1 × 10⁶ T2‐A24 cells were incubated in 1 mL RPMI‐1640 medium with each peptide at a concentration of 10 µM at 27°C for 16 h, followed by additional incubation at 37°C for 3 h. Then, surface HLA‐A24 molecules were labeled with an anti‐HLA‐A24 moAb (One Lambda, Canoga Park, CA, USA) using standard procedures. The T2‐A24 cells were examined by flow cytometry (FACScan; Becton Dickinson, San Jose, CA, USA), and the mean fluorescence intensity (MFI) was analyzed with CellQuest software (Becton Dickinson). The fluorescence index (FI) was calculated using the following formula:

FI = (MFI with given peptide – MFI without peptide)/ (MFI without peptide).

Cytotoxicity assays.  The standard ⁵¹Cr‐release assay was carried out as described previously.^{( 14 )} Briefly, 5.0 × 10³ target cells that had been labeled with ⁵¹Cr (Na₂ ⁵¹CrO₄; New England Nuclear, Boston, MA, USA) were suspended in 100 µL RPMI‐1640 with 10% FCS, and then were seeded in round‐bottomed microtiter wells and incubated with or without target peptide for 90 min. To define the HLA restriction element, the target cells were incubated with an anti‐HLA class I framework moAb (w6/32; American Type Culture Collection, Manassas, VA, USA) or an anti‐HLA‐DR moAb (L243; American Type Culture Collection) at an optimal concentration (10 µg/mL) for 30 min before the addition of effector cells. Target cells were incubated with effector cells at various target‐to‐effector ratios in 200 µL assay medium in each well for 4 h, and then 100 µL supernatant from each well was collected and its radioactivity was measured with a scintillation counter. The percentage of specific lysis was calculated as follows:

Specific lysis (%) = (cpm experimental release – cpm spontaneous release)/(cpm maximal release – cpm spontaneous release) × 100.

Enzyme‐linked immunospot assay.  Enzyme‐linked immunospot (ELISPOT) assays were carried out as described previously.^{( 15 )} Briefly, 96‐well flat‐bottomed MultiScreen‐HA plates with a nitrocellulose base (Millipore, Bedford, MA, USA) were coated with 10 µg/mL anti‐interferon (IFN)‐γ moAb (R & D Systems, Minneapolis, MN, USA) and incubated overnight at 4°C. After being washed with phosphate‐buffered saline (PBS), the plates were blocked with the assay medium for 1 h at 37°C. C1R‐A24 cells (5.0 × 10⁴/well) were pulsed with CML66 peptide at a concentration of 10 µmol/L or with PBS alone, and incubated in RPMI‐1640 medium with 10% FCS for 1 h at room temperature. Then, CD8⁺ responder cells were seeded into each well to mix with the target peptide‐loaded C1R‐A24 cells, and the plates were incubated in a 5% CO₂ incubator at 37°C for 20 h. After incubation, plates were washed vigorously with PBS containing 0.05% Tween 20. A polyclonal rabbit anti‐IFN‐γ antibody (Endgen, Woburn, MA, USA) was added to each well and the plates was left for 90 min at room temperature, followed by exposure to peroxidase‐conjugated goat anti‐rabbit IgG (Zymed, San Francisco, CA, USA) for an additional 90 min. To reveal IFN‐γ‐specific spots, 100 µL of 0.1 mol/L sodium acetate buffer (pH 5.0) containing 3‐amino‐9‐ethylcarbazole (Sigma‐Aldrich Japan, Tokyo, Japan) and 0.015% H₂O₂ were added to each well. After 40 min, the color reaction was interrupted by washing with water, and the plates were dried. Diffuse large spots were counted under a dissecting microscope.

Quantitative real‐time polymerase chain reaction for CML66 mRNA expression.  Total RNA was extracted from each sample with an Rneasy Mini kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. Quantitative real‐time polymerase chain reaction (QRT‐PCR) was carried out in a final volume of 20 µL with the One‐Step Reverse Transcription–Polymerase Chain Reaction Master Mix Reagents kit (Applied Biosystems, Foster City, CA, USA) according to the manufacturer's instructions. The reaction was carried out with 0.1 µg total RNA from each sample by using the ABI Prism 7700 Sequences Detection System (Applied Biosystems). Reverse transcription of the RNA was achieved at 48°C for 30 min, and polymerase chain reaction was carried out with an enzyme‐activation step at 95°C for 10 min, followed by 40 cycles of denaturation (95°C for 15 s) and annealing and extension (60°C for 1 min). Primers and probe for CML66 were synthesized by the TaqMan Gene Expression Assays service (Applied Biosystems; the license does not permit disclosure of detailed information). Simultaneous quantitative analysis of mRNA for glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH) as the standard control was carried out with TaqMan assay reagent target kits (Applied Biosystems). All experiments were carried out in duplicate. To normalize differences in RNA loading for the reverse transcription–polymerase chain reaction procedure, the CML66 mRNA expression level for a particular sample was defined relative to the expression level of the GAPDH gene in that sample. The CML66 gene expression level in the K562 leukemia cell line, which expresses CML66 strongly, was designated as 1.0 and the levels for the experimental samples were calculated relative to this level (comparative ΔCt method).

Results

CML66‐derived peptide, CML66‐76, can bind to the HLA‐A*2402 molecule and elicit peptide‐specific CTL.  Of the four algorithmically predicted epitopes derived from CML66, CML66‐70 and CML66‐76 bound substantially bind to the HLA‐A*2402 molecule with FI values of 3.04 and 3.34, respectively, compared with 0.13 for the negative‐control peptide WT1‐7 and 3.76 for the positive‐control peptide WT1‐235Y (Table 1). Using the methods described above, we successfully generated a CTL line specific for the CML66‐76 peptide. The target epitope specificity of the CML 66‐76‐specific CTL line was determined by the ⁵¹Cr‐release assay. At the initial screening, CTL lines that lysed the target peptide‐loaded autologous LCL to various degrees were generated from four out of four HLA‐A*2402‐positive healthy individuals. Among them, the CTL line, which comprised 20.1% IFN‐γ‐producing CD8⁺ T cells in response to stimulation with CML66‐76 peptide determined by ELISPOT assay, was established and used for the further experiments. As shown in Figure 1, the CTL line lysed CML66‐76 peptide‐loaded autologous LCL cells in a dose‐dependent manner, but did not show any cytotoxicity against peptide‐unloaded autologous LCL or CML66‐76 peptide‐loaded HLA‐A*2402‐negative allogeneic LCL. These findings indicate that the CML66‐76 peptide can bind to the HLA‐A*2402 molecule and elicit peptide‐specific CTL in a HLA‐A*2402‐restricted fashion.

Table 1.

Binding affinities of synthetic peptides

Name (start position)	Sequence	HLA	Predictive binding score (BIMAS)	Predictive binding score (SYFPEITHI)	Fluorescence intensity
CML66‐217 (217–225)	KYEIIKRDI	HLA‐A*2402	210	22	0.54
CML66‐22 (22–30)	GYKLSLEPL	HLA‐A*2402	200	20	0.56
CML66‐70 (70–78)	WYQDSVYYI	HLA‐A*2402	90	22	3.04
CML66‐76 (76–84)	YYIDTLGRI	HLA‐A*2402	90	23	3.30
WT1‐7 (7–15)	DLNALLPAV	HLA‐A*0201	0.18	0	0.13
WT1–235Y (235–243)	CYTWNQMNL	HLA‐A*2402	200	20	3.76

The binding affinities of synthetic peptides for HLA molecules were predicted by computer algorithms available at the BioInformatics and Molecular Analysis Section (BIMAS) website (http://bimas.dcrt.nih.gov/molbio/hla_bind) and the SYFPEITHI website (http://www.syfpeithi.de/home.htm). The binding affinities of synthetic peptides for HLA molecules were evaluated using an MHC stabilization assay.

Figure 1.

CML66‐76 peptide‐specific and concentration‐dependent cytotoxicity of a cytotoxic T lymphocyte (CTL) line. CML66‐76‐specific CTL were generated and their cytotoxicity against CML66‐76 peptide‐loaded autologous lymphoblastoid cell lines (LCL) (open circles), CML66‐76 peptide‐loaded HLA‐A24‐negative allogeneic LCL (closed circles), and peptide‐unloaded autologous LCL (open triangles) was determined by ⁵¹Cr‐release assays at an effector : target cell ratio of 5:1. Target cells were preincubated with and without CML66‐76 peptide at various concentrations for 1 h.

HLA‐A*2402‐restricted cytotoxicity mediated by CML66‐76 peptide‐specific CTL.  The restriction element of CML66‐76 peptide‐specific CTL generated in the present study appeared to be HLA‐A*2402, because HLA‐A*2402‐positive, but not HLA‐A*2402‐negative, allogeneic LCL were lysed by CTL. To confirm HLA‐A*2402 restriction of CML66‐76 peptide‐specific CTL, we examined their cytotoxic activity against the HLA‐A*2402 gene‐transfectant cell line C1R‐A*2402. CTL were cytotoxic to C1R‐A*2402 cells only in the presence of CML66‐76 peptide (Fig. 2a), and this cytotoxicity was significantly attenuated by anti‐HLA‐class I moAb but not by anti‐HLA‐DR moAb (Fig. 2b). These data show that the cytotoxicity of the CML66 peptide‐specific CTL generated in the present study was restricted by HLA‐A*2402.

Figure 2.

HLA‐A*2402‐restriction of cytotoxicity mediated by a cytotoxic T lymphocyte (CTL) line. (a) The cytotoxicity of CML66 peptide‐specific CTL against CML66‐76 peptide‐loaded and ‐unloaded autologous lymphoblastoid cell lines (LCL), HLA‐A*2402‐positive and ‐negative allogeneic LCL, and peptide‐loaded and ‐unloaded C1R‐A24 cells was examined by ⁵¹Cr‐release assays at effector : target cell ratios of 5:1. Target cells were preincubated with or without CML66‐76 peptide at a concentration of 10 µmol/L for 1 h. Asterisks indicate values less than 0%. (b) The cytotoxicity of CML66‐76 peptide‐specific CTL against C1R‐A24 cells pulsed with CML66‐76 peptide at a concentration of 10 µmol/L was determined by ⁵¹Cr‐release assay at an effector : target cell ratio of 5:1 in the presence or absence of anti‐HLA class I monoclonal antibody or anti‐HLA‐DR monoclonal antibody.

CML66‐76 peptide‐specific CTL line can efficiently lyse HLA‐A*2402‐bearing human leukemia cells.  QRT‐PCR for CML66 mRNA revealed that all leukemia cell lines examined appeared to express CML66 mRNA abundantly compared with normal cells (Fig. 3). As shown in Figure 4a, CML66‐76 peptide‐specific CTL showed cytotoxicity against HLA‐A*2402‐positive human leukemia cell lines (OUN‐1, MEG01) but not against HLA‐A*2402‐negative leukemia cell lines (KT‐1, KG‐1, and K562). In contrast to the susceptibility of leukemia cells to CML66‐specific CTL, HLA‐A*2402‐positive normal PBMC and CBMC appeared to be resistant to cytotoxicity mediated by CML66‐specific CTL. These results indicate that CML66‐76 peptide‐specific CTL exert cytotoxicity against leukemia cells bearing HLA‐A*2402 molecules but not against normal cells bearing HLA‐A*2402 molecules. The cytotoxicity of CML66‐76‐specific CTL against leukemia cells was significantly inhibited by anti‐HLA‐class I moAb, but not by anti‐HLA‐DR moAb, indicating that the cytotoxicity of CML66‐specific CTL against peptide‐unloaded leukemia cells is HLA‐class I restricted (Fig. 4b). These data indicate that the CML66 protein is processed in leukemia cells to a CML66‐derived peptide, which forms a complex with the HLA‐A*2402 molecule that can be recognized by HLA‐A*2402‐restricted CTL.

Figure 3.

Expression of CML66 mRNA in leukemia cell lines and normal cells. Expression levels of CML66 mRNA in various leukemia cell lines and normal peripheral blood mononuclear cells (PBMC) and normal cord blood mononuclear cells (CBMC) were determined by quantitative real‐time polymerase chain reaction. The CML66 mRNA expression level in the K562 leukemia cell line is shown as 1.0 and the expression levels in leukemia and normal cells were calculated relative to this value.

Figure 4.

Cytotoxicity of CML66‐specific cytotoxic T lymphocytes (CTL) against leukemia and normal cells. (a) The cytotoxicity of CML66‐76 peptide‐specific CTL against HLA‐A24‐positive and ‐negative leukemia cells, normal peripheral blood mononuclear cells (PBMC), and cord blood mononuclear cells (CBMC) was determined by ⁵¹Cr‐release assays at various effector : target cell ratios. (b) The cytotoxicity of CML66‐76 peptide‐specific CTL against leukemia cells (OUN‐1 and MEG01) was determined by the ⁵¹Cr‐release assay in the presence or absence of anti‐HLA class I monoclonal antibody or anti‐HLA‐DR monoclonal antibody. Experiments were carried out at an effector : target cell ratio of 10:1.

Primary AML cells, ALL cells, and CML cells in blast crisis express high levels of CML66 mRNA.  QRT‐PCR revealed that AML cells, acute lymphocytic leukemia (ALL) cells, chronic myelomonocytic leukemia cells, and CML cells in blast crisis all expressed CML66 mRNA abundantly compared with normal cells. Although the expression level of CML66 mRNA in whole bone marrow cells isolated from patients with CML in chronic phase was relatively low, CML CD34⁺ cells appeared to express CML66 abundantly. In contrast, CD34⁺ cells isolated from normal CBMC expressed a low amount of CML66 mRNA compared with CML CD34⁺ cells (Fig. 5a). With respect to the French‐American‐British (FAB) classification of AML, all subtypes from M0 to M4, particularly M2, expressed CML66 abundantly, suggesting that all types of AML in addition to ALL might be suitable for CML66‐targeting immunotherapy (Fig. 5b).

Figure 5.

Expression of CML66 mRNA in primary leukemia cells. (a) Expression levels of CML66 mRNA in various types of freshly isolated leukemia cells, CD34⁺ cells isolated from bone marrow mononuclear cells of the patients with chronic myeloid leukemia in chronic phase, and CD34⁺ cells isolated from normal cord blood mononuclear cells were determined by quantitative real‐time polymerase chain reaction. (b) Expression levels of CML66 mRNA in various types of leukemia cells classified by FAB classification. The CML66 mRNA expression level in the K562 leukemia cell line, which strongly expresses CML66, is shown as 1.0 and the expression levels in samples were calculated relative to this value.

CML66‐specific CTL precursors are present in patients with AML.  Finally, we examined whether CML66‐specific CTL precursors can be detected in the peripheral blood of patients with acute leukemia and healthy individuals. PBMC samples from six patients with acute leukemia in complete remission, including three cases of AML (two cases of FAB M2 and one case of FAB M1) and three cases of ALL (all FAB L2), and three healthy individuals were used for the ELISPOT assays. In response to stimulation with CML66‐76 peptide, IFN‐γ‐producing CD8⁺ T lymphocytes were detected clearly in peripheral blood from five out of six patients with acute leukemia and one out of three healthy individuals (Fig. 6a). For example, the frequencies of IFN‐γ‐producing T lymphocytes in peripheral blood from a patient with AML‐M2 were 1.15 and 0.35% following stimulation with and without CML66‐76 peptide, respectively. In another case of AML‐M2, the frequencies of IFN‐γ‐producing T lymphocytes in response to stimulation with and without CML66‐76 peptide were 0.54 and 0.1%, respectively (Fig. 6b). These data clearly show that CML66‐specific CTL precursors are present in patients with leukemia.

Figure 6.

Detection of CML66‐specific cytotoxic T lymphocyte precursors in peripheral blood of patients with acute leukemia and healthy individuals. (a) Summary of enzyme‐linked immunospot (ELISPOT) assays for interferon (IFN)‐γ production using peripheral blood mononuclear cells (PBMC) from six patients with acute leukemia and three healthy individuals. (b) Representative data of ELISPOT assays using PBMC from patients with acute leukemia. ELISPOT assays for IFN‐γ production were carried out by incubating PBMC with C1R‐A24 cells pulsed with or without CML66‐76 peptide. These figures show examples of results from two patients with AML‐M2.

Discussion

In the present study, we demonstrated for the first time that CML66, previously identified by SEREX using serum from a patient with CML, is a novel TAA recognized by HLA‐A*2402‐restricted CTL. The physiological function of CML66 remains unknown. There are no significant homologies with the cDNA sequence of CML66 in GenBank. The cDNA sequence of CML66 contains no defined functional motifs and thus the function of this molecule is entirely unknown. Despite its undetermined physiological role, overexpression of CML66 is widespread among human cancers, including leukemia and some solid tumors.^{( 4} , ^{5 )} The CML66 gene is located on human chromosome 8q23.^{( 4 )} Balanced and unbalanced chromosomal abnormalities involving chromosome 8q23 have been found in some types of leukemia and non‐Hodgkin's lymphoma. In addition, genes on the 8q arm are also amplified frequently in advanced prostate cancer.^{( 16} , ^{17 )} Although, to the best of our knowledge, there is no direct evidence that the overexpression of CML66 can induce tumorigenesis, these findings strongly suggest that CML66 plays an important role in the development and progression of malignancies, and further studies will be necessary to determine the mechanism for overexpression of CML66 in human tumors and to identify the function of this molecule.

Both in chronic phase and in blast crisis of CML, CML66 expression in bone marrow appears to be exclusive to CD34⁺ cells, which comprise presumed leukemic stem cells.^{( 4 )} CML66‐specific CTL efficiently lysed leukemia cell lines established from CML cells in blast crisis, presumably originating from CML stem cells. Although there remains no direct evidence that CML66‐specific CTL actually suppress the growth of CML stem cells, CTL directed against CML66 might exert an antileukemic effect against leukemic stem cells in CML as well as in AML. The successful anti‐CML molecular target drug imatinib cannot totally eradicate leukemic stem cells in CML,^{( 18 )} whereas Abl kinase inhibitor in combination with vaccination using CML66 peptide might be clinically effective.

It was previously reported that both normal bone marrow CD34⁺ cells and CML CD34⁺ cells express CML66.^{( 5 )} When considering the feasibility of CML66‐targeting immunotherapy, this may raise a concern that normal CD34⁺ hematopoietic progenitor cells might be damaged by CML66‐specific CTL. In the present study, QRT‐PCR revealed that normal CD34⁺ cells do express CML66 mRNA; however, the expression level of CML66 mRNA in normal CD34⁺ cells appears to be significantly lower than in leukemic CD34⁺ cells. In addition, lower expression of HLA class I molecules on hematopoietic progenitor cells in steady state has been reported.^{( 19 )} These data strongly suggest that CML66‐specific CTL may not damage normal hematopoiesis and immunotherapy targeting CML66 must be safe. According to the online database SymAtlas (http://symatlas.gnf.org), CML66 mRNA is expressed at relatively high levels in the atrioventricular node and normal ciliary ganglion. Therefore, unexpected expression of CML66 in normal tissues should be considered at the initiation of immunotherapy targeting CML66.

CML66 is known to have two spliced isoforms: CML66L and CML66S.^{( 20 )} The epitope of CML66 identified in the present study, CML66‐76, is preserved in both CML66L and CML66S (data not shown). In addition, no amino acid substitutions have been reported in this epitope, as described previously.^{( 4 )} These properties might be advantageous for induction of cancer‐specific CTL, because the CML66‐76 epitope may be preserved broadly in most cancer cells in addition to leukemia.

In conclusion, this is the first report showing that a CML66‐derived peptide has the potential to induce CTL that can lyse human leukemia cells in the context of HLA‐A*2402, which is the most common HLA class I type in the Japanese population. Although further studies are needed before initiation of immunotherapy targeting CML66, the present findings may contribute to the development of novel immunotherapeutic strategies for leukemia, and suggest that vaccination with CML66‐derived peptides may provide an effective treatment option for leukemia. As CML66 is also expressed abundantly in many solid tumors, immunotherapy targeting CML66 may be effective for various cancers as well as leukemia.