An HLA-A3-binding prostate acid phosphatase-derived peptide can induce CTLs restricted to HLA-A2 and -A24 alleles

Abstract

We previously reported peptide vaccine candidates for HLA-A3 supertype (-A3, -A11, -A31, -A33)-positive cancer patients. In the present study, we examined whether those peptides can also induce cytotoxic T lymphocyte (CTL) activity restricted to HLA-A2, HLA-A24, and HLA-A26 alleles. Fourteen peptides were screened for their binding activity to HLA-A*0201, -A*0206, -A*0207, -A*2402, and -A*2601 molecules and then tested for their ability to induce CTL activity in peripheral blood mononuclear cells (PBMCs) from prostate cancer patients. Among these peptides, one from the prostate acid phosphatase protein exhibited binding activity to HLA-A*0201, -A*0206, and -A*2402 molecules. In addition, PBMCs stimulated with this peptide showed that HLA-A2 or HLA-A24 restricted CTL activity. Their cytotoxicity toward cancer cells was ascribed to peptide-specific and CD8⁺ T cells. These results suggest that this peptide could be widely applicable as a peptide vaccine for HLA-A3 supertype-, HLA-A2-, and -A24-positive cancer patients.

Introduction

The field of peptide vaccines for cancer patients is currently in an active state of clinical investigation [1–6]. Peptide-based immunotherapy, however, is highly restricted by HLA-A alleles, and ongoing clinical trials are mostly limited to HLA-A2- or -A24-positive patients, primarily because of the higher worldwide frequencies of these alleles [7]. Subsequently, cancer patients with alleles other than HLA-A2 or -A24 have been out of the scope of the current peptide-based immunotherapy, and candidate peptides widely applicable to such patients are needed.

It has become clear that approximately 20% of cytotoxic T lymphocyte (CTL) epitope peptides could induce CTL activity restricted to more than one HLA-class IA allele [8]. We recently reported that one HLA-A24-restricted CTL epitope peptide could induce such CTL activity restricted to multiple HLA-class IA alleles [9]. In the present study, we examined whether or not the 14 different peptides that can induce HLA-A3 supertype-restricted CTL activity as reported previously [10–14] can also induce CTL activity restricted to HLA-A2 and HLA-A24 alleles in peripheral blood mononuclear cells (PBMCs) of prostate cancer patients. In addition, we have reported one such peptide derived from prostate acid phosphatase (PAP) protein, an appropriate target molecule for a cancer vaccine because of its overexpression in prostate cancer and also colon and gastric cancer, as reported previously [15].

Materials and methods

Patients

Peripheral blood mononuclear cells were obtained from prostate cancer patients who had provided written informed consent. None of the participants were infected with HIV. A total of 20 mL of peripheral blood was obtained, and PBMCs were prepared by the Ficoll-Conray density (1.077 at room temperature) centrifugation method. All of the samples were cryopreserved until used for the experiments. The expression of HLA-A2 and HLA-A24 molecules on PBMCs of cancer patients was determined by flow cytometry using the anti-HLA-A2 monoclonal antibody (mAb) and anti-HLA-A24 mAb, while HLA-A2 subtypes of HLA-A2⁺ patients were determined using DNA-based HLA typing with the Luminex Multi-Analyte Profiling system (xMAP) with the WAKFlow HLA typing kit (Wakunaga, Hiroshima, Japan) [16].

Cell lines

The cell lines used as target cells for cytotoxicity were COLO205 (HLA-A*0101/A*0201) and COLO320 (HLA-A*2402) colon cancer cell lines and PC3 (HLA-A*2402) and LNCaP (HLA-A*0201) prostate cancer cell lines. All of these cell lines highly express PAP antigens, as reported previously [15]. LNCaP cells weakly express major histocompatibility complex (MHC) class I molecules on the surface [17]. Subsequently, we used COLO205 or COLO320 tumor cells to measure HLA-A2-restricted CTL activity as relevant or irrelevant target cells, respectively. In addition, PC3 and LNCaP tumor cells were used to measure HLA-A24-restricted CTL activity as relevant and irrelevant target cells, respectively. All tumor cell lines were maintained in RPMI-1640 (Invitrogen, Carlsbad, CA) medium supplemented with 10% fetal calf serum (MP Biomedicals, Inc, Germany). For pulsing a peptide to measure interferon-γ (IFN-γ) production by peptide-stimulated PBMCs, as reported previously [15], we used T2 cells (HLA-A2, T-B hybridoma) or C1R-A*2402 cells (HLA-A*2402 gene-stable transfectant of the B lymphoblastoid cell line, kindly provided by Dr. M. Takiguchi, Kumamoto University, Japan), as also reported previously [18]. RMA-S cells were derived from a mouse mutant cell line deficient in antigen processing and that showed decreased cell surface expression of MHC class I molecules. The HLA-A*0201, -A*0206, -A*0207, -A*2402, and -A*2601 genes were also individually transfected into RMA-S cells using the FuGENE transfection reagent (Roche, Mannheim, Germany). Stably transfected cloned cells were established from a separate well in the presence of genetecin (0.5 mg/mL). The detailed methods for establishing these transfectants have been reported previously [11].

Peptides

Peptides with more than 90% purity were purchased from Hokkaido System Science (Sapporo, Japan) or Genenet (Fukuoka Japan), dissolved in dimethyl sulfoxide (DMSO) at a concentration of 10 μg/mL. Fourteen peptides (Table 1) capable of inducing HLA-A3 supertype (-A3, -A11, -A31, and -A33)-restricted CTL, as reported previously [10–14], were used in this study. In addition, Epstein-Barr virus (EBV)-derived and HIV-derived peptides were used as controls binding to HLA-A2 and -A24 alleles, as reported previously [19–22]. Furthermore, the positive-control peptide used for HLA-A26 was NS3_1582–1590 [23].

Table 1.

Peptide sequences used in this study

Peptide	Amino acids sequence	Binding score (BIMAS)						References
		A3	A*1101	A*3101	A*3302	A*0201	A24
SART3511–519	WLEYYNLER	24.000	0.160	4.000	9.000	0.004	0.017	[12]
SART3734–742	QIRPIFSNR	2.700	0.080	4.000	15.000	0.000	0.020	[12]
Lck90–98	ILEQSGWWK	60.000	0.800	1.000	0.300	0.067	0.015	[14]
Lck449–458	VIQNLERGYR	0.120	0.080	2.000	15.000	0.002	0.015	[14]
PAP248–257	GIHKQKEKSR	0.600	0.120	1.000	15.000	0.002	0.010	[10]
PSA16–24	GAAPLILSR	0.540	0.240	0.800	3.000	0.000	0.014	[10]
IEX147–56	APAGRPSASR	0.090	0.040	0.200	3.000	0.000	0.010	[13]
β-tubulin5154–162	KIREEYPDR	1.800	0.240	6.000	4.500	0.003	0.024	[11]
β-tublin5309–318	RYLTVAAVFR	0.006	0.036	3.600	4.500	0.000	1.500	[11]
CGI37-72	KFTKTHKFR	0.006	0.060	0.900	0.900	0.000	0.100	[11]
PAP155–163	YLPFRNCPR	4.000	0.080	2.000	9.000	0.069	0.015	[10]
PSMA207–215	KVFRGNKVK	15.000	6.000	0.900	0.150	0.007	0.020	[10]
IEX61–69	RSRRVLYPR	0.135	0.024	0.600	4.500	0.000	0.028	[13]
Lck450–459	IQNLERGYR	0.036	0.120	2.000	3.000	0.002	0.015	[14]
EBV-A2	GLCTLVAML	5.400	0.012	0.040	0.300	49.134	4.800	[18]
EBV-A24	TYGPVFMCL	0.008	0.008	0.024	0.500	0.006	403.200	[19]
HIV-A2	SLYNTVATL	9.000	0.008	0.120	0.300	157.227	4.000	[20]
HIV-A24	RYLRQQLLGI	0.003	0.036	0.144	0.150	0.012	150.000	[21]
NS31582–1590	DNFPYLVAY	0.081	0.000	0.001	0.030	0.000	0.120	[22]

The peptide-binding score was calculated based on the predicted half time of dissociation from HLA-class I molecules, as obtained from a website (Bioinformatics and Molecular Analysis Section, Computational Bioscience and Engineering Laboratory, Division of Computer Research and Technology, the National Institutes of Health)

EBV Epstein-Barr virus, HLA human leukocyte antigen

HLA stabilization assay

Binding of peptides to HLA-A*0201, -A*0206, -A*0207, -A*2402, and -A*2601 molecules was examined using the stabilization assay according to a previously reported method with several modifications [9]. Briefly, RMA-S-A*0201, -A*0206, -A*0207, -A*2402, and -A*2601 (5 × 10⁵ cells per well in a 24-well plate) were cultured at 26°C for 20 h in 500 μL RPMI 1640 supplemented with 20% fetal bovine serum (FBS) followed by incubation with 500 μL Opti-MEM (Invitrogen) containing 0.1–100 μM peptides and human β2 microglobulin (2 μg/mL) at 26°C for 2 h and then at 37°C for 3 h. After being washed with PBS, the cells were incubated for 30 min on ice with an appropriate dilution of anti-HLA-A24 or BB7.2 supernatant (anti-HLA-A2). After two more washes with PBS, the cells were incubated for 30 min on ice with Alexa Fluor 488 goat anti-mouse IgG (Invitrogen). Peptides that, when incubated with cells at a concentration of 100 μM, gave more than 25% of the mean fluorescence intensity of cells cultured at 26°C, were defined as binding peptides.

Induction of peptide-specific CTLs from PBMCs

The assay for peptide-specific CTLs from prostate cancer patients was carried out according to a previously reported method with several modifications [10]. PBMCs (1 × 10⁵ cells per well in a 96-well U-bottom-type plate) were incubated with 10 μg/mL of each peptide in culture medium in a quadruplicate assay. The culture medium consisted of 45% RPMI-1640, 45% AIM-V medium (Life Technologies, Gaithersburg, MD), 10% FBS, 100 U/mL interleukin-2, and 0.1 mM MEM nonessential amino acid solution (Life Technologies). On day 15 of culture, cells were separated into four wells. Two wells were used for the culture with the corresponding peptide-pulsed T2 or C1R-A*2402 cells, and the other two wells were used for the culture with the irrelevant (HIV) peptide. After an 18-h incubation, IFN-γ production was determined by enzyme-linked immunosorbent assay. The induction of peptide-specific CTLs was judged to be successful when the P value was less than 0.05 and when the difference in IFN-γ production compared to that of the control HIV peptide exceeded 50 pg/mL.

Cytotoxicity assay

Peptide-stimulated PBMCs were tested for their cytotoxicity against PC3, LNCaP, COLO205, and COLO320 cells, which were pre-pulsed with either a corresponding peptide or the HIV peptide by a standard 6-h ⁵¹Cr-release assay [13]. Phytohemagglutinin (PHA)-activated T cells were used as a negative control. A total of 2,000 ⁵¹Cr-labeled cells per well were cultured with effector cells in 96 round-well plates at the indicated effector/target ratio. Immediately before the cytotoxicity assay, CD8⁺ T cells were positively isolated using a CD8-positive Isolation Kit (Dynal, Oslo, Norway). The specific ⁵¹Cr release was calculated according to the formula: %specific lysis = (test sample release − spontaneous release) × 100/(maximum release − spontaneous release). Spontaneous release was determined by the supernatant of the sample incubated with no effector cells, and the total release was then determined by the supernatant of the sample incubated with 1% Triton 100-X (Wako Pure Chemical Industries, Osaka, Japan). In some experiments, 10 μg/mL of either anti-HLA class I (W6/32: mouse IgG2a), anti-HLA-DR (L243: mouse IgG2a), or anti-HLA-B, C (B1-23, IgG2a; kindly donated by Dr. Pierre G. Coulie, Catholique de Louvain University, Brussels, Belgium) were added into the wells at the initiation of the culture.

The specificity of peptide-stimulated CTLs was confirmed by a cold inhibition assay. In brief, ⁵¹Cr-labeled target cells (2 × 10⁴ cells per well) were cultured with the effector cells (2 × 10⁴ cells per well) in 96 round-well plates with 2 × 10⁴ cold target cells. T2 or C1R-A*2402, pulsed with either the HIV peptide or a corresponding peptide, were used as cold target cells.

Statistical analysis

The results of the absorption test were analyzed by a two-tailed Student’s t test. All tests of significance were two sided.

Results

HLA stabilization assay

We first screened the binding activity of each of the 14 different HLA-A3 supertype peptides (100 μM) to HLA-A*0201, -A*0206, -A*0207, -A*2402, and -A*2601 alleles, by means of an HLA stabilization assay using RMA-S cells expressing each HLA molecule. Only the PAP_155–163 peptide showed binding activity to HLA-A*0201 or -A*0206 molecules, while PAP_155–163 peptide and β-tubulin 5_309–318 peptide were positive for binding to HLA-A*2402 molecules (data not shown). Subsequently, we focused on the PAP_155–163 peptide; the study of the β-tubulin 5_309–318 peptide is described in another study.

The surface expression of HLA-A*0201 molecules on RMA-S-A*0201 cells was stabilized in a dose-dependent manner when cells were cultured with either a positive control or the PAP_155–163 peptide (MFI increases at 0.1, 1, 10, and 100 μM were 7, 22, 69, and 106%, respectively), whereas the expression was not stabilized when cells were cultured with DMSO alone (Fig. 1a). Similarly, HLA-A*0206 and -A*2402 molecules on RMA-S-A*0206 and -A*2402 cells, respectively, were stabilized when cultured with PAP_155–163 peptide (Fig. 1b, c). However, this peptide failed to stabilize either HLA-A*0207 or -A*2601 molecules on RMA-S-A*0207 or -A*2601 cells, respectively (Fig. 1d, e).

Fig. 1.

Stabilization assay of the PAP_155–163 peptide for various HLA alleles. The binding activities of the PAP_155–163 peptide to HLA-A*0201 and -A*2402 were examined using stable transfectant cell lines RMA-S-A*0201 (a), -A*0206 (b), -A*2402 (c), -A*0207 (d), and -A*2601 (e) with a positive-control peptide and negative control (DMSO). The positive-control peptides used for each HLA were HIV-A2 (a, b), HIV-A24 (c), EBV-A2 (d), and NS3 _1582–1590 (e). Mean fluorescence intensity (MFI) was recorded at 0.1, 1, 10, and 100 μM of the peptide or DMSO. The MFI increase induced by the PAP_155–163 peptide compared with DMSO was calculated and is shown in a, b, and c. However, the MFI levels induced by the PAP_155–163 peptide were comparable in degree to DMSO (d), (e). Representative results from at least three separate experiments are shown. Asterisks statistically significant at P < 0.05

Induction of peptide-specific CTL activity

We attempted to determine whether or not the PAP_155–163 peptide has the potential to generate peptide-specific CTLs from prostate cancer patients by the IFN-γ production assay. PBMCs with different HLA alleles were stimulated in vitro with PAP1_55–163 peptide, positive (EBV) control peptide, or negative (HIV) control peptide, followed by measurement of IFN-γ production in response to the appropriate peptide-pulsed cells. As a result, this peptide induced peptide-specific CTL activity from the PBMCs of both HLA-A*0201 (patients 1, 3, and 6) and HLA -A*0206/A*0207 (patient 2) patients among six prostate cancer patients tested (Table 2). Three of the six patients showed CTL activity reactive to EBV-derived peptide but not to HIV-derived peptide. The PAP_155–163 peptide-induced CTL activity in four out of six HLA-A*2402 prostate cancer patients tested (Table 2). Three of the six patients showed CTL activity reactive to EBV-derived peptide but not to HIV-derived peptide.

Table 2.

Induction of peptide-specific CTL from prostate cancer patients

Pt	HLA allele	Target cells	IFN-γ production (pg/mL) in response to
			PAP155–163	EBV-A2	HIV-A2	EBV-A24	HIV-A24
HLA-A*0201/A*0206
1	A0201/-	T2	50	155	23	–	–
2	A0206/A0207	T2	308	10	45	–	–
3	A0201/-	T2	50	578	29	–	–
4	A0201/-	T2	30	5	0	–	–
5	A0201/A0206	T2	18	126	2	–	–
6	A0201/A0206	T2	139	35	22	–	–
HLA-A*2402
7	A2402/-	C1R-A*2402	90	–	–	71	0
8	A2402/-	C1R-A*2402	51	–	–	51	0
9	A2402/-	C1R-A*2402	78	–	–	0	21
10	A2402/-	C1R-A*2402	37	–	–	14	44
11	A2402/-	C1R-A*2402	13	–	–	539	1
12	A2402/-	C1R-A*2402	70	–	–	44	8

The PBMCs from HLA-A*0201, -A*0206, and -A*2402 prostate cancer patients were induced in vitro with PAP_155–163 peptide. On day 15, the cultured PBMCs were tested for their reactivity to C1R-A*2402 and T2 cells, which were pre-pulsed with a corresponding peptide or the HIV peptide. The peptide-specific CTL was measured by means of IFN-γ release assay with ELISA. Representative results are shown. Background IFN-γ production in response to the HIV peptide was subtracted from that in response to PAP_155–163 peptide

The induction of peptide-specific CTLs was judged to be successful when the P value was less than 0.05 and when the difference in IFN-γ production compared to that of the control HIV peptide exceeded 50 pg/mL. In this table, the value is shown as bold only when there is a significant difference between the two groups

Cytotoxicity assay

We then determined whether or not the CTLs induced by in vitro stimulation with the PAP_155–163 peptide could show cytotoxicity against cancer cells. The PBMCs from HLA-A*0201 (patients 1 and 3) or -A*0206/A*0207 (patient 2) patients, which were stimulated in vitro with PAP_155–163 peptide, exhibited significant levels of cytotoxicity against COLO205 cells but not against COLO320 cells or HLA-A2⁺ PHA-stimulated T-cell blasts. Similarly, the PBMCs from HLA-A*2402 patients showed that this peptide possessed the ability to induce PC3-reactive CTLs from the PBMCs of HLA-A*2402 patients, but not from LNCaP cells or HLA-A2402⁺ PHA-stimulated T-cell blasts (Fig. 2). We further examined CTL activity by means of the chromium release assay in PBMCs from four patients whose samples responded negatively in the IFN-γ assay, as shown in Table 2. None of the patients showed detectable levels of CTL activity with this assay.

Fig. 2.

Cytotoxicity of peptide-stimulated PBMCs from HLA-A*0201, -A*0206, and -A*2402 prostate cancer patients. Peptide-stimulated PBMCs were tested for their cytotoxicity toward three different targets by a 6-h ⁵¹Cr-release assay. Phytohemagglutinin (PHA)-stimulated T-cell blasts were included as HLA-A2⁺ or -A24⁺ normal cells. Asterisks statistically significant at P < 0.05

Taken together, these results indicate that the PBMCs stimulated in vitro with this peptide can exhibit cytotoxicity against cancer cells in HLA-A*0201-, -A*0206-, and -A*2402-restricted ways.

We further attempted to identify the cells responsible for the cytotoxicity of PAP peptide-stimulated PBMCs. Purified CD8⁺ T cells were used in the following experiments. The cytotoxicity of CD8⁺ T cells purified from the peptide-stimulated PBMCs against PC3 and COLO205 was significantly decreased by the addition of anti-HLA class I mAb (W6/32) but not by the addition of either anti-HLA class II (HLA-DR) or anti-HLA-B,C (B1-23, IgG2a) mAbs (Fig. 3). In addition, cytotoxicity was significantly inhibited by the addition of corresponding peptide-pulsed unlabeled C1R-A*2402 or T2 cells but not by the addition of HIV peptide-pulsed unlabeled C1R-A*2402 or T2 cells. These results suggested that the cytotoxicity of peptide-stimulated PBMCs against PAP-expressing cancer cells depended mainly on HLA class I-restricted and PAP peptide-specific CD8⁺ T cells (Fig. 4).

Fig. 3.

Inhibition assay of peptide-stimulated PBMCs with Abs. Peptide-stimulated PBMCs from two HLA-A*0201, one -A*0206, and three -A*2402 patients were tested for their cytotoxicity against COLO205 and PC3 cells expressing each of the HLA-A*0201 and -A*2402 alleles, respectively in the presence of the indicated monoclonal antibodies. Asterisks statistically significant at P < 0.05

Fig. 4.

Inhibition assay of peptide-stimulated PBMCs with peptides. Peptide-stimulated PBMCs from six patients were tested for their cytotoxicity against COLO205 and PC3 cells expressing the HLA-A*0201 and -A*2402 alleles, respectively in the presence of unlabeled T2 and C1R-A*2402 cells, which were preloaded with either the corresponding peptide or the HIV peptide. Asterisks statistically significant at P < 0.05

Discussion

Several HLA-class I-A alleles, such as HLA-A1, are relatively dominant outside of Japan. In this study, since we could not obtain HLA-A1⁺ PBMCs, we were unable to test the ability of HLA-A3 supertype peptides to induce CTLs restricted to those alleles. Also, we could not test the binding activity to HLA-A*2602, -A*2603, or the HLA-A2 subtypes other than -A*0201, -A*0206, and -A*0207 because of the difficulty of obtaining samples. These issues shall be addressed in the near future.

Peptides used as cancer vaccines usually consist of nine or ten amino acids that can bind to particular MHC molecules and activate CTLs reactive to cancer cells or viral infection in a particular MHC-restricted manner. Although peptide–MHC binding was previously thought to be very specific, recent studies have shown that it is neither as narrowly specific nor as unique as originally believed. Sidney and colleagues have pointed out that, among 945 different HLA-A and -B alleles examined to date, over 80% can be assigned to one of the nine subtypes, whereas the remaining alleles might be associated with repertoires that overlap multiple supertypes [8, 24, 25]. The molecular mechanisms of these phenomena are presently unclear, though the stereophonic structure of both the B and F pockets plays a crucial role for binding. Crystallographic studies have revealed that residues 7, 9, 24, 34, 45, 63, 66, 67, 70, and 99 are considered to make up the B pocket, which engages the peptide residue in position 2. The residues that constitute the F pocket, which engages the peptide C-terminal residue, were defined as 74, 77, 80, 81, 84, 95, 97, 114, 116, 123, 133, 143, 146, and 147 [24]. Further studies will clarify these phenomena.

The current results suggest the possibility that one peptide can bind to several different HLA alleles. In fact, this phenomenon has been reported by many other researchers [24]. In this study, we have shown that the PAP_155–163 peptide, a previously reported peptide capable of inducing HLA-A3 supertype-restricted CTLs [10], could bind to multiple HLA molecules including HLA-A*0201, -A*0206, and -A*2402, the dominant HLA types in the Japanese population.

The binding scores of HLA-A*0201 and -A24 of PAP_155–163 (YLPFRNCRP) on BioInformatics and Molecular Analysis Section (BIMAS) are very low compared to those of the HLA-A3 supertypes (Table 1). However, we showed that the PAP_155–163 peptide can bind only to HLA-A*0201-, -A*0206-, and -A*2402-transfected cells. The reason for this discrepancy is unclear, but the binding of this peptide to HLA-A*0201-, -A*0206-, and -A*2402-transfected cells could be expected; the binding motifs of HLA-A2 binding peptides are characterized by the presence of I or L residues at amino acid position 2 and V, F, or M residues at the C terminus [26]. However, merely belonging to the HLA-A2 genealogy does not ensure belonging to the supertype, as exemplified by HLA-A*0207. Guercio et al. [27] have shown that the HLA-A2 supertype binding peptide does not bind -A*0207. Therefore, the binding motif of HLA-A*0207 could be largely different from that of HLA-A*0201 or -A*0206.

HLA-A24 binding peptides are characterized by the presence of Y or F residues at amino acid position 2 and L, F, I, or W residues at their C terminus [28]. Analysis of published epitopes recognized by HLA-A*2301 suggests that its peptide-binding specificity might overlap with that of HLA-A*2402. Thus, an HLA-A24 supermotif incorporating residues commonly recognized by HLA-A24 supertype alleles may be defined as (FWYLVIMT) and (FIYWLM) [8]. These findings suggest the possibility that the PAP_155–163 peptide (YLPFRNCPR) can bind to HLA-A*0201, -A*0206, and -A*2402 molecules and cannot bind to HLA-A*0207.

Prostate acid phosphatase is a prostate-related antigen capable of inducing anti-tumor responses when injected as a cancer vaccine into prostate cancer patients [29]. However, several reports have shown PAP expression in tissues other than those of the prostate and prostate cancer [15, 30–36]. PAP was reported to be immunohistologically positive in neuroendocrine tumors in the pancreas and in small intestine tumors, pancreas endocrine tumors, hindgut-origin tumors, pancreatic islet cell carcinomas, and adenocarcinomas of the urinary bladder [30–32]. Recently, PAP was suggested to be a tumor marker of intravascular large B-cell lymphoma [33]. We have also reported that PAP is expressed in esophageal and lung squamous cell carcinoma cells and PAP-specific CTLs can show cytotoxicity to these tumor cells [34]. In addition, several reports suggest that PAP is expressed in normal tissues, including the normal anal glands of the male and the urethral glands of both sexes [35, 36]. We incidentally found that normal pyloric glands in gastric mucosa express PAP [15].

Small et al. performed a phase 1 clinical trial for patients with hormone-refractory prostate cancer using autologous dendritic cells that were loaded with a recombinant fusion protein consisting of PAP and granulocyte–macrophage colony-stimulating factor. They revealed that the specific immunotherapy targeting PAP is safe and the time to disease progression correlates with the development of the anti-PAP immune response [29, 37]. We also reported that a PAP-derived peptide capable of inducing HLA-A24-restricted CTL activity at positions 213–221 was one of the key peptides inducing clinical responses in prostate cancer patients receiving tailor-made peptide vaccines [5, 6]. These results suggest that PAP is one of the appropriate target molecules of a cancer vaccine.

We previously reported that the employed CTL assay with a 14-day incubation with 5× stimulation could not detect CTL precursors by de novo sensitization to PAP [38]. The sensitivity of this CTL assay was 1 out of 3,000 to 1 out of 5,000 CTL precursors. Thus, the immune response against the PAP-155 peptide is suggested to be relatively specific to PAP because naïve T cells from healthy donors are more difficult to induce when compared to those from prostate cancer patients. Indeed, it was relatively easy to induce a PAP-155-specific CTL in prostate cancer patients, although CTL precursors were not detectable in 4 of the 12 patients tested. This failure might be in part due to immune suppression associated with advanced prostate cancer. On the other hand, more than seven attempts were needed to induce CTL in the cases of healthy donors (data not shown).

In conclusion, HLA-A2, -A24, and -A3 supertypes constitute 98% of the Asian population, 74% of Caucasians, 72% of Spaniards, 76% of Asian Indians, and 59% of Black Africans. Thus, this peptide could be useful as a cancer vaccine for the majority (~60%) of prostate cancer patients as well as for colon and gastric cancer patients.