Enhancement of antitumour immunity by a novel chemotactic antigen DNA vaccine encoding chemokines and multiepitopes of prostate-tumour-associated antigens

Abstract

DNA vaccines provide an attractive technology against cancer because of their safety record in humans and ease of construction, testing and manufacture. In this study, several DNA fragments encoding multiple cytotoxic T lymphocyte (CTL) and T helper cell epitopes were selected from human prostate-specific membrane antigen (hPSM), mouse prostatic acid phosphatase (mPAP), and human prostate-specific antigen (hPSA). These DNA fragments were ligated together to form a novel fusion gene, termed the 3P gene. The secondary lymphoid tissue chemokine (SLC), 3P and human immunoglobulin G Fc genes were inserted into pcDNA3.1 to construct a DNA vaccine, designated pSLC-3P-Fc. After vaccination, the DNA is taken up by cells that produce and secrete the SLC-3P-Fc fusion proteins, termed chemotactic antigen (chemo-antigen). The secreted chemo-antigens, in addition to promoting the co-localization of naive, non-polarized memory T cells and dendritic cells, are efficiently captured and processed by dendritic cells via receptor-mediated endocytosis and then cross-presented to both major histocompatibility complex class I and class II in a cognate manner. The results of this study demonstrate that vaccination with pSLC-3P-Fc by gene gun inoculation induced a strong antitumour response in a mouse tumour model, which significantly inhibited tumour growth and prolonged the survival time of the tumour-bearing mice. In vitro, the secreted SLC-3P-Fc fusion protein can attract lymphocytes from human peripheral blood mononuclear cells (PBMC); when human lymphocytes were stimulated by pSLC-3P-Fc-transfected autologous PBMC, CTLs were induced which could specifically kill hPSM-, hPAP-, or hPSA-expressing tumour cells. These observations provide a new vaccine strategy for cancer therapy through promoting the co-localization of lymphocytes and the concomitant enhancement of antigen-specific CD4⁺ helper and CD8⁺ cytotoxic T-cell responses against tumour.

Introduction

Prostate cancer remains a difficult clinical problem; it is the most diagnosed cancer and the second leading cause of mortality from cancer among American men.¹ Prostate cancer is also a significant and growing health problem in the Chinese population. Although surgery and radiation remain the treatments of choice for the early (localized) stages of prostate cancer, there is no clear effective treatment for patients who develop recurrences or for those who have metastatic disease at the time of diagnosis. T-cell-based immunotherapy has been seriously considered as a promising novel non-invasive treatment option that could be used to prevent metastatic spread or delay recurrences of prostate cancer. However, the development of such vaccines relies on the identification of relevant target antigens as well as approaches to deliver antigens to generate immunity. Prostate tissue and prostate cancer express several potential target antigens, including prostate-specific membrane antigen (PSA), prostate-specific membrane antigen (PSM), and prostatic acid phosphatase (PAP). Over the past years, studies have been initiated to assess both the immunogenicity and the therapeutic potential of vaccination with prostate tumour antigen-derived peptides or gene products. In most studies performed thus far, single tumour antigen-derived peptides or gene products have been administered. Nevertheless, prostate tumour cells can express multiple tumour antigens and the expression of each of these can vary within the tumour. Therefore, a vaccine containing peptides derived from several of these antigens could be more effective than a vaccine directed against a single antigen.² In addition, elicitation of immune responses directed against several antigens can minimize the risk of selecting antigen-loss tumour variants.^3,4

The likelihood of success in cancer immunotherapy is determined by the efficacy of the therapy in inducing an immune response against the tumour antigens. Many tumours escape immunological destruction by down-regulation or loss of immunogenic epitopes. Therefore, it is generally accepted that vaccination with multiple epitopes will be advantageous over single epitope-based vaccines⁴ and these responses can be additionally enhanced by the co-administration of immune-potentiating cytokines. ^5–7 It has been suggested that effective antitumour immunity may be achieved by recruiting professional host antigen-presenting cells (APCs) for tumour antigen presentation to promote specific T-cell activation.⁸ The secondary lymphoid tissue chemokine (SLC) is a CC chemokine expressed by high endothelial venules in lymph nodes and in T-cell zones of the spleen and of lymph nodes that strongly attracts T cells and dendritic cells (DCs).^9,10 SLC recruits both naive lymphocytes and antigen-stimulated DC into the T-cell zones of secondary lymphoid organs, co-localizing these early immune response constituents and culminating in cognate T-cell activation.⁹ It has been reported that intratumoral injection of DC-Ad-SLC was effective in generating systemic antitumour responses.^11,12 Thus, SLC could serve as a potent agent in cancer immunotherapy.

DNA-based vaccination opens the possibility of combining epitopes of different proteins into one vaccine.⁴ DNA vaccination has become an attractive immunization strategy against tumour, because it has the ability to induce both cellular and humoral immune responses. However, a major problem of DNA vaccination is its limited potency to induce immune responses. It is known that DNA applied either intramuscularly or intradermally is primarily taken up by muscle cells or keratinocytes, respectively.^13,14 However, these transfected cells expressing the encoded proteins are unable to initiate primary immune responses. Although the mechanism by which DNA vaccination induces immune responses is still poorly understood, accumulating evidence indicates the critical role of DCs,⁸ the most potent APCs, in inducing the immune responses of DNA vaccines.^13,14 Thus, enhancement of antigen presentation by DC offers an attractive strategy by which to increase the potency of DNA vaccines.

DCs express receptors for the Fc portion of immunoglobulin G (IgG; (FcγRs), which mediate the internalization of antigen–IgG complexes and promote efficient major histocompatibility complex (MHC) class II-restricted antigen presentation. This process is 1000–10 000-fold more efficient than fluid-phase pinocytosis.^15,16 FcγR-mediated endocytosis can also cross-present the internalized antigen to MHC class I.^16,17 In addition, the interaction of Fc with its FcγRs activates DC by up-regulating surface molecules and the cytokines involved in antigen presentation.¹⁶ Thus, FcγRs represent a privileged antigen internalization route for efficient MHC class I- and II-restricted antigen presentation by DC. ^15–17

It has been reported that human PAP (hPAP, GenBank accession no. NM_001099)¹⁸ is 81% homologous to mouse PAP (mPAP, GenBank accession no. NM_019807)^19,20 at the amino acid level. Immunization with a xenogeneic PAP (hPAP into rat) generated cytotoxic T lymphocytes (CTL) and prostate-specific autoimmunity, suggesting that xenoantigens may be capable of overcoming tolerance against the homologous self-antigen.²⁰ In this study, using the bimas computer program (National Institutes of Health) and syfpeithi (http://www.syfpeithi.de/), we searched the entire sequences of the genes for human PSM (hPSM, GenBank accession no. M99487),²¹ mPAP and human PSA (hPSA, GenBank accession no. M26663)²² for identifying regions that contain the anchor motif residues required for binding to MHC class I and class II molecules. DNA fragments which encode multiple CTL and T helper (Th) cell epitopes, were selected from the hPSM, mPAP and hPSA genes and fused to create a novel hPSM-mPAP-hPSA fusion gene (named 3P gene).²

Using 3P as a model antigen, we developed a novel DNA vaccination strategy that relies on co-localization of lymphocytes, encoding multiple epitopes, and the enhancement of DC antigen presentation. Specifically, the 3P gene is linked to SLC (containing a leader sequence) at its 5′ end and to the human IgG Fc gene at its 3′ end. The DNA vaccine was constructed to express SLC-3P antigen fused to a cell-binding domain (human IgG Fc fragment) for receptor-mediated internalization. The current gene gun delivery approach may provide a new methodology for effective and functional delivery of multiple, candidate therapeutic genes for experimental and potential clinical applications. After vaccination via gene gun, cells that produce and secrete SLC-3P-Fc fusion proteins take up the DNA molecules. The secreted fusion proteins strongly attract T cells and DCs and are efficiently captured and processed by DC via receptor-mediated endocytosis and presented to MHC class II and I (cross-priming). The results of this study demonstrate the broad enhancement of antigen-specific CD4⁺ helper and CD8⁺ cytotoxic T-cell responses by this DNA vaccination strategy.

Materials and methods

Cell lines and mice

C57BL/6 melanoma cell line B16F10, mouse T-lymphoma cell line YAC-1, human prostate cancer cell line LNCaP (HLA-A0201⁺, hPSA⁺, hPAP⁺, hPSM⁺),²³ human prostate cancer cell line PC-3M (HLA-A2^–, hPSA⁺, hPAP⁺, hPSM⁺), ^24–26 and human breast adenocarcinoma cell line MCF-7 (HLA-A0201⁺, hPSA^–, hPAP^–, hPSM^–) ^27–29 were kindly provided by L. Chen (Department of Immunology, Mayo Graduate and Medical Schools, Mayo Clinic, Rochester, MD). The T2 cell line was a kind gift from W. Chen, Peking University, Beijing, China. All cells were cultured in RPMI-1640 medium (Gibco-BRL, Gaithersburg, MD) supplemented with 10% fetal calf serum (hereafter, referred to as complete medium).

Male C57BL/6 mice (8–10 weeks old) were purchased from the Experimental Animal Institute of Peking Union Medical College.

Plasmid DNA vaccine constructs and preparation

Polymerase chain reaction (PCR) was performed with the following primers: primer A1: 5′-AGAATTCATGATGAATGATCAACTCATG-3′; primer A2: 5′-AGCACTCATCAAAGTCCTGGCCTTGGAAGGGTCCAC-3′; primer B1: 5′-AGGGCAGTCTCTGAAAGGCAG-3′; primer B2: 5′-AGGACTTTGATGAGTGCTATG-3′ (contains hPSM overlapping sequence); primer C1: 5′-CCTTTCAGAGACTGCCCTGGCGGTGTTCTGGTGCAC-3′ (contains mPAP overlapping sequence); primer C2: 5′-AGATATCGAGCAGCATGAGGTCGTG-3′; primer C3: 5′-TGATATCTCAGAGCAGCATGAGGTCGTG-3′ (contains stop codon); primer D1: 5′-TGGTACCACAGACATGGCTCAGTCACTG-3′; primer D2: 5′-CGAATTCGTCCTGTGCCCCTCCATC-3′

Human PSM, hPAP and hPSA were amplified from LNCaP cells by reverse transcription (RT)-PCR. Murine PAP was amplified by RT-PCR from mouse prostate gland. SLC was amplified by RT-PCR from human lymph nodes. The amplified products were inserted into pUCm-T (Sangon Company, Shanghi, China) to construct pUCm-T-hPSM, pUCm-T-hPAP, pUCm-T-hPSA, pUCm-T-mPAP, pUCm-T-SLC and pUCm-T-SLC (contains the stop codon).

The 3P fusion gene was constructed using a two-step procedure: first, using pUCm-T-hPSM, pUCm-T-mPAP, or pUCm-T-hPSA as template, one strand of hPSM, mPAP, or hPSA was amplified by asymmetric PCR using primers A1 and A2 for hPSM, primers B1 and B2 for mPAP, and primers C1 and C2 or C3 for hPSA. The asymmetric PCR products (hPSM + mPAP or mPAP + hPSA) were combined and PCR was performed to amplify the hPSM-mPAP or mPAP-hPSA fusion gene. Second, using hPSM-mPAP or mPAP-hPSA as template, one strand of hPSM-mPAP or mPAP-hPSA was amplified by asymmetric PCR using primers A1 and B1 for hPSM-mPAP, and primers B2 and C2 or C3 for mPAP-hPSA. The asymmetric PCR products were combined and PCR was performed to amplify hPSM-mPAP-hPSA fusion gene (namely 3P). The 3P fusion gene was cloned into the pUCm-T to construct pUCm-T-3P or pUCm-T-3P (contains the stop codon).

The signal sequence of SLC (abbreviated as sig) was amplified with primers D1 and D2, and the PCR product was cloned into the pUCm-T to generate pUCm-T-sig.

The human IgG1 Fc fragment of pHIgV-ICOS-Fc (a kind gift from L. Chen) was inserted into pcDNA3f, which altered the multiple cloning site of pcDNA3.1 (Invitrogen, Carlsbad, CA), to generate pcDNA3f-Fc.

The pUCm-T-SLC was digested with KpnI and EcoRI and cloned into pFc to produce pSLC-Fc. The EcoRI-EcoRV-digested 3P fragment derived from pUCm-T-3P was ligated into EcoRI-EcoRV-digested pSLC-Fc to generate pSLC-3P-Fc (Fig. 1).

Figure 1.

Schematic representation of expression vectors. The SLC-3P-Fc fusion gene, SLC-3P fusion gene, SLC-Fc fusion gene, 3P-Fc fusion gene, SLC gene, 3P (secretory) gene, Fc cDNA fragment with a signal sequence (secretory) were cloned into the pcDNA3.1 vector under the CMV promoter control. ▪ indicates signal sequences.

The 3P, sig, or SLC derived from pUCm-T-3P, pUCm-T-sig, or pUCm-T-SLC was cloned into pcDNA3f or pcDNA3f-Fc to generate psig-3P (p3P), psig-Fc (pFc), pSLC, pSLC-3P, pSLC-Fc, and psig-3P-Fc (p3P-Fc) (Fig. 1). The sequences of 3P were as follows: 5′-hPSM-mPAP-hPSA-3′; PSM(1–186), mPAP(187–342), hPSA(343–564).

All recombinant constructs were confirmed by sequence analysis. As a control, pure pcDNA3.1 plasmid was used as the empty vector (pC).

The expression of p3P, pFc, pSLC, pSLC-3P, pSLC-Fc, p3P-Fc, or pSLC-3P-Fc was confirmed in the transfected cells by RT-PCR and Western blot analysis.

Plasmids for DNA vaccination were purified from a large-scale culture by alkaline lysis and polyethylene glycol precipitation.³⁰

Target cell preparation

The 3P, hPSA, hPAP, or hPSM derived from pUCm-T-3P, pUCm-T-hPSA, pUCm-T-hPAP, or pUCm-T-hPSM was inserted into pcDNA3.1 to construct the expression vectors, pcDNA3.1-3P, pcDNA3.1-hPSA (pPSA), pcDNA3.1-hPAP (pPAP), and pcDNA3.1-hPSM (pPSM), respectively.

The constructs were transfected into B16F10 or MCF-7 cell lines by lipofection using Lipofectamine Plus reagent (Invitrogen Life Technologies) according to the manufacturer's instructions. B16F10 was selected with 600 μg/ml G418 and MCF-7 with 400 μg/ml G418. Clones were isolated using limiting dilutions. Positive clones were screened by RT-PCR.

The resulting target cells were B16F10-pC, B16F10-3P, B16F10-PSA, 16F10-PAP, B16F10-PSM, MCF-7-pC, MCF-7-PSA, MCF-7-PAP and MCF-7-PSM.

The cell proliferation was determined by 3-(4,5-dimethylthiazol-2)-2,5-diphenyl-tetrazolium bromide (MTT) assay before use in cytotoxicity assays.

Western blot analysis

B16F10 cells were transfected with pC, pSLC, pFc, pSLC-3P, pSLC-Fc, p3P-Fc or pSLC-3P-Fc using Lipofectamine Plus reagent (Invitrogen) according to the manufacturer's instructions. The culture medium was harvested, and cells were lysed in phosphate-buffered saline (PBS) buffer containing 2 μg/ml aprotinin, 100 μg/ml phenylmethylsulphonyl fluoride, 2 μg/ml leupeptin and 1% nonidet P-40. The proteins were separated by 12·5% sodium dodecyl sulphate–polyacrylamide gel electrophoresis, and then transferred onto a nitrocellulose membrane (Bio-Rad Laboratories, Hercules, CA). The membranes were blocked by incubation in 5% non-fat dried milk, then washed and incubated with a goat anti-human IgG Fc antibody (Santa Cruz Biotechnology, Santa Cruz, CA) or goat anti-human SLC antibody (Santa Cruz Biotechnology), followed by incubation with a horseradish peroxidase labelled rabbit anti-goat IgG (Sigma, St Louis, MO). The membrane was visualized using the enhanced chemiluminescence Western blotting system (Amersham, Little Chalfont, Buckinghamshire, UK).

Chemotaxis assay

Peripheral blood mononuclear cells (PBMC) were isolated from heparinized blood by Ficoll–Hypaque density gradient centrifugation. The chemotactic responses of human PBMC or mouse splenocytes to SLC were examined using a chemotaxis microchamber technique (48-well Boyden microchamber; NeuroProbe, Inc., Gaithersburg, MD). The culture supernatants from B16F10 or pSLC-3P-Fc-, pSLC-, p3P-, or pC-transfected B16F10 were added to the lower wells of the chemotaxis chamber. A total of 10⁵ cells/well in 50 μl RPMI-1640 medium was applied to the upper wells of the chamber, with a standard 5- μm pore polyvinylpyrrolidone-free polycarbonate filter (NeuroProbe Inc.) separating the lower wells. After incubation for 4 hr at 37°, the cells that had not migrated into the membranes were wiped from the upper surfaces of the membranes, which were then fixed in 70% methanol and stained using Giemsa. The lymphocytes or monocytes that were associated with the membranes were enumerated by direct counting in a blinded fashion of at least nine 40× objective fields per well. The results were expressed as the fold increase in cells migrating in response to chemoattractant versus the medium control (chemotaxis index). All assays were performed in triplicate.

DNA vaccination

The Helios gene gun system (Bio-Rad) was used for intradermal gene delivery. Bullets containing 1·25 μg of DNA/shot were generated according to the manufacturer's protocols. Briefly, 75 μg DNA was precipitated on 30 mg of 1·6 μm gold particles in the presence of 100 μl of 0·05 m spermidine (Sigma) using 100 μl of 1 m CaCl₂/preparation. The gold was washed three times in fresh absolute ethanol and was resuspended in 2·4 ml of 0·1 mg/ml polyvinylpyrrolidone in 100% ethanol. The gold was then loaded into the tubing using the tubing preparation station (Bio-Rad), and the gold-loaded tubing was cut into 1·27-cm (i.e. 0·5-inch) pieces to load into the cartridges. The bullet-containing cartridges were loaded into the gene gun and delivered into the mouse dermis at a helium pressure of 380 p.s.i.

Histology

Mice were given 1·25 μg pSLC-3P-Fc or pC via a gene gun, or were left unimmunized (negative control). After 72 hr, mouse dermis tissues from the injection sites of treated mice were surgically excised, fixed for 12–24 hr in 10% neutral buffered formalin, embedded in paraffin wax, sectioned at 5- μm thickness, and stained with haematoxylin and eosin (H & E) for histological examination.

Tumour protection and therapy assay

For in vivo tumour prevention experiments, male C57BL/6 mice (n = 5) were immunized with 1·25 μg pSLC-3P-Fc, pSLC-3P, pSLC-Fc, p3P-Fc, pSLC, p3P, pFc, or pC via a gene gun, and then boosted 7 days later. Seven days after the booster, mice were inoculated in the left flank with 5 × 10⁴ B16F10-3P cells/mouse.

For in vivo therapeutic experiments, male C57BL/6 mice (n = 8) were injected with 5 × 10⁴ B16F10 3P cells/mouse and then immunized with DNA vaccine by gene gun on days 4, 9 and 14 after the inoculation of B16F10-3P cells. Tumours were monitored every 2 days, and tumour dimensions were determined by measuring with calipers (length × width × height), and the values were inserted into the formula: tumour volume (mm³) = 0·52 (length × width × height).³¹

Cytotoxicity assay

Male C57BL/6 mice (n = 3) were challenged subcutaneously with 5 × 10⁴ B16F10-3P cells/mouse in the right flank. Four days later, mice were given 1·25 μg pSLC-3P-Fc, pSLC-3P, pSLC-Fc, p3P-Fc, or pC via a gene gun, or were left unimmunized (negative control). These mice were boosted twice with the same regimen as the first vaccination at days 9 and 14. Ten days after the last booster, splenocytes from immunized mice or control mice were used as effector cells in a non-radioactive cytolytic analysis.³² B16F10, B16F10-pC, B16F10-3P, B16F10-PSM, B16F10-PAP, B16F10-PSA and YAC-1 cells were used as target cells. Effector cells were added to target cells at a ratio of 10 : 1, 20 : 1 and 40 : 1 (tested in triplicate). Four hours after inoculation, supernatants were pooled and measured for the release of lactate dehydrogenase (LDH) using a CytoTox 96^® Non-Radioactive Cytotoxicity Assay kit (Promega, Madison, WI) according to the manufacturer's instructions. Specific lysis was calculated according to the formula: % cytotoxicity = [(E–Se–St)/(Mt–St)] × 100, where E stands for the experimental LDH release in effector plus target cell cocultures; Se is the spontaneous release by effector cells alone; St is the spontaneous release by target cells alone; and Mt is the maximal release by target cells.

Depletion of T-cell subsets in vivo

T-cell subsets were depleted as described elsewhere.^2,31,33 Briefly, mice (n = 5) were vaccinated three times at 5-day intervals with pSLC-3P-Fc, or pC via a gene gun, or unimmunized and challenged with 5 × 10⁴ B16F10-3P cells/mouse on day 5 after the third immunization. Mice were injected intraperitoneally with 500 μg of either the anti-CD4 (clone GK 1.5, rat IgG), anti-CD8 (clone 2.43, rat IgG) or isotype controls (normal rat IgG) 1 day before or 8 days after the first immunization and then twice per week for 3 weeks. Depletion was terminated on day 21 after tumour cell injection.

Cell lines, GK 1.5 and 2.43, were kindly provided by L. Chen. Tumour volume was monitored every 2 days. Mice were killed on day 21 after tumour cell injection. Splenocytes from immunized mice were used in cytotoxicity assay as described above. The depletion of CD4⁺ and CD8⁺ T cells was consistently greater than 95% as determined by flow cytometry (Becton Dickinson, San Jose, CA).

Induction of HLA-A0201-restricted CTL

PBMC were isolated from the heparinized blood of HLA-A0201⁺ healthy donors by Ficoll–Hypaque density gradient centrifugation and 1 × 10⁶ cells/ml in upright 25-ml flasks were stimulated with lipopolysaccharide (LPS) (10 μg/ml) for 24 hr in complete medium. LPS-stimulated PBMC were washed twice with serum-free RPMI-1640 to eliminate LPS. Both pSLC-3P-Fc and pC were transfected into PBMC (1 × 10⁷), respectively. Transfections were performed using Lipofectamine Plus reagent (Invitrogen) according to the manufacturer's specifications. PBMC were stimulated with transfected PBMC in the presence of human recombinant interleukin-2 (hrIL-2) (25 U/ml). On day 7, PBMC were restimulated under the same conditions as the initial stimulation. Cultures were fed by adding fresh complete medium and 25 U/ml of hrIL-2 at days 3, 6, 9 and 12. The cells were used on day 14 in a cytotoxicity assay as described above. Briefly, stimulated HLA-A0201⁺ PBMC were used as effector cells. T2, PC-3M, MCF-7, MCF-7-pC, MCF-7-PSM, MCF-7-PAP, MCF-7-PSA, and LNCaP cells were used as target cells. Target cells were cocultured with effector cells at a ratio of 1 : 15, 1 : 30 and 1 : 60 (tested in triplicate).

Statistical analysis

For comparison of individual time-points, anova was used for the comparisons among three or more groups. Student's t-test was used to compare means between the two groups. Survival curves were compared by the log-rank test. Differences were considered significant when P < 0·05. Statistical analysis was performed using commercially available software (spss 11·0).

Results

Expression of DNA constructs and characterization of transfected target cells

The Fc portion of human IgG1 can efficiently bind to human DC as well as to murine DC.³⁴

Total RNA extracted from pSLC-3P-Fc-, pSLC-3P-, pSLC-Fc-, p3P-Fc-, pSLC, p3P, pFc, or pC-transfected cells and target cells was used to verify the expression of SLC-3P-Fc, SLC-3P, SLC-Fc, 3P-Fc, SLC, 3P, Fc, PSM, PAP, or PSA by RT-PCR (Fig. 2a) and the primers used for amplification were as follows: SLC-3P-Fc (primer F, 5′-TGGTACCACAGACATGGCTCAGTCACTG-3′ and primer R, 5′-GGGCCCTCATTTACCCGGAGAC-3′); SLC-3P (primer F and primer S, 5′-TGATATCTCAGAGCAGCATGAGGTCGTG-3′); SLC-Fc (primer F and R); 3P-Fc (primer F and R); SLC (primer F and 5′-GAATTCTATGGCCCTTTAGG-3′); sig-3P (primer F and primer S); Fc (primer F and R); 3P (5′-AGAATTCATGATGAATGATCAACTCATG-3′ and primer S); PSM (5′-ACTCAGGATGAAGACATACAGTGTATC-3′ and 5′-TGATATCTTAGGCTACTTCACTCAAAG-3′); PAP (5′-ATGAGAGCTGCACCCCTC-3′ and 5′-TGATATCTAATCTGTACTGTCCTC-3′); PSA (5′-TCTCGAGGGCGGTGTTCTGGTGCA-3′ and 5′-AGATATCATGTCCAGCGTCCAGCAC-3′); glyceraldehyde-3-phosphate dehydrogenase (5′-ACCACAGTCCATGCCATCAC-3′ and 5′-TCCACCACCCTGTTGCTGTA-3′). Secreted proteins and cellular proteins were analysed by Western blotting (Fig. 2b). These results demonstrate that the constructs were expressed in the eukaryotic cells.

Figure 2.

(a)RT-PCR detection of gene expression. DNA marker (M), B16F10-pC (lane 1), pSLC-3P-Fc-transfected B16F10 (lane 2), pSLC-Fc-transfected B16F10 (lane 3), p3P-Fc-transfected B16F10 (lane 4), pSLC-3P-transfected B16F10 (lane 5), pSLC-transfected B16F10 (lane 6), p3P-transfected B16F10 (lane 7), pFc-transfected B16F10 (lane 8). (b) Western blotting analyses of DNA construct expression in B16F10 transfectants. Secreted proteins (lane 1), cellular proteins (lane 2), control vector (lane 3).

Equivalent numbers (2 × 10³/well) of various gene-transfected B16F10 or MCF-7 cells were seeded in 96-well plates. The cell proliferation was assayed by MTT assay and did not significantly differ between the two cell lines (P > 0·05). Each assay was performed in triplicate. Proliferative activity was measured in terms of optical density at 570 nm on each given day. These results indicated that plasmid DNA had no effect on transfected target cell proliferation.²

Chemotactic activity of pSLC-3P-Fc transfection

Although the sequences of SLC cDNA and SLC-3P-Fc in pSLC or pSLC-3P-Fc were identical, whether SLC or SLC-3P-Fc expressed by pSLC or pSLC-3P-Fc was functional was unknown. To this end, B16F10 cells were transfected with pSLC-3P-Fc, pSLC, p3P, or pC and the culture supernatants were harvested 48 hr later. Chemotactic activity for lymphocytes was assessed by quantifying lymphocyte migration across 5-μm pore polycarbonate membranes in a Boyden cell culture chamber assay to measure the response to chemotactic stimuli placed into the lower chamber.

As shown in Fig. 3(a), supernatants from pSLC- or pSLC-3P-Fc-transfected cells attracted lymphocytes above control levels. No difference in chemotactic response of lymphocytes from mouse spleen or PBMC was observed. Our results demonstrated that human SLC also show chemotactic efficacy in mouse model.

Figure 3.

(a) Chemotactic attraction of lymphocytes from human PBMC or mouse spleen by pSLC-3P-Fc-encoded SLC-3P-Fc. Chemotactic activity of culture supernatants of B16F10 cells transfected with pSLC-3P-Fc, pSLC, p3P, or pC, measured as the fold increase in lymphocytes migrating in response to chemoattractant versus medium control (chemotaxis index). * indicates significant differences at P < 0·01 according to one-way anova. (c–e) Transfection with pSLC-3P-Fc chemoattracts lymphocytes to the immunization site. H&E staining of dermis tissue sections from C57BL/6 mice, untreated mice (b), mice treated with pC (c) or with pSLC-3P-Fc (d). Images are representative of multiple microscopic fields observed in at least three mice per group.

As shown in Fig. 3(b–d) H&E staining of dermis tissue sections from untreated mice or pC-treated mice showed infiltration of few lymphocytes into the immunization region. Tissues from mice that received pSLC-3P-Fc showed a prominent infiltration of lymphocytes, which may be chemoattracted by SLC-3P-Fc.

Induction of the antitumour immunity

To investigate the protective antitumour immunity, we immunized mice with plasmid (pSLC-3P-Fc, pSLC-3P, pSLC-Fc, p3P-Fc, pSLC, p3P, pFc, or pC) by gene gun, and then challenged mice with tumour cells. As shown in Fig. 4(a,b), tumours grew progressively in pSLC-3P-, pSLC-Fc-, p3P-Fc-, pSLC-, p3P-, pFc-, or pC-immunized mice, but there was apparent protection from tumour growth in pSLC-3P-Fc-immunized mice. The results are expressed as mean tumour volume ± SEM. The survival of the mice treated with pSLC-3P-Fc was also significantly greater than that of pSLC-3P-, pSLC-Fc-, p3P-Fc-, pSLC-, p3P-, pFc-, or pC-immunized mice (P < 0·01, by log-rank test).

Figure 4.

Induction of the protective antitumour immunity. (a) Tumour volume (mm³) in mice treated with pSLC-3P-Fc and controls is shown. (b) Percentage survival of mice treated. The survival rate of the pSLC-3P-Fc-treated mice was 80% at day 60 for B16F10-3P melanoma cells. (c,d) Induction of the therapeutic antitumour immunity. (c) Tumour volume (mm³) in mice treated with pSLC-3P-Fc and controls is shown. (d) Percentage survival of mice treated. The survival rate of the pSLC-3P-Fc-treated mice was 63% at day 60 for B16F10-3P melanoma cells.

The therapeutic efficacy of the DNA vaccine encoding SLC-3P-Fc was next tested in the established tumours. The mice were treated starting at day 4 after the injection of B16F10-3P melanoma cells. As shown in Fig. 4(c), mice vaccinated with pSLC-3P-Fc demonstrated the lowest average tumour volume on day 26 compared with mice vaccinated with pSLC-3P, pSLC-Fc, p3P-Fc, pSLC, p3P, pFc, pC, or p3P plus pSLC plus pFc (one-way anova, P < 0·001, Data are expressed as mean tumour volume ± SEM). These data clearly show that therapeutic vaccination did significantly reduce tumour growth as compared with the controls. Furthermore, the survival of the mice treated with pSLC-3P-Fc was also significantly longer than that of pSLC-3P-, pSLC-Fc-, p3P-Fc-, pSLC-, p3P-, pFc-, pC-, or p3P plus pSLC plus pFc-immunized mice (P < 0·001, by log-rank test; Fig. 4d). In summary, these results showed that vaccination with pSLC-3P-Fc could induce a strong antitumour response in a mouse tumour model, which in turn may contribute to greatly reduce tumour growth and significantly prolong the longevity of the tumour-bearing mice. These results also indicated that fusion of SLC, 3P and Fc was required for enhancement of antitumour activity, because pSLC + p3P + pFc failed to inhibit tumour progression and increase the survival time of mice.

Induction of CTLs

To determine whether the antitumour response generated by vaccination of pSLC-3P-Fc vaccine is tumour-specific, splenocytes isolated from immunized mice were used in a cytotoxicity assay. Results from this assay indicated a significant increase in tumour-specific lysis of B16F10-3P, B16F10-PSM, B16F10-PAP, or B16F10-PSA in splenocytes from pSLC-3P-Fc-treated mice compared to pSLC-3P-, pSLC-Fc-, p3P-Fc-, or pC-treated counterparts (P < 0·01; Fig. 5). In addition, this cytolytic activity appeared to be specific for B16F10-3P, B16F10-PSM, B16F10-PAP, or B16F10-PSA, not B16F10, or B16F10-pC (P < 0·01). These results suggest that CTL responses from pSLC-3P-Fc-immunized mice are specific for PSM, PAP and PSA antigen. In addition, there is no increase in natural killer (NK) cell activity against NK-sensitive YAC-1 target cells by the sensitized spleen cells.

Figure 5.

Representative experiment of CTL-mediated cytotoxicity in vitro. Splenocytes from C57BL/6 mice, unimmunized mice (a) and mice immunized with pC (b), pSLC-Fc (c), pSLC-3P (d), p3P-Fc (e), or pSLC-3P-Fc (f) were tested for cytolytic activity against YAC-1, B16F10, B16F10-pC, B16F10-hPSA, B16F10-hPSM, B16F10-hPAP, and B16F10-3P cells. Data are expressed as mean ± SE.

Role of T-cell subsets in pSLC-3P-Fc-induced antitumour activity

To determine the subset of lymphocytes that are important for the rejection of B16F10-3P, we depleted CD8⁺ or CD4⁺ T-cell subsets in vivo before tumour challenge. As observed previously,² if depletions were started after immunization with pSLC-3P-Fc (as shown in Fig. 6), pSLC-3P-Fc-induced cytolytic activity against B16F10-3P cells could be blocked by anti-CD8, but not by anti-CD4, control IgG, or PBS in a cytotoxicity assay. In addition, pSLC-3P-Fc-immunized mice treated with anti-CD4 showed the antitumour activity against B16F10-3P, whereas the treatment with anti-CD8 partially abrogated the protective effects of pSLC-3P-Fc. Furthermore, if depletions were started before immunization with pSLC-3P-Fc (as shown in Fig. 7), in vivo depletion of CD4⁺ T lymphocytes could completely abrogate the antitumour activity, whereas the depletion of CD8⁺ lymphocytes showed partial abrogation of the antitumour activity in vivo. In addition, the treatment with rat IgG, or PBS showed no effect. Mice depleted of CD4⁺ T lymphocytes or CD8⁺ T lymphocytes did not develop detectable CTL activity. These data suggest that CD8⁺ T cells are involved in a role of direct tumour killing, whereas CD4⁺ T lymphocytes are required for the induction of the CD8⁺ CTL response to the immunization with pSLC-3P-Fc vaccine.

Figure 6.

Abrogation of CTL-mediated cytotoxicity and of antitumour activity by in vivo depletion of the T-cell subsets after immunization with pSLC-3P-Fc vaccine. pSLC-3P-Fc-induced cytolytic activity against B16F10-3P cells can be blocked by anti-CD8 in a cytotoxicity assay (a). Whereas the depletion of CD8⁺ T lymphocytes showed the partial abrogation of antitumour activity in vivo (b). The results are expressed as means and error bars represent ± SEM. *Indicates significant difference in tumour volume (P < 0·05) between CD8-depleted mice and the other groups. Data represent day 21 after tumour cell injection.

Figure 7.

Abrogation of antitumour activity and of CTL-mediated cytotoxicity by in vivo depletion of the T-cell subsets before immunization with pSLC-3P-Fc vaccine. Depletion of CD4⁺ T lymphocytes showed complete abrogation of the antitumour activity with the immunization of pSLC-3P-Fc, whereas the depletion of CD8⁺ T lymphocytes showed the partial abrogation (a). The results are expressed as means and error bars represent ± SEM. * indicates significant difference in tumour volume (P < 0·05) between CD8-depleted and other groups. Data represent day 21 after tumour cell injection. Splenocytes from pSLC-3P-Fc-immunized mice were tested for CTL-mediated cytotoxicity against B16F10-3P cells at different effector to target (E : T) ratios by a cytotoxicity assay (b). In vivo depletion of CD4⁺ or CD8⁺ T cells could abrogate the induction of CTL-mediated cytotoxicity with the vaccination of pSLC-3P-Fc.

Antitumour response of HLA-A0201-restricted CTLs

The pSLC-3P-Fc-transfected HLA-A0201⁺ PBMC were used to stimulate autologous lymphocytes in vitro. The effector function of pSLC-3P-Fc-induced CTL was analysed in cytotoxicity assays. The results shown in Fig. 8 demonstrate that the CTLs could recognize and lyse LNCaP, MCF-7-PSM, MCF-7-PAP, and MCF-7-PSA. Whereas control target, MCF-7, MCF-7-pC, T2 (HLA-A0201⁺), or PC-3M were not killed by the CTLs. On the basis of these experiments, it appears that tumour cell lysis was PSM-, PAP-, or PSA-specific and HLA-A0201-restricted.

Figure 8.

Representative experiment of CTL-mediated cytotoxicity in vitro. The CTLs induced by pSLC-3P-Fc-transfected PBMC (a) or pC-transfected PBMC (b) were tested for cytolytic activity against T2, PC-3M, MCF-7, MCF-7-pC, MCF-7-PSM, MCF-7-PAP, MCF-7-PSA, and LNCaP.

Discussion

Generation of an antitumour immune response is a complex process, dependent on coordinate interaction of different subsets of effector cells. Host APCs are critical for the cross-presentation of tumour antigens.³⁵ However, tumours have the capacity to limit the maturation of APCs, their function, and their infiltration of the tumour site.^36–38 Thus, molecules that attract host APCs and T cells and generate both antigen-specific CD4⁺ Th and CD8⁺ CTL responses in a cognate manner may provide optimal immunization against tumours. Here we describe a novel therapy strategy that can efficiently recruit T cells and DCs and cross-present tumour antigen to both MHC class II and class I by DCs in a cognate manner, leading to the activation of both antigen-specific Th and CTL responses.^{9–12,15–17} Thus, this unifying antigen presentation strategy that can induce broad and potent antitumour immunity could be used to improve the efficacy of tumour vaccines and immunotherapies.

According to the strategy described above, our results showed that vaccination with pSLC-3P-Fc could induce strong antitumour activity in a mouse tumour model. We compared the antitumour efficacy of pSLC-3P-Fc with that of p3P-Fc. This data demonstrated that pSLC-3P-Fc might induce stronger antitumour immune responses than p3P-Fc. Consistent with the in vivo experiments, splenocytes from both pSLC-3P-Fc- and p3P-Fc-immunized mice specifically recognized and lysed 3P, PSM, PAP, or PSA, but not B16F10 and B16F10-pC (see Fig. 5) in vitro. The level of CTL activity induced by pSLC-3P-Fc was higher than that of p3P-Fc, the difference was significant (P < 0·01). Furthermore, the effect of the vaccine was severely impaired in mice depleted in vivo of CD8⁺ T cells throughout the effector phase. Depletion of CD4⁺ cells was without effect. Thus, the antitumour activity of pSLC-3P-Fc vaccine may result from the induction of CTL-mediated killing of tumour cells. In the protective and therapeutic experiments, our data demonstrated that vaccination with pSLC-3P-Fc shows considerable retardation in tumour growth and significantly prolonged the survival of the tumour-bearing mice. The antitumour activity could be abrogated by the depletion of CD4⁺ T lymphocytes in the priming phase. Since we found no increase in the NK activity of the sensitized spleen cells, we may rule out the possibility that the antitumour activity with pSLC-3P-Fc vaccine may result from the NK cells against tumour growth in host mice.

Antitumour immunity depends on CD8⁺ T lymphocytes in some mouse models, whereas CD4⁺ T lymphocytes often have little, if any, function.^39,40 Some molecular targets of tumour-specific CD8⁺ T lymphocytes have been identified in human and mouse systems.^39,40 CD8⁺ T lymphocytes have been the focus of recent efforts in the development of a therapeutic antitumour vaccine. However, in this study, we found that, if depletions were started before immunization, the lack of CTL activity as a result of the depletion of CD4⁺ lymphocytes was found to be associated with the complete abrogation of the antitumour activity, whereas the CTL activity as a result of the depletion of CD8⁺ lymphocytes was associated with partial abrogation of the antitumour activity. These results suggest that CD4⁺ T lymphocytes are responsible for the antitumour activity by the vaccination of pSLC-3P-Fc vaccine. It has been reported that CD4⁺ T lymphocytes can steer and amplify the immune response through the secretion of cytokines and the expression of surface molecules such as costimulatory molecules.^41,42 It has been reported that antitumour immunity could be induced by DNA immunization against human gp75/tyrosinase-related protein-1 or tyrosinase-related protein-2 (the slaty locus protein) and has depended on CD4⁺ T lymphocytes in melanoma models.^43,44 These suggestions were further supported by the important roles of CD4⁺ T lymphocytes in the antitumour immunity.^45,46

In the present study, we analysed whether pSLC-3P-Fc-transfected PBMC are capable of stimulating CTL that specifically recognized and lysed PSM-, PAP-, or PSA-expressing tumour targets in cytotoxicity assays. We analysed the effector function of pSLC-3P-Fc-induced HLA-A0201-restricted CTL because the population frequency of HLA-A0201 in several ethnic groups is nearly 50%, allowing future treatment of a significant proportion of the patient population.⁴⁷ Our data demonstrate that pSLC-3P-Fc-transfected HLA-A0201⁺ PBMC are remarkably effective in stimulating potent HLA-A0201-restricted CTL responses against a broad repertoire of antigens, which apparently included CTL subsets specific for PSM, PAP, and PSA. In addition, our findings suggest that the immune response against the hPAP-positive tumour cells may be provoked in a cross-reaction with the plasmid DNA encoding xenogeneic mPAP. The results were similar to the findings with p3P-Fc.²

The novel approach we have developed has several unique and advantageous features.

1. The capacity of SLC to facilitate the co-localization of both DC and T cells may reverse tumour-mediated immune suppression and orchestrate effective cell-mediated immune responses.

2. This approach can efficiently allow DCs to cross-present antigens as exogenous to both class II via the endosomal class II pathway and class I via the cross-priming pathway in a cognate manner, leading to the generation of both antigen-specific Th and CTL responses.

3. The receptor-mediated antigen internalization activates DCs, which is important because an optimal DC antigen presentation requires the antigen processing and a maturation signal to DCs.⁴⁸

4. SLC acts in synergy with IgG1 Fc in SLC-3P-Fc fusion protein (see Figs. 4 and 5).

5. Secreted tumour antigens can be efficiently captured by DCs in both autocrine and paracrine modes to further enhance Th and CTLs responses;

Taken together, the findings in the present study may provide a new vaccine strategy for the treatment of prostate cancer through the recruitment of T cells and DCs to the immunization site and enhancement of DC antigen presentation by the immunization with the plasmid DNA encoding SLC-3P-Fc as vaccine, and may be of importance to the further optimization and refinements of the SLC-Ag-Fc fusion gene strategy with the ultimate goal of developing therapeutic vaccines for prostate cancer patients.

1	atgatgaatg atcaactcat gtttctggaa agagcattta ttgatccatt agggttacca gacaggcctt tttataggca tgtcatct
89	atgctccaag cagccacaac aagtatgcag gggagtcatt cccaggaatt tatgatgctc tgtttgatat tgaaagcaaa gtgg
173	acccttccaa ggccaggact ttgatgagtg ctatgacaaa ccttgcagcc ctgtttcctc cagaggggat cagcatctgg aatc
257	ctagactgct ctggcagccc atcccagtgc acaccgtgtc tctctctgag gatcggttgc tgtacctgcc tttcagagac tgccc
342	tggcggtgtt ctggtgcacc cccagtgggt cctcacagct gcccactgca tcaggaacaa aagcgtgatc ttgctgggtc ggc
425	acagcctgtt tcatcctgaa gacacaggcc aggtatttca ggtcagccac agcttcccac acccgctcta cgatatgagc ctcc
509	tgaagaatcg attcctcagg ccaggtgatg actccagcca cgacctcatg ctgctc