The molecular hallmark of antigens preferentially expressed by tumor cells has attracted numerous interests in the development of tumor antigen based vaccinations. The ability of heat shock proteins (HSPs) to bind cellular peptides makes them the attractive candidates for cancer vaccines.1, 2, 3, 4, 5 HSPs obtained from tumors or virus-infected cells have been shown to induce CTL responses in vitro and in vivo against a variety of antigens expressed in the cells from which HSPs were purified. The immunogenicity of HSP preparations has been attributed to peptides bound to HSPs.1 HSPs are able to chaperon antigenic peptides into antigen-presenting cells (APCs) by binding to specific receptor on APCs.6, 7, 8, 9 The binding of HSP-peptide complex to APCs triggers a cascade of events, including re-presentation of the chaperoned peptides by the major histocompatibility complex, secretion of proinflammatory cytokines and maturation of DCs.10, 11, 12 These properties make HSPs-based vaccination a powerful therapeutic approach to produce tumor antigen specific T cells.
Besides HSPs-based vaccination, many current attempts at harnessing the immune system are capable of eliciting strong T-cell responses against tumor antigens, as high frequencies of tumor antigen specific T cells are often observed in peripheral blood, tumors and draining lymph nodes. However, such responses do not always coincide with tumor rejection.13, 14, 15 In addition, high frequencies of T cells targeting at tumor-associated antigens have often been observed in individuals with malignant tumor, some of these patients nonetheless had progressively growing tumor.16, 17 Thus, it is clear that the residual tumors develop active mechanisms of immune resistance to evade immune attack, although the underlying mechanisms are not wholly clear.
Emerging evidences suggest that B7 homolog 1 (B7-H1) is important in the mechanisms of resistance against tumor-associated antigen-specific immunity.18, 19, 20, 21, 22 B7-H1 (also known as PD-L1) is a cell surface glycoprotein belonging to the B7 family of costimulatory molecules with a profound effect on the regulation of T-cell response. Although B7-H1 mRNA has been found in various organs, including the heart, lung, thymus, spleen, kidney, placenta, skeletal muscle and liver, B7-H1 protein was not detectable in normal tissue, but detectable in many tumors, including carcinoma of breast, lung, ovary, head and colon, as well as glioma. Furthermore, B7-H1 could be further upregulated by some factors in tumor microenvironment, such as IFN-γ, IL-4 and VEGF, and this expression has important functional significance.19, 23 Tumor-associated B7-H1 is thought to contribute to immune escape of cancer by interacting with PD-1 receptor, which is expressed on the activated lymphocytes. Blockade of B7-H1 by specific monoclonal antibodies can enhance tumor-specific CTL response and cause tumor rejection.21, 22, 23, 24
In the present study, we explored the possibility of the involvement of B7-H1 in tumor evasion in the process of immunotherapy with HSP70 vaccine. The significant upregulation of B7-H1 was found in tumor environment, which was responsible for the immune escape. The pulmonary metastasis of melanoma was significantly inhibited by using recombinant soluble PD-1 to blockade B7-H1, following HSP70-peptide complexes vaccination. The finding that B7-H1 blockade synergizes with HSP70-peptide complexes vaccination may provide promising strategy for specific tumor immunotherapy.
Material and methods
Mice and cell lines
Female C57BL/6 mice (6–8-week-old) were purchased from Center of Experimental Animals of Chinese Academy of Medical Science (Beijing, China) and housed in specific pathogen-free conditions. All studies involving mice were approved by the institute's Animal Care and Use Committee. Chinese hamster ovary cell line (CHO), murine melanoma B16F1 cell line and EL-4 thymic lymphoma cell line were purchased from China Center for Type Culture Collection (Wuhan, China). Both CHO cells and tumor cells were cultured in complete medium DMEM supplemented with 10% FCS, 10 mM HEPES, 2 mM L-glutamine, 100 μg/ml penicillin and 100 U/ml streptomycin. To obtain extracellular domain of murine PD-1 (sPD-1) expressing cells, CHO cells were transfected with plasmid pPD-1A using Dosper liposomal transfection reagent (Boehringer Mannheim, Mannheim, Germany) following the manufacturer's protocol. The stably transfected cells were generated by G418 resistance and the expressed sPD-1 in the supernatants was confirmed by Western blot.
The eukaryotic expression plasmid, pPD-1A, of the sPD-1 was constructed previously.25 The expression plasmid pMSH70 of human HSP70 was kindly provided by Dr. Richard Morimoto and Ji-zhong Cheng (The University of Texas Medical Branch). Plasmid DNA was prepared by selective compaction with spermine (Sigma) and endotoxin levels were less than 3 pg/μg of plasmid DNA as determined by Limulus amebocyte lysate assay (BioWhittaker, Walkersville, MD). Spectrophotometric analysis of the plasmid DNA preparation revealed 260/280 nm ratios ≥ 1.80. Purity and conformation of the prepared DNA were confirmed by agarose gel electrophoresis.
Preparation of HSP70-melanoma peptide complex
B16F1 melanoma peptides were prepared as described previously.25, 26 Briefly, melanoma cells at a concentration of 5 × 107/ml subjected to 2 cycles of freeze–thaw, and mixed with a 2-fold volume of distilled water and centrifugated to remove cell debris. The supernatant was incubated in a boiling water bath for 10 min, then in an ice bath for 30 min and centrifugated to remove the denatured high MW proteins. The pH value of the supernatant was adjusted to pH 3.0 with HCl, followed by another centrifugation to remove residual denatured high MW proteins and RNA, and the pH value of the supernatant was adjusted to 7.0 with NaOH solution. Recombinant HSP70 was prepared from engineered bacteria carrying pMSH70.27 The purity of HSP70 proteins was ≥95% as confirmed by silver-stained SDS-PAGE analysis, and the endotoxin was ≤3 pg/μg HSP70 as determined by Limulus amebocyte lysate assay. HSP70 and tumor peptides mixture were mixed to bind each other as described.28 Briefly, tumor peptides and HSP70, at the concentrations of 75 and 250 μg/ml respectively, were mixed and incubated at 37°C for 2 hr in the presence of 1 mM of ADP and 1 mM of MgCl2.
Analysis of B7-H1 expressed on B16F1 cells after treatment with IFN-γ
B16F1 melanoma cells (5 × 105/ml) were cultured in the presence of 5, 10 or 20 ng/ml murine IFN-γ (eBioscience, san Diego, CA) in DMEM complete medium for 48 hr, and then, the cells were washed 3 times with PBS containing 0.2% FCS. The expression of B7-H1 on melanoma cells were examined by FACS analysis as described previously.25
Analysis of gene expression by real-time PCR
Total RNA was extracted using TRIzol reagent (Invitrogen Life Technologies, Carlsbad, CA), and samples were incubated with RNase-free DNase I (Promga) at 37°C for 30 min to avoid amplification/detection of contaminating genomic DNA. RNA concentration was measured spectrophotometrically and equal amounts of RNAs were reverse-transcribed. Real-time quantitative PCR was performed in an ABI PRISM 7700 sequence detection system using the 5′-nuclease method (TaqMan). Primers and TaqMan probes were designed using the primer design software Primer Express (PE Applied Biosystems), except those for β-actin, which were available commercially. The primer and probe sequences were as follows: B7-H1, sense 5′-GGAATTGTCTCAGAATGGTC-3′, antisense 5′-GTAGTTGCTTCTAGGAAGGAG-3′ and probe 5′-CACCAAACCAGCTCTATTCCCTCAGCCTAT-3′. IFN-γ, sense 5′-CAGCAACAGCAAGGCGAAA-3′, antisense 5′-TTCCTGAGGCTGGATTCCG-3′ and probe 5′-CCAGCGCCAAGCATTCAATGAGCT-3′. IL-2, sense 5′-TGGAGCAGCTGTTGATGGAC-3′, antisense 5′-C AATTCTGTGGCCTGCTTGG-3′ and probe 5′-ACTCCCCAGGATGCTCACCTTC-3′. foxp3, sense 5′- TGCCACCTGGGATCAATGT-3′, antisense 5′-CCAGCAGTGGGTAGGATCCTT-3′ and probe 5′-CTCTACTCTGCACCTTCCCACGCT-3′. IL-10, sense 5′-GGTTGCCAAGCCTTATCGGA-3′, antisense 5′-ACCTGCTCCACTGCCTTGCT-3′ and probe 5′-TGAGGCGCTGTCGTCATCGA-3′. TGF-β, sense 5′-TGGCTTCTAGTGCTGACGC-3′, antisense 5′-TAGTTTGGACAGGATCTGGC-3′ and probe 5′-CCACCTGCAAGACCATCGACA -3′. 20 ng of each cDNA samples, except for 10 ng of β-actin cDNA, was mixed with primers and TaqMan Universal PCR Master Mix in a total volume of 25 μl as described in the manufacturer's directions (protocol 4304449; PE Applied Biosystems). The PCR was conducted using the following parameters: 50°C for 2 min, 95°C for 10 min, and 40 cycles at 95°C for 15 sec and 60°C for 1 min. Quantification of the expression of genes was performed using the comparative CT method (Sequence Detector User Bulletin 2; Applied Biosystems). The expression level of each mRNA was normalized to β-actin mRNA and expressed as n-fold difference relative to the control (calibrator). All PCR assays were performed in triplicate at 2 RNA concentrations, and results are represented by the mean values.
Activation of spleen cells by HSP70-peptide complexes in vitro
Spleen cells were prepared and stimulated by HSP70-melanoma peptide complexes (HSP70 vaccine) in vitro as previously described.25 The splenocytes were cultured at the concentration of 1 × 107/ml in RPMI 1640 supplemented with 20 U/ml IL-2 (PeproTech, London, UK) in a 96-well culture plate in the presence of 0.75 μg/ml of HSP70 vaccine. The cells were passaged 48 hr later and cultured for another 5 days, and then used for cytotoxicity assay.
Pulmonary metastasis model
To establish pulmonary metastasis model, C57BL/6 mice were injected i.v. with 5 × 105 B16F1 melanoma cells suspended in 100 μl of PBS. Metastatic foci appeared on lungs as discrete black pigmented foci that were easily distinguishable from normal lung tissue. The number of pulmonary metastatic foci in treatment and control groups was counted in a double-blind fashion on day 14 or 28 after tumor inoculation. Metastatic foci, which were too consolidated and numerous to count, were assigned an arbitrary value of ≥600.
For the treatment of mice with HSP70 vaccine, each mouse received a subcutaneous injection of 100 μl of HSP70 vaccine into the armpit skin starting 1 day after tumor inoculation, once a day for 7 days. Mice of the control group received subcutaneous injection of an equal volume of saline. In B7-H1 blockade experiments, in vivo transfection with pPD-1A was performed by the hydrodynamics-based gene delivery technique.29, 30, 31 Naked plasmid pPD-1A or pcDNA3.1 dissolved in 2 ml of saline was injected intravenously via the tail vein on day 2 after tumor injection. Dosage and injection frequency of pPD-1A or pcDNA3.1 are indicated in the figure legends.
The method of standard Western blot for the detection of protein expression has been previously described.32, 33 For detection of sPD-1 expression in vivo, livers from the mice transfected with pPD-1A by the hydrodynamics-based gene delivery technique were excised 24 hr later and digested with collagenase and EDTA, the collected hepatic cells were cultured at the concentration of 5 × 106 cells/ml. The proteins in 48-hr culture supernatants or sera collected from the mice 48 hr after transfection were separated by electrophoresis in a 12% polyacrylamide gel and then transferred onto a nitrocellulose membrane (Biorad, Ivry sur Seine, France). Membranes were blocked with 3% BSA in TTBS, and then incubated with 0.2 μg/ml anti-mouse PD-1 ployclonal Ab (AF1021, R&D systems, Minneapolis, MN). After washing, membranes were incubated with peroxidase-labeled anti-goat IgG Ab (Amersham Pharmacia Biotech, Piscataway, NJ), the bound antibodies were detected using the ECL Western Blotting Detection System (Amersham Biosciences) according to the manufacturer's instructions.
The levels of IFN-γ and IL-2 in serum were assessed by enzyme-linked immunoadsorbent assay (ELISA), using murine IFN-γ or IL-2 ELISA kits (eBioscience, san Diego, CA), according to the manufacturer's protocol.
Standard 51Cr-release assays were performed. Briefly, B16F1 melanoma target cells, either untreated or treated with IFN-γ sPD-1, were labeled with Na51CrO4 (0.1 μCi/106 cells; Amersham Pharmacia Biotech) at 37°C for 1 hr. After extensive washing, target cells were incubated with effectors at different E:T ratios in triplicate for 4 hr at 37°C, and 51Cr release (cpm) into the supernatants was measured in a γ-counter to calculate specific lysis. Specific lysis was determined as follows: percent specific release = 100 × (experimental cpm − spontaneous release)/(maximum release − spontaneous release). Spontaneous release was ≤ 20% of maximum release in all experiments.
Isolation of tumor-infiltrating cells
Tumor-infiltrating cells from lung parenchyma were isolated as described.34, 35, 36 Lungs were flushed in situ with PBS via right atria to remove residual intravascular blood pools, then, the lung tissues were minced and digested in PBS containing 0.1% collagenase (sigma), 0.01% hyaluronidase (sigma) and 0.002% DNase I (Promga) for 90 min at 37°C. The tissue debris that was not digested was allowed to settle and the released cells were then filtered through stainless-steel mesh screens, and the cells were washed 3 times with RPMI 1640 containing 5% FCS. Cells were separated on a Percoll (Pharmacia Biotech AB, Uppsala, Sweden) density gradient by centrifugation for 30 min at 1500g at room temperature. The dense layer containing enriched lymphocytes was collected and washed. Total RNA was extracted from the isolated lymphocytes using TRIzol reagent (Invitrogen Life Technologies, Carlsbad, CA) and used for quantitative real-time PCR analysis.
Lung tissues were fixed in situ with 4% paraformaldehyde injected via right atria on day 14 after tumor inoculation. The tissues were further fixed in 4% paraformaldehyde, balanced in 20% sucrose, embedded in optimum cutting temperature solution and cut into 10-μm sections. Slides were incubated with anti-CD3 antibody (145-2C11 eBioscience) with 1:100 dilution, and then with biotinylated mouse anti-hamster IgG antibody (Proteintech Group, New York, CA). The reaction product was visualized with the peroxidase-conjugated streptavidin system, using 3, 3-diaminobenzidine (Serva, Heidelberg, Germany) as a substrate.
Results were expressed as mean values ± SD and the differences were determined by ANOVA test, except for survival rate determined by Wilcoxon's rank-test. A value less than 0.05 (p < 0.05) was used for statistical significance.
HSP70 vaccine failed to control the development of metastatic melanoma at later stage
B16F1 cells (5 × 105) injected through tail vein into mice grew rapidly and formed about 117–190 pulmonary metastatic foci within 14 days. All mice in the saline control group presented a large number (range, 117–190) of metastatic foci in their lungs (Fig. 1a). The treatment with HSP70 vaccine significantly reduced metastasis, with the number of metastatic foci ranging from 8 to 20 (p < 0.05 vs. control group). On day 14 after tumor inoculation, mice were sacrificed and lung samples as well as serum samples were collected. Immunohistochemical analysis confirmed the extensive infiltration of T lymphocytes in the area around tumor cells in HSP70 vaccine-treated group compared with that of control group (Fig. 1b). The levels of IL-2 and IFN-γ in sera of mice treated with HSP70 vaccine were significantly higher than those of control mice (Fig. 1c, p < 0.05). These data indicated that the strong immune responses were indeed generated after HSP70 vaccine therapy. However, when the animals treated with HSP70 vaccine were observed for up to 28 days, plenty of metastatic foci were observed (shown below in Fig. 3b) and all animals had to be sacrificed within 48 days (Fig. 1d).
B7-H1 mediates tumor resistance to immune attack
We then asked why the presence of large number of activated T cells at tumor site failed to control progressive melanoma growth finally. Previous study reported that TH1 cytokine IFN-γ contributed to upregulated expression of B7-H1 in tumor. To test whether IFN-γ could induce the expression of B7-H1 on melanoma B16F1 cells in our model system, we first incubated B16F1 cells with different doses of IFN-γ in vitro. FACS analysis revealed that B7-H1 was hardly detectable on the untreated B16F1 cells. By contrast, the percentage of B7-H1-positive B16F1 cells was significantly increased by IFN-γ stimulation (p < 0.05, Fig. 2a), and reached 91.8% in the presence of 20 ng/ml of IFN-γ. Consistent with this result, the upregulatd expression of B7-H1 in vivo in tumor microenvironment was also confirmed by real-time PCR. The expression of B7-H1 was increased by 4.86-fold in Hsp70 vaccine group, but only 0.93-fold in control group, suggesting that the residual melanoma cells in HSP70 vaccine group altered their phenotype rapidly to upregulate the expression of B7-H1.
To test whether B7-H1 on melanoma cells is responsible for the resistance of tumor cells to immune attack, the sensitivity of B16F1 cells, treated or untreated with IFN-γ in vitro, to the cytotoxicity of the activated lymphocytes was determined by incubation with the spleen cells activated by HSP70 vaccine in vitro. As shown in Figure 2c, IFN-γ-treated B16F1 cells were significantly less susceptible to the cytotoxicity of the activated spleen cells than that of B16F1 cells without treatment with IFN-γ. When IFN-γ-treated B16F1 cells were pretreated with sPD-1, which could bind specifically and efficiently to B7-H1 and block its function,25 their susceptibility to the cytotoxicity of the activated spleen cells was almost completely recovered, indicating that immune resistance was not because of other properties of IFN-γ-treated B16F1 but rather because of a specific resistance mediated by tumor-associated B7-H1.
To evaluate whether the blockade of tumor B7-H1 by sPD-1 could be applied to enhance the antitumor efficacy of HSP70 vaccine in vivo, the mice challenged with B16F1 cells by tail vein injection were treated with HSP70 vaccine in combination with in vivo transfection with naked pPD-1A, which was performed by the hydrodynamics-based gene delivery technique.29, 30, 31 The expressed product of pPD-1A was first confirmed by Western blot analysis (Fig. 3a). Treatment with either HSP70 vaccine alone or pPD-1A alone or HSP70 vaccine combined with control plasmid pcDNA3.1 (HSP70 vaccine/pcDNA3.1) produced similar therapeutic effect. As described above, treatment with HSP70 vaccine resulted in significant inhibition on the formation of metastatic foci, but there were still many metastatic foci formed on day 28 after tumor inoculation (Fig. 3b), whereas the treatment with HSP70 vaccine in combination with pPD-1A (combination therapy) resulted in remarkable improvement of efficacy in a significant fraction of animals. The mean number of lung metastatic foci in mice receiving combination therapy was significantly lower than that in pPD-1A-treated mice (5.16 ± 2.63 vs. 222.16 ± 69.15; p ≤ 0.01), that in HSP70 vaccine-treated mice (5.16 ± 2.63 vs. 193.33 ± 53.54; p ≤ 0.01) and that in HSP70 vaccine/pcDNA3.1-treated mice (5.16 ± 2.63 vs. 188.00 ± 54.18; p ≤ 0.01). Survival studies were performed to further evaluate the therapeutic efficacy of HSP70 vaccine in combination with sPD-1. As shown in Figure 3c, the mean survival times of mice receiving either HSP70 vaccine alone or pPD-1A alone were 41.00 ± 3.28 and 42.67 ± 4.13 days, respectively. And the mean survival times of mice receiving HSP70 vaccine/pcDNA3.1 were 43.30 ± 4.13. Whereas, the mean survival time of mice receiving combination therapy was significantly prolonged (p < 0.01) with 83.3% of mice surviving to day 80. All these results indicated that sPD-1 and HSP70 vaccine produced a significant synergistic therapeutic effect.
More efficient induction of tumor-specific CTLs by combination therapy
To analyze antitumor immunity in vivo after the therapy with HSP70 vaccine in combination with sPD-1, splenocytes were prepared from mice on day 28 after tumor inoculation and antitumor CTL activity was tested. As shown in Figure 4a, the cytotoxicity of splenocytes from the combination treatment group to melanoma target cells was significantly augmented, compared with HSP70 vaccine treatment group, pPD-1A treatment group or HSP70 vaccine/pcDNA3.1 group. Meanwhile, no specific cytotoxicity to EL-4 target cells was observed. Thus, combination therapy resulted in the generation of more tumor-specific CTLs in vivo.
To assess the immunological reactivity in tumor microenvironment in the process of combination therapy, we serially examined the expression of several genes in tumor-infiltrating lymphocytes (TILs) by real-time quantitative PCR. Resulting TILs received no in vitro culture or stimulation so that a true representation of in vivo gene expression and immune reactivity could be ascertained.37 As shown in Figure 4b, the expression levels of IFN-γ and IL-2 in TILs from HSP70 vaccine treated mice and HSP70 vaccine/pcDNA3.1 treated mice were significantly increased on day 7; however, there was a slight trend toward significant decrease in the following weeks. The influence of IFN-γ and IL-2 expression by sPD-1, produced after transfection with pPD-1A alone, was not as strong as that by HSP70 vaccine. However, the expression of IFN-γ and IL-2 in TILs from combination therapy group was increased by 3.2-fold and 2.5-fold, respectively, on day 7, and maintained at high levels to day 28.
As the negative regulatory molecules were concerned, there was a predominant decrease of IL-10 expression in pPD-1A-treated group, but not in HSP70 vaccine group and not in HSP70 vaccine/pcDNA3.1 group. The expression of IL-10 was further reduced, by 8.5-fold on day 14, in TILs from combination therapy group. Furthermore, TILs from combination therapy showed further decrease of expression levels of TGF-β and foxp3. Therefore, HSP70 vaccine in combination with sPD-1 resulted in higher and more durable expression of TH1 cytokines and dramatic decrease of the expression of negative regulatory molecules by TILs.
Dose–effect relationship of pPD-1A transfection in combination therapy
To investigate the dosage and regimen of administration of pPD-1A required to obtain antitumor efficacy, the expression pattern of therapeutic gene after the hydrodynamics-based gene delivery was first detected by using Western blotting approach. After a single i.v. injection of 10 μg of pPD-1A expression vector, sPD-1 in the serum increased rapidly and reached peak serum level 1 day after injection. The serum sPD-1 level was gradually decreased thereafter (Fig. 5a). We then set the injection frequency on the basis of this expression pattern. As shown in Figure 5b, in vivo transfection by hydrodynamic injection of 10 μg of pPD-1A, twice a week or once a week, resulted in a significant decrease of tumor metastatic foci on day 28 after tumor inoculation. As few as 3 injections of 10 μg of pPD-1A administered once a week had a significant antitumor effect in combination with HSP70 vaccine. But the 5-μg dose of pPD-1A revealed a dose–response based on the frequency of injection treatment twice a week with 5 μg of pPD-1A resulted in a significant reduction of tumor foci, whereas treatment once a week with 5 μg of pPD-1A did not produce a significant improvement of antitumor response.
It is critical for the immunologic destruction of tumor cells in vivo that the immune cells at the tumor site manifest appropriate effector mechanisms such as direct lysis or the secretion of cytokines capable of causing tumor destruction.38 Recent research has emphasized the importance of suppressor mechanisms arising from the tumor that can inhibit antitumor immune reactions in vivo. Along with the development of tumor vaccines, it is not difficult to induce a strong immune response against tumor in vivo, but it is still a big problem that tumor cells could not be eliminated by the activated immune cells. The immune cells may be functionally altered after they enter the tumor microenvironment, where tumor cells can evade immune attack by developing active suppressor mechanisms.23, 24, 39 Therefore, the reasons for the paradoxical dissociation between vaccine-induced T-cell responses and the lack of their effectiveness in inducing tumor regression should be sought by complementing the analysis of the systemic effects of vaccines with the simultaneous study of tumor–host interactions at the tumor site.40
In the present study, the immunization with HSP70 vaccine led to an effective inducement of specific T-cell responses against melanoma antigen, which was indicated by the significant infiltration of T cells into the tumors, a profile of the expression of TH1 cytokines IFN-γ and IL-2, as well as the successful inhibition of B16F1 melanoma in vivo. Nonetheless, HSP70 vaccine-induced tumor-specific T cells did not coincide with tumor clearance, indicating that the residual tumor cells were capable of escaping from immune attack.
The results of the present study indicated that the inducible expression of B7-H1 on melanoma profoundly contributed to the tumor's resistance to the attack of T cells. B7-H1 on tumor cells inhibited antigen-specific CTL, and this inhibition was associated with progressive tumor growth in the face of existing TAA-specific immunity. In support of this concept, blockade of B7-H1 by recombinant sPD-1 following HSP70 vaccine promoted the inhibitory effect on tumor growth and extended the survival of tumor-bearing mice. Thus, upregulation of B7-H1 on tumors may be an important mechanism for tumor cells to escape from the immune attack induced in the process of immunotherapy.
The inhibitory effect of HSP70 vaccine on the formation of metastatic foci suggested that the effect of tumor-associated B7-H1 on tumor resistance was little at activation stage. The high levels of B7-H1 expressed by residual melanoma at the effector phase could engage PD-1 receptor on effector T cells so as to result in preferentially delivering an inhibitory signal, and may be play a more critical role in suppressing the execution of T-cell effector function.23, 41 Since the constitutive and inducible expression of B7-H1 has been widely found in most tumor cells, our studies provided a possible explanation for the limited success of vaccination and immunotherapy trials. It is possible that blockade of B7-H1 may prevent the escape of tumor cells from the attack of T cells and augment the efficacy of vaccination and immunotherapies.
IFN-γ is crucial for establishing tumor-specific immunity and TH1 polarization of many immunotherapy approaches.42, 43, 44 But on the other hand, IFN-γ contributes, at least in part, to the up-regulation of B7-H1 on melanoma, which was further supported by our results in this study. Since the important role of IFN-γ in immune response, it is unfeasible to inhibit the production of IFN-γ so that the up-regulation of B7-H1 in the process of immunotherapy is an inevitable event. Therefore, we explored the possibility of blockade of B7-H1 by sPD-1, without influencing the production of IFN-γ.
In this study, we used soluble PD-1 to block the function of tumor B7-H1, which differed considerably from previous studies of blocking B7-H1 with antibodies against B7-H1 or PD-1.22, 23, 24 The soluble receptors, such as sFas and sCTLA-4, represent truncated forms of the membrane-bound receptors. They bind the same ligands as those of their membrane-bound counterparts, and play an important role in the regulation of immune response.45, 46 At present, native soluble form of soluble PD-1 has not been found. We generated sPD-1 eukaryotic expression plasmid (pPD-1A) by inserting cDNA encoding the extracellular domain of PD-1 into pcDNA3.1. Our previous studies have shown that the expression of pPD-1A, also used in the present study, could produce biologically active sPD-1.25 The expression of sPD-1 in vivo alone could inhibit the metastatic melanoma to some extent, but it was more important that the gene therapy with sPD-1 significantly improved the efficacy of HSP70 vaccine, indicating that it should be a very important strategy to block the function of B7-H1 in the process of immunotherapy, and sPD-1 was suitalbe for this purpose.
The serial gene analysis of TILs allows a direct molecular analysis of immune responses within tumor microenvironment. The treatment with HSP70 vaccine resulted in the increased expression of TH1 cytokine IFN-γ and IL-2 genes, but had little impact on the expression of immunosuppressive cytokines. These results demonstrated the presence of, perhaps the limited, effectiveness of HSP70 vaccine-induced T-cell response within the tumor environment. But sPD-1 was capable of reducing the expression of IL-10 gene. The combination therapy not only increased the gene expression of TH1 cytokines, but also significantly inhibited the expression of the genes of negative regulatory molecules, including IL-10, TGF-β and foxp3. The reasonable explanation for the decrease of TGF-β and foxp3 gene expression should be that the stronger immunotherapeutic effect reduced the production of TGF-β by tumor cells so to reduce the expression of foxp3 gene, which could be induced by TGF-β in combination with antigens. These data indicated that the therapy with HSP70 vaccine in combination with sPD-1 established a better microenvironment that favored the antitumor immune response and the attack of tumor cells by immune cells.
Hydrodynamics-based gene therapy using pPD-1A in this report has important advantages over traditional antibody treatments. Unlike antibody therapies, the lower, but sustained, serum levels of a therapeutic protein produced by the therapeutic vector may reduce side effects while still maintain an ideal therapeutic effect. A single i.v. injection of expression vector in mice resulted in the existence of the expressed products in serum over a 7-day period. Treatment with pPD-1A in mice bearing melanoma tumors was effective even when the vector was given as infrequently as once a week for a total of only 3 injections. By avoiding the sharp peaks and drops in serum levels that occur after bolus protein injection, gene therapy may sustain a low but effective concentration of therapeutic molecules in the serum, possibly reducing the severity of side effects and prolonging the therapeutic effect.
The immunotherapy with tumor vaccines is one of the most promising approaches in tumor therapy. But there are still many key points that are needed to be found and controlled in the process of immunotherapy of tumor. The improved efficacy of immunotherapy with HSP70 vaccine in combination with sPD-1 revealed one of these key points, which should be controlled or regulated for a better efficacy of immunotherapy. The findings presented in this report have important implications for designing of cancer immunotherapy. The blockade of B7-H1 with sPD-1 will be a valuable approach in antitumor immunotherapeutic strategy.
Acknowledgements: We are grateful to Yan Yang, Zhi-Yong Gong, Yi-Nong Zhang and Mei-Rong Zheng for technical assistance.