Highly optimized DNA vaccine targeting human telomerase reverse transcriptase stimulates potent antitumor immunity

Abstract

High levels of human Telomerase Reverse Transcriptase (hTERT) are detected in over 85% of human cancers. Immunological analysis supports hTERT is a widely applicable target recognized by T cells and can be potentially studied as a broad cancer immune therapeutic, or a unique line of defense against tumor recurrence. There remains an urgent need to develop more potent hTERT vaccines. Here, a synthetic highly optimized full-length hTERT DNA vaccine (phTERT) was designed and the induced immunity was examined in mice and non-human primates. When delivered by electroporation, phTERT elicited strong, broad hTERT-specific CD8 responses including induction of T-cells expressing CD107a, IFN-γ and TNF-α in mice. The ability of phTERT to overcome tolerance was evaluated in a NHP model, whose TERT is 96% homologous to that of hTERT. Immunized monkeys exhibited robust (average 1834 SFU/10⁶ PBMCs), diverse (multiple immunodominant epitopes) IFN-γ responses and antigen-specific perforin release (average 332 SFU/10⁶ PBMCs), suggesting phTERT breaks tolerance and induces potent cytotoxic responses in this human relevant model. Moreover, in an HPV16-associated tumor model, vaccination of phTERT slows tumor growth and improves survival rate in both prophylactic and therapeutic studies. Lastly, in vivo cytotoxicity assay confirmed that phTERT-induced CD8 T cells exhibited specific CTL activity, capable of eliminating hTERT-pulsed target cells. These findings support that this synthetic EP-delivered DNA phTERT may have a role as a broad therapeutic cancer vaccine candidate.

Introduction

Immunotherapy of cancer through induction of anti-tumor cellular immunity has recently re-emerged as an important experimental therapy for the treatment of nonresponsive cancers. However, most tumor-associated antigens (TAAs) are expressed in one or a few tumor types as tumors generally exhibit tissue-specific features (1). In contrast, human telomerase reverse transcriptase (hTERT), a catalytic subunit of telomerase, is highly expressed in more than 85% of human tumors from diverse cancer phenotypes, with little or no expression in normal somatic cells (2-6). Expression of hTERT correlates with telomerase activity, which may be a requirement for tumor survival (7). Telomerase activation/hTERT expression is associated with little loss of telomere length and accounts for the unlimited proliferative capacity of cancer cells. As expression of hTERT is directly linked to tumor cell growth and contributes crucially to the long-term survival of tumor cells, loss of telomerase activity will lead to hTERT-positive tumor cell death by apoptosis (8). In addition, targeting hTERT may have the potential to eliminate cancer stem cells as recent studies have suggested that cancer stem or stem-like cells express hTERT (9-11). These findings collectively point to hTERT as an attractive TAA and provide the basis of developing hTERT-based universal vaccine for cancer immunotherapy (12-14).

Therapeutic hTERT vaccines have been widely studied because of their potential to stimulate the killing of tumor cells by enhancing the activity of telomerase-specific cytotoxic CD8 T cells (15). Many studies have been performed to develop hTERT peptide vaccines containing motifs that either bind to class I (I540 and 572Y) or II molecules (GV1001) (16-19). Moreover, multiple strategies are being explored to use full-length hTERT recombinant constructs targeting multiple CD8 and CD4 epitopes simultaneously. As a result, the potency of autologous dendritic cells transduced with hTERT mRNA (20, 21) and viral vector-based vaccines (22, 23) have been reported with partial success in animal models. However, findings from initial clinical trials of hTERT vaccines in cancer patients have shown that these approaches suffer from limited induction of CD8⁺ T cells and had limited impact on overall survival (13). Therefore, there remains an urgent need to develop more potent hTERT therapeutic vaccines with a broader T cell footprint.

Several clinical trials have evaluated the efficacy of DNA vaccination against a variety of cancers (24). Although there are some indications of limited immune responses in vaccinated melanoma or prostate cancer patients (25, 26), previous DNA vaccines generally appear to induce weak cellular immunity in humans. Recent synthetic DNA design strategies, such as codon/RNA optimization, the addition of highly efficient immunoglobulin leader sequences (27-29), use of more efficient DNA delivery methods including in vivo electroporation (EP) (30), have been applied to improve the immune responses induced by DNA vaccines in humans, with recent significant success (31). However, these new approaches have not been combined to test a new hTERT DNA vaccine.

In this report, we attempt to extend this improved immune potency to construct a synthetic DNA vaccine expressing a full-length hTERT with modifications using a combination of approaches in gene optimization. The hTERT DNA was delivered by EP and its immunogenicity and antitumor effect were evaluated in NHP and mice. These data strongly support further study of the hTERT DNA vaccine in combination with electroporation delivery as a potential immunotherapy platform against an array of human and animal malignancies.

Materials and Methods

Immunogen design and expression

A synthetic hTERT DNA vaccine was generated using the hTERT sequence retrieved from GenBank (accession number: AF018196) with several modifications (Fig. 1A). The full-length hTERT gene was 3512 bp and subcloned into an expression vector pGX0001.

Figure 1.

Design and expression of hTERT DNA vaccine. A, Schematic of hTERT antigen. B, Map of phTERT. C, Detection of phTERT expression by in vitro translation. The gene product was immunoprecipitated using an anti-HA monoclonal antibody, visualized by SDS-PAGE and autoradiography. D, Immunofluorescence assay of phTERT. Transfected RD cells expressing hTERT protein showed typical FITC-fluorescence using a commercial hTERT (C-term) monoclonal antibody.

In vitro hTERT expression was detected utilizing TNT® Quick Couple Transcription/Translation System (Promega). The gene product was immunoprecipitated using an anti-HA monoclonal antibody (Sigma-Aldrich) and analyzed by SDS-PAGE. The synthesized protein was detected by autoradiography.

An indirect immunofluorescent assay was conducted to confirm hTERT expression as previously described (32). Briefly, human rhabdomyosarcoma (RD) cells were transfected with phTERT and pGX0001 (1 ug/well) using TurboFectin™8.0 Transfection Reagent (OriGene). Forty-eight hours later, the cells were fixed and incubated with anti-hTERT (C-term) monoclonal antibody (Millipore) overnight at 4°C. The slides were then incubated with fluorescein isothiocyanate (FITC)-conjugated secondary antibody (Cell Signaling Technology), and analyzed by fluorescent microscopy (Leica Microsystems, Inc.) using the SPOT Advanced software (SPOT™ Diagnostic Instruments, Inc).

Mice studies

Mice and Immunization

Female 8-week-old C57BL/6 mice were purchased from Jackson Laboratory. Their care was in accordance with the guidelines of NIH and University of Pennsylvania Institutional Animal Care and Use Committee (IACUC). Mice were divided into two groups and immunized with 50 ug of hTERT DNA by intramuscular injection (IM) into the quadriceps followed by electroporation (EP) using the CELLECTRA® adaptive constant current device (Inovio Pharmaceuticals Inc.)(33). The mice received 4 immunizations, 2 weeks apart. One week after the last immunization, the mice were sacrificed and splenocytes were isolated for immunology studies.

ELISpot assay

Mouse IFN-γ ELISpot assay was performed as previously described (32). Peptides spanning the entire hTERT protein, each containing 15 amino acids overlapping by 8 amino acids, were synthesized by GenScript. The entire set of peptides was pooled at a concentration of 2 ug/ml/peptide into 4 pools for stimulation of the IFN-γ release.

CD8⁺ T cell depletion

CD8 lymphocytes were depleted from splenocytes using Dynabeads Mouse CD8 (Lyt2) (Life Technologies). After CD8⁺ depletion, IFN-γ ELISpot was performed as described above.

Intracellular cytokine staining

Intracellular cytokine staining was performed as described previously (33). Briefly, splenocytes from vaccinated and naïve mice were stimulated with hTERT peptides, stained with FITC anti-mouse CD107a, and followed by ViViD Dye (Invitrogen). Cells were then stained with the following extracellular antibodies: APC-Cy7 anti-mouse CD3e, PerCP-Cy5.5 anti-mouse CD4, and APC anti-mouse CD8a (BD Biosciences). Intracellular cytokines were subsequently stained with the following antibodies: Alexa Fluor 700 anti-mouse IFN-γ, and PE-Cy7 anti-mouse TNF (BD Biosciences).

Cell line

TC-1 cell line is a well-characterized lung epithelial cell line immortalized with HPV16 E6/E7, ,and transformed with the c-Ha-ras oncogene (34). TC-1 cells were purchased from ATCC and maintained in RPMI 1640 medium supplemented with 10% fetal bovine serum, 100 U/ml penicillin and 100 ug/ml streptomycin at 37°C.

In vivo tumor challenge study

Ten female 8-week-old C57BL/6 mice were immunized with 50 ug of phTERT four times biweekly. One week after the last immunization, each mouse was challenged with 5×10⁴ TC-1 cells injected subcutaneously including ten naïve mice that served as a control. Tumors were measured twice weekly with digital calipers spanning the shortest (width) and longest surface diameters (length) (35). Tumor volumes were calculated according to the formula: V = length × width² × π /6 (36). Mice were sacrificed when tumor diameter reached 20mm in compliance with our Institutional Animal Care and Use Committee protocol.

In vivo tumor treatment study

Female C57BL/6 mice were separated into two groups of 10 mice: naïve and hTERT group. On day 0, all mice were injected subcutaneously with 5×10⁴ TC-1 cells in the right flank. All mice in the hTERT group were immunized with 50 ug of phTERT on days 3, 10, 17 and 24. Tumors measurement was performed as described above.

In vivo cytotoxicity study

An in vivo cytotoxicity assay was performed as previously described (37, 38). Splenocytes from naïve mice were stained with carboxyfluorescein diacetate succinimidyl ester (CFSE) at a concentration of 1 μM or 1 nM. CFSE^hi (1 μM) labeled cells were pulsed with the relevant peptides (hTERT peptides), while CFSE^lo (1 nM) labeled cells were pulsed with the irrelevant peptides (HPV6 E6/E7 peptides). Equal frequency of CFSE^hi and CFSE^lo cells were combined and 10⁷ cells were intravenously injected into naïve or phTERT-immunized mice. Forty-eight hours later, splenocytes were isolated and analyzed by flow cytometry. The percent killing was calculated as follows: 100 - ([(% relevant peptide pulsed in immunized / % irrelevant peptide pulsed in immunized) / (% relevant peptide pulsed in naïve / % irrelevant peptide pulsed in naive)] × 100).

Rhesus monkey studies

Immunization and PBMC Isolation

Four rhesus macaques were vaccinated with phTERT, four times intramuscularly followed by EP using CELLECTRA® adaptive constant current EP device, six weeks apart, at 2 mg DNA/each immunization. Blood were collected two weeks post each immunization and PBMCs were isolated by standard Ficoll-Hypaque density gradient centrifugation.

IFN-γ and Perforin ELISpot assay

Monkey IFN-γ and Perforin ELISpot were performed as previously described (39, 40). Antigen-specific responses were determined by subtracting the number of spots in the negative control wells from the wells containing peptides. After subtracting the negative control, the mean value in the wells with the PBMCs collected post vaccination had to exceed 50 SFU/10⁶ PBMCs and be at least four times higher than pre-vaccination reactivity to be considered as a positive response.

Statistical analysis

Standard and paired student's t-tests were applied to analyze statistical significance of all quantitative data produced in this study, and p < 0.05 was considered statistically significant.

Results

Design and Construction of the full-length hTERT DNA vaccine

As indicated in Fig. 1A, the hTERT immunogen was developed with several modifications, including codon/RNA optimization and the addition of a highly efficient leader sequence, to enhance the expression and immunogenicity of phTERT. Two mutations (R589Y and D1005Y) were incorporated into the hTERT sequence to assist in breaking tolerance (41). The modified gene was subcloned into pGX0001 and named as phTERT for further study (Fig. 1B)

Expression of hTERT

In vitro expression of hTERT was verified by T7 coupled transcription and translation reaction. After immunoprecipitation with the anti-HA monoclonal antibody, hTERT expression was analyzed by 12% SDS-PAGE. The hTERT protein migrated to the corresponding molecular weight at approximately 130 kDa (Fig. 1C). No protein band was detected in the pGX0001 vector lane. An indirect immunofluorescent assay was performed to further confirm hTERT expression. As shown in Fig. 1D, the cells expressing hTERT protein showed typical FITC-fluorescence, supporting the hTERT was expressed with a relatively native conformation. As a control, expression was not detected in pGX0001-transfected cells.

Vaccination with phTERT induces strong CD8 - mediated hTERT-specific responses in mice

IFN-γ ELISpot was performed to assess antigen-specific cellular immune responses induced by phTERT after four immunizations (Fig. 2A). The total response against four pools of hTERT peptides in phTERT-immunized mice was 1817 ± 211 SFU/10⁶ splenocytes, which was significantly greater than the immune responses in the naive group (20 ± 6 SFU/10⁶ splenocytes) (p = 0.01) (Fig. 2B). The C-terminus of hTERT protein (peptide pool 4) was the most immunogenic region, accounting for about half the response elicited by phTERT (970 ± 113 SFU/10⁶ splenocytes).

Figure 2.

phTERT elicited robust CD8⁺ immune response in C57BL/6 mice. A, Immunization schedule. B, Total IFN-γ responses induced by phTERT. Frequencies of IFN-γ-secreting cells/10⁶ splenocytes after four immunizations with phTERT were determined by IFN-γ ELISpot assay. Splenocytes from each mouse (10 mice per group) were stimulated with hTERT peptides. Results were presented as mean ± SEM. C, Total IFN-γ responses induced by phTERT after CD8 depletion. CD8 T cells were depleted from splenocytes and IFN-γ ELISpot assay was performed. D, Schematic diagram of gating strategy of intracellular cytokine staining. E, Secretion of CD107a, IFN-γ, and TNF-α post vaccination in both CD4⁺ and CD8⁺ cells. Splenocytes were stimulated with hTERT peptides for 5 hours prior to surface and intracellular staining. Background-subtracted percentages of hTERT-specific CD4+ or CD8+ T cells producing CD107a, IFN-γ, and TNF-α were calculated. The experiment shown is representative of three different experiments using 5 mice per group.

Another ELISpot assay was performed after CD8 depletion to determine whether CD8+ cells were responsible for the detected robust IFN-γ responses. The results indicated that the spot number was reduced to 244 + 78 in the phTERT-vaccinated mice (a 85% decrease in the frequency of IFN-γ producing cells) after CD8+ depletion (Fig. 2C), supporting that the vaccine-induced IFN-γ production is mediated mainly by CD8+ T-cells.

The breadth of T cell response was suggested to be important for anti-tumor effect. Accordingly, an additional ELISpot assay was performed against 26 hTERT matrix peptide pools. There were 9 matrix pools showing more than 50 spots, indicating phTERT could elicit a broad range of T-cell immune responses (Supplementary Fig. S1A). There were eight epitope-comprising peptides in the region from aa 288 to 869 (Table 1). The C-terminus (aa 862-1158) had the most epitope-comprising peptides (8 peptides), which was consistent with the result in Fig. 2B. Individual peptides in pool 4 were used to confirm the matrix mapping result against the C-terminus of hTERT, and two immunodominant epitope-containing peptides (RKTVVNFPVEDEALG and KNPTFFLRVISDTAS) were identified (Supplementary Fig. S1B). As shown in Supplementary Fig. S1C, some of the identified peptides were conserved between mTERT and hTERT. All peptides listed in Table 1 were confirmed to contain one H2-D^b-restricted epitope by using IEDB analysis resource Consensus tool (http://tools.immuneepitope.org), suggesting effective processing of this antigen.

Table 1.

Identified epitope-comprising peptides in mice immunized with phTERT

	Number of epitope-comprising peptides	Sequence of epitope-comprising peptides
aa 1 – 296 (pool 1)	None	N/A
aa 288 – 582 (pool 2)	4	TGARRLVETIFLGSR
		ETIFLGSRPWMPGTP
		RPLFLELLGNHAQCP
		LGNHAQCPYGVLLKT
aa 575 – 869 (pool 3)	4	DGLRPIVNMDYVVGA
		NMDYVVGARTFRREK
		SVLNYERARRPGLLG
		ARRPGLLGASVLGLD
aa 862 -1158 (pool 4)	8	RKTVVNFPVEDEALG
		PVEDEALGGTAFVQM
		DTRTLEVQSDYSSYA
		QSDYSSYA RTSIRAS
		NSLQTVCTNIYKILL
		TNIYKILLLQAYRFH
		KNPTFFLRVISDTAS
		RVISDTASLCYSILK

Vaccination with phTERT in mice enhances magnitude of IFN-γ, CD107a and TNF-α production in CD8 T cells

Intracellular cytokine staining (ICS) assays were performed to further characterize the responses induced by phTERT. The secretion of IFN-γ, CD107 and TNF-α in both CD4+ and CD8+ T cells were determined (Fig. 2D). As shown in Fig. 2E, the average frequency of CD8+IFN-γ + cells in immunized mice (0.8%) was significantly higher than that of the naive group (0.1%) (p < 0.05). The percentage of TNF-α secreting cells from the total CD8+ T cell population in phTERT-immunized mice was approximately 1.2% on average, whereas the naive group produced only 0.1% (p < 0.05). After demonstrating that antigen-specific CD8 cells had the ability to secrete IFN-γ and TNF-α, we investigated whether these CD8 cells exhibited a phenotype of putative CTLs. CD107a, a marker of cytolytic degranulation on lymphocytes, such as CD8+ cells, was used to evaluate the CTL potential of vaccine-induced T cells. Following stimulation with hTERT peptides, the percentage of CD107a-positive CD8 cells was 1.8%, which was significantly higher than the percentage in the naïve group. There was a trend showing the increased production of IFN-γ, TNF-α and CD107a in CD4 cells in the immunized mice. However, the differences in average frequencies of CD4⁺IFN-γ⁺, CD4⁺ TNF-α⁺ and CD4⁺ CD107a⁺ cells were not significant compared to what were observed in the naïve mice. The immune responses elicited by phTERT are heavily skewed towards driving CD8+ lymphocytes with the potential to lyse hTERT-expressing tumor cells.

Vaccination with phTERT is capable of breaking tolerance and generating robust hTERT-specific CTLs in rhesus macaques

The induction of T-cell immunity against the tumor antigen hTERT could be controlled by mechanisms of central and peripheral tolerance. Sequence homology analysis indicated that hTERT shares 64% identity with mouse TERT, and 96% identity with rhesus macaque TERT. Consequently, immune tolerance is expected to play a major role in testing the efficacy of an hTERT vaccine in non-human primates. Furthermore, rhesus T cell immunity is much closer to human T cell immunity serving as a highly relevant model for immunotherapeutic vaccine development. Therefore, we moved forward to determine whether phTERT is able to break tolerance and induce cellular responses in rhesus macaques. The prebleed blood samples were studied to establish the background level of immune response of each individual animal in the study.

The results showed that the hTERT-immunized monkeys exhibited very low background level of immune response (5 SFU/10⁶ PBMCs) with a dramatic increase in vaccine-induced responses following each immunization. The average numbers of IFN-γ producing cells after the first, second, third and fourth immunization were 11, 177, 834 and 1834 SFU/10⁶ PBMCs (Fig. 3A), respectively. These results showed that immunization with phTERT could elicit strong boostable hTERT-specific cellular responses. Epitope mapping was also performed to investigate the diversity of the observed immune responses in rhesus macaques. Except for animal 5015 with only two matrix pools showing ≥ 50 SFU/10⁶ PBMCs, the rest of immunized animals exhibited responses to 9 (M4628), 13 (M5012) and 10 (M5021) matrix pools with ≥ 50 spots out of 26 pools, suggesting there were multiple dominant and subdominant epitopes in response to the vaccine (Supplementary Fig. S2A). Most identified epitope-comprising peptides are either 100% conserved between NHP and humans or exhibit only a single aa difference (Supplementary Fig. S2B and Table 2), supporting that this approach can break tolerance in relevant species to epitopes with importance to human immune therapy.

Figure 3.

phTERT elicited robust CTLs in rhesus macaques. A, Total IFN-γ responses induced by phTERT following each immunization. Four rhesus macaques were vaccinated with phTERT, four times intramuscularly followed by EP, six weeks apart, at 2 mg DNA/each immunization. PBMCs were isolated and stimulated with hTERT peptides for 24 hours. Frequencies of hTERT-specific IFN-γ secreting cells/10⁶ PBMCs were determined by IFN-γ ELISpot assay. B, Enhanced perforin release in immunized monkeys. Prebleed and PBMCs post last immunization were stimulated with hTERT peptides for 24 hours. Frequencies of hTERT-specific perforin-secreting cells/10⁶ PBMCs were determined by perforin ELISpot assay. Results are presented as mean ± SEM.

Table 2.

Summary of epitope mapping results in rhesus macaques immunized with phTERT

Monkey ID	No. of epitope-comprising peptides	No.of epitope-comprising peptides with 100% identity between hTERT and RhTERT	No. of epitope-comprising peptides with single aa difference between hTERT and RhTERT
M4628	17	5	9
M5012	41	19	14
M5015	1	1	0
M5021	24	12	7

As perforin is known to be a major cytolytic protein and key effector molecule for T-cell mediated cytolysis, Perforin ELISpot was employed to determine whether hTERT-specific T cells were capable of releasing perforin from cytotoxic granules. Results showed that the average numbers of perforin producing cells were 332 SFU/10⁶ PBMCs (Fig.3B), indicating vaccination elicited cytotoxic T cells that could destroy hTERT-expressing target cells.

Assessment of physiological parameters in NHP

In order to examine if there is any CTL-mediated host toxicity, we assessed a number of physiological parameters in NHP (Supplementary Table 1). No significant weight loss was observed and WBC counts remained within normal range. No elevation of alkaline phosphatase (ALK P), alanine aminotransferase (ALT), aspartate aminotransferase (AST) and total bilirubin (TBIL) indicated that induction of hTERT-specific immune responses did not cause significant damage to the liver. No evidence of impaired kidney function was seen, as Creatinine and Blood Urea Nitrogen (BUN) remained within normal limits. Creatine phosphokinase (CPK) was evaluated to determine if EP or induction of immune responses negatively influenced skeletal or cardiac muscle. Elevation of CPK was not detected. Overall, we did not observe any vaccine-induced adverse effects in NHP despite evidence of strong hTERT-specific CTLs in vivo.

Vaccination with phTERT elicits anti-tumor immunity and delays the E6/E7 expressing tumor growth

It is important to determine whether the robust phTERT-induced T cells could actually exhibit anti-tumor immunity in an in vivo challenge model. As hTERT may have value as a broad immunogen to prevent tumor recurrence post treatment or in identified high-risk individuals, we first studied immunization of mice followed by tumor challenge. High level of mTERT expression is detected in TC-1 tumor cells (42), therefore, an in vivo TC-1 tumor challenge study was performed to assess whether vaccination with phTERT could mediate anti-tumor immunity (Fig. 4A). The data showed that immunized mice exhibited significantly smaller tumors compared to those in the naïve group (Fig. 4B and 4D) at all days post-challenge out to day 35 (p < 0.05). Thirty-seven days post tumor challenge, the mice in naïve group either died or were euthanized because the diameters of tumors reached 20 mm. In contrast, about 70% of phTERT-immunized mice still survived 37 days post tumor implantation (Fig. 4C). These data indicated that phTERT induced potent antitumor immunity, and immunized animals exhibited delayed tumor growth and improved survival.

Figure 4.

Preventitve antitumor immunity induced by phTERT. A, Experimental design. Ten mice were immunized with 50 ug of phTERT four times biweekly. Each mouse was challenged with 5×10⁴ TC-1 cells injected subcutaneously including ten naïve mice which served as a control. Tumors were measured twice weekly. B, Delayed tumor growth in immunized mice. Tumor measurements for each time point are shown only for surviving mice. C, Vaccination with phTERT extended survival. D, Representative image of tumor size in naïve or phTERT-vaccinated group at day 34 post TC-1 implantation. E, Percent killing of a representative phTERT-immunized mouse. CFSE-labeled splenocytes were pulsed with either hTERT or HPV6 E6/E7 peptides and adoptively transferred into naïve or phTERT immunized mice through the tail vein. Forty-eight hours later, CFSE-labeled cells were recovered and analyzed by FACS to quantify percent killing. F, Average percent killing of five phTERT-immunized mice. Results were presented as mean ± SEM. The experiment shown is representative of two repeated experiments.

hTERT-specific CD8 T cells induced by vaccination eliminated target cells in vivo

While we demonstrated the up-regulation of CD107a and increased release of perforin in hTERT-specific CD8 T cells, we thought it would be important to confirm CTL activities in vivo. Therefore, an in vivo cytotoxicity assay was conducted to evaluate the ability of vaccine-induced CD8 T cells to eliminate target cells. CFSE-labeled splenocytes were pulsed with either hTERT or HPV6 E6/E7 peptides and adoptively transferred into either naïve or phTERT-immunized mice. The killing activity was evaluated by gating on CFSE-labeled splenocytes (Fig.4E). As shown in 4F, the average percent killing observed in five immunized mice was about 73%, indicating a strong antigen-specific killing of target cells. No killing of irrelevant T cells was observed. The result confirmed that the vaccine-induced CD8 T cells had killing capacity to initiate target cell death in vivo.

Vaccination with phTERT slows the tumor growth in tumor-bearing mice

Given the results obtained from the prophylactic tumor study, an in vivo tumor therapy study was performed to analyze in an initial fashion the therapeutic effect of vaccination with phTERT. We initiated the study by challenging mice with 5x10⁴ TC-1 cells on day 0. Three days after TC-1 cells implantation, 10 mice in the hTERT group were immunized with phTERT and boosted on day 10, 17 and 24 (Fig. 5A). All mice exhibited tumor growth, however, the tumors in phTERT-immunized mice were significantly smaller than those in the naive group at day 39 (p<0.05) (Fig. 5B). Six out of ten phTERT-immunized mice still survived thirty-nine days post tumor implantation while all mice in naïve group were either dead or euthanized (Fig. 5C). Therefore, vaccination with phTERT slows tumor growth and improve survival rate of tumor-bearing C57BL/6 mice.

Figure 5.

Therapeutic effect induced by phTERT in tumor-bearing C57BL/6 mice. A, Experimental design. Mice were separated into two groups of 10 mice, naïve and hTERT group. On day 0, all mice were injected subcutaneously with 5×10⁴ TC-1 cells. Starting on day 3, all mice in the hTERT group were immunized with 50 ug of phTERT at weekly intervals four times. B, Vaccination with phTERT delayed tumor growth in tumor-bearing mice. Tumor measurements for each time point are shown only for surviving mice. C, Vaccination with phTERT extended survival in tumor-bearing mice. The experiment shown is representative of two repeated experiments.

Discussion

Overexpression of hTERT has been linked to development and progression of more than 85% of cancer types in a variety of species. Peptides derived from hTERT can be processed by tumors and presented in the context of major histocompatibility complex class I molecules, thus triggering hTERT-specific T cells (43-45). While there has been a great deal of important activity, there remains a need for improved hTERT immunogens.

DNA vaccines have emerged as an attractive approach for antigen-specific immunotherapy. This technology has significant potential, compared to traditional protein and peptide vaccines, in terms of generating CTL responses (46). However, few studies have been performed to develop hTERT DNA vaccines. Previously, the DNA platform was studied in prime-boost strategies due to the low immunogenicity of hTERT DNA vaccines (47-49). Here, we developed a novel hTERT DNA vaccine using gene optimization strategies and demonstrated the vaccine delivered by EP induced strong cellular immune responses in both mice and NHP. The immune responses observed in monkeys were much higher than what has been previously described for other hTERT DNA vaccines. The T cell immune responses after just three immunizations with 2 mg of phTERT by EP in monkeys (834 SFU/10⁶ PBMCs) were already comparable to these induced by a DNA/EP prime Ad6 boost strategy (which included 5 mg dose DNA delivered with EP 5 times followed by 10¹¹ VP of Ad6 boost 2 times) (49). It is likely that the combination of better construct optimization strategies and a more potent CELLECTRA® electroporation delivery system accounts for the increased vaccine-induced responses. Recently, by using a similar combination of the approaches in gene optimization and gene delivery, we have successfully demonstrated that a novel HPV therapeutic DNA vaccine could induce robust cellular immune responses to E6/E7 of HPV16 and 18 with cytolytic functionality in women previously treated for high-grade cervical dysplasia (31). The immune responses boosted significantly further when the fourth immunization of phTERT was performed (1834 SFU/10⁶ PBMCs), suggesting enhanced antigen-specific immune responses may be obtained by multiple vaccinations. This ability to boost T cell responses without anti-vector responses or other limitations could be an important advantage in clinical studies.

Cytotoxic CD8 T cells are considered crucial components of antitumor immunity that attack tumor cells presenting tumor-associated antigen peptide with MHC class I on their surface (50, 51). As a result, one major focus in the field of cancer immunotherapy has been on the stimulation of antigen-specific CD8+ CTL responses. Research has indicated that breast cancer patients who mounted an hTERT-specific CTL exhibit significantly longer median overall survival (52). Here, we first showed vaccination of mice with the hTERT DNA vaccine significantly enhanced the numbers of CD107a, IFN-γ and TNF-α producing cells, indicating the generation of putative antigen-specific CTLs. Secondly, the vaccine-induced T cells exhibited ability to release perforin in the monkeys immunized with phTERT. Moreover, instead of performing the ⁵¹Cr-release assay, we applied an in vivo cytoxicity assay to confirm the CTL function in vivo by flow cytometry. This assay allowed us to measure cytotoxicity quantitatively by counting the loss of antigen-specific target cells. Our data demonstrated vaccination elicited cytotoxic T cell that could destroy target cells (the average percent killing observed in the immunized mice was about 73%). The clinical benefit of the vaccine-induced CTL responses needs to be further investigated.

Effective cancer therapeutic vaccines that medicate clinical responses in cancer patients may require generation of broadly targeting CTLs against multiple epitopes to limit tumor immune escape. Recently, a therapeutic vaccine for renal cell cancer consisting of multiple tumor-associated peptides was able to induce T cells associated with longer patient survival (53). Expanded diversity of the T cell responses in these trials was associated with clinical benefit, indicating that targeting multiple epitopes in immunotherapy improved clinical efficacy. Liao et al demonstrated that a peptide vaccine including four hTERT HLA-A0201-restricted CTL epitopes could elicit stronger antitumor immunity than their corresponding linear peptides (17). Therefore, multiepitope-containing DNA vaccines may represent promising tools for inducing antitumor responses in cancer patients. By delivering large numbers of CTL epitopes, DNA vaccines may avoid T-cell epitope restriction to a particular MHC haplotype and possible immunoselection of epitope loss variants (54). Here, we confirmed that vaccination with phTERT could not only induce multiple H2-D^b-restricted epitopes in mice, but also could elicit multiple dominant and subdominant epitopes in monkeys. A broad spectrum of T cell responses were induced by vaccination, implying a possible clinical benefit and suggesting potential improved performance in the clinic.

It has been reported that hTERT antigen-specific CTLs are effective in targeting human cancers in vivo (8, 23). As hTERT T cells could be considered effective immune surveillance to prevent recurrence post treatment or to limit tumor development in identified high-risk individuals, we studied immunization of mice followed by tumor challenge for their ability against tumor challenge. Previous studies have shown that xenogeneic melanoma-associated DNA vaccines can elicit effective anti-tumor immunity against murine melanoma (55, 56). Yamano et al found a human C-terminal TERT DNA vaccine induced effective anti-tumor immunity against TS/A murine breast cancer (47). Since the amino acid homology of NTE, RT and CTE regions between hTERT and murine TERT are 61.8%, 65.5% and 69.9%, respectively, the hTERT has regions of homology which may induce testable anti-tumor immunity against murine TERT-expressing tumor cells. Hence, we utilized a TC-1 tumor challenge model to evaluate the antitumor effect of hTERT-specific CTL. The results show that vaccination with phTERT conferred delayed tumor growth and longer overall survival. As this antitumor effect was potentially associated with T cell tumor infiltration, it will be interesting to perform more studies to investigate the phenotype of these tumor-infiltrating CD8 cytotoxic T cells. Based on the data we observed in the NHP study we would expect that closer matching of a mouse TERT vaccine to the native mouse sequence would further improve the effectiveness.

A major challenge with regard to hTERT and other TAA immunotherapy vaccines is to develop therapies capable of generating a robust CTL against this self-antigen in a safe manner. Although hTERT is over expressed in most tumor cells, its expression can also be detected in rare normal cells such as hematopoietic progenitor cells, spermatogonia in the testis, activated lymphocytes and certain epithelial cells (57). Consequently, the question of whether vaccine-induced hTERT-specific CTLs carries the risk of inducing autoimmune responses with pathological consequence was raised. Many studies have shown the hTERT-specific CTLs have no detectable effect on hTERT-positive CD34⁺ hematopoietic progenitor cells or activated T cells (44, 45) and do not result in autoimmune responses which target normal hTERT-expressing cells (23, 58). In addition, clinical studies in cancer patients using hTERT-based vaccines have not shown toxicity (8, 59, 60). These findings may reflect relatively low levels of hTERT expression or ineffective processing of hTERT peptides in normal cells. In the present study, several important physiological parameters were evaluated and no vaccine-induced adverse effects were detected in phTERT-immunized monkeys despite evidence of strong hTERT-specific CTLs.

Taken together, we report that administration of a synthetic highly optimized hTERT DNA vaccine in combination with adaptive constant current electroporation delivery platform was capable of breaking immune tolerance and eliciting robust and diverse CTLs in mouse and NHP models. These vaccine-induced CTLs appeared not to be associated with any major toxicities or organ damage, and were effective in mounting a potent antitumor response. These data support further study of phTERT in the setting of cancer immunotherapy, to improve tumor immune surveillance in high-risk individuals and prevention of disease recurrence.