Comparison of HPV DNA vaccines employing intracellular ... - Nature

Gene Therapy (2004) 11, 1011–1018 & 2004 Nature Publishing Group All rights reserved 0969-7128/04 $30.00 www.nature.com/gt

RESEARCH ARTICLE

Comparison of HPV DNA vaccines employing intracellular targeting strategies JW Kim1,8, C-F Hung1,8, J Juang1, L He1, T Woo Kim1, DK Armstrong2, SI Pai3, P-J Chen4, C-T Lin1,5, DA Boyd1 and T-C Wu1,2,6,7 1

Department of Pathology, The Johns Hopkins Medical Institutions, Baltimore, MD, USA; 2Department of Oncology, The Johns Hopkins Medical Institutions, Baltimore, MD, USA; 3Department of Otolaryngology-Head & Neck Surgury, The Johns Hopkins Medical Institutions, Baltimore, MD, USA; 4Department of Medicine, National Taiwan University, Taipei, Taiwan; 5Department of Obstetrics and Gynecology, Chang Gung Memorial Hospital, Taipei, Taiwan; 6Department of Obstetrics and Gynecology, The Johns Hopkins Medical Institutions, Baltimore, MD, USA; and 7Department of Molecular Microbiology and Immunology, The Johns Hopkins Medical Institutions, Baltimore, MD, USA

Intradermal vaccination via gene gun efficiently delivers DNA vaccines into dendritic cells (DCs) of the skin, resulting in the activation and priming of antigen-specific T cells in vivo. In the context of DNA vaccines, we previously used the gene gun approach to test several intracellular targeting strategies that are able to route a model antigen, such as the human papillomavirus type-16 (HPV-16) E7, to desired subcellular compartments in order to enhance antigen processing and presentation to T cells. These strategies include the use of the sorting signal of lysosome-associated membrane protein (LAMP-1), Mycobacterium tuberculosis heat-shock protein 70 (HSP70), calreticulin (CRT) and the translocation domain (dII) of Pseudomonas aeruginosa exotoxin A (ETA). Vaccination with DNA vaccines encoding E7 antigen linked to any of these molecules all led to a significant enhancement of E7-specific CD8 þ T-cell immune responses and strong antitumor effects against an E7-expressing tumor, TC-1.

However, we were interested in identifying the most potent DNA vaccine for our future clinical trials. Thus, we performed a series of experiments to directly compare the potency of the various DNA vaccines. Among the DNA vaccines we tested, we found that vaccination with pcDNA3-CRT/E7 generated the highest number of E7-specific CD8 þ T cells and potent long-term protection and treatment effects against E7-expressing tumors in mice. Interestingly, we observed that pcDNA3-CRT/E7 is also capable of protecting against an E7-expressing tumor with downregulated MHC class I expression, a common feature associated with most HPV-associated cervical cancers. Our data suggest that the DNA vaccine linking CRT to E7 (CRT/E7) may be a suitable candidate for human trials for the control of HPV infections and HPV-associated lesions. Gene Therapy (2004) 11, 1011–1018. doi:10.1038/ sj.gt.3302252; Published online 26 February 2004

Keywords: DNA vaccines; immunotherapy; human papillomavirus (HPV); E7

Introduction DNA vaccines have emerged as an attractive approach for antigen-specific immunotherapy. DNA vaccines offer many advantages over other conventional vaccines such as peptide or attenuated live pathogens. For example, DNA vaccines are relatively stable and can be easily prepared and harvested in large quantities. Additionally, naked plasmid DNA is relatively safe and can be repeatedly administered without adverse effects. Moreover, DNA is able to be maintained in cells for long-term expression of the encoded antigen; therefore, maintenance of immunologic memory is possible (for reviews, see Donnelly et al1; Pardoll and Beckerleg2; Robinson and Torves3). Correspondence: Dr T-C Wu, Department of Pathology, The Johns Hopkins University School of Medicine, Ross Research Building, Room 512 H, 720 Rutland Avenue, Baltimore, MD 21205, USA 8 Contributed equally to this paper Received 27 January 2003; accepted 24 December 2003; published online 26 February 2004

Intradermal administration of DNA vaccines via gene gun can be used to efficiently deliver genes of interest to professional antigen-presenting cells (APCs) in vivo.4 The skin contains numerous bone marrow-derived APCs (the Langerhans cells) that are able to move through the lymphatic system from the site of injection to draining lymph nodes, where they prime antigen-specific T cells.5 Gene gun immunization therefore provides the opportunity to test vaccine strategies that require direct delivery of DNA to APCs. Based on this system, we have developed many intracellular targeting strategies to improve DNA vaccine potency against the human papillomavirus type-16 (HPV-16) E7 antigen that has been associated with over 99% of all cervical cancers. For example, we have developed strategies to enhance MHC class I presentation of HPV-16 E7 antigen to E7-specific CD8 þ T cells by linking the E7 antigen to calreticulin (CRT),6 to Mycobacterium tuberculosis heat-shock protein 70 (HSP70),7 and to the translocation domain of

Comparison of HPV DNA vaccines JW Kim et al

1012

Pseudomonas aeruginosa exotoxin A (ETA(dII)).8 We have also developed an intracellular targeting strategy to improve both MHC class I and class II presentation of the antigen by linking the antigen to the sorting signal of the lysosome-associated membrane protein 1 (LAMP-1).9 In testing these strategies, we have shown that mice vaccinated with any of these DNA vaccines can augment E7-specific T-cell responses and antitumor effects when compared to mice vaccinated with E7 DNA alone. In order to translate the research to the clinical domain, it would be important to directly compare the efficacy among the various DNA vaccines to identify the best candidate for human clinical trials. In this study, we reported that mice vaccinated with pcDNA3CRT/E7 generated the highest quantity of E7-specific CD8 þ T-cell precursors as well as the highest number of E7-specific CD8 þ memory T cells, resulting in potent tumor treatment and long-term protection effects. Furthermore, the CRT/E7 DNA vaccine is capable of protecting against an E7-expressing tumor that has downregulated MHC class I expression, a common feature shared by many human cancers, including cervical cancers. Since the pcDNA3-CRT/E7 DNA vector contains an ampicillin-resistance gene and a wild-type E7 gene, it raises a concern for potential transformation of transfected cells. We therefore created a modified vaccine, pNGVL4a-CRT/E7(detox) DNA, which would be more suitable for future human clinical trials. The backbone for the pNGVL4a-CRT/E7(detox) DNA vaccine was obtained from the NIH National Gene Vector Laboratory and is identified as pNGVL-4a. This vector is a secondgeneration plasmid derived from pNGVL-3 which has been approved for human vaccination trials. Similar to its parental vector, pNGVL-4a encodes a kanamycinresistance gene and a transcription unit consisting of a CMV promoter, multiple cloning site, followed by a poly-A tail. However, novel to its parental vector, pNGVL-4a contains two short immunostimulatory DNA sequences (ISS) in the noncoding region of the backbone. The ISS sequences are derived from a bacterial ampicillin-resistance gene and consist of tandem repeats of a CpG dinucleotide in a particular base context, specifically 50 -AACGTT-30 . It has been demonstrated that an ISS containing pDNA can elicit the production of IFN-g and IL-12 in transfected keratinocytes and dermal APCs, which results in a potent T helper cell type 1 response.10 This vector also contains a transcription unit that yields high levels of antigen expression and its DNA adjuvant unit elicits strong immunologic Th1 type responses against the pDNA-encoded protein. Thus, the pNGVL-4a vector is a desirable DNA vector for human clinical trials. For the E7 oncogenic protein, we have substituted the amino acids at positions 24 (cysteine to glycine) and 26 (glutamic acid to glycine) of E7, which has been shown to inhibit protein binding to pRb 11 and would abate E7’s ability to transform cells. Our data indicated that mice vaccinated with pNGVL4a-CRT/E7(detox), either through intramuscular injection or gene gun administration, were capable of generating strong E7-specific CD8 þ T-cell immune responses. The potential usage of pNGVL4a-CRT/E7(detox) DNA for human clinical trials is discussed.

Gene Therapy

Results pcDNA3-CRT/E7 DNA vaccine generates the highest number of E7-specific CD8 þ T cells in vaccinated mice Several lines of evidence suggest that CD8 þ T-cell-mediated immune responses are important in controlling both HPV infections and HPV-associated neoplasms. To assess the quantity of E7-specific CD8 þ T-cell precursors generated by pcDNA3, pcDNA3-E7, pcDNA3-CRT/E7, pcDNA3-E7/HSP70, pcDNA3-ETA(dII)/E7, and pcDNA3-Sig/E7/LAMP-1 vaccine constructs, we performed intracellular cytokine staining with flow cytometric analysis using splenocytes derived from vaccinated mice 1 week after the last vaccination. As shown in Figure 1a and b, mice vaccinated with pcDNA3-CRT/E7 DNA exhibited the highest number of E7-specific IFN-g þ CD8 þ T-cell precursors (6557155/3 105 splenocytes) compared to mice vaccinated with pcDNA3-E7/HSP70, pcDNA3ETA(dII)/E7, pcDNA3-Sig/E7/LAMP-1, pcDNA3-E7, or pcDNA3 (Po0.05).

Figure 1 Flow cytometry analysis of IFN-g-secreting E7-specific CD8 þ T-cell precursors in mice vaccinated with various recombinant DNA vaccines. Mice (four per group) were immunized with pcDNA3-CRT/E7, pcDNA3-E7/HSP70, pcDNA3-ETA/E7, pcDNA3-Sig/E7/LAMP-1, pcDNA3E7, and pcDNA3 as described in the Materials and methods section. Splenocytes from vaccinated mice were harvested 7 days after a booster vaccination, cultured in vitro with MHC class I-restricted E7 (aa 49–57) peptide overnight, and stained for both CD8 and intracellular IFN-g. (a) Representative figure of the flow cytometry data. (b) Bar graph depicting the number of antigen-specific IFN-g-secreting CD8 þ T-cell precursors/ 3 105 splenocytes (mean7s.d.). The data presented in this figure are from one representative experiment of two performed.


1013

Mice treated with pcDNA3-CRT/E7 vaccine generate potent antitumor response We then investigated the therapeutic potential of the various chimeric DNA constructs in treating an E7expressing tumor, TC-1, using a previously described lung hematogenous spread model.9.As shown in Figure 2, mice receiving the pcDNA3-CRT/E7 vaccine exhibited significantly lower numbers of pulmonary nodules compared to mice vaccinated with pcDNA3 or pcDNA3-E7 after TC-1 challenge (Po0.05). When comparing pcDNA3-CRT/E7 to pcDNA3-E7/HSP70, pcDNA3-ETA(dII)/E7, or pcDNA3-Sig/E7/LAMP-1 vaccines, pcDNA3-CRT/E7-immunized mice displayed lower mean numbers of pulmonary nodules than the others (Po0.95). Vaccination with pcDNA3-CRT/E7 generates the highest number of E7-specific CD8 þ memory T cells and provides long-term tumor protection in vaccinated mice To assess the quantity of E7-specific CD8 þ memory T-cell precursors generated by pcDNA3, pcDNA3-E7, pcDNA3-CRT/E7, pcDNA3-E7/HSP70, pcDNA3-ETA (dII)/E7, and pcDNA3-Sig/E7/LAMP-1 DNA constructs, we performed intracellular cytokine staining with flow cytometric analysis using splenocytes derived from vaccinated mice 8 weeks after the initial vaccination. As shown in Figure 3, mice vaccinated with pcDNA3-CRT/E7 DNA exhibited the highest number of E7-specific IFN-g þ CD8 þ memory T-cell precursors compared to mice vaccinated with the other DNA constructs (Po0.05). To determine whether the pcDNA3-CRT/E7 vaccine can also provide the best long-term tumor protection against E7-expressing tumors in vaccinated mice among all the vaccines, we performed a long-term tumor protection experiment comparing the following vaccines: pcDNA3, pcDNA3-E7, pcDNA3-CRT/E7, pcDNA3-E7/ HSP70, pcDNA3-VP22/E7, pcDNA3-ETA(dII)/E7, and

Figure 3 Flow cytometry analysis of IFN-g-secreting E7-specific CD8 þ T-cell precursors generated by various DNA vaccine constructs in vaccinated mice 8 weeks after initial vaccination. Mice (four per group) were immunized with the various DNA constructs as described in Figure 1. Splenocytes from vaccinated mice were harvested 8 weeks after the initial vaccination, cultured in vitro with MHC class I-restricted E7 (aa 49-57) peptide overnight, and stained for both CD8 and intracellular IFN-g. (a) Representative figure of the flow cytometry data. (b) Bar graph depicting the number of antigen-specific IFN-g-secreting CD8 þ T-cell precursors/ 3 105 splenocytes (mean7s.d.). The data presented in this figure are from one representative experiment of two performed.

Figure 4 Long-term in vivo tumor protection experiments to compare the antitumor effect generated by various DNA vaccine constructs in vaccinated mice 8 weeks after initial vaccination. Mice (five per group) were immunized and challenged with 1 105 cells/mouse TC-1 tumor cells as described in the Materials and methods. Data are expressed as the mean number of lung nodules; bars,7s.d. The data presented in this figure are from one representative experiment of two performed. Figure 2 In vivo tumor treatment experiments to compare the antitumor effect generated by various DNA vaccine constructs in mice. Mice (five per group) were challenged with 1 104 TC-1 tumor cells and immunized with various DNA constructs seven days later. Data are expressed as the mean number of lung nodules; bars,7s.d. The data presented in this figure are from one representative experiment of two performed.

pcDNA3-Sig/E7/LAMP-1 DNA. As shown in Figure 4, mice vaccinated with the pcDNA3-CRT/E7 DNA exhibited significantly lower numbers of pulmonary nodules compared to mice vaccinated with pcDNA3-Sig/E7/ Gene Therapy


1014

LAMP-1, pcDNA3-E7, or pcDNA3 (Po0.05). Although not statistically significant, the pcDNA3-CRT/E7vaccinated mice also displayed lower mean numbers of pulmonary nodules than mice vaccinated with pcDNA3-E7/HSP70, pcDNA3-VP22/E7, or pcDNA3ETA(dII)/E7 (Po0.70).

Vaccination with CRT/E7 DNA controls E7-expressing tumors with downregulated MHC class I expression Since most cervical cancers are associated with downregulated MHC class I expression, it is important to determine the effectiveness of the CRT/E7 DNA vaccines in protecting against tumors with low MHC class I expression. Thus, we developed an E7-expressing murine tumor model with downregulated MHC class I expression, TC-1 P3 (A15). To assay for MHC class I expression in TC-1 P0 and TC-1 P3 (A15) tumor cells, cells were harvested and prepared for flow cytometric analysis. As shown in Figure 5a, TC-1 P3 (A15) exhibited distinct downregulation of MHC class I expression relative to TC-1. We then performed a tumor protection experiment in C57BL/6 mice whereby mice were vaccinated with 2 mg of CRT/E7 DNA, CRT DNA, E7 DNA, or vector alone, followed by challenge 1 week later with 5 104 TC-1 P3 (A15) tumor cells. As shown in Figure 5b, mice vaccinated with pcDNA3-CRT/E7 DNA demonstrated 100% protection against tumor challenge with TC-1 P3 (A15) up to 45 days after tumor challenge. In contrast, all mice vaccinated with pcDNA3-CRT, pcDNA3-E7, or pcDNA3 vector alone all developed tumors within 14 days. We have previously shown that IFN-g is essential to the antitumor effect generated by DNA vaccines employing M. tuberculosis HSP70 linked to E7 against an E7expressing tumor cell line with downregulated MHC class I expression.12 To determine if IFN-g is essential for the antitumor effect generated by vaccination with CRT/ E7 DNA, we vaccinated wild-type and IFN-g KO C57BL/6 mice followed by challenge with TC-1 P3 (A15). As shown in Figure 5c, while 100% of wildtype mice vaccinated with pcDNA3-CRT/E7 were protected against challenge with TC-1 P3 (A15), only 20% of IFN-g KO mice vaccinated with pcDNA3-CRT/E7

Figure 5 MHC class I expression of TC-1 P3 (A15) and in vivo tumor protection experiment using TC-1 P3 (A15) tumor cells. (a) Flow cytometry analysis was performed to characterize MHC class I expression on TC-1 P0 and the TC-1 P3 (A15) subclone. B16 was used as a negative control (dotted line). TC-1 P0 cells are MHC class I positive (thick line), while TC-1 P3 (A15) exhibits downregulated MHC class I expression (filled region). (b) In vivo tumor protection experiments using TC-1 P3 (A15) tumor cells. Mice (five per group) were vaccinated with 2 mg of pcDNA3-E7, pcDNA3-CRT, pcDNA3-CRT/E7 DNA, or pcDNA3 empty plasmid. At 1 week after the last vaccination, mice were challenged with 5 104 TC-1 P3 (A15) tumor cells by subcutaneous injection in the right leg using a previously described model.25 Mice vaccinated with CRT/E7 DNA provided 100% protection against TC-1 P3 (A15) when compared to mice vaccinated with other DNA vaccines (one-way ANOVA, Po 0.01). (c) In vivo tumor protection experiments using IFN-g KO mice. Wild-type C57BL/6 mice and IFN-g-depleted C57BL/6 mice (five per group) were vaccinated with 2 mg of pcDNA3-CRT/E7 DNA. At 1 week after the last vaccination, mice were challenged with 5 104 TC-1 P3 (A15) tumor cells by subcutaneous injection in the right leg as described above. Note: While 100% of wild-type C57BL/6 mice remained free of tumors, only 20% of IFN-g-depleted C57BL/6 mice remained free of tumors. Gene Therapy

were protected against TC-1 P3 (A15) challenge. These results indicate that the CRT/E7 DNA is capable of controlling E7-expressing tumors with downregulated MHC class I expression and that IFN-g is essential for the antitumor effect.


Intramuscular and gene gun-mediated immunization of pNGVL4a-CRT/E7(detox) DNA are capable of generating enhanced E7-specific CD8 þ T-cell immune responses Since the pcDNA3-CRT/E7 DNA vector contains an ampicillin-resistance gene and the wild-type E7 has concerns for transformation, the pcDNA3-CRT/E7 DNA vaccine is not suitable for human clinical trials. We therefore created pNGVL4a-CRT/E7(detox) DNA for future clinical trials. Since we plan to perform our HPV vaccine trials through intramuscular immunization, it is important to demonstrate that vaccination with pNGVL4a-CRT/E7(detox) intramuscularly can also enhance E7-specific CD8 þ T-cell immune responses. As shown in Figure 6a and b, intramuscular and gene gun immunization of mice with pNGVL4a-CRT/E7(detox) DNA significantly increased the number of E7-specific CD8 þ T-cell precursors compared to vaccination with pNGVL4a-E7(detox) or pNGVL4a vector alone (Po0.05). We also observed that gene gun-mediated immunization of pNGVL4a-CRT/E7(detox) generated more E7specific CD8 þ T-cell precursors than intramascular (i.m.) immunization of pNGVL4a-CRT/E7(detox) (Po0.05).

Figure 6 Intracellular cytokine staining with flow cytometry analysis to demonstrate the number of E7-specific CD8 þ T-cell precursors in mice vaccinated with pNGVL4a-CRT/E7(detox). One group of C57BL/6 mice were immunized intramuscularly with 50 mg of the various DNA vaccines and received a booster with the same regimen one week later. Another group of C57BL/6 mice were immunized intradermally via gene gun with 2 mg of the various DNA vaccines and received a booster with the same regimen 1 week later. Splenocytes were collected one week after the last vaccination. The number of E7-specific IFN-g-secreting CD8 þ T-cell precursors was analyzed using intracellular cytokine staining followed by flow cytometry as described previously.6 (a) Representative figure of the flow cytometry data. (b) The data presented in this figure are from one representative experiment of two performed. Data are expressed as number of interferon-secreting T cells per 3 105 splenocytes.

Furthermore, in the tumor protection experiments using E7-expressing tumors with downregulated MHC class I expression (TC-1 P3 (A15)), our results indicated that mice vaccinated with pNGVL4a-CRT/E7(detox) DNA demonstrated 100% protection against TC-1 P3 (A15) up to 45 days after tumor challenge (data not shown). These results indicate that the pNGVL4a-CRT/E7(detox) DNA, like the pcDNA3-CRT/E7 DNA, is capable of generating marked increase in the number of E7-specific CD8 þ T cells and is capable of controlling E7-expressing tumors with downregulated MHC class I expression. Thus, the pNGVL4a-CRT/E7(detox) DNA may be a suitable candidate vaccine to advance into human trials for control of HPV infections and HPV-associated lesions.

1015

Discussion In this study, we compared the E7-specific CD8 þ T-cell immune responses and antitumor effects generated by five effective DNA vaccines (pcDNA3-CRT/E7, pcDNA3-E7/HSP70, pcDNA3-ETA(dII)/E7, pcDNA3Sig/E7/LAMP-1) previously developed in our lab to determine the most potent DNA vaccine as a candidate for use in future human clinical trials. We found that mice vaccinated with pcDNA3-CRT/E7 generated the highest quantity of E7-specific CD8 þ T-cell precursors as well as the highest number of E7-specific CD8 þ memory T cells, resulting in potent tumor treatment and longterm protection effects. We also observed that the CRT/ E7 DNA vaccine is capable of controlling tumors with downregulated MHC class I expression, suggesting that the vaccine may also be useful for treating patients with cervical lesions with downregulated MHC class I expression that normally would facilitate tumor immune evasion. We have also created and tested pNGVL4aCRT/E7(detox) DNA vaccine for future clinical trials. Our results indicated that the pNGVL4a-CRT/E7(detox) DNA vaccine, administrated via gene gun or intramuscularly, was able to generate an enhancement of E7specific CD8 þ T-cell immune responses in vaccinated mice. Thus, we believe that these data provide a strong rationale for carrying this DNA vaccine into a clinical setting. Our data indicated that CRT/E7 DNA is capable of controlling tumors with downregulated MHC class I expression. One concern related to the efficacy of immunotherapy is that tumors can evade immune responses through various mechanisms, including MHC class I downregulation.13,14 A number of human cancers have been shown to downregulate MHC class I expression, including melanoma,15 lung cancer,16 prostate cancer,17 breast cancer,18,19 ovarian cancer,20 colon cancer,20 and cervical cancer.21–23 By downregulating MHC class I expression, tumor cells escape immune recognition; thus avoid killing by antigen-specific CD8 þ T cells. In this study, we demonstrated the effectiveness of the CRT/E7 DNA vaccine against tumors with low MHC class I expression, suggesting that the CRT/E7 vaccine may be useful for treating patients with advanced cervical cancer with downregulated MHC class I expression. Our previous study determined that a DNA vaccine encoding M. tuberculosis HSP70 linked to E7 was able to generate antitumor effects against the TC-1 P3 (A15) Gene Therapy


1016

tumor cell line.12 In that study, we showed that IFN-g was required for tumor protection and that IFN-g could upregulate MHC class I expression on TC-1 P3 (A15) tumor cells to levels equivalent to P0 cells. This indicated that IFN-g secreted by CTLs generated as a result of vaccination with HSP70/E7 DNA may protect against TC-1 P3 (A15) challenge by upregulating MHC class I expression of tumor cells. In this study, we found that our CRT/E7 DNA vaccine provides protection against a tumor cell line with downregulated MHC class I expression. In order to elucidate the role of IFN-g in this protection, we studied the ability of vaccination with CRT/E7 DNA to protect against TC-1 P3 (A15) tumor challenge in IFN-g KO C57BL/6 mice and wild-type C57BL/6 mice. We found that IFN-g was essential to the observed antitumor response. Thus, the tumor protection generated by CRT/E7 DNA against TC-1 P3 (A15) tumor occurs via an IFN-g-dependent mechanism. We observed that both gene gun and i.m. administration of the pNGVL4a-CRT/E7(detox) DNA were capable of generating enhanced E7-specific CD8 þ T-cell immune responses in vaccinated mice when compared to mice vaccinated with pNGVL4a-E7(detox) and pNGVL4a vector alone, suggesting their potential to elicit E7specific CD8 þ T-cell immune responses in vaccinated humans. Our data indicated that gene gun-mediated vaccination of pNGVL4a-CRT/E7(detox) generated more E7-specific CD8 þ T cells than i.m. administration of pNGVL4a-CRT/E7(detox) in vaccinated mice. The route of administration likely plays an important role in the observed differences mediated by DNA vaccines encoding CRT/E7. Intradermal immunization via gene gun directly targets antigen to professional APCs, Langerhans cells,4,5 allowing the intracellular strategy to further improve direct presentation of antigen to T cells by DNAtransfected DCs. In comparison, intramuscular immunization likely targets antigen to myocytes, and the antigen encoded by DNA vaccine is eventually presented through bone marrow-derived APCs through the cross-priming mechanism. Vaccination with pNGVL4aCRT/E7(detox) likely led to secretion of chimeric CRT/ E7(detox) protein or lysis of cells expressing CRT/ E7(detox) antigen, releasing the chimeric protein exogenously to be taken up and processed by other APCs via the MHC class I-restricted pathway. The linkage of CRT to E7 may facilitate the cross-priming of E7 antigen. One recent study found that CD91, an a2 macroglobulin receptor commonly expressed on professional APCs, serves as a receptor for HSPs, including CRT, and facilitates cross-priming.24 Intramuscular immunization of DNA vaccines encoding CRT/E7(detox) may allow for prolonged release of CRT/E7(detox) protein from transfected cells to target, concentrate CRT/E7(detox) to professional APCs, and facilitate the cross-priming of E7 antigen. Therefore, different routes of administration may generate different degrees of immune response elicited by the same vaccine. In summary, our studies indicated that CRT/E7 DNA vaccine may be an attractive therapeutic vaccine not only because of its ability to generate an effective antitumor response against E7-expressing tumors and its ability to control E7-expressing tumors with downregulated MHC class I expression in vaccinated mice, but also because of its ability to prevent or delay tumor growth by targeting

Gene Therapy

tumor vasculature. The application of the CRT/E7 DNA is not limited to treating patients with E7-expressing cervical cancers. The vaccine may also provide general antiangiogenesis benefits to other cancer patients through the CRT molecule encoded by the vaccine.

Materials and methods Plasmid DNA construction The generation of pcDNA3, pcDNA3-E7, pcDNA3-CRT/ E7,6 pcDNA3-E7/HSP70,7 and pcDNA3-ETA(dII)/E78 has been described previously. To generate pcDNA3Sig/E7/LAMP-1, Sig/E7/LAMP-1 was cut at the EcoRI/ BamHI sites from pCMV(neo)-Sig/E7/LAMP-19 and cloned into pcDNA3. For generation of pNGVL4a-E7(detox), the E7 gene was cloned into pNGVL4a (National Gene Vector Laboratory) using the EcoRI and KpnI restriction sites. Using site-directed mutagenesis, two point mutations, which had previously been found to reduce Rb binding,11 were introduced into the E7 gene. The primers used to introduce these mutations were as follows: E7(detox) forward: 50 -ctgatctctacggttatgggcaattaaatga cagctc 30 and E7(detox) reverse: 50 gagctgtcatttaattgc ccataaccgtagagatca 30 . For generation of pNGVL4aCRT/E7(detox), CRT was PCR amplified by primers (50 aaagtcgacatgctgctatccgtgccgctgc 30 and 50 -gaattcgttgtctggccgcacaatca 30 ) using a human CRT plasmid as a template (which was kindly provided by Dr David Llewellyn of Department of Medical Biochemistry at University of Wales College of Medicine at Cardiff, UK). The PCR product was cut with SalI/EcoRI and cloned into the SalI/EcoRI sites of pNGVL4a-E7(detox). The accuracy of DNA constructs was confirmed by DNA sequencing. Mice Female C57BL/6 mice (6–8 weeks old) were purchased from the National Cancer Institute (Frederick, MD, USA) and kept in the oncology animal facility of the Johns Hopkins Hospital (Baltimore, MD, USA). IFN-g knockout mice were obtained from the Jackson Laboratory (Bar Harbor, ME, USA). All animal procedures were performed according to approved protocols and in accordance with the recommendations for the proper use and care of laboratory animals. Generation of TC-1 and TC-1 P3 (A15) tumor cell line The production and maintenance of TC-1 cells has been described previously.25.For the generation of TC-1 P3 (A15), Vac-Sig/E7/LAMP-1-vaccinated mice were challenged with TC-1 tumor cells. Vaccination with Vac-Sig/ E7/LAMP-1 elicits E7-specific antitumor responses against HPV-16 E7-expressing tumors (TC-1), although the vaccine fails to prevent tumor formation in approximately 20% of the vaccinated mice.25 The outgrown TC-1 tumors from Vac-Sig/E7/LAMP-1-vaccinated mice were explanted, cut into pieces of less than 1 mm in diameter, digested with collagenase at a concentration of 1 mg/ml in DMEM (GIBCO BRL, Rockville, MD, USA), and expanded in vitro. These expanded cell lines were called TC-1 P1. Vac-Sig/E7/LAMP-1 vaccinated mice were then challenged with TC-1 P1 tumor cells. Approximately 40% of vaccinated mice developed tumors (data


not shown). The outgrown tumors from these vaccinated mice were then explanted and expanded in vitro to create the TC-1 P2 tumor cell line. Vac-Sig/E7/LAMP-1vaccinated mice were then challenged with TC-1 P2 tumor cells. This time approximately 60-80% of vaccinated mice developed tumors (data not shown). The outgrown tumors from these vaccinated mice were further explanted and expanded in vitro to generate the TC-1 P3 tumor cell line. TC-1 P3 clones (50) were generated by limiting dilution. Among the TC-1 P3 clones, a representative clone with marked downregulation of MHC class I expression was isolated and expanded, creating the TC-1 P3 (A15) tumor cell line. We have determined that more than 90% of the TC-1 P3 (A15) cells exhibited downregulated MHC class I expression. Less than 10% of the TC-1 P3 (A15) cells expressed any MHC class I molecules. Anti-H-2Kb/H2Db (clone 28-8-6) (BD Bioscience, San Diego, CA, USA) monoclonal antibody was used to detect MHC class I expression. Both TC-1 and TC-1 P3 (A15) cells were grown in RPMI 1640, supplemented with 10% (vol/vol) fetal bovine serum, 50 Un/ml penicillin/streptomycin, 2 mM L-glutamine, 1 mM sodium pyruvate, 2 mM nonessential amino acids, and 0.4 mg/ml G418 at 371C with 5% CO2. On the day of tumor challenge, tumor cells were harvested by trypsinization, washed twice with 1X Hank’s buffered salt solution (HBSS), and finally resuspended in 1 HBSS to the designated concentration for injection.

DNA vaccination For the gene gun-mediated intradermal vaccination, DNA-coated gold particles (1 mg DNA/bullet) were delivered to the shaved abdominal region of C57BL/6 mice using a helium-driven gene gun (BioRad, Hercules, CA, USA) with a discharge pressure of 400 p.s.i., according to a previously described protocol.7 C57BL/6 mice were vaccinated via gene gun with either 2 mg of pcDNA3, pcDNA3-E7, pcDNA3-CRT/E7, pcDNA3-E7/ HSP70, pcDNA3-ETA(dII)/E7, pcDNA3-Sig/E7/LAMP1, pNGVL4a, pNGVL4a- E7(detox), or pNGVL4a-CRT/ E7(detox). These mice received a booster with the same regimen 1 week later. For the i.m.-mediated DNA vaccination, 50 mg/mouse of pNGVL4a, pNGVL4a- E7(detox), and pNGVL4aCRT/E7(detox) DNA vaccines were delivered intramuscularly by syringe needle injection. These mice received a booster with the same regimen 1 week later. Intracellular cytokine staining and flow cytometry analysis Cell surface marker staining of CD8 and intracellular cytokine staining for IFN-g as well as FACScan analysis was performed using conditions described previously.7 Prior to FACScan, splenocytes from different vaccinated groups of mice were collected and incubated for 20 h with 1 mg/ml of E7 peptide (aa 49–57, RAHYNIVTF)26 containing an MHC class I epitope for detecting E7specific CD8 þ T-cell precursors. Golgistop (Brefeldin A) was added 6 h before harvesting the cells from the culture. Triplicate experiments were performed using a pool of spleen cells from the group of mice vaccinated with the same vaccine construct. The number of IFNg-secreting CD8 þ T cells was analyzed using flow

cytometry. Analysis was performed on a BectonDickinson FACScan with CELLQuest software (BectonDickinson Immunocytometry System, Mountain View, CA, USA).

1017

In vivo tumor treatment experiment using TC-1 tumors For in vivo tumor treatment experiments using an E7expressing tumor (TC-1), mice (five per group) were intravenously challenged through the tail vein with 1 104 TC-1 cells/mouse. At seven days after tumor challenge, mice were administered 2 mg of pcDNA3, pcDNA3-E7, pcDNA3-CRT/E7, pcDNA3-E7/HSP70, pcDNA3-ETA(dII)/E7, or pcDNA3-Sig/E7/LAMP-1 DNA via gene gun. At 1 week after the first vaccination, these mice were boosted with the same regimen. Mice were killed and lungs were explanted on day 28. The pulmonary nodules on the surface of the lungs in each mouse were counted by experimenters blinded to sample identity as described previously.9 (8-week) long-term in vivo tumor protection experiment For the 8-week long-term tumor protection experiment, mice (five per group) were vaccinated via gene gun with 2 mg of pcDNA3, pcDNA3-E7, pcDNA3-CRT/E7, pcDNA3-E7/HSP70, pcDNA3-ETA(dII)/E7, or pcDNA3Sig/E7/LAMP-1 DNA. After 1 week, mice were boosted with the same regimen as the first vaccination. At 8 weeks after the initial vaccination, mice were intravenously (i.v.) challenged with 1 105 TC-1 cells/mouse via tail vein. Mice were killed 28 days after the tumor challenge and lung surface pulmonary nodules in each mouse were counted by experimenters blinded to sample identity. In vivo tumor protection against TC-1 P3 (A15) MHC class I down-regulated tumors C57BL/6 mice (five per group) were vaccinated with pcDNA3, pcDNA3-E7, pcDNA3-CRT, pcDNA3-CRT/E7, pNGVL4a, pNGVL4a- E7(detox), or pNGVL4a-CRT/ E7(detox) DNA vaccines mediated by gene gun injection as described above. At 1 week after the last vaccination, mice were challenged with 5 104 TC-1 P3 (A15) tumor cells by subcutaneous injection in the right leg. Tumor growth was monitored by visual inspection and palpation twice weekly as described previously.25 For the experiment performed with wild-type C57BL/ 6 mice and IFN-g KO C57BL/6 mice, mice (five per group) were vaccinated via gene gun with 2 mg of pcDNA3-CRT/E7 DNA and then boosted with the same regimen 1 week later. At 1 week after the last vaccination, mice were challenged with 5 104 TC-1 P3 (A15) tumor cells by subcutaneous injection in the right leg. Tumor growth was monitored by visual inspection and palpation twice weekly as described previously.25 Statistical analysis All data expressed as means7s.d. are representative of at least two different experiments. Data for intracellular cytokine staining with flow cytometry analysis and tumor treatment experiments were analyzed by analysis of variance (ANOVA). Comparisons between individual data points were made using a Student’s t-test. Kaplan– Meier survival curves for tumor protection experiments were applied; for differences between curves, P’values Gene Therapy


1018

were calculated using the log-rank test. The value of Po0.05 was considered significant.

Acknowledgements We thank Drs Robert J Kurman, Keerti V Shah and Drew M Pardoll for helpful discussions. We also thank Drs Ralph Hruban, Ken-Yu Lin, and Richard Roden for critical review of the manuscript. This work was supported by the National Cancer Institute, the American Cancer Society, and Genencor International Inc. T-C Wu received consultation fee, stock option, and sponsored research support from Genecor International Inc. Under separate licensing agreements between Genencor International and the Johns Hopkins University and Cerus Corporation and the Johns Hopkins University, Dr Wu is entitled to a share of royalty received by the University on sales of products described in this article. Dr Wu is a paid consultant to Genencor International and Cerus Corporation. The terms of this arrangement are being managed by the Johns Hopkins University in accordance with its conflict of interest policies.

References 1 Donnelly JJ, Ulmer JB, Shiver JW, Liu MA. DNA vaccines. Annu Rev Immunol 1997; 15: 617–648. 2 Pardoll DM, Beckerleg AM. Exposing the immunology of naked DNA vaccines. Immunity 1995; 3: 165–169. 3 Robinson HL, Torres CA. DNA vaccines. Semin Immunol 1997; 9: 271–283. 4 Condon C et al. DNA-based immunization by in vivo transfection of dendritic cells. Nat Med 1996; 2: 1122–1128. 5 Porgador A et al. Predominant role for directly transfected dendritic cells in antigen presentation to CD8+ T cells after gene gun immunization. J Exp Med 1998; 188: 1075–1082. 6 Cheng WF et al. Tumor-specific immunity and antiangiogenesis generated by a DNA vaccine encoding calreticulin linked to a tumor antigen. J Clin Invest 2001; 108: 669–678. 7 Chen CH et al. Enhancement of DNA vaccine potency by linkage of antigen gene to an HSP70 gene. Cancer Res 2000; 60: 1035–1042. 8 Hung CF et al. Cancer immunotherapy using a DNA vaccine encoding the translocation domain of a bacterial toxin linked to a tumor antigen. Cancer Res 2001; 61: 3698–3703. 9 Ji H et al. Targeting human papillomavirus type 16 E7 to the endosomal/lysosomal compartment enhances the antitumor immunity of DNA vaccines against murine human papillomavirus type 16 E7-expressing tumors. Hum Gene Ther 1999; 10: 2727–2740.

Gene Therapy

10 Sato Y et al. Immunostimulatory DNA sequences necessary for effective intradermal gene immunization. Science 1996; 273: 352–354. 11 Munger K et al. Complex formation of human papillomavirus E7 proteins with the retinoblastoma tumor suppressor gene product. EMBO J 1989; 8: 4099–4105. 12 Cheng WF et al. CD8+ T cells, NK cells and IFN-gamma are important for control of tumor with downregulated MHC class I expression by DNA vaccination. Gene Therapy 2003; 10: 1311–1320. 13 Restifo NP et al. Identification of human cancers deficient in antigen processing. J Exp Med 1993; 177: 265–272. 14 Restifo NP et al. Molecular mechanisms used by tumors to escape immune recognition: immunogenetherapy and the cell biology of major histocompatibility complex class I. J Immunother 1993; 14: 182–190. 15 Ferrone S, Marincola FM. Loss of HLA class I antigens by melanoma cells: molecular mechanisms, functional significance and clinical relevance. Immunol Today 1995; 16: 487–494. 16 Korkolopoulou P et al. Loss of antigen-presenting molecules (MHC class I and TAP-1) in lung cancer. Br J Cancer 1996; 73: 148–153. 17 Sanda MG et al. Molecular characterization of defective antigen processing in human prostate cancer. J Natl Cancer Inst 1995; 87: 280–285. 18 Cabrera T et al. High frequency of altered HLA class I phenotypes in invasive breast carcinomas. Hum Immunol 1996; 50: 127–134. 19 Vitale M et al. HLA class I antigen and transporter associated with antigen processing (TAP1 and TAP2) down-regulation in high-grade primary breast carcinoma lesions. Cancer Res 1998; 58: 737–742. 20 Vegh Z, Wang P, Vanky F, Klein E. Selectively down-regulated expression of major histocompatibility complex class I alleles in human solid tumors. Cancer Res 1993; 53: 2416–2420. 21 Connor ME, Stern PL. Loss of MHC class-I expression in cervical carcinomas. Int J Cancer 1990; 46: 1029–1034. 22 Cromme FV et al. MHC class I expression in HPV 16 positive cervical carcinomas is post-transcriptionally controlled and independent from c-myc overexpression. Oncogene 1993; 8: 2969–2975. 23 Koopman LA et al. Multiple genetic alterations cause frequent and heterogeneous human histocompatibility leukocyte antigen class I loss in cervical cancer. J Exp Med 2000; 191: 961–976. 24 Basu S, Binder RJ, Ramalingam T, Srivastava PK. CD91 is a common receptor for heat shock proteins gp96, hsp90, hsp70, and calreticulin. Immunity 2001; 14: 303–313. 25 Lin KY et al. Treatment of established tumors with a novel vaccine that enhances major histocompatibility class II presentation of tumor antigen. Cancer Res 1996; 56: 21–26. 26 Feltkamp MC et al. Vaccination with cytotoxic T lymphocyte epitope-containing peptide protects against a tumor induced by human papillomavirus type 16-transformed cells. Eur J Immunol 1993; 23: 2242–2249.