Plasmid DNA vaccines represent a relatively new approach to immunization against antigens present on cancer cells. DNA vaccines are relatively simple to prepare and administer and present multiple epitopes within the complete coding sequence of an antigen. Immunostimulatory sequences naturally occurring in bacterial DNA may add to the ability of DNA vaccines to induce both antibody and T-cell responses. The results of experiments using mouse melanoma models have suggested that DNA immunization is a promising new approach to cancer vaccine therapy.
A. Cancer Vaccines: Introduction
A marked evolution has taken place in the field of cancer vaccines since the 1970s. A brief history of the field reveals that the earliest attempts to vaccinate against cancer were based on seminal observations of tumor rejection in inbred strains of mice by Prehn, Main, and others. These early models demonstrated that mice bearing a particular type of chemically induced tumor, which was subsequently surgically excised, could reject the same type of tumor when injected at a later time. However, injection of a different type of tumor was not observed, suggesting a tumor-specific immune response. It is then quite understandable that interest developed in the possibility of creating cancer vaccines for human clinical use, especially given the remarkable safety and efficacy of vaccination for infectious diseases.
B. Melanoma as a Model for Cancer Vaccine Investigation
The majority of clinical trials for cancer vaccines have involved melanoma, based on several lines of circumstantial evidence implicating the immune system in the biology of the disease. The natural history of melanoma in individual patients can be unpredictable, with some patients surviving many years with relatively stable metastatic disease. Melanoma is one of the few cancers in which rare, but well-documented, instances of spontaneous regression occur.
Serum and T cells from individual melanoma patients have been intensively studied and found to contain antibodies and CD4+ and CD8+ T lymphocytes that recognize melanoma cells, and numerous antigens capable of immune recognition have been identified on melanoma cells (summarized in Table I). Also, the most active current therapies for metastatic melanoma utilize the immunologic agents interferon- α and interleukin-2. Although vaccines are also being investigated for many other malignancies, including breast, prostate, colorectal, and lung cancers, this article focuses on melanoma, as most of preclinical as well as clinical trial data are derived from studies of this disease.
TABLE I Types of AntigensCateogry Examples Mutated proteins CDK4, β-catenin Differentiation antigens Tyrosinase family, MART-1, gp100, gangliosides Cancer-testis antigens MAGE, BAGE, NY-ESO-1
C. Cellular Vaccines
Initial clinical trials have involved whole cell, cell lysate, or shed antigen vaccines from either autologous or allogeneic melanomas. Many preliminary trials have been completed with such vaccinees, and current phase III trials are underway for the allogeneic CancerVax and Melacine, as well as the NYU polyvalent shed antigen vaccine. These vaccines have the advantage of containing many potential antigenic targets. Disadvantages include an inability to specifically quantify response, as the exact molecular components of such vaccines are unknown. In addition, immunity to irrelevant allogeneic antigens may be induced.
D. Ganglioside Vaccines
The identification of potential antigenic targets was facilitated initially by the intensive study of sera from melanoma patients. Commonly identified antigens from this serologic screening program included the gangliosides. One of the earliest trials of a monoclonal antibody for cancer utilized R24, a mouse monoclonal antibody (mAb), which recognizes GD3 ganglioside. Clinical trials using R24 and other mAbs demonstrated a reproducible response rate of 10% in patients with advanced metastatic melanoma, confirming the importance of gangliosides in immunotherapy of melanoma and the ability of mAbs to shrink bulky tumors. The logical next step was to develop vaccines for active immunization against gangliosides. One such vaccine is a conjugate of the GM2 ganglioside with keyhole limpet hemocyanin as a carrier (GMK vaccine), which has been studied in a phase III trial as an adjuvant therapy for patients at high risk for recurrence of a surgically resected melanoma.
E. Peptide Vaccines
Advances in molecular biology since the 1980s have allowed for an unprecedented increase in understanding the events involved in the generation of an immune response. By application of such techniques to individual patient samples, several groups have identified peptides derived from antigens recognized by T cells from patients with cancer. Several different categories of antigens have been recognized, which are presented in Table I. In the area of melanoma, tyrosinase and tyrosinase-related protein-2 (TRP-2), components of normal and transformed melanocytic cells, have been demonstrated to be targets for CD8+ T cells. These antigens, along with Melan-A/MART-1 and gp100, represent examples of melanocytic differentiation antigens. Antigens such as the MAGE family constitute the so-called cancer-testis antigens, as their expression is found only on normal spermatogonia and malignant cells. Other antigenic targets found include mutated self proteins such as CDK4 and β-catenin. Peptides that bind to specific HLA class I molecules can be readily synthesized; however, peptide vaccination is limited by the need for potent immunologic adjuvants.
II. DNA VACCINES
All of the vaccination strategies just described have relative advantages and disadvantages. Cell-derived vaccines have a composition that is difficult to precisely define and with that comes risk of transmission of unknown pathogens. However, these vaccines contain numerous potential antigenic targets, which provide an advantage over the carbohydrate vaccines. Peptide vaccines are simple to produce but require adjuvants and are only effective in certain HLA types.
Several years ago, researchers interested in vaccines for infectious disease began to experiment with the possibility of injecting DNA coding for antigens of interest directly into animals. Advances in molecular biology had made available DNA plasmids containing constitutively active promoters, and such plasmid DNA is simple and inexpensive to produce. Intramuscular injection of plasmid DNA had already been shown to result in uptake of the DNA and expression of the encoded gene of interest by myocytes. The work of Ulmer and colleagues demonstrated the first successful immunization of mice with DNA encoding influenza A nucleoprotein, resulting in the production of antigen-specific antibodies, cytotoxic T lymphocytes (CTLs), and protection from subsequent viral challenge. The use of plasmid DNA for vaccination was therefore practical and effective.
DNA immunization has several advantages over more conventional means of vaccination (summarized in Table II). In contrast to peptide vaccines, which are MHC restricted and only present a single epitope, full-length DNA vaccines contain numerous epitopes, both known and unknown. Because the antigen is being transcribed and translated by host cells, it is more likely to be presented in the proper context of MHC and costimulatory molecules. Most methods for delivery of DNA vaccines (described later) involve the transfection of bone marrow-derived antigen-presenting cells, which contributes to the ability of DNA vaccines to generate potent immune responses. The precise mechanism of immunization following introduction of DNA remains to be elucidated, but it is known that at least a small percentage of antigen-presenting cells are directly transfected. Other surrounding cells may release antigen through cell death or secretion, facilitating a process known as cross-priming during which preformed antigen is captured by antigen-presenting cells and transported to draining lymph nodes for presentation to naïve T cells.
TABLE II Relative Advantages of DNA-Based VaccinesCapable of eliciting both antibody and T-cell responses Simple to prepare Long shelf life Relatively inexpensive Multiple potential epitopes: HLA restriction not required Adjuvant effect of immunostimulatory sequences (CpG motifs)
Another property unique to DNA vaccines is the presence of immunostimulatory sequences (ISS) within bacterial DNA. It has long been known that unmethylated CpG motifs in prokaryotic DNA are highly inflammatory and lead to the activation of B cells and the release of cytokines, including IL-6, IL- 12, and both type I and II interferons. This inherent immunogenicity of DNA vaccines allows for their delivery without the need for immunologic adjuvants. Based on the exciting initial studies of DNA vaccines for infectious diseases, several groups began to investigate the possibility of using DNA vaccines to generate antitumor immunity. The first studies performed with DNA vaccination in the authors' laboratory used a model mouse tumor antigen system in which P13.1 tumor cells express β-galactosidase. Mice were immunized with a plasmid encoding β-galactosidase by a gene gun system in which DNA conjugated to gold particles is propelled into depilated epidermis using helium gas. The results of these experiments confirmed the ability of DNA immunization to generate potent CTL responses and also demonstrated that mice could be protected from challenge with tumor expressing a model antigen. In addition, established tumors (3 or 7 days after subcutaneous injection) could be rejected after as few as one immunization with the β-galactosidase gene.
B. Plasmid DNA Vaccines: Preclinical Mouse Melanoma Studies
Data just presented clearly demonstrate the ability of DNA immunization to generate antitumor immunity using a model antigen system. In an attempt to apply this technique to more biologically relevant situations, attention was turned to a mouse melanoma system. The components of melanocytic cells that constitute the melanin synthetic machinery (tyrosinase and tyrosinase-related proteins -1 and -2 (gp75TRP-1 and TRP-2), gp100) have been identified a targets for both antibodies and T cells in patients with melanoma. Immunization of mice with xenogeneic human gp75 results in the production of Th2 antibodies, which recognize both human gp75 and mouse gp75, as well as protection from tumor challenge with a syngeneic B16 mouse melanoma. Tumor immunity was accompanied by an autoimmune depigmentation of coat, which had been previously observed in studies using injection of an anti-gp75 monoclonal antibody. Tumor protection was shown to require natural killer cells, CD4 cells, and intact Fcγ receptors. No CTL response to gp75 has been detected, and CD8 cells are not required. An immune response could only be generated with the use of xenogeneic human gp75. Injection of syngeneic mouse gp75 DNA did not elicit an antibody response and did not result in tumor protection or depigmentation.
TRP-2 is another melanosomal membrane glycoprotein that shares approximately 50% homology with gp75. Immunization of mice with human TRP-2 results in rapid depigmentation and tumor protection from syngeneic B16 challenge. In contrast to gp75, TRP-2 elicits a potent CTL response, which requires CD4 cells, CD8 cells, and perforin to achieve tumor protection. Although Th2 antibodies are generated by immunization with both human gp75 and TRP-2, only gp75 requires antibodies to mediate tumor protection. Thus, experiments with TRP-2 show that DNA immunization can result in a Th1 CTL response, as well as a Th2 antibody response. This ability to mobilize multiple arms of the immune system represents an additional advantage of DNA vaccines. These mouse melanoma studies also document the ability of DNA vaccines to mediate potent immune responses to otherwise poorly immunogenic “self ” molecules.
C. DNA Vaccines: Other Tumor Systems
In addition to the melanoma model described earlier, DNA vaccines have been investigated using antigens from other types of cancer. Carcinoembryonic antigen (CEA) is a 200-kDa glycoprotein produced in large quantities by the first and second trimester fetus. In the adult, elevated levels of CEA can be detected in patients with adenocarcinomas as well as chronic obstructive pulmonary disease and cirrhosis. Immunization of mice with human CEA DNA results in the production of both antibodies and T cells, which recognize human CEA. The gene encoding MUC-1, an epithelial mucin expressed on a variety of adenocarcinomas, has also been used in DNA vaccination studies and was shown to induce protection from tumor challenge with a syngeneic MUC-1 expressing tumor.
B-cell lymphomas express distinct immunoglobulin molecules on their surface (idiotypes), which can serve as true tumor-specific antigens. Idiotypes from individual lymphomas can be cloned and inserted into expression vectors for use as DNA vaccines. Various constructs containing heavy chain, light chain, or single chain Fv (scFv) can be used, and some groups have even produced fusion proteins composed of idiotype and granulocyte-macrophage colony stimulating factor (GM-CSF). Immunization of mice with idiotype DNA vaccines results in an idiotype-specific antibody response, which can provide protection from challenge with the original lymphoma.
D. DNA Vaccines: Delivery Systems
The first studies of DNA vaccines were done simply by injecting plasmid intramuscularly (im) in mice. Subsequently, other routes for needle/syringe injection have been explored, including intradermal, intravenous, intraperitoneal, and subcutaneous injection. There is inconsistency in the literature as to which route is optimal for antibody versus CTL responses; however, most investigators have focused on intradermal and intramuscular injections. Administration of DNA using jet injection devices has also been investigated in both preclinical and clinical trials. Such devices may lead to more potent immune responses and have the important safety advantage of needle-free design. Application of DNA to oral, vaginal, and rectal mucosal surfaces for the purpose of inducing mucosal immunity has also been explored.
The development of the gene gun has popularized the technique of particle bombardment for the delivery of DNA vaccines. DNA conjugated to micronsized gold particles can be delivered to the epidermis by high-pressure helium gas. Relatively small quantities of DNA are needed for gene gun immunization (1 μg vs 100 μg for im). This method of immunization has the added advantage of accessing epidermal dendritic (Langerhans) cells, which are among the most potent antigen-presenting cells.
A recently described method of DNA immunization uses a live attenuated strain of Salmonella typhimurium to deliver plasmid orally. The bacteria is used as a carrier for DNA transfer and has the advantage of acting as an adjuvant through the release of numerous cytokines. Both antibody and T-cell responses have been documented with this system, and tumor protection using a model tumor antigen system was also demonstrated.
E. DNA Vaccines: Clinical Trials
As of the writing of this volume, several clinical trials for DNA vaccines have been completed in the area of infectious disease (malaria and HIV). These trials have documented the safety and immunogenicity of DNA vaccines. Study of the malaria circumsporozoite vaccine clearly demonstrated the advantages of DNA vaccines, especially the ability to generate CTL responses to multiple MHC-restricted epitopes. In the area of DNA-based vaccines for human cancer, the only completed studies are those involving Allovectin-7, a liposomally encapsulated plasmid containing the HLA-B7 gene. The goal of the Allovectin-7 trials was to stimulate an immune response to injected tumors based on recognition of the allogeneic HLA-B7 molecule. These trials documented the safety of intratumoral plasmid DNA injection, as well as suggesting efficacy of the local antitumor effect. Ongoing clinical trials of DNA vaccines for cancer include idiotype DNA vaccines at Stanford University and a gp100 DNA vaccine study at the National Cancer Institute. Further trials are expected to begin shortly for other melanosomal antigens, as well as for CEA.
Jedd D. Wolchok
Alan N. Houghton
Memorial Sloan-Kettering Cancer Center and Weill Medical College of Cornell University
ANTIBODIES IN THE GENE THERAPY OF CANCER ; ANTIIDIOTYPIC ANTIBODY VACCINES ; CANCER VACCINES: GENE THERAPY AND DENDRITIC CELL-BASED VACCINES ; CANCER VACCINES: PEPTIDE- AND PROTEIN-BASED VACCINES ; CARBOHYDRATE-BASED VACCINES ; MELANOMA: CELLULAR AND MOLECULAR ABNORMALITIES ; RESISTANCE TO ANTIBODY THERAPY ; TUMOR ANTIGENS
cross-priming The process during which cells surrounding the site of DNA immunization become transfected and release preformed antigen through cell death or secretion. The antigen is then captured by antigen-presenting cells and is transported to draining lymph nodes for presentation to naïve T cells.
immunostimulatory sequences Unmethlyated cytosine - guanine sequences present normally in bacterial DNA, which result in the production of proinflammatory cytokines in mice and humans.
particle bombardment Process of introducing DNA vaccines during which plasmid conjugated to microscopic particles is delivered under high pressure to the skin.
plasmid DNA vaccine A circular piece of DNA purified from bacteria that contains elements required for the expression of genes for antigens of interest in eukaryotic cells.
xenogeneic immunization Vaccination with a homologous gene or protein derived from an organism of a different species.
Bowne, W. B., Srinivasan, R., Wolchok, J. D., Hawkins, W. G., Blachere, N. E., Dyall, R., Lewis, J. J., and Houghton, A. N. (1999). Coupling and uncoupling of tumor immunity and autoimmunity. J. Exp. Med. 190, 1717-1722.
Donnelly, J. J., Ulmer, J. B., Shiver, J. W., and Liu, M. A. (1997). DNA vaccines. Annu. Rev. Immunol. 15, 617-648.
Huygen, K., Content, J., Denis, O., Montgomery, D. L., Yawman, A. M., Deck, R. R., DeWitt, C. M., Orme, I. M., Baldwin, S., D'Souza, C., Drowart, A., Lozes, E., Vandenbussche, P., Van Vooren, J. P., Liu, M. A., & Ulmer, J. B. (1996). Immunogenicity and protective efficacy of a tuberculosis DNA vaccine. Nature Med. 2, 893-898.
Ugen, K. E., Nyland, S. B., Boyer, J. D., Vidal, C., Lera, L., Rasheid, S., Chattergoon, M., Bagarazzi, M. L., Ciccarelli, R., Higgins, T., Baine, Y., Ginsberg, R., Macgregor, R. R., and Weiner, D. B. (1998). DNA vaccination with HIV-1 expressing constructs elicits immune responses in humans. Vaccine 16, 1818-1821.
Ulmer, J. B., Donnelly, J. J., Parker, S. E., Rhodes, G. H., Felgner, P. L., Dwarki, V. J., Gromkowski, S. H., Deck, R. R., DeWitt, C. M., Friedman, A., et al. (1993). Heterologous protection against influenza by injection of DNA encoding a viral protein. Science 259, 1745-1749.
Wang, R., Doolan, D. L., Le, T. P., Hedstrom, R. C., Coonan, K. M., Charoenvit, Y., Jones, T. R., Hobart, P., Margalith, M., Ng, J., Weiss, W. R., Sedegah, M., de Taisne, C., Norman, J. A., and Hoffman, S. L. (1998). Induction of antigen-specific cytotoxic T lymphocytes in humans by a malaria DNA vaccine. Science 282, 476-480.
Weber, L. W., Bowne, W. B., Wolchok, J. D., Srinivasan, R., Qin, J., Moroi, Y., Clynes, R., Song, P., Lewis, J. J., and Houghton, A. N. (1998). Tumor immunity and autoimmunity induced by immunization with homologous DNA. J. Clin. Invest. 102, 1258-1264.