With so much known about Ad biology, it was a natural step to use recombinant Ad for vaccination purposes. As a first example, [E1-]Ad vectors have been engineered that express genes encoding epitopes (or whole proteins) derived from a number of pathogens, including (but not limited to): malaria, bovine herpesvirus (type 1), foot-and-mouth disease virus (FMDV), measles virus, and HIV (human immunodeficiency virus) (5-7). Vaccination with these Ad vectors has shown positive efficacy in preventing viral infection upon challenge with each of the respective viruses.
The potential benefits provided by the use of Ad vectors as vaccines for the prevention of viral illnesses coincide with simultaneous efforts to use Ad vectors to vaccinate against various types of cancer. The first use of Ad against cancer occurred shortly after its discovery in 1953. In this experiment, concentrated wild-type Ad were intravenously administered into subjects affected by cervical cancer. The hypothesis was that the lytic properties of Ad infection of cancer cells might slow or halt the cancer's progression, a popular notion of the period (8). Of the subjects treated, 65% reportedly had evidence of necrosis of cancerous tissue, without appreciable side effects. Autopsy findings demonstrated that the Ad had effects only on the cervical tumors and was not able to affect growth and metastases of pelvic tumors. Thus, although showing some promise, the lytic properties of Ad did not seem sufficient to effectively fight disseminated cancer.
A resurgence however has occurred in the use of Ad to treat cancer, as a result of both advancements in understanding of cancer biology combined with the ability to generate Ad vectors. For example, Ad vectors have been constructed that express a variety of immunostimulatory genes, tumor-suppressor genes, tumor repair genes, ribozymes, antisense oncogenes, apoptosis-inducing or toxic genes, and other genes thought to be critical in combating the growth of cancerous cells. Prodrug-sensitizing and tumor-specific, semireplicative Ad vectors have also been developed, representing other approaches to Ad-based cancer therapy. In these approaches, tumor nodules are typically directly injected with the respective vector, to achieve at least localized, if not systemic anticancer effects. For a more detailed discussion of these types of strategies, a number of reviews are recommended (9,10).
An alternative approach to using recombinant Ad vectors to combat cancer involves its use as a traditional vaccine, to immunize individuals against cancerous cells. This strategy was made possible by the discovery of various tumor-associated antigens (TAAs), which may provide immunological targets to a number of types of cancer via a traditional vaccine approach. One of the first demonstrations of this approach utilized an [E1-]Ad vector expressing the EBVgp340/220, in an attempt to induce protective immunity in cottontop tamarins challenged with Epstein-Barr virus (EBV)-induced lymphomas (11). In the study, three intramuscular injections of the recombinant Ad vector were sufficient to protect all immunized animals against subsequent tumor challenge. This tumor-specific protective effect was also seen when a recombinant Ad encoding beta-galactosidase was used in a vaccination of BALB/c mice (12). In this study, the ex vivo infection of splenocytes with an [E1-]Ad expressing the bacterial beta-galactosidase gene could induce a dramatic regression of established pulmonary metastases in the mice. A caveat of this study was that the tumor cells (CT26.CL25) were previously engineered to also express the bacterial P-gal-expressing gene. The bacterial beta galactosidase gene may have enhanced the repertoire of immune targets in this scenario, acting as another foreign epitope to the host animal's immune system. Other limitations to the approach were also noted; for example, mice bearing established tumors, subsequently vaccinated with the same [E1-]Ad vector, experienced no significant regression of tumor, even when simultaneously treated with interleukin-2 (IL-2). Though limited by the immu-nologic assays of the day, and use of tumor cells lines expressing potentially foreign transgenes, both of these early studies showed the potential of Ad as an effective cancer vaccine.
Another obvious cancer target can be found in cervical cancers induced by the human papilloma virus (HPV). Approximately 95% of all cervical cancers are caused by HPV infection and subsequent oncogenesis. In these types of cancer, HPV-derived expression of oncogenic proteins (i.e., the E5, E6 and E7 genes) results in the transformation of normal cervical cells. Thus the E6 and E7 are TAAs, potentially allowing for specific induction of immunological effector cells specific for HPV-transformed cervical epithelium. For example, He et al. created both vaccinia virus and [E1-]Ad-based vectors that both encoded the HPV E6 and E7 genes (13). These constructs were both tested for their ability to induce cancer immunity in both BALB/c and C57Bl/6 mice. The results of these studies demonstrated that depending on the strain of mouse model tested, either CD4+ or CD8+ T cells were stimulated by the viral vaccines. The class of immune effector cells elicited resulted in strikingly different responses to tumor cell challenge. The best results were obtained using the AdE7 vaccine in C57Bl/6 mice (~60% survival), a response found to be completely CD8+ dependent. No other vaccinations were successful in the C57Bl/6 mice. In the BALB/c mice, the AdE6 and VacE7 vaccines were equally efficacious (~40% survival), and both vectors induced CD4+-dependent responses. Though surprising, these results led the authors to conclude that [E1-]Ad vectors induce a more potent immune response compared to vaccinia virus-based vectors, and that a potent CD8 T-cell response was needed to eradicate HPV-16 tumor cells.
A similar study conducted using recombinant adenovirus encoding HPV-16 E5 found significant protection against tumor challenge after a single intramuscular injection of the vector into C57Bl/6 mice (14). Utilization of gene-specific knock-out mice confirmed that the effect was also CD8 T-cell dependent in these animals. Both studies reflect favorably upon the principle of using a recombinant Ad vaccination against particular/ specific cancer genes, and that tumor rejection by CD8 T cells is possible once peripheral tolerance has been broken, although the studies suffer from the fact that the targeted TAA gene is also a potentially foreign antigen.
Another TAA target for cancer immunotherapy approaches is the prostate-specific antigen (PSA), a protein expressed only in prostate epithelial cells. Additionally, since this gene is normally expressed in healthy prostate epithelium, this antigen is a target that will induce auto-immunity against the entire prostate (which is, at some level, a nonessential organ). Since 95% of all prostate tumors overexpress PSA, it is a potent target for cancer vaccine strategies in general, and Ad vector-based cancer immunotherapy strategies specifically (15). In Ad vaccine studies involving this antigen, a PSA-specific T-cell response could be demonstrated in Ad-PSA-immunized mice (15). This Ad-PSA-induced immunity was both cellular and humoral, and could also protect mice against a subsequent subcutaneous challenge with PSA-expressing cancer cells. The antitumor immunity was predominantly mediated by CD8+ T lymphocytes, but this was only a protective immunity, as Ad5-PSA alone was unable to control growth of established, preexisting PSA tumors. Existing tumors (500-1000 mm3 in size) could only be eliminated if Ad5-PSA priming was followed 7 d later through an intratumoral injection of several canarypox viruses encoding different immunostimulatory factors (IL-12, IL-2, and TNF-a). In these instances, the immune response was dominated by CD8+ T cells, but natural killer (NK) cells were also necessary for effective tumoricide. Human phase I studies for specific prostate cancer antigens (prostate-specific membrane antigen [PSMA]) have yielded inconclusive results about the effectiveness of Ad vector PSA-
vaccination therapy (possibly because of subjects' concomitant hormone therapies and subject population heterogeneity), but have shown excellent safety profiles after viral administrations (16). Phase II trials to evaluate the efficacy are currently under way.
The specific immunoglobulin protein expressed by a B-cell lymphoma is another tumor-specific antigen, and as a self-antigen, is most likely a weak immunogen. In this, it is a specific and realistic target to test the efficacy of cancer vaccines in their ability to break tolerance and effectively target-specific cancer cells. In several studies, recombinant adenoviruses have been constructed that encode variable regions of the immunoglobulin light and heavy chains with or without a xenogenic Fc fragment (17-20). In vitro studies have shown cellular immunization with variable fragments alone can stimulate the production of T cells specific for the lymphoma (18). When taken to mouse models, the investigators were unable to demonstrate specific T-cell generation after injection of Ad vectors expressing variable region antigens alone (18,19). It was only after the variable regions were recombinantly linked to the xenogenic human Fc fragment (in the Ad vector construct) was the vaccination able to significantly inhibit tumor growth in mouse models. In these cases, a single immunization provided protection for over 40% of mice challenged with a lethal dose of the specific lymphoma line, a level of protection that was greater than could be previously achieved with optimized protein- or plasmid-based vaccines (18). When combined with chemotherapy, Ad vector vaccination significantly prolonged the survival of mice with preexisting tumor; however, vaccination alone wasn't sufficient to clear preexisting tumor loads (17). Again, these responses were noted to be largely CD8+ T-cell dependent, although significant antibody titers were also achieved against the targeted epitopes.
The antigen CO17-1A/GA733 has been consistently identified as a highly expressed antigen in various colorectal cancers and, as such, has been a useful target in both passive and active immunotherapy approaches using both monoclonal antibodies and antiidiotype antibodies. In addition to these strategies, the vaccination approach in targeting this human TAA has been tested using baculovirus, vaccinia, and Ad-based vectors to induce tumor-specific immunity (21). In that study, each of these human GAA733-expressing recombinant viruses induced antigen-specific antibodies and delayed-type hypersensitivity (DTH) responses in mouse models, but only the Ad vaccine was found to induce antigen-specific cytotoxic T lymphocytes (CTLs) and regression of established tumors. In a similar study using an Ad vector expressing human GA733, protected mice were also found to resist subsequent challenge with antigen-negative colorectal cancer tumors (the original tumors were antigen-positive colorectal cancer tumor lines), indicating that an immunologic crosslinking of tumor antigens after a successful vaccination could also be achieved (22). The results demonstrate the potency of successful Ad vaccination in colorectal cancer models, but were hampered by the fact that the recombinant Ad constructs encoded a xenogenic tumor gene (human GA733) delivered into a murine model system. When the murine homolog of the human GA733 antigen was used (murine epithelial glycoprotein, or mEGP) a recombinant Ad was unable to protect vaccinated mice against tumor challenge (21). Only when combined with IL-2 was recombinant Ad-mEGP able to significantly inhibit growth of established mEGP-positive tumors. Thus, although the significance of the immunological crosslinking in these studies cannot be overstated, use of an Ad-based vector alone did not appear sufficient to break native TAA tolerance unless xenogenes or other immunological stimulators (such as certain cytokines) were employed.
Another type of cancer presenting attractive targets for cancer vaccination approaches is melanoma. The melanosomal proteins TRP2, MART, and gp100 are all highly expressed in melanoma, and have been studied in experiments involving recombinant Ad vaccines. TRP2 is highly homologous in mice and humans, and although attempts at plasmid and gene gun vaccination have failed to induce immunity, administration of an Ad vector expressing a xenogenic human TRP2 caused mice to mount an immune response against their native melanocytes, resulting in coat depigmentation (23). These mice were also significantly protected against metastatic growth of B16 melanoma, an immunity associated with the presence of TRP2-reactive antibodies and CD8 T cells. In vitro studies of HLA-A2+ cell lines confirmed this association, in showing generation of melanoma-specific CD8 T cells for the other melanoma antigens, gp100 and MART-1 (24). Studies conducted in C57Bl/6 mice mirror results found with TRP2, showing a protective effect in mice vaccinated with melanoma tumor antigen (24). Furthermore, this protective effect is mediated principally through the generation of CD8+ T cells specific for the tumor antigen.
Based on this success, recombinant Ads encoding MART and gp100 were taken to human phase I trials. In these trials, 54 subjects with metastatic melanoma were given escalating doses (up to 1x 1011 infectious units) of the Ad-based vectors with or without IL-2 (25). Of all the subjects receiving the Ad vector alone, only one achieved a complete response. Other subjects achieved their objective responses, but only with coadministration of IL-2. Immunological assays of subjects showed no consistent immunization to either MART-1 or gp100, although the authors concluded that this may have been attributable to the high level of neutralizing antibodies found in the subjects' sera prior to treatment, and/or the advanced state of the subjects' disease. Importantly, none of the escalating doses proved excessively toxic. The single clinical successes achieved in subjects through the coadministration of IL-2, demonstrate the potential of newer combined vaccine therapy approaches.
New vaccination strategies involve the ex vivo vaccination of a subjects' own dendritic cells (DCs) as a way to more effectively "educate" the immune system to target cancer antigens as well as minimize any possible side effects involved with administering high titers of recombinant adenovirus vectors directly in vivo. DCs are currently recognized as the key professional antigen-presenting cells (APCs) of the immune system and, as such, manipulation of DCs is a very active area of investigation in vaccine-based therapies. Antigen-loaded DCs efficiently interact with and mobilize the effector T cells of the immune system. These interactions are facilitated by DC coexpression of stimulatory adhesion molecules, and DC elaboration of cytokines such as IL-12 (26). Since one can generate clinically relevant numbers of DCs (i.e., from expansion of CD34+ progenitors in cord blood, bone marrow, or the peripheral circulation), the capability exists to "load" DCs with relevant proteins (typically tumor-specific antigens) ex vivo, and reinfuse the antigen-loaded DCs back into subjects to initiate and direct therapeutic immune responses (CTLs) to cells in the body expressing the antigen, i.e., cancer cells.
A number of methods are currently available to load DCs with antigens. Though several methods of DC transduction of DNA or RNA have been explored, many methods have been found to be significantly less efficient than similar attempts utilizing Ad vectors. For example, Arthur et al. demonstrated that Ad vectors encoding a variety of antigens could efficiently transduce 95% of the DC exposed to high titers of the vector whereas Dietz et al. report that, when used in combination with liposomes, low titers of Ad vectors can effectively transduce more than 90% of exposed DCs (27). Importantly, increasing levels of gene expression were noted in the DCs with increasing multiplicities of infection (MOIs) with the Ad vector (28). It has been demonstrated that DCs infected with Ad vectors encoding a variety of antigens (including the tumor antigens MART-1, MAGE-A4, DF3/MUC1, p53, hugp100 melanoma antigen, polyoma virus middle-T antigen) have the ability to induce antigen-specific CTL responses, have an enhanced antigen presentation capacity, and have an improved ability to initiate T-cell proliferation in mixed lymphocyte reactions (27,29-35). Immunization of animals with DCs previously transduced by Ad vectors encoding tumor-specific antigens has been demonstrated to result in significant levels of protection for the animals when challenged with tumor cells expressing the respective antigen (36,37). Interestingly, intratumoral injection of Ads encoding IL-7 was less effective than injection of DCs transduced with IL-7 encoding Ad vectors at inducing antitumor immunity, further heightening the interest in ex vivo transduction of DCs by Ad vectors (38).
Ad vector capsid interaction with DCs (independent of Ad transcription) appears to trigger several responses, which may be enhancing the ability of DCs to present antigens. Controversy exists as to whether this effect is a result of an Ad-mediated induction of DC maturation (26,39). Studies of immature bone marrow-derived DCs from mice suggest that Ad vector infection of these cells resulted in upregulation of cell-surface markers normally associated with DC maturation (MHC I and II, CD40, CD80, CD86, and ICAM-1) as well as downregulation of Co11c, an integrin known to be downregulated upon myeloid DC maturation. In some instances, Ad vector infection triggers IL-12 production by DCs, a marker of DC maturation (26). These events may possibly be due to Ad-triggered activation of an NF-kB pathway (40,41). Similar studies in mature CD83+ human DCs (derived from peripheral blood monocytes) demonstrated that mature human DCs were well tranduced by Ad vectors, and did not lose their functional ability to stimulate the proliferation of naïve T cells when lower numbers of infectious units per DC were utilized; however, some studies suggest that mature DCs are less infectable than immature ones (39,42).
Furthermore, as DC transduction methods improve, DC abilities to stimulate antigen-specific T cells also improve. These studies suggest that, if one can improve tumor gene-encoded antigen expression from within DCs, one might improve the ability to generate T-cell-mediated immune responses. Unfortunately, at higher MOIs, DC cytotoxicity and DC death has been shown to be induced by [E1-]Ad vector infections, as well as causing a "blunted" ability of the mature DCs to stimulate naïve T cells, as compared to lower MOI infections of the mature DCs (42). The latter prompted the authors to suggest that residual gene expression derived from the [E1-]Ad vector backbone may have been the cause for this effect. This hypothesis has yet to be proven, but studies with fully deleted Ad vectors have suggested that residual gene expression may not contribute to DC activation induced by Ad vector infections (43-46). One trial currently under way involves the use of a recombinant Ad vector encoding MART-1 to transduce DCs ex vivo, and reinjecting these modified ("educated") DCs as a potent cancer vaccine. This trial represents what will likely become the current and future thrust in Ad vaccine therapy approaches for cancer patients.
Was this article helpful?
Complete Guide to Preventing Skin Cancer. We all know enough to fear the name, just as we do the words tumor and malignant. But apart from that, most of us know very little at all about cancer, especially skin cancer in itself. If I were to ask you to tell me about skin cancer right now, what would you say? Apart from the fact that its a cancer on the skin, that is.