In the cancer immunotherapy arena, we solved a major paradox in the role of natural killer T cells (NKT cells) in tumor immunosurveillance, as NKT cells have been reported by others to protect, and by us to suppress tumor immunosurveillance. Type I NKT cells (carrying the invariant Va14Ja18 T cell receptor) contribute to tumor immunity, whereas type II NKT cells (with diverse receptors) are sufficient to suppress tumor immunosurveillance. We showed that the type II NKT cell is the main cell suppressing immunosurveillance in 4 tumor models in which the classic CD4+CD25+ T regulatory cell does not play a role. Importantly, we discovered that protective type I NKT cells and suppressive type II NKT cells cross-regulate each other, defining a new immunoregulatory axis between these polar opposites. The balance between these 2 cell types may determine the immune response to tumors. Blockade of interleukin-13 (IL-13), a mediator of type II NKT cells, also delayed cancers, even in Her-2/neu transgenic mice. We discovered that blockade of tumor necrosis factor alpha (TNFalpha) inhibits tumor growth by blocking synergy between IL-13 and TNFalpha to induce transforming growth factor-beta (TGF-beta) production. We also found that blockade of the downstream mediator of this NKT regulatory pathway, TGF-beta, can synergize with anti-cancer vaccines in mice, increase the T cell response, and more completely inhibit growth of an established tumor (cervical cancer model) than the vaccine alone. The protection is CD8 T cell-dependent. We have completed a phase I clinical trial of a human anti-TGF-beta monoclonal antibody in melanoma patients as part of a CRADA with Genzyme Corp. and a phase II trial is planned. The intent is to use this agent in conjunction with a cancer vaccine. We also found that a recombinant adenovirus expressing Her-2 can prevent appearance of spontaneous autochthonous breast cancers in BALB-neuT Her-2 transgenic mice. The mechanism is antibody mediated and requires CD4 T cells only for help, and does not require CD8 cells. However, the protection is not FcR dependent, unlike that of Herceptin. The antibody induced by the vaccine blocks phosphorylation of Her-2 and downmodulates this oncoprotein, rather than initiating killing by antibody-dependent cellular cytotoxicity. It inhibits tumor growth in vitro in the absence of other cells. This vaccine works therapeutically to cure large (larger than 1 cm) established Her-2 positive tumors and large established lung metastases in mice. This vaccine may be more effective for breast cancer than the monoclonal Herceptin, because the patient would make her own antibodies rather than needing repeated costly infusions and these would not be dependent on FcR-mediated mechanisms. The polyclonal antibodies elicited may be more resistant to escape mutation than a monoclonal. We are making an adenovirus expressing the corresponding domains of human Her-2 in preparation for a clinical trial in human breast cancer, and have had pre-investigational new drug discussions with the FDA. We have mapped a new epitope presented by HLA-B7 from the PAX-FKHR fusion protein that is unique to and expressed by 85% of alveolar rhabdomyosarcoma tumors. Also, we have mapped HLA-A2-restricted epitopes from 4 new cancer antigens in collaboration with Ira Pastan (Laboratory of Molecular Pharmacology, CCR), all expressed in prostate cancer but several also expressed in breast cancer and some other tumors. We have modified the sequences of these antigens by epitope enhancement to make them more immunogenic, and have shown killing of human tumor cells by T cells specific for the first two of these agents. The first of these to be developed, TARP, will be used in a clinical trial of prostate cancer patients. In the human immunodeficiency virus (HIV) and viral vaccine arena, in studies also applicable to cancer, we found that interleukin-15-(IL-15) expression by a vaccine leads to induction of higher avidity cytotoxic T lymphocytes (CTLs) that more effectively clear virus infections (or kill tumor cells). We also found that IL-15 overcomes the lack of CD4 T cell help in CD4-deficient animals, allowing induction of long-lived memory CTL that would not be induced otherwise. This may be a critical finding for therapeutic HIV and cancer vaccines, because it will be necessary to immunize patients without assistance from CD4 T cells. Indeed, IL-15 prevented the TNF-related apoptosis-inducing ligand (TRAIL)-induced death of CD8 T cells in the absence of CD4 help, and was necessary as well as sufficient for help, implying that induction of IL-15 production by dendritic cells is an important mechanism of T cell help. The vaccinia-IL-15 was also a more effective smallpox vaccine than a control vaccinia. IL-15 also induced novel CD8aa T cells distinct from intestinal epithelial lymphocytes. Further, we found that CD8 T cell avidity was a major factor in determining which antigen epitopes were immunodominant, more so than the level of epitope expression. We also enhanced HIV helper and CTL epitopes presented by human class II and class I HLA molecules, to improve HIV vaccine constructs. Using one of these epitopes with the anti-retroviral drug lamivudine (3TC), we developed a vaccine that provides CTL-mediated counter-selective pressure against drug-resistant mutants of HIV. This approach is now being tested in a clinical trial in collaboration with Bob Yarchoan (HIV and AIDS Malignancy Branch, CCR). We also showed that the TLR9 ligand CpG oligonucleotides could enhance the efficacy of a live viral vector vaccine, modified vaccinia Ankara (MVA), thereby reducing the dose needed to protect and induce protective CD8 CTL. Strikingly, the CpG oligos used with the MVA also allowed induction of CTL and protection in CD4-deficient mice. This may be another approach to a therapeutic vaccine for CD4-deficient patients infected with HIV. Further, we showed in mouse and macaque studies that clearance of virus from gut mucosal tissues, where HIV predominantly replicates, is dependent on high avidity CTL in the local gut mucosa. Protection correlates with high avidity but not low avidity CTL in the gut mucosa, and not with CTL in the peripheral blood. Because the gut mucosa is a major reservoir for HIV replication, inducing mucosal high avidity CTL to eradicate this reservoir may be critical to keeping HIV under control. We found, in both mice and macaques, that local mucosal immunization induces higher avidity CTL than immunization at more distant sites, and that CTL avidity is correlated with proximity to the site of vaccination, whether mucosal or systemic. We also showed, in macaques, that it is possible for a mucosal AIDS vaccine to induce sufficient CTL in the mucosa to significantly delay dissemination of virus to the bloodstream and beyond. Although this vaccine did not completely prevent transmission, the delay indicates that a substantial fraction of the virus was eradicated at the initial mucosal site of infection. An improved version of this approach may be able to eradicate the initial nidus of infection and thus abort the AIDS infection before it becomes established. To improve this vaccine, we are studying other TLR ligands as mucosal vaccine adjuvants, and have found a synergistic tr [summary truncated at 7800 characters]
Showing the most recent 10 out of 53 publications
