Our Science – Berzofsky Website

Jay A. Berzofsky, M.D., Ph.D.

Vaccine Branch Head, Molecular Immunogenetics and Vaccine Research Section Branch Chief Building 10 Room 6B-04 10 Center Drive (MSC#1578) Bethesda, MD 20892 Phone:   301-496-6874 Fax:   301-480-0681 E-Mail:   berzofsk@HELIX.NIH.GOV

Biography

Dr. Jay Berzofsky was appointed Chief of the new Vaccine Branch, Center for Cancer Research, National Cancer Institute, in 2003, after being Chief of the Molecular Immunogenetics and Vaccine Research Section, Metabolism Branch, National Cancer Institute, NIH, since 1987. He graduated Summa cum Laude from Harvard (1967), and received a Ph.D. and M.D. from Albert Einstein College of Medicine. After interning at Massachusetts General Hospital, he joined NIH in 1974. Dr. Berzofsky's research has focused on antigen processing and presentation by MHC molecules, the structure of antigenic determinants, cytokine and regulatory cell control of T cell function and avidity, and translation to the design of vaccines for AIDS, malaria, cancer, and viruses causing cancer. He has over 390 scientific publications. Dr. Berzofsky has received a number of awards, including the U.S. Public Health Service Superior Service Award, the 31st Michael Heidelberger Award, the McLaughlin Visiting Professorship, the Australasian Society for Immunology Visiting Lectureship, and the Tadeusz J. Wiktor Memorial Lectureship. He is past President of the American Society for Clinical Investigation, and a Fellow of the American Association for the Advancement of Science, and was elected Distinguished Alumnus of the Year for 2007 by the Albert Einstein College of Medicine. He was also just elected Chair-Elect of the Medical Sciences Section of the American Association for the Advancement of Science (AAAS).

Research

T Lymphocyte Recognition of Antigens and Applications to Vaccines for AIDS and Cancer
We are studying mechanisms by which T cells recognize antigens presented by major histocompatibility complex (MHC)-encoded molecules (such as HLA in humans) and factors determining which structures are likely to be recognized, as well as the factors and mechanims that regulate the quantity and the quality of T cell responses, and applying these principles to the design of synthetic vaccines for AIDS, cancer, and viruses that cause cancer.

We have been characterizing the helper and cytotoxic T lymphocyte (CTL) responses to HIV envelope and reverse transcriptase, mapping the key epitopes, and defining the role of individual residues in these epitopes to be able to modify the structures to make more potent immunogens as vaccines. Currently, we are developing new approaches in mouse models to develop second generation vaccine constructs. We have shown proof of principle that we can modify the sequence of a helper epitope of HIV to make it more immunogenic and also much more potent, when coupled to a CTL epitope, in eliciting CTL and protecting against viral infection. We have applied this 'epitope enhancement' approach to conserved HIV helper and CTL epitopes from env, gag, and pol, presented by human class II and class I HLA molecules, as well as to hepatitis C virus (HCV) epitopes and cancer antigen epitopes presented by human HLA-A2.1 (see below). We have developed an enhanced HIV reverse transcriptase epitope which is the subject of a clnical trial being carried out with Dr. Robert Yarchoan, HAMB, CCR, NCI, and we have characterized human responses to an envelope helper epitope, which we have also enhanced. We have discovered ways of increasing CTL, helper, and antibody responses and steering them toward desired phenotypes, such as Th1 or Th2 or particular antibody isotypes, by incorporating cytokines into the emulsion adjuvant with the antigen. We have even further enhanced combinations of cytokines by a push-pull approach in which a potent combination of cytokine and costimulatory molecule is complemented by blocking a suppressive pathway with an inhibitor of IL-13, optimizing the vaccine-induced CTL response and protection. We have previously shown that high avidity CTL specific for HIV-1 envelope peptide are much more effective at clearing a recombinant vaccinia virus expressing HIV gp160 from SCID mice than are low avidity CTL specific for the same peptide-MHC complex, and have worked out two complementary mechanisms. We recently found that immunodominance depends more on CTL avidity than on the level of the epitope expressed. We have finally developed methods to preferentially elicit higher avidity CTL with a vaccine, by using costimulatory molecules to allow a response at lower antigen dose, and by incorporating IL-15 in the vaccine construct, selecting for cells that have higher IL-15Ralpha, and upregulating CD8. Indeed, we found that IL-15 incorporated in a vaccine induces CTL of a different character that are longer-lived memory cells, with higher levels of IL-15Ralpha, greater responsiveness to IL-15 in vitro and greater homeostatic proliferation in vivo, and higher avidity for antigen. We have shown that high avidity CTL express higher levels of IL-15 receptor alpha chain (IL-15Ralpha) and that this allows these cells to undergo more homeostatic proliferation in response to endogenous levels of IL-15 and survive, whereas low avidity CTL with lower levels of this receptor die out over time, providing an explanation for the long-standing enigma of T cell avidity maturation. We also found that IL-15 upregulates expression of the CD8 coreceptor. Thus, IL-15 mediates two mechanisms of T cell avidity maturation, selection at the population level and instruction at the single cell level. These mechanisms can explain the longstanding enigma of T cell avidity maturation. Use of this cytokine in vaccines should allow induction of longer-lived, higher avidity CTL that are more efficacious. Recently, we hypothesized that one explanation for the ability of CD4 T cell help to improve CTL memory was the induction of IL-15 production by dendritic cells activated by CD4 T cells, so that the same cell presenting the antigen to CD8 T cells was presenting IL-15 at the same time, as we have done with our vaccine. We have recently confirmed this hypothesis. This has important implications for therapeutic vaccines for HIV infected patients who have diminished CD4 T cell help. Expression of IL-15 by the vaccine vector might bypass the need for such help and allow a more effective vaccine response.

We have shown for the first time that protection against mucosal transmission of virus can be mediated by CD8 CTL without antibodies, but requires that the CTL be present at the mucosal site of transmission, whereas systemic CTL are not sufficient. The protection can be accomplished by intrarectal immunization with a peptide vaccine and increased by inclusion of IL-12 and GM-CSF with the vaccine. Using this strategy, we immunized MamuA*01-positive Rhesus macaques intrarectally with a similar peptide AIDS vaccine and induced CTL in the colon and mesenteric lymph nodes that have impacted the clearance of virus after intrarectal challenge with pathogenic AIDS virus SHIV-Ku. Intrarectal immunization was more effective than subcutaneous immunization with the same peptide vaccine at protecting against SHIV, in part because we found the induction of mucosal CTL provided for greater clearance from a major site of virus replication, the gut mucosa, which was seeding the bloodstream. We recently completed a second Rhesus macaque mucosal vaccine study using mucosal peptide priming, with a combination of cytokines as adjuvants, and using mucosal boosting with a recombinant poxvirus, to prevent AIDS virus transmission across a mucosal barrier. We used delay in appearance of viremia as a measure of dissemination of virus from the initial mucosal site of infection after intrarectal challenge with pathogenic SHIV-ku2. Peak viral load was delayed by 2-3 weeks, and viral load early after intrarectal challenge correlated with the CTL response before challenge to 2 different CTL epitopes, especially for high avidity CTL. Such a correlation was seen better for CTL in the colonic lamina propria and mesenteric lymph nodes than in the blood. We conclude that CTL in the mucosa can impact dissemination of virus from the mucosal site of transmission, and that the mechanism involves partial eradication of the cells infected in the initial nidus of infection by high avidity mucosal CTL. This finding is important for HIV vaccine design. In addition, we have shown that a new method of transcutaneous immunization can induce CTL in the gut mucosa, providing a potentially more practical route of immunization to induce the needed mucosal immunity. We are also examining the ability of different ligands for toll-like receptors (TLR) to act as mucosal adjuvants, compared to their effect in the systemic immune system, since the mucosal surfaces are constantly exposed to such TLR ligands from commensal flora. Already we have found some interesting synergies and antagonisms among TLR ligands.

With regard to cancer, we have carried out epitope enhancement of a new prostate and breast cancer antigen, TARP (see below), as presented by HLA-A2.1. We are developing a clinical trial to test these TARP peptides as a vaccine in prostate cancer patients who are HLA-A2 positive. We also developed a model of immunosurveillance of cancer in which tumors are rejected by CD8 T cells, but the rejection is incomplete in the presence of normal CD4 regulatory cells, and an escape variant of the tumor recurs. However, depletion of CD4 cells allows complete eradication of the tumor by CD8 cells. Using receptor knock-out mice, we found that the key regulatory cytokine inhibiting immunosurveillance against cancer was IL-13, acting through the IL-4 receptor/STAT6 pathway, although IL-4 itself was neither necessary nor sufficient. We discovered that the major source of IL-13 was NKT cells, not CD25+ “T reg� cells, and that absence of these NKT cells in CD1-knockout mice prevented tumor recurrence in these mice. Recently, we found that the regulatory NKT cell was a non-classical NKT cell restricted by CD1d but not using the invariant Valpha14Jalpha18 T cell receptor. This finding resolves the paradox that NKT cells have been found to protect against cancer as well as inhibit protection, in that the protective cells have been shown to be the classic NKT cells with the invariant receptor. We have now found that this regulatory pathway (or components thereof) applies also to murine models of breast cancer, colon cancer, and osteosarcoma, and that in each case we can prevent tumor growth by abrogating this immunoregulatory pathway. Indeed, in the absence of a vaccine, abrogation of this pathway unmasked immunosurveillance in a non-regressor colon carcinoma that was not apparent without the treatment, greatly reducing the number of tumor metastases. Blockade of IL-13 also delayed spontaneous breast cancer formation in Her-2/neu-transgenic mice. We also determined the mechanism by which IL-13 indirectly inhibits CD8 T cell-mediated immunosurveillance when the CD8 T cells do not have IL-13 receptors. We have identified an intermediate non-lymphoid cell that responds to IL-13 and then suppresses the anti-tumor CTL response by production of TGF-beta. We characterized this cell as a CD11b+Gr-1+ myeloid cell. We recently showed that blockade of TGF-beta could enhance the efficacy of an anti-cancer vaccine in a mouse tumor model. We are also developing clinical trial approaches to amplify immunotherapy of cancer by inhibiting IL-13 or TGF-beta. A trial of a monoclonal human anti-TGF-beta antibody in melanoma and renal cell carcinoma patients is under way.

We also developed a recombinant adenoviral vector vaccine expressing HER-2/neu that induced protection against spontaneous breast cancer in HER-2/neu-transgenic BALB/c mice. We showed that the mechanism of protection was dependent on antibodies that did not need to bind Fc receptors, but could block tumor growth directly in vitro as well as in vivo. Antibodies were necessary and sufficient for protection, whereas CD4 T cells were required only in the first 36 hours after immunization to provide help, and CD8 T cells did not seem to play a role. Strikingly, this adenoviral vaccine could cause complete regression of large established mammary tumors (> 1 cm) in mice. We would like to develop a similar vaccine for human breast cancer. We also induced murine CTL against fusion proteins from chromosomal translocations in pediatric tumors, synovial sarcoma, alveolar rhabdomyosarcoma and Ewing's sarcoma. We also identified novel epitopes spanning these fusion protein junctions in these sarcomas that could bind to several human HLA molecules, HLA-A1, A3, B7 and B27. We were able to map a minimal epitope in synovial sarcoma presented by HLA-B7 and one in alveolar rhabdomyosarcoma presented by HLA-B7, and elicit human CTL that could kill human sarcoma tumor cells in both cases, proving that these fusion proteins are promising tumor antigens for cancer immunotherapy. Recently we have defined epitopes in two newly discovered prostate cancer antigens, TARP and XAGE1, in collaboration with Ira Pastan (CCR, NCI), and have obtained HLA-A2-restricted T cell responses in HLA- transgenic mice. In the case of TARP, we have developed human TARP-specific CTL lines from a cancer patient that kill human breast cancer cells in vitro. Also, both breast and prostate cancer patients show such TARP-specific CTL in their peripheral blood. We have modified one of the TARP epitopes by epitope enhancement to make it more effective. A clinical protocol is written to treat prostate cancer patients who are HLA-A2 positive, representing about half of patients, as TARP is overexpressed in 95% of prostate cancer. 38 patients were treated in a phase I/II clinical trial of the mutant p53/ras peptide vaccine approach to treating cancer, and a large fraction have made CTL or cytokine responses, and no adverse effects were seen. Importantly, we found that a cellular immune response to the mutant peptide was statistically significantly associated with longer survival. We have also started a trial of immunization of cervical cancer patients with peptides from the E6 and E7 oncoproteins of human papillomavirus type 16 that bind to HLA-A2.1 in patients who express this HLA molecule. Many patients made CTL responses, and some had unexpectedly stable disease. Prolonged survival was again associated with response to the vaccine E6 peptide.

This page was last updated on 6/11/2008.