With >36 million people infected and almost 2 million new infections in 2016, the need for a prophylactic HIV-1 vaccine is as urgent as ever. Approximately 20-30% of HIV-1 infected individuals generate antibodies capable of neutralizing diverse heterologous viral strains (broadly neutralizing antibodies or bNAbs). Because they protect from experimental infection, it is thought that bNAbs will be an important part of an HIV-1 vaccine. Yet immunization of humans with recombinant Env leads to the production of antibodies with very narrow breadth of neutralization which fail to block infection of diverse circulating viral isolates. The isolation of monoclonal bNAbs from HIV-1+ individuals has provided valuable information on how bNAbs develop during infection, and on the epitope specificities on the HIV-1 Envelope protein (Env) that an effective vaccine should aim to elicit.
One of the first critical steps in an effective antibody response is the recognition of a foreign antigen by a membrane anchored B cell receptor (BCR) on the surface of a naïve B cell. Naive BCRs arise through the random recombination of germline-encoded immunoglobulin genes during B cell development. Antigen recognition triggers a cascade of events that lead to B cell clonal expansion and somatic hypermutation; critical processes that stochastically diversify the BCR repertoire by introducing mutations into the BCR-encoding DNA. Expanded pools of diverse B cell clones compete for antigen and T cell help in a Darwinian process that ultimately selects for B cells with higher affinity receptors which can be secreted as soluble molecules (antibodies). As a result of these processes many of the bNAbs are highly mutated and highly divergent from their ancestral BCRs from which they are derived. As a consequence of bNAb precursors fail to bind most recombinant HIV-1 Envs (such as those that would be used in HIV-1 vaccines). Thus, one potential reason for the failure of HIV-1 vaccines to elicit bNAbs is that the Env immunogens tested failed to engage bNAb precursor B cells, and thus failed to start the process of bnAb production. As an alternative to using HIV-1 Env as a vaccine immunogen, we are pursuing a novel vaccine strategy where we generate anti-idiotypic antibodies: antibodies raised in a non-human animal model that bind with high affinity and specificity to bNAb precursors. Importantly, because these immunogens are not Env-derived, they should not stimulate non-neutralizing B cell lineages which are readily elicited by conventional Env-vaccines. We are currently generating, optimizing and testing these immunogens in various humanized mouse models.
Epstein-Barr virus (EBV) is one of the most common human viruses. It is a causative agent of infectious mononucleosis and is associated with ~200 000 new cases of cancer and ~140,000 deaths annually. It is thought that a vaccine that prevents EBV infection and/or associated pathologies would have a significant clinical benefit. However, the kinds of immune responses a protective EBV vaccine would need to elicit have not been defined. Consequently it is not clear which viral antigens are ideal candidates for an EBV vaccine.
To get a better understanding of what types of antibodies a protective EBV vaccine should elicit, we isolate monoclonal antibodies targeting multiple viral proteins from people naturally infected with EBV and evaluate their ability to neutralize, or block infection of multiple cell types, and to block experimental EBV infection in animal models. We couple this with structural information to define critical sites of vulnerability on the virus. We then use this information to design and optimize vaccine antigens that can focus the immune response on key antigenic sites and evaluate their ability to elicit protective antibodies in animal models.
Many EBV-associated cancers express virally-encoded transmembrane proteins. Two such proteins Latent Membrane Protein1 (LMP1) and Latent Membrane Protein 2 (LMP2) that play roles in oncogenesis. Both are multi-spanning transmembrane proteins, with a small extracellular region that is exposed on the surface of the infected cell. Our hypothesis is that the exposed extracellular regions of LMP1 and LMP2 could be targeted by cytotoxic T cells expressing chimeric antigen receptors (CARs). However, the development of CAR based therapies to target LMP-expressing malignancies are currently hampered by a lack of antibodies that recognize the extracellular domains of these proteins. Using a humanized mouse model, we are working to isolate human antibodies that specifically recognize the extracellular domains of membrane-anchored LMP1 and LMP2, and to convert these antibodies into chimeric antigen receptors that can be used to direct potent T-cell-mediated killing to EBV+ cancer cells.