Full-length natural human antibodies thwart HIV in mice

The war on the HIV pandemic is being won slowly but surely, one battle at a time. An important battle in that war may have been won by a California Institute of Technology (Caltech) research team led by Alejandro Balazs, Ph.D., and David Baltimore, Ph.D. Although numerous antibodies that can neutralize most HIV strains have been identified, a conventional immunogen-based vaccine that elicits stable antibody production has not been developed. Taking an alternative, gene-based therapy approach, the Caltech team demonstrated in 2011 an optimized vector engineered to encode full-length, natural human HIV-neutralizing antibodies prevents HIV infection in mice (Balazs et al. 2011). The novel approach, called vectored immunoprophylaxis (VIP), works very differently than conventional vaccines and may be just the weapon needed to significantly roll back the HIV pandemic.

Building a better antibody-encoding vector

Previously constructed adeno-associated virus (AAV)-based transgene vectors contained heparin-binding capsids, which elicit an immune response from cytotoxic T cells. This immune response can destroy the cells transduced with the virus and eliminate therapeutic gene expression. To circumvent this response, the Caltech team constructed a vector with a capsid from a non-heparin-binding serotype 8 AAV. To reduce immune responses against the transgene, the researchers used a transgene that encodes full-length heavy- and light-chain variable regions of the naturally occurring human HIV-neutralizing b12 antibody. To ensure good muscle tissue expression, the researchers engineered the transgene to be driven by a novel synthetic “CASI” promoter that includes CMV and ubiquitin C enhancer elements and the chicken b-actin promoter.

To test the antibody-generating ability of this construct, the researchers injected it into the gastrocnemius muscle of four mouse strains. Two of them, C57BL/6J (000664) and BALB/cJ (000651), are immunocompetent; the other two, NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ (005557, NSG) and a Rag2-deficient mouse, are immunodeficient. The Caltech team tested the construct in these four strains to see if the absence or presence of the endogenous mouse immune system would confound its effects. The vector and its encoding transgene lived up to expectations: They induce exceptional and long-lasting amounts of encoded antibody in all four mouse strains.

New vector prevents HIV infection in mice

To determine if their optimized VIP vector protects mice from an HIV infection, the researchers injected it into PBMC-humanized NSG mice. These are mice engrafted with mature human peripheral blood mononuclear cells. They then infected these humanized mice with HIV. As a control, they infected some NSG mice that had been injected with a luciferase-expressing construct instead of a b12-expressing construct. The Caltech team found that, following HIV challenge, the luciferase-expressing controls exhibit the expected dramatic loss of human CD4+ T cells. In contrast, the b12-expressing mice exhibit no CD4+ T cell loss, even when challenged with very high HIV doses.

Humanized NSG mice are one of the best model organisms for studying the human immune system in vivo. They are being widely used to research multiple human infectious diseases as well as human hematopoiesis.

The researchers then compared the efficacy of the b12-expressing construct to similar constructs that express one of three historically known HIV-neutralizing antibodies: 2G12, 4E10, and 2F5. They found that humanized NSG mice injected with constructs that express any one of the three latter antibodies exhibit a range of CD4 cell loss, indicating that they only partly protected T cells from an intravenous HIV challenge.

Additionally, eight weeks after an HIV challenge, humanized mice injected with the 2G12, 4E10 and 2F5 antibody-expressing constructs showed variable levels of HIV-expressed p24 antigen in the spleen consistent with only partial protection against HIV. In contrast, p24 HIV antigen was not detected in the spleens of the b12-expressing humanized mice.

Finally, the Caltech team compared the efficacy of the b12-expressing construct to one expressing VRC01, one of several more recently identified and very effective HIV-neutralizing antibodies. The team found that both constructs induce dose-dependent antibody expression in NSG mice. The minimally effective dose required for the VRC01-expressing construct to completely protect the mice from HIV infection, however, is approximately four-fold lower than for the b12-expressing construct. Suboptimal concentrations of either antibody are only partly protective, with several mice exhibiting no detectable CD4 cell loss and others exhibiting delayed CD4 cell loss.

In summary, the Balazs-Baltimore team developed vectored immunoprophylaxis (VIP), a novel, gene transfer-based therapy that protects humanized mice from HIV infection. VIP consists of an AAV-derived vector that contains non-heparin-binding capsid proteins and encodes genes that express full-length human HIV-neutralizing antibodies. When injected intramuscularly in humanized mice, VIP vectors confer long-lasting protection against even very high-dose HIV infections. The team believes that a similar VIP approach could confer broad, powerful HIV protection to humans. The approach may also be very effective in protecting against other infectious diseases and in therapies that require continuous production of monoclonal antibodies.