Pepe Alcamí
IDIBAPS researcher and scientific director of the HIV Unit at Hospital Clínic de Barcelona
The development of a preventive vaccine against HIV infection is one of the great unfinished business in HIV research and, unfortunately, so far it is a story of failure.
The objective of any preventive vaccine is to induce the production of neutralizing antibodies by the immune system and usually the antigen used must include or consist of the envelope or surface proteins of the virus. It is these proteins that interact with the cell entry receptors, so their blocking by antibodies neutralizes the virus infection.
The difficulty in obtaining a vaccine is due to the structure of the HIV envelope, which makes it very inaccessible to the action of neutralizing antibodies: it is a folded envelope that hides the domains of interaction with the HIV receptors and also has a high degree of glycosylation on its surface that creates a 'glycan shield' that blocks the access of antibodies.
In the last decade, 'Achilles heels' have been identified in the virus, which are domains of the envelope that are accessible. However, these antibodies are very much in the minority because of a minority of antibodies that, in addition to being directed against these Achilles heels, must have a structure capable of crossing these barriers. This structure is an extension of one of its parts, the so-called HCDR3 domain, which only a minority of antibodies have. Moreover, those that are generated have a very low affinity and require a maturation process that takes months or years to generate potent neutralizing antibodies.
Based on these difficulties, we have learned two lessons: first, that an HIV vaccine relies on highly selective envelope structures or domains that immunogenically expose the Achilles' heels of the virus. To guide us in this design, we need to rely on the structural and genetic characteristics of the antibodies we want to induce to design immunogens capable of activating the B lymphocyte (LB) precursors of those antibodies. The second lesson is that, to generate potent antibodies, we need sequential immunizations with slightly different prototypes to induce maturation of the antibodies initially produced by the germline B precursor.
In four papers published in the Science group, two of them this week, immunogens/vaccines are generated that are targeted to activate B lymphocytes capable of producing antibodies with these characteristics and that, in particular, are directed against two areas of the viral envelope, the CD4 binding domain and the V3 region.
In this week's two papers in Science, William Schief's group at the Scripps Institution generates a protein (N332-GT5) that specifically activates germline B lymphocytes in primates, precursors of neutralizing antibodies of the BG18 type (a potent neutralizing antibody). These antibodies, which are directed against the V3 domain of the viral envelope, have a long HCDR3, despite the low proportion of this type of structure. Meanwhile, the group of Andrew Ward, also from Scripps, and Facundo Batista from Harvard University, go a step further and, after immunization with the same prototype (N332-GT5) in a mouse model, make two memories with slightly different immunogens (mRNA or proteins) (B11 and B16) to accelerate the somatic mutation that makes the antibodies gain affinity, and observe it. They conclude that in this model it is possible that the memories are increasing LB proliferation or recruitment of LB memory to germinal centers.
Although in very restrictive preclinical models, these papers represent a proof of concept that it is possible, on the one hand, to selectively activate germline precursor LBs that induce broad-spectrum antibodies and, on the other hand, by a booster strategy with slightly different proteins or mRNAs, to accelerate somatic maturation of antibodies. Phase I trials in volunteers will confirm whether the data from animal models are extrapolable to humans, something that has not always been observed.