HIV vaccine strategy -- targeting the CD4 binding site

Image Credit

Recently, an HIV vaccine paper describes the vaccination of four cows with an Env based vaccine. It suggests that if you have germline antibodies with long enough third heavy chain complementarity determining regions (HCDR3), things which are shorter in humans, a very simple immunization scheme can result in some very nice broadly neutralizing antibodies. How do we encourage the immune system's processes of somatic hypermutation and affinity maturation to favor inserts in the heavy chain or deletions in the light chain which might also facilitate antibody binding, thereby promoting the creation of broadly neutralizing antibodies?

This strategy is inspired in part by the same piece of information that inspired the glycosylation strategy of our HIV microbicide design. It is different in concept from other strategies being pursued to target the CD4 binding site while utilizing a sequential immunization strategy.

Some of the strongest phenotypic properties of HIV's Env structure linked to sexual transmission are shorter hypervariable domains and fewer potential N-linked glycosylation sites. This deficit in glycosylation sites in pioneer strains of HIV may be allowing better access to the conserved part of the CD4 binding site that the heavy chain of broadly neutralizing antibodies such as N6 bind to. During the course of an infection, the glycan coat becomes more full. This may create stress on a bound antibody that may serve to encourage the insertion or deletion mutations that might lead to the broadly neutralizing antibodies that are sometimes observed.

First we would like the heavy chain to get the best possible grip on the conserved part of the CD4 binding site in the absence of glycans that may be obscuring it. This is to maximize the probability that binding will survive the stress of adding back the glycans later. We utilize the principle of complimentary surfaces to create a heterodimeric immunogen designed to encourage somatic hypermutation to sharply focus and improve binding to the conserved part of the CD4 binding site. This heterodimer will likely require some additional designed-in glycans to prevent off-target immunogenicity. We expect that the heavy chain will be the part of the antibody that will be focused on that binding site in the majority of cases.

Now that the heavy chain has a good grip on things, we add back the glycans gradually so that the heavy chain doesn't lose its grip on the CD4 binding site. If things go well, one or another of the binding loops on the heavy chain will have evolved additional length, or maybe the light chain has elements that will experience deletions and become shorter, such that the antibody will then be able to efficiently bind the CD4 binding site in the presence of all of the glycans.

This is a fairly speculative plan. The insertion and deletion mutations are rare and could take a significant amount of time to observe. Something I find fascinating is the relationship between autoimmunity and the induction of broadly neutralizing antibodies in HIV patients. Elicitation of these kinds of antibodies may be facilitated by a compromised immune system, and it may be difficult to elicit them in a completely healthy person. An effective vaccine regimen may involve artificially inducing this kind of compromise.

So maybe one would like to do the whole thing in vitro, and once you have the B-cells that you want, put them back. Of course if you're going to pull out the B-cells and mess with them, you might as well just engineer the antibodies you want into them. Unfortunately, either way, this sort of defeats the purpose of a low cost solution.