Wild-type vs. modification: which is better?

From Stamatatos Laboratory, Vaccine and Infectious Disease Division

Human immunodeficiency virus (HIV) attacks the body’s immune system, thus weakening one’s ability to fight infection. There is currently no vaccine against HIV, but there are now drugs that keep the virus in check and limit symptoms. Even with the availability of drug treatment, researchers are still seeking to develop a vaccine to stop the spread of virus. One method to achieve this goal is to elicit broad neutralizing antibodies (bNAbs), which recognize multiple strains of the virus and block entry into cells. There have been many attempts to elicit bNAbs against HIV, however development is ongoing. One strategy has been to use multiple doses of differing antigens to guide the immune system to generate bNAbs. It is thought that a naïve B cell will need to be coaxed to develop a bNAb, since known bNAbs exhibit many mutations from predicted naïve sequences. Achieving this development requires multiple vaccinations with varying antigens. In order to inform this process, Dr. Stamatatos and colleagues sought to better understand the B cell population that expands after vaccination with two HIV envelope (Env) variants. The findings, recently published in PLOS Pathogens, provide information that can be used to inform future immunogen design.

Rhesus macaques were immunized with two DNA- and two protein-based vaccinations, with either a wild-type class C Env (WT) or a form of the same Env missing the variable loops and three N-linked glycosylation sites (Core). The modifications of WT to Core were done to increase antibody binding at the CD4 binding site (CD4bs) location on Env, with the goal of triggering a more neutralizing B cell response. Since the virus binds to CD4 cells using this site, blocking it with an antibody may lead to neutralization. In this study, all vaccinated animals developed serum antibodies that recognized autologous virus. Most antibodies appeared to bind the gp120 fragment of Env, and a fraction of the Core antibodies recognized the CD4bs, which was not seen for WT. In neutralization assays using WT virus and viruses lacking the glycosylation sites (viruses lacking the variable loops are nonfunctional), no antibody was able to neutralize the WT virus or heterologous viruses. However, the viruses lacking glycosylation could be neutralized by the Core immunized group. This suggests that the Core-induced antibodies can bypass the variable loops and neutralize the virus when some glycosylation sites are removed.

To probe further into the B cell repertoire, Illumina MiSeq deep sequencing analysis was performed. These data demonstrated that the WT-vaccinated animals had a skewing of the IGHV (variable domains of the heavy chain) towards the IGHV3 family and to a lesser extent IGHV1, IGHV4, and IGHV7. Interestingly, the Core-vaccinated animals did not show an enrichment for particular IGHV genes, which suggests that there was not specific stimulation of one gene family and that vaccination with Core stimulates a broad response rather than an enriched, low diversity response (see figure).

Immunoglobin (Ig) heavy and light chain gene usage between immunogens. Graphs highlight the number of genes with a significant positive log fold change for each group. Image provided by Dr. Stamatatos.

To determine if Core immunization produced antibodies better suited for neutralization (or development into future neutralizers), antibodies of either the CD4bs group or non-CD4bs group were compared. Of total Core antibodies, 4.5-8.6% were CD4bs antibodies and only about 0.003-0.02% were neutralizing. Within this CD4bs population, 79 heavy chain and 31 light chains were fully sequenced from non-neutralizing binders. The majority of these sequences were IGHV3 or IGHV4 alleles. In the neutralizing population, 14 paired heavy and light chain sequences were isolated. Of the 14, six antibodies were clonally related (IGHV2-Korf1) and two had unique sequences. Looking closer, the CDRH3 or CDRL3 lengths were between 12-16  and 8-11 amino acids respectively. They also showed 2-9% somatic hypermutation, and like the serum data, did not neutralize WT virus but only the modified versions. From epitope mapping data, the group hypothesized that the neutralizing antibodies bound to gp120 around the CD4bs.

Overall, this study demonstrates that the B cell repertoire that is elicited by vaccination is dependent on the antigen given. The alleles selected for by each vaccination demonstrate that the WT antigen probably contains more epitopes than the modified Core antigen. However, Core seems to better present the CD4bs pocket, which is a good neutralizing target, as compared to WT. This knowledge and technique suggest that immunogen designs that favor reducing non-neutralizing epitopes on antigens may increase the affinity of the antibodies for neutralizing epitopes.

Yacoob C, Lange MD, Cohen K, Lathia K, Feng J, Glenn J, Carbonetti S, Oliver B, Vigdorovich V, Sather DN, Stamatatos L. 2018. B cell clonal lineage alterations upon recombinant HIV-1 envelope immunization of rhesus macaques. PLoS Pathog, 14(6), e1007120.

Funding provided by the National Institutes of Health.

Fred Hutch/UW Cancer Consortium faculty member Dr. Leonidas Stamatatos contributed to this research.