Periodic Reporting for period 4 - HIV B Cell Function (Understanding HIV-specific B cell function and viral immunogenicity)
Reporting period: 2022-12-01 to 2024-05-31
This project aimed to address two key interrelated research questions. Firstly, why do broadly neutralizing antibodies only develop in certain HIV-positive individuals? Secondly, do non-neutralizing antibodies limit the development of broadly neutralizing antibodies in HIV infection and immunization? These questions are important as while effective antiviral therapy can result in good health for people living with HIV, there is as yet no effective vaccine. Moreover, we lack a clear understanding of the immunological parameters which enable an effective broadly neutralizing response to develop. Therefore, there is no clear picture of the immune landscape HIV vaccines need to induce to reach the end goal of producing broadly neutralizing antibodies. Previous work in the field of HIV broadly neutralizing antibodies has shown that their development is a rare event both at the population level and within the B cell repertoire of an individual with broadly neutralizing sera. It has been noted that some virological parameters and peripheral T cell phenotypes are associated with increased broadly neutralizing antibody development, but none of these has been found to be predictive. This is most likely due to widespread immune dysregulation in untreated HIV infection where broadly neutralizing antibodies develop, and to the rarity of broadly neutralizing antibody B cells within the total repertoire. Therefore, the first key objective was to investigate the immunophenotypes of B cells associated with broadly neutralizing antibodies. To achieve this, we used a single cell antigen-specific approach to address why broadly neutralizing antibodies only develop in certain cases. Our second objective was to characterise non-neutralizing and strain-specific antibodies from the same HIV exposure as broadly neutralizing antibodies. This was because prior work in the HIV broadly neutralizing antibody field has focused on isolation of antibodies with remarkable breadth, little attention has been given to non-neutralizing or strain-specific antibodies from these individuals. In this project we directly compared the affinity of antibodies of different functionality, raised within the same individual, to decipher whether other epitopes act as decoys to distract the immune system from making broadly neutralizing antibodies. Finally, our third objective was to compare B-cell receptor (BCR) activation of broadly neutralizing antibodies, strain-specific neutralizing antibodies and non-neutralizing antibodies by HIV envelope following infection and in vaccination. Basic immunology research into B cell activation has shown that B cells encoding antibodies of higher affinity are more readily activated and so are the ultimate output of affinity maturation. However, these studies were mainly performed using model antigens/systems, so how this relates to complicated HIV antigens with multiple epitopes was unclear. Hence, in this project we examined the relationship between antibody affinity, epitope specificity and B cell activation in the context of HIV.
Our initial goal was to identify samples from people living with HIV with the best broadly neutralizing antibody activity and to map the epitopes underlying their neutralization breadth. The first step was to screen samples for breadth by assessing serum activity in a luciferase-based pseudotype neutralization assay against viral strains selected to represent global HIV diversity. We then identified a new broadly neutralizing antibody called ELC07 and produced a high resolution cryo-EM structure of it binding to the interface of the HIV envelope protein. In parallel we used flow cytometry to assess cell surface markers CD27 and CD21 to identify changes in memory B cell phenotypes when broadly neutralizing antibodies were produced. Surprisingly, memory B cells were predominantly resting memory (CD27+ CD21+, approximately 60%), with atypical memory (CD27- CD21-) the most infrequent phenotype. This finding gave rise to the concept that broadly neutralizing antibodies develop only in some individuals because their B cells are protected from the widespread effects of inflammation triggered by HIV viraemia, which normally disrupts the B cell compartment and prevents effective affinity maturation. These findings have been released as a pre-print on BioXriv and presented at the 13th International Symposium of IFReC. We have also successfully characterised broadly neutralizing antibodies, strain-specific and non-neutralizing antibodies from the same HIV exposure. Our work suggests that non-neutralizing and strain-specific antibodies are found alongside broadly neutralizing antibodies but that these functionally inferior antibodies do not compete for the same epitopes, nor have higher affinity for antigen. We generated B cell lines expressing non-neutralising, strain-specific and broadly neutralising antibodies as both IgM and IgG BCRs. To assess the relationship between affinity and activation, release of intracellular calcium was measured by flow cytometry following antigen stimulation of B cell lines. The broadly neutralizing antibodies were selected to cover the major broadly neutralizing epitopes on the HIV envelope proteinand all demonstrate nanomolar affinity for both Env antigens, with affinity constants ranging between 10-9 M and10-12 M, but activation of the cells showed no correlation with overall affinity. Thus, our results indicate that the expected plateau in the affinity/activation relationship is above the previously postulated affinity ceiling or that some other aspect of this antigen-BCR interaction is modulating B cell activation. These data are being prepared for publication and have been presented at conferences. We then went on to assess B cell activation after vaccination in HIV by considering the downstream effects of activation – namely, class switch, B cell phenotype, serological output and clonal diversity. We found that B cell responses can be functionally impaired despite effective treatment. Our data showed that the early delay in the production of neutralizing responses was associated with a higher level of atypical memory B cells, decreased switching to IgG and less clonal diversity. These data have been published in iScience and presented at a Keystone meeting.
Our finding that memory B cell phenotype can be unaltered during the development of broadly neutralizing antibodies despite the presence of high levels of systemic HIV (which normally disrupt memory B cell homeostasis) advances the field beyond the state of the art. It was thought plausible that the B cell dysfunction seen in uncontrolled HIV (when broadly neutralizing antibodies develop) was needed to produce these unusually highly mutated antibodies. Our work disproves this, showing B cell dysfunction is not necessary. Similarly, the data we generated showing that greater affinity isn’t always better for broadly neutralizing antibody development advances the field beyond the state-of-the-art. Generally, to date, vaccine design aims to generate higher affinity binding interactions between candidate vaccines and broadly neutralizing antibodies. Our data show that there is no linear relationship between higher affinity for HIV antibodies and B cell activation after a sufficient level of affinity has been achieved. This, finding suggests evaluation of vaccine candidates should focus more on activation of B cell activation and less on affinity.