Identification of CDRH3 loops in the B cell receptor repertoire that can be engaged by candidate immunogens.

A major goal for the development of vaccines against rapidly mutating viruses, such as influenza or HIV, is to elicit antibodies with broad neutralization capacity. However, B cell precursors capable of maturing into broadly neutralizing antibodies (bnAbs) can be rare in the immune repertoire. Due to the stochastic nature of B cell receptor (BCR) rearrangement, a limited number of third heavy chain complementary determining region (CDRH3) sequences are identical between different individuals. Thus, in order to successfully engage broadly neutralizing antibody precursors that rely on their CDRH3 loop for antigen recognition, immunogens must be able to tolerate sequence diversity in the B cell receptor repertoire across an entire vaccinated population. Here, we present a combined experimental and computational approach to identify BCRs in the human repertoire with CDRH3 loops predicted to be engaged by a target immunogen. For a given antibody/antigen pair, deep mutational scanning was first used to measure the effect of CDRH3 loop substitution on binding. BCR sequences, isolated experimentally or generated in silico, were subsequently evaluated to identify CDRH3 loops expected to be bound by the candidate immunogen. We applied this method to characterize two HIV-1 germline-targeting immunogens and found differences in the frequencies with which they are expected to engage target B cells, thus illustrating how this approach can be used to evaluate candidate immunogens towards B cell precursors engagement and to inform immunogen optimization strategies for more effective vaccine design. Author summary: The development of an HIV vaccine is a global health priority. An effective vaccine will need to elicit broadly neutralizing antibodies that can protect against diverse HIV strains. However, precursors of such broadly neutralizing antibodies that can be stimulated by a vaccine and subsequently matured into protective antibodies, are infrequent in the human immune repertoire. This study presents a combined experimental and computational approach to evaluate HIV vaccine candidates for their potential to engage precursors of broadly neutralizing antibodies. In particular, we focus on the ability of candidate vaccines to recognize precursors with distinctive features in their CDR H3 loop, a functional site that is essential for the neutralization function of some anti-HIV antibodies. Two HIV vaccine candidates were analyzed with the methods developed in this study. One of them was predicted to recognize precursors with diverse CDR H3 loops, while the other was found to only engage CDR H3 loops with a particular amino acid composition, which may limit its applicability. Therefore, this study presents a way to evaluate candidate immunogens and to inform molecular optimization strategies for more effective vaccine design. [ABSTRACT FROM AUTHOR]