Evaluating the role of non-HIV antigens in the initiation of QA013.2 HIV bnAb lineage

Delphine M. Depierreux¹, Kevin Sung², Lindsay A. Bergan³, Mackenzie Shipley¹, Paul Lampe³, Frederick A. Matsen IV^2,5, Julie Overbaugh^1,4*

¹Division of Human Biology, Fred Hutchinson Cancer Center, Seattle, WA, 98109, USA
²Computational Biology, Public Health Sciences Division, Fred Hutchinson Cancer Center, Seattle, WA, 98109, USA
³Translational Research Program, Public Health Sciences Division, Fred Hutchinson Cancer Center, Seattle, WA, 98109, USA
⁴Public Health Sciences Division, Fred Hutchinson Cancer Center, Seattle, WA, 98109, USA
⁵Howard Hughes Medical Institute
^*Correspondence

Abstract

Broadly neutralizing antibodies (bnAbs) are a major goal of HIV vaccine design due to their ability to recognize diverse viral strains. Antibodies typically achieve breadth through affinity maturation, a process where somatic mutations accumulate and increase affinity for a specific antigen. A central challenge in HIV, however, is that for some bnAbs, their unmutated naïve precursors often fail to bind the HIV epitopes recognized by their mature counterparts. Consequently, the antigenic stimuli that initiate these lineages remain elusive, complicating efforts to design vaccines capable of engaging precursor B cells and guiding their maturation towards breadth. Building on our previous identification of the HIV bnAb QA013.2 and evidence that its lineage acquired somatic mutations prior to HIV seroconversion, we explore the hypothesis that QA013.2 maturation is initiated by non-HIV antigens. Here, we test the reactivity of both the naive and mutated forms of this antibody lineage present prior to HIV infection against a broad, diverse panel of glycans, human proteins, and viral peptides. There was no evidence of clear binding to this collection of >114000 antigens. Our findings narrow the landscape of potential initiating antigens, suggesting that if early maturation is indeed driven by non-HIV exposure, the required antigenic motif is likely not a common antigen or it is characterized by low-affinity interactions that remain challenging to identify in standard high-throughput screens.

Overview

The code to reproduce the VirScan data analysis is provided in this repository. The input data, after alignment pipeline processing, is provided on Zenodo. Download the data.tar.gz file and untar in the repository root directory. Generated dataframes and plots are saved to the results/ directory.

Environment Setup

The software environment can be set up with mamba and pip:

git clone https://github.com/matsengrp/QA013.2-HIV-bnAb-analysis.git
cd QA013.2-HIV-bnAb-analysis
mamba env create -f environment.yaml
mamba activate qa013.2-env
git clone https://github.com/matsengrp/phippery
cd phippery/
git checkout 1.3.0
pip install .
cd -

Finding peptide hits for each antibody

python peptide_hits.py

CSV dataframes tabulating enrichment values and annotations for associated peptide sequence and viral species are saved to results/.

The tables in TableSX are

results/447-52D_enrichment_5uL_annotations.csv
results/c68-59_naive_enrichment_5uL_annotations.csv
results/QA013-2_naive_enrichment_5uL_annotations.csv
results/QA013-2_int_VHint6-VLint5_enrichment_5uL_annotations.csv
results/QA013-2_mature_enrichment_5uL_annotations.csv

Generate enrichment scatter plots of peptide hits

python plot_hits.py

A scatter plot for each antibody is saved (SVG, PNG, and PDF formats) to results/. These plots are presented in Figure 5.