by This comparison shows that our analysis provides more detailed mapping of sites that may be relevant to the overall bNAb sensitivity than has been previously assembled. Abstract == == Highlights == HIV-1 bNAb sensitivity signatures from 4 large virus panels mapped across 4 Ab classes Non-contact hypervariable region characteristics are critical for bNAb sensitivity HIV-1 Env 459C used alone as a vaccine can elicit modest tier 2 NAbs in guinea pigs V2 bNAb signature-guided modifications in 459C enhanced neutralization breadth HIV-1 Env amino acid signatures associated with sensitivity to broadly neutralizing antibodies were systematically defined from large neutralization panels. V2 signatures were incorporated in a trivalent vaccine to enhance epitope exposure and to include common epitope variants, resulting in increased neutralization breadth against heterologous viruses in a guinea pig model. == Introduction == Vaccine induction of broadly neutralizing antibodies (bNAbs) against diverse global tier 2 HIV-1 strains remains an unsolved challenge for the HIV-1 vaccine field. There has been some progress in animal models (Escolano et al., 2016,Saunders et al., 2017), but human trials have yet to elicit bNAbs, although NAbs with varying levels of breadth arise during natural infection (Hraber et al., 2014b). bNAbs typically develop slowly during chronic infection as the virus diversifies under immune pressure and B cell lineages adapt to the evolving virus (Bonsignori et al., 2017b,Doria-Rose et al., 2014,Liao et al., 2013,Wu et al., 2015). bNAb breadth and potency are evaluated using large panels of HIV-1 Envelope (Env) pseudoviruses that sample global HIV-1 diversity (Hraber et al., 2014a) or the C clade diversity of Southern Africa (Rademeyer et al., 2016). We used data from 4 large neutralization panels for a SL-327 more comprehensive mapping of viral signatures associated with bNAb sensitivity than undertaken previously (Chuang et SL-327 al., 2013,Evans et al., 2014,Ferguson et al., 2013,West et al., 2013). Signature sites were identified using a strategy that incorporates a phylogenetic correction (Gnanakaran et al., 2010) for amino acids (AAs) and potential N-linked glycosylation sites (PNGSs) (Crispin and Doores, 2015), and we also explored the impacts of hypervariable region characteristics and clades. Recurrent signature patterns were found among bNAbs with shared specificities (Burton and Hangartner, 2016). We next used variable V2 apex (V2) epitope bNAb signatures to inform HIV-1 Env immunogen design in a proof-of-concept exploration of an approach we call signature-based epitope targeted (SET) vaccines. Other vaccine design strategies include engaging bNAb germline precursors (Steichen et al., 2016), using polyvalent sets to capture diversity (Korber et al., 2017), lineage-based designs (Bonsignori et al., 2017b), and engineered native-like Envs (e.g., SOSIPs) (Sanders et al., 2013,Steichen et al., 2016). SOSIP vaccines elicit robust autologous NAbs that have limited breadth in rabbits and non-human primates (NHPs) (Pauthner et al., 2017,Sanders et al., 2015). Our SET vaccine design started with the Env 459C (Bricault et al., 2015), as 459C alone elicited modest neutralization of some tier 2 heterologous strains in guinea pigs. V2-SET immunogens are a trivalent combination of 459C wild-type (WT) plus two additional proteins designed by modifying 459C to include V2 bNAb signatures intended to both enhance V2 epitope exposure and include relevant variation. V2-SET vaccines expressed as SL-327 either gp140 SOSIP trimers or foldon trimers elicited increased iNOS (phospho-Tyr151) antibody NAb breadth compared to 459C alone in guinea pigs, suggesting the potential utility of bNAb signatures in vaccine design. == Results == == Neutralization Data == Four datasets measuring the sensitivity of bNAbs against panels of HIV-1 Envs were analyzed. Three panels sampled global viral diversity (Hraber et al., 2014a), and the other sampled only C clade, which dominates in Southern Africa (Rademeyer et al., 2016).Table S1summarizes bNAb dataset inclusion, relationships, and provenance. bNAbs are grouped by epitope class: V2, V3 glycan (V3), CD4 binding site (CD4bs), and membrane proximal external region (MPER) (Burton and Mascola, 2015). V2 and V3 bNAbs often have great potency but limited breadth, CD4bs have expanded breadth, and MPER has high breadth but low potency (Figures 1,S1, andS2). Heatmaps displaying inhibitory concentrations of 50% (IC50) data illustrate shared pattern sensitivity across bNAb classes (Figures 1andS1), enabling the definition of common bNAb class signatures. == Figure 1. == Heatmaps Showing IC50Neutralization Titers for Dataset 4 Darker red hues indicate more potent neutralization and blue indicates undetected responses. Rows represent pseudoviruses, ordered differently in each panel to highlight commonalities in neutralization profiles across bNAbs in each class. The clade with the strongest clade effect is separated and indicated in green to the left. Key PNGSs are indicated by magenta. Among MPER bNAbs 2F5 is considered separately as it has a unique epitope. == Neutralization.
