Optimization of the Solubility of HIV-1-Neutralizing Antibody 10E8 through Somatic Variation and Structure-Based Design.

UNLABELLED: Extraordinary antibodies capable of near pan-neutralization of HIV-1 have been identified. One of the broadest is antibody 10E8, which recognizes the membrane-proximal external region (MPER) of the HIV-1 envelope and neutralizes >95% of circulating HIV-1 strains. If delivered passively, 10E8 might serve to prevent or treat HIV-1 infection. Antibody 10E8, however, is markedly less soluble than other antibodies. Here, we describe the use of both structural biology and somatic variation to develop optimized versions of 10E8 with increased solubility. From the structure of 10E8, we identified a prominent hydrophobic patch; reversion of four hydrophobic residues in this patch to their hydrophilic germ line counterparts resulted in an ∼10-fold decrease in turbidity. We also used somatic variants of 10E8, identified previously by next-generation sequencing, to optimize heavy and light chains; this process yielded several improved variants. Of these, variant 10E8v4 with 26 changes versus the parent 10E8 was the most soluble, with a paratope we showed crystallographically to be virtually identical to that of 10E8, a potency on a panel of 200 HIV-1 isolates also similar to that of 10E8, and a half-life in rhesus macaques of ∼10 days. An anomaly in 10E8v4 size exclusion chromatography that appeared to be related to conformational isomerization was resolved by engineering an interchain disulfide. Thus, by combining a structure-based approach with natural variation in potency and solubility from the 10E8 lineage, we successfully created variants of 10E8 which retained the potency and extraordinary neutralization breadth of the parent 10E8 but with substantially increased solubility. IMPORTANCE: Antibody 10E8 could be used to prevent HIV-1 infection, if manufactured and delivered economically. It suffers, however, from issues of solubility, which impede manufacturing. We hypothesized that the physical characteristic of 10E8 could be improved through rational design, without compromising breadth and potency. We used structural biology to identify hydrophobic patches on 10E8, which did not appear to be involved in 10E8 function. Reversion of hydrophobic residues in these patches to their hydrophilic germ line counterparts increased solubility. Next, clues from somatic variants of 10E8, identified by next-generation sequencing, were incorporated. A combination of structure-based design and somatic variant optimization led to 10E8v4, with substantially improved solubility and similar potency compared to the parent 10E8. The cocrystal structure of antibody 10E8v4 with its HIV-1 epitope was highly similar to that with the parent 10E8, despite 26 alterations in sequence and substantially improved solubility. Antibody 10E8v4 may be suitable for manufacturing.

Eliminating antibody polyreactivity through addition of N-linked glycosylation.

Antibody polyreactivity can be an obstacle to translating a candidate antibody into a clinical product. Standard tests such as antibody binding to cardiolipin, HEp-2 cells, or nuclear antigens provide measures of polyreactivity, but its causes and the means to resolve are often unclear. Here we present a method for eliminating antibody polyreactivity through the computational design and genetic addition of N-linked glycosylation near known sites of polyreactivity. We used the HIV-1-neutralizing antibody, VRC07, as a test case, since efforts to increase VRC07 potency at three spatially distinct sites resulted in enhanced polyreactivity. The addition of N-linked glycans proximal to the polyreactivity-enhancing mutations at each of the spatially distinct sites resulted in reduced antibody polyreactivity as measured by (i) anti-cardiolipin ELISA, (ii) Luminex AtheNA Multi-Lyte ANA binding, and (iii) HEp-2 cell staining. The reduced polyreactivity trended with increased antibody concentration over time in mice, but not with improved overall protein stability as measured by differential scanning calorimetry. Moreover, glycan proximity to the site of polyreactivity appeared to be a critical factor. The results provide evidence that antibody polyreactivity can result from local, rather than global, features of an antibody and that addition of N-linked glycosylation can be an effective approach to reducing antibody polyreactivity.

Recognition of synthetic glycopeptides by HIV-1 broadly neutralizing antibodies and their unmutated ancestors.

Current HIV-1 vaccines elicit strain-specific neutralizing antibodies. Broadly neutralizing antibodies (BnAbs) are not induced by current vaccines, but are found in plasma in ∼20% of HIV-1-infected individuals after several years of infection. One strategy for induction of unfavored antibody responses is to produce homogeneous immunogens that selectively express BnAb epitopes but minimally express dominant strain-specific epitopes. Here we report that synthetic, homogeneously glycosylated peptides that bind avidly to variable loop 1/2 (V1V2) BnAbs PG9 and CH01 bind minimally to strain-specific neutralizing V2 antibodies that are targeted to the same envelope polypeptide site. Both oligomannose derivatization and conformational stabilization by disulfide-linked dimer formation of synthetic V1V2 peptides were required for strong binding of V1V2 BnAbs. An HIV-1 vaccine should target BnAb unmutated common ancestor (UCA) B-cell receptors of naïve B cells, but to date no HIV-1 envelope constructs have been found that bind to the UCA of V1V2 BnAb PG9. We demonstrate herein that V1V2 glycopeptide dimers bearing Man5GlcNAc2 glycan units bind with apparent nanomolar affinities to UCAs of V1V2 BnAbs PG9 and CH01 and with micromolar affinity to the UCA of a V2 strain-specific antibody. The higher-affinity binding of these V1V2 glycopeptides to BnAbs and their UCAs renders these glycopeptide constructs particularly attractive immunogens for targeting subdominant HIV-1 envelope V1V2-neutralizing antibody-producing B cells.

Chemical synthesis of highly congested gp120 V1V2 N-glycopeptide antigens for potential HIV-1-directed vaccines.

Critical to the search for an effective HIV-1 vaccine is the development of immunogens capable of inducing broadly neutralizing antibodies (BnAbs). A key first step in this process is to design immunogens that can be recognized by known BnAbs. The monoclonal antibody PG9 is a BnAb that neutralizes diverse strains of HIV-1 by targeting a conserved carbohydrate-protein epitope in the variable 1 and 2 (V1V2) region of the viral envelope. Important for recognition are two closely spaced N-glycans at Asn(160) and Asn(156). Glycopeptides containing this synthetically challenging bis-N-glycosylated motif were prepared by convergent assembly, and were shown to be antigenic for PG9. Synthetic glycopeptides such as these may be useful for the development of HIV-1 vaccines based on the envelope V1V2 BnAb epitope.

Structural basis for diverse N-glycan recognition by HIV-1-neutralizing V1-V2-directed antibody PG16.

HIV-1 uses a diverse N-linked-glycan shield to evade recognition by antibody. Select human antibodies, such as the clonally related PG9 and PG16, recognize glycopeptide epitopes in the HIV-1 V1-V2 region and penetrate this shield, but their ability to accommodate diverse glycans is unclear. Here we report the structure of antibody PG16 bound to a scaffolded V1-V2, showing an epitope comprising both high mannose-type and complex-type N-linked glycans. We combined structure, NMR and mutagenesis analyses to characterize glycan recognition by PG9 and PG16. Three PG16-specific residues, arginine, serine and histidine (RSH), were critical for binding sialic acid on complex-type glycans, and introduction of these residues into PG9 produced a chimeric antibody with enhanced HIV-1 neutralization. Although HIV-1-glycan diversity facilitates evasion, antibody somatic diversity can overcome this and can provide clues to guide the design of modified antibodies with enhanced neutralization.

Mining the antibodyome for HIV-1-neutralizing antibodies with next-generation sequencing and phylogenetic pairing of heavy/light chains.

Next-generation sequencing of antibody transcripts from HIV-1-infected individuals with broadly neutralizing antibodies could provide an efficient means for identifying somatic variants and characterizing their lineages. Here, we used 454 pyrosequencing and identity/divergence grid sampling to analyze heavy- and light-chain sequences from donor N152, the source of the broadly neutralizing antibody 10E8. We identified variants with up to 28% difference in amino acid sequence. Heavy- and light-chain phylogenetic trees of identified 10E8 variants displayed similar architectures, and 10E8 variants reconstituted from matched and unmatched phylogenetic branches displayed significantly lower autoreactivity when matched. To test the generality of phylogenetic pairing, we analyzed donor International AIDS Vaccine Initiative 84, the source of antibodies PGT141-145. Heavy- and light-chain phylogenetic trees of PGT141-145 somatic variants also displayed remarkably similar architectures; in this case, branch pairings could be anchored by known PGT141-145 antibodies. Altogether, our findings suggest that phylogenetic matching of heavy and light chains can provide a means to approximate natural pairings.

Antigenicity and immunogenicity of transmitted/founder, consensus, and chronic envelope glycoproteins of human immunodeficiency virus type 1.

Human immunodeficiency virus type 1 (HIV-1) vaccine development requires selection of appropriate envelope (Env) immunogens. Twenty HIV-1 Env glycoproteins were examined for their ability to bind human anti-HIV-1 monoclonal antibodies (MAbs) and then used as immunogens in guinea pigs to identify promising immunogens. These included five Envs derived from chronically infected individuals, each representing one of five common clades and eight consensus Envs based on these five clades, as well as the consensus of the entire HIV-1 M group, and seven transmitted/founder (T/F) Envs from clades B and C. Sera from immunized guinea pigs were tested for neutralizing activity using 36 HIV-1 Env-pseudotyped viruses. All Envs bound to CD4 binding site, membrane-proximal, and V1/V2 MAbs with similar apparent affinities, although the T/F Envs bound with higher affinity to the MAb 17b, a CCR5 coreceptor binding site antibody. However, the various Envs differed in their ability to induce neutralizing antibodies. Consensus Envs elicited the most potent responses, but neutralized only a subset of viruses, including mostly easy-to-neutralize tier 1 and some more-difficult-to-neutralize tier 2 viruses. T/F Envs elicited fewer potent neutralizing antibodies but exhibited greater breadth than chronic or consensus Envs. Finally, chronic Envs elicited the lowest level and most limited breadth of neutralizing antibodies overall. Thus, each group of Env immunogens elicited a different antibody response profile. The complementary benefits of consensus and T/F Env immunogens raise the possibility that vaccines utilizing a combination of consensus and T/F Envs may be able to induce neutralizing responses with greater breadth and potency than single Env immunogens.