Distinct neutralization sensitivity between adult and infant transmitted/founder HIV-1 subtype C viruses to broadly neutralizing monoclonal antibodies

Abstract

Broadly neutralizing antibodies (bnAbs), passively administered or elicited through vaccination, are a promising strategy for novel HIV prevention, treatment or inducing ART-free remission. However, HIV diversity and evolution are a barrier to the efficacy of bnAbs and there is therefore an urgent need for continuous virus surveillance to identify bnAbs with optimal neutralization breadth and potency against transmitted/founder (TF) viruses, especially in high-burden regions. We determined the neutralization sensitivity of TF viruses isolated within seven days after first detection of heterosexually acquired infection from young women 18-23 years old (n = 39) and within 1 month after birth from in-utero infected infants (n = 21) from FRESH and Baby Cure cohorts respectively, in KwaZulu-Natal, South Africa, where HIV-1 subtype C predominates. Neutralization sensitivities of 47 viruses from FRESH and 21 viruses from Baby Cure were assessed against nine bnAbs targeting different regions on the HIV-1 Env trimer. HIV-1 env sequences within and between bnAb epitopes were compared with database. The bnAbs VRC07-523LS, CAP256-VRC26.25, PGDM1400, 10E8 and PGT151 displayed higher neutralization breadth and potency than other bnAbs against FRESH TF viruses (>70% coverage, starting concentration of 10 μg/ml). Furthermore, VRC07-523LS showed higher neutralization breadth and potency than other bnAbs against Baby Cure TF viruses (p = 0.02). Interestingly, CAP256-VRC26.25 and PGT151 had lower neutralization coverage against infant TF viruses (<60% coverage). Moreover, 40% of infants TF had escape mutations within the V2 loop compared to 28% observed in FRESH and these mutations may explain the observed differences in neutralization sensitivities. However, few mutations were observed in gp120-gp41 interface in both adults and infants. Our findings suggest that intervention studies may have to consider different antibody combinations in adult versus paediatric settings. Moreover, high transmission of escape variants in both vertical and heterosexual transmissions is of concern. This information may be important in the selection of bnAbs that will undergo clinical testing in subtype C settings.

Fig 1. The differences in viral load between women in the FRESH cohort, in-utero infected infants and their transmitting mothers from the Baby Cure cohort.

Viral load was measured in the first week of infection in the FRESH cohort, at birth or one month respectively in infants and during labour in transmitting mothers. Differences in Log viral load between FRESH participants and in-utero infected infants (A), FRESH and transmitting mothers (B), infants and transmitting mothers from Baby Cure cohort (C). Differences in CD4⁺ T cell counts between FRESH and transmitting mothers in Baby Cure cohort (D). Median viral load and CD4⁺ T cell counts were compared using Mann Whitney statistical test and * indicates p < 0,001.

Fig 2. Phylogenetic analysis of full-length HIV-1 Env transmitted/founder sequences generated from (A) FRESH participants (n

= 39) and (B) Baby Cure infants (n = 19). A maximum likelihood phylogenetic tree was constructed using 383 single genome sequences and a total of 46 env clones generated from representative single genome sequences. SGA sequenced are shown as red circles, TOPO clones shown as orange circles and codon-optimised clones are shown green circles. (*) denotes participants infected with more than one variant at the time of infection. All HIV-1 reference sequences (blue circles) representing genetic subtypes were obtained from the Los Alamos National Laboratory HIV Sequence Database included: Ref. A1.RW, Ref. A1.UG, Ref. A2.CM, Ref. A2.CD, Ref. D.CD, Ref. D.CM, Ref. B.FR and Ref. C.ZA.

Fig 3. Neutralization sensitivity of the transmitted/ founder viruses to nine broadly neutralizing antibodies that target the CD4 binding site (3BNC117, VRC01, VRC07-523LS), the V1V2 loop (CAP256-VRC26.25, PGDM1400), V3-supersite (PGT121, 10-1074), MPER (10E8) and gp120-gp41 interface (PGT151) in the FRESH cohort (A) and Baby Cure study (B).

The IC₅₀ titers are indicated as percentages. Red coloured squares indicate neutralization IC50 scores <0,1, orange coloured squares indicate neutralization IC₅₀ between 0,1- 0,9, Yellow squares indicate IC₅₀ titers between 1,0-3,0 and light yellow colour indicates IC₅₀ titers between 3,1-10µg/ml. Clear squares indicate IC₅₀ titers>10. The tier phenotyping for neutralization sensitivity or resistance of the viruses tested are not available.

Fig 4. Neutralization coverage of transmitted/founder viruses produced from the FRESH cohort indicated by F and Baby Cure as indicated by B.

Each virus is represented by an individual dot, the highest concentration of each antibody tested against all the pseudoviruses was 10μg/ml. Solid lines represent the median titers. Percentage IC₅₀ neutralization coverage is shown above each scatter plot. Significant differences were observed in VRC07-523LS and 10E8, statistical significance was determined using Mann Whitney t-test, which does not correct for multiple comparisons. * indicates p < 0.05.

Fig 5. Predicted neutralization coverage of triple bnAb combinations against FRESH (A-C) and BABY CURE (D-F) transmitted/ founder viruses (IC80 < 1

μg/ml). Neutralization coverage was predicted using the Bliss Hill model, neutralization coverage on single bnAbs is shown in black, while neutralization on triple combinations is shown in different shades of blue. Neutralization coverage on single “active” bnAb is shown in dark blue, two “active” bnAbs is shown in blue and three “active” bnAbs is shown in light blue. The dotted lines show neutralization coverage at 50% and 90%.

Fig 6. Sequence logo plots T/F viruses from FRESH and Baby Cure studies showing the important amino acid positions within the neutralizing epitopes targeting the CD4 binding site (A, B), V2-glycan (C, D); V3-glycan (E, F), proximal (G, H) and distal MPER (I, J) regions and Gp120-gp41 interface (K, L).

The size of each AA in the logogram is proportional to its frequency and the “O” represents N-linked glycans. The numbering of each AA corresponds to the residue positions of HXB2. Residues associated with a sensitive neutralization phenotype are in blue, residues in red represent resistant residues while green represents differences in AAs between antibodies of the same class.

Fig 7. Amino acid length of the V1 Loop and Potential N-linked glycosylation sites (PNGS) in the V1 Loop on HIV-1 TF variants isolated from FRESH and Baby Cure cohorts in the first week or one month of infection.