Antibody 10-1074 suppresses viremia in HIV-1-infected individuals

Abstract

Monoclonal antibody 10-1074 targets the V3 glycan supersite on the HIV-1 envelope (Env) protein. It is among the most potent anti-HIV-1 neutralizing antibodies isolated so far. Here we report on its safety and activity in 33 individuals who received a single intravenous infusion of the antibody. 10-1074 was well tolerated and had a half-life of 24.0 d in participants without HIV-1 infection and 12.8 d in individuals with HIV-1 infection. Thirteen individuals with viremia received the highest dose of 30 mg/kg 10-1074. Eleven of these participants were 10-1074-sensitive and showed a rapid decline in viremia by a mean of 1.52 log10 copies/ml. Virologic analysis revealed the emergence of multiple independent 10-1074-resistant viruses in the first weeks after infusion. Emerging escape variants were generally resistant to the related V3-specific antibody PGT121, but remained sensitive to antibodies targeting nonoverlapping epitopes, such as the anti-CD4-binding-site antibodies 3BNC117 and VRC01. The results demonstrate the safety and activity of 10-1074 in humans and support the idea that antibodies targeting the V3 glycan supersite might be useful for the treatment and prevention of HIV-1 infection.

Conflict of interest statement

Competing Financial Interests Statement

The authors declare no competing financial interests.

Figures

Figure 1. Study design and pharmacokinetics of 10-1074 in healthy and HIV-1-infected individuals

(a) Schematic representation of the study design. (b) Serum levels of 10-1074 as determined by TZM.bl assay. Mean values for each dose group with SEM for uninfected (left) and HIV-1-infected individuals (right). Dotted grey lines at the bottom indicate lower limit of detection of the assays when performed with HIV-1 strains Du422.1 (top, 1.59 μg/ml), 3103.v3.c10 (middle, 0.43 μg/ml), and 3103.v3.c10 in an antiretroviral therapy-resistant backbone (bottom 0.26 μg/ml). Each sample was measured in duplicate. Serum half-life of 10-1074 between HIV-1-uninfected and -infected individuals was significantly different (p<0.0001). Samples demonstrating nonspecific activity were excluded from the analysis (Supplementary Table 3).

Figure 2. Viral load dynamics after 10-1074 infusion in HIV-1-infected participants

(a) 10-1074 dose of either 10 or 30 mg/kg is indicated on the left. Left graphs show absolute viral loads in HIV-1 RNA copies/ml (y-axis) vs. time in days after infusion (x-axis). Middle graphs demonstrate log10 changes in HIV-1 RNA levels from day 0. Right graphs show average log10 change in viremia after 10-1074 (red line) or 3BNC117 infusion (dotted black; curves were fitted by robust lowess regression with 40% of the data using MATLAB_R2016a). At the 30 mg/kg dose level, viremia was significantly suppressed from about day 3 to day 27 post infusion compared to viral load at day 0. The window of significant viral suppression was assessed by computing simultaneous confidence bands and determined when these excluded zero (Supplementary Figure 5). Stars indicate initiation of antiretroviral therapy. (b) Plot shows 10-1074 serum levels at the time of rebound for individuals receiving 10 or 30 mg/kg of 10-1074 (black or red circles, respectively). Dotted lines indicate mean time to rebound after 10-1074 infusion (black, 10 mg/kg, and red, 30 mg/kg). (c) Maximum log10 decline in viremia as measured by RNA copies/ml vs. 10-1074 IC80 (Spearman coefficient rho = −0.05, p=0.88) of primary culture virus from samples obtained 557 to 61 days prior to infusion as determined by TZM.bl assay. No sensitivity data were obtained from 1HC2, 1HD2 and 1HD10K prior to enrollment. Colors as in (b). (d) Maximum log10 decline in viremia in 10-1074-sensitive individuals vs. initial viral load as measured by RNA copies/ml (Spearman coefficient rho = −0.19, p=0.52). Colors as in (b).

Figure 3. Viral evolution after 10-1074 infusion in HIV-1-infected participants

(a) Maximum-likelihood phylogenetic tree of single genome-derived env gene sequences obtained from plasma before (day 0, grey) and 4 weeks after 10-1074 infusion (Week 4, red). Asterisks indicate nodes with bootstrap support of 100% (100 replicates). (b) Pie chart showing the frequency of resistance mutations found in circulating viruses by SGS before infusion for each subject. Grey indicates absence of potential resistance mutation at positions 325, 332 and 334. Colors correspond to mutations indicated in (c). For all pie charts the number of analyzed sequences is shown in the center. (c) Pie chart showing amino acid frequencies at three recurrently mutated 10-1074 contact sites for all pooled circulating virus sequences obtained by SGS 4 weeks after infusion. Outer rings indicate position of mutation (orange, 325; blue, 332; green, 334; grey, unmutated). (d) As in (c) but for each individual. Colors indicate the type of mutation. For 1HD6K and 1HD10K both week 4 and week 8 were included in (c) and (d). (e) Graph showing sensitivity to the indicated anti-HIV-1 antibodies of 114 different viral isolates obtained from 11 individuals before (grey, 55 isolates) and 4 weeks after 10-1074 infusion (red, 59 isolates) with IC80 values (μg/ml) on the y-axis (log10 scale). Each dot represents one viral isolate. Lines indicate geometric mean. Samples were run in duplicate.

Figure 4. Temporal evolution of escape from 10-1074 over time in subjects 1HB1, 1HB3, 1HC1, 1HD1, 1HD6K, and 1HD10K

Plots display relative frequencies of escape mutations observed in SGS data at envelope positions 325, 332 and 334 as shaded areas over time. Sequencing was performed on day 0 (all subjects), and at week 1 (1HB3, 1HD1), week 2 (1HB3), week 4 (all subjects), week 8 (1HD6K, 1HD10K), week 12 (1HD6K, 1HD10K), week 16 (1HB3), week 20 (1HD6K) and week 24 (1HB1, 1HC1, 1HD1) (see Supplementary Table 5 for absolute numbers). White line indicates serum concentration of 10-1074 as determined by TZM.bl assay (Supplementary Table 3). White circles without border depict 10-1074 serum levels below the limit of detection.

Figure 5. SMRT sequencing analysis

(a) Maximum-likelihood phylogenetic tree of full-length plasma envelope sequences obtained on day 0 and week 4 after 10-1074 infusion from subject 1HD1. Branches show high-quality consensus sequences (HQCSs) with their respective copy number visualized as the size of the colored circle (day 0, grey, and week 4, red). (b) Insets highlight two day 0 minority variants that carry 10-1074 escape mutations 324GKIR327 (blue, mutation in yellow) and 332NIR334 (green, mutation in yellow). (c) Table shows the number of filtered reads for the indicated sequence variants at positions 324–334 and the relative frequency of each variant. Residues 325 and 332–334 are shaded in grey. Deviations from the day 0 majority variant are highlighted in bold.