Polyfunctional Tier 2-Neutralizing Antibodies Cloned following HIV-1 Env Macaque Immunization Mirror Native Antibodies in a Human Donor

Abstract

Vaccine efforts to combat HIV are challenged by the global diversity of viral strains and shielding of neutralization epitopes on the viral envelope glycoprotein trimer. Even so, the isolation of broadly neutralizing Abs from infected individuals suggests the potential for eliciting protective Abs through vaccination. This study reports a panel of 58 mAbs cloned from a rhesus macaque (Macaca mulatta) immunized with envelope glycoprotein immunogens curated from an HIV-1 clade C-infected volunteer. Twenty mAbs showed neutralizing activity, and the strongest neutralizer displayed 92% breadth with a median IC₅₀ of 1.35 μg/ml against a 13-virus panel. Neutralizing mAbs predominantly targeted linear epitopes in the V3 region in the cradle orientation (V3C) with others targeting the V3 ladle orientation (V3L), the CD4 binding site (CD4bs), C1, C4, or gp41. Nonneutralizing mAbs bound C1, C5, or undetermined conformational epitopes. Neutralization potency strongly correlated with the magnitude of binding to infected primary macaque splenocytes and to the level of Ab-dependent cellular cytotoxicity, but did not predict the degree of Ab-dependent cellular phagocytosis. Using an individualized germline gene database, mAbs were traced to 23 of 72 functional IgHV alleles. Neutralizing V3C Abs displayed minimal nucleotide somatic hypermutation in the H chain V region (3.77%), indicating that relatively little affinity maturation was needed to achieve in-clade neutralization breadth. Overall, this study underscores the polyfunctional nature of vaccine-elicited tier 2-neutralizing V3 Abs and demonstrates partial reproduction of the human donor's humoral immune response through nonhuman primate vaccination.

Graphical abstract

FIGURE 1.

Schematic of Env immunogen curation and rhesus sample generation. Vanderbilt Center for AIDS Research cohort study participant VC10014 developed modest neutralization breadth <2 y after primary infection (33). Env immunogens cloned from VC10014 were previously curated in rabbits as described (34). In brief, 50 full-length env genes capable of producing infectious pseudovirus were cloned from VC10014 plasma collected at nine time points spanning almost 6 y following primary infection. The evolution of viral quasispecies driving emerging plasma breadth was analyzed in silico and select Env variants were tested for immunogenicity in four groups of rabbits by coimmunizing i.m. with trimeric gp140 and intradermally with gp160 plasmid DNA. Two of the rabbit groups were immunized with either a single Env clone or with six Envs present immediately prior to and concurrent with the development of VC10014 plasma breadth (Early Breadth group) using the same immunogens throughout. The other rabbit immunization groups were sequentially immunized with quasispecies’ Env diverging longitudinally with a total of four immunizations. Binding and neutralization titers were strongest in the Early Breadth group, and this vaccine scheme was then employed on six rhesus macaques as described in Hessell et al. (40). Following four immunizations, macaque 25257 displayed the highest binding and neutralization titers, including against heterologous tier 2 viruses, and robust levels of Env-responsive lymph node germinal center T follicular helper cells (∼0.2% of CD3⁺CD4⁺ICOS⁺PD-1^hi) and was selected for Ab cloning.

FIGURE 2.

Vaccine-elicited mAbs efficiently neutralize heterologous clade B pseudoviruses. (A) Abs cloned from macaque 25257 single IgG⁺ B cells collected 4 wk post–final immunization (week 24) were screened for neutralization against autologous F8 and heterologous tier 1A SF162 pseudoviruses using the standard TZM-bl reporter assay. The proportion of mAbs showing neutralization activity with IC₅₀s ≤50 μg/ml against either virus is shown. (B) Heat map of IC₅₀ values of mAbs tested against a panel of 13 pseudoviruses in the TZM-bl assay. Broadly neutralizing mAb VRC01 is shown as an intraassay reference control. All neutralization assays were performed with serial dilutions in duplicate and select mAbs were repeated twice to ensure accuracy. Cloned mAbs O.21–O.58 showed no activity against autologous F8 or heterologous SF162 or JRCSF and were not further tested. (C) Percentage breadth and (D) median IC₅₀ value against all detected susceptible viruses shown in (B).

FIGURE 3.

Potent NAbs target the CD4bs or V3 region. (A) All 58 mAbs were screened by ELISA for linear epitopes using a clade B consensus peptide pool. mAbs in positive wells were then screened against overlapping 15-mer consensus peptides to map the linear epitope as detailed in Supplemental Fig. 2. The remaining neutralizing mAbs O.2 and O.11 were scanned against gp41 and gp120, and epitopes for both were exclusive to gp120. (B) ELISA binding of VRC01, 2219, and O.2 to SF162 gp140 without blocking (solid lines) or after blocking with O.2 Fab′2. The complete competition binding panel is shown in Supplemental Fig. 3. (C) mAbs directed to the V3 region were tested for binding by ELISA against previously developed cradle (2219-like binding) and ladle (447-52D–like binding) mimotopes. For ELISAs against the cradle mimotope, a C4-binding mAb was used as a negative control.

FIGURE 4.

Neutralizing mAbs bind soluble stabilized trimers and infected rhesus PBMCs. (A) Heatmap of mAb EC₅₀ values determined by ELISA against the indicated gp140 (F8 and SF162) or gp120 (BaL and JRCSF). (B) Neutralizing mAb-binding kinetics to autologous (F8.V3.N7) and clade A heterologous (BG505) SOSIP trimers measured by biolayer interferometry. Values were calculated from kinetic traces of serial dilutions starting at 120 nm of analyte shown in Supplemental Fig. 4 as described in the Materials and Methods. (C) Comparative biolayer interferometry traces showing each mAb with 60-nm analyte measured simultaneously. (D). mAb binding of infected cells categorized by neutralization sensitive virus (red bars) or neutralization resistant virus (gray bars). Rhesus splenocytes were CD4 enriched by MACS, then stimulated with PHA and spinoculated with SHIV_BaL or SHIV_SF162P3. Seven days postinfection, mAb surface binding was examined by incubating cells with each cloned mAb or an isotype control followed by flour-conjugated secondary mAb, then intracellular stained with anti-p27/FITC and analyzed by flow cytometry. The gating scheme is presented at left and the dotted line indicates median fluorescence intensity of the isotype control. Symbols are colored to indicate Env targeting of each mAb. Significance was evaluated with a paired t test.

FIGURE 5.

Effector function correlates with neutralization potency. (A and B) ADCC of macaque plasma at the time of cloning (week 24, left) and of the indicated cloned mAbs (right). NKR24 luciferase reporter cells were infected with SHIV_BaL (A) or SHIV_SF162P3 (B) and incubated with CD16⁺ KHYG-1 effectors without mAb (background) or with serial dilutions of the indicated mAbs. KHYG-1 cells expressing rhesus or human CD16 were used for plasma and mAb assays, respectively. ADCC activity is depicted as the normalized loss of luminescence, and dotted lines indicate the intraassay threshold for activity (50% for plasma, 80% for mAbs). (C) ADCP of HIV_SF162 gp120–coated beads. THP-1 monocytes were incubated with Ag-conjugated fluorescent beads without mAb (background) or with serial dilutions of the indicated mAbs. The phagocytosis score is shown as the percent bead⁺ cells multiplied by the median fluorescence intensity (percentage of phagocytosing × no. of internalized beads). (D and E) Correlations of mAb surface binding to infected cells, ADCC, and ADCP activity with neutralization IC₅₀ to the corresponding pseudovirus. Gray shading designates the experimental cutoff where no activity was detected. (A–E) Individual mAb data are colored to indicate corresponding Env epitope.

FIGURE 6.

Allele assignments, CDR3 regions, and SHM levels for O.1–O.58. IgH and IgL allele assignments, CDR3 regions, and SHM rates obtained by IgBlast assignment to an individualized IGHV database of 25257 and IMGT’s IGKV and IGLV databases. In L chain assignments, parenthesis indicate alternative assignments with the same homology percentage. Symbols denote mAbs belonging to the same clonotype.

FIGURE 7.

Gene assignments and V region SHM rates. (A) Alleles for the 58 cloned mAbs are shown with corresponding neutralizing activity against at least one of 13 tested pseudoviruses and (B) against their cognate epitopes. The number of mAbs belonging to each clonotype is shown as an insert, where unlabeled slices in the pie equal 1 mAb. Each clonotype is defined as having identical V and J allele assignments, identical HCDR3 length, and at least 80% HCDR3 aa homology. IgH gene assignments were determined using an individualized IgH germline database generated for macaque 25257 by IgDiscover. (C) Prevaccine “baseline” allele usage in macaque 25257 inferred from frequency of barcoded total RNA compared with mAb-encoding alleles. (D) Number of amino acids comprising the CDR-3 (CDR3) of IgH or (E) percent V region nucleotide SHM for each mAb categorized by neutralization activity and compared with a multiple t test (**p < 0.01). Statistical conclusions were not made for the IgL because CDR3 and SHM rates were approximated from the IMGT database instead of an individualized database.