Ablation of the complementarity-determining region H3 apex of the anti-HIV-1 broadly neutralizing antibody 2F5 abrogates neutralizing capacity without affecting core epitope binding

Abstract

The identification and characterization of broadly neutralizing antibodies (bnAbs) against HIV-1 has formed a major research focus, with the ultimate goal to help in the design of an effective AIDS vaccine. One of these bnAbs, 2F5, has been extensively characterized, and residues at the apex of its unusually long complementarity-determining region (CDR) H3 loop have been shown to be crucial for neutralization. Structural studies, however, have revealed that the (100)TLFGVPI(100F) apex residues of the CDR H3 loop do not interact directly with residues of its core gp41 epitope. In an attempt to gain better insight into the functional role of this element, we have recombinantly expressed native 2F5 Fab and two mutants in which either the apical Phe100B(H) residue was changed to an alanine or the CDR H3 residues (100)TLFGVPI(100F) were replaced by a Ser-Gly dipeptide linker. Isothermal titration calorimetry (ITC) and competitive-binding enzyme-linked immunosorbent assays (ELISAs) rendered strikingly similar affinity constants (K(d) [dissociation constant] of approximately 20 nM) for linear peptide epitope binding by 2F5 Fabs, independent of the presence or absence of the apex residues. Ablation of the CDR H3 apex residues, however, abolished the cell-cell fusion inhibition and pseudovirus neutralization capacities of 2F5 Fab. We report competitive ELISA data that suggest a role of 2F5 CDR H3 apex residues in mediating weak hydrophobic interactions with residues located at the C terminus of the gp41 membrane proximal external region and/or membrane components in the context of core epitope binding. The present data therefore imply an extended 2F5 paratope that includes weak secondary interactions that are crucial for neutralization of Env-mediated fusion.

FIG. 1.

Recombinant 2F5 Fab constructs. (A) 2F5 light and heavy chain sequences of the variable region. “FWR” labels framework regions, whereas “CDR” marks complementarity-determining regions. 2F5 residues interacting with the core linear epitope inferred from the crystal structure (PDB ID 3D0L) are marked by asterisks. The residues mutated in 2F5 Fab mutants used in this study [negative-control T15(H)A and F100B(H)A and delta CDR H3] are indicated by arrows and a box, respectively. (B) Design of the 2F5 CDR H3 mutants. For the delta CDR H3 mutant, the Ser-Gly dipeptide linker bridges the distance between the two structured endpoints of amino acids T₉₉ and A_100G. The figure was generated with the PyMOL program (13) using the PDB crystal structure 3D0L. (C) SDS-PAGE gels and Western blots showing the levels of expression and purification of the 2F5 Fab constructs. HRP, horseradish peroxidase. (D) Circular dichroism measurements of the 2F5 Fab constructs show the correct folding of all Fabs, with the expected signal for a predominantly β-sheet secondary structure. Fab-s, Fabs; [θ], molar ellipticity; WT, wild type.

FIG. 2.

Isothermal titration calorimetry was performed. Isotherms of 2F5ep (⁶⁵⁶NEQELLELDKWASLWN⁶⁷¹) binding to wild-type (WT) 2F5 Fab (A) and 2F5 Fab delta CDR H3 (B) are shown. Upper panels, heat released upon consecutive injections of 12 μl of the 2F5ep solution (26 μM) into native or delta CDR H3 2F5 Fab (3 μM) in the calorimetric cell. Lower panels, integrated heats (squares) and nonlinear least-squares fitting (lines) of the data, with a 1:1 binding stoichiometry. The derived thermodynamic parameters of binding for these complexes, as well as the other 2F5 Fab mutants, are listed in Table 1.

FIG. 3.

Neutralization and cell-cell fusion inhibition assays were performed with the different 2F5 Fab constructs. In these assays, pseudovirus or effector cells were preincubated with the recombinant 2F5 Fab constructs and fusion events were monitored after incubation with TZM-bl target cells. In both cases, wild-type 2F5 Fab and negative-control T15(H)A are able to inhibit fusion events effectively, with approximate IC₅₀s of 10⁻⁷ M and 10⁻⁶ M in the pseudovirus infection and cell-cell mediated fusion assays, respectively. 2F5 Fab CDR H3 apex mutant F100B(H)A inhibits fusion 20-fold less than the native 2F5 Fab, whereas mutant delta CDR H3 is almost completely unable to inhibit Env-mediated fusion in the concentration range tested. (A) Pseudovirus infection assays. The infection of TZM-bl target cells was monitored by flow cytometry of the GFP fluorescence signal of infected cells. (B) Cell-cell fusion assays. CHO-Env effector cells were allowed to interact with TZM-bl target cells, and fusion events were measured by the number of nuclei in syncytia after incubation. Graphs in both panels display means ± standard errors of 6 measurements in 3 independent experiments. WT, wild type.

FIG. 4.

Direct and competitive ELISAs measuring the binding of recombinant 2F5 Fab constructs to two gp41 constructs were performed. (A) Direct ELISAs of the binding of the 2F5 Fab constructs to a glycosylated gp41 construct spanning residues 541 to 682 were performed. Clear differences in binding are observed for the 2F5 CDR H3 mutants compared to the binding of the native 2F5 Fab, with the delta CDR H3 mutant almost unable to bind this construct in the concentration range tested. (B) Direct ELISAs of the binding of the 2F5 Fab constructs to a nonglycosylated gp41 construct spanning residues 535 to 669 were performed. Differences in binding to this gp41 construct are also observed for the 2F5 CDR H3 mutants, although to a smaller extent than for the glycosylated gp41 construct. (C and D) Competitive ELISAs measuring the binding affinities of the different 2F5 Fab constructs for a glycosylated gp41 construct spanning residues 541 to 682 and a nonglycosylated gp41 construct spanning residues 535 to 669 were performed. In both cases, although saturation is not reached with the concentrations used and so an exact IC₅₀ cannot be determined, it is possible to observe that all 2F5 Fab constructs have similar apparent binding affinities. This discrepancy between direct and competitive ELISA results is thought to arise from the binding artifacts of the epitope adsorbed to the solid-phase ELISA plate (see text for a more complete discussion). From these results, we believe that competitive ELISA is a better experimental system for investigating the relative binding affinities of the 2F5 Fab constructs for various epitopes. Graphs show means ± standard errors. Fab-s, Fabs; WT, wild type.

FIG. 5.

Competitive ELISAs of the binding of the 2F5 Fab constructs with various epitopes were performed. Top panels show sample binding curves obtained from competitive ELISAs of the different 2F5 Fab constructs from which IC₅₀s are calculated; means ± standard errors are shown. Bottom panels summarize the IC₅₀s calculated from multiple experiments, which allow for a statistical analysis; the horizontal lines show the mean IC₅₀ values for multiple independent experiments, each indicated by a symbol. (A) The linear peptide ELDKWAS was used as a competitor. As seen from both binding curves and derived IC₅₀s, wild-type 2F5 Fab and its mutants bind similarly to this core epitope. (B) The linear peptide 2F5ep (⁶⁵⁶NEQELLELDKWASLWN⁶⁷¹) was used as a competitor. All 2F5 Fab constructs bind with similar apparent affinities to this epitope, confirming the ITC data. (C) In the experiment whose results are shown here, the hydrophobic linear epitope 2F5preTM (⁶⁵⁶NEQELLELDKWASLWNWFNITNWLWYIK⁶⁸³) was used as a competitor. In this case, significant differences in binding affinities are observed for the 2F5 Fab CDR mutants in comparison to the binding affinity of the wild-type 2F5 Fab. This suggests a possible role for CDR H3 residues in interacting with residues located at the C terminus of the gp41 MPER. Whether these CDR H3 apex-mediated interactions are specific or not cannot be inferred from these results (see text for a more complete discussion). (D) An extended epitope consisting of the 2F5preTM peptide inserted in POPC/Chol/PA LUVs was used as a competitor. Significant differences in binding affinities were observed for all 2F5 Fab CDR H3 constructs when interacting with this more complex epitope. Altogether, these results imply a role for residues located at the apex of the 2F5 CDR H3 in mediating interactions with residues located at the C terminus of the gp41 MPER or/and with membrane components in the context of core epitope binding. A one-way analysis of variance (ANOVA) followed by a Bonferroni post analysis test was used for the statistical analysis of IC₅₀s obtained for the different Fab constructs, and the significance levels where the P value are <0.05 are indicated on the graph. WT, wild type.

FIG. 6.

Direct and competitive ELISAs measuring the binding of recombinant 2F5 Fab constructs to membrane components were performed. (A) Direct ELISAs of binding of native 2F5 Fab and its mutants to lipid-coated plates (POPC, Chol, and PA) were performed. All Fabs show interaction with the coated membrane components. Interestingly, deletion of the 2F5 Fab CDR H3 loop appears to slightly increase the ability of this Fab to interact with coated membrane components (see text for discussion). These results are representative of three measurements performed in triplicate on different ELISA plates coated with different amounts of membrane components. (B) Competitive ELISAs probing the binding of Fabs to gp41-coated plates after a preincubation with POPC/Chol/PA (2:1:0.6) LUV competitor were performed. No competition for this membrane bilayer competitor is observed up to the high μM range for any 2F5 Fab constructs, demonstrating a low affinity of 2F5 Fab for membrane components independent of core epitope binding. Graphs show means ± standard errors. WT, wild type.