In-solution virus capture assay helps deconstruct heterogeneous antibody recognition of human immunodeficiency virus type 1

Abstract

Human immunodeficiency virus type 1 (HIV-1) envelope glycoprotein (Env) on whole virions is heterogeneous, so molecular analysis of Env with monoclonal antibodies (MAbs) is challenging. Virus capture assays (VCAs) involving immobilized MAbs are typically used, but these assays suffer from immobilization artifacts and do not provide binding constants. Furthermore, we show here that certain HIV-1 neutralizing MAbs, including 2G12, 4E10, 2F5, Z13e1, and D5, will capture virion particles completely devoid of Env. We modified the VCA such that MAbs and virions are incubated in solution, and unbound MAbs are removed prior to the capture step. This modification nearly eliminated evidence of Env-independent binding by MAbs to virions and allowed determination of apparent affinity constants in solution. Three important qualitative observations were further revealed. First, neutralizing MAbs 2F5, 4E10, and Z13e1 against the membrane-proximal external region (MPER) of HIV-1 gp41 were found to capture virions efficiently only if a significant amount of uncleaved gp160 or synthetic MPER peptide was present. Second, we show how non-native forms of Env vary by Env genotype and that Env from HIV-1(JR-FL) is more homogeneously trimeric than that from HIV-1(JR-CSF). Third, we determined that Env containing all or parts of gp41, including uncleaved gp160, binds spontaneously to free virions. This exogenous Env is an indiscriminate molecular "bridge" between Env-specific Ab and virions and can affect VCA analyses, particularly using pseudotyped virions. Heterogeneity in Env from endogenous and exogenous sources might also subvert humoral immunity to HIV-1, so in-solution VCAs may help to dissect this heterogeneity for vaccine design purposes.

FIG. 1.

Env-deficient virus particles are captured by Ab using the classical (immobilized) VCA but not using an in-solution VCA. A panel of MAbs was used to capture HIV-1 virion particles that were pseudotyped with either cleavage-competent HIV-1_JR-FL gp160 plus VSV-G [Env(+)] or VSV-G alone [Env(−)]. The virions were used either unconcentrated in the immobilized VCA (A), 100-fold concentrated in the immobilized VCA (B), or 100-fold concentrated in the in-solution VCA (C). The MAbs used for capture include anti-HIV-1 MAbs 2G12, b12, 4E10, 2F5, Z13e1, D5 (αNHR), 8K8 (αNHR), and 7B2 (αPID). Virus equivalents captured were quantified using overlay of TZM-bl target cells and measuring luciferase activity in relative light units (RLUs). Background RLUs, typically, ∼10³ RLUs, were determined using DEN3 (an irrelevant, negative control MAb).

FIG. 2.

Dependence of virion capture efficiency on Ab concentration. HIV-1 Env(−) virus particles were captured using the immobilized (A) and in-solution (B) VCAs using incremental concentrations of MAbs. The x axis of the left panel shows the concentration of the capture MAb solution used to coat microtiter wells, and the actual amount of MAb immobilized on wells was not quantified.

FIG. 3.

Capture of Env-deficient virions by biotinylated 4E10 and 2G12 is not blocked using high concentrations of untagged MAb counterpart. Env(−) virions and various concentrations of untagged competitor MAb were mixed and then incubated with corresponding biotinylated capture MAb. The capture MAbs 2G12 (left) or 4E10 (right) were added at 50 μg/ml, a concentration previously determined to mediate a low level of Env(−) virion capture using the in-solution VCA with an immobilized secondary Ab (α-biotin). DEN3 was included as a competitor MAb to control against nonspecific competition from high concentrations of IgG.

FIG. 4.

Env heterogeneity on HIV-1 virion particles produced in 293T cells using molecular clone (MC) versus pseudotyped virions (PSV). (A) Relative levels of gp160 and gp120 associated with virions produced by transfection using an MC vector (pLAI-JRFL), an Env complementation vector (pcDNA-JRFL) plus HIV-1 backbone vector (pSG3ΔEnv), or pcDNA-JRFL alone were analyzed by SDS-PAGE and Western blotting, in the presence (+) or absence (−) of 100 mM DTT. The percent cleavage was calculated as the percentage of gp120 divided by the total Env in the lane, as determined by using a densitometer. (B) BN-PAGE analysis of samples from panel A. Samples were pretreated for 1 h at 37°C, or 57°C (to denature native Env trimers). Gels were blotted and stained with an anti-gp120 (b12, 2G12, and B4e8) MAb cocktail. (C) Molar ratio of p24 to gp120 in the samples, as estimated from ELISA using p24 and gp120_JR-FL standard curves. The results shown are the averages of six replicates. (D) In-solution virus capture of MC (pLAI-JRFL) and PSV (pcDNA-JRFL) virion particles. Virus samples, normalized for p24 content, were captured using broadly neutralizing MAbs (b12, 2G12, 4E10, 2F5, and Z13e1), a neutralizing anti-V3 MAb (B4e8), and a coreceptor binding site MAb (X5), as well as non-neutralizing MAbs to the PID (7B2), CD4bs (b6), and gp41 CHR (D50).

FIG. 5.

Env heterogeneity and Ab capture of HIV-1 produced in 293T cells using MCs LAI-JRFL and LAI-JRCSF. (A) Freshly prepared, virion-associated Env proteins, produced by transfection of pLAI-JRFL and pLAI-JRCSF MC plasmids, were analyzed using SDS-PAGE, as in Fig. 3, in the presence (+) or absence (−) of DTT. (B) BN-PAGE analysis of the virions used in panel A. Samples were treated for 1 h at either 37°C or 57°C, and blots were stained using anti-gp120 (b12, 2G12, and B4e8) and anti-gp41 (4E10, 2F5, and Z13e1) MAb cocktails. (C) In-solution VCA of the above virus samples using neutralizing (IC₅₀s < 300 nM) and non-neutralizing (IC₅₀s > 1,340 nM) MAbs. Virus samples were normalized for p24 content.

FIG. 6.

Fractionation of virions using iodixanol density gradient centrifugation diminishes virion-associated uncleaved gp160. PSVs produced by transfection of 293T cells using pcDNA-JRFL and pSG3ΔEnv were centrifuged over a 6 to 18% iodixanol gradient, and the collected fractions were assayed for infectivity (A) and for efficiency of Env cleavage by SDS-PAGE (B). The percent cleavage of gp160 was calculated as the percentage of cleaved gp120 relative to total Env. “ud” denotes lanes for which Env cleavage could not be determined because the signal was too weak.

FIG. 7.

Soluble Env and MPER peptides can associate with Env-deficient virion particles and mediate MAb-specific virion capture. (A) Purified, soluble gp120_JR-FL, gp41_HxB2, and gp140_CN54 were preincubated at 2 μM with Env(−) virions; the virions pelleted to remove unbound protein; and the in-solution VCA was performed with MAbs b12 and 7B2 and the irrelevant MAb, DEN3. (B) Soluble Env-containing supernatant from 293T cells transfected with pcDNA-gp160_JR-FL (Env only; no backbone vector) was incubated with unconcentrated Env(−) or “Bald” virions for 1 h at 37°C. Virions were pelleted, and the in-solution VCA was performed using MAbs b12, 7B2, and the irrelevant MAb, DEN3. The background level of virus capture, determined using DEN3, was subtracted from each value. (C) MPER peptides were preincubated with Env(−) virions at various concentrations as in panel A, and the in-solution VCA performed using MPER MAbs Z13e1 and 4E10.

FIG. 8.

Cross-competition of MAbs for virus capture reveals distinct Env species. (A) HIV-1_JR-FL PSVs and MCs were mixed with a constant concentration of untagged (competitor) MPER MAbs (200 μg/ml), the mixture was added to the biotinylated (“bio”) MAb counterpart, and the in-solution VCA was performed. The data are plotted as the percent inhibition of capture. (B) PSVs, produced using pcDNA-JRFL, were incubated at 37 or 57°C for 1 h. The in-solution competition VCA was performed with the gp120 MAbs b6 (non-neutralizing) and b12 (neutralizing), and captured virus equivalents were quantified by using p24 ELISA. The percent inhibition was calculated relative to the negative control MAb (DEN3).

FIG. 9.

Schematic showing virion-level phenomena that can comediate binding by Env-specific Abs to HIV-1 virions without interaction with the native trimer. (A) Avidity effects caused by overlaying concentrated virions onto clustered, immobilized MAbs can amplify weak off-target binding, leading to the capture of virions via the membrane or host cell proteins incorporated into virions. (B) HIV-1 with Env genotypes X and Y display different levels (and/or forms) of non-native Env that preexist on the surface of freshly produced virions, which are otherwise identical and bear efficiently processed Env with similar trimer stabilities. (C) Uncleaved gp160, released by transfected or infected cells, can spontaneously associate with the virion membrane via lipophilic sequences in gp41 and can act as a molecular “bridge” to mediate Ab binding to virions. Lipophilic regions of gp41 are denoted with an asterisk. A similar phenomenon occurs with gp41 and synthetic peptides corresponding to the MPER of gp41 and does not require transmembrane anchorage of Env components.