Inhibition of HIV-1 infectivity and epithelial cell transfer by human monoclonal IgG and IgA antibodies carrying the b12 V region

Abstract

Both IgG and secretory IgA Abs in mucosal secretions have been implicated in blocking the earliest events in HIV-1 transit across epithelial barriers, although the mechanisms by which this occurs remain largely unknown. In this study, we report the production and characterization of a human rIgA(2) mAb that carries the V regions of IgG1 b12, a potent and broadly neutralizing anti-gp120 Ab which has been shown to protect macaques against vaginal simian/HIV challenge. Monomeric, dimeric, polymeric, and secretory IgA(2) derivatives of b12 reacted with gp120 and neutralized CCR5- and CXCR4-tropic strains of HIV-1 in vitro. With respect to the protective effects of these Abs at mucosal surfaces, we demonstrated that IgG1 b12 and IgA(2) b12 inhibited the transfer of cell-free HIV-1 from ME-180 cells, a human cervical epithelial cell line, as well as Caco-2 cells, a human colonic epithelial cell line, to human PBMCs. Inhibition of viral transfer was due to the ability of b12 to block both viral attachment to and uptake by epithelial cells. These data demonstrate that IgG and IgA MAbs directed against a highly conserved epitope on gp120 can interfere with the earliest steps in HIV-1 transmission across mucosal surfaces, and reveal a possible mechanism by which b12 protects the vaginal mucosal against viral challenge in vivo.

FIGURE 1

Schematic of the IgA₂ b12 expression vector. Plasmid pSM102 encodes the IgA₁ b12 V_H and V_L (κ) chain, GS, as a selectable-amplifiable marker in mammalian cells, and β-lactamase for selection in Escherichia coli.

FIGURE 2

Expression and purification of monomeric and dimeric forms of IgA2 b12. A, Nonreducing SDS-PAGE analysis of monomeric, dimeric, and polymeric forms of IgA₂ b12, as compared with IgG1 b12. Ab preparations were size-fractionated by nonreducing SDS-PAGE and stained with Coomassie blue. Monomeric IgA₂ b12 migrated with a molecular mass of ~150 kDa (filled arrowhead), whereas higher molecular mass forms (>220 kDa; open arrowheads) were present in the dimeric and polymeric IgA b12 preparations. A protein band of ~50 kDa was present in all IgA b12 preparation (*), and likely corresponds to κ L chain dimers that dissociate from intact Ig (52). IgG1 b12 migrated with a molecular mass of ~160 kDa. B, Separation of dimeric and monomeric forms of IgA₂ b12 by FPLC gel filtration. As described in Materials and Methods, Abs were eluted from by isocratic elution on HiPrep 16/60 Sephacryl S-300 High Resolution column at a flow rate of 0.5 ml/min. Peaks containing dimeric IgA₂ b12 (50.02 peak) and monomeric IgA₂ b12 (61.52 peak) were determined by ELISA and SDS-PAGE. The following molecular mass standards were used to calibrate the column: dextran 2000 kDa, ferritan 440 kDa, catalase 220 kDa, and RNase A 10 kDa. C, Anti-J chain Western blot analysis. Monomeric, dimeric, and polymeric forms of IgA₂ b12, as well as IgG1 b12, were subjected to nonreducing SDS-PAGE, transferred to nitrocellulose, and probed with an anti-human J chain Ab. The minor band observed in the IgG1 b12 lane is probably due to spill over from the adjacent pIgA₂ b12 sample.

FIGURE 3

SC associates with polymeric IgA₂ b12. Microtiter plates were coated with polyclonal goat anti-human IgA Abs, and then overlaid with human secretory IgA (positive control), monomeric IgA₂ b12 (negative control), or dimeric IgA₂ b12 from CHO cell clones J2 or J10. The plates were then probed with human SC and then developed using anti-human SC Abs. Each data point represents the average of two microtiter wells tested in parallel.

FIGURE 4

Reactivity, avidity, and affinity of IgA₂ b12 for gp120. The relative reactivity, avidity, and affinity of different forms of purified IgA₂ b12 were compared with IgG1 b12. A, ELISA analysis examining the reactivity IgA₂ and IgG1 b12 with gp120_MN. Anti-human κ chain Abs were used as secondary reagents to allow direct comparison of reactivity of different b12 isotypes with gp120. B, Functional avidity of IgA₂ and IgG1 b12 for gp120_MN as determined by ammonium thiocyanate dissociation (see Materials and Methods). The AI (inset) is defined as the amount of ammonium thiocyanate (M) required to elute 50% of the Ab from gp120. C, The relative affinities of IgA₂ and IgG1 b12 for gp120_MN were determined indirectly by competitive inhibition. Abs were mixed with increasing concentrations of soluble gp120_MN before being applied to ELISA plates coated with the same Ag. The amount of soluble gp120 required to reduce Ab binding by 50% was defined as the competitive inhibition index (inset, left). The approximate K_D (inset, right) was determined as described in Materials and Methods.

FIGURE 5

Neutralization of HIV-1 infection of PBMCs by IgA₂ b12. Monomeric, dimeric, and polymeric (i.e., a mixture of monomers, dimers, and higher molecular mass polymers) IgA₂ b12, as well as IgG1 b12, at indicated concentrations were incubated with (A) HIV-1_BaL or (B) HIV-1_IIIb for 30 min before being mixed with activated PBMCs. Levels of p24 in PBMC culture supernatants were determined 6 days later by ELISA, as described in Materials and Methods. Viral p24 in the control wells were consistently >14 ng/ml, confirming that the control cells were productively infected with HIV. The mAb concentration (micrograms per milliliter) necessary to inhibit 50% of viral infection, referred to as the IC₅₀, is indicated in the inset.

FIGURE 6

Inhibition of HIV-1 transfer from epithelial cells to PBMC target cells by b12 mAbs. Cell-free HIV-1_BaL or HIV-1_IIIb preparations were incubated with IgG1 or IgA₂ b12, and then applied to confluent monolayers of ME180 cells (left, A and B) or Caco-2 cells (right, C and D). Transfer was assessed by the addition of activated PBMCs and measuring p24 levels in culture supernatants by ELISA. The IC₅₀ (micrograms per milliliter) is indicated in the inset table. Isotype-matched control Abs of irrelevant specificities showed no inhibition of HIV-1 transfer (data not shown).

FIGURE 7

b12 mAbs block inhibition of HIV-1 transfer from epithelial cells to PBMCs. Cell-free HIV-1_BaL or HIV-1_IIIb preparations applied to confluent monolayers of ME180 cells (left, A and B) or Caco-2 cells (right, C and D) were incubated with IgG1 or IgA₂ b12: 1) before epithelial exposure (+/−); 2) after epithelial exposure and washing but before addition of PBMCs (−/+), or both times (+/+), and then transfer was assessed by the addition of activated PBMCs and measuring p24 levels in culture supernatant by ELISA.

FIGURE 8

Inhibition of HIV-1 BaL internalization in ME180 cells by b12 mAbs. Increasing concentrations of IgG1 or IgA₂ b12 were mixed with HIV-1_BaL before application to ME180 cells. After 1 h at 37°C, the cells were washed to remove unbound virus and then lysed. Levels of p24 Ag in cell lysates were considered representative of the total amount of virus bound specifically to cell surfaces, plus the amount of virus that had been internalized (□). Alternatively, ME180 cells were treated with trypsin-EDTA (0.05%) for 8 min before lysis to effectively remove virus bound to the cell surfaces. Levels of p24 Ag in cell lysates from trypsin-treated cells were considered representative of the only amount virus internalized (■).