Placental Hofbauer cells assemble and sequester HIV-1 in tetraspanin-positive compartments that are accessible to broadly neutralizing antibodies

Abstract

Introduction: Within monocyte-derived macrophages, HIV-1 accumulates in intracellular virus-containing compartments (VCCs) that are inaccessible to the external environment, which implicate these cells as latently infected HIV-1 reservoirs. During mother-to-child transmission of HIV-1, human placental macrophages (Hofbauer cells (HCs)) are viral targets, and have been shown to be infected in vivo and sustain low levels of viral replication in vitro; however, the risk of in utero transmission is less than 7%. The role of these primary macrophages as viral reservoirs is largely undefined. The objective of this study is to define potential sites of viral assembly, accumulation and neutralization in HCs given the pivotal role of the placenta in preventing HIV-1 infection in the mother-infant dyad.

Methods: Term placentae from 20 HIV-1 seronegative women were obtained following caesarian section. VCCs were evaluated by 3D confocal and electron microscopy. Colocalization R values (Pearson's correlation) were quantified with colocalization module of Volocity 5.2.1. Replication kinetics and neutralization studies were evaluated using p24 ELISA.

Results: We demonstrate that primary HCs assemble and sequester HIV-1(BaL) in intracellular VCCs, which are enriched in endosomal/lysosomal markers, including CD9, CD81, CD63 and LAMP-1. Following infection, we observed HIV-1 accumulation in potentially acidic compartments, which stained intensely with Lysotracker-Red. Remarkably, these compartments are readily accessible via the cell surface and can be targeted by exogenously applied small molecules and HIV-1-specific broadly neutralizing antibodies. In addition, broadly neutralizing antibodies (4E10 and VRC01) limited viral replication by HIV-1-infected HCs, which may be mediated by FcγRI.

Conclusions: These findings suggest that placental HCs possess intrinsic adaptations facilitating unique sequestration of HIV-1, and may serve as a protective viral reservoir to permit viral neutralization and/or antiretroviral drug entry in utero.

Keywords: HIV-1; Hofbauer cells; MTCT; VCCs; bNAbs; placenta.

Figure 1

HCs exhibit a distinct and variable morphology in culture. HCs were maintained in complete medium over six days, with or without HIV-1_BaL infection. The morphology of uninfected HCs were analyzed over time and presented as (a) bright field images, and (b) the proportion of each phenotype is displayed. Scale bar in panel a=11 µm. To analyze the subcelluar localization of HIV-1 in HCs, at day 0, along with day 2 and 6 post-infection, cells were fixed, permeabilized and labelled with primary antibodies against HIV-1 Gag (green, anti-CA183) and CD9 (red, anti-CD9) (c). Scale bar in panel c for day 0 and day 2=4.30 µm. Scale bar in panel c for day 6=11 µm. Image acquisition was performed with an Applied Precision Deltavision deconvolution microscope. Sections and bars shown represent a minimum of 30 cells for each condition from 10 donors.

Figure 2

In HIV-1-infected HCs, virus assembles in intracellular tetraspanin-rich compartments. Human Hofbauer cells infected with HIV-1_BaL were analyzed by confocal immunofluorescence microscopy three days post-infection. Cells were fixed, permeabilized and labelled with primary antibodies against HIV-1 Gag (green, anti-CA183), (a) tetraspanins (red, anti-CD9, anti-CD81, and anti-CD63) and (b) anti-LAMP-1 (red). (b) Cells were also incubated at 37°C with Lysotracker Red (LT-Red) and fixed, permeabilized and labelled with primary antibodies against HIV-1 Gag (green, anti-CA183). Scale bar=9.0 µm. (c) HIV-1 Gag colocalization and partial colocalization R values (using Pearson's correlation) were quantified with the colocalization module of Volocity 5.2.1. Image acquisition was performed with an Applied Precision Deltavision deconvolution microscope. Sections shown represent a minimum of 30 cells for each condition from 10 donors. Data in bar graphs are expressed as the mean+SE (R value) of triplicate sections from 10 donors. Asterisks (*p<0.01) indicate values that are significantly higher when compared to TfR (control).

Figure 3

Electron microscopy of HIV-1 infected HCs reveals viral budding at plasma membrane and within intracellular compartments. HIV-1_BaL infected HCs were fixed and analyzed by standard electron microscope processing procedures five days post-infection. Sections show intracellular virus-containing compartments (VCCs) with mature virions (red asterisk, a–h), budding profiles (solid arrows, c–e, i) and immature virions (open arrows, b). Mature viral particles were seen under folds (f and h) of the plasma membrane (PM) and in the extracellular space adjacent to the PM (g). Virus budding profiles were observed on the PM (i). Bars represent 0.6 µm for a, b, e and g; 1.5 µm for c and f; 0.6 µm for d and h; and 0.3 µm for i. Sections shown represent a minimum of 20 cells for each condition from four donors.

Figure 4

Low-molecular-weight dextrans can access the VCCs in HIV-1-infected HCs. HIV-1-infected HCs were incubated with Texas Red Dextran, 3000 MW, at 4°C or 37°C for 30 minutes. Cells were then fixed and stained for Gag (green). (a) Representative images of cells incubated at 4°C. (b) Representative image of cells incubated at 37°C. Scale bar=11 µm. Sections shown represent a minimum of 30 cells for each condition from 10 donors.

Figure 5

HC internal VCCs are accessible to antibodies. Human HCs were infected with HIV-1_BaL for three days. Infected HCs were immunolabelled at 4°C (a) or 37°C (b) prior to fixation/permeabilization and labelled with primary antibodies against tetraspanins (red, anti-CD9 and anti-CD81). Labelled HCs were then permeabilized and immunolabelled for HIV-1 Gag (green, anti-CA183). Image acquisition was performed with an Applied Precision Deltavision deconvolution microscope. Sections shown represent a minimum of 30 cells for each condition from 10 donors. Scale bar=7 µm. (c) HIV-1 Gag colocalization and partial colocalization R values (using Pearson's correlation) were quantified with the colocalization module of Volocity 5.2.1. in non-permeabilized cells at 4°C (white) and 37°C (grey), in comparison to permeabilized cells (black). Data in bar graphs are expressed as the mean+SE (R value) of triplicate sections from 10 donors. Asterisks (*p<0.01) indicate significant differences between the R values of non-permeabilized cells compared to permeabilized cells.

Figure 6

HIV-1-neutralizing antibodies access VCCs and inhibit HIV-1 replication in HCs in an FcγRI-dependent manner. (a) Human HCs were infected with HIV-1_BaL for three days. Infected HCs were immunolabelled at 37°C prior to fixation/permeabilization and labelled with primary antibodies against (a) FcγRI (red, anti-CD64), (b) 4E10 (red) and VRC01 (red). Labelled HCs were then permeabilized and immunolabelled for HIV-1 Gag (green). Scale bar=7 µm. Image acquisition was performed with an Applied Precision Deltavision deconvolution microscope. Sections shown represent a minimum of 30 cells for each condition from eight donors. Colocalization and partial colocalization R values (using Pearson's correlation) were quantified with the colocalization module of Volocity 5.2.1. (c) Neutralization activities of anti-HIV-1 neutralizing antibodies, VRC01 and 4E10, were determined in HCs at day 10 post-infection. Infected HCs were exposed to bNAbs alone. In competition experiments, HIV-1-infected HCs were also pre-incubated with the monoclonal antibody against FcγRI (anti-CD64). The percent of inhibition is defined as the reduction of p24 release in the supernatant of Ab-treated HIV-1-infected HCs compared with the control untreated HIV-1-infected HCs. Data in bar graphs are expressed as the mean+SE (% inhibition) of triplicate sections from eight donors. Asterisks (*p<0.01) indicate values that are significantly higher in VRC01- and 4E10-treated HCs compared to hIgG-treated controls. Symbols (*p<0.01) indicate significant differences due to the addition of anti-CD64.