Cell-cell transmission enables HIV-1 to evade inhibition by potent CD4bs directed antibodies

Abstract

HIV is known to spread efficiently both in a cell-free state and from cell to cell, however the relative importance of the cell-cell transmission mode in natural infection has not yet been resolved. Likewise to what extent cell-cell transmission is vulnerable to inhibition by neutralizing antibodies and entry inhibitors remains to be determined. Here we report on neutralizing antibody activity during cell-cell transmission using specifically tailored experimental strategies which enable unambiguous discrimination between the two transmission routes. We demonstrate that the activity of neutralizing monoclonal antibodies (mAbs) and entry inhibitors during cell-cell transmission varies depending on their mode of action. While gp41 directed agents remain active, CD4 binding site (CD4bs) directed inhibitors, including the potent neutralizing mAb VRC01, dramatically lose potency during cell-cell transmission. This implies that CD4bs mAbs act preferentially through blocking free virus transmission, while still allowing HIV to spread through cell-cell contacts. Thus providing a plausible explanation for how HIV maintains infectivity and rapidly escapes potent and broadly active CD4bs directed antibody responses in vivo.

Figure 1. Mode of virus transmission differentially steers susceptibility to entry inhibition.

(A) DEAE-Dextran is not required for effective cell-cell transmission of HIV-1_JR-FL to TZM-bl cells. Serial dilutions of JR-FL infected PBMC were incubated with TZM-bl cells in presence (black circles) or absence (red circles) of 10 µg/ml DEAE-Dextran. Infectivity was measured by enzymatic activity of the luciferase reporter (relative light units (RLU)). Each infected cell input was probed in triplicate. Error bars represent SD (standard deviation). One of four independent experiments is shown. (B) Omission of DEAE-Dextran as media supplement abolishes cell-free JR-FL infection of TZM-bl cells. Serial dilutions of cell-free JR-FL virus were used to infect the luciferase reporter cell line TZM-bl in presence (black squares) or absence (red squares) of 10 µg/ml DEAE-Dextran. Infectivity was measured by induction of the luciferase reporter (relative light units (RLU)). Each virus dilution was probed in quadruplicates. Bars represent SD . One of four independent experiments is shown. (C) Cell-cell transmission but not enforced contact between virus and target cell overcomes entry restriction. The infectivity of cell-free virus without enforced attachment to TZM-bl target cells (gravity sedimentation), or upon spinoculation, magnetic bead virus adsorption and during cell-cell transmission was assessed in presence (solid lines) or absence (dotted lines) of 10 µg/ml DEAE-Dextran. Infection was determined by measuring luciferase production after 48 h (recorded as RLU). Each virus dilution was probed in duplicates. Bars represent SD. One of three independent experiments is shown. (D) Inhibitory profiles of CD4-IgG2 and T-20 during cell-cell and cell-free transmission. TZM-bl target cells were either cocultivated with JR-FL infected PBMC (red circles, no DEAE) or cell-free virus (black squares, with 10 µg/ml DEAE) in the presence of increasing doses of CD4-IgG2 (left panel) or T-20 (right panel). Infection was determined by measuring luciferase production after 48 h and recorded as RLU. Red and black values denote IC50 (nM) of during cell-cell and cell-free transmission, respectively. Data points represent means of duplicates from three independent inhibition experiments. Bars represent SEM. Lines depict fitted dose response curves. (E) Decreased CD4-IgG2 sensitivity during cell-cell transmission is due to an inherent feature of cell-cell transmission. TZM-bl target cells were mixed with replication competent infected JR-FL^rc PBMC in the presence of CD4-IgG2 or T-20 (red bars) in medium lacking DEAE Dextran. Cell-free JR-FL^rc was either spinoculated (hatched bars), adsorbed by magnetic beads (checkered bars) or added without enforced adsorption (grey bars) onto TZM-bl target cells in medium containing DEAE Dextran in the presence of the inhibitor. Fold increases in IC50 of cell-cell compared to cell-free infection are indicated on top of the respective bars. Bars depict means of three independent experiments in duplicates. Lines denote SD. Inhibition of cell-cell transmission by CD4-IgG2 and T-20 (red bars) was significantly less efficient than blocking of cell-free virus (grey bars) infection (Student t-test, p<0.0001 in both cases). (F) Single round infection is highly resistant to CD4-IgG2 inhibition during cell-cell transmission. TZM-bl target cells (no DEAE) were co-cultivated with JR-FL pseudovirus transfected 293-T cells in the presence of CD4-IgG2 or T-20. Cell-free JR-FL^pp-lucAM was added to the TZM-bl (with 10 µg/ml DEAE) in the presence of both inhibitors. Fold increases in IC50 of cell-cell compared to cell-free infection are indicated on top of the respective bars. Bars depict means of three independent experiments performed in duplicates. Lines denote SD. Inhibition of cell-cell transmission by CD4-IgG2 and T-20 (red bars) was significantly less efficient than blocking of cell-free virus (grey bars) infection (Student t-test, p<0.0001 in both cases).

Figure 2. Markedly decreased sensitivity of HIV entry to gp120 directed inhibitors during cell-cell transmission.

(A–C) TZM-bl target cells were either infected with cell-free JR-FL^rc (black squares, with DEAE) or cocultivated with JR-FL^rc infected PBMC (red circles; no DEAE) and inhibition by cell directed (A), gp120 directed (B) and gp41 directed (C) antibodies and inhibitors studied. Infection was determined by measuring luciferase production after 48 h (recorded as RLU). Lines depict fitted results derived from three to five independent experiments in which each sample condition was performed in duplicates. Error bars depict SEM. Dotted lines indicate 50% inhibition levels. (D) Loss of inhibitory activity during cell-cell transmission. Loss of inhibitory activity during cell-cell transmission compared to cell-free transmission is depicted as fold difference of IC50 values determined from data depicted in Fig. 2A–C. A star (*) denotes where the respective inhibitor did not reach a 50% inhibition level at the highest concentration used. The highest concentration probed was used in these cases as minimum estimate to derive the fold differences in IC50 values.

Figure 3. CD4bs directed inhibitors loose while gp41 directed agents maintain activity during cell-cell transmission across divergent HIV-1 isolates.

(A) TZM-bl target cells were either infected with cell-free, replication competent viruses (black squares, with DEAE) or cocultivated with infected PBMC (red circles; no DEAE) and inhibition by the indicated antibodies and inhibitors studied. Virus isolates used (ADA, ZA110, ZA015 and ZA016) are indicated on top of the respective columns, inhibitors on the left of the respective rows. Infection was determined by measuring luciferase production after 48 h (recorded as RLU). Lines depict fitted results derived from two to three independent experiments in which each sample condition was probed in duplicates. Error bars depict SEM. Dotted lines indicate IC50 values. (B) Loss of inhibitory activity during cell-cell transmission. Loss of inhibitory activity during cell-cell transmission compared to cell-free transmission is depicted as fold difference of IC50 values determined from data depicted in Fig. 3A. A star (*) denotes where the respective inhibitor did not reach a 50% inhibition level at the highest concentration used. The highest concentration probed was used in these cases as minimum estimate to derive the fold differences in IC50 values.

Figure 4. Efficient inhibition of T-cell to T-cell transmission by gp41 directed inhibitors.

(A) Inhibition of T-cell to T-cell transmission. A3.01-CCR5 infected with JR-CSF or uninfected controls were co-cultured with A3.01-CCR5^rhTRIM5α target cells (GFP positive) in the presence of the indicated inhibitors or medium alone. Infection of target cells was assessed by intracellular Gag staining by flow cytometry. Percentages of rhTRIM5α expressing, HIV infected cells are indicated. One representative of two independent experiments is shown. (B) Comparison of cell-free and cell-cell inhibition in rhTRIM5α restricted A3.01-CCR5 cells. Inhibition of cell-cell (cc, red and orange symbols) and cell-free (cf, black symbols) transmission of virus isolates JR-CSF, SF162 and NL4-3 by inhibitors (CD4-IgG2, b12 and 447-52D: 50 µg/ml, 2F5, 4E10: 100 µg/ml, T-20: 5 µg/ml, see also Table S2) was studied. To probe cell-cell transmission infected A3.01-CCR5 were cocultured with A3.01-CCR5^rhTRIM5α target cells. To study free virus transmission cell-free virus preparations were used to infect non-restricted A3.01-CCR5 cells. Infection of target cells was assessed by intracellular Gag staining by flow cytometry as described in A). Infection achieved in absence of inhibitor was set to 100% and inhibitor activity expressed in relation to this value. Data depicted are means of two to seven independent experiments. (C) Comparison of cell-free and cell-cell inhibition in rhTRIM5α restricted PBMC. Inhibition of cell-cell (cc, red and orange symbols) and cell-free (cf, black symbols) transmission of virus isolates SF162 and NL4-3 by inhibitors (CD4-IgG2, VRC01, b12 and 447-52D: 50 µg/ml, 2F5, 4E10: 100 µg/ml, T-20: 5 µg/ml) was studied. To probe cell-cell transmission infected PBMC were cocultured with PBMC^rhTRIM5α target cells. To study free virus transmission cell-free virus preparations were used to infect non-restricted PBMC cells. Infection of target cells was assessed by intracellular Gag staining by flow cytometry as described in A). Infection achieved in absence of inhibitor was set to 100% and inhibitor activity expressed in relation to this value. Data depicted are means of two independent experiments in duplicates.

Figure 5. Attachment of virus is blocked by preventing gp120-CD4 interaction.

(A) Schematic illustration of the experimental set up used to analyze virus attachment. (B) Attachment of virus is driven by binding to CD4. Attachment of HIV to CD4 negative (HeLa, A2.01) and related CD4 positive cells (TZM-bl, A3.01-CCR5) as well as stimulated, CD8 depleted PBMC was studied using GFP-labeled virus (JR-FL^ppiGFP). The gray-shaded areas represent the fluorescent signal obtained by flow cytometric analysis of the respective cell line in the absence of HIV. The black lines indicate fluorescence intensity of bound JR-FL^ppiGFP. (C) Influence of entry inhibitors on HIV attachment. Activity of 2F5 (100 µg/ml), 4E10 (100 µg/ml) and CD4-IgG2 (50 µg/ml) to block attachment of GFP-labeled virus (JR-FL^ppiGFP) to A3.01-CCR5 cells is shown. Histograms of one representative of three independently performed experiments are shown. (D) Inhibition of HIV attachment by CD4bs and gp41 directed agents. Attachment (MFI of GFP signal) achieved in absence of inhibitor was set to 100% and inhibitor activity expressed in relation to this value. Data depicted are means of three independent experiments, error bars denote SEM. Left panel: Attachment of Vpr-GFP labeled TN8 virus (NL4-3 envelope) to PBMC. Middle panel: Attachment of GFP-labeled virus (JR-FL^ppiGFP) to A3.01-CCR5. Right panel: Attachment of GFP-labeled virus (NL4-3^ppiGFP) to A3.01-CCR5 cells (individual inhibitor concentration are listed in Table S2).

Figure 6. Post attachment activity of entry inhibitors.

(A) Schematic illustration of the entry assay using BlaM-Vpr labeled virions. (B) JR-FL^pp-BlaM entry into PBMC was studied in presence and absence of CD4-IgG2 (50 µg/ml). One of three individual experiments is shown. Fluorescence of uncleaved CCF2/AM was recorded at 520 nm, β-lactamase cleaved CCF2/AM denoting HIV entry at 447 nm. (C) Inhibition of virus entry. Fusion of JR-FL^pp-BlaM and NL4-3^rc-BlaM with PBMC or A3.01-CCR5 cells was monitored in presence and absence of the indicated entry inhibitors (Inhibitor concentration: CD4-IgG2 and b12: 50 µg/ml, 2F5 and 4E10 : 100 µg/ml, T-20 10 µg/ml). Grey and orange bars correspond to pre- and post-attachment conditions respectively as depicted in (A). Data shown are means of three independent experiments, error bars denote SEM.

Figure 7. Gp41 specific inhibitors have broad post-attachment activity.

(A) Schematic illustration of the luciferase reporter assay utilized to assess post attachment activity of inhibitors. (B) Post-attachment activity of virus directed inhibitors. A3.01-CCR5 cells were infected and treated with inhibitors before or after virus attachment as indicated in (A). (Inhibitor concentration listed in Table S2). Infection of env-pseudotyped, luciferase reporter viruses JR-FL^pp-lucAM (diamonds), SF162^pp-lucAM (circle), 6535^pp-lucAM (star), NL4-3^pp-lucAM (triangle) was determined after 48 h of culture by measuring luciferase production (recorded as RLU). Data depict means of pre attachment (black symbols) and post attachment activity (red and orange symbols) as % inhibition compared to untreated control. Means of three to six independent experiments are shown. (C) Post-attachment activity of cell directed inhibitors. A3.01-CCR5 cells were infected and treated with inhibitors before or after virus attachment as indicated in (A). (Inhibitor concentration listed in Table S2). Infection of env-pseudotyped, luciferase reporter viruses JR-FL^pp-lucAM (diamonds) and SF162^pp-lucAM (circle)was determined after 48 h of culture by measuring luciferase production (recorded as RLU). Data depict means of pre attachment (black symbols) and post attachment activity (red and orange symbols) as % inhibition compared to untreated control. Means of two to four independent experiments are shown.