HIV-1 exploits LBPA-dependent intraepithelial trafficking for productive infection of human intestinal mucosa

Abstract

The gastrointestinal tract is a prominent portal of entry for HIV-1 during sexual or perinatal transmission, as well as a major site of HIV-1 persistence and replication. Elucidation of underlying mechanisms of intestinal HIV-1 infection are thus needed for the advancement of HIV-1 curative therapies. Here, we present a human 2D intestinal immuno-organoid system to model HIV-1 disease that recapitulates tissue compartmentalization and epithelial-immune cellular interactions. Our data demonstrate that apical exposure of intestinal epithelium to HIV-1 results in viral internalization, with subsequent basolateral shedding of replication-competent viruses, in a manner that is impervious to antiretroviral treatment. Incorporation of subepithelial dendritic cells resulted in HIV-1 luminal sampling and amplification of residual viral replication of lab-adapted and transmitted-founder (T/F) HIV-1 variants. Markedly, intraepithelial viral capture ensued an altered distribution of specialized endosomal pathways alongside durable sequestration of infectious HIV-1 within lysobisphosphatidic acid (LPBA)-rich vesicles. Therapeutic neutralization of LBPA-dependent trafficking limited productive HIV-1 infection, and thereby demonstrated the pivotal role of intraepithelial multivesicular endosomes as niches for virulent HIV-1 within the intestinal mucosa. Our study showcases the application of primary human 2D immune-competent organoid cultures in uncovering mechanisms of intestinal HIV-1 disease as well as a platform for preclinical antiviral drug discovery.

Fig 1. Human primary intestinal epithelium transmit infectious HIV-1.

(A) Schematic representation of the establishment of confluent and non-permeable intestinal epithelium (IE) culture model using standardized trans-epithelial electrical resistance (TEER) and FITC-dextran paracellular quantitative assays, respectively, to study HIV-1 infection of human intestine in vitro. (B) Representative confocal images of top view (xz) and side view (xy) of the IE cultures resting on a 3.0 micron pore 24-well cell culture insert. F-actin (phalloidin) is shown in grey and nuclei (DAPI) in blue. Scale bar = 20 micron. Representative of n = 2 donors. (C) Integrity of IE cultures, determined by longitudinal TEER measurements in IE cultures at days 7, 10, and 14. Open circles represent individual gut donors, n  =  5. TEER value ≥200 Ohm.cm2 (dashed line) indicative of a confluent monolayer prior to HIV-1 inoculation. (D) Paracellular permeability of IE cultures at day 14, determined by 4 kDa FITC-conjugated dextran (FD4) permeation rate of monolayers 4h post-FD4 addition. Permeability is expressed as FD4 permeation rate: FD4 basolateral_{t = 4}(μg)/FD4 apical_{t = 0}(μg), empty inserts included as negative control. Data are mean ± SD of n = 5 donors, Student’s paired t-test, **** p<0.0001. (E,F) HIV-1 transmission by IE cultures, determined by intracellular p24 staining of U87.CD4.CCR5 cell line by flow cytometry analyses. IE cultures were apically exposed to HIV-1 NL4.3BaL for 72h. Subsequently, basolateral supernatant was collected and co-cultured with HIV-1 permissive U87.CD4.CCR5 target cells for 96h. (E) Representative flow cytometry plots and (F) quantification. n  =  5 donors, Student’s paired t-test, **p<0.01.

Fig 2. Prophylactic antiretroviral treatment does not suppress HIV-1 transmission by intestinal epithelial cells ex vivo.

(A). Schematic representation of lamina propria-derived cells or epithelium-derived cells harvested from ex vivo human intestine tissue to study the effect of direct-acting antiretrovirals on HIV-1 transmission. (B-G) HIV-1 transmission by primary intestinal lamina propria-derived or epithelium-derived cells, determined by intracellular p24 staining of U87.CD4.CCR5 cell line by flow cytometry analyses. (B,C) Lamina propria-derived cells or (D,E) epithelium-derived CD45^- sorted cells were pre-treated with nucleotide reverse-transcriptase inhibitor Tenofovir (TFV, 50, 25, 12.5 or 6.25 μM) for 2h or left untreated prior to 48h exposure to HIV-1 NL4.3BaL. (F) Lamina propria-derived cells or (G) epithelium-derived CD45^- sorted cells were pre-treated with CCR5-antagonist Maraviroc (MCV, 30 μM), nucleoside reverse-transcriptase inhibitor Emtricitabine (FTC, 1 μM), protease inhibitor Indinavir (IDV,1 μM), or a combination of TFV (25 μM) with either FTC (0.2 μM) or IDV (0.2 μM) for 2h or left untreated prior to exposure to HIV-1 NL4.3BaL for 48h. (B-G) Subsequently, cells were washed extensively to remove input virus, and then co-cultured with permissive U87.CD4.CCR5 cells for 72h. Representative flow cytometry plots of HIV-1 transmission by TFV-treated lamina propria-derived cells (B) or TFV-treated epithelium-derived cells (D), and quantification (C,E,F,G). n  =  3 donors, One-way ANOVA, **p<0.01, ***p<0.001. (F,G) Data are mean ± SD.

Fig 3. Co-culture of human intestinal epithelium with dendritic cells results in enhanced HIV-1 transmission.

(A) Schematic representation of the experimental protocol to generate intestinal epithelium-DC (IEDC) co-culture model. Confocal microscopy images are z-stacks of IE culture and IEDC co-culture. DC surface marker CD11c (yellow), F-actin (grey), and nuclei (blue), insert membrane (white dashed line). (B) Integrity of intestinal epithelial monolayers, determined by TEER measurements pre- (open circles) and post- (closed circles) co-culture with DCs, n = 4 donors, Student’s paired t-test, ns = non-significant. (C) Visualization of IEDC co-culture model by confocal microscopy analyses at the apical (top panel) or basolateral (bottom panel) side, and cross-sectional view (z-stack, middle panel). DC surface marker CD11c (yellow) and nuclei (blue). Dashed lines indicate the separation of apical and basolateral compartments by the insert membrane. Scale bar = 20 micron, representative of n = 3 donors. (D) Cross-sectional view (z-stack) and (E) 3D rendered visualization of colocalization of DC protrusions with viral capsid in IEDC co-culture model (white arrowheads), upon apical exposure to HIV-1 NL4.3BaL for 72h. DC surface marker CD11c (yellow), HIV-1 p24 capsid (red) and nuclei (blue). Scale bar = 20 micron, representative of n = 2 donors. (F) Schematic representation of enzymatic (Accutase)-based treatment of basolateral compartment from IEDC co-cultures, representative flow cytometry plots depict the immune cell fraction (CD45⁺), non-immune epithelial cell fraction (CD45^-) and DC fraction (CD45⁺CD11c⁺DC-SIGN⁺) detached from IEDC co-cultures, n = 3 donors, for downstream analyses of (G) DC activation and (H) HIV-1 infection of DCs. (G,H) IEDC co-cultures were apically exposed to HIV-1 NL4.3BaL for 72h followed by enzymatic treatment of the basolateral compartment to retrieve DCs. (G) Representative histograms showing the expression of canonical activation markers CD83 and CD86 of CD45⁺CD11c⁺DC-SIGN⁺ DCs upon HIV-1 infection as compared to uninfected, n = 3 donors. (H,I) HIV-1 infection of basolateral DC fraction (CD11c⁺p24⁺) and basolateral non-DC fraction (CD11c^-p24⁺), determined by flow cytometer. Data are mean ± SD of n  =  6 donors. (J,K) HIV-1 transmission by IE cultures and IEDC co-cultures, determined by intracellular p24 staining of U87.CD4.CCR5 cell line by flow cytometry analyses. IE cultures (open circles) or cognate IEDC co-cultures (closed circles) were apically exposed to HIV-1 NL4.3BaL for 72h. Subsequently, basolateral supernatants were collected and co-cultured with HIV-1 permissive U87.CD4.CCR5 target cells for 96h. (J) Representative flow cytometry plots and (K) quantification. n  =  5 donors, Student’s paired t-test, *p<0.05, **p <0.01.

Fig 4. Incorporation of DCs results in increased propagation of multiple HIV-1 variants across IEDC co-cultures.

HIV-1 transmission by IE cultures and IEDC co-cultures, determined by luciferase activity of TZM-bl reporter cell line. IE cultures (open circles) or cognate IEDC co-cultures (closed circles) were apically exposed to CCR5-tropic (R5) HIV-1 NL4.3BaL, CXCR4-tropic (X4) NL4.3, transmitter/founder (T/F) virus CHO58 or T/F virus THRO for 72h. Subsequently, basolateral supernatants were collected and co-cultured for 48h with R5/X4 HIV-1 permissive TZM-bl luciferase reporter cell line. Data were normalized to non-infected (NI) sample. n  =  4–5 donors, Ratio paired t-test, *p<0.05.

Fig 5. HIV-1 is internalized by intestinal epithelial cells without disruption of intestinal barrier function.

(A) Junctional complexes in either untreated IE cultures or upon 72h exposure to HIV-1 NL4.3BaL by confocal microscopy analyses. Tight junction protein zonula occludens-1 (ZO-1, violet) and adherens junction adhesion protein e-cadherin (E-Cad, yellow). Scale bar = 20 micron, representative of n = 2 donors. (B) Paracellular permeability of either non-infected (NI) intestinal epithelial monolayers or upon 72h exposure to multiple HIV-1 variants (R5) HIV-1 NL4.3BaL, CXCR4-tropic (X4) NL4.3, transmitter/founder (T/F) virus CHO58 or T/F virus THRO determined by 4 kDa FITC-conjugated dextran (FD4) permeation rate of empty insert or IE cultures at 4h post-FD4 addition. Permeability is expressed as FD4 permeation rate: FD4 basolateral_{t = 4}(μg)/FD4 apical_{t = 0}(μg). Dashed line (0.62) indicates the average of FD4 permeation rate from empty inserts. Data are mean ± SD of n = 4–5 donors, One-way ANOVA, ns = non-significant. (C,D) HIV-1 uptake by intestinal epithelial cells, determined by monitoring % GFP fluorescence by flow cytometry analyses. (C) Representative flow cytometry plots and (D) quantification of epithelial monolayers exposed for 24h or 48h to replication-competent, fluorescently-tagged HIV-1 (HIV-1 Gag-iGFP). Data are mean ± SD of n  =  3 donors. (E) Confocal imaging of internalized HIV-1 Gag-iGFP (yellow arrowheads) within intestinal epithelial cells 48h post-exposure. Representative of n = 2 donors. F-actin (grey), HIV-1 p24 capsid (green) and nuclei (DAPI, blue). Scale bar = 10 micron.

Fig 6. Intestinal epithelial cells harbor HIV-1 within enriched multivesicular late endosomes.

(A) Distribution of endocytic vesicles in either untreated IE cultures or upon 24h exposure to HIV-1 Gag-iGFP by confocal microscopy analyses. Endocytic vesicle markers (red); Early endosomes (EEA1, Early endosome antigen 1), Autophagosomes (LC3, Microtubule-associated protein 1A/1B-light chain 3), Recycling endosomes (Rab11a, Ras-related protein Rab11a), Multivesicular late endosomes (LBPA, lysobisphosphatidic acid), and nuclei (blue). Scale bar = 20 micron, representative of n = 3 donors. (B) Colocalization of endocytic vesicle markers in IE cultures upon 24h exposure to HIV-1 Gag-iGFP by confocal microscopy analyses. Endocytic vesicle markers (EEA1, LC3, Rab11a, LBPA, red), HIV-1 (green, filled arrowheads). Scale bar = 20 micron, representative of n = 3 donors. (C) Histograms of EEA1, LC3, Rab11a or LBPA and HIV-1 fluorescence intensities depicting the degree of overlap between HIV-1 and respective endocytic vesicle along the indicated 6 micron-long Region of Interest (ROI, as specified in B). (D) Pearson’s correlation coefficient for colocalization analyses between each endocytic vesicle marker and HIV-1, n = 3 donors, Two-way ANOVA, **** p <0.0001.

Fig 7. Neutralization of LBPA-mediated endocytic pathway curbs intestinal HIV-1 transmission in vitro.

(A,B) Distribution of multivesicular/late endosomes in IE cultures pre-treated with either neutralizing anti-LBPA antibody (15 μg/ml), or left untreated for 16 h, followed by 24h exposure to HIV-1 Gag-iGFP visualised by confocal microscopy analyses. (A) Multivesicular/late endosome marker (CD63, red) and nuclei (blue). Scale bar = 20 micron, representative of n = 2 donors. (B) Analyses of fluorescence intensity were performed at original magnification by measuring mean grey value with ImageJ software [22]. Open circles represent averages derived from 5 different fields of view, n  =  2 donors. (C-F) HIV-1 transmission by IE cultures or IEDC co-cultures prophylactically-treated with neutralizing anti-LBPA antibody. IE cultures and IEDC co-cultures were apically pre-treated with either 15 μg/ml anti-LBPA antibody, 15 μg/ml isotype antibody control or left untreated for 16 h, followed by exposure to HIV-1 NL4.3BaL for 72h. Subsequently, basolateral supernatants were collected and co-cultured with HIV-1 permissive U87.CD4.CCR5 target cells for 96h. HIV-1 transmission was determined by intracellular p24 staining of U87.CD4.CCR5 cell line by flow cytometry analyses. (C) HIV-1 transmission by IE cultures. Data were normalized to untreated samples and are mean ± SD of n  =  5 donors, One sample t-test, *p = 0.0196. (D) Schematic representation of IEDC co-culture model as drug screening platform for intestinal HIV-1 infection in vitro. (E) Representative flow cytometry plots and (F) quantification of HIV-transmission by IEDC co-cultures. n  =  5 donors, Two-way ANOVA, *p<0.05.