HIV Exposure to the Epithelia in Ectocervical and Colon Tissues Induces Inflammatory Cytokines Without Tight Junction Disruption

Abstract

Epithelial cells in human cervical and colonic mucosa do not express HIV receptor. However, HIV transmission occurs across the unbreached epithelia by an unknown mechanism. In this study, the effect of HIV exposure on tight junction (TJ) and cytokine production in ectocervical and colon mucosal epithelia in tissue biopsies was investigated in an organ culture model. After HIV exposure, the distribution patterns and quantities of epithelial TJ and adherens proteins were evaluated by immunofluorescence staining followed by confocal microscopy. Cytokine mRNA in the mucosal epithelia was also evaluated by real-time reverse transcription-polymerase chain reaction (RT-PCR). HIV transmission was evaluated by measuring p24 production in culture supernatant. Our results showed there were no significant changes in the distribution and quantities of epithelial TJ/adherens junction (AJ) proteins after exposure to HIV. However, higher levels of CXCL10 and CXCL11 mRNA expression were detected in HIV-exposed ectocervical epithelia. In case of colon mucosa, higher levels of CXCL10 and IL-6 mRNA expression were detected in HIV-exposed colon mucosa. Our study suggests that HIV induces cytokine production in epithelial cells, which may facilitate HIV transmission by recruiting HIV target cells in the submucosal region. Furthermore, HIV transmission may not occur through epithelial TJ/AJ disruption.

Keywords: HIV transmission; inflammatory cytokines; mucosal epithelia; tight junctions.

FIG. 1.

Effect of HIV exposure on the integrity of the cervical/colon mucosal epithelia: H&E staining of (A) ectocervical tissues and (B) colon tissues exposed to HIV (10⁶ infectious viral units) or control supernatant in the organ culture for 24 and 6 h, respectively. E, epithelium; H&E, hematoxylin and eosin; L, lumen of ectocervix. Magnification: ectocervical tissues 10×, colon tissues 40×.

FIG. 2.

Characterization of TJ and AJ in ectocervical tissues: (A) H&E image and (B) confocal image of ectocervical tissues incubated with fluorescent-labeled 3 kDa dextran for 1 h with/without prior exposure to EDTA for 2 h. The arrows on the right indicate that the dextran penetrated into epithelial layer of the ectocervical tissues exposed to EDTA. E, epithelium; L, lumen of ectocervix. Images were captured by confocal microscope. (C) Ectocervical epithelia were either exposed to control media or EDTA (10 mm) for 2 h. Tissue sections were stained for either ZO-1 or Claudin-1. Images were captured by confocal microscope. Original magnification of H&E image ×20 and confocal image ×40. AJ, adherens junction; TJ, tight junction; ZO, zona occludens.

FIG. 3.

Effect of HIV exposure on TJ proteins and AJ proteins in ectocervical tissues. (A) Ectocervical epithelia were either exposed to HIV (10⁶ infectious viral units) or control supernatant for 24 h. Ectocervical tissue sections were stained for ZO-1, Claudin-1, Claudin-4, E-cadherin, or Na/K/ATPase protein. Images were captured by confocal microscope. Original magnification ×40. The images shown are representatives of four independent experiments with different donors. Each donor had two control and two HIV biopsies with 5–10 random images obtained from each biopsy. (B) Ectocervical epithelia were exposed to HIV or control supernatant in an organ culture for 24 h. The tissues exposed to control supernatant were treated with or without 10 mM EDTA in the final 2-h incubation time. HIV- or control supernatant-exposed tissues (with or without EDTA) were further exposed to FITC-labeled dextran (3 kDa) for 1 h at 37°C in a CO₂ incubator. Images were captured by Nikon eclipse E600 microscope. Original magnification ×20. The arrows on the right indicate that the dextran penetrated into epithelial layer of the ectocervical tissues exposed to EDTA. E, epithelium; L, lumen of ectocervix. The images shown are representatives of three independent experiments with different donors. Each donor had two controls, two HIV-exposed, and two EDTA-exposed biopsies.

FIG. 4.

Quantitation of TJ and AJ proteins in ectocervical tissues. Fluorescence intensity of ZO-1 (A), Claudin-4 (B), E-cadherin (C), or Na/K/ATPase (D) proteins in ectocervical epithelia was normalized based on number of nuclei in the selected fields. Data shown are the average of fluorescence intensity in control supernatant- and HIV-exposed epithelium. (E) ZO-1 nuclear colocalization (shown as Pearson correlation coefficient between DAPI and ZO-1) in ectocervical epithelium. Data shown are the average of ZO-1 nuclear colocalization in control supernatant- and HIV-exposed epithelium. Bars represent mean ± standard error of four independent experiments with different donors.

FIG. 5.

Effect of HIV exposure on TJ proteins and AJ proteins in colon tissues. Human colon tissues were either exposed to HIV (10⁶ infectious viral units) or control supernatant for 6 h. Colon tissue sections were stained for ZO-1, Claudin-4, E-cadherin, or Na/K/ATPase protein. Images were captured by confocal microscope. Original magnification ×40. The images shown are representatives of four independent experiments with different donors. Each donor had two control and two HIV biopsies with 5–10 random images obtained from each biopsy.

FIG. 6.

Quantitation of TJ and AJ proteins in colon tissues. Fluorescence intensity of ZO-1 (A), Claudin-4 (B), E-cadherin (C), or Na/K/ATPase (D) proteins in colon mucosa was normalized based on number of nuclei in the selected fields. Data shown are the average of fluorescence intensity in control supernatant- and HIV-exposed tissues. Bars represent mean ± standard error of four independent experiments with different donors.

FIG. 7.

Visualization of epithelial junctions in the human colon tissues by transmission electron microscopy. (A) Colon tissues exposed to control supernatant for 6 h. (B) Colon tissues exposed to HIV for 6 h. Dark arrows show the TJs in colon columnar epithelia. White arrows show the HIV-like particles. Original magnification ×25,000.

FIG. 8.

HIV replication in colon/ectocervical tissue after exposure to HIV in vitro. (A) Human colon tissues (n = 3) or (B) human ectocervical tissues (n = 10) were inoculated ex vivo with HIV (10⁶ infectious viral units) for 6 or 24 h, respectively, then washed, and cultured for 16 days. The culture supernatant was tested for HIV p24 antigen production at different time points. Bars represent mean ± standard error of independent experiments with different donors (n = 3 for colon tissues, n = 10 for ectocervical tissues).

FIG. 9.

Cytokine gene expression in epithelial cells following HIV exposure in the ectocervical and colon tissues. (A) Human colon tissues (n = 7) were exposed to either HIV (10⁶ infectious viral units) or control supernatant for 6 h. (B) Human ectocervical tissues (n = 9) were exposed to either HIV (10⁶ infectious viral units) or control supernatant for 24 h. (C, D) Ectocervical tissues (n = 9–22) were exposed to HIV, AT-2-inactivated HIV, or GP120 for 24 h. Total RNA was extracted from microdissected mucosal layers followed by real-time RT-PCR for cytokine gene expression. Krt13, a housekeeping gene, was measured for internal control. Horizontal lines represent mean values of the fold changes in the samples. RT-PCR, reverse transcription–polymerase chain reaction.