Disruption of outer blood-retinal barrier by Toxoplasma gondii-infected monocytes is mediated by paracrinely activated FAK signaling

Abstract

Ocular toxoplasmosis is mediated by monocytes infected with Toxoplasma gondii that are disseminated to target organs. Although infected monocytes can easily access to outer blood-retinal barrier due to leaky choroidal vasculatures, not much is known about the effect of T. gondii-infected monocytes on outer blood-retinal barrier. We prepared human monocytes, THP-1, infected with T. gondii and human retinal pigment epithelial cells, ARPE-19, grown on transwells as an in vitro model of outer blood-retinal barrier. Exposure to infected monocytes resulted in disruption of tight junction protein, ZO-1, and decrease in transepithelial electrical resistance of retinal pigment epithelium. Supernatants alone separated from infected monocytes also decreased transepithelial electrical resistance and disrupted tight junction protein. Further investigation revealed that the supernatants could activate focal adhesion kinase (FAK) signaling in retinal pigment epithelium and the disruption was attenuated by FAK inhibitor. The disrupted barrier was partly restored by blocking CXCL8, a FAK activating factor secreted by infected monocytes. In this study, we demonstrated that monocytes infected with T. gondii can disrupt outer blood-retinal barrier, which is mediated by paracrinely activated FAK signaling. FAK signaling can be a target of therapeutic approach to prevent negative influence of infected monocytes on outer blood-retinal barrier.

Fig 1. Infected monocytes disrupt tight junction protein.

Human monocytic cells, THP-1, were labeled with deep red fluorescent cell tracker dyes and exposed to viable tachyzoites of Toxoplasma gondii expressing green fluorescent protein (PTG strain). (A) After the prepared inserts were turned upside down, THP-1 cells (10⁵ cells/well) were added to the basal side of the inserts. After incubation for 1 h, the inserts were placed back into the 24-well plates and incubated for another 1 h, and then transwell membranes were fixed, permeabilized, and immunostained with ZO-1 (red). The results of (B) THP-1 cells that were not infected (magenta) and (C) THP-1 cells that were infected with tachyzoites (magenta with green fluorescent dots around) were examined under the fluorescence microscope. Figures were selected as representative data from three independent experiments. Scale bar = 20 μm.

Fig 2. Infected monocytes decrease transepithelial electrical resistance.

THP-1 cells that were not infected (Uninfected) and THP-1 cells infected with T. gondii tachyzoites (Infected) were added to RPE monolayer (10⁵ cells/well). Transepithelial electrical resistance was measured before the treatment (0 h), and 3, 6, 12 and 24 hours after the treatment. Values of uninfected group at 0 h were normalized to 100%. Data were presented as the mean ± SEM of three independent experiments. *P<0.05.

Fig 3. Conditioned media from infected monocytes can also disrupt outer BRB.

(A) Transepithelial electrical resistance was measured before the treatment (0 h), and 3,6,12 and 24 h after the treatment of supernatants from THP-1 cells that were not infected (SN_Uninfected), THP-1 cells that were exposed to heat-inactivated tachyzoites (SN_Inactivated T. gondii) and THP-1 cells infected with live tachyzoites of T. gondii (SN_Live T. gondii). In addition, supernatants from infected THP-1 cells were boiled at 100°C for 20 min, and utilized (SN_ Live T. gondii (Boiled)). (B-D) Expression of ZO-1 was evaluated by immunocytochemical staining of ZO-1 (green) 6 h after the treatment of supernatants from THP-1 cells that were not infected (B), from THP-1 cells that were exposed to heat-inactivated tachyzoites (C) and from THP-1 cells infected with live tachyzoites of T. gondii (D). Data were presented as the mean ± SEM of five independent experiments. Figures were selected as representative data from three independent experiments. *, P<0.05. Scale bar = 20 μm.

Fig 4. FAK is activated by conditioned media from infected monocytes.

Representative Western blots of anti-pFAK, anti-FAK and anti-β-actin obtained with RPE cell lysates after treatment of conditioned media. RPE cells were incubated with standard medium, supernatants from uninfected THP-1 cells (SN_Uninfected) for 0.5 h, or supernatants from infected THP-1 cells (SN_Infected) for 0.5 and 4 h. β-actin served as loading control. Figures were selected as representative data from three independent experiments.

Fig 5. Inhibition of FAK signaling attenuates the disruption of outer BRB.

(A) Transepithelial electrical resistance was measured before the treatment (0 h) and 6 h after the treatment of supernatants from THP-1 cells that were not infected (SN_Uninfected) or from THP-1 cells infected with T. gondii (SN_Infected) with or without FAK inhibitor (PF-573228, 1μM). (B-D) Expression of ZO-1 was evaluated by immunocytochemical staining of ZO-1 (green) 6 h after the treatment of supernatants from uninfected THP-1 cells with FAK inhibitor (B), supernatants from infected THP-1 cells without FAK inhibitor (C) or with FAK inhibitor (D). (E) Representative Western blots of anti-pFAK, anti-FAK, anti-ZO-1, anti-occludin and anti-β-actin obtained with RPE cell lysates after treatment of conditioned media with or without FAK inhibitor for 6 h. β-actin served as loading control. Data were presented as the mean ± SEM of five independent experiments. Figures were selected as representative data from three independent experiments. *, P<0.05. Scale bar = 20 μm.

Fig 6. Blockade of CXCL8 can partly rescue disruption of outer BRB.

(A) Transepithelial electrical resistance was measured before the treatment (0 h) and 6 h after the treatment of supernatants from uninfected THP-1 (SN_Uninfected) or from infected THP-1 cells (SN_Infected) with a control IgG or with a neutralizing antibody against CXCL8 (Anti-CXCL8, 1 μg/mL). (B-D) Expression of ZO-1 was evaluated by immunocytochemical staining of ZO-1 (green) 6 h after the treatment of supernatants from infected THP-1 cells (B), and additional treatment with control IgG (C) or neutralizing antibodies against CXCL8 (D). Data were presented as the mean ± SEM of three independent experiments. Figures were selected as representative data from three independent experiments. Scale bar = 20 μm.