Dual contribution of TRPV4 antagonism in the regulatory effect of vasoinhibins on blood-retinal barrier permeability: diabetic milieu makes a difference

Abstract

Breakdown of the blood-retinal barrier (BRB), as occurs in diabetic retinopathy and other chronic retinal diseases, results in vasogenic edema and neural tissue damage, causing vision loss. Vasoinhibins are N-terminal fragments of prolactin that prevent BRB breakdown during diabetes. They modulate the expression of some transient receptor potential (TRP) family members, yet their role in regulating the TRP vanilloid subtype 4 (TRPV4) remains unknown. TRPV4 is a calcium-permeable channel involved in barrier permeability, which blockade has been shown to prevent and resolve pulmonary edema. We found TRPV4 expression in the endothelium and retinal pigment epithelium (RPE) components of the BRB, and that TRPV4-selective antagonists (RN-1734 and GSK2193874) resolve BRB breakdown in diabetic rats. Using human RPE (ARPE-19) cell monolayers and endothelial cell systems, we further observed that (i) GSK2193874 does not seem to contribute to the regulation of BRB and RPE permeability by vasoinhibins under diabetic or hyperglycemic-mimicking conditions, but that (ii) vasoinhibins can block TRPV4 to maintain BRB and endothelial permeability. Our results provide important insights into the pathogenesis of diabetic retinopathy that will further guide us toward rationally-guided new therapies: synergistic combination of selective TRPV4 blockers and vasoinhibins can be proposed to mitigate diabetes-evoked BRB breakdown.

Figure 1

TRPV4 is expressed in both the inner and outer BRB, and its pharmacological inhibition prevents the streptozotocin-induced increase of BRB permeability in rats, similar to the effect of vasoinhibins. (A) Representative confocal stack image of transverse sections of Trpv4 ^+/+ and Trpv4 ^−/− mouse retinas showing TRPV4 (red), RPE-65 (green), and merge immunofluorescence. The anti-TRPV4 sc-98592 and LS-C94498 antibodies (Ab) were used as indicated and cell nuclei were stained with DAPI. Magnification bars were as indicated. Retinal pigment epithelium (RPE), outer segments (OS), outer nuclear layer (ONL), outer limiting membrane (OLM), outer plexiform layer (OPL), inner nuclear layer (INL), inner plexiform layer (IPL), and ganglion cell layer (GCL). Images were captured in three different regions of the same retina section (n = 3). Both retinas of three animals per group were analyzed. (B) Representative confocal image and corresponding projection in z-x (Z₁, right) and z-y (Z₂, bottom) of transverse sections of wild-type choroid-RPE (full-size image is shown in supplemental Fig. 3) showing RPE-65 (green) and TRPV4 (red) antibody immunofluorescence. Projections in z correspond to the area indicated by the yellow lines and show that TRPV4 and RPE-65 did not colocalize. (C) Representative confocal image and projection in z-x (Z₁, right) and z-y (Z₂, bottom) of transverse sections of wild-type OPL and INL (full-size image is shown in supplemental Fig. 3) stained with the anti-TRPV4 sc-98592 or LS-C94498 antibodies (red), the anti-mouse antibody coupled to fluorescein isothiocyanate as a marker of blood vessels (green), as previously reported, and DAPI (blue). (D) Representative images of whole mounts of Trpv4 ^+/+ and Trpv4 ^−/− mouse RPE stained with the LS-C94498 anti-TRPV4 antibody and DAPI. Projection in z-x (right) and z-y (bottom) is also shown. (E) Evaluation of the Evans blue dye content in retinas from control rats intravitreously injected with PBS, the TRPV4 antagonists RN1734 (100 µM) and GSK2193874 (100 nM) or vasoinhibins (Vi, 1 µM) for 24 h and from streptozotocin (STZ)-induced diabetic rats intravitreously injected with PBS, RN1734, GSK2193874 or Vi 24 h before the end of the 4 weeks of diabetes. Values are mean ± s.e.m. normalized to the control (n = 8–14 per group; *P < 0.05). n.s., not significant.

Figure 2

TRPV4 expression in ARPE-19 monolayers. (A) Representative phase-contrast microscopy and confocal images of ARPE-19 cultured on Transwell membrane inserts with pore sizes of 0.4 µm (see Methods) stained with DAPI and anti-TRPV4 sc-98592 and LS-C94498 antibodies (Ab), under control conditions (day 0) and after exposure to high glucose (D-glucose) for 28 days. Projection in z-x (right) and z-y (bottom) showed nucleus alignment and TRPV4 localization. Magnification bar was as indicated. (B) ARPE-19 cells cultured in 5.5 mM D-glucose (control), 5.5 mM D-glucose plus 19.5 mM mannitol (mannitol), and 25 mM D-glucose (D-glucose) for 28 days were analyzed for TRPV4 protein. Total β-actin served as loading control. Extracts from three independent ARPE-19 cell cultures in each condition were analyzed (N = 3); MWM, molecular weight markers. (C) Densitometric analysis of the TRPV4 fragment normalized to β-tubulin (control) expressed in arbitrary units (AU). Values correspond to mean ± s.e.m. for three independent experiments. n.s., not significant.

Figure 3

Dual contribution of TRPV4 antagonism and vasoinhibins to regulate the high glucose-induced effects on ARPE-19 monolayer resistance. (A) Time course of trans-electrical resistance (TER) in ARPE-19 cell monolayers cultured in 5.5 mM D-glucose (control), 5.5 mM D-glucose plus 19.5 mM mannitol (mannitol), and 25 mM D-glucose (D-glucose) over 28 days. (B) Quantification of viability levels in ARPE-19 monolayers treated as previously indicated at day 28. ARPE-19 cells were cultured on inserts with pore sizes of 0.4 µm. TER and MTT signals were normalized to the untreated condition. (B and D) Quantification of TER values in ARPE-19 monolayers after 3 (C) and 28 (D) days of culture in control, mannitol, and D-glucose conditions, in the presence and in the absence of GSK2193874 (50 nM), vasoinhibins (Vi, 10 nM) or both. *P < 0.05 from 3 independent experiments. n.s., not significant.

Figure 4

Vasoinhibins block TRPV4-induced BRB breakdown in rat retinas and TRPV4-induced Ca²⁺ transients and TER reduction in endothelial cell systems in endothelial cell systems. (A) Evaluation of the Evans blue dye content in retinas of control rats intravitreously injected with PBS or one of the TRPV4 agonists RN1747 (100 µM) and GSK1016790A (100 nM) in the presence and in the absence of vasoinhibins (Vi, 1 µM) for 4 h. Values are mean ± s.e.m. normalized to the control (n = 8–14 per group; *P < 0.05). (B) Epifluorescence and confocal images from HMVECs transfected with TRPV4-eGFP, where one can appreciate that TRPV4 is located at the membrane level. Magnification bars, 10 µm. (C) Representative measurements of the change in intracellular Ca²⁺ measured by the change in the fluorescence ratio (R340/380) in TRPV4-transfected HMEC with or without 10 µM 4α-PDD and 10 nM vasoinhibins (Vi). (D) Corresponding distribution of responding cells (% ± s.e.m.). *P < 0.05. (E) MRCEC cells were analyzed for TRPV4 protein. Total β-actin served as loading control. Extracts from three independent MRCEC cell cultures were analyzed (N = 3); MWM, molecular weight markers. (F) Time course of TER in MRCEC monolayers cultured in complete medium (control) with or without 10 µM 4α-PDD and 10 nM Vi. n.s., not significant.

Figure 5

Vasoinhibins block TRPV4-induced actin cytoskeleton redistribution in MRCEC cells in a NO-dependent manner. (A) MRCEC were cultured in complete medium (control) with or without 10 µM 4α-PDD and 10 mM L-NAME, and with the NO donor DETANONOate (10 µM) in the presence and in the absence of 4α-PDD and Vi (10 nM) for 5 min (corresponding to the peak TER values measured in panel D in the presence of 4α-PDD), and then actin cytoskeleton (F-actin) distribution was determined using rhodamine-phalloidin. Representative fields are shown. Scale bar, 10 µm. (B) Corresponding quantification of phalloidin/DAPI fluorescence. Values are mean ± s.d. (*P < 0.05). n.s., not significant.