The Histone Deacetylase Inhibitor AN7, Attenuates Choroidal Neovascularization in a Mouse Model

Abstract

: Choroidal neovascularization (CNV) is a complication of age-related macular degeneration and a major contributing factor to vision loss. In this paper, we show that in a mouse model of laser-induced CNV, systemic administration of Butyroyloxymethyl-diethyl phosphate (AN7), a histone deacetylase inhibitor (HDACi), significantly reduced CNV area and vascular leakage, as measured by choroidal flatmounts and fluorescein angiography. CNV area reduction by systemic AN7 treatment was similar to that achieved by intravitreal bevacizumab treatment. The expression of vascular endothelial growth factor (VEGF), fibroblast growth factor (FGF-2), and the endothelial cells marker CD31, was lower in the AN7 treated group in comparison to the control group at the laser lesion site. In vitro, AN7 facilitated retinal pigmented epithelium (RPE) cells tight junctions' integrity during hypoxia, by protecting the hexagonal pattern of ZO-1 protein in the cell borders, hence reducing RPE permeability. In conclusion, systemic AN7 should be further investigated as a possible effective treatment for CNV.

Keywords: AN7; bevacizumab; choroidal neovascularization; histone acetylation; histone deacetylase inhibitor; hypoxia; mouse model; retinal pigmented epithelium; vascular endothelial growth factor.

Figure 1

AN7 treatment elevates Histone H3 acetylation levels in laser-applied eyes. Representative images of cryosections from naïve eyes (without laser applications) of mice treated by intraperitoneal (IP) AN7 or saline (A,B,E,F) and lesions sites of laser-applied eyes of mice treated with IP AN7 or saline (C,D,G,H), from day 7 post laser application. (A–D) Hematoxylin and Eosin (H&E). Scale bar, 100 µm. Yellow arrows mark the same blood vessel in the H&E and the corresponding immunostaining image to allow orientation. (E–H) Immunostaining for CD31 (green), acetylated histone H3 (AC-H3; red) and cells nuclei, DAPI (blue). GCL, Ganglion Cells Layer; INL, Inner Nuclear Layer; ONL, Outer Nuclear Layer; RPE, Retinal Pigmented Epithelium; C, Choroid. Scale bar, 50 µm. (I) Quantification of AC-H3 staining in laser applied eyes.

Figure 2

Systemic AN7 treatment reduces choroidal neovascularization (CNV) area. (A) Representative images of choroidal flatmounts from day 7 post laser application, with CNV lesions sites from mice treated with saline, AN7 or bevacizumab. Fluorescein isothiocyanate dextran (FITC-dextran) (green) perfused through the blood vessels of the eyes and is seen at the laser lesion site, indicative of CNV formation. Scale bar, 100 µm. (B) Quantification of FITC area in choroidal flatmounts (indicative of CNV area) on day 7 from laser photocoagulation. Three laser applications were performed on the right eyes and mice were randomized to intraperitoneal (IP) 20 mg/kg AN7 or 10 mg/kg AN7 or IP saline-control groups, administered immediately following laser photocoagulation and for a total of three times a week thereafter. One-way ANOVA followed by Sidak post hoc test was used for statistical analysis. n = number of eyes per group. (C) Quantification of FITC area in choroidal flatmounts (indicative of CNV area) on day 7 post laser photocoagulation. Three laser applications were performed on the right eyes. IP injections of AN7 were compared to intravitreal (IVT) injection of bevacizumab and to corresponding saline controls. IP injections of AN7 or saline were administered immediately following laser applications and for a total of three times a week thereafter. IVT injections of bevacizumab or saline were administered once, immediately following the laser applications. One-way ANOVA followed by Sidak post-hoc test was used for statistical analysis. n = number of eyes per group.

Figure 3

AN7 treatment reduces CD31, vascular endothelial growth factor (VEGF) and fibroblast growth factor 2 (FGF-2). Representative images of laser lesion sites from mice treated with IP 20 mg/kg AN7 or saline (control), from day 3 post laser photocoagulation. Sequential cryosections are stained for endothelial cells marker CD31 (green; A,D), VEGF (red; B,E), and FGF-2 (purple; C,F). Cells nuclei are stained with DAPI (blue). GCL, Ganglion Cells Layer; INL, Inner Nuclear Layer; ONL, Outer Nuclear Layer; RPE, Retinal Pigmented Epithelium; C, Choroid. Scale bar, 50 µm. (G) Quantification of CD31 staining in laser applied eyes. (H) Quantification of VEGF staining in laser applied eyes. (I) Quantification of FGF-2 staining in laser applied eyes.

Figure 4

AN7 treatment reduces fluorescein leakage from laser lesions. Representative images of color fundus (A,D) and fluorescein angiography (FA; B,C,E,F) from day 7 post CNV induction of mice treated by intraperitoneal (IP) 20 mg/kg AN7 or saline. Time of imaging post fluorescein injection is indicated at the bottom right side of each angiogram. (B,C) Early (1 to 2 min post fluorescein injection) and late (4 to 5 min post fluorescein injection) angiograms of an eye from IP saline-control group. Lesion one was classified as stained, while lesions two and three were classified as leaky. (E,F) Early and late angiograms of an eye from IP 20 mg/kg AN7-treated group. All three lesions were classified as stained. (G) The percentage of leaky lesions of total lesions (six mice with 18 lesions total for each group) on days 2 to 7 post CNV induction (the proportions of leaky and stained lesions in each group were tested using the Fisher’s exact test; *Day 3, p = 0.041; **Day 4, p = 0.021; ***Day 6, p = 0.002; ****Day 7, p = 0.0003).

Figure 5

AN7 treatment stabilizes the retinal pigmented epithelium (RPE) monolayer during hypoxia. RPE cells were grown in normoxic or hypoxic conditions for 24 h, in the presence or absence of AN7 in the media. (A–D) Representative images of RPE cells stained for acetylated histone H3 (AC-H3; red) and cells nuclei, DAPI (blue). Scale bar, 20µm. (E) Quantification of the percentage of AC-H3 stained cells of total cells, comparing normoxic and hypoxic conditions in the presence or absence of AN7 in the cell media. One-way ANOVA followed by Tukey’s multiple comparisons test was used for statistical analysis. (F–M) Representative images of RPE cells stained for the tight junctions associated protein, Zonula Occludens-1 (ZO-1; red) and cells nuclei, DAPI (blue). Scale bar, 5µm. (E–H) Z-axis images are showing the distribution pattern of ZO-1 through the borders of the cells. (N) Quantification of FITC-dextran leakage through the RPE cell layer grown on Polyethylene Terephthalate (PET) membranes, comparing normoxic and hypoxic conditions in the presence or absence of AN7 in the cell media. One-way ANOVA followed by Tukey’s multiple comparisons test was used for statistical analysis.