Cytochrome P450 oxidase 2J inhibition suppresses choroidal neovascularization in mice

Abstract

Introduction: Choroidal neovascularization (CNV) in age-related macular degeneration (AMD) leads to blindness. It has been widely reported that increased intake of ω-3 long-chain polyunsaturated fatty acids (LCPUFA) diets reduce CNV. Of the three major pathways metabolizing ω-3 (and ω-6 LCPUFA), the cyclooxygenase and lipoxygenase pathways generally produce pro-angiogenic metabolites from ω-6 LCPUFA and anti-angiogenic ones from ω-3 LCPUFA. Howevehr, cytochrome P450 oxidase (CPY) 2C produces pro-angiogenic metabolites from both ω-6 and ω-3 LCPUFA. The effects of CYP2J2 products on ocular neovascularization are still unknown. Understanding how each metabolic pathway affects the protective effect of ω-3 LCPUFA on retinal neovascularization may lead to therapeutic interventions.

Objectives: To investigate the effects of LCPUFA metabolites through CYP2J2 pathway and CYP2J2 regulation on CNV both in vivo and ex vivo.

Methods: The impact of CYP2J2 overexpression and inhibition on neovascularization in the laser-induced CNV mouse model was assessed. The plasma levels of CYP2J2 metabolites were measured by liquid chromatography and tandem mass spectroscopy. The choroidal explant sprouting assay was used to investigate the effects of CYP2J2 inhibition and specific LCPUFA CYP2J2 metabolites on angiogenesis ex vivo.

Results: CNV was exacerbated in Tie2-Cre CYP2J2-overexpressing mice and was associated with increased levels of plasma docosahexaenoic acids. Inhibiting CYP2J2 activity with flunarizine decreased CNV in both ω-6 and ω-3 LCPUFA-fed wild-type mice. In Tie2-Cre CYP2J2-overexpressing mice, flunarizine suppressed CNV by 33 % and 36 % in ω-6, ω-3 LCPUFA diets, respectively, and reduced plasma levels of CYP2J2 metabolites. The pro-angiogenic role of CYP2J2 was corroborated in the choroidal explant sprouting assay. Flunarizine attenuated ex vivo choroidal sprouting, and 19,20-EDP, a ω-3 LCPUFA CYP2J2 metabolite, increased sprouting. The combined inhibition of CYP2J2 with flunarizine and CYP2C8 with montelukast further enhanced CNV suppression via tumor necrosis factor-α suppression.

Conclusions: CYP2J2 inhibition augmented the inhibitory effect of ω-3 LCPUFA on CNV. Flunarizine suppressed pathological choroidal angiogenesis, and co-treatment with montelukast inhibiting CYP2C8 further enhanced the effect. CYP2 inhibition might be a viable approach to suppress CNV in AMD.

Keywords: Age-related macular degeneration; Choroidal neovascularization; Cytochrome P450 oxidase 2J; Lipid metabolism; Long-chain polyunsaturated fatty acid; Tumor necrosis factor-α.

Fig. 1.

CYP2J2 overexpression increased CNV and increased plasma levels of 19,20-EDP. (A) Representative images of CNV lesion areas in WT and CYP2J2 TG mice on ω–6 or ω–3 LCPUFA feed. (B) The CNV lesion areas were significantly larger in CYP2J2 TG mice than in WT. WT; wild type, TG; Tie2-Cre; CYP2J2-overexpressing mice. Scale bar, 200 μm. n = 41–67; *, P < 0.05; ***, P < 0.001.

Fig. 2.

Inhibition of CYP2J2 activity decreased CNV. (A) Representative images of CNV lesion areas in WT and CYP2J2 TG mice with or without flunarizine treatment and fed ω–6 or ω–3 LCPUFA diets. (B, C) Flunarizine treatment significantly reduced CNV lesion area in both WT and CYP2J2 TG mice on both ω–6 (B, n = 26–45) and ω–3 (C, n = 49–66) LCPUFA feed. WT; wild type, TG; Tie2-Cre (CYP2J2-overexpressing mice). Scale bar, 200 μm. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Fig. 3.

Schematic diagram of CYP2C/CYP2J and sEH products from ω–6 and ω–3 LCPUFA. The plasma levels of 19,20-EDP and 14,15-EET in WT and CYP2J2 TG mice with or without flunarizine treatment and on ω–6 or ω–3 LCPUFA feed. WT; wild type, TG; Tie2-Cre; CYP2J2-overexpressing mice. sEH; soluble epoxide hydrolase, EET; epoxydocosapentaenoic acids, EDP; docosahexaenoic acids, DHET; dihydroxyeicosatrienoic acids, n = 5 each group; *, P < 0.05; **, P < 0.01.

Fig. 4.

19,20-EDP abolished inhibition of angiogenesis by flunarizine ex vivo. (A) Representative images of aortic ring sprouting from WT and CYP2J2 TG mice with or without flunarizine and EDP treatment. (B) Representative images of choroid explant sprouting from WT and CYP2J2 TG mice with or without flunarizine and EDP treatment. (C) 19,20-EDP abolished aortic ring sprouting inhibited by flunarizine. (n = 5, each explant) (D) 19,20-EDP rescued choroid explant sprouting inhibition by flunarizine (n = 6, each explant), WT; wild type, TG; Tie2-Cre; CYP2J2-overexpressing mice. Scale bar, 1 mm. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Fig. 5.

Inhibition of both CYP2J2 and CYP2C8 had synergistic effects on CNV inhibition. (A) Representative images of CNV lesion areas in WT mice with both flunarizine and montelukast treatment on normal feed. (B) Double treatment of flunarizine and montelukast further reduced CNV lesion areas than a single treatment. Scale bar, 200 μm. n = 58–66; *, P < 0.05. (C) qRT-PCR analysis of Vegfa, Il-1β, Tnfα, and Il-6 mRNA expression level in double treatment vs. control (n = 9–11 mice in each group).