The role of cytochrome P450 epoxygenases in retinal angiogenesis

Abstract

Purpose: The purpose of this study was to investigate the role(s) of cytochrome P450 epoxygenases (CYPs) and their products, the epoxyeicosatrienoic acids (EETs), in hypoxia-induced VEGF production and pathologic retinal angiogenesis.

Methods: Human retinal astrocytes, Müller cells, and retinal microvascular endothelial cells (HRMEC) were exposed to hypoxia, and relative CYP2C expression was measured by RT-PCR. Astrocyte and Müller cell VEGF production was measured by ELISA after exposure to hypoxia and treatment with the general CYP inhibitor, SKF-525a. Human retinal microvascular endothelial cells were treated with the CYP product, 11,12-epoxyeicosatrienoic acid [EET], or SKF-525a in the presence or absence of VEGF. Proliferation of HRMEC and tube formation were assayed. Oxygen-induced retinopathy (OIR) was induced in newborn rats. Retinal CYP2C11 and CYP2C23 expression were measured by RT-PCR. The OIR rats received SKF-525a by intravitreal injection and preretinal neovascularization (NV) was quantified. Retinal VEGF protein levels were measured by ELISA.

Results: Human retinal astrocytes were the only cells to exhibit significant induction of CYP2C8 and CYP2C9 mRNA expression by hypoxia. Astrocytes, but not Müller cells, exhibited reduced hypoxia-induced VEGF production when treated with SKF-525a. 11,12-EET induced HRMEC proliferation and tube formation, and SKF-525a inhibited VEGF-induced proliferation. Oxygen-induced retinopathy induced expression of CYP2C23, but had no effect on CYP2C11. SKF-525a inhibited retinal NV and reduced retinal VEGF levels in OIR rats.

Conclusions: The CYP-derived 11,12-EET may exhibit a proangiogenic biological function in the retina following stimulation by hypoxia in astrocytes. Inhibition of CYP may provide a rational therapy against retinal NV, because it can reduce VEGF production and VEGF-induced angiogenic responses in endothelial cells.

Keywords: angiogenesis; cytochrome P450; epoxyeicosatrienoic acid; retinopathy of prematurity.

Figure 1

CYP2C11 and CYP2C23 mRNA expression in rat OIR. Two days after removal from oxygen exposure (14[2]), The CYP2C23 expression is significantly increased, while CYP2C11 is unaffected in OIR compared to RA control. Data are presented as mean ± SEM. *P < 0.0001, n = 6.

Figure 2

The effect of 24-hour hypoxic exposure of CYP-expressing retinal cell types on CYP2C8 and CYP2C9 expression. (A) Hypoxia significantly induces CYP2C8 and CYP2C9 expression in human retinal astrocytes. However, hypoxia had no effect on (B) human Müller cells or (C) HRMEC. Data are presented as mean ± SEM. *P < 0.001, **P < 0.05, n = 7 to 9.

Figure 3

The effect of SKF-525a on hypoxia-induced VEGF production in retinal glial cells. (A) Human retinal astrocytes and (B) human Müller cells were exposed to either normoxia (20.9% O₂) or hypoxia (0.1% O₂) for 24 hours. The VEGF levels were measured in the culture medium and normalized to total protein. SKF-525a significantly decreased VEGF levels in cultures of hypoxic astrocytes, but not in cultures of hypoxic Müller cells. Representative figures are shown with mean ± SEM. *P < 0.01, n = 3 to 9.

Figure 4

The effect of 11,12-EET or SKF-525a on VEGF-induced proliferation in HRMEC. Treatment of HRMEC with 11,12-EET led to a dose-dependent increase in proliferation. Treatment with SKF-525a inhibited VEGF-induced proliferation at the highest concentration. Data are presented as mean ± SEM. *P < 0.01, **P < 0.001, ***P < 0.0001, n = 9.

Figure 5

The effect of 11,12-EET or SKF-525a on VEGF-induced tube formation in HRMEC. (A) Representative images from vehicle, 50 ng/mL VEGF or 11,12-EET (0.5 μM). Scale bars: 500 μm. (B) HRMEC treatment with 11,12-EET significantly induced tube formation at all doses used. SKF-525a did not significantly inhibit VEGF-induced tube formation. Data are presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, n = 9.

Figure 6

The effect of intravitreal injection of SKF-525a on OIR-induced pre-retinal NV. (A) Representative images from vehicle-treated and 5.0 μM SKF-525a-treated OIR retinas. (B) At the highest injected concentration (5.0 μM), NV was significantly inhibited. Data are presented as mean ± SEM; *P < 0.05, n = 12.

Figure 7

The effect of intravitreal injection of SKF-525a on VEGF production in rat OIR. Injection of 5.0 μM SKF-525a significantly reduced retinal VEGF protein level at 14(2) compared to vehicle injection in rat OIR. Data are presented as mean ± SEM. *P < 0.05, n = 18.