Parstatin suppresses ocular neovascularization and inflammation

Abstract

Purpose: Parstatin is a 41-mer peptide formed by proteolytic cleavage on activation of the PAR1 receptor. The authors recently showed that parstatin is a potent inhibitor of angiogenesis. The purpose of the present study was to evaluate the therapeutic effect of parstatin on ocular neovascularization.

Methods: Choroidal neovascularization was generated in mice using laser-induced rupture of Bruch's membrane and was assessed after 14 days after perfusion of FITC-dextran. Oxygen-induced retinal neovascularization was established in neonatal mice by exposing them to 75% O(2) at postnatal day (P)7 for 5 days and then placing them in room air for 5 days. Evaluation was performed on P17 after staining with anti-mouse PECAM-1. The effect of parstatin was tested after intravitreal administration. The effects of subconjunctival-injected parstatin on corneal neovascularization and inflammation in rats were assessed 7 days after chemical burn-induced corneal neovascularization. Retinal leukostasis in mice was assessed after perfusion with FITC-conjugated concanavalin A.

Results: Parstatin potently inhibited choroidal neovascularization with an IC(50) of approximately 3 μg and a maximum inhibition of 59% at 10 μg. Parstatin suppressed retinal neovascularization with maximum inhibition of 60% at 3 μg. Ten-microgram and 30-μg doses appeared to be toxic to the neonatal retina. Subconjunctival parstatin inhibited corneal neovascularization, with 200 μg the most effective dose (59% inhibition). In addition, parstatin significantly inhibited corneal inflammation and VEGF-induced retinal leukostasis. In all models tested, scrambled parstatin was without any significant effect.

Conclusions: Parstatin is a potent antiangiogenic agent of ocular neovascularization and may have clinical potential in the treatment of angiogenesis-related ocular disorders.

Figure 1.

Intravitreal injections of parstatin suppress CNV. Laser-induced ruptures of Bruch's membrane were performed in mice. Intravitreal injections of indicated doses of parstatin (parst) or vehicle (control) or scrambled parstatin (scrambled, 10 μg) were administered immediately after laser treatment and 7 days after laser treatment. CNV was assessed 14 days after laser treatment. Mice were perfused with FITC-labeled dextran, and choroidal flatmounts were prepared and examined by fluorescence microscopy. Compared with control eyes (A), those injected with 1 μg (B) or 10 μg (C) parstatin showed proportionally smaller areas of CNV. CNV in eyes injected with scrambled parstatin (D) was similar to that in control mice. (E) The area of CNV at each rupture site was measured by image analysis. Results are expressed as mean areas (mm²) of CNV ± SE for each group calculated from the indicated number (n) of eyes. Statistical analysis was performed compared with the control group. *P < 0.05.

Figure 2.

Intravitreal injections of parstatin suppress oxygen-induced retinal neovascularization. Newborn mice were placed in 75% oxygen at P7. At P12, they were returned to room air. Intravitreal injections of indicated doses of parstatin (parst), vehicle (control), or scrambled parstatin (scrambled, 10 μg) were administered on P12 and P15. At P17, mice were treated with G. simplificolia isolectin B4–589. Compared with control retinas (A, E), those treated with 0.5 μg (B, F) or 3 μg (C, G) parstatin showed proportionally fewer areas of retinal neovascularization. Retinal neovascularization in retinas treated with scrambled parstatin (D, H) was similar to that observed in control mice. (E–H) Higher magnification images of the boxes in (A) to (D), respectively. Scale bar, 200 μm. (I) Total area of neovascularization (NV) at each retina site was measured by image analysis. Results are expressed as mean areas (mm²) of retinal neovascularization ± SE for each group calculated from indicated number (n) of eyes. Statistical analysis was performed compared with control group. *P < 0.05; **P < 0.01.

Figure 3.

Subconjunctival injections of parstatin suppress corneal neovascularization. Chemical burn–induced corneal trauma was performed by the application of 75% silver nitrate and 25% potassium nitrate to the centers of the rat corneas. Two subconjunctival injections per eye of indicated doses of parstatin (parst) or vehicle (control) or scrambled parstatin (scrambled, 2 × 100 μg) were administered immediately after cauterization. Corneal neovascularization was assessed 7 days after cauterization. Compared with control corneas (A), those treated with 2 × 75 μg (B) or 2 × 100 μg (C) parstatin showed proportionally reduced areas of corneal neovascularization. Corneal neovascularization in eyes treated with scrambled parstatin (D) was similar to that observed in control mice. (E) Total area of neovascularization (NV) in each cornea was measured by image analysis. Results are expressed as the mean percentage of area covered by vessels to the total corneal area ± SE for each group calculated from the indicated number (n) of eyes. Statistical analysis was performed compared with the control group. **P < 0.01.

Figure 4.

Corneal histopathology. Rats were treated as described in Figure 3. Seven days after cauterization, eyes were excised, embedded in paraffin, and sectioned. Sections were stained with hematoxylin-eosin, and vascular vessels (A–C) or neutrophils (D–F) in the corneas were counted at 200× or 400× magnification, respectively. Compared with control corneas (A, D), those treated with 2 × 100 μg (B, E) parstatin had fewer blood vessels or infiltrating neutrophils. Blood vessel (C) or neutrophil (F) density in corneas treated with scrambled parstatin were similar to those observed in control rats. (G) The total number of blood vessels was measured in seven sections from each eye. Results are expressed as mean number of vascular vessels per section ± SE for each group calculated from indicated number (n) of eyes. (H) The total number of neutrophils was measured in seven sections from each eye. Results are expressed as mean number of inflammatory cells per section ± SE for each group calculated from indicated number (n) of eyes. Statistical analysis was performed compared with the control group. **P < 0.01.

Figure 5.

Parstatin reduces VEGF-induced retinal leukostasis. Intravitreal injections of vehicle (control), VEGF (10 μΜ), parstatin (parst, 10 μg), or the combination of VEGF with parstatin were administered to assess their effect on retinal leukostasis. After 6 hours, mice were perfused with FITC-conjugated concanavalin A. Retinal flatmounts were prepared, and the total number of leukocytes adhering to the retinal vessels (arrows) was counted at 200× magnification. Few leukocytes adhered to the retinal vessels after injections of vehicle or parstatin (A, 200×; inset, 400×), whereas pronounced leukostasis was observed after injections of VEGF (B, 200×; inset, 400×). The combination of parstatin with VEGF resulted in a significant reduction in VEGF-induced leukostasis (C, 200×). Scale bar, 200 μm. (D) Total numbers of adherent leukocytes were measured in each retina. Results are expressed as mean number of leukocytes per section ± SE for each group calculated from the indicated number (n) of eyes. Statistical analysis was performed compared with the control group. *P < 0.05.