Purchase this article with an account.
H.I. Salti, E. Wong, A. Nuthi, A. Clermont, L. Goddard, S. Bursell, j. Takahashi, G. King, L. Aiello; Bacterial Permeability Increasing Protein (rBPI21) Inhibits VEGF and Diabetes-induced Retinal Vascular Leakage in vivo . Invest. Ophthalmol. Vis. Sci. 2003;44(13):3909.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose: BPI is a 55Kd protein produced by human neutrophils which binds and neutralizes heparin. BPI inhibits VEGF-stimulated cell growth. rBPI21 is an N-terminal fragment of BPI containing 3 heparin binding sites. In this study, we examine whether rBPI21 can inhibit VEGF and diabetes-induced retinal vascular leakage. Methods:In study 1, rBPI was injected intravitreally at 1 or 75 µg/ml in one eye and vehicle alone into the contralateral eye. After 30 minutes, VEGF (25 ng/ml) was injected intravitreally into both eyes. 15 minutes later, 65µL of sodium fluorescein was infused through a jugular catheter. Vitreous fluorescein photometry was performed on both eyes after 30 minutes to determine VEGF induced retinal leakage. In study 2, SD rats had diabetes induced with 55 mg/kg STZ. On days 11 and 13 after diabetes onset, intravitreal injections of BPI (1, 20, and 75 µg/ml) were performed in one eye and vehicle alone in the contralateral eye. Retinal permeability was assessed by evan's blue albumin permeation on day 14. Results: In study 1, rBPI21 (1 and 75µg/ml) inhibited VEGF-induced leakage by 45±92% (3.8±1.4 vs 5.2±1.6, n=8, p=0.062) and 67±25% (3.1±0.5 vs 5.3±1.5, n=8, p=0.014), respectively. Thaumatin (75µg/ml), a similarly sized and charged protein to BPI with no heparin binding activity, had no effect on VEGF induced leakage (4.5±0.9 vs 4.6±0.9, n=4). In study 2, rBPI21 (75µg/ml) had no effect on baseline leakage in non-diabetic rats (5.5±2.8 vs 5.3±1.8 uL/g/hr). After 2 weeks, diabetic retinal vascular leakage in vehicle treated eyes was increased 104±59% (10.8±3.1 µL/g/hr, p < 0.001) as compared to control. In diabetic rats, rBPI inhibited leakage in a dose-dependent manner. Leakage was inhibited at 1, 20, and 75 µg/ml by 23±39% (10.6±2.6 vs 12.1±4.2 uL/g/hr, n=6), 45±55% (7.5±2.6 vs 9.7±2.5 µL/g/hr, n=5, p=0.044), and 64±47% (7.2±2.5 vs 10.5±2.6 µL/g/hr, n=11, p=0.002), respectively. Conclusions: Both VEGF and diabetes-induced retinal vascular leakage were inhibited by rBPI21 in a dose dependent manner in vivo. The inhibition observed by rBPI21 supports further studies to elucidate the therapeutic utility of its anti-permeability activity in ocular disorders such as diabetic macular edema.
This PDF is available to Subscribers Only