MICROVASCULAR PERMEABILITY WITH SULFONYLUREAS IN NORMAL AND DIABETIC HAMSTERS

The hamster cheek pouch is an experimental model in which quantitative studies of macromolecular permeability can be made by direct observat ion of extravasated fluorescein isothiocyanate (FITC)-dextran (leaks). The advantage of this model is that simultaneous light and fluorescen t-light microscopy observations can be performed with instantaneous co rrelations between the site of FITC-dextran extravasation and the vess el morphology. The aims of our study were to compare, using the cheek pouch preparation, the effects of two sulfonylureas, gliclazide and gl ibenclamide, on the macromolecular permeability increase induced by hi stamine using control (normoglycemic) hamsters. In these studies, FITC -labeled dextran 150,000 daltons was administered intravenously and qu antified by UV-light microscopy, and the drugs used were applied topic ally at therapeutic concentrations. Gliclazide and glibenclamide dose- dependently decreased the macromolecular permeability increase induced by histamine. This effect of gliclazide could be blocked by nifedipin e (Ca2+ channel blocker) and not by diazoxide (K+ channel opener), whe reas for glibenclamide it could be blocked by diazoxide and not by nif edipine. To better characterize the antioxidant capacity of gliclazide and glibenclamide, their effect on the macromolecular permeability in crease induced by ischemia/reperfusion was also compared with the effe ct of vitamin C in diabetic hamsters (glycemia > 240 mg/dL). Total isc hemia of the preparation was obtained with a cuff placed around the ne ck of the everted pouch. Diabetes was induced by three intraperitoneal injections of streptozotocin 50 mg/kg/d in 3 days. In diabetic hamste rs during ischemia/reperfusion, gliclazide was more effective in inhib iting the macromolecular permeability increase than glibenclamide (136 .0 +/- 5.8 leaks/cm(2) for placebo; 68.0 +/- 2.9 for 1.2 x 10(-6) mol/ L gliclazide; 55.3 +/- 3.5 for 1.2 x 10(-5) mol/L gliclazide; 89.2 +/- 5.7 for 8 x 10(-8) mol/L glibenclamide; 107.0 +/- 3.8 for 8 x 10(-7) mol/L glibenclamide; 56.7 +/- 3.4 for 10(-6) mol/L vitamin C; and 20.5 +/- 0.6 for 10(-5) mol/L vitamin C). Our results suggest that (1) the inhibition of the permeability increase induced by histamine elicited by gliclazide may be mediated by Ca2+ channels, while that of glibenc lamide may be mediated by K+ channels, and (2) gliclazide appears to h ave an antioxidant capacity in ischemia/reperfusion injury similar to that of 10(-6) mol/L vitamin C. Improvement in the microcirculation wa s independent of the hypoglycemic properties of the drug. Copyright (C ) 1997 by W.B. Saunders Company.