Hyperglycemia stimulates a sustained increase in hydraulic conductivity in vivo without any change in reflection coefficient.

Objective: Increased microvascular permeability contributes to the development of diabetic microvascular complications and diabetic vasculopathy is correlated with blood glucose levels. The mechanisms underlying increased permeability, however, are poorly understood.

Methods: The Landis-Michel technique was used to measure water permeability (hydraulic conductivity, Lp) and macromolecular permeability (reflection coefficient, sigma) of exchange capillaries in frogs and rats.

Results: Dialysed normoglycemic plasma from diabetic patients had no effect on Lp. The same plasma with 20 mM glucose increased hydraulic conductivity from (mean +/- SEM x 10(-7) cm . s(-1). cm H2O(-1)) 5.73 +/- 2.01 to 13.09 +/- 2.67 (P < .01). Nondiabetic control plasma did not affect Lp, but addition of 20 mM glucose increased L(p) to a similar degree. The effect of glucose alone was examined. Glucose at 20 mM increased Lp, from 2.82 +/- 0.61 to 4.71 +/- 1.35 x 10(- 7) cm . s(- 1). cm H2O(-1) (P = .002, n = 13). A similar increase was seen in rat mesenteric microvessels, from 1.04 +/- 0.40 in control perfusions to 2.18 +/- 0.56, P < .05. The microvascular macromolecular reflection coefficient in all the above experiments was unaltered. The use of specific inhibitors indicated that the glucose-induced increased Lp did not appear to be mediated through protein kinase C (PKC), free radical generation, glucose metabolism, or albumin glycation.

Conclusions: These data suggest that hyperglycemia induced increased apparent protein permeability may be secondary to a glucose-mediated change in macromolecular convective flux rather than any change in protein permeability per se. The authors speculate that the increased microvascular permeability to water in vivo is mediated by direct interaction of glucose with the endothelial cells (perhaps with the glycocalyx).