Microperfusion studies on the permeability of retinal vessels. A new model demonstrating organic anion transport and a reabsorptive fluid flux.

We developed an experimental model to study the permeability of individual retinal vessels in vitro using microperfusion techniques adapted from kidney tubule studies. The retinal vessels were isolated by freehand dissection and mounted on a microperfusion apparatus. When inulin was perfused luminally, it was diluted to 80.2 +/- 2.3% of its initial concentration. However, no radioactive leak into the bath side was observed, suggesting that the dilution was due to fluid flux from bath to lumen. The dilution of fluorescein (81.9 +/- 3.8%) was in the same range as that of inulin, the reference marker. The extremely low lumen-to-bath fluorescein flux, 0.5 +/- 0.9 X 10(-12) mol/min/mm, increased by 68% when probenecid was added to the perfusate and by 210% when probenecid was placed in the bath. The effect was concentration-dependent. When placed in the bath, fluorescein moved rapidly across the retinal vessel walls, accumulating in the lumen to concentrations 40 times higher than in the bath. This movement from bath to lumen, which was much higher (13.6 +/- 0.3 X 10(-12) mol/min/mm) than the lumen-to-bath fluorescein flux for the same fluorescein concentration, decreased by adding probenecid to the bath. The kinetics of this unidirectional movement of fluorescein were consistent with a saturable active transport process. The fluid flux from bath to lumen across the retinal vessels, which was 6.3 +/- 1.0 nl/min/mm for perfusion rates of 6.6 +/- 0.2 nl/min, was temperature-dependent and was coupled to the fluorescein transport. Fluorescein stimulated the fluid flux by 17% when added to the perfusate and by 60% when added to the bath, and this effect could be reversed by probenecid. Our results showed an active transport of fluorescein in the rabbit retinal vessels coupled with net fluid flux from outside the vessels into the lumen.