Regulation of fluid flow through corneal stroma in the bullfrog

Abstract

Regulation of fluid flow through corneal stroma was investigated in the bullfrog. Corneal specimens were mounted by clamping their limbal sciera between the two chambers of a Ussing-type chamber. The epithelial surface was covered with Ringer's solution, while the endothelial surface was superfused with Ringer's solution at various pressures ranging from 0–60 mm Hg. At 0 mm Hg, the cornea swelled, while at 10 mm Hg the corneal thickness remained unchanged. Further elevation of the hydrostatic pressure of the endothelial superfusion solution caused a decrease in corneal thickness, suggesting that the hydrostatic pressure in the in vivo frog corneal stroma is about 10 mm Hg. At 10 mm Hg of endothelial superfusion pressure, piercing the epithelial cell layer with a 30-gauge needle caused only slight corneal swelling. Removing glucose from the epithelial perfusion solution induced a slowly progressing increase in corneal thickness. Iodoacetate did not interfere with the swelling of the pierced cornea after the removal of glucose from the epithelial perfusion solution. To examine the possibility that the interstitial fluid flows across the stroma-scleral boundary, corneal specimens having undamped sclera were incubated in Ringer's solution containing 3 mmol/L dextran of various molecular weights ranging from 8800–162000, and the volume of the preparation was monitored by sequential measurement of the weight. In the presence of dextran with a molecular weight higher than 70000, the corneal volume decreased at the beginning of incubation, and after reaching the minimal volume it slowly increased, indicating that the stroma-scleral boundary is permeable to dextran of even a molecular weight of 162000, although dextran molecules diffuse much more slowly than water, and the concentration of unfilterable solutes in the stroma is lower than 3 mmol/L. In experiments using the Ussing-type chamber at 10 mm Hg of endothelial superfusion pressure, a decrease of NaCl in the superfusing solution to 1/2 caused rapid corneal swelling followed by slow recovery. Adding NaCl to the 1/2 NaCl Ringer's solution caused a further corneal thinning in a concentration-dependent manner. The same extent of decrease in corneal thickness as induced by adding NaCl was achieved by the same concentration of glucose as of NaCl, implying that the value of the reflection coefficient of the endothelial cell layer to either Na + or Cl − is about half that of glucose. Our results show that even a small difference in the concentration of low molecular weight solutes (e.g., Na + and Cl −) exerts a force that draws water from the cornea.