The regulation of corneal hydration by a salt pump requiring the presence of sodium and bicarbonate ions.

1. The use of polyacrylamide gel salt bridges enables trans-membrane potentials to be measured to an accuracy of 20 muV over long periods.2. The technique is applied to measure electrical potentials across corneal endothelia of rabbits.3. In de-epithelialized corneas which translocate water, a spontaneous potential of 550 muV is found across the endothelium (tissue resistance 20 Omega cm(2)).4. This electrical potential (and water translocation) is reduced to zero when sodium is absent from the Ringer, and by about 80% when bicarbonate ions are absent. Removal of chloride has no such effect.5. Under a variety of conditions, the potential correlates with the observed translocation of fluid across corneal endothelium. The translocated fluid is shown to be isotonic with sodium in the Ringer and therefore the potential correlates with ;active' sodium transport.6. The potential and water translocation are abolished in the presence of ouabain at concentrations greater than 10(-5)M.7. The potential (lens-side negative) is of the wrong polarity to explain the net sodium transport (into the lens-side) by a sodium ion ;pump'.8. The current does not equal the net sodium flux under short circuit conditions. They differ in magnitude and polarity.9. A model is proposed where the endothelium ;pumps' salt out of the corneal stroma into the aqueous humour.10. Flux equations are derived for a condition where the membrane (corneal endothelium) separates an ion exchanger (corneal stroma) from free solution (aqueous humour), where the usual relationship for free-free solutions Deltapi = c(s)Deltamu(s) does not apply.11. The model is of use only when the stroma is well stirred. It may be used in whole corneas retaining their epithelium but it may not be used in de-epithelialized corneas.12. The model predicts that the presence of an ;active' salt flux out across the endothelium would create passive water and salt fluxes. The passive water flux would also travel out of the stroma across the endothelium; the passive salt flux would travel, in the opposite direction, into the stroma across the endothelium.13. The kinetics of the passive water efflux, as a swollen cornea reverts to physiological hydration (the temperature reversal phenomenon) are predicted extremely well if the ;active' salt flux is chosen at 3.3 x 10(-7) m-mole. cm(-2) sec(-1).14. The value of the active salt flux which cannot be measured directly is extrapolated to be somewhat greater than 2.8 x 10(-7) m-moles. cm(-2) sec(-1); in good agreement with that required by the model to explain the temperature reversal phenomenon.15. The model is further used to calculate the salt concentration difference across the endothelium (which drives salt passively into the stroma) at various stromal hydrations.16. When an appropriate salt concentration is applied across the endothelium of de-epithelialized cornea, it generates a potential of the same polarity and similar magnitude to that found across the endothelium of equilibrated whole cornea. The endothelium acts like a cation exchange membrane.17. Additionally the calculated salt concentration difference across the endothelium correlates well with the measured transendothelial potentials in whole cornea as the corneal hydration varies.18. It is concluded that the model of an endothelial neutral salt ;pump' regulating corneal hydration is self consistent. The spontaneous potential found across the endothelium could be caused by the consequential passive flux of salt in the opposite direction.