PubMed1987
1. Rabbit corneal thickness changes were measured after some of the NaCl in the bathing Ringer solution was substituted by a neutral sugar. 2. The response had three phases which could be closely modelled by three exponentials of different amplitudes and rate constants, originating from the time of the substitution. 3. The first, fastest, phase was interpreted as being driven by the pure osmotic pressure difference developed across the corneal endothelium by the difference between the removed NaCl and the added sugar in the bathing Ringer solution; the second was driven by the diffusion of NaCl out of the stroma; and the third, slowest, phase was driven by the diffusion of the added sugar into the stroma. 4. Consistent with the interpretation, only the third, slowest, phase had its rate constant dependent upon the nature of the substituting neutral sugar. 5. The amplitude of the pure osmotic phase was a linear function of the added sugar. Its amplitude was zero when an equal osmolarity of sugar was substituted for NaCl. 6. It was concluded that the reflexion coefficient of rabbit corneal endothelium to NaCl is the same as that to sucrose, raffinose and stachyose, i.e. about 1. 7. The calculated hydraulic conductivity of the endothelium plus stroma was about the same as that of stroma alone, and it was concluded that the hydraulic conductivity of corneal endothelium is large compared to corneal stroma. 8. It is proposed that most of the hydraulic flow in response to osmotic gradients passes through the cells, whereas salt diffuses through the paracellular route, resulting in an apparent reflexion coefficient of 1. 9. Thus, corneal endothelium is a 'leaky' (12 omega cm2) salt-permeable epithelium and, according to the present study, simultaneously a semi-permeable membrane. To resolve the terminology, we suggest the term 'bi-permeable' for such epithelia which have a high salt permeability but such a very high water permeability as to give apparent reflexion coefficients of 1.