The forces, molecular mechanisms, and solute gradients which drive bidirectional solute-coupled water transport across epithelia have not been unequivocally described, let alone quantified. We report high-resolution transepithelial volume flow measurements across primary cultures of bovine tracheal epithelial cells on permeable supports in experiments designed to (1) estimate the driving forces for fluid absorption/secretion, (2) test correlation of net volume flux with the calculated short-circuit current, and (3) test the effects of counterion transport number on net solute flow under symmetric conditions.

Bovine tracheal epithelial cells obtained from abattoir material were grown to confluence in primary culture on permeable filters (Costar 3401) as described (Kondo et al. 1993). Filters (0.5 cm²) were mounted in mammalian Ringer solution at 36 °C in a Kel-F chamber (Jiang et al. 1993) for simultaneous measurement of transepithelial volume flow with a capacitance probe and transepithelial potential and resistance. The entire experiment was done with high precision control of ambient parameters, particularly temperature. By layering the meniscus of each hemichamber with hexadecane, background volume losses were totally eliminated and the capacitance probe tracked transepithelial volume flow with high stability. Student’s paired t test was used to calculate significance.

Serosal ouabain (> 10 µM) abolished absorption (from 17.2 ± 1.1 to -1.1 ± 2.6 µl cm^-2 h^-1 (S.D.; n = 2) as did 10 µM apical amiloride (from 14.0 ± 3.9 to 0.6 ± 2.4 µl cm^-2 h^-1 (S.E.M.; n = 10, P < 0.001). Simultaneous bilateral substitution of Cl with gluconate changed spontaneous absorption from 5.2 ± 1.0 µl cm^-2 h^-1 to a secretion of 0.8 ± 2.6 µl cm^-2 h^-1 (S.E.M.; n = 4, P < 0.05) without a significant change in the calculated short-circuit current. Serosal hyperosmolarity (sucrose 20-25 mmol, normalised to 20 mmol) in either Cl^–-containing or Cl^–-free solutions increased absorptive flow from 4.0 ± 2.8 to 18.4 ± 2.4 µl cm^-2 h^-1 (S.E.M.; n = 10, P < 0.001).

We conclude that (1) absorptive flows are generated osmotically by Na⁺ absorption with gradients of the order of 20 mmol, (2) a Cl^–-independent secretory flow exists which is not accounted for electrically, and (3) this epithelium is a useful model for both fluid absorption and secretion.

This work was supported by NIH and MIBRS.