We have adapted the vibrating probe extracellular recording technique to use on an epithelium under voltage clamp in an Ussing chamber. The vibrating probe allows very low drift measurements of current density immediately over the epithelial surface. These measurements allowed sites of electrogenic transport in the epithelium to be localized with a spatial resolution of 5 micrometers. The technique was applied to the opercular membrane of the teleost fish, the tilapia, Sarotherodon mossambicus. The mitochondrion-rich "chloride cells" were shown to be the only sites of electrogenic ion transport in this heterogeneous epithelium. Cell sampling experiments demonstrated variable negative short-circuit currents associated with nearly all of approximately 300 chloride cells examined, which appeared to account for all of the tissue short-circuit current. Current-voltage relations for individual cells were also measured. Conductance associated with chloride cells (i.e. cellular and junctional pathways) accounted for all but 0.5 mS/cm2 of the tissue conductance, with the balance apparently accounted for by leak pathways near the edge of the tissue. Current and conductance associated with other cell types was at least 50-fold smaller than for the chloride cell. Chloride-free solutions reduced chloride cell current and conductance by 98 and 95%, respectively.

We have utilized the vibrating probe technique to examine transport by individual chloride cells in the short-circuited fish opercular epithelium. Variability in the steady state and in response to rapid perturbations, including fast-acting hormones and ion replacement, was analyzed. Negative short-circuit currents, corresponding to chloride secretion, were associated with the apical crypts of all but five of 386 chloride cells sampled. Average chloride cell short-circuit current and conductance were 2.7 +/- 0.1 nA and 87.7 +/- 3.8 nS, respectively, or 19 mA cm-2 and 620 mS cm-2 (resistance = 1.6 omega cm2) when normalized to apical crypt surface area. Exposure to 1 microM epinephrine rapidly inhibited the tissue short-circuit current by inhibiting the current pumped by all chloride cells, i.e. all chloride cells have adrenergic receptors. The time course of inhibition for each cell mirrored that of the whole tissue. Reversal of epinephrine inhibition of the tissue short-circuit current by glucagon and phosphodiesterase inhibition was by reversal of epinephrine's inhibition of individual chloride cells, and not by turning on cells which were previously inactive or uninhibited, or by stimulating nonchloride cells. A great amount of variability existed among chloride cells in the ability of these agents to reverse epinephrine-inhibited current. Likewise, considerable variability in the response of chloride cell conductance to these perturbations was observed, and in many instances a clear dissociation between current and conductance was noted. In the steady state, variability among cells in a single tissue always defined a linear relationship between chloride cell current and conductance with zero-current conductance intercept at zero. Equivalent circuit modeling indicates that the leak conductance of chloride cells within a single tissue always contributes the same proportion to the total individual chloride cell conductance, such that the ratio between the conductances of the active and leak pathways of chloride cells is constant. The leak pathway is almost certainly dominated by a sodium-selective paracellular pathway. The results suggest that these cells control the permeability of their paracellular pathway. A possible mechanism for this control is discussed.