Transport-dependent alterations of membrane properties of mammalian colon measured using impedance analysis

Summary

Direct current (DC) measurement methods have been commonly used to characterize the conductance properties of the mammalian colon. However, these methods provide no information concerning the effects of tissue morphology on the electrophysiological properties of this epithelium. For example, distribution of membrane resistances along narrow fluid-filled spaces such as the lateral intercellular spaces (LIS) or colonic crypts can influence DC measurements of apical and basolateral membrane properties. We used impedance analysis to determine the extent of such distributed resistance effects and to assess the conductance and capacitance properties of the colon. Because capacitance is proportional to membrane area, this method provides new information concerning membrane areas and specific ionic conductances for these membranes.

We measured transepithelial impedance under three conditions: (1) control conditions in which the epithelium was opencircuited and bathed on both sides with NaCl−HCO₃ Ringer's solutions, (2) amiloride conditions which were similar to control except that 100 μm amiloride was present in the mucosal bathing solution, and (3) mucosal NaCl-free conditions in which mucosal Na and Cl were replaced by potassium and sulfate or gluconate (“K⁺ Ringer's”). Three morphologically-based equivalent circuit models were used to evaluate the data: (1) a lumped model (which ignores LIS resistance), (2) a LIS distributed model (distributed basolateral membrane impedance) and (3) a crypt-distributed model (distributed apical membrane impedance). To estimate membrane impedances, an independent measurement of paracellular conductance (G _s ) was incorporated in the analysis. Although distributed models yielded improved fits of the data, the distributed and lumped models produced similar estimates of membrane parameters. The predicted effects of distributed resistances on DC microelectrode measurements were largest for the LIS-distributed model. LIS-distributed effects would cause a 12–15% underestimate of membrane resistance ratio (R _a /R _b ) for the control and amiloride conditions and a 34% underestimate for the “K Ringer's” condition. Distributed resistance effects arising from the crypts would produce a 1–2% overestimate ofR _a /R _b .

Apical and basolateral membrane impedances differed in the three different experimental conditions. For control conditions, apical membrane capacitance averaged 21 μF/cm² and the mean apical membrane specific conductance (G _a-norm) was 0.17 mS/μF. The average basolateral membrane capacitance was 11 μF/cm² with a mean specific conductance (G _b-norm) of 1.27 mS/μF.G _a-norm was decreased by amiloride or “K⁺ ringer's” to 0.07 mS/gmF and 0.06 mS/μF, respectively. Basolateral conductance was also reduced by amiloride, whereas capacitance was unchanged (G _b-norm=0.97 mS/μF). For the “K⁺ Ringer's condition, both basolateral conductance and capacitance were greatly increased such thatG _b-norm was not significantly different from the control condition.