Ever since the introduction of the short-circuit technique for studies of electrolyte transport across frog skin epithelium (Ussing & Zerahn, 1951) the ‘Ussing chamber’ has evolved and become adapted to a variety of different applications in epithelial physiology. However, most systems retain the key features of the original design: symmetrical hemi-chambers allowing the two surfaces of the epithelium to be bathed with physically separate solutions; paired electrodes (or salt bridges) placed close either side of the epithelium for measuring the transepithelial PD; and paired current-passing electrodes placed more distally for voltage clamp experiments.

With the increasing use of cultured epithelial cell lines grown on permeable filter supports, modified Ussing chamber systems have become commercially available that are designed specifically to accommodate these filter ‘inserts’. In principle this is an excellent idea which makes it possible to transfer a confluent, polarized monolayer of cells directly from the culture plate into the Ussing chamber with minimal disturbance and no risk of edge damage. In practice however we have found that repetitive draining and refilling of the hemi-chambers introduces significant artifacts and may lead to irreversible changes in transepithelial resistance (TER). In our hands, Caco-2 cells grown on polycarbonate filters (Snapwell, Corning Costar Inc.) showed both a transient stimulation of the short-circuit current and a loss of TER every time the solution on one or both sides of the monolayer was replaced with an identical solution. Such artifacts led to serious problems in measuring and interpreting the small diffusion potentials that are used to evaluate tight junction permeability.

To circumvent these problems, which are presumably due to the pressure changes and other mechanical forces associated with draining and refilling a conventional Ussing chamber, we have devised a simple system in which the two surfaces of the filter are continuously superfused using a peristaltic pump. In order to keep the chamber volumes small, and therefore achieve rapid solution changes at modest perfusion rates, we have based our design around a 12-mm filter insert (Transwell, Corning Costar Inc.) located in situ in one well of a 12-well culture plate. Inlets and outlets for the solutions are provided by syringe needles, the voltage electrodes are agarose-KCl bridges linked to Ag-AgCl electrodes and the current electrodes are chloride-coated Ag wire loops located in the upper and lower compartments. Using this system, solution changes can be achieved without any mechanical disturbance of the cells or loss of TER. Furthermore, by replacing the bottom of the culture plate well with a glass coverslip it is relatively simple to transfer the system on to the stage of an inverted fluorescence microscope for simultaneous measurement of intracellular ions by microfluorometry.

This work was supported by the Wellcome Trust.