Integrating Continuous Transepithelial Flux Measurements into an Ussing Chamber Set-Up

Abstract

Fluorescently labelled compounds are often employed to study the paracellular properties of epithelia. For flux measurements, these compounds are added to the donor compartment and samples collected from the acceptor compartment at regular intervals. However, this method fails to detect rapid changes in permeability. For continuous transepithelial flux measurements in an Ussing chamber setting, a device was developed, consisting of a flow-through chamber with an attached LED, optical filter, and photodiode, all encased in a light-impermeable container. The photodiode output was amplified and recorded. Calibration with defined fluorescein concentration (range of 1 nM to 150 nM) resulted in a linear output. As proof of principle, flux measurements were performed on various cell lines. The results confirmed a linear dependence of the flux on the fluorescein concentration in the donor compartment. Flux depended on paracellular barrier function (expression of specific tight junction proteins, and EGTA application to induce barrier loss), whereas activation of transcellular chloride secretion had no effect on fluorescein flux. Manipulation of the lateral space by osmotic changes in the perfusion solution also affected transepithelial fluorescein flux. In summary, this device allows a continuous recording of transepithelial flux of fluorescent compounds in parallel with the electrical parameters recorded by the Ussing chamber.

Keywords: Ussing chamber; automatized flux measurement; claudin; osmotic stress; paracellular fluorescent marker; tight junction; transepithelial flux.

Figure 1

(a) Original recording of a calibration experiment. Fluorescein was added cumulatively (green arrows) from a 10 µM stock solution to 10 mL of Ringer’s solution, and the output voltage of our device was recorded on a flatbed chart recorder. Directly prior to each addition of fluorescein, the volume was adjusted to 10 mL by removing the previously added volume. The trace shows that even an increment of 1 nM can be reliably detected. (b) Plotting the output voltage of our device (red) and of a plate reader (blue) against the concentration (mean ± SEM, n = 4 to 7 for individual data points) demonstrated linearity of the calibration for our device within the tested concentration range of 1 nM and 150 nM (dotted line, linear regression; R² = 0.996), whereas the plate reader signals deviated from linearity at low concentrations.

Figure 2

Linear dependence of fluorescein flux on its gradient. (a) Apical to basolateral fluorescein flux across a HT-29/B6 cell layer (treated with EGTA to increase permeability) was recorded in the presence of 100 µM fluorescein on the donor side. As indicated by green arrows, fluorescein concentration on the donor side was subsequently increased to 200 and 300 µM. The step-wise increase in the fluorescein gradient caused a proportional step-wise increase in the slope s (dotted lines, V/min), i.e., a linear increase in the fluorescein flux. (b) Average of three such experiments (fluorescein flux relative to the flux observed at 100 µM apical fluorescein, plotted against the apical fluorescein concentration; mean ± SEM). Dotted line: linear regression, R² = 0.957.

Figure 3

(a,b) Original recording of fluorescein flux measurement (increase in fluorescein concentration on the acceptor side per time unit) across cell layers of claudin quin KO overexpressing Cldn4 (a) and Cldn8 (b). Slope s (V/min) of the linear parts of the recordings were determined (dotted line) and converted into fluxes (nmol/cm²/h) using the calibration shown in Figure 1. Under control conditions in standard Ringer’s solution, flux was in the order of 0.1 nmol/cm²/h in a Cldn4-overexpressing cell layer and about 3 nmol/cm²/h in a Cldn8-overexpressing cell layer. In both experiments, flux increased distinctly after the addition of EGTA (final concentration 1.3 mM) to the apical and basolateral side of the cell layer (5.5 and 7.5 nmol/cm²/h, respectively). (c) Averaging the fluxes of three such experiments shows a highly significant (p < 0.01) difference between fluorescein fluxes across Cldn4- and Cldn8-overexpressing claudin quin KO cell layers. After application of EGTA, fluorescein flux increased considerably. Under these conditions, there was no significant difference between Cldn4- and Cldn8-overexpressing claudin quin KO cell layers, indicating a complete loss of barrier function. Statistical analysis was performed with an unpaired, two-sided Student’s t-test, ** p < 0.01, ns—not significant.

Figure 4

Fluorescein flux is not affected by a reduction in TER caused by the stimulation of a transcellular transport pathway. TER and apical to basolateral fluorescein flux (induced by application of 100 µM fluorescein to the apical compartment of an Ussing chamber) were recorded simultaneously on a HT-29/B6 cell layer before and after induction of transcellular chloride secretion by apical application of forskolin (10 µM final concentration). Forskolin caused a reduction in TER, but fluorescein flux remained unaltered. Subsequently, the paracellular pathway was manipulated by a step-wise addition of EGTA to the apical and basolateral compartments (final concentration of 1.2 mM). Similar to that displayed in Figure 3, this caused a further drop in TER, which was accompanied by an increase in fluorescein flux.

Figure 5

(a,b) Original recording of TER and apical to basolateral fluorescein flux measurement across cell layers of claudin quin KO stably transfected to overexpress claudin-2. The Ussing chamber was filled with 10 mL on each of the apical and basolateral sides. Where indicated, 10 µL of a 100 mM fluorescein stock solution was added to the apical side, inducing an increase in signal on the basolateral side. (a) Under control conditions in standard Ringer’s solution, flux was 0.96 nmol/cm²/h. For mannitol application, 5 mL of the basolateral Ringer’s solution was replaced by 5 mL of a stock solution containing 1 M mannitol in Ringer’s solution. This decreased the fluorescein concentration by 50%, causing an instantaneous drop in the recorded signal. Following the application of mannitol, TER dropped, and fluorescein flux increased distinctly to 2.17 nmol/cm²/h. (b) Under control conditions in standard Ringer’s solution, flux was 1.64 nmol/cm²/h. For mannitol application, 5 mL of the apical Ringer’s solution was replaced by 5 mL of a stock solution containing 1 M mannitol and 100 µM fluorescein in Ringer’s solution, ensuring constant fluorescein concentration on the apical side. Following the application of mannitol, TER increased, and fluorescein flux decreased distinctly to 0.92 nmol/cm²/h.

Figure 6

Digital design and 3D printout of components. (a) Digital design of flow-through chamber, with position of LED, optical filter, and photodiode indicated. (b) Digital design of light-shielding encasement, with position of flow-through chamber indicated. (c) 3D printout of flow-through chamber. Windows covered with coverslips. (d) 3D printout of light-shielding encasement with compartments for the optical filter and the flow-through chamber and a separate lid.

Figure 7

Electric circuit of the unit energizing LED and photodiode, and amplifying photodiode output. The following parts were used: BPW 34 Photodiode, TL061 operational amplifier, OP07 operational amplifier, 12 × 3 MΩ resistors, 10 kΩ resistor, 2.2 kΩ resistor, 80 Ω resistor, 5 mm LED (blue, 11,000 mcd, 30°), and 24 V power supply.

Figure 8

(a) Ussing chamber with integrated flux measurement device. (b) Detail of (a). Component a has the same dimensions as the flow-through chamber used for flux measurements and serves the volume adjustment on the donor side. b is the light shielding for the measuring device. It encloses the flow-through chamber (direction of flow indicated by blue arrows), the optical filter, the LED, and the photodiode. c is the inlet for the LED, d shows the connecting wires for the photodiode, the inlet of which is on the back side.