Paris-Lodron University Salzburg, PLUS
Department of Biosciences and Medical Biology

Paul Wätzig, Gernot Posselt, Silja Weßler

Background

Caco-2 cells grown on permeable Transwell inserts at an air–liquid interface (ALI) are commonly used to model the intestinal epithelial barrier. In this setup, the cells form a polarized monolayer that separates an apical, lumen-facing side from a basolateral, tissue-facing side. However, standard culture conditions do not reproduce the physiological oxygen gradient found in the gastrointestinal tract, where the tissue side is oxygenated while the lumen is microaerophilic or anaerobic. Establishing such an oxygen gradient in vitro could make the model more physiologically relevant and support infection studies with oxygen-sensitive bacteria such as Helicobacter pylori.

Set-Up/Sample

The experiment compared three culture conditions: a fully oxygenated Caco-2 monolayer as the normoxic reference control, a completely oxygen-deprived culture as the negative control, and a newly developed gradient model. In the gradient model, the apical compartment was maintained at low oxygen levels while oxygen remained available from the basolateral side, thereby closely resembling the physiological situation in the gastrointestinal tract (Figure 1).

Oxygen levels were monitored independently on both sides of the Transwell membrane throughout the experiment. In parallel, cell survival and metabolic activity were assessed to determine whether the epithelial monolayer remained viable and metabolically active under these conditions.

Results

Figure 2 shows the placement of the oxygen sensors during measurements in the gradient system. Oxygen levels in the basolateral medium were measured contactless in triplicate with oxygen sensor spots, while the apical oxygen concentration was monitored by inserting a needle-type oxygen probe into the sealed apical compartment. The resulting oxygen profiles for the normoxic control, oxygen-deprived control, and gradient model are presented in Figure 3. In the gradient model, the apical oxygen concentration was maintained at approximately 5% O₂, while the basolateral medium remained at around 25% air saturation; the normoxic and oxygen-deprived controls showed the expected oxygen profiles. To assess how the different oxygen conditions affected the cells, cell damage was quantified by measuring LDH release, while metabolic activity was determined using the CellTiter-Blue assay (Figure 4). In the gradient system, LDH release indicated approximately 30% cytotoxicity, accompanied by an approximately 30% reduction in metabolic activity after 48 hours, while both control conditions showed the expected responses.

Conclusion

Oxygen measurements in the ALI system were successfully established using PyroScience sensor spots and a needle-type oxygen probe. The custom gradient system enabled separation of the apical and basolateral compartments, which was confirmed by the distinct oxygen profiles measured on both sides. These oxygen measurements were consistent with the biological readouts: cells under fully oxygen-deprived conditions showed strong loss of viability and metabolic activity, whereas normoxic cultures remained largely unaffected. In the gradient system, metabolic activity decreased by approximately 30% and moderate cell damage was observed, indicating that the current setup still requires optimization. Guided by the PyroScience measurements, future improvements will focus on increasing oxygen supply to the basolateral medium to support long-term culture under a stable and physiologically relevant oxygen gradient.