The intestinal epithelium and its resident microbiota form a dynamic interface that is central to gut homeostasis, but this interface is difficult to model with conventional tools. Static Transwell cultures lack physiological fluid shear, whereas animal models are costly and poorly suited to quantitative, high-content imaging. Organ-on-a-chip technology offers a way to recreate key physical and biological features of the gut, yet many existing gut-on-chip systems remain technically demanding or poorly standardized for routine use. Here, we developed and validated a practical gut-on-chip workflow that enables reproducible, imaging-based quantification of epithelial barrier dynamics during short-term co-culture with Escherichia coli. Using a commercial two-channel microfluidic chip and Caco-2 cells, we first established a perfused intestinal model under low-flow conditions (0.5-1 µL/min) and confirmed long-term stability by phase-contrast imaging, LIVE/DEAD staining, and immunofluorescence for ZO-1, F-actin, and wheat germ agglutinin (WGA). This protocol generated a confluent, polarized monolayer with continuous tight junctions and an apical glycoprotein mucus layer. We then introduced HADA-labeled E. coli and systematically varied the perfusion rate during co-culture. At 0.5 µL/min, the epithelial layer, although stable in monoculture, underwent rapid thinning and focal detachment within 4 days of co-culture, indicating that low shear is insufficient to buffer bacterial overgrowth and metabolite accumulation. In contrast, Caco-2 monolayers tolerated perfusion up to 10 µL/min without structural damage, and at this higher flow rate we achieved a stable, 24-h analytical window in which epithelial integrity and bacterial distribution could be imaged reliably. Structural disruption only became evident when co-culture was extended to four and 7 days. Taken together, these results identify perfusion rate as the decisive parameter for stabilizing host-microbe co-culture on a gut-on-chip platform. The workflow described here couples a functionally validated epithelial barrier with an experimentally simple, yet robust, perfusion regime. It provides an accessible test-bed for quantitative studies of early host-microbe interactions and can be readily extended with orthogonal barrier readouts, such as TEER or tracer permeability, in future work.
Chiang et al. (Wed,) studied this question.