In Vitro Comparison of Two Python-Based Programs for the Automated Analysis of Tight-Junction Phenotype in Brain Endothelium During Bacterial Infection

Abstract

Tight junction complexes are crucial features of brain endothelial cells, as they restrict the paracellular route across the blood–brain barrier. Tight junction disruption has been observed in conjunction with numerous diseases of the CNS. In such cases, the organization or integrity of cell–cell junctions may be analyzed with a variety of automated computer programs that quantitatively assess junction images. Here, we directly compare two previously developed python-based programs—JAnaP and IJOQ— for the semi- or fully automated analysis of tight junctions in human stem cell-derived brain-like endothelial cells. Cells were infected with S. pneumoniae and S. agalactiae to initiate junction disruption, and occludin and ZO-1 were analyzed in mock and infected groups via JAnaP and IJOQ. JAnaP and IJOQ both yielded comparable results for the quantification of tight junction disruption in brain endothelial cells. While JAnaP rendered data at the cellular level and gave more information regarding junction phenotype, IJOQ significantly reduced user time and eliminated potential user bias. Our results suggest that JAnaP and IJOQ are both appropriate for quantifying tight junction integrity in brain endothelial cells, and both may offer distinct advantages depending on their context of use.