Junction-level imaging and modeling of transient leaks in epithelia predicts a dynamic tight junction strand network
Abstract
Epithelial tight junctions restrict paracellular permeability, but how small solutes leak through the barrier is not fully understood, because permeability is usually measured globally. Here, we combined fast zinc-based barrier imaging (ZnUMBA) with computational modelling of the dynamic tight junction strand network to study transient leaks at single bicellular junctions in MDCK II cells. Leaks were short-lived, with a median duration of 17 s in wild-type cells, and occurred in all cell lines studied, including ZO-1 knockout, ZO-2 knockout, and ZO-1-overexpressing cells. In ZO-2 knockout and ZO-1-overexpressing cells, leaks lasted longer without changes in their frequency or transepithelial electrical resistance (TER). ZO-1 knockout increased the junctional ZnUMBA signal and reduced TER and the number of detected leaks. Simulations showed that stochastic breaking and sealing of tight junction strands is sufficient to explain the observed leak frequencies and durations, and that a single pair of strand break formation and sealing rates reproduces both junction-level leaks and global permeability. Whereas leak frequency mainly reflects the ratio of these rates, leak duration provides independent information on strand sealing that global measurements cannot resolve. Modulating myosin II activity had little effect on leaks, whereas actin depolymerization compromised the barrier. Together, these results link junction-level leaks of ions and small solutes to tight junction strand dynamics.
Related
- Computational modeling of VEGF-A signaling predicts lymphatic junction state
- Welcome to The Junction: Polymeric extracellular interfaces and barriers in plant life
- A framework for designing splice-junction experiments in deep 3' single-cell RNA sequencing
- Iron Excess Stress Promotes Adventitious Root Formation at the Root–Hypocotyl Junction in Arabidopsis
- Quantification of compensatory motor neuron sprouting at the Drosophila larval neuromuscular junction
BibTeX
@misc{porsti2026junctionlevel,
title = {{Junction-level imaging and modeling of transient leaks in epithelia predicts a dynamic tight junction strand network}},
author = {A. Pörsti and A. Tervonen and H. Tran and S. Nymark and T. O. Ihalainen},
howpublished = {bioRxiv},
year = {2026},
month = nov,
doi = {10.64898/2026.11.06.764832},
url = {https://www.biorxiv.org/content/10.64898/2026.11.06.764832v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - Junction-level imaging and modeling of transient leaks in epithelia predicts a dynamic tight junction strand network
AU - A. Pörsti
AU - A. Tervonen
AU - H. Tran
AU - S. Nymark
AU - T. O. Ihalainen
PY - 2026
DA - 2026/11/06
PB - bioRxiv
DO - 10.64898/2026.11.06.764832
UR - https://www.biorxiv.org/content/10.64898/2026.11.06.764832v1
AB - Epithelial tight junctions restrict paracellular permeability, but how small solutes leak through the barrier is not fully understood, because permeability is usually measured globally. Here, we combined fast zinc-based barrier imaging (ZnUMBA) with computational modelling of the dynamic tight junction strand network to study transient leaks at single bicellular junctions in MDCK II cells. Leaks were short-lived, with a median duration of 17 s in wild-type cells, and occurred in all cell lines studied, including ZO-1 knockout, ZO-2 knockout, and ZO-1-overexpressing cells. In ZO-2 knockout and ZO-1-overexpressing cells, leaks lasted longer without changes in their frequency or transepithelial electrical resistance (TER). ZO-1 knockout increased the junctional ZnUMBA signal and reduced TER and the number of detected leaks. Simulations showed that stochastic breaking and sealing of tight junction strands is sufficient to explain the observed leak frequencies and durations, and that a single pair of strand break formation and sealing rates reproduces both junction-level leaks and global permeability. Whereas leak frequency mainly reflects the ratio of these rates, leak duration provides independent information on strand sealing that global measurements cannot resolve. Modulating myosin II activity had little effect on leaks, whereas actin depolymerization compromised the barrier. Together, these results link junction-level leaks of ions and small solutes to tight junction strand dynamics.
ER -