Abstract
The intestinal damage can be caused by physical, infectious, and immune-mediated injury. Rapid repair, which is essential for a return to mucosal homeostasis, depends on rapid epithelial migration, i.e., restitution, as well as epithelial proliferation and differentiation. Barrier restoration is a critical component of this repair process, but the contributions of structural proteins that form the barrier to mucosal repair have not been defined.
Aim: To determine the contributions of the intestinal epithelial tight junction protein zonula occludens-1 (ZO-1) to mucosal repair.
Results: Intestinal epithelial ZO-1 transcript and protein expression are reduced in biopsies from inflammatory bowel disease (IBD) patients relative to healthy controls. To determine if this ZO-1 loss contributes to disease pathogenesis, we created intestinal epithelial ZO-1-deficient Tjp1f/f x villin-Cre+ mice. When stressed, for example, with 2% DSS, ZO-1-deficient mice displayed much greater mucosal damage and weight loss relative to ZO-1-sufficient mice and failed to recover fully, even 4 weeks after DSS discontinuation. To better define the defect, mice were injected with nucleoside analogs (BrdU and EdU) to label proliferating cells 1 and 3 days after DSS discontinuation. When assessed 2 hours after labeling, numbers of labeled nuclei were significantly reduced in ZO-1-deficient mice relative to ZO-1-sufficient mice. More striking, however, was the nuclear fragmentation and staining for cleaved caspase-3 at 24 hours after labeling. In vitro studies of colonoids were used to better define the mechanisms of epithelial loss. Colonoids from ZO-1-deficient mice were smaller, had reduced numbers of buds (crypt domains) and Lgr5+ stem cells, and were nonresponsive to enhanced Wnt signaling induced by the GSK3 inhibitor CHIR99021, relative to ZO-1-sufficient colonoids. Live imaging of mitosis showed misoriented mitotic spindles in ZO-1-deficient colonoids that caused one daughter cell to lose contact with the extracellular matrix. Misoriented mitotic spindles were also present in tissues from DSS-treated ZO-1-deficient, not ZO-1-sufficient, mice. Live imaging of colonoids from mRFP1-ZO-1 transgenic mice detected transient accumulation of ZO-1 at the cleavage furrow, suggesting that ZO-1 interactions at this site required for accurate mitotic spindle orientation.
Conclusion: ZO-1 has a previously unappreciated, noncanonical function in mitotic spindle orientation that is independent of barrier maintenance but central to epithelial proliferation and repair. We postulate that, in the absence of ZO-1, loss of contact with the basement membrane leads to anoikis, i.e., detachment-induced apoptosis, and an abortive proliferative response that compromises repair. We speculate that ZO-1 downregulation in IBD may similarly interfere with mucosal healing.
Support: NIH (DK068271, DK061931)
Holocrine Secretion Occurs outside the Tight Junction Barrier in Multicellular Glands: Lessons from Claudin-1–Deficient Mice
