Distinct steady-state properties and TNF responses of epithelial long myosin light chain kinase (MLCK) splice variants

Intestinal epithelia express two long myosin light chain kinase (MLCK) splice variants, MLCK1 and MLCK2. Unlike MLCK2, MLCK1 is concentrated at the perijunctional actomyosin ring and this localization is enhanced by tumor necrosis factor (TNF) signaling. Here we sought to identify and characterize the domain(s) that direct basal and TNF-induced MLCK1 subcellular localization. Quantitative morphometry demonstrated specific increases in MLCK1 expression and perijunctional localization in Crohns disease patient biopsies, relative to controls. TNF induced perijunctional recruitment of MLCK1-EGFP but did not affect localization of MLCK2-EGFP, which was predominantly associated with basal stress fibers. Recombinant N-terminal MLCK1 and MLCK2 regions accelerated actin polymerization in vitro but were not different from one another. In contrast, the affinity of N-terminal MLCK1 binding to F-actin was greater than that of MLCK2. Perijunctional MLCK1 and MLCK2 domain recruitment in intestinal epithelial cells paralleled in vitro F-actin binding. The unique MLCK1 Ig3 domain was necessary, but not sufficient, for both F-actin binding and perijunctional recruitment, but, nevertheless, displaced perijunctional MLCK1, enhanced steady-state barrier function, and limited TNF-induced MLCK1 recruitment and barrier loss. These data demonstrate selective perijunctional MLCK1 recruitment in Crohns disease, suggest that F-actin binding contributes to perijunctional recruitment, and show that Ig3 can act as a dominant negative effector that limits TNF-induced MLCK1 recruitment and barrier loss. These results data provide key mechanistic detail that will enable development of therapeutics that target Ig3, or its intercellular binding partners, to reverse inflammation-induced barrier loss and limit disease progression.