A Novel Approach to Restore Epithelial Barrier Homeostasis to Treat Inflammatory Bowel Disease

SUMMARY Intestinal barrier function is compromised in enteric and systemic diseases, including infectious enterocolitis, food allergy, celiac disease, graft versus host disease (GvHD), and inflammatory bowel disease (IBD). The co-founders of Thelium Therapeutics discovered the central role of myosin light chain kinase (MLCK) in barrier regulation and demonstrated that targeted intestinal epithelial MLCK inhibition limits experimental IBD and GvHD. Unfortunately, severe toxicities associated with barrier-independent MLCK functions in epithelia and other tissues, e.g., smooth muscle, preclude therapeutic targeting of MLCK enzymatic activity. We recently reported (Graham et al., Nature Medicine, 2019) that a specific MLCK splice variant, MLCK1, is central to barrier regulation and depends on interactions mediated by immunoglobulin-cell adhesion molecule domain 3 (IgCAM3). We solved the IgCAM3 crystal structure, identified a drug binding pocket unique to IgCAM3, and screened a library of ~140,000 drug-like molecules. Our tool compound, Divertin, bound IgCAM3, prevented cytokine-induced MLCK1 recruitment, myosin II regulatory light chain phosphorylation, and barrier dysfunction. Critically, Divertin did not inhibit MLCK enzymatic function, epithelial wound healing, or smooth muscle contraction, and in vivo toxicity studies failed to identify adverse effects. Divertin prevented acute TNF-induced barrier loss in vivo (mice) and ex vivo (human intestinal biopsies), and restored immune-mediated barrier loss in vivo (IL-10 knockout mice). Finally, Divertin delayed onset and prevented progression of experimental immune-mediated (T cell transfer) IBD, as indicated by barrier preservation and restoration, reduced mucosal immune activation, and enhanced survival. In a Phase I SBIR, we advanced this program through the discovery of hit compounds with improved activities. Hit compounds were discovered through a rigorous rank order screening funnel to identify compounds with suitable MLCK1 binding affinities, efficacy in preserving and restoring epithelial barrier function, and absence of enzymatic inhibitory activity. In this Phase II proposal, these hit compounds will be optimized with iterative structure activity relationship studies and tested for toxicity and efficacy in mouse models of IBD. This work will facilitate IND-enabling studies for a first-in-class and best-in-class barrier-restorative therapy to manage gastrointestinal and systemic diseases.