LOSS OF CGNL1 CONTRIBUTES TO VERY EARLY ONSET IBD

Abstract INTRODUCTION Very early onset (VED)-IBD is often associated with a severe disease course and unique characteristics including an underlying monogenic defect. We identified new pathogenic mutations in CGNL1 associated with VEO-IBD. CGNL1 encodes an epithelial junctional protein whose function was previously unknown in GI disease. METHODS Whole exome sequencing (WES) was performed on the affected patient with VEO-IBD and all first-degree relatives with IBD and other inflammatory disease. Human intestinal organoids (HIO) from the patient and controls were characterized by transepithelial electrical resistance (TEER), FITC-dextran permeability assay, immunofluorescence (IF), RNA-seq, and transmission electron microscopy (TEM). We generated a new Cgnl1-/- mouse model, which was characterized using RNA-seq, flow cytometry, permeability assay, histology/IF, TEM, colitis/infection models, 16S rRNA-seq, etc HIOs and mice were also treated by chemical chaperones TUDCA and 4-PBA to reduce endoplasmic reticulum (ER) stress. RESULTS In the affected patient with VEO-IBD who required early total colectomy and pouch formation with coexistent primary sclerosing cholangitis (PSC) and celiac disease, WES identified two rare, new, pathogenic variants in CGNL1. Both pathogenic variants were also present in the patient’s younger sister, who has a similar history including early-onset, fulminant UC requiring total colectomy and PSC-autoimmune hepatitis. CGNL1 protein diminished in patient’s intestinal epithelium that is associated with spontaneous ER stress. Patient’s HIO form smaller 3D spheroids with altered cell composition, skewed microvilli, impaired junctional complexes and barrier function (Fig 1). Similarly, we observed shortened microvilli, disrupted junctional complexes, enhanced epithelial ER stress, and elevated intestinal permeability in Cgnl1-/- mice at steady state. Cgnl1-/- mice show Paneth cell metaplasia in the colon and increased fecal lipocalin-2, a marker of early intestinal inflammation. Moreover, Cgnl1-/- mice exhibit substantial vulnerability to DSS-induced colitis and C. difficile infection that directly target intestinal epithelium. Treatment of TUDCA or 4-PBA was sufficient to alleviate ER stress and restore epithelial barrier in both patient’s HIO and Cgnl1-/- mice; they also rescued Cgnl1-/- mice’s susceptibility to colitis and infections (Fig 2). CONCLUSION Pathogenic CGNL1 variants contribute to VEO-IBD by disrupting gut barrier integrity and activating epithelial ER stress, as demonstrated by experiments using patient’s specimens, HIOs, and a new murine model. This study expands our understanding of epithelial barrier impairment in VEO-IBD and highlights intestinal epithelium as a therapeutic target. Chemical chaperones as novel therapies for IBD caused by primary intestinal barrier defects warrant further investigations.