Caspase-8 regulates TNF-α-induced epithelial necroptosis and terminal ileitis

Abstract

Dysfunction of the intestinal epithelium is believed to result in the excessive translocation of commensal bacteria into the bowel wall that drives chronic mucosal inflammation in Crohn's disease, an incurable inflammatory bowel disease in humans characterized by inflammation of the terminal ileum. In healthy individuals, the intestinal epithelium maintains a physical barrier, established by the tight contact of cells. Moreover, specialized epithelial cells such as Paneth cells and goblet cells provide innate immune defence functions by secreting mucus and antimicrobial peptides, which hamper access and survival of bacteria adjacent to the epithelium. Epithelial cell death is a hallmark of intestinal inflammation and has been discussed as a possible pathogenic mechanism driving Crohn's disease in humans. However, the regulation of epithelial cell death and its role in intestinal homeostasis remain poorly understood. Here we demonstrate a critical role for caspase-8 in regulating necroptosis of intestinal epithelial cells (IECs) and terminal ileitis. Mice with a conditional deletion of caspase-8 in the intestinal epithelium (Casp8(ΔIEC)) spontaneously developed inflammatory lesions in the terminal ileum and were highly susceptible to colitis. Casp8(ΔIEC) mice lacked Paneth cells and showed reduced numbers of goblet cells, indicating dysregulated antimicrobial immune cell functions of the intestinal epithelium. Casp8(ΔIEC) mice showed increased cell death in the Paneth cell area of small intestinal crypts. Epithelial cell death was induced by tumour necrosis factor (TNF)-α, was associated with increased expression of receptor-interacting protein 3 (Rip3; also known as Ripk3) and could be inhibited on blockade of necroptosis. Lastly, we identified high levels of RIP3 in human Paneth cells and increased necroptosis in the terminal ileum of patients with Crohn's disease, suggesting a potential role of necroptosis in the pathogenesis of this disease. Together, our data demonstrate a critical function of caspase-8 in regulating intestinal homeostasis and in protecting IECs from TNF-α-induced necroptotic cell death.

Figure 1. Casp8^ΔIEC mice spontaneously develop ileitis and lack Paneth cells

(a) Representative endoscopic pictures and (b) H&E stained cross sections showing villous erosions in the terminal ileum of Casp8^ΔIEC. (c) RT-PCR showing increased level of inflammatory markers in the terminal ileum of Casp8^ΔIEC mice (6 mice per group +SEM, relative to HPRT). (d) Gene ontology analysis of genes significantly downregulated in gene chip analysis of IEC from 3 control and 3 Casp8^ΔIEC mice. (e) Ileum cross sections stained with H&E and lysozyme for Paneth cells (inset = single crypt at higher magnification). Arrows indicate crypt bottom with Paneth cells.

Figure 2. Increased caspase-8 independent cell death within crypts of Casp8^ΔIEC mice

(a) Representative pictures of gut organoids. Inset = Eosin staining indicating Paneth cells. Graph: Number of Paneth cells (pc) per organoid crypt (n=24) +SEM. (b) Crypt cross sections from the small intestine of control and Casp8^ΔIEC mice stained with H&E. TUNEL ( Inset: condensed nuclei) and caspase-3. (c) Quantification of necrotic cells per crypt + SD of control (n=9) and Casp8^ΔIEC (n=14) mice. (d) Electron microscopic pictures of dying crypt cells in Casp8^ΔIEC mice and inducible Casp8^ΔIEC. Asterisks, Paneth cell granules; “n”, nuclei; “l”, crypt lumen; Arrows: mitochondrial swelling.

Figure 3. Inhibition of TNF-α induced epithelial necroptosis in Casp8^ΔIEC mice

(a) Representative RIP3 staining colocalizing with condensed nuclei at the crypt bottom of Casp8^ΔIEC mice. (b) Western blot for RIP3 and cleaved caspase-3 of IEC lysates isolated from TNF-α treated control and Casp8^ΔIEC mice. Actin served as a control. (c) Representative microscopic pictures and (d) cell viability of Casp8^ΔIEC organoids treated for 24 h with TNF-α +/- necrostatin-1. (e) Survival and (f) H&E stained small intestine cross sections of control (n=5), Casp8^ΔIEC (mock pretreated, n=7) and Casp8^ΔIEC (nec-1 pretreated, n=8) mice after intravenous injection of TNF-α. Asterisks show significance level relative to Casp8^ΔIEC without nec-1. All experiment were performed at least 3 times with similar results.

Figure 4. RIP-mediated necroptosis of Paneth cells in patients with Crohn's disease

(a) Representative RIP3 immunostaining of the terminal ileum (healthy patient). TOP: RIP3 expression in human Paneth cells. BOTTOM: Colocalization of lysozyme and RIP3 in Paneth cells. (b) H&E staining of crypts in the terminal ileum. Arrows indicate crypt cells with shrunken eosinophilic cytoplasm and pyknotic nuclei. GRAPH: Number of Paneth cells (+SD) per crypt in control patients (n=7) and patients with active Crohn's disease (n=4). (c) Electron microscopy of the terminal ileum of control and Crohn's disease patient. Asterisks highlight Paneth cell granules, “n” indicates nucleus. (d) Number of crypt cells (+SD) showing organelle swelling but regular nuclei as signs of necroptosis. EM pictures of 4 patients were analyzed. (e) Representative immunofluorescence staining for TUNEL and active caspase-3 in crypts of the terminal ileum of a CD-patient. (f) H&E staining of biopsies from the small intestine of control patients stimulated in vitro with either DMSO (mock), TNF-α alone or in combination with necrostatin-1. GRAPH: Quantitative expression level of the Paneth cell marker lysozyme relative to HPRT. Data from one representative experiment out of 2 is shown. Arrow indicate Paneth cells (a,e) or mitochondrial swelling (c).