Intestinal Inflammation and Dysregulated Immunity in Patients With Inherited Caspase-8 Deficiency

Caspase-8 (CASP8) is an initiator cysteinyl aspartate-specific protease critically involved in mediating cell death but also controls several nonapoptotic functions.1Oberst A. et al.Nat Rev Mol Cell Biol. 2011; 12: 757-763Crossref PubMed Scopus (134) Google Scholar In mice, CASP8 has been shown to regulate inflammatory conditions of the skin, liver, and gut in response to injuries or infections.2Ben Moshe T. et al.Hepatology. 2007; 45: 1014-1024Crossref PubMed Scopus (68) Google Scholar, 3Kovalenko A. et al.J Exp Med. 2009; 206: 2161-2177Crossref PubMed Scopus (156) Google Scholar, 4Gunther C. et al.Nature. 2011; 477: 335-339Crossref PubMed Scopus (628) Google Scholar The critical role of CASP8 in human immunity has been demonstrated by the identification of 2 siblings with autoimmune lymphoproliferative syndrome (ALPS).5Chun H.J. et al.Nature. 2002; 419: 395-399Crossref PubMed Scopus (564) Google Scholar Dysfunction of caspases has been implicated in inflammatory bowel disease (IBD) and colorectal cancer but the clinical relevance remains unclear.6Becker C. et al.Gastroenterology. 2013; 144: 283-293Abstract Full Text Full Text PDF PubMed Scopus (72) Google Scholar Here, we identified CASP8 deficiency as a novel cause for very early onset IBD (VEO-IBD) in 3 patients from unrelated consanguineous kindred using whole-exome sequencing. Patients P1 and P2 had a novel homozygous CASP8 mutation (c.710A>G, p.Q237R) segregating with the disease phenotype and causing impaired protein expression, whereas P3 carried the previously reported mutation (c.793C>T, p.R265W5Chun H.J. et al.Nature. 2002; 419: 395-399Crossref PubMed Scopus (564) Google Scholar) (Figure 1A and Supplementary Figures 1 and 2). In contrast to CASP8-deficient patients with ALPS-like disorder,5Chun H.J. et al.Nature. 2002; 419: 395-399Crossref PubMed Scopus (564) Google Scholar our patients presented with VEO-IBD characterized by failure to thrive, diarrhea, perianal disease, and discontinuous severe structuring and fistulizing proctocolitis with deep ulcerations and epithelial degeneration (Figure 1B). The colitis was refractory despite exclusive enteral nutrition with elemental formula, anti-inflammatory treatment (steroids, azathioprine, infliximab), and ileostomy (P1) or right hemicolectomy (P3). All patients presented with increased susceptibility to bacterial and viral infections and P2 died of septic complications. Further clinical details are provided in the supplementary material. Immunophenotyping of peripheral blood mononuclear cells from P1 revealed an altered distribution of T cells with decreased CD45RO+CCR7+ central memory, CD45RO+HLA-DR− memory regulatory, and CCR6+CXCR3− Th17 CD4+ cells, whereas CCR6−CXCR3+ Th1 cells were increased (Figure 1C and Supplementary Figure 3A). T-cell proliferation and activation were reduced in P1 and P3 on stimulation with anti-CD3 (Supplementary Figure 3B–D). In addition to T-cell dysfunctions, P1 showed reduced CD19+IgD+CD27+ marginal zone and CD19+IgD−CD27+ class-switched B cells but increased CD19+IgMhiCD38++ transitional B cells, indicating impaired B-cell maturation (Figure 1C and Supplementary Figure 4). Studies on conditional Casp8−/− mice showed that dendritic cells exhibit increased NLRP3-mediated inflammasome activity.7Kang T.B. et al.Immunity. 2013; 38: 27-40Abstract Full Text Full Text PDF PubMed Scopus (326) Google Scholar Human CASP8-deficient monocytes from P1 secreted increased interleukin (IL)-1β level in response to lipopolysaccharide (LPS) priming (Figure 1D), whereas addition of ATP failed to induce robust NLRP3 inflammasome activation. Cytotoxicity of LPS-stimulated patient monocytes was comparable to healthy donors (HDs). Immunoblotting revealed reduced expression of key mediators of inflammasome activation (NLRP3, ASC, cleaved CASP1) in cell lysates and the active form of IL-1β in supernatants of patients’ cells (Figure 1E). Collectively, our studies suggested increased secretion of IL-1β on LPS-induced inflammasome activation in primary patients’ cells but could not provide unambiguous insights into the molecular mechanisms of CASP8 dysfunction due to limited access to patient material. To assess the effects of the CASP8 mutations on inflammasome biology in greater detail, we used a BLaER1 monocyte transdifferentiation model.8Gaidt M.M. et al.Immunity. 2016; 44: 833-846Abstract Full Text Full Text PDF PubMed Scopus (434) Google Scholar BLaER1 cells with CASP8 knockout or lentiviral overexpression of the CASP8 mutant Q237R showed increased IL-1β secretion on LPS priming, as compared with wild-type CASP8 reconstituted cells (Figure 1F). No measurable alteration of inflammasome activity could be observed in BLaER1 cells with expression of the CASP8 variant R265W. Elevated IL-1β levels in CASP8-deficient macrophages were associated with increased lactate dehydrogenase release. Blockade of NLRP3-mediated inflammasome activity (MCC950) or necroptosis (NSA) with small molecule inhibitors resulted in decreased LPS-triggered IL-1β production, suggesting that both biological pathways might mediate increased proinflammatory responses in the absence of CASP8. Correspondingly, CASP8-deficient BLaER1 cells showed a distinct pattern of MLKL oligomerization after LPS priming (Supplementary Figure 5), indicating altered necroptosome signaling. Thus, our studies in patients’ and BLaER1 cells confirmed that CASP8 deficiency alters inflammasome activity. Deletion of mouse Casp8 in the intestinal epithelium leads to tumor necrosis factor (TNF)-α–mediated epithelial cell destruction.4Gunther C. et al.Nature. 2011; 477: 335-339Crossref PubMed Scopus (628) Google Scholar In contrast, we could not detect a difference in the frequency of cleaved CASP3-positive cells on stimulation with TNF-α between intestinal organoids from HDs and P1 (Figure 2A and B). Whereas TRAIL stimulation induced CASP3 cleavage and cell death in HD intestinal organoids, CASP8-deficient cells were unresponsive to TRAIL, confirming abrogated CASP8-mediated signaling. To study whether the identified mutations affect the necroptotic pathway, we engineered HT-29 colon carcinoma cells with CRISPR/Cas9-mediated knockout of CASP8 and reconstitution of wild-type or mutant variants. HT-29 cells with CASP8 knockout and expression of the CASP8 variant Q237R exhibited an enhanced pattern of MLKL oligomerization and cell death on treatment with TNF-α and the SMAC mimetic BV6, suggesting increased necroptosis (Figure 2C and D). In contrast, no difference in the frequency of Annexin V+ cells could be detected in response to TNF-α (Figure 2D). Taken together, studies on patient-derived intestinal organoids and heterologous colon carcinoma cells suggest that human CASP8 dysfunctions perturb cell death responses of the intestinal epithelium. Whereas CASP8-deficient mice developed a TNF-α–induced terminal ileitis associated with elevated RIPK3 expression,4Gunther C. et al.Nature. 2011; 477: 335-339Crossref PubMed Scopus (628) Google Scholar our patients presented with colonic inflammation and TNF-α blockade failed to induce clinical remission. CASP8 might represent another example that genetic alterations of human and mouse orthologs result in phenotypic discrepancies.9Liao B.Y. et al.Proc Natl Acad Sci U S A. 2008; 105: 6987-6992Crossref PubMed Scopus (186) Google Scholar Residual CASP8 protein expression in our patients and redundant functions by CASP10,10Fischer U. et al.Oncogene. 2006; 25: 152-159Crossref PubMed Scopus (86) Google Scholar a human homolog of CASP8, might explain phenotypic variabilities and compatibility with life in contrast to constitutive Casp8 knockout mice.11Varfolomeev E.E. et al.Immunity. 1998; 9: 267-276Abstract Full Text Full Text PDF PubMed Scopus (1029) Google Scholar Moreover, our patients showed a strikingly different disease from patients with ALPS-like disorder,5Chun H.J. et al.Nature. 2002; 419: 395-399Crossref PubMed Scopus (564) Google Scholar, 12Niemela J. et al.J Clin Immunol. 2015; 35: 348-355Crossref PubMed Scopus (25) Google Scholar and we observed genotype-dependent variabilities in cellular assays with respect to inflammasome activity and cell death. Further studies are required to assess whether the differences in the phenotype of CASP8 deficiency are due to the type of mutation, genetic modifiers, microbiome composition, inflammation, infections, or medication. In conclusion, our study reveals CASP8 deficiency as a novel cause for VEO-IBD associated with lymphocyte dysfunction, impaired inflammasome activation, and defective epithelial cell death responses. Our findings underline the critical apoptotic and nonapoptotic functions of human CASP8 in maintaining intestinal immune homeostasis and might warrant awareness about potential toxicities of therapeutic strategies targeting CASP8. We are very grateful to our patients and their parents for allowing us to study their diseases. Our work is dedicated to the patients, in particular P2, who died during the course of the study. We thank the interdisciplinary medical staff at the Dr. von Hauner Children’s Hospital, Sheba Medical Center, and American University of Beirut Medical Center. We acknowledge the assistance of the Flow Cytometry and Care-for-Rare Genomics Core Facility at the Dr. von Hauner Children’s Hospital. We kindly acknowledge Drs T. Meitinger and T. M. Strom for help with exome sequencing (Institute of Human Genetics, Technische Universität München, and SyNergy, Munich Cluster for Systems Neurology, Ludwig Maximilians Universität München, Germany). The sequencing of family C was conducted at the Sheba Medical Center. We kindly thank Genentech (South San Francisco, CA) for supplying BV6, Dr T. Graf (Center for Genomic Regulation, Barcelona, Spain) for providing BLaER1 cells, P. Dinse for assisting with histology of organoids, and Prof Dr Bartenstein and his team for providing positron emission tomography–computed tomography images (Department of Nuclear Medicine, LMU Munich, Germany). Further, we gratefully acknowledge our deceased bioinformatician Dr Jacek Puchalka. Members of the VEO-IBD Consortium are as follows: Hans Clevers, Dror S. Shouval, Aleixo M. Muise, Scott B. Snapper, Sibylle Koletzko, Christoph Klein, and Daniel Kotlarz. Author contributions: Anna S. Lehle designed and conducted experiments and analyzed the data. Henner F. Farin, Birgitta E. Michels, and Hans Clevers supported organoid culture experiments. Benjamin Marquardt helped in cloning and lentiviral transduction. Thomas Magg performed T-cell assays, Maryam Ghalandary conducted B-cell proliferation assays, and Raffaele Conca supported immunophenotypical analysis. Yue Li assisted in apoptosis assays and Yanshan Liu supported the CRISPR/Cas9-mediated genetic engineering. Katja Lammens performed structural analysis. Olaf Groß provided protocols and supported analysis of inflammasome activity. Moritz M. Gaidt and Veit Hornung provided the CASP8-deficient BLaER1 model and experimental expertise. Batia Weiss, Atar Lev, Amos J. Simon, Rima Hanna-Wakim, Nadine Yasbeck, Dror S. Shouval, Raz Somech, Fabian Hauck, Aleixo M. Muise, Scott B. Snapper, Philip Bufler, and Sibylle Koletzko recruited and clinically characterized patients and were critical in the interpretation of the human data. Christoph Walz performed histological analysis. Meino Rohlfs conducted whole-exome sequencing in the Next-Generation Sequencing facility at the Dr. von Hauner Children’s Hospital, and Sebastian Hollizeck performed the bioinformatics analysis of sequencing data. Christoph Klein and Daniel Kotlarz conceived the study design, supervised Anna S. Lehle, and recruited study participants. Anna S. Lehle, Christoph Klein, and Daniel Kotlarz wrote the draft of the manuscript. All authors interpreted the data and approved the final version of manuscript.