Telomere dysfunction instigates inflammation in inflammatory bowel disease

Abstract

Inflammatory bowel disease (IBD) is a chronic inflammatory condition driven by diverse genetic and nongenetic programs that converge to disrupt immune homeostasis in the intestine. We have reported that, in murine intestinal epithelium with telomere dysfunction, DNA damage-induced activation of ataxia-telangiectasia mutated (ATM) results in ATM-mediated phosphorylation and activation of the YAP1 transcriptional coactivator, which in turn up-regulates pro-IL-18, a pivotal immune regulator in IBD pathogenesis. Moreover, individuals with germline defects in telomere maintenance genes experience increased occurrence of intestinal inflammation and show activation of the ATM/YAP1/pro-IL-18 pathway in the intestinal epithelium. Here, we sought to determine the relevance of the ATM/YAP1/pro-IL-18 pathway as a potential driver of IBD, particularly older-onset IBD. Analysis of intestinal biopsy specimens and organoids from older-onset IBD patients documented the presence of telomere dysfunction and activation of the ATM/YAP1/precursor of interleukin 18 (pro-IL-18) pathway in the intestinal epithelium. Employing intestinal organoids from healthy individuals, we demonstrated that experimental induction of telomere dysfunction activates this inflammatory pathway. In organoid models from ulcerative colitis and Crohn's disease patients, pharmacological interventions of telomerase reactivation, suppression of DNA damage signaling, or YAP1 inhibition reduced pro-IL-18 production. Together, these findings support a model wherein telomere dysfunction in the intestinal epithelium can initiate the inflammatory process in IBD, pointing to therapeutic interventions for this disease.

Keywords: DNA damage; Yap1; inflammatory bowel disease; pro-IL-18; telomere dysfunction.

Fig. 1.

Telomere shortening in inflammatory bowel disease patients correlates with disease severity. (A) FISH for telomeres in tissue from healthy controls or IBD patient biopsies (control, n = 17; UC, n = 20; CD, n = 23). Fluorescence was quantified as detailed in Materials and Methods indicating telomere length. (Scale bars, 10 µm.) Insets show magnification of the epithelial cells. (B) Quantification of telomere FISH from healthy and IBD biopsies (control, n = 17; UC, n = 20; CD, n = 23). (C) Kendall’s correlation test was performed indicating significant association between telomere length and DAI (disease activity index). (Kendall’s tau = –0.49, Kendall’s P = 5.3e-05). *Statistically significant, P < 0.05 by unpaired Student’s t test. n, number of mice or patient biopsies used in the study. Data are represented as mean ± SEM. Experiments were conducted at least two independent times. Also refer to SI Appendix, Table S1.

Fig. 2.

DNA damage in the IBD patient epithelium activates YAP1-mediated inflammation. (A) Differentially expressed genes from human IBD RNA-seq and mouse RNA-seq (G0 vs. G4) overlapped to identify similarities in the genes and pathways that are deregulated in IBD patients. Top 25 pathways are shown. (B) Representative images of γH2AX immunohistochemistry in colonic biopsies from either healthy (control) or IBD patients showing representative images from each group. (Scale bars, 100 µm.) (C) Histogram representing quantification of staining intensity for γH2AX of colon biopsies from either healthy or IBD patients (control, n = 20; CD, n = 20; UC, n = 20). control vs. UC, P = 0.0001 and control vs. CD, P = 4.088e-06. (D) Immunohistochemistry on the colonic epithelium from healthy (control) and IBD patient biopsies for YAP1 showing representative images from each group (control, n = 20; UC, n = 20; CD, n = 20). (Scale bars, 30 µm.) (E) Histogram depicting the quantification of staining intensity for YAP1 in the colonic epithelium of healthy (control) and IBD patients. Control vs. UC, P = 0.0031 and control vs. CD, P = 0.0001. (F) IL-18 immunohistochemistry in colonic biopsies from either healthy (control, n = 20) or IBD patients (n = 40) showing representative images from each group. (Scale bars, 100 µm.) (G) Histogram shows the quantification of staining intensity for IL-18 from the colon biopsies of both healthy and patients with IBD. Control vs. UC, P = 2.097e-05 and control vs. CD, P = 2.933e-05. *n represents number of patient biopsies used in the study. Each experiment was conducted at least two times.

Fig. 3.

Induction of telomere dysfunction mediates immune responses in human intestinal epithelium. (A) Western blotting performed with the indicated antibodies on protein lysates from organoids transfected with the DOX-inducible mutant TRF2 (iTRF2ΔBΔM) plasmid and treated with or without doxycycline for 7 d. (B) Micrographs of TIF assessment of the cultured organoids treated with or without doxycycline for 7 d. (C) Measurement of TIFs per nucleus counted in 50 cells of the organoids treated with or without doxycycline for 7 d. (Scale bar, 10 μm.) (D) RNA-seq analysis identifying differentially regulated pathways in the mutant organoids compared with the controls. Top 10 up-regulated (red) and down-regulated (blue) pathways are depicted. (E) GSEA graphs for the indicated pathways up-regulated in organoids expressing mutant TRF2 (TRF2ΔBΔM). *Statistically significant, P < 0.05 by unpaired Student’s t test. Data are represented as mean ± SEM. Experiments were conducted at least two independent times.

Fig. 4.

Genetic and pharmacologic activation of telomerase or inhibition of YAP1 suppresses immune activation pathways in IBD epithelium. (A) Micrograph of TIFs in CD duodenal organoids with or without overexpression of telomerase. (Scale bar, 10 µm.) (B) Quantification of TIFs in duodenal organoids from CD patient with or without overexpression of telomerase. Fifty cells from each group were quantified for TIFs. (C) Western blotting of lysates from the CD organoid line transduced with an inducible hTERT overexpression plasmid treated with or without (−DOX) doxycycline for 1 or 2 wk. (D) Western blotting of lysates from control (n = 2), CD (n = 2), and UC (n = 2) organoids with the indicated antibodies, indicating up-regulation of DNA damage and the ATM/YAP1/IL18 axis. (E and F) Western blotting for the indicated antibodies of lysates from older-onset CD organoids treated with or without TA-65 or NAD for 10 d (E) and quantification (F). (n = 2). (G and H) Western blotting for the indicated antibodies of lysates from older-onset UC organoids treated with or without TA-65 or NAD for 10 d (G) and quantification (H). (n = 2). (I and J) Western blotting with the indicated antibodies of lysates from YAP inhibitor treated or untreated CD organoids (I) and quantification (J). (n = 2). (K and L) Western blotting with the indicated antibodies of lysates from YAP inhibitor treated or untreated UC organoids (K) and quantification (L). (n = 2). n represents number of organoid lines used in the study. Each experiment was conducted at least two times.