Telomere dysfunction activates YAP1 to drive tissue inflammation

Abstract

Germline telomere maintenance defects are associated with an increased incidence of inflammatory diseases in humans, yet whether and how telomere dysfunction causes inflammation are not known. Here, we show that telomere dysfunction drives pATM/c-ABL-mediated activation of the YAP1 transcription factor, up-regulating the major pro-inflammatory factor, pro-IL-18. The colonic microbiome stimulates cytosolic receptors activating caspase-1 which cleaves pro-IL-18 into mature IL-18, leading to recruitment of interferon (IFN)-γ-secreting T cells and intestinal inflammation. Correspondingly, patients with germline telomere maintenance defects exhibit DNA damage (γH2AX) signaling together with elevated YAP1 and IL-18 expression. In mice with telomere dysfunction, telomerase reactivation in the intestinal epithelium or pharmacological inhibition of ATM, YAP1, or caspase-1 as well as antibiotic treatment, dramatically reduces IL-18 and intestinal inflammation. Thus, telomere dysfunction-induced activation of the ATM-YAP1-pro-IL-18 pathway in epithelium is a key instigator of tissue inflammation.

Fig. 1. Telomere dysfunction in the gut epithelia drives inflammation.

a Hematoxylin and eosin staining of colon tissue from 8-month old G0 and G4 mice treated with and without tamoxifen (n = 15), scale bar 100 µm. Arrows indicate cryptitis, crypt abscess and submucosal inflammation. b Immunohistochemical staining for γH2AX (arrows) in the colon of mice at 8-month old G0 and G4 mice treated with and without tamoxifen (n = 3), scale bar 50 µm. c Quantification of the γH2AX stained cells in 100 colonic crypts from G0 and G4 mice treated with and without tamoxifen (n = 3). P values were calculated using two sided t test between G0 and G4 (p < 0.0002), between G4 and G4 + TAM (p < 0.0003). d Change in body weight (g) at time of death of telomere proficient G0 and telomere deficient G4 mice treated with and without tamoxifen (G0, n = 18, G4, n = 15, G0 = TAM, n = 16, G4 + TAM, n = 17). P values were calculated using two sided t test between G0 and G4 (p = <0.0001), between G4 and G4 + TAM (p < 0.0003). e Survival curves of mice of the indicated genotypes treated with or without tamoxifen. Dotted line indicates 50% survival of G4 mice (54.6 weeks) and G4 + TAM mice (75 weeks) (G0 mice, n = 15; G4, n = 24; G0 + TAM, n = 15; G4 + TAM, n = 15). P values were calculated using Mantel-cox test, between G0 and G4 (p < 0.0001) and between G4 and G4 + TAM (p < 0.0001). f–i Percentage of lamina propria and intestinal epithelial immune cells present in the colon of the mice of the indicated genotypes treated with or without tamoxifen (n = 6). P values were calculated using two sided t test between G0 and G4, G4 and G4 + TAM. f CD45, p value between G0 and G4, p = 0.0008, between G4 and G4 = TAM, p = 0.0451. g CD3, p value between G0 and G4, p = 0.0097, between G4 and G4 + TAM, p = 0.0079. h CD4, p value between G0 and G4, p = 0.0022, between G4 and G4 + TAM, p = 0.0551. i CD8, p value between G0 and G4, p < 0.0001, between G4 and G4 + TAM, p = 0.0045. j Hematoxylin and eosin staining of colonic biopsies from patients with mutations in TERT or proteins of the shelterin complex, TIN2 (n = 3). Colonic biopsies from two patients show cryptitis, increased apoptosis, increased chronic inflammation in the lamina propria, and mild crypt distortion. k Immunohistochemistry with the colonic epithelium from healthy pediatric patients/(controls) and patients with telomeropathies for γH2AX (Control, n = 10; telomere dysfunctional patients, n = 3). l Histogram depicting the quantification of staining intensity for γH2AX in the colonic epithelium of healthy (control) and telomere dysfunctional patients. γH2AX, p = 0.0001. P values were calculated using Fisher’s exact test. *statistically significant, p < 0.05 by unpaired Student’s t test, two tailed. 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 Supplementary Fig. 1, Supplementary Fig. 2, and Supplementary Fig. 3 and Supplementary Table 1 and 2.

Fig. 2. Identification of YAP1 as a key regulator of immune pathways including IL-18.

a RNA-seq analysis and pathway analysis of colonic crypts from G0 and G4 mice identified several inflammation-related pathways (n = 2). b Gene set enrichment analysis curve from the RNA-seq analysis in panel (a) depicting the innate immune system pathway with a significant p value of 0.02. (n = 2). c Gene set enrichment analysis curve depicting the inflammasome pathway that was identified as the topmost deregulated pathway by RNA-seq from the GFP⁺ sorted colonic epithelial cells of G0 and G4 mice (n = 3). d Top cell process pathways identified by GSEA analysis from the GFP⁺ colonic sorted cells of G0 and G4 mice. (n = 2). e ELISA for IL-18 with the colonic lysates from the mice of the indicated genotypes (n = 5). P values were calculated using two sided t test between G0 and G4, G4 and G4 + TAM. P value between G0 and G4, p < 0.0001, between G4 and G4 + TAM, p = 0.0031. f Immunoblotting with caspase-1 and IL18 antibody of colonic epithelial lysates from G0 and G4 mice treated with or without tamoxifen (n = 2). g Immunoblotting of enteroid lysates from G0 and G4 mice treated with or without tamoxifen with the indicated antibodies (n = 2). h qRT-PCR analysis of the inflammasome pathway genes from enteroids derived from the G0 and G4 mice treated with or without tamoxifen, pro-IL-18 (n = 3). P values were calculated using two sided t test between G0 and G4, G4 and G4 + TAM. P value between G0 and G4, p = 0.0067, between G4 and G4 + TAM, p = 0.0063. i GSEA plot showing an enrichment of Yap1 gene signature in the G4 epithelium compared to the G0 epithelium (p = 0.006). j Overlap of genes identified with YAP1 antibody in G0 and G3 crypt epithelium. k Gene enrichment analysis for the differentially bound genes by YAP1 in the G3 epithelium (n = 4). l ChIP-seq data showing increased YAP1 occupancy at the promoters and the gene body of the inflammasome pathway genes in G3 telomere dysfunctional epithelium compared to the G0 telomere proficient epithelium. m Chromatin immunoprecipitation for YAP1 or IgG in the G0 and G3 enteroids treated with or without tamoxifen. qRT-PCR for the indicated genes was performed with primers directed towards the specific site boxed in red and a non-specific site 1Kb away from the YAP1 binding site. (n = 3). P values were calculated using two sided t test between G0 and G4, G4 and G4 + TAM. P value between G0 and G4, p = 0.0013, between G4 and G4 + TAM, p = 0.0003. *statistically significant, p < 0.05 by unpaired Student’s t test, two-tailed and Fisher’s Exact Test. n, number of mice used in the study. Data are represented as mean ± SEM. Experiments were conducted at least two independent times. Also refer to Supplementary Fig. 2.

Fig. 3. Telomere dysfunction activates YAP1.

a Immunoblotting of intestinal enteroid lysate from G0 and G4 mice treated with or without tamoxifen for the indicated antibodies (n = 2). b Immunohistochemistry for YAP1 in the colonic crypts from the G0 and G4 mice treated with or without tamoxifen (n = 3). Scale bars, 50 µm. Insets are shown to illustrate the higher nuclear localization of YAP1 in the G4 mice than in others. c Immunohistochemistry with the colonic epithelium from healthy pediatric patients(control) and patients with telomeropathies for YAP1 and IL-18. (Control, n = 10; telomere dysfunctional patients, n = 3) d Histogram depicting the quantification of staining intensity for YAP1 and IL-18 in the colonic epithelium of healthy (control) and telomere dysfunctional patients. YAP1, p = 0.0035 IL-18, p = 0.0001. (Control, n = 10; telomere dysfunctional patients, n = 3). e qRT-PCR for the indicated genes from enteroids isolated from the YAP1(S127A) transgenic mouse treated with or without doxycycline to inhibit degradation and increase stabilization and nuclear localization (n = 3). P values were calculated using two sided t test between dox untreated and dox treated organoids. The following are the p-values for pro-Il18, p < .0001, Vgll3, p = 0.0067, Cyr61, p < 0.0001. f Microscopic images of the enteroids from a control vector transduced or YAP1 shRNAs transduced enteroids (n = 2). Scale bars, 30 µm. g qRT-PCR for the indicated genes from G4 enteroids transduced with either a control or YAP1 shRNAs (n = 2). Histogram shown with individual bar for each sample. h Western blot for the indicated proteins in CRL-1831 cells overexpressing TRF2ΔBΔM plasmid or control vector 1-day post selection with puromycin. i Cell fractionation performed with CRL-1831 cells overexpressing TRF2ΔBΔM plasmid or control vector 1-day post selection with puromycin. j Immunohistochemistry for YAP1 in irradiated or control G0 colon crypts showing higher nuclear positive staining for YAP1 after irradiation. Scale bars, 50 μm. k Western blot for the YAP1 and pYAPY357 phosphorylated proteins in irradiated or control G0 colon crypts showing higher levels of phosphorylation after irradiation (n = 3). l qRT-PCR of RNA from control or irradiated colonic G0 mouse crypts for pro-IL-18 (n = 3). P values were calculated using two sided t test, p = 0.0122. *statistically significant, p < 0.05 by unpaired Student’s t test, two-tailed and Fisher’s Exact Test. n represents number of mice used in the study. “Ab” denotes antibody and “I” denotes Input samples. Each experiment was conducted at least two times. Data are represented as mean ± SEM. Also refer to Supplementary Fig. 3.

Fig. 4. Inhibition of ATM and YAP1 reduces IL-18 secretion and ameliorates inflammation.

a Immunoblots of intestinal enteroid lysate from G4 enteroids treated with or without ATM or YAP1 inhibitors for the indicated antibodies (n = 2). b qRT-PCR for the indicated genes with RNA from enteroids treated with DMSO only or inhibitors for ATM or YAP1 (n = 4). P values were calculated using two sided t test. The following are the p-values, pro-IL18, p = 0.0030, Nlrp1b, p = 0.0018, Nlrp6, p = 0.0015, Nlrc4, p = 0.0010. c Model representing the telomere dysfunction-mediated inflammasome pathway and the nodes at which the inhibitors for YAP1, caspase-1 or antibiotics inhibit the pathway leading to reduced secretion cleavage of caspase-1 and secretion of IL-18 by the colonic epithelium. d Hematoxylin and eosin staining of colon tissue from G4 mice treated with or without the YAP inhibitor, verteporfin. (control, n = 5; inhibitor treated, n = 6). Scale bars, 50 μm. e Histopathological score of the intestinal phenotype of the G4 mice treated with or without verteporfin (Yapi) (control, n = 5; inhibitor treated, n = 6). P values were calculated using two sided t test, p = 0.0048. f ELISA for the quantification of mature IL-18 from the colonic epithelium of the G4 mice treated with or without verteporfin (Yapi) (control, n = 5; inhibitor treated, n = 6).P values were calculated using two sided t test, p < 0.0001. g Western blot of colonic lysate from the colonic epithelium of the G4 mice treated with or without verteporfin (Yapi), for the indicated antibodies (n = 3). *statistically significant, p < 0.05 by unpaired Student’s t test, two-tailed. n represents number of mice used in the study. Data are represented as mean ± SEM. Experiments were conducted at least two independent times.