Gastrointestinal (GI) epithelium is highly proliferative and its integrity relies on the regenerative capacity of local stem cells (Barker et al., 2010). Telomeres are DNA-protein structures that protect chromosome ends. With cell division, telomeres shorten and short dysfunctional telomeres provoke apoptosis and/or senescence (Armanios & Blackburn, 2012). Telomerase synthesizes new telomere repeats to offset the loss that occurs during DNA replication(Greider & Blackburn, 1985, 1987). Mice that are null for telomerase show progressive telomere shortening and provide a model for understanding the consequences of telomere dysfunction (Armanios & Blackburn, 2012). In these mice, the highly proliferative tissues, such as the skin and bone marrow, display degenerative phenotypes due to stem-cell failure (Lee et al., 1998; Rudolph et al., 1999; Hao et al., 2005; Rossi et al., 2007). They also develop intestinal villous atrophy and enterocolitis, which contribute to their limited lifespan (Lee et al., 1998; Rudolph et al., 1999; Hao et al., 2005; Armanios et al., 2009). The question of how short telomeres affect the human GI tract has not been systematically examined.
Syndromes caused by mutant telomerase and telomere genes span the age spectrum, and their severity depends on the extent of telomere shortening (Armanios & Blackburn, 2012). Mutations in one of four genes account for the major subset (Armanios & Blackburn, 2012). Mutant telomerase core enzyme genes TERT, the telomerase reverse transcriptase, and TR, its RNA component, cause autosomal dominant disease. Mutations in DKC1, which encodes the telomerase-associated dyskerin protein, cause X-linked disease, and mutations in TINF2, which encodes the telomere protein TIN2, primarily cause de novo disease. In infancy, telomere-mediated disease is recognized as Hoyeraal-Hreidarsson (HH) syndrome, a rare disorder characterized by developmental delay, immunodeficiency, and cerebellar hypoplasia. In children, it is recognized as dyskeratosis congenita (DC) defined by a triad of oral leukoplakia, nail dystrophy and skin hyperpigmentation (Savage & Alter, 2009). Adult-onset telomere disease is heterogeneous and manifests as isolated or syndromic clustering of bone marrow failure, pulmonary fibrosis, and liver cirrhosis (Armanios, 2009). Because the clinical presentation of telomere syndromes is diverse, lymphocyte telomere length measurement is used diagnostically to identify affected individuals (Armanios & Blackburn, 2012). We sought to determine whether telomere syndromes cause GI disease. We show here that they manifest in discrete patterns, and report their prevalence, spectrum, and natural history.
The Johns Hopkins Telomere Syndrome Registry had 38 assessable individuals from 20 families (Methods S1). Of these, six (16%) required evaluation by a gastroenterologist and endoscopy. Clinical findings are summarized in Table 1 and Fig. 1, and the detailed histories are in Data S1. The DKC1, TERT, and TR mutations identified have been reported or shown to compromise telomerase activity (Knight et al., 2001; Marrone & Mason, 2003; Parry et al., 2011). In two cases, no mutations were identified. Lymphocyte telomere length was below the age-adjusted 1st percentile in the five assessable cases (Fig. 1A). Cases were on average 15 years (15 months-34 years), and the GI disease was most severe in the youngest subjects. A 15-month old presented with bloody diarrhea, was diagnosed with enterocolitis, and had severe B cell lymphopenia (Fig. 1B–D). The enterocolitis was refractory to immunosuppression, and she required colectomy and parenteral nutrition support. The GI symptoms persisted after bone marrow transplantation, and led to premature death within a year of diagnosis. A 3-year-old boy with severe swallowing difficulties was found to have a nearly obstructing proximal esophageal web. Three additional cases presented with difficulty gaining weight and abdominal pain, and two were diagnosed with celiac enteropathy based on profound villous atrophy and intraepithelial lymphocytosis (Fig. 1E–G). Case 6 presented with chronic dysphagia and had a proximal esophageal web (Fig. 1H,I).
|Case||Age||Gender||Clinical/ Genetic diagnosis||Telomere phenotypes||GI history||Esophageal and gastric findings||Small and large bowel findings|
|1||15 months||F||Hoyeraal-Hreidarsson|| |
Failure to thrive
|Gastric lamina propria-mild fibrosis|| |
Pancolitis, epithelial sloughing
Gland dropout No plasma cells
|2||3 years||M|| |
|Dysphagia-solids||Esophageal stenosis-Cricopharynx||No data available|
|3||7 years||M||Dyskeratosis congenita|| |
Failure to thrive
Abdominal pain post-prandial
|Esophageal IEL|| |
Villous blunting ↑ Apoptosis
Duodenal neutrophil infiltrates
|4||16 years||F|| |
Failure to thrive
|Esophagus- inflammatory changes|| |
Normal duodenal biopsya
Colonic biopsy not performed
|5||34 years||F|| |
Parietal Cell Dropout
Duodenal and colonic IEL
|6||34 years||M|| |
Lacrimal duct stenosis
Abdominal pain post-prandial
GI mucosal biopsies in symptomatic cases revealed severe epithelial defects. In the case with enterocolitis, there was profound mucosal sloughing, crypt dropout, increased apoptosis, and absent plasma cells. In milder enteropathy cases, there was villous atrophy, intraepithelial lymphocytosis, and increased mitotic errors such as rings and anaphase bridges (Fig. 1G). There was also an increase in epithelial apoptotic bodies (Fig. 1F). We compared epithelial apoptotic bodies in cases and controls and found a significant increase in the duodenum [28.7 ± 7.2 SEM/100 crypts (n = 4) vs. 5.4 ± 1.3 SEM/100 crypts (n = 7), P < 0.01, unpaired t-test] and colon [70.0 ± 30.0 SEM (n = 2) vs. 6.0 ± 1.0 SEM/100 crypts (n = 8), P < 0.01, Fig. 1J,K]. Gastric epithelial apoptosis was seen in one case with parietal cell dropout.
To extend our findings, we reviewed 591 Pubmed entries, which fulfilled prespecified criteria (Methods S1), and identified 43 additional HH and DC cases with GI disease. Table S1 summarizes the 24 upper GI (Addison & Rice, 1965; Robledo Aguilar et al., 1974; Russo et al., 1990; Dokal et al., 1992; Paul et al., 1992; Sawant et al., 1994; Berezin et al., 1996; Herman et al., 1997; Krishnan et al., 1997; Baselga et al., 1998; Elliott et al., 1999; Ghavamzadeh et al., 1999; Knight et al., 1999; de Roux-Serratrice et al., 2000; Yaghmai et al., 2000; Arca et al., 2003; Sznajer et al., 2003; Kanegane et al., 2005; Utz et al., 2005; Handley & Ogden, 2006; Demirgunes et al., 2008; Sasa et al., 2012), and Table S2, the 23 lower GI cases (Steier et al., 1972; Kehrer et al., 1992; Paul et al., 1992; Berthet et al., 1994; Berezin et al., 1996; Knight et al., 1999; Cossu et al., 2002; Arca et al., 2003; Sznajer et al., 2003; Borggraefe et al., 2009; Touzot et al., 2010; Jyonouchi et al., 2011; Sasa et al., 2012). Dysphagia was the most common upper GI complaint, and esophageal stenosis the most prevalent diagnosis (23 of 24 cases, 96%). Mean age at diagnosis was 11.4 years (1 month–27 years). Strictures localized to the proximal esophagus (eight of nine, 89%), and symptoms improved after dilatation, but repeat procedure was at times required (Addison & Rice, 1965; Paul et al., 1992; Baselga et al., 1998). Stenosis was congenital in four cases with poor feeding history reported since birth (Russo et al., 1990; Knight et al., 1999; Yaghmai et al., 2000; Sznajer et al., 2003), while older children had milder long-standing swallowing difficulties. In most cases (21 of 24, 88%), DC was the underlying telomere disorder, indicating that esophageal stenosis may occur at higher frequency in this population.
Diarrhea due to severe enteropathy was the most common lower GI diagnosis (n = 14, 61%) presenting at a mean of 6.7 years (1 month–21 years). HH syndrome was the predominant underlying telomere disorder (17 of 23, 71%). Findings included pancolitis and atrophic mucosa, and pathology showed gland dropout, lamina propria fibrosis, intraepithelial lymphocytosis, and apoptosis, similar to one of our index cases. Esophageal stenosis preceded or followed lower GI disease at times (Berezin et al., 1996; Knight et al., 1999; Arca et al., 2003; Sznajer et al., 2003; Sasa et al., 2012). Intestinal disease presented earlier in HH than DC (mean 1.4 vs. 17 years, respectively). In all HH cases with intestinal disease, there was a concurrent B-cell lymphopenia and/or hypogammaglobulinemia (16 of 16, 100%). Intestinal disease caused significant morbidity in children requiring colectomy or parenteral nutrition support (Paul et al., 1992; Berthet et al., 1994; Knight et al., 1999; Cossu et al., 2002; Sznajer et al., 2003). Therefore, telomere-mediated intestinal disease can be life-threatening, especially in HH patients.
We show here that short telomere length disrupts GI mucosal integrity in telomere syndromes. Disease affected 16% of our Registry subjects and was at times the first and most severe presentation. The collective experience we report indicates nonmalignant telomere-mediated GI disease manifests in three discrete categories: esophageal stricture, enteropathy, and enterocolitis. Of these, enterocolitis is the most severe and occurs in young children with telomere-related B cell immunodeficiency. Enteropathy has a milder course and is associated with villous atrophy. Esophageal stenosis represents one of several luminal stenotic defects that occur in DC, such as lacrimal duct and urethral stenosis, and likely reflects developmental defects.
The convergence of findings in human telomere syndromes with those seen in telomerase null mice suggests the GI pathology we see is telomere-dependent. Villous atrophy in mice represents a telomere-mediated stem-cell failure (Rudolph et al., 1999), while the enterocolitis is thought to represent a compound defect in the epithelium-immune barrier (Armanios et al., 2009). Mice with short telomeres also develop intestinal microadenoma (Hao et al., 2005), and DC patients have an increased incidence of esophageal, rectal, and possibly gastric cancer, although the overall incidence is relatively low (Alter et al., 2009).
Our findings have clinical implications. A heightened index of suspicion for the GI processes described here in patients with known telomere syndromes, and conversely of telomere disorders in individuals with GI pathology can facilitate early diagnosis, prevent unnecessary work-up, and anticipate/avert complications. Telomere length testing in these cases can be a critical diagnostic tool. We note that the GI disease patterns we describe share features of poorly understood processes such as celiac and inflammatory bowel disease. It is possible that telomere length may be a relevant genetic modifier of disease severity in these disorders where the GI mucosa is disrupted. Because short telomere length is acquired with age, the processes we describe may also point to yet-unrecognized age-dependent disease patterns in the GI tract.