Eosinophilic esophagitis (EoE) is a complex, polygenic disorder caused by genetic predisposition and environmental exposures. Because of the recent emergence of EoE as a bona fide global health concern, a paucity of available therapeutic and diagnostic options exists. However, rapid progress has been made in an effort to rectify this lack and to improve understanding of the factors that cause EoE. This article highlights key advances in elucidating the genetic (and epigenetic) components involved in EoE.
Key points
- •
Eosinophilic esophagitis (EoE) is a complex, polygenic disorder.
- •
Disease risk variants and an altered esophageal transcriptional profile underlie the genetic origin of EoE.
- •
Emerging epigenetic modifications link environmental exposures to the genetic dysregulation in EoE.
Introduction
Early evidence for a genetic origin of eosinophilic esophagitis (EoE) came in the form of several epidemiologic studies showing a high prevalence of disease in specific genders and races, with nearly three-quarters of patients being men and almost all (≈90%) being of European descent, respectively. Moreover, an increased disease risk is seen among familial cases, which typically demonstrate a non-Mendelian inheritance pattern. Expression profiling from esophageal biopsies acquired during routine endoscopic procedures has provided molecular insight into genetic dysregulation occurring within the inflamed esophagus. These transcriptional changes affect both coding and noncoding (microRNA) transcripts and underscore consistent, disease-specific alterations in the levels of select molecules expressed by activated immune cells and structural cells of the esophagus. These dysregulated transcripts and associated biologic pathways represent potential targets for novel therapeutics and diagnostic methods.
In addition to the genetic elements, a role for environmental factors in EoE has been established through both clinical and basic research. Patients with EoE are often hypersensitized to multiple food antigens, making directed dietary modification one of the most effective therapies for EoE. Several early life exposures, including Cesarean birth, antibiotics, and formula feeding, have been identified to influence the risk of pediatric EoE. In addition, geographic location, industrialized environments, history of Helicobacter pylori infection, and seasonal variations in disease implicate environmental antigens. These clinical findings have been supported through multiple basic research studies showing that epidermal and pulmonary exposure to various antigens can induce EoE-like symptoms in mice. Emerging epigenetic data are now beginning to provide clues as to how these environmental factors may be intricately intertwined with the genetic dysregulation in EoE and thus act in a concerted fashion to affect disease pathophysiology.
Introduction
Early evidence for a genetic origin of eosinophilic esophagitis (EoE) came in the form of several epidemiologic studies showing a high prevalence of disease in specific genders and races, with nearly three-quarters of patients being men and almost all (≈90%) being of European descent, respectively. Moreover, an increased disease risk is seen among familial cases, which typically demonstrate a non-Mendelian inheritance pattern. Expression profiling from esophageal biopsies acquired during routine endoscopic procedures has provided molecular insight into genetic dysregulation occurring within the inflamed esophagus. These transcriptional changes affect both coding and noncoding (microRNA) transcripts and underscore consistent, disease-specific alterations in the levels of select molecules expressed by activated immune cells and structural cells of the esophagus. These dysregulated transcripts and associated biologic pathways represent potential targets for novel therapeutics and diagnostic methods.
In addition to the genetic elements, a role for environmental factors in EoE has been established through both clinical and basic research. Patients with EoE are often hypersensitized to multiple food antigens, making directed dietary modification one of the most effective therapies for EoE. Several early life exposures, including Cesarean birth, antibiotics, and formula feeding, have been identified to influence the risk of pediatric EoE. In addition, geographic location, industrialized environments, history of Helicobacter pylori infection, and seasonal variations in disease implicate environmental antigens. These clinical findings have been supported through multiple basic research studies showing that epidermal and pulmonary exposure to various antigens can induce EoE-like symptoms in mice. Emerging epigenetic data are now beginning to provide clues as to how these environmental factors may be intricately intertwined with the genetic dysregulation in EoE and thus act in a concerted fashion to affect disease pathophysiology.
Genetic variants
Several candidate gene approaches have identified a handful of genetic risk variants in EoE. For instance, a common single-nucleotide variant (minor allele frequency [MAF] = 0.25 in the HapMap population of European descent) located in the 3′ untranslated region of the chemokine (C-C motif) ligand 26 ( CCL26 ) was overrepresented in patients with EoE in both a case-control and a family-based analysis. Furthermore, 2 coding variants (R501X and 2282del4) in the epidermal barrier gene filaggrin ( FLG ), which is negatively regulated by interleukin (IL) 13 and is decreased in the esophageal mucosa of patients with EoE, associate with EoE risk. Lastly, in a small cohort of patients with steroid-treated EoE, a genetic variant in the promoter of the transforming growth factor beta 1 ( TGF – β1 ) gene, was associated with steroid unresponsiveness and correlated with increased TGF-β1–positive cells in the esophagus. The genetic link between the TGF-β pathway and EoE identified in this study is remarkable given the evidence showing a high rate of EoE, other eosinophilic gastrointestinal disorders, and atopic disease in patients with connective tissue disorders, such as Loeys-Dietz syndrome, which has been associated with variants in the TGF-β receptors 1 and 2 ( Table 1 ).
| Target | Modification | Potential Biologic Effect | ||
|---|---|---|---|---|
| Genetic | Disease risk variants | CCL26 | SNV in 3′ UTR | Enhanced mRNA stability; increased expression |
| TGFB1 | SNV in promoter | Increased expression; nonresponsiveness to topical steroid therapy | ||
| FLG | Nonsense and missense SNVs | Loss of function; reduced barrier function | ||
| TSLP | SNVs in promoter region and introns | Increased expression; correlates with esophageal levels of basophil and GMP-like cells | ||
| CRLF2 | Missense SNV | Male-specific association; enhanced TSLP signaling | ||
| DSG1 | Missense SNVs in patients with SAM syndrome a | Loss of function; reduced epithelial integrity; increased IL-5 and TSLP | ||
| TGFBR1/TGFBR2 | Missense SNVs in patients with LDS b | Increased TGF-β signaling; elevated CD4+ Th2 cells | ||
| PTEN | Missense SNVs, insertions, and deletions in patients with PHTS c | Loss of function; hyperproliferation | ||
| Transcriptome | CCL26 | Increased expression (esophagus) | Promotes eosinophil trafficking into the esophagus | |
| POSTN | Increased expression (esophagus) | Increased eosinophil adhesion; promotes esophageal remodeling; increased TSLP expression | ||
| DSG1 | Decreased expression (esophagus) | Reduced barrier function; increased POSTN expression | ||
| Epigenetic | Histones | H3 | Acetylated | Enhanced CCL26 promoter activity |
| H3 | Methylated (lysine 4) | Enhanced CCL26 promoter activity | ||
| DNA | CCL26 | Hypomethylation in promoter region | Enhanced CCL26 promoter activity | |
| MicroRNAs | miR-21 | Increased expression (esophagus) | Skewed Th2 response; increased eosinophil survival | |
| miR-223 | Increased expression (esophagus, blood) | Increased eosinophil progenitors | ||
| miR-375 | Decreased expression (esophagus) | Enhanced IL-13 transcriptional responses | ||
a EoE was a comorbidity in 1 of 3 patients with SAM syndrome.
b High prevalence of EoE and other eosinophilic gastrointestinal disorders (n = 6) in 58 patients with LDS.
c Significant enrichment of eosinophilic gastrointestinal disorders in patients with PTEN hamartoma tumor syndromes (odds ratio, 272; CI, 89–831; P <10 −4 ).
To identify disease risk variants in a more unbiased fashion, a genome-wide association study (GWAS) was performed in which 351 patients with EoE and 3,104 healthy controls were genotyped for more than 550,000 common variants. On chromosome 5q22, a single locus spanning the thymic stromal lymphopoietin ( TSLP ) and WD repeat domain 36 ( WDR36 ) genes showed a significant association with EoE susceptibility. TSLP is a potent Th2-promoting cytokine involved in the development of multiple allergic diseases. Expression analyses showed increased TSLP in EoE and a genotypic effect of the top associated variant on TSLP expression, with patients carrying the risk allele having elevated TSLP expression. In addition, TSLP risk genotypes correlated with increased levels of basophils, which have a key role in promoting EoE-like disease in mice, and with granulocyte-monocyte progenitor-like cells in the esophagus.
A secondary candidate gene approach also identified variants within the TSLP locus that were significantly associated with EoE risk. In this study assessing more than 700 variants in epithelial-derived genes linked to atopy, TSLP variants were the most significant genetic hits linked to EoE that, importantly, showed a stronger association with disease risk when compared with controls with atopic diseases (atopic dermatitis and asthma). Moreover, a coding variant in the cytokine receptor–like factor 2 ( CRLF2 ) gene, which encodes for the receptor for TSLP, showed a sex-specific association with EoE risk in men only. These cumulative data support aberrant regulation affecting the TSLP pathway as a specific genetic origin in EoE. Given the established role of TSLP in the initiation of allergic diseases, the fact that WDR36 was not differentially expressed in EoE underscores TSLP as the most likely gene involved in driving the esophageal inflammatory responses in EoE. Importantly, however, variants in WDR36 have been linked with peripheral blood eosinophil levels and atopic asthma. Thus, further studies are needed to identify the precise causal variants of EoE and to fully investigate a potential nonesophageal role for WDR36 that may contribute to disease.
The EoE transcriptome
A total of 574 highly dysregulated esophageal genes were identified, termed the EoE transcriptome , which distinguishes patients with EoE from healthy controls and, importantly, from patients with noneosinophilic forms of esophagitis. Despite the patchiness of EoE and phenotypic diversity within the patients analyzed with EoE, the EoE transcriptome is surprisingly well conserved across patient age, gender, atopic status, and nonfamilial relationship. A large-scale screen based on 94 signature EoE transcriptome genes has shown promise as a diagnostic tool capable of discriminating patients with EoE from those with noneosinophilic forms of esophagitis and patients with active EoE from those with EoE in remission (inactive EoE). The cytokine IL-13 is capable of inducing an esophageal epithelial cell gene signature that represents 22% of the EoE transcriptome. This article discusses 3 key genes within the EoE transcriptome, their regulation by IL-13, and their influence on disease pathophysiology.
Chemokine (C-C Motif) Ligand 26
Expression of CCL26 , which encodes the eosinophil chemoattractant eotaxin-3, was upregulated 53-fold in EoE, making it the most highly induced gene of the EoE transcriptome. CCL26 is believed to be the main driver for eosinophil recruitment into the esophagus, because the upregulation of CCL26 was unique among other closely related chemokines from the eotaxin family ( CCL11 and CCL24 ) ; however, other studies have indicated that CCL11 and CCL24 are induced at low levels in EoE. The levels of CCL26 in patients with EoE correlated significantly with the esophageal levels of eosinophils and mast cells. The induction of CCL26 in patients with EoE was determined to be largely because of the influence of IL-13 on esophageal epithelial cells, as CCL26 was also the most highly induced gene in IL-13–treated cells (279-fold when compared with untreated cells). Molecular analyses defined 2 STAT6 binding sites in the CCL26 promoter that were necessary for the induction of CCL26 by IL-13 and by IL-4. Furthermore, several coactivators, including poly-ADP ribosyl polymerase 14 (PARP14), have been shown to act on the CCL26 locus. PARP14 was identified as a specific coregulator of STAT6 signaling, and its overexpression in esophageal epithelial cells enhanced IL-13–induced CCL26 expression in a STAT6-dependent manner. Finally, exposure of esophageal epithelial cells to acidic pH enhances eotaxin-3 release, providing a potential mechanism by which proton pump inhibitor (PPI) therapies could have some anti-inflammatory effects in EoE. A role for the eotaxins and their receptor CCR3 is supported by studies in mice that have shown attenuated eosinophil levels and/or tissue remodeling in eotaxin and/or CCR3 -deficient mice.
Periostin
Periostin ( POSTN ) is a matricellular protein capable of interacting with multiple extracellular matrix molecules and cell surface receptors such as type 1 collagen and Notch1, respectively. Periostin is directly involved in regulating multiple cellular processes, including cell migration and adhesion. Its influence on metastasis, tissue remodeling, and wound healing has made it a highly studied molecule in the context of various human diseases, such as cancer, asthma, and atopic dermatitis. A 47-fold induction of periostin mRNA was observed in EoE, making it the second most highly upregulated gene in the EoE transcriptome. The periostin protein was also increased in EoE, primarily localized within the lamina propria, indicating fibroblasts as the main cellular source of periostin induction. TGF-β and IL-13 induced greater levels of periostin expression in esophageal fibroblasts than in esophageal epithelial cells. Using periostin-deficient mice, Blanchard and colleagues showed that periostin promotes allergic inflammatory responses in the lung and esophagus, partly through enhancing eosinophil adhesion. In skin keratinocytes, periostin can induce the expression of TSLP. These collective findings suggest a molecular loop among TGF-β, periostin, and TSLP, which act synergistically to drive the esophageal pathophysiology associated with EoE. Circulating periostin levels help identify patients with asthma who will experience response to biologic therapeutics, such as anti-IgE and anti–IL-13, and seem to identify eosinophilic asthmatic phenotypes, extending the significance of the eosinophil/periostin connection from EoE to other common atopic disorders.
Desmoglein 1
Desmoglein 1 (DSG1) is a transmembrane molecule belonging to the family of desmosomal cadherins that has an essential role in maintaining epithelial integrity through calcium-dependent intercellular adhesion. The focus on DSG1 as an etiologic component in human disease stemmed from observations linking DSG1 alteration to various dermatologic disorders, in which epithelial integrity and barrier function are compromised. In EoE, DSG1 mRNA is specifically downregulated in the esophageal mucosa of patients with active disease. This specific decrease in DSG1 was shown to result from IL-13 stimulation of differentiated esophageal epithelial cells. Functionally, DSG1 -deficient esophageal epithelial cells exhibited greater cell dissociation, weaker adhesive properties, and reduced capacity to form an intact epithelial barrier. Moreover, the loss of DSG1 triggered epithelial gene expression changes reflective of those in EoE biopsies, including increased POSTN expression.
A key contributory role for DSG1 dysregulation in the allergic disease process was independently demonstrated in genetic studies that identified loss-of-function mutations in DSG1 in consanguineous individuals with severe atopic dermatitis, multiple allergies, and metabolic wasting (SAM) syndrome. All 3 patients with SAM who were analyzed had failure-to-thrive diagnoses and multiple food allergies; notably, one patient also had an EoE diagnosis (see Table 1 ). Skin biopsies from patients with SAM showed reduced DSG1 expression and acantholysis, whereas isolated keratinocytes showed increased expression of IL-5 and TSLP. An intronic mutation in DSG1 showed a suggestive association with EoE risk. Given this potential association and the findings in SAM syndrome, further investigation into EoE risk variants in DSG1 is warranted ( Fig. 1 ).
