Persistent desmoglein-1 downregulation and periostin accumulation in histologic remission of eosinophilic esophagitis

Hannes Hoelz,Tim Faro,Marie-Luise Frank,Ignasi Forné,Daniela Kugelmann,Anja Jurk,Simon Buehler,Kolja Siebert,Monica Matchado,Tobias Straub,Annett Hering,Guido Piontek,Susanna Mueller,Sibylle Koletzko,Markus List,Katja Steiger,Martina Rudelius,Jens Waschke,Tobias Schwerd
DOI: https://doi.org/10.1016/j.jaci.2024.09.016
2024-09-27
Abstract:Background: Patients with eosinophilic esophagitis (EoE) require long-lasting resolution of inflammation to prevent fibrostenosis and dysphagia. However, the dissociation between symptoms and histologic improvement suggests persistent molecular drivers despite histologic remission. Objective: We characterized persisting molecular alterations in pediatric patients with EoE using tissue transcriptomics and proteomics. Methods: Esophageal biopsy samples (n = 247) collected prospectively during 189 endoscopies from pediatric patients with EoE (n = 36, up to 11 follow-up endoscopies) and pediatric controls (n = 44, single endoscopies) were subjected to bulk transcriptomics (n = 96) and proteomics (n = 151). Intercellular junctions (desmoglein-1/3, desmoplakin, E-cadherin) and epithelial-to-mesenchymal transition (vimentin:E-cadherin ratio) were assessed by immunofluorescence staining. Results: Active EoE (≥15 eosinophils per high-power field [eos/hpf]), inactive EoE (<15 eos/hpf), and deep-remission EoE (0 eos/hpf) were diagnosed in 107 of 185, 78 of 185, and 41 of 185 biopsy samples, respectively. Among the dysregulated genes (up-/downregulated 310/112) and proteins (up-/downregulated 68/16) between active EoE and controls, 17 genes, and 6 proteins remained dysregulated in inactive EoE. Using persistently upregulated genes (n = 9) and proteins (n = 3) only, such as ALOX15, CXCL1, CXCL6, CTSG, CDH26, PRRX1, CLC, EPX, and periostin (POSTN), was sufficient to separate inactive EoE and deep-remission biopsy samples from control tissue. While 32 differentially expressed genes persisted in deep-remission EoE compared to controls, the proteome normalized except for persistently upregulated POSTN. Epithelial-to-mesenchymal transition normalized in inactive EoE, whereas desmosome recovery remained impaired as a result of desmoglein-1 downregulation. Conclusion: The analysis of molecular changes shows persistent EoE-associated esophageal dysregulation despite histologic remission. These data expand our understanding of inflammatory processes and possible mechanisms that underlie tissue remodeling in EoE.
What problem does this paper attempt to address?