Society: AGA

Background: Eosinophilic Esophagitis (EoE) is one of the most predominant causes of chronic esophageal symptoms in children. Pathological changes of the esophageal epithelium in EoE are characterized by basal cell hyperplasia (BCH), infiltration of eosinophils and elevated levels of cytokines (e.g. IL-4 and IL-13). Our knowledge of EoE pathogenesis has mostly relied on adult animal models, despite unique clinical phenotypes of pediatric EoE. Accumulation of the extracellular matrix protein tenascin-C (TNC) is associated with asthma; however, its role in EoE pathogenesis has not been investigated. Using proteomics, our preliminary data show that TNC is dramatically increased in IL-13 treated human esophageal basal cells (EPC2 cells). We hypothesize that IL-13 induced pediatric EoE progression is dependent on TNC which modulates BCH. Methods: We adapted our previously established transgenic mouse model to generate Krt5-rtTA; tetO-IL13 mice pups. To examine the role of TNC deletion in our pre-weaning mouse model of pediatric EoE, we generated Krt5-rtTA; tetO-IL-13; TNC^-/- pups to determine the effects on the development of EoE. We also investigated the effect of TNC knockdown in IL-13 treated EPC2 cells using siRNA. Furthermore, we examined human esophageal biopsies from EoE patients and healthy controls to evaluate TNC and IL-13 levels, BCH and esophageal eosinophilia and their correlation with EoE disease stage. Results: IL-13 overexpression was detected in hyperplastic basal cells of Krt5-rtTA; tetO-IL13 transgenic pre-weaning mice upon doxycycline water feeding. Keratin 5 (Krt-5), TNC and major basic protein (MBP) immunostaining confirmed TNC expression in hyperproliferative basal cells in addition to eosinophils in the esophageal epithelium of these pre-weaning mice. These mice weighed less than controls; resembling a failure to thrive phenotype as seen in human infants with EoE as well as skin lesions similar to atopic dermatitis. Deletion of TNC in the transgenic pre-weaning mouse esophagus reduces the thickness of basal cell layers (p63⁺ Krt5⁺), proliferation (p63⁺ Ki67⁺) and differentiation (Krt13⁺) of basal cells. TNC knockdown reduces the number of proliferating (p63⁺ Ki67⁺) EPC2 cells. Consistently, TNC expression is increased in esophageal biopsy specimens from EoE patients as compared to healthy controls. Conclusions: For the first time, we demonstrate that our novel transgenic pre-weaning EoE mouse model shows that TNC is not only critical but also plays a conserved role in epithelial remodeling in the early stages of EoE pathogenesis. Future directions include whether the mechanism involving TNC may be also used in other allergen associated diseases including atopic dermatitis. Keywords: eosinophilic esophagitis, IL-13, basal cell hyperplasia, Tenascin-C, pediatric mouse model

Introduction: Eosinophilic Esophagitis (EoE) involves epithelial barrier dysfunction characterized by basal zone hyperplasia (BZH). Mast cells (MC) are increased in EoE. MC activation leads to release of pre-formed proteases such as carboxypeptidase A3 (CPA3), and translocation of LAMP3 (CD63) to the cell surface. CPA3 has restricted expression in MCs and little is known regarding the role of CPA3, a zinc-dependent metalloproteinase, on epithelial barrier function. We hypothesized that activated MCs in EoE release CPA3 which drives mucosal barrier dysfunction through loss of barrier proteins.
Methods: Flow cytometry was performed on single cell suspensions of esophageal biopsies to detect activated MC and correlated to BZH . CPA3 ELISA was performed on biopsy lysates from patients with active and inactive EoE. IF was performed for CPA3 and SOX2 on biopsies from EoE patients and controls with varying degrees of BZH. Total intraepithelial CPA3+ MCs were counted along with CPA3+ MCs nearby and far from SOX2+ cells. Single cell RNA-Sequencing of esophageal biopsies was performed to assess protease expression among MC subsets between patients with varying degrees of BZH along with DSG1 among epithelial cell subsets. Air-liquid interface (ALI) culture of human esophageal epithelial cells (EPC2-hTERT) was generated and recombinant, mature CPA3 was exposed to the basal ALI layer for 48 hours with or without a carboxypeptidase inhibitor (PCI). Transepithelial electrical resistance (TEER) was measured at baseline and 48 hours after CPA3. Morphologic assessment of H&E-stained sections was performed along with DSG1 IHC.
Results: Increased CD63+ MCs were present in active compared to inactive EoE (p<0.01) and moderately correlated with BZH (p<0.01) severity. Patients with active EoE had significantly more CPA3 protein in biopsy lysates compared to inactive EoE (p<0.01). CPA3+ MCs were increased among patients with severe BZH (p<0.01), and spatially located near SOX2+ basal epithelial cells (p<0.01). MCs from patients with severe BZH and furrows had the highest CPA3 RNA (p<0.0001). Transient-type MCs, found only in patients with severe BZH, had the greatest amount of CPA3 compared to resident MCs (p<0.0001). DSG1 gene expression was significantly reduced only among transitioning epithelium (p<0.001). CPA3 significantly reduced TEER (p<0.001) and was reversible with a carboxypeptidase-specific inhibitor (p<0.05). CPA3-treated ALI showed increased basal cell layers with loss of squamous cells and DSG1.
Conclusion: Activated MCs are present in EoE and correlate with BZH severity. CPA3 is increased in transient-type MCs in EoE, and CPA3+ MCs are located near basal epithelial cells. CPA3 is sufficient to induce epithelial barrier disruption and BZH associated with loss of DSG1 in transitioning epithelial cells, and is reversible with a carboxypeptidase inhibitor.

Background: Mechanistic studies suggest eosinophilic esophagitis (EoE) originates from dysfunction of the esophageal barrier. Environmental exposures have been implicated in EoE pathophysiology, but the specific factors that initiate epithelial barrier dysfunction and eosinophilic inflammation are unknown. Sodium dodecyl sulfate (SDS) is a detergent found in toothpaste where it may be used at concentrations up to 3% w/v (30 mg/mL). Humans regularly ingest SDS (0.01-0.9 mg/kg/day) through swallowed residual toothpaste. We recently demonstrated that SDS induces barrier disruption of the esophageal epithelium in vitro and eosinophilic inflammation following oral exposure in mice. Mucosal impedance (MI) has been used to show compromised mucosal integrity in gastroesophageal reflux disease and EoE and correlates well with transepithelial electrical resistance and dilated intercellular spaces.
Aim: Our objective is to determine the effect of swallowing an SDS-containing toothpaste on esophageal MI. We hypothesize that SDS-containing toothpaste induces esophageal barrier dysfunction.
Methods: We are performing a prospective pilot study to assess the effects of SDS-containing toothpaste on esophageal barrier function. We are enrolling healthy adults and exposing them to SDS-containing toothpaste. High resolution impedance manometry (HRIM) using a multichannel luminal impedance catheter is performed to assess baseline MI. With the transnasal catheter still in place, subjects brush their teeth and tongue for 2 min with 2 g of toothpaste. Subjects expectorate once without rinsing, and HRIM is monitored continuously for 1 hr. For this preliminary analysis, MI of the proximal and distal esophagus (mean of 4 sensors each) was extracted using ManoView™ ESO v3.0.1 acquisition and analysis software. Impedance sensors were selected relative to the upper and lower esophageal sphincters. Longitudinal comparisons of MI were made between baseline vs. 15, 30, 45 and 60 min using one-way ANOVA.
Results: Preliminary results (n = 4 subjects) showed a mean MI ± SEM, 2480Ω ± 252 vs. 1860Ω ± 407 (p = 0.24) in the proximal vs. distal esophagus, respectively. Esophageal MI was significantly decreased in the proximal esophagus at all timepoints after toothpaste exposure (absolute mean MI ± SEM, 15 min 1666Ω ± 325, 30 min 1448Ω ± 174, 45 min 1434Ω ± 154, 60 min 1267Ω ± 129, p=0.01). MI changes in the distal esophagus were more variable (absolute mean MI ± SEM, 15 min 1210Ω ± 328, 30 min 1462Ω ± 338, 45 min 1630Ω ± 265, 60 min 1317Ω ± 177, p = 0.62).
Conclusions: SDS-containing toothpaste decreases barrier function in the proximal esophagus, approaching a 50% reduction in MI at 60 min. Toothpaste-induced decreases in MI may be location dependent and recovery is variable. Additional studies are needed to assess the relationship between chronic daily oral SDS exposure and EoE.

<b>SDS-containing toothpaste decreases MI in the proximal esophagus.</b> MI (absolute mean ± SEM) in the proximal (<b>A</b>) and distal (<b>B</b>) esophagus at baseline, 0, 15, 30, 45, and 60 min. (n=4)

SDS-containing toothpaste decreases MI in the proximal esophagus. MI (absolute mean ± SEM) in the proximal (A) and distal (B) esophagus at baseline, 0, 15, 30, 45, and 60 min. (n=4)

Background Eosinophilic esophagitis (EoE) is an immune-mediated disease of the esophagus characterized by eosinophilic inflammation and epithelial remodeling including basal zone hyperplasia (BZH) and loss of terminal differentiation. BZH correlates with disease severity and persistent symptoms in patients in clinical remission. Yet, the molecular processes driving BZH remain poorly defined. Here, we investigate the cellular identity and transcriptional processes unique to BZH in EoE. Methods Single-cell RNA sequencing (scRNA-seq) was performed on esophageal biopsies from 6 adult EoE patients and 6 healthy donors. Differential expression, upstream regulator prediction, expression-based computational scoring and trajectory-based analysis were used to investigate EoE transcriptional processes across epithelial populations. Results Although histological scoring of adjacent biopsies identified BZH across all EoE patients, no increased proportion of basal cells was seen in EoE patients by scRNA-seq. Instead, EoE patients exhibited a decreased pool of KRT15+ COL17A1+ quiescent cells, a minor increase in KI67+ dividing basal cells and a dramatic increase in KRT13+ IVL+ suprabasal cells. An unsupervised approach scoring basal or terminally differentiated cell identity was used to quantify cell identity changes across all epithelial cell compartments in EoE. Confirming published findings, the superficial compartment exhibited loss of terminally differentiated identity in EoE. Interestingly, suprabasal and superficial clusters showed increased basal cell identity scoring in EoE, but this was not correlated with increased proliferation in these clusters. Pathway analysis of differentially expressed genes in the suprabasal and superficial compartments in EoE identified 'Signaling pathways regulating pluripotency of stem cells' as a top changed term. Enrichment analysis identified SOX2, KLF5 and TP63 as potential drivers of the increased basal identity using Enrichr databases. Trajectory analysis identified unique genes involved in EoE suprabasal to superficial lineage commitment that are unaltered in controls. Most of these genes are published epithelial targets of SOX2, KLF5 or the SOX2-KLF5 interaction. Conclusion This study investigated the cellular identities that expand in the hyperplastic remodeling of the EoE esophageal epithelium. BZH can now be more precisely understood as an expansion of non-proliferative cells committed to suprabasal differentiation but retaining stem-like transcriptional programs. We identified SOX2 and KLF5 as potential master transcriptional regulators driving this retention of stem-like identity in differentiated cells in EoE. A deeper understanding of the molecular mechanisms underlying BZH will lead to the development of better treatment options for patients with EoE.

Background: Increased oxidative stress is thought to contribute to the pathophysiology of Eosinophilic Esophagitis (EoE). Thus, our aim was to determine whether N-acetylcysteine (NAC), an antioxidant known to reduce oxidative stress, could reduce esophageal inflammation in experimental EoE.

Methods: Immortalized human esophageal epithelial cells (EPC2-hTERT) were pre-treated with NAC for 1 hour and then with IL-13, a known activator of STAT6, (except for control) for 30 minutes. Cells were collected for immunoblot of phosphoSTAT6 (pSTAT6) and total STAT6. This experiment was repeated but IL-13 and NAC remained on the cells for 24 hours, and quantitative real-time PCR (qPCR) was done to assess relative gene expression of CCL26, a known target of STAT6 activation. In the absence of cells, recombinant IL13Rα1 was mixed with control and increasing doses of NAC and assayed on a Nanotemp Tycho to assess for qualitative structural changes. For in vivo experiments, eosinophilic esophagitis was induced in C57/BL6 mice with 1 µg recombinant IL-33 intraperitoneal (IP) injections daily for 7 days with or without NAC. Five groups of adult mice were tested: 1) Sucrose water with saline IP; 2) 3% NAC/Sucrose water with saline IP; 3) Sucrose water with IL-33 IP; 4) 3% NAC/Sucrose water with IL-33 IP; 5) 300 mg/kg NAC topical suspension with IL-33 IP. At the end of the 7 days, esophageal tissue was collected for histology.

Results: Pre-treatment of EPC2-hTERT cells with NAC prior to IL-13 led to a significant reduction in pSTAT6/STAT6 in a dose-dependent manner as compared with IL-13 alone (Fig. 1A-B). After 24 hours, CCL26 expression was significantly reduced with NAC + IL-13 as compared with IL-13 alone (Fig. 1C). As previously reported, IL-33 induced esophageal basal cell hyperplasia (BCH) in mice compared to controls (Fig. 2A, 2C). Consumption of NAC in drinking water or the administration of a NAC topical suspension reduced IL-33 induced BCH (Fig. 2A). Furthermore, 3% NAC in drinking water significantly reduced MBP counts with a similar trend when administering NAC topical suspension (Fig. 2B, 2C). When recombinant IL13Rα1 was mixed with NAC, inflection temperatures measured on the Nanotemp Tycho were lowered in a dose-dependent fashion compared to IL13Rα1 alone.

Conclusion: NAC reduces STAT6 activation and CCL26 expression induced by IL-13. NAC consumption via drinking water or by administration of an esophageal topical suspension attenuates IL-33 induced BCH and esophageal eosinophilia in mice. Targeting oxidative stress may serve as a novel pharmacologic pathway for the treatment of EoE.

<u>Figure 1</u>: NAC reduces IL-13 activation of STAT6 and downregulates CCL26 expression. (A) Representative images from immunoblotting of EPC2-hTERT cells under NAC and IL-13 conditions probed with pSTAT6, STAT6 and β-actin. (B) Quantification of immunoblot expression as a ratio of pSTAT6/STAT6. When these cells were pretreated with NAC, a significant reduction in pSTAT6/STAT6 was noted in a dose-dependent manner as compared with IL-13 treatment alone (vs 1 mM, 47%, P<0.05; vs 10 mM, 76%, P<0.01, ANOVA test). (C) After 24 hours of treatment with NAC and IL-13, CCL26 gene expression induced by IL-13 was significantly reduced with 1- (54%, P<0.05, ANOVA) and 10-mM NAC (45%, P<0.01, ANOVA).

Figure 1: NAC reduces IL-13 activation of STAT6 and downregulates CCL26 expression. (A) Representative images from immunoblotting of EPC2-hTERT cells under NAC and IL-13 conditions probed with pSTAT6, STAT6 and β-actin. (B) Quantification of immunoblot expression as a ratio of pSTAT6/STAT6. When these cells were pretreated with NAC, a significant reduction in pSTAT6/STAT6 was noted in a dose-dependent manner as compared with IL-13 treatment alone (vs 1 mM, 47%, P<0.05; vs 10 mM, 76%, P<0.01, ANOVA test). (C) After 24 hours of treatment with NAC and IL-13, CCL26 gene expression induced by IL-13 was significantly reduced with 1- (54%, P<0.05, ANOVA) and 10-mM NAC (45%, P<0.01, ANOVA).

<u>Figure 2</u>: NAC attenuates basal cell hyperplasia and esophageal eosinophilia in an IL-33 experimental esophagitis in vivo murine model. (A) Treatment with IL-33 induced basal cell hyperplasia (20.72 µm) when compared to the negative controls (9.06 µm, 10.56 µm, P<0.0001, Kruskal-Wallis test). A significant reduction in basal cell thickness was noted with NAC drinking water and NAC suspension compared with no NAC treatment (vs NAC water, 14.37 µm, P=0.0010; vs NAC suspension, 12.07 µm, P=0.0016, Kruskal-Wallis test). (B) Relative ratio of major basic protein (MBP) cell counts normalized to experimental control group. IL-33 induced esophageal eosinophilia (1.00 vs 0.04 vs 0.06, P<0.0001, Kruskal-Wallis test). 3% NAC in drinking water significantly reduced MBP counts (1.00 vs 0.51, p<0.05, Kruskal-Wallis test) with a similar trend when administering NAC topical suspension (1.00 vs 0.56, p=0.19, Kruskal-Wallis test). (C) Representative images from H&E and MBP staining of mouse esophagus.

Figure 2: NAC attenuates basal cell hyperplasia and esophageal eosinophilia in an IL-33 experimental esophagitis in vivo murine model. (A) Treatment with IL-33 induced basal cell hyperplasia (20.72 µm) when compared to the negative controls (9.06 µm, 10.56 µm, P<0.0001, Kruskal-Wallis test). A significant reduction in basal cell thickness was noted with NAC drinking water and NAC suspension compared with no NAC treatment (vs NAC water, 14.37 µm, P=0.0010; vs NAC suspension, 12.07 µm, P=0.0016, Kruskal-Wallis test). (B) Relative ratio of major basic protein (MBP) cell counts normalized to experimental control group. IL-33 induced esophageal eosinophilia (1.00 vs 0.04 vs 0.06, P<0.0001, Kruskal-Wallis test). 3% NAC in drinking water significantly reduced MBP counts (1.00 vs 0.51, p<0.05, Kruskal-Wallis test) with a similar trend when administering NAC topical suspension (1.00 vs 0.56, p=0.19, Kruskal-Wallis test). (C) Representative images from H&E and MBP staining of mouse esophagus.

Background: Eosinophilic esophagitis (EoE) is histologically characterized by esophageal eosinophilia and epithelial remodeling, including basal zone hyperplasia (BZH) and dilated intercellular spaces (DIS). Esophageal epithelial remodeling is thought to contribute to esophageal barrier dysfunction and clinical manifestations of EoE, however, the molecular pathways that drive esophageal epithelial remodeling remain largely unexplored.
Objective: To define the role of IL-13-induced transcriptional programs in esophageal epithelial proliferation in EoE.
Methods: We performed RNA sequencing, western blot, RT-qPCR and histological analyses on esophageal biopsies from healthy control and EoE patients, primary esophageal cells (PECs) derived from patients with EoE, and IL-13 stimulated esophageal epithelial keratinocytes (EPC2-ALI). We also performed in silico analyses of bulk and scRNA seq analysis of publicly available validation cohorts. We utilized STAT3 shRNA (genetic approach), a STAT3 degrader (SD-36; pharmacologic approach) and pharmacological agonist and antagonist of SFRP1 to define the role of STAT3 and SFRP1 in esophageal epithelial proliferation and gene expression. We employed a mouse model of IL-13-induced esophageal epithelial remodeling (Krt5-rtTA x tetO-IL-13Tg) to define the relationship between STAT3 and esophageal epithelial proliferation in vivo.
Results: RNAseq analysis of esophageal biopsies (healthy control vs. EoE) and EPC2-ALI (+/- IL-13) revealed 82 common differentially expressed genes (DEGs) that were enriched for putative STAT3 target genes (32 DEGs). IL-13-stimulation of EPC2-ALI and PECs induced STAT3 and STAT6 phosphorylation. STAT3 was required for esophageal epithelial proliferation in vitro, and in vivo mouse modeling validated the link between STAT3 and epithelial proliferation in the esophagus. In silico analyses elucidated a relationship between the putative STAT3 target Secreted frizzled-related protein 1 (SFRP1) and esophageal epithelial cell proliferation. Indeed, IL-13-stimulation increased SFRP1 mRNA expression. IL-13-induced esophageal epithelial proliferation was enhanced with pharmacologic inhibition of SFRP1 and attenuated with recombinant SFRP1 stimulation. Employing a validation dataset, we determined that SFRP1 mRNA is increased in esophageal biopsies from patients with active EoE compared to healthy controls or patients in remission, and that the cellular source of SFRP1 is predominantly in an esophageal suprabasal epithelial cell subpopulation that is characterized by expression of core EoE pro-inflammatory transcriptome (CCL26, ALOX15, CAPN14, ANO1, TNFAIP6).
Conclusions: These studies identify SFRP1 as a key regulator of IL-13-induced and STAT3-dependent esophageal epithelium proliferation in EoE, and as a novel candidate for therapeutic intervention for attenuation of esophageal remodeling in EoE.

Mirna Chehade

Mount Sinai School of Medicine

Appearances