EOSINOPHILIC ESOPHAGITIS ALTERS THE EPITHELIAL AND IMMUNOLOGICAL LANDSCAPE TO LIMIT CARCINOGENESIS IN THE MURINE ESOPHAGUS

Introduction: Patients with the allergic disease eosinophilic esophagitis (EoE) have been shown to have a diminished risk of developing esophageal malignancy. Using murine models of EoE and esophageal squamous cell carcinoma (ESCC) via MC903/ovalbumin (OVA) and 4-nitroquinoline-1-oxide (4NQO), respectively, our prior studies have demonstrated that EoE inflammation limits esophageal carcinogenesis and that EoE and ESCC promote distinct epithelial trajectories. How EoE directly impacts ESCC epithelium and immunological environment remains unclear. Here, we perform single nuclei RNA- and ATAC-sequencing on mice with both EoE and ESCC to illuminate potential tumor-limiting changes in the esophagus.

Methods: C57B6 mice were treated with MC903/OVA for 32 days to induce EoE (n=3), 4NQO for 24 weeks to induce ESCC (n=3), MC903/OVA for 32 days then 4NQO for 24 weeks to induce both EoE and ESCC (n=2), or vehicle treatment for control (n=5). Nuclei were isolated from whole esophageal single cell suspensions and sequenced by Illumina. In R, Unique Molecular Identifier tools identified cell barcodes from 10x Genomics libraries, which were aligned by Spliced Transcripts Alignment to Reference and assigned to genes. Count matrices analysis, dimensionality reduction, and population clustering was performed with Seurat. CD45 was used to subset immune cells for separate clustering.

Results: Unsupervised analysis identified 13 cell populations, including 11 epithelial, 2 fibroblast, and 1 endothelial population across conditions (Fig1A). Epithelial subpopulations were defined as proliferative basal, basal, and superficial by expression of gene markers Top2a, Krt5, and Krtdap, respectively; fibroblast and endothelial populations were resolved by expression of Dcn and Pecam1, respectively (Fig1B). Notably, cell cycle analysis between conditions revealed substantial suppression of S phase cell proportions in mice with EoE and ESCC in combination as compared to those with ESCC (Fig1C). 6 unsupervised clusters were further identified among immune cells, with myeloid, T, and B cells identified by established markers (Fig2A, B). EoE was found to restore myeloid cells that were depleted in mice with ESCC and suppress the expansion of cancer-enriched population 1 (Fig2C).

Conclusions: These data reveal considerable effects of EoE inflammation on cancer-augmented proliferative capacity in the esophagus, as well as shifts in the immune environment that may suppress pro-tumorigenic inflammation. Continuing analysis will validate these findings in vivo and examine alterations in activation states of immune cells between conditions that may contribute to anti-tumorigenic activity associated with EoE inflammation. Furthermore, complete evaluation of differential gene expression as well chromatin availability can identify novel targets for therapeutic action in esophageal carcinoma.