Establishment and multi-resolution transcriptomic landscape profiling of diverse equine intestinal organoid models

Abstract Background Equine intestinal organoid (EIO) models can recapitulate the in vivo structure and function of the small intestinal epithelium, making them ideal tools for studying the equine intestine. However, the specific marker genes of equine intestinal epithelial cells remain incompletely understood. These issues pose challenges to assessing the differentiation status and functional integrity of the EIO model in practical applications. Results In the present study, four types of equine intestinal in vitro models were successfully constructed, including conventional undifferentiated 3D EIOs, differentiated 3D EIOs, submerged 2D EIO monolayers, and air-liquid interface (ALI) monolayer models. All models were established using jejunal crypts isolated from three healthy 3-year-old male Yili horses. A total of 54,495 high-quality nuclei obtained from conventional and differentiated 3D EIOs were profiled via single-nucleus RNA sequencing (snRNA-seq). Five major intestinal epithelial cell subtypes were identified, including enterocytes, proliferative crypt cells, SOX9 + crypt cells, goblet cells, and proliferative transit-amplifying (TA)-like cells, with unique equine-specific cell marker expression patterns. Quantitative analysis showed that conventional EIOs consisted of 54.91% enterocytes, 24.72% TA-like cells, 17.54% proliferative crypt cells, and 2.82% SOX9 + crypt cells, with no goblet cell clusters detected under the applied analytical criteria. In contrast, differentiated EIOs exhibited a dominant proportion of enterocytes exceeding 91%, and goblet cells accounted for 2.41%, while the proportions of other epithelial subtypes were significantly decreased. RNA in situ hybridization was performed to verify the spatial localization of subtype-specific marker genes in native equine jejunal tissues. Additionally, bulk RNA sequencing was employed to compare transcriptomic differences between submerged 2D EIO monolayers and ALI models. Transcriptomic results confirmed that compared with traditional submerged monolayers, the ALI culture system retained a higher proportion of crypt cells and effectively promoted the maturation and terminal differentiation of equine intestinal epithelium. Conclusion This study established and characterized four EIO models via morphological observation, snRNA-seq, and bulk RNA-seq. Rare epithelial subsets were not identified, partly because of the major limitation that snRNA-seq detects only nuclear RNA and interspecies differences in intestinal cell marker genes. These results advance our knowledge of equine intestinal epithelial heterogeneity and supply standardized in vitro models for subsequent veterinary nutrition, pharmacology and intestinal disease research.