Maintaining a functioning epithelial barrier is required for human survival, combining protection from the outside world with tissue-specific functions. The esophageal epithelium ensures continued integrity by endlessly generating new epithelial cells which undergo a stereotyped and coordinated process of differentiation, replacing the entire epithelium within days. Rapid tissue turnover requires the proliferation of esophageal progenitor cells, residing strategically within the epithelium to be able to respond swiftly to changes in the microenvironment. Failure to regulate progenitor cell behavior leads to esophageal dysfunction, often manifested with barrier defects and hyperplasia.

In contrast to other epithelial tissues, differences in esophageal tissue architecture are striking when comparing species. Species-specific distinctions may be the result of an evolutionary adaptation to a combination of factors, including the texture of the food and efficiency of the masticatory system, which collectively impact the softness of the food bolus passing through the esophageal tube. Herein, we explore parallels and highlight differences in the adult mouse and human esophageal epithelium, underscoring areas where future research is required to propel development of promising tools and strategies targeting esophageal disease.

The adult mouse esophageal epithelium is a three-to-four cell layer thick squamous keratinized epithelium (Figure 1A), which is gently folded around the tube circumference to enable expansion upon passing of the food-bolus. The epithelial basal layer contains proliferating (KI67⁺) progenitor cells, marked by K14, K5 and P63 (Rosekrans et al., 2015). Once basal layer progenitor cells upregulate KLF4 (McGinn et al., 2021) (Figure 1A), they commit to differentiation and delaminate out from the basal layer to eventually shed off into the esophageal lumen. The continuous proliferation of basal cells is estimated to renew the esophageal epithelium every 3.5 days (Doupe et al., 2012). While the overall mouse esophageal architecture is strikingly similar to the mouse interfollicular epidermis (Piedrafita et al., 2020), the esophageal epithelium lacks epithelial appendages like the hair follicles and sweat glands of the skin. Recent work, however, describes the presence of rare specialized epithelial taste buds in the upper mouse esophagus (Vercauteren Drubbel and Beck, 2023) indicating that the mouse epithelium is more complex than previously thought.

In contrast to the mouse epithelium, the human esophageal epithelium is up to 40 cell layers thick and displays prominent stromal invaginations, which folds the epithelium into structures called papillae (Geboes and Desmet, 1978) (Figure 1B), reminiscent of the patterning of the human epidermis. The human basal layer is marked by K14, P63, COL17A1 and PDPN (Busslinger et al., 2021; Rochman et al., 2022; Rosekrans et al., 2015) and is largely quiescent (Figure 1B). Proliferating P63-positive progenitor cells locate up to the first six suprabasal cell layers, suggesting a more complex tissue organization than what is found in the mouse epithelium. Epithelial cell layers above the proliferating zone express KLF4, K4 and K13, markers associated with lineage commitment and terminal differentiation. Although the keratin expression profile shifts with human esophageal epithelial differentiation, the keratinized cell layer - present in the mouse epithelium - is absent in the human esophagus. Instead, the human esophagus contains submucosal glands (SMGs) (Figure 1B), which safeguard the surface squamous epithelium by secreting mucus (Long and Orlando, 1999). The turnover time of the human esophageal epithelium is estimated to be around 11 days (Pan et al., 2013).

Due to its simple architecture, the adult mouse esophagus presents an ideal system to uncover mechanisms of epithelial development and homeostasis, and has contributed significantly to our understanding of adult epithelial tissue maintenance and fitness (Abby et al., 2023; Doupe et al., 2012; Frede et al., 2016; Piedrafita et al., 2020). However, the increased complexity of the human epithelial architecture limits translational efforts, especially when probing cell- and signaling-mechanisms related to the development of human esophageal disease.

Stem and progenitor cell behavior is dictated by the local environment, or niche (Lane et al., 2014). The mouse esophageal epithelium, being structurally simple, has no obvious anatomical landmarks that would allow discrete progenitor cell - niche cell interactions. Hence, work using strategies to genetically label an unbiased fraction of basal progenitor cells demonstrate that the esophageal epithelium is maintained by a single progenitor cell population, dividing symmetrically twice per week (Doupe et al., 2012). Daughter cells retain progenitor cell properties or commit to lineage differentiation in a stochastic, random, manner (Doupe et al., 2012). Although this model describes the mouse esophageal tissue well, it does not formally exclude the existence of subset(s) of progenitor cells which behave differently from the bulk population.

Despite the lack of clear anatomical epithelial niches, gene expression within the mouse epithelium is not uniform. Irregular epithelial expression of signaling molecules, including Bmp4, Gli1, Igfbp2 and Igfbp5 suggest that restricted signaling environments could act to diversify behavior in discrete subsets of progenitor cells (Grommisch et al., 2025; Jiang et al., 2015; van Dop et al., 2013). In support of this idea, several studies targeting specific progenitor subpopulations have indicated that heterogeneity within the mouse esophageal progenitor pool exists (Croagh et al., 2007; DeWard et al., 2014; Giroux et al., 2017; Grommisch et al., 2024; Kalabis et al., 2008), reporting differences in in vivo basal clone sizes and in in vitro organoid-forming potential. In addition, recent single-cell profiling reports that local non-epithelial niche cell architectures differ in the upper (proximal) and lower (distal) esophagus (Grommisch et al., 2025), and demonstrates a distinct, distally enriched, epithelial progenitor population (Grommisch et al., 2025; Grommisch et al., 2024). These findings indicate that the regulation of progenitor cell states and behaviors in the mouse esophageal epithelium could be more complex than previously appreciated. The rapid development of in situ sequencing technologies will likely shed new light on thus far hidden progenitor niches and local cellular networks, enhancing our understanding of how progenitor cell behavior is restricted and diversified in structurally simple tissues such as the mouse esophageal epithelium.

The nature of the human esophageal epithelial progenitor pool is less characterized. With the availability of high-resolution sequencing data (Busslinger et al., 2021; Chen et al., 2021; Ding et al., 2024; He et al., 2020; Madissoon et al., 2019; Rochman et al., 2022), we have gained in-depth information about the transcriptional states defining basal and suprabasal cell populations in the human esophagus, differences which may correspond to distinct cellular behaviors. Translating transcriptional cell states into lineage hierarchies and cell behaviors has however proven difficult. Attempts to identify and functionally compare potential progenitor cell populations in the human epithelium include fractionation based on cell surface markers. Separating basal-to-suprabasal populations using cell surface markers ITGA6 or PDPN (Jeong et al., 2016; Rochman et al., 2022) demonstrates high clonogenic capacity in the basal, normally slow cycling, cell population. Taken together, this suggests that basal cells, although largely quiescent in vivo, represent a progenitor population that efficiently forms 2D colonies in vitro. It is possible that proliferating suprabasal cells represent a more transient cell state already en route towards differentiation, as indicated by their low colony forming ability. This would suggest that the human esophageal epithelium is maintained by two distinct progenitor cell populations which can be identified based on their location (basal vs. suprabasal) and proliferation kinetics (slow vs. actively cycling). How they are related hierarchically, and how long each population resides in the epithelium, is still to be addressed.

In addition, basal cells located to folded papillary regions find themselves in a different local environment compared to basal cells in the interpapillary regions (Figure 1B), potentially affecting their progenitor cell potential and behavior. In the human epidermis, topographical location affects basal cell clonogenicity (Jones et al., 1995), suggesting that the same could be true in the esophagus. However, attempts to subset the basal layer into ITGB1^high papillary and ITGB1^low interpapillary cell populations and compare in vitro clonogenicity were largely inconclusive (Seery and Watt, 2000). Available single-cell transcriptional data could be explored to investigate if additional heterogeneity within the basal layer, indicative of distinct papillary-interpapillary cell states, exist.

To protect the non-keratinized human epithelium from damage, submucosal glands (SMGs) are scattered throughout the human esophagus (Figure 1B). Each SMG is composed of multiple exocrine acini which produce mainly mucus, but also bicarbonates and EGF (Epidermal Growth Factor), to facilitate food bolus transport, neutralize stomach acids and promote epithelial regeneration (von Furstenberg et al., 2017). Each acinus is drained by a small duct, which coalesce into one large duct transversing the submucosa and epithelium to allow the SMGs to release their content into the esophageal lumen. Duct cells near the SMG are cuboidal in shape, but transition into a squamous appearance as the duct approaches the lumen of the esophageal tube (Garman, 2017; Garman et al., 2015). Proliferation in SMGs and associated ducts is low during homeostasis but increases in pig and canine models of epithelial damage, including experimentally induced acid reflux (Gillen et al., 1988; Li et al., 1994; Van Nieuwenhove and Willems, 1998; von Furstenberg et al., 2017). Data from animal models also suggest that submucosal glands and/or their connecting ducts may contribute to esophageal epithelial repair upon injury (Gillen et al., 1988; Kruger et al., 2017; Owen et al., 2018), paralleling findings in mice describing activation and contribution of sweat gland duct progenitor cells to the re-epithelialization of the epidermis (Lu et al., 2012). Duct progenitor cells could thus represent a so-far uncharacterized reservoir of latent progenitor cells which can be activated and recruited to contribute to the healing of the squamous esophageal epithelium.

SMG duct cells have attracted attention as a potential cell-of-origin in Barrett’s metaplasia of the esophagus (BE), a condition where the distal part of the human squamous epithelium is replaced by a mosaic of columnar gastric and intestinal cell types, which can progress to dysplasia and adenocarcinoma formation (Jung et al., 2011). The development of BE is linked to repetitive stomach acid reflux and thus initiates at the gastro-esophageal junction. Due to differences in mouse and human gastro-esophageal junctional anatomy (Caron et al., 2015), the cellular origin of BE has proven challenging to formally verify in humans. Although mice, lacking SMGs, can develop metaplastic lesions upon injury (Vercauteren Drubbel et al., 2021), clonal genetic analysis has recorded matching somatic mutation profiles in manually dissected human SMG duct cells and neighboring Barrett’s glandular tissue, indicating a common cellular origin (Leedham et al., 2008). Single cell transcriptional profiling also concluded that the SMG duct cells are transcriptionally closer to cells found in BE, when compared to cells normally located in the intestine or gastric cardia (Owen et al., 2018). However, more recent in-depth chromatin analysis excluded the SMG duct and instead pinpointed the gastric cardia as the cell-of-origin in BE (Nowicki-Osuch et al., 2021). Considering the possibility that multiple progenitor populations could drive the histogenesis of BE would perhaps reconcile these seemingly opposing findings.

Much of our mechanistic understanding of how the human esophageal epithelium is maintained is extrapolated from work in the mouse. Efforts identifying and functionally characterizing key signaling pathways in the adult human epithelium come from mapping somatic disease-causing mutations (Lin et al., 2018; Liu et al., 2017; Martincorena et al., 2018; Zhang et al., 2015) and establishing culture conditions permissive for maintenance and differentiation of esophageal epithelial cells (Kalabis et al., 2012; Laczko et al., 2017; Milne et al., 2024; Mou et al., 2016; Sachdeva et al., 2021; Urano et al., 2025). In addition, reactivation of important developmental pathways is not uncommon in esophageal disease (Vercauteren Drubbel et al., 2021), some of which have successfully and sequentially been exploited to generate esophageal epithelial basal-like cells from human pluripotent stem cells (hPSCs) (Bailey et al., 2019; Ferrer-Torres et al., 2022; Nakagawa et al., 2020; Trisno et al., 2018). Despite these recent advances, a robust, reproducible and readily available esophageal epithelial cell model is still lacking.

Here we provide a short overview of the current knowledge of key pathways known to affect either development of esophageal disease, or expansion and maintenance of esophageal epithelial cells in culture. The suboptimal conditions used for primary and hPSC cell models suggest that additional, yet unidentified growth factors or signalling cues may be required to establish a cell platform that accurately captures in vivo epithelial progenitor heterogeneity and models adult disease initiation.

The last two decades have recorded groundbreaking insights in the cell biological mechanisms regulating tissue development and homeostasis. For reasons unknown, interest in the esophageal epithelium has been comparably limited. We therefore still lack many of the pieces required to complete the puzzle of esophageal tissue maintenance. Ambitious transcriptional, largely descriptive, data is emerging allowing for further in-depth mechanistic studies, both in terms of understanding epithelial progenitor biology, but also detailing cell-to-cell interaction networks and disease-associated transcriptional profiles. In parallel, available human cell models required to test new hypothesis are continuously improving, allowing researchers to ask more targeted questions. The recent hPSC-derived esophageal mesenchymal cells (Han et al., 2020; Kishimoto et al., 2022) are but one example of a valuable tool for understanding human esophageal development and disease. We hope that this review article will provide inspiration to future work dedicated to advancing our understanding of this intriguing epithelial tube.