Background:
Barrett’s esophagus (BE) is a precancerous condition defined as replacement of the normal esophageal squamous epithelium with metaplastic columnar intestinal epithelium caused by chronic gastroesophageal reflux disease. There is a deficiency of knowledge regarding the molecular pathogenesis of BE development. To this end, a thorough elucidation of the development context by which BE develops will uncover novel avenues to detect BE and revert it back to normal esophagus to ultimately halt the development of cancer. At embryonic day 9, the dorsal foregut is marked by expression of SOX2, a transcription factor of the Sry-like HMG box family, which is required for esophageal development. SOX2 continues to be expressed in the esophageal epithelium throughout adulthood as an important homeostatic maintenance factor. I hypothesize that SOX2 functions to maintain foregut squamous epithelial identity, and its loss is a critical step during BE development.

Methods/Results:
Here, we show using human BE tissue microarrays that SOX2 is strongly expressed in the normal squamous epithelium and its expression is significantly decreased during BE development. To assay the molecular effects of SOX2 expression loss, we have conducted a series of novel experiments involving a foregut-specific inducible Sox2 knockout mouse model, Krt5^CreER/+; Sox2^flox/flox; ROSA26^{LSLTdTomato/+}, and assessed the histologic and transcriptional effects of SOX2 loss on squamous epithelium homeostasis. Specifically, we show in vivo that loss of SOX2 induces increased basal layer proliferation, decreased mature squamous structural protein expression, and elicited a squamous damage response program. We have recapitulated and further explored these aspects of SOX2 loss in vitro though 3D, 2D transwell “proliferative” condition-based, and 2D transwell “differentiation” condition-based squamous organoid cultures. In addition, we have performed Cleavage Under Targets and Release Using Nuclease (CUT&RUN) experiments to determine the direct transcriptional targets of SOX2 at homeostasis in the adult murine foregut. Finally, we have treated WT and Krt5^CreER/+; Sox2^flox/flox; ROSA26^{LSLTdTomato/+}mice with bile acid (deoxycholate) and found that SOX2 loss uniquely sensitizes mice to increased death upon exposure. Interestingly, long-term Krt5^CreER/+; Sox2^flox/flox; ROSA26^{LSLTdTomato/+}survivor mice display expansion of columnar glands at the squamocolumnar junction that exhibit intestinal characteristics.

Conclusions:
In summary, it is possible that in BE there is a stepwise transcriptional progression towards a more posterior phenotype with loss of SOX2 expression being an important initial step. Together, these experiments will elucidate novel molecular pathways involved in BE maintenance and may reveal novel therapeutic avenues to treat BE and prevent esophageal cancer.

Ramon Jin

Appearances