Epithelial Dclk1+ cells are not neural crest derived
We have read with great interest the editorial by Koga and colleagues in this journal. In their publication, the authors bring forward a fascinating hypothesis as to why Doublecortin-like kinase 1 (Dclk1) might be expressed in some neuroendocrine tumors. Dclk1 is known to be a marker for radial glial cells, which are precursors for neural stem cells. In addition, Dclk1 also labels rare epithelial cells outside of the central nervous system (e.g., tuft cells), some of which may possess progenitor or cancer stem and initiating properties. Based on the fact that Dclk1 is expressed in melanoma, neuroblastoma and some neuroendocrine tumor cells, the authors suggest that the Dclk1+ lineage might share a common origin, namely, the neural crest (1). As neural crest cells are known to undergo epithelial to mesenchymal transition leading them to exit the neural tube, migrate and invade the developing embryo (2,3), this hypothesis of a neural crest origin could in theory explain the high degree of plasticity and tumorigenicity found in Dclk1 cells in the pancreas and the colon, as well as in pancreatic cancer cell lines (4-7). While the degree of plasticity of the crest-derived cells is reduced once they migrate, some cells have been postulated to be multipotent when they arrive in the bowel (8,9) and could therefore be candidates for quiescent stem or progenitor cells in the adult individual.