Enteric neurospheres retain the capacity to assemble neural networks with motile and metamorphic gliocytes and ganglia

BACKGROUND: Neurosphere medium (NSM) and self-renewal medium (SRM) were widely used to isolate enteric neural stem cells (ENSCs) in the form of neurospheres. ENSCs or their neurosphere forms were neurogenic and gliogenic, but the compelling evidence for their capacity of assembling enteric neural networks remained lacking, raising the question of their aptitude for rebuilding the enteric nervous system (ENS) in ENSC therapeutics. It prompted us to explore an effective culture protocol or strategy for assembling ENS networks, which might also be employed as an in vitro model to simplify the biological complexity of ENS embedded in gut walls. METHODS: NSM and SRM were examined for their capacity to generate neurospheres in mass culture of dispersed murine fetal enterocytes at serially diluted doses and assemble enteric neural networks in two- and three-dimensional cell culture systems and ex vivo on gut explants. Time-lapse microphotography was employed to capture cell activities of assembled neural networks. Neurosphere transplantation was performed via rectal submucosal injection. RESULTS: In mass culture of dispersed enterocytes, NSM generated discrete units of neurospheres, whereas SRM promoted neural network assembly with neurospheres akin to enteric ganglia. Both were highly affected by seeding cell doses. SRM had similar ENSC mitosis-driving capacity to NSM, but was superior in driving ENSC differentiation in company with heightened ENSC apoptosis. Enteric neurospheres were motile, capable of merging together. It argued against their clonal entities. When nurtured in SRM, enteric neurospheres proved competent to assemble neural networks on two-dimensional coverslips, in three-dimensional hydrogels and on gut explants. In the course of neural network assembly from enteric neurospheres, neurite extension was preceded by migratory expansion of gliocytes. Assembled neural networks contained motile ganglia and gliocytes that constantly underwent shapeshift. Neurospheres transplanted into rectal submucosa might reconstitute myenteric plexuses of recipients’ rectum. CONCLUSION: Enteric neurospheres mass-produced in NSM might assemble neural networks in SRM-immersed two- or three-dimensional environments and on gut explants, and reconstitute myenteric plexuses of the colon after rectal submucosal transplantation. Our results also shed first light on the dynamic entity of ENS and open the experimental avenues to explore cellular activities of ENS and facilitate ENS demystification. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s13287-023-03517-y.