Normal neuromuscular function of the esophagus, stomach, small bowel, and colon requires the coordination of the tubular neuromuscular structures and the relevant sphincters. Hirschsprung's disease results from a congenital absence of ganglion cells in some part of the distal gut. The recent progress in using stem cells to restore defective ganglion cells offers new hope for the potential cure of Hirschsprung's disease.

Sources of ENS stem cells can be classified in various ways: Pluripotent stem cells can be harvested as embryonic stem cells, or induced pluripotent stem cells. Multipotent stem cells can be harvested as central nervous system (CNS)-derived neural stem cells, non-CNS-derived neural stem cells or adult (tissue) ENS stem cells.

ENS neurosphere is defined as cellular aggregates containing both stem cells and their differentiated derivative cell types. Neurospheres were harvested from embryonic tissue [13], from post-natal human gut full-thickness tissue and post-natal human gut mucosal tissue (neonates, children, and adults) [14,15]. The most distinguished achievement in harvesting neurospheres is from autologous gut mucosal tissue by endoscopy. This allows for avoiding life-long immune suppression and provides a renewable source of ENS stem cells. Neurospheres were expanded with ex vivo culture and transplanted into a model of the aganglionic gut. Expansion and differentiation into neuron and glial cells was demonstrated [13-15]. Furthermore, neurons from transplanted neurospheres form synapses and regulate the contractility of the developing gut [16]. Intraperitoneally injected selected enteric neural crest stem cells engrafted diffusely throughout the postnatal gut of Hirschsprung's disease rats and differentiated into neurons and glia. Engraftment was not uniform, likely related to age-dependent changes in the gut mesenchyme [17]. Intraperitoneal injection is easily performed in sick neonates and may be developed as a technique for supplying exogenous ENS cells to the diseased postnatal gut. Smooth muscle protein from the aganglionic bowel increased neuronal survival and network formation of myenteric neurons and neural crest derived stem cells [18]. This implies that the microenvironment of stem cells is significant.

ICCs are cells of mesodermal origin and generate unitary potential and slow waves. ICCs provide pacemaker activity to orchestrate gastrointestinal peristalsis and mixing. Tyrosine kinase receptor (c-Kit) and ANO 1 (TMEM16A), a membrane protein associated with calcium-dependent chloride channel are immunohistochemical markers of ICCs. The ultra-structural gold standard of ICCs includes close contacts established with nerve varicosities and the formation of numerous gap junctions, both with each other and with smooth muscle cells. The c-Kit signaling pathway, activated by stem cell factor (SCF), is the critical pathway associated with the control of ICCs survival and proliferation. In humans, the loss of ICC in motility disorders is nearly always associated with the loss of another cell type, including enteric nerve cells [12]. Neuronally derived NO (from neural nitric oxide synthase [nNOS]) modulates ICC numbers and network volume in the mouse gastric body. In NOS-/- mice, the remaining ICC cell bodies were often less well-formed and the processes blunted and decreased resulting in a disrupted network; this was most pronounced in the myenteric plexus region. Neuronally derived NO in particular is required for the maintenance of ICCs. NO appears to be a survival factor for ICCs [19]. Gastric relaxation in diabetics is hampered mainly by impaired NOS expression in the gastric myenteric plexus [20]. Long-standing diabetes mellitus may be associated with a decrease in a number of ICCs and a decrease in inhibitory innervations, associated with an increase in excitatory innervations. In murine diabetic gastroparesis, reduced SCF links smooth myopathy and loss of ICCs [21]. The restoration of ICC numbers and jejunal electrical rhythm, resulting from the blockade of the c-Kit signaling pathway, could be facilitated by local SCF administration in mice [22]. Exogenous SCF partially reversed the pathologic changes of ICC in the colon of diabetic mice [23].

Up to now, three resources for epithelial/mucosal tissue regeneration have been identified, as follows: 1) from a single stem cell, stem cells can be regenerated with ex vivo expansion, 2) from a single intestinal crypt, intestinal crypts can be regenerated with ex vivo expansion, 3) from tissue engineering, intestinal tissue can be regenerated with a scaffold and stem cells.

The above-mentioned two methods usually focus on epithelial regeneration. However, SBS is a tragic status affecting all of the components of the intestine. For SBS, more than epithelial regeneration is needed. In tissue engineering, stem cells or organoid units (multicellular clusters with predominantly epithelial contents in case of SBS) were loaded onto a scaffold, and part of the organ could be regenerated. In brief, animal organs are decellularized to obtain acellular extracellular matrix (ECM) scaffolds (all molecular components [collagens, elastin, fibronectin, laminin, glycosaminoglycans, etc.] preserved, as well as essential growth factors that are present within the ECM scaffold). Cells are harvested from the patient, expanded, and seeded onto or into the scaffold. This construct is allowed to mature in a bioreactor. After maturation, it is implanted in the patient [31].

Outcomes of transplantation of the whole liver or pancreas are far better than those of transplantation of liver cells or pancreatic islets alone [31]. The significance of the stem cell niche (signals provided by these cellualr and acellular components appear to be integrated by stem cells to inform their fate decision, self renewal or differentiation, migration or retention, and cell death or survival) should be evaluated from this perspective.

Autologous tissue-engineered small intestine from ex vivo expansion enables avoiding the problems of transplants, immunosuppression, and donor supply. In a rat study, a tissue- engineered small intestine was formed by transplanting donor organoid units on a polymer scaffold into a host [32]. SBS-induced rats regained a higher percentage of preoperative weight with intestinal organoids using tissue-engineered small intestine than those without tissue engineered small intestine [33]. The tissue-engineered small intestine from autologous tissue in a large animal model showed successful differentiation into goblet cells, enteroendocrine cell, smooth muscle and ganglion cells in Auerbach's and Meisner's plexus and stem cells [34].

Animal experimental trials for esophageal replacement with tissue engineering have been performed. A tissue engineered esophageal construct may be created by the combination of a scaffold and stem cells [35]. In a canine study, an implanted tissue-engineered esophagus showed distensibility, but no peristaltic contractions. Shrinkage of the keratinocytes and smooth muscle also resulted in a shorter segment of restored esophagus [36]. This can be applied to esophageal atresia with insufficient length for primary anastomosis.

A tissue-engineered stomach can replace a native stomach in a rat model. Replacement of the native stomach by a tissue-engineered stomach had beneficial effects on the formation of neomucosa and smooth muscle layers in the tissue-engineered stomach [37]. A tissue-engineered stomach has the potential to function as a food reservoir following total gastrectomy in a rat model [38].

Transplantation of neural stem cells to the pylorus showed improved relaxation of pylorus muscle strips and improved gastric emptying in nNOS deficient mice [39].

A tissue-engineered rat large intestine can be successfully produced with fidelity to the native architecture and in vitro function from neonatal syngeneic tissue, adult tissue, and tissue-engineered colon itself [40].

The ring-shaped construction of the IAS was bioengineered in vitro from isolated smooth muscle cells in rabbits [41], in humans [42], and in mice [43], and bioengineered IAS rings demonstrate physiological functionality [41-43]. Bioengineered IAS rings were implanted subcutaneously and successfully grew and survived with respect to viability and functionality in mice [44]. Human IAS circular smooth muscle was co-cultured with immortomouse fetal enteric neurons. Implanted, intrinsically innervated bioengineered human IAS tissue developed neovascularization, myogenic tone, and normal contraction and relaxation characteristics in response to testing in mice [45]. These studies might pave the way for combining the enteric nervous and gastrointestinal smooth muscle components that will be critical in providing treatment for clinical neuromuscular diseases such as fecal incontinence.