Gut 56:459-461 doi:10.1136/gut.2006.107748
  • Commentary

Transplanting the enteric nervous system: a step closer to treatment for aganglionosis

  1. Michael D Gershon
  1. Correspondence to:
    Professor M D Gershon
    Department of Pathology and Cell Biology, Columbia University College of Physicians and Surgeons, New York, NY 10032, USA; mdg4{at}

    Autologous transplantation can be used to treat Hirschsprung’s disease by implantation and proliferation of the crest-derived stem cells in vitro

    One might think that Hirschsprung’s disease (congenital megacolon) should be passé as a medical problem. After all, many genes, including RET, GDNF, NRTN, EDNRB, EDN3, ECE1, PHOX2b, SOX10, PAX3 and SMADIP1 (SIP1, ZBFX1B),1–7 have successfully been linked to its pathogenesis. Knowledge of the actions and interactions of these genes and their products has enabled the processes by which the bowel is colonised to be, if not completely understood, at least comprehended in general terms.3,8,9,10 Effective treatment for Hirschsprung’s disease, moreover, exists in the surgical removal of the aganglionic segment of bowel.11–13 Unfortunately, the medical problems posed by Hirschsprung’s disease continue despite the lengthy list of genes implicated in its generation and the progress that has recently been made in understanding enteric nervous system (ENS) development. Unresolved medical problems continue because advances made in comprehending genes and pathogenesis have not been translated into new and improved methods of treatment; moreover, although surgical techniques are evolving and associated morbidity is decreasing,12,13 the surgical treatment of Hirschsprung’s disease essentially converts an otherwise lethal defect into a chronic condition with which many, if not most, patients must learn to cope.14,15

    Hirschsprung’s disease occurs when a variable length of terminal bowel is congenitally aganglionic. Because the reflexes and behaviours mediated by the ganglionated plexuses of the ENS are essential for propulsive motility and normal secretion,16 aganglionosis results in a pseudoobstruction that, if left untreated, is incompatible with life. Reliable modern statistics on untreated Hirschsprung’s disease are not available because failure to treat it is immoral; however, aganglionosis is lethal to animals with genetic defects that model the condition.17–23 Because the aganglionic region of the bowel lacks the inhibitory neurotransmitter nitric oxide some authors have speculated that the aganglionic zone goes into spasm and narrows to become obstructive24; however, it is more likely that motor patterns simply fail to propel luminal contents through the aganglionic zone so that the ganglionated bowel proximal to the aganglionic segment dilates. Removal of the aganglionic portion of the gut is thus obviously necessary in the treatment of Hirschsprung’s disease, but in many patients it is not sufficient.

    The greatest problems faced by patients after the definitive surgical correction of the aganglionosis of Hirschsprung’s disease include faecal soiling,15 constipation and postoperative enterocolitis.25 Studies vary in the reported incidence of these complications, and the type of surgery used to carry out the repair undoubtedly matters; however, soiling has been reported in as many as 76% of patients.15 A transanal one-stage pull-through operation may be advantageous for rectosigmoid aganglionosis13 even though it carries a high risk of postoperative enterocolitis because a single surgical procedure is preferable to two12,25; however, a modified Duhamel procedure has been advocated as superior to any other for total colonic aganglionosis.26 Whatever procedure is used, faecal soiling is a serious risk, which over the long term causes surprisingly less psychiatric morbidity than would be expected, given the social stigma attached to that particular defect; nevertheless, faecal soiling gives rise to a great deal of concern in families and is highly distressing to patients.15 The outcomes of treatment are worse for patients with total colonic aganglionosis than for those with short-segment disease, and patients with total colonic aganglionosis tend to perceive themselves as less well adjusted than their matched pairs with a shorter aganglionic region of gut.27 Additional surgeries, including posterior myotomy/myectomy, can be undertaken to lessen the effect of persistent soiling, but these procedures are not invariably successful.28 Clearly, no treatment that carries a high risk of faecal soiling can be considered to be perfect, and one has to examine what we know about the pathogenesis of Hirschsprung’s disease and the development of the ENS to determine whether anything can be done that is better than what is now being done to treat Hirschsprung’s disease.

    The gut is colonised by precursors that migrate to it from the neural crest. The premigratory crest is a heterogeneous structure, in that it seems to contain both pluripotent and fate-restricted precursor cells.29–34 The population of postmigratory cells that colonises the gut is multipotent when it arrives in the bowel,35–37 although individual cells within this population may already be committed to a single fate. The observation that a subset of the serotonergic neurones of the mouse bowel are born as early as embryonic day 8.5, which precedes the migration of any crest-derived cells into the gut, strongly suggests that some cells of the colonising population of crest-derived cells are committed and even postmitotic when they enter the bowel.38 Despite the fact that the crest-derived cell population that colonises the gut contains members that are postmitotic, the group also contains cells that are self-renewing, multipotent stem cells. In fact, it is of particular interest that stem cells are still present in the postnatal bowel.39–41 Both hope and logic suggest that these neural crest-derived stem cells may ultimately provide a solution to the unsatisfactory current status of treatment for Hirschsprung’s disease.

    Major problems must be overcome before enteric neural crest-derived stem cells can be successfully used to treat Hirschsprung’s disease, and at least some of these problems have successfully been dealt by Almond et al42 (see page 489). First, the crest-derived stem cells have to be isolated and then expanded to obtain enough cells for autologous transplantation, which would be the goal. Once instilled into the bowel wall, moreover, these cells would have to migrate to the correct destinations and form appropriate connections with one another so that they can reconstitute the reflexes and integrative neural activity of the normal ENS. It is obviously not enough just to get neurones to form, or even to migrate to correct destinations in the bowel wall; neurones must be functional and in sufficiently good control of effectors so that reflexes are rescued and the pseudo-obstruction of Hirschsprung’s disease can be corrected. Almond et al42 have adapted the technique of producing neurospheres from single-cell suspensions to obtain enriched populations of crest-derived stem cells43–45 and they have been able to expand the size of the original population by maintaining the proliferation of stem cells in vitro. The investigators, furthermore, succeeded in implanting crest-derived stem cells into an aganglionic mouse gut and, after doing so, they observed that the cells fortunately migrate along pathways that are appropriate for cells derived from the neural crest. Strikingly, the implanted crest-derived stem cells differentiate within the aganglionic zone to give rise to end-stage cells that express phenotypic markers identifying them as enteric glia and neurones. The neurones, furthermore, express some of the molecules that characterise the chemical code used for identifying enteric neurones,46 such as vasoactive intestinal peptide and nitric oxide synthase.

    This report is undoubtedly a major step forward, which, for the first time, implies that autologous transplantation of neural crest-derived enteric stem cells is realistic as a prospective treatment for the aganglionosis of Hirschsprung’s disease. Of course, one must be cognizant, as are Almond et al,42 of the enormity of the remaining problems that must still be overcome before autologous transplantation replaces pull-through operations as routine treatment for Hirschsprung’s disease. Although Almond et al demonstrate that some of the correct markers are expressed by the neurones that develop from grafts of neural crest-derived stem cells, they have not shown that the full chemical code46 is acquired. The minimum number of neurotransmitters and neuromodulators necessary for function is unknown, because it is unclear as to whether or not a serviceable ENS might be formed if some of the elements of the normal chemical code failed to develop. More importantly, Almond et al were not yet able to determine whether synaptic connections developed between enteric neurones, and between these neurones and their effectors. In the absence of that information, one can only speculate about whether the newly formed neurones fashion themselves into the complex microcircuits responsible for regulating propulsive and secretory activity. The bottom line restoration of function and clearing of the pseudo-obstruction, furthermore, are still to be demonstrated. There are, however, always many hurdles in the path that leads from scientific discoveries to successful treatment. Although Almond et al have not cleared all of them in showing that autologous transplantation can be used to treat Hirschsprung’s disease, they have leaped across some daunting hurdles and have thus started the race to the cure.

    Autologous transplantation can be used to treat Hirschsprung’s disease by implantation and proliferation of the crest-derived stem cells in vitro


    • Funding: The author’s research is supported by grants NS12969 and NS15547 from the National Institutes of Health, USA, and a research grant from Novartis.

    • Competing interests: None.


    Relevant Article