In vivo transplantation of fetal human gut‐derived enteric neural crest cells

Abstract The prospect of using neural cell replacement for the treatment of severe enteric neuropathies has seen significant progress in the last decade. The ability to harvest and transplant enteric neural crest cells (ENCCs) that functionally integrate within recipient intestine has recently been confirmed by in vivo murine studies. Although similar cells can be harvested from human fetal and postnatal gut, no studies have as yet verified their functional viability upon in vivo transplantation. We sought to determine whether ENCCs harvested from human fetal bowel are capable of engraftment and functional integration within recipient intestine following in vivo transplantation into postnatal murine colon. Enteric neural crest cells selected and harvested from fetal human gut using the neurotrophin receptor p75NTR were lentivirally labeled with either GFP or calcium‐sensitive GCaMP and transplanted into the hindgut of Rag2 − /γc − /C5 −‐immunodeficient mice at postnatal day 21. Transplanted intestines were assessed immunohistochemically for engraftment and differentiation of donor cells. Functional viability and integration with host neuromusculature was assessed using calcium imaging. Transplanted human fetal gut‐derived ENCC showed engraftment within the recipient postnatal colon in 8/15 mice (53.3%). At 4 weeks posttransplantation, donor cells had spread from the site of transplantation and extended projections over distances of 1.2 ± 0.6 mm (n = 5), and differentiated into enteric nervous system (ENS) appropriate neurons and glia. These cells formed branching networks located with the myenteric plexus. Calcium transients (change in intensity F/F0 = 1.25 ± 0.03; 15 cells) were recorded in transplanted cells upon stimulation of the recipient endogenous ENS demonstrating their viability and establishment of functional connections.


| INTRODUCTION
Enteric neuropathies are a diverse and clinically important range of conditions characterized by aberrant or absent propulsive contractile activity secondary to loss or malfunction of the enteric nervous system (ENS). [1][2][3][4] Such disorders range from those in which there is a congenital or acquired absence of intrinsic ENS (e.g. Hirschsprung disease and esophageal achalasia), to conditions such as intestinal pseudoobstruction and slow transit constipation which result from more subtle neuronal deficits that remain to be better defined. 5 treatments are unsatisfactory and limited to surgical interventions, which are associated with high levels of morbidity and mortality. 1 Alternative therapeutic strategies are required for these conditions. 8 The ENS is formed during embryogenesis by migratory neural crest cells arising in the vagal and sacral regions of the neural tube (reviewed in Ref. 9). Enteric neural crest cells (ENCC), which give rise to enteric neurons, glia, and stem cells, can be isolated from both embryonic and postnatal murine intestine. 10,11 Furthermore, upon transplantation, they are able to colonize both ganglionic and aganglionic murine bowel in vivo, forming neural networks that display functional integration with the host neuromusculature. 12,13 The potential to develop a clinically applicable cell replacement therapy for enteric neuropathies has been validated by the isolation of ENCC from fetal and postnatal human gut 8,[14][15][16] To date, human gut-derived ENCCs have been transplanted within in vitro models of aganglionosis, showing physical integration and differentiation. 14,15,17 However, their functional integration remains to be demonstrated within a postnatal in vivo model. Here, we report the efficient isolation of ENCC from human fetal gut and successful engraftment and functional integration following in vivo transplantation into immunodeficient murine gut.

| Animals
Animals were maintained, and experiments were performed, in accordance with local approvals and the UK Animals (Scientific Procedures) Act 1986 under license from the Home Office (PPL70/7500). Rag2 − /γc − / C5 − -immunodeficient mice, deficient in innate immunity and lacking all lymphocytes, were used as recipients for ENCC transplantations. 18,19

| Human cells and tissues
Fetal human gut (age 12-15 weeks) was obtained from the Human Developmental Biology Resource, UCL Institute of Child Health, London, UK, with informed, written consent and under ethical approval from the Health Research Authority (08/H0712/34+5). Studies were performed according to the Declaration of Helsinki. Enteric neural crest cells were isolated, sorted, and cultured as described previously. 14 They generated primary neurospheres after approximately 1 week, and these were transplanted within 15-30 days.

| Lentiviral labeling of human ENCC
Lentiviral constructs expressing enhanced green fluorescent protein (EGFP) 20 and GCaMP (Addgene plasmid 42168:pJMK019, Adam Cohen; Addgene, Cambridge, MA, USA) were used to label cells with high levels of GFP or a calcium-sensitive GFP construct, respectively.
Isolation of lentiviral particles and subsequent transduction of cells was conducted according to a protocol described previously. 20 Mice were maintained for 4 weeks posttransplantation before they were killed and analyzed.
Images were acquired on a Zeiss LSM 710 confocal microscope (Zeiss, Cambridge, UK) and processed using ImageJ 21 and Adobe Photoshop CS3 software (Adobe, San Jose, CA, USA).

| Calcium imaging of transplanted ENCC
Colonic gut samples were prepared as described previously. 13 The endogenous ENS was stimulated via an electrode placed approximately 200 μm from the transplanted cells. Changes in the fluorescence intensity of GCaMP elicited by calcium transients within transplanted cells were then imaged and processed as described previously. 13   ganglia-like structures (Fig. 1C) reminiscent of the endogenous ENS.

Four weeks after transplantation, GFP+ cells expressed neuronal
and glial markers (TuJ1, Fig. 1A-C and S100, Fig. 1D, respectively). Cell bodies from transplanted cells were located adjacent to those of endogenous neurons (Fig. 1C) and projections from transplanted cells followed endogenous ENS fiber tracts of the myenteric plexus, showing physical integration within the endogenous network. Projections were visible over a distance of 1.2 ± 0.6 mm (n = 5), after 4 weeks ( Fig. 1A and B).

| DISCUSSION
A number of studies using murine cells support the idea of using stem cell transplantation for the treatment of enteric neuropathies such as Hirschsprung disease. [12][13][14] As a first step to explore the potential of human gut-derived ENCC for in vivo transplantation, we sought to study fetal enteric neural stem cells. We have previously shown that neurospheres containing such cells are formed faster and with higher efficiency than their postnatal counterparts. 16 We wished to confirm that commit- Previously, we reported higher cell engraftment outcomes following in vivo transplantation of mouse-derived ENCC (90.3%) in contrast to 53.3% in this study. This is likely to reflect both our experience that human cells appear somewhat less robust under experimental conditions than their mouse counterparts as well as the expected inherent variability of human samples utilized in our study (e.g., the age and condition of the sample received, gut regions available for culture, and genetic variability between samples). There was no evidence of host immune response in the examined gut (data not shown) to explain the reduced engraftment of human-derived cells. reducing the risk of uncontrolled proliferation. This is supported by our previous work using mouse-derived ENCC, which provided evidence that this cell type is safe to transplant within an in vivo environment. 13 Such safety data must be further extended to human gut-derived ENCC. Here

CONFLICTS OF INTEREST
The authors have no competing interests.

AUTHOR CONTRIBUTION
NT

SUPPORTING INFORMATION
Additional Supporting Information may be found online in the supporting information tab for this article.