Human enteric nervous system progenitor transplantation restores functional responses in Hirschsprung Disease patient-derived tissue

Stem cell culture and differentiation

Use of hESCs has been approved by the Human Embryonic Stem Cell UK Steering Committee (SCSC15-23). The following hPSC lines were employed: WA09/H9 (20), H9-RFP (21), SFCi55-ZsGr (22), MasterShef11 (23) and SOX10:GFP (24). All cell lines were cultured routinely in feeder-free conditions in mTeSR1 (Stem Cell Technologies) medium on Geltrex LDEV-Free reduced growth factor basement membrane matrix (Thermo Fisher). Cells were passaged twice a week after reaching approximately 80% confluency using ReLesR (Stem Cell Technologies) as a passaging reagent. Cells were screened for mycoplasma using Lookout Mycoplasma PCR detection kit (Sigma-Aldrich) or Mycostrip detection kit (Invivogen and were routinely screened for indicators of pluripotency (OCT4, NANOG, Table S1) and SSEA4 (25, 26). Neural crest differentiation and enteric neurons were generated as previously described (19) and detailed in the supplementary methods.

Intracellular staining and flow cytometry

Cells were dissociated using TrypLE Select (Gibco), fixed in Paraformaldehyde (PFA, 4% w/v) for 10 min at room temperature (RT) and permeabilised/blocked with blocking buffer containing 0.1% Triton X-100 and bovine serum albumin (BSA, 1% w/v) in phosphate-buffered saline (PBS) for 1hr at RT. Primary antibodies were diluted in blocking buffer and cells were incubated with anti-SOX10 (Table S1) followed by anti-rabbit 488 Alexa fluorophores (A- 21206, Invitrogen). Cells were analysed on a FACS Jazz cell sorter (BD). Data were analysed with FlowJo software (BD).

Single cell RNA sequencing and analysis

Single cell RNA sequencing was performed by Single Cell Discoveries (Utrecht, The Netherlands), Cells were differentiated as described above. At days 0, 4 and 6 of differentiation, cells were dissociated using TrypLE Select (Gibco) and cryopreserved in STEM-CELLBANKER (AMSBIO). Library preparation was carried out using the 10x Genomics 3’ v3.1 kit followed by sequencing on an Illumina Novaseq 6000. Approximately 10,000 cells per time point were analysed at 30,000 reads/cells, 150 bp, paired end. Detailed RNA sequencing analysis can be found in supplemental methods.

Calcium imaging

hESC-derived enteric neurons (H9 background) were loaded for 20 min using the Ca²⁺ indicator Fluo-4AM (Invitrogen). Cells were grown on 18mm glass coverslips and transferred to a specifically designed recording chamber. Images were taken using TILLVISION software (TILL Photonics) with a Zeiss Axiovert 200M microscope equipped with a monochromator (Poly V) and a cooled CCD camera (Imago QE), both from TILL Photonics. Fluo-4AM was excited at 470nm and its fluorescence emission collected at 525nm using a 20X (NA 0.75) objective. Images were collected and analysis was performed using custom written macros in IGOR PRO (Wavemetrics). All recordings were performed at RT. To analyse the recorded images Regions Of Interest (ROIs) were drawn over each cell and fluorescence intensity was calculated and normalized per ROI to its baseline starting value. Changes in fluorescence intensity were calculated and expressed as a fraction of the baseline fluorescence, as F/F0. For neuron depolarization, a high K⁺ solution was applied. Cells were also stimulated in an electric field using a 2 second stimulation at 20Hz (40 stim in total) or upon treatment with the nicotinic agonist DMPP (10^-5M, Sigma Aldrich).

Processing of patient-derived HSCR colonic surgical discard tissue

Human colonic surgical discard tissue from HSCR patients was obtained with informed consent through Great Ormond Street Hospital, under local REC approval (04/Q0508/79 and 18/EE/0150). Tissue samples were processed under sterile conditions. Briefly, tissue was washed in PBS with Primocin antibiotic (1:500 v/v, InvivoGen) and denuded of the mucosa using fine forceps and micro-scissors. The serosa was removed along with any fatty tissue/blood vessels. From each patient sample, dependent on sample size, up to a maximum of 4 sections of tissue (approx. 1cm², harvested from representative points along the oral-anal axis) were isolated for transplantation alongside an equivalent segment (radially adjacent) as a control for sham transplantation. Transplant and sham tissue samples were pinned in sylgard-coated dishes filled with culture media (DMEM F12 with HEPES and glutamine, supplemented with N2 (Invitrogen), B27 (Invitrogen) and Primocin (500mg/mL, Invivogen)) and placed in a humidified incubator (37°C, 5% CO2) overnight.

Transplantation of ENS Progenitors to human HSCR colonic tissue

Frozen vials of ENS progenitors were thawed and resuspended in culture media (approx. 125,000 cells/µL). Cell viability and density was assessed using trypan blue dye (Thermo Fisher Scientific Inc.) and an automated cell counter (TC20 Automated Cell Counter, Biorad). Injections into human tissue samples were made using a blunt 5µL Hamilton syringe attached to a pulled glass capillary needle with a tip diameter of 60µm. 4µL cell solution (500,000 cells in total) was injected at a rate of 0.5µl/min. Following injection, the capillary was kept in place for 1 minute before withdrawal. Sham samples received an injection of vehicle-only under the same conditions. Samples were returned to the incubator and maintained for 3 weeks, with media changed every other day.

Organ bath contractility

Human tissue samples were transferred from culture media to oxygenated Krebs solution and mounted in organ baths (10ml, SI-MB4; World Precision Instruments Ltd). Samples were connected to force transducers (SI-KG20, World Precision Instruments Ltd) via sutures (size 4.0, Fine Science Tools) under an initial tension of 1g. Organ baths were maintained at 37°C and received periodic perfusion of oxygenated Krebs solution. After a 1hr equilibration period contractile activity was recorded using a Lab-Trax-4 data acquisition system (World Precision Instruments Ltd). Tissue samples were subjected to electrical field stimulation (EFS) for 30s (5Hz; 40V; 0.3ms pulse duration) via platinum electrode loops located at both ends of the tissue sample using a MultiStim System (D330, World Precision Instruments Ltd), in the absence or presence of TTX (1µM). Following contractility analysis, samples were harvested for fixation and sectioning.

Contractility data were collected, stored and analysed via Labscribe 2-software (World Precision Instruments Ltd) including analysis of baseline contractile responses (frequency and amplitude) and response to electrical stimulation (Area under the curve (AUC) for the duration of electrical stimulation). To account for tissue variability all raw responses were normalised to wet tissue weight (g).

Statistics and data presentation

GraphPad Software (GraphPad Prism) was used to generate all graphs and conduct statistical analysis. Data are expressed as mean ± standard error of the mean. The significance of functional data was assessed using a paired Student’s t-test, with p-values <0.05 taken as significant. The individual ex vivo tissue samples examined are reported as “segments”. The ‘n values’ reported refer to the number of patients from which tissue samples were harvested. Inter-patient sample pairwise comparisons, between sham-transplanted and ENS progenitor-transplanted human HSCR tissue, were analysed using paired student’s t-tests. To enable analysis across the transplanted cohort (i.e., comparison between responses across all patient tissue samples) and to account for variability between patient samples (age range of 2 months to 4.5 years; 6 males and 1 female) functional contractility data were normalised within individual patient groups using feature scaling (min-max): Normalised results (Z) were then analysed using an unpaired Student’s t-test, with p-values <0.05 taken as significant.

Additional methods are provided in online supplementary information.

Robust generation of early ENS progenitors from hPSCs

To evaluate the robustness of our previously published vagal NC/ENS differentiation protocol (Fig. 1A) (18), we examined its compatibility with two distinct human embryonic (hESC) lines and one induced pluripotent (iPSC) stem cell line. We employed (i) the female hESC line WA09 (H9), including genetically modified versions containing either a constitutive RFP or GFP reporter under the transcriptional control of SOX10 (21, 24), (ii) the male hESC cell line MasterShef11 (MShef11) (23) and (iii) the female iPSC line (SFCi55ZsG) marked by constitutive expression of the fluorescent tag ZsGreen, thus enabling cell tracking and visualisation in transplantation experiments (18, 22). Our protocol initially produces an anterior NC progenitor (ANC) population through combined stimulation of WNT signalling, TGFβ signalling inhibition and moderate BMP activity for 4 days (27), followed by addition of retinoic acid (RA) for a further 2 days (Fig.1A). Culture of all 3 cell lines under these conditions efficiently induced the consistent expression of the pan-NC/ENS progenitor marker SOX10 (28, 29) by day 6 of differentiation: 60-80% of all cells were marked by protein expression of SOX10 (Fig. 1B). qRT-PCR analysis of the resulting cultures further demonstrated that the upregulation of SOX10 was accompanied by high expression levels of HOX paralogous group (PG) members indicative of a vagal axial identity (HOXB4, HOXB5 and HOXB7) (Fig. 1C). Furthermore, flow cytometry analysis of day 6 cultures derived from H9 background hESCs revealed widespread positivity for the NC/ENS progenitors cell surface markers p75 (NGFR) and CD49d (97.9% and 82.3%, respectively; Fig. 1D). Moreover, we detected minimal presence of ‘contaminating’ central nervous system neurectoderm and non-neural ectoderm cells, marked by the expression of PAX6 and GATA3 proteins respectively (<1% of total cells; Fig. S1A). Together, these data confirm that our protocol efficiently and rapidly gave rise to high yields of vagal NC/early NC progenitor cells from different hPSC lines.

• Download figure
• Open in new tab

Figure 1:

High yields of homogenous vagal NC can be robustly generated from hESCs and iPSCs (A) Schematic representation of the treatment conditions used to generate ENS progenitors in vitro. (B) Flow cytometry-based quantification of SOX10 expressing cells in day 6 cultures derived from the indicated hPSC lines. Error bars = SEM (n=3-7 independent differentiations). (C) qPCR expression analysis of SOX10 and key vagal HOX genes carried out in the same D6 cultures as those shown in B (Error bars = SEM; n=3-7 independent differentiations). (D) Quantification of p75 and CD49d-positive cells in day 6 cultures following immunostaining and image analysis (Error Bars = SEM; n=3). Each biological repeat is represented by a unique symbol. (E) UMAP visualization of 17928 cells and their distribution in three samples (D0, D4 and D6) corresponding to unbiasedly defined clusters. (F) UMAP plots showing expression of a set of selected marker genes. (G) Dot plot visualization of the selected expressed marker genes for each cluster. (H) Selective re-clustering of D4 and D6 samples. (I) RNA velocity analysis of D4 and D6 samples (J) Cytotrace and pseudotime analysis of D4 and D6 samples.

We next sought to map the differentiation trajectory of hPSCs toward vagal NC/early ENS progenitors in more detail. To this end, we analysed differentiating hESCs (H9-RFP) at days (D) 0, 4 and 6 by single-cell RNA sequencing (scRNA-seq). We obtained 17,928 cells, which passed quality control (Fig. S1B-F), that were allocated to three distinct clusters corresponding to the analysed differentiation timepoints (Fig. 1E). The D0 cluster was enriched in pluripotency-associated transcripts (e.g., POU5F1, NANOG), while the vast majority of cells in the D4 cluster exhibited expression of bona fide NC markers such as SOX10, NGFR, LMO4, ETS1, MEF2C and TFAP2A/B (Fig. 1F&G). Expression of these transcripts, most of which also are also present in the developing ENS, was largely maintained in the D6 cluster (Fig. 1F, G). Cells within the latter uniquely exhibited upregulation of markers specifically denoting a vagal NC/early ENS progenitor state with gliogenic characteristics (PLP1, ASCL1, S100β, HOXB2) (30) (Fig. 1F, G). Examination of cell differentiation status was achieved by assigning the developmental stage of cells to clusters connected by trajectory using RNA velocity (31) and also CytoTRACE, a computational framework for predicting differentiation states based on transcriptional diversity (32). This revealed a single stable developmental trajectory from D4 anterior NC cells to D6 vagal NC (Fig. 1H-J). Collectively, these findings confirm that cultures generated using our protocol are predominantly composed of early ENS progenitors; these appear to emerge exclusively upon RA-driven “posteriorisation” of a HOX-negative cranial NC state, in line with our previous observations (33).

To assess the capacity of our D6 early ENS progenitors (H9-RFP) to generate enteric- like neurons in vitro, we subjected them to ENS differentiation conditions, as previously described (18, 19). By D21, progenitors gave rise to complex neuronal networks widely expressing enteric neuronal (PGP9.5, HUC/D, TH, RET, nNOS) and ENS progenitor/enteric glial markers (SOX10) (Fig. 2A-L, S2A). Similar results were obtained using the SFCi55ZsG iPSC line (Fig. S2B). Moreover, we found that D6 ENS progenitors could efficiently generate enteric neurons and glia even after cryopreservation and storage in liquid nitrogen for 6 months despite a decrease in viability (Fig. S2C, D). The functionality of in vitro-derived neurons was further examined via calcium imaging. Ca²⁺ transients in ENS progenitor derivatives were observed following application of either high K⁺, electrical field stimulation or treatment with the nicotinic agonist DMPP (Fig. 2M-O). Based on these data, we conclude that D6 hPSC-derived vagal NC/early ENS progenitors can be directed to produce in vitro cells that exhibit the hallmarks of functional enteric neurons.

• Download figure
• Open in new tab

Figure 2:

Vagal NC can be differentiated in vitro to enteric-like neurons A-L) Representative images of Day 21 enteric neurons immunolabeled with indicated ENS markers following in vitro differentiation of hESCs (H9 background). Scale bar = 100µm. (M- O) Analysis and quantification of hESC-derived (H9 background) neuronal Ca²⁺ response to depolarization with high K⁺ (M), activation with DMPP (N) and electrical stimulation (O). Left: overview of the regions of interest (ROI) employed for analysis. Right: individual line traces of the responding cells, the change in fluorescence (F/F0) is plotted over time (s). The traces are randomly chosen for illustration. The application of high K⁺/DMPP/electrical stimulation is represented by the grey bar in the graph.

hPSC-derived ENS progenitors integrate and form neurons in human HSCR colon following ex vivo transplantation

We next tested the ability of our hPSC-derived ENS progenitors to integrate and form neurons in human HSCR colon following ex vivo transplantation. HSCR patient-derived colonic samples (obtained from surgical discard tissue under informed consent) were transplanted with freshly thawed vials of frozen ENS progenitors generated from either ZsGreen⁺ SFCi55 iPSCs or H9-RFP hESCs and maintained in organotypic culture ex vivo (Fig. 3A-C). At the time of transplantation (D0), ZsGreen⁺ ENS progenitors were readily observed at the transplant site (Fig. 3D). At D21, ZsGreen expression was maintained and observed in all transplanted segments that were imaged prior to further processing (15/15), with donor cells visualised both at the transplant site and migrating through the surrounding tissue (Fig. 3E). Interestingly, donor ENS progenitors appeared to migrate away from the transplant site as individual cells or in “chains” reminiscent of NC cells/ENS progenitors colonising the embryonic murine GI tract in vivo (34). These migrating transplanted cells were often detected in close association with the presumptive endogenous tissue vasculature (Fig. 3F).

• Download figure
• Open in new tab

Figure 3:

ZsGreen-labelled ENS progenitors survive transplantation into Hirschsprung disease surgical discard tissue, spread and differentiate into neurons (A-C) Schematic representation of ex vivo transplantation procedure conducted on patient- derived HSCR colonic tissue. (D-F) Representative stereoscopic images of ZsGreen⁺ donor ENS progenitor cells following ex vivo transplantation into patient derived HSCR tissue at day 0 (D0; D) and D21 (E). High magnification imaging revealed donor cells migrating from the site of transplantation, apparently following blood vessels (F). (G-O) Representative confocal images at D21. ZsGreen⁺ ENS progenitors could be seen clustered around the initial transplant site and migrating from the transplant site in streams. Numerous instances of ZsGreen⁺, TUBB3⁺ cells were detected at the presumptive transplant site (G-I; arrows). Higher magnification revealed ZsGreen⁺, TUBB3⁺ cells with typical neuronal morphology, including the projection of axon-like processes (J-L; hollow arrows). S100+, TUBB3⁺ cells were also detected (M-O; arrows). Scale bars: D,E – 2mm; F – 500µm; G-I – 100µm; J-O – 50µm.

To determine whether ENS progenitors transplanted into the HSCR colonic tissue microenvironment retained the neuroglial potential demonstrated in vitro, immunohistochemistry was performed for the neuronal marker TUBB3 and glial marker S100. At D21, ZsGreen⁺, TUBB3⁺ donor cells were readily detected within HSCR colonic tissue (Fig. 3G-I, arrows) with differentiated donor cells displaying a neuronal morphology, including extension of ZsGreen⁺, TUBB3⁺ axon-like processes (Fig. 3J-L, hollow arrows). S100⁺, TUBB3⁺ donor cells were also detected (Fig. 3M-O, arrows). Collectively, these results indicate that hPSC-derived ENS progenitors can efficiently migrate, integrate and differentiate within human HSCR colonic tissue ex vivo.

Transplantation of hPSC-derived ENS progenitors restores functional responses in HSCR colonic tissue

The inability of aganglionic colonic tissue to contract and relax in a controlled manner is a key feature of HSCR. As transplanted ENS progenitors were found to integrate and differentiate within the explanted HSCR colon, we next examined their ability to alter the functional responses of the host tissue. Organ bath contractility, performed on transplanted HSCR colonic segments and spatially adjacent sham-injected controls, revealed significant improvements in baseline contractility and response to electrical field simulation (Fig. 4A-G). Basal contractile frequency was comparable between ENS progenitor-transplanted segments and sham-transplanted control tissue (10.7 and 16.9 contractions/min, respectively, p=0.2305, Fig. 4B, n=7 patients, pairwise comparison of 21 ENS progenitor-transplanted and sham- transplanted segments). Similarly, we detected no difference in basal normalised contractile amplitude between ENS progenitor-transplanted segments and sham-transplanted control tissue (3.2 and 5.2g, respectively, p=0.0887; Fig. 4C, n=7 patients, pairwise comparison of 21 paired segments). However, complex basal motor contractions were observed more frequently in hPSC-derived ENS progenitor-transplanted tissue compared to sham- transplanted tissue (61.9% and 42.9%, respectively, n=7 patients, 21 paired segments). This resulted in a significant increase in the cumulative magnitude of basal contractions in ENS progenitor-transplanted tissue compared with sham-transplanted tissue (normalised area under the curve (AUC): 5.8 and 2.7g.s, respectively, p=0.015; Fig. 4D, n=7 patients, 21 paired segments). Additionally, ENS progenitor-transplanted tissue elicited a significantly increased contractile response to electrical field stimulation (EFS) when compared to sham-transplanted controls (Fig. 4E-G). Normalised contractile amplitude was significantly increased in ENS progenitor-transplanted tissue compared to sham-transplanted controls (9.6 and 4.6g, respectively, p=0.0062; Fig. 4F, n=7 patients, 19 paired segments). Similarly, we observed a significant increase in contractile magnitude in ENS progenitor-transplanted tissue compared to sham-transplanted controls (normalised AUC: 157.4 and 68.0g.s, respectively, p=0.0141; Fig. 4G, n=7 patients, 19 paired segments). Notably, following application of tetrodotoxin (TTX, 1μM), a potent neurotoxic sodium channel blocker, ENS progenitor-transplanted tissue and sham-transplanted tissue displayed comparable responses to EFS (normalised AUC: 99.5 and 58.6g.s respectively, p=0.0887; Fig. 4H&I, n=7 patients, 17 ENS paired segments). We next sought to examine if the observed improvement between paired tissue samples during EFS could be extrapolated at an inter-patient level, taking into account variability in patient age, gender, disease severity and resected tissue size. To achieve this, we employed feature scaling whereby functional responses were normalised across each patient. This revealed a positive increase in the response to EFS in 68.4% of ENS progenitor- transplanted HSCR segments compared to sham-transplanted controls (Fig. 5A-G, n=7 patients, 19 paired segments). Promisingly, using this analysis, an overall improvement in the response of ENS progenitor-transplanted segments compared to sham-transplanted controls could be detected at a group level (average slope gradient = +0.3573) which was found to be significant when compared to zero (i.e., no effect, p=0.0108; Fig. 5H, n=7 patients). These data suggest an average positive improvement in contractile response to EFS across the entire patient cohort in our study. Collectively, these findings indicate that transplantation of hPSC-derived ENS progenitors triggers neuronally-mediated functional responses in HSCR patient-derived colonic tissue.

• Download figure
• Open in new tab

Figure 4:

Transplantation of ENS progenitors into Hirschsprung disease surgical discard tissue increases both baseline contractility and response to electrical stimulation via neuronal activity (A) Representative traces showing basal contractile activity in HSCR patient derived tissue 21 days after sham (grey) or ENS progenitor-transplantation (green). (B-D) Summary data of baseline contractile frequency (B), amplitude (C) and cumulative Area Under the Curve (AUC; D). (E) Representative contractility traces of sham (grey) or ENS progenitor transplanted (green) tissues in response to electrical field stimulation (EFS; red bar) in control conditions. (F,G) Summary data of amplitude (F) and AUC (G) in response to EFS in control conditions. (H) Representative contractility traces of sham- (grey) or ENS progenitor-transplanted (green) tissues in response to electrical field stimulation (EFS; red bar) in the presence of 1μM TTX. (I) Summary data of AUC in response to EFS in TTX. * = p<0.05 Error bars = SEM.

• Download figure
• Open in new tab

Figure 5:

ENS progenitor transplantation increases the response of Hirschsprung disease surgical discard tissue samples to electrical stimulation To account for patient variability (in terms of age, gender, disease severity and positioning of the sample along the oro-anal axis) samples were normalised by patient using feature scaling (Z = [X – min(X)] / [min(X) – max(X)]). (A-G) Slope analysis for each pair of tissue samples under examination by patient. (H) Summary data showing scaled AUC analysis of ENS progenitor-transplanted samples compared to sham-transplanted. * = p<0.05 Error bars = SEM.