Combinatorial multiomic analysis from a pedigree of Sox10Dom Hirschsprung mice identifies multiple high confidence candidate modifiers of Enteric Nervous System development

Joseph T. Benthal; Justin A. Avila; Jeffrey R. Smith; E. Michelle Southard-Smith

doi:10.1371/journal.pcbi.1014424

Abstract

Hirschsprung disease (HSCR) is characterized by absence of enteric ganglia (aganglionosis) along variable lengths of the distal intestine. This disorder results from deficient colonization of fetal intestine by enteric neural crest-derived cells (ENCDCs). HSCR exhibits complex, multifactorial inheritance with penetrance and severity varying widely even within families. SOX10 is among causal genes that predispose to aganglionosis. Yet, how gene interactions influence severity of HSCR aganglionosis is not understood. Prior mapping of aganglionosis modifiers was achieved in a standard F₁-intercross utilizing the Sox10^Dom HSCR mouse model. Here we deploy a novel strategy of genotyping an extended pedigree pedigree of Sox10^Dom mice on a mixed genetic background. GWAS in this pedigree points to novel aganglionosis modifier intervals with replication and refinement of prior modifier regions. Complementary omics analysis of the developing Enteric Nervous System (ENS) enabled identification of multiple high-priority candidate genes within these modifier intervals based on gene expression, chromatin accessibility, and presence of conserved SOX10 binding motifs. We implemented a prioritization pipeline for ranking potential modifiers that generated candidate lists including several well-known for effects on ENS development as well as multiple novel genes. Among the novel genes, Dach1 ranked as a top priority candidate gene for modifying migration of ENCDCs and thus influencing aganglionosis severity. The results identify genome intervals with intrinsic genes that are logical candidates for modifying Sox10^Dom aganglionosis severity. We also note that several human orthologs to aganglionosis modifier candidate genes are within linkage disequilibrium blocks containing genetic variants associated with human gut motility disorders, which offers opportunity for gaining biological insight into human HSCR severity.

Author summary

Hirschsprung disease is a complex genetic neurodevelopmental disorder that produces absence of neurons in the distal bowel. The length of gut lacking neurons, called “aganglionosis”, in HSCR patients can vary widely even between affected siblings. Multiple genes are Mendelian causative for HSCR, but little is known about the gene interactions responsible for the notable variation in aganglionosis severity. In this study, we use a mouse model of HSCR to identify genomic regions, “modifiers”, associated with length of aganglionosis. Genes active within these regions are then identified in RNA-sequencing and open chromatin data from progenitor cells, which form the enteric nervous system. These omics approaches identify both known and novel genes that can affect enteric neuron development and may underly HSCR severity in patients. Dach1, already known for effects on neuronal progenitor proliferation and migration in other aspects of the nervous system, emerged as a top priority gene. The findings greatly expand the gene network that influences HSCR aganglionosis.

Citation: Benthal JT, Avila JA, Smith JR, Southard-Smith EM (2026) Combinatorial multiomic analysis from a pedigree of Sox10^Dom Hirschsprung mice identifies multiple high confidence candidate modifiers of Enteric Nervous System development. PLoS Comput Biol 22(7): e1014424. https://doi.org/10.1371/journal.pcbi.1014424

Editor: Mark Ziemann, Burnet Institute, AUSTRALIA

Received: November 3, 2025; Accepted: June 9, 2026; Published: July 6, 2026

Copyright: © 2026 Benthal et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All Code, fragment files, and snATAC-seq Seurat Object are shared at Zenodo: https://doi.org/10.5281/zenodo.20671852 SnATAC-seq data is accessible at GEO under accession GSE298406 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE298406).

Funding: This work was funded by NIH award R01 DK60047 and R01 DK127178 to EMS², with support for JTB on T32-HD007502 and NIDDK 1F31DK137637-01. JAA was supported by T32-HD007502, HHMI Gilliam Fellowship GT11505 (Awarded to EMS² and JAA) and R01-DK127178-03S1. A flow sort shared resource scholarship fund was provided by the Vanderbilt University Medical Center Digestive Disease Research Center (P30DK058404). NIH (https://www.nih.gov/) and HHMI (https://www.hhmi.org/) had no role in study design, data collection or analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

The ENS is essential for normal gastrointestinal (GI) motility and analyses of mouse models have identified genes that are essential for ENS development. ENS neurons and glia that make up the myenteric and submucosal ganglia along the entire length of the intestine are formed by ENCDCs that colonize the fetal gut during development [41]. Gut colonization begins at 9.5 days post coitus (dpc) in mice as neural crest cells invade the foregut and then migrate along the full length of the gut by 14.5dpc [41]. The wavefront of migrating enteric neural crest-derived cells (ENCDCs) leaves behind cells that differentiate into neurons and glia. Disruption of initial ENCDC migration can produce GI motility disorders such as HSCR or Waardenburg-Shah syndrome while disrupted differentiation of ENCDCs is thought to contribute to chronic intestinal pseudo-obstruction [54,67]. The characteristic phenotype for HSCR is aganglionosis—lack of enteric ganglia—of the distal colon at varying lengths, caused by deficient colonization of ENCDCs [41]. Multiple mouse models mimic human HSCR, including Sox10^Dom [42]. The Sox10^Dom HSCR model also recapitulates the variability of aganglionosis length that remains a poorly understood characteristic of human HSCR [72]. Mouse models on controlled genetic backgrounds offer opportunity to identify contributing genes to the variability of HSCR aganglionosis.

Genetic mapping studies in HSCR mouse models to find genes contributing to aganglionosis severity have had limited success due to large genomic intervals from standard crosses and lack of ENCDC expression data for candidate gene prioritization. Initial work to define how genetic background affects aganglionosis was performed by Cantrell et al. who compared lengths of intestinal aganglionosis in congenic C3HeB/FeJ and C57BL/6J.Sox10^Dom strains (2004). C57BL/6J.Sox10^Dom mice exhibit notably more severe aganglionosis than C3HeB/FeJ.Sox10^Dom mice [9]. Owens et al. mapped five broad genomic modifier intervals (ranging from 8 to 30 cM) associated with aganglionosis via genome-wide linkage scan of Sox10^Dom/+ F₂ progeny derived from C3HeB/FeJ and C57BL/6J congenic lines (2005). Further evidence that gene interactions can alter extent of aganglionosis was reported by Maka et al. [53], who found that loss of Sox8 increases extent of aganglionosis in crosses with Sox10^lacZ knockout mice (2005). Multiple gene defects contributing to the severity of HSCR aganglionosis were also demonstrated for Ret^+/-;Ednrb^S/S mutants [56]. Despite the genetic evidence of genome regions that modify aganglionosis phenotype, the ability to identify potential candidate genes within these regions that produce these effects has been limited by lack of gene expression and chromatin accessibility data for the developing fetal gut and specifically the migrating ENCDCs that form mature neurons and glia of the ENS.

There has been recent success in bulk and single cell sequencing of ENS progenitor populations that offers new avenues to prioritize candidate genes within modifier intervals. Stavely and colleagues recently utilized bulk RNA-seq of ENCDCs at the migrating wavefront and the cells behind the wavefront to identify differentially regulated genes at 11.5dpc [74]. In addition, Zhao and colleagues performed scRNA-seq on entire mouse gut over the time course of ENCDC migration (2022). This dataset allows detection of genes that are expressed either within the ENCDCs or in the surrounding gut environment, an important aspect, since interactions between progenitors and the fetal gut mesenchyme are known to influence extent of migration. Such datasets offer unique opportunities to prioritize candidate genes that might be functional among cells that contribute to aganglionosis phenotype.

Prior genetics analyses of human HSCR have focused primarily on Mendelian causes. Relatively limited efforts have pursued the characteristic variability of HSCR disease penetrance and severity. Multiple case-control GWAS have identified variants associated with human HSCR [22,37,76,17]. These studies identified common genetic variation at RET, NRG1, and SEMA3C/D loci that are significantly associated with HSCR. A study investigating additional phenotypic anomalies that can accompany HSCR implicated copy number variation at the SOX2, MAPK10, ZFHX1B, PHOX2B, and SEMA3A loci; severity of aganglionosis in these patients was not considered [30]. To date, six human studies identified variants or genes that influence the penetrance of HSCR [4,64,13,22,30,46,78]. To our knowledge, no human study has undertaken quantitative analysis to identify genetic variation modifying the extent of HSCR aganglionosis length, which is responsible for clinical severity. In contrast, mouse models of HSCR enable precise measure of aganglionosis length on controlled genetic backgrounds to dissect genetics underlying severity.

Given the ability to control genetic background in mice and combine transcriptomic resources to prioritize candidate genes, we sought to refine Sox10^Dom/+ modifier intervals and utilize omics datasets to identify candidate genes that could influence the length of aganglonosis. We investigated a 10-generation pedigree of 830 heterozygous affected Sox10^Dom mice that yielded improved resolution, capitalizing upon the increased number of meiotic recombination events relative to standard F₂ intercrosses. The pedigree analysis accomplished this by reintroducing the greater-effect alleles of the B6 genetic background at each generation [9]. The Sox10^Dom line was maintained by crosses of affected heterozygous males with wild type (WT) F₁ B6C3Fe-a/a females. Pedigree individuals were genotyped with a linkage mapping SNP set appropriate for its genetic resolution. Genome-wide association in this pedigree improved the resolution of known loci, and identified novel loci that modify aganglionosis length, including some with sex-biased effects. Modifiers intervals were still very large containing hundreds to thousands of genes. Thus, candidate genes from each modifier interval were prioritized based on gene expression among ENCDCs and fetal gut during ENCDC migration, chromatin accessibility in fetal enteric neuronal progenitors, and conserved SOX10 transcription factor binding motifs. From the many genes within modifier intervals, 19 met multiple genetic and omics criteria. Among these we observed both novel and known genes involved in ENS development including Ednrb, Nrg1, Col1a1, and Phox2b. In the pedigree, Phox2b and Mcm3 appeared to be sex-biased modifiers. A top candidate gene identified through the omics analyses was Dach1, which has not previously been associated with aganglionosis, yet is expressed in migrating ENCDC, is flanked by SOX10 conserved binding motifs, and resides within accessible chromatin in ENS neuronal progenitors. The modifier genes identified from this analysis are likely to influence ENCDC migration or differentiation in the developing gut.

Results

Genome-wide SNP analysis identifies Sox10^Dom aganglionosis modifier intervals

While maintaining the B6C3Fe-a/a.Sox10^Dom/+ strain, we observed notable phenotype variability among Sox10^Dom/+ pups produced by iterative outcrosses to B6C3Fe-a/a wildtype females with some pups dying from severe aganglionosis at postnatal day (P)7 while others survived beyond a year of age. To understand genetic variation associated with severity of aganglionosis, we quantitatively assessed aganglionosis for 830 P7-10 B6C3Fe-a/a.Sox10^Dom/+ pups. Sox10^Dom/+ pups were distinguished from their WT littermates via characteristic white ventral spotting, white feet, and confirmed by direct genotyping of the mutation [9,63]. Extent of aganglionosis was assessed by whole-mount acetylcholinesterase staining collected over multiple pedigree generations (Fig 1A and 1B and S1 Table) using established methods [9,63]. Because Sox10^Dom pedigree maintenance relied on breeding Sox10^Dom males to WT F₁ B6C3Fe-a/a female mice, alleles for greater severity/aganglionosis were reintroduced at each generation. Reduced severity/protective alleles were selected for by mating Sox10^Dom males that survive to sexual maturity. A standard mouse linkage mapping panel, including 1449 SNPs of which 876 were informative in the pedigree, was used to genotype each Sox10^Dom/+ pup (S2 Table). No significant difference between males and females were detected in length of intestine or length of aganglionosis (length of intestine t-test p = 0.084; length of aganglionosis Wilcoxon p = 0.64; Fig 1B–1E). Rather than being normally distributed, we observed that the aganglionosis phenotype in Sox10^Dom/+ pups was zero-inflated due to animals that lacked detectable aganglionosis (Fig 1D), which replicates prior reports of aganglionosis distribution in this Hirschsprung model [9,63].

Download:

Fig 1. Phenotype distribution of Sox10^Dom pedigree mice exhibits no significant separation of phenotype by sex.

(A) Schematic of the breeding strategy used for the Sox10^Dom mouse extended pedigree. (B) Schematic illustrating quantitation of aganglionosis extent as a proportion of total gut length affected by aganglionosis in Sox10^Dom pups. Collection of gut from the pyloric sphincter to the anus, followed by Acetylcholinesterase wholemount staining, with measurement of total gut length and the region of aganglionosis allowed calculation of the percent gut length affected by aganglionosis. (C,D) Length of intestine split by sex displayed on histograms and a box plot shows normal distribution and no significant difference between males and females. (E,F) Length of intestinal aganglionosis split by sex on histograms and a box plot shows a non-normal, zero-skewed distribution with no significant difference between males and females. Panel B was partially generated in Biorender (Southard-Smith, M. (2026) https://BioRender.com/7osgcym).

https://doi.org/10.1371/journal.pcbi.1014424.g001

To identify genomic intervals associated with aganglionic length in the B6C3Fea.Sox10^Dom pedigree, we tested association between SNPs distinguishing genomic intervals originating from the parental C3HeB/FeJLe-a/a (C3Fea) or C57BL/6J (B6) strains with the variable aganglionosis phenotype. Our approach employed Genome-wide Efficient Mixed Model Analysis (GEMMA), which mitigates false positives that could otherwise arise due to relatedness of inbred mouse strains [94]. Measured agangionic length was evaluated as a proportion of total intestinal length to avoid potential artifacts due to pup size. Sex was included as a covariate (443 females, 387 males). We prioritized consideration of associated loci that A) were significant after multiple testing correction, B) had a logarithm of the odds (LOD) score of ≥3, or C) replicated observations of an independent F₁-intercross study [63]. Even so, we also discuss additional nominally significant loci due the pleiotropic nature of the Sox10^Dom mutation as shown by Owens et al. [63]. In the main GEMMA analysis, an interval on chromosome 5 (LOD score = 3.9) was significantly associated with aganglionic length after false discovery (FDR) rate correction (Figs 2A and S1A and S3 Table). The greatest significance was observed at rs13478309 (chr5:6714120 of mm10, p-Wald = 8.310292e-05); the B6 allele G was associated with increased aganglionosis. This SNP is ~ 48 kb from the transcription start site of Phox2b (mm10; [66]). This finding is consistent with the well-known role of Phox2b in ENS development [65]. While detection of significant association near Phox2b serves as an internal control, we sought to identify other novel candidate modifier genes with the potential to influence aganglionosis extent. Other nominally significant loci were present on multiple chromosomes illustrated in Fig 2A. Four of these are novel by comparison to the prior F₁-intercross analysis [63]. One of these nominally significant loci was detected on chromosome 15 and does not overlap with Sox10 (Fig 2A and S3 Table).

Download:

Fig 2. GEMMA genome-wide scans using phenotype variation and a binary phenotype of aganglionosis in the Sox10^Dom pedigree identify genome regions associated with aganglionosis.

(A) Manhattan plot visualizing association analysis of quantitative aganglionic length with inclusion of all chromosomes. (B) Analysis of quantitative aganglionic length as in A, with exclusion of chromosomes 5 (harboring Phox2B), 15 (harboring Sox10), and X (with potential to accentuate alternative loci). (C) Alternative analysis comparing Sox10^Dom mutation carriers that are unaffected to carriers with any affected aganglionic length (binary analysis). Alternating blue and black traces represent the plotted LOD values for distinct chromosomes with corresponding chromosome number indicated below each trace on the x-axis. Blue lines on Manhattan plots indicate marginal significance (unadjusted P-Wald<0.05), dotted red lines indicate Bonferroni-adjusted significance (adjusted P-Wald<0.05), and the solid red line indicates false discovery rate significance (adjusted P-Wald<0.05). Gene positions shown are approximate and are candidate modifiers derived from omics evidence summarized in Table 6 after the prioritization process and are not meant to indicate genome wide significant SNPs at these genes.

https://doi.org/10.1371/journal.pcbi.1014424.g002

To uncover additive effects of variants on other chromosomes and to account for the zero-inflated aganglionosis phenotype distribution, we evaluated two alternative models. First, given Phox2b’s known role in ENS development, the known issues that can arise when performing analyses on the X chromosome, and the previously established “leave one chromosome out” method, we opted to perform our analysis excluding the SNPs on chromosomes 5, X, and 15, where Phox2B and Sox10 reside, respectively [65,8,47,85]. This approach can augment the statistical signal of loci on alternative chromosomes. Using this approach, we observed additional nominally significant modifier loci on chromosomes 7 and 10 (Figs 2B and S1B and S4 Table). Second, many Sox10^Dom mice in our pedigree do not have detectable aganglionosis, which could obscure SNPs associated with length versus the presence of aganglionosis (Fig 1C). To account for this, we converted aganglionosis measurements to a pseudo case-control binary phenotype. We then compared Sox10^Dom mice who exhibited any aganglionosis to mutation carriers without any detectable aganglionosis. Nominally significant genetic association with absence or presence of aganglionosis was observed at the previously detected loci as well as additional regions (chromosomes 1, 3, 5, 7, 9, 13, and 15) (Figs 2C and S1C and S5 Table). Third, to assess SNP effects on only the variance in aganglionosis without effects of those with no detectable deficits, we performed GEMMA analysis using only mice with measurable aganglionosis (S2 Fig and S6–S8 Tables). This approach produced an approximately 30 percent decrease of animal numbers available for the analysis, and therefore we do not pursue this aspect of the analysis further.

These pedigree-based association scan results expanded upon the previously published F₁-intercross loci modifying Sox10^Dom aganglionosis [63]. The previously observed modifiers on chromosomes 5, 8, 11, and 14 were again detected (Table 1; [63]). Additionally, the pedigree analysis detected multiple novel modifiers on chromosomes 1, 2, 3, 13, and 19 that were nominally significant (Table 1). The chromosome 3 loci observed in this pedigree-based analysis were distinct from that of the intercross (Table 1; [63]). Directions of effect (DOE) for loci replicating in the prior F₂-intercross study were concordant (S3A, S3B and S3F–S3H Fig, S1 and S2 Tables, see [63] Fig 4). However, several of the loci appeared to have sex-biased effects on aganglionosis in the pedigree, which had not previously been observed in the F₁-intercross (S3A–S3E and S3I–S3N Fig, S1 and S2 Tables). Given that Hirschsprung disease exhibits a sex bias with males more frequently affected than females [7,18], the potential sex-bias effect on aganglionosis was investigated further in sex-stratified analysis.

Download:

Table 1. Sox10^Dom aganglionosis modifiers compared to Owens et al.

https://doi.org/10.1371/journal.pcbi.1014424.t001

Download:

Fig 3. Sex-specific GEMMA genome-wide association for aganglionosis modifiers in the Sox10^Dom pedigree.

Manhattan plots visualizing GEMMA genome-wide scan results of male- (A) and female-specific (B) association analysis of quantitative aganglionosis length. Manhattan plots visualizing GEMMA genome-wide scan results of male- (C) and female-specific (D) analysis comparing Sox10^Dom mice that are affected by any length of aganglionosis to those mice that are unaffected (binary analysis). Blue lines on Manhattan plots indicate marginal significance (unadjusted P-Wald<0.05) and dotted red lines indicate Bonferroni-adjusted significance (adjusted P-Wald<0.05). Gene positions shown are approximate and are candidate modifiers derived from omics evidence summarized in Table 6 after the prioritization process and are not meant to indicate genome wide significant SNPs at these genes.

https://doi.org/10.1371/journal.pcbi.1014424.g003

Assessing potential sex-bias among modifiers of Sox10^Dom aganglionosis

Because some of the SNP variants in our pedigree associations exhibited marginally significant sex-biased DOE (S3A–S3E and S3I–S3N Fig, see methods for shorthand terms used for each genome-wide analysis, S1 and S2 Tables), we conducted sex-stratified genome-wide analyses for both quantitative and binary aganglionosis phenotypes using GEMMA [94]. The male-specific genome-wide scan replicated the Phox2b on chromosome 5, which was again genome-wide significant (Figs 3A, S1D and S9 Table). This locus was notably attenuated in the female-specific scan. Several other loci were nominally significant in the male-specific scan, including on chromosome 14 near Ednrb, a critical gene for ENS development (Fig 3A and S9 Table). The female-specific quantitative scan only yielded nominally significant loci that did not remain significant upon multiple testing correction (Figs 3B, S1E and S10 Table).

The sex-specific binary association analysis only identified nominally significant loci. Male-specific binary phenotype analysis loci mirrored those of the quantitative analysis, but were less significant (Figs 3C, S1G and S1F and S11 Table). Female-specific binary phenotype analysis also yielded only nominally significant variants (Fig 3D and S12 Table). Altogether, these findings suggest that some Sox10^Dom modifiers of aganglionosis severity may exert sex-biased effects (Tables 2 and 3).

Download:

Table 2. Sex-biased Sox10^Dom aganglionosis modifier locations.

https://doi.org/10.1371/journal.pcbi.1014424.t002

Download:

Table 3. Top 2 GEMMA peak SNPs per Sox10^Dom aganglionosis modifier interval definition.

https://doi.org/10.1371/journal.pcbi.1014424.t003

In aggregate, the genomic intervals identified from the separate GEMMA association analyses in the Sox10^Dom pedigree (all chromosomes, leave one out, and sex-specific) shared overlap in genome location and localized to chromosomes 1, 5, 8, 11, and 14 (Fig 4A and S13 Table). The span of these intervals in the genome is quite large with some genome intervals exceeding 10Mb. To identify relevant candidate genes within modifier intervals, we implemented a multipronged omics approach to aid in filtering for genes most likely to be affecting ENS development (Fig 4B).

Prioritization of candidate genes based on expression in the fetal mouse intestine

To identify relevant candidate genes potentially underlying the variability of Sox10^Dom aganglionosis, we first assessed which genes within modifier intervals are expressed in ENS precursor cells and the fetal gut environment during ENCDC migratory stages. Genes within modifier intervals (listed in S13 Table) were identified from the UCSC Table Browser by inputting positions of modifier intervals as elaborated further in Methods. This extraction identified a total of 6216 unique genes (Table 4) after GEMMA runs were consolidated. We first leveraged scRNA-seq data from total mouse fetal gut collected over development from 9.5 to 15.5dpc that profiled ENCDCs and surrounding gut mesenchyme (Fig 5A–5C; [92]). After reprocessing this scRNA-seq data, we filtered for genes within modifier intervals that have log-normalized pseudobulk expression greater than 1 (Tables 4 and S14). Differential gene expression analysis for the expressed genes by age and cell type determined which genes are expressed in specific cell types and timepoints in the developing mouse gut (Tables 4 and S15). We detected differential expression of Phox2b and Ednrb (Figs 5C and S4 and S15 Table), genes that are highly expressed in ENCDCs at all timepoints and which are well known genes that influence ENS development [9,15,25,73,61]. Identification of genes expressed in fetal gut reduced the total number of candidate modifier genes to 732 (Table 4).

Download:

Table 4. Number of genes within each modifier interval subsetting those that are expressed and differentially expressed in scRNA-seq of the fetal gut.

https://doi.org/10.1371/journal.pcbi.1014424.t004

We similarly evaluated expression of genes located nearest the top two female associated SNPs from the sex-specific GWAS. The top female associated SNP had a LOD score of >3, while the second most significant female SNP had a higher LOD score than other modifiers that were replicated from the F₂ study [63]. The closest genes to these SNPs include Mcm3, Il17f, and Isl1, of which Mcm3 and Isl1 are expressed in ENCDCs (Figs 5C and S4).

Identification of ENCDC expressed genes allowed us to assess whether pairs of genes (ligands – receptors) that could participate in cell-cell communication during migratory stages are present in modifier intervals. Communication between ENCDCs and the surrounding gut mesenchyme is an essential aspect for normal colonization of the developing fetal gut [62]. To evaluate the potential of candidate genes in modifier intervals to either signal from the gut mesenchyme to ENCDCs or affect communication from ENCDCs to the surrounding gut microenvironment during bowel colonization, we applied CellChat. CellChat is a computational tool that quantifies communication between two cell groups using a reference database of over 3300 ligand receptor interactions and reliance on random permutation to estimate ligand-receptor communication among cell groups in scRNA-seq datasets [31,32]. Utilizing the wildtype scRNA-seq dataset [92], CellChat analysis identified predicted ligand-receptor communication between ENCDCs and other cell types in the wildtype scRNA-seq dataset for five genes present in modifier intervals (S5A Fig and S16 Table). This analysis points to biological relevance of Ppia, Col1a1, Lgals9, Ednrb, and Nrg1 for communication between ENCDCs and other gut cells during migration along the bowel. These genes are within nominally significant modifier intervals that are not significant after multiple testing correction. However, these genes do fall within nominally significant intervals that were detected in the prior F₁-intercross study [63]. CellChat-identified modifier interval genes and the associated binding partner show at least one of each are expressed in ENCDCs in the Zhou scRNA-seq dataset (S5B Fig). Several of the genes present within modifier intervals that CellChat identifies include genes that are either associated with ENS development (Ednrb) or are known risk genes for HSCR like Nrg1 [29]. Our analysis found that while single genes from a ligand-receptor pair were present within modifier intervals no instances were observed where both genes for a ligand and receptor pair resided in modifier intervals. Taken together, the predicted cell communication pathways detected for these modifier interval genes adds another prioritization filter based on the potential to influence ENCDC-fetal gut environment communication during bowel colonization.

Differential expression of modifier interval genes in the migrating wavefront of enteric neural crest-derived cells further prioritizes candidate genes

Colonization of the fetal intestine by ENCDCs relies heavily on migratory capacity of the leading progenitor cells at the advancing edge of the migrating wavefront [86,88,71,89,91]. At 11.5dpc, ENCDCs are transitioning around and across the midgut fold to colonize the hindgut. Prior bulk-RNA seq profiling of leading-edge cells compared to the residual cells further back in this migrating population found these leading wavefront cells exhibit distinct transcriptional features [74]. Because migration is an essential process for hindgut colonization and these leading-edge progenitors exhibit a promigratory expression profile that may be relevant for modifying the migration defects of Sox10^Dom mutants, we filtered for genes within modifier intervals that were differentially expressed in the ENCDC wavefront versus lagging cells (Fig 6A). We overlapped those differentially expressed genes with the genes in our modifier intervals to identify any genes that exhibited differential expression either up- or down-regulated in the leading-edge cells versus cells further back in the migrating population. Of the 732 candidate modifier genes, five genes were downregulated (St18, Hoxb5, Prph, Slc10a4, and Uchl1) and 9 genes were upregulated (Myh10, Col1a1, Alkbh5, Pdgfra, Ptprg, Sh2b3, Dach1, Hs3st3b1, and Tgfbi) in the migrating wavefront (Fig 6B and S17 Table). Several of these genes including Hoxb5, Pdgfra, Ptprg, Alkbh5, Dach1, and Tgfbi have been implicated in various aspects of neural crest development [20,51,81,95,38,59,82,48].

To validate expression in a separate dataset and determine if these differentially expressed genes at the migrating ENCDC wavefront are expressed in specific groups of cells within the ENCDC population, we subset the whole gut developmental scRNA-seq dataset to focus just on ENCDCs (Fig 6B and 6C). Expression of Sox10 and Phox2a identify progenitors and neuronal cells, respectively (Fig 6D). We expected modifier candidate genes upregulated in the migrating wavefront of the ENCDCs might be expressed in progenitor cells which are highly migratory, while downregulated genes would be expressed in more mature neuronal ENCDC cells (Fig 6B and 6D). Downregulated genes in the ENCDC wavefront like Uchl1 and Prph are most highly expressed in more mature, neuronal populations (Fig 6E—Lagging). However, wavefront upregulated genes are expressed in either an intermediate population (Col1a1, Tgfbi, Pdgfra) which also expresses Ascl1, a known marked of enteric neuroblasts, or increase in expression over time (Ptprg, Dach1; Fig 6E —Migrating Wavefront). Mapping expression for these candidate genes in the scRNA-seq across early and later migratory stages provides insight into ENCDC cell type and temporal-specific expression.

Evolutionarily conserved SOX10 binding sites within Sox10^Dom modifier intervals overlap or are near genes differentially expressed in the migrating wavefront

To further distill relevant candidate genes within Sox10^Dom aganglionosis modifier intervals potentially regulated by SOX10 DNA binding, we leveraged a dataset of dimeric SOX10 binding motifs conserved across chick, mouse, and human genomes [24]. We located 24 conserved SOX10 binding motifs within modifier intervals (Tables 5 and S18). We located the closest gene to each conserved SOX10 binding motif, yielding 17 unique genes, some of which are nearby multiple SOX10 binding motifs (Table 5). Most of these binding motifs were intronic to the annotated gene with only one motif positioned outside the nearest gene (~1.3kb away from Sox2; Table 5). Each 1Mb region surrounding the modifier interval-contained SOX10 binding motifs was manually examined on the mm10 genome to determine whether there were nearby genes that were already identified via other data modalities or already known to affect ENS development [66]. This added 6 genes to this list (Table 5). Several of these genes have already been identified in this work by alternate means, including Mcm3, St18, Pdgfra, Ptprg, and Dach1 (Table 5). Also identified are several genes that have two or more intronic or close conserved SOX10 binding motifs including Dach1, Bcas3 (close to Tbx2), Tenm2, Ranbp17 (close to Tlx3), and Mecom with four intronic motifs (Table 5). Lastly, other genes not previously identified through other omics were near or overlap with conserved SOX10 binding motifs within modifier intervals, including Bai3 (Adgrb3), Cyp7b1, Nlg1, Slit2, Ppargc1a, Prdm8 (close to Antxr2), Fam204a, and Sox2 (Table 5). Several of these genes are expressed in the ENCDCs (Figs 6 and S6; [92]). Altogether, these 24 conserved SOX10 binding motifs within modifier intervals highlight potential targets of SOX10 binding that could influence the aganglionosis phenotype caused by the Sox10^Dom mutation.

Download:

Table 5. Conserved SOX10 TF binding sites within Sox10^Dom aganglionosis modifier intervals.

https://doi.org/10.1371/journal.pcbi.1014424.t005

Download:

Table 6. Cumulative omics evidence supporting priority candidate genes within Sox10^Dom aganglionosis modifier intervals.

https://doi.org/10.1371/journal.pcbi.1014424.t006

Chromatin accessibility within enteric neuronal progenitors highlights putative regulatory regions within aganglionosis modifier intervals

Modifiers can consist of variants in cis-regulatory elements that affect gene expression [69,34]. To assay open chromatin, that might harbor sites of gene regulatory activity within Sox10^Dom modifier intervals, we utilized single nucleus assay for transposase-accessible chromatin with sequencing (snATAC-seq) within mouse fetal ENS cells (S7A Fig). We captured ENCDCs at 16.5dpc based on expression of a Phox2b H2B-CFP transgene shown to mark glial, neuronal progenitor, and neuronal cells [11,55]. By 16.5dpc, the gut has been fully colonized, yet enteric neurogenesis is ongoing. We processed 13431 Phox2b H2B-CFP+ nuclei and performed unsupervised clustering to obtain 17 clusters (Fig 7A). We then utilized a pre-existing scRNA-seq dataset flow-sorted from WT 15.5dpc ENS to annotate snATAC-seq nuclei via estimated gene expression (see Supplementary Methods; S7B Fig). This approach assigned six main groups (progenitors, neuroblast1 and 2, and neuronal branches A, B, and C; Figs 7B and S7C). Differential chromatin accessibility analysis was used to localize 71521 differentially accessible chromatin regions (DARs) for each of these six main groups (S19 Table). Genomic annotations did not deviate across groups except for Neuroblast2, which we postulate could be due to fewer nuclei assigned as compared to the other groups (n = 102 Neuroblast2 nuclei/ 13431 total nuclei; S7D Fig, top panel). We filtered these DAR to those within Sox10^Dom modifier intervals, yielding 9151 unique regions (S20 Table). Modifier interval DAR regions in the Neuroblast2 nuclei are limited to promoter and distal intergenic categories, and this might be, again, due to fewer nuclei assigned to this group (S7D Fig, bottom panel).

Download:

Fig 4. Modifier interval positions and gene level prioritization strategy.

(A) Schematic displaying overlap and relative position of modifier intervals from each of the GEMMA association studies for aganglionosis severity. Thick gray lines represent the length of each chromosome in basepairs. Relative position and size of modifier intervals derived from each GEMMA analyses that have greater than 0 genomic scores are displayed as colored vertical lines to the left of each chromosome bar. Colors represent assigned modifier interval scores based on the following criteria to minimize false positives: 1 point assigned to peak SNPs of modifier intervals significant after multiple testing correction; 1 point assigned to peak SNPs of modifier intervals with LOD score of ≥3; and 1.5 points assigned to intervals replicated from the prior F₁-intercross study. (B) Schematic illustrating the process of filtering for high priority candidate genes within modifier intervals. Top section depicts significance of SNPs in a GEMMA genome scan with detection of modifier intervals based on genetic scoring criteria. Second level depicts evaluation of genes for expression in relevant tissue (enteric nervous system in this study). Third level depicts overlay of differential chromatin from snATAC-seq. Fourth level depicts relative position for presence or enrichment of TF binding motifs. This strategy was applied to filter for those genes within the modifier intervals that are logical high priority candidates for further study.

https://doi.org/10.1371/journal.pcbi.1014424.g004

Download:

Fig 5. Differential gene expression in the developing gut within aganglionosis modifier intervals filters a first subset of candidate genes.

(A) Pipeline for prioritizing Sox10^Dom aganglionosis modifier interval genes by differential gene expression. (B) Diagram depicting migration of enteric neural crest-derived cells down the length of the fetal mouse gut. (C) UMAP of reprocessed scRNA-seq data from [92] split by developmental timepoint and colored by their cell type definitions. The red asterisk indicates the location of the enteric neural crest-derived cells (ENCDCs). (D) Dot plot showing expression of differentially expressed genes near the top two LOD peaks per GEMMA GWAS (out of the nearest 10 genes per top 2 LOD peaks, those that are differentially expressed). Cell types have been consolidated to those that are known to be relevant for migrating enteric neural crest-derived cells and those with shared names (numbered) are combined to one identifier. Dots to the left of each gene correspond to which GEMMA genome-wide scan each gene originates.

https://doi.org/10.1371/journal.pcbi.1014424.g005

Download:

Fig 6. Overlap between differentially expressed genes at the enteric neural crest-derived migrating wavefront and genes within modifier intervals identifies migratory candidate genes.

(A) Venn diagram visualizing pipeline for prioritizing Sox10^Dom aganglionosis modifier interval genes by differential gene expression at the migrating wavefront. Red asterisk indicates overlap between modifier intervals and upregulated genes in the migrating wavefront. (B) Volcano plot of all differentially expressed genes of the migrating wavefront enteric neural crest-derived cells at 11.5dpc versus the lagging/stationary cells. Red and blue dots indicate genes upregulated and downregulated in the wavefront, respectively. Genes labeled are those within modifier intervals. (C) UMAP of scRNA-seq from [92] enteric neural crest-derived cells (ENCDCs) based on source data annotations isolated from the main dataset and split by timepoint. (D) Expression of Sox10, labels progenitors (cyan) while Elavl4 and Phox2a expression labels neuronal cells (salmon). Asterisks mark transitional neuroblast populations that express Aschl and Sox10. (E) Feature plots of [92] ENCDCs colored by expression of genes differentially downregulated (left) and genes differentially upregulated (right) in the migrating wavefront. UMAPs visualizing expression genes differentially downregulated in the wavefront are not split by time point while UMAPs for genes differentially upregulated are split by time. Dots next to gene names indicate from which modifier interval set each gene falls within.

https://doi.org/10.1371/journal.pcbi.1014424.g006

Download:

Fig 7. Whole gut 16.5dpc Phox2b H2B-CFP + snATAC-seq-derived transcription factor binding motif enrichment in differentially accessible chromatin identifies modifier interval-contained enriched TF motifs.

(A) UMAP of snATAC-seq nuclei colored by unsupervised clusters. (B) UMAP displaying the result of label transfer of cell types from scRNA-seq to snATAC-seq. (C) Scatterplot of enriched TF binding motifs from the modifier interval-contained differentially accessible chromatin from progenitor and neuroblast clusters filtered by expression of their corresponding genes in the fetal gut ENS. The scatterplot points are colored by which modifier interval set each TF motif is enriched, with duplicates resulting from the analysis being performed by specific interval (“assay”). (C’) Scatterplot of enriched TF binding motifs from panel C plotted only for the BothSexAllChrom Chr5 modifier interval. The percentage difference between the number of TF motifs in the DAR and the background open chromatin (“percent.diff”) is indicated by the size of the point.

https://doi.org/10.1371/journal.pcbi.1014424.g007

To determine whether conserved SOX10 binding sites from Gopinath and colleagues localized to DAR regions within modifier intervals, we examined the overlap between these datasets [24]. Two of the conserved SOX10 binding motifs overlapped with modifier interval DAR. These were intronic to Ranbp17 (close to Tlx3; chr11:33400347–33400362) and Dach1 (chr14:97891443–97891462; Tables 5 and S21).

Transcription factor (TF) binding to cis-regulatory elements is important for the transcription of genes, and prominent candidate genes identified in this study are TFs including Phox2b, that have established roles in enteric neurogenesis [65]. We performed TF binding motif enrichment on modifier interval-contained DAR from progenitor and neuroblast groups to evaluate potential regulators of ENS development in Sox10^Dom modifier intervals. We focused on data from the progenitors and neuroblasts groups because these cell states enable collection of sufficient numbers of nuclei to pursue accessibility studies that otherwise would have been extremely difficult with the low numbers of migrating ENCDCs at early stages of gut colonization. Our approach examined both open chromatin in progenitors and neuroblasts as well as regions that are closed in differentiating enteric neuronal cells. This resulted in 797 enriched TF binding motifs in modifier interval DAR compared to the rest of accessible chromatin (S22 Table). We filtered these TF binding motifs to select for those whose encoding genes are expressed in the ENCDC scRNA-seq data, revealing 367 motifs (S23 Table). This enrichment analysis coupled with expression identifies the following TFs from modifier intervals: Phox2b, Rbpj, Hoxb2/3/4/5/9, Sox2, Pou5f1, Tlx2, Tbx3, Pou3f1, and Mecom (Fig 7C and S23 Table). Several of these TFs participate in ENS development [6,77,49,58,83,2].

Prioritization of top aganglionosis modifier candidate genes across data modalities

To rank candidate genes within Sox10^Dom aganglionosis modifier intervals (S13 Table) we assessed both genetic and multiomics data in a prioritization pipeline (Fig 8). In this process, we applied two scoring metrics: a modifier prioritization score and an omics evidence score. For our modifier interval prioritization score, we assigned 1 to 3.5 points to each modifier interval and genes within these intervals inherited this score based on the following criteria to minimize false positives: 1) peak SNPs of modifier intervals significant after multiple testing correction were assigned one point, 2) peak SNPs of modifier intervals with LOD score of ≥3 were assigned one point, and 3) intervals replicated from the prior F₁-intercross study were assigned 1.5 points (Tables 6 and S24; [63]). We defined the preponderance of evidence omics score over a range of 0–5 corresponding to the number of separate lines of evidence implicating candidate genes. This incorporated single cell differential expression, estimated cell communication (CellChat), differential expression at the ENCDC wavefront, snATAC-seq-derived DAR regions in ENCDCs, ENCDC-expressed genes whose TF binding motifs are enriched within DAR regions in modifier intervals, and proximity to evolutionarily conserved SOX10 binding motifs (Fig 8A and S25 Table). Genes from each modifier interval were assessed independently through our prioritization pipeline to identify relevant candidates with potential to influence aganglionosis (Fig 8A and S25 Table). Thirty genes emerged that appear in three or more of our omics analyses. Of these, 11 with modifier interval prioritization scores of 0, indicating lack of strong genetic evidence, including Mecom and Sox2, were excluded from further analysis. Nineteen genes exhibited modifier prioritization scores >0 and multiple lines of omics evidence (Fig 8B and Table 6). One top candidate modifier gene, Dach1, exhibits five lives of omics evidence. Three additional genes have four supporting lines of evidence (Col1a1, Pdgfra, and Tbx2; Tables 6 and S25). Several genes that emerged from the pipeline are already known to be involved in ENS development and HSCR including Phox2b, Ednrb, and Nrg1, which effectively serve as internal controls.

Download:

Fig 8. Analysis pipeline used to identify candidate genes from Sox10^Dom pedigree-derived aganglionosis modifier intervals.

(A) Outline for candidate gene analysis pipeline following GEMMA genome-wide scans to identify genomic regions that modify the aganglionosis phenotype in Sox10^Dom mice, then use bulk and single cell sequencing strategies to identify candidate causal genes. After identifying candidate genes, these are then used to find gene body variants in C3Fe as compared to mm39 (C57BL/6J), then compared to human phenotype-genotype association data. DGE: differential gene expression. (B) Bar chart plotting the relative genetic modifier priority score and the total omics evidence score for 19 candidate genes with modifier prioritization scores >0 and multiple lines of omics evidence.

https://doi.org/10.1371/journal.pcbi.1014424.g008

Given the transcription of several candidate modifier genes within ENCDCs and the presence of conserved SOX10 binding motifs nearby several of these candidates, we evaluated co-expression of genes with a priority score of four or greater with Sox10 (Table 6). We undertook this analysis in the subset ENCDC populations from the [92] data set during migratory stages (13.5 dpc) when the hindgut is being colonized. This effort identified co-transcription of Sox10 with Phox2b, Ednrb, Dach1, and Tbx2 in large numbers of ENCDCs (Fig 9). Similarly, Ednrb is expressed in progenitors and early neurons coincident with Sox10 transcription. Tbx2 and Sox10 are also co-expressed among large numbers of progenitors and transitional neuroblasts. While Dach1 has the highest omics score (5), the gene is most frequently transcribed in transitional neuroblasts and developing neurons (S8 Fig). This distribution results in the most frequent co-transcription of Sox10-Dach1 occuring in transitional states between progenitors and neurons (Fig 9). In contrast, Pdgfra and Col1a1 exhibited infrequent co-transcription with Sox10 that was sparse among progenitors and observed primarily in a single transitional population. These patterns of co-expression suggest the potential for direct interaction between Sox10 and several candidate modifer genes in ENCDC cells.

Download:

Fig 9. Co-expression of candidate modifier genes with Sox10 during migration stages highlights potential for interaction.

Potential gene interaction within the same cells were examined in feature plots from the [92] data evaluated for co-incident transcription with Sox10. (A) All ENCDCs from Zhao et al. colored to show progenitor and neuronal states (left) and normalized expression of Sox10 at 13.5dpc. Normalized expression not limited by “max.cutoff=0.5” is shown. (B) UMAPs showing co-expression (yellow) at 13.5dpc of top candidate genes (green) with an omics score of 4+ and well established markers of ENS subtypes (Dach1, Phox2b, Ednrb, Elavl4, Pdgfra, Tbx2, and Col1a1) relative to Sox10 transcription in Zhao et. al. 2022 data. Feature plots show normalized expression limited by “max.cutoff=0.5” to highlight cells that may lowly express, but still express candidate genes. (B’) Normalized expression of candidate genes corresponding to those in B without a max cutoff of 0.5. (C) Co-expression of neuronal marker Elavl4 (red) and candidate gene Dach1 (green). Expression shown uses a max cutoff of 0.5 normalized expression to highlight cells that may lowly express, but still express Dach1. (C’) Normalized expression of Elavl4 without a max cutoff of 0.5.

https://doi.org/10.1371/journal.pcbi.1014424.g009

After comparison across data modalities to identify candidate genes with the highest relevance scores, we probed each candidate modifier gene sequence for C3Fe variants, as compared to B6, with predicted high impact on expression [57]. Of these candidate genes, Phox2b and Col1a1 contained C3Fe variants with predicted high functional impact (S26 Table). Phox2b contained a C3Fe variant (T > TC insertion, chr5:67255011, transcript ENSMUST00000174251.2 UCSC Genome Browser, mm39) located at a splice junction boundary that could affect splicing efficiency and overall gene expression levels [66]. Col1a1 also contained a C3Fe variant predicted to impact splicing (S26 Table). Identification of sequence variants between the C3Fe and B6 strains for Phox2b and Col1a1 candidate genes adds further evidence to the likelihood that strain differences may exacerbate severity of Sox10^Dom aganglionosis.

Candidate modifiers of aganglionosis related to human GI phenotypes

To relate our Sox10^Dom aganglionosis modifier candidate genes to human GI and HSCR disease loci, we assessed whether these genes’ human orthologs were in linkage disequilibrium (LD) with SNPs identified as significant from four human HSCR GWAS and GWAS of stool frequency [22,37,76,17,5]. Two separate HSCR GWAS summary statistics had significant SNPs that were in LD with genes identified in this study, including Phox2b, Col1a1, Rbpj, Nrg1, Ednrb, Hoxb2/5/13, and Uchl1 (Tables 7 and S27–S29; [22,76]). The GWAS of stool frequency had significant SNPs that were in LD for Col1a1, Rbpj, Nrg1, and Elf1 (S30 Table; [5]). While not in an LD block with significant SNPs, DACH1 is ~ 2.6Mb away from an associated SNP in the stool frequency GWAS [5].

Download:

Table 7. Human genome-wide significant SNPs that are in LD with candidate gene orthologs identified in Sox10^Dom modifier analyses.

https://doi.org/10.1371/journal.pcbi.1014424.t007

To further relate our candidate Sox10^Dom aganglionosis modifier genes to other human GI, neural, and neural crest phenotypes, we took an exploratory approach to probe the publicly available exome-based PheWAS sever Genebass. We evaluated whether variants in human orthologs of our candidate mouse genes might be associated with ENS-related traits. We utilized a text/term-based search among the top 20 phenotypes that were returned for each of the orthologs (Supplemental Methods; [35,23]). Of the orthologs examined in PheWAS, each had phenotypes based on partial string searches that captured terms based on neural crest, nervous system, or GI (Supplemental Methods). Among these is DACH1 with four different phenotypes relating to GI surgery (S31 Table). Interestingly, among the terms found for DACH1 is “Anterior resection of rectum and exteriorisation of bowel”, a process often used to treat HSCR.

Discussion

While HSCR patient studies have largely focused on causal genes, animal models, such as Sox10^Dom, are advantageous for defining the role of genetic background on aganglionosis severity. Prior Sox10^Dom modifier mapping studies identified very large genomic intervals and relatively few genes emerged as modifiers of aganglionosis severity and penetrance [63]. Omics from relevant cell and tissue types can help prioritize large numbers of genes in modifier intervals to select candidates with highest potential to influence a trait. In this study, we conducted genome-wide analyses on an extended pedigree of the Sox10^Dom mouse model of HSCR to refine aganglionosis modifier intervals and identify candidate modifier genes within those intervals. Several nominally significant modifier intervals are associated in a sex-biased manner. Utilizing multiple omics datasets including scRNA-seq, bulk RNA-seq, snATAC-seq, and TF binding motifs, we analyzed genes within modifier intervals to prioritize those candidate genes with the greatest relevance for aganglionosis. This strategy revealed multiple highly relevant candidate modifier genes that are expressed in developing gut, and which either have known roles in ENS development or exert effects on other aspects of cell migration. Our analysis confirms the interval near Phox2b is a candidate modifier of Sox10^Dom aganglionosis and identifies multiple novel candidate genes likely to influence aganglionosis severity. Among these, Dach1 stands out as the highest priority candidate based on omics data and prior evidence of this gene affecting cell migration [12,84,38].

Quantitative trait mapping in mice has blossomed with the availability of mouse genetic resources like the collaborative cross [50,14]. However, challenges remain for identifying disease modifiers that rely on inclusion of a mutant allele. We took advantage of a multi-generational Sox10^Dom mouse pedigree for GEMMA genome-wide analyses anticipating that the iterative crosses of Sox10^Dom males with B6C3Fe-a/a wildtype females would greatly narrow the associated genome regions. This approach compliments our previous F₁-intercross strategy to map Sox10^Dom aganglionosis modifiers by taking advantage of additional recombination in pedigree generations. Aganglionosis modifiers identified in this study replicate intervals located in a prior F₁-intercross including a genome-wide significant modifier on chromosome 5 and marginally significant modifiers on chromosomes 3 (shifted), 8, 11, and 14 [63]. We also detect novel marginally significant modifiers, including some that are sex-biased (Tables 1 and 2). However, the modifier intervals we detected were still quite large, some exceeding 10Mb. The large size of modifier intervals produced in aggregate thousands of candidate genes. This result necessitated that we develop strategies to prioritize and filter for candidate genes that are most likely to be influencing severity of aganglionosis. This multistep process, moving from genome interval to a limited number of high priority candidate genes, relies upon accumulation of evidence to pinpoint those genes that are most likely to be influencing phenotype and is a prerequisite for further functional testing of individual genes [1].

Our dual score approach—confidence of modifiers and lines of evidence through omics analysis—allowed us to prioritize candidate genes from large genomic intervals. Among the total 6216 annotated genes in modifier intervals, 683 were expressed in either ENCDCs or the surrounding gut mesenchyme through which these cells migrate [92]. Subsequent gene filtering identified 5 cell communication candidate genes, 26 candidates near conserved SOX10 binding motifs, 9151 DAR regions enriched for motifs of 367 ENCDC-expressed TFs from snATAC-seq, and 14 genes differentially expressed at the ENCDC migrating wavefront. Thirty candidate genes with cumulative evidence aggregated from multiple modalities emerged as putative modifiers of Sox10^Dom aganglionosis, however only 19 of these also exhibited significant modifier scores >0 (Table 6). Of the 19 candidate genes, Phox2b, Rbpj, and Uchl1 were located in modifier intervals that 1) had peak SNPs that were significant past the threshold for multiple comparisons, 2) had peak SNPs that had LOD scores ≥3, or 3) overlapped with intervals derived from the prior F₁-intercross study (Tables 6, S24 and S25; [63]). Pdgfra and Antxr2 met two of these criteria (Tables 6, S24 and S25). Dach1 was within a replicated modifier interval (Tables 6, S24 and S25; [63]). We then examined these candidate genes for variants of predicted high impact in the C3Fe genome versus C57BL/6J. This analysis revealed potential function-altering C3Fe variants for Phox2b and Col1a1 that could affect splicing [57].

Our analysis was challenging due to the zero-inflated, non-normal distribution of the phenotype resulting from approximately thirty percent of animals in the pedigree that exhibited no detectable aganglionosis. This zero-inflated phenotype required that, in addition to treating the phenotype as-is, we implement a binary phenotype conversion approach. We considered several alternatives including a rank-based transformation. Transformation of zero-inflated data would break the transformation assumption by collapsing all the zeros into a single rank or small rank set, which would ignore the structure of the trait distribution and could lead to spurious associations. We also considered analysis approaches based on converting the trait to a categorical classification (e.g., none, short, intermediate, long), which would also produce loss of biological variation like the rank-based transformation. Additionally, categorical classification would require imposing cutoffs for each category, which could bias the results. We were unable to locate an appropriate tool that could be utilized specifically for mouse pedigrees and would account for the inherent relatedness of inbred strains as GEMMA does while also accounting for the zero-inflated, non-normal distribution of the aganglionosis phenotype [60]. Therefore, we proceeded with GEMMA analysis despite the complication due to zero-inflation and 1) compensated with treating the trait as a binary phenotype in addition to quantitative and 2) assessed how the resulting associations were significant or marginally significant relative to the prior [63] study. This approach allowed us to identify modifier intervals that were replicated across studies. The presence of zero-inflated phenotypes remains a major challenge in quantitative traits analysis that would benefit from future methods/software package development capable of also dealing with strain relatedness and pedigree structures.

The mouse aganglionosis modifier candidate genes identified here may have relevance as modifiers of human GI disease. Therefore, we compared the mouse Sox10^Dom aganglionosis candidate genes to significant SNPs identified from HSCR and stool frequency GWAS [22,37,76,17,5]. Of the candidate genes, nine of these genes were in LD blocks with significant HSCR GWAS SNPs and four were in LD blocks with significant stool frequency GWAS SNPs (Table 7). In addition, we use exome-based PheWAS study results from the UKBiobank (Genebass) to find GI, neuronal, and neural crest phenotypes associated with variants in our candidate genes [35,23]. These comparisons suggest that our candidate genes in Sox10^Dom aganglionosis modifier intervals are likely relevant for human HSCR phenotype variation or other GI phenotypes.

In this study, we generated and mined a novel snATAC-seq dataset of 16.5dpc fetal ENS collected from Phox2b H2B-CFP mice [11, 55]. We used this dataset to prioritize modifier interval candidate genes based upon differentially accessible chromatin and enriched TF binding motifs. Because this dataset contains chromatin accessibility for populations that include progenitors, neuroblasts, and maturing neuronal linages, it has the potential to be highly informative for researchers interested in regulatory regions that control neuronal diversification and maturation in the fetal ENS. This dataset can be used further for analyses of TF binding motif accessibility changes over neurogenesis for each neuronal trajectory relative to analogous scRNA-seq datasets. Future work combining the snATAC-seq data with single cell gene expression profiles at 16.5dpc has the potential to identify putative target genes of predicted cis-regulatory elements that could emerge from combined snATAC-seq/scRNA-seq analytics. These may be relevant to HSCR or other gastrointestinal motility disorders that arise from disrupted ENS development.

Several of the high priority candidate modifier genes, identified here are co-transcribed with Sox10, which offers a means for direct interaction between genes. Prior genetic studies suggested the potential for direct interaction between Sox10 and Phox2b as well as Ednrb [63,9]. Our analysis reveals co-transcription of Phox2b with Sox10 in progenitors and enteric glia as well as in early neuronal transitional states and is consistent with prior studies reporting Phox2b transcription in these cell types, [11,Young et al., 1998]. Similarly, co-transcription of Ednrb occurs with Sox10 in ENS progenitors and early developing neurons. The novel candidate modifiers identified by our omics analysis including Dach1, Col1a1, Tbx2, and Pdgfra are expressed in ENCDCs and exhibit intriguing patterns of expression across ENCDC populations. Dach1 expression is observed in progenitors and in early transitioning neurons, where high levels of Sox10 transcript are prevalent. Moreover, recent identification of SOX10 protein persistence in developing neurons further extends the potential timeline for interaction between SOX10 and the Dach1 locus [2]. Co-expression of Dach1 in Sox10 + cells during critical migratory stages (13.5dpc) when the hindgut is being colonized has potential relevance for extent of aganglionosis. TBX2 has been reported in postnatal neurons of the ENS [83], although to our knowledge there have not been reports of TBX2 protein or transcription in fetal ENS other than current scRNA-Seq data sets. Our analysis identifies high levels of Tbx2 in numerous ENCDC progenitors and early transitioning neuronal states with frequent co-transcription with Sox10 (Fig 9). Tbx2 exhibits several of the same features as Dach1 with differential expression over ENS development, conserved SOX10 TFBMs, and differential chromatin accessibility between developing neurons and glia although it does not have the same wavefront differential expression that Dach1 does. In contrast, Pdgfra and Col1a1 exhibit very sparse co-expression with Sox10 that is mostly limited to one of the several transitional neuronal states. Whether there is direct interaction between Sox10 and these candidate modifier genes identified in our analysis awaits further experimental study.

There is tremendous temporal and spatial complexity in ENCDC colonization that hinges upon migration, proliferation, and differentiation. Given that HSCR aganglionosis is a consequence of failed colonization, we hypothesize these top modifier candidates regulate different aspects of ENCDC migration (Fig 10). There is substantial prior evidence of Dach1 influencing migration in neural crest, embryonic fibroblasts, and breast cancer [12,84,38]. In Xenopus, Dach1 is specifically associated with early neural crest migration [38]. Similarly, Col1a1 in zebrafish is implicated downstream of retinoic acid (RA) signaling, which has been shown to be important for ENS development and ENCDC migration [68,44,45,79,21]. Pdgfra affects both migration and survival in oligodendrocytes as regulated by Sox9 and Sox10 while PDGFR signaling has been shown to affect nitrergic enteric neuron specification in vitro [19,52]. In zebrafish, tbx2a/b CRISPR/Cas9 targeting caused reduced ENS cell density, suggesting function for differentiation or proliferation of ENCDCs [40]. Two genes that fell off our omics candidate list due to lack of strong genetic evidence include Sox2 and Mecom. However, these two genes exhibit four lines of omics evidence. Mecom affects neural crest-derived chondrocyte differentiation, orientation, and polarity in mouse and zebrafish models [70]. Sox2 has been shown to contribute to maintenance of progenitor states of neuronal progenitors and has been identified as a modifier of human HSCR via copy number variation [30,80]. These genes may exert small additive effects on Sox10^Dom aganglionosis despite their failure to reach significance in our genetic analysis. The depth of prior analyses assist in understanding how these different genes might contribute to the aganglionosis phenotype in Sox10^Dom mice.

Download:

Fig 10. Models illustrating developmental processes by which candidate Sox10^Dom modifier interval genes may alter ENCDC migration.

Schematic diagrams displaying migration of enteric neural crest-derived cells (ENCDCs), differentiation of ENCDCs towards more mature ENS fates, and proliferation of ENCDCs at the migrating wavefront. In blue are the typical functions, and in red are cells and functions that are the hypothesized results of perturbation of the candidate genes. Candidate genes hypothesized to be involved in each function are in smaller boxes in the upper right of each function box, with superscript sex symbols for those genes that are within sex-specific modifier intervals.

https://doi.org/10.1371/journal.pcbi.1014424.g010

There are several limitations to our study. The omics data analyzed in this study are sourced from the fetal gut and therefore does not include genes whose expression during initial migration of vagal neural crest from neural tube to the foregut, which also could influence Sox10^Dom aganglionosis. Ideally future analysis will capture expression profiles of early migrating vagal neural crest and deeply sequence migration stages to capture lowly expressed transcription factors. We utilized CellChat, which estimates ligand-receptor interactions, and misses any sort of transcription factor-chromatin interaction. There is also inherent stochasticity of ENCDC development even in congenic inbred backgrounds which can contribute to variation in aganglionosis [33]. Resolution for detecting modifiers on chrX is limited both due to the cross structure and the analysis tools we applied. GEMMA does not include algorithms that can account for copy number or mosaicism that are needed for analysis of chrX SNPs [94]. We expect since only male Sox10^Dom mice are bred to propagate the pedigree, and WT B6.C3Fe F₁ female mice are introduced into the pedigree at each generation, there is minimal recombination on chrX. Even with the low density of SNPs in the genotyping panel and the high variability of aganglionosis phenotype, we replicated the chr5 interval containing Phox2b, which surpassed genome-wide significance after multiple testing correction. Other loci were nominally significant, including a novel locus on chr1 with a LOD score > 3 (rs6404446, chr1:20784226; Fig 4A and Table 3), and several other loci that replicated prior modifier regions [63]. Candidate genes in these regions were supported by our omics analysis. Lastly, future work validating the effects of individual modifier genes identified here, particularly Dach1 and Tbx2, will be required. Analysis of individual modifier gene loss on ENCDP migration or the extent of aganglionosis will require complementation test crosses with Sox10^Dom mutants or knockdown studies in homogenous genetic backgrounds.

In conclusion, we demonstrate the ability to improve resolution of Sox10^Dom/+ aganglionosis modifier intervals from mapping in an extended pedigree. Combinatorial omics analysis identifies a high priority list of modifier genes that meet multiple genetic and genomic metrics. Detection of Phox2b as a well-known gene in ENS development confirms the pipeline approach. Dach1 emerges as a novel modifier of Sox10^Dom aganglionosis. Cumulative evidence for Dach1 includes genetic interval replication across studies, prior reports of Dach1 mutation on neural crest migration, and omics evidence of SOX10 binding sites in accessible chromatin around this gene. Our analysis enriches the gene network that impacts ENS development and reveals important genes for future variant analyses in human HSCR and other GI motility disorders.

Methods

Ethics statement

All animal experimental protocols were approved by the Institutional Animal Care and Use Committee (IACUC) at Vanderbilt University Medical Center.

Mouse husbandry

All mice were housed in a modified barrier facility on a 14-hour on, 10-hour off-light cycle in high-density caging (Lab Products Inc., #10025) with breeders and all Sox10^Dom mice on (LabDiet 5LJ5) and water ad libitum. The B6C3Fe-a/a.Sox10^Dom line originated at Jackson Lab (Jackson Stock 000290) and was maintained by crosses to B6C3Fe-a/a WT female mice generated on site by crosses of C3FeLe.B6-a/J (Stock # 000198) females crossed to C57BL/6J (Stock # 000664) males. Pedigree offspring over 10 generations were euthanized at postnatal days 7–10 for GI tract collection that was dissected intact from stomach to anus.

The Tg(Phox2b-HIST2H2BE/Cerulean)1Sout mice (MGI: 5013571), Phox2b H2B-CFP, was maintained by crosses with female C3FeB6F1 [11]. Timed matings between female C3FeB6F1 and male Phox2b H2B-CFP mice were conducted to produce fetal intestines at 16.5dpc used for snATAC-seq.

Acetylcholinesterase histochemistry

Whole-mount acetylcholinesterase enzyme histochemistry of all Sox10^Dom mutant offspring from the pedigree collected for this study was performed as described [16,9]. Briefly, intestines were dissected from postnatal 7–10 day old pups extending from the pyloric sphincter to the anus and gently flushed with 1X PBS to remove luminal contents. Tissues were fixed in 4% paraformaldehyde between 90–120 minutes on ice then rinsed twice before transfer to saturated sodium sulfate solution for an overnight incubation at 4^oC. Tissues were stained with a solution of 0.2mM Ethopropazine, 4mM Acetylthiocholine Iodide, 10mM Glycine, 3mM Cupric Sulfate pentahydrate, 6mM Sodium Acetate Trihydrate to reveal patterning of the enteric ganglia within the gut wall as previously illustrated [9]. Total gut length and extent of aganglionosis was measured on a ruler to 1mm increments using a Leica MZ12.5 microscope to view ganglia distribution. All samples were scored independently by two researchers. The percent gut length aganglionic was calculated as a percentage of gut length lacking enteric ganglia over the total gut length to account for differences in body size due to differences in nutrition status of Sox10^Dom pups.

Mouse genotyping

All pedigree offspring were genotyped for the Sox10^Dom allele using established methods [9].

Genome-wide SNP genotyping data was generated by the Center for Inherited Disease Research at Johns Hopkins University using the Illumina Mouse Linkage Panel (GoldenGate GS0006826-OPA), on a Sentrix array (1449 SNPs). Array data was analyzed with de novo clustering in GenomeStudio 2.0 with 1369 SNPs meeting quality metrics for analysis of which 876 were informative between C3HeBFeJLe-a/a and C57BL6/J.

Analysis of Sox10^Dom extended pedigree using Genome-wide Efficient Mixed Model Analysis (GEMMA)

GEMMA was used generally on the population and in sex-specific analyses to generate relatedness matrices [94]. We then performed SNP-based aganglionosis association analyses using GEMMA. Analysis code and further details can be found in Supplemental Materials and in supplementary files located on our Zenodo repository.

Modifier intervals for downstream candidate gene analysis

Sox10^Dom aganglionosis modifier intervals were defined as each interval of adjacent significant SNPs (p-value not adjusted for multiple tests) +/- 0.5 megabase (Mb). The + /-0.5Mb window was chosen to account for associations that may not contain the causal SNP(s), gene(s), or locus (loci) and is consistent with many long-range regulatory elements that are within a 1Mb interval of regulated genes (Fig 7). We utilize shorthand for each GEMMA association run. Shorthand for each GEMMA analysis and modifier interval set used as input for our analysis pipeline (Fig 7) is as follows: sex-regressed quantitative aganglionosis phenotype with zeroes included, BothSexAllChrom; sex-regressed quantitative aganglionosis phenotype with zeroes included with removal of SNPs from chromosomes 5, 15, and X, BothSexNo15X5; sex-regressed with pseudo case-control “binary” phenotype, BinBothSexAllChrom; male- and female-specific quantitative aganglionosis phenotype with zeroes included, Male and Female, respectively; male- and female-specific pseudo case-control “binary” phenotype, BinMale and BinFemale, respectively. Genes within modifier intervals were found using the UCSC Table Browser Tool [36] based on the above modifier interval genomic regions as input (S13 Table). Briefly, BED files were generated using the chromosome number, start, and end of the modifier intervals. These positions were input into the UCSC Table Browser, selecting for gene symbol output. Output gene lists were downloaded in.csv format and assessed for expression, then differential expression in scRNA-seq as described in the section below and our Supplemental Methods.

Reprocessing of whole gut 9.5-15.5 days post coitus scRNA-seq data

ScRNA-seq data were downloaded from the Gene Expression Omnibus at accession GSE186525 [92]. Seurat and SCTransform version 2 were utilized for quality control, clustering, and integration across age and sample [26,28,10]. Seurat’s FindAllMarkers function was used for differential gene expression analysis. Analysis code and further details for scRNA-seq processing can be found in Supplemental materials and on our Zenodo repository.

Enteric neural crest-derived cell wavefront differential gene expression (DGE) overlapping with modifier intervals

We downloaded differential gene expression data from Stavely et al. on the Gene Expression Omnibus at accession GSE217757. This file was imported into R and filtered by multiple testing correction-adjusted p-value of 0.05 and a log2FoldChange of <-0.5 and >0.5. These genes were then filtered by those that were within modifier intervals using the same strategy as performed in the methods for imprinted genes (see Supplemental information section “Determining whether modifier intervals contained imprinted genes”). This was done for each GEMMA association-based modifier intervals, both non-sex specific and sex specific.

SOX10 conserved binding motifs within modifier intervals

SOX10 binding motifs from Gopinath et al. were downloaded and ported into UCSC Genome Browser’s LiftOver tool [27,24]. Motif regions were processed sequentially from hg18 to mm10 in LiftOver. The binding motifs were then assessed to determine which overlapped with modifier intervals from all GEMMA associations using R code.

CellChat analysis of cell-cell communication in ENCDCs

We imported the reprocessed developing fetal gut scRNA-seq dataset/Seurat object from [92] into CellChat version 2.1.2 in R [31,32]. Per timepoint excluding 15.5dpc (when the gut is fully colonized with ENCDCs), we estimated ligand-receptor interactions between ENCDCs (“Neural crest”) and other cell types across the signaling genes expressed. We filtered then filtered to ligand-receptor interactions that passed interaction probability thresholds of 0.25 and were statistically significant. Of those that passed this filter, we subset those that were within modifier intervals.

Generation and processing of fetal ENS single nucleus assay for transposase-accessible chromatin-sequencing (snATAC-seq) data

Isolation of nuclei from fetal mouse tissue.

At 16.5 days post coitus (dpc), C3FeB6.Phox2b-H2BCFP+ fetal intestine (from stomach to anus) was used as source tissue to isolate ENS cells relying on cold-active protease dissociation, as previously described [2]. Tissue was pooled and cells were isolated using fluorescence-activated cell sorting (FACS) to select for viable cells that excluded 7-aminoactinomycin D+ stain. Differential sorting for CFP+ high (neuronal cells) and separately CFP + low (progenitors, glial cells) was performed as reported [55]. Nuclei were produced from post-FACS cell suspensions using the 10x Genomics protocol CG000209_Rev D. Nuclei counts were obtained via hemocytometer or Countess Automated Cell Counter (Thermofisher) followed by encapsulation on the 10x Genomics Chromium Next GEM single cell platform. Libraries for snATAC-seq were produced from two replicates with the first CFP-high and -low sorted populations produced from 6 pooled fetal guts and the second produced from 12 pooled fetal guts. Libraries were sequenced on an Illumina NovaSeq 6000 using an S4 flow cell with custom read lengths to support ATAC-seq samples. Sequencing depth targeted greater than 70,000 paired-end reads per nucleus.

Analysis of snATAC-seq data.

Sequence FASTQs were processed with CellRanger ATAC pipeline version 1.2.0 by the Vanderbilt Technologies for advanced genomics (VANTAGE) shared resource [93]. Sequences were aligned to mm10. One CFP+ high-intensity sample was excluded based on CellRanger output showing an overabundance of nuclei with approximately 10 times fewer fragments per cell than other replicates (see supplementary methods). The remaining single CFP + -high and CFP + -low intensity samples were imported into R using Seurat and Signac. Default Signac processing was used to generate Seurat objects per replicate, including LSI dimensionality reduction [26,75; See Supplementary methods]. Harmony was used to integrate the two low-intensity replicates. Seurat and Signac’s LabelTransfer was used to identify cell types according to scRNA-seq data from similar cell types at 15.5dpc (GSE262898), resulting in snATAC-seq counterpart clusters to the identities seen in scRNA-seq [2]. This approach included both enteric neuronal and glial lineages as genes marking both cell types are expressed at this stage [87]. Differential chromatin accessibility analysis was performed using MACS2-derived peaks and Seurat’s FindAllMarkers function for both unsupervised and LabelTransfer clusters [90]. The resulting differential chromatin accessibility datasets were filtered by Bonferroni-adjusted p-value < 0.05 and average log2 Fold Change > 1.5 and <-1.5. To find overlap for differentially accessible chromatin loci and Sox10^Dom aganglionosis modifier intervals, we first added 25 base pairs on either side of each differentially accessible locus, then assessed for overlap via the same method described in Supplementary Methods.

TFBM enrichment for differentially accessible loci within modifier intervals was performed using Signac’s built-in function “FindMotifs”. Briefly, the R package TFBSTools function “getMatrixSet” was used in combination with the JASPAR 2024 R package to get position frequency matrices (PFM) from each validated (core) vertebrate TFs [76,Rauluseviciute et al. 2024]. Since DACH1 binding motif’s PFM was not in the validated core, this PFM was manually added from the unvalidated section of JASPAR2024 [Rauluseviciute et al. 2024]. Signac’s “AddMotifs” function was used to add the motifs to the snATAC-seq Seurat object. Differentially accessible peaks for progenitor and neuroblast clusters, which were then split by each modifier interval per interval set (in other words, for each modifier interval set, the differentially accessible chromatin for each modifier interval were split, such as “Male_Chrom5_Interval1”). These were then input into the “FindMotifs” function to find enriched TF binding motifs per modifier interval. These were then filtered by TFs that are expressed via Seurat’s “AverageExpression” function greater than 0.1 (grouped by timepoint) in the [92] subset ENCDC (“Neural Crest”) scRNA-seq dataset [92]. The enriched TF binding motifs in modifier intervals were further filtered to those TFs whose genes were found through one of the other previous modalities (scRNA-seq-based differential gene expression, differentially expressed at the migrating wavefront of ENCDCs, conserved SOX10 binding motifs).

Sox10^Dom aganglionosis modifier interval candidate gene prioritization pipeline from mouse datasets

Omics analyses used in prioritization were merged into a final table that was filtered to only candidate genes that were in three or more lines of omics evidence (Tables 6 and S24). A score of 0-3.5 was assigned to each candidate gene based on whether the modifier loci in which the candidate resides were 1) significant after multiple testing correction assigned 1 point, 2) LOD score of ≥3 assigned one point, and 3) overlapped with the modifier intervals from the F₁-intercross study assigned 1.5 points (S25 Table; [63]).⁷ We do this to consider whether modifiers in which candidate genes reside could be false-positive associations.

Filtering Sox10^Dom aganglionosis modifier interval candidate genes for those with intron or exon variants with predicted high impact

C3HeB/FeJ variants (VCF; mm39) for 30 candidate genes were used as input for the online Ensembl Variant Effect Predictor (VEP) for Mus musculus and were filtered for high predicted impact [27,43,57].

Overlap of Sox10^Dom aganglionosis modifier interval candidate genes with human HSCR and stool frequency GWAS summary statistics

Summary statistics or analogous tables were downloaded for stool frequency and four different HSCR GWAS [22,37,76,17,5]. If all SNPs were available, false discovery rate (FDR) p-value adjustment was used, and SNPs were filtered based on FDR significance [22,76,5]. Each summary statistics table was verified to use SNP coordinates on hg19. SNPs were tested for LD with candidate genes via precalculated LD blocks [3]. These genes were then overlapped with their mouse orthologs from Table 7.

Extraction of relevant phenotypes from Genebass-sourced Sox10^Dom aganglionosis modifier interval candidate gene-based PheWAS

Exome-based PheWAS from Genebass for variants in each candidate gene were filtered to the top 20 associated phenotypes per candidate [35,23]. These phenotypes were then filtered by common GI and nervous system terms as described in Supplementary Materials.

Supporting information

S1 Fig. QQ Plots of Non-Sex-specific and Sex-specific GEMMA genome-wide scans on the Sox10^Dom mice population.

QQ plots visualizing GEMMA genome-wide scan results with inclusion of all chromosomes (A) and exclusion of chromosomes 5, 15, and X (B) for the total quantitative aganglionosis percentage phenotype. (C) QQ plot visualizing results from a GEMMA genome-wide scan in which a binary phenotype—either the individual mouse has or does not have aganglionosis measured—was used. QQ plots are also shown visualizing GEMMA genome-wide scan results of female- (D) and male-specific (E) runs for the total quantitative percentage aganglionosis phenotype. QQ plots visualizing GEMMA genome-wide scan results of female- (F) and male-specific (G) runs for the binary phenotype—either the individual mouse has or does not have aganglionosis measured.

https://doi.org/10.1371/journal.pcbi.1014424.s001

(TIFF)

S2 Fig. GEMMA genome-wide scans using only those Sox10^Dom mice that exhibit detectable aganglionosis.

(A) Manhattan and QQ plots visualizing association analysis of quantitative aganglionic length, excluding unaffected Sox10^Dom mutation carriers. (B) Analysis as in A but only including male Sox10^Dom mice. (C) Analysis as in A but only including female Sox10^Dom mice.

https://doi.org/10.1371/journal.pcbi.1014424.s002

(TIFF)

S3 Fig. Stratification of allele-specific directions of effect by sex reveal sex effects.

A-E,I-N The top 2 most significant SNPs per sex-specific genome-wide scans comparing the percent aganglionosis across individuals by genotype split by sex. F Chromosome 3’s top hit in the first GEMMA run shows a larger effect on males than females with rs13477043, and larger effects in females for genotype combinations for rs3667738, chromosome 8 (G). Similar differences in males and females are observed for rs13481145, the top hit for chromosome 11 (H). Each box plot shows distribution of percentage of aganglionosis and comparative statistics for each allele combination split by sex. Color of dots indicate the GEMMA runs with which each SNP is the peak associated SNP. See methods for the shorthand key for GEMMA association runs, which are the labels used here. Overall p: Kruskal-Wallis; internal p: Wilcoxon test. *, p < 0.05; **, p < 0.005; ***, p < 0.0005; ****, p < 0.00005; ns, not significant. A ‡ beside a SNP indicates significance of a SNP past multiple testing correction.

https://doi.org/10.1371/journal.pcbi.1014424.s003

(TIFF)

S4 Fig. Expression of genes in fetal gut near top associated SNP of nominally sex-biased modifier intervals across developmental time.

Feature plots display expression via presence and intensity of purple split by developmental timepoint in the reprocessed [92] scRNA-seq dataset for candidate genes nearest to the top 2 associated SNPs or the most likely candidate gene based on known ENS developmental biology (Ednrb).

https://doi.org/10.1371/journal.pcbi.1014424.s004

(TIFF)

S5 Fig. CellChat identifies modifier interval candidates via estimated cell signaling with neural crest cells.

(A) Dot plot split by time point visualizing probability of activity of significant signaling pathways (y-axis) by cell type (x-axis). Shape of the dot indicates which modifier interval each signaling gene is within. Signaling genes within modifier intervals are underlined. Significance is represented by the size of the dot. Cell types on the left (orange bracket) represent signaling into neural crest cells, while types on the right (blue bracket) represent signaling out of neural crest cells to those cell types. (B) Dot plot showing expression of genes within aganglionosis modifier intervals and their ligand-receptor partners. Cell types have been consolidated to those that are in Fig 5, and each cell type has split expression for developmental time in order, excluding 15.5dpc.

https://doi.org/10.1371/journal.pcbi.1014424.s005

(TIFF)

S6 Fig. Expression of Sox10^Dom aganglionosis modifier interval candidate genes from wavefront ENCDC bulk RNA-seq and near conserved SOX10 binding motifs.

(A) UMAP of the neural crest cells from [92] highlighting neuronal and progenitor cells. UMAPs of the neural crest cells from [92] showing expression in purple of candidate genes that are upregulated in the migrating wavefront of ENCDCs (B) or are near or overlapping with conserved SOX10 binding motifs grouped by prior evidence (C), other data modalities supporting gene as a candidate) and number of binding motifs (D, two binding motifs; E, one binding motif).

https://doi.org/10.1371/journal.pcbi.1014424.s006

(TIFF)

S7 Fig. Whole gut 16.5dpc Phox2b H2B-CFP + snATAC-seq label transfer and derived putative cis-regulatory elements within modifier intervals.

(A) Flow chart of analysis pipeline for differentially accessible chromatin contained within modifier intervals. (B) UMAP of scRNA-seq of 15.5dpc whole gut enteric nervous system cells annotated by supervised clustering used as a template for estimation of cell types in the snATAC-seq. (C) Estimation of cell types via Seurat and Signac’s LabelTransfer function used to annotate Fig 6B. Clusters from B are on the x-axis and clusters from Fig 6B are on the y-axis. (D) Annotations of differentially accessible (DA) peaks from all DA peaks (top) and those within modifier intervals split by cluster (bottom).

https://doi.org/10.1371/journal.pcbi.1014424.s007

(TIFF)

S8 Fig. Expression of progenitor, neuronal and candidate modifier genes across ENCDCs migration stages.

Top column shows the Zhao et al. ENCDCs split by timepoint (columns) in chronological order and all cells (rightmost column) colored by cell state (progenitor, neuronal). Violin plots showing expression of marker genes (Sox10, Phox2b, Ednrb, Elavl4) relative to candidate modifier genes (Dach1, Col1a1, Pdgfra, and Tbx2) by cell state corresponding to each timepoint in the top row.

https://doi.org/10.1371/journal.pcbi.1014424.s008

(TIFF)

S1 Table. Tab-separated table with phenotypes of each mouse in the study listed by column, including length of intestine, length of hypoganglionosis and aganglionosis, percentages of hypoganglionosis and aganglionosis, sex of the mice, and mouse individual IDs.

https://doi.org/10.1371/journal.pcbi.1014424.s009

(TSV)

S2 Table. Tab-separated genotype file in a PED-like format (PLINK) with the first columns being individual IDs for each mouse and columns per SNP genotype.

https://doi.org/10.1371/journal.pcbi.1014424.s010

(TSV)

S3 Table. CSV file containing GEMMA scan results for all mice using sex as a covariate and the quantitative percentage aganglionosis phenotype.

https://doi.org/10.1371/journal.pcbi.1014424.s011

(CSV)

S4 Table. CSV file containing GEMMA scan results for all mice using sex as a covariate and the quantitative percentage aganglionosis phenotype but excluding chromosomes 5, 15, and X.

https://doi.org/10.1371/journal.pcbi.1014424.s012

(CSV)

S5 Table. CSV file containing GEMMA scan results for all mice using sex as a covariate and the binary aganglionosis phenotype (either no aganglionosis or yes aganglionosis).

https://doi.org/10.1371/journal.pcbi.1014424.s013

(CSV)

S6 Table. CSV file containing GEMMA scan results for only mice that had aganglionosis using sex as a covariate and the quantitative percentage aganglionosis phenotype.

https://doi.org/10.1371/journal.pcbi.1014424.s014

(CSV)

S7 Table. CSV file containing GEMMA scan results for only female mice that had aganglionosis using sex as a covariate and the quantitative percentage aganglionosis phenotype.

https://doi.org/10.1371/journal.pcbi.1014424.s015

(CSV)

S8 Table. CSV file containing GEMMA scan results for only male mice that had aganglionosis using sex as a covariate and the quantitative percentage aganglionosis phenotype.

https://doi.org/10.1371/journal.pcbi.1014424.s016

(CSV)

S9 Table. CSV file containing GEMMA scan results for male mice only using the quantitative percentage aganglionosis phenotype.

https://doi.org/10.1371/journal.pcbi.1014424.s017

(CSV)

S10 Table. CSV file containing GEMMA scan results for female mice only using the quantitative percentage aganglionosis phenotype.

https://doi.org/10.1371/journal.pcbi.1014424.s018

(CSV)

S11 Table. CSV file containing GEMMA scan results for male mice only using the binary aganglionosis phenotype (either no aganglionosis or yes aganglionosis).

https://doi.org/10.1371/journal.pcbi.1014424.s019

(CSV)

S12 Table. CSV file containing GEMMA scan results for female mice only using the binary aganglionosis phenotype (either no aganglionosis or yes aganglionosis).

https://doi.org/10.1371/journal.pcbi.1014424.s020

(CSV)

S13 Table. CSV file containing comprehensive information of modifier interval definitions derived from significant SNPs in S3–S5 and S9–S12 Tables.

https://doi.org/10.1371/journal.pcbi.1014424.s021

(CSV)

S14 Table. CSV file that lists the genes within each modifier interval that are expressed above a normalized expression threshold of 1 in the Zhao et al. scRNA-seq dataset.

https://doi.org/10.1371/journal.pcbi.1014424.s022

(CSV)

S15 Table. CSV file containing the significant (Bonferroni-correction p-value) differential gene expression results (Seurat’s FindAllMarkers function) for genes within the modifier intervals in the Zhao et al. scRNA-seq dataset.

https://doi.org/10.1371/journal.pcbi.1014424.s023

(CSV)

S16 Table. Excel file containing significant results of the CellChat analysis in the Zhao et al. scRNA-seq dataset.

https://doi.org/10.1371/journal.pcbi.1014424.s024

(XLSX)

S17 Table. TSV file containing migrating wavefront versus lagging enteric neural crest-derived cell significant differential gene expression results from Stavely et al. filtered for those within modifier intervals.

https://doi.org/10.1371/journal.pcbi.1014424.s025

(TSV)

S18 Table. CSV file containing cross-species conserved SOX10 transcription factor binding motifs from Gopinath et al. that fall within modifier intervals, along with their closest genomic feature found through the ClosestFeature function from the R package Signac.

https://doi.org/10.1371/journal.pcbi.1014424.s026

(CSV)

S19 Table. CSV file containing significant differential chromatin accessibility results (Seurat’s FindAllMarkers function) from 16.5dpc Phox2b H2B-CFP + ENS snATAC-seq.

Cluster definitions were derived from transferring labels from 15.5dpc ENS scRNA-seq onto the snATAC-seq data.

https://doi.org/10.1371/journal.pcbi.1014424.s027

(CSV)

S20 Table. Results in S18 Table but filtered for those differentially accessible chromatin regions that lie within modifier intervals.

https://doi.org/10.1371/journal.pcbi.1014424.s028

(CSV)

S21 Table. Overlap of the results in S17 and S19 Tables; those differentially accessible chromatin regions that overlap with conserved SOX10 transcription factor binding motifs.

https://doi.org/10.1371/journal.pcbi.1014424.s029

(CSV)

S22 Table. CSV file containing significantly enriched transcription factor binding motifs derived from differentially accessible chromatin regions contained within modifier intervals (S19 Table) found via the R package Signac’s FindMotifs function.

https://doi.org/10.1371/journal.pcbi.1014424.s030

(CSV)

S23 Table. CSV file containing the results from S21 Table filtered for those transcription factors that are expressed in the Zhao et al. “neural crest” subset of cells.

https://doi.org/10.1371/journal.pcbi.1014424.s031

(CSV)

S24 Table. Excel file containing, for each omics candidate gene, the genetics score information.

https://doi.org/10.1371/journal.pcbi.1014424.s032

(XLSX)

S25 Table. TSV file containing, for each omics candidate gene, the results from each omics prioritization method including: the p-value, method of quantitation/statistics, cell type (if applicable), age (if applicable), number of loci (if applicable), the modifier interval(s) in which a gene resides, the dataset from which the gene was prioritized, and the omics score (filtered for those omics scores greater than 3).

https://doi.org/10.1371/journal.pcbi.1014424.s033

(TSV)

S26 Table. TSV containing Ensembl Variant Effect Predictor results for omics candidate genes whose C3HeB/FeJ variants were predicted to have high impact on the corresponding mRNA or protein.

https://doi.org/10.1371/journal.pcbi.1014424.s034

(TSV)

S27 Table. CSV file containing Hirschsprung GWAS summary statistics from Tang et al. filtered by those within EUR (European ancestry group) LD of candidate genes.

https://doi.org/10.1371/journal.pcbi.1014424.s035

(CSV)

S28 Table. CSV file containing Hirschsprung GWAS summary statistics from Tang et al. filtered by those within ASN (East Asian ancestry group) LD of candidate genes.

https://doi.org/10.1371/journal.pcbi.1014424.s036

(CSV)

S29 Table. CSV file containing Hirschsprung GWAS summary statistics from Garcia-Barcelo et al. filtered by those within ASN (East Asian ancestry group) LD of candidate genes.

https://doi.org/10.1371/journal.pcbi.1014424.s037

(CSV)

S30 Table. CSV file containing Stool Frequency GWAS summary statistics from Bonfiglio et al. filtered by those within EUR (European ancestry group) LD of candidate genes.

https://doi.org/10.1371/journal.pcbi.1014424.s038

(CSV)

S31 Table. TSV file containing concatenated Genebass.org PheWAS results from this study’s list of candidate genes’ human orthologs.

https://doi.org/10.1371/journal.pcbi.1014424.s039

(TSV)

Acknowledgments

We are grateful to Ashley Cantrell, Elizabeth Stein, and Emily Ferguson for maintenance of the C3FeB6.Sox10^Dom pedigree and assisting with wholemount AChE gut staining; to Joan Breyer for helping grid DNA samples; and to Kevin M. Bradley for assistance with managing genotyping data. Genotyping was performed at the Center for Inherited Disease Research (Johns Hopkins University) via the NIH CIDR Program. We thank the Vanderbilt Technologies for Advanced Genomics (VANTAGE) shared resource for sequencing and CellRanger analysis for the snATAC-seq samples. We appreciate the suggestions made by Dr. Karl Broman on using a split aganglionosis phenotype approach to account for the zero-inflated data for the pedigree analysis. We thank Dr. Laura Reinholdt and colleagues for sharing the C3HeB/FeJ whole genome sequencing dataset that was vital for identifying strain variants in our study. We are grateful to Drs. Rebecca Ihrie and Mary Chalkley for assisting with use of the Countess Automated cell counter. We are grateful for support of a Flow sort Scholarship fund from the Vanderbilt University Medical Center Digestive Disease Research Center. HuC/D antibody was a kind gift from Vanda Lennon (Mayo Clinic).

References

1. Abiola O, Angel JM, Avner P, Bachmanov AA, Belknap JK, Bennett B, et al. The nature and identification of quantitative trait loci: a community’s view. Nat Rev Genet. 2003;4(11):911–6. pmid:14634638
- View Article
- PubMed/NCBI
- Google Scholar
2. Avila JA, Benthal JT, Schafer JC, Flaherty DK, Southard-Smith EM. Single cell profiling in the Sox10Dom Hirschsprung mouse implicates Hox genes in enteric neuron trajectory allocation. Cell Mol Gastroenterol Hepatol. 2025;19(12):101590. pmid:40701368
- View Article
- PubMed/NCBI
- Google Scholar
3. Berisa T, Pickrell JK. Approximately independent linkage disequilibrium blocks in human populations. Bioinformatics. 2016;32(2):283–5. pmid:26395773
- View Article
- PubMed/NCBI
- Google Scholar
4. Bolk S, Pelet A, Hofstra RM, Angrist M, Salomon R, Croaker D, et al. A human model for multigenic inheritance: phenotypic expression in Hirschsprung disease requires both the RET gene and a new 9q31 locus. Proc Natl Acad Sci U S A. 2000;97(1):268–73. pmid:10618407
- View Article
- PubMed/NCBI
- Google Scholar
5. Bonfiglio F, Liu X, Smillie C, Pandit A, Kurilshikov A, Bacigalupe R, et al. GWAS of stool frequency provides insights into gastrointestinal motility and irritable bowel syndrome. Cell Genom. 2021;1(3):None. pmid:34957435
- View Article
- PubMed/NCBI
- Google Scholar
6. Borghini S, Di Duca M, Santamaria G, Vargiolu M, Bachetti T, Cargnin F, et al. Transcriptional regulation of TLX2 and impaired intestinal innervation: possible role of the PHOX2A and PHOX2B genes. Eur J Hum Genet. 2007;15(8):848–55. pmid:17505528
- View Article
- PubMed/NCBI
- Google Scholar
7. Bradnock TJ, Knight M, Kenny S, Nair M, Walker GM, British Association of Paediatric Surgeons Congenital Anomalies Surveillance System. Hirschsprung’s disease in the UK and Ireland: incidence and anomalies. Arch Dis Child. 2017;102(8):722–7. pmid:28280094
- View Article
- PubMed/NCBI
- Google Scholar
8. Broman KW, Sen S, Owens SE, Manichaikul A, Southard-Smith EM, Churchill GA. The X chromosome in quantitative trait locus mapping. Genetics. 2006;174(4):2151–8. pmid:17028340
- View Article
- PubMed/NCBI
- Google Scholar
9. Cantrell VA, Owens SE, Chandler RL, Airey DC, Bradley KM, Smith JR, et al. Interactions between Sox10 and EdnrB modulate penetrance and severity of aganglionosis in the Sox10Dom mouse model of Hirschsprung disease. Hum Mol Genet. 2004;13(19):2289–301. pmid:15294878
- View Article
- PubMed/NCBI
- Google Scholar
10. Choudhary S, Satija R. Comparison and evaluation of statistical error models for scRNA-seq. Genome Biol. 2022;23(1):27. pmid:35042561
- View Article
- PubMed/NCBI
- Google Scholar
11. Corpening JC, Cantrell VA, Deal KK, Southard-Smith EM. A Histone2BCerulean BAC transgene identifies differential expression of Phox2b in migrating enteric neural crest derivatives and enteric glia. Dev Dyn. 2008;237(4):1119–32. pmid:18351668
- View Article
- PubMed/NCBI
- Google Scholar
12. Davis RJ, Shen W, Sandler YI, Amoui M, Purcell P, Maas R, et al. Dach1 mutant mice bear no gross abnormalities in eye, limb, and brain development and exhibit postnatal lethality. Mol Cell Biol. 2001;21(5):1484–90. pmid:11238885
- View Article
- PubMed/NCBI
- Google Scholar
13. de Pontual L, Pelet A, Clement-Ziza M, Trochet D, Antonarakis SE, Attie-Bitach T, et al. Epistatic interactions with a common hypomorphic RET allele in syndromic Hirschsprung disease. Hum Mutat. 2007;28(8):790–6. pmid:17397038
- View Article
- PubMed/NCBI
- Google Scholar
14. Dorman A, Binenbaum I, Abu-Toamih Atamni HJ, Chatziioannou A, Tomlinson I, Mott R, et al. Genetic mapping of novel modifiers for ApcMin induced intestinal polyps’ development using the genetic architecture power of the collaborative cross mice. BMC Genom. 2021;22(1):566. pmid:34294033
- View Article
- PubMed/NCBI
- Google Scholar
15. Elworthy S, Pinto JP, Pettifer A, Cancela ML, Kelsh RN. Phox2b function in the enteric nervous system is conserved in zebrafish and is sox10-dependent. Mech Dev. 2005;122(5):659–69. pmid:15817223
- View Article
- PubMed/NCBI
- Google Scholar
16. Enomoto H, Araki T, Jackman A, Heuckeroth RO, Snider WD, Johnson EM Jr, et al. GFR alpha1-deficient mice have deficits in the enteric nervous system and kidneys. Neuron. 1998;21(2):317–24. pmid:9728913
- View Article
- PubMed/NCBI
- Google Scholar
17. Fadista J, Lund M, Skotte L, Geller F, Nandakumar P, Chatterjee S, et al. Genome-wide association study of Hirschsprung disease detects a novel low-frequency variant at the RET locus. Eur J Hum Genet. 2018;26(4):561–9. pmid:29379196
- View Article
- PubMed/NCBI
- Google Scholar
18. Feng W, Liu Y, Die X, Hou J, Guo Z, Liu W, et al. Sex differences in postoperative outcomes of Hirschsprung disease: propensity score matching analysis. J Pediatr Surg. 2025;60(12):162655. pmid:40930386
- View Article
- PubMed/NCBI
- Google Scholar
19. Finzsch M, Stolt CC, Lommes P, Wegner M. Sox9 and Sox10 influence survival and migration of oligodendrocyte precursors in the spinal cord by regulating PDGF receptor alpha expression. Development. 2008;135(4):637–46. pmid:18184726
- View Article
- PubMed/NCBI
- Google Scholar
20. Fu M, Lui VCH, Sham MH, Cheung ANY, Tam PKH. HOXB5 expression is spatially and temporarily regulated in human embryonic gut during neural crest cell colonization and differentiation of enteric neuroblasts. Dev Dyn. 2003;228(1):1–10. pmid:12950074
- View Article
- PubMed/NCBI
- Google Scholar
21. Gao T, Wright-Jin EC, Sengupta R, Anderson JB, Heuckeroth RO. Cell-autonomous retinoic acid receptor signaling has stage-specific effects on mouse enteric nervous system. JCI Insight. 2021;6(10):e145854. pmid:33848271
- View Article
- PubMed/NCBI
- Google Scholar
22. Garcia-Barcelo M-M, Tang CS-M, Ngan ES-W, Lui VC-H, Chen Y, So M-T, et al. Genome-wide association study identifies NRG1 as a susceptibility locus for Hirschsprung’s disease. Proc Natl Acad Sci U S A. 2009;106(8):2694–9. pmid:19196962
- View Article
- PubMed/NCBI
- Google Scholar
23. Genebass: gene-based association summary statistics. [cited 2024 Nov 14]. Available from: https://app.genebass.org/
- View Article
- Google Scholar
24. Gopinath C, Law WD, Rodríguez-Molina JF, Prasad AB, Song L, Crawford GE, et al. Stringent comparative sequence analysis reveals SOX10 as a putative inhibitor of glial cell differentiation. BMC Genom. 2016;17(1):887. pmid:27821050
- View Article
- PubMed/NCBI
- Google Scholar
25. Hakami RM, Hou L, Baxter LL, Loftus SK, Southard-Smith EM, Incao A, et al. Genetic evidence does not support direct regulation of EDNRB by SOX10 in migratory neural crest and the melanocyte lineage. Mech Dev. 2006;123(2):124–34. pmid:16412618
- View Article
- PubMed/NCBI
- Google Scholar
26. Hao Y, Stuart T, Kowalski MH, Choudhary S, Hoffman P, Hartman A, et al. Dictionary learning for integrative, multimodal and scalable single-cell analysis. Nat Biotechnol. 2024;42(2):293–304. pmid:37231261
- View Article
- PubMed/NCBI
- Google Scholar
27. Hinrichs AS, Karolchik D, Baertsch R, Barber GP, Bejerano G, Clawson H, et al. The UCSC Genome Browser database: update 2006. Nucleic Acids Research. 2006;34(Database issue):D590–8.
- View Article
- Google Scholar
28. Integrative analysis in Seurat v5. [cited 2024 Nov 14]. Available from: https://satijalab.org/seurat/articles/seurat5_integration
- View Article
- Google Scholar
29. Jiang M, Li C, Cao G, Yang D, Zhang X, Yang L, et al. Effects of NRG1 polymorphisms on Hirschsprung’s disease susceptibility: a meta-analysis. Sci Rep. 2017;7(1):9913. pmid:28855726
- View Article
- PubMed/NCBI
- Google Scholar
30. Jiang Q, Ho Y-Y, Hao L, Nichols Berrios C, Chakravarti A. Copy number variants in candidate genes are genetic modifiers of Hirschsprung disease. PLoS One. 2011;6(6):e21219. pmid:21712996
- View Article
- PubMed/NCBI
- Google Scholar
31. Jin S, Guerrero-Juarez CF, Zhang L, Chang I, Ramos R, Kuan C-H, et al. Inference and analysis of cell-cell communication using CellChat. Nat Commun. 2021;12(1):1088. pmid:33597522
- View Article
- PubMed/NCBI
- Google Scholar
32. Jin S, Plikus MV, Nie Q. CellChat for systematic analysis of cell-cell communication from single-cell transcriptomics. Nat Protoc. 2025;20(1):180–219. pmid:39289562
- View Article
- PubMed/NCBI
- Google Scholar
33. Kaern M, Elston TC, Blake WJ, Collins JJ. Stochasticity in gene expression: from theories to phenotypes. Nat Rev Genet. 2005;6(6):451–64. pmid:15883588
- View Article
- PubMed/NCBI
- Google Scholar
34. Kalra S, Lanno S, Sanchez G, Coolon JD. cis- and trans-regulatory contributions to a hierarchy of factors influencing gene expression variation. Commun Biol. 2024;7(1):1563. pmid:39587248
- View Article
- PubMed/NCBI
- Google Scholar
35. Karczewski KJ, Solomonson M, Chao KR, Goodrich JK, Tiao G, Lu W, et al. Systematic single-variant and gene-based association testing of thousands of phenotypes in 394,841 UK Biobank exomes. Cell Genom. 2022;2(9):100168.
- View Article
- Google Scholar
36. Karolchik D, Hinrichs AS, Furey TS, Roskin KM, Sugnet CW, Haussler D, et al. The UCSC Table Browser data retrieval tool. Nucleic Acids Res. 2004;32(Database issue):D493-6. pmid:14681465
- View Article
- PubMed/NCBI
- Google Scholar
37. Kim J-H, Cheong HS, Sul JH, Seo J-M, Kim D-Y, Oh J-T, et al. A genome-wide association study identifies potential susceptibility loci for Hirschsprung disease. PLoS One. 2014;9(10):e110292. pmid:25310821
- View Article
- PubMed/NCBI
- Google Scholar
38. Kim Y-K, Lee H, Ismail T, Kim Y, Lee H-S. Dach1 regulates neural crest migration during embryonic development. Biochem Biophys Res Commun. 2020;527(4):896–901. pmid:32430182
- View Article
- PubMed/NCBI
- Google Scholar
39. Korsunsky I, Millard N, Fan J, Slowikowski K, Zhang F, Wei K, et al. Fast, sensitive and accurate integration of single-cell data with Harmony. Nat Methods. 2019;16(12):1289–96. pmid:31740819
- View Article
- PubMed/NCBI
- Google Scholar
40. Kuil LE, MacKenzie KC, Tang CS, Windster JD, Le TL, Karim A, et al. Size matters: Large copy number losses in Hirschsprung disease patients reveal genes involved in enteric nervous system development. PLoS Genet. 2021;17(8):e1009698. pmid:34358225
- View Article
- PubMed/NCBI
- Google Scholar
41. Lake JI, Heuckeroth RO. Enteric nervous system development: migration, differentiation, and disease. Am J Physiol Gastrointest Liver Physiol. 2013;305(1):G1-24. pmid:23639815
- View Article
- PubMed/NCBI
- Google Scholar
42. Lane PW, Liu HM. Association of megacolon with a new dominant spotting gene (Dom) in the mouse. J Hered. 1984;75(6):435–9. pmid:6512238
- View Article
- PubMed/NCBI
- Google Scholar
43. Lawrence M, Gentleman R, Carey V. rtracklayer: an R package for interfacing with genome browsers. Bioinformatics. 2009;25(14):1841–2. pmid:19468054
- View Article
- PubMed/NCBI
- Google Scholar
44. Li N, Kelsh RN, Croucher P, Roehl HH. Regulation of neural crest cell fate by the retinoic acid and Pparg signalling pathways. Development. 2010;137(3):389–94. pmid:20081187
- View Article
- PubMed/NCBI
- Google Scholar
45. Li C, Hu R, Hou N, Wang Y, Wang Z, Yang T, et al. Alteration of the retinoid acid-CBP signaling pathway in neural crest induction contributes to enteric nervous system disorder. Front Pediatr. 2018;6:382. pmid:30560112
- View Article
- PubMed/NCBI
- Google Scholar
46. Li Z, Lui KN-C, Lau S-T, Lai FP-L, Li P, Chung PH-Y, et al. Transcriptomics of Hirschsprung disease patient-derived enteric neural crest cells reveals a role for oxidative phosphorylation. Nat Commun. 2023;14(1):2157. pmid:37061531
- View Article
- PubMed/NCBI
- Google Scholar
47. Lippert C, Listgarten J, Liu Y, Kadie CM, Davidson RI, Heckerman D. FaST linear mixed models for genome-wide association studies. Nat Methods. 2011;8(10):833–5. pmid:21892150
- View Article
- PubMed/NCBI
- Google Scholar
48. Liu SJ, Casey-Clyde T, Cho NW, Swinderman J, Pekmezci M, Dougherty MC, et al. Epigenetic reprogramming shapes the cellular landscape of schwannoma. Nat Commun. 2024;15(1):476. pmid:38216587
- View Article
- PubMed/NCBI
- Google Scholar
49. López SH, Avetisyan M, Wright CM, Mesbah K, Kelly RG, Moon AM, et al. Loss of Tbx3 in murine neural crest reduces enteric glia and causes cleft palate, but does not influence heart development or bowel transit. Dev Biol. 2018;444 Suppl 1(Suppl 1):S337–51. pmid:30292786
- View Article
- PubMed/NCBI
- Google Scholar
50. Lorè NI, Sipione B, He G, Strug LJ, Atamni HJ, Dorman A, et al. Collaborative cross mice yield genetic modifiers for Pseudomonas aeruginosa infection in human lung disease. mBio. 2020;11(2):e00097-20. pmid:32127447
- View Article
- PubMed/NCBI
- Google Scholar
51. Lui VCH, Cheng WWC, Leon TYY, Lau DKC, Garcia-Barcelo M-M, Miao XP, et al. Perturbation of hoxb5 signaling in vagal neural crests down-regulates ret leading to intestinal hypoganglionosis in mice. Gastroenterology. 2008;134(4):1104–15. pmid:18395091
- View Article
- PubMed/NCBI
- Google Scholar
52. Majd H, Samuel RM, Cesiulis A, Ramirez JT, Kalantari A, Barber K. Engrafted nitrergic neurons derived from hPSCs improve gut dysmotility in mice. Nature. 2025.
- View Article
- Google Scholar
53. Maka M, Stolt CC, Wegner M. Identification of Sox8 as a modifier gene in a mouse model of Hirschsprung disease reveals underlying molecular defect. Dev Biol. 2005;277(1):155–69. pmid:15572147
- View Article
- PubMed/NCBI
- Google Scholar
54. Mallory SB, Wiener E, Nordlund JJ. Waardenburg’s syndrome with Hirschsprung’s disease: a neural crest defect. Pediatr Dermatol. 1986;3(2):119–24. pmid:3952027
- View Article
- PubMed/NCBI
- Google Scholar
55. May-Zhang AA, Tycksen E, Southard-Smith AN, Deal KK, Benthal JT, Buehler DP, et al. Combinatorial transcriptional profiling of mouse and human enteric neurons identifies shared and disparate subtypes in situ. Gastroenterology. 2021;160(3):755-770.e26. pmid:33010250
- View Article
- PubMed/NCBI
- Google Scholar
56. McCallion AS, Stames E, Conlon RA, Chakravarti A. Phenotype variation in two-locus mouse models of Hirschsprung disease: tissue-specific interaction between Ret and Ednrb. Proc Natl Acad Sci U S A. 2003;100(4):1826–31. pmid:12574515
- View Article
- PubMed/NCBI
- Google Scholar
57. McLaren W, Gil L, Hunt SE, Riat HS, Ritchie GRS, Thormann A. The Ensembl Variant Effect Predictor. Genome Biol. 2016;17(1):122.
- View Article
- Google Scholar
58. Memic F, Knoflach V, Morarach K, Sadler R, Laranjeira C, Hjerling-Leffler J, et al. Transcription and signaling regulators in developing neuronal subtypes of mouse and human enteric nervous system. Gastroenterology. 2018;154(3):624–36. pmid:29031500
- View Article
- PubMed/NCBI
- Google Scholar
59. Mo J, Long R, Fantauzzo KA. Pdgfra and Pdgfrb genetically interact in the murine neural crest cell lineage to regulate migration and proliferation. Front Physiol. 2020;11:588901. pmid:33224039
- View Article
- PubMed/NCBI
- Google Scholar
60. Moghimbeigi A. Two-part zero-inflated negative binomial regression model for quantitative trait loci mapping with count trait. J Theor Biol. 2015;372:74–80. pmid:25728790
- View Article
- PubMed/NCBI
- Google Scholar
61. Nagashimada M, Ohta H, Li C, Nakao K, Uesaka T, Brunet J-F, et al. Autonomic neurocristopathy-associated mutations in PHOX2B dysregulate Sox10 expression. J Clin Invest. 2012;122(9):3145–58. pmid:22922260
- View Article
- PubMed/NCBI
- Google Scholar
62. Nagy N, Goldstein AM. Enteric nervous system development: a crest cell’s journey from neural tube to colon. Semin Cell Dev Biol. 2017;66:94–106. pmid:28087321
- View Article
- PubMed/NCBI
- Google Scholar
63. Owens SE, Broman KW, Wiltshire T, Elmore JB, Bradley KM, Smith JR, et al. Genome-wide linkage identifies novel modifier loci of aganglionosis in the Sox10Dom model of Hirschsprung disease. Hum Mol Genet. 2005;14(11):1549–58. pmid:15843399
- View Article
- PubMed/NCBI
- Google Scholar
64. Parisi MA, Kapur RP, Neilson I, Hofstra RMW, Holloway LW, Michaelis RC, et al. Hydrocephalus and intestinal aganglionosis: is L1CAM a modifier gene in Hirschsprung disease? Am J Med Genet. 2002;108(1):51–6. pmid:11857550
- View Article
- PubMed/NCBI
- Google Scholar
65. Pattyn A, Morin X, Cremer H, Goridis C, Brunet JF. The homeobox gene Phox2b is essential for the development of autonomic neural crest derivatives. Nature. 1999;399(6734):366–70. pmid:10360575
- View Article
- PubMed/NCBI
- Google Scholar
66. Perez G, Barber GP, Benet-Pages A, Casper J, Clawson H, Diekhans M, et al. The UCSC Genome Browser database: 2025 update. Nucleic Acids Res. 2024.
- View Article
- Google Scholar
67. Pingault V, Bondurand N, Kuhlbrodt K, Goerich DE, Préhu MO, Puliti A, et al. SOX10 mutations in patients with Waardenburg-Hirschsprung disease. Nat Genet. 1998;18(2):171–3. pmid:9462749
- View Article
- PubMed/NCBI
- Google Scholar
68. Sato Y, Heuckeroth RO. Retinoic acid regulates murine enteric nervous system precursor proliferation, enhances neuronal precursor differentiation, and reduces neurite growth in vitro. Dev Biol. 2008;320(1):185–98. pmid:18561907
- View Article
- PubMed/NCBI
- Google Scholar
69. Shih C-H, Fay J. Cis-regulatory variants affect gene expression dynamics in yeast. Elife. 2021;10:e68469. pmid:34369376
- View Article
- PubMed/NCBI
- Google Scholar
70. Shull LC, Lencer ES, Kim HM, Goyama S, Kurokawa M, Costello JC, et al. PRDM paralogs antagonistically balance Wnt/β-catenin activity during craniofacial chondrocyte differentiation. Development. 2022;149(4):dev200082. pmid:35132438
- View Article
- PubMed/NCBI
- Google Scholar
71. Simpson MJ, Zhang DC, Mariani M, Landman KA, Newgreen DF. Cell proliferation drives neural crest cell invasion of the intestine. Dev Biol. 2007;302(2):553–68. pmid:17178116
- View Article
- PubMed/NCBI
- Google Scholar
72. Southard-Smith EM, Angrist M, Ellison JS, Agarwala R, Baxevanis AD, Chakravarti A, et al. The Sox10(Dom) mouse: modeling the genetic variation of Waardenburg-Shah (WS4) syndrome. Genome Res. 1999;9(3):215–25. pmid:10077527
- View Article
- PubMed/NCBI
- Google Scholar
73. Stanchina L, Baral V, Robert F, Pingault V, Lemort N, Pachnis V, et al. Interactions between Sox10, Edn3 and Ednrb during enteric nervous system and melanocyte development. Dev Biol. 2006;295(1):232–49. pmid:16650841
- View Article
- PubMed/NCBI
- Google Scholar
74. Stavely R, Hotta R, Guyer RA, Picard N, Rahman AA, Omer M, et al. A distinct transcriptome characterizes neural crest-derived cells at the migratory wavefront during enteric nervous system development. Development. 2023;150(5):dev201090. pmid:36779913
- View Article
- PubMed/NCBI
- Google Scholar
75. Stuart T, Srivastava A, Madad S, Lareau CA, Satija R. Single-cell chromatin state analysis with Signac. Nat Methods. 2021;18(11):1333–41.
- View Article
- Google Scholar
76. Tang CS-M, Gui H, Kapoor A, Kim J-H, Luzón-Toro B, Pelet A, et al. Trans-ethnic meta-analysis of genome-wide association studies for Hirschsprung disease. Hum Mol Genet. 2016;25(23):5265–75. pmid:27702942
- View Article
- PubMed/NCBI
- Google Scholar
77. Taylor MK, Yeager K, Morrison SJ. Physiological Notch signaling promotes gliogenesis in the developing peripheral and central nervous systems. Development. 2007;134(13):2435–47. pmid:17537790
- View Article
- PubMed/NCBI
- Google Scholar
78. Tilghman JM, Ling AY, Turner TN, Sosa MX, Krumm N, Chatterjee S, et al. N Engl J Med. 2019;380(15):1421–32.
- View Article
- Google Scholar
79. Uribe RA, Hong SS, Bronner ME. Retinoic acid temporally orchestrates colonization of the gut by vagal neural crest cells. Dev Biol. 2018;433(1):17–32. pmid:29108781
- View Article
- PubMed/NCBI
- Google Scholar
80. Wakamatsu Y, Uchikawa M. The many faces of Sox2 function in neural crest development. Dev Growth Differ. 2021;63(1):93–9. pmid:33326593
- View Article
- PubMed/NCBI
- Google Scholar
81. Wang F, Hu W, Xian J, Ohnuma S, Brenton JD. The Xenopus Tgfbi is required for embryogenesis through regulation of canonical Wnt signalling. Dev Biol. 2013;379(1):16–27. pmid:23246513
- View Article
- PubMed/NCBI
- Google Scholar
82. Wang B, Fang X, Sun X, Du C, Zhou L, Lv X, et al. m6A demethylase ALKBH5 suppresses proliferation and migration of enteric neural crest cells by regulating TAGLN in Hirschsprung’s disease. Life Sci. 2021;278:119577. pmid:33961858
- View Article
- PubMed/NCBI
- Google Scholar
83. Wright CM, Schneider S, Smith-Edwards KM, Mafra F, Leembruggen AJL, Gonzalez MV, et al. scRNA-seq reveals new enteric nervous system roles for GDNF, NRTN, and TBX3. Cell Mol Gastroenterol Hepatol. 2021;11(5):1548-1592.e1. pmid:33444816
- View Article
- PubMed/NCBI
- Google Scholar
84. Wu K, Katiyar S, Li A, Liu M, Ju X, Popov VM, et al. Dachshund inhibits oncogene-induced breast cancer cellular migration and invasion through suppression of interleukin-8. Proc Natl Acad Sci U S A. 2008;105(19):6924–9. pmid:18467491
- View Article
- PubMed/NCBI
- Google Scholar
85. Yang J, Zaitlen NA, Goddard ME, Visscher PM, Price AL. Advantages and pitfalls in the application of mixed-model association methods. Nat Genet. 2014;46(2):100–6. pmid:24473328
- View Article
- PubMed/NCBI
- Google Scholar
86. Young HM, Jones BR, McKeown SJ. The projections of early enteric neurons are influenced by the direction of neural crest cell migration. J Neurosci. 2002;22(14):6005–18. pmid:12122062
- View Article
- PubMed/NCBI
- Google Scholar
87. Young HM, Bergner AJ, Müller T. Acquisition of neuronal and glial markers by neural crest-derived cells in the mouse intestine. J Comp Neurol. 2003;456(1):1–11. pmid:12508309
- View Article
- PubMed/NCBI
- Google Scholar
88. Young HM, Bergner AJ, Anderson RB, Enomoto H, Milbrandt J, Newgreen DF, et al. Dynamics of neural crest-derived cell migration in the embryonic mouse gut. Dev Biol. 2004;270(2):455–73. pmid:15183726
- View Article
- PubMed/NCBI
- Google Scholar
89. Young HM, Bergner AJ, Simpson MJ, McKeown SJ, Hao MM, Anderson CR, et al. Colonizing while migrating: how do individual enteric neural crest cells behave? BMC Biol. 2014;12:23. pmid:24670214
- View Article
- PubMed/NCBI
- Google Scholar
90. Zhang Y, Liu T, Meyer CA, Eeckhoute J, Johnson DS, Bernstein BE, et al. Model-based analysis of ChIP-Seq (MACS). Genome Biol. 2008;9(9):R137. pmid:18798982
- View Article
- PubMed/NCBI
- Google Scholar
91. Zhang D, Rollo BN, Nagy N, Stamp L, Newgreen DF. The enteric neural crest progressively loses capacity to form enteric nervous system. Dev Biol. 2019;446(1):34–42. pmid:30529057
- View Article
- PubMed/NCBI
- Google Scholar
92. Zhao L, Song W, Chen Y-G. Mesenchymal-epithelial interaction regulates gastrointestinal tract development in mouse embryos. Cell Rep. 2022;40(2):111053. pmid:35830795
- View Article
- PubMed/NCBI
- Google Scholar
93. Zheng GXY, Terry JM, Belgrader P, Ryvkin P, Bent ZW, Wilson R, et al. Massively parallel digital transcriptional profiling of single cells. Nat Commun. 2017;8:14049. pmid:28091601
- View Article
- PubMed/NCBI
- Google Scholar
94. Zhou X, Stephens M. Genome-wide efficient mixed-model analysis for association studies. Nat Genet. 2012;44(7):821–4. pmid:22706312
- View Article
- PubMed/NCBI
- Google Scholar
95. Zhu JJ, Kam MK, Garcia-Barceló M-M, Tam PKH, Lui VCH. HOXB5 binds to multi-species conserved sequence (MCS+9.7) of RET gene and regulates RET expression. Int J Biochem Cell Biol. 2014;51:142–9. pmid:24794774
- View Article
- PubMed/NCBI
- Google Scholar

[ref1] 1. Abiola O, Angel JM, Avner P, Bachmanov AA, Belknap JK, Bennett B, et al. The nature and identification of quantitative trait loci: a community’s view. Nat Rev Genet. 2003;4(11):911–6. pmid:14634638
View Article
PubMed/NCBI
Google Scholar

[2] View Article

[3] PubMed/NCBI

[4] Google Scholar

[ref2] 2. Avila JA, Benthal JT, Schafer JC, Flaherty DK, Southard-Smith EM. Single cell profiling in the Sox10Dom Hirschsprung mouse implicates Hox genes in enteric neuron trajectory allocation. Cell Mol Gastroenterol Hepatol. 2025;19(12):101590. pmid:40701368
View Article
PubMed/NCBI
Google Scholar

[6] View Article

[7] PubMed/NCBI

[8] Google Scholar

[ref3] 3. Berisa T, Pickrell JK. Approximately independent linkage disequilibrium blocks in human populations. Bioinformatics. 2016;32(2):283–5. pmid:26395773
View Article
PubMed/NCBI
Google Scholar

[10] View Article

[11] PubMed/NCBI

[12] Google Scholar

[ref4] 4. Bolk S, Pelet A, Hofstra RM, Angrist M, Salomon R, Croaker D, et al. A human model for multigenic inheritance: phenotypic expression in Hirschsprung disease requires both the RET gene and a new 9q31 locus. Proc Natl Acad Sci U S A. 2000;97(1):268–73. pmid:10618407
View Article
PubMed/NCBI
Google Scholar

[14] View Article

[15] PubMed/NCBI

[16] Google Scholar

[ref5] 5. Bonfiglio F, Liu X, Smillie C, Pandit A, Kurilshikov A, Bacigalupe R, et al. GWAS of stool frequency provides insights into gastrointestinal motility and irritable bowel syndrome. Cell Genom. 2021;1(3):None. pmid:34957435
View Article
PubMed/NCBI
Google Scholar

[18] View Article

[19] PubMed/NCBI

[20] Google Scholar

[ref6] 6. Borghini S, Di Duca M, Santamaria G, Vargiolu M, Bachetti T, Cargnin F, et al. Transcriptional regulation of TLX2 and impaired intestinal innervation: possible role of the PHOX2A and PHOX2B genes. Eur J Hum Genet. 2007;15(8):848–55. pmid:17505528
View Article
PubMed/NCBI
Google Scholar

[22] View Article

[23] PubMed/NCBI

[24] Google Scholar

[ref7] 7. Bradnock TJ, Knight M, Kenny S, Nair M, Walker GM, British Association of Paediatric Surgeons Congenital Anomalies Surveillance System. Hirschsprung’s disease in the UK and Ireland: incidence and anomalies. Arch Dis Child. 2017;102(8):722–7. pmid:28280094
View Article
PubMed/NCBI
Google Scholar

[26] View Article

[27] PubMed/NCBI

[28] Google Scholar

[ref8] 8. Broman KW, Sen S, Owens SE, Manichaikul A, Southard-Smith EM, Churchill GA. The X chromosome in quantitative trait locus mapping. Genetics. 2006;174(4):2151–8. pmid:17028340
View Article
PubMed/NCBI
Google Scholar

[30] View Article

[31] PubMed/NCBI

[32] Google Scholar

[ref9] 9. Cantrell VA, Owens SE, Chandler RL, Airey DC, Bradley KM, Smith JR, et al. Interactions between Sox10 and EdnrB modulate penetrance and severity of aganglionosis in the Sox10Dom mouse model of Hirschsprung disease. Hum Mol Genet. 2004;13(19):2289–301. pmid:15294878
View Article
PubMed/NCBI
Google Scholar

[34] View Article

[35] PubMed/NCBI

[36] Google Scholar

[ref10] 10. Choudhary S, Satija R. Comparison and evaluation of statistical error models for scRNA-seq. Genome Biol. 2022;23(1):27. pmid:35042561
View Article
PubMed/NCBI
Google Scholar

[38] View Article

[39] PubMed/NCBI

[40] Google Scholar

[ref11] 11. Corpening JC, Cantrell VA, Deal KK, Southard-Smith EM. A Histone2BCerulean BAC transgene identifies differential expression of Phox2b in migrating enteric neural crest derivatives and enteric glia. Dev Dyn. 2008;237(4):1119–32. pmid:18351668
View Article
PubMed/NCBI
Google Scholar

[42] View Article

[43] PubMed/NCBI

[44] Google Scholar

[ref12] 12. Davis RJ, Shen W, Sandler YI, Amoui M, Purcell P, Maas R, et al. Dach1 mutant mice bear no gross abnormalities in eye, limb, and brain development and exhibit postnatal lethality. Mol Cell Biol. 2001;21(5):1484–90. pmid:11238885
View Article
PubMed/NCBI
Google Scholar

[46] View Article

[47] PubMed/NCBI

[48] Google Scholar

[ref13] 13. de Pontual L, Pelet A, Clement-Ziza M, Trochet D, Antonarakis SE, Attie-Bitach T, et al. Epistatic interactions with a common hypomorphic RET allele in syndromic Hirschsprung disease. Hum Mutat. 2007;28(8):790–6. pmid:17397038
View Article
PubMed/NCBI
Google Scholar

[50] View Article

[51] PubMed/NCBI

[52] Google Scholar

[ref14] 14. Dorman A, Binenbaum I, Abu-Toamih Atamni HJ, Chatziioannou A, Tomlinson I, Mott R, et al. Genetic mapping of novel modifiers for ApcMin induced intestinal polyps’ development using the genetic architecture power of the collaborative cross mice. BMC Genom. 2021;22(1):566. pmid:34294033
View Article
PubMed/NCBI
Google Scholar

[54] View Article

[55] PubMed/NCBI

[56] Google Scholar

[ref15] 15. Elworthy S, Pinto JP, Pettifer A, Cancela ML, Kelsh RN. Phox2b function in the enteric nervous system is conserved in zebrafish and is sox10-dependent. Mech Dev. 2005;122(5):659–69. pmid:15817223
View Article
PubMed/NCBI
Google Scholar

[58] View Article

[59] PubMed/NCBI

[60] Google Scholar

[ref16] 16. Enomoto H, Araki T, Jackman A, Heuckeroth RO, Snider WD, Johnson EM Jr, et al. GFR alpha1-deficient mice have deficits in the enteric nervous system and kidneys. Neuron. 1998;21(2):317–24. pmid:9728913
View Article
PubMed/NCBI
Google Scholar

[62] View Article

[63] PubMed/NCBI

[64] Google Scholar

[ref17] 17. Fadista J, Lund M, Skotte L, Geller F, Nandakumar P, Chatterjee S, et al. Genome-wide association study of Hirschsprung disease detects a novel low-frequency variant at the RET locus. Eur J Hum Genet. 2018;26(4):561–9. pmid:29379196
View Article
PubMed/NCBI
Google Scholar

[66] View Article

[67] PubMed/NCBI

[68] Google Scholar

[ref18] 18. Feng W, Liu Y, Die X, Hou J, Guo Z, Liu W, et al. Sex differences in postoperative outcomes of Hirschsprung disease: propensity score matching analysis. J Pediatr Surg. 2025;60(12):162655. pmid:40930386
View Article
PubMed/NCBI
Google Scholar

[70] View Article

[71] PubMed/NCBI

[72] Google Scholar

[ref19] 19. Finzsch M, Stolt CC, Lommes P, Wegner M. Sox9 and Sox10 influence survival and migration of oligodendrocyte precursors in the spinal cord by regulating PDGF receptor alpha expression. Development. 2008;135(4):637–46. pmid:18184726
View Article
PubMed/NCBI
Google Scholar

[74] View Article

[75] PubMed/NCBI

[76] Google Scholar

[ref20] 20. Fu M, Lui VCH, Sham MH, Cheung ANY, Tam PKH. HOXB5 expression is spatially and temporarily regulated in human embryonic gut during neural crest cell colonization and differentiation of enteric neuroblasts. Dev Dyn. 2003;228(1):1–10. pmid:12950074
View Article
PubMed/NCBI
Google Scholar

[78] View Article

[79] PubMed/NCBI

[80] Google Scholar

[ref21] 21. Gao T, Wright-Jin EC, Sengupta R, Anderson JB, Heuckeroth RO. Cell-autonomous retinoic acid receptor signaling has stage-specific effects on mouse enteric nervous system. JCI Insight. 2021;6(10):e145854. pmid:33848271
View Article
PubMed/NCBI
Google Scholar

[82] View Article

[83] PubMed/NCBI

[84] Google Scholar

[ref22] 22. Garcia-Barcelo M-M, Tang CS-M, Ngan ES-W, Lui VC-H, Chen Y, So M-T, et al. Genome-wide association study identifies NRG1 as a susceptibility locus for Hirschsprung’s disease. Proc Natl Acad Sci U S A. 2009;106(8):2694–9. pmid:19196962
View Article
PubMed/NCBI
Google Scholar

[86] View Article

[87] PubMed/NCBI

[88] Google Scholar

[ref23] 23. Genebass: gene-based association summary statistics. [cited 2024 Nov 14]. Available from: https://app.genebass.org/
View Article
Google Scholar

[90] View Article

[91] Google Scholar

[ref24] 24. Gopinath C, Law WD, Rodríguez-Molina JF, Prasad AB, Song L, Crawford GE, et al. Stringent comparative sequence analysis reveals SOX10 as a putative inhibitor of glial cell differentiation. BMC Genom. 2016;17(1):887. pmid:27821050
View Article
PubMed/NCBI
Google Scholar

[93] View Article

[94] PubMed/NCBI

[95] Google Scholar

[ref25] 25. Hakami RM, Hou L, Baxter LL, Loftus SK, Southard-Smith EM, Incao A, et al. Genetic evidence does not support direct regulation of EDNRB by SOX10 in migratory neural crest and the melanocyte lineage. Mech Dev. 2006;123(2):124–34. pmid:16412618
View Article
PubMed/NCBI
Google Scholar

[97] View Article

[98] PubMed/NCBI

[99] Google Scholar

[ref26] 26. Hao Y, Stuart T, Kowalski MH, Choudhary S, Hoffman P, Hartman A, et al. Dictionary learning for integrative, multimodal and scalable single-cell analysis. Nat Biotechnol. 2024;42(2):293–304. pmid:37231261
View Article
PubMed/NCBI
Google Scholar

[101] View Article

[102] PubMed/NCBI

[103] Google Scholar

[ref27] 27. Hinrichs AS, Karolchik D, Baertsch R, Barber GP, Bejerano G, Clawson H, et al. The UCSC Genome Browser database: update 2006. Nucleic Acids Research. 2006;34(Database issue):D590–8.
View Article
Google Scholar

[105] View Article

[106] Google Scholar

[ref28] 28. Integrative analysis in Seurat v5. [cited 2024 Nov 14]. Available from: https://satijalab.org/seurat/articles/seurat5_integration
View Article
Google Scholar

[108] View Article

[109] Google Scholar

[ref29] 29. Jiang M, Li C, Cao G, Yang D, Zhang X, Yang L, et al. Effects of NRG1 polymorphisms on Hirschsprung’s disease susceptibility: a meta-analysis. Sci Rep. 2017;7(1):9913. pmid:28855726
View Article
PubMed/NCBI
Google Scholar

[111] View Article

[112] PubMed/NCBI

[113] Google Scholar

[ref30] 30. Jiang Q, Ho Y-Y, Hao L, Nichols Berrios C, Chakravarti A. Copy number variants in candidate genes are genetic modifiers of Hirschsprung disease. PLoS One. 2011;6(6):e21219. pmid:21712996
View Article
PubMed/NCBI
Google Scholar

[115] View Article

[116] PubMed/NCBI

[117] Google Scholar

[ref31] 31. Jin S, Guerrero-Juarez CF, Zhang L, Chang I, Ramos R, Kuan C-H, et al. Inference and analysis of cell-cell communication using CellChat. Nat Commun. 2021;12(1):1088. pmid:33597522
View Article
PubMed/NCBI
Google Scholar

[119] View Article

[120] PubMed/NCBI

[121] Google Scholar

[ref32] 32. Jin S, Plikus MV, Nie Q. CellChat for systematic analysis of cell-cell communication from single-cell transcriptomics. Nat Protoc. 2025;20(1):180–219. pmid:39289562
View Article
PubMed/NCBI
Google Scholar

[123] View Article

[124] PubMed/NCBI

[125] Google Scholar

[ref33] 33. Kaern M, Elston TC, Blake WJ, Collins JJ. Stochasticity in gene expression: from theories to phenotypes. Nat Rev Genet. 2005;6(6):451–64. pmid:15883588
View Article
PubMed/NCBI
Google Scholar

[127] View Article

[128] PubMed/NCBI

[129] Google Scholar

[ref34] 34. Kalra S, Lanno S, Sanchez G, Coolon JD. cis- and trans-regulatory contributions to a hierarchy of factors influencing gene expression variation. Commun Biol. 2024;7(1):1563. pmid:39587248
View Article
PubMed/NCBI
Google Scholar

[131] View Article

[132] PubMed/NCBI

[133] Google Scholar

[ref35] 35. Karczewski KJ, Solomonson M, Chao KR, Goodrich JK, Tiao G, Lu W, et al. Systematic single-variant and gene-based association testing of thousands of phenotypes in 394,841 UK Biobank exomes. Cell Genom. 2022;2(9):100168.
View Article
Google Scholar

[135] View Article

[136] Google Scholar

[ref36] 36. Karolchik D, Hinrichs AS, Furey TS, Roskin KM, Sugnet CW, Haussler D, et al. The UCSC Table Browser data retrieval tool. Nucleic Acids Res. 2004;32(Database issue):D493-6. pmid:14681465
View Article
PubMed/NCBI
Google Scholar

[138] View Article

[139] PubMed/NCBI

[140] Google Scholar

[ref37] 37. Kim J-H, Cheong HS, Sul JH, Seo J-M, Kim D-Y, Oh J-T, et al. A genome-wide association study identifies potential susceptibility loci for Hirschsprung disease. PLoS One. 2014;9(10):e110292. pmid:25310821
View Article
PubMed/NCBI
Google Scholar

[142] View Article

[143] PubMed/NCBI

[144] Google Scholar

[ref38] 38. Kim Y-K, Lee H, Ismail T, Kim Y, Lee H-S. Dach1 regulates neural crest migration during embryonic development. Biochem Biophys Res Commun. 2020;527(4):896–901. pmid:32430182
View Article
PubMed/NCBI
Google Scholar

[146] View Article

[147] PubMed/NCBI

[148] Google Scholar

[ref39] 39. Korsunsky I, Millard N, Fan J, Slowikowski K, Zhang F, Wei K, et al. Fast, sensitive and accurate integration of single-cell data with Harmony. Nat Methods. 2019;16(12):1289–96. pmid:31740819
View Article
PubMed/NCBI
Google Scholar

[150] View Article

[151] PubMed/NCBI

[152] Google Scholar

[ref40] 40. Kuil LE, MacKenzie KC, Tang CS, Windster JD, Le TL, Karim A, et al. Size matters: Large copy number losses in Hirschsprung disease patients reveal genes involved in enteric nervous system development. PLoS Genet. 2021;17(8):e1009698. pmid:34358225
View Article
PubMed/NCBI
Google Scholar

[154] View Article

[155] PubMed/NCBI

[156] Google Scholar

[ref41] 41. Lake JI, Heuckeroth RO. Enteric nervous system development: migration, differentiation, and disease. Am J Physiol Gastrointest Liver Physiol. 2013;305(1):G1-24. pmid:23639815
View Article
PubMed/NCBI
Google Scholar

[158] View Article

[159] PubMed/NCBI

[160] Google Scholar

[ref42] 42. Lane PW, Liu HM. Association of megacolon with a new dominant spotting gene (Dom) in the mouse. J Hered. 1984;75(6):435–9. pmid:6512238
View Article
PubMed/NCBI
Google Scholar

[162] View Article

[163] PubMed/NCBI

[164] Google Scholar

[ref43] 43. Lawrence M, Gentleman R, Carey V. rtracklayer: an R package for interfacing with genome browsers. Bioinformatics. 2009;25(14):1841–2. pmid:19468054
View Article
PubMed/NCBI
Google Scholar

[166] View Article

[167] PubMed/NCBI

[168] Google Scholar

[ref44] 44. Li N, Kelsh RN, Croucher P, Roehl HH. Regulation of neural crest cell fate by the retinoic acid and Pparg signalling pathways. Development. 2010;137(3):389–94. pmid:20081187
View Article
PubMed/NCBI
Google Scholar

[170] View Article

[171] PubMed/NCBI

[172] Google Scholar

[ref45] 45. Li C, Hu R, Hou N, Wang Y, Wang Z, Yang T, et al. Alteration of the retinoid acid-CBP signaling pathway in neural crest induction contributes to enteric nervous system disorder. Front Pediatr. 2018;6:382. pmid:30560112
View Article
PubMed/NCBI
Google Scholar

[174] View Article

[175] PubMed/NCBI

[176] Google Scholar

[ref46] 46. Li Z, Lui KN-C, Lau S-T, Lai FP-L, Li P, Chung PH-Y, et al. Transcriptomics of Hirschsprung disease patient-derived enteric neural crest cells reveals a role for oxidative phosphorylation. Nat Commun. 2023;14(1):2157. pmid:37061531
View Article
PubMed/NCBI
Google Scholar

[178] View Article

[179] PubMed/NCBI

[180] Google Scholar

[ref47] 47. Lippert C, Listgarten J, Liu Y, Kadie CM, Davidson RI, Heckerman D. FaST linear mixed models for genome-wide association studies. Nat Methods. 2011;8(10):833–5. pmid:21892150
View Article
PubMed/NCBI
Google Scholar

[182] View Article

[183] PubMed/NCBI

[184] Google Scholar

[ref48] 48. Liu SJ, Casey-Clyde T, Cho NW, Swinderman J, Pekmezci M, Dougherty MC, et al. Epigenetic reprogramming shapes the cellular landscape of schwannoma. Nat Commun. 2024;15(1):476. pmid:38216587
View Article
PubMed/NCBI
Google Scholar

[186] View Article

[187] PubMed/NCBI

[188] Google Scholar

[ref49] 49. López SH, Avetisyan M, Wright CM, Mesbah K, Kelly RG, Moon AM, et al. Loss of Tbx3 in murine neural crest reduces enteric glia and causes cleft palate, but does not influence heart development or bowel transit. Dev Biol. 2018;444 Suppl 1(Suppl 1):S337–51. pmid:30292786
View Article
PubMed/NCBI
Google Scholar

[190] View Article

[191] PubMed/NCBI

[192] Google Scholar

[ref50] 50. Lorè NI, Sipione B, He G, Strug LJ, Atamni HJ, Dorman A, et al. Collaborative cross mice yield genetic modifiers for Pseudomonas aeruginosa infection in human lung disease. mBio. 2020;11(2):e00097-20. pmid:32127447
View Article
PubMed/NCBI
Google Scholar

[194] View Article

[195] PubMed/NCBI

[196] Google Scholar

[ref51] 51. Lui VCH, Cheng WWC, Leon TYY, Lau DKC, Garcia-Barcelo M-M, Miao XP, et al. Perturbation of hoxb5 signaling in vagal neural crests down-regulates ret leading to intestinal hypoganglionosis in mice. Gastroenterology. 2008;134(4):1104–15. pmid:18395091
View Article
PubMed/NCBI
Google Scholar

[198] View Article

[199] PubMed/NCBI

[200] Google Scholar

[ref52] 52. Majd H, Samuel RM, Cesiulis A, Ramirez JT, Kalantari A, Barber K. Engrafted nitrergic neurons derived from hPSCs improve gut dysmotility in mice. Nature. 2025.
View Article
Google Scholar

[202] View Article

[203] Google Scholar

[ref53] 53. Maka M, Stolt CC, Wegner M. Identification of Sox8 as a modifier gene in a mouse model of Hirschsprung disease reveals underlying molecular defect. Dev Biol. 2005;277(1):155–69. pmid:15572147
View Article
PubMed/NCBI
Google Scholar

[205] View Article

[206] PubMed/NCBI

[207] Google Scholar

[ref54] 54. Mallory SB, Wiener E, Nordlund JJ. Waardenburg’s syndrome with Hirschsprung’s disease: a neural crest defect. Pediatr Dermatol. 1986;3(2):119–24. pmid:3952027
View Article
PubMed/NCBI
Google Scholar

[209] View Article

[210] PubMed/NCBI

[211] Google Scholar

[ref55] 55. May-Zhang AA, Tycksen E, Southard-Smith AN, Deal KK, Benthal JT, Buehler DP, et al. Combinatorial transcriptional profiling of mouse and human enteric neurons identifies shared and disparate subtypes in situ. Gastroenterology. 2021;160(3):755-770.e26. pmid:33010250
View Article
PubMed/NCBI
Google Scholar

[213] View Article

[214] PubMed/NCBI

[215] Google Scholar

[ref56] 56. McCallion AS, Stames E, Conlon RA, Chakravarti A. Phenotype variation in two-locus mouse models of Hirschsprung disease: tissue-specific interaction between Ret and Ednrb. Proc Natl Acad Sci U S A. 2003;100(4):1826–31. pmid:12574515
View Article
PubMed/NCBI
Google Scholar

[217] View Article

[218] PubMed/NCBI

[219] Google Scholar

[ref57] 57. McLaren W, Gil L, Hunt SE, Riat HS, Ritchie GRS, Thormann A. The Ensembl Variant Effect Predictor. Genome Biol. 2016;17(1):122.
View Article
Google Scholar

[221] View Article

[222] Google Scholar

[ref58] 58. Memic F, Knoflach V, Morarach K, Sadler R, Laranjeira C, Hjerling-Leffler J, et al. Transcription and signaling regulators in developing neuronal subtypes of mouse and human enteric nervous system. Gastroenterology. 2018;154(3):624–36. pmid:29031500
View Article
PubMed/NCBI
Google Scholar

[224] View Article

[225] PubMed/NCBI

[226] Google Scholar

[ref59] 59. Mo J, Long R, Fantauzzo KA. Pdgfra and Pdgfrb genetically interact in the murine neural crest cell lineage to regulate migration and proliferation. Front Physiol. 2020;11:588901. pmid:33224039
View Article
PubMed/NCBI
Google Scholar

[228] View Article

[229] PubMed/NCBI

[230] Google Scholar

[ref60] 60. Moghimbeigi A. Two-part zero-inflated negative binomial regression model for quantitative trait loci mapping with count trait. J Theor Biol. 2015;372:74–80. pmid:25728790
View Article
PubMed/NCBI
Google Scholar

[232] View Article

[233] PubMed/NCBI

[234] Google Scholar

[ref61] 61. Nagashimada M, Ohta H, Li C, Nakao K, Uesaka T, Brunet J-F, et al. Autonomic neurocristopathy-associated mutations in PHOX2B dysregulate Sox10 expression. J Clin Invest. 2012;122(9):3145–58. pmid:22922260
View Article
PubMed/NCBI
Google Scholar

[236] View Article

[237] PubMed/NCBI

[238] Google Scholar

[ref62] 62. Nagy N, Goldstein AM. Enteric nervous system development: a crest cell’s journey from neural tube to colon. Semin Cell Dev Biol. 2017;66:94–106. pmid:28087321
View Article
PubMed/NCBI
Google Scholar

[240] View Article

[241] PubMed/NCBI

[242] Google Scholar

[ref63] 63. Owens SE, Broman KW, Wiltshire T, Elmore JB, Bradley KM, Smith JR, et al. Genome-wide linkage identifies novel modifier loci of aganglionosis in the Sox10Dom model of Hirschsprung disease. Hum Mol Genet. 2005;14(11):1549–58. pmid:15843399
View Article
PubMed/NCBI
Google Scholar

[244] View Article

[245] PubMed/NCBI

[246] Google Scholar

[ref64] 64. Parisi MA, Kapur RP, Neilson I, Hofstra RMW, Holloway LW, Michaelis RC, et al. Hydrocephalus and intestinal aganglionosis: is L1CAM a modifier gene in Hirschsprung disease? Am J Med Genet. 2002;108(1):51–6. pmid:11857550
View Article
PubMed/NCBI
Google Scholar

[248] View Article

[249] PubMed/NCBI

[250] Google Scholar

[ref65] 65. Pattyn A, Morin X, Cremer H, Goridis C, Brunet JF. The homeobox gene Phox2b is essential for the development of autonomic neural crest derivatives. Nature. 1999;399(6734):366–70. pmid:10360575
View Article
PubMed/NCBI
Google Scholar

[252] View Article

[253] PubMed/NCBI

[254] Google Scholar

[ref66] 66. Perez G, Barber GP, Benet-Pages A, Casper J, Clawson H, Diekhans M, et al. The UCSC Genome Browser database: 2025 update. Nucleic Acids Res. 2024.
View Article
Google Scholar

[256] View Article

[257] Google Scholar

[ref67] 67. Pingault V, Bondurand N, Kuhlbrodt K, Goerich DE, Préhu MO, Puliti A, et al. SOX10 mutations in patients with Waardenburg-Hirschsprung disease. Nat Genet. 1998;18(2):171–3. pmid:9462749
View Article
PubMed/NCBI
Google Scholar

[259] View Article

[260] PubMed/NCBI

[261] Google Scholar

[ref68] 68. Sato Y, Heuckeroth RO. Retinoic acid regulates murine enteric nervous system precursor proliferation, enhances neuronal precursor differentiation, and reduces neurite growth in vitro. Dev Biol. 2008;320(1):185–98. pmid:18561907
View Article
PubMed/NCBI
Google Scholar

[263] View Article

[264] PubMed/NCBI

[265] Google Scholar

[ref69] 69. Shih C-H, Fay J. Cis-regulatory variants affect gene expression dynamics in yeast. Elife. 2021;10:e68469. pmid:34369376
View Article
PubMed/NCBI
Google Scholar

[267] View Article

[268] PubMed/NCBI

[269] Google Scholar

[ref70] 70. Shull LC, Lencer ES, Kim HM, Goyama S, Kurokawa M, Costello JC, et al. PRDM paralogs antagonistically balance Wnt/β-catenin activity during craniofacial chondrocyte differentiation. Development. 2022;149(4):dev200082. pmid:35132438
View Article
PubMed/NCBI
Google Scholar

[271] View Article

[272] PubMed/NCBI

[273] Google Scholar

[ref71] 71. Simpson MJ, Zhang DC, Mariani M, Landman KA, Newgreen DF. Cell proliferation drives neural crest cell invasion of the intestine. Dev Biol. 2007;302(2):553–68. pmid:17178116
View Article
PubMed/NCBI
Google Scholar

[275] View Article

[276] PubMed/NCBI

[277] Google Scholar

[ref72] 72. Southard-Smith EM, Angrist M, Ellison JS, Agarwala R, Baxevanis AD, Chakravarti A, et al. The Sox10(Dom) mouse: modeling the genetic variation of Waardenburg-Shah (WS4) syndrome. Genome Res. 1999;9(3):215–25. pmid:10077527
View Article
PubMed/NCBI
Google Scholar

[279] View Article

[280] PubMed/NCBI

[281] Google Scholar

[ref73] 73. Stanchina L, Baral V, Robert F, Pingault V, Lemort N, Pachnis V, et al. Interactions between Sox10, Edn3 and Ednrb during enteric nervous system and melanocyte development. Dev Biol. 2006;295(1):232–49. pmid:16650841
View Article
PubMed/NCBI
Google Scholar

[283] View Article

[284] PubMed/NCBI

[285] Google Scholar

[ref74] 74. Stavely R, Hotta R, Guyer RA, Picard N, Rahman AA, Omer M, et al. A distinct transcriptome characterizes neural crest-derived cells at the migratory wavefront during enteric nervous system development. Development. 2023;150(5):dev201090. pmid:36779913
View Article
PubMed/NCBI
Google Scholar

[287] View Article

[288] PubMed/NCBI

[289] Google Scholar

[ref75] 75. Stuart T, Srivastava A, Madad S, Lareau CA, Satija R. Single-cell chromatin state analysis with Signac. Nat Methods. 2021;18(11):1333–41.
View Article
Google Scholar

[291] View Article

[292] Google Scholar

[ref76] 76. Tang CS-M, Gui H, Kapoor A, Kim J-H, Luzón-Toro B, Pelet A, et al. Trans-ethnic meta-analysis of genome-wide association studies for Hirschsprung disease. Hum Mol Genet. 2016;25(23):5265–75. pmid:27702942
View Article
PubMed/NCBI
Google Scholar

[294] View Article

[295] PubMed/NCBI

[296] Google Scholar

[ref77] 77. Taylor MK, Yeager K, Morrison SJ. Physiological Notch signaling promotes gliogenesis in the developing peripheral and central nervous systems. Development. 2007;134(13):2435–47. pmid:17537790
View Article
PubMed/NCBI
Google Scholar

[298] View Article

[299] PubMed/NCBI

[300] Google Scholar

[ref78] 78. Tilghman JM, Ling AY, Turner TN, Sosa MX, Krumm N, Chatterjee S, et al. N Engl J Med. 2019;380(15):1421–32.
View Article
Google Scholar

[302] View Article

[303] Google Scholar

[ref79] 79. Uribe RA, Hong SS, Bronner ME. Retinoic acid temporally orchestrates colonization of the gut by vagal neural crest cells. Dev Biol. 2018;433(1):17–32. pmid:29108781
View Article
PubMed/NCBI
Google Scholar

[305] View Article

[306] PubMed/NCBI

[307] Google Scholar

[ref80] 80. Wakamatsu Y, Uchikawa M. The many faces of Sox2 function in neural crest development. Dev Growth Differ. 2021;63(1):93–9. pmid:33326593
View Article
PubMed/NCBI
Google Scholar

[309] View Article

[310] PubMed/NCBI

[311] Google Scholar

[ref81] 81. Wang F, Hu W, Xian J, Ohnuma S, Brenton JD. The Xenopus Tgfbi is required for embryogenesis through regulation of canonical Wnt signalling. Dev Biol. 2013;379(1):16–27. pmid:23246513
View Article
PubMed/NCBI
Google Scholar

[313] View Article

[314] PubMed/NCBI

[315] Google Scholar

[ref82] 82. Wang B, Fang X, Sun X, Du C, Zhou L, Lv X, et al. m6A demethylase ALKBH5 suppresses proliferation and migration of enteric neural crest cells by regulating TAGLN in Hirschsprung’s disease. Life Sci. 2021;278:119577. pmid:33961858
View Article
PubMed/NCBI
Google Scholar

[317] View Article

[318] PubMed/NCBI

[319] Google Scholar

[ref83] 83. Wright CM, Schneider S, Smith-Edwards KM, Mafra F, Leembruggen AJL, Gonzalez MV, et al. scRNA-seq reveals new enteric nervous system roles for GDNF, NRTN, and TBX3. Cell Mol Gastroenterol Hepatol. 2021;11(5):1548-1592.e1. pmid:33444816
View Article
PubMed/NCBI
Google Scholar

[321] View Article

[322] PubMed/NCBI

[323] Google Scholar

[ref84] 84. Wu K, Katiyar S, Li A, Liu M, Ju X, Popov VM, et al. Dachshund inhibits oncogene-induced breast cancer cellular migration and invasion through suppression of interleukin-8. Proc Natl Acad Sci U S A. 2008;105(19):6924–9. pmid:18467491
View Article
PubMed/NCBI
Google Scholar

[325] View Article

[326] PubMed/NCBI

[327] Google Scholar

[ref85] 85. Yang J, Zaitlen NA, Goddard ME, Visscher PM, Price AL. Advantages and pitfalls in the application of mixed-model association methods. Nat Genet. 2014;46(2):100–6. pmid:24473328
View Article
PubMed/NCBI
Google Scholar

[329] View Article

[330] PubMed/NCBI

[331] Google Scholar

[ref86] 86. Young HM, Jones BR, McKeown SJ. The projections of early enteric neurons are influenced by the direction of neural crest cell migration. J Neurosci. 2002;22(14):6005–18. pmid:12122062
View Article
PubMed/NCBI
Google Scholar

[333] View Article

[334] PubMed/NCBI

[335] Google Scholar

[ref87] 87. Young HM, Bergner AJ, Müller T. Acquisition of neuronal and glial markers by neural crest-derived cells in the mouse intestine. J Comp Neurol. 2003;456(1):1–11. pmid:12508309
View Article
PubMed/NCBI
Google Scholar

[337] View Article

[338] PubMed/NCBI

[339] Google Scholar

[ref88] 88. Young HM, Bergner AJ, Anderson RB, Enomoto H, Milbrandt J, Newgreen DF, et al. Dynamics of neural crest-derived cell migration in the embryonic mouse gut. Dev Biol. 2004;270(2):455–73. pmid:15183726
View Article
PubMed/NCBI
Google Scholar

[341] View Article

[342] PubMed/NCBI

[343] Google Scholar

[ref89] 89. Young HM, Bergner AJ, Simpson MJ, McKeown SJ, Hao MM, Anderson CR, et al. Colonizing while migrating: how do individual enteric neural crest cells behave? BMC Biol. 2014;12:23. pmid:24670214
View Article
PubMed/NCBI
Google Scholar

[345] View Article

[346] PubMed/NCBI

[347] Google Scholar

[ref90] 90. Zhang Y, Liu T, Meyer CA, Eeckhoute J, Johnson DS, Bernstein BE, et al. Model-based analysis of ChIP-Seq (MACS). Genome Biol. 2008;9(9):R137. pmid:18798982
View Article
PubMed/NCBI
Google Scholar

[349] View Article

[350] PubMed/NCBI

[351] Google Scholar

[ref91] 91. Zhang D, Rollo BN, Nagy N, Stamp L, Newgreen DF. The enteric neural crest progressively loses capacity to form enteric nervous system. Dev Biol. 2019;446(1):34–42. pmid:30529057
View Article
PubMed/NCBI
Google Scholar

[353] View Article

[354] PubMed/NCBI

[355] Google Scholar

[ref92] 92. Zhao L, Song W, Chen Y-G. Mesenchymal-epithelial interaction regulates gastrointestinal tract development in mouse embryos. Cell Rep. 2022;40(2):111053. pmid:35830795
View Article
PubMed/NCBI
Google Scholar

[357] View Article

[358] PubMed/NCBI

[359] Google Scholar

[ref93] 93. Zheng GXY, Terry JM, Belgrader P, Ryvkin P, Bent ZW, Wilson R, et al. Massively parallel digital transcriptional profiling of single cells. Nat Commun. 2017;8:14049. pmid:28091601
View Article
PubMed/NCBI
Google Scholar

[361] View Article

[362] PubMed/NCBI

[363] Google Scholar

[ref94] 94. Zhou X, Stephens M. Genome-wide efficient mixed-model analysis for association studies. Nat Genet. 2012;44(7):821–4. pmid:22706312
View Article
PubMed/NCBI
Google Scholar

[365] View Article

[366] PubMed/NCBI

[367] Google Scholar

[ref95] 95. Zhu JJ, Kam MK, Garcia-Barceló M-M, Tam PKH, Lui VCH. HOXB5 binds to multi-species conserved sequence (MCS+9.7) of RET gene and regulates RET expression. Int J Biochem Cell Biol. 2014;51:142–9. pmid:24794774
View Article
PubMed/NCBI
Google Scholar

[369] View Article

[370] PubMed/NCBI

[371] Google Scholar

This is an uncorrected proof.

Figures

Abstract

Author summary

Introduction

Results

Genome-wide SNP analysis identifies Sox10Dom aganglionosis modifier intervals

Assessing potential sex-bias among modifiers of Sox10Dom aganglionosis

Prioritization of candidate genes based on expression in the fetal mouse intestine

Differential expression of modifier interval genes in the migrating wavefront of enteric neural crest-derived cells further prioritizes candidate genes

Evolutionarily conserved SOX10 binding sites within Sox10Dom modifier intervals overlap or are near genes differentially expressed in the migrating wavefront

Chromatin accessibility within enteric neuronal progenitors highlights putative regulatory regions within aganglionosis modifier intervals

Prioritization of top aganglionosis modifier candidate genes across data modalities

Candidate modifiers of aganglionosis related to human GI phenotypes

Discussion

Methods

Ethics statement

Mouse husbandry

Acetylcholinesterase histochemistry

Mouse genotyping

Analysis of Sox10Dom extended pedigree using Genome-wide Efficient Mixed Model Analysis (GEMMA)

Modifier intervals for downstream candidate gene analysis

Reprocessing of whole gut 9.5-15.5 days post coitus scRNA-seq data

Enteric neural crest-derived cell wavefront differential gene expression (DGE) overlapping with modifier intervals

SOX10 conserved binding motifs within modifier intervals

CellChat analysis of cell-cell communication in ENCDCs

Generation and processing of fetal ENS single nucleus assay for transposase-accessible chromatin-sequencing (snATAC-seq) data

Isolation of nuclei from fetal mouse tissue.

Analysis of snATAC-seq data.

Sox10Dom aganglionosis modifier interval candidate gene prioritization pipeline from mouse datasets

Filtering Sox10Dom aganglionosis modifier interval candidate genes for those with intron or exon variants with predicted high impact

Overlap of Sox10Dom aganglionosis modifier interval candidate genes with human HSCR and stool frequency GWAS summary statistics

Extraction of relevant phenotypes from Genebass-sourced Sox10Dom aganglionosis modifier interval candidate gene-based PheWAS

Supporting information

S1 Fig. QQ Plots of Non-Sex-specific and Sex-specific GEMMA genome-wide scans on the Sox10Dom mice population.

S2 Fig. GEMMA genome-wide scans using only those Sox10Dom mice that exhibit detectable aganglionosis.

S3 Fig. Stratification of allele-specific directions of effect by sex reveal sex effects.

S4 Fig. Expression of genes in fetal gut near top associated SNP of nominally sex-biased modifier intervals across developmental time.

S5 Fig. CellChat identifies modifier interval candidates via estimated cell signaling with neural crest cells.

S6 Fig. Expression of Sox10Dom aganglionosis modifier interval candidate genes from wavefront ENCDC bulk RNA-seq and near conserved SOX10 binding motifs.

S7 Fig. Whole gut 16.5dpc Phox2b H2B-CFP + snATAC-seq label transfer and derived putative cis-regulatory elements within modifier intervals.

S8 Fig. Expression of progenitor, neuronal and candidate modifier genes across ENCDCs migration stages.

S1 Table. Tab-separated table with phenotypes of each mouse in the study listed by column, including length of intestine, length of hypoganglionosis and aganglionosis, percentages of hypoganglionosis and aganglionosis, sex of the mice, and mouse individual IDs.

S2 Table. Tab-separated genotype file in a PED-like format (PLINK) with the first columns being individual IDs for each mouse and columns per SNP genotype.

S3 Table. CSV file containing GEMMA scan results for all mice using sex as a covariate and the quantitative percentage aganglionosis phenotype.

S4 Table. CSV file containing GEMMA scan results for all mice using sex as a covariate and the quantitative percentage aganglionosis phenotype but excluding chromosomes 5, 15, and X.

S5 Table. CSV file containing GEMMA scan results for all mice using sex as a covariate and the binary aganglionosis phenotype (either no aganglionosis or yes aganglionosis).

S6 Table. CSV file containing GEMMA scan results for only mice that had aganglionosis using sex as a covariate and the quantitative percentage aganglionosis phenotype.

S7 Table. CSV file containing GEMMA scan results for only female mice that had aganglionosis using sex as a covariate and the quantitative percentage aganglionosis phenotype.

S8 Table. CSV file containing GEMMA scan results for only male mice that had aganglionosis using sex as a covariate and the quantitative percentage aganglionosis phenotype.

S9 Table. CSV file containing GEMMA scan results for male mice only using the quantitative percentage aganglionosis phenotype.

S10 Table. CSV file containing GEMMA scan results for female mice only using the quantitative percentage aganglionosis phenotype.

S11 Table. CSV file containing GEMMA scan results for male mice only using the binary aganglionosis phenotype (either no aganglionosis or yes aganglionosis).

S12 Table. CSV file containing GEMMA scan results for female mice only using the binary aganglionosis phenotype (either no aganglionosis or yes aganglionosis).

S13 Table. CSV file containing comprehensive information of modifier interval definitions derived from significant SNPs in S3–S5 and S9–S12 Tables.

S14 Table. CSV file that lists the genes within each modifier interval that are expressed above a normalized expression threshold of 1 in the Zhao et al. scRNA-seq dataset.

S15 Table. CSV file containing the significant (Bonferroni-correction p-value) differential gene expression results (Seurat’s FindAllMarkers function) for genes within the modifier intervals in the Zhao et al. scRNA-seq dataset.

S16 Table. Excel file containing significant results of the CellChat analysis in the Zhao et al. scRNA-seq dataset.

S17 Table. TSV file containing migrating wavefront versus lagging enteric neural crest-derived cell significant differential gene expression results from Stavely et al. filtered for those within modifier intervals.

S18 Table. CSV file containing cross-species conserved SOX10 transcription factor binding motifs from Gopinath et al. that fall within modifier intervals, along with their closest genomic feature found through the ClosestFeature function from the R package Signac.

S19 Table. CSV file containing significant differential chromatin accessibility results (Seurat’s FindAllMarkers function) from 16.5dpc Phox2b H2B-CFP + ENS snATAC-seq.

S20 Table. Results in S18 Table but filtered for those differentially accessible chromatin regions that lie within modifier intervals.

S21 Table. Overlap of the results in S17 and S19 Tables; those differentially accessible chromatin regions that overlap with conserved SOX10 transcription factor binding motifs.

S22 Table. CSV file containing significantly enriched transcription factor binding motifs derived from differentially accessible chromatin regions contained within modifier intervals (S19 Table) found via the R package Signac’s FindMotifs function.

S23 Table. CSV file containing the results from S21 Table filtered for those transcription factors that are expressed in the Zhao et al. “neural crest” subset of cells.

S24 Table. Excel file containing, for each omics candidate gene, the genetics score information.

S26 Table. TSV containing Ensembl Variant Effect Predictor results for omics candidate genes whose C3HeB/FeJ variants were predicted to have high impact on the corresponding mRNA or protein.

S27 Table. CSV file containing Hirschsprung GWAS summary statistics from Tang et al. filtered by those within EUR (European ancestry group) LD of candidate genes.

S28 Table. CSV file containing Hirschsprung GWAS summary statistics from Tang et al. filtered by those within ASN (East Asian ancestry group) LD of candidate genes.

S29 Table. CSV file containing Hirschsprung GWAS summary statistics from Garcia-Barcelo et al. filtered by those within ASN (East Asian ancestry group) LD of candidate genes.

S30 Table. CSV file containing Stool Frequency GWAS summary statistics from Bonfiglio et al. filtered by those within EUR (European ancestry group) LD of candidate genes.

S31 Table. TSV file containing concatenated Genebass.org PheWAS results from this study’s list of candidate genes’ human orthologs.

Acknowledgments

References

Genome-wide SNP analysis identifies Sox10^Dom aganglionosis modifier intervals

Assessing potential sex-bias among modifiers of Sox10^Dom aganglionosis

Evolutionarily conserved SOX10 binding sites within Sox10^Dom modifier intervals overlap or are near genes differentially expressed in the migrating wavefront

Analysis of Sox10^Dom extended pedigree using Genome-wide Efficient Mixed Model Analysis (GEMMA)

Sox10^Dom aganglionosis modifier interval candidate gene prioritization pipeline from mouse datasets

Filtering Sox10^Dom aganglionosis modifier interval candidate genes for those with intron or exon variants with predicted high impact

Overlap of Sox10^Dom aganglionosis modifier interval candidate genes with human HSCR and stool frequency GWAS summary statistics

Extraction of relevant phenotypes from Genebass-sourced Sox10^Dom aganglionosis modifier interval candidate gene-based PheWAS

S1 Fig. QQ Plots of Non-Sex-specific and Sex-specific GEMMA genome-wide scans on the Sox10^Dom mice population.

S2 Fig. GEMMA genome-wide scans using only those Sox10^Dom mice that exhibit detectable aganglionosis.

S6 Fig. Expression of Sox10^Dom aganglionosis modifier interval candidate genes from wavefront ENCDC bulk RNA-seq and near conserved SOX10 binding motifs.