Gene–environment interactions are believed to play a role in multifactorial phenotypes, although poorly described mechanistically. Cleft lip/palate (CLP), the most common craniofacial malformation, has been associated with both genetic and environmental factors, with little gene–environment interaction experimentally demonstrated. Here, we study CLP families harbouring CDH1/E-Cadherin variants with incomplete penetrance and we explore the association of pro-inflammatory conditions to CLP. By studying neural crest (NC) from mouse, Xenopus and humans, we show that CLP can be explained by a 2-hit model, where NC migration is impaired by a combination of genetic (CDH1 loss-of-function) and environmental (pro-inflammatory activation) factors, leading to CLP. Finally, using in vivo targeted methylation assays, we demonstrate that CDH1 hypermethylation is the major target of the pro-inflammatory response, and a direct regulator of E-cadherin levels and NC migration. These results unveil a gene–environment interaction during craniofacial development and provide a 2-hit mechanism to explain cleft lip/palate aetiology.
The concept of gene–environment interactions has been present in genetics for decades1,2,3,4,5,6, however with scarce demonstration of how this interaction occurs in human traits, especially in regards to development. Gene–environment interactions studies have mostly focused on environmental insults during development leading to maladaptive development and metabolic disorders2,3,4, with rare contributions to human malformation.
Cleft lip with or without cleft palate (CLP), the most frequent craniofacial malformation in humans (1:700 live-births) has been widely described as a multifactorial condition affecting craniofacial development7,8,9. Although CLP heritability studies corroborate the role of both genetic and environmental factors in its aetiology, with heritability ranging from 45 to 85% depending on the analysed population10,11, most studies rely either on associating genetic variants to CLP by sequencing strategies or on epidemiological association of environmental variables to CLP8,9,12,13.
Candidate genes associated with CLP tend to fall into the Epithelial-to-Mesenchymal transition (EMT) pathway and pathogenic variants in genes belonging to the cadherin-catenin complex have been linked to familial CLP14,15,16,17,18. CDH1 (E-cadherin), a member of the cadherin-catenin complex, is an example in which pathogenic variants can lead to CLP and/or gastric cancer. Despite the progress made on the identification of predisposing genetic variation to CLP, which includes both rare pathogenic variants as well as common single nucleotide polymorphisms (SNPs), those variants fail to completely explain CLP heritability, as in many complex disorders19,20,21. Alongside this missing heritability scenario, a similar phenomenon is also displayed in familial CLP harbouring rare pathogenic variants, in which incomplete penetrance is often observed and variants do not fully explain the phenotype, which remains an old question still not fully answered22,23,24,25,26,27,28. Incomplete penetrance is also the case for CDH1-linked CLP17, suggesting additional factors may play a role in this malformation.
On the other hand, epidemiological studies have associated pro-inflammatory conditions as diabetes, maternal infections, and obesity to CLP29,30,31,32,33,34,35,36,37, however, with no experimental demonstration of how those risk factor could affect craniofacial development. Concurrently, there is evidence that pro-inflammatory activation modulates CDH1/E-cadherin levels in multiple cell models38,39,40,41,42,43,44,45.
Therefore, we have explored the gene–environment interface leading to impairments in craniofacial development taking into account four key observations: (1) CLP families harbouring CDH1/E-cadherin pathogenic variants exhibit incomplete penetrance; (2) epidemiological associations linking maternal pro-inflammatory conditions to a higher risk of CLP in the offspring12,13,14,15,16,17,18,19,20; (3) cellular and molecular evidence of CDH1/E-cadherin downregulation under pro-inflammatory cytokine or bacteria exposure21,22,23,24,25,26; and (4) CDH1-linked familial gastric cancer, in which CDH1 promoter hypermethylation acts as a second hit for penetrance27,28. We therefore hypothesised a 2-hit model in which CLP penetrance is dependent on a genetic hit (CDH1 loss-of-function) and an environmental hit (pro-inflammatory activation) leading to CDH1 epigenetic variation. In this study, we investigated craniofacial development-associated phenotypes resulting from the combination of gene loss-of-function with environmental insults using both in vitro and in vivo models.
Neural crest E-cadherin loss leads to clefting phenotype
Based on the observation that families harbouring CDH1/E-cadherin pathogenic variants show CLP with incomplete penetrance (~53%) (Fig. 1a, Supplementary Fig. 1, Supplementary Table1) we decided to explore the role of CDH1/E-cadherin on CLP ethology. Although most of craniofacial development is dependent on neural crest46,47,48 a requirement for E-cadherin during cranial neural crest migration has remained controversial49,50. We asked whether a lack of E-cadherin in the neural crest is compatible with CLP phenotypes. Using mice expressing Cre recombinase under the neural crest-specific Wnt1 promoter (Wnt1-Cre2) and floxed Cdh1-LoxP, we found cleft palate (100% penetrance, n = 5) and wider lip gaps (80% penetrance, 4 out of n = 5) in Wnt1-Cre2/Cdh1flox/flox E15.5 mouse embryos (Fig. 1b; Supplementary Fig. 2). Double heterozygous (Wnt1-Cre2/Cdh1flox/+) or floxed homozygous lacking Wnt1-Cre2 (Cdh1flox/flox) did not shown any observable craniofacial phenotype. Such findings show that oral clefting phenotypes linked to CDH1 loss-of-function could be neural crest related. To study neural crest development in vivo we used a previously characterised E-cadherin morpholino knockdown in Xenopus laevis50 by injecting blastomeres which give rise mainly to neural crest. Neural crest depleted of E-cadherin also exhibited craniofacial phenotypes affecting mainly upper and lower jaw cartilages in 93% of morphant Xenopus larvae (Fig. 1c). These phenotypes in neural crest derivatives could be due to problems in neural crest formation, migration, or differentiation. In order to test whether such phenotypes were neural crest migration related, we performed E-cadherin morpholino knockdown in the neural crest in Xenopus and observed significant reduction in in vivo migration (Fig. 1f–h). Additionally, E-cadherin morpholino co-injected with human wild type E-cadherin mRNA was able to rescue neural crest migration, whereas co-injection with E-cadherin mRNAs bearing the familial CLP variants here reported fail to rescue neural crest migration in vivo (Supplementary Fig. 3). To test whether these observations in animal models can be reproduced in human neural crest cells, we performed in vitro migration assays using human induced neural crest cells (hiNCCs) from iPSCs with CRISPR-Cas9 edited for CDH1 loss-of-function. Using scratch assays in hiNCCs and CDH1 loss-of-function (ECAD KO/21del ECAD), we observed that hiNCCs migration was significantly impaired by CDH1 loss (Fig. 1d–e), consistent with the reduced migration observed in Xenopus neural crest (Fig. 1f–h). Therefore both in vivo (Xenopus embryos) and in vitro (human cells) manipulations of CDH1/E-cadherin in the neural crest presented significant neural crest migration reduction (Fig. 1d–h). Taken together, these results demonstrate that E-cadherin loss-of-function affects neural crest migration and that such neural crest E-cadherin loss is compatible with craniofacial phenotypes such as CLP.
Pro-inflammatory activation impairs neural crest migration
We next evaluated whether pro-inflammatory activation could affect in vivo neural crest migration. Importantly, neural crest cells have been reported to respond to and produce pro-inflammatory cytokines both in vitro and in vivo37,51,52,53. Exposing pre-migratory neural crest Xenopus embryos (stage 15) to a well-characterised pro-inflammatory stimulus combination of lipopolysaccharide (LPS) and adenosine triphosphate (ATP)54,55) impairs neural crest migration (Fig. 2a, b; Supplementary Fig. 4a). We tested the pro-inflammatory response by showing an upregulation of tnfa mRNA in the neural crest of LPS + ATP treated embryos, as well as pro-inflammatory cytokines il1b, il6 and NFkB component nfkb1 (Fig. 2c, Supplementary Fig. 5a). Similar reduction in neural migration is also observed when TNFa protein is injected into the neural crest region of Xenopus embryos (Fig. 2d; Supplementary Fig. 4b, c). To test the role of the TNFa pathway in neural crest migration, Xenopus embryos depleted of TNFa or TNFR were treated with LPS and ATP (Fig. 2e). The inhibition of neural crest migration induced by LPS/ATP treatment was reversed by TNFa/TNFR inhibition (Fig. 2f, g), showing that the proinflammatory response is TNFa-dependent. In our in vivo pro-inflammatory model, we also observed that LPS + ATP is more efficient in upregulating tnfa mRNA than LPS or ATP alone (Supplementary Fig. 5b). To test if the in vivo migratory phenotypes in the neural crest were cell autonomous, we directly treated isolated neural crest with LPS + ATP. While control neural crest explants dispersed as expected, LPS + ATP treated cells show reduced dispersion and speed of migration (Fig. 2h–k), indicating that neural crest cells can directly respond to the proinflammatory stimulus. Both in vivo and in vitro phenotypes led us to conclude that the neural crest can respond to external pro-inflammatory stimuli by cytokine upregulation and changes in migratory behaviour.
Neural crest E-cadherin is downregulated by pro-inflammation
Pro-inflammatory activation either by bacterial infection or direct pro-inflammatory cytokine exposure can lead to E-cadherin downregulation in a variety of cellular models38,39,40,41,42,43. Therefore, we evaluated cdh1/E-cadherin expression in the Xenopus migrating neural crest and observed a reduction in both protein E-cadherin and cdh1 mRNA levels upon the pro-inflammatory treatment (Fig. 3a–d), with no effect on the expression of the neural crest marker snai2 and upregulation of the pro-inflammatory factor tnfa, as expected (Fig. 3d). Because E-cadherin knockdown in the Xenopus neural crest led to reduced migration and LPS + ATP treatment resulted in E-cadherin downregulation, we tested whether pro-inflammatory phenotypes could be E-cadherin related. We therefore overexpressed cdh1 by cdh1 mRNA injection in the neural crest precursor blastomeres and exposed embryos to LPS + ATP, followed by neural crest migration analysis. cdh1 mRNA overexpression was able to rescue the neural crest migratory phenotype in comparison to non-injected embryos (Fig. 3e, f). This result indicates that E-cadherin expression is the major downstream target of pro-inflammatory activation with respect to neural crest migration; however, we cannot rule out that other genes not analysed in this study could also play a role in the pro-inflammatory phenotype, as other cadherins are known to be dysregulated upon pro-inflammatory activation, including CDH2 (coding for N-cadherin)56,57. We did not, however, observe any significant changes in neural crest cdh2 mRNA upon LPS + ATP exposure (Supplementary Fig. 5a). Taken together, our results demonstrate that pro-inflammatory activation hampers neural crest migration through downregulation of CDH1/E-cadherin, the absence of which is associated with CLP.
A double hit model affecting neural crest migration
Our hypothesis predicts a 2-hit interaction between a CDH1 loss-of-function allele plus pro-inflammatory activation leading to a further reduction in CDH1/E-cadherin levels, together causing impaired neural crest migration. To test this hypothesis, we applied three different approaches, analysis of Xenopus neural crest in vivo, human hiNCCs in vitro and analysis of mouse neural crest in vivo. Firstly, treatment of Xenopus embryos with reduced concentrations of LPS + ATP or knockdown of E-cadherin via injection with a morpholino did not affect neural crest migration (Fig. 4a, LPS + ATP low, ECAD MO low), whereas a combination of the same levels of ECAD MO and LPS + ATP significantly impaired neural crest migration in vivo (Fig. 4a–c). Secondly, using the in vitro hiNCCs with either TNFa exposure, CDH1 heterozygous knockout (ECAD KO/21del ECAD) or a combination of both (ECAD KO + TNFa), showed that the combination of ECAD KO + TNFa resulted in reduced cell migration (wound-healing closure) and lower CDH1 mRNA levels (Fig. 4d–f). To confirm the requirement of CDH1/E-cadherin on neural crest migration we performed transwell assays of wild type and three different CDH1 heterozygous knockouts (21del ECAD, 7del ECAD, 11del ECAD) hiNCCs, showing decreased migration when cells are exposed to TNFa, and even further reduction in mutant compared to wild type cells (Supplementary Fig. 6). Concordantly, E-cadherin protein levels are significantly reduced in CDH1 heterozygous knockouts hiNCCs in comparison to control, with lower levels when those cells are exposed to TNFa (Supplementary Fig. 7). No effect in the specification of the cells as neural crest was observed under our treatments as all hiNCCs exhibited >95% SOX10 and SOX9 positive nuclei (Supplementary Fig. 8). In addition, no significant change in caspase 3/7 activity was observed in those cells with or without TNFa exposure, in comparison to control (Supplementary Fig. 9), indicating that our treatments are not inducing apoptosis. To further characterise the effects of exposing hiNCCs to TNFa, the expression of several genes was analysed by qPCR. Pro-inflammatory cytokines TNFa and IL6 are upregulated upon TNFa treatment during the differentiation of hiPSC to hiNCC, and CDH1 is reduced; neural border (PAX3 and TFAP2a), EMT (ZEB2, TWIST1 and SNAIL2) and neural crest genes (SOX9 and ETS1) are mostly unperturbed (Supplementary Fig. 10), confirming that the pro-inflammatory treatment is not affecting specification or differentiation of neural crest cells.
Lastly, we explored the 2-hit model using Wnt1-Cre2/Cdh1flox/+ mouse and pro-inflammatory activation through maternal LPS injections. If true, our hypothesis predicts that embryonic lack of E-cadherin together with maternal pro-inflammatory activation would lead to impaired neural crest migration. To gain insight in this model, following the mating of Wnt1-Cre2 females and Cdh1flox/flox males, pregnant females were injected with LPS at 8 and 9 days post-coitum inducing cytokine overexpression in pregnant females and embryos collected at E10.5 for neural crest analysis. First, homozygous embryos lacking E-cadherin (Wnt1-Cre2 Cdh1flox/flox) exhibited reduced neural crest migration compared with wild type embryos (Fig. 5a, top panels); whereas heterozygous E-cadherin embryos exhibited normal neural crest migration (Fig. 5a, Wnt1-Cre2 Cdh1flox/+). However, when these heterozygous embryos were exposed to maternal immune activation a reduced neural crest migration was observed (Fig. 5, Wnt1-Cre2 Cdh1flox/+ + LPS), supporting the 2-hit model.
Taken together, our results support the 2-hit model in which perturbation of CDH1/E-cadherin during development can be achieved by both genetic and environmental insults in a complementary fashion. Although not fully explored here, it is possible that pro-inflammatory factors could modulate phenotype penetrance in CDH1-linked CLP families. In addition, our results support the notion that a precise level of E-cadherin is required for neural crest migration, consistent with observations on other developmental processes and in cancer cells58,59,60,61,62.
E-cadherin hypermethylation is induced by pro-inflammation
The modulation of genome function in response to environmental changes is frequently established via epigenetic modifiers63. In this regard, pro-inflammatory activation is known to modulate gene function via epigenetic changes, including DNA methylation64, 65. Furthermore, we have previously demonstrated that penetrant individuals in CDH1-linked CLP families display higher CDH1 promoter DNA methylation levels in comparison to non-penetrant and family control individuals and reanalysed this previous data (Supplementary Fig. 11)66. Taking that into account, along with CDH1 promoter hypermethylation under infectious/pro-inflammatory conditions previously reported43, we set out to investigate if pro-inflammatory activation in the neural crest could induce such epigenetic changes. Using neural crest explants exposed to LPS + ATP, we observed a significant increase in Xenopus cdh1 promoter methylation in comparison to untreated samples, in which the pro-inflammatory condition induced almost full methylation of the cdh1 promoter (Fig. 6a, b, Supplementary Fig. 12a). Such increased DNA methylation upon pro-inflammatory activation is accompanied with DNA methyltransferase dnmt3a upregulation in the neural crest (Supplementary Fig. 13). Accordingly, hiNCCs exposed to TNFa also displayed increased CDH1 promoter methylation, both in wild-type and ECAD KO cells (Fig. 6c, Supplementary Fig. 12b). To test the effect of this methylation on E-cadherin expression and neural crest migration we induced cdh1 hypermethylation targeting the neural crest using dCas9-DNMT3A-GFP mRNA and cdh1 promoter sgRNA injections in vivo (Fig. 6d, Supplementary Fig. 14). dCas9-DNMT3A-GFP plus sgRNA injected neural crest presented higher cdh1 promoter methylation levels, as well as reduced cdh1 mRNA, in comparison to controls and dCas9-DNMT3A-GFP without sgRNA (Fig. 6e, f). Using the same experimental design, we evaluated in vivo neural crest migration under cdh1 promoter targeted hypermethylation and we observed significant decrease in neural crest migration under dCas9-DNMT3A-GFP plus sgRNA injections (Fig. 6g, h). We therefore concluded that cdh1 promoter hypermethylation is sufficient to induce neural crest migratory deficiency.
E-cadherin demethylation rescues pro-inflammatory migration
If pro-inflammatory activation modulates neural crest cdh1 levels via DNA methylation, inhibiting cdh1 promoter methylation during LPS + ATP exposure would rescue the neural crest migratory phenotype. To test this, we used a dCas9-TET1-GFP construct for targeted cdh1 demethylation with the same sgRNA previously used (Fig. 7a, Supplementary Fig. 14). Under LPS + ATP pro-inflammatory activation, dCas9-TET1_GFP plus sgRNA injections are efficient in reducing cdh1 methylation levels in the neural crest, as well as upregulating cdh1 transcript levels (Fig. 7b, c). Most importantly, dCas9-TET1-GFP plus sgRNA injections were able to rescue in vivo neural crest migration under pro-inflammatory conditions (Fig. 7d, e). Hence, both cdh1-targeted hypermethylation and hypomethylation assays led us to conclude that cdh1 epigenetic control via DNA methylation is likely the major consequence of pro-inflammatory activation in the neural crest (Fig. 7f).
We demonstrated a 2-hit model for CDH1-linked CLP in which gene loss-of-function, in combination with an environmental hit, namely pro-inflammatory activation, leads to CDH1 hypomorphism in the neural crest, affecting its migratory properties and resulting in craniofacial malformations.
Pro-inflammatory activation during mouse pregnancy, or maternal immune activation, has been reported to affect neuronal development in mouse embryos, which has been linked to a model of higher susceptibility to the autistic spectrum disorder in humans67,68,69,70. Despite several contributions in this field, the study of the effects of pro-inflammatory activation in craniofacial development is very limited and this work links pro-inflammatory activation during development affecting the neural crest. We have hypothesised CDH1/E-cadherin as a central node affected by pro-inflammatory activation and our results corroborate this idea. We believe, however, that pro-inflammatory activation might affect other genes and pathways that may also contribute to the phenotypes we observed, and such ideas should be investigated.
Importantly, our study corroborates the need of E-cadherin during proper neural crest migration, both in Xenopus as in mice embryos, as well as in human iNCCs. We do not address, however, which mechanism the lack of E-cadherin affects neural crest migration by. It has been reported that E-cadherin lack in migratory cells could compromise cell cortex tension, cell polarity or focal adhesion stability50, 59,71 and such mechanisms could be involved in the phenotypes here reported. Noteworthily, the experiments conducted here using Xenopus embryos aim to target the neural crest by microinjections in specific blastomeres which give rise mainly, but not exclusively, to the neural crest. This raises the question on whether the observed in vivo migratory phenotypes were due to neural crest specific effects or to affected surrounding tissues. To shed light into this question, we have also observed similar reduction in neural crest dispersion using neural crest explants, which demonstrates that neural crest can directly respond to pro-inflammatory factors, reenforcing the idea that such phenotypes are neural crest specific.
Apart from the cellular and developmental phenotypes observed, by using both Xenopus and mouse in vivo models for E-cadherin loss-of-function in the neural crest we were able to reproduce craniofacial phenotypes comparable to those in humans. While lip and palatal structures arise from distinct embryonic structures in the frog than in mammals, our results suggest that the lack of E-cadherin affecting neural crest migration can generate malformed craniofacial cartilages in Xenopus larvae. Similarly, mouse oral clefting here described are comparable to those previously reported in models for the Cadherin-catenin complex mutations18.
Gene–environment interactions are frequently manifested via epigenetic variation63,72,73, and here we demonstrated that CDH1 promoter hypermethylation acts as a mechanism through which pro-inflammatory activation affects gene expression and ultimately cell behaviour during development. Interestingly, we have observed increased DNA methyltransferase dnmt3a expression in Xenopus neural crest under LPS-ATP treatment, suggesting pro-inflammatory activation can modulated such genes as reported elsewhere74,75 and could be part of the mechanism by which CDH1 promoter is hypermethylated. Although we acknowledge the limitation of not evaluating other loci than CDH1 promoter for DNA methylation upon pro-inflammatory activation, our data again suggest a central role for CDH1 methylation, as revealed by our targeted DNA methylation and demethylation experiments Xenopus neural crest, in which cdh1 methylation control was sufficient to either produce or rescue migratory phenotypes.
Despite widely considered a multifactorial malformation, gene–environment interactions reported in CLP are rarely demonstrated. This is the case of Msx1 mutation in mice, which has been reported to interact with hypoxia and phenytoin during development and increase penetrance of CLP in the offspring76. In this sense, our work explores this multifactorial component of CLP by the investigation of E-cadherin hypomorphism and pro-inflammatory activation during development.
CLP is one of the most common human malformations, which opens the questioning on whether causal factors, either genetic and/or environmental, should be also common in our population. From this perspective, pro-inflammatory conditions during pregnancy could be a common environmental factor, in which maternal infections, diabetes and obesity are potential common causes. We here validate the effects of inflammation on craniofacial development by neural crest migration impairment, opening a route for further investigation of the impacts of pro-inflammatory conditions during development.
Not less important, this study approaches gene–environment interactions by exploring CLP incomplete penetrance. Incomplete penetrance, a frequent phenomenon in autosomal dominant traits, is usually attributed to undetected genetic variants or environmental modulation22,77. Here we demonstrate that pro-inflammatory activation could be an environmental factor modulating CLP penetrance. Such findings therefore shed light on how multifactorial conditions can arise from the combination of genetic and environmental factors as well as how incomplete penetrance could be explained for CDH1-linked CLP.
The use of human data and samples in this study was approved by the Human Research Ethics Committee from the Biosciences Institute (University of Sao Paulo, Brazil) under the protocol 363.876. Biological samples and data regarding from all human research participants were obtained under informed and signed consents from those individuals or legal guardians. The study design and conduct complied with all relevant regulations regarding the use of human study participants and was conducted in accordance with the criteria set by the Declaration of Helsinki.
Mice experiments and procedures were approved by the Animal Research Ethics Committee from the Biosciences Institute (University of Sao Paulo, Brazil) under the protocol 353/2019.
Animal licenses were approved by the Animal Welfare and Ethical Review Board (WERB) at University College London and the UK Home Office. All animal procedures were performed under the ethics standards established by the UK Home Office.
CDH1 sequencing analysis
Targeted sequencing was performed for families F10626 (individuals II-2, III-1 and III-2), F1842 (III-2) and F8250 (III-2 and IV-1; Supplementary Fig. 1). Libraries were prepared with Ampliseq™ custom panel (Illumina), according to the manufacturer’s instructions, and sequenced using the Illumina MiSeq platform, with MiSeq® Reagent Kit v2 (300 cycle). A mean exon coverage of 503x was achieved (±218 standard deviation), with 98.9% of bases presenting coverage above 20x. Exome and Sanger sequencing performed for family F3788 (Fig. 1a) are described elsewhere17. Human CDH1 isoform NM_004360.3 was adopted as reference for variant nomenclature. Penetrance in those families was estimated using a model previously described78.
Neural crest Cdh1 conditional knockout mice
Mice were kept in cages with water and food ad lib, in rooms with controlled temperature and illumination (12 h/12 h light dark cycle), until the moment of euthanasia. Neural crest specific Cdh1 knockout mice was achieved by the use of Wnt1-Cre2 mice (129S4.Cg-E2f1Tg(Wnt1-cre)2Sor/J), which expresses Cre-recombinase in neural crest cells (Strain #:022137, The Jackson Laboratory, USA) and LoxP flanked Cdh1 mice (B6.129-Cdh1tm2Kem/J; Strain #:005319, The Jackson Laboratory, USA). Both mice strains were crossed to obtain Wnt1-Cre2/Cdh1fl/+. Female and male mice were Wnt1-Cre2/Cdh1fl/+crossed to obtain Wnt1-Cre2/Cdh1fl/fl embryos. Observation of copulatory plugs were made and counted as embryonic day 0.5 (E 0.5). For clefting phenotype, once pregnant mice reached E 15.5, females were euthanized, and embryos collected in PBS and fixed in formalin for further analyses. For pro-inflammatory activation via maternal immune activation in pregnant females, at days 8 and 9 post-coitum females were injected with 100 μg/Kg LPS and females were euthanized at day 10, and embryos collected in PBS and fixed in formalin for further analyses. Adult mice and embryos were genotyped by PCR following The Jackson Laboratories recommendations. As controls, wild-type JB6 or Wnt1-Cre2 E15.5 embryos were used.
Xenopus laevis embryo collection
Xenopus laevis embryos were obtained as previously described79. In brief, ovulation of mature 2–5-year-old females (obtained from NASCO, USA) was induced by injecting 100 IU pregnant mare serum gonadotropin (Intervet) subcutaneously into the dorsal lymph sac. Then, 72 h later, a second injection of 200-300 IU human chorionic gonadotropin (Chorulon, Intervet) was performed. Eggs were fertilised in vitro by mixing with a sperm solution. Testes were provided by the European Xenopus Resource Centre. Embryos were staged according to Nieuwkoop and Faber80. Fertilised eggs were dejellied in a solution of 1 g l-cysteine (Sigma-Aldrich) and 500 μl 5 N NaOH in 50 ml H2O and maintained in 0.1× MMR or 3/8 normal amphibian media (NAM) at 14.5 °C.
Morpholino, mRNA and sgRNA microinjections
E-cadherin translation blocking morpholino (ECAD MO) was obtained from GeneTools (USA) as previously described49. TNFa and TNFR translation blocking morpholinos were also obtained from GeneTools (USA) and sequences are available at Supplementary Table 2. cdh1 mRNA was obtained using a cloned Xenopus cdh1 cDNA construct previously published50 and human CDH1 construct (Addgene, plasmid #28009) were used, and in vitro transcription of 5′ capped mRNAs was performed using mMessage mMachine T7 transcription kit (ThermoFisher Scientific). Human CDH1 pathogenic variants c.760G>A and c.2351G>A were added to the original human CDH1 construct by using Q5® Site-Directed Mutagenesis Kit (New England Biolabs), following the fabricant recommendations and confirmation by Sanger sequencing. dCas9-DNMT3A-GFP and dCas9-TET-GFP constructs were obtained from Addgene (pdCas9-DNMT3A-EGFP, plasmid #71666 and pPlatTET-gRNA2, plasmid #82559, respectively). In brief, inserts from both plasmids were subcloned into pCS2+ backbones and in vitro transcription was carried out using T7 RiboMAX™ Express Large Scale RNA Production System (Promega). Transcripts were in vitro 5′capped and polyadenylated using Vaccinia Capping System (NEB) and E. coli Poly(A) Polymerase (NEB), respectively. sgRNA for Xenopus cdh1 promoter was designed using ChopChop (https://chopchop.rc.fas.harvard.edu/) and synthesised using Precision gRNA Synthesis Kit (ThermoFisher Scientific). cdh1 sgRNA sequence is available at Supplementary Table 2.
Microinjections were performed using 5 nl solution in a calibrated needle at 8-cell stage embryos into 2 dorsal blastomeres on the right side, while embryos were into NAM 3/8. Eight nanograms of either ECAD MO, TNF MO or TNFR MO were used for microinjections. For ECAD MO low conditions, 2 ng of morpholino was used. For rescue experiments, 50 pg of either cdh1 mRNA or human CDH1 mRNAs were used. For cdh1 promoter targeted methylation and demethylation assays, 300 pg of dCas9-DNMT3A-GFP or dCas9-TET-GFP were injected, with or without 10 pg of cdh1 sgRNA. Xenopus embryos or incubated at 14.4 °C until developmental stages 20, 24, 25 or 52 for further analyses.
Microtomography (uCT) scanning and 3-dimensional reconstructions
E15.5 mouse embryos or stage 52 Xenopus larvae were used for uCT scan and observation of craniofacial structures. In both cases, embryos were fixed in 4% PFA for 2 h at room temperature and serialy dehydrated in ethanol/PBS. Next, embryos were incubated in 100% Lugol solution (Sigma-Aldrich) overnight at 4 °C. After a 5 min wash in distilled water, embryos were placed in 50 ml plastic tubes and scanned on a Nikon CT Scanner, using 45 mV and 7 mA. Reconstructions were made using CT Pro Scan software and analysed using Voxel software. Measurement of structures was performed on FIJI v.1.53k. Images were coloured using Adobe Photoshop.
Induced neural crest cells (iNCCs), immunostaining and caspase 3/7 activity measures
iNCCs were generated from healthy induced-pluripotent stem cells (iPSC) and characterised as previously described81. iPSCs used in this study were NIBSC8, obtained with the National Institute for Biological Standards and Control (UK). Briefly, iPSCs grown in E8 media (ThermoFisher Scientific) were dissociated as single cells using Tryple Select (ThermoFisher Scientific) and plated at 20.000 cells/cm2 and incubated on E6 media (ThermoFisher Scientific) for 4 days to generate neural border cells (NBCs). Next, NBCs were dissociated to single cell and plated at 20.000 cells/cm2 density in E6 media added with 1 μM CHIR99021 (Sigma-Aldrich), 20 μM SB431542 (Tocris) and 8 ng/ml basic FGF (ThermoFisher Scientific) for 15 days. Rock-inhibitor Y-27632 (Sigma-Aldrich) was always added at 5uM when cells were harvested.
For CDH1 heterozygous knockouts (21del ECAD, 11del ECAD and 7del ECAD) iNCCs generation, iPSCs were transfected with dual expressing Cas9 and sgRNA vector pSpCas9(BB)−2A-Puro (Addgene plasmid #62988) containing a sgRNA for CDH1 exon 3 (Supplementary Table 2). Transfected iPSCs were selected with 1 μg/ml Puromycin for 5 days and survivor colonies were manually isolated for expansion. iPSCs were then Sanger sequenced and frameshift deletions were identified in 3 of the colonies and therefore selected for further assays. iNCCs differentiation was performed as described above.
For immunostaining of those cells, cells were fixed on PFA 4% for 15 min at room termperature and blocked with 2% BSA (PBST) for 1 h at room temperature. Antibody incubation was performed overnight at 4 °C in blocking solution and then cells were washed three times with PBST prior to secondary antibody incubatior. Secondary antibody (AlexaFluor 488 or AlexaFluor 594, Thermofisher Scientific) incubation was performed in blocking solution for 1 h at room temperature using 1:1000 dilution. Primary antibodies used for NCCs were goat anti human/mouse E-cadherin (1:50, AF748, R&D Systems), goat anti human Sox9 (1:100, AB5535, Sigma), mouse anti human Sox10 (1:50, PCRP-SOX10-1D8, DSHB), used following supplier recommendations. Cells were imaged using an EVOS m7000 microscope (Thermofisher Scientific). For apoptotic level quantification in the mentioned iNCCs, we have quantified caspase 3/7 activity by pro-luminescence using the Caspase-Glo® 3/7 Assay System (Promega) following the fabricant recommendations.
In situ hybridisation and immunostaining
Colorimetric whole-mount in situ hybridisation was performed as previously described82. Briefly, snai2 or twist (for Xenopus embryos) and Sox10 (for mouse embryos, a gift from Vassilis Pachnis lab) digoxigenin riboprobes was transcribed using the Riboprobe in vitro Transcription System (Promega, P1420). Embryos were fixed in MEMFA, bleached in 6% hydrogen peroxide, and then incubated with probes for snai2 or twist overnight in hybridisation buffer. Embryos were then washed, blocked with 2% blocking reagent, incubated with 1:3000 anti-digoxigenin-AP antibodies (Roche), and then revealed using NBT/BCIP (Roche) with AP buffer. Embryos were imaged using the Nikon SMZ800N attached to the DS-Fi3 camera (Nikon DS-L4 v.184.108.40.206).
Fluorescent in situ hybridisation was performed as previously described81. In brief, embryos were fixed in MEMFA and incubated with snai2 probe overnight in hybridisation buffer. Embryos were washed, bleached in 3% hydrogen peroxide, incubated with 1:1000 anti-digoxigenin-POD antibody. After washing, embryos underwent the fluorescent POD reaction with Cy5-tyramide solution. For subsequent immunostaining, embryos were dehydrated overnight in 15% sucrose and embedded in 15% sucrose/30% fish gelatin solution, frozen on dry ice and cryosectioned into 30 μm slices. Slides were incubated at 37 °C for 1 h and then at room temperature overnight, washed with PBS, blocked in 10% NGS and then incubated in anti-E-cadherin antibody (1:50; DSHB, 5D3) overnight and then with AlexaFluor 488 anti-mouse (1:1000) and DAPI (1:1000) for 1 h at room temperature. Sections were mounted and imaged on the Leica SP8 confocal microscope where Z-stacks were obtained. Z-stack images were visualised with FIJI v.1.53k.
Neural crest explants and dispersion assays
Neural crest was dissected as previously described82. In brief, using a hair knife, the overlying ectoderm was lifted, and the neural crest removed. Dissection was performed in 3/8 NAM. Neural crest explants were plated on fibronectin-coated plastic dishes with DFA media. Time-lapse microscopy was performed using a Nikon microscope and images were analysed using FIJI v.1.53k in combination with Dispersion Tool for triangulation, dispersion area quantification and speed analysis.
Pro-inflammatory activation in embryos and cells
Xenopus embryos at stage 18 and neural crest explants were exposed to 500 μg/ml lipopolysaccharide (LPS, Sigma-Aldrich) and 1 mM ATP (Sigma-Aldrich) overnight on 0.1 MMR or DFA media, respectively, referred as LPS + ATP treatment. LPS + ATP low conditions were composed by 250 μg/ml LPS and 0.1 mM ATP. Human TNFa protein (ThermoFisher Scientific) was also used for pro-inflammatory activation in Xenopus embryos or iNCCs. In Xenopus embryos at stage 18, 50 ng of TNFa protein or PBS 1X (control) were injected directly under the epidermis on the neural crest region of one side of the embryo. In iNCCs, 50 ng/ml TNFa protein was added in the final media, while controls were added with the equivalent PBS volume.
Scratch and transwell assays
For scratch assays, 200,000 iNCCs were plated in 24-well plastic plates coated with Matrigel (Corning) and incubated for 24 h. A central vertical scratch was made in each well using a 20 μl tip and cells were washed with PBS twice and full iNCC media was added. Images were taken immediately after replenishing media (time = 0 h) using a EVOS m7000 microscope (ThermoFisher Scientific) and after 24 h. Images were analysed using FIJI v.1.53k and the automated MRI Wound Healing Tool. Transwell assays were performed using 24-well Thincerts 8um (GBO), where 50.000 cells were plated on matrigel coated Thincerts and left to migrate to the other side for 48 h. Next the cells were fixed on 4% PFA, membranes carefully removed from Thincerts and cells stained with DAPI. After PBS washes, the membranes were imaged on a SP8 confocal and nuclei counted from both sides of membranes using FIJI ImageJ.
RNA extractions and quantitative PCR (qPCR) analysis of gene expression
RNA extractions were performed using Monarch Total RNA Miniprep Kit (NEB) following the fabricant recommendations. For whole embryo RNA, embryos were individualised in 1.5 ml tubes and lysed using the kit’s Lysis Buffer. For neural crest RNA, pools of 10 neural crest explants were combined in a 1.5 ml tube and lysed as above reported. For iPSCs to NCC differentiation, cell pellets containing 100.000 cells were used, while for iNCC used in the scratch assays the whole well was recovered and lysed in the Lysis buffer. cDNAs were synthesised using SuperScript™ VILO™ Master Mix (ThermoFisher Scientific) using 1 μg of RNA per sample and cDNAs were diluted 1:10 in nuclease-free water. qPCRs reactions were set up in triplicates using 2 μl of cDNA, 10 μl of PowerUp™ SYBR™ Green Master Mix (ThermoFisher Scientific), 100 nM forward and reverse primers and nuclease free water in a 20 μl reaction. Primers for targets used in this study were designed in exon-exon junctions using Primer Blast tool (NCBI) and are described in Supplementary Table 2. Reactions were performed in a QuantStudio 3 qPCR System (ThermoFisher Scientific) using standard parameters. Relative expression values (Delta Ct) were calculated as previously reported83, using ef1a for Xenopus samples and TBP for human samples as endogenous controls.
DNA methylation analysis
Bisulfite sequencing was used for assessing Xenopus cdh1 and human CDH1 promoter methylation levels as previously described84. In brief, DNA was extracted from neural crest explant pools or iNCC pellets using Monarch® Genomic DNA Purification Kit (NEB) and 500 ng of DNA per sample undergone bisulfite conversion using EpiJET Bisulfite Conversion Kit (ThermoFisher Scientific). Bisulfite-specific PCR was conducted for human CDH1 promoter and Xenopus cdh1 promoter, in which primers were designed using MethPrimer85 and adaptors were added at the 5’ portion of each primer for peak normalisation83 (Supplementary Table 2). Sanger sequencing of amplicons was performed and peak values for cytosines in CpG positions were called using Chromas Lite Version 2.1.1. Cytosine signals were normalised by a normalisation factor (NF), based on the ratio of the signals for the C and T encoded by the tails of primers. Then, the peak height of each C (Ci) included in the target sequence was corrected for this NF as follow: Cnorm = Ci/NF. Finally, the normalised C signals were used to determine the methylation percentage as in 100 ∗ Cnorm/(Cnorm + T).
Normality in the spread of data for each experiment was tested using the Kolmogorov–Smirnov in Prism 9 (GraphPad). Significances for datasets displaying normal distributions were calculated in Prism 9 with an unpaired two-tailed Student’s t tests or One-way or Two-way analysis of variance with post-hoc Tukey’s test for multiple comparisons. Significances for non-normal distributed data were calculated in Prism 9 using two-tailed Mann–Whitney U-tests or Kruskal–Wallis tests with host hoc Dunn’s test for multiple comparisons.
All asterisks of statistical significance or lack or statistical significance (NS) refer to a comparison with the control, unless otherwise indicated.
The authors were not blinded to embryos or cells. The criteria for selection were survival and correct delivery of the injected treatments. Embryos and cells were allocated into experimental groups randomly. No predetermination of sample sizes was performed.
Xenopus laevis illustrations were either adapted from Xenbase (www.xenbase.org RRID:SCR_003280)86 using Adobe Illustrator or directly created on Adobe Illustrator.
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
All data produced in this manuscript can be found throughout figures, Supplementary figures and Supplementary tables and in the Source Data file. Source data are provided with this paper.
Chop chop sgRNA design: https://chopchop.rc.fas.harvard.edu/.
Human CDH1 accession code: RefSeq NM_004360.
Xenopus laevis cdh1 accession code: RefSeq NM_001172232.1.
Garrod, A. E. The incidence of alkaptonuria: a study in chemical individuality. Lancet 160, 1616–1620 (1902).
Hunter, D. J. Gene-environment interactions in human diseases. Nat. Rev. Genet. 6, 287–298 (2005).
Boyce, W. T., Sokolowski, M. B. & Robinson, G. E. Genes and environments, development and time. Proc. Natl Acad. Sci. USA 117, 23235–23241 (2020).
Esposito, G., Azhari, A. & Borelli, J. L. Gene × environment interaction in developmental disorders: where do we stand and what’s next? Front. Psychol. 9, 2036 (2018).
Baye, T. M., Abebe, T. & Wilke, R. A. Genotype-environment interactions and their translational implications. Per. Med. 8, 59–70 (2011).
Czamara, D. et al. Integrated analysis of environmental and genetic influences on cord blood DNA methylation in new-borns. Nat. Commun. 10, 2548 (2019).
Kobayashi, G. S., Brito, L. A., Meira, J. G. C., Alvizi, L. & Passos-Bueno, M. R. in Cleft Lip and Palate Treatment (eds. Alonso, N. & Raposo-Amaral, C. E.) 25–35 (Springer International Publishing, 2018). https://doi.org/10.1007/978-3-319-63290-2_4.
Ludwig, K. U. et al. Imputation of orofacial clefting data identifies novel risk loci and sheds light on the genetic background of cleft lip ± cleft palate and cleft palate only. Hum. Mol. Genet. 26, 829–842 (2017).
Leslie, E. J. & Marazita, M. L. Genetics of cleft lip and cleft palate. Am. J. Med. Genet. C: Semin. Med. Genet. 163C, 246–258 (2013).
Brito, L. A. et al. Genetic contribution for non-syndromic cleft lip with or without cleft palate (NS CL/P) in different regions of Brazil and implications for association studies. Am. J. Med. Genet. A 155A, 1581–1587 (2011).
Grosen, D. et al. Risk of oral clefts in twins. Epidemiology 22, 313–319 (2011).
Mossey, P. A., Little, J., Munger, R. G., Dixon, M. J. & Shaw, W. C. Cleft lip and palate. Lancet 374, 1773–1785 (2009).
Dixon, M. J., Marazita, M. L., Beaty, T. H. & Murray, J. C. Cleft lip and palate: understanding genetic and environmental influences. Nat. Rev. Genet. 12, 167–178 (2011).
Vogelaar, I. P. et al. Identification of germline mutations in the cancer predisposing gene CDH1 in patients with orofacial clefts. Hum. Mol. Genet. 22, 919–926 (2013).
Frebourg, T. et al. Cleft lip/palate and CDH1/E-cadherin mutations in families with hereditary diffuse gastric cancer. J. Med. Genet. 43, 138–142 (2006).
Benusiglio, P. R. et al. Cleft lip, cleft palate, hereditary diffuse gastric cancer and germline mutations in CDH1. Int. J. Cancer 132, 2470 (2013).
Brito, L. A. et al. Rare variants in the epithelial cadherin gene underlying the genetic etiology of nonsyndromic cleft lip with or without cleft palate. Hum. Mutat. 36, 1029–1033 (2015).
Cox, L. L. et al. Mutations in the epithelial cadherin-p120-catenin complex cause mendelian non-syndromic cleft lip with or without cleft palate. Am. J. Hum. Genet. 102, 1143–1157 (2018).
Manolio, T. A. et al. Finding the missing heritability of complex diseases. Nature 461, 747–753 (2009).
Maher, B. Personal genomes: The case of the missing heritability. Nature 456, 18–21 (2008).
Basha, M. et al. Whole exome sequencing identifies mutations in 10% of patients with familial non-syndromic cleft lip and/or palate in genes mutated in well-known syndromes. J. Med. Genet. 55, 449–458 (2018).
Wright, C. F. et al. Assessing the Pathogenicity, Penetrance, and Expressivity of Putative Disease-Causing Variants in a Population Setting. Am. J. Hum. Genet. 104, 275–286 (2019).
Gowans, L. J. J. et al. The prevalence, penetrance, and expressivity of etiologic IRF6 variants in orofacial clefts patients from sub-Saharan Africa. Mol. Genet. Genom. Med. 5, 164–171 (2017).
Aylward, A. et al. Using whole exome sequencing to identify candidate genes with rare variants in nonsyndromic cleft lip and palate. Genet. Epidemiol. 40, 432–441 (2016).
Rahimov, F., Jugessur, A. & Murray, J. C. Genetics of nonsyndromic orofacial clefts. Cleft Palate Craniofac. J. 49, 73–91 (2012).
Beiraghi, S. et al. Autosomal dominant nonsyndromic cleft lip and palate: significant evidence of linkage at 18q21.1. Am. J. Hum. Genet. 81, 180–188 (2007).
Xu, T. et al. Identification of a novel TP63 mutation causing nonsyndromic cleft lip with or without cleft palate. BMC Med. Genomics 14, 53 (2021).
Stanier, P. & Moore, G. E. Genetics of cleft lip and palate: syndromic genes contribute to the incidence of non-syndromic clefts. Hum. Mol. Genet. 13, R73–R81 (2004).
Bhagirath, A. Y. et al. Role of maternal infections and inflammatory responses on craniofacial development. Front. Oral. Health 2, 735634 (2021).
Carter, T. C. et al. Maternal self-reported genital tract infections during pregnancy and the risk of selected birth defects. Birth Defects Res. A: Clin. Mol. Teratol. 91, 108–116 (2011).
Peterka, M., Tvrdek, M., Likovský, Z., Peterková, R. & Fára, M. Maternal hyperthermia and infection as one of possible causes of orofacial clefts. Acta Chir. Plast. 36, 114–118 (1994).
Ács, L. et al. Maternal-related factors in the origin of isolated cleft palate-A population-based case-control study. Orthod. Craniofac. Res. 23, 174–180 (2020).
Waller, D. K. et al. Maternal report of fever from cold or flu during early pregnancy and the risk for noncardiac birth defects, National Birth Defects Prevention Study, 1997-2011. Birth Defects Res. 110, 342–351 (2018).
Blanco, R., Colombo, A. & Suazo, J. Maternal obesity is a risk factor for orofacial clefts: a meta-analysis. Br. J. Oral. Maxillofac. Surg. 53, 699–704 (2015).
Cedergren, M. & Källén, B. Maternal obesity and the risk for orofacial clefts in the offspring. Cleft Palate Craniofac. J. 42, 367–371 (2005).
Kutbi, H. et al. Maternal underweight and obesity and risk of orofacial clefts in a large international consortium of population-based studies. Int. J. Epidemiol. 46, 190–199 (2017).
Li, H. et al. The double-edged sword role of TGF-β signaling pathway between intrauterine inflammation and cranial neural crest development. FASEB J. 36, e22113 (2022).
Jee, Y. S., Jang, T. J. & Jung, K. H. Prostaglandin E(2) and interleukin-1β reduce E-cadherin expression by enhancing snail expression in gastric cancer cells. J. Korean Med. Sci. 27, 987–992 (2012).
Debelec-Butuner, B. et al. TNFα-mediated loss of β-catenin/E-cadherin association and subsequent increase in cell migration is partially restored by NKX3.1 expression in prostate cells. PLoS ONE 9, e109868 (2014).
Saito, T. et al. Inflammatory cytokines induce a reduction in E-cadherin expression and morphological changes in MDCK cells. Res. Vet. Sci. 96, 288–291 (2014).
Zhou, J., Zhang, C., Pan, J., Chen, L. & Qi, S.-T. Interleukin‑6 induces an epithelial‑mesenchymal transition phenotype in human adamantinomatous craniopharyngioma cells and promotes tumor cell migration. Mol. Med. Report. 15, 4123–4131 (2017).
Gu, K. et al. Interleukin-17-induced EMT promotes lung cancer cell migration and invasion via NF-κB/ZEB1 signal pathway. Am. J. Cancer Res. 5, 1169–1179 (2015).
Qian, X. et al. E-cadherin promoter hypermethylation induced by interleukin-1beta treatment or H. pylori infection in human gastric cancer cell lines. Cancer Lett. 263, 107–113 (2008).
Machado, J. C. et al. E-cadherin gene (CDH1) promoter methylation as the second hit in sporadic diffuse gastric carcinoma. Oncogene 20, 1525–1528 (2001).
Grady, W. M. et al. Methylation of the CDH1 promoter as the second genetic hit in hereditary diffuse gastric cancer. Nat. Genet. 26, 16–17 (2000).
Mayor, R. & Theveneau, E. The neural crest. Development 140, 2247–2251 (2013).
Bronner, M. E. & LeDouarin, N. M. Development and evolution of the neural crest: an overview. Dev. Biol. 366, 2–9 (2012).
Minoux, M. & Rijli, F. M. Molecular mechanisms of cranial neural crest cell migration and patterning in craniofacial development. Development 137, 2605–2621 (2010).
Huang, C., Kratzer, M.-C., Wedlich, D. & Kashef, J. E-cadherin is required for cranial neural crest migration in Xenopus laevis. Dev. Biol. 411, 159–171 (2016).
Scarpa, E. et al. Cadherin switch during EMT in neural crest cells leads to contact inhibition of locomotion via repolarization of forces. Dev. Cell 34, 421–434 (2015).
Zhu, Y. et al. Migratory neural crest cells phagocytose dead cells in the developing nervous system. Cell 179, 74–89.e10 (2019).
Bayless, N. L., Greenberg, R. S., Swigut, T., Wysocka, J. & Blish, C. A. Zika virus infection induces cranial neural crest cells to produce cytokines at levels detrimental for neurogenesis. Cell Host Microbe 20, 423–428 (2016).
Pallocca, G. et al. Impairment of human neural crest cell migration by prolonged exposure to interferon-beta. Arch. Toxicol. 91, 3385–3402 (2017).
Xie, Q. et al. Lipopolysaccharide/adenosine triphosphate induces IL‑1β and IL-18 secretion through the NLRP3 inflammasome in RAW264.7 murine macrophage cells. Int. J. Mol. Med. 34, 341–349 (2014).
Sanz, J. M. & Di Virgilio, F. Kinetics and mechanism of ATP-dependent IL-1 beta release from microglial cells. J. Immunol. 164, 4893–4898 (2000).
Na, Y.-R., Lee, J.-S., Lee, S.-J. & Seok, S.-H. Interleukin-6-induced twist and N-cadherin enhance melanoma cell metastasis. Melanoma Res. 23, 434–443 (2013).
Kim, D. W. et al. Interleukin-2 alters distribution of CD144 (VE-cadherin) in endothelial cells. J. Transl. Med. 12, 113 (2014).
Cousin, H. Cadherins function during the collective cell migration of xenopus cranial neural crest cells: revisiting the role of E-cadherin. Mech. Dev. 148, 79–88 (2017).
Cai, D. et al. Mechanical feedback through E-cadherin promotes direction sensing during collective cell migration. Cell 157, 1146–1159 (2014).
Grimaldi, C. et al. E-cadherin focuses protrusion formation at the front of migrating cells by impeding actin flow. Nat. Commun. 11, 5397 (2020).
Elisha, Y., Kalchenko, V., Kuznetsov, Y. & Geiger, B. Dual role of E-cadherin in the regulation of invasive collective migration of mammary carcinoma cells. Sci. Rep. 8, 4986 (2018).
Padmanaban, V. et al. E-cadherin is required for metastasis in multiple models of breast cancer. Nature 573, 439–444 (2019).
Feil, R. & Fraga, M. F. Epigenetics and the environment: emerging patterns and implications. Nat. Rev. Genet. 13, 97–109 (2012).
Bayarsaihan, D. Epigenetic mechanisms in inflammation. J. Dent. Res. 90, 9–17 (2011).
Ligthart, S. et al. DNA methylation signatures of chronic low-grade inflammation are associated with complex diseases. Genome Biol. 17, 255 (2016).
Alvizi, L. et al. Differential methylation is associated with non-syndromic cleft lip and palate and contributes to penetrance effects. Sci. Rep. 7, 2441 (2017).
Lombardo, M. V. et al. Maternal immune activation dysregulation of the fetal brain transcriptome and relevance to the pathophysiology of autism spectrum disorder. Mol. Psychiatry 23, 1001–1013 (2018).
Fatemi, S. H. et al. Maternal infection leads to abnormal gene regulation and brain atrophy in mouse offspring: implications for genesis of neurodevelopmental disorders. Schizophr. Res. 99, 56–70 (2008).
Garbett, K. A., Hsiao, E. Y., Kálmán, S., Patterson, P. H. & Mirnics, K. Effects of maternal immune activation on gene expression patterns in the fetal brain. Transl. Psychiatry 2, e98 (2012).
Solek, C. M., Farooqi, N., Verly, M., Lim, T. K. & Ruthazer, E. S. Maternal immune activation in neurodevelopmental disorders. Dev. Dyn. 247, 588–619 (2018).
Serrels, A., Canel, M., Brunton, V. G. & Frame, M. C. Src/FAK-mediated regulation of E-cadherin as a mechanism for controlling collective cell movement: insights from in vivo imaging. Cell Adh. Migr. 5, 360–365 (2011).
Tammen, S. A., Friso, S. & Choi, S.-W. Epigenetics: the link between nature and nurture. Mol. Asp. Med. 34, 753–764 (2013).
Jaenisch, R. & Bird, A. Epigenetic regulation of gene expression: how the genome integrates intrinsic and environmental signals. Nat. Genet. 33, 245–254 (2003).
Liao, M. et al. Aging-elevated inflammation promotes DNMT3A R878H-driven clonal hematopoiesis. Acta Pharm. Sin. B 12, 678–691 (2022).
Leoni, C. et al. Dnmt3a restrains mast cell inflammatory responses. Proc. Natl Acad. Sci. USA 114, E1490–E1499 (2017).
Nakatomi, M. et al. Msx1 deficiency interacts with hypoxia and induces a morphogenetic regulation during mouse lip development. Development 147, dev189175 (2020).
Goodrich, J. K. et al. Determinants of penetrance and variable expressivity in monogenic metabolic conditions across 77,184 exomes. Nat. Commun. 12, 3505 (2021).
Otto, P. A. & Hiromoto, A. R. V. R. Penetrance rate estimation in autosomal dominant conditions. Genet. Mol. Biol. 35, 583–588 (2012).
Theveneau, E. et al. Chase-and-run between adjacent cell populations promotes directional collective migration. Nat. Cell Biol. 15, 763–772 (2013).
Gerhart, J. & Kirschner, M. Normal Table of Xenopus Laevis (Daudin) A Systematical and Chronological Survey of the Development from the Fertilized Egg Till the End of Metamorphosis (Garland Science, 2020). https://doi.org/10.1201/9781003064565.
Nani, D. A., Hsia, G. S. P., Passos-Bueno, M. R. & Kobayashi, G. S. Modeling early neural crest development via induction from hiPSC-derived neural plate border-like cells. Methods Mol. Biol. 2549, 281–298 (2022).
Barriga, E. H., Shellard, A. & Mayor, R. In vivo and in vitro quantitative analysis of neural crest cell migration. Methods Mol. Biol. 1976, 135–152 (2019).
Pfaffl, M. W. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res. 29, e45 (2001).
Brisotto, G. et al. An improved sequencing-based strategy to estimate locus-specific DNA methylation. BMC Cancer 15, 639 (2015).
Li, L.-C. & Dahiya, R. MethPrimer: designing primers for methylation PCRs. Bioinformatics 18, 1427–1431 (2002).
Zahn et al. Normal table of xenopus development: a new graphical resource. Development 149, dev200356 (2022).
The authors thank CLP families involved in this study and Operation Smile Brazil. We thank Naila Lorenzo and Simone Ferreira for assistance in processing human samples. We thank Heloisa Bueno for assistance with mice. We thank Centro de Estudos do Genoma Humano e Celulas-Tronco sequencing team for support in sequencing analysis. We are also very thankful to Jonas Hartmann and Adam Shellard for discussions and assistance with this manuscript. This study was supported by the Sao Paulo Research Foundation (LA FAPESP, 2017/11430-7) and CEPID-FAPESP (MRPB 2013/08028-1). Work in RM’s laboratory is supported by grants from the Medical Research Council (MR/S007792/1), Biotechnology and Biological Sciences Research Council (M008517) and Wellcome Trust (102489/Z/13/Z).
The authors declare no competing interests.
Peer review information
Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Alvizi, L., Nani, D., Brito, L.A. et al. Neural crest E-cadherin loss drives cleft lip/palate by epigenetic modulation via pro-inflammatory gene–environment interaction. Nat Commun 14, 2868 (2023). https://doi.org/10.1038/s41467-023-38526-1
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.