Apport des approches à haut débit au développement et aux pathologies de la crête neurale / Integrating High Throughput Approaches in Neural Crest Development and Disease

Séminaire organisé par Anne-Hélène Monsoro-Burq (Centre de recherche Institut Curie et Université Paris Sud, France) et Lukas Sommer (University of Zurich, Institute of Anatomy, Switzerland) du 13 au 18 novembre 2017

Participants

Nadège Bondurand, Marianne Bronner-Hansen, Aravinda Chakravarti, Julien Debbache, Olivier Delattre, Paul Kulesa, Carole LaBonne, Nicole Le Douarin, Stanislas Lyonnet, Anne-Hélène Monsoro-Burq, Angela Nieto, William J. Pavan, Filippo Rijli, Tatjana Sauka-Spengler, Pier Luigi Scerbo, Marcos Simoes Costa, Lukas Sommer, Philippe Soriano, Paul Trainor, Joanna Wysoka

Résumé

La crête neurale est une population cellulaire clé, spécifique des embryons de vertébrés. Au cours du développement, la formation de la crête neurale à la frontière de la plaque neurale implique l’activation contrôlée d’un réseau de régulateurs transcriptionnels et de leurs effecteurs, résultant en une transition épithélio-mésenchymateuse stéréotypée (EMT), suivie de la migration et de la différenciation des cellules de la crête neurale en multiples dérivés tels que les neurones et la glie périphériques, les cellules pigmentaires, os et cartilage craniofaciaux, des cellules cardiaques et des dérivés surrénaliens.

Étant donné l’importance de la crête neurale pour l’évolution des vertébrés, le développement embryonnaire, les maladies congénitales et la formation de tumeurs, il est essentiel de comprendre les mécanismes qui sous-tendent la spécification de cette population cellulaire, sa migration et sa différenciation. Depuis des années, divers laboratoires abordent ces problèmes au moyen de l’embryologie moléculaire, de la biologie cellulaire et de la manipulation fonctionnelle de l’expression des gènes chez les embryons d’oiseau, d’amphibien ou de rongeurs. Ce n’est que récemment que la crête neurale est devenue la cible de recherches plus globales et systématiques. En effet, plusieurs équipes ont développé des stratégies à haut débit, utilisant la génomique, l’épigénomique et les approches d’imagerie adaptées à l’analyse de cette population cellulaire embryonnaire complexe. Ce séminaire a stimulé des interactions entre des chercheurs établis dans ces divers domaines complémentaires de la biologie de la crête neurale, qui créeront et consolideront des interfaces entre différents projets.

Par exemple, les réseaux géniques (GRN) actifs dans les cellules souches multipotentes de la crête neurale sont partagés avec ceux présents dans les cellules pluripotentes embryonnaires précoces et dans les cellules de mélanome. Les discussions ont mis en évidence la nécessité d’une analyse comparative et fonctionnelle approfondie de ces réseaux. De même, les groupes qui étudient les programmes génétiques et épigénétiques dans la migration des cellules de la crête neurale obtiennent probablement des informations pertinentes pour l’invasion des cellules métastatiques dans les tumeurs dérivées de la crête neurale étudiées par d’autres laboratoires. Enfin, plusieurs laboratoires analysent les paysages génomiques des gènes régulateurs de la crête neurale. Combinées avec l’analyse des transcriptomes des pathologies associées, ces nouvelles approches pourraient identifier de nouvelles voies de signalisation et coopérations géniques associées à la maladie comme au développement de la crête neurale, et permettre le criblage de potentiels médicaments, utilisant des cellules de crête neurale dérivées de cellules souches pluripotentes induites (iPSC) établies par d’autres groupes.

Compte-rendu (en anglais)

A deep understanding of neural crest biology is essential for fundamental research as well as for analysis of human disease. One third of human congenital malformations results from defective events in neural crest development or in the biology of neural crest derivatives. The most aggressive cancers, with high metastatic potential, often come from the malignant transformation of neural crest derivatives (glioma, melanoma, childhood cancers…). The recent advances in neural crest biology, in modeling human neural crest-linked disease, and the similarities in the mechanisms of cancer metastasis and neural crest migration, have spotted light on the neural crest gene regulatory network more than ever. Stem cell biology and cancer biology are great assets to devise novel assays using the recent findings in neural crest biology and develop new tools for potential therapies. Neural crest cells, or vertebrate embryonic tissues in general, are available in very limited amounts and, therefore, difficult to study by standard cell and molecular biology assays. This limitation has hindered the use of many large-scale approaches for several years. However, using cutting edge and complementary strategies, a few laboratories across the world have recently devised novel approaches allowing neural crest analysis in a more global and also a more precise fashion at a genome scale (transcriptomics across development, single cell sequencing, epigenomics). This meeting has gathered prominent scientists and junior researchers, studying neural crest from different angles, including various combinations of experimental and molecular embryology, transcriptomics, epigenomics, genetics, evolution, imaging, human neurocristopathies and cancer biology.

The neural crest gene regulatory network: an embryological and evolutionary perspective.

Neural crest development is modeled by a complex multi-step gene regulatory network (NC-GRN) that is common to all vertebrates, and serves as a scaffold to integrate novel regulators and compare the modifications of the network during evolution and disease. This GRN initiates with signaling and transcriptional modules at the edge of the neural plate that cooperate to turn on the neural crest specification program. Several participants have studied specific modules of the network. Marianne Bronner described the neural crest GRN in the context of the differences in neural crest cells along the body axis, such that only cranial but not trunk neural crest cells form craniofacial cartilage and bone in vivo. In amniotes, they uncovered a cranial neural crest-specific transcriptional circuit unique to this population and missing from the trunk neural crest. In examining the evolutionary conservation of this subcircuit, they find that many elements are missing from the premigratory cranial neural crest of the basal vertebrate, lamprey. Moreover, the origin of enteric nervous system in lamprey was discussed. Anne-Hélène Monsoro-Burq presented new bioinformatics strategies developed to define the spatial transcriptome signatures of the neural border and dorsal ectoderm, and compared to preliminary data on neural crest single cell transcriptome datasets. They defined a minimal neural crest-inducing module using two transcription factors, Pax3 and Zic1, necessary and sufficient to activate neural crest in pluripotent blastula cells. They showed that this module directly triggers an epithelium-to-mesenchyme transition (EMT) signature in premigratory neural crest. Tatiana Sauka-Spengler has explored novel Crispr/Cas9 -mediated strategies to manipulate the epigenome of chick embryos and study the role of the regulatory elements controlling the key genes in the NC-GRN in vivo.

Neural crest multipotency is essential to form its diverse derivatives, and was discussed by several participants. Carole LaBonne talked about the high conservation of the NC-GRN components between pluripotent blastula stem cells and neural crest cells, and a revised model for neural crest development and evolution whereby neural crest cells evolved via retention of potency possessed by blastula cells rather than an induced regaining of potential. She also discussed a major difference between these two cell populations – that neural crest cells deploy SoxE rather than SoxB1 factors to control potency. A role for FGF signaling in retention of pluripotency was described, with a novel switch in effector pathway (MAPK or AKT) utilization as cells progress from pluripotency. In a related analysis, on blastula cells, Pierluigi Scerbo has shown that asymmetric degradation of pluripotency factors during mitosis of pluripotent cells is the key step for cell competence to differentiation in vivo. Mechanistically, he found that MEK1, part of MAPK pathway, regulates the asymmetric distribution of the pluripotency factor Ventx2. Inhibition of MEK1 activity led to symmetric distribution of Ventx2 protein, prolonged expression of pluripotency regulators (like pou5f3/oct4) and loss of embryonic cell competence to lineage restriction. Marcos Simoes-Costa discussed novel mechanisms that control the silencing of neural crest multipotency during cell differentiation. The complex NC-GRN program has to be repressed during cell commitment so that multipotent progenitors can adopt their terminal fates. They characterized a new regulatory circuit that actively silences the neural crest GRN at the late stages of migration, during the transition from multipotency to differentiation. Their results show that, as cells migrate away from the dorsal neural tube, crucial nodes of the GRN are targeted, resulting in the collapse of the network.

Neural crest epithelial mesenchymal transition and migration, parallels with cancer metastasis.

The epithelial to mesenchymal transition (EMT) is crucial for the delamination and migration of neural crest cells. Angela Nieto gave an overview of the importance of this process in both embryonic development and adult diseases. She presented data on a novel developmental EMT process that drives heart looping in vertebrates and also discussed why the organism needs so many EMT inducing transcription factors (i.e. Snail, Zeb, Twist, etc.). She showed that when epithelial cells are forced to express individual EMT-TFs they all undergo EMT but their migratory behaviour is different. As both embryonic and invasive cancer cells express a combination of EMT-TFs, their final behavior would be the integration of that impinged by the particular combination and levels of each EMT-TF. Paul Kulesa presented how expression of such EMT-TFs and other gene sets change in neural crest cells after their delamination and during migration. He shared the results of the group’s single cell RNA-seq analysis of migrating chick cranial neural crest cells. The data revealed that the majority of neural crest cells within the invasive front change their gene expression in a consistent manner during migration. However, a small subpopulation of cells narrowly confined to the invasive front have a conserved so-called trailblazer signature. Of note, neural crest cells newly exited from the dorsal neural tube clustered into two subpopulations with profiles distinct from the trailblazer signature and indicative of a rapid switch from EMT to directed migration. Key aspects of the trailblazer signature were validated by a novel method of combining multiplexed fluorescence in situ hybridization with immunohistochemistry and tissue clearing. Finally, Olivier Delattre discussed how EMT-like properties, i.e. the capacity to acquire a mesenchymal phenotype, reflect an intrinsic plasticity of two neural crest-derived tumors, neuroblastoma and Ewing sarcoma. Likely, this may account for the strong ability of cells from these malignancies to metastasize and resist to treatment.

The neural crest derivatives: formation, evolution, and disease of craniofacial structures.

The laboratory of Joanna Wysoka uses cranial neural crest cells (CNCCs) as a paradigm to study how genetic information harbored by cis-regulatory elements is decoded into a diversity of functions, behaviors and morphologies during development and evolution. The group recently introduced ‘cellular anthropology’, a strategy of using iPSCs from humans and great apes to in vitro-derive evolutionarily important cell types for the systematic discovery of cell type-specific regulatory changes. With this approach, epigenomic profiling of human and chimpanzee CNCCs was used to annotate evolutionary divergence of craniofacial cis-regulatory landscapes. In the second part of her talk, she reported how mutations in general regulators of housekeeping cellular functions contribute to many craniofacial disorders, revealing a previously unappreciated mechanism linking nucleolar dysfunction, rDNA damage and craniofacial malformations. Likewise, Paul Trainor discussed the development of mammalian neural crest cells and their importance in neurocristopathies affecting the head and face. Using mouse and zebrafish as model systems, he showed how genes involved in human craniofacial disorders are essential for neural crest generation, proliferation and survival as a consequence of the critical roles they play in rDNA transcription and ribosome biogenesis. Another approach to understand the genetic and molecular bases of neurocristopathies was presented by Stanislas Lyonnet. His laboratory is engaged in a forward genetic screen based on patients with rare genetic disorders resulting in neural crest anomalies of development. In parallel, his lab studies disease-relevant pathways using in vitro and in vivo models, culture techniques (iPS), RNAseq and CRISPR/Cas9 technology. Philippe Soriano discussed the roles of FGF signaling in craniofacial development. While conditional disruption of Fgfr1 in neural crest cells leads to facial clefting, conditional deletion of Fgfr2 does not lead to an overt phenotype. Conditional deletion of both receptors leads to agenesis of the frontal bones and most of the mandible. He further discussed signaling pathways downstream of FGR signaling. How CNCCs can possibly cope with altered signaling during development or in disease was illustrated by Filipo Rijli. He demonstrated that CNCCs remain plastic until postmigratory stage by establishing a Polycomb-dependent bivalent chromatin epigenetic signature allowing Hox-negative CNCCs to maintain broad patterning competence and plasticity through migration, while being poised to respond to local cues and induce position-specific transcriptional subprograms. Epigenetic poising of promoters and enhancers may allow CNCCs to rapidly adapt their response to local variations in environmental signaling during migration. However, CNCCs not only receive environmental signals during development of craniofacial structures, they themselves are instrumental in instructing proper development of other structures, notably the forebrain, as presented by Nicole Le Douarin. Intriguingly, deletion of anterior Hox-free cranial neural crest results in drastic brain malformations, due to reduced signaling by FGF8. Accordingly, FGF8 can rescue brain defects caused by neural crest ablation. Thus, the Hox-free cranial neural crest acts as a kind of ‘organizer’, influencing shape and size of developing anterior brain structures.

The neural crest derivatives: formation and pathologies of melanocytes and enteric nervous system.

Neurocristopathies not only affect craniofacial structures, but also several other neural crest derivatives. Such defects might be caused, for instance, by aberrant metabolic processes, as introduced by Lukas Sommer. He has recently identified Yy1 as a master transcription factor regulating multiple metabolic processes and protein translation. Deletion of this factor causes severe defects in all neural crest derivatives analyzed due to decreased proliferation and increased cell deaths. Strikingly, Yy1-regulated mechanisms are also crucial for melanoma initiation and growth, indicating that formation of tumors with a neural crest origin might involve molecular programs related to those active during embryonic development. The laboratory of William Pavan has been using variation in skin and hair color of animals to identify genes needed for the development and function of neural crest-derived melanocytes and to understand how these genes are important in human health and disease. More recently, the lab has developed genomic (chromatin marks and RNAseq) resources aimed at identifying non-coding regions of the genome needed for neural crest derived melanocytes and has determined how transcription factor binding changes with altered developmental states and environmental perturbations. Comprehensive genetics analyses are also used by Aravinda Chakravarti to understand the molecular basis of Hirschsprung disease (HSCR), or congenital intestinal aganglionosis. HSCR demonstrates a complex inheritance pattern. Indeed, a diversity of genes with a diversity of mutations characterizes the disorder, and individual patients harbor different collections of these mutations. However, despite this diversity, there is functional coherence in the genetic defects, with the majority affecting a single GRN. Complementing these strategies, Nadège Bondurand described the different approaches they have used to expand the knowledge of the molecular and cellular bases of Waardenburg syndrome and HSCR, focusing on the role of SOX10 and its network in neural crest development. Combined double mouse mutants and in vitro studies were applied to show that differentiation of enteric nervous system progenitor cells is controlled by endothelin 3. She then discussed the importance of high throughput technologies to identify new SOX10 target genes and new genes involved in Waardenburg syndrome. Finally, Julien Debbache presented data showing that pathological conditions associated with neural crest cells may also involve paracrine signaling elicited from neural crest-derived cells on the surrounding tissue. Using a murine wound healing model, he discovered that physical injury activates peripheral glia by promoting their de-differentiation, cell-cycle re-entry and dissemination of the cells into the wound bed. Injury-activated glia upregulate a wound healing secretome, which promotes myofibroblast differentiation by paracrine modulation of TGF-β signaling. In contrast, ablation of injury-activated glia counteracts efficient wound healing. Hence, the secretome of injury-activated glia presents therapeutic potential towards clinical wound healing disorders.

Conclusions and perspectives

Given the relevance of the neural crest for vertebrate evolution, embryonic development, congenital diseases, and tumor formation, it is important to gain insights into the mechanisms underlying neural crest specification, migration, and multi-lineage differentiation. For several years, various laboratories have addressed this issue by means of molecular embryology, cell biological approaches, and functional manipulation of gene expression in avian, frog, and rodent embryos. Only recently, the neural crest became the target of more global and systematic research activities. Indeed, several teams have been developing tools and collections of data by high throughput strategies, using genomics, epigenomics, and imaging approaches. This workshop has fostered interactions between established researchers active in complementary fields of neural crest biology, which will create and consolidate interfaces between different projects in neural crest formation.

For instance, GRNs active in multipotent neural crest stem cells are shared with those present in early embryonic pluripotent cells and in melanoma cells. Discussions have highlighted how a thorough comparative and functional analysis of these networks remains to be done. Likewise, groups investigating genetic and epigenetic programs in neural crest cell migration likely obtain information also relevant for invasiveness of metastatic cells in neural crest-derived tumors studied by other laboratories. Finally, several labs in the field perform genome-wide enhancer and mutational landscape analyses of neural crest regulators. Combined with disease-related transcriptomics, this could lay the grounds for the identification of novel disease-associated pathways and drug-screening approaches in ESC- and iPSC-derived neural crest cells established by other groups.