PHD2022-01
Quantification and modeling of embryonic lymph node organogenesis at the single cell scale
Host laboratory and collaborators
Léo GUIGNARD (LIS) / leo.guignard@univ-amu.fr
Serge VAN DE PAVERT (CIML) / vandepavert@ciml.univ-mrs.fr
Abstract
Lymph node (LN) formation requires hematopoietic lymphoid tissue inducer cells (LTi) to interact with mesenchymal cells at precise locations within the embryo, where they subsequently form aggregates. We have postulated that the peripheral nervous system outgrowth initiates the earliest events in LN formation. Indeed, preliminary data show that LTi aggregate morphology and cell density is affected in whole-mount stained mouse embryos lacking neuronal subsets. To understand the relationship between neuronal outgrowth and lymph node formation, we will develop new computational methods to reconstruct and quantify LTi aggregates and peripheral nervous system morphology at the single cell scale in the whole embryo. The reconstructions will then be used to develop a machine learning framework to systematically quantify phenotypes in perturbed mouse embryos. These quantifications will in turn allow to model the effects of neuronal outgrowth on LN formation.
Keywords
Quantitative embryogenesis, whole-mount analysis, peripheral nervous system, immune system, image analysis, machine learning, big data analysis
Objectives
- To build a library of whole-mount mouse embryo images
- To develop computational methods for the analysis of the generated library, including:
- Reconstruction and mapping of the neuronal network
- Quantification of LTi aggregate morphologies and positions
- To develop computational methods to automatically stage mouse embryos
- To model LN formation function of the neuronal network morphology
Proposed approach (experimental / theoretical / computational)
Different mouse models in which specific neuronal subsets are deleted are already available. Embryos will be isolated at the time during which LN are initiated and whole-mount stained, cleared and acquired on the UltramicroscopeII in the CIML. In the Guignard lab, the datasets generated will be computationally segmented using algorithms to constitute a quantitative single cell atlas of mouse LTi morphology. Machine learning algorithms will then be developed to quantify LTi morphology phenotypes for the set of mouse models. The quantification will drive potential new experiments and the development of a model of LTi aggregate formation. Both partners have published on the proposed methods before and thus have the necessary background to carry out this project. The labs will meet frequently to discuss progress and ensure the alignment between the biological and the computational questions.
Interdisciplinarity
The proposed project aims at developing computational algorithms and methods to answer a precise biological question for which new experiments will be necessary. While the van de Pavert lab at the CIML harbors the mouse lines, isolate, stain and acquire the mouse embryos, the Guignard lab at the LIS will be responsible for developing the algorithms which will quantify and compare the aggregates and neurons morphologies. The analysis of the quantitative outputs while help designing and orienting future experiments that will be carried out in the van de Pavert lab. The collaboration between both labs is essential to propose a model for LN formation.
Expected profile
Enthusiastic, creative and ambitious, good communication skills and eager to learn.
Master degree with major or minor in computer science. Affection for developmental biology. Some experience in developmental biology is preferred but not required.
Note: exception can be made for students who have not studied computer science if the student can prove coding skills.
Is this project the continuation of an existing project or an entirely new one? In the case of an existing project, please explain the links between the two projects
Data generated on this project thus far was not part of a specific project, but conducted within internship projects. Since we have now convincing preliminary data that neurons affect LN formation, we need to quantify their effect on LN formation by computational modeling.
2 to 5 references related to the project
- Simic M. et al. Cell Reports, 32(6):108004
- Wang et al., Developmental Cell, 56(22), 3128-3145.e15
- Van de Pavert S.A. et al., Nature Immunology 10(11):1193-9
- Van de Pavert S.A. et al. Nature 508(7494):123-7
- McDole K., Guignard L. et al. Cell, 175(3):859-876
3 main publications from each PI over the last 5 years
Léo Guignard:
- Mouse embryonic stem cells self-organize into trunk-like structures with neural tube and somites. Veenvliet J., Bolondi B., et al., Science, 370(6522):aba4937
- Contact area-dependent cell communication and the morphological invariance of ascidian embryogenesis. Guignard L., Fiuza U.-M., et al., Science, 369(6500):eaar5663
- In Toto Imaging and Reconstruction of Post-Implantation Mouse Development at the Single-Cell Level. McDole K., Guignard L. et al., Cell, 175(3):859-876
Serge van de Pavert:
- 1-deoxysphingolipids bind to COUP-TF to modulate lymphatic and cardiac cell development. Wang et al., Developmental Cell, 56(22), 3128-3145.e15
- Distinct Waves from the Hemogenic Endothelium Give Rise to Layered Lymphoid Tissue Inducer Cell Ontogeny. Simic, M. et al. Cell Reports, 32(6):108004
- The evolution of innate lymphoid cells. Vivier E., van de Pavert SA., Cooper MD., Belz GT. Nature Immunology 17(7):790-4
- Identification of natural RORγ ligands that regulate the development of lymphoid cells. Santori F., Huang P., van de Pavert SA., et al. Cell Metabolism 21(2):286-2