Skip to main content
Dryad logo

scRNAseq datasets of cranial myogenic progenitors using Mesp1 and Myf5 lineages

Citation

Grimaldi, Alexandre; Comai, Glenda; Mella, Sébastien; Tajbakhsh, Shahragim (2022), scRNAseq datasets of cranial myogenic progenitors using Mesp1 and Myf5 lineages, Dryad, Dataset, https://doi.org/10.5061/dryad.gf1vhhmrs

Abstract

How distinct cell fates are manifested by direct lineage ancestry from bipotent progenitors, or by specification of individual cell types is a key question for understanding the emergence of tissues. The interplay between skeletal muscle progenitors and associated connective tissue cells provides a model for examining how muscle functional units are established. Most craniofacial structures originate from the vertebrate-specific neural crest cells except in the dorsal portion of the head, where they arise from cranial mesoderm. Here, using multiple lineage-tracing strategies combined with single cell RNAseq and in situ analyses, we identify bipotent progenitors expressing Myf5 (an upstream regulator of myogenic fate) that give rise to both muscle and juxtaposed connective tissue. Following this bifurcation, muscle and connective tissue cells retain complementary signalling features and maintain spatial proximity. Disrupting myogenic identity shifts muscle progenitors to a connective tissue fate. The emergence of Myf5-derived connective tissue is associated with the activity of several transcription factors, including Foxp2. Interestingly, this unexpected bifurcation in cell fate was not observed in craniofacial regions that are colonised by neural crest cells. Therefore, we propose that an ancestral bi-fated program gives rise to muscle and connective tissue cells in skeletal muscles that are deprived of neural crest cells.

Methods

Methods as described in associated article (Grimaldi et al. 2022)

For E10.5 to E12.5 embryos, the cranial region above the forelimb was dissected in ice-cold 3% FBS in PBS and mechanically dissociated with forceps and pipetting. The same procedure was applied at E14.5 but the dissection was refined to the pharyngeal and laryngeal regions. Tissues were then digested in TrypLE (ThermoFisher Cat #: 12604013) during 3 rounds of 5-min incubation (37°C, 1400 RPM), interspersed with gentle pipetting to further dissociate the tissue. Cells were resuspended in FBS 3%, filtered, and incubated with Calcein Blue (eBioscience, Cat #: 65-0855-39) and Propidium Iodide (ThermoFisher Cat #: P1304MP) to check for viability. Viable cells were sorted on BD FACSAria™ III and manually counted using a hemocytometer. RNA integrity was assessed with Agilent Bioanalyzer 2100 to validate the isolation protocol prior to scRNAseq (RIN>8 was considered acceptable). 4000 to 13000 cells were loaded onto 10X Genomics Chromium microfluidic chip and cDNA libraries were generated following manufacturer’s protocol. Concentrations and fragment sizes were measured using Agilent Bioanalyzer and Invitrogen Qubit. cDNA libraries were sequenced using NextSeq 500 and High Output v2.5 (75 cycles) kits. Genome mapping and count matrix generation were done following 10X Genomics Cell Ranger pipeline. scRNAseq datasets were preprocessed using Seurat in R (https://satijalab.org/seurat/) (Butler et al., 2018). Cells with more than 20% of mitochondrial gene fraction were discarded. The number of genes expressed averaged to 4000 in all 4 datasets. Dimension reduction and UMAP generation were performed following Seurat workflow. Doublets were inferred using DoubletFinder v3 (McGinnis et al., 2019). Cell cycle genes, mitochondrial fraction, number of genes, number of UMI were regressed in all datasets following Seurat dedicated vignette. We noticed that cell cycle regression, although clarifying anatomical diversity, seemed to induce low and high UMI clustering. For the E10.5 and E11.5 datasets, 2 replicates were generated from littermates and merged after confirming their similitude. For subsequent datasets (E12.5 and E14.5), no replicates were used. Annotation and subsetting were also performed in Seurat. “Myogenic” and “Non-myogenic” annotations were based on Pdgfa and Pdgfra expression and myogenic genes Myf5, Myod, and Myog. Cells not expressing Pdgfa were annotated as “non-myogenic” unless they express myogenic genes. Cells expressing Pdgfa were annotated as “myogenic”.

Usage Notes

This dataset contains 4 zip files:

-A zip file containing raw and filtered scRNAseq matrices, as well as preprocessed Seurat objects (.Rdata) of both whole dataset and subset of Mesp1-derived cells of the anterior part of 2 E10.5 mouse embryos (embryo 1 and 2). The subset is comprised of the "Anterior somites" and "Cardiopharyngeal mesoderm" clusters.

-A zip file containing raw and filtered scRNAseq matrices, as well as preprocessed Seurat objects (.Rdata) of whole dataset of Myf5-derived cells of the anterior part of 2 E11.5 mouse embryos (embryo 1 and 2).

-A zip file containing raw and filtered scRNAseq matrices, as well as preprocessed Seurat objects (.Rdata) of whole dataset Myf5-GFP cells of the anterior part of an E12.5 mouse embryo.

-A zip file containing raw and filtered scRNAseq matrices, as well as preprocessed Seurat objects (.Rdata) of both whole dataset Myf5-GFP cells of the anterior part of an E14.5 mouse embryo.

"cellType" and "myoVSnonmyo" clustering information can be found in seurat object metadata.