Identification and characterization of intermediate states in mammalian neural crest cell epithelial to mesenchymal transition and delamination

Abstract

Epithelial to mesenchymal transition (EMT) is a cellular process that converts epithelial cells to mesenchymal cells with migratory potential in both developmental and pathological processes. Although originally considered a binary event, EMT in cancer progression involves intermediate states between a fully epithelial and a fully mesenchymal phenotype, which are characterized by distinct combinations of epithelial and mesenchymal markers. This phenomenon has been termed epithelial to mesenchymal plasticity (EMP), however, the intermediate states remain poorly described and it's unclear whether they exist during developmental EMT. Neural crest cells (NCC) are an embryonic progenitor cell population that gives rise to numerous cell types and tissues in vertebrates, and their formation is a classic example of developmental EMT. An important feature of NCC development is their delamination from the neuroepithelium via EMT, following which NCC migrate throughout the embryo and undergo differentiation. NCC delamination shares similar changes in cellular state and structure with cancer cell invasion. However, whether intermediate states also exist during NCC EMT and delamination remains unknown. Through single cell RNA sequencing, we identified intermediate NCC states based on their transcriptional signature and then spatially defined their locations in situ in the dorsolateral neuroepithelium. Our results illustrate the progressive transcriptional and spatial transitions from premigratory to migratory cranial NCC during EMT and delamination. Of note gene expression and trajectory analysis indicate that distinct intermediate populations of NCC delaminate in either S phase or G2/M phase of the cell cycle, and the importance of cell cycle regulation in facilitating mammalian cranial NCC delamination was confirmed through cell cycle inhibition studies. Additionally, transcriptional knockdown revealed a functional role for the intermediate stage marker Dlc1 in regulating NCC delamination and migration. Overall, our work identifying and characterizing the intermediate cellular states, processes, and molecular signals that regulate mammalian NCC EMT and delamination furthers our understanding of developmental EMP and may provide new insights into mechanisms regulating pathological EMP.

Keywords: Dlc1; EMP; EMT; cell cycle; epithelial to mesenchymal plasticity; epithelial to mesenchymal transition; mouse; neural crest cells.

Figure 1.

Single-cell RNA seq analysis of mouse early E8.5 cranial tissues. (A) Schematic of experimental design. Wnt1-Cre;RosaeYFP and Mef2c-F10N-LacZ embryos with between 7-9 somites (6 each) were dissected and cranial tissues anterior to rhombomere 3 were collected. Tissues were dissociated into single cell suspensions before being processed through the 10X Genomics pipeline. The final dataset used for analysis consisted of 21,190 cells (12,498 cells from Wnt1-Cre;RosaeYFP and 8,692 from Mef2c-F10N-LacZ) and 29,041 genes. (B) YFP and LacZ staining of E8.5 Wnt1-Cre;RosaeYFP and Mef2c-F10N-LacZ embryos and 10um cranial transverse sections. YFP (green) labels cells located in the dorsal neuroepithelium and their lineages. As a result, both premigratory and migratory NCC are marked by YFP expression. LacZ (blue) labels migratory NCC. (C) Uniform Manifold Approximation and Projection (UMAP) and clustering of 6 major tissue types in the cranial region of E8.5 mouse embryos: cranial NCC, neuroectoderm, non-neural ectoderm, mesoderm, endothelial cells, and embryonic blood cells. (D) Dotplot showing the expression of tissue specific markers used for cluster identification. Dot size indicates the percentage of cells in each corresponding cluster (y-axis) that expresses a specific gene (x-axis). Dot color intensity indicates the average expression level of a specific gene in a cell cluster.

Figure 2.

Expression of NCC development related genes and EMT functional genes identifies NCC EMT intermediate populations. (A) UMAP and re-clustering of the cranial NCC cluster into 5 smaller subclusters at a resolution of 0.26. (B) Dotplot showing the expression of NCC development related genes in 5 cranial NCC subclusters. (C) UMAP and re-clustering of the early migratory NCC subclusters 0, 1 and 4 into smaller subclusters at a resolution of 2.0. (D) Heatmap showing the expression of NCC development related genes in the smaller early migratory NCC subclusters at a resolution of 2.0 shown in (C). High levels of expression are indicated in yellow, and low levels of expression are indicated in pink. Based on the gene expression profile of each subcluster, subcluster 17’ was determined to be premigratory NCC; subcluster 2’ and 10’ are EMT intermediate NCC; the remaining subclusters are migratory NCC. (E) Dotplot showing the expression of EMT functional genes in premigratory NCC subcluster 17’ and intermediate NCC subclusters 2’ and 10’. EMT intermediate NCC display reduced expression of adherens junction, tight junction and apical basal polarity genes compared to premigratory NCC, whereas protrusion related genes are upregulated in intermediate NCC.

Figure 3.

Mouse cranial NCC delaminate in S phase or G2/M phase cell cycle independently. (A) Dotplot showing the expression of G2/M phase cell cycle genes in premigratory NCC (PM) and EMT intermediate NCC. G2/M phase cell cycle genes are expressed in PM and intermediate subcluster 10’ cells. (B) Dotplot showing the expression of S phase cell cycle genes in PM and EMT intermediate NCC. S phase cell cycle genes are expressed in PM and intermediate subcluster 2’ cells. (C) Pseudotime analysis of the cranial NCC cluster reveals the temporal relationship between intermediate NCC subcluster 2’ and 10’. Dark color indicates early NCC development, and light color indicates later NCC development. PM and intermediate NCC subclusters represent the earliest developmental timepoints among all cranial NCC. (D) Trajectory analysis of the cranial NCC cluster reveals lineage/fate relationship between PM and intermediate NCC subcluster 2’ and 10’. Two intermediate NCC subclusters develop simultaneously and independently from premigratory NCC. Apart from their cell cycle status, early migratory NCC formed from the different intermediate subclusters are transcriptionally indistinguishable. Color coding of the cell population is consistent with the re-clustering of the cranial NCC cluster into 5 smaller subclusters at a resolution of 0.26 as shown previously in Figure 2A.

Figure 4.

Cell cycle regulation plays an important role in mouse cranial NCC delamination and EMT. (A) Cell cycle marker staining of early E8.5 mouse embryonic cranial and trunk tissues reveals differences in cell cycle status between cranial delaminating premigratory NCC and trunk neural plate border cells. EdU (magenta) and pHH3 (cyan) staining were performed on 10um transverse sections of early E8.5 (5-7 somites) Wnt1-Cre;RosaeYFP mouse embryo cranial and trunk tissues. (B) E8.0 CD1 mouse embryos treated with Aphidicolin exhibit reduced migratory NCC that primarily express pHH3. Cranial sections of treated embryos were stained with Sox10, EdU and pHH3 (magenta) and arrowheads indicate migratory NCC expressing pHH3. Most remaining migratory NCC in Aphidicolin treated samples express pHH3. In contrast, a small proportion of migratory NCC in control DMSO treated samples express pHH3. (C) Cell cycle staining quantification of delaminating premigratory NCC in the cranial neural plate border shows that most cells express cell cycle markers. Staining and quantification were performed on delaminating premigratory NCC in the cranial neural plate border of 5-7 somite Wnt1-Cre;RosaeYFP mouse embryos (n=3). The neural plate border region was manually selected in the most dorsolateral domain of the neural plate. EdU+%=the percentage of EdU positive cells within eYFP positive delaminating premigratory NCC in the selected neural plate border domain. pHH3+%= the percentage of pHH3 positive cells within eYFP positive delaminating premigratory NCC. EdU+pHH3+%= the percentage of EdU and pHH3 double positive cells within eYFP positive delaminating premigratory NCC. EdU-pHH3-%= the percentage of EdU and pHH3 double negative cells within eYFP positive delaminating premigratory NCC. (D) Cell cycle staining quantification of trunk neural plate border cells shows that a significant proportion of cells do not express any cell cycle markers. Staining and quantification were performed on trunk neural plate border cells of 5-7 somite Wnt1-Cre;RosaeYFP mouse embryos (n=3). The neural plate border region was manually selected in the most dorsolateral domain of the neural plate. EdU+%=the percentage of EdU positive cells within DAPI positive neural plate border cells at the trunk axial level. pHH3+%= the percentage of pHH3 positive cells within DAPI positive trunk neural plate border cells. EdU+pHH3+%= the percentage of EdU and pHH3 double positive cells within DAPI positive trunk neural plate border cells. EdU-pHH3-%= the percentage of EdU and pHH3 double negative cells within DAPI positive trunk neural plate border cells. (E) Quantification of Sox10 expressing migratory NCC upon Aphidicolin and control treatment reveals fewer cranial migratory NCC in Aphidicolin treated embryos. Sox10 staining and quantification were performed on cranial sections of 4-6 somite CD1 mouse embryos post treatment (n=3 per treatment; ****p<0.0001). For quantification, we calculated the ratio of Sox10 positive migratory NCC over DAPI positive neural plate/neuroepithelial cells.

Figure 5.

SABER-FISH of EMT intermediate stage markers pinpoints the location of EMT intermediate NCC within the dorsal most region of the neural fold. (A) Dotplot showing the expression of selected EMT intermediate NCC markers in early migratory NCC subclusters (resolution 2.0). (B) SABER-FISH staining of premigratory, EMT intermediate stage and migratory NCC marker genes on the same section. Higher magnification insets of the left side neural fold (box) showing that Wnt1 is expressed in the neuroepithelium and Sox10 is expressed in migratory NCC populating the underlying mesenchyme. Dlc1, Sp5 and Pak3 are expressed in the dorsolateral most region of the neuroepithelium. (C) 2D map showing the number of transcripts per cell, calculated from the SABER-FISH staining. To evaluate the expression of each gene within and across tissues, a polyline kymograph was generated along the track indicated by the arrows at a width of 100 pixels. The polyline kymograph can be seen to the right of each neural fold map it depicts. At the beginning of the track, Wnt1 expression is highest, demarcating the dorsal lateral domain of the neuroepithelium. Towards the middle of the track, at the location of the most dorsolateral region of the neuroepithelium, Wnt1 is expressed along with the intermediate stage markers Dlc1, Sp5 and Pak3. As the track progresses to just outside of the neuroepithelium, Sox10 expression appears and increases as the track continues through the migratory NCC population.

Figure 6.

Dlc1 plays a regulatory role in mouse cranial NCC EMT and delamination. (A) Sox10 immunostaining was performed on cranial sections of E8.5 control and Dlc1 knockdown mouse embryos. (B) Dlc1 knockdown significantly reduced the number of migratory NCC compared to the control. The number of Sox10+ migratory NCC was quantified in control (n=4) and all Dlc1 knockdown (n=12) embryos. All data points in Dlc1 knockdown samples were normalized to the control samples. ****p<0.0001. (C) Dlc1 shRNA-based lentiviruses achieved an average of 30% reduction of Dlc1 expression in all Dlc1 knockdown embryos based on qRT-PCR analysis. *p<0.05. (D) TUNEL staining showed minimal cell death in Dlc1 knockdown samples.