May 1, 2021;
A temporally resolved transcriptome for developing "Keller" explants of the Xenopus laevis dorsal marginal zone.
BACKGROUND: Explanted tissues from vertebrate embryos reliably develop in culture and have provided essential paradigms for understanding embryogenesis, from early embryological investigations of induction, to the extensive study of Xenopus animal caps, to the current studies of mammalian gastruloids. Cultured explants of the Xenopus dorsal marginal zone ("Keller" explants) serve as a central paradigm for studies of convergent extension cell movements, yet we know little about the global patterns of gene expression in these explants.
RESULTS: In an effort to more thoroughly develop this important model system, we provide here a time-resolved bulk transcriptome for developing Keller explants.
CONCLUSIONS: The dataset reported here provides a useful resource for those using Keller explants for studies of morphogenesis and provide genome-scale insights into the temporal patterns of gene expression in an important tissue
when explanted and grown in culture.
[+] show captions
Figure 1. Dissections, culturing, and preparation of explant RNA-Seq libraries. A, Diagram identifying where cuts were made for dorsal marginal zones (DMZ), ventral marginal zones (VMZ), and animal caps. B, Diagram showing flow of dissection and culturing until RNA extraction. C, Diagram showing DMZ explant culture and RNA collection timeline
Figure 2. Cultured DMZs can be temporally differentiated by their global transcriptomes. A, Venn diagram comparing genes enriched in the DMZ with respect to the VMZ in three independent studies. Differential expression analysis for each data set are reported followed by the genes representing the union of all three data sets. B, Heatmap showing gene expression of genes captured by the union of all three data sets. Color represents expression of gene scaled by row. C, PCA plot clustering all samples. Blue samples are cultured DMZ, green samples are cultured VMZ, and pink samples are cultured animal caps. D, Correlation heatmap demonstrating the calculated distance between samples. Hierarchical clustering was performed by Pheatmap
Figure 3. Variation in differential expression analysis pipelines can lead to low recall in differentially expressed genes. A, Venn diagram showing the overlap in the number of gene called enriched in DMZ stage 11 samples over VMZ samples. Box below venn diagram indicates main differences in differential expression analysis for each approach. B, Heatmap showing expression of 100 union genes between analysis workflows. Color represents expression of gene scaled by row
Figure 4. Cultured DMZs cluster to 5 k-means clusters. A, Heatmap of all differentially expressed genes clustered to 5 k-means clusters. B, Gene ontology analysis results from GO analysis on all genes per cluster. C, Heatmap of top 10 genes by FDR from each cluster. Color in heatmaps represent expression of gene scaled by row. D, Heatmap of the same genes in C, plotted from published data in the whole embryo at complimentary developmental timepoints.
Figure 5. Canonical markers used to study development of the Xenopus embryo. Heatmaps showing gene expression of canonical markers of development in (left) explants and (right) whole embryos (Session 2016). Color represents expression of gene scaled by row.
Figure 6. L and S homoeologues are differentially represented and may have differential expression between cognate pairs. A, Pie chart demonstrating the distribution of only L homoeologues, only S homoeologues, and genes with L and S homoeologues across all transcripts in our dataset. B, Distribution of homoeologues in only differentially expressed genes. Genes in the L and S category must have both L and S homoeologues be differentially expressed. C, Correlation plot for log(counts) of L vs S homoeologues for cognate pairs in all samples. The red number represents the Pearson correlation coefficient. D, Correlation of randomly sampled L and S homoeologue log(counts) from all samples. E, Histogram of Pearson correlation coefficients calculated for each differentially expressed gene with and L and S homoelogues across DMZ time. F, Heatmap representing the number of genes with L and S homoeologues with each unique cluster call pairing. G, Heatmap showing examples of congruent expression patterns of homoeologues called in the same cluster. H, Heatmap showing examples of expression patterns of homoeologues that were called in different clusters. Color represents expression of gene scaled by row in heatmaps from (G/H).
Figure 7. Known PCP pathway genes have little transcriptional concordance. A, Heatmap plotting gene expression of PCP pathway genes in (left) explants and (right) the whole embryo. B, Heatmaps plotting septin genes downstream of PCP. Heatmap chunks represent septins representing different protein complex subunits. Color represents expression of gene scaled by row.