Johnson K et al. (2022), Quantitative analysis of transcriptome dynamics...

XB-ART-59405

BMC Genomics 2022 Oct 23;231:723. doi: 10.1186/s12864-022-08953-3.

Show Gene links Show Anatomy links

Quantitative analysis of transcriptome dynamics provides novel insights into developmental state transitions.

Johnson K , Freedman S , Braun R , LaBonne C .

Abstract
BACKGROUND: During embryogenesis, the developmental potential of initially pluripotent cells becomes progressively restricted as they transit to lineage restricted states. The pluripotent cells of Xenopus blastula-stage embryos are an ideal system in which to study cell state transitions during developmental decision-making, as gene expression dynamics can be followed at high temporal resolution. RESULTS: Here we use transcriptomics to interrogate the process by which pluripotent cells transit to four different lineage-restricted states: neural progenitors, epidermis, endoderm and ventral mesoderm, providing quantitative insights into the dynamics of Waddington's landscape. Our findings provide novel insights into why the neural progenitor state is the default lineage state for pluripotent cells and uncover novel components of lineage-specific gene regulation. These data reveal an unexpected overlap in the transcriptional responses to BMP4/7 and Activin signaling and provide mechanistic insight into how the timing of signaling inputs such as BMP are temporally controlled to ensure correct lineage decisions. CONCLUSIONS: Together these analyses provide quantitative insights into the logic and dynamics of developmental decision making in early embryos. They also provide valuable lineage-specific time series data following the acquisition of specific lineage states during development.

PubMed ID: 36273135
PMC ID: PMC9588240
Article link: BMC Genomics

Species referenced: Xenopus laevis
Genes referenced: a2m bix1.1 bmp2 bmp4 bmp7.1 dand5 dlx3 eomes evx1 foxi1 foxi2 grhl1 grn gsc gtpbp2 id3 krt12.4 mix1 msx1 otx1 otx2 post pou5f3.3 pygm smad1 smad2 sox11 sox17a sox17b.1 sox2 sox3 tgfb1 vegt ventx2 ventx2.2 wnt8a zic1
Morpholinos: dand5 MO1

GEO Series: GSE198598: NCBI

Article Images: [+] show captions

	Fig 1. Xenopus blastula explants can be reprogrammed to adopt any lineage. A Schematic of Waddington’s Landscape portraying lineage specification process. B Embryos stage 9,10,10.5,11,12, and 13 used to confirm developmental stages of blastula explants. C Schematic showing signaling molecules used to generate all 4 lineages. D-I RNA Seq TPM expression over time of (D) maternally provided pluripotency marker Pou5f3.3, (E) maternally provided Foxi2, (F) epidermal marker EpK, (G) neural marker Otx1, (H) mesoderm marker Brachyury(T) (I) endoderm marker Sox17. H Graphs are sums of S + L allele. Width of lines represents SEM of three biological replicates
	Fig 2. PCA and Time Series Analysis Reveal Novel Lineage-Specific Dynamics. A PCA for each individual lineage with the coordinates of PC1 and PC2 for each lineage plotted against developmental time. B PCA performed on all four lineages simultaneously, with plot showing clustering of all lineages for PC1 vs PC2. C Scree plot of the variance explained by the top10 principal components for PCA done on all lineages. D Distances from stage 9 to 13 for all 74 PCs for each lineage, error bars are SEM of all 9 stage 9–13 distances for each lineage, (*P < 0.005, P < .01). E Number of genes differentially expressed between successive stages in each lineage, padj ≤ 0.05
	Fig 3. Gene Expression Dynamics Provide Novel Insights into the Neural Default State. A Number of differentially expressed genes between successive developmental stages of the epidermal and neural lineage, padj ≤ 0.05 B Venn Diagrams for the total number of DE genes between stages for the epidermal and neural lineages, as well as the number of genes increasing and decreasing over time. C TPM of Foxi1, Foxi2 and Grhl1, revealing epidermal onset dynamics. Graph shows sum of S + L allele. Width of the line represents SEM of three biological replicates. D Number of differentially expressed genes between lineages at each developmental stage (padj ≤ 0.05). E Graphs of the TPM of three pluripotency markers maintained in the neural lineage (E) Sox3, Sox11, and Zic1. Graphs are sums of S + L allele. Width of the line represents SEM of three biological replicates
	Fig 4. Robust BMP Signaling is initiated in Explants Around the Onset of Gastrulation. A Western blot analysis of lysates of developing epidermal (WT) and neural (20uM K02288) explants for pSmad1/5/8 and Smad1 with Actin loading control. Gel was cropped to minimize empty space, uncropped gel is Supplemental Fig. 8A. B Significance of BMP overrepresentation (hypergeometric p-value) in temporally differentially expressed genes. C-D Graphs of BMP responsive genes in epidermal and neural lineages (C) Ventx2.1 (D) Id3. Graphs are sums of S + L allele. Width of the line represents SEM of three biological replicates
	Fig 5. Early Response to Activin and BMP4/7 Reveals Unexpected Overlap. A Number of differentially expressed genes between successive developmental stages of the endoderm and ventral mesoderm lineage, padj ≤ 0.05. B Venn Diagrams for the total number of DE genes between stages for the epidermal and neural lineages, as well as the number of genes increasing and decreasing over time. C-F Heatmaps of the top 50 genes differentially increased in response to (C) BMP4/7 at stage 10 and/or 10.5 (D) BMP4/7 at stage 11 and/or 12. (E) Activin at stage 10 and/or 10.5 (F) Activin at stage 11 and/or 12. All genes in heatmaps are ranked by Log2FC of differential expression of either mesoderm (BMP4/7) or endoderm (Activin) compared to the wildtype epidermal lineage. Only genes increasing in response to BMP4/7 or Activin with a minimum expression of 10TPM at the relevant stages and higher expression at the relevant stage than at stage 9 were included in the heatmap. Colors represent z-scores of TPM
	Fig 6. Time Series Data Provides Novel Insights into Mesendoderm GRN. A Heatmaps of genes in the published mesendoderm GRN across time in the endoderm and ventral mesoderm lineage with colors representing z-score of TPM expression in both lineages [106]. (B) Schematic of the timing of genes from published Mesendoderm GRN in both the endoderm and ventral mesoderm lineage, as defined by expression of at least 30 TPM in the L and S allele combined for the average of three biological replicates. C TPM of Bix1.3, Evx1, and Mix1 revealing mesoderm onset dynamics. D-G TPM of genes proposed as novel mesendoderm GRN members based on DESeq2 and Limma analysis (D) Tmcc1 (E) Nptx2 (F) Ca14 (G) Pygm. Graphs are sums of S + L allele. Width of the line represents SEM of three biological replicates
	Fig 7. Early BMP signaling drives ventral mesoderm rather than early epidermal divergence. A Western Blot Analysis of lysates for developing epidermal (WT) and ventral mesoderm (BMP4/7 20 ng/uL) explants for pSmad1/5/8 and Smad1 with Actin loading control. Gel was cropped to minimize empty space, uncropped gel is Supplemental Fig. 8B. B Number of differentially expressed genes between successive developmental stages of the epidermal and ventral mesoderm lineage, padj ≤ 0.05. C Venn Diagrams for the total number of DE genes between stages for the epidermal and ventral mesoderm lineages, as well as the number of genes increasing and decreasing over time. D-I TPM of genes representative of different expression dynamics in response to early BMP (D) Grhl1 (E) Ventx2.1 (F) Post (G) Vegt (H) Wnt8a (I) Mix1. Graphs are sums of S + L allele. Width of the line represents SEM of three biological replicates
	Fig 8. BMP Signaling is Restrained Until stage 10.5 by Dand5 Activity. A-B TPM of genes involved in BMP signaling (A) BMP heterodimer ligands BMP7.2, BMP4, BMP2 and primary BMP target Smad, Smad1 (B) maternally provided BMP antagonists Dand5 and Gtpbp2. Graphs are sums of S + L allele. Width of the line represents SEM of three biological replicates. C Western Blot Analysis of lysates for stage 10 epidermal (WT), Dand5 MO injected (40 pmol/embryo), ventral mesoderm (BMP4/7 20 ng/uL), and endoderm (160 ng/uL Activin) explants for pSmad1/5/8 and pSmad2 with Actin loading control. Gels were cropped to minimize empty space, uncropped gel is Supplemental Fig. 8C. D qRT-PCR of animal pole explants examining the fold change from stage 9 to 13 of expression of mesodermal marker Bra(T) for epidermis(WT), ventral mesoderm(BMP4/7 20 ng/uL) and Dand5MO(300pmo/embryo) injected(***P < 0.005)
	Supplemental Figure 1. Xenopus blastula explants can be reprogrammed to adopt any lineage. (A-E) RNA Seq TPM expression over time of (A) average of all monotonically decreasing genes with a >25fold decrease between stage 9 and 13 and maximum TPM of 20 at stage 13, (B) epidermal marker Dlx3, (C) neural marker Otx2, (D) mesoderm marker Evx1 (E) endoderm marker Endodermin. Graphs are sums of S+L allele. Width of lines represents SEM of three biological replicates.
	Supplemental Figure 2. Transcription Factor UpSet Plots and Zygotic PCA. (A-E) UpSet plots of transcription factors expressed at a minimum of 10 TPM at (A) stage 10, (B) stage 10.5, (C) stage 11, (D) stage 12, (E) stage 13. X-axes show genes unique to each lineage and overlapping in all different combinations of lineages ordered from largest number of genes to smallest. (F) PCA on only zygotically expressed genes.
	Supplemental Figure 3. Temporal DESeq UpSet Plots. (A-J) UpSet plots of Differentially Expressed Genes (A) increased between stages 9 and 10, (B) decreased between stages 9 and 10, (C) increased between stages 10 and 10.5, (D) decreased between stages 10 and 10.5, (E) increased between stages 10.5 and 11, (F) decreased between stages 10.5 and 11, (G) increased between stages 11 and 12, (H) decreased between stages 11 and 12, (I) increased between stages 12 and 13, (J) decreased between stages 12 and 13. X-axes show genes unique to each lineage and overlapping in all different combinations of lineages ordered from largest number of genes to smallest.
	Supplemental Figure 4. Epidermal and Neural Lineages Diverge at St 10.5 (A) Number of enriched Neural GO Terms in genes significantly higher in the neural lineage (blue) and epidermal lineage (red) at each developmental stage. (B) Heatmap of the top 20 DE genes by Log2FC with a minimum expression of 10TPM between the epidermal and neural lineages at stage 10.5 (C) KEGG enrichment analysis of genes differentially expressed between epidermal and neural lineage at each developmental stage. Genes significantly increased in the epidermal lineage are enriched for TGF-beta genes, as defined by KEGG database from stages 10.5-13.
	Supplemental Figure 5. BMP Ventralizes Mesoderm. (A)Western blot analysis of lysates of developing mesoderm (20ng/uL BMP4/7) and endoderm (160ng/uL Activin) explants for pSmad2 and Smad2 with Actin loading control. Gels were cropped to minimize empty space, uncropped gels are Supplemental Figure 9A. (B) PCA of published Activin induced mesoderm data (Satou-Kobayashi et al. 2021) with our ventral mesoderm and endoderm data shows that activin induced mesoderm clusters with our activin induced endoderm rather than BMP4/7 induced mesoderm. (C) Hierarchical clustering of Dorsal/Ventral genes (Ding et al. 2017) shows that endoderm and ventral mesoderm cluster on opposite sides of the heat map, suggesting BMP4/7 effectively ventralizes the mesoderm.
	Supplemental Figure 6. Mesendoderm Analysis. (A) Number of differentially expressed genes between the endoderm and ventral mesoderm lineages at each developmental stage (padj ≤ 0.05). (B) KEGG enrichment analysis of genes differentially expressed between ventral mesoderm and endoderm lineage at each developmental stage. Genes significantly increased in the endoderm lineage are enriched for TGF-beta genes, as defined by KEGG database from stages 10-10.5 and genes significantly higher in the ventral mesoderm lineage are enriched for TGFbeta genes for stages 11-12. (C) WGCNA on stages 10,10.5 and 11 in the mesoderm and endoderm lineages identifies 22 gene modules, the blue module demonstrating increasing correlation over time to the endoderm lineage and the brown module demonstrating increasing correlation over time to the mesoderm lineage.
	Supplemental Figure 7 Early not Exogenous BMP Causes Mesoderm Formation. (A) Number of differentially expressed genes between the epidermal and ventral mesoderm lineages at each developmental stage (padj ≤ 0.05). (B) Number of enriched Mesoderm GO Terms (solid line) and Epidermis GO Terms (dashed line) in genes significantly higher in the ventral mesoderm lineage (purple) and epidermal lineage (red) at each developmental stage. (C) Western Blot Analysis of lysates for epidermal (WT) and BMP4/7 treated at stage 9 (BMP4/7 20ng/uL) explants collected at stage 10 and epidermal (WT) and BMP4/7 treated at stage 10.5 (BMP4/7 20ng/uL) explants collected at stage 11 for pSmad1/5/8 with Actin loading control. Gel was cropped to show one replicate, uncropped gel is Supplemental Figure 9B. (D) qRT-PCR of animal pole explants examining the fold change from stage 9 to 13 of expression of mesodermal marker Bra(T) for epidermis(WT), treated with BMP4/7(20ng/uL) at stage 9 and treated with BMP4/7 (20ng/uL) at stage 10.5 (***P<0.005).
	Supplemental Figure 8. Raw Images of Main Figure Western Blots. (A) Figure 4A Western blot analysis of lysates of developing epidermal (WT) and neural (20uM K02288) explants for pSmad1/5/8 and Smad1 with Actin loading control. Blot was cut prior to antibody hybridization to conserve antibody, unedited scan showing blot edges provided. (B) Figure 7A Western Blot analysis of lysates for developing epidermal (WT) and ventral mesoderm (BMP4/7 20ng/uL) explants for pSmad1/5/8 and Smad1 with Actin loading control. Blot was cut prior to antibody hybridization to conserve antibody. Initial scan showing blot edges (top) and high resolution unedited scan of relevant size (bottom) both provided. (C) Figure 8C Western Blot of lysates for stage 10 epidermal (WT), Dand5 MO injected (40pmol/embryo), ventral mesoderm (BMP4/7 20ng/uL), and endoderm (160ng/uL Activin) explants for pSmad1/5/8 with Actin loading control. Blot was cut prior to antibody hybridization to conserve antibody. Initial scans showing blot edges and high resolution unedited scans of relevant size for all three replicates are provided. Figure 8C is replicate 1.
	Supplemental Figure 9. Raw Images of Supplemental Western Blots. (A) Supplemental Figure 5A Western blot analysis of lysates of developing mesoderm (20ng/uL BMP4/7) and endoderm (160ng/uL Activin) explants for pSmad2 and Smad2 with Actin loading control. Blot was cut prior to antibody hybridization to conserve antibody. Initial scan showing blot edges (top) and unedited high resolution scan of relevant size bands (bottom) both provided. (B) Supplemental Figure 7C Western Blot Analysis of lysates for epidermal (WT) and BMP4/7 treated at stage 9 (BMP4/7 20ng/uL) explants collected at stage 10 and epidermal (WT) and BMP4/7 treated at stage 10.5 (BMP4/7 20ng/uL) explants collected at stage 11 for pSmad1/5/8 with Actin loading control. Blot was cut prior to hybridization with antibody to conserve antibody. Initial scan showing blot edges (top) and unedited high resolution scan of revelant size bands (bottom) both provided. All three replicates are shown, Supplemental Figure 6A is Rep 1.