Early lineage restriction in temporally distinct populations of Mesp1 progenitors during mammalian heart development

Abstract

Cardiac development arises from two sources of mesoderm progenitors, the first heart field (FHF) and the second (SHF). Mesp1 has been proposed to mark the most primitive multipotent cardiac progenitors common for both heart fields. Here, using clonal analysis of the earliest prospective cardiovascular progenitors in a temporally controlled manner during early gastrulation, we found that Mesp1 progenitors consist of two temporally distinct pools of progenitors restricted to either the FHF or the SHF. FHF progenitors were unipotent, whereas SHF progenitors were either unipotent or bipotent. Microarray and single-cell PCR with reverse transcription analysis of Mesp1 progenitors revealed the existence of molecularly distinct populations of Mesp1 progenitors, consistent with their lineage and regional contribution. Together, these results provide evidence that heart development arises from distinct populations of unipotent and bipotent cardiac progenitors that independently express Mesp1 at different time points during their specification, revealing that the regional segregation and lineage restriction of cardiac progenitors occur very early during gastrulation.

Figure 1. Mesp1-rtTA transgenic mice faithfully recapitulates Mesp1 endogenous expression.

a. Macroscopic analysis of a Mesp1-Cre/Rosa-tdTomato embryo at E14.5. Scale bars: 500μm. b-c. Confocal analysis of Rosa-tdTomato (b) and Mesp1-Cre/Rosa-tdTomato heart sections (c) at E14.5 co-stained with anti-cardiac troponin T (cTnT) antibody. d-g. Confocal analysis of Mesp1-Cre/Rosa-tdTomato heart sections at E14.5 co-stained with epicardial (Wt1) (d), EC (endoglin) (e), pace-maker (Hcn4) (f) and SMC (smMHC) (g) markers. Scale bars: 20 μm. lu: lumen, V: ventricle, A: atria, OFT, outflow tract, IFT, inflow tract. h. Scheme of the genetic strategy used for the characterization of the Mesp1-rtTA transgenic mice. DOX administration leads to the activation of the Cre recombinase between E6.25 and E7.5 in Mesp1-rtTA/TetO-Cre/Rosa-tdTomato but no activation of the Cre recombinase was detected when DOX was administrated later (E8.5). i-j. Confocal analysis of Rosa-tdTomato (i) and Mesp1-rtTA/tetO-Cre/Rosa-tdTomato heart sections (j) at E14.5 co-stained with anti-cardiac troponin T (cTnT). k-n. Confocal analysis of Mesp1-rtTA/TetO-Cre/Rosa-tdTomato heart sections at E14.5 co-stained with epicardial (Wt1) (k), EC (endoglin) (l), pace-maker (Hcn4) (m) and SMC (smMHC) (n) markers. Scale bars: 20 μm. lu: lumen, V: ventricle, OFT, outflow tract, SAN, sino-atrial node. o. Temporal analysis of the activation of the Mesp1-rtTA transgene. While absence of Dox administration did not induce GFP expression in the embryos, GFP positive cells could be detected only 5h after Dox injection in the primitive streak (PS) and nascent mesoderm. A, anterior; P, posterior. p. Temporal analysis of the recombination of the Rosa-tdTomato locus investigated by PCR following Dox administration. The Rosa-tdTomato locus was recombined as soon as 6h following Dox administration in Mesp1-rtTA/TetO-Cre/Rosa-tdTomato embryos at E6.25 and E7.25, similarly as found with Mesp1-Cre/Rosa-tdTomato embryos at the same time points. Negative controls including WT tail and Rosa-tdTomato tail show PCR amplification corresponding to the unrecombined Rosa-tdTomato locus (around 1,000bp) and Mesp1-Cre/Rosa-tdTomato heart at E12.5 (positive control) show recombined Rosa-tdTomato locus (about 180bp).

Figure 2. Two temporally distinct populations of Mesp1 progenitors contribute to the FHF and SHF development.

a. Scheme of the genetic strategy used for the clonal tracing of Mesp1 expressing progenitors with different fluorescent proteins to assess their regional contribution. b. Low dose of doxycycline (DOX) was injected between E6.25 and E7.25. Induced Mesp1-rtTA/tetO-Cre/Rosa-Confetti unicolour embryos were analyzed at E8.5 and E12.5. c. Proportion of the fluorescent proteins in unicolour-labelled hearts. (n=7 unicolour hearts at E8.5 and n=37 unicolour hearts at E12.5). d-e. Examples of Mesp1-rtTA/tetO-Cre/Rosa-Confetti unicolour labelled hearts at E8.5. f-g. Examples of Mesp1-rtTA/TetO-Cre/Rosa-Confetti unicolour labelled hearts at E12.5. Note that each patch is localized within either the FHF or the SHF but no unicolour patches that encompassed derivatives of the FHF and the SHF were observed. OFT, outflow tract; RV, right ventricle; LV, left ventricle; RA, right atrium; LA, left atrium; IFT, inflow tract. Scale bars: 200 μm. h-j. Examples of E12.5 unicolour hearts induced at E6.25 (H) and E6.75 (I) showing the labelling of FHF derived progenitors, while Dox administration at E7.25 shows preferential labelling of SHF progenitors (J). Scale bars: 200 μm. k. Graph depicting in all unicolour hearts the regional contribution of the labelled cells and the number of clusters of labelled cells per chamber according to the developmental time of Dox administration. * asterisks indicates that labelling was also detected in the epicardial layer. l. Quantification of the regional (FHF and SHF) contribution of patches of Mesp1 labelled cells in unicolour hearts shows the preferential labelling of the FHF (red) during Dox administration at the early time points (E6.25 and E6.75), while Dox administration in the late stage of cardiac progenitor specification (E7.25) shows the preferential labelling of Mesp1 progenitors that contribute to the SHF (green) derivatives. The number on the upper right in each panel refers to the ID of the labelled heart.

Figure 3. Bio-statistical modeling of the the multicolour labelled hearts.

a. Scheme of the genetic strategy used for the clonal tracing of Mesp1 expressing progenitors with different fluorescent proteins b. Low dose of doxycycline (DOX) was injected at E6.25, E6.75 or E7.25. Multicolour induced hearts were analyzed at E12.5 and classified according to their regional contribution. c. Upon Dox administration, Mesp1 expressing cells are stochastically labelled in different colours. During early development, cells migrate and are rearranged such that growing clones may fragment into disconnected clusters. d. Statistical analysis of uni- and multicolour hearts was performed to infer induction frequency (pN) and the fragmentation rate (f). e. The stochastic nature of the lineage labelling and fragmentation results in a broad distribution of fragment numbers (squares). With an induction frequency, pN=1.3, and the fragmentation rate, f=1.6, the statistical model (solid line) is in excellent agreement with the experimental data. n=263 hearts by colour. f. Statistical analysis, allows to restrict the analysis to fragments that are likely to be monoclonal with a known error rate of 12% (Supplementary Fig. S5c and Theory). g-h. Examples of E12.5 multicolour hearts induced at E6.25 (g), or E7.25 (h). Scale bars: 200 μm. In the right corner is indicated which colour is considered as clonal, based on the statistical analysis. We compare the probability L(m = 1|k) that k fragments stem from a single clone (black line) with the probability L(m > 1|k) that these fragments stem from more than one cell (solid blue line). The latter is given by the sum contributions of clones with multiple cell origin (dashed blue lines). We consider k fragments as monoclonal, if L(m = 1|k) > L(m > 1|k), which leaves us with a threshold value of k = 3 (dashed grey line). The circles denote fragment numbers of the three fluorescent markers in examples shown. i. Regional contribution of FHF and SHF progenitors in monoclonal datasets (n=89), showing the contribution of the FHF and SHF progenitors to other cardiac regions. j. Temporal appearance of FHF and SHF progenitors inferred from all datasets at each induction time (n=263 hearts by colour). The number on the bottom right in each panel refers to the ID of the labelled heart. Error bars indicate one sigma Poisson confidence intervals.

Figure 4. Clonal analysis of lineage differentiation of Mesp1 derived progenitors in vivo.

a. Scheme of the genetic strategy used for the clonal tracing of Mesp1 expressing progenitors with different fluorescent proteins to assess their fate. b. Low dose of doxycycline (DOX) was injected to the pregnant female between E6.25 and E7.25 and induced Mesp1-rtTA/tetO-Cre/Rosa-Confetti embryos were analyzed at E12.5 for the expression of markers specific of the different cardiovascular lineages of the heart: CMs (cTnT), ECs (Endoglin) and SMCs (smMHC). c. Fate of the labelled cells in the different sectioned hearts is assessed by confocal analysis of co-immunostaining of the three markers in a given cluster. The localization of the patches within the different heart chambers and their FHF and SHF origin are indicated below. OFT, outflow tract; RV, right ventricle; LV, left ventricle; RA, right atrium; LA, left atrium. d-i. Confocal analysis of serial sections of fluorescently labelled hearts co-stained for CM and EC markers show that clones in the LV differentiated only into either CM (d) or EC fate (f), and no FHF progenitors show clones positive for CM and EC markers. h-i. In contrast, bipotent clones presenting the ability to differentiate at the clonal level into either CMs (h) and ECs (h’) or CMs (i) and SMCs (i’) can be observed in the SHF. Arrowheads point to double marked cells. Scale bars: 20 μm. j. Percentage of labelling in the epicardium in unicolour hearts depending on the time of induction. k-l. Examples of E12.5 unicolour hearts showing labelling in the epicardial layer only (k) or in the epicardium and myocardium (l). Scale bars: 200 μm. The number on the upper right in each panel refers to the ID of the labelled heart.

Figure 5. Molecular signature of early and late Mesp1 expressing cells in vivo.

a. Genetic and cell-sorting strategy used to assess the molecular signature of early and late Mesp1 expressing cells in vivo. Induced Mesp1-rtTA/TetO-H2B-GFP embryos at E6.25 or E7.25 were dissected 6h after Dox administration. GFP positive (GFP+) and negative (GFP-) cells were isolated by FACS and microarrays analyses were performed in two independent biological experiments. b. GSEA of Mesp1-GFP signature at E6.5 showing the distribution of genes upregulated by Mesp1 overexpression in ESC 6 (left) or the genes upregulated in ES Mesp1-GFP 7 (right). Genes are shown within the rank order list of all the microarray probe sets of E6.5 GFP+ cells. The highly significant enrichment score (ES) and normalized enrichment score (NES) are shown for each analysis. c. Gene ontology enrichment in Mesp1-GFP expressing cells at E6.5 (black) or E7.5 (grey). d. Expression of early mesodermal markers, Mesp1, EMT markers such as Snai1 and cardiac progenitor markers in E6.5 Mesp1 GFP+ cells as measured by microarrays. The fold change is presented over the GFP-population in duplicate samples. e. Surface marker expression in E6.5 Mesp1 GFP+ cells as measured by microarrays. f-i. FACs analysis showing GFP expression in E6.75 Mesp1-rtTA/TetO-H2B-GFP embryos 6 hours following Dox administration (f). FACs analysis of the combined expression of Cxcr4 (blue), Pdgfra and Flk1 expression in the all living cells (g), in GFP- (h) and Mesp1 GFP+ (i) populations, shows that the GFP+ population is enriched in triple positive (TP) cells. The percentage of cells in each quadrant is shown and the percentage of Pdgfra+/Flk1+/ Cxcr4+ cells is shown in brackets. j. FACs analysis of E6.75 Mesp1-rtTA/TetO-H2B-GFP embryonic cells showing that the Flk1+/Pdgfra+ double positive (DP) cells (red) and Flk1+/Pdgfra+/Cxcr4+ (TP) triple positive cells (blue) are highly enriched in Mesp1-GFP expressing cells. k. Comparison od Mesp1 expressing cells at E6.5 and E7.5. Dot plot representing the signal of each probe (merge of the two duplicates) showing that some key developmental genes are differentially expressed between E6.5 and E7.5. l-m. mRNAs expression at E6.5 and E7.5, as defined by microarray analysis. Genes upregulated at E6.5 (l) and at E7.5 (m).

Figure 6. Different temporal expression of Mesp1 direct target genes

a. qRT-PCR analysis of Mesp1 target genes 24h after Dox administration in Dox inducible Mesp1 expression cells at D2 of ESC differentiation. The fold change is presented over the unstimulated cells (n=2 duplicates). b-h. Mesp1-Chip-Seq for Snai1 (b), Gata6 (c), Gata4 (d), Aplnr (e), Myl7 (f), Hoxb1 (g) and Foxc2 (h), showing that these genes are direct target genes of Mesp1 in ES cells. Red bars indicate significant peaks. i-j. Single cell RT-PCR analysis of Snai1, Gata6, Gata4, Aplnr, Myl7, Hoxb1 and Foxc2 as well as Etv2 in Mesp1 GFP+ cells at E6.5 (i) and E7.25 (j). β-actin and Mesp1 were used as internal positive controls. A dark colour indicates strong expression while a light colour indicates a weak expression (Supplementary Fig. S6). Blank cells indicate that no PCR amplification of the genes was detected. Percentages of cells expressing the markers are indicated on the right.

Figure 7. Revised model of the early step of cardiovascular progenitor specification and lineage commitment during mouse development.

Clonal and molecular analysis of Mesp1 progenitors shows the existence of temporally distinct Mesp1 progenitors that contribute to the heart development. Mesp1 progenitors first gives rise to the FHF (in red) and then to the SHF (in green) progenitors with an overlapping expression of Mesp1 in the two populations at E6.75. FHF progenitors are unipotent and give rise to either CMs or ECs. SHF progenitors are either unipotent or bipotent. Epicardial and epicardial derived cells (EPDCs) arises as an independent Mesp1 derived lineage at the early time points.