Reprogramming Axial Level Identity to Rescue Neural-Crest-Related Congenital Heart Defects

Abstract

The cardiac neural crest arises in the hindbrain, then migrates to the heart and contributes to critical structures, including the outflow tract septum. Chick cardiac crest ablation results in failure of this septation, phenocopying the human heart defect persistent truncus arteriosus (PTA), which trunk neural crest fails to rescue. Here, we probe the molecular mechanisms underlying the cardiac crest's unique potential. Transcriptional profiling identified cardiac-crest-specific transcription factors, with single-cell RNA sequencing revealing surprising heterogeneity, including an ectomesenchymal subpopulation within the early migrating population. Loss-of-function analyses uncovered a transcriptional subcircuit, comprised of Tgif1, Ets1, and Sox8, critical for cardiac neural crest and heart development. Importantly, ectopic expression of this subcircuit was sufficient to imbue trunk crest with the ability to rescue PTA after cardiac crest ablation. Together, our results reveal a transcriptional program sufficient to confer cardiac potential onto trunk neural crest cells, thus implicating new genes in cardiovascular birth defects.

Keywords: aorticopulmonary septum; cardiac crest subcircuit; cardiac neural crest; congenital birth defects; ectomesenchymal fate; heart development; outflow tract; persistent truncus arteriosus; reprogramming; specification.

Fig. 1:. Chick cardiac neural crest ablation results in cardiovascular abnormalities.

(A) Dorsal view of a whole-mount stage HH9+ embryo (inset). Cardiac domain stained with neural crest marker Pax7 (Magenta) and neural tube marker Sox2 (Green). Dotted line indicates level of sections in B-D. (B) Transverse section through A shows cardiac neural crest residing in the dorsal neural tube in wild type embryo. (C-D) Transverse section through embryos after unilateral (C) and bilateral (D) dorsal neural fold ablation. (E-G) Ventral view of E3 control (E), unilaterally ablated (F), and bilaterally ablated (G) primary heart tubes. (E’-G’) Transverse sections through E-G stained with the muscle marker MF20 show uniform MF20 labeling in wild type (E’) heart but patchy expression (arrows) in ablated embryos (F’-G’). (H) Whole-mount image of an E6 chick embryo. Dotted line shows angle of sectioning in I-K. (I) Cross-section through the outflow tract of a wild type E6 embryo shows complete septation, with the aorticopulmonary septum (AoP) separating the aorta (Ao) from pulmonary trunk (PT). (J-K) Unilateral (J) and bilateral (K) cardiac crest ablation results in the failure of outflow tract septation, resulting in a single vessel emerging from the heart. (SMA; red, DAPI; blue) ot-otic placode, s-somites, At-atrium, V-ventricle, OFT-outflow tract, ect-ectoderm, nc-notochord, Hb-hindbrain, ov-otic vesicle, H-heart, F-forelimb. See also Fig.S1.

Fig. 2:. Bulk and single cell transcriptional profiling of cardiac neural crest.

(A-B) Pure populations of cardiac or trunk neural crest cells labeled with the FoxD3-NC2 enhancer were isolated using FACS at HH12 (A) and HH18 (B), respectively. (C-C’) Volcano plot showing fold change and significance of genes enriched in cardiac and trunk neural crest (logFoldChange>1). Transcription factors examined for expression in migrating cardiac crest are highlighted in (C’). (D-I) In situ hybridization of HH12 embryos shows expression of selected transcription factors in migratory cardiac crest (dorsal view; arrows). (J) UMAP plot depicting clustering of 156 single cardiac neural crest cells that were profiled using smart-seq v2. (J’-J”) Cluster C1 and C2 exhibit high expression of neural crest markers Sox10 (J’) and FoxD3 (J”) and differed subtly in the expression of proliferation genes. (K) GO term analysis on differentially upregulated genes in each cluster confirms identity of each subpopulation within the migrating cardiac crest (Fisher’s exact test, adjusted-p<0.05). (L) Heatmap illustrating hierarchical clustering of single cells that passed filtering parameters and expression levels of selected neural crest, ectomesenchymal, and neuronal genes. A few cells from the cardiac crest progenitor clusters C1 and C2 grouped together with cells from the ectomesenchymal cluster C3. (M-N) Pseudotime analysis (M) on migratory cardiac crest cells. Cells are labeled according to their pseudotime values (N). ot-otic vesicle, cnc-cardiac neural crest, tnc-trunk neural crest, Hb-hindbrain. See also Fig.S1; table S1.

Fig. 3:. Tgif1 is critical for cardiac neural crest specification and outflow tract septation.

(A-D) Spatiotemporal expression pattern of Tgif1 in cardiac neural crest between stages HH9+ and HH12 (dorsal view). (E) Transverse section through (B) shows Tgif1 expression in delaminating cardiac crest cells in the dorsal neural tube. (F-F’) Transverse section through (D) shows Tgif1 expression in migrating cardiac crest which overlaps with the expression of the neural crest marker HNK1 (F’). (G) Diagram depicting ex ovo electroporation strategy for Tgif1 knockout in gastrula stage embryos. (H) An embryo transfected with gRNAs targeting Tgif1 (magenta) and control gRNA (green) (dorsal view). (I-N) Following CRISPR-Cas9-mediated knockout of Tgif1, expression of the indicated transcription factors was significantly reduced (dorsal view). Arrows indicate normal gene expression on control side. (O) Experimental strategy to investigate the role of Tgif1 in outflow tract septation. Cardiac neural folds from a transgenic embryo where Tgif1 was bilaterally knocked out at the gastrula stage were grafted in place of the ablated cardiac crest in a stage-matched HH9 wildtype host. (P) Cardiac crest cells from control gRNA-electroporated transgenic graft are observed within the condensed mesenchyme of the properly septated outflow tract. (Q) The OFT of embryos grafted with the Tgif1 knockout cardiac neural fold failed to septate. Hb-hindbrain, ot-otic placode, ect- ectoderm, nt-neural tube, nc-notochord, cnc-cardiac neural crest, Ao-aorta, PT-pulmonary trunk, OFT-outflow tract. See also Fig.S2.

Fig. 4:. Tgif1 is coexpressed with Sox8 and Ets1 in migrating cardiac neural crest.

(A-B”) HCR against Sox8 (B), Tgif1 (B’), and Ets1 (B”) shows their overlapping expression in the migrating cardiac crest cells. (C-F) Expression levels of Sox8 (C), Tgif1 (D), and Ets1 (E) in individual cardiac crest cells following deep single-cell profiling. In progenitor cardiac crest clusters, the three genes had overlapping expression (orange) in 93% of the cells (F). (G-I) Spatiotemporal expression pattern of Sox8 (G-G’”), Tgif1 (H-H’”), and Ets1 (I-I’”) in the cardiac crest region between stages HH9+ and HH12 (dorsal view). (J) Schematic diagram of HH12 embryo shows overlapping expression domains of Tgif1, Ets1, and Sox8 in the cardiac neural crest. (K) The hierarchy of temporal expression dynamics of cardiac crest subcircuit genes from HH9 to 12. Ot-otic vesicle, Hb-hindbrain, nt-neural tube, cnc-cardiac neural crest, r4-rhombomere 4 migrating crest stream. See also Fig.S2,S3.

Fig. 5:. Sox8r Tgif1 and Ets1 comprise a transcriptional cascade important for cardiac crest identity.

Dorsal view of HH11 embryos where Sox8 (A, E, J), Tgif1 (B, F, K), or Ets1 (C, G, L) was knocked out on the right side at gastrula stage. Sox8 knockout resulted in loss of Tgif1 (E) and Ets1 (J), Tgif1 knockout resulted in reduced Ets1 (K), but no noticeable change in Sox8 (B); Ets1 knockout had no effect on either Sox8 (C) or Tgif1 (G). (D-D’) Expression of Tgif1 on the right side in Sox8 knockout embryos (D’) was sufficient to partially rescue Ets1 expression (D). (H-I) Mutation of Sox8 binding sites in a Tgif1 enhancer resulted in reduced enhancer activity on the right side (I). H2B-RFP (H, M) was used as a transfection control on both sides. (N) Mutation of Tgif1 binding site in an Ets1 enhancer resulted in reduced enhancer activity in cardiac but not cranial neural crest. (O) Functional relationships between Sox8, Tgif1, and Ets1 in a transcriptional subcircuit based on the results in A-N. (P-S) Pseudotime lineage trajectory with each subcluster labeled (P) according to scRNA-seq analysis. Expression of cardiac neural crest subcircuit genes Sox8, Tgif1, and Ets1 was overlaid on the trajectory. (T) Heatmap showing changes in gene expression profiles of reprogrammed trunk neural crest compared to wildtype cardiac and trunk crest. (U-V) Dorsal view of an HH9+ embryo where Tgif1, Ets1, and Sox8 were overexpressed at gastrula stage on the right side (U). Ex ovo culturing of the embryos resulted in ectopic expression of cardiac crest gene Nacc2 in cranial neural crest (V) and surrounding naïve ectoderm. (W-X) Dorsal view of the trunk neural crest of an HH18 embryo where cardiac crest subcircuit genes were transfected on the right side of the neural tube in ovo (W). Expression of trunk neural crest marker Hes6 was reduced in resident trunk neural crest cells (X). Ot-otic vesicle, Hb-hindbrain, tnc-trunk neural crest, rep-reprogrammed. See also Fig.S4;table S2.

Fig. 6:. Reprogrammed trunk neural crest cells exhibit cardiac crest-like migratory behavior.

(A) Grafting strategy to test behavior of transplanted cells. Stage-matched cardiac neural fold, unperturbed trunk neural fold, or trunk neural fold reprogrammed by the ectopic expression of Sox8, Tgif1, and Ets1 was grafted in place of ablated cardiac neural folds in a wildtype host at HH9+. (B) Embryo with a unilateral transgenic implant spanning the cardiac crest domain immediately after grafting (dorsal view). (B’-B’”) Transverse section through the hindbrain of a chimera showing successful incorporation of the GFP⁺ implant (B”). The graft was electroporated with expression constructs for cardiac neural crest subcircuit genes Sox8, Tgif1, and Ets1 (B’”). (C-H) Whole-mount and cross-section images of chimeric embryos grafted with dorsal cardiac neural folds (C,F,F’), non-transfected dorsal trunk neural folds (D,G,G’), and ‘reprogrammed’ dorsal trunk neural folds (E,H,H’) harvested 2 days post grafting. Transgenic cells from the cardiac neural fold graft migrated into branchial arches III-IV, whereas cells from the trunk neural fold graft exhibited restricted migration. Ectopic expression of the cardiac crest subcircuit was sufficient to change the migration behavior of these cells, with transgenic cells populating branchial arches III and IV. Hb-hindbrain, ot-otic placode, Cv-cardinal vein, DAo-dorsal aorta, BA-branchial arch.

Fig. 7:. Reprogrammed trunk neural crest acquires cardiac ectomesenchymal potential.

(A-C) Transverse sections through embryos grafted with GFP⁺ cardiac neural folds show cardiac crest-derived cells in the aorticopulmonary septum (A, arrowheads in A’), tunica media surrounding the branchial arch arteries (B, arrowheads in B’), and neurons and satellite cells of the jugular ganglion of the vagus nerve (C,C’; orange arrowhead-satellite cells, yellow arrowhead-neurons). (D-F) Embryos grafted with non-transfected trunk neural fold cells exhibit PTA (D,asterisk), abnormally constricted right-fourth branchial arch artery (E,asterisk; arrowhead in E’), and improper gangliogenesis (F, asterisk). Transplanted cells were found associated with ectopic cervical nerves (F’, orange arrowheads). (G-I) GFP⁺ cells from the reprogrammed trunk implant migrate into the outflow tract and contribute to the aorticopulmonary septum (G, arrowheads G’). GFP⁺ cells were found surrounding vessels (H, arrowheads H’), and in neurons and satellite cells of the jugular ganglion (I,I’; orange arrowhead-satellite cells, yellow arrowhead-neurons). Ao-aorta, PT-pulmonary trunk, AoP-aorticopulmonary septum, OFT-outflow tract, Hb-hindbrain, nc-notochord.