Induced pluripotent stem cell modelling of HLHS underlines the contribution of dysfunctional NOTCH signalling to impaired cardiogenesis

Abstract

Hypoplastic left heart syndrome (HLHS) is among the most severe forms of congenital heart disease. Although the consensus view is that reduced flow through the left heart during development is a key factor in the development of the condition, the molecular mechanisms leading to hypoplasia of left heart structures are unknown. We have generated induced pluripotent stem cells (iPSC) from five HLHS patients and two unaffected controls, differentiated these to cardiomyocytes and identified reproducible in vitro cellular and functional correlates of the HLHS phenotype. Our data indicate that HLHS-iPSC have a reduced ability to give rise to mesodermal, cardiac progenitors and mature cardiomyocytes and an enhanced ability to differentiate to smooth muscle cells. HLHS-iPSC-derived cardiomyocytes are characterised by a lower beating rate, disorganised sarcomeres and sarcoplasmic reticulum and a blunted response to isoprenaline. Whole exome sequencing of HLHS fibroblasts identified deleterious variants in NOTCH receptors and other genes involved in the NOTCH signalling pathway. Our data indicate that the expression of NOTCH receptors was significantly downregulated in HLHS-iPSC-derived cardiomyocytes alongside NOTCH target genes confirming downregulation of NOTCH signalling activity. Activation of NOTCH signalling via addition of Jagged peptide ligand during the differentiation of HLHS-iPSC restored their cardiomyocyte differentiation capacity and beating rate and suppressed the smooth muscle cell formation. Together, our data provide firm evidence for involvement of NOTCH signalling in HLHS pathogenesis, reveal novel genetic insights important for HLHS pathology and shed new insights into the role of this pathway during human cardiac development.

Figure 1

Characterization of HLHS patient specific iPSC lines. (A) Brightfield images of two clones derived from each patient using the non-integrative RNA based Sendai virus; (B) Representative example of flow cytometric analysis of key pluripotent cell markers, TRA-1-60 and NANOG. A representative example is shown from HLHS1 clone 1, however similar results were obtained for all HLHS and unaffected iPSC lines; (C) Elimination of Sendai viral vectors from derived iPSC lines (KOS stands for KLF4, OCT4 and SOX2); (D) Immunofluorescent images showing antibody staining of cells derived from all three germ layers: neuronal class β-III-tubulin staining neuronal cells derived from ectoderm; α-Smooth muscle actin staining smooth muscle cells derived from mesoderm αFP (alpha-fetoprotein) staining endodermal cells. A representative example is shown from HLHS2 clone 1; however similar results were obtained for all HLHS and unaffected iPSC lines. (E) In vivo differentiation of HLHS1-clone 1 as a representative example of teratoma forming ability of HLHS patient specific iPSC lines. (A) low power showing heterogeneous structure of the teratoma; (B) neuroepithelium; (C) cartilage; (D) intestinal epithelium. (A: scale bar = 400 µm, B-D: scale bar = 50 µm).

Figure 2

Differentiation of iPSC lines to cardiomyocytes. (A) Overview of the protocol used for differentiation of control and HLHS patient specific iPSC to cardiomyocytes; (B) Representative immunocytochemical staining of cardiomyocytes derived from control and HLHS patient specific iPSC lines using antibodies raised against alpha actinin and cardiac troponin. One clone from each patient and unaffected control is shown; however similar data were obtained for the second clone. Scale bar = 100 µm.

Figure 3

HLHS iPSC lines show an impaired ability to give rise to cardiomyocytes. (A) Analysis of the percentage and beating frequency of EBs that generate contraction during differentiation of control and HLHS patient specific iPSC. At least 100 EBs were assessed for each biological replicate; (B) Flow cytometry data analysis demonstrating a lower ability of HLHS-iPSC lines to give rise to mesodermal and cardiac progenitors when compared to control derived cells at day 7 of the differentiation time-course; (C) Flow cytometry data analysis demonstrating that HLHS-iPSC lines have a lower ability to give rise to cardiomyocytes but have an enhanced ability to differentiate to smooth muscle cells when compared to control derived cells at day 14 of the differentiation time-course. One way Anova analysis with Dunnett multiple comparison tests was performed. ***P < 0.001; **P value between 0.001 and 0.01; *P value between 0.01 and 0.05.HLHS iPSC lines: N = 6 (2 clones × triplicate biological repeats); control iPSC lines: N = 12 (2 clones × triplicate biological repeats × 2 unaffected control iPSC lines).

Figure 4

(A–C) Transmission electron microscopy ultrastructure of cardiomyocytes derived from an age matched control iPSC line (SB-Neo3) shows abundant myofibrillar bundles, regular transverse Z bands with intervals of around 2 µm (black arrows) and large amount of mitochondria (labeled ‘m’) similar to those observed in fetal or neonatal cardiomyocytes. (D-R) A more random arrangement of myofibrils with poorly defined Z bands (black arrows) was observed in HLHS-iPSC-derived cardiomyocytes. The interval between individual Z bands is frequently irregular and smaller (0.5–1.0 µm). Compared with control iPSC-CMs, HLHS-iPSC derivatives had less mitochondria (labeled ‘m’) of smaller size and mal-formed inner membrane. One clone from each HLHS patient is shown; however similar data were obtained from the second clone.

Figure 5

All HLHS patients harbour deleterious variants in genes involved in the NOTCH signalling pathway. (A) Summary of NOTCH receptor variants identified in HLHS patients. Published NOTCH variants associated with HLHS are shown in cyan, published variants associated with congenital heart disease are shown in green and published variants not associated with heart disease are shown in red. Novel variants uncovered by our exome analysis and not reported to date are shown in black font. (B) Confirmation of variants by direct sequencing in HLHS patient and unaffected controls (representative examples). These mutations were not present in the controls or other HLHS samples.

Figure 6

Reduced expression of key components of NOTCH signalling pathway in HLHS iPSC lines. Quantitative RT-PCR analysis performed on day 14 iPSC- derived cardiomyocytes showing decreased expression of NOTCH receptors, NOTCH ligands and targets in HLHS-iPSC-derived cardiomyocytes when compared to unaffected controls. Data are presented as mean+/- SEM. The values for fetal heart sample were set to 1 and all other values were normalised against this. One way Anova analysis with Dunnett multiple comparison tests was performed. ****P < 0.0001; ***P < 0.001; ** P value between 0.001 and 0.01 *P value between 0.01 and 0.05. HLHS iPSC lines: N = 6 (2 clones × triplicate biological repeats), control iPSC lines: N = 12 (2 clones × triplicate biological repeats × 2 unaffected control iPSC lines).

Figure 7

Activation of NOTCH signalling pathway restores the ability of HLHS- iPSC lines to give rise to cardiomyocytes. (A) Analysis of the percentage of EBs that generate contraction during differentiation of Jagged and scrambled peptide treated HLHS-iPSC line; (B) Analysis of beating frequency of EBs that generate contraction during differentiation of Jagged and scrambled peptide treated HLHS-iPSC line; (C) Flow cytometry data analysis demonstrating that Jagged treatment enhances the ability of HLHS-iPSC to give rise to cardiomyocytes and reduces the ability to differentiate to smooth muscle cells when compared to scrambled peptide treated control cells at day 14 of the differentiation time-course. T-test analysis was carried out, ***P < 0.001; **P value between 0.001 and 0.01; *P value between 0.01 and 0.05. Data are presented as mean+/- SEM, n = 3+.