Inducible apelin receptor knockdown reduces differentiation efficiency and contractility of hESC-derived cardiomyocytes

Abstract

Aims: The apelin receptor, a G protein-coupled receptor, has emerged as a key regulator of cardiovascular development, physiology, and disease. However, there is a lack of suitable human in vitro models to investigate the apelinergic system in cardiovascular cell types. For the first time we have used human embryonic stem cell-derived cardiomyocytes (hESC-CMs) and a novel inducible knockdown system to examine the role of the apelin receptor in both cardiomyocyte development and to determine the consequences of loss of apelin receptor function as a model of disease.

Methods and results: Expression of the apelin receptor and its ligands in hESCs and hESC-CMs was determined. hESCs carrying a tetracycline-inducible short hairpin RNA targeting the apelin receptor were generated using the sOPTiKD system. Phenotypic assays characterized the consequences of either apelin receptor knockdown before hESC-CM differentiation (early knockdown) or in 3D engineered heart tissues as a disease model (late knockdown). hESC-CMs expressed the apelin signalling system at a similar level to the adult heart. Early apelin receptor knockdown decreased cardiomyocyte differentiation efficiency and prolonged voltage sensing, associated with asynchronous contraction. Late apelin receptor knockdown had detrimental consequences on 3D engineered heart tissue contractile properties, decreasing contractility and increasing stiffness.

Conclusions: We have successfully knocked down the apelin receptor, using an inducible system, to demonstrate a key role in hESC-CM differentiation. Knockdown in 3D engineered heart tissues recapitulated the phenotype of apelin receptor down-regulation in a failing heart, providing a potential platform for modelling heart failure and testing novel therapeutic strategies.

Keywords: Apelin receptor; Cardiomyocyte; Cardiovascular disease; Stem cell.

Graphical Abstract

Effect of early or late apelin receptor inducible knock-down in cardiomyocytes derived from human embryonic stem cells

Figure 1

Human embryonic stem cells (hESCs) and hESC-derived cardiomyocytes (hESC-CMs) express the apelin receptor and its endogenous ligands. (A) Schematic representation of hESC to cardiomyocyte differentiation protocol. (B) Comparison of relative expression of (i) APLNR, (ii) APLN, and (iii) APELA in hESCs and hESC-derived cardiomyocytes determined by qRT-PCR. hESCs n = 4, hESC-derived cardiomyocytes n = 8. Expression displayed relative to mean hESC expression, means compared by unpaired, two-tailed Student’s t-test. For (iii) P = 0.005. (C) Comparison of reads per million (RPM) for APLNR mRNA in adult human left ventricle (LV, n = 6) and hESC-CMs (n = 3) by RNA-sequencing. Expression levels compared by unpaired, two-tailed Student’s t-test. (D) Representative fluorescent confocal images of endogenous expression of apelin receptor in hESC-CMs using an apelin receptor antibody (i), or hESC-CMs treated with an IgG isotype control (ii). Scale bars as indicated in figure. (E) Saturation radioligand binding in (i) hESCs and (ii) hESC-CMs using [¹²⁵I]apelin-13, with measures of density and affinity compared with that of adult cardiomyocytes (inset table). (F) Comparison of concentration of (i) ELA and (ii) apelin peptides in conditioned supernatant from hESCs (n = 3 for ELA, n = 5 for apelin) and hESC-CMs (n = 8 for ELA, n = 6 for apelin). Means compared by unpaired, two-tailed Student’s t-test. (G) Cardiac troponin T (TnT) positive percentage from flow cytometric analysis of control hESC-CMs and hESC-CMs cultured in the presence of 10 nM [Pyr¹]apelin-13 throughout differentiation (n = 4 for both). Means compared by unpaired, two-tailed Student’s t-test, P = 0.03. Data represent mean ± SEM.

Figure 2

Generation and validation of a novel apelin receptor single step inducible knockdown system (shAPLNR). (A) Transgene generated containing APLNR shRNA targeted to the AAVS1 locus. 5′-HAR/3′-HAR, upstream/downstream homology arm; SA, splice acceptor; T2A, self-cleaving T2A peptide; Puro, puromycin resistance; pA, polyadenylation signal; H1, H1 promoter; CAG, CAG promoter; TO, Tet operon; tetR, tetracycline-controlled repressor. (B) Expression of apelin receptor in hESCs cultured with or without tetracycline for 4 days. Comparison of relative expression (wrt shB2M + Tet) of apelin receptor gene (APLNR) for hESCs expressing shRNA directed against the (i) apelin receptor (n = 5, P < 0.001) or (ii) control line expressing shRNA directed against the B2M gene (n = 3). Expression levels compared by unpaired, two-tailed Student’s t-test. (iii) Saturation specific [¹²⁵I]apelin-13 binding in hESCs expressing shAPLNR. (iv) Specific binding of fixed concentration of [¹²⁵I]apelin-13 in hESCs expressing shB2M transgene. Specific binding levels compared by unpaired, two-tailed Student’s t-test. n = 3 for all. (C) Expression of apelin receptor in early knockdown hESC-CMs. Comparison of relative expression (wrt shB2M + Tet) of APLNR for hESC-CMs expressing shRNA directed against the (i) apelin receptor (control n = 4, +Tet n = 5, P = 0.04) or (ii) control line expressing shRNA directed against the B2M gene (n = 4). Expression levels compared by unpaired, two-tailed Student’s t-test. (iii) Saturation specific [¹²⁵I]apelin-13 binding in hESC-CMs carrying the shAPLNR transgene. (iv) Specific binding of fixed concentration of [¹²⁵I]apelin-13 in hESC-CMs expressing shB2M transgene. Specific binding levels compared by unpaired, two-tailed Student’s t-test. n = 3 for all. (D) Expression of apelin receptor in late knockdown hESC-CMs. Comparison of relative expression (wrt shB2M + Tet) of APLNR for hESC-CMs expressing shRNA directed against the (i) apelin receptor (n = 3, P = 0.01) or (ii) control line expressing shRNA directed against the B2M gene (n = 4). Expression levels compared by unpaired, two-tailed Student’s t-test. (iii) Saturation specific [¹²⁵I]apelin-13 binding in hESC-CMs carrying the shAPLNR transgene. (iv) Specific binding of fixed concentration of [¹²⁵I]apelin-13 in hESC-CMs expressing shB2M transgene. Specific binding levels compared by unpaired, two-tailed Student’s t-test. n = 3 for all. Inset table displays maximum binding in the three knockdown conditions for control and tetracycline treated cells. Data represent mean ± SEM.

Figure 3

Early APLNR knockdown reduces differentiation efficiency of hESC-CMs and increases number of cells with fibroblast like identity. (A) Representative brightfield images of hESC-CMs carrying shAPLNR transgene cultured (i) without or (ii) with tetracycline (+Tet) throughout differentiation. (iii) and (iv) represent control cells carrying shB2M transgene cultured without or with tetracycline throughout differentiation. Scale bar = 200 µm. (B) Representative flow cytometry plot of control and APLNR knockdown hESC-CMs co-stained for cardiac troponin T (TnT, APC) and Thy1 (PE). (C) Quantification of TnT and Thy1 relative expression from flow cytomteric co-stain for (i) APLNR knockdown hESC-CMs relative to control (P = 0.002 for TnT, P = 0.005 for Thy1) and (ii) shB2M with tetracycline relative to control. n = 3 for all, compared by unpaired, two-tailed Student’s t-test. (D) TnT positive percentage from flow cytometric analysis of control, APLNR knockdown hESC-CMs and APLNR knockdown hESC-CMs cultured in the presence of 10 nM [Pyr¹]apelin-13 throughout differentiation (+Apelin), n = 3. Means compared by one-way ANOVA followed by Tukey’s post hoc test. Data represent mean ± SEM.

Figure 4

APLNR knockdown throughput differentiation (early knockdown) has functional consequences on hESC-CMs. (A) Workflow for bulk RNA-sequencing of APLNR knockdown hESC-CMs. (B) Plot representing number of up- and down-regulated differentially expressed genes (DEGs) in APLNR knockdown hESC-CMs compared with control. (C) Change in expression of SCN5A gene encoding the cardiac sodium channel Na_v1.5 with APLNR knockdown. Control n = 3, APLNR knockdown n = 4, means compared by two-tailed Student’s t-test, P = 0.028. (D) (i) Time to peak (TTP) (P < 0.001) and (ii) time to 90% decay (T90) (P = 0.006) of voltage-sensitive dye in paced control and APLNR knockdown hESC-CMs. n = 5, means compared by unpaired, two-tailed Student’s t-test. (E) (i) TTP and (ii) T90 of calcium-sensitive dye in paced control and APLNR knockdown hESC-CMs. n = 5, means compared by unpaired, two-tailed Student’s t-test. Data represent mean ± SEM.

Figure 5

APLNR knockdown has detrimental effects on contractility in 3D engineered heart tissues (EHT). (A) Schematic of workflow used to produce APLNR knockdown EHTs. (B) (i) Time to peak (TTP) and (ii) time to 90% decay (T90) of voltage-sensitive dye in paced control and APLNR knockdown EHTs. n = 4, means compared by unpaired, two-tailed Student’s t-test. (C) (i) TTP and (ii) T90 of calcium-sensitive dye in paced control and APLNR knockdown hESC-CMs. n = 3, means compared by unpaired, two-tailed Student’s t-test. (D) (i) Active and passive force produced by control and APLNR knockdown EHTs as measured by force transducer in response to increasing strain. (ii) Linear regression of force produced to generate Frank-Starling curve of active and passive force. (iii) Slope of generated Frank-Starling curve of active and passive force. n = 4, means compared by unpaired, two-tailed Student’s t-test. For active force P < 0.001, for passive force P = 0.02. (E) Representative images of (i) control and (ii) APLNR knockdown EHTs imaged using second-harmonic imaging microscopy to visualize collagen. (iii) Quantification of mean pixel intensity of control and APLNR knockdown EHT collagen signal. n = 3, means compared by unpaired, two-tailed Student’s t-test, P = 0.02. Data represent mean ± SEM.