Distinct roles of microRNA-1 and -499 in ventricular specification and functional maturation of human embryonic stem cell-derived cardiomyocytes

Abstract

Background: MicroRNAs (miRs) negatively regulate transcription and are important determinants of normal heart development and heart failure pathogenesis. Despite the significant knowledge gained in mouse studies, their functional roles in human (h) heart remain elusive.

Methods and results: We hypothesized that miRs that figure prominently in cardiac differentiation are differentially expressed in differentiating, developing, and terminally mature human cardiomyocytes (CMs). As a first step, we mapped the miR profiles of human (h) embryonic stem cells (ESCs), hESC-derived (hE), fetal (hF) and adult (hA) ventricular (V) CMs. 63 miRs were differentially expressed between hESCs and hE-VCMs. Of these, 29, including the miR-302 and -371/372/373 clusters, were associated with pluripotency and uniquely expressed in hESCs. Of the remaining miRs differentially expressed in hE-VCMs, 23 continued to express highly in hF- and hA-VCMs, with miR-1, -133, and -499 displaying the largest fold differences; others such as miR-let-7a, -let-7b, -26b, -125a and -143 were non-cardiac specific. Functionally, LV-miR-499 transduction of hESC-derived cardiovascular progenitors significantly increased the yield of hE-VCMs (to 72% from 48% of control; p<0.05) and contractile protein expression without affecting their electrophysiological properties (p>0.05). By contrast, LV-miR-1 transduction did not bias the yield (p>0.05) but decreased APD and hyperpolarized RMP/MDP in hE-VCMs due to increased I(to), I(Ks) and I(Kr), and decreased I(f) (p<0.05) as signs of functional maturation. Also, LV-miR-1 but not -499 augmented the immature Ca(2+) transient amplitude and kinetics. Molecular pathway analyses were performed for further insights.

Conclusion: We conclude that miR-1 and -499 play differential roles in cardiac differentiation of hESCs in a context-dependent fashion. While miR-499 promotes ventricular specification of hESCs, miR-1 serves to facilitate electrophysiological maturation.

Figure 1. Characterization of miRNAs Present in hESC and Ventricular CMs.

(A) A heatmap showing the results of ANOVA differential expression analysis; 63 miRs were differentially expressed between undifferentiated human ESC and hESC-derived VCMs (p<0.05). Red to green indicates high to low expression. Right panels: Average signal intensities of the 29 miRs differentially expressed in hESC (top), and the 23 miRs that displayed a “plateau” pattern across E-, F- and A-VCMs (bottom). (B) “Plateau” miRs validated with quantitative RT-PCR. (C) After normalized with the values of hESCs, miR-1, -133, and -499 displayed the highest differential expression between VCMs and hESCs. (D) MiR-188 and -296 were identified as stably expressed across developmental stages, at levels comparable to standard endogenous controls RNU38B and RNU48. Conversely, the expression of miR-1, -499, and -133 varies greatly across developmental stages. The box plot for each sample summarizes the median, first and third quartiles and range of raw CT values for each miRNA or endogenous control in all cell types assayed. (E) Between-group analysis (BGA) was carried out using principle component analysis (PCA) to provide a visual means to observe the variance of data in specified groups (G) along G-1 number of axes. Data points corresponding to individual miRs are distributed along the axes according to discriminating eigenvectors. Data points and groups that are strongly associated will be clustered within the graph due to their similar projection (both direction and distance) from the origin. Details on the calculations used in BGA can be found in Culhane et al. (2002) .

Figure 2. Electrophysiological and Molecular Properties of Control and miR-1, -133, and -499 Transduced hE-VCMs.

A) Representative tracings of ventricular, atrial and pacemaker action potentials (APs) of control hESC-CMs. B) The percentage distribution of ventricular, atrial and pacemaker phenotypes before and after LV-miR-1 or -miR-499 transduction. C) Transcriptional expression of cardiac sarcomeric genes in LV-miR-1-, 133- and 499-transduced hESC-CMs. *, p<0.05.

Figure 3. AP Parameters of Control and miR-1 and -499 Transduced hE-VCMs.

A) Representative AP tracings of Control, LV-miR-1- and -miR-499-transduced hESC-derived ventricular derivatives as labeled. B-G) Bar graphs summarizing the AP parameters APD₅₀, APD₉₀, repolarization velocity, upstroke velocity, resting membrane potential (RMP), and maximum diastolic potential (MDP) of the groups in A).

Figure 4. Effects of miR-1 transduction on electrophysiological and molecular properties of hESC-derived CMs.

A) Transcriptional expression of sarcolemmal ion channels (Kir2.1, HCN4, Kv1.4, Kv4.3, HERG, SCN5A, DHPR) in hESC-CMs after LV-miR-1 transduction. B) Representative tracings of I_Ca,L of control and LV-miR-1-transduced hE-VCMs as labeled. Currents were elicited by the electrophysiological protocol given in the inset. Currents shown were nifedipine-sensitive. C–D) The corresponding current-voltage (I–V) relationship, steady-state activation (circle) and inactivation curves (square) of I_Ca,L. E) Representative tracings of I_f recorded from control and LV-miR-1-transduced hE-VCMs. Currents were elicited by the electrophysiological protocol given in the inset. Currents shown were ZD7288-sensitive. F–G) The corresponding current-voltage (I–V) relationship and steady-state activation curve of I_f. * p<0.05.

Figure 5. Currents and Current-Voltage Relationships in Control and miR-1 Transduced hE-VCMs.

A, C and E) Representative tracings of I_to, I_Kr and I_Ks recorded from WT, LV-miR-1-transduced E-VCMs as labeled. Currents were elicited by the electrophysiological protocol given in the inset. Currents shown were 4-Aminopyridine-, Chromanol 293B- and E4031–sensitive, respectively. B, D and F) The corresponding current-voltage (I–V) relationships. *, p<0.05.

Figure 6. Calcium Handling in Control and miR-1, -133, and -499 Transduced hE-VCMs.

A) Representative tracings of Ca²⁺ transients recorded from control, LV-miR-1-, 133- and 499-transduced hESC-CMs. B–D) Comparison of amplitude (B), maximum upstroke velocity (V_max-upstroke, C), and maximum decay velocity (V_max-decay, D) of Ca²⁺ transients in control and LV-miR-transduced hESC-CMs as indicated. E) Transcriptional expression of Ca²⁺-handling proteins. *, p<0.05.

Figure 7. Analysis of Putative Transcriptomic Effects of miR-1 and -499 Expression in CM Development and Maturation.

A) Transcriptomic analysis of predicted targets for miR-1 and -499 involved in select pathways as indicated. All genes were log₂ transformed and normalized. Cluster analysis was performed with Cluster 3.0 using centroid linkage. Pie charts for B) miR-1 and C) -499 summarizing the number of genes (in parentheses) in each selected pathway or Gene Ontology (GO) classification that were found to agree with our transcriptomic data. See text for further details. The color gradient corresponds to a percentage (% = number of genes in each group meeting criteria/total number of genes in pathway or GO classification * 100). D) Summary of the proposed sequence of biological processes that occur during human cardiogenesis and the actions of miR-1 and -499.