Mitochondrial oxidative metabolism is required for the cardiac differentiation of stem cells

Abstract

Cardiogenesis within embryos or associated with heart repair requires stem cell differentiation into energetically competent, contracting cardiomyocytes. While it is widely accepted that the coordination of genetic circuits with developmental bioenergetics is critical to phenotype specification, the metabolic mechanisms that drive cardiac transformation are largely unknown. Here, we aim to define the energetic requirements for and the metabolic microenvironment needed to support the cardiac differentiation of embryonic stem cells. We demonstrate that anaerobic glycolytic metabolism, while sufficient for embryonic stem cell homeostasis, must be transformed into the more efficient mitochondrial oxidative metabolism to secure cardiac specification and excitation-contraction coupling. This energetic switch was programmed by rearrangement of the metabolic transcriptome that encodes components of glycolysis, fatty acid oxidation, the Krebs cycle, and the electron transport chain. Modifying the copy number of regulators of mitochondrial fusion and fission resulted in mitochondrial maturation and network expansion, which in turn provided an energetic continuum to supply nascent sarcomeres. Disrupting respiratory chain function prevented mitochondrial organization and compromised the energetic infrastructure, causing deficient sarcomerogenesis and contractile malfunction. Thus, establishment of the mitochondrial system and engagement of oxidative metabolism are prerequisites for the differentiation of stem cells into a functional cardiac phenotype. Mitochondria-dependent energetic circuits are thus critical regulators of de novo cardiogenesis and targets for heart regeneration.

Figure 1

Embryonic stem cell cardiac differentiation is coordinated with metabolic transcriptome reprogramming and mitochondrial oxidative metabolism. (A) ES have no contractile activity on linescans and MEF2C, a cardiac transcription factor, and cardiac α-actinin, a tissue-specific protein of the contractile apparatus, are absent. (B) Cardiomyocytes derived from ES have distinct structures on light microscopy (upper panel) and electron microscopy (inset), contractile activity on linescans, and MEF2C and cardiac α-actinin are abundant in the nucleus and sarcomeres, respectively, on immunofluorescent microscopy. (Bars 10 μm [A,B left], 2 μm [A,B insets], 2 s [B middle], and 5 μm [A,B right]). (C) The basal respiratory rate and maximum respiratory capacity of cardiomyocytes were markedly higher, while the lactate production from anaerobic glycolysis was lower than in ES, underscoring distinct metabolic identities of the progeny compared with the embryonic source. A higher ADP:ATP ratio reflected an increased rate of energy use in cardiomyocytes than ES. (D) High-membrane-potential mitochondria (red) in ES on confocal microscopy. (E) Mitochondria have lower membrane potential (green) in cardiomyocytes, associated with increased energy use. (F) Intercellular cardiomyocyte connections with mitochondrial traffic (mitotrail) indicate cell–cell metabolic cross-talk in cardiogenesis. (D–F, bars 20 μm, sample >3.) (G) Microarray analysis of total mRNA in ES and cardiomyocytes revealed specific changes in genetic programming of the cellular energetic system. Genes were hierarchically clustered as mRNA copy numbers of cardiomyocytes versus ES transcripts (n = 3 in each group). Redundant probe sets were included to illustrate the individual dynamics of expression profiles. ^aP<0.05; n=3–12. Abbreviations: CM, cardiomyocytes; ES, embryonic stem cells; mRNA, messenger RNA.

Figure 2

Development and maturation of mitochondrial network in stem cell cardiac differentiation. (A) Gene array analysis of selected genes related to mitochondrial fission, fusion, and/or cristae maturation. Compared with the embryonic stem cell source, genes involved in mitochondrial fission and membrane structure remodeling in cardiomyocytes (Dnm1l, Mtp18, Opa1, and DAP3) were downregulated, whereas those involved in mitochondrial fusion and cristae maturation were typically upregulated (Mfn2 and IMMT). (B) Transmission electron microscopy revealed spherical mitochondria with underdeveloped cristae in embryonic stem cells versus elongated, cristae-rich mitochondria in cardiomyocytes. (C) Live-cell imaging showed discrete organelles with no apparent pattern of mitochondrial arrangement in ES versus an expanded network of aligned mitochondria in cardiomyoctyes. (D) Tracking cardiomyocyte development revealed an organization of the mitochondrial network (upper panels) ranging from random (left), to perinuclear (center), to transcellular (right), along with concomitant maturation of myofibrillar structure (lower panels). Mitochondria were visualized with MitoTracker Red, myofibrils with α-actinin staining (green), and nuclei with DAPI (blue). (E) Development of mechanoenergetic coupling through intercalation of mitochondria with myofibrils in cardiomyocytes observed by confocal microscopy (upper panel). Profile of fluorescence intensity (lower panel corresponds to arrow in upper panel) indicates an alternating distribution of mitochondria (red) and myofibrils (green). (F) Integration of mitochondrial (green) and electrical (red) activities in a beating area of an embryoid body (EB). Cardiac beating area delineated by the tight correlation of electrical activity staining with RH237, a probe for plasma membrane potential, and mitochondrial imaging with JC-1 (upper panel). Profiles of overlapping fluorescence intensity for both signals within the cardiac beating area are depicted in the lower panel (lower panel corresponds to line in upper panel). ^aP <0.05. Abbreviations: CM, cardiomyocytes; ES, embryonic stem cells.

Figure 3

Disturbed execution of the cardiac differentiation program with inhibition of the mitochondrial respiratory chain. (A) After antimycin (50 nmol/l) or rotenone (250 nmol/l) treatment the percentage of beating EBs (n>80), α-actinin expression (n=6–10) and sarcomeric content (n = 7–10) were reduced compared with those in untreated controls. (B) Imaging of embryoid bodies with mitochondrial and plasma-membrane-potential probes (JC-1 and RH237, respectively) revealed that respiratory chain inhibition caused mitochondrial fragmentation and defective assembly of cardiac beating areas. (C) Cardiomyocytes isolated from EBs treated with antimycin or rotenone and cultured in the presence of the corresponding inhibitors had aberrant mitochondrial content, localization, and network architecture as detected by JC-1. (D) Sarcomere formation was compromised with mitochondrial inhibition. The intensity of α-actinin staining was decreased and fewer cardiomyocytes had detectable sarcomeric structures. (Bars 20 μm.) (E) Effects of respiratory chain inhibition on contractions in cardiac beating areas. Contractions were measured as changes in cell edge position using confocal microscopy, and analyzed with the region of interest function in the software. Abbreviations: CM, cardiomyocytes; EBs, embryoid bodies.