Why Don't They Beat?

• embryonic stem cells
• cardiomyogenesis
• apelin/APJ
• cripto
• MAPK

The convoluted maze of transcription and signaling factors that underpin cardiomyocyte differentiation has taken 2 major paths of inquiry that inform and sometimes puzzle each other: embryonic in situ development of cardiomyocytes and differentiation of stem cells to cardiomyocytes. Cardiomyocyte differentiation in embryonic development has fascinated scientists for millennia. In the middle of the 20th century, embryonic cardiomyocytes were tracked down in the embryo using extirpations and in vitro cultures of either cardiogenic or noncardiogenic mesoderm. Cultures that beat were relegated to the cardiogenic region of the anterior lateral plate mesoderm.¹ Maps created in this way have been verified and refined in vivo. In the case of noncardiogenic mesoderm explants, treatment with “cardiogenic” factors resulted in beating cultures.^2,3 These studies allowed us to begin to understand where the specified cardiogenic cells resided in the mesoderm and, later, to determine what signaling factors promoted myocardial specification in noncardiogenic mesoderm and subsequent differentiation. In an era when stem cell transplantation has become a reasonable goal of adult medical regenerative treatment, the ability to specify and differentiate large numbers of cardiomyocytes (or other pure populations of differentiated cells) from stem cells has become something of a holy grail.

In this regard, a novel signaling pathway promoting cardiomyocyte differentiation in embryonic stem cells is introduced in this issue of Circulation Research. D'Aniello et al, using cripto-null embryonic stem cells, which are unable to differentiate as cardiomyocytes, show that the apelin ligand and its receptor APJ (explained below) are able to partially rescue cardiomyocyte differentiation.⁴ I say “partially” because the cripto-null, apelin/APJ-treated stem cells begin to express myocardial contractile proteins which are not expressed by untreated cripto-null stem cells.⁵ Although the cells express contractile proteins, they do not beat.

The apelin/APJ signaling pathway regulates adult blood pressure and positive inotropy in the heart.^6–9 The apelin ligand is translated as a 77-aa prepropeptide, expressed primarily in endothelial cells, that is cleaved to shorter activated C-terminal peptide fragments.^10–12 Apelin peptides activate a 7-transmembrane G protein-coupled receptor called APJ (angiotensin II receptor-like 1, Agtrl1, Xmsr in Xenopus).¹³ The APJ receptor is expressed in endothelial cells, vascular smooth muscle cells, and cardiomyocytes.^14,15 Recent evidence suggests that the APJ receptor may function, in some contexts, independent of the apelin ligand.¹⁶

The role of apelin during embryonic development is not well studied, and this new report showing that apelin/APJ signaling is involved in the cascade of signaling factors that promote myocardial differentiation makes it imperative to reevaluate the apelin/APJ studies that have been done in embryonic development. In zebrafish, apelin expression is confined to unidentified cells in the embryonic midline, whereas its receptor (agtrl1b) is expressed in lateral plate mesoderm.¹⁷ Modulation of either the ligand or receptor (grinch mutant) causes a reduced number of cardiac progenitors to move inefficiently to the lateral plate mesoderm and subsequently to make a disorganized heart-like structure.¹⁷ Interestingly, the apelin receptor in zebrafish does not colocalize with lateral plate mesoderm cells expressing cmlc2,^17,18 suggesting that expression may be extinguished as the cells differentiate. Knockdown experiments in frog have shown that apelin is required for normal vascular and cardiac development.¹⁹ In frog, by the time the heart forms, expression of apelin is confined to the endothelium and endocardium.¹³ In mouse embryos, APJ is expressed from embryonic day 8 in lateral plate mesoderm.²⁰ Apelin-deficient mice are viable and fertile,²¹ suggesting that other factors can rescue in vivo cardiac development in the absence of apelin.

In the case of the embryonic stem cell cultures reported by D'Aniello et al, apelin/APJ rescues expression of contractile proteins in cripto-null cells.⁴ Cripto is a member of the EGF-CFC family of signaling factors named for Cripto, Frl1 (Xenopus), and Cryptic.²² EGF-CFC genes encode extracellular proteins that share several domains including an N-terminal signal sequence, a variant epidermal growth factor (EGF)-like motif, a novel conserved cysteine-rich domain named the CFC (Cripto, FRL-1, and Cryptic) motif, and a C-terminal hydrophobic region.²² Crypto is transcribed from a gene called teratocarcinoma-derived growth factor-1,²³ which was identified and cloned from an embryonal carcinoma cell line. Cripto is an extracellular protein that is tethered to the surface of cells by a glycosyl-phosphatidylinositol linkage to the cell membrane which appears to be important for its activity.²²

In embryonic day 8.5 embryos, cripto is expressed in the myocardium of the developing heart tubes and in the outflow tract of the heart at embryonic day 9.5 to 10. The expression pattern suggests a role in cardiac morphogenesis and sure enough, cripto-null mice never show any signs of cardiomyocyte differentiation.²⁴

The article by D'Aniello et al⁴ builds on the fact that cripto-null cells express markers that indicate successful cardiomyocyte specification, such as Nkx2.5, Mef2C, Gata4, ehand, dhand,^4,5 but the cells fail to differentiate and so don't express any of the cardiomyocyte contractile proteins and never beat even when cultured for an extended period of time. Wild-type clones give rise to a higher percentage of beating clones than the heterozygous cells, suggesting some dose dependence.⁵ Because other mesodermal cell types differentiate normally, this suggests a specific defect in cardiac differentiation.⁵ Addition of crypto protein to stem cell cultures at 0 to 2 days restores the cardiomyocyte differentiation potential of crypto-null stem cells.²⁵ This is somewhat odd because it indicates that cripto signaling is required very early, perhaps even before specification of cardiomyocytes.

After establishing that apelin/APJ signaling is downstream of cripto, D'Aniello et al⁴ go on to show that apelin/APJ induces activation of extracellular signal-regulated kinases and AKT, leading to activation of p70S6 kinase. Blocking activation of MAPK prevents apelin rescue of cardiomyocyte differentiation.

So does any of this explain why they don't beat? The simplest explanation is that cripto promotes cardiomyocyte differentiation via at least 2 different pathways: one through apelin/APJ to promote contractile protein expression and a second (or more) to promote expression of the membrane components needed for automaticity. Even though the study by D'Aniello et al⁴ shows that a few of the major contractile proteins are transcribed, it is unclear whether they are all translated and if the whole battery needed for construction of normal sarcomeres is made. I think we can all agree that beating has been the traditional measure of cardiomyocyte differentiation, and even though we are now able to identify molecular cardiomyocyte markers, beating must be included in considering differentiation of myocardial cells. That is why I have referred to the rescue reported by D'Aniello et al⁴ as partial. In any case, this study brings us to some interesting new thoughts about cardiomyocyte differentiation.

How many different signaling pathways converge to promote all the elements that lead to beating? Much has to happen, including contractile protein expression, sarcomeregenesis, synthesis of components that make an excitable membrane and excitation-contraction coupling. The fact that apelin/APJ represents one of these pathways is something of a surprise, and there are likely more surprises to come.