*3.3. Embryonic Stem Cells Versus Induced Pluripotent Stem Cells*

The primary intended purpose of reprogramming of somatic cells and generation of iPSCs was to wipe the initial cellular identity and drive them back to the embryonic inner mass state, and hence serve as a surrogate for embryonically derived cells, i.e., ESCs. Indeed, iPSCs greatly resemble conventional ESCs in terms of growth characteristics, gene-expression profiles, epigenetic status, and developmental potential, which were shown in earlier studies by Yamanaka and colleagues, as well as others [79,84–86]. However, upon comparison of various undifferentiated cell lines, reports argued that iPSCs may not be perfectly identical to conventional ESCs. This is largely attributed to the unique epigenetic signatures of their parent somatic cells. Despite previous studies showing that somatic cells undergo epigenetic remodeling upon reprogramming, studies have shown that iPSCs indeed retain epigenetic

patterns of their donor cells, e.g., CpG island methylation [86–90]. Furthermore, gene and miRNA expression signature were also shown to trail along with iPSCs (reviewed in [91]). Upon differentiation to CMs, further comparison of mature CMs differentiated from ESCs and iPSCs can be insightful. In this regard, CMs of either origin were reported to display similar ultrastructural phenotypes, upon electron microscopic examination [92]. In line with these findings, a study by Gupta et al. revealed that global transcriptional profiles of mature CMs derived from either human iPSCs or ESCs are highly similar [93]. However, iPSC-CMs were more likely to share some somatic cell signature with their undifferentiated iPSC-parents. Thus, identification of these variations between iPSC- and ESC-CMs, as well as the interline variability of either type of PSCs, is essential before they are utilized for disease modeling or clinical application.

Unlike ESCs, iPSCs derivation does not involve destruction of embryos, and hence does not fall into the same ethical pitfalls. However, other ethical considerations arose with hiPSCs, especially with regard to the possibility of reproductive cloning, the risk of generating genetically engineered human embryos, and, more extremely, human–animal chimeras [94]. Furthermore, and like ESCs, iPSCs are subject to safety concerns due to their ability to form tumors, even with rigorous protocols of differentiation and selection [95].

In recent years, substantial developments in stem cell technology in terms of reprogramming efficiency and enhancing their clinical applicability have prompted scientist to utilize pluripotent stem cells (PSCs), not only to regenerate, but also to model the human heart for basic research purposes. Furthermore, some countries have tentatively started to loosen their tight regulations, especially on hESCs; a step that coincided with the establishment of stem-cell registries in the US and Europe [96–98]. This has led to several initiatives on stem cell therapy for many disease conditions, including CVDs [99]. As promising as this may sound, several challenges, however, preclude the full realization of PSC-based therapy. In the following, we shall focus on PSCs by addressing efforts made over the past decades to optimize their generation, differentiation, and maturation for CVD research, as well as efficient delivery methods for late clinical and/or translational purposes.

#### **4. Cardiac Stem-Ness**

Embryology is the fundament for generation of cardiac cells from PSCs in the laboratory. The heart is the first organ to develop and function during embryogenesis [100]. In the lateral mesoderm, cardiac specification takes place, a process initiated by two T-box transcription factors, Eomesodermin and Brachyury(T), which have been shown to induce the expression of yet another critical factor, namely mesoderm posterior 1(MesP1) [101,102]. MesP1 is a basic helix-loop-helix (bHLH) transcription factor considered to be the master regulator orchestrating the differentiation and commitment of cardiac precursors [101,102]. Cardiac precursors then assume a crescent-shaped structure known as the cardiac crescent, at which cells are irreversibly committed to the cardiac lineage. This is marked by the expression of key transcription factors, namely Nkx2.5, GATA4 and Tbx5 [103]. Two waves of Nkx2.5 expression ensue, depicting the formation of two regions known as the first and second heart fields, which subsequently give rise to different heart chambers, as well as the cardiac outflow tract [76,103].

After all, heart development is a dynamic three-dimensional process governed by an intricate network of signals and gene transcription [104,105]. Howbeit, three major signaling pathways converge to drive the process, from early cardiac tissue specification of mesoderm progenitors to subsequent differentiation into cardiac progenitors, namely BMP (bone morphogenic protein) and Nodal/Activin, both being members of the TGF-β (transforming growth factor beta) cytokine family, and the Wnt/β-catenin [106,107]. Paracrine signals responsible for the fine-tuning of those pathways is crucial for heart development. For example, signals activating the Wnt/β-catenin pathways are essential for early mesoderm induction, whereas inhibitors of the same pathway are subsequently required for precardiac specification [76].

#### *4.1. Generation*/*Di*ff*erentiation of Pluripotent Stem Cells*

Protocols for in vitro generation of cardiac stem cells (CSCs) from PSCs, either from ESCs or iPSCs, rely primarily on simulating the signaling microenvironment, which induces the aforementioned rudimental pathways, starting by initial epithelial to mesenchymal transition, mesodermal specification, and subsequent cardiogenic differentiation, followed by selection for cardiac markers [108–110]. The initially reported protocols relied simply on serum in culture medium as a source of inducing factors, observing spontaneous formation of aggregates called embryoid bodies (EBs) when cells are plated in suspension [111]. These EBs would later show contractions and positive staining for cardiac markers. This method was first reported in ESCs, however, with very low efficiency [111]. Nevertheless, the EB-based differentiation remained a standard protocol and was also the first differentiation method applied to generate CMs from mouse iPSCs only a couple of years after their first introduction in 2006 [112,113]. The first CMs generated from iPSCs were reported by a team of researchers from Leibniz institute in Germany, with few refinements introduced to the protocol, which led to the differentiation of typical CMs comparable to those generated from ESCs [112]. Interestingly, precisely at the same time and in the same journal issue, the iPSCs-founding team from Kyoto also published a systematic differentiation protocol of mouse iPSCs into cardiac lineages [113]. Nevertheless, and as mentioned before, the efficiency of the EB-based protocols was low, mainly due to the uncontrolled differentiation cues in the supporting media. One of the earliest and most cited protocols to differentiate ESCs to beating CMs was reported by Mummery et al. in 2003, where they delegated the differentiation cues to paracrine signaling of murine visceral endoderm-like cells (END-2) [114]. They compared their generated CMs to primary human fetal CMs, as well as primary human adult CMs, and reported comparable structural and functional properties. Improvements to differentiation protocols by temporal application of cytokines, as well as small molecule inhibitors (e.g., inhibitors of the Wnt pathway) to simulate the developmental processes have also been successful introduced to generate CMs from PSCs [108–110,115–117]. Furthermore, several groups have sought to simplify the differentiation protocols by using chemically defined culture media consisting of only a few components [115–117].

Nevertheless, differentiation of PSCs by using standard protocols usually yields a mixed population [118]. Thus, identification of selection markers is crucial for the purification of cardiomyocyte progenitors. Pioneering studies by Moretti and colleagues have greatly contributed to the refinement of selection protocols for cardiac organogenesis from PSCs [119]. From an embryological standpoint, myocyte progenitors are distinguished from nonmyocytes (vascular progenitors) by consistent expression of Isl1-1 transcription factor, along with Nkx2.5, whereas co-expression of Isl-1 and CD31 is a marker for endothelial progenitors [119]. Among myocytes, cardiomyocytes can be further distinguished from smooth muscle cells (SMCs) by expression of vascular cell adhesion molecule 1 (VCAM 1) and signal regulator protein alpha (SIRPα), both of which were reported to be reliable selection markers in culture conditions, yielding as much as 98% pure-cardiomyocyte populations by antibody-based sorting from PSCs [120,121]. Successful differentiation can be further confirmed by expression of other cardiomyocyte markers, such as cardiac troponins, e.g., TNNI1 [121,122]. Using lentiviral vectors, expression of selection markers, e.g., antibiotic-resistant genes or fluorescent proteins, under control of cardiomyocyte-specific promoter, has also been reported to purify cardiomyocytes [123,124]. Importantly, documented biochemical disparity between CMs and non-CMs in energy metabolism was also exploited for the so-called "metabolic purification" of CMs. In this regard, manipulation of culture conditions by altering the composition of the culture medium (e.g., glucose depletion, lactate, and glutamine supplementation) was found to be crucial for such nongenetic purification of CMs [125].

Finally, studies have pointed out the important role of MicroRNAs (small, noncoding RNAs that regulate gene expression by degradation of messenger RNAs) in CM phenotype differentiation [126–128].

#### *4.2. Maturation of Pluripotent Stem Cells*

To be utilized for disease modeling or regenerative medicine, one might expect PSC-CMs to recapitulate the structural and functional characteristics of adult CMs (Figure 1). Nevertheless, CM differentiation of PSCs usually yields immature cells, resembling the embryonic or fetal state [129]. This manifests in their morphology, gene expression, and electrophysiology. More recently, single-cell-transcriptomic analyses have proven to be a powerful tool to understand the transcriptional roadmap of in vitro CM differentiation, and therefore enable a better design of differentiation and maturation protocols [130,131]. The following highlights the major differences between immature PSC-derived CMs and mature and/or adult ones.

Morphologically, PSC-CMs are significantly smaller in size, compared to their adult or matured counterparts. Upon maturation, cells assume an elongated shape, reminiscent of adult CMs [132]. Sarcomeres are much less organized in immature PSC-CMs and become much more organized upon maturation, which usually correlates with isoform switch of sarcomeric proteins. A good example is troponin I, wherein different isoforms distinguish embryonic CMs from adult ones [133,134]. Stoichiometric replacement of the fetal troponin *TNNI1*, encoding slow skeletal troponin I (TnIs), gene with the adult *TNNI3*, encoding adult cardiac troponin I (TnIc), gene was reported in a study by Bedada et al. as a quantifiable marker for maturation in PSC-CMs [135]. Another well-characterized hallmark of mature CMs is the isoform switch of myosin heavy chain (MHC). Two isoforms exist, the alpha isoform (encoded by *MYH6*), also known as the faster isoform, and the beta isoform (encoded by *MYH7*), also known as the slower isoform [136]. Importantly, differences exist between rodents and humans in this regard. In small rodents (mice and rats) with faster heart rates, alpha-MHC isoform predominates and increases upon maturation, whereas, in bovine and human hearts, despite the presence of the alpha-MHC isoform, the beta MHC isoform usually predominates, regardless of the state of development, and increases with age [136,137]. However, most differentiation protocols of human PSC yield CM with both isoforms, but studies have shown that long-term cultures, especially on stiff substrates, lead to a greater shift toward the beta-isoform, reflecting maturation [138]. Titin is another key component of the sarcomere that undergoes isoform switch during maturation. Fetal titin isoforms N2BA 1 and 2 are more compliant, but they switch to the N2B isoform in postnatal and adult cardiomyocytes [139]. Genes encoding structural and force-generating myofibrillar proteins are much poorly expressed in in vitro maturated PSC-CMs when compared to adult- and fetal-heart samples [140]. This might be attributed to the absence of biomechanical stresses in vitro, which are normally present upon heart development in vivo [141].

Electrophysiological, and similar to contractile components, ion-transport related genes, such as those for voltage-gated potassium channels, e.g., *KCNJ2* and Ryanodine receptor *RYR2*, were poorly expressed in immature CMs [142]. The lower expression level of the *KCNJ2*-encoding membrane protein of the inward-rectifier current, as well as genes encoding beta-subunit members of the voltage-gated potassium channels, such as *KCNIP2*, *KCNAB1,* and *KCND3*, all affect both the inward-rectifier (Ik1) and the transient-outward (Ito) currents, respectively, leading to the characteristic "less negative" resting membrane potential in PSC-CMs (~–60 mV) compared to adult CMs (~–90 mV) [129,132,142,143]. Furthermore, studies have shown that PSC-CMs have few to no T-tubules, which are key components of excitation–contraction coupling (ECC) and a hallmark of mature and/or adult CMs; this is typified by unsynchronized Ca2<sup>+</sup> transients in immature CMs [144].

Metabolically, immature CMs have few and underdeveloped mitochondria, accounting for a small fraction of the cell volume. Adult CMs, on the other hand, show highly developed, well-distributed, and dense mitochondria, accounting for ~20–40% of the adult myocyte volume. During development, hypoxia is an early trigger for mesoderm cardiac specification [145]. The growing heart, thus, resorts to glycolysis as a major source (80%) of energy. As CMs mature and become terminally differentiated, mitochondrial oxidative capacity increases, with fatty acid β-oxidation (80%) becoming a major source of energy [146]. PSC-CMs recapitulate both mitochondrial structure and glycolytic dependence of embryonic-state CMs [92,147]. Recent studies have shown that tweaking the culture media composition

to mimic these metabolic changes, e.g., replacing high-carbohydrate, high-insulin, glucose-based, with low-carb, low-insulin fatty-acid-based media-enhanced maturation [148]. Table 1 summarizes the major differences between human-PSC-derived and adult CM.


**Table 1.** Major structural, electrophysiological and metabolic differences between human PSC-derived CMs and maturated/adulte CMs.

Over the years, efforts have been made to enhance the maturation of PSC-CMs, and these include prolonged cultures, using stiff gel micro-patterned substrates, and application of electrical and/or biochemical stimuli [132,149–152]. The overall goal was to simulate the in vivo environment of the myocardium, where CMs are under constant physical, topographic, and humoral stimuli leading to their structural and functional maturation.

#### *4.3. Engineered Heart Tissue*

Importantly, the accumulated knowledge of cardiac stem cell biology and maturation has culminated in the so called "Engineered Heart Tissue" (EHT), a milestone achievement. The nascent EHT is attributed to work done by Zimmermann and Eschenhagen in the early 2000s [153,154]. Ever since, EHT technology has rapidly progressed through refinements in mechanical loading, electrical stimulation, medium supplementation, and miniaturization. The result was a 3D cardiac tissue structure with mature CMs and near-physiological contractile forces [155]. The pioneering work of these scientists has opened the doors for more revolutionary developments, such as 3D bioprinting, organ-on-chip platforms, and laser-cut decellularized myocardium, all with ample opportunities for both basic research and clinical applicability [156–158].

### **5. Applications of PSCs in Cardiovascular Research**
