The heart is a vital organ that maintains every other organ in the human body through the pumping and circulation of blood. Therefore, heart problems pose a great threat to human morbidity and mortality. In this context, cardiovascular disease, including ischemic heart disease, cardiomyopathy, and hypertensive heart disease, is one of the leading causes of death for mankind, and the global death toll and consequential burdens have continued to increase (1). To overcome these diseases, it is necessary to identify the pathological factors of disease progression and their mechanism, and to further develop effective treatments suitable for each disease. For this purpose, diverse three-dimensional (3D) human cardiac tissue models that mimic the human heart have been actively developed. These tissue models are free from the ethical issues which usually surround conventional animal models and can also eliminate gaps present between different species. Depending on which cardiac cells and engineering platforms are used, cardiac tissue models showing various shapes and characteristics can be developed (Fig. 1). Moreover, some biophysical cues such as mechanical and electrical stimulation have been applied to cardiac tissue models to further generate functionally matured ones.
The developed cardiac tissue models can be utilized for the modeling of various heart diseases. By using patient-derived induced pluripotent stem cells (iPSCs), cardiac tissue models expressing disease-associated genes can be developed (2). This model helps us to understand the differences in characteristics and phenotypes caused by a particular gene. Recently, such models can also be produced by gene editing methods, such as the CRISPR-Cas9 system, which manipulate specific target genes, and the role of genes can be investigated through comparison with normal models (3). Finally, heart disease models can be established by induction of the causative pathway of disease or by exposure to unfavorable environments (
Cardiomyocytes (CMs) are the basic beating units of the heart, and there are several types such as pacemaker cells, ventricular CMs, and atrial CMs. For decades, fully differentiated primary CMs extracted from the heart have been used as
There are two methods for acquiring CMs, the differentiation method from human iPSCs and the direct reprogramming technique. The differentiation method from iPSCs first generates mesoderm cells, and then CMs are generated through cardiac progenitor cells. This process mimics normal embryonic development, for which several signaling pathways, such as Wnt/β-catenin and Activin A/BMP4 signaling, are regulated at each stage (5). With this method, CMs can be obtained without burden to the patient and an unlimited supply may be possible if iPSCs are cultured adequately. However, the differentiation method still has limitations such as yield, purity, variability, reproducibility, and cost-effectiveness (6). Direct reprogramming, which enables the acquisition of CMs without going through pluripotent stem cells, is gaining attention as an alternative to produce cardiac cells (7). Directly induced CMs were first produced using three transcription factors (
Since cardiac tissue models, composed only of CMs, lack maturity compared to the native heart, the cell composition of cardiac heart tissue models is currently being highlighted (12). An adult human heart contains many types of non-myocytes, including endothelial cells (ECs), cardiac fibroblast (CFs), and leukocytes, and their amount is even greater than that of CMs (13). Each plays a different role in the human heart and affects the maturity of CMs, which will eventually be important in creating disease models (14).
The ECs distributed inside the vessel form the major population of the heart, and serve to transport oxygen and nutrients to maintain the heart. For implementation of vascularized networks in cardiac tissue models, human umbilical vein endothelial cells (HUVECs), human cardiac microvascular endothelial cells (HMVECs), and human adipose-derived stem/stromal cells (hASCs) have widely been used (15-17). Although vascular structures can be produced using these cells, studies on the differentiation of human iPSCs to ECs are also actively underway, due to the strengths of reprogramming techniques, such as the possibility of mass production, development of patient-specific models, and simulation of interaction in developmental stages. ECs can be differentiated simultaneously with CMs from the cardiac mesoderm, and such differentiated ECs are advantageous for reconstituting vascular networks in heart (18). In recent years, more efficient EC differentiation methods have continued to be developed (19, 20). For example, the method of generating functional ECs with high purity from cardiogenic and hemogenic mesoderm was developed (20). For the same reason, the CF differentiation methods from human iPSC have also been developed (21, 22). Until now, CMs, ECs, and CFs have been recognized as the three major cell types for cardiac tissue models, and other non-myocytes, such as smooth muscle cells, have been occasionally added (23). Furthermore, as the immune response is a critical factor for cardiac homeostasis and disease pathophysiology, it is necessary to implement immune cells in 3D cardiac models (24). For example, macrophages in heart are known to interact with CMs via connexin 43 for improved electrical conduction (25).
CMs and non-myocytes play important roles in the transmission of signals through interactions with other cells, and these interactions are also involved in disease progression and normal development. The ECs contribute to cardiac development, remodeling, and regeneration by interacting with CMs (26). They not only secrete several paracrine factors, such as nitric oxide (NO), neuregulin-1 (NRG-1), and apelin (APLN), to improve function and contractility of CMs during normal development, as well as to enhance the cardio-protection in a disease environment (26). ECs actively communicate with other non-myocytes. ECs resident in endocardium contribute to CF generation through endothelial-to-mesenchymal transition during normal embryonic development (27). ECs also participate in immune responses, helping immune cells migrate, by secreting cytokines, and even acting as an antigen presenter (28). The main role of CFs, generally located between CMs, is to construct the extracellular matrix (ECM) environment and maintain the homeostasis of ECM in the heart. In normal conditions, they communicate with CMs via gap junctions, membrane nanotubes, and paracrine signaling, and ultimately enhance the structural maturation and electrophysiological function of CMs (22, 29). However, excess activation and accumulation of CFs could be one of the phenotypes in cardiac diseases, such as myocardial infarction, cardiac fibrosis, and hypertensive heart disease. From this point of view, the implementation of CFs is important in establishing both the functionally advanced normal cardiac models and the heart disease models. The interactions between CFs and other non-myocytes have been examined in previous studies. When tested in the co-culture model, CFs assisted the proliferation of ECs and their sprouting (30). IL-1β expression in CFs, under a disease condition of myocarditis, has been shown to recruit leukocytes and induce the inflammatory process (31). Therefore, the incorporation of non-myocytes to cardiac tissue models is essential to replicate their interactions in a human heart tissue and predict the disease environment more precisely. In addition, it is required to develop culture platforms that can realize their interactions, including the secretion and absorption of cytokines, while each cell type can be maintained in optimal conditions.
Various types of cardiac tissue models, possessing different characteristics and merits, can be produced depending on technologies used for fabrication. The simplest form of the model is a cardiac spheroid, which can be mainly prepared using devices that enable the collection of single cells, including a microwell, hanging drop plate, and V-bottom plate (22, 32, 33). The multicellular cardiac spheroids are fabricated by combination of CMs, ECs, and CFs. For instance, multicellular spheroids comprising human iPSC-derived CMs (70%), ECs (15%), and CFs (15%) showed enhanced electrical maturation and contractile phenotypes, compared with multicellular spheroids generated with other fibroblasts (
Engineered heart tissue (EHT) models, first established by Zimmerman
Organ-on-a-chip technology has been applied to develop
In recent years, the development of cardiac organoids, which mimic early heart development in the human body, has been demonstrated. The first developed cardiac organoids derived from aggregates of human iPSCs recapitulated the cardiomyogenesis of embryonic development and the expression of other surrounding regions and cells, such as mesenchymal cells and ECs (45). As cardiac organoids are generated via intrinsic development program, they can represent the
Establishing ECM environments artificially in cardiac tissue models has been conducted using hydrogels prepared from synthetic materials (
Cardiac tissue models currently tend to be manufactured using human iPSCs. However, the maturity and functionality of these tissue models are still limited when compared to the
Cardiomyopathy is a heart disease indicated by enlarged, thickened, or stiffened heart muscle, which is associated with several intrinsic and extrinsic factors, but has been mainly recognized as an inherited disorder (66). Hypertrophic cardiomyopathy (HCM) and dilated cardiomyopathy (DCM), characterized by a thickening of the left ventricle and dilated left ventricle, respectively, are representative examples among the several types of cardiomyopathies (67). HCM is caused, in the majority of cases, by mutations of genes that encode the sarcomere proteins, while DCM can be caused by more diverse factors such as mutation of sarcomere and cytoskeletal proteins, systemic pathology, maternal mitochondrial DNA, and X-linked mutations (68).
For cardiomyopathies caused by genetic mutations, disease models can be produced using either patient-derived iPSCs or gene editing systems like CRISPR/Cas9 if the genetic information is known. In 2015, EHT was developed for DCM modeling using patient-derived iPSC-CMs which have mutations in the
Cardiomyopathies occur often without genetic mutation and family history, such as nonfamilial HCM. These cases can be modeled through the provision of external factors or stimuli. Zhao
Arrhythmia is a common disease in which an irregular heartbeat appears for various reasons. Arrhythmia is caused predominantly by problems in the electrical circuit in the heart (77). One of the major inherited arrhythmic disorders is long QT syndrome (LQTS), which is caused by mutation of cardiac ion channels. LQTS is named due to its clinical feature of a prolonged QT interval, and there are several types depending on which ion channel is affected (78). Thus, LQTS cardiac tissue models could be developed by incorporation of LQTS patient-derived cardiac cells or by the treatment of drugs that block specific cardiac ion channels. For instance, ring-shaped EHT models were generated with LQT2 patient-derived iPSC-CMs (79). These models exhibited a prolonged action potential duration, an abnormal calcium transient, and arrhythmic responses (79). The same research group demonstrated ring-shaped atrial arrhythmia in the EHT models and tested therapeutic strategies of electrical stimulation or anti-arrhythmic drugs (Fig. 3B) (80). QT prolongation in the heart could also occur due to the adverse action of the drugs (81). Previous study demonstrated the development of two types of LQTS EHT models (LQT1 and LQT2) through the treatment of drugs which inhibit each ion channel (HMR-1556 and E-4031 for blocking
Myocardial ischemia is characterized by a limited oxygen supply due to a blood vessel occlusion, which leads to the deterioration of heart function. Myocardial infarction is a life-threatening disease in which necrosis of the heart occurs due to extended ischemia or an acute blockage of the coronary artery (86). Models to simulate these events can be developed through culture in hypoxic conditions. For example, a previous study cultivated a cardiac tissue model fabricated using human iPSC-CMs in three types of hydrogels with different stiffness (0.8, 8, and 30 kPa) under conditions of a 1% oxygen concentration and investigated the effects of cell age and tissue stiffness on viability and reactive oxygen species (ROS) production (87). In another study, a myocardial infarction model was constructed by gradually decreasing the oxygen concentration towards the inside of the multicellular spheroids comprising 50% human iPSC-CMs and 50% non-myocytes (CFs, HUVECs, and hASCs) under conditions of a 10% oxygen concentration (88). This spherical disease model exhibited several pathological phenotypes, such as a fibrotic response and impaired calcium handling, and was used to test anti-fibrotic drugs and to identify the exacerbation of cardiotoxicity by drugs (88).
Reperfusion therapy is widely used to treat myocardial infarction, however, it can cause additional injury, called ischemia-reperfusion injury (IRI). IRI cardiac tissue models were first developed in the form of EHT in 2019 via a method of restoring oxygenation from ischemic conditions (89). These IRI models can be used for testing several therapeutic approaches targeting ischemic preconditioning, intracellular pH, the opening of the mitochondrial permeability transition pore (MPTP), and oxidative stress (89). Spherical IRI cardiac tissue models were also developed in another study using a similar inducing method, and these models showed some hallmarks of disease phenotypes such as cell death and increased secretion of cytokines associated with inflammation, angiogenesis, and cell migration (90). The angiogenic effect of IRI was confirmed through enhanced tube formation of HUVECs cultured with conditioned medium from spherical IRI tissue models (90). ECs are known to be an important mediator in myocardial dysfunction after IRI (91). To accurately predict pathological responses in IRI, it is necessary to implement the interaction between ECs and CMs. In this context, the cardioprotective roles of endothelial extracellular vesicles and their mechanism were investigated using an IRI heart-on-a-chip model (92). In a separate study, an IRI chip model was constructed through co-culture of human iPSC-CMs and iPSC-ECs in microfluidic chip, and changes in TSG101 and CD63 subunit expression of exosomes secreted in chip models were detected after ischemic injury and IRI, respectively (93).
Severe heart defects caused by myocardial infarction and ischemia evoke cardiac fibrosis, which eventually leads to heart failure. CF, known to maintain the homeostasis of ECM, is a key cell type in cardiac fibrosis characterized by excessive ECM production and disrupted balance of ECM homeostasis (94). In this regard, cardiac tissue models based on CFs have been developed for modeling cardiac fibrosis, and various fibrosis induction methods have been used. To simulate a fibrotic response in cardiac tissue models, the application of biomechanical cues could be used. In many organs including the heart, overexpression of transforming growth factor-β (TGF-β) signaling is known to accelerate fibrosis after injury by activating fibroblasts, remodeling ECM, and promoting myofibroblast conversion (95). Accordingly, activation of TGF-β signaling resulted in a fibrotic response of the cardiac tissue model with CMs and CFs encapsulated in gelatin methacryloyl (GelMA) hydrogels (96). Moreover, cyclic mechanical compression was applied to a microdevice containing CFs encapsulated with GelMA hydrogels, and this triggered a phenotypic remodeling of CFs to myofibroblasts (97). In another study, cardiac fibrosis-on-a-chip was designed and fabricated in the form of EHT containing human iPSC-CMs and CFs, and fibrosis was induced by activation of TGF-β signaling (98). This fibrosis EHT model in chip showed hallmarks of fibrosis such as collagen deposition and increased stiffness, and therapeutic efficacy of anti-fibrotic drugs like pirfenidone was examined (98).
The fibrotic cardiac tissue model could be optimized by tuning the ratio of CMs and CFs. Wang
Owing to the development of stem cell and reprogramming technologies and the inaccuracy and ethical issues in animal experiments, 3D cardiac model production is in the spotlight. Numerous cardiac tissue models, each having their own advantages, have been developed and shown the benefits of modeling heart diseases. The cardiac tissue models that incorporate genetic diseases have been realized by using reprogramming and gene editing techniques. Modeling heart diseases with various risk factors in addition to genetic factors, such as myocardial infarction and cardiac fibrosis, has been achieved via creation of disease-specific microenvironments. Despite significant improvement in heart disease modeling, the complexity and maturity of the cardiac tissue models are still insufficient to simulate the complicated features and pathophysiology of heart diseases. Therefore, application and incorporation of tissue-engineering platforms (
This work was supported by the Bio & Medical Technology Development Program (2022M3A9B6082675) of the National Research Foundation of Korea (NRF) funded by the Korean government, the Ministry of Science and ICT (MSIT). This work was also supported by the Yonsei Signature Research Cluster Program of 2021-22-0014.
The authors have no conflicting interests.