Human iPSC-Based In Vitro Cardiovascular Tissue Models for Drug Screening Applications

Abstract

Purpose of Review:

To provide an overview of human induced pluripotent stem cell (hiPSC)-derived cardiovascular lineages and describe their impact on drug testing in vitro.

Recent Findings:

hiPSCs have garnered tremendous interest over the last decade due to their potential for unlimited proliferation and differentiation into cardiovascular lineages. Technologies using tissue engineering, 3D bioprinting, and organ-on-a-chip platforms composed of hiPSC derivatives can produce cardiovascular tissue mimetics that enhance drug screening applications.

Summary:

hiPSC-derived cardiovascular lineages advance drug screening efforts by using autologous cells that are more therapeutically relevant. Established approaches to reproducibly generate hiPSC-derived cardiovascular lineages and their subsequent organization into 3D constructs more accurately mimic the physiological organization of cardiac tissue, leading to improved identification of potential drug targets for therapeutic testing.

1. Introduction

Cardiovascular diseases remain a major cause of mortality worldwide. According to the American Heart Association, more than 2300 deaths per day are attributed to them, and over $300 billion in treatment costs are incurred, with this cost projected to increase to over $1000 billion by 2030 [1]. Owing to limitations in mimicking the pathophysiology of clinical cardiac pathologies using preclinical models, drug development has been challenging. Therefore, it is crucial to develop robust biomimetic models of human heart diseases for accelerating drug discovery and development.

Human pluripotent stem cells (hPSCs), including human induced pluripotent stem cells (hiPSCs) and human embryonic stem cells (hESCs), have garnered immense attention for their applications in cell therapy. The unique versatility of hPSCs to differentiate into cardiovascular lineages, combined with recent advances in biomaterials, microfluidics, and organoid technologies, has paved the way for sophisticated in vitro cardiovascular tissue models, ranging from cardioids and heart-on-a-chip models to transplantable cardiac patches and 3D-bioprinted hearts. Since their successful generation in 2006 [2], hiPSC technology has garnered considerable interest for in vitro drug screening and personalized medicine, owing to the autologous nature of the cells. Human iPSC-based cellular models have advanced to organ-on-a-chip and organoid models that recapitulate salient aspects of native cardiac tissue architecture and cell-cell interactions. Despite the potential of PSC-derived cardiomyocytes (PSC-CMs), their immaturity and heterogeneity are limitations that have been addressed through the use of external stimuli, such as mechanical and electrical stimulation. Multi-cellular interactions, facilitated by the inclusion of vascular lineages and fibroblasts, further enhance the quality of resulting 3D cardiovascular tissues [3].

Bioengineering techniques, such as microfluidics and 3D bioprinting, have been developed to leverage extracellular matrix (ECM) signaling for cellular organization in 3D culture systems. The ECM plays an important role in supporting interactions between cells and providing structural support for tissue morphogenesis and homeostasis [4]. Since the ECM undergoes significant remodeling during heart development, postnatal cardiac maturation, and regeneration [5], it is a critical component in the generation of 3D cardiovascular tissue models.

Here, we overview the salient aspects of hiPSCs differentiation and maturation of iPSC-CMs in vitro. Furthermore, we describe the application of hiPSC derived cells for generating 3D in vitro cardiovascular tissue models. Finally, we discuss the application of these 3D in vitro models for drug screening and development, along with a critical assessment of the emerging technologies and concepts for future improvements.

2. Generation and Maturation of hiPSC-CMs

Highly efficient differentiation protocols have been developed for generating hiPSC-CMs. These protocols are based on the time course of important developmental signaling pathways during cardiac development. The key steps involve stage-specific application of soluble factors to induce iPSC mesodermal induction and specification, followed by cardiac specification and CM differentiation, which have been previously reviewed in detail [6,7]. Until 2012, mesodermal induction in these approaches was predominantly guided by growth factors such as activin A, bone morphogenetic protein 4, and basic fibroblast growth factor, while subsequent differentiation was achieved using vascular endothelial growth factor-A, fibroblast growth factor 2, and Dickkopf-related protein 1 [8]. However, in 2012, Lian et al. demonstrated that Wnt modulation via small molecules facilitates higher yields of CMs with superior purity [9,10], which was later supported by Burridge et al. [11,12].

Based on these landmark publications, most present-day cardiac differentiation protocols rely on the programmed activation and inhibition of Wnt signaling pathways [13]. These methods initially convert iPSCs into cardiac mesoderm by inducing β-Catenin signaling pathway, a specific canonical Wnt pathway, using a glycogen synthase kinase inhibitor such as CHIR99021 [14]. Subsequently, this mesoderm population is differentiated into cardiac progenitor cells and then into CMs using broad-spectrum Wnt inhibitors, such as the porcupine inhibitor IWP-2 or the tankyrase inhibitor IWR-1. Over the years, numerous research groups have adopted and refined this Wnt modulation strategy by optimizing the concentrations, duration, and timing of these small molecule treatments to improve consistency and efficiency in the hiPSC-CM differentiation protocols. Recent studies advocate a 24-hour culture with 7 μM CHIR99021 on day 0, followed by 48 hours with 5 μM IWR-1 on day 2, and continued maturation until day 15 [13].

Despite the ability to generate hiPSC-CMs that express CM-specific phenotypic markers, such as cardiac troponin, myosin heavy chain, and sarcomeric actinin, while functionally capable of undergoing spontaneous contraction, these cells are generally immature in phenotype and function compared to adult primary CMs (Table 1). As a result, biochemical strategies have been tested to enhance hiPSC-CM maturation. For example, the addition of thyroid hormone triiodothyronine (T3) to hiPSC-CM cultures for 7 days promoted CM maturation, as evidenced by enhanced cell size and sarcomere length, as well as improved calcium handling kinetics, mitochondrial activity, and contractile force [15]. Additionally, supplementation with fatty acids has been shown to induce an adult-like CM morphology. Whereas glucose is the source in most cardiac differentiation protocols, adult CMs utilize energy produced by oxidative phosphorylation using lipids. Accordingly, supplementation of the culture medium with fatty acids rather than glucose can promote CM maturation, characterized by increased cell size, myofibril density and alignment, as well as improved contractile force, calcium action potential, and mitochondrial oxidative capacity [16].

Table 1.

Summary of the key differences between fetal, adult, and iPSC-CMs.

Characteristics	Fetal CMs	Adult CMs	iPSC-CMs
Morphology
Cell volume	Circular/Irregular	Regular	Circular/Irregular
Shape	Small	Large	Small
Structure
MLC-2v/β-MHC	Low	High	Low
MLC-2a/α-MHC	High	Low	High
Sarcomeres	Disorganized	Organized	Disorganized
T-tubules	Absent	Isoform	Absent
Troponin I	Fetal isoform	Adult	Fetal isoform
Gap junctions	Circumferential	Present	Circumferential
Electrical
Max diastolic potential (mV)	~ −40	~ −90	~ −60
Action potential (mV)	70 – 90	110 –120	70 – 90
Metabolism
Metabolic pathways	Glycolysis	FA	Glycolysis
Substrates	Glucose, lactate	Oxidation	Glucose, lactate, FA
Oxidative phosphorylation	Low	High	Low
Mitochondria number	+++	+	+++

CM: cardiomyocytes; iPSC-CMs: induced pluripotent stem cell cardiomyocytes; FA: fatty acids.

Adapted from Bizy et al. [28].

Besides soluble factors for inducing CM maturation, biomechanical and bioelectrical approaches have also proven effective. For example, Ribeiro et al. studied the effect of hiPSC-CM maturity on polyacrylamide, generating more force on surfaces with 10 kPa physiological stiffness than on stiffer surfaces [17]. Furthermore, biophysical regulation using rectangular micropatterns having a 7:1 aspect ratio led to higher alignment of sarcomeric myofibrils and displayed more mature electrophysiology, T-tubule formation, calcium flow and higher contractility. Vunjak-Novakovic and coworkers highlighted that, since the responsiveness of hiPSC-CMs to physical stimuli declines over time, such conditioning should be initiated early, while the cells still exhibit high plasticity [18]. In this regard, they reported that electromechanical stimulation applied with gradually increasing intensity to cardiac tissues derived from early-stage hiPSC-CMs resulted in well-differentiated cardiac muscle, accompanied by marked expression of genes associated with adult-like conduction, maturation, ultrastructure, energetics, and calcium handling. Therefore, in addition to early initiation, they demonstrated that the increasing stimulation intensity mimics the mechanical loading during the fetal–postnatal transition and promotes cardiac maturation.

Following differentiation, the resulting iPSC-CMs typically comprise a heterogeneous mix of atrial CMs (ACMs), ventricular CMs (VCMs), and nodal cells [19]. However, these mixed populations have been considered unsuitable for disease modeling and for testing drugs that target specific CM subtypes [20]. The two myocyte subtypes can be distinguished based on their distinct electrophysiological properties, gene expression profiles, and biomarker expression [21,22]. Studies, in this regard, have shown that VCMs originate from a different mesoderm population than the one that gives rise to ACMs [23]. As a result, multiple studies have investigated strategies to generate chamber-specific CM subtypes from hiPSCs by modulating signaling pathways involving Wnt, fibroblast growth factors, activin, and retinoic acid [24–26]. More recently, Radisic and coworkers demonstrated that, in addition to using directed differentiation protocols, applying electrical conditioning can further facilitate the specialization of ACM versus VCM [27].

3. Three-dimensional Cardiovascular Tissue Models for Drug Screening

The field of cardiac tissue engineering gained prominence in the late 20^th century following the foundational work of Langer and Vacanti, who established tissue engineering as a new research discipline [29]. Vunjak-Novakovic and colleagues were the first to report optimal cell seeding and culture parameters for the development of in vitro cardiac tissues [30]. Soon thereafter, in the early 2000s, multiple strategies including the use of 3D scaffolds [31], perfusion bioreactors [32], and electrical stimulation [33] were shown to enhance the functional performance of engineered cardiac constructs. Concurrently, Eschenhagen and his team demonstrated that advanced culture environments involving mechanical stretch could promote the differentiation of cardiac myocytes into myocardium-like tissue [34].

The discovery of hPSCs in 2006 led to their rapid integration into cardiac tissue engineering [2]. Initially, conventional tissue culture polystyrene dishes were utilized for high-efficiency production of hiPSC-CMs and hiPSC-ECs. However, conventional dishes are 2D and do not simulate the 3D architecture, biomechanical cues, and intercellular interactions of native tissues. Consequently, 3D cell culture platforms composed of hiPSC-CMs, hiPSC-derived endothelial cells (ECs), and other support cells have been utilized for engineering complex cellular models. Such 3D platforms include engineered heart tissues (EHTs), 3D bioprinting-based tissue engineering, organoids, and heart-on-a-chip systems (Table 2). These platforms are amenable to personalized drug discovery and screening to obviate drug cardiotoxicity or to identify potential therapeutic drug targets.

Table 2.

Key advantages and disadvantages of 3D cardiovascular tissue models for human iPSC-based in vitro systems.

Cardiovascular Model	Advantages	Disadvantages
Engineered Heart Tissue	Mimics native contractile functionality and force generation; reproducible and scalable	Poor oxygen and nutrient supply in larger constructs; limited spatial complexity
3D Bioprinting	Allows precise spatial control and complex geometries; offers potential for vascular and multicellular integration	Challenging fabrication with respect to bioink limitations and cell viability; resource-intensive; poor scalability
Cardiovascular Organoids	Recreates intricate cardiac microenvironment; can be generated without exogenous materials; cost-effective	Limited control over internal organization; high batch-to-batch variability; poor vascularization and perfusion
Heart-on-a-Chip	Enables real-time monitoring; suited for integrating multiple cell types; supports high-throughput studies	Complex fabrication; requires integration with other models to recapitulate the 3D environment; high cost

iPSC: induced pluripotent stem cell; 3D: three-dimensional.

3.1. Engineered Heart Tissue

Engineered heart tissue (EHT) models, first established by Zimmerman et al. [35–37], have been developed in desired shape by mixing hydrogels, such as collagen and fibrin, with hPSC-CMs, often accompanied by supportive cell types like fibroblasts, stromal cells, and/or ECs. The hydrogel facilitates cell-ECM interactions, allowing the material to compact into a 3D tissue unit that can contract spontaneously or in response to mechanical or electrical stimulation [38]. Consequently, the EHT models can closely mimic key aspects of in vivo 3D organization, contractile functionality, and cell–cell interactions. Over the past two decades, multiple groups have reported a wide range of tissue models grown from pluripotent stem cells, including those having diverse geometries and formats [39]. Some of the EHT shapes reported in this regard include patch [40], biowire [41], microwire [42], and ring [43].

Goldfracht et al. reported the generation of chamber-specific EHTs by embedding hPSC-derived VCMs and ACMs in collagen-based, ring-shaped 3D structures (Figure 1A) [43]. The study demonstrated that the ventricular and atrial EHTs expressed several chamber-specific markers at both the RNA and protein levels, while exhibiting distinct electrophysiological and contractile properties. Furthermore, they validated the clinical relevance of their EHT model for pharmacological testing by showing that vernakalant, an atrial-selective antiarrhythmic drug, modulated atrial electrophysiological response and prevented new atrial arrhythmias. Previously, Lemoine et al. used fibrin-based EHTs [44] as a testing platform for ventricular arrhythmias triggered by drug-induced and genetic long-QT syndrome [45]. The implemented platform allowed the application of sharp microelectrodes to measure action potentials without enzymatic dissociation, thereby serving as a sensitive model to study the response to I_Kr blockers, which are antiarrhythmic medications that slow repolarization by blocking a potassium channel.

Figure 1. Three-dimensional cardiovascular tissue models.

(A) (i) Immunostaining of atrial and ventricular engineered heart tissues (EHTs) was performed using cardiac troponin I (cTnI) together with either the atrial marker sarcolipin (SLN) or the ventricular-specific marker MLC2v, confirming the formation of chamber-specific EHTs. (ii) Real-time qPCR analysis of atrial-specific marker, GJA5, and ventricular-specific marker, MYL2. (iii) Representative activation maps of the arrhythmogenic atrial EHTs showing arrhythmic activity and restoration of normal rhythm following drug treatment with 30 μM vernakalant. Reproduced from [43]. (B) (i) Schematic illustration of processing omentum-derived cells and matrix into iPSC-based 3D-printed, vascularized heart tissues. (ii) A printed heart model within a support bath, which upon extraction was injected with red and blue dyes to demonstrate the left and right hollow ventricles. (iii) Confocal images of the printed heart showing cardiomyocytes (CMs) in pink, and endothelial cells (ECs) in orange. Reproduced from [61]. (C) (i) Schematic illustration showing production of self-organizing cardioids from human pluripotent stem cells. (ii) Maximum intensity projections demonstrating successful CMs (MYL7-GFP and TNNT2) and ECs (CDH5) specification of the cardioids. Scalebar: 200 μm. (iii) Using single-cell RNA sequencing to classify cardioid cells based on markers determined by Cui et al. [77]. CM: cardiomyocytes; EC: cardiac endothelial cells; EP: epicardial cells [65]. (Fig 1C Reprinted from: Hofbauer P et al. Cell. 2021;184:3299-3317.e22, with permission from Elsevier) [56]. (D) (i) The implemented seeding procedure for the microfluidic Heart-on-Chip model. (ii) Representative confocal image of a full μEHT composed of CM, fibroblasts, smooth muscle cells, and EC (CFSE). (iii) Conduction velocity of μEHTs composed of CM and fibroblasts (CF) and CFSE stimulated at a frequency of 2 Hz. (iv) Gene ontology (GO) chord diagram of three upregulated GO terms in the μEHTs within the Heart-on-Chip model. Reproduced from [73].

Despite the key advantages of 3D EHTs in cardiovascular drug screening, one of their critical limitations includes poor oxygen and nutrient supply, especially when the thickness exceeds 100 μm, ultimately leading to an ischemic core [46]. Therefore, an effective EHT model requires simultaneous vascularization as the CMs mature. Caspi et al. reported the construction of the first 3D vascularized human cardiac tissue by combining CMs with ECs and embryonic fibroblasts [47]. More recently, Giacomelli et al. demonstrated the successful integration of CMs and ECs in 3D microtissue systems through their simultaneous differentiation from hiPSCs following initial cardiac mesoderm induction [48]. Concurrently, Cheng et al. showed that hiPSC-derived ECs promote a superior maturation and functionality of hiPSC-CMs in vitro and in vivo [49]. These studies have paved the way for the development of high-throughput drug screening methods utilizing 3D cardiovascular tissues.

3.2. 3D Bioprinted Cardiovascular Tissues

3D bioprinting strategies now achieve sub-100 μm spatial resolution, allowing researchers to sculpt heart-like tissues that faithfully mimic native structure [50]. Different techniques were applied in 3D bioprinting, including extrusion, inkjet bioprinting, laser-assisted bioprinting, coaxial, and volumetric photopolymerization [51], as well as the latest FRESH v2.0. Each offering unique control over cell alignment and matrix chemistry [52]. These techniques employ cell-encapsulated bio-inks derived from natural or synthetic ECMs to form defined 3D geometries, thereby enabling precise cellular placement. The principles of each bioprinting technique have been reviewed extensively elsewhere [44, 45]. A 3D bioprinting platform for direct embedding of hydrogel inks within a sacrificial hydrogel was developed for assembly of highly perfusable cardiac tissues by the bioprinting of vascular channels into a living 3D matrix of densely packed iPSC-derived cardiac organoids [55]. The resulting bioprinted cardiac tissue demonstrated desired 3D shapes and tunable mechanical properties, enabling the formation of 3D organ-like cultures. In another example, 3D printing has been utilized to generate cardiovascular tissue devices using biomaterials of high conductivity for the integration of strain sensors within micro-grooves to facilitate cell organization in an anisotropic pattern [56]. The sensors provide a platform for non-invasive, electronic measurement of tissue-generated contractile forces.

Relying on their precisely controlled geometry parameters, 3D bioprinting cardiovascular tissues can build nature-like micro-tissues or organoids for drug screening. Cantilever-printed micro-tissues that occupy standard 96-well plates [57], dense FRESH-printed strips (more than 6×10⁸ M cells per mL) for 24-well screening [58], endothelial-lined FRESH channels that speed CM alignment and sharpen drug gradients [58], and spheroidal droplet models molded by nozzle-free extrusion [59] illustrate how extrusion, FRESH and droplet bioprinting can yield heart tissues ready for assays. Flow-conditioned vascular grafts, printed as coaxial smooth muscle with fibroblast tubes and lined with endothelium, replicate platelet adhesion under shear, thus enabling antithrombotic screens [60]. Noor et al. reported the fabrication of 3D-printed, thick, vascularized, and perfusable cardiac patches using biopsy of an omental tissue, from which the cells were reprogrammed into iPSCs and then differentiated into CMs and ECs. At the same time, the ECM was processed into a personalized hydrogel (Figure 1B) [61].

Most of these studies demonstrate that incorporating ECs cells enhances tissue alignment, provides a living barrier model for studying diffusion-limited drugs, and supplies paracrine cues that make pharmacodynamic readouts more faithful to real human physiology. These studies demonstrate the potential of 3D bioprinting for cardiovascular drug discovery.

3.3. Cardiovascular Organoids

Cardiac organoids are tissue constructs generated by the self-assembly of cells. hiPSC-derived organoids possess both differentiation and self-organization potential. Compared to other organs, such as the brain, lungs, and liver, cardiovascular organoids have lagged behind, but in recent years have made marked progress. Two commonly used strategies for generating cardiac organoids are direct assembly and self-assembly [62]. Direct assembly involves aggregation of starting cell types in hydrogels, whereas in self-assembly, hPSCs or derivative cells aggregate together into spheroids in the absence of exogenous ECM. The addition of soluble factors within the culture media further promotes tissue morphogenesis and vascularization within the organoid [63]. An example of direct assembly was reported by Drakhlis et al. in 2021, who generated 3D heart-forming organoids by embedding hPSC aggregates and inducing differentiation through Wnt pathway modulation using CHIR and IWP2 [64]. The platform was validated by investigating genetic defects, where NKX2.5-knockout organoids were revealed to recapitulate analogous aspects of the cardiac phenotype observed in transgenic mice. In the same year, Hofbauer et al. reported the development of the first self-organizing cardiac organoids from hPSCs, called cardioids, which can intrinsically specify, organize, and transform into chamber-like structures (Figure 1C) [65]. Subsequently, the authors performed single-cell RNA sequencing analysis of these cardioids to reveal the signaling pathways guiding the endothelial and myocardial self-organization, which is a critical principle of human cardiogenesis. Although groundbreaking, their protocol relied heavily on complex mixtures of growth factors. To overcome this limitation, Lewis-Israeli et al. introduced a comparatively more straightforward small molecule-based approach using sequential Wnt modulation to produce complex heart organoids [66].

Overall, organoids can mimic numerous physiological properties of tissue development in vitro; however, one of their key drawbacks is the frequent absence of crucial cooperative paracrine interactions between neighboring tissues. To investigate the role of such multi-tissue interactions, Silva et al. reported a multilineage iPSC-derived organoid that mimicked the cooperative development and maturation of cardiac and gut tissues [67]. The simultaneous development of both tissues was observed to enhance the phenotypic and functional maturation of the CMs, whose ultrastructural features, as seen under electron microscopy, resembled those of the native third-trimester human fetal heart. Another example of multicellular cardiac organoids was highlighted by Voges et al., who developed a one-step differentiation protocol for vascularized human cardiac organoids [68]. The authors demonstrated that the inclusion of ECs enhanced the expression of maturation markers and improved function in the organoid structures.

Unfortunately, most currently reported protocols for cardiovascular organoids suffer from limited reproducibility and scalability. To address this, Prondzynski et al. recently described the first cardiac organoid model generated entirely in a bioreactor [13]. The single-cell RNA sequencing of the resulting organoids primarily revealed ventricular-like CMs, along with non-CMs such as fibroblasts and ECs. In summary, these studies highlight the growing potential of organoid models for disease modeling and the discovery of cardiovascular drugs.

3.4. Heart-on-a-chip

A heart-on-chip refers to more complex systems that incorporate 3D cardiovascular models within micro- or even nano-systems formed from microfluidic channels. Heart-on-a-chip platforms provide well-controlled flow and can integrate biosensors to study cellular physiology. They also allow for multi-cellular culture in separate channels or chambers. Heart-on-a-chip platforms are becoming increasingly more physiologically relevant for modeling cardiovascular responses. The EHT model can be integrated using microfluidic chips to form a cardiovascular tissue model that can be electrically stimulated by 3D electrodes embedded into the chip. For example, engineered heart tissues paced at 2 Hz through soft conductive-hydrogel pillar electrodes exhibited a 45% increase in twitch force and a faster decay of calcium transients [69].

Heart-on-a-chip platforms can now incorporate other technologies, such as optical mapping and genetically encoded calcium and voltage sensors, for the simultaneous analysis of calcium dynamics, electrophysiological properties, and contractile force [70]. Soft hydrogel-pillar electrodes built into the chamber pace the tissue at physiological rates while streaming continuous force and field-potential data [69]. High-resolution optogenetic modules enable light pulses to steer conduction in hiPSC-derived tissues and reveal subtle anisotropies in real-time [71]. Two-photon micro-printing has also been used to create a miniature ventricular pump whose cell-driven pressure-volume loops mimic those of a native heart [72]. Heterotypic constructs that mix CMs with ECs, smooth muscle cells, and fibroblasts self-assemble into capillary-lined μEHTs and show adult-like conduction velocities (Figure 1D) [73]. Perfusable extracellular-matrix lattices now double twitch force and extend culture life beyond a month [51]. Finally, coupling the cardiac chamber to a liver module lets researchers watch a drug’s hepatic metabolites alter functionalities (beat rate and QT interval) in the same circuit [74]. Taken together, these diverse platforms point toward self-sensing, multimodally stimulated micro-organs that capture in vivo biomechanical clues.

Recent drug-screening studies add functional depth to these chips. A multi-lineage device that cultures hiPSC-CMs and hiPSC-ECs in parallel channels detects loss of contractility and endothelial barrier breakdown induced by sorafenib, capturing dual-cell toxicity that cannot be achieved by 2D assays [75]. An endothelialized “μEHT-on-chip” built from CMs, ECs, smooth muscle cells, and fibroblast cells self-organized into capillary-lined tissues and reproduced verapamil dose-responsiveness with adult-like conduction profile [73]. A tumor-heart dual-chip quantified doxorubicin IC₅₀ in cancer spheroids, meanwhile tracking beat-rate slowing and conduction block in the adjacent endothelialized cardiac chamber [76]. Endothelial-free strips integrated with hydrogel-pillar electrodes resolved force-frequency shifts to isoproterenol within minutes, illustrating high-throughput ionotropic profiling [69]. Finally, a perfusable cardiac-microenvironment chip exposed to propranolol showed flow-dependent drug uptake and a 30% greater negative chronotropy than static cultures, underscoring the role of convection in pharmacokinetics [51]. These examples illustrate the application and potential of heart-on-a-chip technology for cardiovascular drug discovery.

4. Future Directions

The advent of somatic cell reprogramming has led to the emergence of new frontiers in modeling human physiology by enabling the induction of pluripotency through complex cellular mechanisms. In vitro iPSC-derived 3D models have advanced the assessment of drug efficacy and pharmacokinetics, establishing them as effective preclinical platforms for drug screening and discovery. These approaches have gained wider appeal in recent years with the 2022 introduction of the FDA Modernization Act 2.0, which allows for the replacement of animal testing with alternative platforms, such as in vitro models, for establishing drug safety and efficacy [78,79].

Nevertheless, before 3D in vitro cardiovascular models can be widely used for drug discovery and screening applications, several bottlenecks need to be overcome. First, reproducibly generating scalable quantities of high-quality and mature patient-specific hiPSC-derived cardiovascular lineages will require rigorous standardization and cell manufacturing quality control measures. Automating the process of cardiovascular differentiation or generation of 3D models would significantly enhance the reproducibility of in vitro systems, thereby making them more suitable for high-throughput studies. Second, heart-on-a-chip platforms can be integrated with other organ or tissue systems, including the lung, liver, and blood vessels, to mimic whole-body microphysiological systems better [80,81]. Such approaches create increasingly complex 3D models with greater capabilities but are susceptible to technical challenges in manufacturing high-quality 3D models for each organ or tissue unit. Third, technological advancements for understanding the molecular mechanisms of drug effects on cell phenotype and function will be important. Recent technological revelations in spatial transcriptomics and proteomics, multi-omics, and CRISPR gene editing can provide fundamental insights that inform drug safety and efficacy [82,83]. Finally, with the increasing use of artificial intelligence and machine learning algorithms, the ability to perform predictive computational modeling of patient-specific drug responses is emerging and will become increasingly impactful in the future [84,85]. Despite these current limitations, with continued progress in this highly attractive field, we anticipate that these challenges can be overcome in the near future.

5. Conclusions

3D models using hiPSC-derived cardiovascular lineages have advanced the progress of drug screening and discovery in well-controlled in vitro environments. Compared to conventional 2D culture, these 3D models better mimic the complex cellular composition, organization, and three-dimensional geometry of native cardiovascular tissues. Such systems can be integrated with mechanical or electrical stimuli, biosensors, and optical imaging systems to create more complex environments and data readouts. The potential benefits of 3D cardiovascular tissue models are high, and we anticipate that these in vitro models can propel the development of new drugs toward clinical translation.

Data Availability:

Data sharing is not applicable to this article as no new datasets were generated or analyzed in this work. All data discussed in this review are available within the published literature cited in the reference list.