Advances in cardiac organoid technology: current trends and standardized quality assessment for toxicity testing in preclinical models
Article information
Abstract
Cardiotoxicity is a prominent cause of drug withdrawal, impacting drug development during preclinical and clinical phases, as well as post-market approval. Traditional 2-dimensional cell culture models, despite being resource-rich, suffer from a suboptimal physiological environment. Animal models provide an in vivo setting; however, their use is declining due to interspecies variability and ethical concerns. Cardiac organoids, which are derived from human induced pluripotent stem cells, incorporate human genetic information and contain various cell types within a 3-dimensional environment that mimics the human heart. As a result, they are becoming increasingly popular in the development of preclinical models for drug testing. To effectively use cardiac organoids as preclinical models, a standardized quality assessment is crucial. This review discusses current trends in cardiac organoid generation and examines recent analytical techniques for quality assessment.
Introduction
Toxicity is a primary cause of adverse drug reactions throughout preclinical and clinical drug development [1–3]. It also plays a significant role in the withdrawal of approved drugs, with cardiotoxicity accounting for 45% of these cases [4]. Drug withdrawals due to cardiotoxicity encompass a spectrum of phenomena, including cardiac arrhythmias, disturbances in electrophysical functions, myocardial ischemia, heart failures, myocardial necrosis, valvular heart disease, hypertension, and more [5–7]. Commonly used preclinical cell models have limitations, notably the absence of a physiological environment, relying predominantly on hERG (human ether-a-go-go-related gene) channel assays to detect cardiac arrhythmias [3,8]. Conversely, animal models offer the advantage of evaluating various types of cardiotoxicity in a physiological environment, but are subject to variations due to interspecies differences [9–11].
To overcome the limitations of these models, cardiac organoid models derived from human induced pluripotent stem cells (hiPSCs) have been proposed [12–17]. Cardiac organoids are composed of multiple cell types and are structured in a 3-dimensional (3D) environment that closely mimics the structure and environment of organs in humans [18]. The methods for generating hiPSC-derived cardiac organoids vary, leading to differences in the composition of cell types and their ratios based on the cell source and reagents used [19]. Therefore, to use the generated cardiac organoids as a cardiotoxicity model in drug development, it is essential to establish quality evaluation criteria [20]. This review presents the current trends in techniques for generating hiPSC-derived cardiac organoids and provides an overview of existing methods for quality analysis.
Trends in the development of cardiac organoids for toxicity testing
Ethics statement: This study was a literature review of previously published studies and was therefore exempt from institutional review board approval.
There are 2 main approaches for generating cardiac organoids using hiPSCs: the self-assembly method involving the formation of embryoid bodies and the reassembly of various 2-dimensional (2D) cell types at a specific ratio to generate a 3D structure [21–23]. Research trends in the differentiation of cardiac organoids from embryoid bodies are presented in Table 1 [14–17,24,25]. Hoang et al. [14] utilized a polydimethylsiloxane (PDMS)-based cell-patterning technique to induce self-organization of embryoid bodies of a specific size, leading to self-tissue organization and the development of a cardiac organoid model mimicking early heart development stages. Hofbauer et al. [15] showed that the regulation of the WNT/BMP signaling pathway facilitated lineage-specific self-tissue organization, resulting in cardiac organoids with structures resembling ventricles. These models exhibited ventricular-like structures and demonstrated heart regeneration and early features of heart disease when subjected to cryoinjury. Lewis-Israeli et al. [16] developed a cardiac organoid model with characteristics similar to human fetal heart employing a 3-step regulation of the WNT signaling pathway. This disease model was validated as reproducing congenital heart disease under conditions similar to maternal gestational diabetes. Drakhlis et al. [24] directly differentiated induced pluripotent stem cell aggregates embedded in Matrigel (Corning, New York, NY, USA), using a 2-step process for WNT signaling regulation, creating cardiac organoids resembling early developmental stages before heart tube formation. These organoid models replicated abnormal cardiac function phenotypes through NKX2.5 knockout. Lee et al. [17] developed a protocol for efficiently enhancing electrophysiological characteristics in cardiac organoids using ZM447439, an Aurora kinase inhibitor, during the differentiation process. Lee et al. [26] developed a protocol for chamber-formed cardiac organoids with cardiac structure and functionality using Matrigel to form embryoid bodies and a 2-step process for WNT signaling regulation. They confirmed vascularization through in vivo transplantation.
Protocols for cardiac organoid differentiation from embryoid bodies
Methods for reassembling cardiac constituent cells into cardiac organoids within a 3D environment are presented in Table 2 [27–33]. Richards et al. [27] engineered organoids replicating the lumenized vascular network during developmental stages by incorporating adult stem cells to promote endothelial cell function, combined with existing myocardial cells, fibroblasts, and endothelial cells. Mills et al. [28] developed a 96-well plate with elastic pillars promoting functional enhancement by providing physical resistance during contraction, resulting in cardiac organoid formation around the pillars. They screened for culture conditions that increased maturity and elucidated the associated mechanisms. Li et al. [29] developed a human ventricle-like cardiac organoid chamber with a pumping chamber structure using a collagen-based hydrogel mold. Li et al. [30] developed a flexible mesh-shaped nanoelectrode array that, when seeded with a mixture of cardiac constituent cells, self-organized into 3D cardiac organoids. The bio-sensor that was distributed uniformly within the organoid enabled the analysis of electrophysiological properties in multiple organoids. In the study of Song et al. [31], cardiac midline-derived cells differentiated into spheroid-shape cardiac organoids through 2D differentiation, exhibiting more mature characteristics than organoids based on differentiated myocardial cells with involvement of the LEFTY-PITX2 signaling pathway. Meier et al. [25] developed self-organizing epicardioids with ventricular myocardium and epicardium using collagen I gel and retinoic acid signaling activation. The trends in cardiac organoid generation methods are gradually shifting towards techniques facilitating easier reconstitution of a consistent cell composition, particularly through lineage differentiation using embryoid bodies [21].
Protocols for cardiac organoid generation from reassembling cardiac constituent cells
Trends in quality assessment methods for cardiac organoids
Quality assessments of cardiac organoids generated through various protocols are deemed imperative [23]. The methodologies proposed for evaluating the quality of cardiac organoids closely resemble those utilized for cardiomyocyte quality assessment. Established approaches include molecular analysis primarily based on genetics and structural characteristics, as well as functional analysis utilizing imaging techniques, as presented in Table 3 [34]. Table 3 summarizes the various methods used to assess the quality and characteristics of cardiac organoids, alongside their observed purposes and corresponding references [14–17,26–28,32,35–47].
Methods and analysis for quality assessment
Immunofluorescence techniques are employed to identify cell type markers, sarcomere arrangement, cardiac structure, interactions between cardiomyocytes and fibroblasts, and extracellular matrix composition. Polymerase chain reaction (PCR) and quantitative reverse transcription PCR (qRT-PCR) are used to evaluate cell type markers, function-related genes, mitochondrial biogenesis, and structure-related genes. Patch clamp and multi-electrode Array (MEA) assays are utilized to assess electrophysiological properties. Voltage-sensitive dye imaging and calcium imaging further contribute to understanding these properties. Optical coherence tomography (OCT) is applied for detailed structural and mitochondrial biogenesis analysis. Transmission electron microscopy (TEM) is used for structural analysis, including sarcomere length measurement. Hematoxylin-eosin staining assesses wall thickness, and image (video) analysis examines contractile properties. The seahorse mitochondrial stress test measures the oxygen consumption rate, providing insights into mitochondrial function. The terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) assay is used to detect apoptosis, while microscopy examines morphological characteristics. RNA in situ hybridization identifies cardiac lineage genes, and single-cell RNA sequencing (scRNA-seq) is used for comprehensive gene expression assays. Various advanced data analysis methods such as principal component analysis (PCA), gene set enrichment analysis (GSEA), differential gene expression analysis, gene set variation analysis, k-mean cluster analysis, and transcriptome analysis are conducted for in-depth data interpretation and validation of gene expression profiles. Proteomics is employed to study protein expression related to the adult heart.
This comprehensive overview highlights the range of techniques applied to assess the structural, functional, and molecular attributes of cardiac organoids, thereby providing a detailed framework for evaluating their quality and potential for research and clinical applications.
Recent advancements have introduced techniques that incorporate visual analysis and algorithmic assessments. Devarasetty et al. [48] developed software for visually analyzing video files of beating cardiac organoids and used it to analyze responsiveness to 5 different drugs affecting cardiac contractions. Sala et al. [49] developed an analysis tool based on imaging to measure the contraction and relaxation of muscular structures, nabling comparisons between single and multilayer cardiac cells, cardiac organoids, engineered cardiac tissues, and adult cardiomyocyte models. The tool’s applicability was confirmed in scenarios using various analysis tools, such as patch clamps. Feng et al. [35] produced chambered cardiac organoids and used RNA-seq analysis to identify the specific cellular composition of the fabricated organoids. Additionally, they developed an scRNA-seq data analysis method based on the random forest classification algorithm using machine learning. Kim et al. [50] developed low mechanical modulus and low impedance electrodes for electrophysiological measurements of 3D organoids, introducing a wet cardiac organoid analysis tool.
While these advanced techniques offer extensive information, they also come with notable limitations and challenges. Immunofluorescence is effective in identifying cell markers and structures, but is hindered by staining variability. PCR/qRT-PCR is useful for evaluating gene function, but requires high-quality RNA, which can be difficult to obtain. Patch clamp and MEA are valuable for measuring electrophysiological properties, but necessitate significant technical expertise and are susceptible to noise interference. TEM and OCT provide detailed structural analysis but may introduce artifacts during sample preparation. Seahorse assays are useful for assessing mitochondrial function, but are highly sensitive to environmental conditions. TUNEL assays are utilized to detect apoptosis but carry the risk of false positives. Lastly, advanced methods such as scRNA-seq, PCA, and GSEA yield comprehensive data, but are computationally intensive, costly, and require specialized expertise.
Additionally, as 3D structures, cardiac organoids pose challenges in traditional experimental approaches. These include difficulties in the adhesion of 3D organoids on 2D MEA plates, measurements of 3D structural characteristics using imaging techniques (e.g., TEM, video analysis), and reflection of the core of the organoid due to reagent permeability constraints (e.g., TUNEL assay, calcium imaging) [36,50–52]. To overcome these challenges, it is essential to develop analysis methods using computer algorithms or specialized analytical techniques tailored for 3D structures [53,54]. Furthermore, establishing quality assessment criteria that accurately reflect the structural characteristics of cardiac organoids is crucial for advancing their effective application.
Conclusion
In conclusion, the production of hiPSC-derived cardiac organoids exhibits variability depending on factors such as cell type, differentiation media conditions, and more [34,55]. Two primary methods, involving the formation of embryoid bodies and the reassembling of 2D heart constituent cells, are employed based on the initial differentiation approach [21,56,57]. The embryoid body formation method offers the advantage of being composed of cells derived from the same origin [21]. However, the reassembly method provides high reproducibility of relative control cell ratios but has a weaker physical structure for heart development [21]. Consequently, quality assessment is essential for the effective utilization of generated cardiac organoids in preclinical evaluations. Various experimental methods are used to assess the quality of produced cardiac organoids, and more suitable experimental methods for organoid quality are being developed. With more accurate and precise quality assessments, cardiac organoids can exhibit structural and physiological reactivity similar to the human heart. These organoids can be used not only as preclinical cardiotoxicity models, but also as models for cardiac development research and patient-specific treatments.
Notes
Conflict of interest
No potential conflict of interest relevant to this article was reported.
Funding
This work was supported by the Technology Innovation Program (#20009774) funded by the Ministry of Trade, Industry, and Energy (MOTIE, Korea) and by grants from the National Research Foundation of Korea (NRF-2022M3A9H1015784, RS-2023-00225239).
Data availability
Please contact the corresponding author for any inquiries related to data availability.