Listen to Your Heart and Engage Your Immunity: A Human Multi‐Organoid Platform for Investigating Heart and Immune System Interactions

Heart disease is a leading cause of death worldwide. It is far more complex than a problem confined to a single organ and often involves dynamic interactions among multiple organ systems. Data published in The Lancet show that cardiovascular diseases claim millions of lives annually and mostly involve multi-organ dysfunction, which calls for multidisciplinary action [1]. Traditional models primarily rely on animal experiments or simplified in vitro systems using isolated cells or tissues, but these approaches present numerous limitations. First, they cannot accurately replicate the electrophysiological characteristics of the human heart due to species differences. Second, the absence of a three-dimensional structure and intercellular interactions prevents them from simulating real heart function. These limitations reduce the authenticity and predictive value of disease models and drug testing. Cardiac organoids represent a breakthrough [2]. They are three-dimensional structures derived from stem cells—including embryonic stem cells, induced pluripotent stem cells (iPSCs), or adult stem cells—that can mimic certain heart functions and characteristics in the body, such as spontaneous beating and electrophysiological properties. However, isolated cardiac organoids often overlook the communication between the heart and other tissues, making it difficult to accurately capture human-specific interorgan interactions. Currently, multi-organ models have made significant advances [3]. These include cardiopulmonary organoids, cardio-renal units, cardiac-intestinal components, and neuro-cardiac chip models [4-7]. Circulating immune cells are crucial for cardiac repair and remodeling during heart injury. A recent paper titled “Human Multi-Organoid Platform to Model Immune Dynamics in Cardiac Injury and Disease,” published in Circulation Research, introduces an innovative multi-organoid system that advances the modeling of immune and cardiac interactions [8]. This platform opens new avenues for understanding the interactions between the heart and other organs, including bone marrow and the immune system. The study developed a fluidic connection device based on 3D printing technology. The device integrates vascularized cardiac organoids (VCOs) with bone marrow organoids (comBOs) within a standard six-well culture plate. This setup enables cellular and molecular exchange solely through the rocking motion of the incubator. It operates without external pumps and mimics key aspects of in vivo circulatory dynamics. The platform uses human iPSC-derived models. The comBOs system generates various innate and adaptive immune cells, while the VCOs simulate vascularized cardiac tissue. In the experiments, ischemia was induced in VCOs by exposing them to 1% O2 for 24 h, followed by co-culturing with comBOs for 5 days. Analysis by flow cytometry, immunofluorescence, and single-nucleus RNA sequencing (snRNA-Seq) revealed the temporal dynamics of the immune response. Notably, on Day 2 after injury, there was marked recruitment of CD45+ immune cells accompanied by fibrosis, as indicated by increased αSMA and collagen deposition, as well as upregulation of inflammatory genes (e.g., HIF1A, TGFB1, IL1β). By Day 5, these acute responses had subsided, but remodeling markers (e.g., IL18, MMP2) emerged, indicating ongoing tissue remodeling. These findings demonstrate that the platform successfully captured the dynamic cellular and molecular crosstalk from acute injury through to resolution. Additionally, snRNA-Seq identified subpopulations of cardiomyocytes, fibroblasts, and immune cells. CellChat analysis showed enhanced intercellular interactions after injury on Day 2, while they gradually declined by Day 5, although some dysregulation of signaling pathways persisted, consistent with the pathological features of human ischemic heart disease. This paper presents an in vitro dynamic platform of human-specific immune crosstalk by generating the first engineered human iPSC-derived cardiac-bone marrow co-culture system. Fluid exchange is achieved through simple shaking to simulate cyclic dynamics, avoiding expensive external equipment and enhancing the platform's accessibility and scalability. This work powerfully captures key pathological processes underlying ischemic injury. Experimental data show that CD45+ immune cells increase on Day 2 after injury (flow cytometry results) and return to baseline by Day 5. Immunofluorescence confirms the infiltration of CD14+ monocytes and CD68+ macrophages, along with a simultaneous increase in fibrosis markers. More importantly, the platform also reveals dynamic molecular events. SnRNA-Seq analysis reveals fibroblast activation, indicated by increased PDGFRB and POSTN expression, and a reduction of vascular smooth muscle cells after injury, while differential gene expression analysis further shows the up-regulation and down-regulation of inflammatory genes (e.g., NFKB1, CXCL8) and fibrosis-related genes (e.g., COL1A1, MMP9), consistent with an acute injury resolution model. Gene set enrichment analysis further confirms the dynamic changes in hypoxia response and inflammatory pathways, reinforcing the reliability of the platform. These findings not only validate the platform's effectiveness in modeling heart-bone marrow crosstalk but also highlight its potential as a scalable and accessible alternative to animal models. Furthermore, qRT-PCR experiments demonstrate that in the absence of immune cell co-culture, the inflammatory response is reduced, emphasizing the central role of crosstalk in disease mechanisms. Although the platform demonstrated the feasibility of using a combined model of VCOs and comBOs to enable human-specific studies of immune-driven cardiac injury, repair, and remodeling, several aspects require further investigation. In particular, chronic remodeling and the associated long-term cellular and molecular crosstalk involved need to be better understood. The data show that fibroinflammatory remodeling markers (e.g., IL18 and SLC8A) persist beyond Day 5; however, the extended dynamics spanning several months remain unexamined. This is crucial for heart failure management, as about 40% of patients experience progressive remodeling. Furthermore, personalizing therapies and addressing disease heterogeneity continue to be challenging tasks due to the complex and variable nature of immune responses among patients. Although human iPSC models are human-specific, they may not fully reflect genetic diversity, especially in immune-related diseases like myocarditis, which hinders the development of personalized therapies. Therefore, future research should expand the platform by incorporating more organ types and AI-driven analysis to develop a comprehensive “human-on-a-chip” system that advances precision medicine and drug testing [9]. As the authors highlight, this approach can be easily adapted to include emerging organoid models, which expands coverage of immune-relevant tissues involved in cardiovascular pathology. Future studies should aim to integrate essential regulatory systems, such as the nervous and endocrine hormones, into the cardiac–bone marrow organoid platform, enabling the reconstruction of more physiologically relevant multi-organ systems. In parallel, there is a critical need to develop standardized evaluation protocols to accelerate clinical translation and investigate novel therapeutic approaches that specifically target interactions between organs involved in cardiovascular disease. Additionally, combining multi-organoid models with organ-on-a-chip technology promises to elucidate complex disease mechanisms. This advancement contributes to more effective prevention and treatment of cardiovascular diseases and offers significant scientific value and clinical potential. Baimei Zhuang: writing – original draft, writing – review and editing. The author has nothing to report. The author has nothing to report. The author has nothing to report. The author declares no conflicts of interest. Data sharing is not applicable to this article as no data sets were generated or analyzed during the current study.