The research, led by a team using a heterotopic heart transplantation model in mice, provides some of the most direct evidence yet that mechanical forces are a central regulator of cardiac regenerative capacity. In the heterotopic model, a donor heart is surgically connected to the recipient’s circulation in parallel with its own heart. The graft becomes vascularized and receives blood, but because it beats in parallel rather than in series with the native heart, it performs no pumping work. The result is a heart that is alive and perfused but mechanically unloaded — a preparation that has existed for decades as a surgical curiosity but has now been repurposed as a powerful tool for probing mechanotransduction in cardiac regeneration.

Using this system, the investigators tracked cell-cycle re-entry in adult cardiomyocytes with two complementary approaches. First, they employed Ki67-based assays, detecting the expression of Ki67, a nuclear protein present in all active phases of the cell cycle but absent from resting cells, to document that cardiomyocytes in unloaded hearts re-entered the cell cycle at rates far exceeding those observed in normally loaded hearts. Second, and more conclusively, they used MADM — mosaic analysis with double markers — a genetic lineage-tracing technique that allows the daughter cells of a single dividing progenitor to be labeled in distinct, distinguishable colors. Because MADM unambiguously distinguishes true cell division from nuclear doubling or DNA synthesis without cytokinesis, a frequent source of false positives in cardiomyocyte proliferation studies, the lineage-tracing data established that unloaded adult hearts generate new cardiomyocytes through genuine cell division, not merely through cell-cycle activity that stalls before completion.

The team then asked whether this load-dependent regenerative potential could be exploited therapeutically. Applying the same heterotopic transplantation approach to hearts that had suffered myocardial infarction, they found that mechanical unloading promoted regeneration specifically within peri-infarct regions — the border zones where viable, damaged myocardium surrounds the dead scar. This finding is clinically evocative: left ventricular assist devices (LVADs), which unload the failing heart in patients awaiting transplantation or as destination therapy, are known to produce reverse remodeling, and reverse remodeling has been associated with increases in cardiomyocyte cell-cycle markers in human explant tissue. The new work provides a mechanistic framework for understanding what those devices might actually be doing at the cellular level.

To identify the molecular signals linking mechanical load to proliferation, the researchers performed single-nucleus RNA sequencing on unloaded hearts, profiling gene expression in individual nuclei to reconstruct the tissue’s cellular ecosystem. The analysis revealed a striking enhancement of communication between the epicardium — the outer epithelial layer of the heart — and cardiomyocytes. Rather than an intrinsic shift within cardiomyocytes themselves, unloading appeared to reprogram the epicardium, which ramped up production of neuregulin 1 (NRG1), a secreted growth factor long known to promote cardiomyocyte proliferation during embryonic development and in neonatal life.

The pathway that carried the signal was equally revealing. NRG1 acts on cardiomyocytes through its receptor ERBB4, a receptor tyrosine kinase whose downstream signaling had previously been implicated in developmental cardiac growth. In the unloaded heart, single-cell analysis and functional experiments traced a signaling route from epicardial NRG1 through cardiomyocyte ERBB4 to the transcription factor STAT3, which became activated and translocated to drive a proliferation program. The causal chain was tested with genetic precision: when the investigators deleted Nrg1 specifically in the epicardium, STAT3 activation in cardiomyocytes was lost, and cardiomyocyte proliferation in the unloaded hearts was abolished. The experiment demonstrates that the epicardium is not a passive covering but an active mechanosensory relay that translates the absence of load into a proliferative instruction.

The final piece of the mechanism addresses a question that has dogged adult cardiomyocyte proliferation research: why do adult cardiomyocytes that enter the cell cycle so often fail to complete division? The answer, according to the new study, lies in metabolism. Using CUT&Tag, a chromatin profiling technique that maps where a transcription factor binds across the genome and what it does to local histone marks, the team showed that STAT3 directly upregulates the gene H6pd, which encodes hexose-6-phosphate dehydrogenase, an enzyme operating within the endoplasmic reticulum that channels glucose-6-phosphate flux through the oxidative branch of the pentose phosphate pathway. Activation of this metabolic arm boosts the production of nucleotide precursors — the raw material for DNA replication — and of reducing equivalents in the form of NADPH, which buffer the oxidative stress that accompanies proliferative metabolism. In other words, mechanical unloading does not merely flip a mitogenic switch; it simultaneously opens a metabolic supply line that allows the cell-cycle program to run to completion.

Taken together, the findings delineate what the authors describe as a mechanotransductive pathway: mechanical load suppresses regenerative signaling, and mechanical unloading restores it, with the signal flowing from the epicardium through NRG1 to ERBB4 and STAT3 in cardiomyocytes, and onward into metabolic reprogramming that sustains cell division. The pathway’s architecture explains a biological paradox — why neonatal mammals can regenerate their hearts while adults cannot — in terms of workload rather than an irreversible loss of potential. At birth, the transition from placental to pulmonary circulation imposes the full hemodynamic burden of systemic pumping on the heart, and cardiomyocytes respond to that load by withdrawing permanently from the cycle. Remove the load, the new data suggest, and the withdrawal can be partially reversed, at least in mice.

The translational implications extend across several domains of cardiovascular medicine. For patients on LVAD support, the study suggests that unloading therapy may be doing more than resting and remodeling the failing heart; it may be actively stimulating endogenous repair through epicardial NRG1 signaling, and the NRG1–ERBB4 axis offers a pharmacological target for amplifying that effect. Exogenous NRG1 has already been tested in clinical trials for heart failure, with mixed results, and the new mechanistic data suggest that timing, delivery, and the concurrent mechanical state of the heart may determine whether neuregulin therapy succeeds. The metabolic arm of the pathway — the STAT3–H6pd link to the pentose phosphate pathway — similarly suggests that proliferative failure in the adult heart may be a bioenergetic problem that could be addressed directly. The study also reframes surgical and device-based strategies: heterotopic approaches, partial unloading, or even temporally staged unloading could, in principle, be designed to open a regenerative window in injured myocardium, allowing peri-infarct tissue to rebuild before the heart resumes full load.

Important caveats remain. The work was performed in mice, and the heterotopic transplant model is an extreme form of unloading that cannot be directly replicated in patients; the degree, duration, and rhythm of unloading that a human heart would tolerate while retaining regenerative benefit are unknown. The magnitude of cardiomyocyte renewal observed, while functionally meaningful in the peri-infarct setting of the mouse, would need to be substantially amplified to rebuild a large human infarct. And the epicardium, thin and largely quiescent in the adult human heart, may respond differently to unloading than the mouse epicardium does. Nevertheless, the study establishes a clear causal pathway from mechanics to metabolism in cardiac regeneration, and it elevates mechanical load from a passive biomechanical variable to a master regulator of cardiomyocyte cell fate. For a field that has spent decades searching for the molecular key to heart regeneration, the message is unexpectedly physical: one of the most powerful signals for renewal may be rest itself.

Heart attacks destroy cardiomyocytes and disrupt the vascular network that supplies oxygen to the heart muscle. Because adult human cardiomyocytes have limited regenerative capacity, researchers have investigated whether cells produced from human pluripotent stem cells could restore damaged tissue or stimulate the heart’s own repair mechanisms. Cardiovascular progenitors are particularly attractive for this purpose because they can develop toward several heart-related lineages, including muscle, vascular and supportive stromal cells. Yet the behavior of transplanted cells after they are delivered into a living heart remains difficult to observe. Conventional molecular analyses often require tissue to be removed and homogenized, erasing the precise locations of cells and the interactions taking place between the graft and the host.

The team addressed this challenge by applying spatial transcriptomics at multiple time points after transplantation into a pig model of myocardial infarction. Spatial transcriptomics combines gene-expression profiling with positional information, allowing researchers to determine not only which genes are active but also where those genes are being expressed within a tissue section. In this study, the approach enabled the investigators to distinguish human xenograft signals from the surrounding porcine myocardium and to examine how both compartments changed as healing progressed. Instead of producing a single molecular snapshot, the time-series design provided a dynamic view of engraftment, revealing how the transplanted progenitors adapted to the hostile, oxygen-poor and fibrotic environment of an infarcted heart.

The analysis showed that the human cardiovascular progenitors gradually increased expression of genes associated with a more mature cardiac state. These included programs involved in oxidative metabolism, the process by which cells generate energy efficiently through mitochondrial respiration. The grafted cells also upregulated genes linked to calcium handling, a central function in cardiac contraction because calcium ions regulate the interaction between contractile proteins inside muscle cells. Such changes suggest that the transplanted progenitors were not remaining in an undifferentiated state. Instead, they appeared to be responding to the cardiac environment and moving toward a phenotype better suited to functioning within heart tissue.

The molecular data also indicated a shift in pathways related to fibrosis. Following myocardial infarction, excessive deposition of extracellular matrix proteins can stiffen the ventricular wall and interfere with electrical conduction and contraction. Although scar formation initially helps prevent the injured heart from rupturing, persistent fibrosis can contribute to long-term heart failure. The observed activation of gene programs associated with fibrosis resolution suggests that the grafted cells may influence the remodeling of the scar environment, either directly or through signals exchanged with host cells. The findings do not establish that the transplanted progenitors alone remove scar tissue, but they provide evidence that their presence is associated with a tissue environment moving toward repair rather than continued damage.

To identify the signals responsible for communication between the human graft and the pig heart, the researchers performed cell–cell communication analysis. This computational strategy examines ligand–receptor relationships: molecules released by one cell type are matched with receptors expressed by another, allowing scientists to predict which cellular conversations may be biologically important. Among the candidate signals, Midkine emerged as a prominent regulator connected to host neovascularization. MDK is a secreted growth factor involved in cell survival, migration and tissue remodeling. In the context of an infarcted heart, its production by the human progenitors suggested a mechanism through which transplanted cells could act at a distance, stimulating neighboring host cells rather than needing to become large numbers of new cardiomyocytes themselves.

The investigators then tested the MDK hypothesis experimentally. Immunohistochemistry was used to examine protein-level patterns in the tissue, providing an independent validation of the transcriptomic findings. They also used lentiviral gene delivery to increase MDK expression in the relevant cells and conducted functional assays to determine whether the factor altered endothelial behavior. Endothelial cells form the inner lining of blood vessels, and their ability to migrate is a key early step in angiogenesis. The experiments demonstrated enhanced endothelial cell migration when MDK activity was increased. In transplanted hearts, MDK overexpression was associated with greater density of CD31-positive vessels, with CD31 serving as a commonly used marker of endothelial cells and vascular structures.

These results place neovascularization at the center of the graft’s therapeutic activity. Restoring blood flow is crucial after infarction because surviving cardiomyocytes at the edge of the injury remain vulnerable to oxygen deprivation. New or remodeled vessels can improve nutrient delivery, remove metabolic waste and create conditions that support tissue stabilization. The findings suggest that human cardiovascular progenitors may function as biological signal generators, releasing factors such as MDK that recruit or activate the host’s own endothelial cells. This paracrine mechanism may be as important as, or potentially more important than, the direct replacement of lost heart muscle. It also creates a possible therapeutic strategy in which MDK or related pathways are targeted to promote vascular repair.

The study’s use of a pig model is significant because porcine hearts resemble human hearts in size, anatomy and aspects of cardiac physiology more closely than many small-animal systems. At the same time, the work remains preclinical. A higher density of CD31-positive structures indicates increased vascularization, but future studies will need to determine whether those vessels are fully functional, carry effective blood flow and improve measurable cardiac performance over the long term. Researchers will also need to assess the durability, safety and electrical behavior of the grafted cells, as well as the potential consequences of altering MDK activity in other organs. The immune response to human cells in a large-animal setting and the reproducibility of the treatment will be important considerations before clinical translation.

To make the extensive dataset accessible, the researchers have created a publicly available interactive Shiny atlas containing spatial and temporal transcriptomic information from pig hearts affected by myocardial infarction and treated with human xenografts. Interactive resources of this kind allow investigators to explore gene-expression patterns across tissue regions and recovery stages without relying solely on static figures. The atlas may help other groups compare candidate pathways, examine host–graft interactions and generate new hypotheses about cardiac repair. Together, the study and its accompanying resource offer a detailed molecular timeline of transplanted cardiovascular progenitors in an injured heart, while identifying MDK as a tractable target for therapeutic angiogenesis. The work shifts attention from asking only whether stem-cell grafts survive to understanding how they communicate with damaged tissue and recruit the body’s own repair machinery.

Subject of Research: Human pluripotent stem-cell-derived cardiovascular progenitors transplanted into infarcted pig hearts, with a focus on host–graft communication and MDK-mediated neovascularization.

Article Title: Spatiotemporal transcriptomics of human cardiovascular progenitors in pig hearts identifies Midkine as a positive regulator of neovascularization.

Article References: Adusumalli, S., Leong, K.S., Lim, S. et al. “Spatiotemporal transcriptomics of human cardiovascular progenitors in pig hearts identifies Midkine as a positive regulator of neovascularization.” Nature Cardiovascular Research 5, 744–762 (2026). https://doi.org/10.1038/s44161-026-00851-1

Image Credits: AI Generated

DOI: August 2026

Keywords: stem cell therapy, cardiac repair, myocardial infarction, spatial transcriptomics, cardiovascular progenitors, Midkine, MDK, neovascularization, angiogenesis, endothelial cells, pig model, regenerative medicine