The study, published in Annals of Biomedical Engineering, focused on mitral valve interstitial cells, or MVICs, the resident cells embedded within the collagen-rich extracellular matrix that gives the valve leaflet its remarkable combination of strength and flexibility. In healthy tissue, these cells quietly maintain the matrix, synthesizing and degrading structural proteins as needed to keep the leaflet functioning across billions of cardiac cycles. In disease, however, MVICs can adopt an activated phenotype, characterized by elevated expression of the contractile protein alpha-smooth muscle actin, along with excess collagen deposition, matrix fragmentation, and increased activity of matrix metalloproteinases, the enzymes that break down structural proteins. The result is a valve leaflet that thickens and stiffens, further distorting the delicate mechanics that keep it closing properly. What remained unknown until now is whether these altered cellular behaviors are intrinsic to the cells themselves, or merely a response to an abnormal tissue environment—and precisely how they change the mechanical landscape around each cell.

To answer that question, the researchers, led by Toni M. West and Micheal S. Sacks, isolated MVICs from three patients undergoing first-time surgery for mitral valve prolapse and from three donors whose valves were anatomically normal. Rather than studying the cells in flat culture dishes or in opaque excised tissue—both of which obscure the true three-dimensional mechanical dialogue between a cell and its surroundings—the team embedded the isolated cells in engineered hydrogels designed to mimic key aspects of the native cellular microenvironment. The gels were built from eight-arm poly(ethylene) glycol functionalized with norbornene groups, crosslinked with peptide sequences that cells can enzymatically degrade, and decorated with CRGDS adhesion peptides that allow cells to grip their surroundings. Suspended within each gel were fluorescent microspheres roughly one micrometer in diameter, serving as fiducial markers whose movements would betray every tug and push exerted by the encapsulated cells.

After three days of culture, each hydrogel was imaged twice on a laser scanning confocal microscope: once in the basal state, with the cells actively contracting, and once in a fully relaxed state induced by cytochalasin-D, a drug that dismantles the actin cytoskeleton. By subtracting the relaxed configuration from the contracted one, the researchers obtained precise three-dimensional displacement fields for thousands of microspheres surrounding each cell, captured at an isotropic voxel resolution of just 0.288 micrometers. The key innovation of the study, however, lay not in the imaging but in the computational machinery applied to those images. Conventional traction force microscopy, a technique dating back to 1999, assumes that the hydrogel surrounding a cell remains mechanically uniform, an assumption that is demonstrably false when cells are actively degrading and rebuilding their matrix. The new study deployed a recently developed framework that dispenses with that assumption entirely.

In the inverse modeling approach, the hydrogel was treated as a compressible Neo-Hookean hyperelastic solid, an appropriate description given the significant compressibility of poly(ethylene) glycol gels, which must be accounted for to recover accurate mechanical properties. The local stiffness of the gel was encoded as a spatially varying material parameter expressed in exponential form, allowing the model to capture changes in modulus spanning large orders of magnitude while remaining numerically stable. The team then solved a nonlinear optimization problem: using the limited-memory Broyden–Fletcher–Goldfarb–Shannon algorithm, they iteratively adjusted the simulated stiffness field until the computed displacements matched the experimentally measured ones throughout a reliable measurement zone the authors call the “event horizon,” defined by a displacement cutoff of 0.38 micrometers corresponding to three standard deviations of imaging noise. Gradient computations performed with adjoint methods and Tikhonov regularization kept the ill-posed inverse problem stable, and convergence was reached in roughly an hour per hydrogel on a 24-core desktop workstation. The output was a full three-dimensional map of how each cell had locally stiffened or softened its surroundings, from which surface traction forces and local strain energy densities could be computed using the first Piola–Kirchhoff stress tensor.

The results delivered a surprise on multiple fronts. Collagen staining confirmed that cells deposited collagen in their immediate vicinity, producing pronounced local stiffening close to the cell membrane, while regions farther from the cell surface actually softened, reflecting enzymatic degradation of the gel’s degradable crosslinkers. Cells from prolapsed valves stiffened their gels more and degraded them less than cells from normal valves, consistent with the known tendency of diseased valve cells to overproduce collagen. But the most striking finding concerned contractility. The prolapsed-valve cells exhibited larger basal contractile displacements, squeezing their surroundings more deeply, yet the actual traction forces they generated and the strain energy they injected into the gel were significantly lower than those of normal cells. In other words, the diseased cells appeared to be working harder visually while doing less mechanical work—an apparent paradox that only became visible because the model accounted for the cells’ remodeling of their own microenvironment.

The interpretation is subtle but consequential. Because the cells from prolapsed valves had remodeled their surroundings so extensively, depositing stiffer matrix nearby, they could achieve large displacements with comparatively modest force. A conventional analysis assuming a uniform gel would have misread this picture entirely, potentially overestimating the force output of the diseased cells. By capturing the true, heterogeneous mechanical state of the gel, the new framework revealed that prolapsed-valve interstitial cells are not simply hypercontractile muscle-like cells, as their elevated alpha-smooth muscle actin expression might suggest, but cells with a fundamentally rebalanced mechanical portfolio: more matrix deposition, less degradation, deeper but weaker contractions.

Crucially, because the experiments were performed on isolated cells in standardized gels, the observed differences cannot be attributed to the altered biochemical environment of a diseased valve. The behaviors are intrinsic to the cell phenotype itself, apparently persisting even after the cells were removed from the patient’s tissue and cultured at low passage on collagen-coated dishes designed to prevent artificial activation. This intrinsic quality carries a sobering implication for surgeons and patients alike: repairing the geometry of a prolapsed valve does not necessarily correct the underlying cellular behavior that drove the disease, and those intrinsic behaviors may continue remodeling the repaired tissue in ways that promote recurrent regurgitation and secondary surgery.

The study’s authors suggest that this new mechanobiological window could reshape how post-repair remodeling is understood and, eventually, treated. If the contractile and matrix-remodeling behaviors of valve interstitial cells can be measured, perhaps from cells harvested during surgery, clinicians might one day stratify patients by risk of recurrent disease, identifying whose valves are likely to remodel adversely after repair. More ambitiously, drugs that modulate the contractility or matrix-synthesis programs of these cells could be developed as adjuncts to surgery or even as alternatives, targeting the cellular engines of the disease rather than its structural consequences. Such therapies have remained elusive precisely because the cellular mechanisms of post-repair remodeling were so poorly characterized; this work supplies the first direct, quantitative evidence that prolapsed-valve cells behave mechanically differently from normal ones in ways that could be pharmacologically targeted.

The technical achievement also extends beyond the mitral valve. The inverse modeling framework, which simultaneously recovers cellular tractions and cell-induced changes in local material properties for compressible, degradable hydrogels, offers a template for studying any cell type whose remodeling behavior complicates conventional traction measurements, including cancer cells, fibroblasts in fibrotic disease, and cells in engineered tissues. For a condition that has historically been managed exclusively with scalpels and sutures, the demonstration that its cellular drivers can be measured, quantified, and distinguished from healthy behavior marks a meaningful step toward understanding—and perhaps one day preventing—the silent remodeling that undermines valve repair. As the researchers conclude, the behaviors of mitral valve interstitial cells may determine how prolapsed valves respond to repair, and for the first time, those behaviors can now be seen with clarity.

The study, led by Guang Li, associate professor in the University of Pittsburgh School of Medicine’s Department of Cell Biology, represents the first reported effort to grow heart valve-like structures on this type of human heart assembloid. The findings, published in Cell Stem Cell, could help researchers investigate congenital and acquired valve disorders without relying exclusively on animal models. The platform may also eventually provide a way to test potential treatments using patient-specific human cells.

Heart valves are highly specialized structures that open and close thousands of times each day, directing blood through the chambers of the heart and preventing it from flowing backward. Their formation depends not only on the correct genetic instructions, but also on mechanical forces generated by blood flow, tissue movement and the contraction of nearby cardiac muscle. These interacting signals are difficult to reproduce in conventional cell cultures, where cells are often grown on flat surfaces under relatively static conditions.

Animal models have provided valuable information about valve development, but they do not perfectly replicate human biology. Heart valves in animals can develop at different speeds and may respond differently to genetic mutations, physical stress or metabolic injury. The researchers therefore sought to build a human model in which valve-like tissue could develop while exposed to carefully controlled physical and biological signals. “Human valves are very different from animal valves,” Li said. “To study human valve diseases, we need human valve models.”

The team began with human induced pluripotent stem cells, or iPSCs. These cells can be generated from adult tissues such as blood or skin and reprogrammed into a flexible state in which they can produce many different cell types. By directing the cells through specific developmental signals, the researchers generated cardiac tissues and combined distinct organoids into an assembloid. This approach is designed to capture the way organs form from multiple interacting tissues rather than from a single uniform cell population.

The resulting heart assembloid incorporated two organoid components made from different types of heart cells. The researchers then encouraged valve-like tissue to form on its surface by adding several features that mimic the environment of a developing heart. A flowing culture medium reproduced aspects of blood movement, while an endothelial cell layer modeled the specialized cells that line the interior of blood vessels and heart valves. The system also included magnetized beads positioned around the tissue and moved by an external magnetic belt. Their motion generated mechanical stimulation intended to imitate the contractions of cardiac muscle.

These physical cues were central to the experiment. In the developing heart, mechanical forces influence how cells organize, change shape and adopt specialized identities. Fluid movement can alter the behavior of endothelial cells, while repeated tissue deformation can affect the maturation and strength of valve-like structures. By integrating these forces into the assembloid, the researchers created a model that was more dynamic than a conventional organoid and closer to the environment in which human valves develop.

After establishing the model, the researchers used it to investigate several forms of valve pathology. One focus was mitral valve prolapse, or MVP, a disorder in which the mitral valve does not close normally and can bulge backward when the heart contracts. MVP is associated with genetic factors and affects an estimated 7 to 8 million people in the United States. When the researchers introduced a mutation linked to the disorder, the developing valve-like structures displayed features consistent with MVP, suggesting that the assembloid could connect a specific genetic change to a measurable developmental defect.

The researchers also used the platform to model acquired valve injuries. These included processes associated with valve calcification, cryo-injury and metabolic complications related to hypoglycemia and diabetes. Such conditions can damage valve tissue over time and interfere with its flexibility and function. Because the assembloids were built from human cells, the team could examine how these stresses altered cellular pathways and tissue development in a controlled setting. The experiments identified biological signaling routes involved in the abnormal development associated with MVP and pointed toward mechanisms that might be corrected therapeutically.

The current system is still a simplified representation of a human heart. The valve-like structures developed on the surface of the assembloid rather than inside a chamber, and the model does not yet reproduce the full circulation, pressure and electrical coordination of a mature organ. Li’s next goal is to increase its complexity by creating assembloids with two chambers and growing valves within them. A more integrated model could provide a closer simulation of how valves form and function inside the heart, while offering researchers a powerful tool for studying disease mechanisms, evaluating drugs and exploring regenerative therapies.

Subject of Research: Lab-produced tissue samples

Article Title: Human iPSC-derived heart valve-like assembloids model valve development and disease pathology

News Publication Date: 11-Aug-2026

Web References: https://www.cell.com/cell-stem-cell/fulltext/S1934-5909(26)00271-7

References: DOI: 10.1016/j.stem.2026.07.011

Image Credits: Yuanhang He/University of Pittsburgh

Keywords

Organoids, assembloids, heart valves, mitral valve prolapse, induced pluripotent stem cells, tissue engineering, biomedical engineering, cardiac function, heart disease, human disease models