Biomechanical regulation of stem cells: Implications for cell therapeutics

Summary

Stem cell biology has rapidly expanded into an interdisciplinary field with potential for next-generation cell-based therapies. However, a critical gap remains in our understanding of how physical forces influence stem cell behavior. Recent studies in mechanotransduction, the process by which cells sense and convert mechanical cues into biochemical signals, have revealed that biomechanical regulation is fundamental for stem cell fate, proliferation, and therapeutic efficacy. This review synthesizes recent findings on the intrinsic and extrinsic parameters of physical cues that govern mechanotransduction and shape stem cell biology, highlights the mechanosensitive responses, and explores how biomedical engineering (BME) can be employed to manipulate these processes to improve translational outcomes in clinical settings. By integrating insights from cell biology, mechanobiology, and engineering, this interdisciplinary field offers strategies to translate benchside discoveries into clinical applications, advancing the development of precise and effective stem cell-based therapies.

Therapeutics; Cell; Biomechanics

Introduction

Stem cell biology is a rapidly evolving interdisciplinary field with great potential for cell-based therapeutics and regenerative medicine. Stem cells are distinctive cellular populations characterized by their capacities for self-renewal and differentiation into multiple specialized cell lineages, making them uniquely suited for repairing and regenerating compromised tissues and organs.^1,2,3 Stem cells are usually categorized based on their potency: totipotent stem cells, pluripotent stem cells, and multipotent stem cells, each of which is equally relevant to therapeutic applications (Figure 1).

Figure 1.

Overview of stem cell hierarchy

The figure illustrates the differentiation potential of stem cells, including embryonic stem cells (ESCs), hematopoietic stem cells (HSCs), mesenchymal stem cells (MSCs), and induced pluripotent stem cells (iPSCs) generated through somatic cell reprogramming. Figure created with Biorender.com.

Most of the adult stem cells are multipotent stem cells, such as hematopoietic stem cells (HSCs) and mesenchymal stem cells (MSCs), which are able to differentiate into tissue-specific lineages. Thus, they have been extensively characterized and have been performing well clinically for treating diseases in specific body systems (such as hematological pathologies).^4,5,6 Embryonic stem cells (ESCs) and induced pluripotent stem cells (iPSCs), on the other hand, exhibit pluripotency, which can differentiate into cell types representative of all three embryonic germ layers, making them a prototypical model for understanding early developmental trajectories as well as fulfilling the needs for specific lineages.^7,8,9 This grants stem cells with great potential in diverse research areas, including regenerative medicine, developmental biology, and the establishment of disease-specific models, providing the platform for personalized patient-specific medicine.^10,11 In recent years, stem cell based therapies have emerged, relying on their regenerative capacities by directly using different types of stem cells to restore and regenerate impaired tissue structures and functions, representing a precise and targeted strategy.^{12,13,14,15,16,17}

A critical determinant of cell behavior is biomechanical regulation, which describes cellular responses elicited by physical and mechanical cues originating from the cellular microenvironment.^18,19 These biomechanical regulations can be either a result of intrinsic cellular properties or extrinsic environmental factors, both having a profound impact on stem cells. The fate of stem cells are intricately modulated by biomechanical parameters, which influence stem cell functions including proliferation, differentiation, and lineage selection. Thus, a comprehensive understanding of these biomechanical interactions is necessary for the optimization of stem cell-based therapies and the subsequent enhancement of their clinical efficacy.

In this literature review, we are going to critically synthesize and assess existing research on the biomechanical regulation of stem cell behaviors and therapeutic outcomes under multiple physiological and pathological settings. The review will thoroughly examine current empirical findings, interdisciplinary technological advancements, and clinical therapeutic implementations, while also exploring emerging trends and future research directions in stem cell biology. By unraveling the scientific mechanisms and clinical applications of biomechanical principles, this review aims to advance the translational capacity and clinical impact of stem cell therapies.

Biomechanical regulation of stem cells

Mechanotransduction is the mechanism by which physical forces such as stretch, pressure, or stiffness are converted into cellular responses.^20,21 When there is a defect in mechanotransduction, failure to generate a proper response and initiation of a cell signaling cascade can result in various pathological phenotypes.^22,23,24 Stem cells are highly susceptible to mechanotransduction because physical stimuli control their ability to differentiate into various cell types due to their reciprocal interaction with their surrounding microenvironment, including local niche and extracellular matrix (ECM).^25,26 The physical stimuli presented in the microenvironment can be generally categorized into two types: intrinsic parameters, which refer to the mechanical and structural properties of the cells themselves, and extrinsic parameters, which involve mechanical cues from the surrounding microenvironment, such as substrate stiffness, shear stress, or topographical features of the ECM. These signals are transduced into biochemical responses through a cascade of mechanosensitive signaling pathways, ultimately guiding cellular behaviors.

Intrinsic biomechanical parameters

The cytoskeletal system, mainly composed of actin filaments, intermediate filaments, and microtubules, makes up the basic cell structure, which not only defines the overall shape of the cell and provides mechanical stability, but also acts as a highly dynamic regulator of cellular function. By organizing the cytoskeleton and associated scaffolding proteins, they serve as mechanosensitive platforms that transmit physical cues into biochemical signals, modulating diverse signaling pathways essential for cell survival and communication.

Microfilaments, the most abundant cytoskeletal structure in most eukaryotic cells, are composed of actin, which are globular protein subunits that polymerize to form actin filaments. They are organized into distinct architectures via branched networks, crosslinked networks, and parallel bundles to regulate cellular motility and shape.²⁷ These assemblies form more complex actin structures, such as stress fibers and cortical actin, each of which has distinct mechanotransduction role.²⁸ Due to the ubiquitous distribution of microfilaments within stem cells, these signals are often related to stem cell fate decisions based on different architectures. For example, the contractile actomyosin-rich stress fibers produce intracellular tension through focal adhesions to upregulate focal adhesion kinase/extracellular signal-regulated kinase (FAK/ERK)²⁹ and Ras homolog family member A/Rho-associated protein kinase (RhoA/ROCK)³⁰ signaling; this simultaneously promotes the nuclear localization of Yes-associated protein/transcriptional coactivator with PDZ-binding motif (YAP/TAZ), in which YAP/TAZ function as transcriptional co-activators of osteogenic differentiation in bone marrow MSCs.^28,31,32,33 In contrast, increased cortical actin causes cytoplasmic localization of YAP/TAZ and alternative lineage specification, including pancreatic differentiation from pluripotent stem cells.³⁴ Other types of actin filament architectures are less directly linked with the YAP/TAZ pathway, but can exert mechanoregulation as well. The branched actin networks, which is generated through Arp2/3-mediated actin nucleation, will form lamellipodial structures central to cell motility, and in human ESCs, current findings suggesting it influences YAP activity and mechanosensing indirectly through effects on cell shape and cytoskeletal organization rather than direct transcriptional control.³⁵ Linear protrusive actin (or filopodia) are slender, linear actin-rich protrusions, which act as dynamic mechanosensors in many stem cell types (mesenchymal, neural, and pluripotent) that probe local cell-cell and cell-ECM contacts, supporting adhesion formation and guidance decisions.^36,37 Thus, mechanotransductive outcomes depend less on actin abundance per se and more on the spatial organization and contractile state of actin networks.

Microtubules also help regulate stem cell lineage differentiation based on the mechanical cues they receive, especially in multipotent cells. For the osteogenic differentiation of MSCs, studies have shown that MSCs cultured on 3D nanofibrous scaffolds display that microtubules are essential in modulating nuclear morphology and lineage commitment. In particular, the presence of an organized microtubule network facilitates flattened, elongated nuclear structure, increasing intracellular tension and profoundly promoting osteogenic gene expression and differentiation.^38,39 Conversely, disruptions to the microtubule cytoskeleton reduce nuclear area and aspect ratio, resulting in more rounded or irregular nuclear morphologies that impede osteogenesis progression.⁴⁰ A similar pattern of microtubule-mediated mechanical signaling is observed in HSCs, where they can control HSC differentiation into either myeloid or lymphoid progenitors,^38,41,42 and sustain proliferation for engraftment through pathways such as Wnt/β-catenin.^43,44,45 Microtubules also contribute to nuclear positioning and transduce mechanical cues that the nucleus receives, mostly influencing transcriptional activities.^46,47 Because of their participation during metaphase and anaphase in mitosis, specifically by forming mitotic spindle fibers and attaching to kinetochores to segregate sister chromatids, they can directly regulate cell division and proliferation rates of various types of stem cells.^38,48,49

Intermediate filaments are diverse rope-like proteins that can be classified into five types: keratins (types I and II), vimentins (type III), desmin (type III), lamins (type V), and nestin (Type VI). These intermediate filaments can be found in various stem cells and serve as sources of structural strength to promote cell function. Nestin is found in neural stem cells, while vimentin mostly in MSCs.^50,51 Vimentin, as a key intermediate filament protein, governs the form and function of stem cells and regulates cell fate and migration patterns.^52,53,54,55 Mechanistically, vimentin integrates cytoskeletal tension with focal adhesions and nucleo-cytoskeletal coupling to modulate nuclear mechanics,^56,57,58 which are linked to epigenetic regulation on mechanotransduction due to its impacts on OCT4-chromatin interactions.^53,59

Nuclear stiffness is another intrinsic parameter that can modulate mechanotransduction, which is largely governed by lamin A/C, a type V intermediate filament protein with a profound connection to the cell cycle.⁶⁰ Lamins in stem cells are directly linked to differentiation and cell lineage selection, and adjust to any mechanical stress they experience. Studies have found that the depletion of lamin A/C in MSCs reduces nuclear stiffness and impairs differentiation.⁶¹ High lamin A levels, often induced by stiffer matrices, promote osteogenic differentiation, while lower levels favor adipogenic or less mechanically demanding lineages.^61,62 Changes in lamin A/C expression, and subsequently changes in nuclear stiffness, are linked to gene regulation and chromatin remodeling. Depletion of lamin A/C results in softer nuclei, thus altering mechanotransduction and impacting differentiation potential and cell migration.^61,63 Moreover, spatial distribution of lamin A within the nuclear envelope also modulates nuclear deformability, influencing how stem cells sense and respond to mechanical cues from their environment.⁶⁴

As the cytoskeleton interacts with other intracellular compartments, such as myosin II, it will generate cytoskeletal tension forces to further help maintain cellular structure and mediate mechanotransduction. In stem cells, elevated cytoskeletal tension promotes differentiation into stiff-tissue lineages such as osteoblasts, while reduced tension favors adipogenic or chondrogenic pathways.^65,66,67 Lineage differentiation is tightly regulated by mechanosensitive pathways, including RhoA/ROCK and YAP/TAZ, which translate mechanical cues into gene expression changes.⁶⁸

Cell shape and mechanical integrity play a critical role in mediating the transduction of physical cues into intracellular biochemical signals. An alteration in cell stiffness or integrity frequently reflects a disorganization of cytoskeletal structure. In stem cells, geometric morphology has been identified as a key factor influencing lineage-specific differentiation.⁶⁹ A well-spread morphology of MSCs promotes actomyosin contractility and activates RhoA/ROCK signaling, which preferentially shifts toward osteogenic differentiation. In contrast, a rounded morphology with low cytoskeletal tension favors adipogenesis.⁷⁰ Defects can lead to a reduction in cell stiffness in HSCs, thereby reducing their adhesion to the bone marrow niche and consequently promoting HSC regression and hyperactivity, as well as impacting hematopoietic regeneration capacity.⁷¹ Beyond cell shape, flattened nuclear morphology and high mechanical tension are directly linked to a higher YAP nuclear/cytosolic (N/C) ratio, thus promoting the differentiation of MSCs and osteogenesis.^72,73

Extrinsic biomechanical parameters

Substrate stiffness measures resistance when a mechanical force is applied, which is given by ECM and influences cell fate and behavior, adhesion and migration.⁷⁴ Stem cells prefer to adhere to stiffer substrates present in their ECM environment under both static conditions, where stiffness is constant over time, and in dynamic conditions, where mechanical cues change temporally, such as in cyclic loading and time-dependent matrix stiffening.^75,76,77 During stem cell differentiation, substrate stiffness balances asymmetrical division via mechanotransduction to maintain populations of both stem cells and differentiated lineage cells.⁷⁸ In addition, increased substrate stiffness levels can facilitate proliferation in ESCs, HSCs, and MSCs.⁷⁷ Stem cell differentiation fate is controlled by the level of stiffness of the ECM.^79,80,81,82 In pluripotent stem cells, specific levels of stiffness in the culturing gel can induce differentiation toward specific lineages - a softer matrix, for example, favors differentiation toward cardiac and neural lineages.⁸³ By modifying the stiffness level of the environment, researchers can control stem cell differentiation processes to obtain desired differentiation fates.⁸⁴ These stem cells’ progeny holds multiple applications in fields such as biomedical device engineering.

Shear stress, a biophysical concept that refers to mechanical force induced by friction, is correlated with cell mobility.^85,86,87 Multiple shear stress mechanosensitive channels and protein complexes on the cell surface have been identified in recent decades, with remarkable significance in regulating stem cell proliferation and differentiation.^{88,89,90,91,92} Mechanosensitive ion channels, such as transient receptor potential ankyrin 1 (TrpA1) and Piezo-type mechanosensitive ion channels (Piezo1/2), mediate these effects primarily through mechanically induced cation influx.⁸⁷ Piezo1/2 are non-selective cation channels, and +activation of these channels leads predominantly to Ca²⁺ entry, triggering Ca²⁺-dependent signaling pathways (e.g., YAP/TAZ) that regulate stem cell fate.⁹³ In human neural stem cells, mechanically activated Piezo1-mediated Ca²⁺ influx influences YAP localization and fate specification.⁹⁴ Similarly, TrpA1 is a non-selective cation channel that conducts Ca²⁺ and Na⁺, with Ca²⁺ influx functioning as a key second messenger in downstream intracellular signaling pathways associated with mechanosensitive responses.⁹⁵

Stem cells are also known to sense the topographical features of ECMs, including characteristics such as roughness, grooves, or nanotexture. These surface patterns profoundly influence the morphology of MSCs, as well as their adhesion, proliferation, and lineage commitment.^{84,96,97,98,99} For example, micro-grooved substrates not only enhance MSC proliferation and pluripotency marker expression, but also shift differentiation toward adipogenic, osteogenic, and chondrogenic lineages through contact-guidance effects and altered cell shape dynamics.¹⁰⁰ Furthermore, recent reviews emphasize that surface topography, alongside elasticity, triggers mechanotransducive signaling cascades in MSCs, influencing adhesion, migration, proliferation, and differentiation by engaging integrins, cytoskeletal remodeling, and downstream biochemical pathways.^{100,101,102,103} These findings underscore the importance of ECM topography not merely as passive substrate features, but as active regulators of stem cell fate with key implications for intentional design of biomaterials and scaffolds in tissue engineering. In particular, nanotopography can modulate the secretion profile of human bone marrow MSCs by growing them on different nanotextured titanium surfaces. The MSCs exhibit distinct profiles in the small extracellular vesicles they secrete, and those produced under nanotopographic conditions have an enrichment in osteogenesis-related microRNAs, indicating that ECM topography can act as a regulatory factor in stem cell paracrine signaling.¹⁰⁴

In addition to interactions with the ECM, cells also interact with other neighboring cells present in the microenvironment, and these cell-cell interactions can shape how they sense and respond to mechanical cues from each other. Through mechanosensing processes, stem cells can adjust to changes in their environment and present with optimal phenotypes in response to those changes.^{54,76,105,106} The mechanical confinement and the geometric context of neighboring cells shape MSC organization, cytoskeletal architecture, and lineage outcome: limited degree of cell-cell contacts among MSCs inhibits the adipogenic, osteogenic, and chondrogenic differentiation, and vice versa.^{105,107,108,109} Moreover, integrin-mediated cell adhesion, actomyosin tension, and mechanoregulatory signaling through entities such as YAP/TAZ, Piezo channels, and cadherins facilitate direct biomechanical coupling between adjacent cells as well as their shared ECM.^24,110 Therefore, the mechanical context between cells is tightly intertwined, enabling stem cells to detect and adapt to complex microenvironmental changes, thereby optimally aligning their phenotypes with evolving physiological demands.

Epigenetic memory

Biomechanical cues not only alter short-term signaling cascades but also exert prolonged influences on epigenetics, providing a mechanism for mechanical forces to leave long-lasting molecular memories in the stem cells.¹¹¹ Epigenetic regulations, such as chromatin remodeling, histone modifications, and DNA methylation, have critical roles in governing stem cell fate decisions.^112,113 Both intrinsic and extrinsic mechanical inputs can alter these epigenetic marks, thereby shaping lineage specification and self-renewal capacity.^114,115

Since epigenetic regulation mostly takes place in the nucleus, nucleus stiffness has become an essential focal point in studying long-term mechano-epigenetic regulation on stem cells.¹¹⁶ Deformation of the nuclear envelope, particularly through the lamin A/C network, influences chromatin condensation and accessibility, thus affecting the gene expression of both serum response factor (SRF) involved in differentiation and a Hippo pathway factor (YAP1) involved in growth.⁶² Recent cryo-EM mapping has revealed direct binding between the lamin A/C tail and nucleosomes, providing structural evidence for how nuclear deformation translates into changes in chromatin condensation and gene accessibility.¹¹⁷ Matrix elasticity also regulates lamin A/C phosphorylation, reinforcing cytoskeletal-nuclear feedback loops that fine-tune chromatin state and lineage bias.¹¹⁸ Stiffer matrices and elevated cytoskeletal tension increase nuclear flattening (Figure 2), which has been correlated with reduced heterochromatin and enhanced transcriptional activity.¹¹⁹

Figure 2.

Matrix elasticity, which is often correlated with nuclear stiffening/flattening, leads to contrastingly different downstream effects in mesenchymal stem cells (MSCs)

Stiff matrix environments increase YAP/TAZ nuclear localization, coordinating signaling crosstalk across various transcription factors and decreasing chromatin condensation, which influences gene expression related to growth and differentiation. Conversely, soft matrix environments have decreased YAP/TAZ localization control and increased chromatin compaction, resulting in alternative stem cell fates. Figure created with BioRender.com.

Conversely, soft microenvironments promote chromatin compaction, biasing cells toward alternative fates.¹¹⁸ These structural changes interface with chromatin-modifying enzymes, such as histone acetyltransferases and deacetylases, which are differentially recruited under distinct mechanical contexts.¹²⁰ Upon nuclear translocation, the factors in YAP/TAZ pathway, along with mechanosensitive activators such as myocardin-related transcription factor A (MRTF-A), drive immediate transcriptional programs as well as recruit histone-modifying complexes, reinforcing lineage bias through persistent acetylation or methylation at specific loci.^120,121 Nuclear YAP/TAZ have also been shown to recruit the bromodomain-containing protein 4 (BRD4) chromatin-reader complex to super-enhancers, amplifying transcriptional outputs in response to substrate stiffness.¹²²

Stem cells are also subject to histone modifications and chromatin accessibility. MSCs cultured on stiff substrates display increased histone H3 lysine 4 trimethylation (H3K4me3) and histone H3 lysine 9 acetylation (H3K9ac) at osteogenic gene promoters,¹²³ thereby locking in osteogenic differentiation programs. In contrast, adipogenic conditions correlate with the deposition of repressive marks such as histone H3 lysine 27 trimethylation (H3K27me3) at osteogenic loci,¹²⁴ highlighting the reciprocal relationship between mechanical inputs and epigenetic patterning. Mechanical forces not only modify histone acetylation and methylation patterns, but can also reprogram chromatin accessibility and nucleosome positioning,^113,125 providing a mechanistic basis for how cells encode “mechanical memory” into stable transcriptional programs.

Furthermore, DNA methylation patterns have also been shown to respond to mechanical cues.¹¹⁴ Substrate elasticity influences methylation at key pluripotency and lineage-specific genes,¹²⁶ while flow-induced shear stress can demethylate endothelial-lineage loci to promote vascular differentiation.^127,128 Stiffer matrices in endothelial cells induce widespread DNA hypomethylation, particularly at angiogenic gene loci.¹¹⁵ This indicates that biomechanical inputs are capable of both initiating and stabilizing long-term epigenetic programs, thereby conferring a form of “mechanical memory” to stem cell populations.

Importantly, this form of “mechanical memory” is conceptually distinct from biochemical memory. Biochemical memory is primarily maintained via sustained ligand-receptor signaling and transcriptional feedback loops; whereas mechanical memory involves the persistence of cellular responses due to prior physical environments and is transmitted through cytoskeletal tension and nuclear deformation—these alterations are then stored and stabilized by chromatin and epigenetic reorganization.¹²⁹

Mechanical memory can result in either reversible or irreversible fate outcomes depending on the magnitude and duration of mechanical exposure. The transition from a transient mechanical signal to a sustained memory depends on mechanical dosing: the threshold of duration, amount, and intensity of a mechanical force a cell experiences that can result in long-term, irreversible cell fates.¹³⁰ In human MSCs, short-term culturing on stiff substrates results in the nuclear localization of mechanosensitive transcription factors, such as YAP/TAZ and runt-related transcription factor 2 (RUNX2), but the effects can be reversed upon transfer to soft matrices, consistent with a reversible mechanical memory state. However, prolonged stiff matrix culture results in sustained YAP/TAZ and RUNX2 activation and biased differentiation toward osteogenesis even after the mechanical stimulus is removed, illustrating irreversible mechanical memory.¹²⁴ Additional studies of hMSCs cultured on stiff and soft hydrogels revealed that persistent chromatin remodeling and histone modification states are dependent on the duration of stiff mechanical exposure, further supporting the idea of dose-dependent reversibility.¹³⁰

Therefore, mechanical memory differs from biochemical memory not only in its physical origin but also in its reliance on nuclear architecture and chromatin mechanics as primary information carriers. While biochemical memory is often mediated via feedback within signaling networks, mechanical memory is encoded through persistent alterations in nuclear shape, chromatin accessibility, histone modifications, and integrated epigenomic states, providing a structural basis for long-term commitment.¹³¹ These findings position epigenetics as a central downstream effector of mechanotransduction.^132,133 By encoding epigenetic marks without altering genetic structure, transient physical cues can be translated into lasting changes in stem cell identity and function.¹³⁴

Conserved mechanotransduction frameworks and cell-type specific outcomes

As stem cells are highly sensitive to both intrinsic cues and extrinsic cues, these biomechanical inputs need to be transduced into biochemical signals through conserved mechanotransduction pathways that regulate their proliferation, self-renewal, and lineage commitment. There are several major signaling cascades that have been identified as critical mediators of mechanobiology (Figure 3). Biomechanical regulation in stem cells emerges from a shared systems architecture rather than from lineage-specific signaling modules, where mechanical information is gathered through adhesion complexes, cytoskeletal tension networks, and nuclear mechanosensors that collectively convert physical inputs into transcriptional outputs. Multiple canonical pathways, including RhoA/ROCK,^135,136 YAP/TAZ,¹³⁷ FAK/ERK,^138,139 Wnt/β-catenin,^140,141 Notch receptor,^141,142 and mechanosensitive ion channels such as Piezo 1/2,^143,144 operate as interconnected nodes within this architecture, with extensive crosstalk and feedback that preclude simple linear signaling models. As tunable control points for directing stem cell fate decisions, all these mechanotransducive pathways may reflect parallel readouts of a common mechanical state rather than independent mechanosensors. And importantly, these interconnected mechanotransduction pathways are highly susceptible to modulation by biomaterials engineering strategies, including substrate stiffness, topography, and ligand presentation. As a result, mechanotransduction outcomes are not dictated solely by pathway activation per se, but by how these pathways form a broadly conserved mechanotransduction architecture that enables stem cells to interpret extracellular and intracellular forces in an integrative and holistic manner.

Figure 3.

Regulatory mechanisms of intrinsic and extrinsic biomechanical parameters on stem cells

Mechanical parameters (in bold) are sensed and transduced through major signaling cascades, including mitogen-activated protein kinase/extracellular signal-regulated kinase (MAPK/ERK), Rho GTPase/Rho-associated protein kinase (Rho/ROCK), focal adhesion kinase/Src family kinase (FAK/Src), the Wnt/β-catenin signaling pathway, the Notch signaling pathway, and Yes-associated protein/transcriptional coactivator with PDZ-binding motif (YAP/TAZ). Actin filaments, microtubules, and intermediate filaments are depicted in pink, purple, and green lines, respectively. Dashed arrow indicates interaction but not direct regulation. Figure created with BioRender.com.

At the intracellular level, actin cytoskeletal tension emerges as a central regulator of stem cell behavior across lineages. Actomyosin contractility facilitates force transmission from the cell surface to the nucleus, coordinating cytoskeletal remodeling with transcriptional responses.^30,32,33,48 Similarly, nuclear mechanosensing through lamin A/C provides a conserved mechanism by which mechanical forces influence chromatin organization and gene expression.^18,53,145 Across stem cell types, increased cytoskeletal tension and nuclear stiffness are generally associated with lineage commitment, while reduced tension favors maintenance of plasticity or quiescence. Microtubules and intermediate filaments further contribute to force distribution, nuclear positioning, and mechanical buffering, reinforcing the integration of cytoskeletal and nuclear mechanics in stem cell regulation.^38,41,50,51 Extrinsic mechanical cues are likewise interpreted through conserved mechanisms. Substrate stiffness modulates integrin clustering and focal adhesion maturation, leading to changes in cytoskeletal tension and downstream transcriptional activity in virtually all stem cell populations; shear stress activates mechanosensitive ion channels, leading to a calcium-dependent signaling that regulates proliferation, migration, and differentiation; ECM topography and cell-cell mechanical coupling further shape mechanotransduction by altering cytoskeletal organization, force anisotropy, and junctional signaling.^18,19,54 Collectively, these physical cues regulate stem cell behavior by converging on shared mechanosensitive pathways rather than acting through lineage-specific receptors alone.

While this core architecture appears broadly conserved, its functional consequences are strongly shaped by cell identity, prior mechanical history, and microenvironmental constraints, highlighting a key limitation in the field: current models frequently emphasize pathway engagement without resolving how mechanical thresholds, timing, or intracellular force distribution govern downstream fate decisions. Induced pluripotent stem cells (iPSCs) exhibit attenuated cytoskeletal tension and low nuclear stiffness, features that support pluripotency and epigenetic plasticity; mechanical cues in these cells primarily regulate self-renewal, lineage priming, and differentiation competence rather than enforcing terminal fate decisions.^53,146 In contrast, in MSCs, increased cytoskeletal tension, stiff substrates, and nuclear YAP/TAZ localization consistently bias differentiation toward mechanically demanding lineages such as osteoblasts.^30,32,52 Additionally, HSCs display yet another interpretation of mechanical signals, wherein excessive tension or stiffness disrupts niche retention and quiescence, promoting mobilization or premature differentiation rather than lineage maturation.^38,44,147 Mechanical memory further amplifies these distinctions, as transient biomechanical cues can be epigenetically encoded through lamin A/C-chromatin interactions, YAP/TAZ-dependent recruitment of chromatin modifiers, and changes in histone modification or DNA methylation.^18,53,145 The persistence and reversibility of this mechanical memory vary across stem cell types, contributing to long-term differences in fate stability and regenerative capacity.

The unifying mechanotransduction model of stem cells shares conserved mechanobiological sensing frameworks, but the interpretation of mechanical inputs will be different based on cell-type-specific transcriptional and epigenetic programs. Recognizing this distinction between conserved sensing and contextual interpretation is essential for reconciling seemingly inconsistent findings in the literature and for designing biomaterials and engineering strategies that appropriately match mechanical cues to specific stem cell applications. Beyond individual mechanical cues, a key conceptual understanding emerging from these studies is that stem cell mechanoregulation operates hierarchically across scales. Extrinsic mechanical inputs such as matrix stiffness, shear stress, and topography act as upstream signals, but their effects are filtered and integrated through intrinsic cellular mechanics, including cytoskeletal tension, nuclear lamina composition, and chromatin organization. In this framework, the cytoskeleton serves as a force-transmission and amplification layer, while nuclear mechanics function as a downstream gatekeeper that constrains transcriptional responsiveness to extracellular cues. This hierarchical coupling helps explain why identical mechanical environments can elicit divergent fate outcomes depending on cell state or mechanical history.

Despite this emerging model, several important controversies remain unresolved. It is unclear to what extent extrinsic mechanical cues can override intrinsic mechanical states, as growing evidence suggests that cytoskeletal and nuclear properties may limit responsiveness to otherwise instructive niche signals. Additionally, conflicting reports persist regarding whether static stiffness or time-dependent properties such as stress relaxation and matrix remodeling are more predictive of fate decisions. Finally, the frequent observation of heterogeneous outcomes under tightly controlled mechanical conditions highlights an unresolved tension between deterministic niche-design paradigms and probabilistic, noise-influenced models of stem cell fate regulation. Addressing these gaps will be critical for advancing predictive and generalizable bioengineered niche design.

Current challenges for clinical translation

Recent clinical investigations have increasingly focused on stem cell-based therapies and pharmacological interventions that modulate stem cell biomechanics. Contemporary trials encompass a broad range of strategies, including mesenchymal stromal cell transplantation, exosome-mediated delivery, and small-molecule inhibitors targeting cytoskeletal or mechanotransduction pathways. Interventions such as microtubule stabilizers, FAK inhibitors, and ROCK inhibitors underscore the growing recognition of biomechanical regulation as a critical determinant of stem cell fate and therapeutic efficacy.

While an increasing number of clinical studies reference biomechanical regulation or mechanotransduction pathways, it is important to distinguish mechanobiology-inspired concepts from direct stem cell-specific clinical implementation. Many pharmacological interventions currently in clinical trials, such as ROCK, FAK, or cytoskeletal inhibitors, modulate mechanotransduction pathways broadly across multiple cell types and tissues, rather than selectively targeting stem cells or engineered stem cell niches. In these cases, biomechanical regulation often serves as a conceptual framework informing therapeutic rationale, rather than a mechanism that is directly or exclusively applied to stem cell populations. As such, existing clinical efforts should be viewed as early translational probes of mechanobiology principles rather than definitive stem cell-based mechano-therapies. Based on clinical trial records from the National Library of Medicine (NIH), this table compiled representative studies from the past 15 years that focus on translating the biomechanical regulation of stem cells into potential therapeutic applications (Table 1).

Table 1.

Clinical studies intersecting with mechanotransduction pathways relevant to stem cell biology

ID	Year	Phase	Intervention	Targetted Mechanotransduction	Clinical Status
NCT01469598	2011	II	Gemcitabine and docetaxel in patients with previously treated metastatic esophageal squamous cell cancer	Cytoskeleton (microtubule depolymerization)	Completed
NCT02152774	2014	II	Rho-kinase inhibitor AR-12286 in patients with chronic angle-closure glaucoma	RhoA/ROCK pathway: (Rho-kinase)	Unknown
NCT02523014	2015	II	Visomodegib, FAK inhibitor GSK2256098, capivasertib, and abemaciclib efficacy on patients with progressive meningiomas	FAK/ERK pathway (FAK)	Ongoing
NCT03574753	2018	II	ABBV-399 effect on patients with C-Met-positive stage IV or recurrent squamous cell lung cancer	Cytoskeleton (microtubule depolymerization)	Completed
NCT03608631	2021	I	MSC-derived exosomes with siRNA against KrasG12D	ECM stiffness	Ongoing (non-randomized)
NCT04879368	2021	III	Reforafenib/nivolumab in patients with AGOC to test whether patients have improved overall survival vs. standard chemotherapy	Cytoskeleton (microtubule depolymerization)	Ongoing (non-randomized)
NCT03792490	2023	IV	Inhibition of Rho kinase (ROCK) with Fasudil as disease-modifying Treatment for ALS (ROCK-ALS)	RhoA/ROCK pathway; Cytoskeleton (actin filament)	Completed

Clinical trials targeting biomechanical regulation in stem cell biology. Summary of selected clinical trials investigating pharmacological agents, cell therapies, and biomaterial-based approaches in the past 10 years that modulate the mechanotransduction of stem cells. Trials are listed with intervention, mechanistic rationale, and clinical status.

While stem cell research has advanced significantly, several key challenges continue to limit its clinical translation. These include technical limitations in replicating biomechanical environments, difficulties in translating into therapeutics, safety, and ethical concerns. One of the main technical challenges is accurately reproducing the physical conditions that stem cells experience in the body. This arises from the oversimplification of mechanical environments in vitro, where individual parameters such as substrate stiffness are often varied in isolation, whereas in vivo tissues present tightly coupled mechanical, biochemical, and cellular cues. For example, changes in matrix stiffness are frequently confounded with alterations in ligand density, porosity, or viscoelasticity, making it difficult to attribute observed stem cell responses to stiffness alone.^{148,149,150,151} In addition, commonly used synthetic substrates may not recapitulate the nonlinear, dynamic, and time-dependent mechanical properties of native extracellular matrices, such as stress relaxation and strain stiffening, which are increasingly recognized as critical regulators of stem cell fate.^152,153,154 Factors such as matrix stiffness, stretch, and shear stress, all influence stem cell fate in a syngeneic manner, but it is often hard to separate them, and replicating all these forces consistently in laboratory settings is difficult. ^37,155 As a result, stem cell behavior observed in vitro often does not reflect what occurs in vivo. Another issue that further contributes to translational discrepancies is the difference in dimensionality, where bench-side studies rely on two-dimensional cultures but not on a three-dimensional tissue environment. Many mechanobiology studies rely on two-dimensional cultures that are able to exaggerate cell spreading, focal adhesion formation, and cytoskeletal tension compared to three-dimensional tissue environments.^156,157 In 3D matrices, stem cells experience spatial confinement, heterogeneous force distributions, and altered nuclear deformation, leading to mechanotransduction responses that diverge markedly from those observed on planar substrates. Cell density represents an additional confounding factor, as variations in cell-cell contact, paracrine signaling, and collective force generation can substantially alter mechanosensitive signaling pathways, including YAP/TAZ and Notch, independently of extracellular mechanics.^{102,158,159,160}

Moreover, in vitro systems often fail to capture the dynamic remodeling of the stem cell niche that occurs in vivo. Native tissues continuously adapt through matrix deposition, degradation, and force feedback between cells and their microenvironment, whereas static culture systems impose fixed mechanical conditions.^161,162This discrepancy is particularly relevant for stem cells with high migratory capacity or those residing in mechanically active niches, such as hematopoietic and vascular progenitors.^163,164 All these limitations give insights into the need for advancing experimental platforms that integrate matrix mechanics with dimensionality, cellular heterogeneity, and temporal dynamics in order to improve the physiological relevance and translational potential of mechanobiology studies. Currently, researchers in BME are working to develop systems that can deliver precise and tunable mechanical inputs, but reliable and scalable models are still limited.¹⁶⁵

In clinical settings, one of the most pressing issues is the lack of standard protocols for delivering mechanical cues during stem cell expansion and differentiation.¹⁶⁶ Without consistent control of these cues, different laboratories or manufacturing centers may produce stem cells with varying properties, which affects reproducibility and safety. Safety concerns are associated with manipulating core mechanosensitive pathways such as RhoA/ROCK, YAP/TAZ, and cytoskeletal dynamics, which are ubiquitously involved in cell proliferation, survival, and tissue homeostasis. Pharmacological or biomaterial-driven modulation of these pathways risks off-target effects, including aberrant differentiation, fibrosis, loss of stem cell quiescence, or tumorigenic potential, particularly when interventions alter nuclear mechanosensing or chromatin accessibility.^167,168,169 For example, a specific safety concern is the risk of teratoma formation, which can result from the presence of undifferentiated pluripotent stem cells.¹⁷⁰ These risks are amplified by the fact that mechanotransduction pathways often operate in a dose- and context-dependent manner, making it difficult to achieve therapeutic specificity without disrupting normal tissue mechanics.

Manufacturing and scalability further complicate translation. Reproducing precise mechanical cues at a clinical scale remains technically challenging, particularly under good manufacturing practice (GMP) conditions.^171,172 Unlike soluble-factor-based differentiation protocols, mechanically engineered cues are difficult to standardize, monitor, and validate under GMP conditions, complicating potency assays and regulatory approval. Small variations in material properties, cell density, or culture geometry can lead to substantial heterogeneity in stem cell fate and function, undermining reproducibility and batch-to-batch consistency. Scaling up these processes to produce large numbers of therapeutic cells while maintaining their quality and function is another major hurdle.¹⁷³ Ensuring that only fully differentiated and functional cells are delivered to patients requires efficient screening and purification methods, which are still being optimized. Even when mechanically primed stem cells demonstrate enhanced function in preclinical models, these benefits may be lost once cells encounter the dynamic and heterogeneous mechanical landscapes of native tissues.

Long-term monitoring for adverse effects is also necessary as therapies move into clinical use, which will pose additional hurdles. Current regulatory frameworks are better suited for chemically defined cell products than for therapies whose function depends on dynamic physical cues or engineered microenvironments.¹⁷¹ Mechanotransduction-based interventions frequently blur the boundary between cell therapy and device-based approaches, complicating safety evaluation, potency assays, and approval pathways. Together, these challenges highlight that successful translation will require not only deeper mechanistic understanding but also advances in biomaterial standardization, manufacturing control, and regulatory strategies tailored to mechanically engineered stem cell therapies.

Stem cell research continues to raise important ethical questions.¹⁷⁴ The use of ESCs has been controversial due to the destruction of embryos, and although iPSCs avoid this issue, they still present concerns related to genetic manipulation and informed consent.¹⁷⁵ Additionally, access to advanced stem cell treatments is not equally available to all populations, raising concerns about healthcare access and equity.¹⁷⁶ Addressing these challenges will require collaboration across disciplines, including cell biology, BME, clinical research, and bioethics. Both scientific and societal barriers must be addressed to translate stem cell research into safe and effective therapies.

Integration of biomedical engineering (BME) into stem cell-based therapeutics

Building on the solid foundations established through wet lab experimentation, researchers are now exploring an innovative direction that focuses on translating these results into practical applications with great promise. Because mechanotransduction relies on both the application and detection of physical forces, advancements in BME have opened the door for meaningful translation. Research groups have already begun integrating BME into stem cell biology, revealing exciting possibilities for improving patient care and medical outcomes.¹⁷⁷ Specifically, stem cell-based therapeutics would rely heavily on ex vivo expansion, a process in which stem cells are cultured and grown outside the body under controlled conditions, exemplifying this translational potential. This process is influenced by the properties of supporting biomaterials, including surface biochemistry, topography, and mechanical properties, which mimic the stem-cell niche and affect the expansion and differentiation fate of stem cells, with a promising application in regenerative medicine and drug discovery.¹⁷⁸ During ex vivo expansion, lineage selection is used to engineer stem cells toward specific target cell types, ensuring both safety and efficacy in therapeutic applications. In this content, many BME techniques can be used to conduct lineage selection in a faster and more specific manner by adjusting the mechanical features the stem cells receive.

Bioengineering strategies increasingly enable rapid, more specific lineage selection from iPSCs by tuning the mechanical features that cells experience ex vivo, rather than relying on soluble factors alone.¹⁷⁹,¹⁸⁰ Based on the major biomechanical parameters that are essential in stem cell behavior, a decision-tree framework is plotted based on the current techniques (Figure 4), where distinct mechanical inputs, such as ECM stiffness, flow/shear stress, and topography/alignment, can be manipulated by established BME platforms (e.g., tunable hydrogels, microfluidics/organ-on-chip, and nanotopography) to rationally control fate choice and maturation toward targeted therapeutic cell types.¹⁸⁰

Figure 4.

A decision tree framework summarizes engineered mechanical features that can bias induced pluripotent stem cell (iPSC) lineage specification

ECM stiffness, shear stress (magnitude, pattern, and delivery mode), and surface topography are shown as representative biophysical inputs that influence cell morphology, cytoskeletal organization, and mechanotransduction pathways, thereby directing lineage tendencies under defined in vitro conditions. Lineage outcomes in this decision tree may arise from the integration of multiple co-existing mechanical, biochemical, and contextual factors and should not be interpreted as exclusive or deterministic pathways. Figure created with BioRender.com.

A well-studied axis of this mechanical control is ECM elasticity, which spans a wide physiological range and can be recapitulated using synthetic or semi-synthetic matricess.^181,182 It has been established that tissue-like elasticities can bias lineage programs: soft matrices support neurogenic signatures, intermediate stiffness supports myogenic programs, and rigid matrices support osteogenic differentiation, providing a foundational “stiffness map” used to guide engineered culture systems.¹⁸³ In engineered iPSC workflows, stiffness is routinely modulated using tunable hydrogels (e.g., PEG- or GelMA-based networks) and micropatterned substrates, which allow independent control of cell spreading/traction and matrix mechanics to improve lineage selectivity and reproducibility.¹⁷⁹ Researchers have discovered a range of ECM elasticity that can guide stem cells into defined lineages (Figure 4), with stiffness ranging from 0.1 to 40 kPa (kPa), highlighting how mechanical elasticity influences fate decisions toward different lineages.^184,185,186 This reduces the risk of heterogeneous or unwanted cell populations, minimizing adverse outcomes during cell therapies. Advanced BME integrated techniques, such as biomaterial engineering of hydrogels, can enable the generation of highly specialized cells for specific needs, with more efficiency and higher yield than previous techniques.^{187,188,189,190} These artificial extracellular matrices enable researchers to observe many factors that impact stem cell interaction within its microenvironment and the subsequent influence of engineered ECM on stem cell fate.¹⁹¹ New advancements in BME have created opportunities to recreate these unique niche environments using advanced peptide and protein engineering so that the physiological model of stem cell niches can be mimicked, allowing for vast potential.¹⁹²

A rapidly expanding toolkit for tunable ECM uses protein-based hydrogels to decouple mechanics from biochemical presentation. Recombinant protein hydrogels are genetically programmable, enabling systematic tuning of crosslinking, viscoelasticity, and incorporation of defined adhesive motifs to control both stiffness and ligand presentation without reliance on animal-derived matrices.^193,194 In parallel, DNA-based hydrogels provide a programmable materials strategy with advantages in precise, customizable nano-or micro-scale patterning and controllable mechanical behavior through sequence design, modular hybridization, and integration with polymer/nanocomposite backbones.^195,196 Together, protein- and DNA-enabled matrices broaden the design space beyond static elasticity, supporting more faithful recapitulation of developmental niches via defined adhesion chemistry and programmable network mechanics.^193,195 Besides ECM stiffness, ligand density is another independent factor that can be manipulated and thus influences stem cells biomechanically, regardless of substrate stiffness. In gelatin methacryloyl (GelMA), a commonly used biopolymer that mimics ECM, the ligand density can be altered by changing the concentration of gelatin, where the RGD (Arg-Gly-Asp) peptide sequence is found there.¹⁹⁷ The RGD sequence mimics the cell-binding motifs found in many ECM proteins, such as collagen and fibronectin. With high ligand density in the ECM that facilitates αVβ3-integrin clustering and the formation of focal adhesions, a substantial increase in the YAP nuclear/cytoplasmic ratio is observed.¹⁹⁸ Higher ligand densities in GelMA generally promote osteogenic or proliferative phenotypes, whereas lower densities can favor stemness maintenance or alternative lineages, depending on ECM identity and integrin usage.¹⁹⁹

Beyond ECM mechanics, shear stress and flow provide another highly controllable cue that can be precisely delivered using microfluidics, perfusion bioreactors, and organ-on-chip platforms based on different flow-conditioning.²⁰⁰ Importantly, the flow magnitude, waveform, and pattern, either laminar or disturbed/oscillatory, can be used as an engineering “dial” to control the preferences toward endothelial or disease-relevant phenotypes for better research.²⁰¹ For instance, unidirectional shear stress has been used to promote hiPSC differentiation toward valvular endothelial-like cells, with mechanosensing mediated by a Piezo1-dependent Ca²⁺ axis.²⁰² Flow-enabled systems can also be applied to developmental-mimetic settings. For example, “dorsal aorta-on-chip” models apply mechanical cues to hemogenic endothelium to support hematopoietic emergence.²⁰³ More broadly, perfusion-induced shear can bias non-vascular outcomes in lineage-specific bioreactors, including reported enhancement of osteogenic differentiation through mechanosensing pathways such as ERK.⁹¹

Topography and alignment, where nanoscale and microscale patterning features, including nanogrooves, aligned fibers, and lithographic features, can also steer cell morphology and cytoskeletal organization, thus targeting specific cell type maturation state. In human iPSC-derived cardiomyocytes, aligned nanotopographic cues promote anisotropy and structural maturation consistent with improved tissue-like organization.²⁰⁴ Whereas neurogenic surface topographies, such as aligned nanofibers, microgrooved or ridge-groove substrates, and anisotropic adhesion or stiffness patterns, preferentially promote neuronal differentiation by inducing elongated cell morphology, enhancing neurite alignment and extension, suppressing YAP/TAZ activity, and activating pro-neural transcriptional programs that mimic features of the native neural extracellular matrix.^205,206,207

Nonetheless, cell-cell interaction architectures, especially those involving 3D spatial arrangements, including controlled ligand matching (e.g., Notch ligands), organoid co-development, vascular niche co-culture, and organoid fusion/assembloids, provide additional decision points by recreating contact-dependent signaling and multicellular context that are difficult to achieve in homogeneous 2D cultures (micropatterned fate patterning),²⁰⁸ gastruloid signaling dynamics,²⁰⁹ and brain organoid enhancement strategies.²¹⁰ Engineered organoids can be developed for further advanced research, and even be applied in real-world scenarios such as transplants or cell and tissue rejuvenation.^211,212,213

These BME-driven mechanical “decision modules” (Figure 4) help reduce heterogeneity and off-target populations while promoting functional maturation, strengthening the manufacturability, safety, and therapeutic potential of iPSC-derived cell products for cell replacement and regenerative applications.^214,215 While current bioengineering strategies primarily focus on externally applied and statically defined mechanical cues, native stem cell niches are increasingly recognized as active, non-equilibrium systems in which cells themselves generate forces and dynamically remodel their microenvironment.²¹⁶ Stem cells exert traction stresses, reorganize extracellular matrices, and modulate local stiffness and architecture through cytoskeletal activity and matrix deposition or degradation, creating feedback loops between cellular mechanics and fate regulation,²¹⁷ which introduce temporal variability and spatial heterogeneity that cannot be fully captured by pre-defined mechanical inputs alone.⁷⁴ Incorporating such active and adaptive behaviors into engineered niches represents an emerging frontier in BME, with implications for controlling stem cell plasticity, reducing culture-induced bias, and improving the physiological relevance of ex vivo differentiation systems.

Other fields in BME support more in-depth research for biomechanical regulations of stem cells, optimizing stem cell-based therapies, and improving the efficiency and reliability of stem cell therapie. This includes nanotechnology, single-cell profiling, and gene-editing technologies, which have promising results for enhancing the result of stem cell lineage specification.^{104,190,218,219,220} BME has also provided advanced technologies to further investigate mechanotransduction in stem cells, such as atomic force microscopy, which enables researchers to apply precisely measured, minimal forces to spatially defined regions and observe the resulting physical changes, such as those that occur at cell-cell interfaces.^{71,211,221,222} Image analysis techniques based mostly on confocal microscopy can provide more information than just a zoomed-in image.²²³ Live-cell imaging with fluorescent labeling enables the visualization of cellular behavior across spatial and temporal dimensions, revealing the dynamic distribution of protein of interest. This BME technique allows researchers to directly visualize what happens in cells using micro-level magnification, with improvements in experimental precision, scalability, and efficiency.^73,211,224

Recent developments in artificial intelligence (AI) and machine learning have also yielded great insights into stem cell biology. AI-based digital image analysis is becoming a common and useful application in science. AI can analyze more detailed characteristics and differences that humans cannot perceive with the naked eye, and make pattern-based predictions.²²⁵ AI-powered image analysis also aids in the early detection of cellular anomalies, reducing risks and improving treatment efficacy.²²⁶ Additionally, some computational methods and algorithms, such as deep-learning genomics, allow researchers to analyze vast amounts of genetic data, identifying structural and functional characteristics in stem cells, and uncovering patterns that drive stem cell behaviors and differentiation.^227,228 By using deep learning networks, including convolutional neural networks (CNNs), a successful prediction can be made on MSC lineage fate based on morphological changes during differentiation.²²⁹ They demonstrated not only that subtle morphological features are tied to fate decisions, but also that the usefulness of machine learning as a future technique to the fields of biomanufacturing and regenerative medicine. AI shows exciting promise to drive advancements in stem cell therapy by enhancing precision and efficiency. By applying AI techniques, researchers can optimize stem cell differentiation, predict patient outcomes, personalize treatments, analyze larger datasets for patterns of side effects, and monitor therapy responses in real-time.^226,230,231

Future directions and prospectives

The stem cell and regenerative medicine field is rapidly evolving from simple cell transplantation, either multipotent or pluripotent stem cells, to sophisticated, personalized engineered therapies, where precise therapeutics are designed according to individual patients’ unique needs.²³² As biomechanical regulation takes up a big part in stem cell behavior, by manipulating the pathways that govern stem cells using different BME techniques, researchers can tailor stem cells to better suit the therapeutic needs of patients, ensuring the optimization of therapies in clinical settings.²³³

Stem cell biology stands at the crossroads of fundamental science and translational medicine, offering unique opportunities for regenerative therapies. Since mechanotransduction has emerged as a critical determinant of stem cell fate, studies have identified both intrinsic biomechanical parameters, such as cytoskeletal dynamics and nuclear stiffness, and extrinsic biomechanical parameters, including substrate stiffness, shear stress, and ECM topography. Together, these factors not only regulate short-term cellular responses through major signaling cascades but also influence long-term epigenetic programs, providing a mechanistic link between physical forces and durable changes in stem cell behavior and functionality.

Integration of BME, and soon, AI and machine learning, will open new avenues for advancing personalized medicine through stem cell research. These innovations promise to significantly enhance both the efficiency and accuracy of comprehensive patient treatments. The integration of BME has further expanded the methods available to study and manipulate these processes, ranging from biomaterial scaffolds and tissue engineering to AI-driven analyses and precision medicine approaches. Despite ongoing challenges, the convergence of mechanobiology, stem cell biology, and BME holds the potential to transform stem cell research from benchwork into clinically viable therapies, ultimately improving patient outcomes across a wide spectrum of diseases and injuries. Through sustained collaboration among biologists, engineers, and clinicians, future research will continue to bridge mechanistic insight with practical innovation, paving the way toward precise, efficient, and patient-specific regenerative solutions.

Acknowledgments

The authors acknowledge support from the National Institutes of Health (HL162725, HL130995, and CA282579).

Author contributions

A.C., C.H., and E.Y. contributed equally to article creation and revision. C.K.Q. conceptualized the idea and supervised the work.

Declaration of interests

The authors declare no competing interests.