The force is strong with this epigenome: chromatin structure and mechanobiology

Abstract

All life forms sense and respond to mechanical stimuli. Throughout evolution, organisms develop diverse mechanosensing and mechanotransduction pathways, leading to fast and sustained mechanoresponses. Memory and plasticity characteristics of mechanoresponses are thought to be stored in the form of epigenetic modifications, including chromatin structure alterations. These mechanoresponses in the chromatin context share conserved principles across species, such as lateral inhibition during organogenesis and development. However, it remains unclear how mechanotransduction mechanisms alter chromatin structure for specific cellular functions, and if altered chromatin structure can mechanically affect the environment. In this review, we discuss how chromatin structure is altered by environmental forces via an outside-in pathway for cellular functions, and the emerging concept of how chromatin structure alterations can mechanically affect nuclear, cellular, and extracellular environments. This bidirectional mechanical feedback between chromatin of the cell and the environment can potentially have important physiological implications, such as in centromeric chromatin regulation of mechanobiology in mitosis, or in tumor-stroma interactions. Finally, we highlight the current challenges and open questions in the field and provide perspectives for future research.

Graphical abstract:

Introduction

Sensing and responding to stimuli are primary characteristics of life forms on earth. Organisms, no matter their size or complexity, experience mechanical stresses throughout their lifespan, such as compression, stretching, and liquid shear stress [1,2]. Therefore, establishment of reliable mechanosensing and mechanotransduction pathways is crucial. Mechanical forces can propagate over considerable distances through cellular structure rapidly [3]. In contrast, diffusible biochemical molecules move slowly and dissipate significantly as the distance increases, although this can be compensated by signal amplification through signaling cascades [4]. In single cells, mechanical forces can be sensed and transduced via two routes: a mechanoreceptor-cytoskeleton direct force transduction pathway, and a mechanoreceptor-coupled biochemical secondary messenger pathway. The former (e.g., integrins and actomyosin) permits direct propagation of forces all the way into the nucleus [5,6]. The latter (e.g., stretch-sensitive ion channels and G-protein coupled receptors) depends on biochemical signaling cascades [7] (Fig. 1). The combination of these two routes allows for fast and sustained cellular mechanoresponses, where gene expression is often altered [8]. However, evolution may prefer more than just speed, magnitude, and duration. Mechanoresponses often require memory and plasticity, to better prepare for future needs, and to help the organisms adapt to environmental changes [9,10]. The memory and plasticity are stored and applied in the form of epigenetic modifications to the chromatin and DNA inside the nucleus, which include chromatin structure alterations [11,12]. It is now widely accepted that chromatin is more than just the passive storage complex of DNA, it also possesses important non-genetic functions, including the active regulation of nuclear mechanical properties (viscoelasticity) [13]. Recent studies have shown that chromatin structure alteration is a common feature of cellular mechanoresponse. However, much of its molecular mechanisms and biological functions remain unresolved. Moreover, it is unclear whether chromatin structure alterations can mechanically affect the cell, and even the extracellular environment.

Figure 1: Overview of mechanotransduction pathways and chromatin structure.

Environmental forces are received by mechanoreceptors (#1–4) on the plasma membrane, which transduce forces to the actomyosin cytoskeleton (#5) or induce calcium (Ca²⁺) influx. Contraction of actomyosin cytoskeleton propagates forces to LINC complex (#6) on the nuclear envelope. The nuclear envelope deforms, and stretch-sensitive ion channels Piezo1 (#7) on the ER-nuclear membrane get activated for calcium influx into the nucleus and the cytoplasm. Nuclear pore complex (#8) gets dilated, increasing nuclear import of YAP/TAZ and mechanosensitive transcription factors. Globular (G-) actin and HDAC3 also import/export through the nuclear pore complex under different mechanical stimulations. LINC complex (#6) transduces forces to the nuclear lamina (#9), which relays forces to the chromatin through lamina-associated domains (#10, yellow). Chromatin structure consists of chromatin states (#11), chromatin compartments (#12), chromatin domains/topologically associating domains (TADs) (#13) formed by CTCF and cohesin (purple rings), chromatin loops (#14) formed by CTCF and cohesin, or by local transcription machinery protein interactions (dashed circle), chromatin fiber (#15), single nucleosomes (#16), and DNA (#17). (Created with BioRender.com)

In this review, we first discuss diverse vs. conserved aspects of mechanobiology and chromatin regulation throughout evolution. We then discuss how chromatin structure emerges to be at the center of cell-environment mechanical feedback pathways, with additional perspectives on centromeric chromatin and tumor therapeutics. Finally, we discuss technical challenges faced in the field and potential ways to overcome them.

Evolutionary perspectives of mechanobiology in the chromatin context

How each individual organism responds to mechanical stresses is highly context-dependent, including changes in their life cycle. A bacteria (Pseudomonas aeruginosa) senses shear stress to decide when to initiate biofilm formation for mechanical protection [14]. A thale cress plant (Arabidopsis thaliana) lacks the capacity to mobilize and evade mechanical stresses, while the Nopili rock-climbing goby (Sicyopterus stimpsoni) and the cave angel fish (Cryptotora thamicola) have the capacity to withstand extensive aquatic pressure when they scale a waterfall. Single cell organisms such as the Stentor ciliate experience unique mechanical stresses, both from their aquatic life form, as well as their ability to stretch extensively in search for food while being physically stuck to one spot in a stream [15]. Another example of mechanosensitive process is cell migration, which is critical for all mobile organisms from single cell organisms like Amoeba proteus, to multi-cellular organisms like Homo sapiens. In amoeba, cell migration dictates all critical aspects of life, such as moving toward nutrients and evasion from danger and mechanical stresses [16]. In humans, cell migration dictates development and organogenesis, tissue maintenance, wound healing, immune surveillance, and cancer metastasis [17].

Current mechanobiology research is heavily biased toward multicellular organisms. In fact, very few studies exist on the impact of mechanobiology on gene expression and chromatin structure in single cell eukaryotes [18], whereas mechanoresponses in fungi/yeast are extensively studied [19]. Works in plants reveal that mechanical stresses impact both cell wall biosynthesis and cytoskeletal genes, affecting fruit ripening, risk of pathogenic infections, and meristem development [20–25]. In a recent perspective, two of the fifteen most pressing open questions in plant biology are mechanobiology-related [26], including “Is the plant cortical microtubule cytoskeleton a mechanosensory apparatus?” and “Can mechanical forces trigger new cell fates in plants?” There are indeed strong clues in Arabidopsis that mechanical forces influence cell fates. During the development of young floral buds in Arabidopsis, tissue is deformed, resulting in compression of cells at the boundary that separates the floral bud from the stem cell niche (meristem) [27,28]. This boundary domain has a specific gene expression profile, including the mechanically induced expression of linker histone H1.3 [23], providing strong evidence that chromatin compaction is modified upon developmentally regulated mechanical stress. Similarly, in fish muscle and micropyle development are dependent on mechanical stress [29,30]. Single-cell differentiation within tissue is commonly dependent on lateral inhibition driven by the mechanosensitive YAP/TAZ signaling pathway across species [31]. In addition, the WNT-β-catenin pathway has also been found central to the formation of body axis in the freshwater organism Hydra [32] and blastoderm of fruit flies [33]. The expression of WNT3 is considered a quantitative genetic read-out for cellular stretching in Hydra [34]. These diverse studies indicate that the conversion of mechanical inputs into cellular differentiation pathways is a conserved principle (Table 1). Moreover, the use of stretch-sensitive ion channels (in both single- and multi-cellular organisms) and cell-cell junctions (in multi-cellular organisms) as mechanosensors are also highly conserved across species (Table 1). Finally, the mechanotransduction pathway through the cytoskeleton and the LINC complex is another well conserved feature that enables force propagation to the nucleus of eukaryotes (Table 1).

Table 1: Key features of mechanobiology throughout evolution.

Conserved features shared across at least 4 species are highlighted with underlines.

Species	Mechanosensors	Mechanoresponses	Key molecular players	References
Bacteria	Stretch-sensitive ion channels Type IV pili Flagella	Osmolarity and cell volume regulation Cell wall regulation Surface adhesion Biofilm formation	Ca2+ Cpx cAMP c-di-GMP DegS-DegU	[1,14,35,36]
Fungi	Stretch-sensitive ion channels Cell wall-plasma membrane sensors Nano-spring extracellular domains	Cell wall regulation Cell polarity regulation Chromatin flow	Ca2+ MAPK Cytoskeleton LINC complex Chromatin structure	[19,37–39]
Plants	Stretch-sensitive ion channels Cell wall-plasma membrane sensors Cell-cell junctions	Osmolarity and cell volume regulation Cell wall regulation Cell fate determination Organogenesis Chromatin structure alterations	Ca2+ Cl− Cytoskeleton LINC complex Chromatin structure (H1.3)	[23,40–43]
Hydra		Cell fate determination	WNT-β-catenin MAPK	[32,34]
C. elegans	Stretch-sensitive ion channels Cell-cell junctions	Cell fate determination Organogenesis	Ca2+ WNT-β-catenin Cytoskeleton LINC complex Nuclear lamina	[7,44–46]
Drosophila	Stretch-sensitive ion channels Cell-cell junctions	Cell fate determination Organogenesis Cell cycle control	H+ WNT-β-catenin YAP/TAZ Cytoskeleton LINC complex Nuclear lamina	[33,47–50]
Fish	Stretch-sensitive ion channels Cell-cell junctions	Cell fate determination Organogenesis Cell cycle control Immune response	Ca2+ WNT-β-catenin YAP/TAZ cPLA2 Cytoskeleton LINC complex Nuclear lamina	[30,51–54]
Mammals	Stretch-sensitive ion channels Cell-cell junctions Cell-matrix adhesion complexes Glycocalyx G-protein-coupled receptors	Cell fate determination Organogenesis Cell cycle control Cell motility Chromatin structure alterations	Ca2+ WNT-β-catenin YAP/TAZ cPLA2 Cytoskeleton LINC complex Nuclear lamina Nuclear actin Chromatin structure (H3K9me3, H3K27me3, H3K4me3, H3ac)	[8–10,55–63]

Taken together, all life forms experience mechanical stresses, and there appears to be conserved features in how they sense and respond to mechanical stimuli (Table 1). One outstanding question is, how well conserved are the transcriptional outcome and chromatin structure alterations in response to specific mechanical stress? Another question is, since mechanical stresses must have been experienced by the earliest organisms, is the link between mechanosensing and gene expression more ancient than the evolution of biochemical signal transduction pathways?

Chromatin structure takes the center stage of mechanobiology

As the field of mechanobiology and its studying methods develop (Box 1), the concept of mechanoreciprocity [64], namely, the mechanical interaction of cells and the environment, gains more attention. Forces are sensed and transduced inside the cell, which elicits cellular mechanoresponses that in turn shape the environment. Emerging evidence suggests that chromatin structure sits at the center of these bidirectional mechanical feedback pathways. In the following section, we highlight the outside-in and inside-out mechanobiology pathways of chromatin structure. Comprehensive mechanisms of mechanotransduction and the effects on chromatin are also nicely summarized in these recent reviews [8,65].

Box 1. Methods studying mechanobiology and chromatin structure.

The fast-developing chromatin and mechanobiology field relies on several powerful interrogative tools. These methods (except for the non-invasive imaging-based methods) provide a means of mechanical perturbation to monitor the mechanoresponse of samples. However, only a subset of them can simultaneously quantify mechanical properties such as the stiffness/elasticity of samples.

Atomic force microscopy (AFM) is one of the classical methods used for probing the nucleus and the chromatin, which applies mechanical force to indent samples while measuring their elasticity and topography. AFM uses flexible cantilevers and tips which can be adapted to samples with a wide range of scales and stiffness, from three-dimensional (3D) tissue [66] to single cells [67] and single nucleosomes [68]. The high-speed scanning feature of AFM also enables the monitoring of chromatin and nucleosome dynamics with minimal perturbation [69,70]. Other classical tools include micromanipulation [71,72], and optical/magnetic tweezers [5,73,74]. Both of them can be used to precisely stretch and mechanically perturb single chromosomes [71,74] or single nuclei [5,72,73] while measuring their elasticity. Optical tweezer-based active microrheology is a variant of the classical optical tweezer, which enables oscillation of beads inside the cell for local displacement and subsequent viscoelasticity measurements [75]. However, the beads cannot enter the nucleus for chromatin-related probing.

Emerging mechanical perturbation methods are now being used in the field, for example, micropatterning, where cells are seeded on different geometric constraints [76,77]; substrate softening/stiffening, where cells are seeded on matrix with tunable stiffness via degradation or crosslinking [58,78]; substrate stretching, where cells are subjected to stretching on a flexible membrane substrate [10], or between aligned suspended nanostrings [79]. Other methods include osmotic shock, where cells are subjected to change in media osmolarity that leads to nuclear compression or swelling [80]; compression, where cells are confined vertically with height defined by microbeads or micropillars [62,63]; and microfluidics, where cells pass through confined microchannels via migration [59] or perfusion [81]. As an alternative to microfluidics, migration through confined transwell assays is used for transient nuclear deformation [82,83]. Microfluidics also include micropipette aspiration [84], and it is the only emerging method that provides quantification of nuclear stiffness and viscosity at the same time. Finally, sending cells on a suborbital ballistic rocket flight or microgravity parabolic flight to test the effect of altered gravity on chromatin [85,86] is a futuristic method that may have substantial implications for the space age.

A few recent techniques enable loci-specific mechanical manipulation in live cells: condensate-based methods, where loci of interest are tagged with intrinsically disordered regions (IDRs) via optogenetics, subsequently forming condensate that mechanically excludes dense chromatin and brings together tagged sites [87]. Magnetic-based methods, where loci of interest are tagged with magnetic nanoparticles and pulled toward an external magnetic pillar, for short or long distances within the nucleus, could be a very powerful method to explore mechanical effects on chromatin [88].

Developments in imaging-based methods enable fully non-invasive mechanobiology, where live cells or tissue are not mechanically perturbed, but instead are imaged by specialized microscopy setup and analyzed to assign mechanical properties. Examples include image-based elastography [89], Brillouin microscopy (for elasticity) [90,91], and fluorescence correlation spectroscopy (for viscosity) [79]. Notably, advances in Brillouin microscopy [92] were highlighted as among the top 10 science stories of the year 2022 [93].

Modern methods of studying chromatin structure can be roughly divided into two main categories: sequencing-based and imaging-based. Sequencing-based methods allow high-throughput identification of chromatin structure in a genome-wide fashion. Examples include Hi-C and its derivatives, where 3D chromatin interaction sites are cross-linked, digested, and ligated for sequencing, to identify chromatin A/B (active/inactive) compartments and topologically associating domains (TADs) [82,83,94,95]; MNase-seq, DNase-seq, and ATAC-seq, where chromatin is either digested (by micrococcal nuclease or DNase I), or attacked and fragmented by transposases for sequencing, to identify nucleosome positioning and chromatin accessibility (open/close state) [59,96–98]; and ChIP-seq/ChIP-qPCR and CUT&RUN, where specific histones (e.g., variants or modifications) or chromatin binding factors are pulled-down along with bound DNA fragments for sequencing (or targeted qPCR), to identify their distribution and relative abundance across the genome [10,55,99].

Imaging-based methods are generally lower-throughput, but they allow direct visualization of chromatin structure inside the cell, with some offering real-time capabilities in live cells. Examples include electron microscopy such as transmission electron microscopy (TEM) and cryo-EM/cryo-ET, where the electron dense property of heterochromatin and single nucleosomes enables their structure and localization inside the nucleus to be imaged [10,100,101]; fluorescence in situ hybridization (FISH) and immunofluorescence staining, where specific parts of the chromatin, histone modifications, or chromatin binding factors can be fluorescently labeled for localization and quantification [9,58,59,76,81,83,102], and single-nucleosome resolution can be achieved when coupled with super-resolution microscopy such as stochastic optical reconstruction microscopy (STORM) [103,104]; and finally, super-resolution single-molecule tracking such as photoactivated localization microscopy (PALM) coupled with oblique illumination microscopy, where nucleosome dynamics of live cells are monitored in real-time, which can also be used to infer chromatin and nuclear mechanics [105,106].

For comprehensive summaries of many of the mechanobiology and chromatin structure methods mentioned and their pros and cons, please refer to these excellent reviews [107–109].

1. Outside-in mechanotransduction alters chromatin structure

How do cells sense forces, and how do forces get transduced into biological mechanoresponses, including chromatin changes? Under physiological conditions, cells encounter environmental forces such as compression (due to cell crowding or cell migration, etc.), stretching (due to tissue architecture, cell spreading, or blood pressure, etc.), or shear stress (due to blood flow, etc.). As illustrated in Figure 1, cell-cell junctions, cell-matrix adhesion complexes, or glycocalyx on the plasma membrane sense and transmit these forces to the contractile cytoskeletons, such as actomyosin [8]. Actomyosin cytoskeleton contraction then relays forces to the nuclear envelope through the LINC complex [110]. Tension on the nuclear envelope can lead to nuclear pore dilation, where shuttling into the nucleus is increased [61]. Increased nuclear import of the mechanosensitive YAP/TAZ coactivators and mechanosensitive transcription factors leads to downstream mechanosensitive gene expression [61]. LINC complex can further propagate forces into the nucleus, where it connects the nuclear lamina and eventually leads to chromatin via lamina-associated domains (LADs) [111]. (Fig. 1) It is unclear whether biochemical secondary messenger-based pathways at the plasma membrane, such as stretch-sensitive ion influx, contributes to chromatin mechanoresponse.

Aside from mechanosensitive gene expression, force-induced chromatin structure alteration is the most prominent chromatin mechanoresponse. Chromatin is organized into a complex multi-layered structure, from chromatin states all the way down to single nucleosomes and DNA (Fig. 1), as demonstrated by modern methods developed in the field (Box 1). Most of these layers are affected by mechanical stimuli. The nature of force-induced chromatin structure alteration is cell type-, force magnitude-, and time frame-specific [10,112]. It is also determined by the baseline nuclear mechanics and chromatin properties of individual cells [9,10,113]. Despite these context-dependent factors, some unifying principles and molecular mechanisms begin to emerge from recent studies (Fig. 2).

Figure 2: Outside-in mechanotransduction alters chromatin structure and leads to biological functions.

Principles of mechanotransduction pathways mediating chromatin structure alterations (top), and their biological functions (bottom). Blue shading (upper left) indicates the traditional mechanoreceptor-cytoskeleton mechanotransduction pathways, while turquoise shading (upper right) indicates the emerging nuclear-deformation related pathways. Purple shading (lower left) indicates regulation of cellular functions, while pink shading (lower right) indicates damage control mechanisms. All mechanotransduction pathways to the chromatin pass through the nuclear lamina, and are likely regulated by their association with it (yellow area). (Created with BioRender.com)

A). Integrin-mediated

Chromatin of Chinese hamster ovary cells responds to twisting/stretching forces applied to integrins, leading to direct chromatin stretching, chromatin decondensation via reduced H3K9me3 heterochromatin (inactive chromatin), and increased gene expression [5,6]. In the same way, chromatin of human mesenchymal stem cells responds to stiffening of substrate, leading to chromatin decondensation via increased H3ac and H3K4me3 euchromatin (active chromatin), redistribution of H3K27me3 heterochromatin away from the nuclear periphery, and ultimately differentiation [58,78,104]. In contrast, chromatin of pig myofibroblasts responds to substrate stiffening with chromatin condensation (reduced H3ac euchromatin), leading to persistent myofibroblast activation [114]. Similarly, leukemic T cells’ adhesion to substrate alone via integrins induces chromatin condensation (increased H3K9me2/3 heterochromatin), facilitating migration through confined three-dimensional (3D) environments [115].

B). Actomyosin contractility-mediated

Force-induced reduction of actomyosin contractility upon cell geometry change or compression of single mouse fibroblasts can induce translocation of HDAC3 into the nucleus, causing chromatin condensation (increased H3K9me3 and H3K27me3 heterochromatin), and the subsequent gene silencing within an hour [76,77,116]. In contrast, when mouse fibroblasts are cultured for days on geometric constraints, they proliferate into laterally confined 3D spheroids with stem cell-like phenotypes, which undergo global chromatin decondensation (increased H3K9ac euchromatin) and enrichment of H3K4me3/H3K27me3 bivalent chromatin [117]. The process is also mediated by reduction of actomyosin contractility [118]. Moreover, pharmacological induction of chromatin decondensation (increasing H3K9ac euchromatin) by trichostatin A (TSA) alone is sufficient to induce such stem cell-like spheroids for tissue rejuvenation [119]. Similarly, chromatin decondensation (increased H3K4me3 euchromatin) and the subsequent nuclear softening in leukemic T cells are mediated by increased actomyosin contractility during 3D migration [57].

C). Nuclear deformation: calcium-mediated and nuclear actin-mediated

Mechanical stimulations (Box 1) that cause nuclear deformation subsequently induce chromatin structure alterations, including altered chromatin states (euchromatin and heterochromatin) [9,10,55,59,79,81,83,101], A/B (active/inactive) compartments [82,83], and TADs [82,83]. While the molecular mechanisms involved in most of the nuclear deformation-induced chromatin structure alterations remain unclear, the Piezo1 calcium-mediated pathway [10] and the nuclear actin-mediated pathway [55] are two primary candidates.

Stretch-sensitive Piezo1 calcium channels on the nuclear-ER membrane are shown to dictate nuclear force sensing, which open for calcium influx (form perinuclear-ER lumen into the cytoplasm and the nucleus) upon unfolding of the nuclear envelope wrinkles [10,62,63]. Nuclear deformation-induced calcium influx leads to chromatin decondensation (decreased H3K9me3 and H3K27me3 heterochromatin) and nuclear softening [10]. Nuclear globular actin (G-actin) is shown to redistribute to the nuclear exterior upon substrate stretching, forming filamentous actin (F-actin) rings [55]. The decrease in nuclear G-actin inhibits transcription, and thus leads to heterochromatin formation [55]. Interestingly, formation of F-actin rings is also driven by increased intracellular calcium [10,120], suggesting that the calcium-mediated pathway may be an upstream mechanism.

Biological functions of nuclear deformation-induced chromatin structure alterations can be divided into two categories: cellular function regulation, and damage control mechanisms (Fig. 2). Cellular function regulation includes cell fate determination [55,81], nuclear stiffening for cellular (mechanosensitive) function determination [9], heterochromatin increase and reorganization for tumor invasion [101] and migration adaptation [59,83]. Damage control mechanisms include protection against disruption of long-range interactions [82], and transient nuclear softening to avoid force-induced DNA damage [10]. These damage control mechanisms of chromatin structure alterations demonstrate how chromatin adapts to mechanical perturbations in a way that minimizes disruption to normal gene expression programs. Notably, Hi-C data shows that short-range interactions within B compartments (inactive heterochromatin) are predominantly disrupted after nuclear deformation in confined migration [82]. Nuclear deformation-induced transient decrease of H3K9me3 heterochromatin (for DNA damage protection) [10] and long-term increase of H3K9me3/H3K27me3 heterochromatin (for migration adaptation/fitness) [59] also occur predominantly at the inactive, gene-poor and less accessible heterochromatin/B compartment. Together, this suggests a “buffering” role of heterochromatin during mechanical perturbation, absorbing disruptions while protecting the genome from large-scale gene expression alteration in the coding regions.

D). DNA damage and nuclear envelope rupture-mediated

The final mechanism of force-induced chromatin structure alteration is a byproduct of nuclear deformation, and more of a negative consequence than a mechanoresponse: DNA damage and nuclear envelope rupture. DNA damage, such as double-strand DNA breaks, alters chromatin structure during DNA damage response [121]. Nuclear deformation by compression or confined migration induces replication stress at S/G2 phase of the cell cycle, leading to DNA damage [122]. Nuclear envelope rupture is the breach of the interphase nuclear envelope under extreme mechanical stress and nuclear deformation, leading to uncontrolled exchange of nucleoplasm and cytoplasm [123–125]. Nuclear envelope rupture can lead to nuclear entry of cytoplasmic nucleases (TREX1), causing DNA damage [126]. It also exposes chromatin to the cytoplasm, leading to local chromatin condensation and activation of the cGAS-STING signaling pathway [101]. cGAS-STING is an evolutionary conserved cytosolic DNA sensing pathway part of the anti-viral immunity mechanism [127]. However, when aberrantly activated by self-DNA due to nuclear envelope rupture, cGAS-STING drives epithelial-to-mesenchymal transition (EMT) and tumor invasion in breast cancer cells [101]. In fact, excessive and chronic DNA damage caused by nuclear envelope rupture alone can trigger EMT and tumor invasion in similar settings [126]. Mechanosensitive translocation of YAP is also shown to drive DNA damage, nuclear envelope rupture, and the subsequent cGAS-STING signaling in melanoma cells [113].

E). Association with the nuclear lamina: a regulatory determinant

The location of chromosome territories (areas occupied by each single chromosome in the interphase nucleus) and their association with the nuclear lamina (through LADs) play a regulatory role in determining the chromatin structure and transcriptional outcome upon mechanical perturbation. Chromatin association with nuclear lamina is generally linked to transcription repression, as tethering loci to the nuclear periphery leads to silencing [128]. Stretch-induced gene expression via integrins is blocked by the silencing heterochromatin at the nuclear periphery [6]. Soft substrate induces chromatin redistribution from the nuclear interior to the nuclear periphery, coinciding with H3K27me3 heterochromatin formation at the nuclear periphery [104]. Likewise, nuclear deformation caused by confined migration reorganizes H3K9me3 heterochromatin to the nuclear periphery. Hi-C data also shows that a higher fraction of constitutive LADs switch to the inactive B compartment, compared to other parts of the genome [83]. Compression from tissue crowding and fluidification (i.e., the process of cell unjamming in densely packed tissue) also induces more H3K27me3 heterochromatin near the nuclear periphery [101]. Under microgravity, chromosome 18, which is predominantly located at the nuclear periphery in most human cell lines [129], increases overall gene expression; while chromosome 19, which is predominantly located away from the nuclear periphery, decreases overall gene expression, likely due to their opposite movement (chromosome 18 moving inward, and chromosome 19 moving outward) in the nucleus [86]. Moreover, Hi-C data shows that smaller chromosomes, which are generally located away from the nuclear periphery [130], have a significant increase of intra- and inter- chromosomal interactions under microgravity [86]. Overall, these studies highlight the regulatory role of association with the nuclear lamina in force-induced chromatin structure alterations (Fig. 2, yellow).

2. Inside-out mechanobiology of chromatin structure

How do chromatin structure alterations affect nuclear and cellular mechanical properties, and how does that alter the mechanosensitivity of the cell? Inside-out mechanobiology is less studied compared to its outside-in counterpart, but concepts have been proposed based on recent studies (Fig. 3).

Figure 3: Inside-out mechanobiology of chromatin structure alterations.

Principles of molecular sources leading to chromatin structure alterations (top), and their mechanical output (bottom). Blue shading (upper left) indicates chromatin reorganization, while turquoise shading (upper right) indicates chromatin in crisis. Purple shading (lower left) indicates cellular mechanical changes, while pink shading (lower right) indicates extracellular mechanical changes. (Created with BioRender.com)

A). Starting from the nucleosomes: histone variants and histone modifications

The most basic structure of chromatin is the nucleosome, consisting of two copies of histones H2A, H2B, H3, and H4 to form the octamer core, which wraps 147 bp of DNA. DNA that links adjacent nucleosomes is often bound to the linker histone H1 [131]. Histone deposition comes in two routes: the replication-coupled canonical histones, and the replication-independent histone variants. Histone variants are incorporated into nucleosomes, replacing the canonical ones due to various reasons, including specialized chromatin structure (e.g., H3 variant CENP-A at centromeres), tissue or developmental stage-specific expression (e.g., H2 variant TH2B at testes), or special genomic functions (e.g., H2A variants H2A.X for DNA damage response) [132–134]. Aside from structural and functional differences, nucleosomes with histone variants are also mechanically different from nucleosomes with canonical histones. CENP-A nucleosomes are softer (more deformable) than the canonical H3 nucleosomes, according to AFM-based nanomechanical force spectroscopy [68]. Moreover, CENP-A nucleosomes can become stiffer (less deformable) upon binding of key components of the kinetochore [68], which potentially regulates the mechanobiology of the centromere and mitosis (Box 2). Several histone variants are also shown to regulate heterochromatin formation, for example, H3.3, macroH2A, and H2A.Z [132–134]. Therefore, histone variant composition and binding partners of nucleosomes, the most basic building block of chromatin, may have greater mechanical implications in higher-order structure of the chromatin and the nucleus. In fact, the specific upregulation of an H1 linker histone variant in plants during leaf development [23] highlights the potential role of histone variants in chromatin mechanoresponses.

Box 2. Centromeric chromatin in mechanobiology of mitosis.

During interphase, chromatin is confined to the nucleus, where it is physically separated from the cytoplasm. In contrast, during mitosis, the nuclear envelope breaks down and chromosomes condense into rod-shaped structures, which subsequently align on the metaphase plate [135]. The sister chromatids are pulled towards the spindle poles, with each pole pulling on one of the two sister chromatids via the spindle microtubules. Once a balance is sensed by all sister chromatids, the spindle assembly checkpoint is satisfied, and chromosome segregation occurs [136]. During cytokinesis, the nuclear envelope reforms, again separating the chromatin from the cytoplasm. During the process of mitosis, chromosomes and specifically the centromeres experience extensive pulling and pushing forces from the spindle apparatus [137–139]. Consequently, cells must evolve machinery that both regulate and resist the mitotic pulling and pushing forces at the chromatin level. For example, chromosome arms undergo extensive compaction via condensing and global histone deacetylation, which provides the mechanical strength critical to resist penetrating forces from the mitotic spindles [140]. Tension at the centromeres is the primary source of information for the spindle assembly checkpoint to ensure error-free chromosome segregation [141–143]. Therefore, mechanical regulation of the centromeric chromatin is crucial for mitosis.

The centromere is where the large protein complex called the kinetochore assembles. The kinetochore forms the physical connection between centromeric chromatin and the mitotic spindles. Centromeric chromatin contains the specialized H3 variant, CENP-A [132–134]. Only the CENP-A nucleosome is known to bind to key components of the kinetochore, namely CENP-C and CENP-N [144]. Computational and experimental work show that CENP-A nucleosomes are softer and more deformable than H3 nucleosomes under the same magnitude of force application [68,145–147], and CENP-A nucleosomes can become stiffer and less deformable via binding to a small fragment of CENP-C [68]. It is estimated that around 400 CENP-A proteins are at the centromere, which would make up around 10–20% of all nucleosomes at the centromere [148]. The total level of CENP-C at the centromere is lower than CENP-A, according to ChIP-seq and chromatin fiber analyses [149,150]. This means that less than 10–20% of all centromeric nucleosomes can be stiffened upon kinetochore assembly. These data prompt the question: if, and how does this small number of stiffened nucleosomes regulate the stiffness of centromeres during mitosis? Does centromere stiffness regulate chromosome resistance to mitotic forces, therefore affecting tension at the centromeres and proper chromosome segregation?

Histone modifications, the post-translational modifications on the N-terminal tails of histones, are another crucial feature at the nucleosome level that correlates with higher-order chromatin structure, specifically, chromatin states. For example, H3K9me3 generally correlates with constitutive heterochromatin, which is located near centromeric and telomeric regions [10,151,152]; and H3K27me3 generally correlates with facultative heterochromatin, which is developmentally regulated [152,153]. In contrast, H3K4me3 and H3ac (H3K9ac/H3K27ac) generally correlate with euchromatin, which is permissive for gene expression [9,104,140,154]. While H3K9me3 precedes heterochromatin formation, as propagation of heterochromatin is dependent on pre-existing H3K9me3 [155], other histone modifications inform chromatin states with more complexity and interplay. For example, bivalent chromatin contains both H3K4me3 and H3K27me3, which leads to an intermediate and poised state between heterochromatin and euchromatin [156]. We will discuss the roles of chromatin states in mechanobiology in the following section.

B). Chromatin structure alters nuclear mechanics and mechanosensitivity

As illustrated in the bottom half of Figure 1, adjacent nucleosomes connected by linker DNA form the classical beads-on-a-string chromatin structure, which form into chromatin fibers [103], and fibers interact with each other to form higher-order structures [157,158]. Gene regulatory elements like enhancers and promoters are brought together into chromatin loops, by local protein interactions of the transcription machinery, or by cohesin-based loop extrusion within CTCF boundaries. Chromatin loops subsequently form larger-scale chromatin domains/TADs that can be either dynamic or stable. Different TADs of the same active/inactive states form an A/B chromatin compartment. Compartments of the same states form euchromatin/heterochromatin. Multiple compartments form individual chromosome territories inside the interphase nucleus [159,160].

How chromatin loops and TADs contribute to mechanical properties of the nucleus remains to be determined, but recent studies have correlated chromatin states with nuclear mechanics more clearly. Chromatin is shown to dictate a regime of mechanical response (smaller scale) separate from that of nuclear lamins (larger scale) [72,161]. Heterochromatin stiffens, while euchromatin softens the nucleus [72,161,162]. Stretch-induced decrease in heterochromatin utilizes its effect in nuclear softening to prevent DNA damage [10]. Increasing baseline heterochromatin renders the nucleus insensitive to deformation, therefore blocking the stretch-induced nuclear softening [10]. Differentiated or mechanically stretched bovine mesenchymal stem cells increase heterochromatin and nuclear stiffness, rendering them more mechanosensitive than undifferentiated cells [9]. Moreover, altered chromatin structure determines confined cell migration efficiency due to its impact on nuclear mechanics [59,163–165]. Together, these findings suggest that not only can chromatin structure change in response to mechanical perturbation, but also altered chromatin structure can affect nuclear mechanics and mechanosensitivity, which helps cells mechanically adapt to the environment. Interestingly, a new study suggests that transcription alone may contribute to regulation of nuclear shape, without affecting nuclear mechanics [166], adding an additional layer to how chromatin alteration can affect nuclear structure and integrity.

With new tools developed to force chromatin loop formation [87,167,168], or to mechanically drag chromatin loops in controlled directions and distances [88] (Box 1), future studies will finally be able to decipher how chromatin loops and/or TADs affect nuclear mechanics and mechanosensing, and how/whether this influences the biology of the organism in a transient or permanent manner.

C). Mechanically altering the cytoskeleton and the environment

How does the cell respond mechanically to alter its environment? The question has been well studied in the cell migration field, where cells respond to mechanical cues by actively deforming, realigning, remodeling, or degrading the ECM [64]. However, it is still unclear if chromatin structure alterations can affect the ability of cells to apply forces to the environment. It has been proposed that chromatin condensation (resulting in a smaller and stiffer nucleus) makes it easier for the actomyosin cytoskeleton to push or pull the nucleus during migration [169,170]. This mechanism may partly explain the observation that chromatin condensation facilitates cell migration, especially through tight spaces [59,101,163,165,169]. On the contrary, reduced heterochromatin increases actomyosin cytoskeleton stiffness as a compensatory response [171], which may lead to increased cellular stiffness that reportedly impairs tumor cell migration ability [172]. Moreover, chromatin condensation is required for differentiation of bovine mesenchymal stem cells and persistent pig myofibroblast activation, directly controlling collagen and fibronectin production, and thus contributes to ECM stiffness and tissue fibrosis [9,114]. Through this proposed inside-out mechanism, chromatin structure alterations can lead to altered cytoskeleton strength and ECM production capability, ultimately mechanically affecting the environment. The potential interplay and feedback between nuclear mechanics via heterochromatin and cellular mechanics via cytoskeleton [171] can have critical implications in the regulation of this inside-out mechanism.

D). Chromatin in crisis: chromosome mis-segregation, micronuclei, and DNA damage

A special source of large-scale chromatin structure alteration comes from chromatin in crisis. During mitosis, chromosomes are exposed to the cytoplasm, and are under tightly regulated spindle tension that determines correct segregation [141–143]. Altered number of spindle poles, or defects of spindle assembly checkpoint and cohesin can cause chromosome mis-segregation, which leads to aneuploidy (gaining or losing chromosomes) and micronuclei formation in daughter cells [173]. Mechanical compression that disrupts cell rounding during mitosis also leads to chromosome mis-segregation [67,174]. The mis-segregated chromosomes in micronuclei are vulnerable to excessive DNA damage and subsequent chromothripsis (massive chromosome rearrangement) [175,176], as the defective micronuclear envelope is prone to irreversible collapse [177]. Chromothripsis of micronuclear chromosomes can drive tumorigenesis [173,175]. Nuclear envelope rupture of micronuclei also triggers cGAS-STING signaling, driving confined migration of breast cancer cells and metastasis in mouse models [178]. Pharmacological induction of excessive DNA damage by doxorubicin alone can trigger EMT and tumor invasion of breast cancer cells [126]. Taken together, these findings suggest that large-scale structure alteration of chromatin in crisis induces long-term cellular changes, including tumorigenesis and cancer metastasis, which can indirectly affect cellular and tissue mechanics in the process.

In summary, chromatin structure is altered through an outside-in mechanotransduction pathway (Fig. 2). At the same time, an inside-out mechanical alteration of the cell and the environment controlled by chromatin structure is proposed (Fig. 3). This bidirectional mechanical feedback pathway can be found in a recent study, which shows that tissue fluidification/unjamming-induced heterochromatin formation and nuclear envelope rupture promotes breast cancer cell invasion, which presumably leads to further tissue fluidification [101]. The mechanical interactions between cells and the environment may have direct implications in tumor therapeutics focusing on tumor-stroma interactions (Box 3).

Box 3. Chromatin structure and mechanobiology in tumor-stroma interactions.

A tumor continuously interacts with and remodels its surrounding tumor microenvironment (stroma) for oxygen, nutrients, and biochemical and mechanical cues. This interaction and remodeling form a dynamic feedback loop [179] termed tumor-stroma interactions [180]. Tumor microenvironments consist of both ECM and a wide range of stromal cells, including fibroblasts, endothelial cells, and immune cells. Cancer cells adapt to mechanical stimulations from the tumor microenvironment via tumor-stroma interactions. An important mechanical stimulation comes from stromal stiffening, a fibrosis process mediated by cancer-associated fibroblasts (CAFs) via increased production and crosslinking of the ECM network [179,181,182]. CAFs are fibroblasts activated by TGF-β signaling in tumor stroma, making them constantly synthesize and remodel the ECM [183]. Moreover, CAFs exhibit a myofibroblast phenotype, which contract the tumor microenvironment [184]. Mechanical forces caused by stiffened ECM and cellular contraction further activates CAFs through increased focal adhesion signaling, leading to even more ECM production and crosslinking [185]. This mechanical positive feedback loop establishes a stiff tumor microenvironment that drives angiogenesis, tumor progression, and metastasis [179,181,182,186]. Can chromatin structure alterations, particularly chromatin condensation under stiff and confined environments [59,114–116] contribute to this mechanical positive feedback loop? Can this be a driving force for tumor invasion and metastasis, as implicated in tissue fluidification-induced chromatin condensation [101]?

Immunosuppression is the major barrier to effective tumor treatment [187]. Tumor-associated macrophages (TAMs) are macrophages stimulated and converted from the pro-inflammatory M1 polarization into the tumor-promoting and anti-inflammatory M2 polarization by the tumor microenvironment [188]. TAMs suppress T cell function and promote CD8⁺ T cell exhaustion (decreased proliferation and decreased effector cytokine production) [187]. Interestingly, mechanical stimulations can alter macrophage polarization. Spatial confinement impairs the normal HDAC3-dependent chromatin structure alterations that lead to M1 polarization, but instead drives the tumor-promoting M2 polarization and decreased phagocytosis, similar to the phenotypes of TAMs [189]. Is it possible that the stiffened and confined tumor microenvironment mechanically drives TAM formation through the same pathway?

Conclusion and future perspectives

With recent advances in mechanobiology assays and a deeper understanding of complex interactions within the 3D folded chromatin fiber, researchers are starting to bridge the gap between mechanobiology and chromatin structure. It is now becoming evident that not only does the environment mechanically alter chromatin structure, but chromatin structure can also mechanically affect the environment.

Many exciting challenges and outstanding questions remain. Distinguishing between the molecular mechanisms mediating force-induced chromatin structure alterations can be challenging, as extensive cross-talks exist among them. For example, nuclear deformation-mediated pathways easily get buried under integrin- and actomyosin-mediated ones. One solution would be to apply suitable bioengineering tools or biochemical assays to separately test the involvement of different pathways, such as the use of wedged AFM cantilever to selectively deform the nucleus or the cytoplasm [62], or the use of dominant negative LINC complex constructs to disrupt nucleo-cytoskeletal connection [5]. The bidirectional mechanical feedback pathway can also lead to cause-and-effect ambiguity. For example, chromatin condensation is observed upon initiation of tissue fluidification [101]. However, chromatin condensation occurs at the same time as tissue fluidification, which may also facilitate tumor invasion and further tissue fluidification [101]. So, is the observed chromatin condensation merely a passive “response”, or an active “driver” of this process? To clearly dissect its biological functions and its position in the bidirectional mechanical feedback pathway, further bioengineering, biochemical, and genetic manipulations are needed. Some other pressing open questions include: what is the contribution of ER force sensing in cellular and chromatin mechanoresponse, if such mechanism exists? How is chromatin structure alteration in “heterochromatin buffering” restricted within heterochromatin regions? It is interesting to consider how active retroviral invasion into the genome [190] might affect chromatin structure as well.

Finally, the vastly different cellular behaviors obtained from 2D and 3D environments, and the extensive cell-cell/cell-matrix interactions in physiological conditions, have pushed the field toward using in vitro assays mimicking in vivo environment (microfluidics tissue-on-a-chip and 3D bio-printing [126,191,192]) or using actual tissue and mouse models for mechanical probing [10,113]. Similarly, recapitulating tumor-stroma interactions and tumor tissue mechanics is critical for future cancer mechanobiology research. Investigating the role of chromatin structure alterations in tumor mechanoresponse in the tumor microenvironment could be a novel route for therapeutic development. In the end, the mechanosensing force is strong with this epigenome, and we must continue to strive to understand it, while potentially harnessing its power for therapeutic purposes. To paraphrase the iconic words of Grand Master Yoda in The Empire Strikes Back [193], sensing the force can be a powerful means to understand and target biological processes occurring inside the cell and in other places, from within the chromatin structure, to beyond the extracellular environment. Seeing through the lens of mechanobiology helps us better understand evolution, and sheds light on future research. “Through the Force, things you will see. Other places. The future. The past.” [193]

Highlights:

• Mechanobiology in the chromatin context shares conserved features across species.

• Chromatin structure is altered via an outside-in mechanotransduction pathway.

• Chromatin structure mechanically alters the environment via an inside-out pathway.

• Bidirectional mechanical feedback of chromatin may affect tumor-stroma interactions.

Abbreviations

ECM

extracellular matrix

cAMP

cyclic adenosine monophosphate

c-di-GMP

bis-(3’-5’)-cyclic dimeric guanosine monophosphate

MAPK

mitogen-activated protein kinase

LINC

linker of nucleoskeleton and cytoskeleton

YAP

yes-associated protein 1

TAZ

transcriptional coactivator with PDZ-binding motif

cPLA₂

cytosolic phospholipase A₂

H3K9me2/3

di-/trimethylated histone H3 lysine 9

H3K27me3

trimethylated histone H3 lysine 27

H3K4me3

trimethylated histone H3 lysine 4

H3K9ac

acetylated histone H3 lysine 9

H3ac

acetylated histone H3

ER

endoplasmic reticulum

LAD

lamina-associated domain

TAD

topologically associating domain

CTCF

CCCTC-binding factor

AFM

atomic force microscopy

MNase

micrococcal nuclease

DNase I

deoxyribonuclease I

ATAC-seq

assay for transposase-accessible chromatin with sequencing

ChIP

chromatin immunoprecipitation

qPCR

quantitative real-time polymerase chain reaction

CUT&RUN

cleavage under targets and release using nuclease

cryo-EM/cryo-ET

cryogenic electron microscopy/tomography

cGAS

cyclic GMP-AMP synthase

STING

stimulator of interferon gene

EMT

epithelial-to-mesenchymal transition

CAF

cancer-associated fibroblast

TGF-β

transforming growth factor-beta

TAM

tumor-associated macrophage