Microscopy methods to visualize nuclear organization in biomechanical studies

Abstract

The mechanical environment plays an important role in influencing cell identity. The nucleus’s organization and mechanical state are essential regulators of cellular function. However, open questions remain about the mechanisms underlying how the physical microenvironment influences nuclear mechanics and organization to drive specific transcriptional and epigenetic shifts. Understanding how biophysical cues change cell behavior provides groundwork to improve medical technologies such as tissue engineering, stem cell therapy, and mitigation of aberrant cell behavior. Microscopy is an indispensable tool that noninvasively explores the cell’s nuclear state, providing valuable measurements on features including nuclear morphology, nuclear mechanical properties, protein localization, and genomic organization. In this review, we discuss notable imaging techniques, such as super-resolution microscopy, examples of how they have recently advanced the field, and how they can further our knowledge of the interplay between nuclear mechanoregulation and cell function.

Introduction

Cells must respond to a broad spectrum of mechanical inputs. External forces can be active and dynamic (e.g., sheer flow, compression, and stretch) or passive and relatively time-invariant (e.g. differences in tissue stiffness) [1]. Several studies have demonstrated that cells sense and react to mechanical perturbations and adapt their phenotype in response to their mechanical environment [1–4]. In the past, studies have focused on pathways where cytoplasmic signaling propagates into the nucleus and incites shifts in gene expression to ultimately produce functional changes [3]. However, the physical environment can also directly affect the nucleus [4,5]. Rather than biochemical signals alone, emerging perspectives emphasize the mechanical state of nuclear components as a key regulator of gene expression and cellular function [2,3,6].

Growing evidence indicates that the nucleus plays a central role in registering the mechanical environment by adapting its structure and genome organization. Physical forces either propagate through the cytoskeleton, which is coupled to the nucleus, or directly deform the nucleus [7]. Mechanical events often correlate with chromatin remodeling and transcriptional changes, but the mechanisms by which cells coordinate specific transcriptional responses based on diverse physical cues are unclear [8–10].

The relationship between biophysical cues, nuclear organization, and cellular function has important implications in developmental, physiological, and pathological processes. Given that biophysical factors can influence cell fate, there is a clear incentive to discern how they influence the nucleus to alter cell function. A better understanding of this relationship will not only offer insights into cell regulation pathways, but also provide insight into poorly understood disease mechanisms and innovations in regenerative medicine [1,3,7,11].

Several tools have been developed in the recent past that allow us to interrogate the transcriptional, epigenetic, and chromatin state of cells. RNA-Seq, ChIP-Seq, ATAC-Seq, Hi-C, and other sequencing-based methods have gained popularity because they provide high-throughput and unbiased information on transcriptional, epigenetic, and chromatin spatial changes in response to perturbations [12–16]. Yet these methods are often invasive and fail to maintain the spatial context of cells and sub-cellular structures, restricting the ability to observe the pattern of force propagation and response throughout a cell. Additionally, these approaches typically require hundreds of thousands of cells and cannot provide information at the single cell level. Another drawback is the high level of difficulty in combining multiple sequencing-based methods together to obtain multimodal transcriptomic, epigenetic, and chromatin information.

Microscopy based methods, although lower in throughput, can address these limitations and provide complimentary information to sequencing approaches. They can non-invasively survey biological structures at multiple levels of organization ranging from individual molecular localizations to cell organization within tissues. Thus, microscopy is a well-suited tool to probe the nucleus and study the relationship between the nuclear and mechanical state of the cell. In recent years, exciting imaging advancements have been developed, such as super-resolution microscopy as well as spatial transcriptomics and genomics imaging methods [17–19], which possess high potential for investigating mechanoregulation in cells. In this review, we provide a brief overview of current knowledge on mechanoregulation. We then focus on imaging strategies especially designed to observe the nucleus under different mechanical contexts or those that we envision will be useful for future investigations into the link between a cell’s mechanical environment and its gene expression.

Mechanoregulation and mechanotransduction

To respond to their environment, cells must be capable of sensing, transmitting, and producing forces. Although an extensive discussion of these pathways is beyond the scope of this review, we highlight some key components of mechanoregulation (Figure 1) to better underscore how microscopy can further the field (for more comprehensive Reviews on mechanotransduction, we refer readers to the following [1,7,27–29]).

Figure 1. Nuclear organization and mechanotransduction.

Mechanical signals are sensed by the cell through focal adhesions and the cytoskeleton [20]. The nuclear envelope is a double lipid bilayer membrane comprised of the outer nuclear membrane (ONM), inner nuclear membrane (INM), and the perinuclear space (PNS). The lamina is a network of lamin proteins that gives the nuclear envelope structural integrity. Nesprins and SUN-domain proteins physically couple the lamina to the cytoskeleton, forming the linker of nucleoskeleton and cytoskeleton (LINC) complex. This complex transmits forces experienced by the cell to the nucleus. In addition to this mechanism, biochemical signals initiated from the cytoplasm in response to mechanical cues can propagate to the nucleus through the nuclear pore complex (NPC) that spans the nuclear envelope. MRTF-A, YAP/TAZ, and β-catenin are notable examples of proteins that localize to the nucleus depending on the cell’s mechanical state to induce changes to gene expression [21,22]. These factors can be regulated via other nuclear proteins, such as the INM protein, emerin, or nuclear actin. Epigenetic changes to chromatin, a complex of DNA wrapped about histone proteins, act as another means of regulating cell identity. Chromatin links to the lamina at lamina-associated chromatin domains (LADs), often as compacted, transcriptionally inactive heterochromatin via proteins such as lamin B receptor. Conversely, transcriptionally active, open euchromatin, typically localizes at the interior of the nucleus. Epigenetic modifications, such as histone acetylation or methylation, or large-scale chromatin reorganization influences the cell’s transcriptional activity. The structure of chromatin is sensitive to changes in lamin proteins [23]. The activity of some histone-modifying enzymes such as HDAC3 and several chromatin remodelers are sensitive to binding to emerin [24] or changes in nuclear actin [25,26]. Created with BioRender.com.

In mammalian cells, the nucleus is the largest and stiffest organelle [1]. The nucleus is encapsulated by the nuclear envelope or membrane, a double lipid bilayer composed of the outer and inner nuclear membranes (ONM, INM). The 30–50 nm perinuclear space (PNS) spans between these two membranes [28]. Channels known as nuclear pore complexes (NPCs) selectively transport cargos through the nuclear envelope [30]. The linker of nucleus and cytoskeleton (LINC) complex physically couples the nucleoskeleton to the cytoskeleton to transmit physical forces experienced by the cytoskeleton directly to the nucleus [2]. At the cytoskeletal interface of the complex, the KASH domain of nesprin family proteins binds to SUN domain proteins, which in turn bind to the nuclear lamina [2]. The lamina resides within the INM and is formed from a thick network of fibrous A/C-type or B-type lamin proteins that reinforce the nuclear envelope [31,32]. Disruption to lamin proteins not only results in altered nuclear structure but also induces changes to cellular identity and altered mechanosensing, suggesting that these proteins play an essential role in regulating the genome response to mechanical environments [33,23].

Within the nucleus lies chromatin, the complex of DNA organized by histone proteins into units known as nucleosomes. Chromatin can be marked either through modifications on DNA (e.g., DNA methylation) or on the histone proteins (e.g., histone tail methylation or acetylation) that ultimately impact the organization of the complex [34]. At a higher level of organization, chromatin collects into discrete compartments with separate boundaries known as topologically associated domains (TADs) that can feature differing levels of transcriptional activity [35]. Depending on the epigenetic marks at a given genomic or regulatory region, chromatin can be in an open, transcriptionally active state known as euchromatin or a closed, transcriptionally silent state known as heterochromatin. Heterochromatin can be bound to the lamina forming lamin-associated chromatin domains (LADs) [36]. These regions play an important role in silencing genes belonging to alternative lineages during development [2,37]. Epigenetic modifications have been demonstrated to be sensitive to changes in cell and tissue geometry [38,39]. Likewise, studies observing altered chromatin organization and cell fate decisions in response to perturbations to the lamina, as well as different mechanical loading conditions, support the concept that the LINC complex plays an essential role in mechanosensing and that the disruption of chromatin organization can lead to aberrant cell behavior [33,23]. Proteins anchored to the INM such as emerin [24,40,41] and lamin B receptor [36], as well as lamin proteins [23], are thought to regulate anchoring of peripheral heterochromatin to the nuclear lamina in response to mechanically regulated signals.

Proteins traditionally associated with the cytoskeleton, such as actin, myosin, Arp2/3, and WASP family proteins, additionally regulate several key nuclear processes, including transcription regulation, chromatin remodeling, splicing, and DNA damage repair [25,42–45]. Depletion of β-actin in the nucleus leads to disruption of compartment-level genome organization [46]. Nuclear actin binds with chromatin remodeling complexes (e.g., BRG1 [47], BAF [46,48], INO80 [49], and TIP60 [50]), often leading to altered remodeling activity [25]. Nuclear actin can also regulate transcription factors, one notable example being myocardin-transcription factor A (MRTF-A, also known as MAL or MKL1). Actin binding to MRTF-A blocks its nuclear localization signal [21], such that MRTF-A becomes concentrated in the cytoplasm. Upon mechanical or mitogenic stimulation, actin polymerizes, reducing levels of monomeric actin. Subsequently, MRTF-A travels through NPCs and localizes to the nucleus where it activates target genes [43]. By regulating actin dynamics, lamins, and emerins comprise another layer of mechanoregulation to MRTF-A activity [51]. Actin dynamics are modulated by other nuclear proteins, such as those in the LINC complex or INM [51]. Yes-associated protein (YAP)/transcriptional coactivator with PDX-binding motif (TAZ) are another family of mechanosensitive transcription factors that are shuttled between the cytoplasm and nucleus to activate select genes depending on mechanical cues [52,22]. Rather than directly binding to actin, they are regulated by actin binding proteins, where their regulation is driven by cytoskeletal reorganization. β-catenin is another mechanosensitive factor involved in several complex pathways [53], one of which is the same signaling pathway as YAP/TAZ [54]. To illustrate the distinct responses elicited by parallel pathways, a recent study revealed that β-catenin, accompanied by nuclear actin, translocated to the nucleus specifically under dynamic strain, rather than static strain. Conversely, nuclear YAP1 levels changed more dramatically with static strain, a response more commonly linked with the reorganization of the cytoskeleton [55]. Lastly, mechanosensitive signals like nuclear actin [42,56] influence the activity of nuclear bodies such as the nucleolus, the RNA-processing center of the nucleus, as well as numerous nucleoplasmic or chromatin-associated RNA and proteins that contribute towards the function of the organelle [2,42,56]. The multifaceted role of actin, just one of the many proteins involved in mechanotransduction, illustrates the complexity of interaction across multiple signaling pathways that must be detangled.

Over time, progress has been made in uncovering mechanosensitive pathways and notable factors regulating chromatin organization and transcription. Although some key regulatory proteins have been identified, a full evaluation of how different mechanical forces influence these relationships is needed. In addition, a complete understanding of how cells organize their chromatin for long-term behavior is still lacking and the direct link between the mechanical force and many downstream transcriptional responses remains obscured [9,25,52,57]. To fully decipher how these players coordinate with each other to integrate mechanical signals to achieve a selected behavior, future experimental approaches must be able comprehensively interrogate the nucleus in response to different mechanical cues.

Immunocytochemistry to capture broad changes in the nucleus and cell in response to mechanical stimuli

A straightforward approach to observe chromatin density is with DNA-binding dyes such as Hoechst or DAPI. The staining pattern of the nucleus provides a measurement of the distribution of heterochromatin versus euchromatin, which can then be compared for cells under different treatments. This approach was used to demonstrate that dynamic tensile loading in mesenchymal stem cells (MSCs) rapidly induces chromatin condensation, and that this response is dependent upon cytoskeletal contractility [58]. Nuclear morphology can also be measured from labeling nuclei. Based on changes in the nuclear aspect ratio of differentiating MSCs, it was observed that differentiation attenuates stretch-induced nuclear deformation [59]. Moreover, single cell chromatin and nuclear features captured from images of stained nuclei were sufficient single-cell biomarkers for phenotypic mapping of tumor progression [60].

Examining other cellular components besides the nucleus provides additional information on mechanical interactions taking place within the cell. Using confocal imaging of cytoskeleton polymerization, chromatin remodeling, and histone acetylation, one study evaluated the influence of matrix mechanics on MSC paracrine activity (i.e., the secretion of signalling factors to nearby cells that promote angiogenesis), which is hypothesized to be an important therapeutic benefit for bone regeneration therapy [61]. Cells cultured on dynamic stiffening materials presented higher degrees of paracrine activity, cytoskeletal polymerization, nuclear deformation, and chromatin remodeling than those cultured on static stiffness material. The findings suggested that the uninterrupted deformation of the nucleus and histone acetylation regulation on dynamic substrates activates chromatin remodeling, promoting angiogenesis-related genes.

A better understanding of nuclear actin was also achieved using fluorescent constructs. Initially, the decreased presence of actin filaments in the nucleus compared to the cytoplasm and a poor performance of traditional actin stains on nuclear filaments led to the belief that canonical actin filaments do not exist in the nucleus [25,62]. However, the development of probes comprised of actin-binding domains fused to nuclear localization signals enabled study of nuclear actin dynamics [63,64]. These probes exemplify the additional information offered by the ability to appropriately image targets of interest.

Conventional microscopy of fluorescent labeled targets is a straightforward, accessible, and versatile strategy. However, due to the low spatial resolution of such methods, capturing changes that occur at length scales relevant for gene activity is not possible. More sophisticated methods can provide deeper insights into changes within the nucleus and cell behavior.

Microscopy methods to measure deformation, strain, and forces

Deformation microscopy

Deformation microscopy is a specialized technique for mechanobiology studies that also employs fluorescently labeled nuclei. Cells are imaged throughout a deformation event in vivo and changes in fluorescence intensity throughout the image sequence are used to generate a displacement map. Using this map in combination with physical models of the material properties of the nucleus, computational analysis extracts relative measurements on the different strains that heterochromatin and euchromatin regions experience (Figure 2a). A study employing deformation microscopy observed that cells with disrupted LINC complexes, lamin A/C deficiencies, or grown in different culture conditions present abnormal strain distributions in their nuclei [65]. In a follow-up study, the method was used together with computational models of cells, generated from confocal microscopy z-stacks, to ascertain the most influential parameters contributing to differences in nuclear strain patterns experienced by chondrocyte cells [69]. Not only did this approach reveal that late-passage cells showed abnormal strain in particular regions of the nucleus, but the parametric analysis was able to attribute these changes to the dedifferentiated cells’ altered 3D aspect ratio. Thus, the combination of deformation microscopy and computational modeling can identify major sources of different strains (e.g., local shear, compressive, tensile, and second deviatoric) within the cell and nucleus and could serve as a useful tool to explore other cellular systems susceptible to aberrant strain patterns.

Figure 2. Fluorescence microscopy to measure deformation, tension, and biomolecular condensates.

(a) Deformation microscopy overview. A cell nucleus labeled with DNA-binding dyes is imaged throughout a deformation event. Computational analysis of these images generates a map of the deformations throughout the nucleus that is then used to calculate strain magnitudes and directions. Figure adapted from Ref. [65]. (b) A conventional AFM setup. The sample is placed on a piezoelectric tube to move a sample relative to a cantilever. The deflection of the cantilever is measured via an optical system where a laser is deflected from the cantilever and sensed by a photodetector [66]. (c) The FliptR probe. The excitation maxima and fluorescence signal lifetime of the reporter is dependent upon the degree of alignment between the two fluorescent groups, where the mean fluorescence lifetime of the reporter is greatest when the groups are planarized (i). This serves as a readout for properties of the membrane, such as distinguishing from a liquid-disordered membrane (green) to a liquid-ordered membrane (red), which is influenced by the composition of the membrane (ii). Structural changes in the membrane with increased tension cause FliptR molecules to change their degree of planarization. Lipid decompression leads to deplanarization while tilting of membrane lipids results in increased planarization. Lastly, phase separation of different membrane lipids can produce discrete regions of FliptR molecules with either high or low lifetimes (iii). Figure adapted from Ref. [67]. (d) CasDrop overview. Cas9 scaffolds target genomic regions of interest. Exposure to blue light drives IDR-containing proteins to bind to these scaffolds, promoting the formation of biomolecular condensates. Figure adapted from Ref. [68]. Partially created with BioRender.com.

Traction force microscopy

Like deformation microscopy, traction force microscopy (TFM) [70] is a well-established method to measure the forces that adherent cells exert on their microenvironment. Deformations transmitted from the cytoskeleton onto the extracellular matrix via focal adhesions are imaged through the displacement of fluorescent beads embedded in the substrate [71]. Analyzing these displacements subsequently allows for computational modeling of strain and stresses. The measurement of cell traction forces is especially amenable for investigating mechanotransduction through cell–matrix interactions. In the past, studying the early response of cells to substrate-based stretch was impractical due to sample drift caused by thermal fluctuations or cell migration. Recently, a real-time, three-dimensional tracking approach with sub-second time resolution has been developed to measure tensile mechanical properties of adherent cells immediately following substrate deformation to capture transient subcellular forces and nuclear deformations. This was used to measure the traction force exerted by cells seeded on a substrate undergoing time-controlled stretch and was paired with imaging of Hoechst-stained nuclei for morphology measurements [72]. The results revealed that adherent cell nuclei respond depending on the applied stretching rate. The study demonstrated that this behavior was not a consequence of focal adhesion remodeling nor regulation of contraction, as the traction between cells stretched at slow and fast rates was observed to be comparable. These changes in the nucleus’s mechanical properties were proposed to be a mechanism to protect genomic material from physical disruptions in a time-dependent manner. Despite its utility for understanding the interaction of a cell and its physical environment, it is important to note that TFM alone is only able to indirectly assess nuclear deformations as the technique relies on measuring interactions that are propagated through the cytoskeleton and focal adhesions, rather than the nucleus itself.

Atomic force microscopy

Atomic force microscopy (AFM) is a popular approach for cellular mechanical studies. A microcantilever force sensor connected to an optical system (Figure 2b) either measures the viscoelastic properties of cells or applies forces onto the cell. By precisely measuring or controlling forces and strains, this tool can map the viscoelastic properties of a cell with high spatial resolution but is low throughput and not amenable for population-based studies [66]. Using a needle penetration technique, it is possible for AFM to characterize the mechanical behavior of the nucleus [73]. Based on nuclear elasticity AFM measurements, achieved by indenting the nuclear surface through the cell cortex, Wickstrom et al. demonstrated that nuclear stiffness decreases under stretch conditions followed by a recovery of elasticity [74]. Using the same procedure, this study also showed that F-actin depolymerization did not impact nuclear stiffness, but lamin A depletion strongly reduced stiffness. Thus, AFM can interrogate elastic properties of the nucleus.

However, conventional AFM cannot isolate the properties of individual nuclear components. AFM-LS combines AFM with light sheet fluorescence microscopy (LSFM) to address this issue [75]. LSFM achieves high speed, high contrast, low phototoxicity 3D imaging of live samples. A laser light sheet illuminates only a thin section of the sample that reduces both phototoxicity and background signal while scanning larger areas in the same time frame compared to point-scanning methods such as confocal microscopy [76,77]. By coupling the two technologies, AFM-LS simultaneously achieves volumetric imaging with high spatiotemporal resolution of the cross-section of the cell (perpendicular to the plane of the slide) as the atomic force microscope dynamically applies forces within a range from 10 pN to 100 nN [75]. Unlike AFM alone, AFM-LS can capture fluorescence images of the cell volume from which the strain on the nuclear surface area can be analyzed separately from the nuclear volume. This tool was applied to cells treated with either siRNAs targeting lamin A/C for knockdown or Trichostatin A (TSA), a histone deacetylase inhibitor which reduces chromatin compaction by promoting histone tail acetylation. A comparison of wild type to treatment groups demonstrated that as cells underwent compression, chromatin decompaction weakened resistance to nuclear volume changes while lamin A/C depletions increased deformations to nuclear surface area [78]. Thus, AFM-LS is a new and powerful tool for volumetric, live-cell imaging of nuclear dynamics under mechanical perturbations.

Tension biosensors

Fluorescence resonance energy transfer (FRET) based biosensors measure forces experienced by proteins in live cells [79,80]. These biosensors comprise of a donor and acceptor fluorophore. When the excited donor is within 1–10 nm of the acceptor, non-radiative energy transfer excites the acceptor, which emits the observed wavelength [80,81]. For mechanobiology studies, FRET systems can signal conformational changes in complexes and measure mechanical forces across linkers connecting the donor to acceptor [80]. In one example, FRET tension sensors were used to reveal a direct mechanotransduction pathway regulating cell fate in keratinocyte differentiation [82]. A FRET system comprised of the mTFP and Venus fluorophores connected by an elastic linker was inserted between the cytoplasmic and transmembrane domain of mini-Nesprin-2 in the LINC complex. This system examined tension across the LINC complex in β1 integrin null keratinocytes as well as keratinocytes grown in colonies either at the periphery where cells undergo cell–matrix adhesions and integrin engagement or at the interior, where cells only engage in cell–cell adhesions. From both experiments, the FRET signal indicated decreased tension across the LINC complex with reduced integrin engagement, regardless of actomyosin contractility levels. Combined with other experiments, this FRET data supports a model where a β1 integrin-engaged actin network transmits tension to the nucleus through the LINC complex and represses epidermal differentiation complex. Hence, FRET systems are especially useful for mechanotransduction studies to visualize tension across select molecular targets that are responsible for transmitting forces throughout the cell or nucleus.

The fluorescent lipid tension reporter (FliptR) is designed to monitor membrane tension at the molecular level in live cells (Figure 2c) [83,67]. The probe features two planar, fluorescent groups that can twist relative to each other, modulating the molecule’s excitation maxima and fluorescence lifetime, or the duration a fluorophore is in the excited state prior to photon emission, which is measured via fluorescence lifetime imaging microscopy (FLIM) [67,84]. Membrane tensions can then easily be quantified based on the linear relationship between the fluorescence lifetime and membrane tension. This system was used to image the stretched perinuclear membrane and showed that mechanical stretch reduces nuclear membrane tension and alters the membrane state [74].

Advanced microscopy methods for visualizing nuclear bodies, regulatory factors, and chromatin

Single molecule tracking

Single molecule tracking (SMT) utilizes a bright, stable fluorescent probe to follow a particle’s trajectory [85]. Typically, a nanomolar concentration of the protein of interest is labeled with a live-cell compatible tag such as Halo-tag. This labeling strategy enables a time course of a sparse population of the protein of interest such that their images are isolated and the center position of the image (corresponding to the protein’s position in space) can be determined with high precision, providing highly resolved spatial and temporal information on local dynamics. Ideal SMT conditions require low fluorescent background and bright, stable fluorophores that do not photobleach quickly [86]. This technique can characterize the kinetics of nuclear proteins and machinery and has brought many developments to the understanding of chromatin dynamics [87]. Live-cell SMT of the Poly-comb repressive complex 1 (PRC1) demonstrated that this transcription repressor sparsely binds to target sequences and is highly dynamic [88]. This work also revealed that altered histone modifications influences the binding frequency of the repressor to chromatin. SMT was further used to measure the dynamics of core histone proteins in different cell types [89,90], showing that chromatin dynamics is cell type dependent, and to determine the accessibility of different transcription factors to low or high mobility chromatin states [91]. Multiple particle tracking (MPT) extends SMT to follow several particles at once. In a study on acute lymphoblastic leukemia, MPTof high-density chromatin regions in Hoechst-stained nuclei revealed that high-risk leukemia cell nuclei exhibited increased viscosity than those of healthy lymphocytes [92]. Thus, live-cell particle tracking approaches could be used to either investigate nuclear protein dynamics as cells are mechanically perturbed or to compare the mechanical properties of nuclei derived from different cell types.

Fluorescent reporter systems for detecting biomolecular condensates

Cells can be engineered such that fluorescent probes act as signals for specific epigenetic or cell state changes. Several such probes have been developed and applied to study nuclear mechanobiology. More recently, specialized probes are being utilized to study a new and emerging concept in cell organization: liquid–liquid phase separation (LLPS) [93]. Intrinsically disordered regions (IDRs) in proteins mediate the formation of a diverse class of membrane-less mixtures of proteins, RNA, and DNA known as biomolecular condensates [94]. Nuclear condensates have been shown to influence spatial nuclear organization and transcriptional activity [95,96]. The CRISPR-Cas9-based optogenetic technology CasDrop was developed to investigate how physical forces impact the formation of condensates at specific genomic loci [68]. Exposure of a cell to blue light drives mCherry-labeled IDR-containing proteins to assemble into a multimeric protein complex at genomic loci targeted by GFP-containing Cas9 scaffolds, resulting in condensate formation. This enables the visualization of condensate formation through the colocalization of mCherry signal to GFP on the seeding sites. Imaging the system with confocal microscopy revealed that IDR-driven droplets preferentially form in regions of low chromatin density and mechanically exclude chromatin not bound to the Cas9 scaffold (Figure 2d). A similar technology, Corelets, demonstrated that a chromatin-dense environment directly inhibits condensate growth dynamics and leads to anomalous coarsening of droplets [97]. The sensitivity of condensates to their mechanical environment also suggests they could be used as a readout of the viscoelastic properties of the nucleus. Thus, microscopy combined with novel engineered probes is a powerful approach in investigating these nuclear components and their role in the mechanical state of the nucleus. For an extensive review on imaging methods for biomolecular condensates, see Ref. [98].

Super-resolution microscopy

Although far-field light microscopy is a practical tool for a variety of biological applications, the diffraction of light limits spatial resolution to ~200–300 nm. Many supramolecular nuclear structures form below this threshold, including chromatin fibers (as ~5–24 nm chains) and the lamina (as a ~10–30 nm thick network) [31,99–101]. Super-resolution microscopy techniques can overcome this barrier. Super-resolution microscopy was first demonstrated in 2000 by the development of simulated emission depletion (STED; Figure 3a) [102], followed by the development of Single Molecule Localization Microscopy (SMLM)-based techniques [104] that include Stochastic Optical Reconstruction Microscopy (STORM; Figure 3b) [105], (fluorescence) Photoactivated Localization Microscopy (f/PALM; Figure 3b) [106,107], and DNA point accumulation in nanoscale topography (DNA-PAINT; Figure 3c) [108]. Depending on the modality and the application, spatial resolution ranging from ~10 to 50 nm can be achieved [17]. Many of these approaches have recently been combined to push the spatial resolution near the molecular scale. For example, MINFLUX and MINSTED combine the concept of STED with single molecule localization [111,112]. ROSE, SIMFLUX, ModLoc similarly combine the concepts of structured illumination with single molecule localization [113–115]. Finally, Expansion Microscopy (ExM) is a conceptually distinct super-resolution modality that achieves high resolution through isotropic and physical expansion of the sample (Figure 3d) [109,110]. As such, expansion microscopy does not require specialized microscopes and is compatible with wide-field, confocal or light sheet microscopy while achieving a resolution of ~70 nm in its most standard form. For a detailed review on these methods, we refer the reader to Ref. [17].

Figure 3. Notable examples of super-resolution microscopy techniques.

(a) In STED, the excitation beam is co-aligned to the donut-shaped STED beam such that fluorophores exposed to the STED beam undergo stimulated emission (i). This leaves the region in the STED beam’s center to emit observable fluorescence and effectively the point spread function of the illumination is reduced (ii). The field of view is scanned with the excitation and STED beams to produce a super-resolved image [102]. Figure adapted from Ref. [103]. (b) The premise of STORM/PALM relies on fluorescent probes bound to the target, which switch between a light-emitting and dark state. A series of images that capture the blinking fluorophores is recorded. The sparse distribution of active fluorophores in each frame enables the precise localization of individual molecules. A range of photoswitchable, photoconvertible, or photoactivatable organic fluorophores or fluorescent proteins are available for STORM/PALM imaging [104–107]. Figure adapted from Ref. [103]. (c) DNA-PAINT labels targets with short DNA oligonucleotides (docking strand; P₁, P₂, … P_n) that are complementary to free floating strands bound to fluorophores (imager strand; P₁*, P₂*, … P_n*; i). Imagers transiently bind to the docking strand, becoming immobilized and allowing the target to be localized (ii). Multiple targets can be labeled with different oligonucleotides and visualized one at a time based on the imager included in the buffer. The buffer can then be washed out for the next target. The result is a set of localizations with high spatial resolution of multiple targets (iii). Figure adapted from Ref. [108]. (d) In ExM, samples are embedded in swellable hydrogels that can isotopically expand. The increased distance between targets enables microscopy methods to achieve higher spatial resolutions [109,110]. Figure adapted from Ref. [110].

Super-resolution microscopy techniques have been extensively used to study nuclear processes and structures. STORM imaging showed that nucleosomes assemble into heterogeneous groups along the chromatin fiber, termed “nucleosome clutches” [99]. Combined with novel quantitative analysis methods, STORM imaging of histones and DNA can give information on both the nanoscale and mesoscale chromatin compaction [116–118] of histone proteins and histone marks (e.g., H2B, H3K4me3, and H3K27me3). One study used this approach to image histone proteins in adult meniscal fibrochondrocytes and MSCs under TSA treatment, revealing not only that the treatment led to smaller H2B nanodomains (consistent with less chromatin folding and a more open chromatin state), but also higher levels of H3K9 acetylation and no significant changes in H3K27me3 levels [119]. This analysis of chromatin state was also paired with an evaluation of deformation and cell migration, finding that although TSA treatment did not change overall cell morphology, contractility, or migration across 2D substrates, these cells experienced increased nuclear deformability via chromatin relaxation enabling improved cell migration through micron-sized pores. In another example, STORM was used to determine that chemo-mechanical changes like substrate stiffness, oxygen availability, and pro-inflammatory cytokines result in abnormal nanoscale and mesoscale chromatin remodeling and changes in methylation in human tenocytes and MSCs [120]. Aside from histones, PALM and STORM imaging have also been applied to study nuclear F-actin populations at mitotic exit [121].

The concept of STORM/PALM can also be modified to achieve super-resolution imaging of protein dynamics in live cells, a technique known as single-molecule tracking photoactivated localization microscopy (sptPALM) [122]. This method has demonstrated that emerin monomers can form local nanoclusters, whose properties can adapt in response to mechanical challenges [40]. Live-cell PALM imaging of RNA Polymerase II (Pol II) revealed enhanced Pol II clusters upon serum stimulation in a nuclear actin-dependent fashion. Two-color STORM of Pol II and nuclear actin further showed that nuclear actin filaments colocalize with Pol II clusters during serum stimulation [56]. SMLM techniques have provided many other insights into Pol II transcription dynamics and transcription factor kinetics [123]. Another group visualized the higher-order chromatin dynamics by multiplexing PALM chromatin imaging and single nucleosome tracking in live mammalian cells [89]. Although this system was originally applied to chromatin behavior throughout the cell cycle, this and other super-resolution methods could be adapted for high resolution imaging of chromatin dynamics as cells undergo active mechanical forces.

ExM was applied to neutrophil cells, which feature a particularly small, heterogeneous, and dynamic nucleus in addition to chromatin bound to various antimicrobial peptides and enzymes. During an immune-defense mechanism known as neutrophil extracellular traps (NETs), the chromatin decondenses to the point where the nuclear envelope ruptures and eventually emerges from the cell to target pathogens. Super-resolution imaging via ExM was used to observe increased spatial information on various nuclear components including chromatin, histone H1, lamins, and the protein nucleophosmin, providing new insights into how these components interact throughout NET formation [124]. This application of ExM to neutrophils demonstrates the advantage of using high resolution techniques to better study complex and dynamic nuclear environments, which could be of great relevance in future studies examining nuclear properties and responses to mechanical cues.

Sequential and multiplexed fluorescence in situ hybridization for spatial genomics and transcriptomics studies

Fluorescence in situ hybridization (FISH) is a technology that visualizes nucleotide sequences (RNA or DNA) through hybridization and imaging of probes to visualize RNA transcripts or genomic sequences [125,126]. FISH has proven a useful tool for examining the spatial organization of a set of DNA or RNA sequences. For instance, DNA-FISH can compare the chromatin organization of cells in different mechanical environments. One study investigated the effect of substrate surface topography on MSC differentiation and found through FISH staining of chromosomal territories that topography stimulates changes in nuclear organization, which correlates with alterations in gene regulation [127]. In another example, single-cell sequencing was complemented with RNA-FISH imaging to study homogeneous populations of non-small-cell lung carcinoma transitioning to heterogeneous subpopulations upon physical compression [128]. The RNA-FISH data measured changes in transcription rate and variability in gene expression and validated a numerical model predicting the transition of single cells towards either an epithelial or mesenchymal cell fate. RNA-FISH has also been applied to study the expression level of chondrogenic genes at the single cell level as they undergo dedifferentiation when expanded in vitro or in 3D collagen gels. Combining RNA-FISH labeling with the labeling of cell matrix proteins in single cells revealed that chondrocyte passaging in vitro forms heterogenous subpopulations whose extracellular matrix production cannot be predicted by canonical marker genes [129]. Moreover, redifferentiation by passaging dedifferentiated chondrocytes in gels demonstrated that absolute transcript levels of aggrecan, the major proteoglycan in articular cartilage, remains relatively constant throughout both dedifferentiation and redifferentiation. RNA-FISH of aggrecan was also used to study the inhomogeneity of fibrocartilage [130]. This technology was used to image fibrous micro-domains (FmDs) and non-fibrous proteoglycan-rich micro-domains (PGmDs) that occur in fibrocartilaginous tissue constructs, revealing that PGmDs feature a significantly higher aggrecan to GAPDH mRNA ratio than FmDs. Thus, RNA/DNA-FISH methods are fruitful tools in exploring the relationship between single cell transcription or genomic state correlated to different mechanical environments or identify differences in the transcriptome or genome organization of heterogeneous cell populations featuring different mechanical properties.

Oligopaint is a DNA-FISH method that uses small (32mer) oligonucleotides with genomic homology to label a wide range of genomic regions ranging from small genes to entire chromosomes [131]. Super-resolution microscopy can be used together with Oligopaint, in an approach named OligoSTORM [132]. In contrast to previously described super-resolution microscopy studies that looked at chromatin globally, OligoSTORM is capable of mapping specific genomic sequences along chromosomal structures [132,133]. For instance, one study leveraged OligoSTORM to compare 46 epigenetic domains that were classified as active, inactive, or Polycomb-repressed based on histone modifications and regulatory protein enrichment [134]. Probes bound to targeted genomic regions were imaged through conventional fluorescence microscopy and STORM, where the latter option offered more structural information of the domains such as physical volume, radius of gyration, and interactions to each other across epigenetic boundaries. These super-resolution approaches hold great promise for understanding how the nuclear bodies, chromatin and specific genomic regions are remodeled in response to physical cues.

Until recently, imaging-based methods like RNA-FISH and Oligopaint suffered from low throughput, only capable of visualizing and quantifying a few genes due to the limited number of spectrally distinct fluorophores for multicolor imaging. These technologies have undergone a revolution in the recent past in which the throughput has been drastically scaled up using barcoding and sequential hybridization approaches [18]. First, unlabeled primary oligonucleotide probes (oligos) recognize short, target sequences (~12–20 nucleotides) and in turn present their own set of non-genomic, barcoded sequences (~12–20 nucleotides) as targets for fluorescently labeled secondary oligos. After the secondary oligos are imaged, the probes are either bleached or removed so that the conjugated dyes no longer emit signal. This prepares the system for the next round of hybridization and imaging [19]. Once the imaging is completed, the barcode belonging to a specific RNA species or genomic region can be decoded, producing a high-throughput spatial map of transcriptional activity of single cells (MERFISH; see Figure 4a [137], seq-FISH [138], seq-FISHþ [139]) or high-resolution spatial map of genomic sequences (DNA-MERFISH [136], ORCA [140], Hi-M [141], and MINA [142]; see Refs. [18,133] for an extensive review on spatial genomics and transcriptomics approaches).

Figure 4. Advanced microscopy methods to visualize nuclear organization.

(a) Multiplexed FISH overview (MERFISH as an illustrative example). Through combinatorial labeling, N-bit binary words target subsets of RNA species. After multiple rounds of imaging and addressing potential errors, the full suite of RNA species can be localized within the sample. Figure adapted from Ref. [135]. (b) Chromatin tracing multiplexed with other imaging modes, in this case, RNA and protein imaging. This demonstrates the advantage microscopy offers over most sequencing technologies, where the relationship between multiple components of a cellular system can be measured at once. In this schematic, the genomic traces, nascent RNA, and subnuclear organelles can all be visualized within the same nucleus. Figure adapted from Ref. [136].

These approaches can be combined with immunofluorescence to achieve multimodal imaging of the 3D genome, nucleome, and transcriptome within the same cell. For example, DNA-MERFISH allows for the simultaneous imaging of over a thousand distinct genomic loci and nascent RNA transcripts in addition to landmark nuclear structures such as nucleoli and nuclear speckles (Figure 4b) [136]. This same multiplexing capability was used for the high throughput in situ imaging of cell type specific organization of developmental control genes in Drosophila embryos [140]. Promoter-enhancer contacts and TADs were examined in cells classified based upon their RNA expression. While this technology has yet to be coupled with mechanical studies, chromatin tracing could be used to evaluate at a single-cell level how a gene, its related transcriptional elements, and larger nuclear components reorganize in response to differing mechanical microenvironments.

Traditional FISH approaches typically involve the use of heat or harsh chemicals to denature double-stranded DNA for hybridization, but this can distort genome organization. Approaches such as Cas9-mediated FISH (CASFISH) and Resolution After Single-strand Exonuclease Resection FISH (RASER-FISH) address this issue by circumventing the need for such disruptive treatment and better preserve chromatin structure [143,144]. One drawback of FISH is that most methods are not compatible with live cell imaging due to fixation and denaturation, which is especially undesirable for studies on active mechanical stressors within short time scales [81,135]. CRISPR live-cell fluorescent in situ hybridization (CRISPR LiveFISH) is a recent live-cell compatible FISH approach where fluorophore-conjugated guide RNAs detect either DNA or RNA using the Cas9 or Cas13 systems, respectively [135]. Thus, this technology enables real-time visualization of transcription that could be more amenable for live cell stretching experiments and is also capable of imaging genome editing and chromosome translocation.

Outlook

Research has revealed the intimate relationship between the external physical cues, internal mechanical state of the nucleus, and transcriptional activity. Microscopy represents a collection of diverse methods that have contributed towards understanding the relationship between physical forces, the nucleus, and ultimately cell fate. While significant progress has been made in the field of mechanoregulation, outstanding questions remain open on the specificity and mechanisms behind these links. Further efforts to decipher these mechanisms will benefit from both higher resolution imaging, live cell imaging and sample throughput to efficiently visualize and quantify how strains applied to a cell translate to changes in genome architecture and transcriptional activity at the single cell level. We also emphasize microscopy’s unique capacity to achieve a more holistic survey of cellular processes via multiplexed imaging. Utilizing such techniques will capture the states of multiple targets simultaneously to provide valuable information as to how different key players in mechanotransduction coordinate with each other to respond to the same cue. Finally, the development of more high-resolution imaging methods compatible with live cells would be a great boon to the field. Because cells may behave differently based on the dynamics of external mechanical forces, it is often necessary to evaluate cellular responses outside of fixed cell systems. However, many established super-resolution technologies either rely on fixation or must compromise between spatial and temporal resolution. To further the field, technologies must overcome such challenges and become more widely available. Overall, novel imaging technologies will help to elucidate how to best apply mechanical cues to engineer cell fate changes in applications such as tissue engineering or treat pathological processes.