How cells respond to physical cues in order to meet and withstand the physical demands of their immediate surroundings has been of great interest for many years, with current research efforts focused on mechanisms that transduce signals into gene expression. Pathways that mechano-regulate the entry of transcription factors into the cell nucleus are emerging, and our most recent studies show that the mechanical properties of the nucleus itself are actively controlled in response to the elasticity of the extracellular matrix (ECM) in both mature and developing tissue. In this Commentary, we review the mechano-responsive properties of nuclei as determined by the intermediate filament lamin proteins that line the inside of the nuclear envelope and that also impact upon transcription factor entry and broader epigenetic mechanisms. We summarize the signaling pathways that regulate lamin levels and cell-fate decisions in response to a combination of ECM mechanics and molecular cues. We will also discuss recent work that highlights the importance of nuclear mechanics in niche anchorage and cell motility during development, hematopoietic differentiation and cancer metastasis, as well as emphasizing a role for nuclear mechanics in protecting chromatin from stress-induced damage.
Evolution has likely driven our tissues and organs to fulfill their roles with optimal efficiency and, of course, viability. Mature tissues need to be particularly resistant to the mechanical demands of an active life. Our bones, cartilage, skeletal muscle and heart tissues are stiff, making them robust and resistant to routine physical exertion, such as walking or running, during which they are subjected to high-frequency shocks, stresses and strains. With every heartbeat, the left ventricular wall experiences a 20% radial strain (Aletras et al., 1999), and local strains of ∼20% also occur in the cartilage of knee joints with every step (Guilak et al., 1995). Tissue-level deformations might even be amplified within cells and their nuclei (Henderson et al., 2013). Our softer tissues have less need for robustness because their function does not require them to bear load. Furthermore, some of our softest tissues, such as brain and marrow, are protected from shocks and stresses by our bones. Nonetheless, when soft tissues are subjected to impact, such as during a collision of heads in American football or rugby, occurrences of rapid straining can cause lasting damage (Viano et al., 2005).
We have recently sought to characterize the composition of cells and extra-cellular matrix (ECM) in tissues of increasing stiffness, and by implication, in tissues that are subjected to the greatest stress (Swift et al., 2013b). A close correlation between the concentration of ECM components and bulk tissue elasticity was discovered for mouse tissues (Swift et al., 2013b). More surprisingly, we also discovered a systematic scaling between tissue elasticity and the concentration of lamins in the nucleoskeleton that was partially re-capitulated in cultured cell systems (Swift et al., 2013b). Corresponding changes to nuclear mechanical properties associated with this tuning of lamina composition suggest that this response might act to protect the chromatin cargo of the nucleus from shocks that are transmitted through the surroundings, across the cytoskeleton and into the nucleus. An active regulation of cell or ECM composition in response to the environment implies feedback into pathways of protein turnover and remodeling, or control of the rate of new protein production. Responsive matching of mechanical properties to physical demands has classically been described as a ‘mechanostat’ in the context of bone regulation (Frost, 1987), but a recent explosion in mechanobiological studies has uncovered a host of other mechanically sensitive cellular phenomena, including contraction (Discher et al., 2005), migration (Hadjipanayi et al., 2009b; Winer et al., 2009; Raab et al., 2012), proliferation (Hadjipanayi et al., 2009a; Klein et al., 2009; Lo et al., 2000), differentiation (Engler et al., 2004; Engler et al., 2006) and apoptosis (Wang et al., 2000). However, despite great recent progress, questions of how mechanical signals are transduced into specific transcriptional or regulatory pathways continue to challenge the field.
The lamina is a network structure formed from intermediate filament lamin proteins, and it lies just inside the nuclear envelope and interacts with both chromatin and the cytoskeleton (Fig. 1A). In the somatic cells of humans, mice and most vertebrates, the main forms of lamin protein are expressed from three genes: lamin-A and lamin-C are alternatively spliced products of the LMNA gene, and they are collectively referred to as ‘A-type’ lamins; lamin-B1 and lamin-B2 are encoded by the LMNB1 and LMNB2 genes, respectively, and are known as ‘B-type’ lamins. The lamins share structural features and, indeed, have some similarity in amino acid sequence, but they differ in their post-translational modification, with B-type lamins permanently appended by a farnesyl group that is cleaved from mature lamin-A (reviewed, for example, by Dechat et al., 2010; Ho and Lammerding, 2012). Like other intermediate filament proteins, such as keratin and vimentin, the lamins form coiled-coil parallel dimers that assemble into higher-order filamentous structures that fulfill important structural roles (Herrmann et al., 2009).
This Commentary aims to summarize recent efforts to characterize the proteins that vary systematically with tissue stiffness. The effects of the composition of the lamina on nuclear mechanical properties will be elaborated in detail, and we will consider the functions of the lamina in transducing mechanical signals from the ECM and the surroundings of the cell into cellular responses, both in terms of the active regulation of the lamina itself and its broader role in linking mechanotransduction pathways. Although we focus on a primarily protective function of nuclear lamin, there are additional regulatory consequences of possessing such a stiff and bulky organelle as the nucleus, and we will summarize recent evidence that such properties limit the freedom of cells to move through tissue. The proximity of the lamina to heterochromatin within the nucleus has led it to be widely associated with epigenetic regulation (e.g. Kim et al., 2011; Meuleman et al., 2013). This Commentary seeks to highlight the pervasive influence of the mechanical role of the lamina and hence proposes that lamin acts as both the guardian and the gatekeeper of chromatin.
Scaling of ECM and lamina components in mature and developing tissue
Collagens and other protein constituents of the ECM are the most prevalent proteins in our bodies, largely determining the mechanical properties of tissue. Collagens are found at higher levels in stiff mature tissues where, consistent with an expectation for proteins to behave as ‘biological polymers’ (Gardel et al., 2004), their increased concentration is the basis of tissue elasticity (Fig. 1B, left panel). By using quantitative label-free mass spectrometry for proteomic profiling (Swift et al., 2013a), we have shown that collagens and other ECM-associated proteins scale with tissue elasticity (Swift et al., 2013b). Mass spectrometry was also used to quantify ∼100 of the most abundant proteins in the cytoskeleton and nucleus, and we found that the elasticity of bulk tissue is strongly correlated with the composition of the nuclear lamina in terms of its content of A-type and B-type lamins (Fig. 1B, right panel). Although primarily characterized by the ratio between these two main families of lamins, the compositional scaling is dominated by a 30-fold increase in the concentration of A-type lamin from brain to bone. Although our recent observations are broadly in agreement with an extensive literature on lamin quantification (e.g. Broers et al., 1997; Cance et al., 1992; Krohne et al., 1981; Röber et al., 1990), they provide a new perspective on systematic variations across many tissues.
The relationship between tissue stiffness and ECM during development has also been determined (Majkut et al., 2013); micropipette aspiration of embryonic chick tissue showed that the homogeneous embryonic disk is initially very soft, with proteomic profiles indicating correspondingly low levels of collagen (Fig. 1C, left and middle panels). However, the properties of different tissues diverge during development, with the brain remaining soft, whereas the heart stiffens as ECM proteins are deposited (Majkut et al., 2013). Cells in stiffening tissues, such as the heart, are also likely to have correspondingly higher levels of lamin-A and lamin-C (Lehner et al., 1987). Nuclei in embryonic stem cells have, indeed, been shown to be very soft and to have low levels of lamin-A and lamin-C (Eckersley-Maslin et al., 2013; Pajerowski et al., 2007). As these cells commit to a lineage-specific fate, the levels of these lamins increase, and the nucleus becomes correspondingly stiffer (Fig. 1C, right panel).
Importantly, despite an apparent role in reinforcing the decisions of animal cell fate (in conjunction with ECM elasticity) (Swift et al., 2013b), lamin-A and lamin-C are not essential to development, because knockout mice that lack A-type lamins still form all tissues (Sullivan et al., 1999). Likewise, lamin-B knockout mice survive embryogenesis (Kim et al., 2011), and embryonic stem cells can be cultured and differentiated without having any lamins (Kim et al., 2013). The most crucial role of lamin might therefore be to fine-tune the properties and regulation of maturing tissues in higher organisms, and its absence can perhaps be compensated for during development. However, we note that the distinction here might be blurred; there is still a need to regulate and maintain nuclear structure during some stages of development, such as during cell migration, and the processes of trafficking and differentiation continue throughout the lifespan. Nonetheless, the fact that lamins are not absolutely essential for cell viability is consistent with the current notion that lamins are not expressed in yeast and plants (Dittmer and Misteli, 2011), despite the latter possessing genomes that are larger and more complex than those of animals. The hard cell walls of these organisms likely protect the chromatin in a manner that is not possible for animal cells with soft cell membranes.
The influence of lamina composition on the mechanical properties of the nucleus
Micropipette aspiration experiments have enabled the detailed study of the mechanical properties of the nucleus, which can be investigated by measuring the rate of deformation under pressure (Fig. 2A,B; Dahl et al., 2005; Pajerowski et al., 2007). By examining nuclei with different lamina compositions, it is thus possible to approximate the characteristic contributions of A-type and B-type lamins to nuclear properties (Fig. 2C; Harada et al., 2014; Shin et al., 2013; Swift et al., 2013b). The nuclear mechanical response has been characterized as a combination of elastic (spring-like) and viscous (liquid-like, flowing) properties, with B-type lamins contributing primarily to the elastic response and A-type lamins contributing to the viscosity (Harada et al., 2014; Shin et al., 2013; Swift et al., 2013b). Thus, the difference between a nucleus in which the lamina is stoichiometrically dominated by A-type versus B-type lamins might be akin to comparing a balloon filled with honey to one filled with water. The importance of A-type lamins in maintaining nuclear structural integrity and cell viability has been appreciated for many years (e.g. Broers et al., 2004; Lammerding et al., 2006), and the influence of A-type lamins on nuclear viscosity has been more recently demonstrated in studies where nuclei are deformed during migration through microfluidic circuits (Rowat et al., 2013) or transwell pores (discussed later; Harada et al., 2014).
Laminopathies are a family of diseases that are caused by truncations or single amino acid substitutions in A-type lamins (reviewed, for example, by Butin-Israeli et al., 2012; Davidson and Lammerding, 2014; Isermann and Lammerding, 2013; Worman, 2012). These disorders include muscular dystrophies (Bonne et al., 1999), cardiomyopathies (Fatkin et al., 1999), lipodystrophies (Hegele et al., 2000; Shackleton et al., 2000; Speckman et al., 2000) and premature ageing (termed ‘progeria’; Merideth et al., 2008). Indeed, one of the confounding aspects of lamin-related disease is the mechanism by which such a widely expressed protein can cause tissue-specific symptoms. Although much further work is needed to resolve this question, it is broadly true that laminopathies cause defects in tissues where A-type lamins are the dominant nuclear lamin (i.e. bone, heart, muscle and fat). However, there are exceptions – Charcot-Marie-Tooth disorder affects the nervous system (De Sandre-Giovannoli et al., 2002). Mouse models of A-type lamin knockout display defects in muscle and connective tissue and typically die from heart failure several weeks after birth (Jahn et al., 2012; Kubben et al., 2011; Sullivan et al., 1999). Despite the apparently constitutive expression of B-type lamins, mouse models with lamin-B1 and lamin-B2 ablation progress through embryogenesis, with their eventual death being due to defects in brain development (Coffinier et al., 2011; Kim et al., 2011).
Mechanisms of lamin regulation
Earlier discussion has posited that lamins are closely regulated to match the mechanical properties of the nucleus with the physical demands of the tissue. In addition to being determined by the epigenetic programming required to make a given tissue or organ, it is also important to note that protein levels vary in response to feedback from their surroundings. Even within bulk tissues, mechanical loading can cause inhomogeneous straining (for example in human articular cartilage, meniscus and ligaments; Chan and Neu, 2012), making it beneficial to have mechanisms of lamin regulation that act at the local level of individual cells.
The many mechanisms by which the levels of lamin-A and lamin-C can be regulated are summarized in Fig. 3A. We have shown that the transcript and protein levels of A-type lamins are highly correlated in tissue (Swift et al., 2013b), a finding that is suggestive of a tight regulatory feedback. A recent study of the proteome and transcriptome in mouse fibroblasts showed that there are around 107 copies of A-type lamin proteins per cell – accounting for about 0.7% of cellular protein mass – and around 200 copies of the LMNA transcript. Half-life in the cell is reported to be ∼4 days for the protein and ∼20 hours for the mRNA, both slightly higher than the cellular average for all proteins and transcripts (Schwanhäusser et al., 2011). Measurements made on proteins in a human lung cancer cell line showed the half-life of A-type lamin to be ∼12 hours, roughly in the middle of the span of protein half-lives recorded in the study (Eden et al., 2011). DNA methylation is an epigenetic mechanism by which gene activity can be regulated, but it was discounted as the foremost means of controlling LMNA levels – no consistent changes were observed in the methylation of the LMNA promoter either in a range of cell lines known to express different levels of A-type lamin protein (Swift et al., 2013b) or in tissues from patients with laminopathic disorders (Cortese et al., 2007). LMNA transcription has been reported to be controlled by transcription factors of the retinoic-acid receptor family (RAR and RXR family proteins; Okumura et al., 2004a; Okumura et al., 2004b; Olins et al., 2001; Shin et al., 2013; Swift et al., 2013b), with the resulting mRNA alternatively spliced to give the lamin-A and truncated lamin-C forms. Soft tissue generally preferentially expresses the lamin-C spliceoform (Swift et al., 2013b), and, in brain, the micro-RNA MIR-9 specifically targets and deactivates the mRNA of the lamin-A spliceoform (Jung et al., 2012; Jung et al., 2013). Little is currently understood about different functional roles of lamin-A and lamin-C, although differences in both physical properties, such as mobility (Broers et al., 2005), and lamina assembly (Kolb et al., 2011; Pugh et al., 1997) have been reported.
Stress-responsive regulation of lamin – ‘use it or lose it’
To better understand how lamin proteins are actively regulated in response to stress, mesenchymal stem cells (MSCs) were cultured on type-I-collagen-coated polyacrylamide hydrogels with stiffnesses that were set to mimic the ECM of either brain (0.3 kPa) or pre-calcified bone (40 kPa) (Buxboim et al., 2010b; Swift et al., 2013b). Images of the cultured MSCs showed that the nuclear envelopes of cells on soft substrate are wrinkled and relaxed, whereas, on stiff substrate, the nuclei are flattened by stress fibers and appear taut and smooth (Fig. 3B, left panel). Accompanying proteomic analyses revealed that the native fold of A-type lamin is maintained on stiff substrate (specifically, that of the globular Ig-like domain), but that the total quantity of A-type lamin protein is upregulated. Furthermore, the extent of phosphorylation at four sites is decreased on stiff substrate. Phosphorylation is recognized as a key mechanism for modulating the solubility, conformation and organization of intermediate filament proteins (Omary et al., 2006), and indeed, lamins are highly phosphorylated during normal mitosis, driving the disassembly of the lamina in preparation for chromosomal separation (Gerace and Blobel, 1980; Heald and McKeon, 1990). Thus, the response we observe to substrate-induced stress is the converse of this process, with decreased phosphorylation acting to decrease A-type lamin solubility and so strengthening the lamina. On soft substrate, A-type lamins are more extensively phosphorylated, more mobile and so, we hypothesize, more susceptible to turnover (as observed following Akt-mediated phosphorylation of S404 in prelamin-A (Bertacchini et al., 2013). These observations hence point to a ‘use it or lose it’ dynamic, whereby non-essential A-type lamin is eventually degraded. The level of A-type lamins has been reported to drive the translocation of the lamin-promoting transcription factor retinoic acid receptor γ (RARG) to the nucleus, pointing to a feedback mechanism by which lamin protein promotes its own transcription (see gene circuit in Swift et al., 2013b).
There are a number of cases in the literature where cell mechanics are regulated by phosphorylation or, by contrast, where external stresses drive changes in phosphorylation state. As an example of the former, changes in the phosphorylation of microtubule associated proteins (MAPs) affect the assembly of microtubules that have important structural roles in the cytoskeleton (Matenia and Mandelkow, 2009). Instances of the second case include both the extension-dependent phosphorylation of the Cas substrate domain protein, p130Cas (also known as BCAR1), which transduces mechanical stretching into biochemical signaling (Sawada et al., 2006), and the phosphorylation of non-muscle myosin-IIa (encoded by MYH9), which is inversely related to substrate stiffness and acts to regulate cell contractility (Raab et al., 2012). Recent work has shown that the nuclear envelope protein emerin is phosphorylated in response to cell contractility, which consequently influences nuclear rigidity and, potentially, the way in which emerin interacts with other nuclear proteins such as lamin-A and lamin-C (Guilluy et al., 2014). Furthermore, phosphorylation events could be key to driving the transport of proteins such as transcription factors to the nucleus, where they have crucial regulatory roles (Jans and Hübner, 1996).
The concept of ‘stress strengthening’ is also found in other biological systems; stretched type I collagen fibrils were found to be less susceptible to enzymatic degradation (Fig. 3B, right panel; Flynn et al., 2010). Although the mode of enzymatic action differs between A-type lamin and type I collagen, both are examples of polymeric structural proteins that are protected from degradation pathways when they are under tension. These parallels should also serve as reminders that, although this review is primarily focused on stress response within the cell, the ECM in living tissue is also being continuously remodeled in response to extracellular cues (Xu et al., 2009).
Mechanotransduction to the nucleus – downstream of ECM and lamin
Lamin is a key component in a system of protein linkages that allows the transmission of signals from the surroundings of a cell into the transcriptional machinery of its nucleus (Fig. 1A; discussed in recent reviews, e.g. Gundersen and Worman, 2013; Rothballer and Kutay, 2013; Simon and Wilson, 2011; Sosa et al., 2013). Cell–cell interactions link to the cytoskeleton through tight junctions and adherens junctions that tether to actin, and through desmosome complexes that interact with cytoplasmic intermediate filaments, such as keratin (Jamora and Fuchs, 2002). Cell–ECM interactions are mediated by integrins and focal adhesion complexes that bind to cytoplasmic actin (Puklin-Faucher and Sheetz, 2009; Watt and Huck, 2013). The appropriately named LINC complex (linker of nucleoskeleton and cytoskeleton) acts as an intermediary between cytoplasmic and nuclear structural proteins; F-actin binds to the nuclear envelope components nesprin-1 and nesprin-2, and intermediate filaments bind to the desmosome protein plectin, which, in turn, binds to nesprin-3. Nesprins can also interact with kinesin and dynein complexes to tether to the microtubule network; nesprins bind to the SUN-domain-containing family of inner nuclear membrane proteins and these, in turn, bind to the lamina on the inside of the nuclear envelope.
Lamins engage in extensive protein–protein interactions within the nucleus (Wilson and Berk, 2010; Wilson and Foisner, 2010), as emphasized by Wilson and Berk in their review: “almost all characterized [inner nuclear membrane] proteins bind to A- or B-type lamins (or both) directly”. These interactions include binding to structural proteins, like actin (Simon et al., 2010), and binding to a range of proteins that interact with the nuclear membrane, including emerin, barrier-to-autointegration factor (BANF; Montes de Oca et al., 2009), lamina-associated polypeptide 2 (LAP2, also known as ERBB2IP) and lamin-B receptor (LBR; Solovei et al., 2013). Of these, emerin has attracted considerable recent interest for its role in mediating changes in the stiffness of isolated nuclei in response to tension applied to nesprin-1 (Guilluy et al., 2014), and its role in mechanosensing through its effect on the translocation of the transcription factor MKL1 (Ho et al., 2013). Furthermore, some transcription factors, such as Oct-1, interact with the lamina directly (Malhas et al., 2009). Many lamin-binding proteins also interact with chromatin, particularly in its silenced heterochromatic form (Wagner and Krohne, 2007), and indeed, the lamins have been shown to bind to DNA directly (Ludérus et al., 1992; Shoeman and Traub, 1990; Stierlé et al., 2003). This chain of interactions thus completes a continuous physical linkage through which deformations can be transmitted from the cell exterior to chromatin (Maniotis et al., 1997). What is missing from this picture, however, is an understanding of the mechanism by which blunt inputs – forces and perturbations acting without microscopic coherence – can be converted from mechanical to biochemical signals to activate individual genes at precise spatial locations within the nucleus. Perhaps specificity can be delivered through changes in binding, local concentration, conformation and modification of cofactors or transcription factors.
As described earlier in this Commentary, mechanical signals from outside the cell are reflected in alterations in protein conformations, modifications and expression levels, and these can broadly affect cell morphology. Taking these changes further, it is of particular interest to understand how external factors drive stem cells to commit to specific lineages, with far-reaching implications for therapeutics and tissue engineering. Populations of MSCs can be expanded in culture in an undifferentiated state, but they can differentiate into mesenchymal lineages, including fat, cartilage, muscle and bone, in a manner that is dependent on external cues, such as the presence of nutrients, growth factors and cytokines, cell density, spatial constraints and mechanical forces (Pittenger et al., 1999). It has been shown that cell shape influences cell fate by modulating the activity of the small GTPase RhoA, with round cells being predisposed towards adipogenesis and well-spread cells more likely to enter the osteogenic lineage (McBeath et al., 2004). Here, RhoA was found to drive commitment to a specific lineage in conjunction with its effector Rho-associated protein kinase (ROCK), through its effect on cytoskeletal tension.
Cell fate can be influenced by the stiffness of the ECM (Engler et al., 2006), and we have recently shown that this effect can be modulated by the nuclear lamina (Fig. 4A; Swift et al., 2013b). MSCs cultured on soft hydrogel substrates are biased towards adipogenesis, but the extent of adipogenesis, as determined by staining with Oil Red O, is doubled when combined with the knockdown of A-type lamins. Likewise, a cooperative effect between stiff substrate and lamin-A overexpression enhances osteogenesis. Stiff substrate has been found to drive the translocation of the transcription factors RARG and Yes-associated protein 1 (YAP1) to the nucleus, thus directly regulating the transcription of LMNA and triggering a differentiation-inducing signaling cascade (Dupont et al., 2011). Culturing MSCs on stiff substrate also increases the protein levels of cytoskeletal components, consistent with greater activity of serum response factor (SRF) (Connelly et al., 2010; Ho et al., 2013), and, in combination with a stiffer nucleus, this further increases cytoskeletal tension. Translocation of molecular signal carriers is a recurring motif in mechanosensitive pathways that has been shown to occur during the transfer of ions and changes in osmotic pressure (Finan et al., 2009; Irianto et al., 2013; Kalinowski et al., 2013), as well as during the migration of protein factors. Protein translocation could be driven by a change in the concentration of binding sites, (e.g. on A-type lamins or emerin; Ho et al., 2013; Swift et al., 2013b) or, conceivably, by protein modifications (e.g. YAP1 nuclear localization can be stimulated by phosphorylation of the protein; Murphy et al., 2014). Additionally, transient breakdown of the nuclear envelope in laminopathic cells, perhaps as a consequence of a reduced robustness under conditions of mechanical stress, has been shown to allow the ingress of transcription factors such as RelA (De Vos et al., 2011).
The ability of lamin and its binding partners to tether to DNA has led to interest in its role in chromatin organization and regulation (Guelen et al., 2008; Kim et al., 2011; Kind et al., 2013; Lund et al., 2013; Meuleman et al., 2013; Zullo et al., 2012). Lamina-associated domains (LADs) located at the nuclear periphery are typically associated with low levels of gene expression, whereas actively transcribed euchromatin is usually found in the nuclear interior. These conserved spatial relationships within the nucleus give rise to defined ‘chromosome territories’ (Cremer and Cremer, 2001; Iyer et al., 2012) and to transcriptional hotspots within specific locations (Fraser and Bickmore, 2007). Although chromatin and DNA are generally considered to make negligible contributions to overall nuclear mechanics (Guilluy et al., 2014; Pajerowski et al., 2007), particular cases are emerging – for example in ESCs passing through a metastable transitional state before differentiation – where the condensation state of chromatin can become mechanically significant (Pagliara et al., 2014). It is not yet fully understood which proteins could give rise to mechanically responsive locally defined structures and organization within the nucleus, nor how a protein as ubiquitously expressed as lamin could play a part in such specificity. Knowledge in this field will continue to improve as new experimental methods and models emerge to allow the study of protein-mediated changes in chromatin organization in response to perturbation of the system (Shivashankar, 2011; Talwar et al., 2013). However, based on current work, we have hypothesized that the effect of lamin on nuclear mechanics could determine the sensitivity and timescale of nuclear reorganization in response to stress (Fig. 4B; Swift et al., 2013b).
Cell migration is slowed by the nuclear stiffness needed to protect chromatin
As the nucleus is generally the largest and stiffest organelle, it can be a limiting factor in the migration of a cell through the three-dimensional ECM. This means that the mechanical properties of the nucleus can have regulatory roles in certain processes, such as development, wound healing, hematopoiesis, cancer metastasis and others (Fig. 5A, left panel). Studies of migration through narrow pores that mimic those in tumor tissue, migration through which requires the deformation of the nucleus, demonstrated a dependence on lamina composition – migration was limited when lamin-A was overexpressed, but was promoted by a ∼50% knockdown of the protein (Harada et al., 2014). However, a more effective knockdown to <10% of normal protein expression levels was found to cause apoptosis, underscoring the importance of lamin in providing physical protection to the nucleus (Fig. 5A, middle panel). Consistent with earlier observations that A-type lamin and B-type lamin contribute primarily viscous and elastic mechanical properties to nuclei, respectively (Fig. 2), nuclei in which high A-type lamin levels dominated the mechanical characteristics were observed to recover slowly following deformation, maintaining an elongated morphology after emerging from the pores (Fig. 5A, upper-right panel). By contrast, nuclei with dominant levels of elastic B-type lamins rapidly returned to their more spheroid pre-migratory shapes following deformation (Fig. 5A, lower-right panel).
Cell migration is an important part of the development process, and it is possible that the elasticity imparted by lamin-B is needed to allow nuclei to recover from the deformation (typically elongation) that occurs during migration, perhaps explaining why the brain fails to develop in lamin-B-knockout mice (Coffinier et al., 2011; Jung et al., 2013; Kim et al., 2011). Neutrophilic cells also have very low levels of nucleoskeletal proteins to allow their deformation as they squeeze into confined spaces (Olins et al., 2009; Rowat et al., 2013), and indeed, the composition of the nuclear lamina is continuously regulated during hematopoiesis (Fig. 5B; Shin et al., 2013). We hypothesize that by downregulating components of the lamina, white blood cells compromise their robustness in favor of mobility, and that this contributes to the short lifetimes of many of these cells in circulation. Cancer metastasis is an equally complex process that depends on factors including the remodeling of the ECM and the capacity of the nucleus to deform (Harada et al., 2014; Wolf et al., 2013). Other work has shown that the ability of myosin-II to deform the nucleus can be a decisive factor in limiting glioma migration into brain tissue (Beadle et al., 2008; Ivkovic et al., 2012), but cancer cells in general show no universal lamina phenotype (reviewed in Foster et al., 2010). Although low levels of A-type lamin have been correlated with increased reoccurrence of colon cancers (Belt et al., 2011), A-type lamin was also found to be upregulated in certain skin and ovarian cancers (Hudson et al., 2007; Tilli et al., 2003), and higher expression of these lamin proteins has been associated with better clinical outcomes in breast cancer (Wazir et al., 2013). Our own studies of tumor expansion in mouse flank have been suggestive of an association between moderately reduced levels of A-type lamin and increased invasiveness into the surrounding tissue, but a more complex relationship between the level of A-type lamin and clinical prognosis might be explained by the tenuous balance between the effect of the lamina on nuclear deformability compared with that on cell survival.
Conclusions and prospects
We have sought to emphasize the underlying importance of nuclear mechanics in the context of tissue function, considering how they determine the protective properties of the lamina, influence cell fate and also regulate cell migration. In understanding that one of the key functions of lamins is to ensure that the mechanical properties of the cell meet the demands of a tissue – either directly or indirectly, by driving broader changes with regard to cell fate – we are essentially allocating lamin a role in stress response (Zuela et al., 2012). The response to cellular stress is classically thought of in terms of how cells mitigate heat shock, which can otherwise result in high levels of unfolded proteins (Hartl et al., 2011), but mechanical stress can also cause protein unfolding with associated loss of function (Swift et al., 2013b). Another recent review highlighted the similarities between cellular responses that involve the expression of heat-shock proteins and those that drive structural intermediate filament proteins (Toivola et al., 2010). Just as the lamina responds to mechanical inputs from the ECM, mechanical stretching is sufficient to induce the expression of the molecular chaperone heat-shock protein HSP70 (Silver and Noble, 2012). However, we are still a long way from fully understanding cellular protection mechanisms and the way in which stress response pathways affect the regulation of structural features within the cell. Nonetheless, it is apparent that nuclear biophysics has an important role to play.
We thank our colleagues P. C. Dave P. Dingal, Irena L. Ivanovska and Stephanie Majkut at the University of Pennsylvania for helpful comments.
We appreciate support from the U.S. National Institutes of Health, the U.S. National Science Foundation and the University of Pennsylvania's research centers (Materials Research Science and Engineering; Nano Science and Engineering; Nano/Bio Interface). Deposited in PMC for release after 12 months.
The authors declare no competing interests.