Mechanical Properties of the Cell Nucleus and the Effect of Emerin Deficiency

Cell culture and materials

Emerin-deficient mouse embryo fibroblasts (MEFs) as well as wild-type controls were immortalized by repeated passage and were kindly provided by C. L. Stewart and S. Kozlov, National Cancer Institute, Bethesda, MD (14). HeLa cells were obtained from American Tissue Cell Culture (ATCC, Manassas, VA). Both HeLa and MEF cells were maintained at 37°C and 5% CO₂ in Dulbecco's Modified Eagle Medium containing 10% fetal bovine serum, 1% glutamine, and 1% penicillin/streptomycin (BioWhittaker, Cambrex, East Rutherford, NJ).

Experiments were performed on intact MEF and HeLa cells, as well as on isolated HeLa cell nuclei. A procedure to isolate nuclei was developed from a previously described protocol (15). When the cells had attained ∼90% confluency, they were harvested by trypsinization, washed 3× with PBS and resuspended in a buffer of 10 mM HEPES-KOH (pH 7.9), 10 mM KCl, and 1.5 mM MgCl₂ containing protease inhibitors (complete tablets, EDTA-free, Roche, Penzburg, Germany). Thereafter cells were incubated on ice for 5–10 min or until the plasma membranes were sufficiently swollen as observed by phase contrast microscopy. Dounce homogenization (typically 10 strokes of the tight pestle in a 7 mL Dounce tissue homogenizer (Wheaton Scientific Products, Millville, NJ)) released the nuclei that were then collected by centrifugation (1000 × g) for 5 min. After isolation, nuclei were resuspended in a physiological buffer formulated to reproduce ionic conditions in the cytoplasm (16): 130 mM KCl, 1.5 mM MgCl₂, 10 mM Na₂HPO₄, 1 mM Na₂ATP, and 1 mM DTT, pH 7.4. For ATP regeneration, 5 mM creatine phosphate and 0.1 mg/mL creatine phosphokinase were added.

The enhanced green fluorescent protein conjugated to Lamin A (GFP-LamA) construct was a generous gift from D. K. Shumaker and R. D. Goldman, Northwestern University, Chicago, IL (17). Transfection was performed with twofold excess carrier DNA (Herring testes DNA, Clontech, San Jose, CA) into HeLa/MEF cells at 70–80% confluency using CaPO₄ (Invitrogen, Carlsbad, CA). The GFP-LamA delineated lamina was visualized in living MEF and HeLa cells as well as in isolated HeLa cell nuclei. Membranes of isolated nuclei were labeled by incubation with 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate (DiIC18, Molecular Probes, Eugene, OR) at a final concentration of 0.01 μM in physiological buffer for 15 min. Nuclei that were not cleanly isolated and remained associated with external membranes from, for example, the endoplasmic reticulum, were easily identified as the membrane staining extended beyond the periphery of the nucleus itself and were excluded from the analysis. SYTOX Orange Nucleic Acid Stain (Molecular Probes, Eugene, OR) was used to stain nucleic acids by incubation with isolated nuclei at a final concentration of 0.1 μM before imaging. The nuclear interior and nucleoli were visualized in living HeLa cells transfected with CFP-NLS-NLS-NLS (Clontech, San Jose, CA) using the same transfection protocol as described above. CFP-NLS-NLS-NLS (CFP-NLS×3) consists of an eCFP fluorophore fused to three consecutive nuclear localization sequences (NLS). The NLS targets proteins to the nucleus (18), and three consecutive NLS sequences are thought to preferentially localize the protein to nucleoli (J. S. Andersen, 2005, personal communication). CFP-NLS×3 has no known specific interactions with nuclear structures.

Bright-field micropipette aspiration

Micropipette aspiration was used to deform nuclei and the induced response was studied by bright-field microscopy. A chamber open at both sides allowing for pipette entry was prepared by affixing a glass coverslip with a few small drops of vacuum grease to the bottom of a home-built, thermostated chamber. The chamber was mounted on the microscope and filled with nuclei resuspended in physiological buffer. A temperature of 25°C was maintained using a water bath. Pipettes were pulled from glass capillaries of 1 mm diameter (World Precision Instruments, Sarasota, FL) using a pipette puller (Sutter Instruments, Novato, CA). Subsequent forging of the pipettes (Microforge MF-900, Narishige, Japan) ensured a flat pipette with a typical final diameter of 2R_p = 2.0–3.0 μm. After pulling and forging the glass capillaries, micropipettes were quickly dipped in silanization solution (Number II, dichlorodimethylsilane solution, Sigma-Aldrich, Copenhagen, Denmark) to prevent sticking of biological material to the glass surface and were thereafter incubated in an oven at 90°C for 120 min. After cooling to room temperature, pipettes were inspected on the microforger for clogging that can result from the silanization treatment. The bottom coverslip of the chamber was also silanized. Pipettes were backfilled with physiological buffer solution using a 1 mL plastic syringe fitted with a 97-mm long, 28-gauge backfiller (Microfil, World Precision Instruments, Sarasota, FL).

Isolated HeLa cell nuclei were visualized using bright-field optics on an inverted microscope (Zeiss Axiovert S100, Zeiss LD Achroplan 40×/0.65, Carl Zeiss, Göttingen, Germany) and images were acquired using a charge-coupled device camera (XC-85000E Donpisha, Sony, Tokyo, Japan) connected to a computer via a frame grabber (Sigma-SLC, Matrix Vision, Oppenweiler, Germany). Micromanipulators (Coherent, Ely, Cambridgeshire, UK) were used to position the micropipette that was connected to a custom-built manometer system (19,20). A pressure transducer (DP1530/N1S4A, Validyne, Northridge, CA) was used to measure the aspiration pressure (ΔP) that ranged from 1 to 7 kPa. After aspirating a nucleus, an equilibration time of 5–10 s preceded image acquisition. Thereafter the pressure was incrementally increased and at each pressure level a snapshot was acquired for analysis of nuclear geometry (Fig. 1, a and b). Additional measurements over timescales up to 300–400 s showed no significant changes in nuclear geometry. The scale factor (μm/pixel) was calibrated with a stage micrometer (Granules, Pyser-SGI, Edenbridge, UK). Image analysis was performed using ImageJ (National Institutes of Health, Bethesda, MD). The shape of a nucleus was approximated as a prolate ellipsoid with equatorial and polar radii, R_a, R_b, and R_c, where the polar radius in the axial dimension, R_c, is assumed to be equal to R_b: R_a > R_b ∼ R_c. Fig. 1 c shows the relevant geometric parameters for a nucleus under aspiration. The shape of an aspirated nucleus was approximated by a spherical endcap with a radius, R₁, which is equal to the radius of the pipette, R_p; a cylinder of length L–R_p; and a prolate ellipsoid external to the pipette with radii R_a and R_b. With this approximation, the estimated volume of an unaspirated nucleus is

(1)

and under aspiration,

(2)

The optically resolvable surface area was approximated as

(3)

and under aspiration

(4)

Extension to confocal imaged microdeformation (CIMD)

Methods for confocal imaged microdeformation (CIMD) were previously described in brief (13). Here we present a more detailed account of the experimental setup. Accommodating the confocal microscope setup for micropipette aspiration required the design of new chambers to facilitate micropipette access. Open chambers that allow for entry of the micropipette at a moderate angle were constructed by carefully using a razor blade to excise the side of a two-well chamber with a borosilicate coverslip bottom (0.17-mm thick, Nalge Nunc, Slangerup, Denmark). Isolated nuclei were visualized in this modified chamber at room temperature.

To adapt CIMD for living cells, cells were plated on coverslips (20 × 0.17 mm) (LaCon, Staig, Germany) before transfection. Before imaging, a thin layer of vacuum grease was applied to the circumference of the holder in which the coverslip was mounted. A rubber O-ring was placed on top to contain the cell medium in this open chamber into which the pipette enters at an angle. Cells were immersed in a CO₂-independent medium (Minimum Essential Medium without phenol red containing 10% fetal bovine serum, 1% glutamine, 1% sodium pyruvate, 1% penicillin/streptomycin, and 1% 1M HEPES (Invitrogen, Carlsbad, CA)) and pipettes were backfilled with this medium. An objective heater (Bioptech, Butler, PA) allowed for imaging at 37°C. Buffer evaporation can result in increased salt concentrations with time and so the experimental timescale was limited to 15 min.

Micropipette aspiration was conducted as described above with the pipette mounted on a Zeiss Axiovert 200M laser scanning confocal microscope model No. 510 equipped with a META polychromatic multichannel detector and Zeiss Apochromat 40×/1.2W corr water immersion objective. The microscope, laser, and micropipette setup were mounted on an air-cushioned table that dampens vibrations, ensuring the pipette remained stationary during image acquisition. GFP was excited using the 488 nm laser line and CFP with the 458 nm laser line of an argon laser. DiIC18 and SYTOX were excited using a He-Ne laser (543 nm). Covisualization of fluorophores was performed using multitrack mode exciting GFP at 458 nm and SYTOX at 543 nm. The pinhole size was typically set to one Airy unit and the confocal slice thickness to 0.75 μm. Three-dimensional reconstructions were generated and analyzed using Zeiss LSM 510 v.3.0 software to circumvent the apparent intensity gradient in a single confocal section that can result from the non-zero angle of the micropipette in the focal plane (Fig. 2 a). Analysis of three-dimensional reconstructions allows for more accurate estimations of nuclear geometry where the polar axis of the ellipsoid in the axial dimension is now defined as R_c. Using CIMD, nuclear volume was thus estimated as

(5)

and under aspiration,

(6)

Geometric parameters were determined with 15% accuracy. All reported values are the mean value ± SD among a population of n nuclei. Each experiment was performed at least three independent times.

Characterizing fluorescence intensity profiles

Inspired by fluorescence-imaged microdeformation (21) and building on previous work (13), we determined nuclear envelope properties by analyzing the fluorescence intensity along the deformed tongue of three-dimensional reconstructions obtained by CIMD. In practice, we made two-dimensional projections of the three-dimensional reconstructions (Fig. 2 a) and constructed pixel intensity profiles toward the tip of the tongue (Fig. 2 c). Analysis of the three-dimensional reconstruction is essential as it allows us to capture the intensity changes of the fluorophore along the deformed tongue that extends through several confocal planes. Investigating the intensity distribution of single confocal slices shows that GFP-LamA was largely localized at the nuclear envelope, and was also found in the inner nucleus (Fig. 4 a). Intranuclear GFP-LamA thus contributes to the overall intensity; however, the signal from these intranuclear lamins (with the exception of bright foci) was typically weak compared to that of the nuclear envelope-associated lamins and remained constant over the experimental timescale. Hence, to a good approximation, the GFP-LamA intensity profile across a three-dimensional reconstruction reflects the intensity at the nuclear envelope.

The observed intensity profiles were interpreted using nonlinear, two-dimensional elasticity theory to obtain information about the material properties of the nuclear envelope. For a two-dimensional elastic shell deformed into a micropipette, the density, ρ, varies along the cylindrical axis of deformation, x (Fig. 2 b; see also Appendix for detailed derivation) (13),

(7)

Here R_p is the radius of the pipette and the parameter a reflects the elastic properties of the underlying material. As the fluorescence intensity is proportional to the underlying density of fluorophore (22), Eq. 7 was fit to this profile (Fig. 2 c) to yield a value of a. Together with the ratio of the relative intensities in the tip of the tongue and the bulk nucleus (unperturbed nuclear envelope external to the pipette), ρ_i/ρ₀ (Fig. 2 d), a is related to the ratio of the area expansion (K) to shear moduli (μ) and is thus a signature of the elastic properties of the underlying material. The relationship between a and K/μ for a characteristic observed intensity ratio, ρ_i/ρ₀ = 0.62, is shown in Fig. A1 (Appendix).

Experimentally, the ratio of the mean intensities in the tip and the bulk, ρ_i/ρ₀, was obtained by constructing pixel intensity histograms for a circular area of the inner nucleus in the tip of the tongue and bulk nucleus (Fig. 2 a). Measurable background intensity was subtracted from the obtained pixel intensities. Pixel intensity histograms typically revealed a main peak with a relatively sharp Gaussian distribution, and an asymmetric tail extending to higher intensities containing contributions from brighter foci. In this way, heterogeneities in intensity could be separated from the mean intensities in the tip, ρ_i, and bulk, ρ₀. The mean intensity of the bulk nucleus, ρ₀, was consistent with the intensity of the unaspirated nucleus, and therefore represents the density of the unperturbed nucleus. Nonlinear regression analysis was used to obtain values for the a parameter.

Characterizing nuclei under micropipette aspiration

We used bright-field micropipette aspiration to determine how isolated nuclei respond to deformation. To investigate the response of specific nuclear structures, we fluorescently labeled particular components of HeLa cell nuclei, both isolated and in intact living cells, and obtained images by CIMD.

Physical properties of the nuclear envelope

To determine the physical properties of the nuclear envelope itself, we analyzed changes in the apparent surface area of nuclei. We also compared how fluorescently labeled nuclear membranes versus lamina respond to deformation using isolated HeLa cell nuclei and CIMD. In addition, we applied the CIMD method (13) to quantitatively analyze nuclear envelope properties in intact, living MEFs lacking the nuclear envelope protein, emerin. Nuclei in emerin-deficient cells serve as a model for abnormal nuclei often found in cells from Emery-Dreifuss muscular dystrophy patients.

A model for nuclear stability under aspiration

To properly interpret deformations of nuclei using micropipette aspiration demands an understanding of the physical forces that stabilize the nucleus in the pipette during aspiration. Here we elaborate on why micropipette aspiration of nuclei cannot be interpreted using conventional methods and describe a model for nuclear stability in the pipette.

In the analysis of classical micropipette experiments (21,29–31), the fluid vesicle or cell (erythrocyte) in the pipette is treated as an elastic shell encapsulating a liquid that has constant volume under aspiration. This provides the basis for relating the pressure difference, ΔP, to the in-plane stress distribution (29). For the case of a fluid membrane, this relationship can be written simply as

(8)

where ΔP = P₁ – P₄ is the pressure difference between the inner pipette and chamber, σ is the surface tension, R₁ is the radius of the pipette, and R₂ the radius of the sphere external to the pipette. The pressure difference, ΔP, is based on the fact that the inner material is liquid; that is, the pressure in the aspirated tongue, P₂, and cell body external to the pipette, P₃, are equal, P₂ = P₃. This model requires that volume remains constant throughout the duration of the experiment. For vesicles and some cells (erythrocytes), this condition can be obtained with osmotically equivalent sugar or buffer solutions internal and external to the membrane (32). Due to volume conservation, the deformation L is uniquely related to ΔP under equilibrium conditions, and ΔP is related to surface stress, or surface tension for a fluid membrane (Eq. 8). A correction for small volume changes (∼1%) in osmotically stabilized cells and vesicles has been considered (19,32).

Our observations show that there are large volume changes (60–70%) of nuclei during micropipette aspiration. For this reason, we cannot relate ΔP to the tension in the nuclear envelope. We have thus characterized nuclear deformations in terms of L/R_p and not ΔP. As conventional micropipette analysis cannot be applied, a new model is needed to describe the stability of the nucleus under aspiration. The phenomenological model presented here is based on our experimental observations and minimal a priori assumptions. We begin by summarizing some of our key experimental observations:

1. First, we show that nuclear volume decreased dramatically with aspiration pressure while nucleic acids remained in the nucleus. For each incremental pressure increase, a new equilibrium volume was obtained within the 5–10 s timescale. The volume reduction upon aspiration indicates that the unaspirated inner nucleus behaves as a largely aqueous, porous material with components that can easily be exchanged with the cytoplasm. These observations suggest that chromatin and other macromolecules remained in the inner nucleus, while water and smaller solutes (e.g., salts) were exchanged with the cytoplasm. Our results are in agreement with the current knowledge of the NPC that allows for free nucleocytoplasmic exchange of molecules smaller than 8 nm (∼60 kDa) and impedes passive transport of larger molecules across the nuclear envelope (33).

2. Nucleoli underwent small changes in shape and position under deformation, but remained relatively fixed in position with respect to each other, suggesting that the inner nucleus has properties of a solid material.

3. Upon aspiration, fluid membrane blebs emerged from the nuclear envelope into the pipette. The membrane blebs had a characteristic tear-drop shape as they detached from the nuclear envelope, indicating that these blebs were subject to a gentle aqueous flow (27) through the nuclear envelope into the pipette. During a series of incremental pressure increases, the general behavior of the blebbing vesicles remained similar, suggesting that flow through the nucleus did not change significantly with increasing aspiration pressure.

4. Volume loss stabilized after nuclear volume was reduced by ∼60–70%. At these large volume reductions, the nucleus was resistant to further compression, preventing the complete collapse of nuclear volume and entry of the nucleus into the pipette (for R_p < R_b).

5. Intensity analysis revealed that the relative intensity drop of GFP-LamA between the tip and the bulk, ρ_i/ρ₀, did not depend significantly on the degree of deformation (L), even for the same nucleus. This indicates that the nuclear envelope was stretched toward the tip region, yet was not maximally extended since the nuclear envelope at the tongue tip remained intact (no ruptures nor tears) on the length scale of optical resolution. In the case of maximal extension of the nuclear lamina meshwork in the tongue tip we would expect to see one of the hallmark signs: 1), ruptures or tears; or 2), a plateau in fluorescence into the tongue tip (34). We do not observe either sign of maximal network extension, indicating that maximal network expansion at the tip is not what limits further aspiration of the nucleus. Within the framework of the elastic model of the nuclear envelope (Appendix), this indicates that the pressure drop over the tip of the deformed tongue did not change significantly with L. Indeed, the presence and shape of the blebbing vesicles remained similar with increasing aspiration pressure, indicating that the pressure difference over the tip did not change significantly with increasing deformation. These observations suggest that the tension in the tongue tip did not change markedly during a series of incremental increases in aspiration pressure. Furthermore, there was no correlation between L/R_p for different cell types with altered nuclear envelope elasticity (Table 1).

Collectively, our data suggest that it was the extent of inner nuclear compression rather than nuclear envelope stretching that determined the degree of nuclear deformation (L/R_p) at different aspiration pressures. Our observations of relatively constant flow with increasing aspiration pressure (Observation 3) are consistent with this picture.

Here we present a new model that describes the relationship between mechanical equilibrium of the nucleus in the pipette, nuclear compression, and aqueous flow through the nucleus during aspiration. Although total nuclear volume is not conserved, a total pressure difference is maintained between the pipette and the exterior chamber. This pressure difference generates aqueous flow through the nucleus into the pipette, which we model through a series of filters that obey Darcy's law for aqueous flow. Fig. 1 c illustrates the pressures associated with the different regions of the micropipette setup, where P₁ > P₂ > P₃ ≥ P₄. We neglected the pressure difference between the bulk inner nucleus and the exterior chamber, ΔP_iii = P₄ – P₃, since the area of the nucleus external to the pipette is relatively large and the total area of NPCs renders it highly permeable. The negligible pressure difference, ΔP_iii, is further supported by the fact that the bulk nuclear envelope was not stretched significantly during aspiration.

In steady state, aqueous flow is described by Darcy's law,

(9)

where V_aq is the aqueous volume transported through the pipette and S₁ and S₂ are the filter coefficients with units [s × m⁴/kg] corresponding to flow through the nucleus. This aqueous flow is driven by the pressure difference ΔP_i = P₁ – P₂ between the pipette and the tongue, and ΔP_ii = P₂ – P₃ between the tongue and the nucleus external to the pipette. The total aspiration pressure can be controlled in the experiments, but neither ΔP_i nor ΔP_ii can be quantified. However, they are related by the pressure balance, , which reflects mechanical equilibrium of the nucleus in the pipette. ΔP_i and ΔP_ii depend on the material properties of the nuclear envelope and inner nucleus, which we do not attempt to model in detail. These pressure differences cause flow through the nucleus that is dictated by S₁ and S₂. The value S₁ depends on NPC architecture and the number of pores in the tip. In contrast, S₂ is characteristic of porous materials and changes significantly as the inner nuclear material is compressed.

Based on our experimental evidence (see Observations 1–3 above) and other studies (25,35), we considered that the inner nucleus contains a loose polymeric structure with aqueous voids that are compacted during compression. Water flow at low pressure gradients through gels and porous media is well described by Darcy's law, and for compressible gels, the filter coefficient changes with compression (36). A simple dimensional analysis shows that transport is limited by the size, ξ, of the aqueous voids in the inner nucleus, , where s₂ is a kinetic prefactor. The value ξ will decrease during compression if flexible macromolecules are a major component of the inner nucleus. From models of flexible polymers in solution, such as semidilute polymer solutions or polymer gels under compression (37,38), , where Π is the osmotic pressure of the inner nucleus. In the pipette setup, Π is controlled by ΔP_ii; thus, in a first approximation, Π(ΔP_ii) = Π(0) + Π′(0)ΔP_ii. Taking these considerations together, a relation for the pressure dependence of the filtration coefficient is

(10)

The relationship , together with Eqs. 9 and 10, can be used to obtain the functional behavior of ΔP_i and ΔP_ii: ΔP_i saturates on s₂k_BT/S₁Π′(0) as ΔP increases, while ΔP_ii grows approximately as ΔP. The total flow through the pipette does not change significantly with increasing ΔP, in agreement with the observations enumerated above. Compression of the inner nucleus also results in an overall stiffening of the nucleus that inhibits its advancement into the pipette, while the pressure over the tip changes very little, as there is no further extension of the lamina.

Physical properties of the inner nucleus

Our experiments and theoretical modeling indicate that the inner nucleus is primarily aqueous and, when compressed by up to 60–70% in volume, becomes resistant to further compression. The remaining, compressed inner nuclear structure that we observe may contain condensed chromatin and/or other nuclear structures including intranuclear lamins, actin, and a ribonucleoprotein network. Such a “nuclear matrix” has been postulated to constitute an insoluble part of the inner nucleus that remains after heavy extraction of isolated nuclei (41–43) (reviewed in (44–46)). The extent to which the nuclear matrix is an artifact of the preparation procedure has been a controversial topic. Our micropipette aspiration studies show that after a series of incremental pressure increases, the remaining inner nuclear structure is highly resistant to deformation and prevents entry of the entire nucleus into the pipette. The excellent agreement between our observations of the nucleus in living cells and isolated nuclei suggests that this structure is not merely an artifact of nuclear isolation.

Consistent with our observations that the nucleus becomes more resistant to deformation with decreasing nuclear volume, a study by Hancock (47) revealed the disassembly of subnuclear compartments upon an increase in nuclear volume. These findings are all consistent with a previous report that isolated Xenopus oocyte nuclei resist compression in the presence of high concentrations of dextran (12). That study attributed the observed compressibility limit to the nuclear lamina, but compression of the bulk nucleus could also cause resistance to compression.

Nuclear envelope mechanics

Here we show that the porous, primarily aqueous inner nucleus is encapsulated by a fluid-solid composite nuclear envelope: nuclear membranes have fluid characteristics and associate with a stable, two-dimensional scaffold of lamins and NPCs that exhibits resistance to shear. We demonstrate this quantitatively using CIMD and two-dimensional elasticity theory and qualitatively with observations of surface crumpling and buckling. Surface crumpling and buckling have previously been observed in membrane systems that have finite shear moduli (48,49). This response to tension reflects that the surface bending rigidity is not sufficient to prevent out-of-plane deformation when the surface is subjected to in-plane stress (50). These observations as well as the observed intensity decay at larger deformations are consistent with the behavior of a two-dimensional solid-elastic shell. Interestingly, an intensity decay in the fluorescently labeled lamina is observed in different cell types from different species, suggesting that the solid-elastic behavior of the nuclear envelope is a universal feature of the eukaryotic cell nuclear envelope during interphase. We observed a decrease in GFP-LamA intensity into the deformed tongue in HeLa cells (derived from human cervical cancer cells) as well as in MEFs both with and without the nuclear envelope protein, emerin. Similar behavior has also been observed in GFP-Lamin-B-labeled nuclei isolated from TC7 cells (subline of African green monkey kidney epithelium cells) (11). Although nuclear envelope behavior is qualitatively similar in different nuclei, we also observe variations in nuclear envelope material properties (K/μ). Why K/μ varies among cell types is not known, but we speculate that these differences reflect cell- and tissue-specific functions.

Interplay between the nuclear envelope and the inner nucleus

The nuclear lamina is linked to both nuclear membranes and chromatin via various binding proteins (3,4), including nuclear lamins, yet the physical nature of this coupling has not been well characterized; discerning the relative contributions of the nuclear envelope and inner nucleus to nuclear mechanics is thus challenging. Our results reveal that the nuclear envelope and inner nucleus have distinctly different physical properties, and that each dominates nuclear response to small versus large deformations. We show that the nuclear envelope behaves as a two-dimensional material comprised of a solid-elastic lamina and fluidlike membranes, while the inner nucleus has characteristics of a largely aqueous, highly compressible gel. Based on our experimental observations and analytical considerations of pressure differences and fluid flow through the nucleus, our integrated understanding of nuclear behavior during microdeformation is as follows: initially at small deformations (values for L/R_p), nuclear response is governed by the behavior of the nuclear envelope, evidenced by surface buckling and crumpling. However, as nuclear volume decreases with increasing values of L/R_p during aspiration (Fig. 3), the extension of the nucleus into the pipette becomes limited by compression of the inner nucleus. As a consequence, differences in the maximal extension, L/R_p, of individual nuclei into the pipette may be attributed to variations in the initial state of inner nuclear architecture and/or nuclear volume, rather than nuclear envelope elastic properties.

To better understand the contributions of the nuclear envelope and inner nucleus to nuclear mechanics requires further experiments, yet this work remains challenging. For example, the physical properties and morphology of isolated nuclei are extremely sensitive to buffer conditions (unpublished observations) (11), suggesting that salts can induce changes in the structure (chromatin packing) (51) and mechanical properties of the inner nucleus.

Nuclear stability and disease

In this study, we applied the CIMD technique to show that emerin-deficient cells have significantly altered nuclear envelope mechanics, as the value of the area expansion to shear moduli ratio (K/μ) is less than half that of wild-type cells. This is direct evidence that loss of emerin alters nuclear envelope elastic properties. Indeed, electron microscopy studies of cells from X-EDMD patients revealed numerous nuclei with nuclear envelope fragmentation (52), and nuclear membrane breakdown in the form of chromatin-filled blebs that extrude from the nucleus (53). These structural abnormalities may be linked to altered nuclear mechanics and/or dynamics as seen by studies of nuclear shape variation and response to biaxial strain of emerin-deficient MEFs (8). Ultimately these observed changes in nuclear shape stability and nuclear envelope material properties reflect altered ultrastructure. The loss of emerin could create holes or vacancies in the lamina meshwork that alter nuclear envelope elastic properties. Emerin itself may also have a structural role in the nucleus organizing a cortical actin network at the nuclear envelope (54). As emerin is a binding partner for lamin proteins, its absence from the nuclear envelope (as in EDMD (55,56)) results in increased lamin solubility (52) and thus potentially a reduction of lamins at the nuclear envelope. It is well established that lamin A/C deficient nuclei are more susceptible to deformation by mechanical stress (1,2). Similarly, diseases of the erythrocyte membrane (e.g., hereditary ovalocytosis and elliptocytosis (26)) have been linked to changes in the connectivity of the membrane and underlying spectrin network and consequent alterations in membrane-spectrin material properties (K, μ) (57). The K/μ values for nuclei with and without emerin that we report here are important for future modeling and simulations of elastic networks to elucidate the mechanisms that govern the elasticity and lateral organization of the nuclear lamina.

How do altered nuclear envelope properties give rise to tissue-specific disease phenotypes? The observed abnormalities in nuclear envelope ultrastructure and material properties that result from, for example loss of emerin, could make nuclei less resistant to deformations in shape, leading to nuclear damage and cell death in mechanically strained tissue. Altered nuclear envelope elasticity may also affect force transmission from the cytoskeleton to the nucleus (39) as well as how forces are sensed within the nucleus, thereby modulating the response of mechanosensitive genes. In addition, emerin-mediated changes in intranuclear organization (58,59), transcriptional regulation, and gene expression (14) may also contribute to the pathological mechanism of nuclear envelope-related disease.