Abstract
Hutchinson–Gilford progeria syndrome, caused by a mutation in the LMNA
gene, leads to increased levels of truncated prelamin A, progerin, in
the nuclear membrane. The accumulation of progerin results in defective
nuclear morphology and is associated with altered expression of linker
of the nucleoskeleton and cytoskeleton complex proteins, which are
critical for nuclear signal transduction via molecular coupling between
the extranuclear cytoskeleton and lamin‐associated nuclear envelope.
However, the molecular mechanisms underlying progerin
accumulation‐induced nuclear deformation and its effects on
intranuclear chromosomal organization remain unclear. Here, the
spatiotemporal evolution of nuclear wrinkles is analyzed in response to
variations in substrate stiffness using a doxycycline‐inducible
progerin expression system. It is found that cytoskeletal tension
regulates the onset of progerin‐induced nuclear envelope wrinkling and
that the molecular interaction between SUN1 and LMNA controls the
actomyosin‐dependent attenuation of nuclear tension. Genome‐wide
analysis of chromatin accessibility and gene expression further
suggests that an imbalance in force between the intra‐ and extranuclear
spaces induces nuclear deformation, which specifically regulates
progeria‐associated gene expression via modification of
mechanosensitive signaling pathways. The findings highlight the crucial
role of nuclear lamin–cytoskeletal connectivity in bridging nuclear
mechanotransduction and the biological aging process.
Keywords: actomyosin contractility, chromatin remodeling,
heterochromatin, Hutchinson–Gilford progeria syndrome, LINC complex,
mechanosensation, nuclear deformation, nuclear tension, nuclear
wrinkling, progerin, SUN1
__________________________________________________________________
The premature aging‐related progerin leads to defective nuclear
morphology and is associated with disrupted molecular coupling between
the extranuclear cytoskeleton and lamin‐associated nuclear envelope. It
is discovered that progerin expression reduces nuclear tension, forms
nuclear wrinkling, and enhances chromatin dynamics, thereby regulating
progerin‐induced mechanosensitive signaling pathways.
graphic file with name ADVS-12-2502375-g003.jpg
1. Introduction
Hutchinson–Gilford progeria syndrome (HGPS), a premature aging disease,
is caused by a de novo point mutation (G608G; GGC > GGT) in the LMNA
gene encoding lamins A and C, creating a truncated prelamin A form
lacking 50 amino acid residues near the C terminus, commonly referred
to as Δ50 lamin A or progerin.^[ [42]^1 ^] These amino acids include a
cleavage site for the zinc metallopeptidase STE24 (ZMPSTE24), which is
crucial in the formation of mature lamin A by removing the farnesyl
group; loss of this cleavage site leads to a persistent farnesylation
of progerin.^[ [43]^2 ^] Accumulation of progerin in the nuclear
envelope (NE) compromises the structural and functional integrity of
the nucleus, resulting in an abnormal nuclear shape, a hallmark of
HGPS.^[ [44]^3 ^] For instance, fibroblasts obtained from patients with
HGPS display defective nuclear shapes, including invaginations,
lobulations, and wrinkles.^[ [45]^3 , [46]^4 ^] Furthermore,
overexpression of Δ50 lamin A in primary dermal fibroblasts induces
abnormalities in nuclear morphology similar to those observed in HGPS
cells.^[ [47]^5 ^]
Progerin expression also leads to altered molecular connections between
the nucleus and cytoskeleton,^[ [48]^6 ^] further resulting in enhanced
sensitivity to mechanical stress^[ [49]^7 ^] and reduced force
propagation from the extracellular matrix (ECM) into the cell.^[ [50]^8
^] ECM stiffness, which reflects tissue‐specific biomechanical
properties, is detected by transmembrane receptor proteins, such as
integrins,^[ [51]^9 ^] which mediates the formation of focal adhesion
complexes. These complexes transmit mechanical signals into
intracellular biochemical pathways, thereby inducing diverse cellular
responses.^[ [52]^10 ^] Moreover, these adhesion complexes promote the
polymerization of intracellular actin cytoskeletal networks, where
actomyosin contractility‐induced mechanical forces allow cells to sense
and respond to the biomechanical characteristics of their surrounding
environment.^[ [53]^11 ^] These force‐transmitting molecular
connections are largely mediated by the linker of the nucleoskeleton
and cytoskeleton (LINC) complexes.^[ [54]^12 ^] The LINC complex is
composed of a SUN (Sad1, UNC‐84) domain located in the inner nuclear
membrane and a KASH (Klarsicht, ANC‐1, and Syne Homology) domain
located in the outer nuclear membrane.^[ [55]^13 ^] Nuclear envelope
spectrin‐repeat proteins (nesprins), consisting of actin‐binding
N‐terminal and KASH domain‐binding C‐terminal domains, also bind to SUN
proteins in the perinuclear space between the nuclear membranes.^[
[56]^14 ^] Thus, the LINC complex transmits biophysical stimuli into
the nucleus through nesprin‐mediated molecular connections between the
actin cytoskeleton and SUN proteins.^[ [57]^15 ^] In particular,
LINC‐mediated cytoskeletal tension plays a key role in cellular
mechanoresponses,^[ [58]^16 ^] where actomyosin contractility and
force‐dependent reorganization of the cytoskeletal architecture remodel
LINC‐associated protein components to transmit extracellular physical
stimuli into the nucleus.^[ [59]^17 ^]
Lamins interact with heterochromatic genomic regions via
lamina‐associated domains (LADs) that are enriched with repressive
histone marks, such as trimethylated histone H3 at lysine 9 (H3K9me2/3)
and H3K27me3.^[ [60]^18 ^] Transcriptional repression of genes within
such heterochromatic regions is attributed to the loss of DNA–nuclear
lamina interactions that contribute to chromatin condensation and the
expression of H3K9me2/3.^[ [61]^19 ^] Thus, lamin‐associated
intranuclear chromatin organization correlates with nuclear stiffening,
which is crucial for maintaining nuclear structural integrity in
response to mechanical stress.^[ [62]^20 ^] Because NE‐associated SUN
proteins directly bind chromatin to the nuclear periphery,^[ [63]^21 ^]
impaired LINC complexes hinder intracellular force transmission, which,
in turn, disrupts perinuclear cytoskeleton‐dependent mechanosensing of
the extracellular microenvironment. Ultimately, mechanosensitive
nuclear deformation compromises intranuclear elasticity distribution,^[
[64]^22 ^] indicating the remodeling of heterochromatin accessibility.
Reduced chromatin mobility by the inhibition of actomyosin
contractility or nuclear detachment from the cytoskeleton further
highlights the critical role of an intact actomyosin apparatus and the
LINC complex in mechanical signal transduction to chromatin.^[ [65]^22
, [66]^23 ^]
Progerin expression disrupts the LINC complex by altering the
expression of SUN proteins without affecting the localization of SUN1
and SUN2 to the nuclear membrane,^[ [67]^24 ^] resulting in altered
cellular mechanotransduction.^[ [68]^25 ^] In NIH 3T3 fibroblasts
expressing myc‐tagged progerin and in HGPS fibroblasts, enhanced SUN1
expression in response to progerin accumulation alters nuclear coupling
to actin filaments (F‐actin) and microtubules (MT) through nesprin 2,
resulting in defective nuclear movement and cell polarity.^[ [69]^26 ^]
SUN1 overexpression in HGPS fibroblasts can further lead to increased
SUN1–nesprin 2 coupling with MT, thereby inhibiting nuclear movement
due to its complementary coupling with the actin cytoskeleton.^[
[70]^26b ^] Consequently, nesprin 2‐mediated actomyosin tension applied
to the NE is reduced,^[ [71]^27 ^] confirming a decrease in nuclear
forces in response to SUN1 overexpression in progerin‐expressing cells.
Moreover, SUN1 silencing in HGPS fibroblasts rescues the deformed
nuclear morphology,^[ [72]^28 ^] whereas SUN2 depletion fails to
restore progerin‐induced nuclear deformation.^[ [73]^4 ^] These
observations highlight the role of SUN1 upregulation in nuclear
deformation as a molecular hallmark of the pathological signature of
HGPS.
Progerin expression increases F‐actin polymerization^[ [74]^29 ^] and
RhoA activation,^[ [75]^30 ^] resulting in cytoskeletal stiffening.^[
[76]^7 ^] For instance, F‐actin polymerization is increased in
mesenchymal stromal/stem cells from the ZMPSTE24 ^−/− HGPS mouse model,
Z24^−/− mesenchymal stromal cells (MSCs), which exhibit a higher
elastic modulus than wild‐type MSCs. Increased RhoA activity in Z24^−/−
MSCs is suppressed by pharmaceutical inhibition of RhoA signaling,
resulting in reduced nuclear deformation.^[ [77]^30 ^] Consistent with
the results from Z24^−/− MSCs, increased F‐actin polymerization and
nuclear deformation in HGPS fibroblasts are reversed by treatment with
a farnesyltransferase inhibitor.^[ [78]^30 , [79]^31 ^] Elevated
cytoskeletal tension in progeria cells arises from mechanical stress
due to both ECM stiffness^[ [80]^32 ^] and increased nuclear rigidity,
as sustained RhoA signaling promotes F‐actin cytoskeletal stiffness.^[
[81]^30 ^] Progerin‐induced alterations in nucleocytoskeletal
connections result in reduced propagation of cytoskeletal forces to the
nucleus,^[ [82]^8 ^] as determined by particle tracking analysis in
exogenous progerin‐expressing HeLa cells, human umbilical vein
endothelial cells, and HGPS fibroblasts.^[ [83]^8 ^]
Accumulating evidence suggests that nucleus‐responsive forces can alter
epigenetic modifications through chromatin remodeling;^[ [84]^20 ^]
mechanosensitive changes in forces applied to the nucleus can alter
intranuclear heterochromatin reorganization and chromatin
accessibility. For instance, dermal fibroblasts derived from patients
with HGPS show changes in DNA methylation, and chromatin accessibility
is enriched in NE‐associated regions, contributing to the abnormal gene
expression patterns observed in HGPS.^[ [85]^33 ^] Therefore, progerin
accumulation leads to widespread alterations in the repressive histone
mark H3K27me3, disrupting the association between heterochromatin and
the nuclear lamina in HGPS fibroblasts, resulting in the loss of
chromosomal compartmentalization.^[ [86]^34 ^]
Previously, studies on HGPS models have consistently demonstrated
increased cytoskeletal tension and RhoA activation,^[ [87]^30 ^] which
are essential to regulate force transmission to the nucleus. In turn,
nuclear forces alter epigenetic modifications through chromatin
remodeling.^[ [88]^20 , [89]^33 ^] However, the causal relationship
between progerin‐induced nuclear–cytoskeletal force transmission and
nuclear deformation in progerin‐expressing cells, as well as the
consequent alterations in gene expression remains to be identified.
While previous studies have shown that nuclear deformation can be
regulated by overexpressed SUN1, the precise mechanism by which SUN1
overexpression induces nuclear deformation remains poorly understood.
Therefore, here, we investigated the biophysical mechanisms that
regulate nuclear morphological changes in response to progerin
expression and their downstream effects on gene expression. To this
end, we developed a human HGPS cell model, where the progerin
expression was precisely controlled using the doxycycline‐induced
Tet‐On system, overcoming the constraints of using the primary cells
from patients with HGPS owing to their limited availability. We
elucidated the mechanism by which progerin expression alters nuclear
morphology over time. We further investigated how progerin accumulation
at the nuclear membrane in HGPS cells affects LINC complex expression
and the physical connectivity between the actin cytoskeleton and
nuclear lamina. Furthermore, we examined the impact of these nuclear
morphological changes on chromatin organization and the aberrant gene
expression profiles in HGPS. Our study improves our understanding of
the progerin‐induced alteration of nuclear tension that drives nuclear
deformation.
2. Results
2.1. Mechanosensing of Substrate Stiffness Modulates Progerin‐Induced Nuclear
Deformation
HGPS, characterized by abnormal nuclear morphology due to the
expression of truncated prelamin A (Δ50 LMNA/progerin),^[ [90]^35 ^]
displays impaired nuclear mechanotransduction.^[ [91]^26b ^] To
elucidate the molecular mechanism by which progerin accumulation alters
mechanical signal transduction through nuclear deformation, we
developed an HGPS model in HeLa cells, where the onset of progerin
expression was precisely controlled by transfection with the XLone
plasmid that combined the piggyBac transposon and Tet‐On 3G
doxycycline‐inducible gene expression system (see the Experimental
Section and Figure [92]S1, Supporting Information).^[ [93]^36 ^]
The treatment of Tet‐On gene‐expressing HeLa cells with 2 µg mL^−1
doxycycline specifically induced progerin expression and dramatically
increased the population of abnormal nuclei with blebs, lobulation, and
wavy surface texture, while the overall cell and nuclear size remained
unchanged (Figure [94]1A–E; Figure [95]S2A, Supporting Information). We
noted that traditional nuclear morphometry measurements, including
nuclear area, circularity, and aspect ratio (defined as nuclear
spreading area, 4π(area)/(perimeter^2), and the ratio of the longest
axis to the perpendicular shortest axis, respectively), were
insufficient to quantitatively assess the nuclear deformation induced
by progerin expression (Figure [96]1B,C; Figure [97]S2B,C, Supporting
Information). Hence, we assessed the NE wrinkling area by measuring the
fraction of the nuclear surface area occupied by NE wrinkles to more
precisely represent progerin‐induced NE‐defective nuclei (Figure
[98]S2D, Supporting Information). To further confirm that
doxycycline‐induced progerin expression controls the progression of
nuclear deformation, we systematically quantified progerin expression
and NE wrinkling at different time points, where we noted that the
time‐dependent elevation of progerin expression preceded the increase
in NE wrinkling (Figure [99]S3, Supporting Information). These results
demonstrate that our doxycycline‐controlled progerin‐expressing Tet‐On
HeLa cell line not only replicates the deformed nuclear morphology
detected in HGPS cells but also enables time‐dependent monitoring of
progerin expression and nuclear deformation.
Figure 1.
Figure 1
[100]Open in a new tab
Substrate stiffness‐dependent differential evolution of nuclear
deformation in doxycycline‐inducible progerin‐expressing HeLa cells.
A–E) Morphological alterations of doxycycline‐controlled Tet‐On HeLa
cells expressing mutant lamin A protein (Δ50 LMNA/progerin).
Representative confocal images depict immunofluorescence staining for
progerin (red), F‐actin (green), and nuclei (DAPI, blue) in
doxycycline‐untreated control cells (−Dox) or doxycycline‐treated (2 µg
mL^−1) progerin‐expressing Tet‐On HeLa cells (+Dox). Hemispherical and
cross‐sectional views of 3D‐rendered nuclei show progerin
expression‐induced formation of abnormal nuclear morphology. Empty and
full arrowheads indicate the absence and presence of progerin
expression, respectively (A). Immunofluorescence intensity‐based
quantifications of cell area (B), nuclear area (C), progerin expression
(D), and the fractional occurrence of abnormal nuclear shapes (E) were
performed in the absence and presence of doxycycline treatment. In
panel B, 810 and 505 nuclei; in panel C, 138 and 153 nuclei; in panel
D, 93 and 72 nuclei were analyzed under −Dox and +Dox conditions,
respectively. For panel E, 70 to 105 nuclei were analyzed, which was
independently repeated three times per each condition. Error bars
indicate the standard error of the mean (S.E.M.); an unpaired t‐test
was applied (***: p < 0.001, NS: not significant). F–G) Substrate
stiffness‐dependent differential expression of progerin. Progerin
(red), lamin B1 (green), and nuclei (DAPI, blue) of progerin‐expressing
Tet‐On HeLa cells were plated on control glass substrates or
polyacrylamide hydrogel (PAG) substrates with elastic moduli of 34 kPa
(stiff) and 1.37 kPa (soft) (F). Control glass and stiff PAG substrates
maintained doxycycline‐inducible progerin expression, while soft PAG
substrates significantly reduced progerin expression (G). In panel G,
>50 cells were tested per condition. Error bars indicate the S.E.M.;
Student's t‐test was applied (***: p < 0.001, NS: not significant).
H–J) Time‐lapse monitoring of substrate stiffness‐dependent progerin
expression and nuclear deformation. Progerin intensity and nuclear
envelope (NE) wrinkling were monitored every 20 min for 36 h in
mCherry‐progerin‐expressing Tet‐On HeLa cells plated on control glass
(H), stiff PAG (I), and soft PAG (J) substrates. Yellow dotted lines
mark the nuclear boundary determined by differential interference
contrast (DIC) imaging, showing nuclear spreading area. Full and empty
arrowheads indicate the presence and absence of progerin (red) or NE
wrinkling (blue), respectively; transparency of the full arrowheads
represents the magnitude of progerin expression and NE wrinkling;
oversaturated fluorescence intensity (white) indicates nuclear surface
wrinkles (H–J). K–P) Quantifying differential onset of progerin
expression and nuclear deformation in response to changes in substrate
stiffness. Progerin expression and NE wrinkling were quantified by
measuring fluorescence intensity (red curves) and the fraction of the
nuclear spreading area occupied by NE wrinkling area (blue curves)
after doxycycline treatment (K–M). All values were normalized using the
formula (x – x [min])/(x [max] – x [min]) to range from 0 (min) to 1
(max). Progerin expression and NE wrinkling followed extended sigmoidal
curves, with inflection points for progerin expression at 20 h and NE
wrinkling at 25 h on control glass and stiff PAG substrates (K,L,N,O),
which were delayed to 22 and 32 h, respectively, on soft PAG substrates
(M–O). The time interval between inflection points of progerin
expression and NE wrinkling extended from 5 h on control glass and
stiff PAG substrates to 10 h on soft PAG substrates (P). In panels K–P,
>20 cells were analyzed per condition. Error bars indicate the S.E.M.;
one‐way analysis of variance (ANOVA) with Tukey's test was used for
comparisons (***: p < 0.001, NS: not significant).
Mechanosensing of matrix rigidity alters protein expression by
remodeling the nucleus–cytoskeletal connection, mediating the
intracellular mechanical balance between cytoskeletal force and nuclear
surface tension.^[ [101]^37 ^] Accordingly, we investigated whether
cells adapted to differential substrate stiffness could modulate
progerin expression in doxycycline‐activated HeLa cells
(Figure [102]1F,G). Cells placed on control glass and polyacrylamide
hydrogel (PAG) substrates, mimicking the in vivo elastic moduli of
rigid and compliant organs,^[ [103]^38 ^] displayed deformed nuclear
shapes in response to doxycycline‐induced progerin expression,
regardless of substrate stiffness (Figure [104]1F). In contrast to
doxycycline‐untreated cells not expressing progerin,
doxycycline‐treated cells placed on rigid PAG substrates (E ≈ 34 kPa)
maintained elevated progerin expression similar to cells placed on
control glass substrates, while those grown on soft PAG substrates (E ≈
1.37 kPa) showed a significant reduction in progerin expression
(Figure [105]1G). This result strongly suggests that substrate
stiffness‐dependent changes in intracellular tension could regulate
progerin expression.
Since doxycycline‐controlled progerin expression led to nuclear
deformation in a time‐dependent manner (Figure [106]S3, Supporting
Information), and progerin expression levels were altered by substrate
stiffness (Figure [107]1F,G), we assessed whether differential
substrate stiffness could modulate the temporal relationship between
doxycycline‐induced progerin expression and nuclear deformation. To
this end, we conducted real‐time live cell monitoring of mCherry‐tagged
progerin‐expressing Tet‐On HeLa cells placed on substrates with varying
stiffness (Figure [108]1H–J; Movies [109]S1 and [110]S2, Supporting
Information). Time‐lapse imaging revealed that doxycycline‐induced
progerin expression exhibited sigmoidal changes, followed by NE
wrinkling, which was most evident on control glass substrates, but
attenuated on reduced substrate stiffness (Figure [111]1H–M).
Assessment of the inflection time points in the curves further revealed
that progerin expression and nuclear deformation were delayed in cells
placed on soft PAG substrates compared to those placed on control glass
or stiff PAG substrates (Figure [112]1K–M). Specifically, the
inflection points for progerin expression and NE wrinkling occurred at
20 and 25 h on control glass and stiff PAG substrates, respectively,
whereas they were observed at 22 and 32 h on soft PAG substrates
(Figure [113]1K–M). Consequently, the time interval between the
inflection points of the curves for progerin expression and NE
wrinkling was extended from 5 h on control glass and stiff PAG
substrates to 10 h on soft PAG substrates (Figure [114]1N–P).
These results indicate that doxycycline‐induced progerin expression,
followed by NE wrinkling, is highly mechanosensitive to changes in
substrate stiffness, which further suggests that progerin‐induced
nuclear deformation can be regulated by intracellular cytoskeletal
tension applied to the nucleus.
2.2. Mechanosensing of Substrate Stiffness Induces the Differential
Attenuation of Nuclear Tension in Progerin‐Expressing Cells
Previously, we demonstrated that substrate stiffness‐dependent spatial
reorganization of lamin A/C is regulated by cytoskeletal tension,^[
[115]^16 ^] and that 3D morphological alterations of the nucleus in
response to mechanical stimuli are regulated by nucleus–cytoskeletal
connections.^[ [116]^39 ^] Because the LINC‐mediated molecular
connection between the cytoskeleton and nuclear lamina facilitates
force transmission across the nuclear membrane,^[ [117]^21 ^] we
hypothesized that the nuclear tension between these elements could
determine the substrate stiffness‐modulated differential onset of
progerin‐induced nuclear deformation.
To directly measure the force applied to the nuclear–cytoskeletal
connection, we used a fluorescence energy transfer (FRET)‐based NE
tension sensor module tagging the inner nuclear membrane SUN‐binding
domain and cytoplasmic F‐actin‐binding domain, mimicking the LINC
complex associating nesprin 2^[ [118]^27 ^] (Figure [119]S4A,
Supporting Information). As an elevated FRET ratio in HGPS cells
implies diminished nuclear tension due to reduced actomyosin
contractility,^[ [120]^27 ^] we tested whether doxycycline‐induced
progerin expression could impinge on the substrate stiffness‐dependent
differential decline of nuclear tension. To monitor time‐dependent
gradual changes in NE tension, we transiently transfected the nesprin
tension sensor into progerin‐expressing Tet‐On HeLa cells before
placing them on substrates of varying stiffness (Figure [121]S4B,C,
Supporting Information). We confirmed that the transfected nesprin
tension sensors were specifically localized along the nuclear membrane
and that their fluorescence intensity was maintained without
significant decay during time‐lapse imaging of the nuclei every 6 h for
36 h (Figure [122]S4D,E, Supporting Information).
To analyze the substrate stiffness‐dependent differential application
of NE tension in response to doxycycline‐induced progerin expression,
we masked FRET signals located along the NE (Figure [123]2A–C), thereby
preventing interference with FRET signals stemming from other regions
of the cell (see the Experimental Section for details). FRET efficiency
was calculated by dividing the fluorescence intensity of the acceptor
by that of the donor after background subtraction within the masked NE
region, which was differentially color‐coded, where approaching purple
indicated a decreased FRET ratio due to enhanced NE tension, and
conversely, approaching red indicated an increased FRET ratio due to
reduced NE tension (Figure [124]2A–C).
Figure 2.
Figure 2
[125]Open in a new tab
Substrate stiffness‐dependent differential NE tension. A–C)
Visualization of substrate stiffness‐dependent changes in NE tension
using the nesprin tension sensor. Tet‐On HeLa cells expressing
fluorescence‐marker‐untagged progerin were transiently transfected with
the nesprin tension sensor and plated on control glass or stiff (E ≈ 34
kPa) or soft (E ≈ 1.37 kPa) PAG substrates. NE tension was analyzed
every 6 h for 36 h after doxycycline treatment. Binary layers outlining
the nesprin tension sensor‐localized nuclear membrane were determined
by creating polygonal hollow masks to exclude fluorescence intensity
outside the nucleus (top rows). FRET signals were differentially
color‐coded (bottom rows). Purple and red indicates high FRET
efficiency. FRET efficiency, representing the inverse of NE tension,
gradually increased in response to doxycycline‐induced progerin
expression, but this rate was reduced in cells on soft PAG substrates
compared to those on control glass and stiff PAG substrates (A and B vs
C). D–G) Quantification of substrate stiffness‐modulated differential
NE tension. Column scatter plots represent the time‐dependent increase
in the FRET ratio in response to doxycycline‐induced progerin
expression on control glass and stiff or soft PAG substrates. The
nesprin tension sensor‐based FRET ratio increased significantly after
24 h in cells on glass (D) and stiff PAG substrates (E), but after 30 h
in cells on soft PAG substrates (F). Accordingly, the FRET signal in
cells on glass and stiff PAG substrates was significantly higher than
in those on soft PAG substrates after 24 h (G). In panels D–G, red bars
represent the mean ± S.D., and one‐way ANOVA with Tukey's test was
applied for group comparisons (***: p < 0.001, **: p < 0.05, NS: not
significant).
The FRET ratio, representing the opposite of NE tension, significantly
increased in response to progerin expression, regardless of substrate
stiffness, after 24 h of doxycycline treatment on control glass and
stiff PAG substrates (Figure [126]2D,E), while the change was detected
after 30 h on soft PAG substrates (Figure [127]2F). Interestingly, the
temporal reduction in NE tension showed no difference between cells
placed on glass and stiff PAG substrates (Figure [128]2D,E,G).
Accordingly, a significant difference between the FRET ratios of cells
placed on rigid and soft substrates was observed after 24 h
(Figure [129]2G), which is consistent with a previous report showing
that HGPS cells displayed lower nuclear tension than normal control
cells.^[ [130]^27 ^]
Together with our results quantifying the substrate stiffness‐dependent
differential onset of progerin expression and nuclear deformation
(Figure [131]1K–P), these data further reveal the temporal relationship
between the progerin‐induced diminution of NE tension and nuclear
deformation. We found that the time point at which nuclear tension was
significantly reduced (i.e., a significantly increased FRET ratio)
appeared between two distinct time points at which an abrupt increase
in progerin expression and NE wrinkling were detected. This finding
strongly supports the hypothesis that progerin induces nuclear
deformation by reducing nuclear tension. In addition, we noted that the
time interval between the increment of progerin expression and the
reduction of nuclear tension, as well as the time interval between the
reduction of nuclear tension and the increase in NE wrinkling, were
doubled by relocating cells from the control glass and stiff PAG
substrates to the soft PAG substrates, i.e., from 4 to 8 h and from 1
to 2 h, respectively.
These data indicated that the time required to decrease nuclear tension
in response to progerin expression was approximately four times longer
than that required to induce nuclear deformation due to reduced nuclear
tension. Further, these findings strongly suggest that cytoskeletal
tension, modulated by changes in substrate stiffness, mediates nuclear
tension‐dependent nuclear deformation in progerin‐expressing cells.
2.3. Progerin‐Induced Reduction in Nuclear Tension Determines
Mechanosensitive Nuclear Wrinkling
The temporal onset of progerin expression, reduction in NE tension, and
nuclear wrinkling in response to doxycycline treatment were delayed by
relocating cells from stiff to soft substrates (Figures [132]1
and [133]2), which suggest that not only gene expression driven by
biochemical stimuli but also its biophysical consequences, such as
alterations in NE tension and nuclear surface remodeling, are sensitive
to the mechanical balance between the nucleus and the physical
environment of the cell. Specifically, we employed a computational
model to determine whether substrate stiffness‐dependent nuclear force
mediates the differential evolution of progerin‐induced nuclear
wrinkling.
To simulate NE wrinkling, a representative nuclear surface deformation
induced by progerin expression, we developed a computational model by
adapting nuclear geometry obtained from 3D‐reconstructed confocal
images to establish a soft, spherical, elastic, thin shell as the
initial model structure (Figure [134]3A). The shell was discretized
into triangular elements using a series of vertices to characterize the
deformation of the nuclear surface. The potential energy generated by
the movement of the vertices allows the shell to resist external forces
(see the Experimental Section for details). Because the FRET analysis
indicated reduced NE tension (tensile force) at the onset of progerin
expression (Figure [135]2), we simplified the force transmitting from
the external environment onto the nucleus, i.e., the cytoskeletal
force, as a net pressure force. Furthermore, by combining our previous
report demonstrating that the actin cytoskeleton exerts elevated
pressure on the nucleus on a stiff substrate,^[ [136]^40 ^] we applied
varying external pressures to model the nuclear response to changes in
substrate stiffness, where pressure increases with increasing substrate
stiffness.
Figure 3.
Figure 3
[137]Open in a new tab
Mechanical model of nucleus wrinkling. A) Construction of a mechanical
model mimicking progerin‐induced deformation of the nuclear surface.
Experimental observations of Tet‐On‐inducible progerin expression are
depicted by 3D reconstructions of confocal images (top) and the
corresponding mechanical model (bottom), showing the smooth spherical
shape of progerin‐absent control nuclei (left), a buckyball‐like
surface configuration at the onset of progerin expression (middle), and
a folded surface texture due to the progression of nuclear wrinkling
(right). The color code represents displacement of the nuclear surface.
B–E) Simulation of time‐dependent nucleus wrinkling in response to
changes in substrate stiffness. Reduced pressure on the nucleus on
compliant substrates delays surface wrinkling. The second‐order
buckling (i.e., from a buckyball‐like pattern to a folded pattern)
occurs at characteristic time scales of 45, 90, and 105 on stiff,
medium, and soft substrates, respectively (B). (Inset) The local force
balance in a membrane microelement, where membrane tension is mimicked
by an equivalent internal pressure due to surface curvature and
mechanical equilibrium. Time‐dependent volume changes of the nucleus on
different stiffness substrates indicate that nuclei on soft substrates
take the longest time, while those on stiff substrates take the
shortest, based on the characteristic time to reach a specific volume
change (C), where ∆V and V [0] represent the volume change and the
initial volume, respectively, and black circles mark the accelerated
collapses along the nucleus triggered by the second‐order buckling. The
nuclear surface tension, calculated from nuclear volume change, shows
that nuclei on stiff substrates have the highest tension value and
largest change rate, while nuclei on soft substrates have the lowest
tension value and smallest change rate (D), where the stretching force
F [e] was applied to characterize the membrane tension F [memb], F [0]
represents the unit characteristic force, and the black cross indicates
the breakdown of computational model due to the contact of the membrane
under large deformation, respectively. Increasing internal pressure
(i.e., enhanced membrane tension), corresponding to a greater 𝐾[𝑉]
value, indicating a stronger resistance to external pressure reduces
nuclear volume change (E), indicating that reduced nuclear tension on
soft substrates delays nuclear wrinkling compared to nuclei on stiff
substrates.
The nuclear mechanical model revealed that shell wrinkling progresses
in two stages: the smooth surface of the control nucleus first wrinkles
into a buckyball‐like structure, followed by a transition from the
buckyball‐like morphology to a labyrinthine pattern with deep
invaginations along the nuclear surface (Figure [138]3A; Movie [139]S3,
Supporting Information). As substrate stiffness decreased, the
buckyball‐to‐labyrinth transition was delayed at characteristic times
(t/t [0]) of 45, 90, and 105 for stiff, medium, and soft substrates,
respectively (Figure [140]3B), as measured by the relative volume
change (Δv/v [0], Figure [141]3C). The nuclear membrane surface tension
(F [memb]) quantitatively assesses the substrate‐stiffness‐dependent
differential nuclear deformation, with the occurrence of the maximum
value delayed by reducing substrate stiffness at t/t [0] values of 30,
75, and 100 for stiff, medium, and soft substrates, respectively
(Figure [142]3D). These results indicate that the shrinkage of the
nuclear shell volume and the elevation of nuclear membrane tension are
delayed by reducing substrate stiffness, consistent with our live‐cell
monitoring, which showed that the nucleus on stiff substrates favors
nuclear wrinkling in response to progerin expression (Figure [143]1).
Since we used an internal pressure to generate membrane tension (inset,
Figure [144]3B), a lower volume modulus (K [V]), corresponding to a
smaller ability to resist volume change, induces lower membrane
tension. This low membrane tension, i.e., low internal pressure,
accelerated nuclear wrinkling (Figure [145]3E; Movie [146]S3,
Supporting Information), confirms that a greater net external pressure
leads to faster and deeper wrinkling.
These results demonstrate that reduced nuclear membrane tension in
cells placed on soft substrates delays the occurrence of nuclear
wrinkling. They further suggest that nuclear wrinkling, induced by the
localized accumulation of permanently farnesylated prelamin A along the
nuclear membrane, could disrupt the nuclear–cytoskeletal connection
that is highly sensitive to changes in substrate stiffness.
2.4. Mechanosensitive Progerin Expression Induces Temporal Alterations in
Gene Profiling
Nuclear deformation is not merely a phenotypic signature of disease but
is strongly indicative of nuclear force‐dependent alteration of genome
architecture and dysregulation of gene expression.^[ [147]^41 ^] Since
progerin‐induced nuclear wrinkling is attributed to reduced NE tension
and this biochemical response is tightly regulated by mechanosensing of
substrate stiffness (Figures [148]1, [149]2, [150]3), we investigated
whether progerin‐induced temporal differential NE wrinkling could also
alter gene expression profiles in response to changes in substrate
stiffness.
To systematically analyze gene expression over time, we performed RNA
sequencing for cells cultured on either a stiff glass substrate or a
soft PAG substrate at 1.37 kPa, followed by doxycycline treatment at 6
h intervals for up to 36 h (Figure [151]4 ). Hierarchical clustering of
gene expression, determined by pairwise complete‐linkage clustering
analysis, revealed that the tested samples were primarily grouped by
doxycycline treatment time, irrespective of substrate stiffness (i.e.,
0, 6, 12, 18, and 36 h). However, only the samples treated with
doxycycline for 24 and 30 h formed distinct groups based on substrate
stiffness, i.e., soft 1.37 kPa substrates at 24 and 30 h versus stiff
glass substrates at the same time points (Figure [152]4A; Figure
[153]S5A, Supporting Information). Further analysis of the number of
differentially expressed genes (DEGs) in experimental conditions
compared to the doxycycline‐untreated control group (denoted as glass 0
h or 1.37 kPa 0 h) indicated that the largest changes in gene
expression levels, including both upregulation and downregulation,
occurred at 24 and 30 h after doxycycline treatment for each substrate
stiffness (Figure [154]4B). Interestingly, these time points coincided
with the reduction of NE tension and the formation of nuclear wrinkles,
occurring at 24 and 25 h on stiff substrates, or 30 and 32 h on soft
substrates, respectively (Figures [155]1 and [156]2). These results
imply that mechanosensitive progerin expression leads to differential
changes in gene expression as NE tension‐dependent nuclear wrinkling
progresses.
Figure 4.
Figure 4
[157]Open in a new tab
Time‐dependent alteration of gene profiling and signaling pathways in
response to mechanosensitive progerin expression. A–D)
Progerin‐induced, time‐dependent differential gene expression in
response to changes in substrate stiffness (stiff glass substrates vs
soft PAG substrates of 1.37 kPa). The similarity in gene expression
profiles was assessed by Euclidean distance and complete linkage
clustering, where the height of the dendrogram represents the Euclidean
distance between clusters, indicating the similarity in expression
profiles (A). Note that gene expression profiles remained similar
between cells on stiff glass and soft PAG substrates at 18 h but
clustered by substrate stiffness at 24 and 30 h (marked by red boxes),
and reclustered at 36 h. The number of differentially regulated genes
is displayed for each substrate stiffness by comparing with the
doxycycline‐untreated control condition (glass 0 h on the left, 1.37
kPa 0 h on the right), where yellow and blue bars indicate upregulated
and downregulated genes, respectively (B). Volcano plots display
doxycycline treatment time‐dependent evolution of log2 fold changes in
LMNA and LINC complex‐associated genes (e.g., SUN1, SUN2, SYNE1, SYNE2,
and SYNE3) on stiff glass substrates (top row) or soft PAG substrates
(bottom row) (C). Note that while LMNA expression increased from 6 h
after doxycycline treatment on both stiff glass and soft PAG
substrates, significant increases in LMNA, SYNE2, and SYNE3 were
observed at 12 h on stiff substrates but at 18 h on soft substrates
(marked by red dotted boxes), indicating delayed expression of nesprin
on soft substrates. Gene ontology (GO) analysis was performed comparing
the control glass and soft PAG substrates at different doxycycline
treatment times (12, 18, 24, 30, and 36 h) for cellular components,
biological processes, and molecular functions, with term sizes between
10 and 500 (D). In panel B, the criteria for significant changes in
gene expression were fold change ≥ |2| and raw p‐value < 0.05. In panel
C, yellow and red dots indicate specific gene expression levels
corresponding to fold change ≥ |1.5|, raw p‐value < 0.5, and fold
change ≤ |1.5| with raw p‐value < 0.5, respectively, with LINC
complex‐associated genes colored blue. In panel D, adjusted p‐values
reported from g:Profiler were derived using a one‐sided hypergeometric
test and corrected by the Benjamini–Hochberg method (***: p < 0.001,
**: p < 0.01, *: p < 0.05). E–H) Heatmap analysis of GO terms related
to mechanosensing of substrate stiffness. Representative signaling
pathways, including the Notch signaling pathway (GO:0007219, E), BMP
signaling pathway (GO:0030509, F), extracellular structure organization
(GO:0043062, G), and tissue homeostasis (GO:0001894, H), were
visualized. Euclidean distance was used as the distance metric, and
complete linkage was applied for hierarchical clustering in the
analysis of each dataset. For further details, refer to the
Experimental Section.
Since the LINC complex bridging between the extra‐nuclear–cytoskeletal
network and nuclear membrane‐associated proteins that physically
interact with chromosomal architecture regulates the transmission of
biophysical stimuli into the nucleus,^[ [158]^21 , [159]^41 ^] we
conducted RNA sequencing to investigate whether progerin expression
regulates the expression of LINC complex‐associated genes encoding SUN
and nesprin proteins. Compared to the doxycycline‐untreated control
conditions, our results showed a significant increase in the expression
of LMNA, SYNE2, and SYNE3 genes in response to increasing doxycycline
treatment across substrate stiffness, where we also observed
doxycycline treatment time‐dependent differential expression of these
genes (Figure [160]4C). For instance, LMNA gene expression was observed
at 6 h in both stiff and soft substrates, whereas SYNE2 and SYNE3
expression increased after 12 h of doxycycline treatment on stiff
substrates but after 18 h on soft substrates (Figure [161]4C). These
findings align with previous results showing that doxycycline‐induced
progerin expression (Figure [162]1), reduction of nuclear tension
(Figure [163]2), and nuclear wrinkling (Figures [164]1, [165]2, [166]3)
were delayed on soft substrates compared to cells on stiff substrates.
To further investigate whether substrate stiffness‐dependent
differential gene expression induced by doxycycline‐inducible progerin
expression could functionally alter mechanosensation‐mediated pathways,
we used the gene set enrichment analysis (GSEA) database (Figure
[167]S5B–E, Supporting Information). Specifically focusing on
actomyosin‐related genes (GSEA C5>GO, 286 genes) and nuclear
membrane‐related genes (GSEA C5>GO, 417 genes) obtained after 24 h of
doxycycline treatment, 13 actomyosin‐related genes and 17 nuclear
membrane‐related genes were identified as DEGs (Figure [168]S5B–D,
Supporting Information). Furthermore, the changes in the expression of
these DEGs occurred 24 h after doxycycline treatment on stiff
substrates and 30 h after doxycycline treatment on soft substrates
(Figure [169]S5E, Supporting Information).
Among the top 20 GO enrichment analysis results, based on adjusted
p‐values, our investigation of cellular components, biological
processes, and molecular functions identified a significant delay in
soft substrates compared to that in stiff substrates (Figure [170]4D).
These results were compared with RNA sequencing data from the gene
expression omnibus public datasets [171]GSE141950 and [172]GSE118633,
which analyzed dermal fibroblasts from healthy individuals and patients
with HGPS (Figure [173]S6, Supporting Information). [174]GSE141950
analysis revealed distinct gene expression patterns in fibroblasts from
patients with HGPS (Figure [175]6A,B, Supporting Information).
Furthermore, pathway enrichment analysis of cellular components,
biological processes, and molecular functions revealed similar changes
in these pathways (Figure [176]4D; Figure [177]S6C, Supporting
Information). Moreover, [178]GSE118633 analysis revealed distinct gene
expression patterns in fibroblasts from patients with HGPS, confirming
changes in the same pathways (Figure [179]4D; Figure [180]S6D–F,
Supporting Information). These results not only confirm that our
developed Tet‐On‐inducible progerin‐expressing HeLa cells effectively
model HGPS but also demonstrate that substrate stiffness‐dependent
differential progerin expression could modify the onset of gene
expression that regulates multiple signaling pathways.
Finally, we evaluated whether mechanosensitive progerin expression
could regulate Notch signaling and bone morphogenetic protein (BMP)
signaling. The Notch signaling pathway, highly sensitive to mechanical
signals, regulates cell and tissue fate in most tissues,^[ [181]^42 ^]
and the BMP signaling pathway, involving nuclear membrane proteins, is
directly regulated by mechanical signal transduction pathways without
autocrine ligands, occurring at the receptor, cytoplasmic, and nuclear
levels.^[ [182]^43 ^] Heatmap analysis confirmed that Notch and BMP
signaling were altered by doxycycline‐induced progerin expression in a
substrate stiffness‐dependent manner (Figure [183]4E,F). Furthermore,
we observed a distinct regulation of gene expression involved in
extracellular structure organization in response to changes in
substrate stiffness (Figure [184]4G), which is necessary for tissue
homeostasis.^[ [185]^44 ^] Heatmap analysis of tissue
homeostasis‐regulating differential gene expression further confirmed
that mechanosensitive progerin expression could functionally regulate
tissue homeostasis (Figure [186]4H).
Together, these results suggest that substrate stiffness‐dependent
differential progerin expression regulates gene expression, ultimately
altering mechanosensitive signaling pathways.
2.5. Mechanosensitive Progerin Expression Modulates the Spatiotemporal
Reorganization of Heterochromatin
Progerin exhibits a strong binding affinity for histone‐lysine
N‐methyltransferase SUV39H1, preventing its proteasomal degradation and
thereby increasing epigenetic modifications to the DNA packaging
protein histones, e.g., H3K9me3. This, in turn, reduces DNA repair
capacity and accelerates senescence.^[ [187]^45 ^] Meanwhile, elevated
substrate stiffness correlates with enhanced levels of H3K9me2/3.^[
[188]^46 ^] Thus, we investigated whether the substrate
stiffness‐dependent mechanosensitive alteration of progerin expression
could modulate H3K9me2/3 levels, indicative of heterochromatin
formation and transcriptional silencing.
To assess H3K9me2/3 protein expression levels, western blot analysis
was performed on cells cultured on either stiff glass substrates or
soft PAG substrates (1.37 kPa), which were treated with doxycycline at
12 h intervals for up to 36 h (Figure [189]5A–D). Cells cultured on
stiff glass substrates showed a significant increase in H3K9me2/3
expression 24 h after doxycycline treatment, with levels doubling by 36
h (Figure [190]5A,C). In contrast, cells cultured on soft PAG
substrates showed increased H3K9me2/3 expression after 24 h of
treatment, but the levels only slightly increased further by 36 h
(Figure [191]5B,D). These results suggest that temporal alteration of
histone modifications is also accompanied by the substrate
stiffness‐dependent differential progerin expression.
Figure 5.
Figure 5
[192]Open in a new tab
Time‐dependent differential epigenetic modifications in response to
mechanosensitive progerin expression. A–D) Quantification of
doxycycline treatment time‐dependent differential expression of
H3K9me2/3 in response to changes in substrate stiffness. H3K9me2/3
expression was quantified by immunoblotting against H3K9me2/3 and GAPDH
antibodies in progerin‐expressing Tet‐On HeLa cells placed on control
stiff substrates (glass, A) and soft PAG substrates (1.37 kPa, B) with
doxycycline treatment every 12 h for up to 36 h. (C,D) Total protein
expression increased with doxycycline treatment in each condition. In
panels C and D, three independently performed experiments were averaged
and normalized to the values in 0 h condition. Error bars indicate the
S.E.M., and one‐way ANOVA using Tukey's test was applied for comparison
between groups (****: p < 0.0001, ***: p < 0.005, *: p < 0.05, NS: not
significant). E–N) Spatiotemporal alterations of heterochromatic
histone modifications in doxycycline‐induced progerin‐expressing cells
placed on varying substrate stiffness. Tet‐On HeLa cells expressing
mCherry‐tagged Δ50 LMNA (red) placed on control glass (E) and PAG
substrates of 1.37 kPa (F) were immunostained for nuclei (DAPI, blue)
and H3K9me2/3 (green) every 12 h after doxycycline treatment.
3D‐rendered nuclei, reconstructed from z‐stacked confocal fluorescent
images, depict that doxycycline treatment time‐dependently increased
progerin expression, inducing H3K9me2/3 clustering in the nuclear
interior (E–N). H3K9me2/3 clusters were detected after 24 h on stiff
substrates (glass, E,G–J) but appeared after 36 h on soft substrates
(1.37 kPa, F,K–N). Yellow dotted lines indicate the nuclear boundary as
determined by DAPI staining; white arrowheads indicate clustered
H3K9me2/3 (E,F). Fluorescence intensity profiles monitored by line
scanning through the maximum intensity projected nuclear images show
that H3K9me2/3 clusters largely alternate with progerin staining. More
intensive peaks were detected in nuclei of cells placed on stiff
substrates compared to those on soft substrates (G–J vs K–N), where red
and green arrowheads indicate fluorescence intensity peaks
corresponding to progerin and H3K9me2/3 expression, respectively.
Given that spatially resolved epigenetic modifications of histones can
induce specific alterations in chromatin accessibility and
transcriptional phenotypes,^[ [193]^46 ^] we also investigated the
intranuclear distribution of H3K9me2/3 (Figure [194]5E–N).
Cross‐sectional analysis of 3D‐reconstructed confocal images of nuclei
costained for progerin and H3K9me2/3 revealed that increased H3K9me2/3
levels were associated with clustering of H3K9me2/3 within the nucleus
24 h after doxycycline treatment on stiff substrates, where H3K9me2/3
clustering was alternatively stained for progerin expression
(Figure [195]5E). Fluorescence intensity profiles obtained by line
scanning through the nuclear interior confirmed that progerin
expression and H3K9me2/3 clustering displayed distinct alternative
staining patterns at 24 h, with fluorescence profiles intensifying by
36 h (Figure [196]5G–J). However, on soft PAG substrates, intranuclear
H3K9me2/3 clustering predominated after 36 h of doxycycline treatment,
coinciding with enhanced progerin expression at the same time
(Figure [197]5F). Line scanning through the nuclear interior revealed
distinct alternative staining for H3K9me2/3 clusters and progerin at 36
h (Figure [198]5K–N).
These results indicate that substrate stiffness‐dependent differential
onset of progerin expression leads to a time‐dependent progression of
heterochromatin levels, accompanied by clustering of H3K9me2/3 within
the nucleus. This strongly suggests that the delayed gene expression in
cells placed on soft substrates, which transmit reduced forces to the
nucleus, can be attributed to incomplete heterochromatin structure.
2.6. Substrate Stiffness‐Dependent Differential Progerin Expression Is
Mediated by Distinct Transcription Factor Binding Motifs
Alterations in gene expression and epigenetic modifications are
regulated by substrate stiffness‐dependent differential progerin
expression (Figures [199]4 and [200]5). In line with previous reports
indicating that lamin A depletion enhances chromatin mobility,^[
[201]^47 ^] these results strongly imply that progerin‐induced
differential gene expression patterns in response to changes in
substrate stiffness can be attributed to variations in transcription
factor binding motifs.
To test this hypothesis, we first examined whether progerin expression
alters internuclear chromatin dynamics by comparing telomere motion in
dermal fibroblasts derived from a 3‐year‐old healthy individual and a
same‐aged patient with HGPS, denoted as 3YR (control) and 3YR (HGPS),
respectively (Figure [202]6A–C). Time‐lapse monitoring of GFP‐labeled
telomeres bound by the telomeric repeat‐binding factor 2 (TRF2)
revealed enhanced telomere movement in HGPS cells compared to control
cells (Figure [203]6A,B). Single‐particle tracking, followed by
calculation of the mean squared displacement (MSD) for the recorded
trajectories, confirmed more diffusible chromatin motion in HGPS cells
than in control cells, representing a confined motion (Figure [204]6C;
Movie [205]S4, Supporting Information).
Figure 6.
Figure 6
[206]Open in a new tab
Differential chromatin accessibility in response to mechanosensitive
progerin expression. A–C) Differential chromatin mobility in response
to progerin expression. Time‐lapse tracking of fluorescence‐tagged
chromatin was performed in TRF2 (telomeric repeat‐binding factor
2)‐transfected human dermal fibroblasts obtained from a three‐year‐old
healthy control (denoted as 3 YR (Control), A) and an HGPS patient
(denoted as 3 YR (HGPS), B), where nine randomly selected chromatin
trajectories are displayed. (C) Quantitative analysis of the mean
squared displacement (MSD) at each time lag indicates enhanced
chromatin mobility in HGPS patients compared to the healthy control.
D–R) ATAC sequencing‐based identification of differential key
transcription factor (TF) binding motifs in response to substrate
stiffness‐dependent progerin expression. Representative de novo TF
binding motifs in Tet‐On HeLa cells expressing progerin were identified
between cells on control stiff substrates (denoted as glass, D) and
cells on soft PAG substrates (denoted as 1.37 kPa, E) at 12 h intervals
after doxycycline treatment for 36 h using Homer software. The bar
graph indicates the percentage of target binding motifs for CTCF,
FOS::JUNB, TEAD family, KLF1, and ZNF331 (F), where each bar represents
fold enrichment, defined as the percentage of target sequences with the
motif divided by the percentage of background sequences with the motif.
Data are normalized to doxycycline‐untreated control groups (denoted as
glass 0 h, 1.37 kPa 0 h). Doxycycline treatment increased fold
enrichment for TEAD (G,H), CTCF (K,L), and NFkB‐p65‐Rel (O,P), but
decreased fold enrichment for JunB (I,J) and KLF1 (M,N) in control
stiff substrates, with these changes diminished in soft PAG substrates.
ATAC‐seq tracking of ZNF331 (Q), a known transcriptional repressor, and
BMP2 (R), a component of mechanosensory pathways, was visualized using
Integrative Genomics Viewer (IGV), where blue and red peaks indicate
stiff glass and soft PAG substrates, respectively.
Because these results strongly imply that nuclear wrinkling, typically
observed in patients with HGPS, is associated with enhanced chromatin
dynamics due to progerin‐induced loss of mechanical integrity along the
nuclear lamina,^[ [207]^47 , [208]^48 ^] we further investigated
whether progerin‐induced differential chromatin dynamics regulate
genome‐wide chromatin accessibility. Specifically, we performed a
high‐throughput assay for transposase‐accessible chromatin using
sequencing (ATAC‐seq) to assess whether progerin‐induced distinct,
time‐dependent gene expression patterns in response to changes in
substrate stiffness are attributable to variations in transcription
factor binding motifs. The calculation of transcription start site
(TSS) enrichment scores revealed that chromatin‐accessible regions are
enriched at TSSs, and the distribution of aligned fragment lengths
obtained from all tested samples confirmed the high quality of ATAC‐seq
data (Figure [209]S7A,B, Supporting Information). We observed that
chromatin‐accessible regions were similarly distributed across the
genome under different experimental conditions (Figure [210]S7C,D,
Supporting Information). Analysis of transcription factor (TF) binding
motifs within chromatin‐accessible regions revealed specific changes in
TF binding motifs in response to doxycycline‐induced progerin
expression on stiff glass substrates (denoted as glass, Figure [211]6D)
and soft PAG substrates (denoted as 1.37 kPa, Figure [212]6E), selected
from the top ten TFs predicted to have binding motifs in
chromatin‐accessible regions for each condition (Figure [213]S7E,F,
Supporting Information).
In particular, as represented by the percentage of targets, we noted
that the expression of differential TF‐binding motifs for CCCTC‐binding
factor (CTCF), FOS::JUNB, TEA domain (TEAD) family, Kruppel‐like factor
1 (KLF1), and zinc finger protein 331 (ZNF331) was delayed on soft
substrates compared to stiff substrates (Figure [214]6F). These results
suggest that progerin‐induced differential gene expression
(Figure [215]4), following enhanced H3K9me2/3 clustering
(Figure [216]5), could be attributed to differences in TF‐binding
motifs. Progerin‐induced alterations in heterochromatin structure could
modulate the configuration of TF binding, ultimately regulating
substrate stiffness‐dependent gene expression.
A systematic comparison of fold enrichment values normalized to the
untreated doxycycline group (0 h) identified differentially regulated
TFs that promote multiple aging‐associated cellular mechanisms, while
all the detected TFs were expressed at lower levels on soft substrates
than on stiff substrates (glass vs 1.37 kPa, Figure [217]6G–P). For
instance, the expression of TEAD (Figure [218]6G,H), CTCF
(Figure [219]6K,L), and NFkB‐p65‐Rel (Figure [220]6O,P), regulating
pathological processes in HGPS, including cell proliferation, chromatin
structure, and inflammatory responses,^[ [221]^49 ^] progressively
increased with doxycycline treatment. In contrast, the expression of
JunB (Figure [222]6I,J) and KLF1 (Figure [223]6M,N), regulating
cellular stress responses (e.g., oxidative stress and inflammation)^[
[224]^50 ^] and gene transcription in erythroid differentiation,^[
[225]^51 ^] respectively, gradually decreased with doxycycline
treatment. Meanwhile, mechanosensitive signaling‐mediated ZNF331, a
transcriptional repressor containing the Kruppel‐associated box (KRAB)
domain,^[ [226]^52 ^] exhibited an increase in ATAC‐seq peak height
following doxycycline treatment, indicating enhanced accessibility of
the ZNF331 gene (Figure [227]6Q). In contrast, the gene accessibility
of BMP2, involved in the BMP signaling pathway, one of the key pathways
identified in mechanosensation (Figure [228]4F), increased
progressively by doxycycline‐induced progerin expression
(Figure [229]6R).
These findings suggest that nuclear wrinkling induced by progerin
expression elevates chromatin dynamics, enhances heterochromatin
clustering, and ultimately regulates chromatin accessibility.
2.7. Progerin‐Induced Nuclear Deformation Is Mediated by the Remodeling of
LINC Complex‐Dependent Molecular Connections with LMNA
Lamin A interacts with the cytoskeleton via the LINC complex, which is
composed of SUN proteins in the inner nuclear membrane and nesprin
isoforms, the cytoplasmic domains of KASH proteins in the outer nuclear
membrane, enabling the transmission of cytoskeletal forces to the
nuclear membrane.^[ [230]^53 ^] We previously demonstrated that
progerin expression reduced nuclear tension by incorporating the
nesprin tension sensor, mimicking nesprin 2 bridging between the actin
cytoskeleton and SUN proteins (Figure [231]2). Based on this, we
hypothesized that progerin‐induced nuclear deformation could be
mediated by the differential interactions between the LINC complex and
LMNA.
To determine whether the evolution of NE wrinkling in response to
progerin expression was associated with chromosomal interactions
between DNA and LINC proteins, we first analyzed chromatin
accessibility of LMNA, SUN1, and SYNE2 using ATAC‐seq (Figure
[232]S8A–F, Supporting Information). After doxycycline treatment, these
genes progressively enhanced chromatin accessibility on both stiff and
soft substrates (Figure [233]7A–F). While LMNA gene accessibility
increased on both substrate types 12 h post‐doxycycline treatment,
however, the accessibility of SUN1 and SYNE2 genes was delayed on soft
substrates, with a similar increase observed on both stiff and soft
substrates at 36 h (Figure [234]7A–F; Figure [235]S8A–F, Supporting
Information). Consistent with hierarchical clustering analysis of
RNA‐seq indicating that samples were clustered by substrate stiffness
at 24 and 30 h, but clustered by doxycycline treatment time at 36 h
(Figure [236]4A), these results suggest that substrate
stiffness‐dependent differential chromatin accessibility was dominant
at 24 h of progerin expression, but diminished at 36 h of doxycycline
exposure. Together with the previous result exhibiting a delayed
reduction of NE tension on soft substrates (Figure [237]2), these
findings suggest that Tet‐On‐inducible progerin expression modulates
LMNA gene expression, simultaneously upregulating SUN1 and SYNE2, while
reduced NE tension induced by progerin expression facilitates LINC
complex‐mediated molecular binding.
Figure 7.
Figure 7
[238]Open in a new tab
LINC complex‐mediated remodeling of LMNA‐associated nuclear tethering
in progerin‐expressing cells. A–F) ATAC‐seq‐based identification of
chromatin accessibility for LMNA‐associated LINC complex components in
progerin‐expressing cells. ATAC‐seq peaks for LMNA (A,B), SUN1 (C,D),
and SYNE2 (E,F) in progerin expression‐induced cells treated with
doxycycline at 12 h intervals for up to 36 h on control glass (A,C,E)
and soft PAG substrates (B,D,F) were visualized by Integrative Genomics
Viewer (IGV). Note that as doxycycline treatment progresses, peaks for
LMNA, SUN1, and SYNE2 in both control glass and soft PAG increase,
indicating enhanced chromatin accessibility, but the peak height in
soft substrates remains lower than that in control glass, approaching a
similar level at 36 h. G–N) Differential expression of NE‐associated
proteins and nuclear wrinkling in response to progerin expression.
Tet‐On HeLa cells expressing mCherry‐tagged LMNA or Δ50 LMNA (progerin)
were immunostained for SUN1 (green) and nucleus (DAPI, blue) (G,H) or
nesprin 2 (green) and nucleus (DAPI, blue) (I,J) before (−Dox) and
after (+Dox) doxycycline treatment. While doxycycline‐induced
expression of mCherry‐tagged LMNA did not alter the SUN1 (G,K) and
nesprin 2 contents (I,L), doxycycline‐induced expression of
mCherry‐tagged Δ50 LMNA significantly increased SUN1 (H,K) and nesprin
2 (J,L). Compared to doxycycline‐induced LMNA expression, which did not
induce changes in nuclear shape (G,I,M), Δ50 LMNA expression
significantly increased NE wrinkling (H,J,M). SUN1 and nesprin 2
expression was more sensitive to Δ50 LMNA expression than to LMNA
expression (N). In panels K–N, >50 nuclei were analyzed for each
condition; error bars indicate the standard error of the mean (S.E.M.);
and Student's t‐test was applied for comparison between two groups
(****: p < 0.0001, ***: p < 0.001, NS: not significant). O–T)
Progerin‐induced differential interaction in LMNA‐associated LINC
proteins. The strength of molecular interaction between SUN1 and
nesprin 2 (O,Q,R) or between SUN1 and LMNA (P,S,T) was estimated by
quantifying the number and total intensity of proximity ligation assay
(PLA) signals (red dots) in DAPI‐stained nuclei (blue). Hemispherical
and cross‐sectional views of 3D‐rendered nuclei showed that punctate
PLA signals were preferentially localized along the nuclear periphery.
Note that doxycycline‐induced progerin expression significantly
increased the PLA signals of SUN1 associated with LMNA (S,T), while PLA
signals of SUN1 associated with nesprin 2 remained unchanged (Q,R). In
panels Q, R, S, and T, >50 nuclei were analyzed for each condition;
error bars indicate the S.E.M.; and an unpaired t‐test was applied to
assess statistical significance (***: p < 0.001, NS: not significant).
To test this notion, we examined the expression levels of SUN1 and its
binding partner, nesprin 2, in response to progerin expression by
comparing mCherry‐tagged LMNA and progerin (denoted as Δ50
LMNA)‐expressing Tet‐On HeLa cells (Figure [239]7G–J). Quantitative
immunofluorescence microscopy revealed that both SUN1 and nesprin 2
were upregulated in progerin‐expressing cells but not in
LMNA‐overexpressing cells (Figure [240]7K,L), which was confirmed by
immunoblotting analysis of total protein levels (Figure [241]S8G–J,
Supporting Information). These results demonstrate that the
upregulation of LINC proteins is specifically induced by progerin
expression and not by the accumulation of intact LMNA. To assess
whether LINC complex proteins regulate progerin‐induced nuclear
deformation, we compared changes in NE wrinkling before and after
doxycycline‐induced expression of LMNA or progerin. As expected, NE
wrinkling remained unchanged before doxycycline treatment and was not
altered by LMNA expression (left two bars, Figure [242]7M). However,
progerin expression resulted in a sixfold increase in NE wrinkling
(right two bars, Figure [243]7M). Moreover, we confirmed that LINC
complex proteins were more tightly regulated by progerin expression
than by intact LMNA expression, i.e., the expression levels of SUN1 and
nesprin 2 increased proportionally with progerin expression but not
with LMNA expression (Figure [244]7N). These results support the
association between NE wrinkling and the upregulation of LINC complex
proteins in response to progerin expression.
Building on previous reports that SUN1 and SUN2 are not functionally
equivalent for nuclear connection to the actin cytoskeleton,^[ [245]^54
^] and that SUN1, but not SUN2, is upregulated in HGPS cells,^[
[246]^28 ^] we investigated whether doxycycline‐activated progerin
expression could remodel the molecular connectivity of LINC‐associated
proteins to the nucleus via upregulation of SUN1. We assessed whether
SUN1 overexpression altered LMNA‐associated molecular interactions in
the nuclear lamina, leading to progerin‐mediated nuclear deformation.
To quantitatively analyze the interaction between SUN1 and LMNA or
nesprin 2 in response to progerin expression, a proximity ligation
assay (PLA) was performed (Figure [247]7O–T). As expected, a majority
of punctate PLA signals were preferentially localized along the nuclear
periphery, where LINC components form a molecular assembly with lamin
proteins (Figure [248]7O,P). By counting the number and total intensity
of fluorescent dots representing molecular interactions, we observed
that doxycycline‐induced progerin expression significantly increased
the association between SUN1 and LMNA (Figure [249]7P,S,T), whereas the
interaction between SUN1 and nesprin 2 remained unchanged
(Figure [250]7O,Q,R).
Together with the upregulation of SUN1 in progerin‐expressing cells,
the preservation of identical molecular interactions between SUN1 and
nesprin 2 further implies that progerin accumulation in intact LMNA
enhances the spatial proximity of SUN1 to LMNA, strongly suggesting
that progerin expression induces nuclear deformation through the
remodeling of SUN1‐dependent molecular connections with LMNA.
2.8. Actomyosin Contractility Regulates Nuclear Deformation by Altering
Nuclear Tension in Progerin‐Expressing Cells
We previously demonstrated that progerin‐induced NE wrinkling was
highly mechanosensitive (Figure [251]1) and that substrate
stiffness‐dependent cytoskeletal tension mediated nuclear force at the
nucleus–cytoskeletal interface (Figure [252]2). In conjunction with
recent studies showing increased F‐actin polymerization in
LMNA‐depleted human retinal pigment epithelial cells^[ [253]^29 ^] and
elevated RhoA activation in Z24^−/− MSCs,^[ [254]^30 ^] we hypothesized
that progerin‐induced nuclear deformation could be regulated by
alterations in nuclear tension driven by actomyosin contractility.
To investigate whether doxycycline‐controlled progerin expression could
alter cytoskeletal tension, we examined F‐actin content and myosin
activity using quantitative immunofluorescence microscopy (Figure
[255]8A–C). As predicted, progerin‐expressing cells exhibited increased
F‐actin organization (Figure [256]8A,B). Furthermore, we observed a
significant increase in myosin II content relative to F‐actin, assessed
by measuring the intensity of phospho‐myosin light chain 2 (pMLC2)
normalized to individual F‐actin fibers (Figure [257]8C). These
findings confirm that myosin‐dependent cytoskeletal tension is elevated
in progerin‐expressing cells.
Figure 8.
Figure 8
[258]Open in a new tab
Actomyosin contractility‐dependent nuclear deformation in
progerin‐expressing cells. A–C) Differential actomyosin contractility
in response to doxycycline‐induced progerin expression.
Doxycycline‐inducible progerin‐expressing HeLa cells were immunostained
for F‐actin (green), phospho‐myosin light chain 2 (pMLC2, red), and
nucleus (DAPI, blue) before (−Dox) and after (+Dox) doxycycline
treatment (A). (Insets) The details of pMLC2 staining along the actin
stress fibers. Doxycycline‐induced progerin expression significantly
increased the F‐actin (B) and pMLC2 (C) contents, which were normalized
to cell area and actin stress fibers, respectively. In panels B and C,
>60 cells were analyzed for each condition; error bars indicate the
S.E.M.; unpaired t‐test was applied (***: p < 0.001). D–F) Differential
formation of F‐actin and pMLC2 in doxycycline‐controlled
progerin‐expressing Tet‐On HeLa cells in response to pharmaceutical
inhibition of myosin‐dependent cytoskeletal tension. Cells were
immunostained for nuclei (DAPI, blue), F‐actin (green), and pMLC2 (red)
before (−Dox) and after (+Dox) doxycycline treatment, where
differential concentrations of myosin‐II inhibiting blebbistatin were
added (D). Compared to −Dox control cells, doxycycline‐induced
progerin‐expressing cells showed significantly enhanced F‐actin, which
remained unchanged in response to specific disruption of myosin
activity (E). Significantly increased pMLC2 content due to
doxycycline‐induced progerin expression was reversed by increasing the
concentration of blebbistatin (F). Treating doxycycline‐induced
progerin‐expressing cells with 15 µm blebbistatin fully restored their
pMLC2 content to the level of doxycycline‐untreated progerin
nonexpressing cells (F). In panels E and F, >150 cells were analyzed
per condition; error bars indicate the S.E.M.; one‐way ANOVA using
Tukey's test was applied (***: p < 0.001, **: p < 0.05, NS: not
significant). G,H) Actomyosin contractility‐dependent differential
changes of nuclear tension. Representative nesprin tension sensor‐based
FRET signals along the nuclear membrane of doxycycline‐inducible
progerin‐expressing cells were captured before (−Dox) and after (+Dox)
doxycycline treatment in the presence of DMSO and 10 or 15 µm
blebbistatin (G). Compared to doxycycline‐untreated control,
doxycycline‐induced progerin expression significantly enhanced the FRET
ratio, which was gradually diminished by increasing the concentration
of blebbistatin and fully restored to the level of
doxycycline‐untreated control condition by 15 µm blebbistatin treatment
(H). I–K) Tight regulation of SUN1 expression and NE wrinkling in
response to changes in actomyosin contractility. Tet‐On HeLa cells
expressing mCherry‐tagged Δ50 LMNA (progerin) were treated with DMSO
and 10 or 15 µm blebbistatin in the absence (−Dox) and presence (+Dox)
of doxycycline before immunostaining for nucleus (DAPI, blue) and SUN1
(green) (I). Compared to the doxycycline‐untreated control, SUN1
expression and NE wrinkling were significantly increased in
doxycycline‐treated progerin‐expressing cells, which was gradually
diminished by increasing the concentration of blebbistatin and fully
restored to the level of the control condition by 15 µm blebbistatin
treatment (J,K). In panels H, J, and K, >20 cells were analyzed per
condition; error bars indicate the S.E.M.; one‐way ANOVA using Tukey's
test was applied for comparison between groups (***: p < 0.001, **: p <
0.05, NS: not significant).
Next, to determine whether actomyosin contractility regulates nuclear
tension‐mediated NE wrinkling, we treated progerin‐expressing cells
with the myosin II inhibitor blebbistatin at varying concentrations
(Figure [259]8D). Consistent with previous results (Figure [260]8A–C),
doxycycline‐induced progerin expression significantly increased F‐actin
and pMLC2 intensities in dimethyl sulfoxide (DMSO)‐treated control
cells (Figure [261]8E,F). However, pMLC2 content gradually decreased
with increasing drug concentration, and 15 µm blebbistatin treatment
fully restored pMLC2 content to the level observed in
doxycycline‐untreated progerin nonexpressing cells (Figure [262]8F),
while F‐actin content remained unchanged (Figure [263]8E).
We assessed actomyosin contractility‐dependent nuclear tension using
the nesprin tension sensor in doxycycline‐induced progerin‐expressing
cells treated with varying concentrations of blebbistatin
(Figure [264]8G). Surprisingly, we found that a gradual reduction in
myosin activity enhanced nuclear tension, as indicated by the FRET
ratio (Figure [265]8H). Treatment of doxycycline‐induced
progerin‐expressing cells with 15 µm blebbistatin reduced pMLC2
expression to levels observed in doxycycline‐untreated progerin
nonexpressing cells (Figure [266]8F) without disrupting the F‐actin
content (Figure [267]8E). Under these conditions, the FRET ratio was
fully restored (Figure [268]8H).
We further confirmed that enhanced nuclear tension (i.e., decreased
FRET ratio) in response to pharmaceutical inhibition of pMLC2
expression suppressed SUN1 expression in progerin‐expressing cells
(Figure [269]8I,J). Consequently, the level of NE wrinkling in 15 µm
blebbistatin‐treated progerin‐expressing cells was fully restored to
that observed in doxycycline‐untreated cells (Figure [270]8K). These
data suggest that inhibition of myosin‐dependent cytoskeletal tension
reverses nuclear deformation by downregulating SUN1 expression and
restoring NE tension.
Taken together, our results demonstrate that actomyosin contractility
regulates nuclear deformation through alterations in nuclear tension in
progerin‐expressing cells.
2.9. Inhibition of SUN1 Recovers Progerin‐Induced Nuclear Wrinkling in HGPS
Cells
Upregulated SUN1 expression in progerin‐expressing cells resulted in
the remodeling of molecular connections with LMNA (Figure [271]7), and
progerin‐induced nuclear deformation was regulated by pMLC2‐dependent
nuclear tension (Figure [272]8). Together with a previous study showing
that SUN1 depletion reduces actomyosin activity without disrupting the
expression of nesprin 2 in vascular smooth muscle cells,^[ [273]^17a ^]
we hypothesized that the level of SUN1 expression in
progerin‐expressing cells could determine myosin‐dependent cytoskeletal
tension, ultimately regulating nuclear tension‐dependent nuclear
deformation.
To directly assess the causality between SUN1 expression and myosin
activity, we inhibited SUN1 expression by transfecting
progerin‐expressing Tet‐On HeLa cells with small interfering RNA
(siRNA) and compared them to control siRNA‐transfected cells (Figure
[274]9A). Consistent with previous data (Figures [275]7K and [276]8C),
doxycycline‐induced progerin‐expressing cells transfected with control
siRNA showed significantly increased levels of SUN1 and pMLC2, whereas
siRNA‐mediated SUN1 depletion reduced these levels comparable to those
observed in doxycycline‐untreated, progerin‐negative cells
(Figure [277]9B,C). Confirming that the changes in pMLC2 expression in
Tet‐On HeLa cells in response to doxycycline‐induced progerin
expression and/or siRNA transfection were largely proportional to
changes in SUN1 expression (Figure [278]9A–C), a strong correlation
between SUN1 and pMLC2 expression was detected, regardless of specific
conditions (Figure [279]S9, Supporting Information). These results
indicate that increased pMLC2 expression in progerin‐expressing cells
is induced by SUN1 upregulation.
Figure 9.
Figure 9
[280]Open in a new tab
SUN1‐mediated nuclear tension regulates progerin‐induced nuclear
deformation. A–C) SUN1‐mediated modulation of actomyosin activity in
progerin‐expressing cells. Tet‐On HeLa cells expressing progerin were
immunostained for F‐actin (green), pMLC2 (red), SUN1 (orange), and
nucleus (DAPI, blue) in the doxycycline‐untreated control condition
(−Dox) and doxycycline‐treated conditions (+Dox) with siControl
(+Dox/+siCon) or siSUN1 (+Dox/+siSUN1)‐mediated knockdown (A).
Doxycycline‐induced progerin expression significantly increased the
expression levels of SUN1 and pMLC2, which were maintained in
siControl‐transfected cells but restored to levels similar to those
observed in doxycycline‐untreated progerin nonexpressing cells after
transfection with siSUN1 (B,C). In panels B and C, >150 cells were
analyzed per condition; error bars indicate the S.E.M.; one‐way ANOVA
using Tukey's test was applied (***: p < 0.001, NS: not significant).
D–E) SUN1 expression‐dependent NE tension. Nesprin tension sensor‐based
FRET signals along the nuclear membrane of doxycycline‐inducible
progerin‐expressing cells transfected with siControl (+siCon) or siRNA
targeting SUN1 (+siSUN1) were captured before (−Dox) and after (+Dox)
doxycycline treatment (D). Doxycycline‐induced enhanced NE tension was
maintained in siControl‐transfected cells but restored to the level of
the doxycycline‐untreated control condition in siSUN1‐transfected cells
(E). In panel E, >20 cells were analyzed per condition; error bars
indicate the S.E.M.; one‐way ANOVA using Tukey's test was applied (***:
p < 0.001, NS: not significant). F–H) Quantification of SUN1‐mediated
NE wrinkling. mCherry‐tagged progerin‐expressing Tet‐On HeLa cells
transfected with siControl (+siCon) or siRNA targeting SUN1 (+siSUN1)
were immunostained for lamin B1 (green), SUN1 (yellow), and nuclear DNA
(DAPI, blue) in the absence (−Dox) or presence (+Dox) of doxycycline
(F). Doxycycline‐induced progerin expression significantly increased
the NE wrinkling, which was maintained in siControl‐transfected cells
but reduced to the level of the doxycycline‐untreated control condition
in siSUN1‐transfected cells (G). The Pearson product‐moment correlation
assessment applied to the merged dataset, including all conditions,
showed a highly correlative relationship between SUN1 expression and NE
wrinkling (r = 0.83) (H). In panels G and H, >50 cells were analyzed
per condition; error bars indicate the S.E.M.; and one‐way ANOVA using
Tukey's test was applied for comparison between groups (***: p < 0.001,
NS: not significant). I–M) SUN1‐mediated restoration of the nuclear
morphology of HGPS fibroblasts. Human dermal fibroblasts obtained from
a three‐year‐old healthy control (denoted by 3 YR) and an HGPS patient
were immunostained for F‐actin (green), pMLC2 (red), SUN1 (orange), and
nuclei (DAPI, blue), where HGPS fibroblasts were transfected with
siControl (HGPS/+siCon) or siSUN1 (HGPS/+siSUN1) (I). Full and empty
arrowheads indicate the smooth and wrinkled nuclear surface,
respectively. Compared to control fibroblasts, HGPS cells displayed a
significantly enhanced expression of SUN1 and pMLC2, which was
maintained in siControl‐transfected cells, but transfection with siSUN1
restored SUN1 and pMLC2 expression to levels similar to those observed
in control cells (J,K). Nuclear wrinkles specifically featured in HGPS
fibroblasts and siControl‐transfected HGPS fibroblasts were recovered
in siSUN1‐transfected cells to levels comparable to those in healthy
controls (L). Pearson correlation analysis applied to the merged
dataset incorporating all experimental conditions showed a strong
correlation between SUN1 expression and pMLC2 expression (red, r =
0.98), SUN1 expression, and NE wrinkling (blue, r = 0.99) (M). In
panels J and K, >50 cells were analyzed per condition; in panel L, >20
cells were analyzed per condition; error bars indicate the S.E.M.;
one‐way ANOVA using Tukey's test was applied for comparison between
groups (***: p < 0.001, **: p < 0.05, NS: not significant).
As reduced nuclear tension (i.e., increased FRET ratio) was restored by
inhibiting pMLC2 expression (Figure [281]8H), we assessed whether
transfection with SUN1‐siRNA could also restore this reduced nuclear
tension (Figure [282]9D). In progerin‐expressing cells transfected with
SUN1‐siRNA (+Dox/+siSUN1), the increased FRET ratio induced by
doxycycline‐mediated progerin expression (+Dox or +Dox/+siCon) was
reduced to the level observed in doxycycline‐untreated cells (−Dox)
(Figure [283]9E). This confirmed that the reduced nuclear tension in
progerin‐expressing cells was mediated by SUN1 upregulation.
Measurement of NE wrinkling indicated that nuclear deformation caused
by progerin‐induced reduction in nuclear tension was reversed by
transfection with SUN1‐siRNA (Figure [284]9F,G), indicating a strong
correlation between SUN1 content and the level of NE wrinkling under
all conditions (Figure [285]9H). Combined with previous data showing
that progerin‐induced NE wrinkling is mediated by a SUN1‐dependent
reduction in nuclear tension, accompanied by upregulated
myosin‐associated cytoskeletal tension (Figures [286]7 and [287]8), and
that gradual inhibition of pMLC2 expression reverses the upregulation
of SUN1 and NE wrinkling (Figure [288]8J,K), these findings suggest
that SUN1 upregulation is responsible for nuclear deformation in
progerin‐expressing cells.
To extend our findings to human patients suffering from an
accelerated/premature aging disorder, we utilized dermal fibroblasts
derived from a 3‐year‐old patient with HGPS and compared them to
control cells from a healthy individual of the same age. Transfection
of HGPS fibroblasts with SUN1–siRNA significantly reduced the
expression of both SUN1 and pMLC2, without disrupting F‐actin
organization, which remained unchanged over 5 days (Figure [289]S10,
Supporting Information). By systematically comparing fibroblasts from
healthy individuals (denoted as 3 YR), we found that both HGPS
fibroblasts and siControl‐transfected HGPS fibroblasts exhibited
significantly increased levels of SUN1 and pMLC2, which was reversed by
transfection with siSUN1 (Figure [290]9I–K). Accordingly, the level of
NE wrinkling, which is specifically featured in HGPS fibroblasts and
siControl‐transfected HGPS fibroblasts, was restored to healthy control
levels by siRNA‐induced depletion of SUN1 (Figure [291]9I,L). In
patients with HGPS, a stronger correlation was observed between SUN1
and pMLC2 intensities, as well as nuclear envelope wrinkling. Notably,
SUN1 knockdown using siRNA reduced these levels to those observed in
healthy controls (Figure [292]9M).
Together with previous results showing that doxycycline‐induced
progerin expression results in SUN1 accumulation in the nuclear lamina,
where nuclear tension along the SUN1–nesprin 2–F‐actin connections is
diminished by increased pMLC2 in response to progerin expression, these
results reconfirm that defective nuclear morphology in HGPS is induced
by reduced nuclear tension, accompanied by SUN1 upregulation‐mediated
pMLC2 expression rather than altered F‐actin connectivity, which
coincides with chromosomal remodeling via modification of
heterochromatin accessibility (Figure [293]10 ).
Figure 10.
Figure 10
[294]Open in a new tab
Schematic summary depicting the functional relationship between
SUN1‐mediated nuclear tension and NE wrinkling in response to progerin
expression. Progerin expression accumulates LINC complex proteins SUN1
and Nesprin 2, reorganizing the actin‐binding Nesprin‐associated LINC
complex at the nuclear envelope, and determining the biophysical
interactions of the nuclear–cytoskeletal connection. Although the
molecular linkages connecting SUN1, Nesprin 2, and F‐actin remain
unchanged in response to progerin expression, nuclear tension along the
SUN1–Nesprin 2–F‐actin connection is reduced by increased pMLC2. In
summary, progerin‐induced morphological defects forming the surface
wrinkling along the nuclear lamina are determined by the accumulation
of LINC complexes proteins at the nuclear envelope and reduced nuclear
tension accompanied by pMLC2 via the SUN1–Nesprin 2 bridge, regulating
the expression of various genes within the nucleus. Ultimately,
progerin‐induced nuclear wrinkling features increased chromatin
dynamics in the heterochromatin‐rich nuclear periphery, resulting in
the misregulation of mechanotransduction signal pathways in the HGPS
model. Doxycycline‐induced progerin expression exhibits mechanical
sensitivity to variations in substrate stiffness. Approximately 10%,
25%, and 28% of delays in onsets of progerin expression, reduction of
nuclear tension, and nuclear wrinkling, respectively, on the soft
substrate identifies the intracellular cytoskeletal force exerted on
the nucleus as the origin of progerin‐induced nuclear wrinkling.
3. Discussion
Increased cytoskeletal tension, determined by its connection to the
nuclear envelope, is crucial in regulating the mechanical forces
transmitted to the nucleus. However, the causal relationship between
progerin‐induced changes in cytoskeletal dynamics and nuclear
deformation in progerin‐expressing cells remains unclear. In this
study, we aimed to address this gap by investigating the molecular
mechanisms underlying nuclear morphological changes in response to
progerin expression and their contribution to chromatin reorganization
and aberrant gene expression patterns observed in HGPS. In this study,
we employed a doxycycline‐inducible Tet‐On system to precisely control
progerin expression in a human HGPS cell model, enabling real‐time
monitoring through mCherry‐tagged proteins. This system facilitates the
spatiotemporal monitoring of nuclear deformation in response to
mechanosensitive progerin expression while overcoming the limitation of
primary cell availability. This approach enabled us to investigate the
molecular mechanisms underlying progerin‐induced nuclear deformation
and its impact on temporal chromatin remodeling and gene expression
profiles. Moreover, this system served as an experimental platform to
explore the relationship between progerin‐induced nuclear deformation
and mechanosensing of substrate stiffness.
In combination with previous reports showing that adhesion‐dependent
cells typically display enhanced cytoskeletal tension on a rigid matrix
compared to that on a compliant matrix to maintain mechanical
integrity,^[ [295]^55 ^] our findings of enhanced myosin activity in
progerin‐expressing cells indicate that the signaling pathways
determining the progerin‐induced nuclear deformation are highly
mechanosensitive (Figure [296]1). Previous results showed that dermal
fibroblasts from a patient with HGPS exhibited abnormal nuclear
morphology and even formed nuclear rupture on stiff substrates ranging
from 10 to 20 kPa and 80 kPa, whereas such abnormalities were less
frequently observed in soft substrates ≈3 kPa.^[ [297]^56 ^] Because
mechanosensitive nuclear deformation is also time‐dependent
(Figure [298]3), These results suggests that the differential onset of
nuclear abnormalities correlates with an increase in cytoskeletal
tension, featuring enhanced cell spreading and elongation that
typically amplify cytoskeletal forces transmitted to the nucleus. Since
progerin expression is regulated by cytoskeletal forces, nuclear
responses to differential expression of progerin vary depending on the
mechanical properties of the extracellular environment. Thus, exposure
to substrates of varying substrate stiffness could modulate
cytoskeletal tension and mechanotransduction, ultimately shaping the
cellular adaptations associated with progerin‐induced phenotypes.^[
[299]^57 ^] As varying matrix rigidity induces distinct differentiation
of MSCs, e.g., neurogenesis on soft substrates ranging from 0.1 to 1
kPa, and osteogenesis on stiff substrates ranging from 25 to 40 kPa,
respectively,^[ [300]^58 ^] 1.37 kPa condition applied in our study
closely mimics physical settings of soft tissues, whereas 34 kPa
condition represents a microenvironment of stiffer tissues
(Figure [301]1). This enables our system to reliably predict the
phenotypic development associated with nuclear deformation observed in
HGPS patients.
We observed that progerin expression induced a critical reduction in
nuclear tension after 24 h, followed by NE wrinkling within 1 h on
stiff substrates, whereas these time intervals doubled on soft
substrates (Figures [302]1 and [303]2). These results indicate that the
synergistic increase in cytoskeletal tension can be attributed to the
combination of progerin expression and plating of cells on the rigid
matrix, which accelerates the individual steps of progerin‐induced
reduction in nuclear tension and nuclear tension‐dependent NE wrinkling
(Figures [304]1 and [305]2). The computational model system further
demonstrated that as the mechanical resistance of the nucleus to
external pressure increased (i.e., on a stiff substrate), the nuclear
volume and NE tension decreased, leading to accelerated nuclear
wrinkling compared to the nucleus on the soft substrate
(Figure [306]3).
Meanwhile, RNA sequencing comparing the number of DEGs following
progerin expression on substrates of varying elastic moduli showed that
the largest shift in gene expression levels (both upregulation and
downregulation) occurred at 24 h after doxycycline treatment on stiff
substrates and at 30 h on soft substrates (Figure [307]4A,B), following
a significant reduction in NE tension (Figures [308]1 and [309]2).
These results indicate that progerin‐induced nuclear wrinkling is not
only mechanosensitive to changes in extracellular mechanical settings,
mainly mediated by actomyosin contractility and nucleus–cytoskeletal
connections,^[ [310]^16 ^] but also induces distinct pathogenetic gene
regulation. The upregulation of LMNA, SYNE2, and SYNE3 supports this
notion that mechanosensitive progerin expression mediates essential
signaling pathways and functional gene regulation (Figure [311]4).
Moreover, consistent with enhanced expression of H3K9me2/3 on rigid
substrates,^[ [312]^46 ^] progerin‐induced epigenetic modification is
highly mechanosensitive (Figure [313]5) and leads to enhanced chromatin
dynamics in fibroblasts from patients with HGPS (Figure [314]6A–C).
A comparative analysis with publicly available RNA sequencing datasets
from primary fibroblasts of healthy individuals and patients with HGPS
([315]GSE141950 and [316]GSE118633) revealed distinct gene expression
profiles in HGPS fibroblasts (Figure [317]S6, Supporting Information).
Pathway enrichment analyses further demonstrated consistent alterations
in cellular components, biological processes, and molecular functions
across models. Notably, the [318]GSE118633 dataset corroborated
pathway‐level changes, reinforcing the validity of our findings. These
results not only support the robustness of our Tet‐On‐inducible
progerin‐expressing HeLa cell model in recapitulating key molecular
features of HGPS but also suggest that progerin expression
differentially regulates gene expression in a substrate
stiffness‐dependent manner by modulating multiple signaling pathways
(Figure [319]4D–H; Figure [320]S6, Supporting Information). In
addition, ATAC‐seq revealed that the progerin‐induced alteration of
transcription factors is also responsive to substrate stiffness
(Figure [321]6D–R), further suggesting that progerin‐induced nuclear
wrinkling amplifies chromatin dynamics by promoting heterochromatin
clustering to regulate chromatin accessibility.
Intact LMNA‐producing normal cells typically transmit cytoskeletal
tension to the nuclear membrane via the stable expression of LINC
complex‐associated molecular components.^[ [322]^16 ^] Therefore, a
strong correlation between cytoskeletal tension and the force applied
to the nucleus has been detected.^[ [323]^59 ^] However, we noted a
mismatch between the reduced nuclear tension and enhanced cytoskeletal
tension in doxycycline‐induced progerin‐expressing cells (Figure [324]2
vs Figure [325]8). This could be due to the formation of the LMNA
mutant‐progerin, which could disrupt the LMNA‐mediated molecular
connection. Consequently, the cytoskeletal forces may not be properly
transmitted into the nucleus, and vice versa. The mechanical imbalance
between outer nuclear cytoskeletal tension and LINC‐mediated nuclear
tension in progerin‐expressing cells further highlights the critical
role of LMNA‐dependent molecular connections between the NE and
cytoskeleton. Notably, we showed that doxycycline‐induced progerin
expression upregulated LINC complex‐associated genes (Figure [326]7).
Previous studies have shown that progerin predominantly associates with
SUN1, but not with SUN2, in LMNA^−/− mouse embryonic fibroblasts,^[
[327]^28 ^] and that the stable expression of progerin proportionally
increases the level of SUN1, but not SUN2, in progerin‐expressing
NIH3T3 fibroblasts.^[ [328]^26b ^] In an in situ PLA assay, we observed
that individual interactions between LMNA and SUN1 were increased in
response to progerin expression, while the molecular interaction
between SUN1 and its binding partner nesprin 2 remained unchanged
(Figure [329]7O–T). Combining these results, we propose a new model
that fills the missing link between myosin activity‐dependent
cytoskeletal tension and SUN1‐mediated nuclear tension (Figure [330]8).
Our findings suggest that progerin‐induced nuclear deformation is
mediated by reduced nuclear tension, accompanied by SUN1
upregulation‐dependent myosin tension, rather than by F‐actin
connectivity through SUN1–nesprin 2 bridging. Validation in both
SUN1‐depleted Tet‐On HeLa cells and HGPS patient‐derived fibroblasts
established that our HeLa cell models are consistent with those from
primary HGPS fibroblasts. This concordance reinforces that
SUN1‐dependent nuclear tension plays a critical role in regulating
progerin‐induced nuclear deformation, highlighting the mechanistic
relevance of our cell model to recapitulate key aspects of pathological
features of HGPS (Figure [331]9).
Taken together, our results suggest that i) progerin expression,
reduced NE tension, and nuclear wrinkling are mechanosensitive to
changes in substrate stiffness, ii) progerin expression disrupts the
LMNA‐mediated force balance between cytoskeletal force and nuclear
tension, iii) the SUN1–LMNA interaction mediates force transmission
from the cytoskeleton to the nuclear interior, and iv) reduced nuclear
tension enhances chromatin dynamics, thereby regulating
progerin‐induced mechanosensitive signaling pathways. The key findings
regarding the SUN1‐dependent nuclear tension and chromatin remodeling
have important translational implications, as they suggest that
manipulation of nuclear tension could provide a novel therapeutic
strategy to mitigate the effects of progerin‐induced nuclear
deformation. These results could pave the way for developing targeted
interventions that modulate nuclear mechanics to treat progeria as well
as diseases associated with defects in nuclear architecture.
4. Experimental Section
Cell Culture and Drug Treatment
HeLa cells (purchased from Korean Cell Line Bank, Seoul, Republic of
Korea) were cultured in T25 rectangular canted neck cell culture flasks
(Falcon, 353108) containing 2 mL of Dulbecco's Modified Eagle's Medium
(DMEM, Corning, 10‐013‐CV) supplemented with 10% fetal bovine serum
(FBS, Merck, TMS‐031‐BKR) and 1% penicillin–streptomycin (Thermo,
15140122) at 37 °C with 5% CO[2] in a humidified incubator. Human
dermal fibroblasts obtained from a 3‐year‐old normal healthy individual
(GM05565) and a 3‐year‐old HGPS patient (AG06917) were purchased from
Coriell Cell Repositories (Camden, NJ). Fibroblast cells were cultured
in DMEM supplemented with 15% FBS and 1% penicillin–streptomycin at 37
°C with 5% CO[2] in a humidified incubator. Culture media used in this
work were refreshed every 2–3 days. To inhibit myosin II activity, HeLa
cells were treated with 10 and 15 µm blebbistatin (Sigma‐Aldrich,
B0560) for 1 h. Control cells, without blebbistatin, were treated with
DMSO (Merck, 317275) and incubated under identical conditions. After
DMSO or blebbistatin treatment, fresh culture medium was added prior to
immunostaining and imaging.
Plasmids and Subcloning
To generate stable cell lines with doxycycline‐inducible expression of
wild type LMNA or Δ50 LMNA tagged with/without mCherry, PCR‐amplified
DNA sequences for mCherry‐LMNA (Addgene, #55068), mCherry‐Δ50 LMNA
(modified from Addgene, #17653), and Δ50 LMNA (modified from Addgene,
#17653) were inserted into PiggyBac XLone‐GFP plasmid (Addgene, #96930)
after digesting with KpnI, SpeI to replace the GFP sequence.^[ [332]^36
^] To deliver genetic cargo into the genome, blasticidin (Thermo
Fisher, [333]R21001)‐resistant gene‐containing PiggyBac transposon was
used.^[ [334]^60 ^] Telomeric repeat‐binding factor 2 (TRF2)‐GFP
plasmid used for analyzing chromatin dynamics was modified.
PCR‐amplified DNA sequences for TRF2 were obtained using TRF2‐IRES‐eGFP
(Addgene, #19798), and PCR‐amplified DNA sequences for pcDNA3‐EGFP
(Addgene, #13031). The TRF2 and pcDNA3‐EGFP were ligated using the
NEBuilder HiFi DNA assembly master mix (NEB, E2621L) to construct the
EGFP‐TRF2 plasmid.
Tet‐On System for Doxycycline‐Inducible Gene Expression
HeLa cells were seeded in T25 rectangular canted neck cell culture
flasks to reach 70–80% confluency after culturing for 24 h in DMEM
supplemented with 10% FBS and 1% penicillin–streptomycin at 37 °C and
5% CO[2] in a humidified incubator. The cells were then transfected
with the generated plasmids (1 µg µL^−1; mCherry‐LMNA, mCherry‐Δ50
LMNA, or Δ50 LMNA) and the PiggyBac transposase (0.5 µg µL^−1; System
Biosciences, PB210PA‐1) using Lipofectamine 3000 (Thermo Fisher,
L3000015) following the supplier's instructions. After 72 h of
incubation, the cells were replated onto 96‐well cell culture plates
(SPL, 30096) and selected with 5 µg mL^−1 blasticidin for 7 d to
generate stable cell lines expressing the respective target genes. To
induce gene expression, HeLa cells transfected with mCherry‐lamin A‐,
mCherry‐Δ50 lamin A‐, or Δ50 lamin A were treated with 2 µg mL^−1
doxycycline (Merck, D9891).
Immunofluorescence and Time‐Lapse Live Cell Monitoring
After growing for 24 h in a glass bottom dish (SPL, 101350) coated with
0.2 mg mL^−1 type‐I rat tail collagen (Corning, 354236) diluted in 0.2
N acetic acid, the cells were fixed with 4% paraformaldehyde
(Biosesang, PC2031‐100‐00) for 10 min at 4 °C, permeabilized with 0.1%
Triton X‐100 (Merck, T8787) for 10 min, and then blocked with
phosphate‐buffered saline (PBS, Corning, 21‐031‐CV) supplemented with
10% FBS for 30 min at room temperature (RT). Fixed cells were incubated
with primary antibodies for 1 h at RT. After washing with PBS three
times, secondary antibodies with 4′,6‐diamidino‐2‐phenylindole
dihydrochloride (DAPI, Sigma‐Aldrich, D1306) and Alexa Fluor 488
phalloidin (Thermo, A12379) were added. Primary antibodies used in this
study are as follows: anti‐Progerin (1:200, Abcam, ab66587), anti‐Lamin
B1 (1:500, Abcam, ab16048), anti‐SUN1 (1:200, Millipore, ABT273),
anti‐Nesprin 2 (1:100, Sigma‐Aldrich, MABC86), anti‐phospho‐myosin
light chain 2 (Ser19) (1:200, CST, 3675), and di/tri‐methyl‐histone H3
(Lys9) (1:100, CST, 5327). Secondary antibodies used are as follows:
Goat anti‐Mouse IgG Heavy and Light Chain Antibody DyLight 488
Conjugated (Bethyl, A90‐116D2), Sheep anti‐Rabbit IgG Heavy and Light
Chain Antibody DyLight 488 Conjugated (Bethyl, A120‐100D2), Goat
anti‐Mouse IgG‐Heavy and Light chain Antibody DyLight 594 Conjugated
(Bethyl, A90‐116D4), Sheep anti‐Rabbit IgG‐Heavy and Light chain
Antibody DyLight 650 Conjugated (Bethyl, A120‐100D5). All samples were
imaged using confocal laser microscopy (A1R, Nikon) through 20x plan
lens with a z‐stack of 0.4 µm or 60× oil lens with a z‐tack of 0.2 µm
or using fluorescence microscopy (Ti2, Nikon). For time‐lapse live cell
monitoring, HeLa cells expressing mCherry‐Δ50 lamin A were seeded onto
a glass bottom dish and imaged through a 20× plan lens using a confocal
microscope equipped with a stage‐top incubator (Okolab, Italy). Live
cell images were captured every 20 min for 36 h to monitor progerin
expression and NE wrinkling, and every 30 min for 16 h to collect
z‐stacked NE wrinkling images. Z‐stacked time‐lapse confocal images
were reconstructed and 3D images were rendered using NIS elements
software (Nikon).
Nuclear Morphometry and Protein Content Measurement
High‐throughput cell phenotyping analysis was employed to assess cell
and nuclear size, nuclear aspect ratio, nuclear circularity, and
protein expression at a single‐cell level.^[ [335]^61 ^] Briefly,
immunostained cells were autofocused through the DAPI channel, and
multiple images were automatically captured on a scale of 7 × 7 in two
or three channels with DAPI, fluorescein‐5‐isothiocyanate (FITC), and
tetramethylrhodamine through a 20× plan lens using a fluorescence
microscope. Image analysis was conducted using a customized program
coded in MATLAB (MathWorks Laboratory). In the acquired images, a
threshold value was applied to each FITC and DAPI channel image to
separate the cell and nuclear regions, prior to calculation of the
fluorescence intensity. To determine the nuclear deformation, the
fractional occupancy of nuclear wrinkles shown as the bright regions in
the nucleus marked by expression of mCherry‐Δ50 lamin A or
immunostaining of Lamin B1 was quantified. Specifically, the degree of
NE wrinkling was calculated using the following equation
[MATH: NEwrinkling%=NuclearwrinklingareaNuclearspreadingarea×100 :MATH]
(1)
where the nuclear wrinkling area and the nuclear spreading area
indicate the nuclear region showing a higher intensity value than the
average intensity value of the entire nucleus and the area of the
fluorescence‐marked entire nucleus, respectively.
Preparation of Polyacrylamide Hydrogel Substrates
The surface of glass bottom dishes was pretreated with (3‐aminopropyl)
trimethoxysilane (Sigma‐Aldrich, 281778) for 5 min and then with 0.5%
glutaraldehyde (Sigma‐Aldrich, G6257) for 30 min. After washing with
deionized water, acrylamide/bis‐acrylamide solution containing ammonium
persulfate (Sigma‐Aldrich, A3678) and tetramethylenediamine
(Invitrogen, 15524‐010) was added onto the surface‐modified glass
bottom dish. Dichlorodimethylsilane (Sigma‐Aldrich, 40140) precoated
cover slips were immediately placed onto the droplets to form a flat
PAG. PAGs were activated with Sulfo‐SANPAH (Thermo Fisher, 22589),
followed by coating with type‐I rat tail collagen overnight at 4 °C. To
modulate the stiffness of the PAG, the concentration ratios of
acrylamide and bis‐acrylamide were varied as follows: 5% acrylamide +
0.06% bis‐acrylamide (E ≈ 1.37 kPa) and 10% acrylamide + 0.3%
bis‐acrylamide (E ≈ 34 kPa). The elastic moduli of the gels were
adapted from previous studies.^[ [336]^62 ^]
Construction of Nuclear Tension Sensor
To measure nuclear tension, a previously developed nesprin tension
sensor was modified (Addgene, #68127).^[ [337]^63 ^] Briefly, the
FRET‐based nesprin tension sensor consisted of mTEP1 and venus, which
were separated by a 40‐amino acid elastic linker flanked by XhoI and
NotI restriction sites. To modify the nesprin tension sensor to
accommodate the confocal microscopy wavelength, the mTEP1 and venus
were replaced with enhanced green fluorescence protein (EGFP) and
DsRed, respectively, using the pQCXI Puro DsRed‐LC3‐GFP plasmid
(Addgene, #31182).
FRET Imaging and Analysis
To determine nuclear tension, cells cultured on glass bottom dishes
were transfected with a modified nesprin tension sensor construct. EGFP
and DsRed were excited by 488 and 561 nm lasers, respectively. FRET
signals of cells expressing the nesprin tension sensor were captured at
488 nm excitation wavelength through a 60× oil lens using the Nikon A1R
confocal laser microscope equipped with a stage‐top incubator. Images
were acquired on the same day at a fixed gain and laser intensities in
each channel, and then analyzed using NIS‐Element software. The FRET
ratio was defined as the ratio of energy transferred from the donor to
the acceptor depending on the distance between the donor and acceptor
proteins, as described by the following equation
[MATH: FRETratio=IntensityofDsRed−Backgr
oundofDsRedIntensityofEGFP−Backgro
undofEGFP :MATH]
(2)
The FRET ratio was color‐coded for better visibility.
Computational Analysis of Nuclear Deformation
A mechanical model of a soft spherical elastic thin shell was developed
to characterize the deformation of nuclear surface. The spherical shell
is discretized into many (5120 in this work) identical triangle face
elements. Each face element contains 3 vertices and 3 sides, for the
entire sphere shell, excluding repeated counted vertices, there exist a
total of 2562 independent vertices. According to Euler's formula of
polyhedral
[MATH: V−E+F=2 :MATH]
(3)
5120 independent sides are defined, where V is the number of vertices,
E is the number of sides, and F is the number of faces.
The deformation and motion of the shell are described by the vertices,
and the motion of each vertex follows the over damped Langevin equation
[MATH: ηdridt=−∇i<
mi>U+Fir+Pie :MATH]
(4)
where η is the damping coefficient, r [i] is the position vector of
vertex i. U is the total shape potential energy of the entire shell,
and ∇ [i] means to derive the gradient in terms of r [i] . F [i] ^r is
the repulsive force exerting onto vertex i by other vertices that are
too close to it and P [i] ^e is the external pressure exerting onto
vertex i from the environment.
Using the expression of charged elastic shell,^[ [338]^64 ^] the shape
potential energy U of the sphere shell is written as
[MATH: U=Ue+Ub+UV :MATH]
(5)
with
[MATH: Ue=∑i=1EKe2li−l<
/mi>02 :MATH]
(6)
[MATH: Ub=∑i=1EKb2n^i,1−n^i,22 :MATH]
(7)
[MATH: UV=∑i=1FKV2Vi−V<
/mi>02 :MATH]
(8)
where U [e] represents the stretching energy, generated from the
extension of shortening of all sides, K [e] is the stretching modulus,
l[i] is the actual length of side i, l [0] is the initial side length
shared by all sides. U [b] represents the bending energy, generated
from the bending of the membrane featured by the dihedral angle between
each two adjacent faces, K [b] is the bending modulus,
[MATH: n^i,1 :MATH]
and
[MATH: n^i,2 :MATH]
are the unit outward normal vectors of the 2 adjacent faces of side i.
U [e] and U [b] describe the mechanical properties of the membrane
itself. U [V] represents the volume energy, generated from the
elasticity of the contents of the shell, K [V] is the volume modulus,
V[i] is volume element i’s volume defined by face i, and V [0] is the
initial volume shared by all elements.
F [i] ^r represents the repulsive forces between 2 very close vertices,
and the traditional 6–12 law is adopted to characterize the molecular
level interactions
[MATH: Fir=−Kr∑jε
Ci1ri−r<
/mi>j6−1ri−r<
/mi>j12
ri−r<
/mi>j :MATH]
(9)
where K [r] is the repulsion coefficient, C[i] represents the set of
other vertices whose distance from vertex i is less than a threshold
distance D.
The cytoskeleton in connected to the nuclear lamina through the LINC
complexes, and the LINC complexes are regarded as a molecular spring.
Once the progerin is expressed, the distance between cytoskeleton and
the nuclear lamina gets reduced (refer to FRET Imaging and Analysis),
leading to a reduced nuclear tension, thus the spring can generate
greater compressive pressure between the cytoskeleton and nucleus
membrane. Hence, a uniform and constant external pressure term P [i] ^e
is used to model the force exerting onto the shell (nucleus membrane)
by the environment (cytoskeleton)
[MATH: Pie=−Pie∇i<
mi>V :MATH]
(10)
where P[i] ^e is the scalar value of P [i] ^e, and V is the total
volume of the shell. With increasing substrate stiffness, the pressure
also increases, therefore, P[i] ^e is set higher on stiffer substrate.
Specifically, P[i] ^e = 500P [0], 600 P [0], and 750 P [0] are set on
soft, medium, and stiff substrates, respectively. All parameters and
values used in the model were listed in Table [339]1 .
Table 1.
Parameters for mechanical model.
Parameter Numerical value Refs.
F 5120 [[340]71]
V 2562
E 7680
K [e] 1000 [[341]64]
K [b] 4
K [V] 3
K [r] 1 × 10^−13
[MATH: Pie :MATH]
500 on soft substrate [[342]72]
600 on medium substrate
750 on stiff substrate
η 1 [[343]73]
dt 0.005
D 0.9
[344]Open in a new tab
RNA Sequencing
Cells cultured on glass and 1.37 kPa PA gel were used to extract total
RNA using the RNeasy Mini Kit (QIAGEN, 74104). A library was
independently prepared with 1 µg of total RNA for each sample by
Illumina TruSeq Stranded mRNA Sample Prep Kit (Illumina, Inc., San
Diego, CA, USA, #20020595). Libraries were quantified using KAPA
Library Quantification kits for Illumina Sequencing platforms according
to the qPCR Quantification Protocol Guide (KAPA BIOSYSTEMS, #KK4854)
and qualified using a TapeStation D1000 ScreenTape (Agilent
Technologies, # 5067–5582). Indexed libraries were then subjected to
Illumina NovaSeq (Illumina, Inc., San Diego, CA, USA), and paired‐end
(2 × 100 bp) sequencing was performed by Macrogen Inc. (South Korea).
Reference genome sequences and gene annotation data were downloaded
from the NCBI Genome Assembly and RefSeq databases, respectively. The
aligned data (SAM file format) were sorted and indexed using SAM tools
v 1.9. After alignment, the transcripts were assembled and quantified
using StringTie v2.1.3b.^[ [345]^65 ^] Gene‐ and transcript‐level
quantification were calculated as the raw read count, Fragments Per
Kilobase of transcript per million mapped reads, and transcript per
million mapped reads. The statistical significance of the DEGs was
determined using the edgeR most exact, and the fold change and p‐value
were extracted from the most exact results. All p‐values were adjusted
using the Benjamini–Hochberg algorithm to control the false discovery
rate. Hierarchical clustering of the log‐transformed values for
significant genes was performed using these parameters (distance metric
= Euclidean distance; linkage method = complete). Gene‐enrichment and
functional annotation analysis for significant genes were carried out
using gProfiler (Raudvere, Uku, et al. 2019,
[346]https://biit.cs.ut.ee/gprofiler/orth) against Gene Ontology (GO)
database. Adjusted p‐values reported from the gProfiler results were
derived using a one‐sided hypergeometric test and corrected using the
Benjamini–Hochberg method.
Assay for Transposase‐Accessible Chromatin Using Sequencing (ATAC Sequencing)
For ATAC sequencing, samples were prepared according to previously
established methods. Briefly, cells were detached from the glass and
1.37 kPa polyacrylamide hydrogel; 100 000 cells were prepared. The
cells were then lysed in a cold lysis buffer. The nuclei concentration
was determined using LUNA‐FL Automated Fluorescence Cell Counter (logos
biosystems) and the nuclei morphology was examined using microscopy.
Immediately after lysis, the transposition reaction was continued using
Tagment DNA TDE1 Enzyme and Buffer Kit (Illumina). Nuclei (50 000
cells) were resuspended in the transposition reaction mixture by
incubating for 30 min at 37 °C. Immediately after transposition, cells
were purified using the Qiagen MinElute PCR Purification Kit. ATAC‐seq
was performed by Macrogen, Inc. (South Korea). The cleaned reads were
aligned to the human genome (GRCh38) using Bowtie2.^[ [347]^66 ^] The
mapped data (in SAM file format) were sorted and indexed using SAM
tools (version 1.9).^[ [348]^67 ^] After removing the reads aligned to
the mitochondrial genome from the indexed BAM file, duplicate reads
were removed using MarkDuplicates in Picard (version 0.118). Peaks in
the aligned sequence data were identified using a model‐based analysis
of ATAC‐seq (MACS2 version 2.1.1.20160309).^[ [349]^68 ^] The algorithm
empirically models the length of ATAC‐seq fragments from sequence data,
considering local genomic biases in the distribution of mapped reads.
The following parameters were used: macs2 callpeak‐t ATAC‐seq. bam‐g
hs–bdg–nolambda–keep‐dup all–broad. Among the called peaks, those that
overlapped with the ENCODE blacklisted regions were excluded.
ChIPseeker (version 1.22.1),^[ [350]^69 ^] a Bioconductor package in R
(version 4.2.2) used to facilitate batch annotation of enriched peaks,
was used to identify nearby genes and transcripts from the peaks
obtained by MACS2. The HOMER software was used to discover de novo TF
binding motifs within ATAC‐seq‐defined regions. An integrative genome
viewer (IGV; [351]http://igv.org/) was used to visualize the genome
track corresponding to the gene locus.
Chromatin Dynamics Analysis
Chromatin dynamics were analyzed by tracking telomere movement in
three‐year normal healthy (GM00565) and HGPS fibroblast cell lines
(AG06917) using the EGFP‐TRF2 plasmid. Briefly, 2.0 µg of EGFP‐TRF2 in
Opti‐MEM (Thermo Fisher, 31985062) was mixed with 2.0 µL of
Lipofectamine 3000 and 4.0 µL of P3000 transfection reagent in
Opti‐MEM. The mixture was gently incubated for 20 min at RT, followed
by a 4 h incubation at 37 °C and 5% CO[2] in a humidified incubator.
Next, transfected EGFP‐TRF2 cells were imaged at 0.07 s intervals for 5
min. Trajectory images were analyzed using the LIM Tracker Plugin in
the Plugins menu of ImageJ, and MSD analysis was performed using a
custom MATLAB script.
Immunoblotting
Cellular lysates were prepared using RIPA buffer (Thermo Fisher
Scientific, 89901) supplemented with a 1% protease inhibitor cocktail
(Sigma, P8340). These preparations were then electrophoresed using
SDS‐PAGE, and the resolved proteins were transferred to a
polyvinylidene difluoride membrane (Thermo Fisher, LC2002). Nonspecific
interactions were blocked using a protein‐based blocking reagent
(Invitrogen, T2015) for 1 h. The membranes were then incubated with
primary antibodies against H3K9me2/3 (CST, 5327), SUN1 (Proteintech,
24568), Nesprin 2 (Abcam, ab233034), and GAPDH (Sigma, G8795) overnight
at 4 °C. Subsequently, membranes were incubated with the appropriate
secondary antibodies, goat anti‐mouse IgG heavy and light chain
antibody HRP Conjugated (Bethyl, A90‐116P), and goat anti‐rabbit IgG
heavy and light chain antibody HRP Conjugated (Bethyl, A120‐101P) at RT
for 1 h. Proteins were visualized using an ECL kit (Thermo Fisher
Scientific, 32106).
Proximity Ligation Assay (PLA) Analysis
Subconfluent cells were fixed on the glass bottom dish with 4%
paraformaldehyde for 10 min at 4 °C and permeabilized with 0.1% Triton
X‐100 for 10 min at RT. After blocking with Doulink Blocking Solution
(Sigma‐Aldrich, DUO82007) for 60 min at 37 °C in a humidity chamber,
samples were incubated overnight at 4 °C with anti‐SUN1 (1:200,
Millipore, ABT273), anti‐Lamin A/C (1:200, Cell Signaling, 4777S), or
anti‐Nesprin 2 (1:100, Sigma‐Aldrich, MABC86). Cells were washed with
1× wash Buffer A (Sigma‐Aldrich, DUO82049) and incubated in preheated
humidity chamber for 1 h at 37 °C with anti‐PLUS and anti‐MINUS PLA
probes diluted 1:5 in the Duolink Antibody Diluent (Sigma‐Aldrich,
DUO92013). Samples were then incubated in a humidity chamber at 37 °C
with ligation solution diluted 1:40 in the 1× ligation buffer. After 30
min, ligation buffer diluted 1:80 in 1× amplification buffer was added
to the samples, followed by incubation for 100 min at 37 °C in
preheated humidity chamber. Samples were mounted using Duolink in situ
mounting medium (Sigma‐Aldrich, DUO82040) with DAPI. The cells were
imaged using confocal laser microscopy through a 60× oil lens. To
confirm the level of protein–protein interaction, the number and total
intensity of red dots per cell were assessed based on the fluorescence
signal.^[ [352]^70 ^]
siRNA‐Mediated Knockdown
To knockdown SUN1, siRNA targeting SUN1 (Santa Cruz Biotechnology,
sc‐106672) and control siRNA (Santa Cruz Biotechnology, sc‐37007) were
delivered to progerin‐expressing HeLa cells or HGPS fibroblast cells
(Coriell Cell Repositories), according to the manufacturer's
instructions. Briefly, 75 pmol siRNA in Opti‐MEM (Thermo Fisher,
31985062) and 7.5 µL Lipofectamine 3000 transfection reagent in
Opti‐MEM were gently mixed and incubated for 5 min at RT.
Progerin‐expressing HeLa cells at 70–80% confluency on glass bottom
dishes were transiently transfected with siControl or siRNA targeting
SUN1 for 24 h. After removing the transfection mixture and replacing it
with fresh growth medium, cells were immunostained for analysis.
Data Processing and Statistical Analysis
All statistical analyses were performed using GraphPad Prism (GraphPad
software, USA), and statistical significance was analyzed using
unpaired t‐test or Student's t‐test when two groups were compared.
One‐way analysis of variance (ANOVA) using Tukey's test or Bonferroni's
post hoc test was used for pairwise comparisons between multiple
groups. Error bars represent the standard deviation (S.D.) or standard
error of the mean (S.E.M.), as indicated. Significance levels are
indicated with asterisks in each figure as follows: *: p < 0.01; **: p
< 0.005; ***: p < 0.001; ****: p < 0.0001.
Conflict of Interest
The authors declare no conflict of interest.
Author Contributions
J.P., J.J., and K.X. contributed equally to this work. J.P. and J.J.
performed and analyzed all experiments unless otherwise specified and
cowrote the manuscript. K.X. developed computational model to simulate
the nuclear wrinkling. S.L. and S.H. analyzed the protein contents and
chromatin dynamics in the human dermal fibroblasts. J.J., Y.L., and
W.C. developed the Tet‐On system. W.C., B.L., and S.H.K. provided a
conceptual design of the experimental settings. K.X. and B.L. cowrote
the manuscript. D.K. supervised the project, designed the experiments,
and wrote the manuscript.
Supporting information
Supporting Information
[353]ADVS-12-2502375-s002.docx^ (8.4MB, docx)
Supplemental Movie 1
[354]Download video file^ (499KB, avi)
Supplemental Movie 2
[355]Download video file^ (610.9KB, avi)
Supplemental Movie 3
[356]Download video file^ (1.1MB, avi)
Supplemental Movie 4
[357]Download video file^ (1.7MB, avi)
Acknowledgements