Cytoskeleton stiffness regulates cellular senescence and innate immune response in Hutchinson–Gilford Progeria Syndrome

Abstract Hutchinson–Gilford progeria syndrome (HGPS) is caused by the accumulation of mutant prelamin A (progerin) in the nuclear lamina, resulting in increased nuclear stiffness and abnormal nuclear architecture. Nuclear mechanics are tightly coupled to cytoskeletal mechanics via lamin A/C. However, the role of cytoskeletal/nuclear mechanical properties in mediating cellular senescence and the relationship between cytoskeletal stiffness, nuclear abnormalities, and senescent phenotypes remain largely unknown. Here, using muscle‐derived mesenchymal stromal/stem cells (MSCs) from the Zmpste24 −/− (Z24 −/−) mouse (a model for HGPS) and human HGPS fibroblasts, we investigated the mechanical mechanism of progerin‐induced cellular senescence, involving the role and interaction of mechanical sensors RhoA and Sun1/2 in regulating F‐actin cytoskeleton stiffness, nuclear blebbing, micronuclei formation, and the innate immune response. We observed that increased cytoskeletal stiffness and RhoA activation in progeria cells were directly coupled with increased nuclear blebbing, Sun2 expression, and micronuclei‐induced cGAS‐Sting activation, part of the innate immune response. Expression of constitutively active RhoA promoted, while the inhibition of RhoA/ROCK reduced cytoskeletal stiffness, Sun2 expression, the innate immune response, and cellular senescence. Silencing of Sun2 expression by siRNA also repressed RhoA activation, cytoskeletal stiffness and cellular senescence. Treatment of Zmpste24− / − mice with a RhoA inhibitor repressed cellular senescence and improved muscle regeneration. These results reveal novel mechanical roles and correlation of cytoskeletal/nuclear stiffness, RhoA, Sun2, and the innate immune response in promoting aging and cellular senescence in HGPS progeria.


| INTRODUC TI ON
Hutchinson-Gilford progeria syndrome (HGPS) is a rare, fatal genetic disorder caused by the mutation of LMNA (lamin A) gene (Schreiber & Kennedy, 2013). As a nucleoskeletal protein at nuclear lamina, lamin A is essential for mechanical support of the nucleus and is required for the structural link between the nucleoskeleton and cytoskeleton (Phillip, Aifuwa, Walston, & Wirtz, 2015). LMNA gene mutation in HGPS leads to the production of mutant prelamin A (progerin), which accumulates at the nuclear envelope (NE) and causes dramatic changes in the nuclear architecture, including thickening of the nuclear lamina, increased nuclear stiffness and nuclear irregularity (nuclear blebbing, a hallmark of HGPS cells), and impaired nucleus deformation capacity (Cao et al., 2011;Dahl et al., 2006;Phillip et al., 2015;Verstraeten, Ji, Cummings, Lee, & Lammerding, 2008;Young, Fong, & Michaelis, 2005). This abnormal nuclear architecture confers additional adverse cellular processes, such as disruption of chromatin anchoring on the laminar structure, inappropriate reorganization of chromatin, telomere dislocation, dysfunction and erosion, delayed response in DNA damage repair and thereby increased DNA damage, and accelerated senescence (Cao et al., 2011;Phillip et al., 2015). Progerin accumulation can also cause natural age-associated increase in nuclear stiffness, abnormal histone modification patterns, global changes in gene expression, and impaired cell function (Cao et al., 2011;Pacheco et al., 2014;Phillip et al., 2015). In contrast to HGPS cells, lamin A knockout (LMNA −/− ) cells do not express lamin A and were reported to develop decreased mechanical stiffness in both the nucleus/nucleoskeleton and cytoskeleton (Broers et al., 2004;Kim et al., 2017;Schreiber & Kennedy, 2013). Although the increased nuclear stiffness in HGPS cells had been previously observed (Booth, Spagnol, Alcoser, & Dahl, 2015;Dahl et al., 2006), changes in cytoskeletal stiffness in progerin-expressing cells and the potential correlation of cytoskeletal stiffness with nuclear abnormalities and progeria phenotypes have not been examined.
The cell nucleus is tightly integrated into the structural network of the cytoplasm through linker of the nucleoskeleton and cytoskeleton (LINC) complexes, which contain Sun1/2 proteins (Isermann & Lammerding, 2013;Phillip et al., 2015). Lamin A/C is required for this structural connection of nucleus and cytoskeleton (Isermann & Lammerding, 2013;Phillip et al., 2015), which is essential for a broad range of cellular functions. Mechanical stimuli to the cells can be transmitted from the extracellular matrix (ECM) to the nucleus via the cytoskeleton (Isermann & Lammerding, 2013;Phillip et al., 2015). The mechanical properties of the ECM, especially the stiffness of the local environment, can have a direct and profound impact on the expression of mechano-responsive genes and the organization of cytoskeletal and nucleoskeletal proteins (Isermann & Lammerding, 2013;Phillip et al., 2015).
Filamentous actin (F-actin), a key component of the cytoskeleton, is a major determinant of a cell's mechanical properties and is critical to stress response. F-actin plays a crucial role in mediating ECM-nuclear mechanical coupling. The stiffness of local environment influences cellular functions via regulating cytoskeletal structure, and cells on stiffer substrates exhibited increased F-actin polymerization and stress fiber formation with an order of orientation. Lamin A/C is connected to F-actin cytoskeleton via Sun proteins (Sun1/2), which can mediate the formation of the perinuclear apical actin cap to regulate the nuclear structural integrity (Hoffman et al., 2020;Kim et al., 2017;Lei et al., 2009). Normal actin cytoskeletal dynamics can protect cells from stress by modulating the stress response and cell death (Baird et al., 2014;Gourlay, Carpp, Timpson, Winder, & Ayscough, 2004). Actin cytoskeletal dynamics is closely regulated by the activity of Rho GTPases, particularly RhoA (Sit & Manser, 2011). RhoA is crucial for regulating cell morphology, migration, adhesion, autophagosome formation and function, and many more events associated with F-actin dynamics (Li et al., 2011). The interaction of RhoA with key mechano-sensing factors located at the cell membrane, cytoplasm/cytoskeleton, or nuclear membrane has been well documented (Li et al., 2011). In particular, our recent study demonstrated the co-activation of RhoA with pro-inflammatory, pro-fibrogenic, and pro-osteogenic signaling factors in skeletal muscles of muscular dystrophic mice (Mu et al., 2013).
Moreover, RhoA also plays a role in determining the commitment of stem cell fate, by regulating cytoskeletal mechanics and interacting with mechanical signaling pathways involved in stem cell differentiation (Li et al., 2011). Cell shape, mechanical cues, and cytoskeletal tension were found to regulate the switch in lineage commitment of MSCs by modulating RhoA activity (Li et al., 2011;McBeath, Pirone, Nelson, Bhadriraju, & Chen, 2004). Abnormal RhoA regulation can lead to the dysregulated adipo-osteogenic balance, which has been linked to various pathological conditions, such as aging, obesity, osteopenia, osteopetrosis, and osteoporosis (Li et al., 2011). Indeed, dysregulated adipo-osteogenic balance also has been reported in HGPS MSCs with increased osteogenic, but decreased adipogenic potential (Meshorer & Gruenbaum, 2008). Thus, it is possible that RhoA signaling may play a role in conferring the dysregulated adipo-osteogenic balance in HGPS cells. and improved muscle regeneration. These results reveal novel mechanical roles and correlation of cytoskeletal/nuclear stiffness, RhoA, Sun2, and the innate immune response in promoting aging and cellular senescence in HGPS progeria.

K E Y W O R D S
accelerated aging, cell nucleus, cellular senescence, skeletal muscle, stem cells Based on these previous findings, we hypothesized that there is increased cytoskeletal stiffness and RhoA activation in HGPS cells, possibly as a response to increased nuclear stiffness, DNA damage, and/or production of reactive oxygen species (ROS). We also hypothesized that increased cytoskeletal stiffness and RhoA activation are associated with accelerated cellular senescence. Thus, we examined cytoskeletal stiffness and RhoA activation in cells from HGPS patients and from a mouse model of HGPS due to a deficiency in Zmpste24 (Z24), a zinc metalloproteinase involved in the formation of mature lamin A. Z24 −/− mice have premature onset of aging-related musculoskeletal changes, similar to those observed in HGPS (Bergo et al., 2002;Fong et al., 2004;Yang et al., 2006). Through a gain and loss of RhoA function, we examined whether changes in cytoskeletal stiffness and RhoA activation in progeria cells correlate with nuclear blebbing, Sun1/2 expression, and micronuclei-induced cGAS-Sting activation, part of the innate immune response that has also been linked to senescence. Because the LINC complex containing Sun2 was found to promote focal adhesion assembly by activating RhoA (Thakar, May, Rogers, & Carroll, 2017), we also examined the effect of siRNA-mediated Sun2 inhibition on cytoskeleton stiffness, perinuclear actin cap formation, nuclear blebbing, and cellular senescence.

| RE SULTS
2.1 | MSCs from muscle of Z24 −/− mice have elevated expression of senescence-associated secretory phenotype (SASP) factors that negatively impacts muscle stem cell function Mesenchymal stromal/stem cells in skeletal muscle are nonmyogenic and express platelet-derived growth factor receptors (PDGFRs).
These PDGFR-α + cells often function as fibro-adipogenic progenitors (FAPs) and have been found as the main cell type that contributes to pathological fibrosis and fat infiltration in diseased and injured F I G U R E 1 Increased expression of SASP factors in Z24 −/− MSCs negatively impacts muscle stem cell function. (a) Gastrocnemius (GM) skeletal muscle tissues were harvested from 5-month-old Z24 −/− and wild-type (WT) mice. Immunostaining analysis of PDGFR-α and CD68 showed increased number of PDGFR-α + MSCs and CD68 + macrophages. Scale bar = 100 µm. (b) Quantification of CD68 + cells is shown. (c) Immunostaining analysis of dystrophin and Pax7 showed a decreased number of Pax7 + muscle stem cell in Z24 −/− mice. (d) Quantification of the ratio of the PDGFR-α + cells to Pax7 + cells is shown. (d) Immunostaining analysis of PDGFR-α and Pax7 demonstrating a close interaction between these two types of cells in stem cell niche. White arrow indicates a Pax7 + cell; orange arrow indicates a PDGFR-α + cell. Scale bar = 50 µm. (e) Quantification of mRNA level of PDGFR-α in muscles is shown. (f) Immunostaining analysis of PDGFR-α and Pax7 to show their relative localization at stem cell niche. Scale bar = 10µm. (g) Immunostaining analysis of PDGFR-α in mesenchymal stem/stromal cells (MSCs) from WT mice and Z24 −/− mice. Scale bar = 100 µm. (h) qPCR results of mRNA from WT MSC and Z24 −/− MSCs. (i) Treatment of WT muscle progenitor cells (MPCs) with conditioned medium from WT or Z24 −/− MSCs to check the impact on myogenesis potential [the formation of fast-myosin heavy chain (f-MHC)-positive myotubes], and level of DNA damage (γ-H2AX). Quantitation of cells positive with f-MHC or γ-H2AX is shown. Scale bar = 100 µm. (j) qPCR results of mRNA from WT MPCs treated with conditioned medium from WT or Z24 −/− MSCs. Data are shown as mean ± standard error. N ≥ 6. * indicates p < .05 muscles (Contreras, Rebolledo, Oyarzun, Olguin, & Brandan, 2016).
Analysis of skeletal muscle from Z24 −/− mice (5-month-old) showed there was increased activation of PDGFR-α + cells and CD68 + cells Immunostaining of MSCs demonstrated that both Z24 −/− and WT MSCs were predominantly PDGFR-α positive (Figure 1g). qPCR analysis further demonstrated that Z24 −/− MSCs specifically expressed higher level of senescence and SASP markers (i.e., p16 INK4a , CXCL1, MCP1, IL-1α, IL-1β, IL-6, and TNFR1) and the pro-fibrotic factor TGF-β1 (Figure 1h), whereas the expression of anti-inflammatory factors (i.e., IL-10 and Klotho) was down-regulated. In addition, conditioned medium (CM) from Z24 −/− MSCs was able to confer repression of F I G U R E 2 Z24 −/− MSCs display increased senescent phenotypes, and enhanced F-actin polymerization and cytoskeletal stiffness is directly associated with increased and nuclear blebbing. MSCs isolated from the skeletal muscle of WT and Z24 −/− mice were compared. (a) Immunostaining analysis of γ-H2AX, p21 Cip1 , and lamin A/C was performed, as well as SA-β-Gal staining for senescence. Quantitation of γ-H2AX + cells, SA-β-Gal + cells, p21 + cells, and cells with nuclear blebbing is shown. Scale bar = 30µm. (b) Immunostaining analysis and quantification of H3K9me3. The increased level of H3K9me3 (red) in the micronuclei in contrast to nucleus indicates the loss of heterochromatin from nucleus to micronuclei (arrows). Scale bar = 2.5 µm. (c) Staining of F-actin with Alexa Fluor 488 Phalloidin and quantification of F-actin polymerization. Scale bar = 20 µm. (d-g). Testing of cytoplasm stiffness using a Bruker AFM probe. H. The cytoplasm stiffness (kPa) calculated by NanoScope analysis. (i) Immunostaining analysis of lamin A/C and F-actin in WT and Z24 −/− MSCs, showing higher level of F-actin and nuclear blebbing in same Z24 −/− cell (arrow). Scale bar = 50 µm. (j) Immunostaining analysis of lamin A/C and F-actin in Z24 −/− MSCs. Quantitation of nuclear blebbing is shown. The number of cells with nuclear blebbing was compared between cells with top 30% of F-actin intensity (Actin-high) and cells with bottom 30% of F-actin intensity (Actin-low). Scale bar = 30 µm. (k) Immunostaining analysis of lamin A/C and F-actin to observe the effect of treatment of Z24 −/− MSCs with F-actin stabilizing JPK (200 nM) or F-actin depolymerizing CyD (100 ng/ml) for 48 hr. Quantitation of nuclear blebbing is shown. Scale bar = 15 µm. Arrows: nuclear blebbing. N ≥ 6. "*" at bar charts indicates p < .05 MSCs, whereas expression of IL-10 was up-regulated ( Figure S5).

| Constitutive activation of RhoA signaling promotes nuclear abnormalities and cellular senescence
In Sun proteins (i.e., Sun1 and Sun2) at the nuclear envelope transduce mechanical stress from the ECM and cytoskeleton into the nucleus (Lei et al., 2009). LINC complexes that contain Sun2, but not Sun1, were found to promote focal adhesion assembly by activating RhoA (Thakar et al., 2017). Thus, we examined whether Sun2 may mediate the effect of RhoA activation in promoting nuclear blebbing in SiRNA reduced RhoA activity ( Figure S14d) and down-regulated expression of senescence-associated genes ( Figure S14e). In addition, treatment of Z24 −/− MSCs with either Sun2 SiRNA or Y-27632 reduced the level of Sun2 protein, but not Sun1 protein (Figure 4j,k).

| Systemic inhibition of RhoA in Z24 −/− mice extended the healthspan and rescued the defective phenotypes in skeletal muscle
To examine the effect of RhoA inhibition in vivo, 10-week-old Z24 −/− mice were treated with Y-27632 (10 mg/kg) by i.p. injection 3 times a week for 12 weeks. Following this treatment regimen, the number of myogenic Pax7 + cells was increased in the skeletal muscles of Z24 −/− mice (Figure 6a,b), whereas the number of SA-β-Gal + cells, pro-inflammatory CD68 + cells, and PDGFR-α + cells was reduced ( Figure 6b). Also, RhoA inhibition improved the muscle regeneration potential of Z24 −/− mice following cardiotoxin-induced injury (7 days after injury) (Figure 6a,c). This suggests that RhoA inhibition improves the function of muscle stem cells and muscle regeneration both by increasing the number of muscle stem cells and repressing the deleterious impact of senescent cells on muscle stem cells. Based on these results, we propose a potential cellular and molecular mechanism of muscle stem cell dysfunction in Z24 −/− mice, involving the over-activation of RhoA signaling, and pro-inflammatory, and pro-fibrogenic factors in MSCs and inflammatory cells ( Figure S16).

| Human HGPS fibroblasts have increased F-actin polymerization, which was reduced by RhoA/ ROCK inhibitor and farnesyltransferase inhibitor (FTI)
To extend the results from the Z24 −/− mouse model of HGPS, fibroblasts from HGPS patients also were examined. Similar to murine  Figure S17). In addition, AFM probe testing of cell stiffness showed that HGPS fibroblasts seemed to have a higher cytoskeletal stiffness than normal fibroblasts ( Figure S18).

The farnesyltransferase inhibitor (FTI) lonafarnib currently is the only approved medication for HGPS patients, effective in revers-
ing the dramatic nuclear structure abnormalities and to alleviate some disease symptoms. FTIs act by inhibiting the farnesylation of prelamin A (Mehta, Eskiw, Arican, Kill, & Bridger, 2011). Intriguingly, the activity of RhoA protein also is regulated by farnesylation. This common farnesylation-dependent regulatory mechanism of both prelamin A and RhoA proteins suggests that RhoA might serve as a promising target for therapeutic treatment of HGPS. We observed that treatment of HGPS fibroblasts with lonafarnib efficiently reduced F-actin polymerization, nuclear blebbing, and Sun2 expression (Figure 6d-g).

| D ISCUSS I ON
The LMNA mutation in HGPS disease causes nuclear abnormalities and cellular fragility in response to mechanical stress; however, correlations between genotypes and mechanical phenotypes in the cells remain unclear. Also, the role of cytoskeletal stiffness and its potential association with nuclear abnormality in progeria cells also remains to be investigated. Progerin production occurs not only in progeria cells, but also in naturally aged human cells (Booth et al., 2015;Cao et al., 2011;Pacheco et al., 2014;Phillip et al., 2015).
The role of cytoskeletal stiffness in mediating the deleterious effect of progerin in promoting cellular senescence and aging is poorly understood. Therefore, we examined the mechanism of cytoskeletal stiffness-mediated nuclear variations in HGPS cells and in a HGPS mouse model. Here, we demonstrate the crucial role of cytoskeletal stiffness in mediating nuclear abnormalities in cells from HGPS patients and a progeria disease mouse model as well as the mechanical roles of RhoA and Sun2 in modulating cytoskeletal stiffness, nuclear blebbing, and senescence.
Our results are consistent with previous reports of increased cytoskeletal stiffness in certain types of senescent cells, which also suggests that cytoskeletal stiffness is coupled with nucleoskeletal stiffness (Nishio, Inoue, Qiao, Kondo, & Mimura, 2001;Schulze et al., 2010). Our results further indicate that activation of the RhoA signaling and Sun2 protein is crucial for mediating the accelerated cellular senescence of MSCs and aging-associated pathologies in the muscle of HGPS patients. Interestingly, a recent chemical screening study identified ROCK as a target for recovering mitochondrial function in HGPS cells. Here, Y-27632 treatment of HGPS cells induced the recovery of mitochondrial function and reduction of abnormal nuclear morphology (Kang et al., 2017), which is consistent with our results. However, our study further revealed the biomechanical mechanism of RhoA/ROCK activation in mediating cellular senescence, including the direct connection of RhoA/ROCK activation with cell stiffness and nuclear blebbing, the interaction of RhoA/ ROCK and Sun2 in response to cytoskeleton stiffness and nuclear blebbing, and the correlation among RhoA/ROCK activation, Sun2 expression, micronuclei formation, and the innate (cGAS/Sting) immune response.
There have been previous reports describing the normal cytoskeletal stiffness in HGPS cells (Verstraeten et al., 2008) and reduced cytoskeletal stiffness in LMNA mutant (LmnaL530P/L530P) mouse cells (Hale et al., 2008). These different results in cytoskeletal stiffness compared to our studies may be partially attributed to the different cell lineages, cell culture conditions, or stiffness testing methods. LmnaL530P/L530P cells do not produce progerin, and there is a decreased level of lamin A in nuclear lamina (Hale et al., 2008). The weaker/softer nuclear lamina in LmnaL530P/L530P cells could be responsible for the reduced cytoskeleton stiffness observed. However, LMNA mutation in HGPS cells leads to progerin accumulation and trapping of lamin A/C in the nuclear lamina, forming orientationally ordered microdomains and stiffer lamina network (Dahl et al., 2006). We propose that RhoA activation is an indirect outcome of the LMNA mutation in progeria cells through two possible mechanisms.
The first mechanism involves the elevated cytoskeletal mechanical stress induced by mutation of LMNA. This leads to progerin accumulation in nuclear lamina, which increases the stiffness of lamina network (Dahl et al., 2006). This mechanical stress from lamina network can be transmitted to cytoskeleton via the LINC complex, and mechanical sensing factors in the cytosol can mobilize RhoA activation to promote F-actin polymerization. The second possibly mechanism involves the increased Sun2 expression induced by Our results suggest that both increased ECM stiffness and nuclear stiffness of progeria cells contribute to increased RhoA activation and cytoskeleton stiffness. Also, elevated ROS production, inflammatory signaling (NF-κB), and DNA damages in progeria cells can promote RhoA activation, resulting in further increased cytoskeleton stiffness. Increased RhoA and Sun2 expression, nuclear blebbing, and cytoplasmic DNA fragments in micronuclei, can all mediate accelerated cellular senescence. N ≥ 6. "*" at bar charts indicates p < .05

The potential mechanism of cytoskeleton stiffness and cellular senescence in HGPS cells
Telomere Telomere LMNA mutation. Recent studies indicates that Sun2 is mechanical responsive (Hoffman et al., 2020), and the expression of Sun2 in LINC complex can be modulated by varied mechanical features of lamina network. Sun2 was found to promote RhoA activation (Thakar et al., 2017), and thus, the LMNA mutation may cause RhoA activation by increasing Sun2 expression in HGPS cells.
Previous observations have demonstrated the role of Sun1 in responding to progerin accumulation in nuclear lamina (Haque et al., 2010;Lei et al., 2009). It was shown that the reduction of Sun1 accumulation in LMNA mutant cells (i.e., Lmna −/− and LmnaΔ9 fibroblasts) corrected nuclear defects and cellular senescence (Chen et al., 2012). We believe that this beneficial effect of Sun1 reduction in progeria cells may be partially due to the regulation of mechanical features of nucleus or cytoskeleton since nuclear lamina network is connected to cytoskeleton by both Sun1 and Sun2 in the LINC complexes.
However, we did not observe the accumulation of Sun1 in our cells, potentially because of the different cell culture condition and genotypes of cell lineages, especially the potential difference in progerin generation in cells. Recent studies have suggested a potential role for Sun2 in mechano-sensing and involvement in mechanical stimulation-induced changes in actin cytoskeleton (Hoffman et al., 2020). Our results have confirmed further the mechanical role of Sun2 in mediating increased nuclear stiffness, nuclear abnormalities, and cellular senescence, as well as a potential mechanical feedback loop of Sun2 with RhoA.
In aged tissues, the extracellular matrix (ECM) usually becomes stiffer because of increased collagen deposition, varied composition of collagen subtypes, and the crosslinking structure of collagen fiber.
Cells adapt to this increased ECM stiffness by modifying the expression of mechano-responsive genes. Mechanical stimuli from the ECM can be transmitted to the nucleus via the cytoskeleton and LINC complex, and directly contribute to the triggering of mechano-responsive genes to adapt to the mechanical environment (Isermann & Lammerding, 2013;Phillip et al., 2015). Both lamin and Sun proteins are required for this structural connection between the nucleus and cytoskeleton (Isermann & Lammerding, 2013;Phillip et al., 2015) and for transducing mechanical stresses from ECM and cytoskeleton into nucleus, which is essential for a broad range of cellular functions.
Furthermore, the expression of both lamin A and Sun2 was found to be mechano-responsive and scales with ECM/substrate stiffness or elasticity. A stiff substrate is associated with increased lamin A and Sun2 expression and strengthened nuclear envelope, whereas a soft substrate for cells is associated with reduced lamin A and Sun2 expression. In aged and progeroid cells, progerin accumulation in the nuclear lamina causes increased nuclear stiffness (Booth et al., 2015;Cao et al., 2011;Dahl et al., 2006;Phillip et al., 2015;Verstraeten et al., 2008;Young et al., 2005). Our current results with Z24 −/− MSCs and HGPS fibroblasts demonstrate the increased accumulation of Sun2 protein in the nuclear envelope, especially in cells with abnormal nuclear architecture (nuclear blebbing). The increased activation of Sun2 in these in vitro cultured progeria cells is not caused by stiffer substrate/ECM, but rather is a result of an intrinsic change in the mechanical properties of the cell nucleus. We propose that Sun2 accumulation in progeria cells is a result of progerin accumulation because there is direct physical interaction between the Sun2 and lamin A proteins. Mechanical stimuli resulting from increased nuclear stiffness could be transmitted to the cytoskeleton via LINC, leading to increased F-actin polymerization. The Sun2 protein also can activate RhoA signaling directly (Thakar et al., 2017), which may further enhance F-actin polymerization and cytoskeletal stiffness. Therefore, increased cytoskeletal stiffness in progeria cells could actually be the results of mechanical stress from both ECM stiffness (Hernandez et al., 2010) and nuclear stiffness (Figure 6c). We have proposed the potential mechanisms of cytoskeletal stiffening in mediating nuclear blebbing, micronuclei formation, and cellular senescence, the involvement of the RhoA and Sun2 factors, and how increased ECM, nucleoskeletal, and cytoskeletal stiffness may be coupled in progeria cells.
Our results also demonstrate that there is sustained activation of RhoA signaling in progeria cells, which promotes F-actin cytoskeletal stiffness and impairs proper F-actin dynamics. RhoA activation in progeria cells could be a result of Sun2 activation as well as increased pro-inflammatory signaling, pro-fibrotic signaling, DNA damage, ROS production, and activated mechano-sensing signaling (Li et al., 2011). Lamin A or Zmpste24-deficiency has been shown to cause dysregulated differentiation capacity of stem cells, including increased osteogenic potential and repressed adipogenic potential.
Our results strongly suggest that increased cytoskeletal stiffness and RhoA activation in progeria cells are determinant for this disrupted adipo-osteogenic balance of cell fate.
Aging is associated with senescence and depletion of functional stem cells in various tissues, leading to impeded tissue homeostasis and regeneration capacity. Senescence and depletion of stem cells can be caused by both cell autonomous mechanisms and noncell autonomous mechanisms (Baar, Perdiguero, Munoz-Canoves, & de Keizer, 2018). Senescent cells have a senescence-associated secretory phenotype (SASP), which alters the microenvironment and impacts the function of nonsenescent cells (Baar et al., 2018).
The SASP can affect the stem cell niche, impacting stem cell function via a noncell autonomous mechanism. In skeletal muscle, mus- whereas MPC proliferation and function declined (Mu et al., 2013;Sohn, Lu, Tang, Wang, & Huard, 2015). Similarly, homeostasis and regeneration of aged skeletal muscle can be impaired by macrophage-mediated chronic inflammation and fibrogenic cell-mediated fibrosis formation. SASP factors from senescent cells also can chronically impair muscle stem cell function and attenuate tissue regeneration in muscle. Therefore, the defective muscle regenerative capacity of muscle stem cells depends intimately on the microenvironment modulated by various types of senescent cells, and senescent Z24 −/− MSCs could contribute to the impaired function of muscle stem cells in Z24 −/− muscle. It is notable that increased intrinsic defects including nuclear abnormalities, DNA damage, and cellular senescence also can be induced by LMNA mutation in Pax7 + muscle stem cells, and the paracrine effect by MSCs is not the only causal factor of muscle stem cell defects.
Importantly, treating Z24 −/− mice with a RhoA inhibitor reduced senescence, suggesting that RhoA inhibitors represent a novel type of senotherapeutic.
The increased micronuclei formation and cytoplasmic DNA induced by DNA damage can activate cGAS-Sting innate immune signaling and contribute to cellular senescence (Dou et al., 2017). The potential correlation of cGAS-Sting signaling with mechanical properties of a cell has not been addressed before our study. Current results suggest that increased RhoA activation, cytoskeletal stiffness, and Sun2 expression in progeria cells are associated with increased generation of nuclear blebs/micronuclei and cytoplasmic DNA fragments, thus mediating the activation of cGAS-Sting signaling and cellular senescence progression.
Taken together, our novel results demonstrate that increased cytoskeletal stiffness and RhoA/Sun2 signaling represents a novel mechanism for promoting aging and cellular senescence and therefore represents a novel therapeutic target for intervention in progeria diseases and for extending healthspan with natural aging.
These results extend our understanding regarding how mutations in lamin A, a structural and mechano-responsive protein in nuclear lamina network, changes the mechanical properties of nuclear and cytoskeleton and contributes to driving cellular senescence and thus aging and age-related diseases.

| Muscle cell isolation and culturing
Muscle-derived mesenchymal stem/stromal cells (MSCs) and muscle stem/progenitor cells (MPCs) were isolated from the skeletal muscle

| Human fibroblasts from HGPS and normal patients
The fibroblasts from HGPS patients and normal humans were kindly shared from the Coriell Institute and were cultured in DMEM medium supplemented with 15% fetal bovine serum (FBS), 1X glutamax, and 1% penicillin-streptomycin. The HGPS fibroblasts cell strains include AG03513, AG06917, and AG08466, and normal fibroblasts include AG08468, AG08469, and AG08470. Living cells gotten from Coriell Institute were cultured to passage 3 to be compared in our studies.

| Atomic force microscopy (AFM) testing of cell stiffness
Cells were fixed 4% paraformaldehyde, and the cytoplasm or nuclear stiffness was measured with AFM system at room temperature (Collum et al., 2017). The force curves measurements were using NanoScope analysis program from the Bruker Corporation.

| Telomere staining
Telomeres in the interphase nuclei were detected using a Cy3conjugated peptide nucleic acid (PNA) probe provided with the Telomere PNA FISH Kit (Agilent). Fluorescence in situ hybridization of telomere was performed according to the manufacturer's instruction. Briefly, the cells were heated at 80°C for 5 min to denature DNA in the presence of the Cy3-conjugated PNA probe followed by hybridization in the dark at room temperature (RT) for 30 min. The hybridization was followed by a brief rinse with a Rinse Solution, and a posthybridization wash with a Wash Solution at 65°C for 5 min. DAPI was used to counterstain DNA.

| In vitro RhoA inhibition assays
Z24 −/− MSCs or HGPS fibroblasts cultured in collagen-coated plastic plates were treated with a specific RhoA/Rho kinase (ROCK) inhibitor Y-27632 (EMD Millipore, Billerica, MA) (10 μM) for 48 hr in cell proliferation medium, and the potential changes of F-actin polymerization, nuclear blebbing, lamin A/C, Sun2, γ-H2AX, and Histone methylation/acetylation markers were observed. RhoA-specific inhibition also was performed with C3 transferase (5 µg/ml) treatment for 48 hr to verify the effect of RhoA inhibition on F-actin polymerization and nuclear blebbing.

| Measurement of cellular senescence
The percent of senescent muscle cells cultured in vitro (10% FBS in DMEM for 4 days) and in skeletal muscle tissue was measured using the senescence-associated β-Galactosidase (SA-β-gal) Staining Kit (Cell Signaling Technology) following the manufacturer's protocol. The number of cells positive for β-gal activity at pH 6, a known characteristic of senescent cells, was determined.

| Preparation of conditioned medium (CM)
The WT MSCs and Z24 −/− MSCs were plated and grown to ~90% confluence in proliferation medium, and the culture medium was changed to low-serum medium (DMEM with 2% FBS) to initiate the conditioning of culture medium. The cells were continued to be cultured for 48 hr to collect the CM by filtering through 0.4-µm filter.
The cell number of WT MSCs and Z24 −/− MSCs was similar when CM was initiated and collected. The CM was mixed with same volume of fresh medium (1:1) when being used for cell treatment.

| In vitro myogenic, osteogenic, and adipogenic differentiation assays
MPCs from 5-month-old WT or Z24 −/− mice (30 000 cells per well of 12-well collage type I-coated plastic plate), maintained in DM, 2% horse serum in DMEM (Invitrogen), were cultured with different conditioned medium (CM) from Z24 −/− MSCs with or without Y-27632 pretreatment. The Z24 −/− MSCs were pretreated with or without Y-27632 for 48 hr and continued to be cultured for another 48 hr to harvest the CM. Z24 −/− MPCs were then treated with the CM for 4 days to observe the progression of myogenic differentiation.

| Constant activation of RhoA in MSCs
Rho activator II (Cytoskeleton Inc.), which can robustly increase the level of GTP-bound RhoA and result in constitutively active Rho (Schmidt et al., 1997), was added to WT MSCs cultured on collagen-coated plates at 200 ng/ml for 24 hr to observe increased RhoA protein and F-actin polymerization. Expression of constitutively active RhoA-GFP was also performed with amplified plasmid of pcDNA3-EGFP-RhoA-Q63L (Addgene, plasmid #12968). Expression of a dominantly negative RhoA-GFP with amplified plasmid of pcDNA3-EGFP-RhoA-T19N (Addgene, plasmid #12967) served as a negative control. Lipofectamine 3,000 (Thermo Fisher) was applied to facilitate the plasmid transfection into WT MSCs.

| Epigenetics assay
The heterochromatin markers histone H3K9me3 and H3K27me3 were immunostained in Z24 −/− MSCs with and without Y-27632 treatment (10 μM for 48 hr) or Sun2 siRNA treatment, and HGPS cells with and without Y-27632 treatment. The anti-H3K9me3 and H3K27me3 antibodies were from Abcam.

| qRT-PCR
Total RNA was obtained from muscle cells or muscle tissues using the RNeasy Mini Kit (Qiagen, Inc.). Reverse transcription was performed using an iScript cDNA Synthesis Kit (Bio-Rad Laboratories, Inc.). The sequences of primers were given in Table   S1 for SASP genes (IL-1α, IL-1β, IL-6, TNF-α, TNFR1, MCP1, and CxCl2), IL-10, Klotho, PDGFRs, GAPDH (glyceraldehyde 3-phosphate dehydrogenase), etc.. PCR reactions were performed using an iCycler thermal cycler (Bio-Rad Laboratories, Inc.). The cycling parameters used for all primers were as follows: 95°C for 10 min; PCR, 40 cycles of 30 s at 95°C for denaturation, 1 min at 54-58°C for annealing, and 30 s at 72°C for extension. Products were separated by size and visualized on a 1.5% agarose gel stained with ethidium bromide. All data were normalized to the expression of GAPDH.

| In vivo RhoA inhibition with Y-27632
Systemic inhibition of RhoA/ROCK signaling was conducted by

| Cardiotoxin-induced muscle injury
In order to compare the muscle regeneration potential in WT mice, Z24 −/− mice, and Z24 −/− mice treated with Y-27632, cardiotoxin (Sigma) (10 μM in 40 μl of PBS) was injected intramuscularly into the GM muscles of mice (~20-week old) to induce muscle injury. Muscle tissues were collected at 7 days after the injection of cardiotoxin to compare fibrosis formation, muscle regeneration, and number of senescent cells.

| Histology and immunofluorescent staining
Frozen tissue sections were fixed with 10% formalin, and cultured cells were fixed with 4% paraformaldehyde. All of the primary an-

| Measurements of results and statistical analysis
Image analysis was performed using Nikon NIS-Elements (Nikon Instruments Inc.) and Image J software (National Institutes of Health). Data from at least three samples were pooled for statistical analysis. Results are given as the mean ± standard deviation (SD).
The statistical significance of any difference was calculated using Student's t test or one-way ANOVA test. P values < .05 were considered statistically significant.
Foundation (to JPC) and by funding from the University of Texas Health Science Center at Houston. We also thank Dr. Mary Hall and Dr. Lavanya Rajagopalan for editorial assistance while completing this manuscript.

CO N FLI C T O F I NTE R E S T
Johnny Huard discloses the fact that he receives royalties from Cook Myosite, Inc. for muscle stem cell technologies.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.