The distinct roles of myosin IIA and IIB under compression stress in nucleus pulposus cells

Inappropriate or excessive compression applied to intervertebral disc (IVD) contributes substantially to IVD degeneration. The actomyosin system plays a leading role in responding to mechanical stimuli. In the present study, we investigated the roles of myosin II isoforms in the compression stress‐induced senescence of nucleus pulposus (NP) cells.


| INTRODUC TI ON
Intervertebral disc (IVD) degeneration and secondary pathological changes are the leading cause of low back pain, which is at the top of the main reasons of chronic disability. 1 The IVD is an avascular tissue that consists of three inter-related structures: the highly hydrated nucleus pulposus (NP) at the centre, the highly fibrocartilaginous annulus fibrosus on the outer periphery and the endplates on the upper and lower faces. 2 The IVD undergoes substantial mechanical stimuli during spinal motion and joint muscle movement. During loading, cells of the IVD undergo compressive, tensile, shearing deformation and fluid flows, all of which play an significant role in cell metabolism of the IVD. [3][4][5][6] NP cells are substantial contributors in the extracellular matrix (ECM) anabolism to maintain the gelatinous property of NP tissue, allowing it to buffer the various mechanical stimuli. 7,8 Increasingly, evidence suggests that the cell-mediated and biological remodelling responses to mechanical stimuli play an important role in IVD degeneration. 9,10 Many studies have sought to elucidate the biological responses of NP cells to mechanical stimuli; however, the mechanical signal sensing and mechanical signal transduction processes in NP cells are still poorly understood.
The actomyosin system plays a leading role in the response to mechanical stimuli, including stretch, compression and shear forces, as well as internally generated tension and strain. 11,12 The actomyosin system consists of actin and myosin, which form stress fibres for generating contractile forces. The members of the myosin family are grouped into more than 30 classes that have different distributions and play diverse functions. 13 Myosin II has emerged as an obvious player in a variety of force processes, such as cytokinesis, cell migration, polarization and adhesion. [14][15][16] In mammalian cells, myosin II heavy chains include three different isoforms: myosin IIA, myosin IIB and myosin IIC encoded by MYH9, MYH10 and MYH14 genes, respectively. Myosin IIA and IIB can be detected in most tissues whereas myosin IIC is absent in some tissues. 17 Although myosin IIA and IIB present a high degree of homology in the amino acid sequences of their heavy chains, they display differences in organization, intracellular distributions and molecular interactions. For example, in epithelial junction assembly, myosin IIA is parallel to the junction and provides mechanical force for the maintenance of adherens junction. Myosin IIB localizes at junctional membranes and organizes the junctional branched actin meshwork. 18 A recent study demonstrated that myosin IIA is dedicated to generating cortex tension for faster cleavage furrow ingression in cell division, while myosin IIB acts as a stabilizing motor by reducing cortex tension. 19 In our previous studies, we have demonstrated that compression stress can induce apoptosis of NP cells and ECM degradation during IVD degeneration. 20,21 However, little is known about the distinct roles of myosin II isoforms in compression stress-induced senescence of NP cells.
Furthermore, the actomyosin cytoskeleton is highly regulated by the RhoA/ROCK signalling pathway. RhoA belongs to the Ras homology proteins family of 22 small GTPases. 22 RhoA can switch between an inactive GDP-bound form and an active GTP-bound state, located in the cell membrane when it is active. 23,24 Its major biological functions include the generation of actomyosin bundles, stress fibres, focal adhesions and lamellipodia. 25 Rho-associated coiled-coil-containing protein kinases (ROCK) are dominating regulators of the actin cytoskeleton, which downstream of the small GTPase RhoA. The ROCK family includes two members, ROCK1 and ROCK2, which have 65% homology in amino acid sequence. 26,27 The ROCK proteins act on target substrates including the myosin regulatory light chains (MLC) and a myosin-binding subunit of the myosin phosphatase through which they regulate myosin contractility, and the LIM kinases through which they control actin depolymerization.
The RhoA/ROCK signalling has been reported to act as a prominent player in responding to mechanical stimuli. 26,28,29 However, the correlation between myosin II and the RhoA/ROCK signalling under compression stress has rarely been studied.
In our study, we investigated the role of myosin II subunits in com-

| Compression treatment
According to the previous studies, a pressure of 1.0 MPa is widely used to induce IVD degeneration. 20,21,32 A compression apparatus contains a mixture of 5% CO 2 and 95% compressed air was used to provide 1.0 MPa compression as described in our previous work. 21 In addition, the apparatus was placed in an incubator at 37°C. NP cells were exposed to 1.0 MPa compression for 0, 12, 24 or 36 hours.

| Total and nuclear-cytosol protein extraction and Western blot analysis
Total protein was extracted from human NP cells by RIPA lysis buffer containing protease inhibitor (Beyotime). Nuclear and cytoplasmic proteins were extracted from NP cells by a Nuclear-Cytosol Extraction Kit (Solarbio) according to the manufacturer's instructions. Western blot procedure was performed as previously described. 33 Primary antibodies against the following proteins were used: myosin IIA

| Quantitative real-time polymerase chain reaction (qRT-PCR)
Total RNA was extracted from human NP cells with TRIzol reagent

| Senescence-associated β-galactosidase (SA-βgal) staining
The senescence of NP cells was determined using a SA-β-gal Staining Kit (Beyotime) following the manufacturer's protocol. Briefly, cells were washed with PBS (pH 7.4) and fixed in an SA-β-gal working solution (pH 6.0) at 37°C without CO 2 overnight. Then, the average percentage of total SA-β-gal-positive cells was calculated for quantitative analysis.

| Immunofluorescence analysis
Immunofluorescence analysis was performed as previously described. 33 Briefly, 4% paraformaldehyde was used to fix attached human NP cells, then 0.2% Triton X-100 in PBS was used to permeabilize. The slides were washed in PBS and blocked with 2% bovine serum albumin (BSA) in PBS for 2 hours at 37°C and then incubated with primary antibodies against: myosin IIA (Abcam, ab138498, 1:200), myosin IIB (Abcam, ab230823, 1:100), MRTF-A (Abcam, ab115319, 1:200). After washed twice, the slides were subsequently treated with secondary goat anti-rabbit antibody (Boster) at 37°C for 2 hours. Nuclei were co-stained for 5 minutes with 0.1 g/mL DAPI (Beyotime, Nantong, China), and images were captured under a microscope (Olympus, BX53). To analyse the nuclear-to-cytoplasmic fluorescent intensity ratio of MRTF, the nuclear intensity and cytoplasmic intensity of 20 cells from each group were calculated and analysed by Image J (NIH). For co-localization analysis of myosin II isoforms with F-actin, the Pearson coefficient was calculated by Image J (NIH) and the Coloc 2 plugin.

| Flow cytometric and cell cycle analysis
The human NP cells were harvested through trypsinization, fixed with 70% cold ethanol overnight at 4°C. Next, the cells were incubated with RNase (50 μg/ml; KeyGEN) for 30 min at 37°C, followed by propidium iodide dye (50 μg/ml; KeyGEN) for a further 30 min. The cells were then analysed using flow cytometry (BD FACSCalibur; BD Biosciences).

| Immunoprecipitation
The Protein A/G Magnetic Beads (MCE) was washed with lysis buffer three times before mixing with antibody. Prior to immunoprecipitation, 5 μg of purified antibodies against myosin IIA, myosin IIB, actin filaments or normal IgG were mixed with 400 μl of lysis buffer with 25 μl beads and incubated 2 hours at 4°C on a rocker table. The prepared antibody-beads complex was added to 1 ml of cell lysates (1 mg/ml) and incubated with rocking for 4 hours at 4°C. Then, magnetic separation was performed, and the supernatant was removed.
Subsequently, the immune complex was mixed with 5 × loading buffer and then boiled for 10 minutes before the analysis by Western blot.

| Immunohistochemistry
NP samples were fixed in 10% formaldehyde for 24 hours and embedded in paraffin. Then, the samples were sliced into 4-μm sections.
Immunohistochemistry was carried out as described previously. 34 The sections were incubated with antibodies against: p-RhoA-

| Statistical analysis
Data are presented as the means ± SD of three independent experiments. Statistical analyses were performed using GraphPad Prism 8.0 software. Differences between groups were evaluated with Student's t test or one-way ANOVA by analysis of variance. Pvalue < .05 was considered statistically significant.

| Compression stress induced the senescence of human NP cells
To investigate the senescence level of human NP cells under compression stress, we used SA-β-gal staining to access cellular senescence. In addition, changes in ECM metabolism and cell proliferation were measured. As shown in Figure 1A over time ( Figure 1C). As the exposure to compression stress was prolonged, there was a progressive reduction in the expression of aggrecan and collagen type II genes ( Figure 1D). Simultaneously, collagen type I expression increased, indicating a fibrotic phenotype in human NP cells. In addition, flow cytometric and cell cycle analysis were used to determine the proliferation of human NP cells. As shown in Figure 1E

| Compression stress regulated the interaction of myosin IIA and IIB with actin
Given that the actomyosin system acts as a prominent player in responding to mechanical stimuli, it is important to explore how it assembles into the necessary structures in human NP cells under compression stress. To understand the role of myosin IIA and IIB in the compression stress-induced senescence of NP cells, the interaction of myosin II subunits and F-actin under compression stress were measured. In normal human NP cells, myosin IIA tended to be broadly peripheral (Figure 2A), while myosin IIB was distributed throughout the cytoplasm ( Figure 3A). Compression stress exposure induced dramatic changes in the reorganization of myosin IIA and IIB with F-actin. Immunofluorescence analysis revealed that myosin IIA exhibited an increased co-localization with F-actin exposed to compression than normal (Figure 2A  Protein interaction between myosin IIB and actin was determined by co-immunoprecipitation. Following treatment, cell lysates were immunoprecipitated with anti-actin antibody. Isotypematched (IgG) served as negative control. Each precipitated sample was detected for the presence of myosin IIB and actin by immunoblot analysis using specific antibodies. Whole cell lysates prior to the immunoprecipitation served as input controls. (E, F) Cell lysates were immunoprecipitated with anti-non-muscle myosin IIB antibody, the next steps are described above. Data were presented as the mean ± SD (n = 3). *P < .05, vs control; **P < .01, vs control | 9 of 17 KE Et al.

| Distinct effects of myosin IIA or IIB knockdown in NP cells under compression stress
To determine the specific roles of myosin IIA and IIB in human NP cells under compression stress, knockdown of myosin IIA or IIB was achieved by transfection with siRNA (siMyosin IIA or siMyosin IIB).

| Compression stress induced RhoA/ROCK1 pathway activation in human NP cells
Given that the actomyosin cytoskeleton is highly regulated by the RhoA/ROCK signalling pathway, we hypothesized that this pathway  Figure 6A). In addition, the expression of ROCK1 increased progressively as the exposure to compression stress was prolonged ( Figure 6B). In contrast, compression stress did not affect the protein level of ROCK2. MLC phosphorylation (p-MLC, a marker of the active form of MLC) increased after 24 hours of compression stress and the ratio of p-MLC to MLC increased with time ( Figure 6C). Furthermore, immunohistochemistry showed an increase in the proportion of p-RhoA-, ROCK1-and p-MLC-positive cells in the IVD degeneration group compared to that in the control group, while no obvious change was detected in ROCK2 expression ( Figure 6D). Together, these findings indicated that compression stress induced RhoA/ROCK1/p-MLC pathway activation in human NP cells.

| Inhibition of the RhoA/ROCK1 pathway attenuated compression stress-induced human NP cells senescence by regulating the actomyosin cytoskeleton remodelling
To explore the role of the RhoA/ROCK1 pathway in compression stress-induced human NP cells senescence, Y27632, a ROCK1 inhibitor, was used to inhibit the RhoA/ROCK1 pathway. Interestingly, immunofluorescence and statistical analysis revealed that Y27632 significantly reversed the co-localization between myosin IIA and actin filaments ( Figure 7A). In addition, inhibition of the RhoA/ ROCK1 pathway rescued the decreased interaction between myosin IIB and actin under compression stress ( Figure 7B).

| D ISCUSS I ON
As the load-bearing structure of the spine, the IVD is subjected to various mechanical stimuli in daily life. 35,36 Inappropriate or excessive compression applied to IVD contributes to IVD degeneration. 9,37 In our previous studies, we demonstrated that compression stress can induce apoptosis of NP cells and ECM degradation during IVD degeneration. 20,21 However, the roles of myosin II in human NP cells under compression stress is poorly understood. Here, we for the first time explored the roles of myosin II isoforms in the compression stress-induced senescence of NP cells (Figure 9).
The actomyosin cytoskeleton is responsible for most forcedriven processes in cells and tissues. 14,38,39 Although the actomyosin is comprised of the same major components, namely actin and myosin II, how these structures are assembled at the right time and place to maintain the normal physiological structure of the cell is an important question. Myosin II has been reported to form polarized cables to maintain tissue elasticity and cell shape upon mechanical stretch. 40 Myosin polarity increased tissue stiffness to protect against fractures and injuries. Another study reported that myosin IIA forms bipolar filaments in red blood cells that are associated with F-actin at the membrane. 41 Myosin IIA activity regulated interactions with the spectrin-actin network to control red blood cell biconcave shape and deformability. In our study, myosin IIA tended to be broadly in peripheral in normal human NP cells, while myosin IIB was distributed throughout the cytoplasm. Compression stress exposure induced dramatic changes in the reorganization of myosin IIA, IIB and F-actin. Myosin IIA had a stronger association with actin filaments under compression stress than normal conditions. In contrast, compression stress reduced the interaction of myosin IIB and F-actin. Overall, our results indicated that compression stress differentially regulated the interaction of myosin IIA and IIB with actin.
Cellular activity, such as growth and motility, is the result of a balance of forces. This balance is regulated by the coordinated operation of molecular motors. [42][43][44] Disorder in the mechanical environment leads to disturbances in various cellular activities. It has been reported that myosin IIA drives cell retraction and maintains tensile adhesion, while myosin IIB drives outgrowth. 45,46 A recent study described the distinct roles of the myosin II submits in regulating actin cortex mechanics during cell division. 19 While myosin IIA was mainly responsible for generating cortex tension, myosin IIB acted as a cortex stabilizer to maintain cortex tension. Myosin IIA depletion decreased cortex tension and slowed furrow ingression.
In contrast, myosin IIA depletion increased intracellular pressure and drove faster cleavage furrow ingression. In our study, we firstly revealed that actomyosin cytoskeleton remodelling participated in the compression stress-induced human NP cells senescence. Moreover, myosin IIA or IIB knockdown generated distinct effects under compression stress. Our results revealed that myosin IIA knockdown rescued the inhibition of proliferation, while myosin IIB knockdown further increased the expression of senescence-related proteins.
Cell division is a delicate process that requires a stable mechanical environment. We speculated that myosin IIA knockdown might provide the environment needed for cell division by reducing the intracellular forces; while myosin IIB exacerbated the imbalance of mechanical environment under compression stress. Thus, our data revealed the diverse effects of myosin IIA or IIB knockdown on compression stress-induced human NP cells senescence.
The actomyosin cytoskeleton is involved in a dynamic cycle of polymerization and depolymerization. When cells are exposed to mechanical stimuli, monomeric G-actin always aggregates into F-actin to maintain tissue shape. 47 Subsequently, MRTF-A relaxes from its complex with G-actin and is translocated into the nucleus where it associates with SRF to activate the transcription of target genes, notably fibroblast-like matrix genes. 48,49 It has been reported that cells growth on a stiff matrix, representative of the pathologic stiffness of Crohn's strictures, expressed increased levels of fibrotic genes and was associated with nuclear localization of the transcriptional cofactor MRTF-A. 50 Another study demonstrated that mechanical strain induced the nuclear translocation of MRTF-A and a pro-fibrotic phenotype in human mitral valvular interstitial cells. 51 A recent study revealed that NP cells in stiff substrate appeared spread and fibrotic shape opposed to the round shape in soft matrix. 52 In addition, matrix stiffness increased nu- cytoskeleton. 53,54 In addition, Rho/ROCK signals regulate the formation of lamellipodia and promote the degradation of ECM. 55 It has been demonstrated previously that substrate stiffness regulated migration and invasion ability of adenoid cystic carcinoma cells via the RhoA/ROCK pathway. 56 Another study also detected increased ROCK activity due to ECM rigidity; however, they found that ROCK1 and ROCK2 differentially regulated invadopodia activity through separate signalling pathways via contractile (myosin II) and non-contractile (LIMK) mechanisms. 57 Interestingly, curved microstructures were found to promote osteogenesis of mesenchymal stem cells by regulating the RhoA/ROCK pathway. 58 In our study, we detected in-

CO N FLI C T O F I NTE R E S T
The authors have declared that no competing interest exists. F I G U R E 8 Inhibition of RhoA/ROCK1 pathway attenuated compression stress-induced human NP cells senescence. (A, B) The level of senescent cells in different groups was assessed by SA-β-gal staining. Scale bar: 50 μm. (C, D) Human NP cells untreated or pretreated with 20 μM Y27632 for 2 h prior to compression stress for 36 h were stained with MRTF-A (green) and DAPI (blue) (n = 20). Scale bar: 10 μm. (E, F) mRNA levels of the ECM remodelling proteinases (MMP3, MMP13 and ADAMTS5) and ECM component (COL1A, COL2A and aggrecan) were measured after the human NP cells were treated withY27632. (G, H) The proportions of cells in each cycle were measured through flow cytometry in different groups. (I, J) CDK4, Cyclin D1, p21 and p53 expression levels in different groups were measured through Western blot analysis and normalized to that of GAPDH. Data were presented as the mean ± SD (n = 3). #No significance, vs compression group; *P < .05, vs compression group; **P < .01, vs compression group; ***P < .001, vs compression group F I G U R E 9 Schematic graph of the role of myosin IIA and IIB in compression stress-induced senescence of NP cells. Compression stress induced the RhoA/ ROCK1 pathway activation, which regulated the interaction of myosin IIA and IIB with actin. The actomyosin cytoskeleton remodelling was involved in the compression stress-induced fibrotic phenotype mediated by MRTF-A nuclear translocation and inhibition of proliferation in human NP cells

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.