Rho-Kinase Is Differentially Expressed in the Adipose Tissue of Rodent and Human in Response to Aging, Sex, and Acute Exercise

Abstract

White adipose tissue (WAT) controls energy storage, expenditure, and endocrine function. Rho-kinase (ROCK) is related to impaired thermogenesis, downregulation of preadipocyte differentiation, and adipokine production. Furthermore, WAT ROCK responds to metabolic stress from high-fat diets or diabetes. However, ROCK distribution in adipose depots and its response to aging and sex remain unclear. Thus, we aim to investigate ROCK function in adipose tissue of rodent and human in response to aging and sex. We observed specific differences in the ROCK1/2 distribution in inguinal WAT (ingWAT), perigonadal WAT (pgWAT), and brown adipose tissue of male and female rodents. However, ROCK2 expression was lower in female ingWAT compared with males, a fact that was not observed in the other depots. In the pgWAT and ingWAT of male and female rodents, ROCK activity increased during development. Moreover, middle-aged female rodents and humans showed downregulation in ROCK activity after acute physical exercise. Interestingly, ROCK levels were associated with several inflammatory markers both in rats and humans WAT (Nfkb1, Tnf, Il1b, Il6, and Mcp1). Induction of cell senescence by etoposide elevates ROCK activity in human preadipocytes; however, silencing ROCK1/2 demonstrates improvement in the inflammatory and cell senescence state. Using public databases, several pathways were strongly associated with ROCK modulation in WAT. In summary, WAT ROCK increases with development in association with inflammatory markers. Further, ROCK activity was attenuated by acute physical exercise, implicating it as a possible therapeutic target for metabolism improvement mediated by adipose tissue inflammatory state changes.

Aging compromises several metabolic and physiological processes in the human body (1,2). In late life and after distinct challenges, our body loses control of basic tasks mediated by the central nervous system (3), skeletal muscle (4), liver (5), and adipose tissue (6). Adipose tissue acts as an energy reservoir, can secrete key endocrine signals to the whole body, contributing to glucose and lipid homeostasis control, energy expenditure, and circulating cytokines (7). All these processes can be altered by aging resulting in drastic changes that have a significant impact on an individual’s life expectancy (6).

Adipose tissue is organized in distinct depots that are distributed in the subcutaneous and visceral regions, each playing a different role in metabolism and in the control of energy expenditure (7). Moreover, these adipose depots can be composed of white, beige, and brown adipocytes. Beige and brown adipocytes are unique in their ability to metabolize glucose and lipids, playing a central role in thermogenesis (7). However, aging leads to a significant dysfunction of these cells and a redistribution of white adipocytes to the visceral cavity and inter/intramuscular area (6). Furthermore, emerging studies have shown that there are sex-specific differences in adipocyte metabolism and distribution in humans which are essential to understanding obesity-related diseases (8). For instance, fat distribution may reflect the differences in the prevalence of diseases, where female humans are protected from metabolic syndrome and type 2 diabetes (9–11). The percentage of body fat in female humans is usually higher than in men but is predominantly accumulated in the subcutaneous region, whereas in men it is more localized to the abdominal cavity (8). This singular distribution implicates the control of energy expenditure, adipokine production, inflammatory profile, free fatty acids release, and other factors that communicate with other organs (12–14).

The process of adipogenesis is driven by the differentiation of adipocytes by a well-described series of processes controlled by pro or anti adipogenic factors whose function could be affected by sex and aging (15–18). In this context, studies have shown a role for Rho-kinase 1 and 2 (ROCK1/2) in controlling metabolic health mechanisms that could be mediated by adipose tissue (19). Several pathological conditions, such as diabetes, obesity, and oxidative stress, are associated with increased ROCK activity in distinct tissues (19), playing a role in glucose homeostasis, adipogenesis, adipocyte differentiation, and inflammation (20–24).

Increased ROCK activity in adipocytes is associated with impaired thermogenic capacity, negative regulation of preadipocyte differentiation, and blunted beige adipocyte differentiation (19). Moreover, a high-fat diet can increase ROCK activity in white adipose tissue (WAT), which contributes to adipocyte hypertrophy, leading to changes in adipocytokine expression, and the recruitment of proinflammatory cells to adipose tissue, which contributes to metabolic syndrome (25). Conversely, the overexpression of an adipocyte-specific dominant-negative RhoA in mice led to ROCK activity attenuation in the adipose tissue, accompanied by decreased WAT hypertrophy and reduced macrophage infiltration (25). Furthermore, ROCK2 may play a dominant role in adipocyte metabolism because adipocyte-specific ROCK1 knockout mice have no differences in WAT hypertrophy and inflammation after a high-fat diet (26), a fact that was observed in mice with partial deletion of ROCK2 (20,23).

Due to the connection of ROCK with these stress conditions mentioned before we speculate that advancing age should increase ROCK content and activity in the white adipose tissue. Consequently, the elevated ROCK activity in the WAT may be a contributing factor to metabolic disorders in advanced ages. Considering that physical exercise is well-known to promote positive metabolic effects through distinct molecular pathways (27), including the ROCK pathway (28–31), we also investigated whether a single bout of physical exercise decreased ROCK activity in the adipose tissue of female humans and female rats. Therefore, the main objective of the study was to investigate whether ROCK1/2 is differentially expressed in the adipose tissue of rodent and human in response to aging, sex, and acute exercise. These findings may contribute to future strategies aimed at suppressing ROCK activity in adipose tissue and preventing the development of metabolic diseases associated with aging.

Experimental Procedures

Experimental Animals and Tissue Collection

The Wistar rats were obtained from the Central Animal Breeding Center of the University of Campinas with 4 weeks of life and maintained in a group cage with 3 animals/cage with free access to water and a standard diet, under thermoneutral conditions (~30 °C). Then, female and male rats were distributed in groups of 5, 8, 11, and 15 mo. The Ethics Committee of UNICAMP approved the experimental procedures (5369-1). When the animals reached the required age, the animals fasted for 6 h, were weighed, and sacrificed after anesthetization (90 mg/kg ketamine chloralhydrate and 10 mg/kg xylazine). The adipose depots (ie, ingWAT, pgWAT, and BAT) were removed for tissue weighting and storing at −80°C. A small fragment of pgWAT, ingWAT, and BAT was fixed in 4% PFA solution for 48 hours and submitted to hematoxylin-eosin (H&E) staining. The images were acquired from different section regions (10−) with an optical microscope (LAB2000, LABORANA, São Paulo, Brazil) coupled with Moticam Pro 282B 5.0 megapixels (Motic, Hong Kong, China). The adipocyte frequency was calculated by measuring the area of the adipocytes for each animal (n = 6/group) following the frequency of the area in 1 adipocyte area interval (0–10 000 µm²).

Molecular Analysis (Immunoblotting)

The fat tissue (~80 mg) was homogenized with protein extraction buffer using TissueLyser II (Qiagen), centrifuged (12 000 rpm at 4 °C for 15 min), followed by supernatant storage. The bicinchoninic acid method measured the sample’s total protein content. Then, ~30 µg/sample was submitted to SDS–PAGE and immunoblotting as described before (28). The nitrocellulose membranes were incubated with the following primary antibodies (4 °C overnight): ROCK1 #sc-5560, GAPDH #47724 from Santa Cruz Biotechnology; ROCK2 #8236, NF-κB p65 #8242, β-actin #3700, α-tubulin #2144 from Cell Signaling Techonology; and pMYPT1^T696 #ABS45 from Millipore. Finally, the images were acquired with G: Box XR5 (Syngene, Frederick, MD, USA).

Acute Treadmill Exercise

Before being submitted to the acute treadmill exercise, female rats at 15 mo. were adapted to a rodent treadmill (5 d, at 10 m/min, for 10 min; AVS Projetos, São Carlos, Brazil). Twenty-four hours after the last adaptation day, the rats performed an incremental load test (3 m/minutes increments every 3 minutes until the animal’s exhaustion.) to determine each rat’s exhaustion velocity (EV). Afterward, the animals were submitted to an acute exercise session at 60% of the EV for 60 minutes on the treadmill. The adipose depots were collected 24 hours after the acute exercise.

Cell Culture

Immortalized mouse preadipocyte cell lines were generated from the stromovascular fraction (SVF) of the subcutaneous fat as described before (47). Immortalized human preadipocytes were obtained from primary SVF from human neck fat as described before (48). The preadipocytes were cultured in high-glucose Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin. Cells were seeded in a 12-well-plate, for the etoposide treatment after reaching total confluence. The cells were treated with Etoposide (E1383, Sigma) with 6.25, 12.5, and 25 µM for 24 hours and collected for immunoblotting. For the siRNA transfection, cells were transfected in a 24-well-plate at 50% confluence using 5 pmol/well of siRNA and Lipofectamine RNAiMAX (13778075, Thermo). The siRNAs were produced by Horizon and the sequences are available in Supplementary Table 1. After 48 hours post-transfection, the human preadipocytes were treated for 24 hours with Etoposide (25 µM), and the cells were collected in TRIzol (15596026, Invitrogen^®) for RT-qPCR. Another plate was used for SA-β-Gal staining following the manufacturer’s instructions (9860, Cell Signaling Techonology).

Quantitative Real-Time Polymerase Chain Reaction (RT-qPCR)

Inguinal WAT, human scWAT, and human preadipocyte cells were processed after the treatments by washing with 1× PBS and adding TRIzol, followed by RNA extraction protocol and complementary DNA (cDNA) synthesis (2 µg) using High-Capacity cDNA Reverse Transcription Kits (4368813, Thermo). Samples from this cDNA were subject to RT-qPCR using 100 ng cDNA, 150 nM primers, and IQ SYBR Green Supermix (1708884, Bio-Rad). The ΔΔCt was calculated and normalized by a housekeeping gene (Actb). The primer sequences are listed in Supplementary Table 1.

Bioinformatics

The public database Genenetwork (www.genenetwork.org) was used to collect the ROCK1 and ROCK2 mRNA levels in the subcutaneous adipose tissue of human samples (GTEXv8). The mRNA levels, expressed as Log2 (Transcripts Count Per Million), were distributed into sex and age according to the subject information. It was also obtained data from BXD mouse strains (33) on GeneNetwork (www.genenetwork.org), dataset: [EPFL/LISP BXD CD Subcutaneous WAT Affy MTA 1.0 Gene Level (Feb16) RMA]. The mRNA levels in BXD samples were obtained as robust multi-array average (RMA) levels. The top5000 ROCK1 or ROCK2 correlated genes in both mouse and human populations were distributed as positive/negative correlated and overlapped in a Venn diagram. Next, the overlapped genes were submitted to Reactome pathway analysis in the DAVID platform V6.8. (https://david.ncifcrf.gov/). The Pearson’s correlations were distributed in different heat maps separated by pathways. The following website was utilized for heat map building (https://software.broadinstitute.org/morpheus/).

Subjects

The subcutaneous adipose tissue (scWAT) from humans used in this study was analyzed from another ongoing project. The volunteers were recruited according to: (a) 20–30 years old female humans; (b) sedentary/inactive; (c) no physical exercise impediments; and (d) 18.5–27 kg/m² body mass index. Individuals with hypertension, diabetes, dyslipidemia, and other pathological conditions were excluded. The interventions were performed with consent terms and the Ethics Committee of the University of Campinas approval (n°: 1.967.450/2017). The samples were obtained from 20–29 years old (n = 6) to 60–73 years old (n = 6) patients. Previous to the acute aerobic exercise protocol and 24 hours after the acute exercise session, the subjects received local anesthesia, and the scWAT samples (~120 mg) were biopsied from the subcutaneous fat depot at the right side of the wait and on the iliac crest. These collections were performed in the morning with a previous fasting period of ~12 hours. For further immunoblotting analysis, the samples were frozen in liquid nitrogen and −80°C.

Acute Aerobic Exercise Protocol

All volunteers performed an exercise protocol on an ergometric treadmill (Quinton, model TM55, United States). A previous cardiorespiratory test (progressive effort protocol on a treadmill, increments of 0.3 km/hour every 30 seconds, with a constant inclination of 1%, until exhaustion) with a continuous collection of expired gases (CPX Ultima, Medical Graphics, United States), was performed to determine the anaerobic threshold (AT). The acute aerobic session was performed based on the individual AT in a single exercise session, in which the volunteers walked and/or ran for 50 minute with varying intensity, 5 minute below the anaerobic threshold (warm-up), 40 minute at the anaerobic threshold and below the respiratory compensation, and 5 minute below the anaerobic threshold (cooldown). The exercise intensity was monitored through the walking/running speed, the heart rate obtained in the cardiorespiratory test, and also through the subjective effort table (49). Twenty-four hours after the exercise session, the post-exercise scWAT samples were biopsied.

Statistical Analysis

The data were analyzed using one-way ANOVA followed by Tukey’s post hoc test or unpaired Student’s t-test. The statistical significance was p < .05. The graph figures illustrate the mean ± standard error of the mean (SEM). The GraphPad Prism 8.0.1 software was utilized for data statistical analysis and graphing.

Results

Adipose Depot Characterization in Male and Female Wistar Rats Across Different Ages

First, we collected samples from pgWAT, ingWAT, and BAT from male and female rats at 5, 8, 11, and 15 months old (mo.) (Figure 1A). Female rats at 11 and 15 mo. had higher body weight compared with 5 mo. (Figure 1B). Male rats at 15 mo. had higher body weight compared with 5, 8, and 11 mo. (Figure 1B). This result reflected the increased abdominal circumference in female and male rats at 15 mo. compared with their controls (Figure 1C). When we analyze the adipose depots of female rats, the ages of 11 and 15 mo. have higher pgWAT than 5 mo., with no differences in the ingWAT and BAT (Figure 1D). Both pgWAT and ingWAT were higher at 11 and 15 mo. in male rats, with no differences in the BAT (Figure 1E). Moreover, we measured the frequency and area of adipocytes in histological samples, demonstrating that after 8 mo of life, the pgWAT was robustly increased in female rats, a fact that did not occur in males (Figure 1F-I). Regarding ingWAT and BAT, no significant adaptations were observed in response to development or sex-specificity (Figure 1J–Q).

ROCK Distribution in Distinct Adipose Depots in Male and Female Rats

In male rats, we observed increased ROCK1 content in the ingWAT compared with pgWAT and BAT and increased ROCK activity (pMYPT^T696) in the BAT compared with pgWAT and ingWAT (Figure 2A). In female samples, ROCK1 was decreased in the pgWAT compared with ingWAT and BAT, ROCK2 was lower in the BAT compared with pgWAT and ingWAT, and ROCK activity (pMYPT^T696) was lower in the ingWAT compared with pgWAT and BAT (Figure 2B). Comparing male and female samples, we observed no differences in the ROCK1, ROCK2, and pMYPT^T696 content in the pgWAT and BAT (Figure 2C and E) but found lower ROCK2 content in the ingWAT of females (Figure 2D).

Acute Exercise Counteracts Age-Related Increased ROCK Activity and Inflammation in the Subcutaneous Adipose Tissue of Rats and Humans

In the pgWAT, ROCK1 content increased in female rats at 11 mo, ROCK2 increased in female rats at 11 and 15 mo, and ROCK activity (pMYPT^T696) increased in 15 mo. female rats (Supplementary Figure 1A). In the ingWAT, ROCK1 and pMYPT^T696 increased in 15 mo. female rats compared with the other groups, and ROCK2 was raised in female rats at 15 mo compared with 5 mo. (Figure 3A). Similarly, at 15 mo. male rats showed higher ROCK1 and higher pMYPT^T696 at 11 and 15 mo. in the ingWAT (Supplementary Figure 1B). In the pgWAT, 11 and 15 mo. male rats had increased content of ROCK1 and ROCK2, accompanied by higher pMYPT^T696 at 15 mo. compared with 5 and 8 mo. (Supplementary Figure 1C). Because we found increased ROCK content and activity in the pgWAT and ingWAT of female rats, we investigated whether an acute physical exercise session can modulate these proteins (Figure 3B). Female rats at 15 mo submitted to acute physical exercise showed lower pMYPT^T696 in the pgWAT (Supplementary Figure 1D) and decreased ROCK1, ROCK2, and pMYPT^T696 in the ingWAT (Figures 3B–C). This modulation of ROCK1/2 content and ROCK activity was associated with higher levels of the Nfkb1 gene, and a trend to increase Tnf, Il1b, and Ikk genes in the ingWAT (Figure 3D). On the other hand, acute physical exercise was efficient to improve this inflammatory signature in 15 mo. female rats (Figure 3D). In sequence, we analyzed samples obtained from the scWAT of adult female humans from 20 to 40 years, which were not submitted to acute exercise protocol, and samples obtained from older female humans from 60 to 70 years before and after acute exercise protocol (Figure 3E). As observed in human samples, ROCK activity (pMYPT^T696) was increased in the scWAT of old individuals but attenuated after acute exercise (Figure 3F), which was also related to an inflammatory state in the scWAT observed by changes in Tnf, Il1b, Il6, and Mcp1 genes (Figure 3G).

ROCK Silencing Improves Cell Senescence Phenotype in Human Preadipocytes

Because we observed an association between the ROCK pathway and the NFκB gene, and considering that NFκB is 1 of the major transcription factors involved in the senescence-associated secretory phenotype (SASP) (32), we induced cell senescence in murine and human preadipocytes using etoposide to understand whether ROCK plays a role in the development of cell senescence. Interestingly, ROCK1, ROCK2, and pMYPT^T696 were upregulated in a dose-dependent manner, accompanied by the enhancement of NFκB p65 portion both in the murine and human preadipocytes (Figure 4A–C). After that, we induced cell senescence with etoposide in human preadipocytes with previous ROCK1 or ROCK2 silencing (Figure 4D). It was observed a lower etoposide-induced Senescence β-Galactosidase Staining (SA-β-Gal) in the preadipocytes that have silenced ROCK1 and ROCK2 compared with etoposide control (siNT) cells (Figure 4E). The cell’s phenotype was accompanied not only by a downregulation in the p21 gene in the siROCK2 cells but also by decreased SASP-related genes (Cxcl1, Cxcl2, Il6) in both siROCK1 and siROCK2 treated cells, decreased Ccl2 only in the siROCK1 cells, and decreased Cxcl10 only in the siROCK2 cells compared with etoposide treated cells (Figure 4F). It is also important to note that in the silencing of ROCK2, but not ROCK1 in human preadipocytes without etoposide stimulation, the NFκB p65 protein was reduced by about 50% compared with control cells (Figure 4D). Thus, the protective effect also observed on ROCK1 silenced cells could be related to other pathways distinct from NFκB transcriptional activity.

Bioinformatic Analysis Suggests Downstream Signaling Pathways Correlated With ROCK Modulation

Following the experimental data obtained in animals and cell culture experiments, we took advantage of public datasets from the Genotype-Tissue Expression (GTEx) project and from a BXD mouse genetic reference population (33) to look for the top5000 genes that are correlated with ROCK1 or ROCK2 gene in the subcutaneous adipose tissue (scWAT). Firstly, when the human samples were distributed by different age ranges (Supplementary Figure 2A–B), we observed increased ROCK1 mRNA in females at 50 and 60 years and males at 60 years compared with 20 years group (Supplementary Figure 2B and E). In contrast, ROCK2 mRNA was attenuated in the scWAT of males at 50 and 60 years, compared with 20 years old (Supplementary Figure 2F). From these human dataset and from BXD mouse strains (Supplementary Figure 2G–I), we analyzed the top 5 000 correlated genes separately with ROCK1 and with ROCK2 in the scWAT (Figure 5A). As observed, besides the majority of these correlated genes being unique for the species, there were 326 genes positively correlated and 139 negatively correlated with ROCK1 that overlapped in both human and mouse samples (Figure 5B). These genes were submitted to the Reactome pathways analysis that revealed pathways related to the “SUMOylation” process, “Cell Cycle”-related pathways, and “Rho GTPases signaling” in the positively correlated genes (Figure 5C). For the negatively correlated genes, it was found “The Citric Acid Cycle and Respiratory Electron Transport,” “Mitochondrial translation”-related pathways, “Translation,” and “Metabolism” (Figure 5D). Part of the genes that compose these pathways were highlighted in Pearson’s correlations plots (Figure 5E–H).

When we looked at ROCK2 correlated genes in the scWAT of mouse and humans, 351 and 145 genes with a positive and negative correlation, respectively, overlapped (Figure 5I). Reactome pathways related to “Membrane Trafficking,” “Signaling by Rho GTPases,” and “Adaptive Immune System” were presented in the list of positively correlated genes (Figure 5J, L–M). On the other hand, the genes with a negative correlation were integrated into pathways associated with “Immune System,” “Cytokine Signaling in Immune System,” “Non-canonical NF-KB signaling,” and “Signaling by Interleukins” (Figure 5K, N–O). Therefore, besides the fact that ROCK proteins can play a role in adipocyte differentiation, adipogenesis, and insulin sensitivity, we also bring new possible pathways that could be directly involved with ROCK1 or ROCK2 modulation in the white adipose tissue.

Discussion

ROCK is a downstream effector of small GTPase RhoA and is involved with cytoskeleton organization, contraction, cell proliferation, and other fundamental cellular functions (34). In addition, several data in the literature showed an important participation of Rho-kinase in whole-body metabolism (19). In our study, we found that even showing a different pattern of protein distribution at 5 mo. in the adipose tissue of both sexes, male and female rats at 15 mo. showed higher ROCK activity in the ingWAT and pgWAT compared with 5 mo. controls, which was associated with higher levels of some inflammatory markers. Importantly, a single bout of exercise was sufficient to reduce ROCK activity in 15 mo. female rats, and also prevented age-related increases in ROCK activity and inflammation in the scWAT of human subjects. Here, using human and mouse preadipocytes we also verified augmented ROCK activity in response to etoposide (a cell-senescence inducer) together with NFκB p65 protein, an important transcriptional driver for senescence-associated secretory phenotype (SASP). However, when ROCK1 and ROCK2 were separately silenced, there was an improvement in the SA-β-Gal staining and a decrease mainly in SASP-related genes after etoposide treatment. To complement these obtained data, we used public datasets (GTEx and BXD families) to highlight a variety of pathways that have a strong correlation with ROCK1/2 modulation in the scWAT of mouse and humans. Then, looking at the fact that ROCK seems to play a detrimental role in adipose tissue metabolism, we observe an important effect of acute exercise in mediating this ROCK activation, that could be associated with downstream signals mediated by this kinase.

The rationale to characterize ROCK1 and ROCK2 in different adipose depots was based on the fact that previous studies have shown a role of this kinase in the regulation of adipogenesis, as well as inflammatory response and insulin resistance. However, most of these studies did not look at the effect on development/aging, but were more focused on the obesity and energy expenditure context. For example, diet-induced or genetic (ob/ob and db/db) obese mice have increased ROCK activity in WAT compared with controls (25,26). However, partial deletion of ROCK2 protects against body weight gain in response to aging (12 mo.) or a high-fat diet in male mice due to its effect on white adipose tissue being (20,23). Wei et al. demonstrated that ROCK2 inhibition initiated a thermogenic program in the adipocytes and improved the whole-body energy expenditure contributing to insulin sensitivity (20). In addition, actomyosin contraction in response to increased ROCK activity was efficient in suppressing peroxisome proliferator-activated receptor gamma (PPARγ) and CCAAT Enhancer Binding Protein Alpha (C/EBPα) activity, leading to impaired beige adipogenesis (20).

The adaptations mentioned earlier are also observed by Soliman et al. which showed increased PPARγ in the adipose tissue of ROCK2^+/− mice under a high-fat diet, which was accompanied by lower inflammatory markers and preserved insulin signaling (23). Consistent with this, Hara et al. generated an adipose-specific mouse model with a dominant-negative form of the ROCK agonist RhoA (DN-RhoA transgenic mice), which had decreased ROCK activity in adipose tissue resulting in protection from the harmful effects of a high-fat diet (body weight gain, glucose intolerance, and inflammation) (25). They proposed that adipocyte hypertrophy in overnutrition increases ROCK activity in the adipocytes, leading to the recruitment of inflammatory cells and consequently causing insulin resistance. This vicious cycle related to ROCK activation, adipocyte hypertrophy, and insulin resistance, could be directly involved in promoting cardiometabolic risk. Regarding the aging effect, we previously reported an increase in the ROCK activity in the pgWAT of middle-aged male Wistar rats but not in the pgWAT of male middle-aged Fischer rats, which are leaner (35). This result could be related to the increased fat mass in the middle-aged Wistar rats, and to the fact that ROCK activity is also stimulated by fat accumulation as commented earlier.

Regarding ROCK mechanisms of action, it is important to highlight that KD025 (ROCK2-specific inhibitor), but not other inhibitors (Y-27632, fasudil, and H-1152P), was able to suppress adipogenesis and improve insulin signaling in adipocytes (22,24), corroborating previous findings (20,23,25,26). The KD025 treatment in adipocytes provided beneficial effects as an antiadipogenic factor occurred independently of ROCK activity or cellular actin cytoskeleton modifications. The specific downstream proteins responsible for the KD025 effects in the adipocytes are unknown, but PPARγ and C/EBPα mediated part of these effects. In addition, other proteins controlling ROCK activity may also play a role in the adipocyte metabolism and insulin resistance state. For example, Dankel et al. proposed that RhoE, the ROCK inhibitor protein, could be involved in the stimulation of lipolysis in adipocytes (21). However, because RhoE is a ROCK inhibitor, it is unclear why both RhoE and ROCK mRNA increase in the adipocytes of ob/ob mice (21). The increased RND3-mediated lipolysis in adipocytes may contribute to an inflammatory condition and metabolism alterations.

It is clear that ROCK inhibition can alter adipocyte metabolism and differentiation. However, it is also worth noting that this can also be achieved by natural strategies like physical exercise. Several studies describe the positive effects of exercise in the improvement of metabolic parameters, attributing it to attenuation of inflammation (36,37), improved insulin sensitivity (38), enhanced energy expenditure (39,40), and several other adaptations that oppose the harmful effects of obesity and aging (27,41). Here, we used an acute exercise protocol to avoid significant alterations in whole-body morphology and to investigate if ROCK activity could be rapidly attenuated in response to a single exercise session. We have previously shown that ROCK activity is responsive to physical exercise in aging and obesity conditions in the skeletal muscle (28,30,31). Notably, ROCK was positively associated with improved glucose uptake in skeletal muscle and was increased by physical exercise. In the present study, we show that ROCK activity is attenuated in adipose tissue in response to acute exercise. This could be due to downstream signaling mediated by ROCK which could be differentially regulated in adipose tissue versus skeletal muscle. This fact also emphasizes that physical exercise mediates metabolic processes in different organs via different mechanisms, ultimately leading to final outcomes that improve whole-body metabolism. Besides the fact that we used an acute exercise protocol, we believe that repeated training sessions can suppress ROCK activity, maintaining the appropriate adaptation of adipose tissue to long-term physical exercise and keeping the adipocytes healthy.

Cellular senescence is a state of irreversible growth arrest where cells lose their ability to divide, but also to perform basic cellular functions, and is associated with pathological conditions (42). On the other hand, senolytic drugs have emerged as a clinical potential strategy to attenuate chronic diseases (43). Another important data found here in this study was the fact that silencing ROCK1 and ROCK2 resulted in a protective effect against cell-senescence and attenuated SASP-related gene expression. Until now, there has been little evidence of an association of ROCK and cell senescence, and this is the first study to demonstrate the function of ROCK on cell senescence in preadipocytes (44,45). Thus, our study provides support for the use of exercise as a senolytic therapeutic (46) whose mechanism could be mediated in part by the downregulation of ROCK activity in preadipocytes.

In conclusion, development, aging, and sex were important determinants of ROCK expression in adipose tissue in rodents. Further, ROCK was stimulated by cell senescence induction in human preadipocytes, and etoposide-mediated induction of cell senescence and SASP in preadipocytes were blunted by ROCK1/2 silencing. Thus, an increase in ROCK activity may be promoting downstream signaling pathways that are driving impaired adipocyte biology, inflammation, and perturbations in whole-body energy homeostasis. On the other hand, acute exercise attenuated ROCK activity in white adipose tissue both in rats and humans, and counteracted cell senescence in adipocyte depots. Thus, our findings implicate ROCK as an important regulator of adipose tissue senescence and metabolism and provide novel insights into its role in the regulation of whole-body metabolism.

Funding

None.

Conflict of Interest

None.

Data Availability

The data supporting the current findings are available from the corresponding author upon reasonable request.

Acknowledgments

We would like to thank Dr. C. Ronald Kahn from Joslin Diabetes Center, Harvard Medical School, for the support with laboratory structure and supplies during the review process.

This work was supported by FAEPEX, the National Council for Scientific and Technological Development (CNPq; Case numbers 303571/2018-7, 140285/2016-4, 442542/2014-3, and 306535/2017-3), the Coordination for the Improvement of Higher Education Personnel (CAPES; finance code 001), and the São Paulo Research Foundation (FAPESP; Case numbers 2015/26000-2, 2016/18488-8, 2018/20872-6, 2019/11820-5, 2020/13443-1, and 2021/08692-5).

Author Contributions

V.R.M. and J.R.P. wrote the paper and were ultimately responsible for the experiments in this study. V.R.M., R.F.L.V., C.K.K., A.P.A.M., and R.C.G., designed and performed experiments with animals. V.R.M and M.L. performed cell culture experiments. V.R.M. performed the bioinformatic analysis. J.C.S.T., R.G.D., M.P.T.C.M., and C.R.C. designed and performed experiments with humans. A.S.R.S., E.R.R., D.E.C., M.P.T.C.M., C.R.C., and J.R.P. contributed to the discussion and laboratory support. All the authors have read and approved this manuscript.

References