Mechano-signaling via Piezo1 prevents activation and p53-mediated senescence of muscle stem cells

Skeletal muscle stem cells (MuSCs), also called satellite cells, are instrumental for postnatal muscle growth and skeletal muscle regeneration. Numerous signaling cascades regulate the fate of MuSCs during muscle regeneration but the molecular mechanism by which MuSCs sense mechanical stimuli remain unclear. Here, we describe that Piezo1, a mechanosensitive ion channel, keeps MuSCs in a quiescent state and prevents senescence. Absence of Piezo1 induces precocious activation of MuSCs, attenuates proliferation, and impairs differentiation, essentially abolishing efficient skeletal muscle regeneration and replenishment of the MuSC pool. Furthermore, we discovered that inactivation of Piezo1 results in compensatory up-regulation of T-type voltage-gated Ca2+ channels, leading to increased Ca2+ influx, which strongly induces NOX4 expression via cPKC. Elevated NOX4 expression in Piezo1-deficient MuSCs increases ROS levels and DNA damage, causing P53-dependent cellular senescence and cell death. The importance of the P53/P21-axis for mediating Piezo1-dependent cellular defects was confirmed by pharmacological inhibition of P53 in Piezo1-deficient mice, which abrogates increased senescence of muscle cells and normalizes muscle regeneration. Our findings uncover an essential role of Piezo1-mediated mechano-signaling in MuSCs for maintaining quiescence and preventing senescence. Reduced mechano-signaling due to decreased physical activity during aging may contribute to the increase of senescent cells and the decline of MuSC numbers in geriatric mice and humans.


Introduction
Skeletal muscle regeneration requires MuSCs, a population of adult stem cells capable of self-renewal and myogenic differentiation. MuSCs are mostly quiescence in adult skeletal muscles under homeostatic conditions and reside between the sarcolemma and basal lamina of myofibers [1]. Quiescent MuSCs in skeletal muscles are marked by expression of PAX7 and undergo activation in response to injury or other stimuli such as intense physical exercise [2][3][4][5]. MuSCs rapid upregulate myogenic transcription factors (e.g. MYF5, MYOD) during activation and differentiate following proliferation by fusing to each other or to existing myofibers [6]. A subpopulation of activated MuSCs escapes differentiation and returns to quiescence, allowing repopulation of the muscle stem cell niche [7][8][9]. Major efforts have been made to understand the mechanism determining MuSC quiescence and activation. Numerous signal pathways were identified that play critical roles in this process, including Notch, mTOR, and P38/MAPK [9][10][11][12]. However, it is still not fully understood how MuSCs sense dynamic changes in the microenvironment and how the decision is made whether to stay in a quiescent state or to initiate activation. Changes within the MuSC niche are considered to affect MuSCs and to direct fate decisions.
Damaged, unwanted, or superfluous MuSCs may submit to programmed cells death or undergo senescence [13]. Cellular senescence is characterized by permanent cessation of cell division, accompanied by distinct phenotypic changes including metabolic dysfunction, chromatin remodeling, increased autophagy, and a pro-inflammatory senescence-associated secretory phenotype (SASP) [14,15]. Acquisition of senescence plays a physiological role during normal development, but enhanced accumulation of senescent cells contributes to age-related pathological processes [16][17][18]. Cellular senescence can be initiated by numerous processes including telomere erosion, DNA damage, mitochondrial dysfunction, and accumulation of oxygen reactive species (ROS) [15,19,20]. Senescence-induced cell cycle arrest is driven by two tumor suppressor pathways, P16/RB and P53/P21. Although increased presence of P16 is usually considered as a hallmark of senescence, some cell types are more responsive to P53/P21 compared to P16/RB for initiation of senescence, depending on the cellular context [19,21]. During aging, MuSCs undergo enhanced rates of cellular senescence, which has been associated with lowered muscle performance and decreased muscle regeneration [22,23] Accumulation of senescent MuSCs is promoted by age-related structural and molecular changes of the MuSC niche, which impede functions of MuSCs [24]. Reduced exercise-dependent mechanical load during the course of aging may contribute to changes of the MuSC niche. However, it is currently not known whether and how MuSCs directly sense exercise-induced mechanical load and whether constant sensing of mechanical signals is indispensable for MuSCs to avoid cellular senescence. Piezo1 is mechanosensitive cation channel, which opens when membrane tension increases [25], a process that is instrumental for responding to mechanical load in various cell types [26][27][28][29]. Numerous in vivo studies demonstrated the critical mechanosensitive function of Piezo1 in various physiological processes, including somatosensation, hearing, bone formation, adipogenesis, vascular development, and the control of blood pressure and heart functions, [25,[28][29][30][31][32][33]. Piezo1 is considered to have irreplaceable roles in tissue homeostasis but relatively little is known about its function in skeletal muscles. Previous in vitro studies, using primary myoblasts and the C2C12 myoblast cell line, unraveled a function of Piezo1 in muscle cell differentiation, while in vivo studies with a focus on MuSCs are lacking so far [34][35][36].
Here, we demonstrate a critical role of Piezo1 for reducing ROS formation and preventing P53-dependent senescence of MuSC. We show that Piezo1 is required to keep MuSCs in a quiescent state but also for myoblast proliferation and differentiation, which makes Piezo1 indispensable for normal skeletal muscle regeneration. In a more general sense, the study uncovers how mechanical signal transduction reduces cellular senescence in a population of somatic stem cells and emphasizes the importance of mechanical cues for skeletal muscle physiology.

Piezo1 is preferentially expressed in quiescent MuSC
To analyze whether MuSCs may sense mechanical load via Piezo1, we performed immunofluorescence staining for Piezo1 on sections of tibialis anterior (TA) muscles (Fig. 1A). Co-staining for Piezo1 and PAX7, a MuSC marker, revealed that most PAX7 + cells express Piezo1 while the sarcolemma of myofibers did not show obvious Piezo1 staining, indicating that expression of Piezo1 within the muscle lineage is confined to MuSCs ( Fig. 1A and B). To confirm this finding, we analyzed freshly isolated single myofibers. Piezo1 expression was detected in PAX7 + cells immediately (0h) after myofiber isolation (Fig. 1C) and to a lower degree in activated, MYOD-expressing MuSCs after 24h of cultivation (Fig. 1C). Next, we determined mRNA levels of Piezo1 by RT-qPCR in freshly isolated MuSC, purified by fluorescence activated cell sorting (FACS), and in MuSCs after cultivation for 24h. Piezo1 was highly expressed in freshly isolated MuSCs but expression declined substantially during the culture period, concomitant with increased expression of the MuSC activation marker MYOD (Fig. 1D). Taken together the results indicate that Piezo1 is specifically expressed in quiescent MuSCs and declines during activation and differentiation.

Piezo1 is required for maintaining quiescence of MuSCs and to avoid depletion of the MuSC pool
The decline of Piezo1 expression in activated compared to quiescent MuSCs suggested that Piezo1 might secure MuSC quiescence, which is required for long-term maintenance and self-renewal of the MuSC pool, allowing efficient muscle regeneration at later stages of life. To analyze this possibility, we treated explanted myofibers carrying attached MuSCs and freshly isolated MuSCs with Yoda1, a specific activator of Piezo1. Treatment with Yoda1 attenuated activation of MuSCs as indicated by reduced ratios of PAX7 + MYOD + /PAX7 + cells in both cultured myofibers and MuSCs (Fig. S1), suggesting that activation of Piezo1 maintains quiescence of MuSCs.
To further test this hypothesis, we inactivated Piezo1 in PAX7 + cells in vivo by breeding Pax7-Cre (Pax7 ICN ) mice with a strain in which the Piezo1 gene is flanked by loxP sites (Piezo1 f/f ) to generate Pax ICN /Pie-zo1 f/f mice (hereafter named Piezo1 mKO ). RT-qPCR analysis of MuSCs isolated from Piezo1 mKO mice confirmed efficient inactivation of Piezo1, reducing expression of Pizeo1 by approximately 94% in FACS-purified, freshly isolated Piezo1 mKO MuSCs compared to control Piezo1 f/f or Pax7 ICN MuSCs (Fig. S2A). Importantly, FACS analysis of hind limb muscles revealed a marked reduction of the number of Piezo1 mKO MuSCs compared to control hind limb muscles (Fig. S2B). Consistent with this observation, we detected an approximately 80% reduction of the numbers of PAX7 + MuSCs in TA muscles of Piezo1 mKO (Fig. S2C) suggesting that failure of mechanosensing due to inactivation of Piezo1 during muscle development depletes the MuSC pool, probably due to precocious activation of MuSCs or the failure to acquire a quiescent state. Only 60% of the remaining PAX7 + MuSCs in Piezo1 mKO TA muscles expressed CALCR, a marker for MuSC quiescence, while virtually all PAX7 + MuSCs showed CALCR expression under sedentary conditions in control muscles (Fig. S2D) To exclude developmental defects caused by deletion of Piezo1, we also inactivated Piezo1 in adult MuSCs using a tamoxifen (TAM)inducible Pax7-Cre allele (Pax7 CreERT2 ). Efficient inactivation of Piezo1 was accomplished by 10 daily injections of TAM into Pax7 CreERT2 /Pie-zo1 f/f mice (hereafter named Piezo1 scKO ), which lowered Piezo1 expression to nearly undetectable levels compared to control mice (Piezo1 f/f ) ( Fig. 2A and B). FACS analysis unveiled a reduction of MuSCs by approximately 50% in Piezo1 scKO compared to control mice, which corresponded to a 60% reduction of PAX7 + MuSCs on sections of TA muscles ( Fig. 2C and D). In addition, we observed increased activation of the remaining MuSCs in Piezo1 scKO muscles as indicated by a 40% reduction of CALCR + PAX7 + relative to all PAX7 + cells while the number of MYOD-expressing PAX7 + MuSCs in Piezo1 scKO muscles increased to approximately 25% ( Fig. 2E and F). Along the same line, analysis of cultured single myofibers immediately after plating (0h) identified MYOD + PAX7 + cells on fibers from Piezo1 scKO but not from Piezo1 f/f control mice (Fig. 2G). Although MYOD + PAX7 + cells emerged on fibers from control mice after cultivation for 24h, the number was substantially higher on fibers from Piezo1 scKO mice (Fig. 2G). MYOD + PAX7 + cells were also more abundant in MuSCs isolated from Piezo1 scKO compared to Piezo1 f/f mice, albeit the difference was less pronounced than in the myofiber cultures (Fig. 2H). Taken together, these results indicate that Piezo1 is instrumental for keeping MuSCs in quiescence and for maintaining the MuSC pool in skeletal muscles.

Piezo1 supports proliferation and differentiation of isolated MuSCs in vitro
Although expression of Piezo1 declines in activated and proliferating MuSCs, we wanted to know whether Piezo1 does also play a role for proliferation and differentiation of MuSC, either due to precocious activation of MuSCs or due to a function during proliferation. Piezo1 was inactivated in adult MuSCs by 10-times injection of TAM, which was verified by immunofluorescence staining of isolated PAX7 + MuSCs with a Piezo1 antibody. We detected a substantial decrease of PAX7 + MuSCs in regenerating TA muscles of Piezo1 scKO compared to Piezo1 f/f control mice, 5 and 7 days following cardiotoxin (CTX) injection to induce muscle injury ( Fig. 3A-C). Similar results were obtained when assessing the proliferation rate of MuSCs in vitro. MuSCs isolated from the hind limbs of adult control and TAM-induced Piezo1 scKO mice were cultured in growth medium for 4 days and then pulsed with the thymidine analog EdU for 3 h (Fig. 3D). EdU incorporation in Piezo1 scKO MuSCs was reduced by approximately 50% compared to Piezo1 f/f control MuSCs ( Fig. 3E and F), which corresponded to a reduced proliferation rate monitored by time-lapse imaging (Fig. 3G). We also observed a substantially decline of the fusion index and percentage of MYOG positive nuclei in Piezo1 scKO MuSCs after placement into differentiation medium, indicating compromised differentiation ( Fig. 3H-K).
Since reduction of PAX7 + MuSCs in regenerating TA muscles of Piezo1 scKO and attenuated proliferation of Piezo1 scKO MuSCs may be caused by precocious activation of MuSCs in vivo, we turned to a different experimental system. We isolated MuSCs by FACS from adult Piezo1 f/f muscles and then treated cultured MuSCs with Adenovirus-CRE (Ad-Cre) to induce inactivation of Piezo1 in vitro (Fig. S4A). RT-qPCR analysis revealed successful inactivation of Piezo1 in cultured MuSCs after 48h exposure to Ad-Cre but not Ad-Null (Fig. S4B). Inactivation of Piezo1 in vitro essentially yielded the same results as observed in vivo and when culturing isolated MuSCs that had lost Piezo1 expression in vivo. We found a substantial reduction of EdU incorporation and a severe reduction of the proliferation rate in adult These results clearly indicate that Piezo1 does not only suppress activation of MuSCs but also plays an important role at later stages to enable proliferation and differentiation of MuSC.

Expression of Piezo1 in MuSCs is necessary for efficient skeletal muscle regeneration
The precocious activation of in Piezo1 scKO MuSC, reduction of MuSC numbers, as well as their reduced proliferation and differentiation strongly suggested that absence of Piezo1 compromises skeletal muscle regeneration. To address this hypothesis, we inactivated Piezo1 in adult MuSCs by 10-times injections of TAM and subjected control and Pie-zo1 scKO TA muscles to cardiotoxin (CTX) injury and subsequent analysis at 7 and 14 days (Fig. 4A). Piezo1 scKO TA muscles showed a strong reduction of TA muscle weight at 7 and 14 days after CTX injury, while CTX-injected Piezo1 f/f control and non-injured Piezo1 scKO TA muscles regained essentially the same mass as non-injured Piezo1 f/f control TA muscles ( Fig. 4B and C).
Histological analysis at day 7 after injury revealed reduced formation of regenerated centronuclear fibers in Piezo1 scKO mice compared to Piezo1 f/f control mice ( Fig. 4D and E). The relative number of eMYHC + (Myh3) regenerating myofibers was higher in Piezo1 scKO mice than in control mice, indicating delayed formation of myofibers (Fig. 4F). The average fiber cross-sectional area (CSA) area was still reduced by approximately 50% in Piezo1 scKO compared to control mice 14 days after CTX-induced injury ( Fig. 4G and H; Fig. S5A). Most regenerated fibers were unusually small 7 and 14 days after injury in Piezo1 scKO mice (<200 μm 2 ) and large fibers were missing (Fig. 4I, Fig. S5B).
To characterize the regenerative capacity of Piezo1 scKO mice after repetitive injury, we subjected TA muscles to 3 rounds of CTX-induced muscle regeneration (Fig. 4J). Repetitive muscle injury nearly abolished regeneration in Piezo1 scKO mice with very few newly forming centronuclear myofibers and a severe reduction of muscle mass 14 days after the last injury (  TA muscles did not show any significant morphological differences during the course of the experiment without CTX-induced muscle injury (Figs. S5E and F). The very strong muscle regeneration defect was associated with a nearly complete loss of PAX7 + MuSCs in Piezo1 scKO mice ( Fig. 4O; Fig. S5G), probably caused by compromised self-renewal and the inability of Piezo1 scKO MuSCs to re-acquire quiescence. These data unequivocally establish a major role of Piezo1 for regeneration of skeletal muscles.

Loss of Piezo1 enhances ROS and P53 accumulation and leads to increased senescence and cell death of MuSCs
To study the molecular mechanism enabling Piezo1-dependent regulation of MuSC quiescence, skeletal muscle regeneration, and MuSC maintenance, we subjected FACS-isolated MuSCs from Piezo1 f/f and Piezo1 scKO mice to RNAseq analysis (Fig. 5A). We identified 1734 differently expressed genes (DEGs) between Piezo1 f/f and Piezo1 scKO (caption on next page) MuSC. 1208 genes were upregulated in Piezo1 scKO MuSC, while 526 were downregulated based on absolute fold-changes ≥2 (Benjamini-Hochberg-corrected p-value ≤ 0.05) (Fig. 5B). Gene set enrichment analysis (GSEA) revealed differential expression of genes related to cell senescence in Piezo1 scKO MuSC, including DNA damage response, increase of ROS production, lysosome activity, and telomerase dysfunction ( Fig. 5C; Fig. S6). Furthermore, Metacore and Ingenuity Pathway Analysis (IPA) indicated that changes in the activity of enriched transcriptional regulators can be traced back to P53 (Fig. 5D). These data suggested that P53 may play a key role in mediating effects in the regulatory network downstream of Piezo1. To validate this assumption, we performed western blot analysis, which revealed a strong accumulation of P53 in Piezo1 scKO MuSCs (Fig. 5E). Furthermore, we detected increased expression of P21, a canonical target of P53, in undamaged and regenerating Piezo1 scKO muscles (Fig. 5F). Increased production of ROS promotes accumulation of P53 in numerous cell types [37,38], which sparked the idea that the enhanced presence of P53 in Piezo1 scKO MuSCs is due to increased ROS production. Measurement of ROS confirmed increased ROS production in Piezo1 scKO MuSCs ( Fig. 5 G), while treatment of MuSCs with the Piezo1-activator Yoda1 had no effects on ROS-levels in WT MuSCs (Fig. S7A). Furthermore, we detected a strong increase of oxidative DNA damage as indicated by accumulation of 8-oxoG (Fig. S7B) and increased numbers of MuSCs with the DNA-damage marker γH2AX in Piezo1 scKO muscles, both in undamaged and in regenerating conditions (Fig. S7C). To analyze whether increased ROS production due to loss of Piezo1 is indeed responsible for accumulation of P53, we treated Piezo1 scKO MuSCs with the ROS scavenger NAC. We observed that scavenging of ROS normalized P53 to control levels (Figs. S7D and E), indicating that enhanced ROS production is responsible for P53 accumulation in Piezo1 scKO MuSC.
DNA damage and accumulation of P53 often results in apoptosis and cellular senescence [39]. To explore whether loss of Piezo1 in MuSCs and the resulting increase in ROS and P53 levels increase the number of apoptotic and senescent MuSCs, we stained for senescence-associated β-galactosidase (SA-β-Gal), a commonly accepted marker for senescent cells [40]. The number of SA-β-Gal + MuSCs was dramatical higher in cultured Piezo1 scKO MuSCs compared to controls ( Fig. 5H and I). Treatment of cultured WT MuSCs with the Piezo1 channel activator Yoda1 did not reduce the number of senescent cells, probably because senescence is very low at baseline conditions (Fig. S8). We also found a strong increase of senescent MuSCs in muscle tissue sections from Piezo1 scKO compared to Piezo1 f/f control mice ( Fig. 5J; Fig. S9A), which was even more evident in regenerating TA muscles 7 days after injury ( Fig. 5K;  Fig. S9B), suggesting that mechanical cues sensed by Piezo1 suppress cellular senescence of MuSC. Moreover, we observed a strong increase of apoptosis in cultured MuSC, in which the floxed Piezo1 alleles were deleted by Ad-Cre infection, based on TUNEL-and cleaved Casp3-staining (Fig. S10).

Inactivation of Piezo1 increases expression of T-Type Ca 2+ channels, elevates Ca 2+ influx, and stimulates ROS production in MuSCs
The increase of ROS levels in Piezo1 scKO MuSCs was surprising, since formation of ROS is usually enhanced by increased but not by decreased Ca 2+ levels [41]. In Piezo1 scKO MuSCs we expected to the see reduced Ca 2+ levels and thus reduced ROS production, because Piezo1 allows Ca 2+ influx in response to different types of external forces [42]. Therefore, we decided to have a closer look at Ca 2+ influx and molecules involved in this process. GSEA of Piezo1 scKO MuSCs revealed a strong enrichment of genes responsible for Ca 2+ import, including voltage-gated Ca 2+ channels (Fig. 6A). Further analysis of RNAseq data uncovered increased expression of the voltage-dependent T-Type Ca 2+ channel genes Cacna1g and Cacna1h in freshly isolated Piezo1 scKO compared to Piezo1 f/f MuSCs (Fig. 6B), which was validated by RT-PCR experiments (Fig. 6C). In addition, we observed a strong elevation of the corresponding CaV3.1 (Cacna1g) and CaV3.2 (Cacna1h) channels in MuSCs, in which the floxed Piezo1 alleles were deleted by Ad-Cre infection (Fig. 6D), strongly suggesting that loss of the Piezo1 Ca 2+ channel leads to a compensatory increase of T-Type Ca 2+ channels. To investigate whether the strong increase of CaV3.1 and CaV3.2 also increases intracellular Ca 2+ levels, we performed Ca 2+ influx measurements in cultured Piezo1 scKO and Piezo1 f/f MuSCs using the Ca 2+ -sensitive Fura-4-AM dye [29]. We detected a clear increase of Ca 2+ influx in Piezo1 scKO compared to Piezo1 f/f MuSCs (Fig. 6E), which was normalized by addition of ML218, a specific inhibitor of T-Type Ca 2+ channels [43] (Fig. 6F).
Next, we investigated whether the increased Ca 2+ influx is indeed responsible for enhanced ROS production in Piezo1 scKO MuSCs. To this end, we measured ROS levels in MuSCs, in which the floxed Piezo1 alleles were deleted by Ad-Cre infection, either in the absence or presence of the Ca 2+ chelator BAPTA-AM. Administration of BAPTA-AM reduced ROS in Piezo1 scKO MuSCs back to control levels while no decrease was observed in Piezo1 f/f MuSCs (Fig. 6G). Strikingly, we found the same normalization of ROS levels, when Piezo1 scKO MuSCs were treated with ML218 (Fig. 6H). To further prove the critical role of T-Type Ca 2+ channel-mediated Ca 2+ influx, we measured P53 levels, which are increased in Piezo1 scKO MuSCs due to raised ROS concentrations. We found that administration of ML218, lowered P53 in Piezo1-mutant MuSCs to control levels (Fig. 6I), confirming that the compensatory increase of Ca 2+ influx via T-Type Ca 2+ channels raises ROS levels and P53 expression. Weights of TA muscles after CTX-induced muscle injury. All the values are relative to Piezo1 f/f + saline. TA muscle weights were normalized to body weights (one way-ANOVA-test: **p < 0.01; ns, p > 0.05, n = 3, 2-months-old male mice, 3 sections per sample). (D) H&E staining of muscle sections from injured TA muscles of Piezo1 f/f and Piezo1 scKO mice 7 days after CTX injection. (E) Numbers of centralized myofibers (C-nuclear fibers) in TA muscles of Piezo1 scKO mice compared to Piezo1 f/f mice (t-test: **p < 0.01, n = 3, 2-months-old male mice, 3 sections per sample). (F) Immunofluorescence staining for embryonic myosin heavy chain, laminin-2 (green) and DAPI (blue) on TA muscle sections from Piezo1 f/f and Piezo1 scKO mice 7 days after CTX injection. Quantifications are on the right (t-test: **p < 0.01, n = 3, 2-months-old male mice, 3 sections per sample).

Increased Ca 2+ influx via T-Type Ca 2+ channels after inactivation of Piezo1 activates cPKC, thereby stimulating NOX4 expression and ROS generation
Although increased Ca 2+ concentrations are known to enhance the activity of ROS-generating enzymes, the strong increase in ROS levels in in Piezo1 scKO MuSCs suggested an additional mechanism. Ca 2+ binds to the C2-domain of conventional PKCs (cPKCs) to induce translocation of cPKCs to the plasma membrane, which enables binding of DAG and phosphatidylserines, classical activators of cPKCs [44]. GSEA of Pie-zo1 scKO MuSCs disclosed enhanced activity of the PKC signaling pathway (Fig. 7A), confirmed by higher enzyme activity of PKC in cultured MuSC, in which the floxed Piezo1 alleles were deleted by Ad-Cre infection (Fig. 7B). Normalization of PKC enzyme activity in Piezo1-deficient MuSCs by the T-Type Ca 2+ channel blocker ML218 confirmed that the increased PKC activity was induced by elevated Ca 2+ influx mediated via CaV3.1 and CaV3.2 (Fig. 7C). Since PKC was recently reported to activate expression of the ROS-generating enzyme NOX4 in kidney cells [45], we investigated the expression of different NOX enzymes in MuSC. We found that increased PKC activity in Piezo1-deficient MuSCs correlated with increased gene expression of NOX4 but not of NOX1, NOX2 and NOX3 (Fig. 7D). Furthermore, western blot analysis revealed strongly elevated levels of NOX4 in MuSCs after inactivation of Piezo1 (Fig. 7E), which was completely normalized by the T-Type Ca 2+ channel blocker ML218 (Fig. 7F). To learn whether activation of cPKC is indeed instrumental for elevated ROS levels in Piezo1 scKO MuSCs, we inhibited cPKC activity pharmacologically. Addition of Go 6983, a pan-PKC inhibitor, strongly reduced NOX4 expression (Fig. 7G) and also normalized ROS in cultured Piezo1-deficient MuSCs to control levels (Fig. 7H) [46]. We also found that inhibition of NOX4 with GLX351322 reduced ROS in cultured Piezo1-deficient MuSCs to control levels, abolished increased P53 expression in Piezo1-deficient MuSCs (Fig. 7I and J) and prevented cellular senescence ( Fig. 7K and L). Taken together, the data indicate that increased Ca 2+ influx via T-Type Ca 2+ channels in Pie-zo1-deficient MuSCs activates cPKC, which stimulates NOX expression, raises ROS and P53 levels, and induces cellular senescence.

Pharmacological inhibition of P53 reduces formation of senescent MuSCs and improves muscle regeneration in Piezo1 scKO mice
The strong accumulation of P53 and P21 in Piezo1 scKO MuSCs suggested a causal role in the formation of senescent MuSCs, which might be the underlying reason for defective skeletal muscle regeneration. To explore this hypothesis, we treated Piezo1 scKO and Piezo1 f/f control mice during muscle regeneration with the P53 inhibitor Pifithrin-α (PFT-α) (Fig. 8A). Immunofluorescence staining 7 days after injection of CTX showed a strong reduction of the numbers of P21 + MuSCs in Piezo1 scKO mice treated with PFT-α, indicating successful inhibition of P53 activity (Fig. 8B). Likewise, we observed a strong decline of γH2AX + MuSCs in Piezo1 scKO mice after PFT-α treatment, while no significant changes were observed in Piezo1 f/f control mice with or without PFT-α treatment (Fig. S11). Furthermore, we observed that the PFT-α treatment not only reduced the number of P21 + and γH2AX + MuSCs but also prevented loss of PAX7 + MuSCs in Piezo1 scKO muscles (Fig. 8C, Fig. S11), indicating that loss of MuSCs in Piezo1 scKO muscles is initiated by aberrant activation of P53.
Next, we investigated whether inhibition of increased P53 activity in Piezo1 scKO MuSCs prevents accumulation of senescent MuSCs and restores muscle regeneration. Strikingly, PFT-α treatment strongly reduced the number of SA-β-Gal + cells in Piezo1 scKO muscles, essentially matching the numbers found in Piezo1 f/f control muscles (Fig. 8D). Similar results were obtained when isolated MuSCs from Piezo1 f/f and Piezo1 scKO mice were treated for 5 days with PFT-α (Fig. 8D). Furthermore, treatment with PFT-α also reestablished muscle regeneration in Piezo1 scKO mice. We detected a major increase of TA muscle mass in PFTα-treated versus untreated Piezo1 scKO mice 7 days after injury (Fig. 8E).
The number of newly formed centronuclear myofibers increased dramatically as well, essentially eradicating the presence of unrepaired areas in damaged muscles of Piezo1 scKO mice (Fig. 8F). The increase of newly formed centronuclear myofibers went along with a decrease of smaller fibers and an increase of larger fibers as well as disappearance of eMYHC-positive myofibers, all indicative of improved skeletal muscle regeneration (Fig. 8G-J). Taken together, we concluded that Piezo1mediated suppression of ROS production and P53 accumulation is critical to prevent formation of senescent MuSC, thus maintaining a healthy MuSC pool allowing efficient skeletal muscle regeneration.

Discussion
MuSCs are situated on contractile myofibers and therefore directly exposed to mechanical signals, resulting from muscle tonus and muscle contractions. Several studies have investigated the impact of mechanical signaling on MuSCs [47,48], but knowledge about the mechanisms by which MuSCs sense mechanical cues is very limited. In this study, we uncovered that the mechanosensitive ion channel Piezo1 is highly expressed in MuSCs, prevents precocious activation of MuSC, and is required for regular MuSC proliferation and differentiation. Furthermore, we demonstrate that loss of Piezo1 increases ROS formation and promotes accumulation of P53, resulting in massive induction of senescence, depletion of the MuSC pool, and eventually in severe defects of skeletal muscle regeneration. Since MuSCs undergo enhanced senescence during aging when physical activity and muscle tonus decline, we speculate that mechanosensing by Piezo1 plays an important role to maintain skeletal muscle homeostasis and to prevent senescence of MuSCs in geriatric muscles.
Quiescence of MuSCs is regulated by numerous signaling pathways, which exert their functions within the MuSC niche, in which the myofiber acts as the primary signal-providing hub. Ligands provided by muscle fibers include delta-like 1 (DLL1), WNT4, Oncostatin M (OSM), and many more (see Ref. [49] for a recent review). In addition, other cells within in the MuSC niche, such as endothelial cells and fibroadipogenic precursor cells (FAPs), support MuSCs to maintain quiescence. Quiescence was initially assumed to reflect a default state of cellular inactivity, but mounting evidence indicates that MuSCs require active input to regulate quiescence, which is pivotal to guarantee stemness of MuSCs and to maintain regenerative functions. It is surprising that mechanosensing via the Piezo1 channel upholds quiescence instead of activating MuSC, since accretion of MuSCs by myofibers is required for sustained hypertrophic growth of myofibers following enhanced mechanical load, despite some controversy about the role of MuSCs for skeletal muscle hypertrophy [50,51]. Apparently, muscles have developed mechanisms to distinguish between strong mechanical load, resulting in activation of MuSCs and hypertrophic growth, and low-level tonic signals, preventing activation of MuSCs. Our study uncovers that active Piezo1 is crucial for arresting MuSCs in quiescence. The biological meaning of mechanosensing for keeping MuSCs in quiescence is currently unclear but it is tempting to speculate that a decline of low-level tonic signaling due to diminished force generation by myofibers awakens MuSCs, allowing renewal of myofibers to regain strength. Similarly, continuously reduced mechanosensing in geriatric muscles with lowered muscle tonus may result in permanent activation of muscle stem cells in an attempt to restore muscle fiber strength but eventually causing depletion of the MuSC pool. The increased propensity of MuSCs in geriatric muscle for activation and differentiation as well as the reduction of MuSC numbers during aging supports such a hypothesis [22].
MuSCs undergo enhanced senescence in muscles of geriatric mice, which is associated with impaired muscle regeneration after injury [23]. Senescence-induced impairment of muscle regeneration during aging or in progeria models has been explained by persistent P38 MAPK activity, downregulation of Slug, a member of the Slug/Snail superfamily of zinc-finger transcriptional factor, downregulation of the cell surface protein CDON [52], reduced mitophagy, TGF-β-induced SMAD3 activation, and over-activation of Notch [53,54], all promoting stem cell exhaustion by increased senescence [22,23,55]. The current study extends this list and establishes Piezo1-mediated mechanosensing as an important suppressor of MuSC senescence in skeletal muscles. However, we currently do not know whether expression of Piezo1 or Piezo1-mediated mechanosensing declines during aging, although the lowered muscle tonus in aged muscles makes such a scenario likely. Loss of Piezo1 in our experimental model has some similarities to P53-dependent senescence in activated Numb-deficient MuSCs during muscle regeneration, resulting in regeneration defects of skeletal muscles [56]. In both cases, MuSC senescence relies on P53 activation but in contrast to Numb-deficient MuSC, the absence of Piezo1 instigates MuSC senescence under homeostatic conditions without prior injury. Increased senescence of MuSCs alone may be sufficient to explain depletion of the MuSC pool and regeneration defects but the loss of  quiescence will certainly contribute. Two different scenarios seem possible: (i) Loss of Piezo1 directly initiates senescence of MuSCs via increased ROS-production and P53 activation, which leads to loss of MuSC. (ii) Loss of Piezo1 causes aberrant activation of MuSC, which subsequently increases ROS-production and P53 activation. Single cell RNA-seq of MuSCs after in situ fixation to avoid artifacts resulting from activation of MuSCs during the isolation procedure may help to answer the question whether senescence-specific gene are already expressed in MuSCs of Piezo1 scKO mice prior to isolation [57].
Our study uncovered the basic mechanisms by which inactivation of Piezo1 increases ROS formation and activates P53, although some questions still remain open. We found that inactivation of the Piezo1 Ca 2+ channel caused up-regulation of the T-Type Ca 2+ channel genes Cacna1g and Cacna1h, most likely by a compensatory mechanism. As net result, Ca 2+ influx into Piezo1-deficient MuSCs was higher, which may activate ROS-generating enzymes and formation of free radicals [41]. More importantly, we also detected a Ca 2+ -dependent increase of cPKC activity, which strongly induced expression of the ROS-generating enzyme NOX4, confirming previous results in kidney cells [45]. Since inhibition of NOX4 essentially normalized ROS levels, we conclude that induction of NOX4 is the main reason for elevated ROS production in Piezo1-deficient MuSC. Changes in Ca 2+ concentrations are associated with senescence of skeletal muscle fibers, further supporting a link between aberrant Ca 2+ handling and senescence but more detailed studies are necessary to understand the precise mechanisms [58].
We found that inhibition of ROS prevents increase of P53 and that inhibition of p53 attenuates senescence of MuSC, partially rescuing the skeletal muscle regeneration defect in Piezo1 scKO mice. These observations clearly suggest that ROS-mediated activation of P53 plays a pivotal role for triggering senescence in Piezo1 scKO MuSC. ROS is well known to activate P53 in different conditions, although it has been debated whether activation of P53 by ROS is solely facilitated through DNA damage response, redox signaling, or both [59]. We observed enhanced levels of the DNA damaged markers 8-oxoG and γH2AX in Piezo1 scKO MuSC, which argues for ROS-induced DNA damage as a critical trigger of P53 activation but we do not want to exclude other ROS-dependent pathways. Moreover, it was reported recently that Piezo1 directly binds to P53, and that knockdown of Piezo1 upregulates expression of P53 in human esophageal carcinoma cell lines, suggesting mutual interaction based on protein-protein interactions [60]. Since we observed that scavenging of ROS prevents increase of P53, we do not believe that such a mechanism is relevant for increased P53 activity and increased senescence in Piezo1 scKO MuSC.
In conclusion, we demonstrate that Piezo1-mediated mechanosensing is essential for maintaining quiescence of MuSCs and for preventing P53-dependent senescence. Our findings underscore the importance of mechanical signal transduction in skeletal muscle homeostasis and suggest tight interactions between mechanical activities, senescence, and age-related decline of skeletal muscle functions. We assume therapeutic targeting of Piezo1 is an interesting option to prevent age-related loss of MuSCs and improve muscle functions at advanced age.

Animal models
The Piezo1 f/f (Piezo1 loxp/loxp ) and Pax7 ICN mouse strains have been described previously [29,61]. The Pax7 CreERT2 mouse strain was obtained from The Jackson Laboratory [62]. Primers used for genotyping are shown in Supplementary Table 1

Pifithrin-α treatment of mice
Pifithrin-α (Pifithrin-α hydrobromide) (1267, Tocris), a pan-P53 inhibitor was used to inhibit P53 activities in vivo [63]. Intraperitoneal injections of Pifithrin-α were performed at a concentration of 2 mg/kg of mice body weight together with tamoxifen treatment for 10 consecutive days, prior to CTX injection. Intraperitoneal administration of Pifithrin-α (2.2 mg per 1 kg body weight) was continued every second day after induction of muscle injury until isolation of the samples as indicated.

MuSCs and myofiber cultures
MuSCs were purified via FACS and cultured in growth medium 1 (DMEM GlutaMAX containing 20% FCS 1%, Penicillin/Streptomycin, and 5 ng/mL basic fibroblast growth factor (bFGF)) [6]. The medium was changed at every second day. MuSC differentiation was induced by changing growth to differentiation medium (DMEM GlutaMAX containing 2% horse serum and 1% Penicillin/Streptomycin), after 6 days culturing in growth medium. Myofibers were isolated by enzymatic digestion of isolated FBD muscle with collagenase P (0.02%, Roche), and cultured in growth medium 2 (DMEM medium with 20% fetal calf serum (FCS) and bFGF (5 ng/mL)) [64]. MuSCs or myofibers isolated from WT mice were cultured in the absence or presence of 1.5 μM, 3 μM or 10 μM Yoda1 (SML1558, Sigma) for 8h (Myofibers) or 48h (MuSCs) and then fixed with 4% paraformaldehyde (PFA) for further analysis. FACS purified MuSCs from Piez1 fl/fl mice were cultured in growth medium and infected with Cre recombinase containing Ad-Cre virus (Vector biolabs) for at least 48h to induce Cre-mediated deletion of Piezo1 in MuSCs in vitro. MuSCs treated with adenoviruses that do not express Cre (Ad-Null) served as controls.

ROS measurements
To detect total cellular ROS levels, MuSCs isolated from Piezo1 scKO and Piezo1 f/f mice were cultured in growth medium. After five days of culture, cellular ROS levels were measured using a Total ROS detection kit (51011, ENZO) following the manufacturer's instructions. Signals were recorded using a fluorescence plate reader (LB 940, Mithras).

In vitro EdU incorporation assay
Growth medium was supplemented by addition of EdU to a final concentration of 10 μM, 3 h before fixation of cultured MuSCs and then analyzed using the Click-iT EdU kit (Invitrogen) according to the manufacturer's protocol.

TUNEL assay
Cells were fixed in 4% PFA for 10 min and then subjected to the TUNEL reaction using the TUNEL Assay Apoptosis Detection Kit (C10246, Invitrogen), following the manufacturer's instructions.

Protein kinase C activity assay
PKC activity in MuSCs was measured with a PKC Kinase Activity Assay Kit (ab139437, Abcam). Cells were washed once with PBS and lysed with lysis buffer containing 50 mM KH2PO4, 1.5 mM MgCl 2 , 10 mM NaCl, and 1 mM EDTA (pH 7.4) on ice. Cells were collected into Eppendorf tubes and centrifuged (18,000×g, 15 min, 4 • C). Supernatants were collected and diluted 1:20 using kinase assay dilution buffer provided in the kit. Diluted samples were pipetted to the 96-well plate included in the kit, and processed according to the manufacturer's instructions. Protein concentrations of samples were determined with BCA to normalize PKC activities to the protein concentrations [65].

Myoblast fusion index and measurement of myotube length
Myotubes were identified by positive MF-20 staining. The fusion index was calculated by dividing the number of nuclei incorporated in myotubes by the total number of nuclei observed in each images. Statistical assessment is based on 60 randomly chosen images for each group.
Myotube length was measured using ImageJ. Five representative measurements of myotube length were collected per viewing field. 50 randomly chosen viewing fields were analyzed for each group. For comparisons, the weights of dissected muscles from individual mice were normalized to the respective body weights.

Calcium measurements
Changes in intracellular Ca 2+ levels were determined by the fluorescent Ca 2+ -sensitive dye Fluo-4 AM (F14201, Invitrogen), following published protocols [66]. MuSCs were incubated for 30 min s at 37 • C with 5 μM Fluo-4 AM in Ca 2+ and Mg 2+ free HBSS-HEPES buffer, after which cells were washed 3 times with HBSS-HEPES buffer. Time-lapse images were acquired at every 1.5s with a Leica SP8 MP Microscope using an immersion 40× objective. Excitation wavelength was set to 488 nm, and emission was recorded at 535 ± 15 nm. Analysis and processing of image sequences was done using the ImageJ/FIJI software.

SA-β-galactosidase detection assay
SA-β-Galactosidase staining was performed as described previously [56]. Cells and sections were fixed for 4 min at RT in a solution of PBS, 1% PFA, 0.2% glutaraldehyde. Samples were washed twice in PBS at pH 7 for 10 min and incubated for 30 min in PBS (pH 6), followed by incubation in X-gal staining solution (4 mM K3Fe(CN)6, 4 mM K4Fe(CN)6, 2 mM MgCl 2 , 0.02% NP-40 and 400 μg/ mL X-gal (15520-018, Sigma) in PBS pH6) at 37 • C. Staining was allowed to proceed overnight for cultured cells and for 48h with a change of the staining solution when sections were used. Samples were washed in PBS, and post-fixed in 1% PFA 5 min for cells and 30 min for sections. After 3 times washing for 10 min with PBS, samples were mounted in PBS, 20% glycerol or processed for immunochemistry.

RNA sequencing
RNAseq was performed as previously described [64]. Samples were generated in biological triplicate and used for RNaseq. Total RNA was isolated using commercial kits and according to the manufacturer's protocols (RNAeasy Mini kit, QIAGEN). RNA integrity of RNA and library preparations were assessed using the LabChip Gx Touch 24 (Per-kinElmer). 300 ng of total RNA was used as input for VAHTS Stranded mRNA-seq. Library preparations were done following manufacture's protocol (Vazyme). Sequencing was performed on a NextSeq500 instrument (Illumina) using v2 chemistry, resulting in an average of 30 M reads per library with 1x75bp single end reads. Raw reads were assessed for quality, adaptor content and duplication rates with FastQC 0.10.1, trimmed by Reaper version 13-100 69 and terminally aligned to the Ensemble mouse genome version mm10 (GRCm38) by STAR 2.4.0a. The number of reads aligning to genes was counted with the featureCounts 1.4.5-p1 tool from the Subread package. Only reads mapping at least partially inside exons were admitted and aggregated per gene. Reads overlapping multiple genes or aligning to multiple regions were excluded. Differentially expressed genes were identified using DESeq2 version 1.62 l. Only genes with an absolute fold change of ≥2, a Benjamini-Hochberg corrected p-value ≤ 0.05, and a minimum combined mean of 5 reads were assumed to be significantly differentially expressed. Ingenuity Pathway Analysis was performed using QIAGEN's Ingenuity Pathway Analysis algorithm (www.qiagen.com/ingenuity, QIAGEN, Redwood City, CA, USA) for genes with minimum fold changes of ±2 and a Benjamini-Hochberg corrected p-value ≤ 0.05. Gene set enrichment analysis was performed to characterize differentially expressed genes (DEGs). Gene sets were extracted from MSigDB. Heatmaps of enriched leading-edge gene clusters were generated using the 'pheatmap' and 'circheatmap' packages of R. Volcano plots of DEGs between Piezo1 scKO MuSCs and controls were generated using the 'ggplot2' package of R. Gene expression data were deposited in the NCBI GEO database under GSE190933 (https://www.ncbi.nlm.nih.gov/geo/ query/acc.cgi?acc=GSE190933).

RT-qPCR assays
RT-qPCR analysis were performed using TaqMan gene expression or SYBR Green-based RT-qPCR assays. Total RNA of samples was isolated using commercial kits, following the manufacturer's protocols (RNAeasy Mini kit, QIAGEN). cDNA was generated with the Prime-Script™ II 1st strand cDNA Synthesis Kit (Takara). TaqMan probe-based RT-PCR was performed with the TaqMan Fast Advanced Master Mix (Thermo Fisher). Sequences of TaqMan probes for detection of Piezo1, MyoD1, Pax7, Calcr genes are listed in Supplementary Table 1. SYBR Green-based RT-qPCR was performed with the Applied Biosystems™ PowerUp™ SYBR™ Green Master Mix (Thermo Fisher). Sequences of primers to detect expression of Cacna1g, Cacna1h, Nox1, Nox2, Nox3, Nox4 are listed in Supplementary Table 1. Relative expression levels of genes of interest were determined using the comparative threshold method, using Gapdh as internal control. Data were analyzed using the ΔΔCt method [64].

Histological analysis
The muscles were dissected at different time points as indicated in each figure, and snap frozen in isopentane cooled with liquid nitrogen.
Sections (8 μm) of muscle samples were prepared with a cryomicrotome and processed for Hematoxylin & Eosin staining as described previously [6].

Statistics
All experiments were performed in biological triplicates unless stated otherwise. Sample sizes for studies was chosen based on observed effect sizes and standard errors. For statistical analysis we either used one-way ANOVA when multiple groups were compared or unpaired two-tailed Student's t-test when two groups were compared. Student's t-test was only used when the variances of the two groups did not differ by factor >3. P values < 0.05 were considered statistically significant (*p < 0.05, **p < 0.01). Values are depicted as ± s.e.m. GraphPad Prism 9 was used for data analysis. Calcium transients were calculated as dF/F = (Ft-F)/F, where F is baseline fluorescence intensity and Ft is the intensity of the fluorescence detected in Piezo1 scKO or Piezo1 scKO MuSC. dF/F of 11 ROI (regions of interest) were calculated for each image.

Declaration of interests
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Author contribution
YP, AS and TB conceived and designed experiments. YP and JD performed most of the experiments, analyzed the data, and prepared figures. SG performed next generation RNA sequencing. YD performed bio-informatics analysis. SW provided transgenic mouse lines and advice. LZ supervised animal experiments, TB, AS, YD and JD wrote the manuscript.