RhoA-mediated G₁₂-G₁₃ signaling maintains muscle stem cell quiescence and prevents stem cell loss

Abstract

Multiple processes control quiescence of muscle stem cells (MuSCs), which is instrumental to guarantee long-term replenishment of the stem cell pool. Here, we describe that the G-proteins G₁₂-G₁₃ integrate signals from different G-protein-coupled receptors (GPCRs) to control MuSC quiescence via activation of RhoA. Comprehensive screening of GPCR ligands identified two MuSC-niche-derived factors, endothelin-3 (ET-3) and neurotensin (NT), which activate G₁₂-G₁₃ signaling in MuSCs. Stimulation with ET-3 or NT prevented MuSC activation, whereas pharmacological inhibition of ET-3 or NT attenuated MuSC quiescence. Inactivation of Gna12-Gna13 or Rhoa but not of Gnaq-Gna11 completely abrogated MuSC quiescence, which depleted the MuSC pool and was associated with accelerated sarcopenia during aging. Expression of constitutively active RhoA prevented exit from quiescence in Gna12-Gna13 mutant MuSCs, inhibiting cell cycle entry and differentiation via Rock and formins without affecting Rac1-dependent MuSC projections, a hallmark of quiescent MuSCs. The study uncovers a critical role of G₁₂-G₁₃ and RhoA signaling for active regulation of MuSC quiescence.

Introduction

Skeletal muscle stem cells (MuSCs) are crucial for postnatal muscle growth, long-term maintenance of muscle mass, and muscle regeneration. Sustained survival of MuSCs and replenishment of the stem cell pool critically depend on the acquisition of a resting, quiescent state. Loss of MuSC quiescence during aging impairs self-renewal and reduces the number of MuSCs¹. Compromised quiescence and self-renewal along with increased senescence of MuSCs in geriatric mice has been associated with decreased muscle mass, although direct evidence that loss of MuSC quiescence enhances sarcopenia is missing so far². The situation is further complicated by findings that nearly complete ablation of MuSC does not cause sarcopenia under base line conditions^3,4.

Quiescent MuSCs are characterized by cell cycle withdrawal, translational suppression, epigenetic inactivation of myogenic regulatory factors (e.g., Myf5, MyoD), and increased expression of Pax7^2,5,6,7. Numerous mechanisms have been described which secure quiescence of MuSCs, ranging from diffusible extracellular cues, signals from the MuSC niche, and transcription factor circuits to epigenetic mechanisms². For example, muscle fiber-derived Wnt4 suppresses MuSC activation through RhoA and increased supply of Wnt4 arrests MuSCs in a deep state of quiescence, delaying muscle repair⁸. Similarly, collagen V (COLV), secreted by MuSCs, is a surrogate ligand for the calcitonin receptor (CalcR) and constitutes an important component of the quiescent niche, as demonstrated by aberrant cell cycle entry of MuSCs after inactivation of COLV⁹. The sheer number of processes regulating quiescence is staggering, raising questions how integration of signals is achieved to control the eventual outcome. Coordination of parallel signaling processes is critical, probably requiring intracellular integration for orchestrating the regulatory network that controls the quiescence state of MuSC¹⁰.

Extracellular signals are often relayed into cells via G-protein-coupled receptors (GPCRs), which belong to a large superfamily of proteins with seven transmembrane domains. GPCRs regulate such diverse processes as growth, proliferation, differentiation, secretion, and metabolism. Canonical GPCR signaling is mediated by different heterotrimeric G-proteins (Gα, Gβ, and Gγ), which are activated by GPCRs, facilitating exchange of GDP by GTP on the Gα subunit upon ligand binding. In addition, unconventional GPCRs exist, such as the WNT receptors, which belong to the Class Frizzled of GPCRs. Frizzled receptors mainly signal via the WNT/β-catenin, WNT/calcium, or planar cell polarity (PCP) pathways, the latter leading to activation of RhoA and ROCK¹¹. Activated GPCRs regulate different subfamilies of heterotrimeric G-proteins, including G_s, G_i-G_o, G_q-G₁₁, and G₁₂-G₁₃, which activate different intracellular signaling pathways¹². Members of the G_s and G_i-G_o family antagonistically regulate adenylyl cyclases and thereby intracellular cAMP concentrations. The prototypical marker of quiescent MuSCs, CalcR, is coupled to G_s and signals via adenylyl cyclase and cAMP^9,13. However, inactivation of Calcr has only moderate effects on MuSC quiescence¹³. Activation of G_q-G₁₁ by GPCRs primarily increases PLC-β activity, leading to PKC and Ca²⁺ signaling. G_q-G₁₁-coupled receptors do not appear to discriminate between G_q and G₁₁ and no differences in the ability to regulate phospholipase C-isoforms were observed. The G-proteins G₁₂-G₁₃ are often activated by receptors that also couple to G_q-G₁₁, but initiate various signaling pathways, including RhoA-dependent processes. Genetic studies indicated a strong functional overlap between G_q and G₁₁ as well as between G₁₂ and G₁₃, requiring the use of Gnaq-Gna11 and Gna12-Gna13 compound mutant mice for functional studies¹².

Several GPCRs such as CalcR and the adhesion GPCR GPR116, also known as AGDRF5, which increases the nuclear localization of β-arrestin, thereby enabling interactions with cAMP response element-binding protein, are known to regulate MuSC quiescence^13,14. However, a systematic analysis of the role of canonical GPCRs in the regulation of MuSCs is missing. Likewise, the potential function of G_q-G₁₁ and G₁₂-G₁₃ for integrating activities of GPCRs in MuSCs or the role of signaling pathways downstream of these G-proteins have not been analyzed so far. In a combination of hypothesis-driven and bias-free approaches, we comprehensively assessed expression of GPCRs in quiescent MuSCs and determined the function of individual GPCRs for preventing activation of MuSCs using a pharmacological screening approach. Two niche-derived GPCR ligands were identified, endothelin-3 (ET-3) and neurotensin (NT), which activate the G₁₂-G₁₃ pathway. Genetic inactivation of Gnaq-Gna11 and Gna12-Gna13 as well as expression of constitutively active RhoA in MuSCs revealed a fundamental role of G₁₂-G₁₃ in integrating signals from different GPCRs to control RhoA activity and MuSC quiescence. Our findings unravel a central role of G₁₂-G₁₃-RhoA signaling in MuSCs and provide new opportunities for pharmacological manipulation of quiescence.

Results

Functional characterization of different GPCRs in quiescent MuSCs

To systematically determine changes in the expression pattern of GPCRs during the transition of MuSC from quiescence to activation, we analyzed five published bulk RNA sequencing (RNA-seq) datasets^{15,16,17,18,19}. We identified 39 GPCRs, which showed higher expression in quiescent compared to activated MuSCs in at least two datasets (Supplementary Fig. S1a). Subsequent RT-qPCR analysis revealed substantially higher expression of 19 out of these 39 GPCRs in quiescent compared to activated MuSCs (foldchange > 2) (Supplementary Fig. S1b). Immunofluorescence staining of MuSCs localized on freshly isolated single myofibers confirmed the presence of seven GPCRs in the plasma membrane of quiescent MuSC (Supplementary Fig. S1c, d). We corroborated the enrichment of CalcR and S1P₃ in quiescent MuSCs^13,14,20, but also detected GPCRs that have not been described in quiescent MuSC before, such as ET_B and NTS₂ (Supplementary Fig. S1c, d). Localization of the remaining GPCRs could not be validated in the plasma membrane, since suitable antibodies were not available.

Expression of a GPCR does not necessarily prove a physiologically relevant function. Thus, we combined the expression analysis with an unbiased pharmacological screen²¹, utilizing a customized GPCR compound library consisting of 259 compounds, targeting 24 subfamilies of GPCRs, including both natural and chemical agonists and antagonists (Supplementary Fig. S2a). We reasoned that manipulation of GPCRs or ligand‒GPCR pairs, functionally important for inducing or maintaining quiescence, should alter activation of MuSC in vitro. To assess MuSC activation, we performed EdU-incorporation assays and quantified the ratio of quiescent PAX7⁺MYOD^‒ to all PAX7⁺ cells. (Fig. 1a). Cut-offs for potential hits were defined by activities of Oncostatin M (OSM) (Fig. 1b, c), which induces quiescence of MuSCs¹⁸, and FGF2 and IGF1 (Fig. 1b, c), which induce proliferation and myogenic differentiation of MuSCs, respectively^22,23. We identified 15 GPCR subfamilies, whose manipulation changed quiescence or activation of MuSCs, including adrenergic receptors (ARs), Cannabinoid receptors (CB1 and CB2), metabotropic glutamate receptors (mGluRs), and Thyroid-stimulating hormone receptor (TSHR) (Supplementary Fig. S2b). Furthermore, we found that ET-3 and NT prevented expansion and myogenic commitment of MuSC at low dosages (Fig. 1d). Similar results were obtained when MuSCs attached to isolated single myofibers from mouse flexor digitorum brevis (FDB) muscles were treated with ET-3 or NT (Supplementary Fig. S2c, d). The strong activity of ET-3 and NT is consistent with the presence of the corresponding receptors ET_B and NTS₂ in the plasma membrane of quiescent MuSCs (Supplementary Fig. S1c).

Fig. 1: GPCR compound screening identifies ET-3 and NT as regulators of MuSC quiescence.

a Schematic representation of the GPCR compound screening strategy. b, c Scatter plots of EdU⁺ (b) and PAX7⁺MYOD^‒ (c) MuSC ratios following 5-day treatment with 259 synthetic and natural GPCR compounds (10 μM, GPCR antagonists (red) and agonists (blue)). Green dashed lines: threshold for positive selection (mouse OSM treatment); red dashed lines: negative selection (FGF2 or IGF1 treatment); black lines: DMSO-treated MuSCs. d Inhibition of proliferation (upper panel) and myogenic activation (lower panel) of MuSCs by ET-3 and NT (1 nM–1 µM). e, f Immunofluorescence (e) and quantification of GFP⁺ nuclei and Dystrophin⁺ myofibers (f) in transverse sections of Dmd^mdx-4Cv/Y TA muscles, engrafted with freshly isolated MuSC (FSC) or cultured in the presence of DMSO, ET-3, and NT for 5 days, after 3 consecutive CTX-induced injuries. The cartoon in e represents the schematic outline of MuSC engraftment (n = 3). g Immunofluorescence for CalcR (red), ET-3/NT (green), and DAPI (blue) on transverse sections of TA muscles. h Western blot analysis for ET-3/NT in uninjured (U) TA muscles and injured TA muscles at D2, D7, and D14 after CTX injury. Quantification is shown in the lower panel (n = 3). The data represent means ± SEM, analyzed by one-way ANOVA with Bonferroni’s multiple comparisons test (f, h). Scale bars: 20 µm in e, 5 µm in g.

We next tested whether ET-3 or NT is able to retain stem cell properties of MuSCs, which are normally lost during expansion in vitro, reducing the ability of expanded MuSC to contribute to skeletal muscle regeneration in a mouse model of Duchenne muscular dystrophy (Dmd^mdx-4Cv/Y mice)^24,25. MuSCs were isolated from Pax7nGFP^+/+ mice, specifically labeling nuclei of PAX7⁺ cells with GFP, and cultured for 5 days with and without the presence of ET-3 or NT. Afterward, cultured MuSCs were grafted into the tibialis anterior (TA) muscle of 18 Gy-irradiated hindlimbs of Dmd^mdx-4Cv/Y mice, shortly after the first three consecutive cardiotoxin (CTX) injections (Fig. 1e). We also used freshly isolated MuSCs, which typically show a higher regenerative capacity than cultured cells^7,26. Self-renewal and regenerative capacity of transplanted MuSCs were examined by counting the number of GFP-positive nuclei and dystrophin-positive myofibers after completion of regeneration. Treatment with ET-3 or NT enabled transplanted MuSCs to massively increase the numbers of dystrophin-positive myofibers and GFP-positive nuclei within the host tissue compared to untreated controls, although the efficiency of freshly isolated MuSCs was not fully reached (Fig. 1e, f). The results suggested that NT or ET-3 treatment promotes engraftment of cultured MuSC as indicated by higher numbers of Pax7-GFP⁺ cells and also improves the contribution of transplanted MuSC to regenerating muscle fibers in mdx muscles.

ET-3 and NT are niche-derived factors, preventing premature activation of MuSCs in vivo

To investigate the origin of ET-3 and NT from cells in the MuSC niche, we first consulted the “scmuscle” database (scRNAseq.org), a recently constructed omics database that provides information about transcript levels in single cells of skeletal muscles under various conditions²⁷. According to the scmuscle database, expression of Edn3 (the gene encoding for ET-3) is highest in quiescent MuSCs compared to other cells in the skeletal muscle, implying an autoregulatory loop to keep MuSCs in quiescence (Supplementary Fig. S2e). In contrast, Nts (the gene encoding for NT) is mainly found in endothelial cells (Supplementary Fig. S2e), consistent with prior findings²⁸. RT-qPCR and immunofluorescence analyses confirmed that Edn3 is expressed in quiescent but not in activated MuSCs, whereas Nts is expressed in lymphatic and blood endothelial cells, directly adjacent to quiescent MuSCs (Fig. 1g; Supplementary Fig. S2f‒i). Specificity of antibodies against NT was verified by western blot analysis of lymph endothelial cells after siRNA-mediated knockdown of NTS (Supplementary Fig. S2j, k). Both ET-3 and NT are strongly downregulated but gradually restored towards the final stages of muscle regeneration (Fig. 1h). Interestingly, we also observed a substantial decline of Edn3, Ednrb, and Ntsr2 expression in aged (24-month-old) MuSCs (Supplementary Fig. S2l, n). The expression of ET-3 and NT in the MuSC niche as well as their decline following muscle injury, suggest a role of ET-3 and NT in regulating stem cell quiescence. Similarly, the reduced expression of Edn3, Ednrb, and Ntsr2 during aging may contribute to the reduced quiescence of aged MuSCs.

Consistent with the results of our initial GPCR expression profiling, subsequent RT-qPCR and immunofluorescence analyses revealed high expression of the receptors for ET-3 and NT, ET_B and NTS₂ in quiescent but not in activated MuSCs (Fig. 2a‒c). We also found that treatment of cultured MuSC with ET-3 and NT increases expression of Ednrb, and Ntsr2, respectively, indicating the existence of a positive feedback loop, which secures enhanced Ednrb and Ntsr2 expression in MuSCs when regeneration is completed (Supplementary Fig. S2o, p). The second receptor for NT, NTS₁, is neither expressed in quiescent nor in activated MuSCs (Fig. 2a‒c). To study the roles of ET_B and NTS₂ in the regulation of MuSC quiescence, we utilized selective antagonists for ET_B (BQ788) and NTS₂ (SR142948A). Treatment of MuSCs attached to isolated single FDB myofibers with BQ788 or SR142948A did not change proliferation of MuSCs or the ratio of PAX7⁺MYOD⁺ to all PAX7⁺ MuSCs, probably due to low concentration of ET-3 and NT in the experimental set-up (Fig. 2d, e). However, treatment with BQ788 after prior administration of ET-3, abrogated inhibitory effects of ET-3 on MuSC proliferation and MYOD expression (Fig. 2f, g). Similar consequences were observed when SR142948A was added to the growth medium of single FDB myofibers treated with NT (Fig. 2f, g). Combined treatment with ET-3 and NT further increased the ratio of PAX7⁺MYOD^‒ relative to all PAX7⁺ MuSCs compared to single treatments, although no synergistic effects were observed. This might be due to the rather artificial experimental set-up, providing higher concentrations of ET-3 and NT than normally available to MuSCs in vivo (Fig. 2g). To validate the roles of ET_B and NTS₂ for maintaining MuSC quiescence in vivo, we injected BQ788 or SR142948A into TA muscles of WT mice. After 7 days of exposure to either BQ788 or SR142948A, 33% of MuSCs in BQ788-treated mice and 65% in NT-treated mice were located outside the basal lamina in the interstitial space, indicating MuSC activation. Likewise, the ratios of KI67⁺PAX7⁺ and MYOD⁺PAX⁺ to all PAX7⁺ cells increased dramatically, whereas the overall numbers of PAX7⁺ cells did not change after treatment with BQ788 or SR142948A (Fig. 2h‒l). We concluded that activation of ET_B and NTS₂ is essential to keep MuSC arrested in quiescence within the niche.

Fig. 2: Pro-quiescence effects of ET-3 and NT on MuSCs are mediated by ET_B and NTS₂.

a RT-qPCR analysis of Ednrb, Ntsr1, and Ntsr2 expression in freshly isolated MuSCs (FSC) and activated stem cells (ASC) (n = 3). b Representative images of immunostaining for ET_B/NTS₁/NTS₂ (red), PAX7 (green) and DAPI (blue) on FDB myofibers of WT mice at 0 h (upper panel) and 24 h (lower panel) of culturing. c The quantification of GPCR staining intensity (n = 3). d, e Quantification of the ratios of PAX7⁺KI67⁺ (d) and PAX7⁺MYOD^‒ (e) MuSCs on FDB myofibers after 24 h culturing in the presence of DMSO, BQ788 or SR142948A (n = 6). f, g Quantification of the ratio of PAX7⁺KI67⁺ (f) and PAX7⁺MYOD⁻ (g) MuSCs on FDB myofibers after 24 h culturing in the presence of DMSO, ET-3, NT, ET-3 + BQ788, NT + SR142948A, or ET-3 + NT (n = 6). h‒l Representative images (h) and quantification of PAX7⁺ (i), the ratio of PAX7⁺KI67⁺ (j), and PAX7⁺MYOD⁺ (k) cells, and quantification of MuSCs outside the basal limina (l) on transverse sections of WT mice TA muscle after intramuscular injection of solvent, BQ788 (1 mg/kg bodyweight), or SR142948A (0.5 mg/kg bodyweight), respectively (n = 3). The cartoon in h depicts the outline of the experimental design. The data represent means ± SEM, analyzed by unpaired t-test (a, c) and one-way ANOVA with Bonferroni’s multiple comparisons test (d‒g and i‒l). ND not detected. Scale bars: 5 µm in b, 10 µm in h.

G₁₂-G₁₃ integrate signals from different GPCRs to maintain quiescence of MuSCs

Upon ligand binding, GPCRs couple to different G-proteins, enabling them to activate various intracellular signaling pathways. Signals from distinct GPCRs may also converge on the same G-protein subtype, which can integrate signaling events initiated by multiple ligands. To investigate which G-protein is employed by NTS₂ and ET_B to promote MuSC quiescence, we focused on G₁₂-G₁₃ and G_q-G₁₁, since either of these G-proteins has been reported to interact with NTS₂ and ET_B in different physiological processes^29,30. Because the functions of G₁₂ and G₁₃ as well as of G_q and G₁₁ strongly overlap and single inactivation of G₁₂, G₁₃, G_q, and G₁₁ often does not yield clear effects¹², we focused on G₁₂-G₁₃ and G_q-G₁₁ compound mutants. Inactivation of Gna12-Gna13 but not of Gna11-Gnaq by adenoviral transduction of Cre recombinase into isolated MuSCs from Gna11^‒/‒Gnaq^fl/fl and Gna12^‒/‒Gna13 ^fl/fl mice (Fig. 3a), followed by treatment with ET-3 or NT, strongly increased the ratio of EdU⁺PAX7⁺ to all PAX7⁺ MuSCs (Fig. 3b, d). Accordingly, the number of PAX7⁺MYOD^‒ to all PAX7⁺ MuSCs declined only in Cre adenovirus-transduced MuSCs from Gna12^‒/‒Gna13^fl/fl but not from Gna11^‒/‒Gnaq^fl/fl mice (Fig. 3c, e). We also used the same experimental model to investigate potential quiescence-promoting effects of S1P₃, a GPCR that mediates sphingosine-1-phosphate signaling, and BK₁, a GPCR that mediates bradykinin signaling. Both GPCRs are enriched in quiescent compared to activated MuSCs. Treatment of MuSCs on isolated myofibers with CYM5541 and des-Arg⁹-Bradykinin (Des-Arg⁹-BK), agonists for S1P₃ and BK₁, respectively, reduced the ratios of KI67⁺PAX7⁺ and increased the ratios of PAX7⁺MYOD^‒ to all PAX7⁺ cells to a similar extent as ET-3 and NT (Supplementary Fig. S3a, b). Likewise, inactivation of G₁₂-G₁₃ abrogated effects of CYM5541 and Des-Arg⁹-BK on MuSCs (Fig. 3f, g). Since only inactivation of G₁₂-G₁₃ but not of G_q-G₁₁ abrogated the responsiveness of MuSCs to ET-3 and NT (Fig. 3b‒e), we concluded that G₁₂-G₁₃ is mandatory for enabling signaling by ET_B and NTS₂, although it is possible that inactivation of G₁₂-G₁₃ prevents acquisition of quiescence irrespective of specific ligands.

Fig. 3: ET-3 and NT prevent MuSC activation through G₁₂-G₁₃ signaling.

a Schematic experimental outline of b‒g. b, c Ratio of PAX7⁺EdU⁺ (b) and PAX7⁺MYOD^‒ (c) MuSCs isolated from Gnaq^‒/‒Gna11^fl/fl mice infected with Ad-Null or Ad-Cre virus and cultured in the presence of DMSO, ET-3, or NT for 5 days (n = 5). d, e Ratios of PAX7⁺EdU⁺ (d) or PAX7⁺MYOD^‒ (e) MuSCs isolated from Gna12^‒/‒Gna13^fl/fl mice infected with Ad-Null or Ad-Cre virus and cultured in the presence of DMSO, ET-3, or NT for 5 days in vitro (n = 5). f, g Ratio of PAX7⁺EdU⁺ (f) and PAX7⁺MYOD^‒ (g) MuSCs isolated from Gna12^‒/‒Gna13^fl/fl mice infected with Ad-Null or Ad-Cre virus and cultured in the presence of DMSO, CYM5541, or des-Arg⁹-Bradykinin (des-Arg9-BK) for 5 days (n = 5). h, i Response curves to ET-3 for activation of G₁₂(134)-RLuc8/Gβ3γ9-GFP2 (h) and G₁₃(126)-RLuc8/Gβ3γ9-GFP2 (i) BRET biosensors in control (vector) or EDNRB-overexpressing HEK cells (3 independent experiments, 3 replicates per group). j, k Response curves to NT for activation of G₁₂(134)-RLuc8/Gβ3γ9-GFP2 (j) and G₁₃(126)-RLuc8/Gβ3γ9-GFP2 (k) BRET biosensors in control (vector) or NTRS2-overexpressing HEK cells (3 independent experiments, 3 replicates per group). The data represent means ± SEM, analyzed by one-way ANOVA with Bonferroni’s multiple comparisons test (b‒g).

To confirm that activation of ET_B or NTS₂ enhances coupling to G₁₂ or G₁₃, we employed TRUPATH biosensors, which are optimized bioluminescence resonance energy transfer (BRET2) Gαβγ biosensors³¹. Dissociation of the G₁₂ or G₁₃ subunits and the Gβγ heterodimer is monitored by measuring the BRET2 signal upon agonist-induced activation of GPCRs. Reduction of the BRET2 signal indicates receptor-mediated G-protein dissociation. Co-expression of full-length ET_B or NTS₂ together with the BRET2 reporters Gβ3, Gγ9-GFP2, and G₁₂(134)-RLuc8 or G₁₃(126)-RLuc8 in HEK293 cells generated stable BRET2 luminescence signals, which declined in a dose-dependent manner after addition of increasing concentrations of either ET-3 or NT (Fig. 3h‒k). The decline of G₁₃-dependent signals was more pronounced and occurred at lower concentrations of ligands compared to G₁₂-dependent signals but coupling of ET_B and NTS₂ was evident for both G₁₂ and G₁₃. Taken together, the results indicate that binding of either ET-3 to ET_B or NT to NTS₂ directly activates G₁₂-G₁₃ signaling.

G₁₂-G₁₃ signaling is indispensable for maintaining quiescence of MuSCs and preventing depletion of the MuSC pool during aging

To further examine the role of G₁₂-G₁₃ as a signaling hub and integrator of GPCR-dependent signaling for inducing quiescence of MuSC, we generated MuSC-specific conditional compound knock-out mice for Gna12-Gna13 (G12/13^scKO) (Supplementary Fig. S3c, d). Ten days after initiation of Gna12-Gna13 inactivation, we observed a massive increase of MuSCs, of which 80% were localized outside the basal laminar (Fig. 4a, b). In addition, we detected a major increase of PAX7⁺KI67⁺ double-positive cells in the TA muscles of G12/13^scKO mice, indicating increased activation and proliferation of MuSCs (Fig. 4c; Supplementary Fig. S3e). The numbers of PAX7⁺MYOD⁺ and PAX7^‒MYOD⁺ were strongly elevated as well, whereas the number of PAX7⁺MYOD^‒ MuSC declined (Fig. 4d; Supplementary Fig. S3f), indicating reduced self-renewal and enhanced myogenic differentiation. This conclusion was further supported by the presence of numerous MyoG-positive nuclei in G12/13^scKO TA muscles (Fig. 4e, f). Intriguingly, the number of centrally located nuclei surged in G12/13^scKO TA muscles (Fig. 4g, h), associated with increased levels of eMyHC (Fig. 4i), indicating that activation of MuSCs due to depletion of G₁₂-G₁₃ results in fusion of MuSC to adjacent myofibers or formation of new fibers. This hypothesis was confirmed by genetic lineage tracing, revealing that all centrally located myonuclei are derived from Gna12-Gna13-deficient mCherry-labeled MuSCs (Supplementary Fig. S3g, h). Numerous mCherry⁺ nuclei were also detected in the periphery of myofibers, demonstrating continuous addition of Gna12-Gna13-deficient MuSCs and subsequent maturation (Supplementary Fig. S3i). Altogether, these findings demonstrate a critical role of active G₁₂-G₁₃ signaling for maintaining quiescence of MuSCs. To analyze whether the aberrant activation of G12/13^scKO MuSCs compromises skeletal muscle regeneration and prevents return of mutant MuSCs to the stem cell niche, we subjected G12/13^scKO mice to one- and three-times CTX-induced muscle injury. As expected, G12/13^scKO mice showed signs of compromised muscle regeneration 20 days after the injuries, reflected by fiber size heterogeneity and increased numbers of mononuclear cells (Supplementary Fig. S3j). The numbers of PAX7⁺ MuSCs were substantially lower in G12/13^scKO muscles after completion of regeneration. Moreover, only very few PAX7⁺ MuSCs were detected under the basal lamina, indicating reduced self-renewal and the inability of G12/13^scKO MuSCs to return to the stem cell niche (Supplementary Fig. S3k, l).

Fig. 4: Inactivation of G₁₂-G₁₃ abrogates MuSC quiescence, depletes the MuSC pool, and enhances sarcopenia during aging.

a, b Immunofluorescence (a) and quantification of MuSCs outside the basal lamina (b) (n = 3). c Ratios of KI67⁺ MuSCs in TA muscles of control and G12/13^scKO mice (n = 3). d Quantification of PAX7⁺MYOD^‒ (green bars), PAX7⁺MYOD⁺ (gray bars), PAX7⁻MYOD⁺ (ivory bars) cells in TA muscles of control and G12/13^scKO mice (n = 3). e, f Immunofluorescence (e) and quantification of MYOG⁺ cells (f) in TA muscles of control and G12/13^scKO mice (n = 3). g H&E staining of TA muscle sections from control and G12/13^scKO mice. h Quantification of centronuclear myofibers (n = 3). i Western blot analysis of eMyHC and GAPDH in TA muscles of control and G12/13^scKO mice (n = 3). j–l Body weight (j), TA (k) and GAS muscle (l) weights of control and G12/13^scKO mice (n = 3). m Distribution of cross-sectional areas (CSA) of myofibers in TA muscles of control (blue) and G12/13^scKO (yellow) mice (n = 3). n Quantification of myofibers on transverse sections of aged control and G12/13^scKO TA muscles (n = 3). o H&E staining of TA muscle sections from control and G12/13^scKO mice. p Quantification of PAX7⁺ cells in TA muscles of aged control and G12/13^scKO mice (n = 3). q Proliferation curve of young (YO, 2-month-old male mice) and old (OD, 80-week-old male mice) control and G12/13^scKO MuSCs (statistical significance pertains to the last time point, n = 3). The data represent means ± SEM, analyzed by unpaired t-test (b‒d, f, h‒l, n and p) and one-way ANOVA with Bonferroni’s multiple comparisons test (m, q). Male 2-month-old mice were used in a‒i and 80-week-old male mice in j‒q. Scale bars: 10 µm in a, e, 20 µm in g, and 50 µm in o.

In contrast to G12/13^scKO mice, no activation of MuSC was observed in single G12^KO (Gna12^‒/‒) and G13^scKO (Pax7^CreERT2 Gna13^fl/fl) mice, indicating that G₁₂ and G₁₃ serve overlapping functions in MuSCs (Supplementary Fig. S4a‒e). Similarly, inactivation of Gnaq-Gna11 in MuSCs did not increase the numbers of PAX7⁺ or PAX7⁺MyoD⁺ MuSCs, the fraction of PAX7⁺ MuSCs outside the basal lamina, or KI67⁺PAX7⁺ cells (Supplementary Fig. S4a‒e). Apparently, G_q-G₁₁ does not play a major role for initiating or maintaining quiescence of MuSC.

To analyze long-term consequences of G₁₂-G₁₃ depletion and loss of MuSC quiescence, we inactivated Gna12-Gna13 in MuSCs of 2-month-old mice and then allowed the mice to age. At 20 months of age, G12/13^scKO mice showed decreased body weight and muscle mass (Fig. 4j‒l). Aged, 20-month-old G12/13^scKO mice experienced a 26% reduction of TA and a 23% reduction of gastrocnemius (GAS) muscle mass compared to aged control mice (Fig. 4j‒l). Increased loss of muscle mass was associated by disorganized myofiber structures and variations in size, but the numbers of muscle fibers did not decline in statistically significant manner (Fig. 4m‒o).

We also noted changes in fiber-type composition, characterized by a decrease in oxidative type I fibers (Supplementary Fig. S4f, g), different from the consequences of Gna13 inactivation in myofibers, which increases type 1 and 2a oxidative fibers³². Moreover, the number of MuSCs in TA muscles of aged G12/13^scKO dramatically declined (Fig. 4p; Supplementary Fig. S4h), associated with a much lower proliferation rate of aged MuSCs from G12/13^scKO mice compared to MuSCs from control mice (Fig. 4q). The changes in fiber-type composition of G12/13^scKO mice skeletal muscles did not rely on conversion of myofibers into a G12/13^scKO state, which might have been caused by continuous accretion of a G12/13^scKO MuSCs. We only observed a minor reduction of G13 protein in aged muscle fibers of G12/13^scKO mice, even in 80-weeks-old mice (Supplementary Fig. S4i). We also genotyped individual aged muscle fibers from G12/13^scKO mice, reasoning that replacement of existing nuclei from mutant MuSC should result in accumulation of the mutant allele. We clearly detected the genomic fragment derived from the Gna13 mutant allele in individual myofibers, but the wild-type band was much stronger, indicating that only a subset of myonuclei was replaced or added over time (Supplementary Fig. S4j). These results are well in line with previous reports, demonstrating that myonuclei are rather stable without prior muscle injury³³. Even in conditions of severe atrophy, the number of myonuclei remain constant without major replacement³⁴. Taken together, our findings demonstrate that G₁₂-G₁₃ signaling is instrumental for maintaining the MuSC pool and ensuring proper MuSC function during aging. Abrogation of G₁₂-G₁₃ signaling within MuSCs decreases the number of type I myofibers, which is correlated with enhanced skeletal muscle sarcopenia during aging and disrupts normal muscle morphology without loss of G13 protein in aged muscle fibers.

G₁₂-G₁₃ signaling in response to ET-3 and NT requires RhoA for suppressing MuSC activation

G₁₂-G₁₃ activate different intracellular signaling pathways, including Jun kinase (JNK) and cyclooxygenase-2 (COX-2), but the main downstream event is direct regulation of RH-RhoGEFs, which activates Rho GTPases³⁵. To gain insights into the signaling pathways activated by G₁₂-G₁₃, we investigated changes in the transcriptional activity of fluorescence-activated cell sorting (FACS)-purified MuSCs treated with ET-3 or NT for 5 days. Principal component analysis (PCA) of RNA-seq data revealed strong differences between treated and non-treated sample but a high correlation between ET-3- or NT-treated MuSCs with an R-value of 0.86 (Fig. 5a). The majorities of upregulated and downregulated genes (69.47% and 79.55%, respectively) were identical between ET-3 and NT treatments, relative to treatment with solvent (Supplementary Fig. S5a). Further Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG), and Gene Set Enrichment Analysis (GSEA) indicated that upregulated and downregulated genes were primarily associated with Rho signaling pathways, suggesting that members of Rho family are the major effector molecules mediating G₁₂-G₁₃ signaling after ET-3 or NT stimulation (Fig. 5b; Supplementary Fig. S5b, c). Next, we examined the levels of active-Rho (Rho-GTP) following 5 days of G₁₂-G₁₃ activation by ET-3 and NT treatment, unraveling a substantial increase of active-Rho levels in MuSCs compared to control (Fig. 5c). Consistent with these findings, we observed a similar upregulation of active-Rho levels in MuSCs on isolated single FDB myofibers after treatment for 30 min with ET-3 or NT. The increase of active-Rho induced by ET-3 or NT was comparable to the effects of WNT4, which has been previously reported to activate RhoA in quiescent MuSCs (Fig. 5d, e). However, unlike ET-3 or NT and despite increased levels of active Rho following WNT4 treatment, WNT4 was unable to limit activation and proliferation of MuSC in vitro (Fig. 5f; Supplementary Fig. S5d). We speculate that effects of WNT4 treatment on RhoA activation is dissenting, less durable, or more indirect compared to ET-3 and NT, although we did not analyze such possibilities in detail. Consistent with these findings, the Rho inactivator C3 transferase (C3) blocked ET-3/NT-induced effects on MuSCs ex vivo (Fig. 5f; Supplementary Fig. S5d), confirming that Rho family members are the pivotal downstream mediators of G₁₂-G₁₃ signaling induced by ET-3 or NT. Analysis of active-Rho levels in MuSCs from G12/13^scKO mice confirmed the dependency of Rho activation on G₁₂-G₁₃ signaling. In Gna12-Gna13-deficient MuSCs on single myofibers isolated of G12/13^scKO mice, we observed significantly lower levels of active Rho and phosphorylated-Myosin Light Chain (pMLC), which is regulated by the Rho/Rho-kinase signaling pathway (Supplementary Fig. S5e‒h).

Fig. 5: ET-3- and NT-induced G₁₂-G₁₃ signaling requires RhoA for suppressing MuSC activation.

a PCA of RNA-seq data from DMSO-, ET-3- and NT-treated MuSCs and Pearson’s values. Each dot represents the mean of three biological samples. b GSEA of upregulated and downregulated genes, overlapping between ET-3- or NT-treated MuSCs compared to those treated with DMSO. c Rho activity of MuSCs measured by the G-LISA Kit after a 5-day culture with DMSO, ET-3, or NT. Active RhoA-GTP was determined by luminescence at 490 nm (n = 8). d, e Immunofluorescence (d) and quantification (e) of active Rho in MuSCs on isolated FDB myofibers of WT mice, after 30 min exposure to DMSO, ET-3 or NT (n = 5). f Ratios of PAX7⁺MYOD^‒ MuSCs on FDB myofibers from WT mice after 24 h exposure to DMSO, ET-3, or NT, with or without Rho inhibitor (C3) (n = 6). g RT-qPCR analysis of Rhoa expression in fresh isolated MuSCs of control (blue), Rhoa^scKO/+ (purple), and Rhoa^scKO (green) mice (n = 3). h, i Immunofluorescence (h) and quantification (i) of MuSCs outside the basal lamina in TA muscles of control, Rhoa^scKO/+, and Rhoa^scKO mice (n = 4). j Ratios of KI67⁺ MuSCs in TA muscles of control, Rhoa^scKO/+ and Rhoa^scKO mice (n = 3). k Quantification of PAX7⁺ (green bars), PAX7⁺MYOD⁺ (gray bars), and MYOD⁺ (ivory bars) cells in TA muscles of control, Rhoa^scKO/+, and Rhoa^scKO mice (n = 4). l, m Immunofluorescence (l) and quantification of MYOG⁺ cells (m) in TA muscles of Rhoa^scKO mice (n = 4). n, o H&E staining of Rhoa^scKO TA muscles sections (n). Quantification of centronuclear myofibers (o) in TA muscles of control, Rhoa^scKO/+, and Rhoa^scKO mice (n = 4). The data represent means ± SEM, analyzed by one-way ANOVA with Bonferroni’s multiple comparisons test (c, e–g, i, j, m and o). Scale bars: 5 µm in d, 10 µm in h, l and 20 µm in n.

Twenty mammalian Rho GTPases have been described, of which RhoA is one of the most prominent members, often activated through G₁₂-G₁₃³⁶. To test whether RhoA is indeed the critical mediator of G₁₂-G₁₃ signaling induced by ET-3 and NT, we conditionally inactivated Rhoa in MuSCs using Rhoa^scKO mice, in which gene inactivation is driven by Pax7-CreERT after tamoxifen administration (Fig. 5g). Homozygous inactivation of Rhoa in MuSCs dramatically increased the number of PAX7⁺ MuSCs outside the stem cell niche (Fig. 5h, i) and the number of proliferating Ki67⁺PAX7⁺ cells 24 days after initiation of gene inactivation (Fig. 5j). Furthermore, Pax7⁺MYOD^‒ MuSCs declined whereas Pax7⁺MYOD⁺ and Pax7^‒MYOD⁺ cells increased (Fig. 5k). Heterozygous inactivation of Rhoa had similar but far less pronounced effects and showed some notable differences to the homozygous mutant state (Fig. 5h‒j). We did not observe a significant increase of Pax7^‒MYOD⁺ cells in heterozygous Rhoa^scKO/+ muscles (Fig. 5k) and no increase of MYOG-positive muscle cells and centrally located myonuclei, which were abundant in homozygous Rhoa^scKO muscles (Fig. 5i‒o). Apparently, full activation of the myogenic program, indicated by expression of MYOG, resulting in the fusion of MuSCs to adjacent fibers and giving rise to centrally located myonuclei, requires nearly complete repression of RhoA activity. Taken together, inactivation of Rhoa in MuSCs fully phenocopies the loss of Gna12-Gna13, indicating that RhoA is necessary and irreplaceable for relaying G₁₂-G₁₃-dependent quiescence signals, probably serving as the primary downstream effector of G₁₂-G₁₃ in MuSCs to secure quiescence.

Activation of RhoA is sufficient to maintain MuSC quiescence in the absence of G₁₂-G₁₃ and has no effects on cytoplasmic projections of quiescent MuSC

To investigate whether RhoA is not only necessary but also sufficient to arrest MuSCs in quiescence, we generated a mouse model enabling conditional, MuSC-specific expression of a constitutively active mutant (Q63L) of human RhoA (Pax7^CreERT2; Tg: R26^LSL-caRhoa). We also fused a GFP-reporter to the N-terminus of caRhoA to facilitate detection (Fig. 6a). The approach was highly effective and specific. Essentially all Pax7⁺ MuSCs showed GFP-fluorescence, without any GFP fluorescence outside Pax7⁺ MuSCs (Supplementary Fig. S6a, b), and active RhoA was strongly increased (Supplementary Fig. S6c). The enhanced RhoA activity strongly suppressed motility of MuSCs, a hallmark of MuSC activation³⁷, in a transwell migration assay (Fig. 6b) and essentially annihilated proliferation of freshly isolated MuSCs (Fig. 6c). Furthermore, enhanced RhoA activity prevented activation of MuSC as indicated by the absence of MYOD expression in MuSCs on isolated myofibers from FDB muscles after 8 h of culture (Fig. 6d; Supplementary Fig. S6d).

Fig. 6: Constitutive activation of RhoA enforces MuSC quiescence and bypasses effects caused by inhibition of G₁₂-G₁₃ signaling.

a Design of the caRhoa overexpression allele. b Motility measurement of control and caRhoa^scOE MuSCs for 5 days in transwell migration assays (n = 5). c Time-lapse imaging of MuSCs proliferation. Gray dots: control MuSCs; red dots: caRhoa^scOE MuSCs (statistical significance at the last time point, n = 3). d Ratio of MYOD⁺PAX7⁺ MuSCs on FDB myofibers, 8 h after isolation from control and caRhoa^scOE mice (n = 6). e Experimental design of f‒h. f‒h Quantification of MuSCs outside the basal limina (f) and ratios of MYOD⁺ (g) and KI67⁺ (h) MuSCs in control and caRhoa^scOE mice in TA muscles intramuscularly injected with saline, BQ788 (1 mg/kg bodyweight), or SR142948A (0.5 mg/kg bodyweight) (n = 3). i Experimental design of j‒l. j‒l Quantification of MuSCs outside the basal lamina (j), ratio of MYOD⁺ MuSCs (k), and number of centronuclear myofibers (l) in control, caRhoa^scOE, G12/13^scKO, and caRhoa^scOE/G12/13^scKO (G12/13^RSU) TA muscles (n = 3). m Immunofluorescence for PAX7 (green), α-TUBULIN (red), DAPI (blue), and GFP (green) for control and caRhoa^scOE EDL myofibers (upper panel). The proportion of MuSCs with or without QPs is shown in the lower panel (n = 3). n, o Immunofluorescence (n) and quantification of Rac-GTP (o) in MuSCs on EDL myofibers (n = 6). The data represent means ± SEM, analyzed by unpaired t-test (b, d, m, o) and one-way ANOVA with Bonferroni’s multiple comparisons test (c, f‒h, j‒l, n, o). Scale bars: 10 µm in m and 5 µm in n.

To analyze whether constitutively active RhoA preserves quiescence even after inhibition of ET-3 and NT signaling, which normally abrogates MuSC quiescence, we administered the ET_B and NTS₂ blockers SR142948A and BQ788, respectively, to TA muscles of caRhoa^scOE mice (Fig. 6e). Expression of caRhoA prevented activation of MuSC induced by SR142948A and BQ788 in control mice, measured by the absence of MuSCs outside the basal lamina, missing increase of MyoD⁺Pax7⁺ MuSCs, and missing increase in proliferating Ki67⁺/Pax7⁺ MuSCs (Fig. 6f‒h; Supplementary Fig. S6g). We also employed the genetic G12/13^scKO mouse model to demonstrate that increased levels of active RhoA are sufficient to arrest MuSC in quiescence, despite the absence of G₁₂-G₁₃-mediated signals. Analysis of caRhoa^scOE/G12/13^scKO triple mutant mice (G12/13^RSU mice) revealed a full rescue of the G12/13^scKO phenotype by expression of caRhoA, completely eliminating premature MuSC activation and normalizing the number of centrally located myonuclei to control levels (Fig. 6i‒l; Supplementary Fig. S6e‒j). We concluded that RhoA is a central signaling hub, in which not only G₁₂-G₁₃-derived signals but also others are integrated.

A recent report described that a Rac1-to-Rho switch is required for exiting quiescence and for retraction of complex cytoplasmic projections of quiescent MuSC, named quiescent projections (QPs). The authors postulated a cross-regulated equilibrium between Rac1 and Rho, in which increased Rho activity breaks quiescence³⁸. The claim that increased Rho/ROCK signaling abrogates quiescence and results in retraction of QPs is clearly in conflict with our findings. We therefore decided to examine the morphology of QPs as a further indicator of MuSC quiescence and analyze RhoA targets including Rac proteins and ROCKs in caRhoa^scOE MuSCs. Immunofluorescence staining of MuSCs for PAX7 and α-Tubulin on isolated single myofibers from control and caRhoa^scOE mice revealed no effects of increased RhoA activity on the formation of QPs, which is consistent with the arrest of caRhoa^scOE MuSCs in quiescence. Neither did the length of QPs change nor the number of MuSCs with visible QPs (Fig. 6m). Moreover, increased expression of active RhoA did not alter the levels of active Rac-GTP, indicating that RhoA does not suppress Rac in MuSCs in vivo (Fig. 6n, o).

After establishing that RhoA does not exert its effects by altering Rac activity, we turned to two well-known targets of RhoA, the Rho-Kinase 1/2 (ROCK1/2) and Diaphanous-related formins (DRFs), which rearrange the cytoskeleton by polymerizing actin³⁹. Proliferation of MuSC was measured by EdU incorporation on isolated myofibers and activation of myogenic differentiation by determining the number of MYOD⁺PAX7⁺ cells. Pharmacological inhibition of ROCK1-ROCK2 by the ROCK inhibitor Y-27632 (Y27) released the proliferation block of caRhoa^scOE MuSCs (Fig. 7a, b) but did not release the inhibition of myogenic differentiation, imposed by expression of caRhoA (Fig. 7c). Vice versa, inhibition of DRFs by the general formin inhibitor SMIFH2 induced expression of MYOD in caRhoa^scOE MuSCs on isolated myofibers but had no effects on proliferation (Fig. 7d, e). The differential effects of Y27 and SMIFH2 on MuSCs expressing caRhoA indicate that RhoA controls two different pathways to enforce quiescence, ROCKs for preventing cell cycle entry and formins to suppress myogenic differentiation. Notably, mDia1, a member of the DRF family, has been reported to inhibit expression of MYOD and MYOG, which confirms our findings^40,41. Since RhoA induces ROCK-mediated phosphorylation of MLC, which suppresses activation and nuclear translocation of YAP in some cell types⁸, we analyzed the presence of YAP in nuclei of control, G12/13^scKO, Rhoa^scKO/+, and Rhoa^scKO MuSCs. Only very few YAP⁺ nuclei were detected in control MuSCs, whereas deletion of Gna12-Gna13, homozygous inactivation of Rhoa, and in particular heterozygous inactivation of Rhoa raised the number of YAP⁺ nuclei (Fig. 7f). However, we would like to emphasize that a large fraction of MuSCs in Rhoa^scKO and G12/13^scKO mice did not show nuclear translocation of YAP (Fig. 7f), suggesting that RhoA does not solely rely on inhibition of YAP to induce quiescence of MuSCs. To confirm the critical role of YAP downstream of RhoA for stem cell activation, we inhibited ROCK with Y27 and formins with SMIFH2 in caRhoa^scOE MuSCs on isolated myofibers, with or without concomitant treatment with the YAP inhibitor verteporfin (VP) (Fig. 7g). Strikingly, administration of VP abrogated stimulation of EdU incorporation instigated by blockage of ROCK with Y27 but had no effect on MyoD expression, induced by SMIFH2 (Fig. 7h, i). Taken together, these data corroborated the hypothesis that RhoA controls two different pathways to synergistically enforce quiescence, a formin-dependent pathway that inhibits myogenic differentiation and a ROCK-dependent pathway, which inhibits proliferation by suppressing YAP activation (Fig. 7j).

Fig. 7: Effects of RhoA to prevent cell-cycle entry and myogenic differentiation are mediated by ROCK1/2-YAP and Formin signaling.

a Experimental design of b‒e. b, c Ratios of EdU⁺ (b) and MYOD⁺ (c) MuSCs in control and caRhoa^scOE EDL myofibers, following 8-h exposure to DMSO and 10 µM Y27 (n = 4). d, e Ratios of MYOD⁺ (d) and EdU⁺ (e) MuSCs in control and caRhoa^scOE EDL myofibers following 8-h exposure to DMSO and 20 µM SMIFH2 (n = 4). f Proportions of YAP⁺ MuSCs on control, G12/13^scKO, Rhoa^scKO/+ and Rhoa^scKO TA muscle sections (n = 3). g Schematic representation of the experimental design of h, i. h, i Quantifications of the ratio of EdU⁺ (h) and MYOD⁺ (i) MuSCs in isolated myofibers from FDB muscles of caRhoa^scOE mice after 8 h treatment with DMSO, 10 µM Y27, or 20 µM SMIFH2 alone or in combination with 5 mΜ VP (n = 4). j Model depicting ROCK-YAP and Formin signaling pathways downstream of RhoA, preventing cell cycle entry and myogenic commitment in quiescent MuSC. The data represent means ± SEM, analyzed by one-way ANOVA with Bonferroni’s multiple comparisons test (b‒e, h, i). Drawings displayed in a, g and j were created with BioRender.

Discussion

The broad array of signals and mechanisms controlling MuSC quiescence is staggering and hard to understand. Several GPCRs, such as the CalcR and GPR116, are already known to contribute to MuSC quiescence but their individual impact on quiescence appears to be limited, suggesting the existence of combinatorial and partially overlapping signaling pathways^8,9,13,14. The discovery that G₁₂-G₁₃ but not G_q-G₁₁ bundle the input of numerous different GPCRs, including ET_B, NTS₂, S1P₃ and BK₁, to arrest MuSC in quiescence answers the question why different ligand‒receptor combinations have similar effects on MuSCs. However, it still remains a mystery why so many diverse ligands are employed, all resulting in G₁₂-G₁₃ activation. The answer for this conundrum may reside in the different origins of ligands and the need for graded responses of MuSCs to comply to changes in the microenvironment. Taking ET-3 and NT as an example: NT is derived from endothelial cells whereas ET-3 serves MuSCs in an autoregulatory loop. Altered expression of NT in endothelial cells, which may happen due to changes in the blood flow or during ischemia, provides a direct communication channel between the vasculature and MuSCs. Altered expression of ET-3 in MuSC will modify responsiveness of MuSC to other signals from the niche, including signals from endothelial cells. Eventually, information is computed by G₁₂-G₁₃ to enhance or lower RhoA activity, allowing proportional responses of MuSCs. Similar principles may apply to other adult stem cells but also during development and in pluripotent stem cells⁴².

We identified RhoA as a critical signaling hub downstream of G₁₂-G₁₃, which integrates signals from different pathways to control initiation of proliferation via ROCK and YAP, and myogenic differentiation via the DRF formins in quiescent MuSCs. We demonstrated that G₁₂-G₁₃ are dominant activators of RhoA in quiescent MuSCs, although other pathways have been described to activate RhoA in quiescent MuSCs, such as WNT4, derived from myofibers⁸. However, the effects of WNT4 inactivation in myofibers on MuSC quiescence are relatively weak, which corresponds to our observations that, unlike ET-3 and NT, WNT4 is unable to limit activation and proliferation of MuSC in vitro. WNT4 activates RhoA via VANGL2 in the PCP pathway in some cell types⁴³. So far, it has not been investigated whether this mechanism is active in MuSCs and whether heterotrimeric G-proteins contribute to WNT-mediated activation of RhoA. Evidence for an involvement of heterotrimeric G-proteins in WNT-signaling is scarce and an involvement of G₁₂-G₁₃ has not been described so far, although interactions of Frizzled receptors with Gα subunits have been reported by different groups^44,45. In our hands, regular levels of WNT4 derived from myofibers were unable to sufficiently activate RhoA in the absence of G₁₂-G₁₃ for maintaining MuSCs quiescence, which might be seen as a hint for involvement of G₁₂-G₁₃ in WNT4-mediated activation of RhoA. Yet, the effects of WNT4 on RhoA activation via the PCP pathway might be simply weaker compared to GPCR-G₁₂-G₁₃-mediated RhoA activation or might require additional, synergistic inputs. It will be possible to distinguish between PCP- and G-protein-dependent signaling downstream of WNT4 in future experiments and analyze whether WNT4 signaling can bypass the block in RhoA activation in G12/13^scKO mice and restore MuSC quiescence.

Our studies uncovered that active RhoA simultaneously employs two different mechanisms to prevent activation of MuSCs: (i) suppression of YAP nuclear translocation via ROCK1/2 and (ii) inhibition of myogenic differentiation via activation of DRF formins, which blocks cell cycle entry and obstructs MyoD expression, respectively. These findings are in full agreement with previous reports, showing that mDia1 (a DRF formin) suppresses expression of MyoD and myogenin^40,41. Likewise, it has been demonstrated that RhoA inhibits YAP activation via ROCK-mediated MLC phosphorylation in certain cell types⁸. We reason that concomitant activation of both pathways is necessary to arrest MuSC in quiescence, since numerous MuSCs in Rhoa ^scKO and G12/13^scKO mice did not show translocation of YAP into the nucleus, despite widespread activation of MuSCs in these mutants.

Increase of active RhoA had no effects on Rac1 activity and on the number and length of QPs, which are considered to be hallmarks of quiescent MuSCs^38,46. Our data are in conflict with a recent report, describing a cross-regulated equilibrium between Rac1 and Rho, in which increased Rho activity breaks quiescence and results in retraction of QPs³⁸. The conflicting data might be explained by different experimental conditions and/or differential roles of activated RhoA during MuSC activation. RhoA is a principal regulator of the F-actin cytoskeletons, controlling F-actin dynamics when quiescent MuSC extend QPs, but also when they retract extensions^38,47,48,49. Quiescent MuSCs are dynamic and exhibit variability in protrusions within their natural environment, probably in response to changes in mechanical force, stiffness, or biochemical cues within the muscle, suggesting that retraction of QPs also happens in quiescent MuSCs when RhoA activity is high⁴⁶. We assume that additional signals are necessary for permanent retraction of QPs, which temporarily take advantage of the role of RhoA as a regulator of F-actin dynamics, independent of its role in regulating MuSC quiescence via ROCK1/2 and DRFs. Additional studies are necessary to unveil details of this fascinating process, focusing on temporal changes of RhoA/Rac1 activities in MuSC in vivo.

We found that loss of MuSC quiescence following inactivation of Gna12-Gna13 results in depletion of the MuSC pool and is associated with enhanced sarcopenia during aging. The correlation between loss of quiescence enhanced sarcopenia during aging is intriguing, but requires further investigations. Resting MuSCs in geriatric mice lose reversible quiescence by switching to an irreversible pre-senescence state, which is a different condition than mere loss of quiescence¹. Surprisingly, diphtheria-mediated depletion of MuSC impairs skeletal muscle regeneration but without causing sarcopenia³. A similar study demonstrated that despite ablation of MuSCs with > 97% efficiency, the average size of cross-sectional areas of myofibers does not decline in aged mice⁴. Compensatory mechanisms within myofibers seem to allow maintenance of fiber sizes even without the influx of new MuSCs, at least to a certain degree and in the absence of severe injury or other challenges. These studies clearly show that it is not necessarily the reduction or absence of MuSC, which causes sarcopenia during aging. Instead, accretion of constantly activated, defective MuSCs might be responsible. This hypothesis is supported by the necessity for tight quality control of MuSCs by programmed cell death to ensure muscle health⁵⁰, but certainly needs further proof. We are speculating that long-term loss of quiescence may lead to the accumulation of persisting chromatin changes or other defects in MuSCs, which may contribute to fiber atrophy and sarcopenia. Careful analysis of defects in MuSCs of aging G12/13^scKO mice and inhibition of MuSC fusion with myofibers will clarify whether such phenomena occur or not. Alternatively, constantly activated or aberrant MuSCs might exert adverse paracrine effects on myofibers, promoting atrophy⁵¹.

At present, we do not know whether loss of MuSC quiescence, the observed changes in the fiber type composition of G12/13^scKO mice, or completely different processes are responsible for enhanced sarcopenia during aging. G12/13^scKO show a strong reduction of type I myofibers, which increases the relative abundance of non-oxidative type II fibers. The decline in skeletal muscle mass with aging has been mainly attributed to a reduction in type II muscle fiber size⁵². Thus, it is entirely feasible that enhanced sarcopenia during aging is promoted by fiber type switching in G12/13^scKO muscles. Surprisingly, inactivation of Gna12-Gna13 in MuSCs reduces the number of type I fibers, whereas inactivation of Gna13 in myotubes increases type I and IIa oxidative fiber³², suggesting involvement of different mechanisms. Accordingly, we did not find a substantial reduction of G13 protein levels in myofibers from G12/13^scKO mice, which was matched by the dominant presence of the intact WT Gna13 allele in individual myofibers. These results suggest that nuclei from Gna12-Gna13-deficient MuSCs in non-regenerating muscles only contribute to a minority of nuclei in myofibers, even in 80-week-old mice, corroborating earlier studies that myonuclei are rather stable under physiological conditions³³.

Taken together the comprehensive characterization of GPCRs, G-proteins, and downstream effectors regulating MuSC quiescence provides novel means for pharmacological restoration of quiescence. We assume that the continuous decline of Edn3, Ednrb, and Ntsr2 expression in MuSCs during aging, marking an age-associated decline in G₁₂-G₁₃-transduced GPCR signaling in MuSCs, contributes to the loss of MuSC quiescency and attenuated regeneration in aged individuals^53,54,55. Pharmacological enhancement of G₁₂-G₁₃ and RhoA signaling in MuSC may pave the way to better muscle health during aging.

Materials and methods

Animals

The Gnaq^‒/‒ (Gnaq knockout), Gna11^fl/fl, Gna12^‒/‒ (Gna12 knockout), Gna13^fl/fl, and Rhoa^fl/fl mouse strains have been described previously^{56,57,58,59,60}. The Pax7^CreERT2 mouse strain was obtained from Kardon lab⁶¹. For engraftment assays, 2-month-old male Dmd^mdx-4Cv/Y mice and male Pax7^nGFP transgenic mice were used^62,63. Transgenic R26^LSL-caRhoa mice (Tg: Rosa26 locus: Loxp-stop-Loxp-caRhoa), expressing a constitutively activate mutant (Q63L) of human RhoA protein with an N-terminal fusion of enhanced GFP (caRhoA) after Cre recombinase-mediated excision of a loxP-stop-loxP sequence, were generated by inserting a cassette derived from a plasmid construct (pcDNA3) containing cytomegalovirus enhancer, cytomegalovirus promoter, promotor for bacteriophage T7 RNA polymerase, and a bovine globin polyadenylation signal (# 12968, Addgene) into the ROSA26 locus of V6.5 mouse embryonic stem (ES) cells. Chimeric mice were generated by injection of ES cells into mouse blastocytes as described⁶⁴. All mice were backcrossed into the C57BL6 genetic background for at least 3 generations. The Primers used for genotyping are shown in Supplementary Table S2.

Animal procedures

2-month-old male mice were used in all experiments, unless specified otherwise. Intraperitoneal administration of tamoxifen (Sigma) at a dosage of 0.05 mg/g of body weight was carried out on the mice. TA muscles were injected with CTX (0.06 mg/mL, Sigma) in a volume of 50 μL. Engraftment of MuSC was done according to a published protocol⁶⁵. Prior to engraftment, donor cells were counted using a haemocytometer, excluding non-viable cells stained with 0.4% Trypan Blue (Gibco). Donor MuSCs were then centrifuged at 1200× g, 4 °C for 20 min and resuspended in DMEM media (Life Technologies) before transplantation into the TA muscle using a 5-µL microcapillary pipette (Drummond). All animal care and handling procedures were conducted in strict accordance with the “Guide for the Care and Use of Laboratory Animals” published by the US National Institutes of Health. The study protocol was approved by the Animal Rights Protection Committee of the State of Hessen (Darmstadt, Germany, approval number: B2/1224), as well as the Institutional Animal Care and Use Committee of Sichuan Agricultural University (Chengdu, China, approval number: 20190241).

MuSC isolation & culturing

MuSCs were purified via FACS using a published method⁶. Isolated MuSCs were cultured in growth medium 1 (GM1: DMEM-GlutaMAX containing 20% fetal calf serum (FCS), 1% Penicillin/Streptomycin, and 5 ng/mL basic fibroblast growth factor (bFGF)) or in screening medium (SM: DMEM-GlutaMAX containing 20% knockout serum (KSR, Giboc), 1%, Penicillin/Streptomycin, and 1% chicken embryo extract). MuSCs were cultured ± ET-3 (1 μM), ± NT (1.5 μM), ± BQ788 (10 nM), ± SR142948A (5 nM), ± Cell Permeable Rho Inhibitor (C3 Transferase) (1 μg/mL) and other compounds listed in the Supplementary Table S1. GM or SM was supplemented by addition of EdU to a final concentration of 10 μM, 3 h before fixation. Fixed MuSCs were analyzed using the Click-iT EdU kit (Invitrogen) following the manufacturer’s protocol. Time-lapse imaging and analysis were performed using an Incucyte Live-Cell Imaging System and software (Essen Instruments).

Adenovirus-mediated Cre deletion of floxed sequences in MuSC

FACS-purified MuSCs from Gna12^‒/‒Gna13^fl/fl or Gna11^‒/‒Gnaq^fl/fl mice were cultured in SM and exposed to Ad-Cre virus (Vector Biolabs) for at least 2 days to facilitate Cre-mediated deletion of Gna13 or Gnaq in MuSCs in vitro. MuSCs treated with adenoviruses that do not express Cre (Ad-Null) served as controls.

Single myofiber isolation & culturing

Myofibers were obtained through enzymatic digestion of FBD or extensor digitorum longus (EDL) muscles using collagenase 2 (0.02%, Roche) without or with prefixation (mice were perfused with 0.5% paraformaldehyde (PFA), 3 times within 20 min). Unfixed freshly isolated myofibers were cultured in growth medium 2 (GM2: DMEM-GlutaMAX medium supplemented with 20% FCS and bFGF at a concentration of 5 ng/mL) or screening medium 2 (SM2: DMEM-GlutaMAX containing 20% KSR, 1% Penicillin/Streptomycin). Drugs listed in Supplementary Table S1 were added immediately after muscle excision already in the digestion tubes and in culture plates, ensuring immediate exposure. Myofibers were cultured for 0, 8, or 24 h in the absence or presence of these drugs. Subsequently, myofibers were fixed with 4% PFA for 10 mins in preparation for further analysis.

GPCR compounds screening

For screening of the customized GPCR compound library (Supplementary Table S3), freshly isolated MuSCs were suspended in the screening medium (SM), detailed in the ‘MuSC Isolation & Culturing’ section. 50 µL of the suspension, containing 400 cells, were added into a single well of a 384-well plate (781097, Greiner), precoated with Matrigel (354234, Corning). After seeding, additional 50 µL of SM, containing either the respective compound, known activators or inhibitors (serving as positive and negative controls, respectively), or DMSO (solvent control; final concentration of 0.5%) were added to the wells. Plates were incubated at 37 °C and 5% CO₂. The different compounds were kept in 10 mM stock solutions in DMSO, ranging from 1%‒100% in water. Compounds were diluted in SM to ensure the final DMSO concentration remained below 0.5% per well. Based on cytotoxicity measurements in a 1‒100 µM range, an initial concentration of 10 µM was selected for primary screening. EdU was added at the fifth day after compound administration to reach a final concentration of 10 μM, three hours before cell fixation. The ratios of EdU⁺/DAPI⁺ and MYOD⁺/PAX7⁺ nuclei were used as a read-out to determine activation of MuSC. Cut-offs for potential hits were defined by the activities of OSM, which induces MuSC quiescence, and FGF2 and IGF1, which promote proliferation and myogenic differentiation of MuSCs. Compounds that showed promise were used for a secondary assay at reduced concentrations.

Immunofluorescence assay

Myofibers, cryosections, or cultured MuSCs were fixated using 4% PFA for 10 min. Samples were permeabilized, following fixation, using 0.1% Triton X-100 for 15 min. Subsequently, specimens were blocked in a 3% bovine serum albumin (BSA) solution for 30 min and then exposed to the respective antibodies (Supplementary Table S1) overnight at 4 °C. Visualization of immunofluorescence signals was achieved by secondary antibodies conjugated with Alexa488 or Alexa594. A fluorescence microscope (AXIO observer Z1, Zeiss) equipped with objective lenses of 63×, 40×, and 20× was employed to detect and record the fluorescence signals. To quantify fluorescence intensity, ImageJ software was utilized, enabling subsequent statistical analyses of data. Specificity of the antibody against NT (GTX37368, GeneTex) was verified in in vitro cultured primary murine lymphatic endothelial cells after siRNA-mediated knockdown of Nts (Supplementary Fig. S2j, k).

RhoA G-LISA assay

To determine RhoA activity in cultured MuSCs, the RhoA G-LISA Activation Assay kit (BK121, Cytoskeleton) was employed, which measures Rho-GTP level. Cultured MuSCs were starved in serum-free DMEM-GlutaMAX medium for a duration of 10‒30 min. Subsequently, they were stimulated with either ET-3 (3 μM) or NT (3 μM) for 10 min. Following stimulation, cells were lysed with the lysis buffer included in the kit for 10 min on ice. Total cell extracts were prepared and adjusted to a protein concentration of 1 mg/mL for quantitative detection of active and total RhoA according to the manufacturer’s instructions. Luminescence intensity recorded using the LB940 Mithras plate reader (Berthold Technologies) at 490 nm.

In situ binding assays for Rho & Rac GTPase activity

Fixed myofibers (see above) were blocked in 3% BSA and incubated for 1 h at room temperature with GST-tagged PAK-P21 binding domain (Cytoskeleton Inc., PAK01) proteins and GST-tagged Rhotekin-Rho binding domain (Cytoskeleton Inc., RT01) which specifically binds to active Rac/Cdc42 and active Rho, respectively. Myofibers were washed 3 times with PBS and then incubated with a fluorophore-conjugated anti-GST antibody (Invitrogen) for an hour at room temperature. DAPI (Life Technologies) was utilized to stain nuclei.

BRET2 assays

HEK 293 T cells were transfected with a plasmid cocktail (1:1:1:1 plasmid DNA ratio of receptor:Gα12(134)-RLuc8/Gα13(126)-RLuc8:Gβ3:Gγ9-GFP2) after reaching a density of 600,000 ~ 800,000 cells per well in 6-well plates following a previously published protocol³¹. Medium of transfected cells was changed daily for 3 days, using selection medium (DMEM containing 10% FCS, 1% Penicillin/Streptomycin, and 200 μg/mL G418). The whole length cDNA of human EDNRB and NTSR2 genes were cloned from EDNRB-Tango (# 66458, Addgene) and NTSR2-Tango (# 66458, Addgene) plasmids, respectively, and inserted into the pcDNA3.1(+) vector subsequently. Selected cells were seeded into 96-well plates, after which the culture medium was replaced with 60 µL of assay buffer (1× Hank’s balanced salt solution (HBSS) + 20 mM HEPES, pH 7.4), followed by addition of 10 µL freshly prepared 50 µM coelenterazine 400a (Nanolight Technologies). After 5 min equilibration, cells were treated with 30 µL of ET-3 or NT for an additional 5 min. Next, luminescence signals were recorded at 395 nm (RLuc8-coelenterazine 400a) and 510 nm (GFP2) emission filters, at integration times of 1 s per well, using the LB940 Mithras plate reader (Berthold Technologies). Fluorescence signals were serially recorded for six times. Measurements from the sixth read were used in all analyses. BRET2 ratios were computed as the ratio of GFP2 to RLuc8 emission. To confirm the specificity of ligand‒receptor interactions in BRET assays, NTSR2-expressing cells were treated with 30 µL of ET-3, and EDNRB-expressing cells with 30 µL of NT. The absence of significant BRET signals in the control groups confirmed the specificity of interactions.

RNA-seq

RNA-seq analysis was conducted following established protocols⁶⁶. Total RNA was extracted using the RNAeasy Mini kit (QIAGEN), following the manufacturer’s guidelines. Quality and integrity of RNA and of library preparations were assessed using the LabChip Gx Touch 24 (PerkinElmer). For VAHTS Stranded mRNA-seq, 300 ng of total RNA was used as input. Library preparations were performed according to the manufacturer’s protocol (Vazyme). Sequencing was carried out on a NextSeq500 instrument (Illumina) with v2 chemistry, resulting in an average of 30 M reads per library using 1× 75 bp single-end reads. Raw reads underwent quality assessment, adapter content analysis, and duplication rate estimation using FastQC 0.10.1. Trimming of reads was performed using Reaper version 13-100. Subsequently, reads were aligned to the Ensemble mouse genome version mm10 (GRCm38) using STAR 2.4.0a. The featureCounts 1.4.5-p1 tool from the Subread package was employed to count reads aligning to genes. Only reads mapping within exons were considered, and counts were aggregated per gene. Reads overlapping multiple genes or aligning to multiple regions were excluded. DESeq2 version 1.62 l was utilized to identify differentially expressed genes (DEGs). Genes were considered significantly differentially expressed if they exhibited an absolute fold change of ≥ 2, a Benjamini-Hochberg corrected P ≤ 0.05, and a minimum combined mean of 5 reads. The “plotPCA function” of R package “DESeq2” was employed to perform PCA on the rlog-transformed count data. Pearson correlation was calculated between the means of CPM normalized expression for each replicate group. Differences in expression were assessed using the lfcShrink function with the apeglm method (v1.6.0). GSEA was performed to characterize DEGs, utilizing gene sets extracted from MSigDB. GO and KEGG analyses were performed using the “clusterProfiler” package of R.

Western blot assay

Freshly isolated or cultured MuSCs were rinsed with PBS and lysed in cell lysis buffer for 10 mins. Lysates were sonicated and proteins were separated through SDS-PAGE and transferred onto nitrocellulose membranes (Millipore, Billerica, MA). The membranes were probed with primary antibodies detecting G₁₂, G₁₃, Histone-3 (H3), eMyHC, ET-3, NT and GAPDH using antibodies listed in the Supplementary Table S1. After overnight incubation at 4 °C, HRP-conjugated secondary antibodies and the ECL detection system (Pierce) were used for signal visualization. Protein quantification was performed using ImageJ software.

RT-qPCR assays

RT-qPCR analyses were performed using SYBR Green or TaqMan probe-based RT-qPCR assays. Total RNA of samples was isolated using RNAeasy Mini kit (QIAGEN), following the manufacturer’s protocols. cDNA was generated with the PrimeScript™ II 1st strand cDNA Synthesis Kit (Takara). TaqMan probe-based RT-qPCR was performed with the TaqMan Fast Advanced Master Mix (Thermo Fisher). The assay IDs of commercial TaqMan probes for detection of genes of interest are listed in Supplementary Table S2. SYBR Green-based RT-qPCR was performed with the Applied Biosystems™ PowerUp™ SYBR™ Green Master Mix (Thermo Fisher). Sequences of primers are listed in Supplementary Table S2. Relative expression levels of genes of interest were determined using the comparative threshold method, using Gapdh as an internal control. Data were analyzed using the ΔΔCt method.

Histological analysis

Muscles were collected at specified time-points, as indicated in each figure, and rapidly frozen by immersion in isopentane cooled with liquid nitrogen. Subsequently, 8 μm thick sections of muscle samples were generated using a cryomicrotome and subjected to Hematoxylin and Eosin (H&E) staining following established protocols.

Statistics

Statistical information related to individual experiments is provided in the corresponding figure legends. Data sets with sample sizes (n) of three or greater were subjected to the Shapiro–Wilk test to assess normality; all data sets presented herein passed this test. For comparison of multiple groups, we used one-way ANOVA and for comparison of two groups, the unpaired two-tailed Student’s t-test was used. A P < 0.05 was considered to be significant (ns: no significant, *P < 0.05, **P < 0.01, ***P < 0.001). Results are presented as means ± SEM. Data analysis was performed using GraphPad Prism 9 software.