Imbalance between pSmad3 and Notch induces CDK inhibitors in old muscle stem cells

Abstract

Adult skeletal muscle robustly regenerates throughout an organism's life, but as the muscle ages, its ability to repair diminishes and eventually fails. Previous work suggests that the regenerative potential of muscle stem cells (satellite cells) is not triggered in the old muscle because of a decline in Notch activation, and that it can be rejuvenated by forced local activation of Notch. Here we report that, in addition to the loss of Notch activation, old muscle produces excessive transforming growth factor (TGF)-beta (but not myostatin), which induces unusually high levels of TGF-beta pSmad3 in resident satellite cells and interferes with their regenerative capacity. Importantly, endogenous Notch and pSmad3 antagonize each other in the control of satellite-cell proliferation, such that activation of Notch blocks the TGF-beta-dependent upregulation of the cyclin-dependent kinase (CDK) inhibitors p15, p16, p21 and p27, whereas inhibition of Notch induces them. Furthermore, in muscle stem cells, Notch activity determines the binding of pSmad3 to the promoters of these negative regulators of cell-cycle progression. Attenuation of TGF-beta/pSmad3 in old, injured muscle restores regeneration to satellite cells in vivo. Thus a balance between endogenous pSmad3 and active Notch controls the regenerative competence of muscle stem cells, and deregulation of this balance in the old muscle microniche interferes with regeneration.

Figure 1 |. TGF-β/pSmad3, but not myostatin, increases in old skeletal muscle.

a, Immunodetection of TGF-β, myostatin or follistatin (green) and laminin (red) in 10-μm skeletal muscle cryosections. Hoechst labels nuclei (blue). Scale bar, 50 μm. b, Western blot on myofibres and satellite cells; quantified in c.d, Immunoprecipitation with anti-Smad3 antibody, followed by western blot with anti-phosphorylated Smad3 antibody; quantified in e. f, After overnight transwell co-culture with young or old myofibres, young satellite cells were analysed by western blotting; data are means ±s.d., n = 3. *P ≤ 0.05 compared with young.

Figure 2 |. Notch removes pSmad3 from the 5’ regulatory regions of CDK inhibitors.

a, Satellite cells treated with TGF-β1, with or without Notch activation, were analysed by western blotting for p15, p16, p21 and p27; quantified in b. *P ≤ 0.05 compared with untreated control (0); **P ≤ 0.05 compared with TGF-β. Smad3 co-precipitated proteins (c) were resolved by western blot; genomic DNA (d) was analysed by RT–qPCR, using primers specific for 5’ regions of CDK inhibitors; data are means ±s.d., n = 3. e, qPCR reactions after 44 cycles revealed fragments of expected molecular weight on agarose gel.

Figure 3 |. Inhibition of endogenous Notch upregulates CDK inhibitors.

a, Compared with untreated cells (−), Notch inhibition by GSI caused prompt upregulation of p15, p16, p21 and p27; western blot assays quantified in b; *P ≤ 0.05 compared with untreated control. In ChIP assay, Smad3 co-precipitated proteins were resolved by western blot (c), genomic DNA (d) was analysed by RT–qPCR, using primers to p15, p16, p21 and p27 5’ regulatory regions; data are means ± s.d., n = 3. e, qPCR reactions after 44 cycles revealed fragments of expected molecular weight on agarose gel.

Figure 4 |. Smad3 shRNA rescues responses of satellite cells in old niche in vivo.

Lentiviral-infected muscle (control transduction (Tdxn), Smad3 shRNAs, non-target shRNA) was analysed 5 days after injury. a, Haematoxylin and eosin 10-μm cryosections of Smad3 shRNA-1; quantified in b; n = 15, *, **P ≤ 0.05; individual Smad3 shRNA-2, −3 data in Supplementary Fig. 10a; immunodetection of eMyHC myofibres with BrdU⁺ nuclei. Scale bar, 100μm; O, old; Y, young. c, Isolated satellite cells analysed by western blot for p15, p16, p21 and p27; quantified in d; *P ≤ 0.05, compared with non-target shRNA, control tdxn; data are means ± s.d., n = 3.