Macrophage-released ADAMTS1 promotes muscle stem cell activation

Abstract

Coordinated activation of muscle stem cells (known as satellite cells) is critical for postnatal muscle growth and regeneration. The muscle stem cell niche is central for regulating the activation state of satellite cells, but the specific extracellular signals that coordinate this regulation are poorly understood. Here we show that macrophages at sites of muscle injury induce activation of satellite cells via expression of Adamts1. Overexpression of Adamts1 in macrophages in vivo is sufficient to increase satellite cell activation and improve muscle regeneration in young mice. We demonstrate that NOTCH1 is a target of ADAMTS1 metalloproteinase activity, which reduces Notch signaling, leading to increased satellite cell activation. These results identify Adamts1 as a potent extracellular regulator of satellite cell activation and have significant implications for understanding the regulation of satellite cell activity and regeneration after muscle injury.Satellite cells are crucial for growth and regeneration of skeletal muscle. Here the authors show that in response to muscle injury, macrophages secrete Adamts1, which induces satellite cell activation by modulating Notch1 signaling.

Fig. 1

ADAMTS1 activates satellite cells. a Representative confocal images of myofibers with associated MyoD-negative (upper) and MyoD-positive (lower) YFP+ (Pax7+) satellite cells isolated from extensor digitorum longus (EDL) muscles of tamoxifen-treated Pax7^CreER/+;R26^eYFP/+mice and treated with control (upper) or rADAMTS1 (1.4 μg/ml) (lower). Activated YFP+ (Pax7+) satellite cells were identified by MyoD (red) staining and nuclei were stained with DAPI (blue). Scale bars = 20 μm. b Quantification of the increased proportion of activated satellite cells on myofibers after exposure to rADAMTS1 medium compared to control (n = 75–104 myofiber-associated satellite cells per condition; n = 4 EDL replicates. The mean of the replicates is graphed. P = 0.0023). c Quantification of the number of YFP+ (Pax7+) satellite cells per myofiber across the conditions. Each point on the graph represents one myofiber. *P < 0.05. Error bars represent s.e.m. Statistical significance tested using paired t-tests

Fig. 2

Adamts1 is induced by muscle injury and expressed in the macrophages of skeletal muscles. a Time-course monitoring the expression level of Adamts1 in tibialis anterior (TA) muscles isolated from wild-type mice before and after muscle injuries. Adamts1 levels were quantified using RT-qPCR and the injured muscles were compared to the contralateral uninjured muscle (n = 5). b, c Western blots and quantification of ADAMTS1 protein levels in cell lysates from TA muscles before and after muscle injuries compared to Actin control (n = 3 for each time point). d Images from IHC performed on frozen sections prepared from TA muscles 1 day post injury (insets are higher magnification images of a F4/80- and ADAMTS1-positive cell that is present in the lower magnification field). Histological slides were stained using antibodies against ADAMTS1 (red) and F4/80 (green); a marker for macrophages. ADAMTS1 co-localized with macrophages (inset). Scale bar = 10 μm. e Flow cytometry over a time-course monitoring Ly6C in F4/80+cells after muscle injuries. f RT-qPCR measuring the expression levels of Adamts1 in Ly6C+ compared to Ly6C− macrophages. g Representative flow cytometry plots showing ADAMTS1 expression in the macrophage population isolated from injured muscle tissue. *P < 0.05, ***P < 0.001. Error bars represent s.d. Statistical significance tested using paired t-tests

Fig. 3

Overexpression of Adamts1 promotes satellite cell activation in vivo. a Images from immunocytochemistry (ICC) performed on satellite cells immediately following isolation by FACS from wild-type (Wt) and Adamts1 (Tg) mice that had received an in vivo pulse of EdU. Cells were stained with DAPI (blue) to identify nuclei and EdU (green) and MyoD (red) to distinguish activated satellite cells (scale bar = 100 μm). b Quantification of the percent of satellite cells that are MyoD (left) and EdU (right) positive in Wt (n = 6) and Tg (n = 5) mice. c (upper) Images from IHC performed on histological sections prepared from TA muscles from Wt and Tg animals that had received an in vivo pulse of EdU and stained for DAPI (blue), EdU (green and arrows) and Dystrophin (white). Scale bar = 100 μm. (lower) Higher power magnification images of IHC showing EdU-positive cells (green) located beneath the myofiber membrane as delineated by Dystrophin (white) staining. d Quantification of the number of EdU-positive myonuclei per myofiber in Wt (n = 3) and Tg (n = 3) mice (>200 myofibers per mouse were examined). e Gross wet weights of tibialis anterior (TA) and quadriceps (Quad) muscles in Wt (n = 3) and Adamts1 transgenic (Tg; n = 4) mice at 1 month of age. f Histological sections prepared from TA muscles of 1-month-old mice and stained with hematoxylin and eosin (H&E). Scale bar = 50 μm. g Quantification of the diameters of myofibersin TA muscles from 1-month-old Tg (n = 4) and Wt (n = 3) mice. *P < 0.05, **P < 0.01. Error bars represent s.d. Statistical significance tested using paired t-tests

Fig. 4

Adamts1 mice have increased satellite cell activity in response to injury. a (left) Images from IHC performed on histological sections prepared from injured TA muscles of 1-month-old Wt and Tg mice. EdU was injected into the mice at the time of muscle injury and mice were sacrificed for IHC on day 1. Histological sections were stained with DAPI (blue) to identify cell nuclei and EdU (green) to distinguish proliferating satellite cells. Scale bar = 100 μm. a (right) Quantification of the number of proliferating satellite cells after a muscle injury in Wt (n = 3) and Tg (n = 3) (>200 myofibers per mouse were examined). b (left) Images from IHC performed on histological sections prepared from injured TA muscle from Wt mice that received bone marrow transplantations from Wt mice (Wt-BMT) or Adamts1 mice (Tg-BMT) donors. Scale bar = 50 μm. b (right) Quantification of the number of proliferating satellite cells after a muscle injury in Wt-BMT (n = 3) and Tg-BMT (n = 4) mice (>200 myofibers per mouse were examined). *P < 0.05, ***P < 0.001. Error bars represent s.d. Statistical significance tested using paired t-tests

Fig. 5

Adamts1 mice have accelerated regeneration when young but a depleted satellite cell pool with aging. a–c (left) Images of H&E-stained histological sections prepared from TA muscle injuries that were induced at 1 a, 4 b and 8 c months of age in Wt and Tg mice (mice were sacrificed 5 days post injury for the 1- and 4 month-old mice and 7 days post injury for the 8-month-old mice). Scale bar = 50 μm. a–c (right) Quantification of the cross-sectional myofiber diameters post injury calculated for a 1-month-old (Wt: n = 6; Tg: n = 3), b 4-month-old (Wt: n = 4; Tg: n = 4) and c 8-month-old (Wt: n = 3; Tg: n = 3) mice. The numbers of myofibers counted is displayed in the frequency axis. The 1-month-old Tg mice had accelerated regeneration compared to Wt littermates (P < 0.0001). The 8-month-old Tg mice had impaired regeneration compared to Wt littermates (P < 0.0001). d, e (left) Flow cytometry plots of CD45−/CD31−/Scal1−/Vcam+ satellite cells in Tg and Wt mice at d 1 and e 4 months of age. d, e (right) Quantification of the number of satellite cells per total live mononuclear cells isolated (~ 5×10⁵ cells from each animal were analyzed) from the hind limb muscles of 1-month-old (Wt: n = 11; Tg: n = 10) and 4-month-old (Wt: n = 11; Tg: n = 9) mice showing a depletion of satellite cells in Tg mice (**P < 0.01 at 1 month; ***P < 0.001 at 4 months). f (left) Images from IHC of histological sections of TA muscles isolated from 4-month old Wt and Tg mice and stained for Pax7 (red), Dystrophin (white) and DAPI (blue). Scale bar = 100 μm. f (right) Quantification of the number of Pax7-positive satellite cells per myofiber show Tg mice have fewer satellite cells than Wt mice (n = 3 for each genotype; >200 myofibers per mouse were examined). **P < 0.01. g, Quantification of the number of Pax7-positive cells on freshly isolated myofibers from 4-month-old mice (each point represents a myofiber, n = 3 mice for each genotype) demonstrated that Tg mice have fewer satellite cells. ***P < 0.001. Error bars represent s.d. Statistical significance tested using χ² for myofiber diameters and paired t-tests for other comparisons

Fig. 6

Adamts1 inhibits Notch signaling. a, b Time-course monitoring the expression levels of Notch target genes Hes1 and Hey1 in TA muscles isolated from wild-type mice before and after muscle injuries. Hes1 and Hey1 levels were quantified using RT-qPCR and the injured muscles were compared to the contralateral uninjured muscles (n = 5 mice per time point). c Representative western blot of lysates from primary myoblasts isolated from Tg and Wt mice. NICD=Notch1 intracellular domain (activated form of NOTCH1). Actin was used as a loading control. d Quantification of the levels of NICD protein in Tg (n = 7) compared to Wt (n = 8). e Quantification of Hes1 and Hey1 expression levels using RT-qPCR on primary satellite cells isolated from Wt (n = 3) and Tg (n = 3) mice by FACS. f Co-immunoprecipitation (IP) of ADAMTS1 and NOTCH1 from primary muscle tissue isolated from Tg mice using anti-HA (transgene has a hemagglutinin (HA) tag), anti-ADAMTS1 or IgG control antibodies. The transmembrane/intracellular fragment of NOTCH1 (TMIC) was immunoprecipitated specifically with ADAMTS1. NS is a nonspecific band. g Schematic of deletion and mutant Adamts1 expression constructs generated. h NOTCH1 co-IPs with ADAMTS1 and sequential ADAMTS1 deletion mutants map the NOTCH1 interaction domain to the metalloproteinase domain. Cont, control empty plasmid; FL, full-length Adamts1. i Western blots on cell lysates from co-cultured cells transfected with Notch1 and either Adamts1 (ADAMTS1) or E402Q Adamts1 (E402) or empty plasmid control (Control) expression plasmids. j (left) IP of processed myc-tagged NOTCH fragments from co-cultures with cells overexpressing Adamts1, Notch1 and Dll1 or empty vector controls. j (right) Quantification of the levels of NICD/TMIC fragments in response to co-culturing with Adamts1 overexpressing cells with and without cells expressing Dll1. P < 0.05, **P < 0.01, ***P < 0.001. Error bars represent s.d. Statistical significance tested using paired t-tests