Author contributions: T.K.: conception and design, collection and/or assembly of data, data analysis and interpretation, manuscript writing; K.H.: manuscript writing, final approval of manuscript.
First published online in STEM CELLS EXPRESS October 19, 2010.
Disclosure of potential conflicts of interest is found at the end of this article.
Satellite cells are skeletal muscle stem cells responsible for growth, maintenance, and repair of postnatal skeletal muscle. Although several studies have demonstrated that Notch signaling plays a critical role in muscle regeneration through promoting proliferation and self-renewal of satellite cells, the function of Notch3 is yet to be elucidated. We analyzed muscle regeneration in Notch3-deficient mutant mice. We found a remarkable overgrowth of muscle mass in the Notch3-deficient mice but only when they suffered repetitive muscle injuries. Immunochemical analysis found that Notch3 was expressed in Pax7+/MyoD− quiescent satellite cells and also in Pax7+/MyoD+-activated satellite cells, but the expression was restricted to around half the population of each cell type. In Notch3-deficient mice, the number of sublaminar quiescent satellite cells was significantly increased compared with those in control mice. We also found that primary cultured myoblasts isolated from the Notch3-deficient mice proliferated faster than those from control mice. Analysis of cultured myofibers revealed that the number of self-renewing Pax7-positive satellite cells attached to the myofiber was increased in the Notch3-deficient mice when compared with control mice. The data obtained in this study suggested that Notch3 pathway might be distinct from Notch1 in muscle regeneration. Because overexpression of Notch3 activated the expression of Nrarp, a negative feedback regulator of Notch signaling, Notch3 might act as a Notch1 repressor by activating Nrarp. STEM CELLS 2010;28:2205–2216
Adult skeletal muscle exhibits a remarkable ability to regenerate in response to various injuries. The primary cellular constituents of mature muscles that are responsible for this regenerative potential are the resident stem cells, termed satellite cells, which are mononucleated cells located between the sarcolemma and basement membrane of terminally differentiated muscle fibers . In normal adult skeletal muscles, satellite cells are in a mitotically quiescent state and replicate very slowly to replenish the satellite cell pool [2, 3]. When muscles are damaged, however, they become activated and undergo multiple rounds of cell division prior to terminal differentiation and fusion to form multinucleated myofibers [4, 5]. Interestingly, the muscle can regenerate even when subjected to repeated severe damage [6, 7], indicating that satellite cells possess the ability to undergo self-renewal, effectively repopulating the quiescent satellite cell compartment after activation [8–11].
The molecular mechanisms controlling activation of satellite cells and maintaining the satellite cell pool are not yet fully understood. Growth and differentiation of satellite cells are considered to be regulated by various trophic factors . Additionally, signals mediated by the Notch pathway have recently been reported to be implicated in regulating satellite cells during muscle regeneration [13, 14] and also during embryonic development [15, 16]. The Notch signaling pathway is an evolutionarily conserved mechanism that controls a broad range of developmental processes, including maintenance of progenitors, cell fate determination, proliferation, and differentiation . Genes of the Notch family encode large transmembrane receptors that interact with membrane-bound ligands encoded by the Delta and Jagged family of genes. In mammals, four Notch genes have been identified (Notch1 , Notch2 , Notch3 , and Notch4 ). Satellite cells in the adult express Notch1, Notch2, and Notch3, together with the Notch ligand Delta-like 1 [10, 13, 22]. Satellite cell activation in mice has been reported to be accompanied by activation of Notch1, which leads to proliferation and, if maintained, prevents satellite cell differentiation . Similarly, maintaining Notch1 activity by targeted disruption of the transcriptional repressor Stra13 results in perturbed satellite cell differentiation and compromised muscle regeneration . Pharmacologically blocking the Notch pathway with a γ-secretase inhibitor or a soluble Jagged-Fc fusion protein, which can block signaling of all Notch receptors, inhibits satellite cell proliferation and self-renewal [10, 14]. Furthermore, the decline of Notch1 signaling with age is thought to be another cause of the decreased regenerative potential of aged skeletal muscle. Indeed, enhancement of Notch1 signaling promotes muscle regeneration in old skeletal muscle [14, 24]. Although these experiments suggest a crucial role for Notch1 signaling in satellite cell function, much remains to be determined, especially regarding the role of Notch3 signaling during muscle regeneration.
In this study, we focused our attention on the role of Notch3 during muscle regeneration because Notch3 was expressed in satellite cells, and various structural and functional differences between Notch3 and Notch1/Notch2 have been reported . As we could not observe any obvious phenotypes in the musculature of Notch3-deficient mice in a previous study , we analyzed the musculature after repeated muscle injuries, which were conducted by generating Notch3 and dystrophin-double-deficient mice and also by repetitive intramuscular administration of cardiotoxin (CTX) to the Notch3-deficient mice. To investigate the function of Notch3 during muscle regeneration, we examined the expression of Notch3 in the satellite cells, explored the target gene of Notch3, and analyzed the effect of Notch3 deficiency on quiescent, activated, and self-renewing satellite cells in vivo and in vitro. The mechanism of satellite cell regulation during muscle regeneration was discussed in relation to the mode of action of Notch3.
MATERIALS AND METHODS
Mice and Muscle Injury
Notch3-deficient mice  were backcrossed with C57BL/6J mice for 5–10 generations. The mdx mice (genetic background C57BL/10) were purchased from CLEA Japan, Inc. (Tokyo, http://www.clea-japan.com) and crossed with Notch3-deficient mice. The mutation in the dystrophin gene was detected by polymerase chain reaction (PCR) analysis as described previously . To induce regeneration of skeletal muscle, we anesthetized mice, and 100 μl of CTX (10 μM in 0.9% NaCl; Sigma-Aldrich, St. Louis, MO, http://www.sigmaaldrich.com) was injected into the left tibialis anterior (TA) muscle with a 30-gauge needle. For each experiment, age- and gender-matched controls were used. The Institutional Animal Care and Use Committee of Kitasato University approved all experimental protocols. Care was taken to minimize the number of animals used, as well as their pain and suffering.
Six-month-old mdx mice received daily intraperitoneal injections of 5-bromo-2-deoxyuridine (BrdU) at 50 mg/kg body weight for four consecutive days. At 24 hours after the last injection, TA muscles were frozen in isopentane.
Histology, Fiber Number, and Fiber Size
Skeletal muscles from each genotype were frozen in isopentane cooled by liquid nitrogen. Sections with comparable areas were used for histological and immunohistochemical analysis. The sections were stained with anti-laminin antibody, and the cross-sectioned area (CSA) of individual myofibers was measured using ImageJ software (National Institutes of Health, Bethesda, MD; http://rsb.info.nih.gov/ij/).
Preparation of Primary Myoblasts
Primary myoblasts were isolated by a method described previously . Briefly, extensor digitorum longus (EDL) muscles of 3- to 4-month-old mice were incubated in Dulbecco's modified Eagle's medium (DMEM) with 0.5% collagenase type I (Sigma-Aldrich) for approximately 90 minutes at 37°C. Myofibers were then cultured in satellite cell growth medium (DMEM with 20% FBS supplemented with 10 ng/ml basic fibroblast growth factor, 2.5% 10-day-old chick embryo extract and 1,000 U/ml mouse leukemia inhibitory factor) for 3 days on Matrigel (BD Biosciences, San Diego, CA, http://www.bdbiosciences.com)-coated 24-well plates. After 3 days in culture, myofibers were aspirated from the plate and adherent cells were detached. Cells were plated onto a polystyrene tissue culture dish (BD Falcon, San Diego, CA, http://www.bdbiosciences.com) in growth medium for 30–60 minutes to remove fibroblasts. The cells that remained in suspension were transferred onto Matrigel-coated dishes. After 3–5 passages, more than 95% of cells were myogenic, as seen by double staining with anti-Pax7 and anti-MyoD antibodies. All assays with primary myoblasts were performed within passages 2–4.
Myofiber Culture and Fiber-Associated Satellite Cell Proliferation Assay
Single myofibers with associated satellite cells were isolated from 3- to 4-month-old mice. To activate the associated satellite cells, myofibers were cultured in suspension in DMEM containing 10% horse serum and 0.5% 10-day-old chick embryo extract at 37°C in 5% CO2 for up to 72 hours and fixed in 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS) for 5–60 minutes. A proliferation assay on the associated satellite cells was performed using a Click-iT EdU Alexa Fluor 594 Imaging Kit (Invitrogen, Carlsbad, CA, http://www.invitrogen.com) following the manufacturer's instructions.
After washing in PBS, samples were blocked with blocking buffer (PBS containing 10% Blocking One [Nacalai Tesque Inc., Kyoto, Japan, http://www.nacalai.co.jp/en]) and 0.05%–0.3% Triton X-100 for 30 minutes at room temperature. Samples were then incubated with the following primary antibodies diluted in blocking buffer: goat anti-Notch3 (Catalogue no. AF1308, Lot: ISY014061, 0.1–0.2 μg/ml; R&D Systems Inc., Minneapolis, MN, http://www.rndsystems.com), rabbit anti-laminin (AB19012, 5 μg/ml; Chemicon, Temecula, CA, http://www.chemicon.com), mouse anti-M cadherin (clone 12G4, 2–5 μg/ml; Abcam, Cambridge, MA, http://www.abcam.com), mouse anti-Pax7 (clone PAX7, 2 μg/ml; Developmental Studies Hybridoma Bank [DSHB], Iowa City, IA, http://dshb.biology.uiowa.edu), mouse anti-myosin heavy chain (MHC; clone MF 20, 2–5 μg/ml; DSHB), mouse anti-myogenin (clone F5D, 1 μg/ml; DSHB), rabbit anti-MyoD (M-318, 0.5–1 μg/ml; Santa Cruz Biotechnology Inc., Santa Cruz, CA, http://www.scbt.com). After incubation with primary antibodies at 4°C overnight, the samples were washed in PBS containing 0.05%–0.3% Triton X-100 and then incubated with the following secondary antibodies diluted in blocking buffer: Cy3-conjugated bovine anti-goat IgG (0.7 μg/ml; Jackson ImmunoResearch Laboratories, West Grove, PA, http://www.jacksonimmuno.com), Cy5-conjugated anti-rabbit IgG (1.5 μg/ml; Jackson ImmunoResearch Laboratories), Alexa Fluor 488-conjugated anti-rabbit IgG (1–2 μg/ml; Molecular Probes, Eugene, OR, http://probes.invitrogen.com), Alexa Fluor 488-conjugated anti-mouse IgG (1–2 μg/ml; Molecular Probes), Alexa Fluor 568-conjugated anti-mouse IgG (1 μg/ml; Molecular Probes), Alexa Fluor 350-conjugated anti-rabbit IgG (10 μg/ml; Molecular Probes). To test the integrity of the Notch3-immunoreactivity, tissues from Notch3-deficient mice were used as a negative control, and the optimal dilution of antibody and image capture conditions were determined (Supporting Information Fig. 1).
Primary myoblasts (1 × 106) were lysed in 200 μl radioimmunoprecipitation assay buffer (150 mM NaCl; 50 mM Tris-HCl, pH 8.0; 1% Nonidet P-40; 0.1% sodium dodecyl sulfate; 0.5% sodium deoxycholate) containing a cocktail of protease inhibitors (Sigma-Aldrich). Proteins (5 μg) were resolved on NuPAGE 3%–8% Tris-Acetate Gels (Invitrogen) and then transferred to a nitrocellulose membrane. The membranes were probed with rabbit polyclonal antibody against the intracellular domain of Notch3 protein (m3; 1 μg/ml; ). Equal protein loading was confirmed by reprobing the blot with an anti α-tubulin antibody (clone DM1A, 0.5 μg/ml; Sigma-Aldrich).
Satellite Cell Number
TA muscle from 4-month-old mice was frozen in isopentane cooled by liquid nitrogen. Frozen sections of 8 μm thickness were fixed in 4% PFA in PBS for 30 minutes. For anti-Pax7 staining, antigen retrieval was conducted by heating slides for 20 minutes at 90°C in 10 mM sodium citrate buffer (pH 6.0). The sections were stained with anti-Pax7 antibody and counterstained with Hoechst. Pax7-positive nuclei were counted from several random fields of view for each animal.
Primary Myoblast Proliferation and Differentiation
For proliferation assays, primary myoblasts were plated at a low density (5 × 103 cells/cm2) on Matrigel-coated 24-well plates and cultured in growth medium. The proliferation assay was performed using a disulfonated tetrazolium salt, WST-8 (Cell Count Reagent SF; Nacalai Tesque Inc., Japan) following the manufacturer's instructions. A spectrophotometer was used to measure the absorbance at 450 nm. For differentiation assays, primary myoblasts were plated at high density (1.25 × 105 cells/cm2) on Matrigel-coated dishes. The cells were cultured in differentiation medium (DMEM with 5% horse serum).
DNA Construct and Transfection
The mouse Notch3 intracellular domain (N3ICD; codons 1664-2318) expression vector (CMV-N3IC-HA) was a kind gift from Dr. U. Lendahl (Department of Cell and Molecular Biology, Karolinska Institute, Stockholm, Sweden). The pCX-EGFP expression vector containing EGFP under the control of the chicken β-actin promoter and the immediate early enhancer of cytomegalovirus was a kind gift from Dr. Okabe (University of Osaka, Japan). The hemagglutinin (HA)-tagged Notch3 IC fragment was amplified by PCR from CMV-N3IC-HA. The PCR product (N3IC-HA) was ligated into an EcoRI-digested pCX-EGFP vector to form the CAG-N3ICD-HA expression vector. This construct was sequenced to ensure that the correct orientation and reading frame were obtained. For transfection experiments, primary myoblasts were plated in growth medium, and transfected with CAG-N3ICD-HA or the control (pCX-EGFP) vector using Lipofectamine LTX reagent (Invitrogen).
Quantitative Reverse Transcription PCR
Total RNA was isolated using the RNeasy Mini kit (Qiagen) and subjected to DNase digestion following the manufacturer's instructions. The cDNA was synthesized using QuantiTect Reverse Transcription kit (Qiagen). Quantitative real-time PCR was performed using the QuantiTect SYBR Green PCR kit (Qiagen). Each sample was amplified in triplicate in a thermal cycler (MxPro-Mx3000P; Stratagene, La Jolla, CA, http://www.stratagene.com) and transcript levels were normalized to glyceraldehyde-3-phosphate dehydrogenase (Gapdh) transcript levels. Relative mRNA levels were calculated using the ddCT method. The primer pairs were obtained from Qiagen (QuantiTect Primer Assay). The catalog numbers were as follows: Hes1, QT00313537; Hey1, QT00115094; HeyL, QT00128954; Nrarp, QT00262199; Gapdh, QT01658692; Notch1, QT00156982; Notch2, QT00153496; Notch3, QT01051729; and Dll1, QT00113239.
Notch3 Deficiency Results in Remarkable Muscle Overgrowth After Repeated Muscle Injuries
We generated Notch3- and dystrophin-double deficient mice by crossing Notch3-deficient mice with mdx mice (designated mdx:Notch3−/−). The mdx mice, carrying a loss-of-function point mutation in the X-linked dystrophin gene, exhibited continuing cycles of severe muscle degeneration followed by regeneration throughout their lifetime [29, 30], which has been confirmed by intraperitoneal administration of BrdU, followed by analysis of satellite cell proliferation (Supporting Information Fig. 2).
Four weeks after birth, there were no obvious gross phenotypic alterations in mdx:Notch3−/− mice. However, at 4 months of age, we observed the mdx:Notch3−/− mice were somewhat larger than their littermate controls and a dramatic increase in the body weight of mdx:Notch3−/− mice became evident with age. This excessive increase in body weight continued up to 12 months of age (Fig. 1A). Adult mdx: Notch3−/− mice also displayed an abnormal body shape, with pronounced shoulders and hips. When the skin was removed, it became apparent that the muscles of the mdx:Notch3−/− mutants were much larger than those of the control mice (Fig. 1B). The increase in skeletal muscle mass appeared to be widespread throughout the body. Individual muscles taken from mdx:Notch3−/− mice weighed approximately 1.3–1.6 times more than those from littermate controls (Fig. 1C). Examination of H&E-stained sections of mdx:Notch3−/− TA muscles revealed that mdx:Notch3−/− mice exhibited similar phenotypes typically observed in mdx mice, including centrally located nuclei, rare pockets of myonecrosis, inflammation, and increased variability in fiber size (Fig. 1D).
To clarify whether the muscle overgrowth observed in the mdx:Notch3−/− mice occurred because of repeated muscle injury, we performed injury-induced regeneration experiments using Notch3-deficient mice. Notch3-deficient mice matured normally, as did wild-type controls during postnatal development (Supporting Information Fig. 3A). The average weight of quadriceps at postnatal day (P) 10 of Notch3-deficient mice was also at the same level as that of control mice (Supporting Information Fig. 3B). To induce muscle injury, the hind limb muscle was exposed by incision, and CTX was administrated at three different intervals (Fig. 2) into the left TA muscles of 3- to 4-month-old mice. At 2–3 months after CTX administration, TA muscles were isolated and analyzed. As shown in Figure 2A, muscles from Notch3-deficient mice were 1.3-fold heavier than those of control mice when TA muscles received a single dose of CTX. When CTX was given intramuscularly three times at intervals of 25 days, TA muscles recovered from Notch3-deficient mice weighed about 1.6-fold more than those from control mice (Fig. 2B). When CTX was administered intramuscularly seven times at intervals of 7 days, the weight of TA muscles in control mice was markedly decreased, but the weight of TA muscles in Notch3-deficient mice was similar to uninjured muscles and twice as heavy as those from control mice (Fig. 2C). These results coincided well with the results obtained from mdx:Notch3−/− mice, and we concluded that a loss of Notch3 resulted in a striking muscle overgrowth after repeated muscle regeneration. To determine whether the muscle overgrowth was caused by hyperplasia or hypertrophy, we analyzed the histological sections of the injured TA muscles (Fig. 2D) to estimate total number and mean CSA of myofibers at the widest region of the muscles. As shown in Figure 2E, the mean CSA of the myofibers was reduced according to the frequency of muscle injuries, and no significant difference was detected between Notch3-deficient and control mice except when CTX was administered seven times. In contrast, the total number of myofibers was consistently increased in the Notch3-deficient mice compared with control mice (Fig. 2F), suggesting that the muscle overgrowth observed in Notch3-deficient mice was mainly due to hyperplasia.
Notch3 Is Expressed in Quiescent and Activated Satellite Cells
Because the muscular overgrowth in Notch3-deficient mice became apparent only when they were subjected to repeated muscle injuries, we focused our attention on the satellite cells and examined the expression pattern of Notch3 in the satellite cells in vivo and in vitro by Western blot and immunocytochemical methods. The primary myoblasts were isolated from single EDL myofibers, and the expression of Notch3 was determined by Western blot. As shown in Figure 3A1, proliferating myoblasts in growth medium expressed Notch3 at relatively high levels. When myoblasts were induced to differentiate, the Notch3 expression continued for 24 hours and was quickly downregulated (Fig. 3A1). Similar results were obtained in quantitative reverse transcription PCR experiments (Fig. 3A2). We also conducted immunocytochemical analysis on individual myoblasts in culture. Almost all of the myoblasts in growth medium were found to be uniformly expressing both Pax7 and MyoD (Fig. 3B2). Strong Notch3 expression was detected in about 48% of cultured cells with the remaining 52% of cells expressing low levels of Notch3 (Fig. 3B1). When the myoblasts were cultured in differentiation medium, strong Notch3 expression was observed in the myoblasts with relatively high levels of Pax7, whereas Notch3 expression was hardly detected in the cells, which weakly expressed Pax7 (Fig. 3C). Notch3 and myogenin were expressed in a mutually exclusive manner (Fig. 3D). These results indicated that Notch3 was expressed in nearly half the population of the primary myoblasts.
We analyzed the expression of Notch3 on histological sections of muscle tissues. Frozen sections of TA muscle from 3-month-old mice were costained with antibodies reactive to Notch3, laminin, and M-cadherin, a marker for quiescent satellite cells. The result shown in Figure 4A1 demonstrated Notch3 expression in mononucleated cells located beneath the basal lamina, which is consistent with expression in quiescent satellite cells. In addition, M-cadherin-expressing quiescent satellite cells were observed to express Notch3 (Fig. 4A2–4A4). We also examined the expression of Notch3 in satellite cells attached to individually isolated myofibers. The myofibers were isolated from the EDL of 3-month old mice, immediately fixed and immunostained for Notch3 and Pax7. The expression of Notch3 was observed in the Pax7-expressing myofiber-attached cells (Fig. 4B1–4B3), confirming that Notch3 was expressed in quiescent satellite cells within resting adult skeletal muscles. However, the expression of Notch3 was limited to about half the population of the myofiber-attached cells and Notch3 expression was barely detected in the remaining cells (Fig. 4B4), indicating that quiescent satellite cells were a heterogeneous population with respect to Notch3 expression.
When the isolated myofibers were cultivated in suspension, most of the attached satellite cells underwent cell division during the first 2 days and eventually formed 5–10 cell clusters within 3 days (Fig. 4C1). At this time, strong immunostaining for Notch3 was detected in many of these clustered cells. Nearly, all of the Notch3-expressing cells were also expressing Pax7 (Fig. 4C2–4C6). The Notch3-negative cells in the cluster always expressed MyoD and most of them also expressed myogenin, suggesting that they were committed to myogenic differentiation .
From these results, we concluded that Notch3 was expressed in Pax7+/MyoD−-quiescent satellite cells and Pax7+/MyoD+-activated satellite cells, but the expression was restricted to around half the population of each cell type.
Notch3 Deficiency Leads to an Increase in Quiescent Satellite Cells and Accelerates Proliferation of Activated Satellite Cells
To explore the reason why muscle overgrowth occurred in repeatedly injured muscles of Notch3-deficient mice, we examined phenotype alteration in the satellite cells of these mice. TA muscle sections from 4-month-old mice were double-stained with antibodies against Pax7 and laminin, and the number of quiescent satellite cells was counted on the histological sections (Fig. 5A). As shown in Figure 5B, the number of quiescent satellite cells was 1.4 times higher in the TA muscles of Notch3-deficient mice compared with control mice. To further confirm this result, we counted the number of satellite cells attached to freshly isolated EDL myofibers of 4-month-old mice. Again, the number of quiescent satellite cells attached to myofibers of Notch3-deficient mice was 1.4 times greater than in controls (Fig. 5C).
We examined whether the population of proliferating satellite cells was affected by Notch3 deficiency. Individual myofibers were isolated from the EDL of 4-month-old Notch3-deficient and control mice and cultured in suspension with 5-ethynyl-2′-deoxyuridine (EdU) for 48 hours, then stained for EdU and Pax7. After 48 hours in culture, the majority of satellite cells were found to be activated and entering the cell cycle, as confirmed by EdU incorporation (Fig. 5D). The number of EdU-positive satellite cells was twofold greater in Notch3-deficient myofibers (Fig. 5D).
The proliferation rate of primary-cultured myoblasts was also determined. Growth of the primary cultured myoblasts was assayed using the WST-8 method. As shown in Figure 6A, the primary myoblasts isolated from Notch3-deficient mice were found to proliferate at a higher rate than those from control mice, and this elevated proliferation rate became apparent after 3 days of culture. We next examined the differentiation potential of Notch3-deficient primary myoblasts. To eliminate the effect of proliferation rate, primary myoblasts were plated at a high density in differentiation medium. After 48 hours in differentiation medium, a large number of cells in both controls and mutants were positive for MHC (Fig. 6B). Measurement of the fusion index revealed that there was no significant difference between the myoblasts from control and Notch3-deficient mice (Fig. 6C).
To further confirm the role of Notch3 in proliferation, the Notch3-deficient myoblasts were transfected with a constitutive active form of Notch3 (N3ICD). These cells were cultivated with EdU. The ratio of EdU-positive cells was greatly reduced when the cells were transfected with N3ICD (Fig. 6D and 6E), indicating that Notch3 blocked proliferation of the primary-cultured myoblasts.
To determine whether Notch target genes were affected by Notch3, we examined the expressions of Hes1, Hey1, HeyL, and Nrarp and found that only the expression of Nrarp was greatly elevated by N3ICD transfection, and the expression of Hes1, Hey1, and HeyL remained at relatively low levels (Fig. 6F). Next, we tested whether expression of Notch target genes might be downregulated by Notch3 deficiency. However, as shown in Figure 6G, no statistically significant differences in the expression of Notch target genes were detected between Notch3-deficient and wild-type myoblasts.
Notch3 Deficiency Leads to an Increase in Quiescent Satellite Cells During Self-Renewal In Vitro
We analyzed the proliferation and differentiation of the Notch3-deficient satellite cells located on isolated myofibers. The myofibers were isolated from EDL muscles of either Notch3-deficient or control mice and then cultured in suspension for 3 days followed by immunostaining for Pax7 and MyoD. The satellite cells attached to the myofibers proliferated to form cell clusters during the cultivation period. The satellite cells prepared from Notch3-deficient mice formed larger clusters than those from control mice (Fig. 7A, 7B). Within the clusters on the Notch3-deficient myofibers, we found a number of quiescent satellite cells (Pax7+/MyoD−), indicating self-renewal of satellite cells can occur without Notch3 (Fig. 7A). The ratio of quiescent satellite cells to total clustered cells was increased in Notch3-deficient mice and the ratio of activated satellite cells to total clustered cells was decreased when compared with controls (Fig. 7C), suggesting Notch3 might negatively regulate self-renewal of activated satellite cells. This result was further confirmed by estimating the population of Pax7-positive cells after repeated muscle injury. As shown in Figures 7D–7E, after seven intramuscular administrations of CTX, the number of Pax7-positive cells were increased in Notch3-deficient mice, suggesting that the satellite cell pool had increased after every cycle of muscle degeneration and regeneration.
Skeletal muscle regeneration in response to injury requires a well-organized orchestration of multistep processes including activation and proliferation of satellite cells, subsequent differentiation into multinucleated myotubes to form new muscle tissue, and a return to quiescent satellite cells to replenish the stem cell pool. These processes are so important for maintaining muscle tissues that they need to be tightly controlled by various regulators. It has been well documented that the Notch signaling pathway, particularly the Notch1 pathway, plays critical roles in the regulation of satellite cells during muscle regeneration. One of the primary roles of Notch1 is to maintain the activated satellite cells in a highly proliferative state and prevent myogenic differentiation. Knockdown of Notch1 expression by RNAi has been shown to inhibit proliferation of satellite cells and promote their differentiation into myogenic cells . Kuang et al. interrupted Notch signal transduction by treatment with a γ-secretase inhibitor for 3 days on freshly isolated myofibers, resulting in a significantly reduced number of Pax7+/MyoD− self-renewing satellite cells , demonstrating that Notch signaling was essential not only for proliferation but also for self-renewal of satellite cells.
The aim of this study was to clarify the role of the Notch3 pathway during skeletal muscle regeneration by analyzing phenotypic alterations that occurred in the Notch3-deficient mice. Because Notch3-deficient mice developed normally and no morphological alterations were observed in the muscle tissues of Notch3-deficient mice (Supporting Information Fig. 3), it seemed unlikely that Notch3 has some role in myogenesis during mouse development. We report here that adult mice lacking Notch3 exhibited uncontrolled muscle growth eventually inducing hyperplasia only when they suffered repeated muscle injuries (Figs. 1, 2). This result, indicating the importance of Notch3 in the regulation of muscle regeneration, was contrary to previous results where it was shown that the Notch signaling pathway was essential for skeletal muscle regeneration [13, 14].
We have found, through analyses of Notch3-deficient satellite cells, that the Notch3 pathway seemed functionally distinct from Notch1. Notch3-deficient satellite cells proliferated faster than control cells and forced expression of N3ICD resulted in reduced proliferation, suggesting Notch3 suppressed proliferation of satellite cells (Fig. 6A, 6D, 6E). By analyzing satellite cells attached to individually isolated myofibers, we have shown that the ratio of Pax7+/MyoD− self-renewing cells was increased in Notch3-deficient mice as compared with control mice (Fig. 7C), indicating that Notch3 suppressed self-renewal of satellite cells. Treatment with a γ-secretase inhibitor, as conducted by Kuang et al., interrupted all Notch pathways including the Notch3 pathway, whereas in our experiments only the Notch3 pathway was blocked, making it difficult to directly compare our data with that of Kuang et al. . It seems likely that Notch3 acts in opposition to Notch1 in the regulation of activated satellite cells. We speculate that the processes of both proliferation and self-renewal in the activated satellite cells are positively regulated by Notch1, whereas Notch3 negatively regulates both of these processes. It was interesting to note that activated satellite cells were a heterogeneous population with 48% of cells expressing Notch3 and Notch3 expression hardly detected in the remaining 52% of cells (Fig. 3B1). In contrast, Notch1 was expressed in nearly all satellite cells . We speculate that the group of activated satellite cells expressing both Notch1 and Notch3 is committed to proceed toward the process of self-renewal. Proliferation and self-renewal of satellite cells might be under tight control of Notch1 and Notch3 so as not to generate too many stem cells. Because Pax7+/MyoD− self-renewing cells appeared in the Notch3-deficient myofibers in culture (Fig. 7A–7C), it is apparent that self-renewal of satellite cells could occur without Notch3. Without Notch3, however, muscle overgrowth was induced by repeated muscle injury, suggesting the critical role of Notch3 in controlling the number of satellite cells.
The mechanisms of the various effects of Notch1 and Notch3 on satellite cell proliferation and self-renewal have not been fully elucidated. Notch1 and Notch2 have the highest homology with each other, while Notch3 is structurally divergent, lacking the transactivation domain that is seen in Notch1 and Notch2 . By analyzing certain cell lines and transgenic mice, previous studies demonstrated that Notch3 was a poor activator of Notch target genes and repressed Notch1-mediated target gene activation [31–33]. If Notch3 acts as a Notch1 repressor in satellite cells, lack of Notch3 should lead to an increase in Notch1 activity, causing enhanced proliferation of satellite cells, similar to the myoblast constitutively expressing Notch1 . Because the expression levels of Notch1 (and also Notch2 and Dll1) in Notch3−/− myoblasts were not significantly different from those in wild-type myoblasts (Supporting Information Fig. 4A–4B), the expression of Notch1 might be independent of Notch3. Alternatively, it has been strongly suggested that Notch3 exerts an effect through regulating a target gene of Notch1. N3ICD overexpression in Notch3−/− myoblasts resulted in a dramatic increase in the expression of Nrarp, although expression of other Notch target genes remained at relatively low levels (Fig. 6F). Nrarp, a member of the Delta-Notch synexpression group and encoding a small protein containing two ankyrin repeats, has been supposed to act as a negative feedback regulator of Notch signaling that attenuates ICD-mediated transcription . Therefore, Nrarp might be a target gene of Notch3 in the myoblasts and Notch3 might act as a negative regulator of the Notch signaling pathway by activating Nrarp. However, this speculation might be too simple because no significant difference in expression levels of Nrarp was detected between Notch3−/− and wild-type myoblasts (Fig. 6G). There might exist a much more complicated regulatory system in myoblasts when they proliferate and differentiate. For example, it has been proposed that activation of the Notch signaling pathway, stimulated by Delta-1 ligand and Notch3 receptor interactions, plays a positive role in the self-renewal of satellite cells . Additional studies are needed to determine whether there are unique combinations of receptor-ligand interactions and/or different signal-transduction cascades downstream of Notch1 and Notch3 in satellite cells.
A remarkable feature of skeletal muscle is its ability to undergo rapid repair in the face of acute damage, with a constant pool of satellite cells maintained by self-renewal. One of our interesting findings is that the population of quiescent satellite cells is increased in resting muscles of Notch3-deficient mice, suggesting that Notch3 signaling is also critical for controlling the pool size of satellite cells under normal physiological conditions. The regenerative ability of skeletal muscles was strengthened by repeated muscle injuries in Notch3-deficient mice, probably corresponding to the increase in Pax7+ satellite cells present in muscle tissues (Fig. 7D–7E). Therefore, it is interesting to speculate that transplantation of Notch3−/− satellite cells may improve the efficiency of muscle regeneration, and that Notch3−/− satellite cells effectively contribute to the satellite stem cell compartment. If this is the case, manipulation of Notch3 signaling in satellite cells might be favorable for cell transplantation therapy for degenerative muscular diseases such as Duchenne muscular dystrophy.
In conclusion, the present results demonstrate Notch3 as an important regulator for muscle repair. The data indicate that Notch3 negatively regulates both proliferation and self-renewal of activated satellite cells after muscle injury. These results provide insight into the mechanisms of self-renewal and homeostatic regulation of satellite cells during muscle regeneration.
This work was supported in part by grants from the Ministry of Health, Labor and Welfare of Japan and from the Ministry of Education, Culture, Sports, Science and Technology of Japan.
DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
The authors indicate no potential conflicts of interest.