MCB Accepts, published online ahead of print on 17 October 2011 Mol. Cell. Biol. doi:10.1128/MCB.05857-11 Copyright © 2011, American Society for Microbiology and/or the Listed Authors/Institutions. All Rights Reserved. 1 O regulates skeletal muscle progenitor differentiation through PI3K/AKT 2 2 signaling 3 4 Amar J. Majmundar1,2, Nicolas Skuli1,2, Rickson C. Mesquita4, Meeri N. Kim4, Arjun G. 5 Yodh4, Michelle Nguyen-McCarty1,2, and M. Celeste Simon1,2,3,* D o 6 w n 7 1Abramson Family Cancer Research Institute, lo a d e 8 2Department of Cell and Developmental Biology, and d f r o 9 3Howard Hughes Medical Institute, Perelman School of Medicine at the University of m h 10 Pennsylvania, Philadelphia, PA 19104, USA tt p : / / 11 4Department of Physics & Astronomy, University of Pennsylvania, Philadelphia, PA m c b 12 19104, USA .a s m 13 . o r g 14 Running title: O2-dependent PI3K/AKT signaling in skeletal muscle differentiation o/ n 15 A p r 16 *Corresponding author: M. Celeste Simon il 4 17 Abramson Family Cancer Research Institute , 2 18 University of Pennsylvania School of Medicine 0 1 19 BRB II/III, Room 456 9 20 421 Curie Blvd b y 21 Philadelphia, PA 19104 g 22 Telephone: (215) 746-5532 u e 23 Fax: (215) 746-5511 s t 24 E-mail: [email protected] 25 26 Introduction, Results and Discussion word count – 5179 27 Material and Methods word count - 1282 28 1 29 ABSTRACT: 30 Skeletal muscle stem/progenitor cells, which give rise to terminally differentiated 31 muscle, represent potential therapies for skeletal muscle disease. Delineating the 32 factors regulating these precursors will facilitate their reliable application in human 33 muscle repair. During embryonic development and adult regeneration, skeletal muscle D o 34 progenitors reside in low O2 environments before local blood vessels and differentiated w n lo 35 muscle form. Prior studies established that low O levels (hypoxia) maintain muscle 2 a d e 36 progenitors in an undifferentiated state in vitro, although it remained unclear if d f r o 37 progenitor differentiation is coordinated with O2 availability in vivo. In addition, the m h 38 molecular signals linking O2 to progenitor differentiation are incompletely understood. ttp : / / 39 Here we show that the muscle differentiation program is repressed by hypoxia in vitro m c b 40 and ischemia in vivo. Surprisingly, hypoxia can significantly impair differentiation in the .a s m 41 absence of Hypoxia Inducible Factors (HIFs), the primary developmental effectors of O . . 2 o r g 42 In order to maintain the undifferentiated state, low O2 levels block the PI3K/AKT o/ n 43 pathway in a predominantly HIF1α-independent fashion. O deprivation affects AKT A 2 p r 44 activity by reducing IGF-I Receptor sensitivity to growth factors. We conclude that AKT il 4 , 2 45 represents a key molecular link between O and skeletal muscle differentiation. 0 2 1 9 b y g u e s t 2 46 INTRODUCTION: 47 Skeletal muscle damage or loss arises in a range of diseases, including inherited 48 Muscular Dystrophies, Critical Limb Ischemia in Peripheral Arterial Disease (PAD), and 49 aging-related sarcopenia (4, 10, 22, 26, 33). Weakened and aberrant muscles 50 contribute significantly to the morbidity and mortality of patients suffering from these D o 51 illnesses (4, 10, 22, 26, 33). Skeletal muscle stem/progenitor cells, which give rise to w n lo 52 embryonic and adult muscle (34, 32), represent potential therapies for human skeletal a d e 53 muscle disease (22, 57). Delineating the pathways controlling the maintenance and d f r o 54 differentiation of these precursors will facilitate their reliable application in muscle repair m h 55 (22, 57). tt p : / / 56 In adult mammals, skeletal muscle stem cells—“satellite” cells—reside in a niche m c b 57 enveloped by differentiated muscle fibers and a layer of basement membrane. .a s m 58 Quiescent satellite cells, expressing the transcription factor PAX7, become activated . o r g 59 after muscle injury and terminally differentiate into new multi-nucleated skeletal muscle / o n 60 fibers (34, 32). These processes depend on several transcription factors known as A p r 61 muscle regulatory factors, or “MRFs”: MYF5, MYOD and Myogenin (Fig. 1A) (34, 32). il 4 , 2 62 MYF5 and MYOD are co-expressed with PAX7 in activated satellite cells but possess 0 1 9 63 distinct functional roles (Fig. 1A) (34, 32). While MYF5 is important for muscle b y g 64 progenitor proliferation, MYOD is required for subsequent differentiation of these u e s t 65 precursors (Fig. 1A) (34, 32). MYOD and its target Myogenin stimulate terminal 66 differentiation through the activation of genes expressed in mature muscle (e.g. Myosin 67 heavy chain) (34, 32). 3 68 Skeletal muscle differentiation, or myogenesis, is tightly regulated and responds 69 to environmental cues (34, 32). For example, Insulin and Insulin-like Growth Factors 70 (IGFs) can act upon cultured muscle progenitors, termed myoblasts, and stimulate their 71 terminal differentiation (18). In agreement with these findings, IGF-I has been shown to 72 promote embryonic skeletal muscle development (36) and adult muscle regeneration in D o 73 vivo (48). A key pathway activated by Insulin and IGFs is Phosphatidylinositol 3-Kinase w n lo 74 (PI3K)/mammalian Target of Rapamycin Complex 2 (mTORC2)/AKT. These molecules a d e 75 are required downstream of Insulin/IGFs for muscle differentiation in vitro (13, 14, 28, d f r o 76 29, 41, 59–62). For example, it was recently demonstrated that mTORC2 component m h 77 RICTOR regulates terminal myoblast differentiation upstream of AKT (54). AKT, tt p : / / 78 furthermore, has been shown to promote embryonic muscle development and adult m c b 79 regeneration in vivo (44, 47). .a s m 80 Skeletal muscle progenitors also respond to the availability of local nutrients such . o r g 81 as glucose (20) and molecular oxygen (O2) (15, 25, 51, 64). In fact, skeletal muscle is o/ n 82 marked by reduced O availability, or hypoxia, during both development and disease. A 2 p r 83 Embryonic somites, where early skeletal muscle progenitors reside, exhibit increased il 4 , 2 84 expression of hypoxic markers (e.g. Hypoxia Inducible Factor 1α) prior to the formation 0 1 9 85 of local blood vessels and embryonic muscle (49, 50). In addition, adult skeletal muscle b y g 86 exhibits severe pathological hypoxia in Peripheral Arterial Disease (4, 7, 24, 26, 33, 45). u e s t 87 Hind limb ischemia—or insufficient blood supply—acutely leads to tissue damage in 88 mouse models of this disease (7, 24, 45). In otherwise healthy animals, skeletal muscle 89 progenitors as well as injured muscle fibers experience O and nutrient deprivation until 2 90 neo-vascularization restores perfusion to the tissue (7, 24, 45). As blood flow returns, 4 91 newly generated fibers re-constitute affected muscle groups (7, 24, 45). Thus, in both 92 embryonic development and adult regeneration, skeletal muscle stem/progenitor cells 93 reside in a hypoxic microenvironment before the formation of local blood vessels and 94 terminally differentiated muscle (7, 24, 45, 49, 50). In severe cases of PAD, however, 95 vascular insufficiency and muscle damage can persist chronically (4, 26, 33). D o 96 O2 may exert a developmental function in these contexts, for low O2 conditions w n lo 97 are known to maintain skeletal myoblasts in an undifferentiated state in vitro (15, 25, 51, a d e 98 64). This suggests that in the hypoxic microenvironment of developing or regenerating d f r o 99 skeletal muscle, O2-dependent pathways may constrain progenitor differentiation until m h 100 there is ample blood supply, thereby conserving the stem/progenitor pool for tt p : / / 101 appropriate circumstances for growth. However, this has not been formally tested in m c b 102 vivo. .a s m 103 While it is established that O regulates myoblast differentiation, the molecular . 2 o r g 104 mechanisms are incompletely understood. In other tissues, Hypoxia Inducible Factors / o n 105 (HIFs) represent the principle developmental effectors of O availability (55). These A 2 p r 106 transcription factors are comprised of an O2-labile α-subunit and O2-independent β- il 4 , 2 107 subunit (37). In hypoxic conditions, the two biologically relevant α-subunits—HIF1α and 0 1 9 108 HIF2α—are stabilized and form dimers with HIF1β to activate the expression of b y g 109 numerous genes (37). The role of the HIFs in myogenesis has been controversial. In u e s t 110 one study, ectopic HIF1α did not affect myoblast differentiation in ambient O conditions 2 111 (64). Another claimed that hypoxia inhibits muscle progenitor differentiation through a 112 novel complex between HIF1α and NOTCH (25). However, neither report showed if 113 endogenous HIFα was essential for the effects of hypoxia on myogenesis (25, 64). 5 114 In the present study, we employed animal and cell culture models to determine if 115 O can influence the myogenic program in vivo and to delineate which factors modulate 2 116 skeletal muscle progenitors in response to low O . We show that low O inhibits muscle 2 2 117 progenitor differentiation and myogenic regulatory factor expression in vitro. In a murine 118 model of PAD, MRF expression is similarly affected by ischemia in vivo. We then D o 119 pursued the mechanism(s) linking O2 to muscle differentiation. Surprisingly, while w n lo 120 HIF1α deficiency has modest effects on myoblast differentiation, hypoxiacan a d e 121 significantly modulate progenitor differentiation in the absence of HIF1α. We d f r o 122 determined that hypoxia regulates muscle differentiation through predominantly HIF1α- m h 123 independent effects on PI3K/mTORC2/AKT signaling. Low O2 levels block PI3K/AKT ttp : / / 124 signaling by reducing IGF-I Receptor sensitivity to growth factors, and restoration of m c b 125 PI3K/AKT activity is sufficient to rescue myoblast differentiation. These findings .a s m 126 suggest that HIF-independent factors may regulate the capacity of progenitors to repair . o r g 127 skeletal muscle in settings of hypoxic/ischemic injury. / o n A p r il 4 , 2 0 1 9 b y g u e s t 6 128 MATERIALS AND METHODS: 129 Cell Culture 130 C2C12 myoblasts (ATCC, CRL-1772) were propagated in 20% FBS in DMEM. To 131 evaluate differentiation, myoblasts were grown to 80-90% confluency and switched to 132 2% horse serum in DMEM. D o 133 Primary mouse myoblasts were isolated from gastrocnemius muscles of 8-12 w n lo 134 week old C57BL/6 as described in (56). Briefly, calf muscles were dissected, minced, a d e 135 and digested with 0.2% type II collagenase. Fibers were subsequently triturated, d f r o 136 washed, and further digested in 1% dispase/0.05% type II collagenase. Satellite cells m h 137 were displaced from fibers by triturating through an 18 gauge needle. Cells were further tt p : / / 138 washed, decanted through a 40 micron strainer, and plated onto collagen coated m c b 139 dishes. Primary cells were expanded in 20% FBS and 10 ng/mL rhFGF (Promega) in .a s m 140 F10/Hams for 7-9 days. For differentiation assays, 7.5 x 103 cells were plated in a 24 . o r g 141 well plate overnight, and the media was changed to 5% horse serum in DMEM. / o n 142 Low oxygen conditions were achieved in a Ruskinn In vivO 400-work station. A 2 p r 143 These inhibitors were used to modulate PI3K and mTORC activities: 10 μM LY294002, il 4 , 2 144 40 nM rapamycin and 250 nM Torin1 (gift from D. Sabatini laboratory). Recombinant 0 1 9 145 IGF-I and NOTCH ligand fusion protein Fc-JAG1 were purchased from R&D sytems. γ- b y g 146 secretase inhibitors DAPT (10 μM) and L-685,458 (1μM) were purchased from Sigma- u e s t 147 Aldrich. 148 149 Virus Preparation 7 150 For shRNA-mediated knockdown of Hif1α and Pten, lentiviral particles bearing pLKO.1 151 shRNA plasmids were generated in HEK-293T cells. 293T cells were transfected 152 overnight with pLKO.1 empty vector, non-specific shRNA, or target-specific shRNA and 153 viral packaging plasmids, according to the Fugene reagent protocol (Roche). The 154 following shRNA pLKO.1 plasmids were employed: pLKO.1 empty (Addgene 8543), D o 155 pLKO.1 scrambled shRNA (Addgene 1864), pLKO.1 Hif1α shRNA (TRCN0000054451), w n lo 156 pLKO.1 Pten shRNA (TRCN0000028991), VSV-G, pMDLG, pRSV-rev. Media was a d e 157 recovered from cultures at 40 hours post-transfection, and virus in supernatant was d f r o 158 concentrated using 10kDa Amicon Ultra-15 Centrifugal Filter units (Millipore). m h t 159 Myoblasts were incubated for with 1/10 of concentrated supernatant and 8 ug/mL tp : / / m 160 polybrene in order to achieve 90-100% transduction efficiency. Because pLKO.1 c b . 161 shRNA plasmids contain a puromycin resistance gene, transduction efficiency was a s m 162 evaluated by puromycin selection. Cells were used for assays at 3 days post- .o r g / 163 transduction. o n A 164 For ectopic expression of myristoylated AKT (gift from Anthony Chi and Avinash p r 165 Bhandoola), retroviral particles bearing migR expression plasmids were generated in il 4 , 2 166 HEK-T293 cells as described above. Viral supernatant was concentrated, as described 0 1 9 167 above, and administered to myoblasts. Myoblasts were transduced, as described b y g 168 above, with 1/10 of concentrated supernatant in order to achieve 80-90% transduction u e s t 169 efficiency. Because migR plasmids facilitate co-expression of GFP, transduction 170 efficiency was evaluated by GFP positivity by IF. Cells were used for assays at 3 days 171 post-transduction. 172 8 173 siRNA transfection 174 For siRNA-mediated knockdown of Hif1α, C2C12 cells were treated with siRNA 175 duplexes (100 nM) according to the HiPerfect protocol (QIAGEN) for 24 hours. After 48 176 hours, cells were changed to differentiation conditions. The following duplexes were 177 used: HIF1α targeting siRNA H1 (SI00193011), HIF1α targeting siRNA H4 D o 178 (SI00193032), and Negative Control siRNA (SI03650325). w n lo 179 a d e 180 Quantitative RT-PCR (qRT-PCR) d f r o 181 Total RNA was isolated from cells using the Trizol reagent protocol (Invitrogen) and m h 182 from skeletal muscle tissue using the RNAeasy Mini kit (QIAGEN). mRNA was reverse tt p : / / 183 transcribed using High Capacity RNA-to-cDNA kit (Applied Biosystems). Transcript m c b 184 expression was evaluated by quantitative PCR of synthesized cDNA using an Applied .a s m 185 Biosystems 7900HT Sequence Detection System. Target cDNA amplification was . o r g 186 measured using Taqman primer/probe sets (Applied Biosystems) for Hif1α, Epas1, / o n 187 MyoD, Myogenin, Pgk1, Hey1, Hey2, HeyL, Hes1, Mxi1, and 18S. A p r 188 il 4 , 2 189 Western Blot Analysis 0 1 9 190 Whole cell and whole tissue lysates were prepared in RIPA buffer (150 mM NaCl/1% b y g 191 NP40/50 mM Tris pH 8.0/0.1% SDS/0.5% NaDeoxycholate). Proteins were u e s t 192 subsequently separated by SDS-PAGE and transferred to nitrocellulose membranes. 193 Membranes were probed using the following antibodies: rabbit anti-HIF1α (Cayman), 194 mouse anti-MYOD (Novus), mouse anti-Myogenin (Santa Cruz), rabbit anti-Myogenin 195 (Novus Biologicals), mouse anti-MHC (MF-20, DSHB), rabbit anti-β-Tubulin (Cell 9 196 Signaling), rabbit anti-PARP (Cell Signaling), rabbit anti-AKT (Cell Signaling), rabbit 197 anti-P-AKT S473 (Cell Signaling), rabbit anti-P-AKT T308 (Cell Signaling), rabbit anti-P- 198 GSK3α/β S21/S9 (Cell Signaling), rabbit anti-GSK3β (Cell Signaling), rabbit anti-P- 199 FOXO1/3A (Cell Signaling), rabbit anti-P-P70 S6K (Cell Signaling), rabbit anti-P70 S6K 200 (Cell Signaling), rabbit anti-P-S6 S240/S244 (Cell Signaling), rabbit anti-S6 (Cell D o 201 Signaling), rabbit anti-P-IGF-IRβ Y1135 (Cell Signaling), rabbit anti-IGF-IRβ (Cell w n lo 202 Signaling), rabbit anti-P-IRS1 S636/S639 (Cell Signaling), rabbit anti-P-IRS1 S307 (Cell a d e 203 Signaling), rabbit anti-P-IRS1 S612 (Cell Signaling), rabbit anti-IRS1 (Cell Signaling), d f r o 204 rabbit anti-IRS2 (Cell Signaling), rabbit anti-P-MEK1/2 S217/S221 (Cell Signaling), m h 205 rabbit anti-MEK1/2 (Cell Signaling), rabbit anti-P-ERK1/2 T202/Y204 (Cell Signaling), tt p : / / 206 rabbit anti-ERK1/2 (Cell Signaling), rabbit anti-PERK (Rockland), rabbit anti-XBP1 m c b 207 (Santa Cruz), rabbit anti-CHOP (Santa Cruz) and rabbit anti-P-RICTOR S1235 (Cell .a s m 208 Signaling). Densitometry was performed using NIH ImageJ software. Representative . o r g 209 Western blot images of multiple independent experiments are presented. / o n 210 A p r 211 Femoral Artery Ligation (FAL) Studies il 4 , 2 212 In 8-12 week old mice (maintained on a mixed B6;129 background), hind limb ischemia 0 1 9 213 was induced by ligating the left femoral artery as previously described (40). Briefly, the b y g 214 femoral artery was exposed at the hip and separated from the femoral vein and nerve. u e s t 215 Silk suture was passed under the artery and tied to occlude it. Limb perfusion 216 measurements were taken before surgery, immediately following surgery, and 48 hours 217 later using Diffuse Correlation Spectroscopy (40). Diffuse Correlation Spectroscopy 218 (DCS) measurements were performed using a home-built instrument with two 10