Smad signaling pathway by short decorin-derived peptides

Experimental Cell Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Contents lists available at ScienceDirect

Experimental Cell Research journal homepage: www.elsevier.com/locate/yexcr

Research Article

Inhibition of the myostatin/Smad signaling pathway by short decorin-derived peptides Nelly El Shafey a, Mickaël Guesnon a, Françoise Simon b, Eric Deprez b, Jérémie Cosette c, Daniel Stockholm c, Daniel Scherman a, Pascal Bigey a, Antoine Kichler a,d,n a

Unité de Technologies Chimiques et Biologiques pour la Santé, CNRS UMR8258-Inserm, U1022 – Université Paris Descartes, Chimie ParisTech, 75006 Paris, France Laboratoire de Biologie et Pharmacologie Appliquée, ENS Cachan, UMR8113 CNRS, IDA FR3242, 94230 Cachan, France c Inserm, UMR 951, Université d’Evry Val d’Essonne, Genethon, 91000 Evry, France d Laboratoire de Conception et Application de Molécules Bioactives UMR7199 CNRS-Université de Strasbourg, LabEx Medalis, Faculté de Pharmacie, 67401 Illkirch, France b

ar t ic l e i nf o

a b s t r a c t

Article history: Received 13 November 2015 Received in revised form 27 January 2016 Accepted 31 January 2016

Myostatin, also known as growth differentiation factor 8, is a member of the transforming growth factorbeta superfamily that has been shown to play a key role in the regulation of the skeletal muscle mass. Indeed, while myostatin deletion or loss of function induces muscle hypertrophy, its overexpression or systemic administration causes muscle atrophy. Since myostatin blockade is effective in increasing skeletal muscle mass, myostatin inhibitors have been actively sought after. Decorin, a member of the small leucine-rich proteoglycan family is a metalloprotein that was previously shown to bind and inactivate myostatin in a zinc-dependent manner. Furthermore, the myostatin-binding site has been shown to be located in the decorin N-terminal domain. In the present study, we investigated the anti-myostatin activity of short and soluble fragments of decorin. Our results indicate that the murine decorin peptides DCN48-71 and 42-65 are sufficient for inactivating myostatin in vitro. Moreover, we show that the interaction of mDCN48-71 to myostatin is strictly zinc-dependent. Binding of myostatin to activin type II receptor results in the phosphorylation of Smad2/3. Addition of the decorin peptide 48-71 decreased in a dose-dependent manner the myostatininduced phosphorylation of Smad2 demonstrating thereby that the peptide inhibits the activation of the Smad signaling pathway. Finally, we found that mDCN48-71 displays a specificity towards myostatin, since it does not inhibit other members of the transforming growth factor-beta family. & 2016 Published by Elsevier Inc.

Keywords: Decorin Small leucine rich proteoglycans Myostatin Peptide TGF-β superfamily

1. Introduction Despite adequate nutritional intake, the loss of skeletal muscle mass is observed in different pathological and non-pathological situations: in sarcopenia resulting from aging, or cachexia which is a wasting syndrome observed in patients with cancer, AIDS, chronic obstructive lung disease, muscular dystrophies such as Duchenne muscular dystrophy (DMD), etc. The important impact of skeletal muscle wasting is exemplified by the fact that it Abbreviations: ActRIIB, activin receptor IIB; DCN, decorin; ECM, extracellular matrix; ED50, median effective dose; FAM, carboxyfluorescein; GDF-8, growth differentiation factor 8; GDF-11, growth differentiation factor 11; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; MSTN, myostatin; SLRP, small leucine-rich proteoglycan; TGF-β, transforming growth factor-beta n Correspondence to: Equipe “Vecteurs: Synthèse et Applications Thérapeutiques” (V-SAT), UMR 7199 CNRS – Université de Strasbourg, Faculté de Pharmacie, 74 route du Rhin, F-67401 Illkirch cedex, France. E-mail address: [email protected] (A. Kichler).

contributes to 20% of cancer deaths [1]. The mechanisms regulating skeletal muscle mass have been identified by studying among others the cattle breeds Belgian Blue and Piedmontese, which display an exceptional muscle development commonly referred to as “double-muscling” phenotype [2,3]. Mutations in the gene of myostatin (MSTN) are responsible for the increase of the muscle mass in these cattles and have also been found in hyper-muscled sheeps [4] and dogs [5,6]. In addition, a loss-of-function mutation in the myostatin gene has been identified in a child that exhibited a large increase in muscle mass and strength [7]. Altogether, the results show that myostatin is a key regulator of skeletal muscle mass. While its deletion or loss of function induces muscle hypertrophy, its overexpression or systemic administration causes muscle atrophy [8,9]. Myostatin is also referred to as growth differentiation factor-8 (GDF-8), and belongs to the transforming growth factor-beta (TGFβ) superfamily. The MSTN gene is highly conserved across species, since the sequences of murine, rat, human, porcine, chicken, and

http://dx.doi.org/10.1016/j.yexcr.2016.01.019 0014-4827/& 2016 Published by Elsevier Inc.

Please cite this article as: N. El Shafey, et al., Inhibition of the myostatin/Smad signaling pathway by short decorin-derived peptides, Exp Cell Res (2016), http://dx.doi.org/10.1016/j.yexcr.2016.01.019i

N. El Shafey et al. / Experimental Cell Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

2

turkey myostatin are 100% identical in the C-terminal region [3]. Myostatin, which is predominantly expressed in skeletal muscle, is synthesized as pre-pro-myostatin. It is secreted as an inactive promyostatin, whose proteolytic cleavage generates an amino-terminal propeptide (28 kDa) and a carboxy-terminal MSTN (12.5 kDa) that is biologically active as dimer. However, even after cleavage, most of the mature MSTN remains bound to the propeptide – and forms the so-called latency associated complex – preventing MSTN binding to its receptor [9,10]. Activation of the latent complex releases the mature myostatin and allows its binding to activin type II receptors (primarily ActRIIB), which initiates an intracellular signaling cascade leading to the phosphorylation and activation of Smad2/3 [11]. The activated Smad complexes then translocate into to the nucleus where they modify the transcription of genes involved in cellular differentiation and proliferation. Myostatin also decreases Akt phosphorylation and signals through FOXO transcription factors, resulting in increased expression of atrophy-related genes and atrophy induction [12]. Considering MSTN negative effect on muscle growth, its inhibition has been proposed as a therapeutic approach in muscledegenerative and wasting conditions, such as muscular dystrophies and cachexia. Several different anti-myostatin approaches have been developed in the last years, in particular inhibitors like MSTN propeptide [13,14], a soluble form of the activin receptor [15,16] and follistatin [17,18]. A clinical trial involving patients with muscular dystrophies has been performed using an antimyostatin antibody (MYO-029 antibody) [19], but the results were not as good as expected and the trial was stopped. Thus, research on the development of new inhibitors of myostatin is still needed. Decorin (DCN) is the best characterized member of the small leucine-rich proteoglycan (SLRP) family [20–22]. It has a core protein of 40 kDa and a single glycosaminoglycan chain with an averaged molecular weight of 22,000 Da covalently attached to a serine residue in the N-terminal part of the protein (Ser34) (Fig. 1). After synthesis, DCN is secreted into the extracellular space where it participates to the organization of the extracellular matrix (ECM) through the interaction with a variety of matrix components including fibronectin and types I, IV, and V collagen [23]. In addition to the interaction with ECM constituents, decorin interacts with fibrinogen [23] as well as with growth factors such as TGF-β [24] and the connective tissue growth factor (CTGF/CCN2) [25]. Finally, decorin interacts also with different types of membrane-located receptors including the epidermal growth factor receptor [26], the insulin growth factor-1 receptor [27], Met which is the receptor for the hepatocyte growth factor [28] and the low density lipoprotein receptor-related protein (LRP-1) [29–31]. Altogether, the capacity of decorin to interact with such a variety of factors and receptors explains how this SLRP can be involved in mechanisms as different as wound healing [32], hepatic fibrosis [33], post-myocardial GAG (Ser34)

SP PP

Cys

N

LRR N-linked Oligosaccharides

2. Materials and methods 2.1. Materials The recombinant mouse myostatin was obtained from R&D Systems. The branched polyethylenimine of 25 kDa (B-PEI) was obtained from Sigma-Aldrich. The DMAPAP cationic lipid was obtained from Dr. V. Escriou (UMR8258, Paris Descartes, France). The p(CAGA)12-Luciferase reporter expression cassette was kindly given by Dr. C.H. Heldin. 2.2. Peptides Peptides were synthesized by Proteogenix and GeneCust. The sequences of the different peptides are shown in Table 1. The 5-carboxyfluorescein (5FAM)-labeled peptides were modified at the N-terminus. 2.3. Cell culture The human embryonic kidney cell line HEK293T cells (ATCC, CRL-1573) was cultured at 37 °C with 5% CO2 using Dulbecco's modified Eagle medium (DMEM; Gibco) supplemented with 100 U/mL penicillin, 100 mg/mL streptomycin and 10% fetal bovine serum (PAA).

Cys

C 1

infarction remodeling [34], and suppression of tumorigenic growth and angiogenesis [35]. Decorin, as other SLRPs such as biglycan, play also an important role in myogenesis [36,37]. Indeed, it was shown that decorin is able to activate the differentiation of skeletal muscle cells [38]. These results can be explained, at least in part, by the suppression of myostatin activity after binding of decorin to mature myostatin in the presence of zinc [39]. Also, it was shown that the direct injection of recombinant DCN efficiently prevents fibrosis and enhances muscle regeneration [40,41]. Based on these findings, we have recently shown that the intramuscular injection of recombinant decorin results in a significant increase of dystrophic muscle mass [42]. In addition, we have identified fragments of the DCN N-terminal region, in particular murine DCN peptide 31-71, that bind to and inactivate myostatin [42]. However, this latter peptide is prone to aggregation and is thus not easy to handle. Also, knowing that decorin can bind to different proteins including to TGF-β [24,43] the issue of the specificity of action of decorin fragments remained unanswered. The three main objectives of the present study were to: i – identify short and well-soluble decorin-derived peptides that are able to inhibit myostatin, ii – characterize the mechanism of inhibition, and iii – demonstrate that the peptides have a specificity towards myostatin and do not inhibit other members of the TGF family, in particular TGF-β, GDF-11 and activin A.

354

Myostatin binding domain Fig. 1. Scheme of the decorin protein. A signal peptide (SP) composed of the first 16 residues is followed by a 14 amino acid long pro-peptide (PP). The protein core consists of leucine-rich repeats (LRR) with conserved Cys-rich N- and C-terminal domains. The N-terminal cysteine cluster has a CX3CXCX6C pattern, which is conserved in class I of the SLRPs. A single glycosaminoglycan (GAG) side chain is attached to serine 34. The binding domain to myostatin is located in the N-terminal domain.

Table 1 List of the mDecorin derived peptides. Peptide mDCN42-65 mDCN48-71 Mutated mDCN48-71

Sequencea DNPLISM PYR Q HLRVVQ SDL M PYR Q HLRVVQ SDLGLDKVP MAPYRAQAHLRVVQASDLGLDKVPb

a The peptides were amidated at the C-terminus and the cluster of cysteines are highlighted in gray; b The alanine residues (in bold and underlined) were used to replace the amino acid cysteine of the wild-type sequence of the mDCN48-71 peptide.

Please cite this article as: N. El Shafey, et al., Inhibition of the myostatin/Smad signaling pathway by short decorin-derived peptides, Exp Cell Res (2016), http://dx.doi.org/10.1016/j.yexcr.2016.01.019i

N. El Shafey et al. / Experimental Cell Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

2.4. Generation of a plasmid containing the neomycin gene and the CAGA-Luc sequence A PCR was realized to isolate the (CAGA)-Luc insert. Briefly, 5 ng of pGL3-(CAGA)-Luc plasmid were mixed with 200 nM of each primer (3′ TTA-GGA-TCC-TTA-CAC-GGC-GAT-CTT-TCC-GC; 5′-CGCATT-ATT-GGT-ACC-GAG-CTC-TTA-CGC-GT), 0.2 mM dNTP mix (New England Biolabs), 2 units Phusion enzyme (New England Biolabs) and 50 mL of water. Then, a double digestion of the plasmid pEGFP-C1, which has the Neomycin resistance gene, was performed using 5 mg peGFP-C1 and 10 U/10 U AseI/BamHI (New England Biolabs). The fragment that contains the Neomycin resistance gene was used as cloning vector. The ligation with the (CAGA)-Luc insert was performed by using 400 U enzyme T4 ligase (New England Biolabs). The products of ligation were then electrophoresed on 1% agarose gels in TAE buffer (Tris base, acetic acid and EDTA) at 90 V for 30 min. Sequencing confirmed that the resulting plasmid contained both the p(CAGA)-Luciferase sequence and the Neomycin resistance gene. 2.5. Generation of the stable cell line 700,000 HEK293T cells were plated in 6-well plates, one day before transfection. Two mg of the plasmid construct containing (CAGA)-LucNeomycin were transfected using the cationic lipid 2-{3-[Bis-(3-amino-propyl)-amino]-propylamino}-N ditetradecylcarbamoyl methylacetamide (DMAPAP) [44]. Cells were sub-cultured in Petri dishes with medium containing antibiotic at day 8. The selection was done over a period of four weeks. Individual clones were taken and cultured with G418 antibiotic at 1.4 mg/mL. Clones were plated in 24-well plates and myostatin response was tested by luciferase activity after 6 h of incubation at 37 °C with 10 ng of myostatin. Cells were lysed and luciferase activity was determined. In order to evaluate the homogeneity of positive clones in luciferase activity we used an optical camera (PhotonIMAGER™ Optima Biospace) to measure luciferase activity in each well. The enzyme substrate (luciferin) was directly added into each well a few minutes before the measurement. Clones with all positive cells were selected, and further tested. Among the clones, the one with the highest increase of luciferase activity after addition of myostatin was selected. 2.6. CAGA-Luc reporter assay with the stable cell line The capacity of decorin peptides to inhibit myostatin activity in vitro was measured by using the principle of p(CAGA)-firefly luciferase expression cassette [42]. Briefly, binding of myostatin to ActRIIB receptor results in the phosphorylation of the receptor specific Smad2/3. The activated Smads form a complex with Smad4 and are translocated to the nucleus where they interact with different cellular partners, bind to DNA-to sequences that have been termed CAGA boxes-and regulate transcription of specific genes. The p(CAGA)12-Luc reporter construct containing 12 copies of the consensus CAGA sequence has been shown to be useful as a myostatin inducible promoter system; indeed, addition of myostatin to cells containing the p(CAGA)12-Luc cassette results in a significant increase of reporter gene expression. The stable cell line HEK293T-(CAGA-Luc) was seeded in 48-well plates at 150,000 cells per well. The day after, myostatin, activin A (R&D Systems, 338-AC), TGF-β2 (R&D Systems, 7346-B2) or GDF11 (R&D Systems, 1958-GD) 7 decorin peptide or myostatin propeptide (1539-PG, R&D Systems) were mixed and incubated for 30 min at 37 °C. This mixture was then diluted in 220 mL of serumfree culture medium. After the addition of 10 mM ZnCl2, this medium was put over the cells. After 6 h, the assay was stopped by adding 100 mL of lysis buffer 1 X (Promega, Luciferase Cell Culture Lysis X Reagent). The luciferase activity was measured after

3

centrifugation of the samples for 5 min at 12,000 rpm. The amount of proteins per well was measured using a BCA assay (Thermo Scientific Pierce, 23225). Calibration curve was established using bovine serum albumin. The luciferase activity was expressed as relative light units/s/mg of protein. 2.7. Western blot Ten mg of protein of each sample were mixed with denaturation buffer (Invitrogen, NuPAGEsSample Reducing Agent 10X, NP0004) and sample buffer (Invitrogen, NuPAGEsLDS Sample Buffer 4X, NP0007). Protein denaturation was performed at 90 °C during 10 min. Protein electrophoresis was performed with a NovexsNuPAGE gel system, 4–12% Bis Tris, placed in a vertical container (XCell SureLock Invitrogen) immersed in migration buffer provided by Invitrogen (50 mM morpholino-propane-sulfonic acid, 0.1% SDS, 1 mM EDTA, 50 mM Tris base, pH 7.7). A molecular weight marker was introduced in one well (SeeBluesPlus2 PreStrained Standard, Invitrogen). Electrophoresis was performed during 2 h at 120 V. Gel system was then electro-transferred to a Hybond ECL membrane (Amersham Pharmacia Biotech) during 45 min at 25 V with pore size of 0.45 mm. Free sites of the membrane matrix were saturated with a blocking buffer (Odyssey, LICOR) completed with Tween-20 0.1%, and incubated with the primary antibody, anti-Phospho-Smad2 (Cell Signaling Technology, 3108) diluted 1:500, or anti-actin (Sigma-Aldrich, A2066) diluted 1:1000, overnight at 4 °C. The membranes were then washed with Tris buffered saline þ Tween-20 0.1%. Next, membranes were incubated with the secondary antibody (donkey anti-rabbit IgG coupled with IRDye 800 CW (LI-COR, 926-32213)) diluted 1:15,000 in blocking buffer completed with Tween-20 at 0.1%. Fluorescence imaging of the membranes was performed with a scanner ODYSSEYsCLx (LI-COR). Fluorescence intensity was measured in region of interest defined on bands using Image J (http:// rsbweb.nih.gov/ij/). 2.8. Steady-state fluorescence anisotropy Fluorescence anisotropy measurements are based on the principle of photoselective excitation of a fluorophore by a polarized light, providing information about rotational diffusion between photon absorption and emission. Light depolarization (decrease of fluorescence anisotropy) mainly originates from the overall rotational motion of the fluorescently labeled molecule or flexibility of the fluorescent moiety. Here, we assessed the binding of a fluorescent version of peptide mDCN48-71 to myostatin by monitoring steady-state fluorescence anisotropy. An increase of the anisotropy value (characterizing the peptide alone) in the presence of myostatin accounts for the existence of a peptide/myostatin complex characterized by a slower rotational diffusion or lower flexibility level. Steady-state anisotropy values were recorded on a Beacon 2000 instrument (PanVera), in a cell thermostatically held at 37 °C. 20 nM of peptide (mDCN48-71 or mutated mDCN48-71 used as a control), labeled with the 5-FAM fluorophore (5-carboxyfluorescein) were incubated in the presence of increasing concentrations of MSTN (0, 133, 267, 347, 451 and 533 nM) 710 mM of ZnCl2 in a final volume of 150 mL HEPES buffer 20 mM pH 7.8. Beforehand, the tubes for anisotropy measurements were coated with 20 mg/mL of BSA (Bovine Serum Albumin Z99%, Sigma-Aldrich) for 20 min and washed so as to limit surface adsorption of the peptide. 2.9. Statistical analysis Results were analyzed with the unpaired non-parametric Mann–Whitney test using one-tailed p-values.

Please cite this article as: N. El Shafey, et al., Inhibition of the myostatin/Smad signaling pathway by short decorin-derived peptides, Exp Cell Res (2016), http://dx.doi.org/10.1016/j.yexcr.2016.01.019i

N. El Shafey et al. / Experimental Cell Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

4

3. Results and discussion

A

3.1. Generation of a CAGA-inducible luciferase expression cell line

3.2. Validation of the cell line To further validate the isolated HEK293T-(CAGA-Luc) clone, we measured the luciferase production after addition of increasing amounts of MSTN. As shown in Fig. 2B, the cells responded in a dose-dependent manner to MSTN addition. Human MTSN maximally activated the p(CAGA)12 promoter activity 5-fold over basal, with an ED50 of 0.4 nM (Fig. 2B). In order to verify that the cell line can be used for the identification of MSTN inhibitors, we performed an assay with MSTN in the presence of increasing amounts of the MTSN propeptide. As shown in Fig. 2C, the propeptide inhibits the myostatin-mediated induction of of p(CAGA)12-Luc with an IC50 of 1 nM, a value that is in accordance with a previous study [11]. 3.3. Decorin-derived peptides as inhibitors of myostatin-mediated p (CAGA)12-Luc induction The decorin fragments able to inhibit MSTN that have been identified until now are mDCN31-71 and mDCN42-71 [42]. With the aim to increase the solubility of the decorin derived peptides, we tested two shorter peptides, namely mDCN42-65 and mDCN48-71 (Table 1). As DCN binding to MTSN is zinc dependent, we added ZnCl2 into the medium, at a physiological body-fluid Zn2 þ concentration [39]. The addition of 2 ng/ml of myostatin (0.36 nM) significantly increased the level of luciferase activity, and, 10 mM Zn2 þ had a slight decreasing effect on the monitored luciferase activity (Fig. 3). The addition of mDCN48-71 (Fig. 3A) or mDCN42-65 (Fig. 3B) in the presence of 10 mM ZnCl2 further decreased the luciferase activity in a dose-dependent manner, the latter peptide being slightly less efficient than the former. The addition of 40 μg/ well decreased the values of luciferase to the basal level for both peptides (i.e. cells without myostatin). We then focused on peptide mDCN48-71, which seemed to be

B Luciferase Luc if er aseActivity act iv it y (light(Runits/s/mg protein) LU /m g pr ot ei n)

1×1007 8×1006 6×1006 4×1006 2×1006 0 0

1

2

3

4 5 6 7 MSTN (ng)

C ns

7

Luciferase Activity (light units/s/mg protein)

The binding of members of the TGF superfamily such as myostatin (MSTN) and TGF-β to their receptors initiates a signaling cascade that leads to phosphorylation and activation of Smad2/3 [11]. The activated Smads then form a complex with Smad4, and the resulting Smad complex translocates to the nucleus and binds to Smad binding elements (SBE), leading to transcription and expression of TGF/Smad responsive genes. Smad3/Smad4 binding sequences, termed CAGA boxes, have been identified within the promoter of the human PAI-1 gene [45]. The p(CAGA)12-Luc reporter construct [45], containing 12 copies of the consensus CAGA sequence has been used as a TGF-β inducible promoter system. Addition of MSTN to cells containing the p(CAGA)12-Luc cassette results in an increase of the luciferase expression [11,46]. An interesting aspect of this system is that it can be used for the identification of MTSN inhibitors [14,47]. Here, p(CAGA)12-Luc was used to study the capacity of decorin fragments to inhibit MTSNmediated signaling. While transient transfections can be used to perform such assays [42] it is not convenient as it lengthens the time of the assay and it requires co-transfection with another reporter plasmid for normalization. For these reasons we generated a HEK293T cell line that stably expresses the p(CAGA)12-Luc reporter construct. To this end, a plasmid containing the neomycin resistance gene and the p(CAGA)12-Luc sequence was created (see Section 2). After 4 weeks of selection, we selected a clone that responded well to the addition of recombinant human MTSN (Fig. 2A).

3.0 ×10

8

9 10

*

7

2.0 ×10

7

1.0 ×10

0 Cells MSTN

5

10

15

20

30

Propeptide (ng) Fig. 2. HEK293T-(CAGA-Luc) clone selection and effect of myostatin addition to the HEK293T-(CAGA-Luc) clone. (A) In order to evaluate the homogeneity of positive clones in luciferase activity we used an optical camera (PhotonIMAGER™ Optima Biospace) to measure luciferase activity in each well. The enzyme substrate (luciferin) was directly added into each well a few minutes before the measurement. The results with the best clone are shown (right image): 10 ng of MSTN/well (0.89 nM) were added to the cells for 6 h before measuring luciferase expression. On the left, the luciferase levels of the clone in the absence of myostatin. (B) HEK293T-(CAGA-Luc) cells were incubated for 6 h with increasing amounts of human MTSN: 1, 3, 5, and 10 ng/well (corresponding to 0.18, 0.55, 0.91 and 1.82 nM). The experiment was performed in triplicate and the luciferase level is expressed in relative light unit (RLU)/s/mg protein. (C) Inhibition of myostatinmediated p(CAGA)12-Luc induction by increasing amounts of myostatin propeptide (5; 10; 15; 20 and 30 ng/well corresponding to 0.38; 0.77; 1.55; 3.10 and 4.65 nM). Experiments were performed in 48-well plates (n¼ 3–6) seeded with the stable cell line HEK293T-(CAGA-Luc). 1.5 ng per well of MSTN (0.27 nM) were used (ns: nonsignificant, *po 0.05). The results are expressed as average value 7 SEM in RLU/s/ mg protein.

more active than mDCN42-65. The addition of mDCN48-71 alone to the cells did not alter the luciferase levels (Supplementary Fig. 1A). Also, a MTT assay showed that the mDCN48-71 peptide did not reduce the luciferase levels due to cytotoxic effects (Supplementary Fig. 1B). The two mDCN-derived peptides contain four cysteine residues

Please cite this article as: N. El Shafey, et al., Inhibition of the myostatin/Smad signaling pathway by short decorin-derived peptides, Exp Cell Res (2016), http://dx.doi.org/10.1016/j.yexcr.2016.01.019i

N. El Shafey et al. / Experimental Cell Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

5

Fig. 3. Evaluation of the anti-myostatin activities of decorin derived peptides. HEK293T-(CAGA-Luc) cells were seeded in 48-well plates. The capacity of DCN peptides to inhibit myostatin-induced p(CAGA)12-Luc activity was measured by mixing MSTN (2 ng/well; 0.36 nM) with increasing amounts of (A) mDCN48-71 (1; 5; 10; 20 and 40 mg/ well corresponding to 1.64; 8.22; 16.4; 32.9 and 65.8 mM, respectively; n¼ 8–33), (B) mDCN42-65 (1; 5; 10; 20 and 40 mg/well corresponding to 1.63; 8.14; 16.3; 32.6 and 65.1 mM, respectively; n¼ 3–13) or (C) mutated mDCN48-71 (CTL peptide) (1; 5; 10; 20 and 40 mg/well corresponding to 1.72; 8.62; 17.2; 34.5 and 69 mM, respectively, n¼ 3– 6), in the presence of 10 mM ZnCl2. The results are expressed as the average value 7 SEM (ns: non-significant, *p o 0.05, **p o 0.01, **p o 0.001). (D) Determination of the IC50 value for mDCN48-71.

with the spacing CX3CXCX6C, which is characteristic of subgroup I of the SLRPs [20,22]. Although the residues implicated in the binding of zinc have not been identified, it is reasonable to suppose that the cysteine cluster is involved in this binding process [42,48]. Thus, we replaced in the 48-71 sequence the four cysteines by alanine residues (Mutated mDCN48-71, noted CTL (control)) and subsequently performed a (CAGA)-Luc assay. As shown in Fig. 3C, the mutated peptide was unable to block the MSTN-mediated luciferase induction. The IC50 of mDCN48-71 was 7  10  6 M (Fig. 3D). Although the experimental conditions are not identical, this value fits well with the IC50 that had been found for the mDCN31-71 peptide (3  10  6 M) [42]. The presence of several cysteine residues in a peptide could lead to the formation of disulfide bridges, and this in turn could reduce the capacity of the peptide to bind zinc and consequently also MSTN. Therefore, the percentage of free thiol groups in the mDCN48-71 sample was determined by using the Ellman assay. We found that about 60% of cysteines contained a free –SH group. This suggests that the IC50 of 7  10  6 M found for

the mDCN48-71 peptide is probably underestimated. Altogether, our results demonstrate that the N-terminal derived peptides mDCN42-65 and 48-71 are sufficient for the binding to and inactivation of myostatin. Of note, both peptides were easily solubilized in water at 1 mg/ml.

3.4. The peptide mDCN48-71 inhibits the activation of the Smad pathway The binding of myostatin to ActRII receptor has been shown to result in the phosphorylation of the receptor specific Smad2/3 [11]. Western blot analysis showed that the inhibition of the MSTNmediated induction of the (CAGA)-Luc reporter construct by the mDCN48-71 peptide results in a dose-dependent decrease of the phosphorylation of the Smad2 protein (Fig. 4). This result confirms that the binding of the decorin-derived peptide to myostatin blocks the capacity of the latter to interact in a productive manner (i.e. leading to Smad2/3 activation) with the ActRII receptor.

Please cite this article as: N. El Shafey, et al., Inhibition of the myostatin/Smad signaling pathway by short decorin-derived peptides, Exp Cell Res (2016), http://dx.doi.org/10.1016/j.yexcr.2016.01.019i

N. El Shafey et al. / Experimental Cell Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

6

A

2 ng MSTN + mDCN48-71 + 10 µM Zn2+ Cells

MSTN 2 ng MSTN (2 ng) + 10 µM Zn2+

Cells

2 ng

+10 M

MSTN

ZnCl 2

1 µg

5 µg

20 µg

40 µg

PSmad2 Actin

PSmad2/Actin

B

1.5

1.0

0.5

0.0 1 g

5 g

20 g

40 g

mDCN48-71 + 10 M ZnCl 2 Fig. 4. Quantification of the amounts of phosphorylated PSmad2. Lysates from a (CAGA)-Luc inhibition assay (with 0.36 nM of MSTN/well) were used to perform a western blot in order to quantify the amounts of phosphorylated PSmad2 in each sample. The actin protein was used for normalization (n¼ 2). The amounts of DCN peptide were: 1; 5; 20 and 40 mg/well corresponding to 1.64; 8.22; 32.9 and 65.8 mM, respectively. (A) Western blot; (B) Results after quantification and normalization of the bands.

concentrations of myostatin, in the presence or absence of 10 μM ZnCl2 (Fig. 5). Anisotropy measurements were also performed using a 5FAM-labeled mutated control peptide, in which the 4 cysteines were replaced by alanine residues (see Table 1). Notably, the 5FAM-mDCN48-71 displayed similar activities than the non-labeled peptide in the MSTN (CAGA)-Luc reporter induction assay, meaning that the fluorescent moiety did not interfere with

3.5. Evaluation of the binding capacity of the mDCN48-71 peptide to myostatin by fluorescence anisotropy To characterize the binding of the N-terminal derived DCNderived peptide 48-71 to MSTN, we conducted steady-state fluorescence anisotropy experiments [49,50]. A 5FAM-labeled version of the DCN-derived peptide was mixed with increasing 0.05

n=4

0.04 n=6

r

0.03 n=3

n=7

n=5

0.02 0.01 0.00

n=3

n=3

133 nM MSTN

267 nM MSTN

347 nM MSTN

451 nM MSTN

533 nM MSTN

20 nM mDCN48-71 + 10 M ZnCl2

0,0175 0.0175

0,0218 0.0218

0,0221 0.0221

0,0279 0.0279

0,0403 0.0403

20 nM mDCN48-71

0,0033 0.0033

0,0043 0.0043

nd nd

nd nd

0,0024 0.0024

-0,0005 -0.0005

0,0005 0.0005

nd nd

nd nd

0,0003 0.0003

20 nM Control + 10 M ZnCl2

Fig. 5. Measurement of the interaction between protein myostatin and 5FAM-mDCN48-71 or 5FAM-control peptide by steady-state fluorescence anisotropy (r). 20 nM of 5FAM-mDCN48-71 or 5FAM-control peptide (mutated mDCN48-71 for the four cysteines) 710 mM ZnCl2 were incubated in the presence or not of increasing concentrations of protein myostatin (MSTN: 133; 267; 347; 451 and 533 nM). Then, Δr values (representing the difference in r values between peptide þ myostatin and peptide alone) were calculated; the r values characterizing MSTN-free peptides were 0.051, 0.044 and 0.033 for mDCN48-71 þZn, mDCN48-71 without Zn and the control peptideþ Zn, respectively. The numbers of repeated experiments are indicated under the bars, lines with n ¼3 mean numbers of experiments for each condition precised; nd: not determined. The results are expressed as average value7 SEM.

Please cite this article as: N. El Shafey, et al., Inhibition of the myostatin/Smad signaling pathway by short decorin-derived peptides, Exp Cell Res (2016), http://dx.doi.org/10.1016/j.yexcr.2016.01.019i

N. El Shafey et al. / Experimental Cell Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

the DCN/MSTN interaction (data not shown). We observed a net increase in the fluorescence anisotropy of the fluorescein-peptide conjugate by increasing the MSTN concentration, suggesting a physical interaction between the DCN peptide and the MSTN protein. This effect strictly required the presence of Zn2 þ . Accordingly, no interaction was observed with the control (mutant) peptide in the presence of Zn2 þ . These results clearly indicate that mDCN48-71 can directly bind to myostatin. In addition, they demonstrate that this interaction requires the presence of Zn2 þ . 3.6. Specificity of the mDCN48-71 mediated inhibition It has been previously reported that the DCN protein is able to bind to different proteins, including TGF-β [24,43]. Therefore, we wanted to address the question as to whether mDCN48-71 is able to inhibit other members of the TGF superfamily, and more particularly TGF-β2, GDF-11 and activin A. The choice of these proteins was made for the following reasons: i – it has been previously shown that the DCN protein is able to bind and inactivate TGF-β [51]. Although, the high affinity binding site to TGF-β is located in the DCN domain Leu155-Val260, a lower binding site is also present in the fragment Asp45-Leu155 [24]; ii – Myostatin and growth differentiation factor 11 (GDF-11) share 90% amino acid identity in the C-terminal region and they also have similar signaling pathways; both bind the activin type IIB receptor and activate the intracellular mediator Smad 2/3 pathway [52], and both are antagonized by follistatin [53]; iii – In muscle, activin A and MSTN bind to the same surface receptor complex and activate the same signaling cascade that leads to Smad2/3 translocation

7

into the nucleus [54]. The activity of activin A is regulated by its latent complex, where its propeptide remains associated and prevents its binding to its receptor, and also by follistatin and follistatin-like proteins, which inhibit its activity. Activin A levels are upregulated in many catabolic disease states, suggesting its involvement in pathogenesis of wasting [55]. In order to address this point, we measured the luciferase production after addition of increasing amounts of the 3 members of the TGF family. As shown in Fig. 6A, the cells responded well, and in a dose-dependent manner, to all three proteins. The maximal induction level with human activin A was 10-fold with an ED50 around 0.08 nM whilst for TGF-β2 and GDF-11, the maximal induction levels were of 5.4and 8.5-fold with an ED50 of 0.09 nM and 0.07 nM, respectively. While using the same experimental conditions than those with MSTN, increasing amounts of mDCN48-71 peptide were unable to decrease in a significant manner the luciferase levels induced either by GDF-11, TGF-β2 or activin A (Fig. 6B). These results show that mDCN48-71 is unable to block the binding of GDF-11, TGF-β2 or activin A to their receptors, suggesting that the activity of the peptide is specific for myostatin. Taken together, our results show that the peptides mDCN48-71 and 42-65 are sufficient for the binding to and inactivation of the negative muscle mass regulator myostatin. An important aspect of reduction of the size of the active peptides is that mDCN48-71 is well soluble in aqueous solution while this was not the case of the previous peptides, in particular for mDCN31-71, which was prone to aggregation. Whether injection of this decorin peptide could help to combat muscle mass loss in the context of sarcopenia or cachexia remains to be shown.

Fig. 6. Specificity of action of mDCN48-71. (A) p(CAGA)12-Luc induction by increasing amounts of GDF-11, TGF-β2 and activin A. HEK293T-(CAGA-Luc) were seeded in 48-well plates (n¼ 3). Increasing quantities of proteins were incubated 6 h in serum-free medium. The amounts indicated in the figures were those used per well: GDF-11 (0.5; 1.5; 2.5; 5 and 10 ng/well corresponding to 0.09; 0.27; 0.45; 0.90 and 1.80 nM, respectively), TGF-β2 (0.1; 0.25; 0.5; 1; 3 and 5 ng/well that correspond to 0.04; 0.09; 0.18; 0.36; 1.07 and 1.79 nM, respectively) and activin A (0.1; 0.25; 0.5; 1; 3 ng/well corresponding to 0.02; 0.04; 0.09; 0.17; 0.52 and 0.87 nM, respectively). The results are expressed as average value7SEM in RLU/s/mg protein. (B) Effect of mDCN48-71 on the luciferase induction mediated by the 3 proteins. Increasing quantities of mDCN48-71þ 10 mM de ZnCl2 were incubated with HEK293T-(CAGA-Luc) cells: 1; 5; 10; 20 and 40 mg/well (1.64; 8.22; 16.4; 32.9 and 65.8 mM, respectively) in the presence of 1.5 ng/well of GDF-11 (¼ 0.27 nM), 0.4 ng/well of TGF-β2 (¼ 0.14 nM), or 0.5 ng/well of Activin A (¼ 0.09 nM). After 6 h of incubation, luciferase levels were measured (n¼ 3; ns for non-significant).

Please cite this article as: N. El Shafey, et al., Inhibition of the myostatin/Smad signaling pathway by short decorin-derived peptides, Exp Cell Res (2016), http://dx.doi.org/10.1016/j.yexcr.2016.01.019i

N. El Shafey et al. / Experimental Cell Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

8

Conflict of interest disclosure The authors declare no competing financial interests.

Acknowledgments We are grateful to Dr. C.H. Heldin for the kind gift of the p(CAGA)12-Luciferase reporter expression cassette. We would like to acknowledge Dr. V. Escriou for helpful discussions. This work was performed with the financial support of the Association Française contre les Myopathies (DdT2 2012 N°16127).

Appendix A. Supplementary material Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.yexcr.2016.01.019.

References [1] K. Burckart, S. Beca, R.J. Urban, M. Sheffield-Moore, Pathogenesis of muscle wasting in cancer cachexia: targeted anabolic and anticatabolic therapies, Curr. Opin. Clin. Nutr. Metab. Care 13 (2010) 410–416. [2] L. Grobet, L.J. Martin, D. Poncelet, D. Pirottin, B. Brouwers, J. Riquet, A. Schoeberlein, S. Dunner, F. Menissier, J. Massabanda, R. Fries, R. Hanset, M. Georges, A deletion in the bovine myostatin gene causes the doublemuscled phenotype in cattle, Nat. Genet. 17 (1997) 71–74. [3] A.C. McPherron, S.J. Lee, Double muscling in cattle due to mutations in the myostatin gene, Proc. Natl. Acad. Sci. USA 94 (1997) 12457–12461. [4] A. Clop, F. Marcq, H. Takeda, D. Pirottin, X. Tordoir, B. Bibe, J. Bouix, F. Caiment, J.M. Elsen, F. Eychenne, C. Larzul, E. Laville, F. Meish, D. Milenkovic, J. Tobin, C. Charlier, M. Georges, A mutation creating a potential illegitimate microRNA target site in the myostatin gene affects muscularity in sheep, Nat. Genet. 38 (2006) 813–818. [5] D.S. Mosher, P. Quignon, C.D. Bustamante, N.B. Sutter, C.S. Mellersh, H. G. Parker, E.A. Ostrander, A mutation in the myostatin gene increases muscle mass and enhances racing performance in heterozygote dogs, PLoS Genet. 3 (2007) e79. [6] G.D. Shelton, E. Engvall, Gross muscle hypertrophy in whippet dogs is caused by a mutation in the myostatin gene, Neuromuscul. Disord.: NMD 17 (2007) 721–722. [7] M. Schuelke, K.R. Wagner, L.E. Stolz, C. Hubner, T. Riebel, W. Komen, T. Braun, J. F. Tobin, S.J. Lee, Myostatin mutation associated with gross muscle hypertrophy in a child, N. Engl. J. Med. 350 (2004) 2682–2688. [8] A.C. McPherron, A.M. Lawler, S.J. Lee, Regulation of skeletal muscle mass in mice by a new TGF-beta superfamily member, Nature 387 (1997) 83–90. [9] T.A. Zimmers, M.V. Davies, L.G. Koniaris, P. Haynes, A.F. Esquela, K. N. Tomkinson, A.C. McPherron, N.M. Wolfman, S.J. Lee, Induction of cachexia in mice by systemically administered myostatin, Science 296 (2002) 1486–1488. [10] S.J. Lee, A.C. McPherron, Regulation of myostatin activity and muscle growth, Proc. Natl. Acad. Sci. USA 98 (2001) 9306–9311. [11] X. Zhu, S. Topouzis, L.F. Liang, R.L. Stotish, Myostatin signaling through Smad2, Smad3 and Smad4 is regulated by the inhibitory Smad7 by a negative feedback mechanism, Cytokine 26 (2004) 262–272. [12] J. Rodriguez, B. Vernus, I. Chelh, I. Cassar-Malek, J.C. Gabillard, A. Hadj Sassi, I. Seiliez, B. Picard, A. Bonnieu, Myostatin and the skeletal muscle atrophy and hypertrophy signaling pathways, Cell. Mol. Life Sci.: CMLS 71 (2014) 4361–4371. [13] H. Collins-Hooper, R. Sartori, R. Macharia, K. Visanuvimol, K. Foster, A. Matsakas, H. Flasskamp, S. Ray, P.R. Dash, M. Sandri, K. Patel, Propeptidemediated inhibition of myostatin increases muscle mass through inhibiting proteolytic pathways in aged mice, J. Gerontol. Ser. A, Biol. Sci. Med. Sci. 69 (2014) 1049–1059. [14] N.M. Wolfman, A.C. McPherron, W.N. Pappano, M.V. Davies, K. Song, K. N. Tomkinson, J.F. Wright, L. Zhao, S.M. Sebald, D.S. Greenspan, S.J. Lee, Activation of latent myostatin by the BMP-1/tolloid family of metalloproteinases, Proc. Natl. Acad. Sci. USA 100 (2003) 15842–15846. [15] S.J. Lee, L.A. Reed, M.V. Davies, S. Girgenrath, M.E. Goad, K.N. Tomkinson, J. F. Wright, C. Barker, G. Ehrmantraut, J. Holmstrom, B. Trowell, B. Gertz, M. S. Jiang, S.M. Sebald, M. Matzuk, E. Li, L.F. Liang, E. Quattlebaum, R.L. Stotish, N. M. Wolfman, Regulation of muscle growth by multiple ligands signaling through activin type II receptors, Proc. Natl. Acad. Sci. USA 102 (2005) 18117–18122. [16] Q. Wang, A.C. McPherron, Myostatin inhibition induces muscle fibre hypertrophy prior to satellite cell activation, J. Physiol. 590 (2012) 2151–2165. [17] H. Amthor, G. Nicholas, I. McKinnell, C.F. Kemp, M. Sharma, R. Kambadur, K. Patel, Follistatin complexes Myostatin and antagonises Myostatin-mediated

inhibition of myogenesis, Dev. Biol. 270 (2004) 19–30. [18] C.E. Winbanks, K.L. Weeks, R.E. Thomson, P.V. Sepulveda, C. Beyer, H. Qian, J. L. Chen, J.M. Allen, G.I. Lancaster, M.A. Febbraio, C.A. Harrison, J.R. McMullen, J. S. Chamberlain, P. Gregorevic, Follistatin-mediated skeletal muscle hypertrophy is regulated by Smad3 and mTOR independently of myostatin, J. Cell Biol. 197 (2012) 997–1008. [19] K.R. Wagner, J.L. Fleckenstein, A.A. Amato, R.J. Barohn, K. Bushby, D.M. Escolar, K.M. Flanigan, A. Pestronk, R. Tawil, G.I. Wolfe, M. Eagle, J.M. Florence, W. M. King, S. Pandya, V. Straub, P. Juneau, K. Meyers, C. Csimma, T. Araujo, R. Allen, S.A. Parsons, J.M. Wozney, E.R. Lavallie, J.R. Mendell, A phase I/IItrial of MYO-029 in adult subjects with muscular dystrophy, Ann. Neurol. 63 (2008) 561–571. [20] P.A. McEwan, P.G. Scott, P.N. Bishop, J. Bella, Structural correlations in the family of small leucine-rich repeat proteins and proteoglycans, J. Struct. Biol. 155 (2006) 294–305. [21] T. Neill, L. Schaefer, R.V. Iozzo, Decorin: a guardian from the matrix, Am. J. Pathol. 181 (2012) 380–387. [22] L. Schaefer, R.V. Iozzo, Biological functions of the small leucine-rich proteoglycans: from genetics to signal transduction, J. Biol. Chem. 283 (2008) 21305–21309. [23] T.A. Dugan, V.W. Yang, D.J. McQuillan, M. Hook, Decorin binds fibrinogen in a Zn2 þ-dependent interaction, J. Biol. Chem. 278 (2003) 13655–13662. [24] E. Schonherr, M. Broszat, E. Brandan, P. Bruckner, H. Kresse, Decorin core protein fragment Leu155-Val260 interacts with TGF-beta but does not compete for decorin binding to type I collagen, Arch. Biochem. Biophys. 355 (1998) 241–248. [25] C. Vial, J. Gutierrez, C. Santander, D. Cabrera, E. Brandan, Decorin interacts with connective tissue growth factor (CTGF)/CCN2 by LRR12 inhibiting its biological activity, J. Biol. Chem. 286 (2011) 24242–24252. [26] R.V. Iozzo, D.K. Moscatello, D.J. McQuillan, I. Eichstetter, Decorin is a biological ligand for the epidermal growth factor receptor, J. Biol. Chem. 274 (1999) 4489–4492. [27] E. Schonherr, C. Sunderkotter, R.V. Iozzo, L. Schaefer, Decorin, a novel player in the insulin-like growth factor system, J. Biol. Chem. 280 (2005) 15767–15772. [28] S. Goldoni, A. Humphries, A. Nystrom, S. Sattar, R.T. Owens, D.J. McQuillan, K. Ireton, R.V. Iozzo, Decorin is a novel antagonistic ligand of the Met receptor, J. Cell Biol. 185 (2009) 743–754. [29] E. Brandan, C. Retamal, C. Cabello-Verrugio, M.-P. Marzolo, The low density lipoprotein receptor-related protein functions as an endocytic receptor for decorin, J. Biol. Chem. 281 (2006) 31562–31571. [30] C. Cabello-Verrugio, E. Brandan, A novel modulatory mechanism of transforming growth factor-beta signaling through decorin and LRP-1, J. Biol. Chem. 282 (2007) 18842–18850. [31] C. Cabello-Verrugio, C. Santander, C. Cofre, M.J. Acuna, F. Melo, E. Brandan, The internal region leucine-rich repeat 6 of decorin interacts with low density lipoprotein receptor-related protein-1, modulates transforming growth factor (TGF)-beta-dependent signaling, and inhibits TGF-beta-dependent fibrotic response in skeletal muscles, J. Biol. Chem. 287 (2012) 6773–6787. [32] H. Jarvelainen, P. Puolakkainen, S. Pakkanen, E.L. Brown, M. Hook, R.V. Iozzo, E. H. Sage, T.N. Wight, A role for decorin in cutaneous wound healing and angiogenesis, Wound Repair Regen.: Off. Publ. Wound Heal. Soc. Eur. Tissue Repair Soc. 14 (2006) 443–452. [33] K. Baghy, K. Dezso, V. Laszlo, A. Fullar, B. Peterfia, S. Paku, P. Nagy, Z. Schaff, R. V. Iozzo, I. Kovalszky, Ablation of the decorin gene enhances experimental hepatic fibrosis and impairs hepatic healing in mice, Lab. Investig.; J. Tech. Methods Pathol. 91 (2011) 439–451. [34] S.M. Weis, S.D. Zimmerman, M. Shah, J.W. Covell, J.H. Omens, J. Ross Jr., N. Dalton, Y. Jones, C.C. Reed, R.V. Iozzo, A.D. McCulloch, A role for decorin in the remodeling of myocardial infarction, Matrix Biol.: J. Int. Soc. Matrix Biol. 24 (2005) 313–324. [35] T. Neill, L. Schaefer, R.V. Iozzo, Decorin as a multivalent therapeutic agent against cancer, Adv. Drug Deliv. Rev. (2015). [36] E. Brandan, C. Cabello-Verrugio, C. Vial, Novel regulatory mechanisms for the proteoglycans decorin and biglycan during muscle formation and muscular dystrophy, Matrix Biol.: J. Int. Soc. Matrix Biol. 27 (2008) 700–708. [37] E. Brandan, J. Gutierrez, Role of skeletal muscle proteoglycans during myogenesis, Matrix Bio.: J. Int. Soc. Matrix Biol. 32 (2013) 289–297. [38] Y. Kishioka, M. Thomas, J. Wakamatsu, A. Hattori, M. Sharma, R. Kambadur, T. Nishimura, Decorin enhances the proliferation and differentiation of myogenic cells through suppressing myostatin activity, J. Cell. Physiol. 215 (2008) 856–867. [39] T. Miura, Y. Kishioka, J. Wakamatsu, A. Hattori, A. Hennebry, C.J. Berry, M. Sharma, R. Kambadur, T. Nishimura, Decorin binds myostatin and modulates its activity to muscle cells, Biochem. Biophys. Res. Commun. 340 (2006) 675–680. [40] K. Fukushima, N. Badlani, A. Usas, F. Riano, F. Fu, J. Huard, The use of an antifibrosis agent to improve muscle recovery after laceration, Am. J. Sports Med. 29 (2001) 394–402. [41] Y. Li, J. Li, J. Zhu, B. Sun, M. Branca, Y. Tang, W. Foster, X. Xiao, J. Huard, Decorin gene transfer promotes muscle cell differentiation and muscle regeneration, Mol. Ther.: J. Am. Soc. Gene Ther. 15 (2007) 1616–1622. [42] S. Guiraud, L. van Wittenberghe, C. Georger, D. Scherman, A. Kichler, Identification of decorin derived peptides with a zinc dependent anti-myostatin activity, Neuromuscul. Disord.: NMD 22 (2012) 1057–1068. [43] H. Hausser, A. Groning, A. Hasilik, E. Schonherr, H. Kresse, Selective inactivity of TGF-beta/decorin complexes, FEBS Lett. 353 (1994) 243–245.

Please cite this article as: N. El Shafey, et al., Inhibition of the myostatin/Smad signaling pathway by short decorin-derived peptides, Exp Cell Res (2016), http://dx.doi.org/10.1016/j.yexcr.2016.01.019i

N. El Shafey et al. / Experimental Cell Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎ [44] M. Khoury, P. Louis-Plence, V. Escriou, D. Noel, C. Largeau, C. Cantos, D. Scherman, C. Jorgensen, F. Apparailly, Efficient new cationic liposome formulation for systemic delivery of small interfering RNA silencing tumor necrosis factor alpha in experimental arthritis, Arthritis Rheum. 54 (2006) 1867–1877. [45] S. Dennler, S. Itoh, D. Vivien, P. ten Dijke, S. Huet, J.M. Gauthier, Direct binding of Smad3 and Smad4 to critical TGF beta-inducible elements in the promoter of human plasminogen activator inhibitor-type 1 gene, EMBO J. 17 (1998) 3091–3100. [46] M.S. Jiang, L.F. Liang, S. Wang, T. Ratovitski, J. Holmstrom, C. Barker, R. Stotish, Characterization and identification of the inhibitory domain of GDF-8 propeptide, Biochem. Biophys. Res. Commun. 315 (2004) 525–531. [47] R.S. Thies, T. Chen, M.V. Davies, K.N. Tomkinson, A.A. Pearson, Q.A. Shakey, N. M. Wolfman, GDF-8 propeptide binds to GDF-8 and antagonizes biological activity by inhibiting GDF-8 receptor binding, Growth Factors 18 (2001) 251–259. [48] V.W. Yang, S.R. LaBrenz, L.C. Rosenberg, D. McQuillan, M. Hook, Decorin is a Zn2þ metalloprotein, J. Biol. Chem. 274 (1999) 12454–12460. [49] K. Carayon, H. Leh, E. Henry, F. Simon, J.F. Mouscadet, E. Deprez, A cooperative and specific DNA-binding mode of HIV-1 integrase depends on the nature of the metallic cofactor and involves the zinc-containing N-terminal domain, Nucl. Acids Res. 38 (2010) 3692–3708.

9

[50] M. Pinskaya, E. Romanova, E. Volkov, E. Deprez, H. Leh, J.C. Brochon, J. F. Mouscadet, M. Gottikh, HIV-1 integrase complexes with DNA dissociate in the presence of short oligonucleotides conjugated to acridine, Biochemistry 43 (2004) 8735–8743. [51] W.A. Border, N.A. Noble, T. Yamamoto, J.R. Harper, Y. Yamaguchi, M. D. Pierschbacher, E. Ruoslahti, Natural inhibitor of transforming growth factorbeta protects against scarring in experimental kidney disease, Nature 360 (1992) 361–364. [52] D.M. Ho, C.Y. Yeo, M. Whitman, The role and regulation of GDF11 in Smad2 activation during tailbud formation in the Xenopus embryo, Mech. Dev. 127 (2010) 485–495. [53] A.L. Schneyer, Y. Sidis, A. Gulati, J.L. Sun, H. Keutmann, P.A. Krasney, Differential antagonism of activin, myostatin and growth and differentiation factor 11 by wild-type and mutant follistatin, Endocrinology 149 (2008) 4589–4595. [54] K. Tsuchida, M. Nakatani, A. Uezumi, T. Murakami, X. Cui, Signal transduction pathway through activin receptors as a therapeutic target of musculoskeletal diseases and cancer, Endocr. J. 55 (2008) 11–21. [55] J.L. Chen, K.L. Walton, C.E. Winbanks, K.T. Murphy, R.E. Thomson, Y. Makanji, H. Qian, G.S. Lynch, C.A. Harrison, P. Gregorevic, Elevated expression of activins promotes muscle wasting and cachexia, FASEB J.: Off. Publ. Fed. Am. Soc. Exp. Biol. 28 (2014) 1711–1723.

Please cite this article as: N. El Shafey, et al., Inhibition of the myostatin/Smad signaling pathway by short decorin-derived peptides, Exp Cell Res (2016), http://dx.doi.org/10.1016/j.yexcr.2016.01.019i