Gene 553 (2014) 130–139

Contents lists available at ScienceDirect

Gene journal homepage: www.elsevier.com/locate/gene

Periostin is temporally expressed as an extracellular matrix component in skeletal muscle regeneration and differentiation Cansu Özdemir a, Uğur Akpulat a, Parisa Sharafi a, Yılmaz Yıldız a, İlyas Onbaşılar b, Çetin Kocaefe a,⁎ a b

Dept. of Medical Biology, Hacettepe University School of Medicine, Ankara, Turkey Laboratory Animal Breeding and Research Unit, Hacettepe University School of Medicine, Ankara, Turkey

a r t i c l e

i n f o

Article history: Received 22 July 2014 Received in revised form 5 October 2014 Accepted 7 October 2014 Available online 8 October 2014 Keywords: Periostin Muscle injury Muscle regeneration Myogenic differentiation

a b s t r a c t The transcriptional events and pathways responsible for the acquisition of the myogenic phenotype during regeneration and myogenesis have been studied extensively. The modulators that shape the extracellular matrix in health and disease, however, are less understood. Understanding the components and pathways of this remodeling will aid the restoration of the architecture and prevent deterioration under pathological conditions such as fibrosis. Periostin, a matricellular protein associated with remodeling of the extracellular matrix and connective tissue architecture, is emerging in pathological conditions associated with fibrosis in adult life. Periostin also complicates fibrosis in degenerative skeletal muscle conditions such as dystrophies. This study primarily addresses the spatial and temporal involvement of periostin along skeletal muscle regeneration. In the acute skeletal muscle injury model that shows recovery without fibrosis, we show that periostin is rapidly disrupted along with the extensive necrosis and periostin mRNA is transiently upregulated during the myotube maturation. This expression is stringently initiated from the newly regenerating fibers. However, this observation is contrasting to a model that displays extensive fibrosis where upregulation of periostin expression is stable and confined to the fibrotic compartments of endomysial and perimysial space. In vitro myoblast differentiation further supports the claim that upregulation of periostin expression is a function of extracellular matrix remodeling during myofiber differentiation and maturation. We further seek to identify the expression kinetics of various periostin isoforms during the differentiation of rat and mouse myoblasts. Results depict that a singular periostin isoform dominated the rat muscle, contrasting to multiple isoforms in C2C12 myoblast cells. This study shows that periostin, a mediator with deleterious impact on conditions exhibiting fibrosis, is also produced and secreted by myoblasts and regenerating myofibers during architectural remodeling in the course of development and regeneration. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Under physiological conditions, skeletal muscle exhibits a stable structure. However, in the case of injury, satellite cells – the resident somatic stem cells of the muscle – are activated to restore structural integrity (Charge and Rudnicki, 2004). Differentiation and repair both pursue cascade of transcriptional events harmonized by common myogenic regulatory factors (MRFs) and signaling pathways (Braun and Gautel, 2011). The transcriptional events responsible for the acquisition of the myogenic phenotype are relatively well understood. Repeated cycles of injury and regeneration impact the tissue architecture of the skeletal muscle, causing fibrosis, fatty infiltration and myofibrillary atrophy, which are the three hallmarks of chronic muscle degeneration. The molecular events that shape the extracellular matrix and cause fibrosis

Abbreviations: MRFs, myogenic regulatory factors; DMEM, Dulbecco's Modified Eagle's Medium; FCS, fetal calf serum. ⁎ Corresponding author at: Department of Medical Biology, Hacettepe University Faculty of Medicine, 06100 Sihhiye, Ankara, Turkey. E-mail address: [email protected] (Ç. Kocaefe).

http://dx.doi.org/10.1016/j.gene.2014.10.014 0378-1119/© 2014 Elsevier B.V. All rights reserved.

under pathological conditions, such as the dystrophies, aging or disuse atrophy are still obscure. High throughput technologies provide insights into the transcriptional events that characterize differentiation, repair and adaptive response of the tissue. Using the open-access transcriptome data, we conducted an in silico survey to pinpoint a conjoint list of genes that were significantly altered in a set of transcriptome observations commonly targeting physiologic and pathologic conditions of skeletal muscle. These include time-course profiling of myoblast differentiation (Moran et al., 2002; Tomczak et al., 2004; Chen et al., 2006), various human skeletal muscle pathologies (Bakay et al., 2006), dystrophindeficient (mdx) mouse diaphragm (Porter et al., 2004) and sarcopenia model in rat (Wang et al., 2012). This approach accentuated a number of commonly altered soluble or exported factors, including periostin. Periostin is a matricellular protein which is ubiquitously expressed during embryonic morphogenesis in the extracellular matrix and connective tissues. It is directly interacting with collagen and fibronectin during fibrillogenesis in early development (Kudo, 2011). In the postnatal life, periostin is abundant in connective tissues exposed to mechanical strains such as tendon, bone, heart valves and skin (Merle and

C. Özdemir et al. / Gene 553 (2014) 130–139

Garnero, 2012). However, periostin upregulation is associated with fibrotic pathological conditions such as scar formation in wound healing, crash-induced bone damage, myocardial infarcts, pulmonary fibrosis, liver cirrhosis and fibrosis of the skeletal muscle (Hamilton, 2008; Rani et al., 2009; Dobaczewski et al., 2010; Merle and Garnero, 2012). An attributed function promoting collagen cross-linking is the association of periostin to pathologic fibrotic events (Maruhashi et al., 2010). In skeletal muscle, periostin function is linked to fibrosis-related conditions such as strain injury evoked by eccentric exercise (Rani et al., 2009). Periostin knockout animals exhibit diminished fibroblastic proliferation in the course of cardiac recovery from myocardial infarction (Oka et al., 2007; Norris et al., 2009). Likewise, diminished fibrosis and enhanced myofiber regeneration are observed in knockout dystrophic mice (Lorts et al., 2012). Besides its role in chronic degenerative conditions, the abovementioned transcriptome results primed us to investigate the spatial and temporal expression of periostin in various models of muscle regeneration and myoblast differentiation.

2. Materials and methods 2.1. Animals and procedures All experiments were conducted on three-month-old male Sprague– Dawley rats (225 g ± 30 g) and all procedures were performed according to institution-approved protocol and under strict biological containment (Approval Decision No.: 2009/30-4 and 2005/40-8). The acute muscle injury model was accomplished via injection of cardiotoxin into the tibialis anterior (TA) muscles of adult rats. Briefly, following appropriate anesthesia and disinfection, 1 nmol of cardiotoxins (Sigma-Aldrich) was infiltrated using a Hamilton syringe. On the pre-determined days (1st, 3rd, 4th, 6th, 8th, 10th, and 14th) the time-course sampling was carried out. The total TA muscle from the right extremity was resected under general anesthesia prior to sacrifice (n = 3 for each time point). The extracted TA muscles were harvested and fresh frozen accordingly for RNA isolation, protein extraction or tissue sectioning. Achilles tenotomy was used as a model to induce fibrosis in skeletal muscle. Following appropriate anesthesia and disinfection, partial resection of the Achilles tendon was performed on the right side by removing approximately 4 mm distal section. Soleus muscles from both the right and left extremities were resected under general anesthesia for a follow-up period of 6 weeks (n = 3 for each time point). The muscles were handled as previously described above.

2.2. Cells and in vitro studies C2C12 mouse embryonic myoblast cells (Yaffe and Saxel, 1977) were used for the documentation of the periostin expression during proliferation, fusion and myotube maturation stages of differentiation. The C2C12 cell line was purchased from the Foot and Mouth Disease Institution's Animal Cell Culture Collection Facility (Ankara, Turkey) and all tissue culture consumables were purchased from Biochrom AG, Germany. Cells were maintained on “proliferation medium” which consisted of Dulbecco's Modified Eagle's Medium (DMEM) supplemented with 10% heat inactivated fetal calf serum (FCS), 2 mM L-Glutamine and 1% (w/v) streptomycin and penicillin. The medium was refreshed every 2 days. Upon reaching confluency, cells were induced for myogenic differentiation using a “differentiation medium,” in which 2% horse serum was substituted for the 10% of FCS. Triplicated samples of 50% and 90% confluent cells under proliferation conditions in the 24th, 48th, and 96th hours following induction of differentiation were harvested using trizol reagent.

131

2.3. Isolation and in vitro differentiation of primary myoblasts Expression kinetics of periostin was further documented in differentiating primary myoblasts (Kocaefe et al., 2010). Lower extremity muscles were dissected and subjected to collagenase type 1 (La Roche Ltd., Basel, Switzerland) and dispase (Sigma) digestion (0.5% w/v). Mononuclear cells were harvested following retrieval through a 30 μm mesh and Ficoll gradient and immediately plated on matrigel (BD Biosciences, NJ) coated tissue culture plates. Primary myoblasts were expanded in DMEM/Ham's Nutrient Mixture F12 medium (1:1) supplemented with 20% FCS and 2% Ultroser G (Pall, NY), 2 mM L-glutamine and 1% of penicillin and streptomycin, and differentiated in a similar medium where the FCS and Ultroser G supplements were replaced with 2% horse serum. Total RNA from primary myoblast cultures proliferating at 50% and 90% confluency and differentiated for the 6th, 24th, 48th, 72nd and 96th hours was collected using trizol. 2.4. RNA isolation and qPCR studies The time-course periostin mRNA expression levels were investigated by reverse transcriptase coupled quantitative real-time PCR (qPCR). 50–75 mg of TA muscle was rapidly disrupted using a bead-beater and the RNA extraction was accomplished using Trizol reagent upon the manufacturer's recommendations. RNA integrity and quality were assessed via denaturing agarose gel electrophoresis and UV absorbance measurements. 1 μg of total RNA was reverse transcribed into cDNA using Improm II (Promega) reverse transcriptase following manufacturer's protocols. The quantitative gene expression analysis was accomplished using the SYBR Green technique. Equal amounts of cDNA were used for the real-time amplification of the target transcripts using Jumpstart SYBR Green mix in 3 mM final Mg++ concentration (Sigma-Aldrich) according to the manufacturer's recommendations on a Rotorgene 6000 (Corbett Life Science, Australia) fluorometric PCR instrument. Following an initial denaturation of 2 min at 94° C, two-step PCR reaction was conducted for 40 cycles that consisted of 5 s at 94 °C for denaturation and 20 s at 60° C for annealing and elongation. Fluorimetric acquisition on green channel was done at the end of the annealing and elongation step. The sequences of primer pairs are provided in Table 1. The amplification of periostin cDNA was accomplished using a primer pair matching both rat and mouse species spanning exons 16 and 23 (Ensembl Exon accession numbers for mouse and rat are ENSMUSE00000172742, ENSMUSE00000804649 and ENSRNOE00000121871, ENSRNOE00000340025, respectively). This approach yielded the amplification of the six putative alternatively spliced exons. The expression of β-actin was used to normalize the periostin expressions. The relative expressions were further normalized to the expression of the control samples and the results were presented as relative fold changes. The one-way ANOVA test was utilized to test the significance of the periostin expression (p N 0.05). PCR products were run on 3% agarose gels to verify the absence of any non-specific products and primer dimers, and alternatively spliced isoforms were documented. A densitometric analysis was performed on agarose gel images using ImageJ software (http://imagej.nih.gov/ij/) (Schneider et al., 2012). The C2C12 and rat skeletal muscle cDNA amplicons were further analyzed using cycle sequencing with Applied Biosystems big dye terminator 3.1 reagents on an ABI 3130 genetic analyzer. In order to determine the splice variants in mouse C2C12 cells, full length cDNA products were also cloned and sequenced using primers targeting 5′UTR and stop codon. 2.5. Immunostaining and histochemistry Tissue samples were sectioned to 8 μm thickness and investigated by standard immunostaining methods for periostin immunolocalization. Primary anti-developmental type myosin heavy chain (Leica, NLCMHCd corresponding to Myh3 isoform, working dilution 1:50) and

132

C. Özdemir et al. / Gene 553 (2014) 130–139

Table 1 The accession numbers of the relevant genes and the sequences of the primer pairs that are utilized for the cDNA amplifications are provided. Gene

Mouse

Periostin

NM_001198765.1

Rat

NM_001108550.1 Tgfb1

NM_011577.1

NM_021578.2

Col1a1

NM_007742.3

NM_053304.1

β-actin

NM_007393.3

NM_031144.3

anti-desmin (Sigma DE-U-10, working dilution 1:80) mouse monoclonal antibodies were co-stained using primary rabbit polyclonal antibody against rodent periostin (Abcam, ab14041, working dilution: 1:300) and appropriate fluorescent-conjugated secondary antibodies (Molecular Probes, A11036, A11029). 1 μg/ml of DAPI was applied to cover slips as a nuclear stain (Sigma-Aldrich, D9542). All immunofluorescent images were captured using a Leica DFC 320 camera (Leica, Switzerland). Standard protocols for hematoxylin/eosin (H&E) and sirius red stains were conducted on tenotomised samples where appropriate (Kiernan, 1999). 2.6. Western blot Periostin expression during injury repair was documented at the protein level. Total protein extraction was conducted from timecourse samples and concentrations were determined by BCA Protein Assay. For each sample, 50 μg of total protein was resolved on a 12% SDS-PAGE gel and transferred onto nitrocellulose protein membrane (Bio-Rad). Membrane probing was done using primary rabbit polyclonal antibody against rodent periostin (Abcam, ab14041, at a final dilution of 0.5 μg/ml) and mouse monoclonal antibody against α-tubulin (Sigma, T6074, at a final dilution of 0.5 μg/ml). Generation of the signal was accomplished following appropriate horseradish-peroxidaseconjugated secondary antibodies (1:8000; Sigma) using chemiluminescence detection system (Thermo Scientific, SuperSignal West Femto maximum sensitivity substrate). 3. Results The expression of periostin in skeletal muscle regeneration was investigated during the injury repair course following acute necrosis. The observations were further extended and confirmed in models of in vitro differentiation. 3.1. Periostin transcription was transiently induced in regenerating muscle tissue Cardiotoxin is a cocktail of phospholipases which is the main constituent of snake venom. Once injected into the muscle tissue, these phospholipases induce rapid disruption of membrane structures, causing extensive damage to myofibers, sparing the satellite cells. Upon the commencement of necrosis, an extensive mononuclear cell infiltration is induced that peaks in 3 to 4 days. The debris is phagocyted and cleared away by the monocytes and macrophages. Concomitantly, activated satellite cells proliferate and migrate to the damaged area and fuse to restore the myofibrillary structure. In h&e micrographs, these young myofibers were observed as basophilic cell clumps at day 4 (Fig. 1B). Following the initiation of the synthesis of contractile machinery, centrally nucleated myofiber structures with thin eosinophilic cytoplasm were observed (day 6). After the cessation of the mononuclear infiltration and the development of the contractile structures, the

Primer pairs 5′-TTGTTCGTGGCAGCACCTTC-3′ 5′-GTGGTGGCTCTTACAATTTCC-3′ 5′-TCGTTCGTGGCAGCACCTTC-3′ 5′-ATGGGGCTCTTATAATTTCC-3′ 5′-TGAACCAAGGAGACGGAATACAG-3′ 5′-GCCATGAGGAGCAGGAAGG-3′ 5′-CAAGATGTGCCACTCTGACT-3′ 5′-TCTGACCTGTCTCCATGTTG-3′ 5′-GTGCTATGTTGCCCTAGACTTCG-3′ 5′-GATGCCACAGGATTCCATACCC-3′

endomysial space was progressively narrowed (day 8). The restoration of the intact muscle structure was accomplished within 10 to 12 days (Zhao and Hoffman, 2004). We utilized this model to observe the expression of periostin during the regeneration process. Basal periostin expression in skeletal muscle rapidly decreased upon the induction of extensive necrosis of the tissue by the first day and was strongly and significantly induced in the course of injury regeneration starting from day 4, peaking at day 6 by 37 fold. The expression declined to less than 10 fold after day 8, indicative of a return to the basal levels (Fig. 1). Western blot analysis was conducted to investigate the kinetics of periostin protein in regenerating skeletal muscle. This analysis revealed a single immunoreactive band corresponding to approximately 100 kDa weight in rat muscle samples. Results revealed a rapid decline in periostin abundance in the necrotic muscle following damage at day 1. A gradual increase in periostin levels was observed along days 4, 6 and 8. Control periostin levels were restored in day 10 and 14 samples (Fig. 1C). Spatial localization of periostin was further investigated in order to pinpoint the source expression. In the course of regeneration, the upregulation gained significance at day 4 and immunostaining deduced that at day 4, periostin expression was restrictively localized to the early regenerating myofibers that were positive for neonatal (developmental) myosin heavy chain (Fig. 2, top panel). At the peak expression (day 6), periostin was localized both to the newly regenerating fibers as well as the MHCd negative stromal cells residing in the endomysial space (Fig. 2, middle panel). However, by day 8, periostin protein was solely observed in the endomysial space (Fig. 2, lower panel). Centrally nucleated and desmin (+) fibers were devoid of any periostin staining. Periostin was present in the endomysial space as deposits in the extracellular matrix, delineating the rim of the desmin positive sarcoplasm (Fig. 2 lower panel higher magnification). Periostin immunolocalization was also strongly detected in some stromal cells residing in the endomysial space that were lacking desmin (nuclei are marked with asterisk). No remarkable periostin expression was observed in infiltrating mononuclear cells. 3.2. Periostin expression was induced in tenotomy immobilized skeletal muscle In order to achieve a comparative evaluation, spatial and temporal expression of periostin was further investigated in a rat model which illustrates skeletal muscle fibrosis. The release of preload by tenotomy induces rapid endomysial fibrosis as well as myofibrillary atrophy (Jozsa et al., 1990). Four weeks of follow-up on periostin expression deduced 15 to 22 fold upregulation (Fig. 3A) in tenotomised soleus muscle. Investigation of periostin localization on tissue sections revealed that periostin was stringently present in the perimysium in control muscle. However, in tenotomy induced fibrosis, periostin deposition was prominent surrounding the fibers as well as increased localization in perimysium (Fig. 3B). Sirius red staining was also employed to demonstrate fibrotic collagen deposition in response to tenotomy.

C. Özdemir et al. / Gene 553 (2014) 130–139

133

Fig. 1. Relative fold change of periostin mRNA expression is plotted in the course of skeletal muscle injury repair normalized to the average of control samples (A). Error bars represent standard deviation of independent biological replicates (n = 3 for each time point) and asterisk indicates statistically significant deviation compared to control samples (p b 0.05). Typical representative H&E staining micrographs of the time-course injury regeneration model are shown (B). Size bar represents 50 μm. Western blot analysis of periostin expression revealed rapid breakdown of periostin at day 1 and gradual restoration to the control levels following day 10 (C). α tubulin is used to demonstrate equal loading.

3.3. Periostin was upregulated in the course of in vitro myoblast differentiation In vitro differentiation of myoblasts to myotubes shared common features of in vivo injury regeneration such as the transcriptional program and the cascade of cellular events. This in vitro assay served to observe the three cardinal differentiation stages; proliferation, fusion and myotube maturation. Thus, the observed transcriptional regulation of periostin in injury regeneration was further confirmed separately on C2C12 mouse myoblast cells and primary rat myoblasts to document the kinetics of mRNA expression in the course of differentiation. In C2C12 myoblasts, periostin expression was detected in the proliferation phase. Periostin expression level was increased the induction of differentiation exhibiting a significant 4.2 fold upregulation (Fig. 4A). At 72 h after differentiation, initiated expression of periostin was more than 50 fold over the baseline (Fig. 4A). The same observation was also valid for primary rat myoblasts. At 48 h following the induction of differentiation, the upregulation reached significance and was confined to 11.5 and 17.1 fold over the proliferating cells in 48 and 72 h respectively (Fig. 4B). 3.4. Expression profile of periostin isoforms during injury repair and differentiation The time-course investigation of periostin mRNA expression was conducted using a primer pair encompassing exons 16 to 23.

Current knowledge implies that these last five exons are subjected to alternative splicing in mice and humans (Ensembl Access Nos.: ENSMUSG00000027750 and ENSG00000133110, respectively). Agarose gel analysis of PCR products revealed concomitant expression of various isoforms during myoblast differentiation (Fig. 5). Expression profiles of these different isoforms were further documented on both rat primary myoblasts (Fig. 5A) and C2C12 cells (Fig. 5B). A densitometric analysis of the gel images was conducted to elucidate the ratio of expression of each band that corresponded to different isoform(s). Including independent biological replicates, this approach revealed that the expression ratios for different isoforms were stable along the progression of differentiation, and splicing patterns were not altered (Fig. 5). Due to the fact that the PCR product sizes of some of these isoforms were very close and agarose gel electrophoresis was not adequate for discrimination between all of the splice variants, cDNA amplicons were subjected to sequence analysis to investigate the dominating periostin isoforms in skeletal muscle. cDNA sequencing confirmed that one single cDNA isoform dominated the rat skeletal muscle transcripts that encompasses exons 16, 18, 19, 20, 22 and 23 (Fig. 6A). Furthermore, in rats, sequencing deduced the presence of a previously unreported 22nd exon also found in mouse and human genes. This rat transcript was the exact homolog of Postn-002 in mouse (Ensembl Access No.: ENSMUST00000117373) and designated as R1 (GenBank Access No.: KM117173). Sequence analysis of the mouse cDNA sequence and full-length clones confirmed the

134

C. Özdemir et al. / Gene 553 (2014) 130–139

Fig. 2. Periostin immunolocalization was investigated along the injury repair process using fluorescent immunostaining. Periostin was stained in red, dMHC (days 4 and 6) and desmin (day 8) in green, the nuclei were counterstained using DAPI in blue. Periostin immunolocalization was confined to the newly forming myofibers expressing dMHC in day 4. Two emerging myofibers with sarcoplasmic periostin content are shown in higher magnification images (magnification area is delineated by a dashed box in the overlay image). By day 6, periostin was detectable in dMHC positive fibers and stromal cells (mid panel). By day 8, young, centrally nucleated desmin positive fibers were devoid of periostin. Immunostaining was strongly localized to the endomysial space. Higher magnification images show periostin deposits in the extracellular matrix, delineating the rim of the desmin positive fibers (arrows). Periostin immunolocalization was strongly detected in stromal cells residing in the endomysial space (nuclei are marked with asterisk). Size bar represents 50 μm.

presence of two clones; one was Postn-002, which corresponds to the 463 bp band in Fig. 5B. The other was a previously unreported isoform with a full-length N-terminal (spanning exons 1 to 15) and a C-terminal excluding exons 17, 20 and 21 that corresponded to 385 bp band on agarose gel (M1 in Figs. 5B and 6B, GenBank Accession No.: KM117171). Sequencing of the full-length clones also revealed a previously unreported isoform lacking exons 20 and 21, corresponding to the 466 bp band in Fig. 5B (designated as M2, GenBank Access No.: KM117172). A reconstruction of the genomic sequence and the splicing patterns are presented in Fig. 6D. 4. Discussion Chronic degenerative conditions of skeletal muscle have distinctive features associated with altered muscle function and morphology. These may be initiated by genetic defects in structural proteins, such as in the dystrophies or diminished regenerative function in sarcopenia. The hallmarks of chronic degeneration and disrupted tissue architecture are myofibrillary atrophy, endomysial fibrosis and fatty infiltration

(Klingler et al., 2012). Among these, fibrosis is one key issue that is gaining higher importance in the development of new therapeutic strategies. Firstly, fibrosis is the restrictive component of degeneration inhibiting kinetic properties of the skeletal muscle. Secondly, disrupted tissue architecture and endomysial fibrosis are deteriorating the satellite cell niche inhibiting reaching of external regenerative signals as well as satellite cell activation (Boldrin et al., 2009). Thirdly, any future remedies attempting gene correction will have to cross the fibrotic barrier in order to achieve a successful delivery into the myofibers. From this viewpoint, molecular modulators, markers or regulatory pathways associated with degeneration and fibrosis are one primary goal of research targeting skeletal muscle disorders (Zhou and Lu, 2010). High throughput transcriptome studies provide a valuable tool for the observation of transcriptional events and regulatory pathways that impact tissue architecture and remodeling. Converging in silico results of different transcriptome studies on the upregulation of periostin directed us to investigate its role in the course of acute and chronic degenerative models as well as in vitro differentiation (Supplementary Fig. 1).

C. Özdemir et al. / Gene 553 (2014) 130–139

135

Fig. 3. Periostin mRNA expression was investigated in tenotomy induced skeletal muscle fibrosis by using qPCR. Periostin expression levels were normalized to the average of the control samples and provided as relative fold change of expression (A). Error bars represent standard deviation of independent biological replicates (n = 3 for each time point). Periostin immunolocalization in control and two weeks tenotomized skeletal muscle samples was investigated using fluorescent immunostaining (B) and sirius red staining was employed to demonstrate collagen deposition. Periostin was stained in red and desmin in green. In control samples, limited periostin deposition could be visualized in the perimysium (arrows). However, in tenotomy, periostin was highly abundant in the endomysium surrounding the fibers (asterisk) as well as the perimysial space. Periostin was not detectable in fibers. Size bar represents 50 μm.

4.1. Myofibers express periostin to shape ECM during regeneration and differentiation Periostin is a matricellular protein that is linking tenascin, fibronectin and collagen in the extracellular matrix (Kudo, 2011). Much of the extracellular matrix remodeling takes place during the development and remodeling in adult life is generally associated with a pathologic process. Likewise other connective tissues that are subjected to mechanical tension, in skeletal muscle periostin also resides in the extracellular matrix. This study primarily addresses periostin expression along the time-course injury repair model in skeletal muscle. Western blot studies show that the total tissue periostin was rapidly destroyed upon the induction of necrosis in the skeletal muscle and the protein levels were gradually restored in about 10 days. The expression studies further deduced that a transient upregulation of periostin mRNA expression during days 4, 6 and 8 was functionally restoring the periostin content of the muscle. Immunostaining could capture the newly emerging young myofibers as the initial source of the periostin expression. Apparently, these young regenerating fibers were expressing (and secreting) periostin to shape their own extracellular matrix while their fusion window was still open. Periostin could be traced around the maturated fibers expressing desmin but lacking MHCd indicating a deposition to the extracellular matrix during the regeneration.

The transcriptional events of satellite cell activation following acute injury that initiate proliferation, migration and fusion are commonly shared by myogenesis. Thus, in vitro differentiation of skeletal muscle precursors or myoblasts provides a model to compare the events during stages that can be summarized as proliferation, fusion and myotube maturation. Upregulation of periostin expression both in rat primary myoblasts and C2C12 embryonic myoblast cells, along with fusion, further support the claim that periostin expression is a function of newly emerging muscle fibers. This induction in transcription was relatively limited in primary myoblasts (16 fold) compared to C2C12 cells (60 fold). One conceivable explanation of this limitation is that the dilution of the upregulation is confined to the myoblast compartment by the contaminating fibroblast cell population in primary cultures. Although it is known that fibroblasts do express periostin (Kudo, 2011), this is only confined to the activated state and not valid for proliferative conditions of the tissue culture environment. This comparative analysis shows that the commonly shared aspects of skeletal muscle injury regeneration and myogenesis extend to involve key extracellular matrix components as well. 4.2. Transient periostin expression during regeneration is not associated with fibrosis Periostin is a member of TGF beta family proteins, with a putative role associated with pathologic fibrotic events throughout the body

136

C. Özdemir et al. / Gene 553 (2014) 130–139

Fig. 4. Periostin mRNA expression was quantified in the course of differentiation of C2C12 (A) and primary rat myoblasts (B). Expression was normalized to the average expression of proliferating cells (50% confluence) and plotted to represent a relative fold change. Error bars represent standard deviation of independent experimental replicates (n = 3 for each time point). Upon switching to the differentiation medium, an induction of expression was observed that reached statistical significance in 24 h in C2C12 cells and 48 h in primary rat myoblasts (* = p b 0.05). For the sake of clarity, typical representative phase contrast micrographs of selected time-points were also provided.

(Conway et al., 2014). Periostin was also shown to be associated with fibrosis in the skeletal muscle and periostin knockout genotypes exhibited diminished fibrosis in dystrophic mice (Lorts et al., 2012). These previous reports have localized periostin expression to the activated fibroblasts associated with irreversible fibrotic events. It is known that cardiotoxin (or notexin) induced skeletal muscle injury regenerates without fibrosis (Zhao and Hoffman, 2004). The results presented in this study show that, besides chronic degenerative conditions, periostin is also associated with the events that assist the remodeling of the extracellular matrix and basement membrane during early skeletal muscle repair. In order to support this postulate further, we also sought to analyze TGF beta 1 and Collagen 1A1 mRNA expressions during the peak

periostin expression days 4, 6 and 8 in the course of injury regeneration. As expected, periostin expression correlated to TGF beta 1 (r = 0.75) but no significant upregulation of collagen expression was observed (Supplementary Fig. 2). This finding indicates that a previously established association between TGF beta and periostin in mesenchymal cells of the lung is also valid for the skeletal muscle (Naik et al., 2012). However, the consequences are not associated with fibrosis. In order to achieve a comparative evaluation of periostin expression in a model with irreversible fibrosis, we utilized tenotomy-induced immobilization. Periostin expression was upregulated to 20 fold and exhibited a steady course along with the establishment of fibrosis. Contrasting to acute injury regeneration, in tenotomised muscle, periostin

C. Özdemir et al. / Gene 553 (2014) 130–139

137

Fig. 5. qPCR products designating various periostin isoforms were separated on 3% agarose gel for primary rat myoblasts (A) and C2C12 myoblast cells (B). Randomly selected independent biological replicates were loaded onto adjacent lanes. The gel images were subjected to densitometric analysis and average expression ratios of the replicates were plotted. A single rat transcript dominated the primary rat myoblasts, while multiple splice variants were observed in C2C12 cells. In the course of differentiation, the relative expression ratio of the variants did not exhibit any significant variation.

was exclusively localized to the endomysial and perimysial fibrotic space, exhibiting a cloudy and patchy appearance in close vicinity of the stromal cells but strongly excluding myofibers. This observation showed that separate events were driving these distinct expression origins along acute injury repair and a chronic degenerative condition. This deviation further suggests that periostin expression originating from regenerating (or maturating) fibers is a temporary process aiding the architectural remodeling of the extracellular matrix; it is not stable and does not convoy any fibrotic events. However, the expression that was confined to the stromal cells is a distinct steady process associated with irreversible fibrosis. These non-myofiber stromal cells that express periostin is of particular interest. During acute injury regeneration, periostin expression was also strongly observed in a number of stromal cells (especially at day 8). Two plausible scenarios may explain these stromal cells. Either these are the activated fibroblasts, shaping the stroma and would become quiescent upon cessation of the activation signal (most likely to be TGF beta) or these cells are of the recently identified heterologous population of the fibro-adipogenic precursors (Uezumi et al., 2011).

homolog include a 22nd exon, and one single transcript is dominantly expressed in muscle. The C-terminal of the periostin protein as well as the corresponding genomic segment is of particular interest. In rat, mouse and human, exons 17 to 22 are of symmetrical nature as well as having similar lengths. Furthermore, these exons share outstanding sequence similarity at the DNA level and are alternatively spliced. Various combinations of three of these six exons depict six alternative splicing variants in mice. However, this C-terminal does not harbor any known functional domain. This is also in line with a previous report describing the tendency of symmetrical exons not to disrupt protein domain structures (Magen and Ast, 2005). We show that the expression kinetics of various mouse (and rat) isoforms were stable throughout differentiation. Thus, speaking for periostin, it is highly likely that co-expression of different alternatively spliced isoforms (such as the mice) or expression of different isoforms among different species may not imply any major impact on the protein function.

5. Conclusion 4.3. Periostin gene structure and isoforms We also sought to illuminate expression variants of periostin in rodent muscle. We show that the rat periostin transcript and the mouse

Periostin was previously identified as a physiological mediator in connective tissue that plays a role in the maturation of extracellular matrix. Moreover, periostin also exerts a deleterious impact on pathological conditions complicated with fibrosis. While this latter condition

138

C. Özdemir et al. / Gene 553 (2014) 130–139

Fig. 6. Sequencing of the cDNA amplicons revealed previously unreported periostin mRNA (R1) in primary rat myoblasts composed of all exons excluding 17, 21, and previously unreported exon 22 (A). In C2C12 myoblasts, the Postn-002 isoform dominated the periostin cDNA pool together with a previously unreported splice variant (M1) lacking exons 17, 20 and 21 (B). Sequencing of the full length clones also demonstrated the presence of a novel isoform (M2) lacking exons 20 and 21 (C). The newly identified rat and mouse transcripts are presented (D).

was described for skeletal muscle, in this study, we report that periostin is also produced and secreted both by myoblasts during differentiation and regenerating myofibers during architectural remodeling. These results show that development and regeneration also share common extracellular matrix components besides transcriptional actors. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.gene.2014.10.014. Acknowledgment This study was supported by the research grant from the Scientific and Technological Research Council of Turkey, TUBITAK (Grant No.: SBAG-105S364). The authors also thank Audrey Richardson for the English language editing of the manuscript. References Bakay, M., Wang, Z., Melcon, G., Schiltz, L., Xuan, J., Zhao, P., Sartorelli, V., Seo, J., Pegoraro, E., Angelini, C., Shneiderman, B., Escolar, D., Chen, Y.W., Winokur, S.T., Pachman, L.M., Fan, C., Mandler, R., Nevo, Y., Gordon, E., Zhu, Y., Dong, Y., Wang, Y., Hoffman, E.P., 2006. Nuclear envelope dystrophies show a transcriptional fingerprint suggesting disruption of Rb-MyoD pathways in muscle regeneration. Brain 129, 996–1013. Boldrin, L., Zammit, P.S., Muntoni, F., Morgan, J.E., 2009. Mature adult dystrophic mouse muscle environment does not impede efficient engrafted satellite cell regeneration and self-renewal. Stem Cells 27, 2478–2487. Braun, T., Gautel, M., 2011. Transcriptional mechanisms regulating skeletal muscle differentiation, growth and homeostasis. Nat. Rev. Mol. Cell Biol. 12, 349–361. Charge, S.B., Rudnicki, M.A., 2004. Cellular and molecular regulation of muscle regeneration. Physiol. Rev. 84, 209–238.

Chen, I.H., Huber, M., Guan, T., Bubeck, A., Gerace, L., 2006. Nuclear envelope transmembrane proteins (NETs) that are up-regulated during myogenesis. BMC Cell Biol. 7, 38. Conway, S.J., Izuhara, K., Kudo, Y., Litvin, J., Markwald, R., Ouyang, G., Arron, J.R., Holweg, C.T., Kudo, A., 2014. The role of periostin in tissue remodeling across health and disease. Cell. Mol. Life Sci. 71, 1279–1288. Dobaczewski, M., Gonzalez-Quesada, C., Frangogiannis, N.G., 2010. The extracellular matrix as a modulator of the inflammatory and reparative response following myocardial infarction. J. Mol. Cell. Cardiol. 48, 504–511. Hamilton, D.W., 2008. Functional role of periostin in development and wound repair: implications for connective tissue disease. J. Cell Commun. Signal. 2, 9–17. Jozsa, L., Kannus, P., Thoring, J., Reffy, A., Jarvinen, M., Kvist, M., 1990. The effect of tenotomy and immobilisation on intramuscular connective tissue. A morphometric and microscopic study in rat calf muscles. J. Bone Joint Surg. (Br.) 72, 293–297. Kiernan, J.A., 1999. Histological and Histochemical Methods: Theory and Practice, 3rd ed. Butterworth Heinemann, Oxford; Boston. Klingler, W., Jurkat-Rott, K., Lehmann-Horn, F., Schleip, R., 2012. The role of fibrosis in Duchenne muscular dystrophy. Acta Myol. 31, 184–195. Kocaefe, C., Balci, D., Hayta, B.B., Can, A., 2010. Reprogramming of human umbilical cord stromal mesenchymal stem cells for myogenic differentiation and muscle repair. Stem Cell Rev. 6, 512–522. Kudo, A., 2011. Periostin in fibrillogenesis for tissue regeneration: periostin actions inside and outside the cell. Cell. Mol. Life Sci. 68, 3201–3207. Lorts, A., Schwanekamp, J.A., Baudino, T.A., McNally, E.M., Molkentin, J.D., 2012. Deletion of periostin reduces muscular dystrophy and fibrosis in mice by modulating the transforming growth factor-beta pathway. Proc. Natl. Acad. Sci. U. S. A. 109, 10978–10983. Magen, A., Ast, G., 2005. The importance of being divisible by three in alternative splicing. Nucleic Acids Res. 33, 5574–5582. Maruhashi, T., Kii, I., Saito, M., Kudo, A., 2010. Interaction between periostin and BMP-1 promotes proteolytic activation of lysyl oxidase. J. Biol. Chem. 285, 13294–13303. Merle, B., Garnero, P., 2012. The multiple facets of periostin in bone metabolism. Osteoporos. Int. 23, 1199–1212. Moran, J.L., Li, Y., Hill, A.A., Mounts, W.M., Miller, C.P., 2002. Gene expression changes during mouse skeletal myoblast differentiation revealed by transcriptional profiling. Physiol. Genomics 10, 103–111.

C. Özdemir et al. / Gene 553 (2014) 130–139 Naik, P.K., Bozyk, P.D., Bentley, J.K., Popova, A.P., Birch, C.M., Wilke, C.A., Fry, C.D., White, E.S., Sisson, T.H., Tayob, N., Carnemolla, B., Orecchia, P., Flaherty, K.R., Hershenson, M.B., Murray, S., Martinez, F.J., Moore, B.B., Investigators, C., 2012. Periostin promotes fibrosis and predicts progression in patients with idiopathic pulmonary fibrosis. Am. J. Physiol. Lung Cell. Mol. Physiol. 303, L1046–L1056. Norris, R.A., Moreno-Rodriguez, R., Hoffman, S., Markwald, R.R., 2009. The many facets of the matricelluar protein periostin during cardiac development, remodeling, and pathophysiology. J. Cell Commun. Signal. 3, 275–286. Oka, T., Xu, J., Kaiser, R.A., Melendez, J., Hambleton, M., Sargent, M.A., Lorts, A., Brunskill, E.W., Dorn II, G.W., Conway, S.J., Aronow, B.J., Robbins, J., Molkentin, J.D., 2007. Genetic manipulation of periostin expression reveals a role in cardiac hypertrophy and ventricular remodeling. Circ. Res. 101, 313–321. Porter, J.D., Merriam, A.P., Leahy, P., Gong, B., Feuerman, J., Cheng, G., Khanna, S., 2004. Temporal gene expression profiling of dystrophin-deficient (mdx) mouse diaphragm identifies conserved and muscle group-specific mechanisms in the pathogenesis of muscular dystrophy. Hum. Mol. Genet. 13, 257–269. Rani, S., Barbe, M.F., Barr, A.E., Litvin, J., 2009. Induction of periostin-like factor and periostin in forearm muscle, tendon, and nerve in an animal model of work-related musculoskeletal disorder. J. Histochem. Cytochem. 57, 1061–1073. Schneider, C.A., Rasband, W.S., Eliceiri, K.W., 2012. NIH Image to ImageJ: 25 years of image analysis. Nat. Methods 9, 671–675.

139

Tomczak, K.K., Marinescu, V.D., Ramoni, M.F., Sanoudou, D., Montanaro, F., Han, M., Kunkel, L.M., Kohane, I.S., Beggs, A.H., 2004. Expression profiling and identification of novel genes involved in myogenic differentiation. FASEB J. 18, 403–405. Uezumi, A., Ito, T., Morikawa, D., Shimizu, N., Yoneda, T., Segawa, M., Yamaguchi, M., Ogawa, R., Matev, M.M., Miyagoe-Suzuki, Y., Takeda, S., Tsujikawa, K., Tsuchida, K., Yamamoto, H., Fukada, S., 2011. Fibrosis and adipogenesis originate from a common mesenchymal progenitor in skeletal muscle. J. Cell Sci. 124, 3654–3664. Wang, I.M., Zhang, B., Yang, X., Zhu, J., Stepaniants, S., Zhang, C., Meng, Q., Peters, M., He, Y., Ni, C., Slipetz, D., Crackower, M.A., Houshyar, H., Tan, C.M., Asante-Appiah, E., O'Neill, G., Luo, M.J., Thieringer, R., Yuan, J., Chiu, C.S., Lum, P.Y., Lamb, J., Boie, Y., Wilkinson, H.A., Schadt, E.E., Dai, H., Roberts, C., 2012. Systems analysis of eleven rodent disease models reveals an inflammatome signature and key drivers. Mol. Syst. Biol. 8, 594. Yaffe, D., Saxel, O., 1977. A myogenic cell line with altered serum requirements for differentiation. Differentiation 7, 159–166. Zhao, P., Hoffman, E.P., 2004. Embryonic myogenesis pathways in muscle regeneration. Dev. Dyn. 229, 380–392. Zhou, L., Lu, H., 2010. Targeting fibrosis in Duchenne muscular dystrophy. J. Neuropathol. Exp. Neurol. 69, 771–776.