Research Article: Hmgb3 Is Regulated by MicroRNA-206 during Muscle Regeneration

Date Published: August 17, 2012

Publisher: Public Library of Science

Author(s): Simona Maciotta, Mirella Meregalli, Letizia Cassinelli, Daniele Parolini, Andrea Farini, Giulia Del Fraro, Francesco Gandolfi, Mattia Forcato, Sergio Ferrari, Davide Gabellini, Silvio Bicciato, Giulio Cossu, Yvan Torrente, Sandro Banfi.


MicroRNAs (miRNAs) have been recently involved in most of human diseases as targets for potential strategies to rescue the pathological phenotype. Since the skeletal muscle is a spread-wide highly differentiated and organized tissue, rescue of severely compromised muscle still remains distant from nowadays. For this reason, we aimed to identify a subset of miRNAs major involved in muscle remodelling and regeneration by analysing the miRNA-profile of single fibres isolated from dystrophic muscle, which was here considered as a model of chronic damage.

The miRNA-signature associated to regenerating (newly formed) and remodelling (resting) fibres was investigated in animal models of muscular dystrophies and acute damage, in order to distinguish which miRNAs are primary related to muscle regeneration. In this study we identify fourteen miRNAs associated to dystrophic fibres responsible for muscle regeneration and remodelling, and confirm over-expression of the previously identified regeneration-associated myomiR-206. In particular, a functional binding site for myomiR-206 was identified and validated in the 3′untranslated region (3′UTR) of an X-linked member of a family of sequence independent chromatin-binding proteins (Hmgb3) that is preferentially expressed in hematopoietic stem cells. During regeneration of single muscle fibres, Hmgb3 messenger RNA (mRNA) and protein expression was gradually reduced, concurrent with the up-regulation of miR-206.

Our results elucidate a negative feedback circuit in which myomiR-206 represses Hmgb3 expression to modulate the regeneration of single muscle fibres after acute and chronic muscle damage. These findings suggest that myomiR-206 may be a potential therapeutic target in muscle diseases.

Partial Text

MiRNAs are a class of short non-coding RNAs that take part in mastering the balance of gene-regulating networks by binding to 3′UTR of target mRNAs and inhibiting their expression [1]. The disclosure of these small RNA molecules introduced a new labyrinthine dimension to gene regulation and gave the opportunity to deepen our understanding on many biological processes. MiRNAs were in fact demonstrated to be involved in most of biological events and to finely regulate them [1], [2]. Even more importantly, they were found dysregulated in many diseases [3], [4], [5], [6], [7], [8] and apart from their specific role, their normalization is considered a potential method of intervention in patients. Focusing on the involvement of miRNAs in muscle development and myogenesis, a restricted group of muscle-enriched miRNAs, also called myomiRs (miR-1, miR-133, miR-206 and miR-208), was demonstrated fundamental for muscle physiology and plasticity [9], [10], [11]. In line with a growing characterization of the myomiR network, researchers started to investigate the role of miRNAs in muscle degeneration, which turns out to be very challenging if considering the absence of an efficient therapy for patients affected by most of primary muscular disorders. A faithful model for studying muscle damage and regeneration, is represented by muscular dystrophies (MDs) since they are a group of diseases characterized by muscle wasting and weakness due to defects in structural proteins expressed in the skeletal muscle [12], [13]. In particular MDs are proposed as two-tiered diseases with acute large amount of myofibre necrosis resulting from growth spurts or damaging exercise, superimposed upon a background of a chronic low level of damage, with different factors contributing to these two situations [14]. When Eisenberg et al. analyzed the miRNome of muscle biopsies from patients affected by 10 major muscular disorders including MDs, 185 miRNAs were found to be differentially expressed in diseased muscle [15]. This evidence introduced an additional dimension to muscle degeneration, denouncing a strong participation of miRNAs in muscle degeneration. Soon after this discovery, Capogrossi and colleagues further investigated the role of miRNAs in muscle damage by studying their expression profile in an animal model of Duchenne Muscular Dystrophy (DMD), the mdx mouse, and of acute ischemia [16]. This study demonstrated a common signature of DMD and ischemic muscle, outlining three different families of DMD-signature miRNAs: inflammatory (miR-222 and miR-223), degenerative (miR-1, miR-29c, and miR-135a) and regenerative (miR-31, miR-34c, myomiR-206, miR-335, miR-449, and miR-494). Finally, Cacchiarelli and colleagues demonstrated the mis-regulation of several miRNAs in Duchenne condition (miR-1, myomiR-133, miR-29 and miR-30) as consequence of differential HCDAC nytrosilation state [17]. It is however important to underline that whole-muscle based-analyses, as studies mentioned above, supply a wide image of the miRNA-dysregulation associated to damaged muscle without performing a deep analysis of the muscular miRNAs involved in these processes. Dystrophic muscle is in fact characterized by inflammatory infiltrations and overcoming fibrotic- and adipose-tissue that progressively substitute skeletal muscle [13]. The present study was addressed to identify muscular miRNAs involved in muscle remodelling and regeneration whose in vivo modulation might help recovering from injury. In order to reach this aim, we initially profiled the miRNome of single muscle fibres isolated from the mdx mouse. The mdx mouse originates from a spontaneous mutation in exon 23 of the dystrophin gene resulting in a cyclic process of fibre necrosis and regeneration, and giving rise to progressive muscle wasting and replacement by connective and adipose tissue [18], [19], [20]. Although the mdx mutants have a belated onset of mild clinical symptoms compared with both Duchenne and Becker muscular dystrophy, they have several histopathological features in common [21], [22]. We thus validate our miRNome results in single muscle fibres isolated from other animal models of MDs such as the alpha-Sgca-null and the FRG1 over-expressing mice, respectively animal models of a limb-girdle 2D (LMGD-2D) [23], [24] and of Facioscapulohumeral dystrophy (FSHD) [25]. The α-Sgca null mouse represents the first engineered animal model for autosomal recessive MD with a primary sarcoglycan gene defect [23]. Like the mdx mouse, the main histological changes of α-Sgca-null muscles include extensive central nucleation, increased variability of muscle fibre diameter, and the persisted presence of degeneration and regeneration [23]. Regarding muscle impairment, absence of α-sarcoglycan was demonstrated to perturb more fast-twitch than slow-twitch fibres. The FRG1 over-expressing mouse is a transgenic mouse characterized by strong up-regulation of human FRG1, causing muscular dystrophy similar to FSHD (abnormal spinal curvature, progressive muscle dystrophy, skeletal muscle atrophy, differential involvement of muscle types similar to FSHD) [25]. Basing on several measurements (including fibre size variation, number of internal nuclei, percentage of connective tissue, number of necrotic fibres and number of regenerating fibres), this animal model demonstrated different distribution of muscle impairment. In particular, trapezius (TPZ) was defined as the most affected muscle in FRG1 high over-expressing mice and FSHD patients [25]. We also verified miRNAs associated to regenerating single muscle fibres after acute damage. Strategies that have been successfully used for this purpose include crush [26], freeze [27], or chemically induced injury [28]. Perhaps the most extensive and reproducible muscle injury is the delivery of cardiotoxin (CTX; purified from the venom of the Naja nigricollis snake) into the skeletal muscle of the mouse [28]. In this condition it is possible to induce local acute muscle damage with a muscle degeneration of about 80–90% [25] and a successive regeneration which restore muscle integrity and homeostasis [29]. We identify fourteen miRNAs associated to dystrophic fibres (miR-15b, miR-17, miR-21, miR-27a, miR-31, miR-128a, miR-142-5p, miR-199a-5p, miR-199b, miR199b*, miR-206, miR-221, miR-223 and miR-335-5p) that may mediate muscle regeneration and remodelling in animal models of MDs and acute muscle damage, and confirm over-expression of the previously identified regeneration-associated myomiR-206. Using a target-gene prediction-analysis with 6 different computational algorithms (PITA, TargetScan, PicTar, ElMMo, miRDb and miRanda), we identified and validated a functional binding site for miR-206 in the 3′ untranslated region (3′UTR) of an X-linked member of a family of sequence-independent chromatin-binding proteins (Hmgb3) [30]. Importantly, Hmgb3 expression was reduced during muscle regeneration concurrent with the up-regulation of myomiR-206 in dystrophic muscle. All these data demonstrate for the first time the involvement of Hmgb3, an Hmg-box family transcription factor, in muscle regeneration.

Following the discovery of miRNAs, their participation was investigated in almost all biological processes and, even more importantly, their central role in gene-expression regulation was implicated in many human diseases [3], [4], [5], [6], [7], [8] Regarding this, in the recent years many efforts were focused to finely characterize the role of miRNAs in myogenesis, so that now the miRNA biogenesis is considered necessary for proper muscle development and a restricted number of miRNAs, known as myomiRs (miR-1, miR-133, miR-181, miR-206, miR-208), is considered as integral part of muscle biology [9], [11]. Starting from these evidences, Eisenberg et al. demonstrated that muscle biopsies affected by 10 major primary muscular disorders are characterized by the dysregulation of several miRNAs, opening the possibility to use these small RNA molecules as therapeutic targets or potential biomarkers [15]. Nevertheless more investigation is required to finely tune myogenesis through the modulation of miRNAs. With the present study we wanted to identify miRNAs that are modulated by muscle fibres in response to damage in order to aloud muscle remodelling and regeneration. This purpose was reached by analysing the miRNome of muscle fibres isolated from dystrophic and CTX-injected muscle. Dystrophic muscle is characterized by progressive muscle wasting and weakness, representing an animal model of chronic muscle damage, while CTX-injected muscle is an animal model of acute damage. Importantly single-fibre based-analysis avoided contamination from non-muscle tissue, focusing the investigation on resting (remodelling) and newly-formed (regenerating) muscle fibres. In the present study three animal models of MDs were taken in consideration: the mdx mouse, animal model of DMD; the α-Sgca null mouse, animal model of LMGD-2D; and the FRG1 high over-expressing mouse, animal model with a dystrophy similar to FSHD. More in particular, we started by the miRNome of single fibres isolated from different muscles (TA, DIA and VA) of 3½ months-old mdx mouse. The dysregulated miRNAs associated to the mdx mouse were then verified during disease progression of this animal model and in single fibres of α-Sgca null mice (SOL, EDL and DIA) and FRG1 over-expressing mice (TPZ, DIA and VA) in order to evaluate their correlation with genetics of tested dystrophic mice. Dystrophic muscle fibres isolated from different animal model of MDs were commonly characterized by the over-expression of several miRNAs (miR-15b, miR-21, miR-27a, miR-31, miR-128a, miR-142-5p, miR-199a-5p, miR-199b, miR199b*, miR-206, miR-221, miR-223 and miR-335-5p) with an expression profile strictly dependent on muscle impairment and damage accumulation (Figure 7). In support to this, when the dysregulated miRNAs were analyzed during disease progression of the dystrophic mdx mouse (newborn, 3½ and 6 months-old), their expression levels were found similar to control levels in newborn mdx mice and heterogeneously modulated among the different muscles of older mdx mice, reflecting the heterogeneous distribution of muscle impairment. Taken together, these evidences also demonstrate no link between dystrophic miRNA-profile and the genetic causative defects of the different MDs, excluding the possibility to use any of the muscular miRNAs mentioned above as biomarker of a specific MD. Otherwise the over-expression of these miRNAs might function as a biomarker of chronic muscle damage. Surprisingly, apart from the regeneration-associated myomiR-206, no known myomiRs were found dysregulated in dystrophic single muscle fibres. Otherwise a group of miRNAs recently correlated to myogenesis (miR-27a, miR-31, miR-221) [33], [37], [38] were confirmed to be mis-modulated in dystrophic single muscle fibres. In comparison with data already published by Eisenberg et al. and Greco et al. [15], [16], single-fibres based-analysis allowed the discovery of new muscular miRNAs whose expression levels are up-regulated in dystrophic murine muscle (miR-15b, miR-27a, miR-128a and miR-199b*). For this reason we verified their expression in single muscle fibres isolated from biopsies of DMD patients. Analyses were also extended to myomiR-206 since its up-regulation in human dystrophic muscle was only confirmed by Greco et al. and not by Eisenberg et al. [15], [16]. MiRNAs were quantified in human muscle fibres of 12 control subjects and 18 DMD patients, confirming the over-expression of miR-17, miR-27a and miR-206 in diseased muscle.