Among the members of class IIa, HDAC4 plays crucial functions in striated muscles. Increased expression of HDAC4 has been detected in skeletal muscle in different diseases, such as Duchenne Muscular Dystrophy (DMD)
[27] and Amyotrophic Lateral Sclerosis (ALS)
[28][29]: importantly, the observations in pre-clinical models have been validated in patients. Despite binding and repressing the activity of two major myogenic factors, i.e., MEF2
[30] and SRF
[31], mice harboring a skeletal-muscle specific deletion of
Hdac4 are viable and do not display obvious defects in skeletal muscle
[32]. While class IIa HDACs play redundant roles in the establishment of the metabolic pattern of skeletal muscle fibers, by repressing MEF2
[33], HDAC4 per se is necessary and sufficient to mediate responses upon different stimuli in skeletal muscle. For instance, deletion of
Hdac4 in differentiating skeletal muscle cells via myogenin:Cre recombinase, hampers muscle regeneration, due to the release of soluble factors that inhibit muscle precursor cell differentiation
[34]; if the deletion of
Hdac4 occurs earlier in the myogenic cells, such as in Pax7
+ cells, it compromises muscle stem cell (MuSC) proliferation and differentiation
[35]. Together with HDAC5, HDAC4 connects neural activity to skeletal muscle transcription upon denervation, via both epigenetic regulation of gene expression
[36][37][38] and by modulating nuclear and cytoplasmic non-histone protein acetylation
[39][40], thereby mediating neurogenic muscle atrophy. Interestingly, deletion of
Hdac4 in skeletal muscle is protective in experimental models of neurogenic muscle atrophy in the early phases after the surgical procedure
[37]; however, it resulted detrimental effects following long-term denervation, causing muscle degeneration due to the impairment in the activation of multiple signaling, including the oxidative stress response, the ubiquitin-proteasome system and the autophagic pathway
[32]. Consistently, deletion of
Hdac4 in skeletal muscle in a mouse model of ALS worsened pathological features, advancing and exacerbating skeletal muscle atrophy and denervation by modulating several biological processes and gene networks
[29]. Similar to ALS, HDAC4 expression is upregulated in DMD skeletal muscles
[27], and, consistently, deletion of
Hdac4 results detrimental in both disease states. Indeed,
mdx mice with a skeletal muscle-specific deletion of HDAC4 show increased muscle damage and hampered muscle regeneration, overall leading to decreased muscle function. HDAC4 prevalently localizes in the cytoplasm of dystrophic muscles, where it mediates activation of the membrane repair mechanism, likely through a deacetylase-independent activity, thereby affecting muscle necrosis, satellite cell survival and myogenic capacity
[27]. Overall, these studies suggest that skeletal muscle up-regulates HDAC4 expression upon stress as a response to a disease state. In the heart, the N-terminal proteolytically derived fragment of HDAC4 finely regulates lipid metabolism and glucose handling through MEF2-dependent gene expression, ultimately protecting from heart failure
[41][42].
2. Histone Deacetylases in Muscular Dystrophies
Several HDAC isoforms have been implicated in skeletal muscle remodeling, both in physiological and pathological conditions
[62][63]. Ample work revealed that HDACs exert pivotal roles in regulating fiber type specification
[63], muscle fiber size and innervation
[29][37][64], metabolic fuel switching
[16][65][66], muscle development
[67], insulin sensitivity and exercise capacity
[68][69][70], thus contributing to the maintenance of skeletal muscle homeostasis. The evidence of a wide variety of HDAC functions in skeletal muscle led to an increasing interest to clarify their roles in skeletal muscle disorders
[29][63], including muscular dystrophies (MDs).
MDs consist of a heterogeneous group of genetic disorders characterized by progressive weakness and degeneration of skeletal muscles resulting in impaired muscle function
[71]. Traditionally classified by a patient’s clinical presentation, muscle group involvement, mode of inheritance, age of onset and overall disease progression, MDs have been linked to a variety of distinct single-gene mutations
[72]. So far, molecular genetic mapping techniques have shown that MDs are caused by numerous mutations in several genes encoding structural and functional muscle proteins, resulting in degeneration or dysfunction of skeletal muscle
[72].
The most severe and the most common adult form of MD is Duchenne Muscular Dystrophy (DMD), which affects 1 in 3500–6000 live male births, and is caused by the lack of functional dystrophin protein due to mutations in the dystrophin gene (
DMD)
[73]. The structural role of dystrophin is closely related to its centrality in assembling the sarcolemmal Dystrophin-Associated Protein Complex (DAPC), which provides the molecular link between the cytoskeleton and the extracellular matrix of skeletal myofibers
[74][75]. Lack of dystrophin results in mechanical instability causing myofibers rupture during contraction. Moreover, being connected with multiple proteins, dystrophin modulates several signal transduction pathways, including Ca
2+ entry, nitric oxide (NO), and reactive oxygen species (ROS) production
[76][77][78]. The
mdx mouse, harboring a nonsense point mutation in the exon 23 that aborts the full-length dystrophin expression, is the most widely used animal model for DMD research
[79]. Despite the loss of dystrophin,
mdx mice show minimal clinical features of the disease, if compared with DMD patients, probably due to compensatory mechanisms. The latter include muscle regeneration, which is more efficient in
mdx mice compared to DMD patients, in part due to differences in telomere shortening and muscle stem cell regenerative capacity
[80]. Among compensatory mechanisms triggered by the absence of dystrophin, the upregulation of
utrophin has been reported in both DMD and
mdx myofibers
[81]. Utrophin is a structural and functional autosomal paralogue of dystrophin, normally located at the neuromuscular and myotendinous junctions in adult skeletal muscle in physiological condition
[82], but enriched at the sarcolemma in dystrophic myofibers where it acts to preserving muscle function and mitigating necrosis
[81]. Importantly, while the exogenous expression of
utrophin attenuated the
mdx dystrophic phenotype, its deletion in
mdx mice worsened the pathology, thus confirming that utrophin protective functions in DMD
[83][84][85].
In addition to dystrophin, another important member of the DAPC is the sarcoglycan complex, which is composed of four sarcoglycan (SG) proteins, α−, β−, δ−, and γ-SG, playing a key role to protect striated muscle membranes against contraction-induced damage
[86][87]. Mutations in one of the four sarcoglycan genes (
SGCA) cause a different form of autosomal recessive sarcoglycanopathies
[88][89], a subgroup of Limb Girdle Muscular Dystrophies (LGMDs). Sarcoglycanopathies are more frequently found among the most severe forms of MDs, and the clinical phenotype closely resembles that of DMD, with onset during childhood
[90][91].
The role of HDACs in MDs is not yet fully identified; indeed, most of our knowledge derives from studies with HDAC inhibitors in dystrophic contexts (discussed below). However, several studies revealed the deregulation of HDAC expression or activity in dystrophic muscles (Figure 1).
Figure 1. Histone deacetylase functions in muscle dystrophy condition. The cellular responses that promote MD progression are indicated in red, while in green those that counteract MD pathological features. AChR: acetylcholine receptor; Ac: acetyl group; Fst: follistatin; Mstn: myostatin; Utrn: utrophin; HDAC: histone deacetylase; SIRT: sirtuin; DYSF: Dysferlin; PGC-1α: peroxisome proliferator-activated receptor, gamma, coactivator 1 alpha; TGF-β: transforming growth factor beta.
Higher global deacetylase activity was first detected in muscles of
mdx mice and in DMD patients
[92][93], accompanied by selectively elevated levels of HDAC2 in MuSCs
[92]. Further investigations revealed a molecular link among the DAPC, NO signaling and HDAC2
[92][93]. Indeed, in
mdx mice, the loss of an essential component of the dystrophin–glycoprotein complex leads to the displacement of the muscle-specific variant of the neuronal nitric oxide synthase (nNOSm) enzyme, which is normally located at the sarcolemma in close contact with the DAPC complex, resulting in reduced generation of NO. In addition to impairing many processes, including mitochondrial biogenesis and glucose metabolism, reduced NO bioavailability alters S-nitrosylation of HDAC2, resulting in increased activity and constitutive inhibition of HDAC2-target genes in dystrophic muscle
[92][94][95]. HDAC2 directly inhibits
follistatin gene transcription in
mdx muscle cells, which in turn blocks a powerful inhibitor of muscle growth, i.e.,
myostatin [96]. Consistently, the follistatin-myostatin axis has been identified as a target to ameliorate MDs; indeed, myostatin blockade at early stages of the disease provides a beneficial effect in both
mdx and α-SG–deficient mice
[97][98]. Moreover, HDAC2 modulates a specific subset of miRNAs, including miR-1 and miR-29, while HDAC1 specifically inhibits miR-206 in dystrophic MuSCs, thereby correlating with several pathogenetic traits of DMD
[95]. HDAC2 downregulation by siRNA or NO-donor led to improved myogenesis of
mdx MuSCs in vitro, in addition to ameliorating functional and morphological parameters in vivo
[92].
Among the members of class I HDACs, HDAC3 has been shown to be directly involved in the pathogenesis of the X-linked Emery–Dreifuss muscular dystrophy (EDMD1)
[99][100]. This disease is caused by mutations in the
emerin gene, which encodes for a nuclear membrane protein that binds to and recruits HDAC3 to the nuclear lamina. The loss of emerin in muscle cells leads to aberrant nuclear envelope architecture and heterochromatin organization, which results in a more open conformation because of the delocalization and loss of activity of HDAC3. As a result, skeletal MuSCs are unable to differentiate, resulting in progressive skeletal muscle wasting and impaired skeletal muscle regeneration
[101]. Moreover, muscles from EDMD1 patients and
emerin-null mice show an increased and improper temporal expression of marker genes involved in muscle regeneration, including Pax7, MyoD, and Myf5
[102]. Importantly, activation of HDAC3 catalytic activity by theophylline treatment rescues myogenic differentiation in
emerin-null mice, confirming HDAC3 as a master regulator in coordinating the spatiotemporal localization of gene loci to the nuclear envelope required for proper differentiation and muscle regeneration
[103].
In a recent study, HDAC8 was found to be overexpressed in DMD human primary myoblasts and myotubes, and in a zebrafish DMD model
[104]. In the same study, the authors clarified the role of HDAC8 in modulating cytoskeletal architecture and stability through the deacetylation of α-tubulin. Moreover, selective inhibition of HDAC8, by PCI-34051 administration, rescues the DMD phenotype in terms of increased human myoblast differentiation and reduced lesion extent in zebrafish embryos, overall restoring skeletal muscle histomorphology and reducing inflammation
[104].
Differently from class I HDACs, which predominantly localize to the nucleus, where they mostly act as epigenetic regulators, class IIa HDACs shuttle between the nucleus and the cytoplasm, regulating numerous stress responses. HDAC4 has been shown to be crucial for proper MuSCs proliferation and differentiation
[35] and muscle regeneration
[34] following acute muscle injury. A recent paper revealed enhanced expression of HDAC4 in
mdx and DMD muscles, characterized by a higher cytoplasmic abundance of HDAC4
[27], thus suggesting a potential role for HDAC4 in this pathology.
Mdx mice carrying a skeletal muscle-specific deletion of HDAC4 developed a more severe MD pathology, with increased muscle damage and reduced muscle regeneration, overall showing decreased muscle performance. The protective role of HDAC4 in the cytoplasm of dystrophic muscles is independent of its deacetylase activity and depends on its involvement in the membrane repair process. Indeed, cytosolic HDAC4 mediates the activation of a compensatory mechanism of membrane repair in
mdx muscles, thus promoting MuSCs survival and differentiation, ultimately improving muscle regeneration and function
[27].
HDAC5 is downregulated in the nucleus of
mdx muscle and MuSCs, compared with normal controls, and has been implicated in the epigenetic control of chromatin landscape during
mdx MuSCs differentiation. Impaired NO-dependent protein phosphatase 2A activity induces a hyperphosphorylation of HDAC5, thus reducing the amount of nuclear HDAC5 in complex with HDAC3, and affecting
mdx MuSCs differentiation
[93].
Regarding class IIb HDACs, two independent groups identified interesting functions for HDAC6 in DMD
[105][106]. HDAC6 exclusively localized in the cytoplasm, where it removes acetyl groups from non-histone proteins such as α-tubulin, modulating microtubule network stability and organization
[49]. HDAC6 also possesses a non-enzymatic zinc-finger ubiquitin-binding domain at its C-terminus, through which HDAC6 interacts with components of the ubiquitin proteasome pathway, thus playing a critical role in the cellular response to misfolded and aggregated proteins
[107]. Moreover, HDAC6 and its endogenous inhibitor paxillin, regulate acetyl choline receptors (AChR) clustering at the neuromuscular junctions, by mediating a fine balance of nonacetylated and acetylated microtubule network
[50]. Increased HDAC6 protein expression has been reported in
mdx muscles, with a concomitant reduction of acetylated α-tubulin, which contributes to the disorganization of microtubule network and to the impairment of the autophagic flux in DMD. The pharmacological inhibition of HDAC6, by tubastatin A administration, restores the microtubule acetylation and rescues the autophagic flux enhancing autophagosome-lysosome fusion in
mdx mice, in addition to improve AChR clustering and distribution
[105][106]. Moreover, HDAC6 inhibition downregulates transforming growth factor beta (TGF-β) signaling, through an increase of SMAD2/3 acetylation, thereby reducing muscle atrophy and fibrosis and improving protein synthesis in
mdx muscles
[105].
Members of class III HDACs rely on NAD
+ to deacetylate their targets, thereby mediating several important functions in skeletal muscle physiology and diseases
[62][108]. Although SIRT1 expression does not change between
mdx and control muscles, the lack of dystrophin abrogates proper diurnal oscillation of SIRT1 mRNA expression
[109]. Moreover, an increased level of phosphorylated SIRT1 (p-SIRT1) was observed in
mdx muscles, with a concomitant increase of histone H3 acetylation at Lys9/Lys14, thus suggesting attenuated SIRT1 activity
[110]. In addition, NAD
+ concentration was found to be reduced in dystrophic muscles, supporting a model in which SIRT1 activity is downregulated in
mdx mice
[111][112]. Functional proof that SIRT1 downregulation contributes to MD pathogenesis comes from gain-of-function studies.
Mdx mice overexpressing SIRT1 in skeletal muscle developed a less severe DMD pathology, with decreased myofiber necrosis, oxidative stress and fibrosis, accompanied by a fast-to-slow myofiber shift, and overall improvement of muscle performance
[111]. Most of the improvements reported in
mdx SIRT1 overexpressing mice have been proven to be mediated by the activation of peroxisome proliferator-activated receptor, gamma, coactivator 1 alpha (PGC-1α), a SIRT1 target known to protect and ameliorate dystrophic muscles
[113][114]. Indeed, increased expression of PGC-1α in dystrophic muscle mimics, in part,
mdx SIRT1 transgenic mice, enhancing mitochondrial biogenesis, improving the oxidative metabolism and driving a fast-to-slow fiber switch, and preventing muscle degeneration
[115]. Skeletal muscle-specific
Sirt1 knockout mice display a mild dystrophic phenotype, being more prone to suffer from exercise-induced muscle injury, probably due to defects in membrane resealing
[116]. However,
Sirt1 loss in skeletal muscle of
mdx mice does not exacerbate the dystrophic phenotype
[116], suggesting redundant protective mechanisms in skeletal muscle under stress conditions.
SIRT2 modulates autophagy signaling, thereby affecting skeletal muscle atrophy and myoblast proliferation
[117][118]. A recent role for SIRT2 in skeletal muscle following injury has been demonstrated.
Sirt2 KO mice showed a delay in muscle regeneration due to a decreased expression of anabolic and cell cycle regulators genes, with a concomitant increase in catabolic genes and muscle atrophy
[119]. Interestingly, a significant upregulation of SIRT2 mRNA has been reported in MuSCs derived from DMD patients
[120]. These recent results illustrate that further research is needed to better understand the role of SIRT2 in MDs, since SIRT2 could be a promising new therapeutic target in those muscular pathologies where regeneration is inefficient. Moreover, SIRT2 has been proposed as new serum dystrophic marker, since it is upregulated in
mdx serum while it is reversed to control levels by overexpressing utrophin in
mdx mice
[121].
SIRT3, SIRT4, and SIRT5 are exclusively localized to mitochondria and regulate a wide range of metabolism-oriented enzymes in skeletal muscle, thereby modulating energy metabolism in response to mitochondrial stress. Mitochondrial dysfunction is a pathological feature of several MDs
[122][123], suggesting a possible involvement of these sirtuins in such pathologies. SIRT3, SIRT4 and SIRT5 mRNA expression have been found to be upregulated in MuSCs derived from DMD patients and
mdx mice
[120], although no further investigations elucidating their potential role in MDs have been performed.
SIRT6 plays a pivotal role in heterochromatin stabilization through deacetylation of H3K9ac, H3K18ac and H3K56ac. In skeletal muscle, SIRT6 has been reported to negatively regulate
myostatin expression via suppressing NF-
κB signaling, in addition to modulating glucose homeostasis and insulin sensitivity
[124][125]. SIRT6 expression has been found to be upregulated in skeletal muscle and in MuSCs of
mdx mice, where it mostly acts on H3K56ac, thereby repressing several genes, including
utrophin and
myostatin [120]. Lack of SIRT6 reduces muscle fragility and damaged myofibers, increasing the physical activity of
mdx mice. Interestingly,
Sirt6-depleted MuSCs showed attenuated activation, characterized by a strong reduction of Pax7/MyoD double-positive cells, reduced proliferation rate, and decreased expression of stress response-related genes
[120]. Overall, these results indicate that reducing the persistent and chronic activation of MuSCs in
mdx muscles is protective, and that inactivating SIRT6 in DMD ameliorates the dystrophic phenotype in mice.
The class IV HDAC11, which is a lysine de-fatty acylase
[126][127][128], is highly expressed in skeletal muscle but it is dispensable for adult muscle growth. Interestingly, its genetic deletion accelerates regeneration in response to muscle injury
[129][130]. The recent study on HDAC11-deficient mice show a more efficient muscle regeneration following acute injury
[129] likely due in part to an increase in IL-10, which allows a faster transition from inflammatory to pro-regeneration environment. Since high levels of IL-10 have been demonstrated to ameliorate the pathology of
mdx mice
[131][132], these new results on HDAC11 functions are promising and open new avenues for the development of more specific HDAC inhibitors, such as specific HDAC11 inhibitors, as an effective approach to treat MDs. Further studies are needed to evaluate whether this HDAC is involved in the persistent and inefficient regeneration in MDs, and to verify whether HDAC11 is a candidate target to improve muscle repair in this pathological condition.
3. Targeting Histone Deacetylases in Muscular Dystrophies
Epigenetic mechanisms controlling transcriptional programs in tissue progenitors are becoming a critical area of interest in medicine. Indeed, current studies are focused on manipulating chromatin targets of individual signaling pathways to provide novel regenerative strategies based on epigenetic drug administration.
Numerous studies have highlighted the fundamental role of HATs and HDACs in regulating muscle gene transcription and therefore, muscle development and differentiation. Moreover, cumulative in vitro and in vivo evidence in the last years has underscored the link between HDAC deregulation and the pathogenesis of several MDs, in particular of the most severe one, the DMD
[133][134][135]. In this context, HDACi have been shown to act in a selective way, potentiating myogenesis through the hyperacetylation of genes regulated during development and resolving their epigenetic bivalency, a characteristic signature that identifies genes poised for transcription that typically are enriched in embryonic stem cells or pluripotent cells
[136]. Starting from this evidence, by inhibiting HDACs and reestablishing the epigenetic events necessary to activate adult stem cells, it represents one of the most powerful approaches to restoring the downstream networks of muscle regeneration and muscle homeostasis, leading to increased functional and morphological recovery of dystrophic muscles.
At first, focusing on the HDACi activity on skeletal muscle cells in vitro, it was observed that the pharmacological treatment targets myogenic differentiation
[64][137]. Indeed, treatment of wild-type myoblasts with pan-HDACi, such as Trichostatin A (TSA), Valproic acid (VPA), or Sodium Butyrate (PhB), increases their differentiation potential and fusion capacity, due to different mechanisms: (i) the upregulation of MyoD acetylation; (ii) the modulation of histone acetylation at specific gene promoters and (iii) the increase of the expression of the pro-myogenic protein follistatin
[64][137].
Several years ago, a link between dystrophin loss and HDAC activity was demonstrated
[92][93]. In
mdx whole muscles and primary myoblasts, an increase in global HDAC activity and HDAC2 expression was observed in association with a reduction in follistatin expression. Inhibition of HDAC2, by using the class I HDAC inhibitor MS-275 or siRNA, restores the level of global HDAC activity similar to healthy control muscles, leading to morphological and functional benefits in dystrophic muscles
[92]. In more recent studies, increased activity of class I, class IIa and class I/IIb HDACs in muscles of 1.5-month-old
mdx mice
[27] and in Fibro-Adipogenic Progenitors (FAPs) isolated from 1.5 month- and 12 month-old
mdx mice has been reported
[138], further suggesting the involvement of HDACs in the pathogenesis of DMD.
Next-generation sequencing studies have focused on the fine regulation of myogenesis by HDACi, paying attention to the epigenetic players that create changes in the epigenome, opening new therapeutic options in muscle diseases. It emerged that most of the beneficial effects of the HDACi on dystrophic muscles arise from their ability to selectively activate a microRNA-SWI/SNF-based epigenetic network in FAPs, a specific population of mesenchymal cells resident in muscle interstitium
[139][140]. FAPs are a muscle cell population that, while in regenerating conditions support MuSCs differentiation, in pathological conditions, such as DMD, contribute to the progression of the disease, affecting fibrotic and fat deposition, decreasing muscle contractility, and altering metabolism
[141][142][143]. Intriguingly, pan-HDACi manipulate cell fate determination that redirects the lineage commitment of FAPs from a fibro-adipogenic toward a myogenic one
[140].
In the context of MDs, it is worth mentioning the sirtuins, which are class III histone/protein deacetylases, are able to modulate several important physiological mechanisms such as inflammation, apoptosis, glucose homeostasis, life span, and neuroprotection. Acting pharmacologically on these enzymes permits modification of the acetylation state of several intracellular messengers, thereby regulating downstream mechanisms. This approach likely has strong therapeutic potential for many human diseases such as metabolic disorders, and degenerative diseases such as MDs. As described above, the most studied of the sirtuins is SIRT1, which is expressed in many tissues, including skeletal muscle and heart, where it deacetylates and activates PGC-1α, a key modulator of muscle metabolism. The activated form of PGC-1α controls mitochondrial biogenesis and homeostasis, and therefore SIRT1 modulation was seen to be associated with muscle pathologies. It is now well known that PGC-1α overexpression in dystrophic
mdx mice leads to milder signs of pathology and an improved function both in normal condition and after intense physical exercise
[61][113]. Other mechanistic roles are attributed to SIRT1 modulation, supporting the beneficial effects on muscle pathologies. It has been described for example that SIRT1 stimulates and restores autophagy in muscle tissue through the deacetylation of autophagy components, including Atg5, Atg7, and Atg8, and activating FoxO3a a transcription factor that regulates autophagy in skeletal muscle
[144][145]. Moreover, SIRT1 may modulate the activity of SMAD transcription factors, key TGF-β signaling components that are involved in myofibroblast differentiation. The activity of SMAD is regulated by lysine acetylation/deacetylation, which plays a critical role in tissue fibrosis
[146].
All these data generated in vitro on cells (Figure 2), together with the in vivo evidence of deregulated activity of HDACs in MDs, have provided the rationale for using pan-HDACi and modulators of sirtuins in preclinical studies, with the aim of assessing the ability of these classes of compounds to improve muscle regeneration and counteract muscle degeneration in models of MD.
Figure 2. In vitro evidence of inhibiting HDACs on myoblast or FAP lineage progression. TSA: Trichostatin A; VPA: Valproic acid; PhB: Sodium Butyrate; Fst: follistatin; SCs: satellite cells; FAPs: Fibroadipogenic progenitors.