The nSMase2/Smpd3 gene modulates the severity of muscular dystrophy and the emotional stress response in mdx mice

Background Duchenne muscular dystrophy (DMD) is a progressive, degenerative muscular disorder and cognitive dysfunction caused by mutations in the dystrophin gene. It is characterized by excess inflammatory responses in the muscle and repeated degeneration and regeneration cycles. Neutral sphingomyelinase 2/sphingomyelin phosphodiesterase 3 (nSMase2/Smpd3) hydrolyzes sphingomyelin in lipid rafts. This protein thus modulates inflammatory responses, cell survival or apoptosis pathways, and the secretion of extracellular vesicles in a Ca2+-dependent manner. However, its roles in dystrophic pathology have not yet been clarified. Methods To investigate the effects of the loss of nSMase2/Smpd3 on dystrophic muscles and its role in the abnormal behavior observed in DMD patients, we generated mdx mice lacking the nSMase2/Smpd3 gene (mdx:Smpd3 double knockout [DKO] mice). Results Young mdx:Smpd3 DKO mice exhibited reduced muscular degeneration and decreased inflammation responses, but later on they showed exacerbated muscular necrosis. In addition, the abnormal stress response displayed by mdx mice was improved in the mdx:Smpd3 DKO mice, with the recovery of brain-derived neurotrophic factor (Bdnf) expression in the hippocampus. Conclusions nSMase2/Smpd3-modulated lipid raft integrity is a potential therapeutic target for DMD.

muscle regeneration [6,7]. This suggests that myofiber instability is not the only cause of dystrophic degeneration, but rather that the phenotype might be caused by multiple factors, including stem cell and myofiber functions.
In addition to the function of dystrophin in the structural integrity of myofibers described above, a novel function of asymmetric cell division in satellite cells (SC) has been revealed, in which SCs lacking dystrophin show a marked increase in abnormal nonpolarized mitotic divisions and reduced asymmetric cell divisions and myogenic progenitors [7][8][9]. Thus, the continuing cycles of degeneration and regeneration in the initial stages of the dystrophic pathology exacerbate the phenotype. This exacerbation is thought to be caused by the misregulation of SC fate between differentiation and self-renewing proliferation during the regeneration of degenerated muscles [10,11]. However, the mechanisms by which SC dysfunction is involved in muscular dystrophy have not yet been elucidated. In addition to muscle degeneration, the loss of dystrophin in the brain has often been associated with nonprogressive cognitive deficits, behavioral disabilities, and enhanced fearfulness [12][13][14]. However, an effective treatment for these abnormalities has not yet been established.
In this study, we test the hypothesis that the pathogenesis in the dystrophic muscles and brains of mdx mice is affected by the nSMase2/Smpd3 protein through an inflammation response, as well as regeneration, differentiation, and signaling pathways. Deletion of the nSMase2/Smpd3 gene from mdx mice resulted in decreased inflammation and increased muscle regeneration in the skeletal muscle (SM) in early stages of the dystrophic process, but caused adverse effects in later stages. Furthermore, loss of the nSMase2/ Smpd3 gene in mdx mice suppressed abnormal emotional behavior, such as the stress-induced anxiety response, as well as the recovery of hippocampal Bdnf expression. Thus, these findings suggest that the nSMase2/Smpd3 protein is a potential therapeutic target for muscular dystrophy and abnormal behavior.

Cell culture
The murine skeletal myoblast C 2 C 12 cell line were maintained in a proliferation medium that Dulbecco's modified Eagle's medium (DMEM; Sigma-Aldrich, St. Louis, MO, USA) supplemented with 10% (v/v) fetal bovine serum (Cell Culture Technologies, Lugano, Switzerland) and 1% (v/v) penicillin/streptomycin (Wako Pure Chemical Industries, Osaka, Japan) at 37°C in controlled humidified air with 5% CO 2 in six-well plates (BM Equipment, Tokyo, Japan). Before reaching confluency, the cells were rinsed with PBS, and the proliferation medium was changed to a differentiation medium (DMEM containing 2% heat-inactivated horse serum (HS) and 1% penicillin/streptomycin) until myotube formation was completed.
Dystrophin-deficient H2K myoblasts, which were derived from H-2 kb-tsA58 transgenic mice [21], were seeded at a density of 5 × 10 4 cells/well in a 75 cm 2 flask and grown at 33°C in DMEM with GlutaMAX, IFNgamma at a concentration of 20 U/mL, and 20% (v/v) fetal bovine serum [22]. After treatment, the cells were differentiated into myotubes by incubating them in DMEM with GlutaMAX containing 5% (v/v) HS at 37°C. At this plating density, the cells must be passaged once or twice a week, when they reach approximately 10 4 per cm 2 , and the medium must be changed twice weekly to fresh medium and fresh IFN-gamma. Four days after the initiation of differentiation, which is induced by a high cell density (≥ 10 4 per cm 2 ), the myotubes were used for experiments [23].

Extraction and quantification of transcripts
As described previously [23], for total RNA extraction, differentiated C 2 C 12 myotube cells at the indicated times and mouse tissue samples were homogenized. DNA-free RNA was obtained by using the Purelink total RNA extraction kit (Ambion, Austin, TX, USA) according to the manufacturer's instructions. Specific complementary DNA (cDNA) was synthesized from the purified total RNA with random primers, using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA, USA). Real-time PCR was performed by the StepOne Real-Time Polymerase Chain Reaction (PCR) System (ABI, Foster City, CA, USA) with genespecific primers, according to the manufacturer's instructions (Additional file 1: Table S1). Relative expression of each genes was calculated using SDS 2.1 realtime PCR data analysis software (ABI) and 2 −ΔΔCt method. Beta-actin and glyceraldehyde-3-phosphate dehydrogenase (gapdh) were used as reference genes for normalization. The gene expression data from triplicate data per one sample are presented as median ± standard error of the mean (SEM).

Measurement of SMase enzymatic activity
SMase enzymatic activity was measured using the Sphingomyelinase Activity Colorimetric Assay Kit according to the manufacturer's protocol (BioVision). Briefly, GAS muscles were removed from wt, mdx, and mdx: Smpd3 −/− (#238 and #11) mice, and then SMase Assay Buffer and SMase Extraction Detergent with Protease Inhibitor Cocktail (Promega, Madison, WI, USA) were added. To extract the tissue lysate, GAS muscles were homogenized on ice, and supernatants were collected after centrifugation at 10,000×g for 5 min. Next, 5 μL of supernatant was added to a 96-well plate with 45 μL of SMase Assay Buffer. The reaction mix, containing 32 μL of SMase Assay Buffer, 2 μL of SMase Enzyme Mix I, 10 μL of SMase Enzyme Mix II, 4 μL of SMase Substrate, and 2 μL of SMase Probe, was added to each well containing the sample solution. After mixing well, the reaction solution was incubated for 30 min at 37°C and the absorbance at 570 nm was measured.

Generation of nSMase2/Smpd3 knockdown mice
Under specific pathogen-free (SPF) condition, all mice lines used in this study were maintained at 21°C on a 12:12 h light to dark cycle at the National Institute of Neuroscience (NCNP), Japan, and treated in accordance with the guidelines provided by the Ethics Committee for the Treatment of Laboratory Animals of the National Center of Neurology and Psychiatry (approval ID: 2015006), which has adopted the three fundamental principles of replacement, reduction, and refinement. Cages were always enriched with crinkle paper and animals were allowed to drink water and eat food ad libitum. Consistent with the protocol stipulated by the research permit, all efforts were made to minimize the suffering and discomfort experienced by the animals. To design sgRNAs against the nSMase2/Smpd3 gene, the Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) Design Tool (http://crispr.mit.edu) was used to search for candidate nucleotide sequences, and we selected three candidate target sequences based on the prediction scores of the CRISPR Design Tool (Additional file 2: Table S2). The plasmids expressing CRIS PR-associated protein 9 (Cas9) and the sgRNAs were constructed by ligating three oligonucleotide pairs against each sgRNA with the BbsI-digested pX330 plasmid (Addgene, Watertown, MA, USA) according to the manufacturer's instructions [24]. In brief, to prepare for the ligation of the sgRNAs into the BbsI site of the pX330 plasmid, three oligonucleotide pairs were added using four tagged nucleotide sequences, i.e., 5′-cacc-3′ at the 5′-end of the forward oligonucleotide and 5′caaa-3′ at the 3′-end of the reverse oligonucleotide, together with 0.1 μM of forward and reverse oligonucleotides in tris-ethylenediaminetetraacetic acid (TE) buffer, and the oligonucleotides were annealed using a thermal cycler, as follows: one cycle of 95°C for 5 min, 60°C for 5 min, and 25°C for 60 min. The ligation solution was prepared as follows: annealed oligonucleotides, BbsIdigested pX330, and Ligation high buffer, and this was incubated for 1 h at 16°C. The plasmid DNA-ligated sgRNA was transformed into a competent Escherichia coli strain (Competent Quick DH5alpha, TOYOBO, Tokyo, Japan), and incubated on Luria-Bertani (LB) containing 100 μg/mL ampicillin at 37°C for 14 h.
Individual colonies on the plate were picked up and incubated in LB liquid culture containing 100 μg/mL ampicillin with shaking overnight at 37°C for 16 h. The plasmid DNA was purified from the liquid cultures using a spin column from a commercially available kit (Pure-LinkR HiPure Plasmid Maxiprep Kit, Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions, and integration of the sgRNA into the BbsI site of the pX330 vector was confirmed by direct sequencing. SgRNA-pX330 plasmid DNA in their circular form were injected directly into the pronuclei of zygotes collected from the oviducts of wt female B6C3F1 mice mated to wt male B6C3F1 mice, to reduce their integration into the host mouse genome, although the transgenic efficiency with the circular form of the plasmid DNA is approximately 10 times lower than that of the linear form [24,25].
To screen for target mutant mice, genomic DNA was extracted from tissue samples taken from the tails of pups that developed from the microinjected eggs, by incubating them in lysis buffer at 50°C for 17 h followed by phenol/chloroform purification. The extracted DNA was then genotyped by PCR amplification, the amplified PCR fragments were purified using DNA purification spin columns, and nucleotides were determined by sequencing with the indicated primer set (Additional file 3: Table S3). To determine the integration of Cas9 nucleotides into the genome DNA of the mice targeted by the sgRNAs, PCR amplification was performed using a pair of primers for the Cas9 nucleotide sequence. To check for off-target effects, a homology search for the three sgRNAs with a mouse genome sequence was performed using BLAST (https://blast.ncbi.nlm.nih.gov/Blast.cgi). The resulting matching genomic sequences were amplified and their nucleotide sequences were determined via PCR amplification and sequenced using region-specific primer sets. To generate nSMase2/Smpd3-null, dystrophin-deficient DKO mice, we crossed nSMase2/ Smpd3 KO mice with mdx mice. The genetic background of the mdx:nSMase/Smpd3 DKO mice used in this study is a mix that was created by mating mice from the B6C3F1 and B6j backgrounds and used after 4-5 generations of inbreeding.

Tissue preparation
Mice were sacrificed by cervical dislocation. Body and wet muscle weight were measured. The TA, GAS, and diaphragm muscles were collected using standard dissection methods, as described previously [26]. Some muscle samples were frozen in isopentane cooled by liquid nitrogen for histological analysis, and the remaining muscle samples were frozen in liquid nitrogen for RNA or protein isolation, and stored at 80°C as described previously [26,27]. Transverse cryosections (20 μm thick) of each muscle were stained with H&E, as described previously [28].

Patients
A total of 119 unrelated Japanese patients with DMD, BMD, myotonic dystrophy 1, distal myopathy with rimmed vacuoles (DMRV), facioscapulohumeral muscular dystrophy (FSHD), limb-girdle muscular dystrophy (LGMD), or limb-girdle muscular dystrophy 2B (LGMD2B) were participated in this study (Additional file 4: Table S4). Informed consent form with the patient's signature was obtained from all participants after the details of the study had been explained to them and prior to the collection of peripheral blood. The protocol was approved by Research Ethics Committee of the NCNP (approval ID: A2011-113), in accordance with the regulations of the Declaration of Helsinki. All results were treated with standard medical confidentiality, and confidentially was maintained to the extent stipulated by the law.

Measurement of CK activity
From the blood samples, the supernatant was removed by microcentrifugation after incubation at room temperature. Serum CK assays were carried out using a commercially available Fuji Dri-Chem system (Fujifilm Medical, Tokyo, Japan) according to the manufacturer's protocol, as described previously [23]. Serum (10 μL) was incubated at 37°C on a Fuji Dri-Chem slide, and the dye absorbance was measured spectrophotometrically for 5 min at 540 nm. The values were calculated according to the installed formula, and data are expressed as units per liter (U/L).

Muscle Evans blue dye uptake experiments
To assess muscle damage, EBD (Nacalai Tesque, Tokyo, Japan) was dissolved in phosphate buffered salts (PBS) and sterilized by filtration. Twenty-four hours before sacrifice, mice were injected intraperitoneally with 1% EBD as described previously [29]. The muscle tissues were collected, and then frozen in melting isopentane. The sectioned muscle was created, and incubated in acetone (ice-cold) for 10 min. After wash the sections three times for 10 min with PBS, the flat-mount muscle was created by Vectashield mounting medium. Fluorescence microscopy was used for evaluation of the presence of EBD in myofibers.

Treadmill and grip strength tests
The treadmill muscle performance test was performed as described previously [30].
Briefly, mice were placed on a motor-driven flat MK-680S treadmill system (Muromachi Kikai, Tokyo, Japan) and forced to run for 5 min at a speed of 5 m/min. After 5 min, the speed was accelerated by 1 m/min every min. The test was stopped when the mouse was exhausted and did not attempt to remount the treadmill, and the time to exhaustion was recorded.
The forelimb grip strength of the mice was monitored using a grip strength meter (MK-380 M, Muromachi Kikai) with the investigator blinded to genotype. The mice were held 2 cm from the base of the tail, allowed to grab a woven metal wire with their forelimbs, and were pulled gently in the horizontal plane until they released their grip. The force at the time of release was recorded as peak tension. Five sequential tests were carried out for each mouse, at 5 s intervals. The average peak tension in these attempts was defined as forelimb grip strength.

Extraction and quantification of miRNA
Total RNA isolation from serum or tissues was performed according to the manufacturer instructions for the PureLinkR RNA Mini Kit (Ambion, Austin, TX, USA) as previously described [31,32]. The cDNA derived from the total RNA was prepared using a TaqMan miRNA Reverse Transcription (RT) kit (ABI, Foster City, CA, USA) and miRNA-specific stem-loop primers (part of the TaqMan miRNA assay kit; ABI) as previously described [23,33]. For the real-time PCR of the miRNA, we used individual miRNA-specific primers (part of the TaqMan miRNA assay kit; ABI) with the StepOne Real-Time PCR System (ABI) according to the manufacturer's protocol. Each miRNA was assayed in triplicate and data are presented as median values with the standard deviation. The relative expression levels for each miRNA were normalized using by endogenous and exogenous controls that are miR-16 and cel-miR-39. SDS 2.1 realtime PCR data analysis software (ABI) was used for the data analysis.

Cell viability analysis
The C 2 C 12 or H2K cell line was seeded in 96-well plates (BM Equipment, Tokyo, Japan). Ten microliters of 2-(2methoxy-4-nitrophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H-tetrazolium salt from the Cell Counting Kit-8 (CCK-8) (DOJINDO Laboratories, Kumamoto, Japan) was added to each well to detect cell proliferation and cell toxicity based on the quantity of formazan dye generated by the dehydrogenases in the cells, as described previously [23]. The cells continued to be cultured for 2 h, and then the absorbance at 450 nm of each well was measured using a Model 680-well microplate reader (Bio-Rad Laboratory).

Proteome profiler cytokine array
The skeletal and diaphragm muscles of mdx:Smpd3 −/− (#238) and mdx mice were excised and homogenized in PBS with protease. After homogenization, Triton X-100 was added to a final concentration of 1%. To remove cellular debris, the tissue lysates were centrifuged at 10, 000×g for 5 min. The protein concentration of the lysates was determined, and the relative expression of the cytokines and chemokines in the lysates was quantified using the Proteosome Profiler Array (Mouse Cytokine Array, Panel A; R&D Systems, Minneapolis, MN, USA) according to the manufacturer's protocol. Briefly, 2.0 mL of Array Buffer was added into each well of the 4-well multi-dish as a blocking buffer. Each membrane was incubated for 1 h in a well of a 4-well multi-dish plate on a platform shaker. 1.5 μL of reconstituted Mouse Cytokine Array Panel A Detection Antibody Cocktail was added to each prepared sample, and the samples were then incubated at room temperature for 1 h. After blocking, the Array Buffer was aspirated from the wells. The sample/antibody mixtures were placed on the 4-well multi-dish and incubated overnight at 4°C on a rocking platform shaker. Each membrane was washed thrice with 1× Wash Buffer for 10 min on a rocking platform shaker. The membrane was incubated with diluted Streptavidin-HPR for 30 min at room temperature on a rocking platform shaker. Then, after the membrane had been washed, signal detection was performed using the Chemi Reagent Mix and analyzed on the LAS-3000 Imager (Fujifilm Corporation).

Measurement of cathepsin B activity using an in vivo imaging system
ProSense 680 is a small peptide substrate for activated cathepsin B (CTSB). When cleaved by the CTSB enzyme, two caged fluorophores are released, with peak excitation at 680 nm and emission in the near-infrared range (700 nm). ProSense 680 (0.75 or 1.5 nmol) was intraperitoneally injected into 5-week-old wt, mdx, mdx: Smpd3 −/− , and Smpd3 −/− mice, which were then placed on a heating pad under anesthesia to keep their body temperature constant. Imaging was performed using a single excitation/emission filter pair optimal for the wavelength of the probe.

Apoptosis assay
Caspase-3 and caspase-9 activities were determined using Caspase-3/CPP32 or Caspase-9 colorimetric Assay Kit (BioVision), respectively, according to the manufacturer's protocol. Briefly, GAS and TA muscles and cerebellum of mdx:Smpd3 −/− , mdx, and wt mice were homogenized, and then lysis buffer (50 mM Tris [pH 7.5], 150 mM NaCl, 0.5% Nonident P-40, and protease inhibitor cocktail) was added. After incubation on ice for 10 min, the tissue lysates were centrifuged at 10,000×g for 5 min for removal of cellular debris. The DEVD-pNA substrate was added into the sample solution adjusted the protein concentration and incubated at 37°C for 2 h. The absorbance of each well at 405 nm was determined using a Model 680-well microplate reader (Bio-Rad Laboratory).

Restraint test in mice
Mice were restrained by the experimenter by placing the neck between the thumb and index finger and putting the tail between the third and little fingers [12]. After 10 s, the mouse was released into a cage (24 cm × 17 cm, surrounded by a 12 cm-high wall) containing wood chips (illuminated at 80 lx). A camera on the ceiling of the cage recorded a video and saved it to a personal computer. Locomotion and freezing were calculated from the image files obtained during the 5 min after the restraint using Image OF (O'Hara & Co., Tokyo, Japan), which is a modified version of the public-domain NIH Image program (developed at the US National Institutes of Health and available from http://rsb.info.nih.gov/nihimage/). Complete immobilization of the mouse, except for respiration, was regarded as a freezing response.

Sample size and statistical power
To determine the effective sample size needed to ensure the required statistical power in tests involving mdx and mdx:Smpd3 +/− or mdx:Smpd3 −/− mice and DMD and BMD patients, we calculated sample sizes based on the means of populations 1 and 2 and the common standard deviation, using the Power/Sample Size Calculator (available from https://www.stat.ubc.ca/~rollin/stats/ssize/n2. html) (Additional files 5, 6: Table S5, S6).

Results
Expression of nSMase/Smpd-family genes during C 2 C 12 differentiation and generation of nSMase2/Smpd3 knockdown mice in the mdx genetic background To investigate the expression patterns of nSMase2/ Smpd3 genes in response to the induction of myogenic differentiation in C 2 C 12 mouse myoblast cells, we quantified the mRNA expression levels of acid sphingomyelinase (aSMase)/Smpd1, nSMase/Smpd2, and nSMase2/ Smpd3 during the course of C 2 C 12 differentiation. All three genes were upregulated as early as 4 days after induction with differentiation medium, when the cells became phase-bright fused myotubes, and the upregulated levels were sustained for 6 days (Additional file 7: Fig.  S1A). The expression level of the adult myosin heavy chain (Myh1) gene was used as a differentiation marker for C 2 C 12 myoblasts, similar to the expression pattern of nSMase2/Smpd3 (Additional file 7: Fig. S1A). Western blotting showed that the nSMase2/Smpd3 protein was upregulated until day six (Additional file 7: Fig. S1B). On the other hand, the Bdnf protein was upregulated at day one, peaked on days two and three, then decreased until day six (Additional file 7: Fig. S1B). These findings indicate that nSMase2/Smpd3 is essential for myoblast differentiation.
To define the role of nSMase2/Smpd3 in the dystrophic process, we generated nSMase2/Smpd3 DKO mice using the CRISPR-Cas9 system with three single guide RNAs (sgRNAs) located in a large insertion region (LIR) between catalytic domain 1 (CD1) and CD2 of the nSMase2/Smpd3 protein (Additional file 7: Fig. S1C: lower panel). To reduce the possibility of off-target effects, we searched the mouse genome for regions homologous to the three sgRNA target nucleotide sequences using a BLAST search. We examined a total of 82 nucleotide sequences from homologous genomic regions for off-target effects using PCR-amplification and direct sequencing of the genomic DNA of each nSMase2/Smpd3 DKO mouse line generated (data not shown). In addition, the genomic DNA was analyzed to ensure that the Cas9 transgene (Tg). After the mouse lines with off-target effects or the Cas9 Tg had been eliminated, six independent nSMase2/Smpd3 DKO mouse lines were established, of which one (#20) had a premature stop codon at the amino acid (AA) at position 224 of the nSMase2/Smpd3 protein, with the last seven AAs included as substitutions (Additional file 7: Fig To generate nSMase2/Smpd3-null, dystrophin-deficient double-mutant (DM) mice, we crossed nSMase2/Smpd3 DKO mice with mdx mice (Additional file 8: Fig. S2A and B). These founders were not born in a Mendelian ratio (Additional file 9: Table S7A and B), but there were no differences in organ weights between them and mdx mice (data not shown). Immunoblot analysis indicated that the nSMase2/Smpd3 protein was expressed in the gastrocnemius (GAS) muscle of both mdx and mdx:Smpd3 −/− DKO (#30) mice at 18 weeks of age (Additional file 10: Fig. S3A), in the cerebellum of wild-type (wt), mdx, and mdx: Smpd3 −/− DKO (#238) mice at 14 weeks of age (Additional file 10: Fig. S3B), and in the hippocampus of both mdx and mdx:Smpd3 −/− DKO (#238) mice at 5 and 9 weeks of age (Additional file 10: Fig. S3C). In addition, the hippocampi of two of the mdx:Smpd3 −/− DKO mouse lines, namely #11 (at 12 weeks) and #30 (at 18 weeks), expressed nSMase2/Smpd3 proteins, but in mouse line #28 at both 6 and 13 months, and in line #20 at 19 weeks, the specific band for nSMase2/Smpd3 could not be detected (Additional file 10: Fig. S3C). In addition, analysis of lysates of the SM and hippocampi of wt, mdx, and mdx:Smpd3 −/− DKO (#238 and #11) mice showed that there was no difference in nSMase enzymatic activity between the #238 line and mdx mice in either the SM or the hippocampus (Additional file 10: Fig. S3C). However, the nSMase activity in the #11 line was significantly lower in the SM (Additional file 10: Fig.  S3D) and somewhat lower in the hippocampus (Additional file 10: Fig. S3E) than in the #238 line and mdx mice. These results indicated that the LIR, in part, regulates nSMase/Smpd3 activity in vivo.

Deletion of the nSMase2/Smpd3 gene in mdx mice reduces inflammation in dystrophic muscles
To characterize the contribution of the nSMase2/Smpd3 gene to muscular inflammation in dystrophic muscles, cross sections of tibialis anterior (TA) muscles from 6week-old mice were stained with hematoxylin and eosin (H&E). The area of inflammatory infiltration was significantly smaller in 6-week-old mdx:Smpd3 +/− (#11) mice than in mdx mice (Fig. 1a, b, p < 0.05). To quantify the inflammation in the limb muscles of live animals, cathepsin-B enzyme activity was measured as an indicator of inflammation. The hindlimb muscles of mdx mice exhibited higher levels of cathepsin-B activity than those of wt mice and mdx:Smpd3 −/− mice, and activity levels in the muscles of Smpd3 −/− mice were significantly lower than in wt mice (Fig. 1c, d, p < 0.01).
To assess the effects of nSMase2/Smpd3 on the phosphorylation of Ser536 of p65, we performed a western blot analysis on the GAS of mdx:Smpd3 −/− and mdx mice. However, we found no obvious differences in the expression of p65 or phosho-p65 between mdx and mdx: Smpd3 −/− mice (data not shown). At 17 weeks, expression of Foxp3 in the GAS of mdx:Smpd3 −/− mice was significantly lower than in mdx mice (Additional file 13: Fig. S6, p < 0.01), whereas at 19 weeks it was significantly higher (Additional file 13: Fig. S6, p < 0.05). Combined, these data suggest that depletion of nSMase2/Smpd3 may reduce inflammatory cytokine expression in monocytes/macrophages in the dystrophic process.
Next, to visualize and quantify the degree of myofiber damage, we analyzed uptake of Evans blue dye (EBD) in the TA hind limb and diaphragm muscles. At 12 weeks of age, there were significantly fewer EBD-positive muscle fibers in the TA of mdx:Smpd3 +/− mice than in mdx mice (Fig. 3b: upper, p < 0.05), but there was no difference between the diaphragm muscles of mdx: Smpd3 +/− and mdx mice (data not shown). The number of EBD-positive muscle fibers in the TA of line #238 mdx:Smpd3 −/− mice at 20 weeks of age was also substantially lower than in mdx mice, but it was higher in line #11 mdx:Smpd3 −/− mice (Fig. 3b: bottom). These results suggest that early in life, nSMase2/Smpd3 ablation may have beneficial effects with respect to myofiber membrane degeneration in mdx mice, but that later on it may have adverse effects.
Genetic ablation of nSMase2/Smpd3 improves muscle performance in mice with dystrophic phenotypes To further assess whether the reduction in muscle membrane permeability caused by nSMase2/Smpd3 KO in mdx mice improved muscle performance, we assessed muscle force and endurance. The grip strength test demonstrated that at 12 weeks of age, the muscle strength of mdx mice was significantly lower than that of wt mice (p < 0.001), but that of mdx:Smpd3 +/− mice was significantly higher than that of mdx mice (Fig. 4a, p < 0.05). However, the mdx:Smpd3 +/− mice also weighed significantly more than the mdx mice (Fig. 4b, p < 0.01), and when grip strength was scaled to body weight, there was no difference between mdx:Smpd3 +/− and mdx mice (data not shown). An inclined treadmill running test indicated that mdx:Smpd3 +/− mice tended to run further than mdx mice, but the difference was not statistically significant (Fig. 4c). However, mdx:Smpd3 −/− mice ran for significantly longer than mdx mice at 16 (Fig. 4d, p < 0.001) and 60 (Additional file 15: Fig. S8, p < 0.05) weeks of age. These data show that loss of the nSMase2/Smpd3 gene may improve dystrophic muscle function in mdx mice.
nSMase2/Smpd3 deletion modulates fiber size in dystrophic muscles of mdx mice We analyzed muscle fiber diameter and central nucleation in wt, mdx, and mdx:Smpd3 +/− (#238 and #11) mice at 6 and 20 weeks of age. The muscle fibers were narrower in both lines of mdx:Smpd3 +/− mice than in mdx mice at 6 weeks (Fig. 5a), but were thicker at 20 weeks (Fig. 5b). At 6 weeks, there were significantly fewer centrally nucleated fibers (CNFs) containing two nuclei per fiber in both lines of mdx:Smpd3 −/− mice than in mdx mice (Fig. 5c, p < 0.05), but significantly more without central nuclei (Fig. 5c, p < 0.05). However, at 20 weeks there were no significant differences in the numbers of CNFs (Fig. 5d). These findings suggest that the loss of the nSMase2/Smpd3 gene may affect fiber size via fiber fusion.
Effects of the nSMase2/Smpd3 gene on the survival, proliferation, and differentiation of myogenic cells in mdx mice Next, we hypothesized that nSMase2/Smpd3 regulates muscle regeneration via the survival and apoptosis of dystrophic myofibers. To test this hypothesis, we measured caspase-3 and caspase-9 activity in the SM and cerebellum of wt, mdx, and mdx:Smpd3 −/− (#238) mice. At 12 weeks, caspase-3 activity in the GAS and TA muscles of mdx:Smpd3 −/− mice was significantly lower than in mdx mice (Additional file 16: Fig. S9A, p < 0.05), but there were no differences between them for the GAS muscle at 9 or 18 weeks or for the cerebellum at 18 weeks (Additional file 16: Fig. S9A). There were also no differences between these lines in terms of caspase-9 activity in the GAS at either 9 or 12 weeks of age (Additional file 16: Fig. S9A). The expression levels of Bdnf protein in the GAS muscles of mdx:Smpd3 −/− mice at 5 (ns), 9 (p < 0.0.5), 12 (ns), and 14 (p < 0.05) weeks of age were lower than in mdx mice (Additional file 16: Fig.  S9B-E), but there was no difference at 18 weeks (Additional file 16: Fig. S9F). We also performed a western blot analysis using an anti-caveolin-3 antibody as a lipid raft marker protein. At 14 weeks, caveolin-3 expression in the GAS muscle of mdx mice was significantly higher than in both wt and mdx:Smpd3 −/− mice (Additional file 16: Fig. S9E, p < 0.05).
The expression of Myf5 and CD34, the earliest markers of myogenic commitment, was slightly higher in mdx:Smpd3 −/− (#238 and #11) mice at 14 weeks of age than in mdx mice (Additional file 16: Fig. S9H). To test the specification of muscle fiber types, the expression of myosin heavy chain (MyHC) isoforms (types I, IIa, IIb, and IIx) were quantified using real-time PCR. At 12 weeks, expression of slow/oxidative type I and fast/oxidative type IIa fibers were unaffected by the loss of nSMase2/Smpd3 in the mdx genetic background, but that of fast/glycolytic type IIb and fast/intermediate type IIx was higher in mdx:Smpd3 −/− mice than in mdx mice (Additional file 16: Fig. S9H). In 9-week-old mdx: Smpd3 −/− mice, levels of Pgc-1alpha and troponin I (Tnnil1) were significantly lower than in mdx mice (Additional file 16: Fig. S9H, p < 0.05).
Combined, these findings suggest that nSMase2/ Smpd3 knockdown might induce the regeneration of dystrophic muscle early in life, but that the opposite effect might be induced later on.

Effects of expression of muscle-specific miRNAs on muscle health
It was recently reported that treatment of 10 DMD patients with an antisense oligonucleotide (NS-065/NCNP-01) that induces skipping of exon 53 in mutated dystrophin transcripts increased the dystrophin/spectrin ratio [34]. Using the sera of 10 DMD patients who had been administered with intravenous NS-065/NCNP-01 for 12 weeks, we quantified their expression of a set of muscle-abundant miRNAs (myomiRs) before and after this treatment. The median expression levels of miR-1, miR-133a, and miR-206 in all 10 patients after treatment were approximately 25%, 67%, and 20% lower than before treatment, respectively, but these results were not significant; only that of miR-133a approached significance (p = 0.08; Additional file 17: Fig. S10A-C).
We then divided the 10 patients into three groups according to the dose of NS-065/NCNP-01 they received (1.25, 5, or 20 mg/kg weekly). We compared the foldchange in the miRNA levels before ("pre") and after ("post") treatment and calculated these as log 2 [post/pre]. At a weekly dose of 5 mg/kg, the levels of all three miR-NAs after treatment were lower than before treatment, and levels of miR-133a were reduced after treatment in all three groups (Additional file 17: Fig. S10D-F).
It has been shown that recovery of the dystrophin protein via the skipping of exon 23 in the mature dystrophin transcript normalizes serum myomiR levels [35,36]. Given that NS-06/NCNP-01 induces the skipping of exon 53 in the dystrophin transcript and increases dystrophin protein expression, our results suggest that the restoration of extracellular myomiR levels is associated with the pathology of dystrophic muscles. To assess the association of serum extracellular vesicle (EV) content with muscular damage, we analyzed serum EV content in patients with seven different types of muscular disorder, and found that it was substantially lower in DMD patients than in Becker muscular dystrophy (BMD) patients (Additional file 17: Fig. S10G).
Next, to analyze whether loss of the nSMase2/Smpd3 gene in the mdx background affects muscular dystrophy via muscle-abundant miRNAs, we quantified the levels of the myomiRs miR-1, miR-133a, and miR-206 in the GAS muscles of mdx and mdx:Smpd3 −/− (#11 and #238) mice. The expression levels of miR-1 and miR-133a in mdx mice were lower than in wt mice, but that of miR-206 was higher (Additional file 18: Fig. S11). In contrast, levels of miR-1 and miR-133a were substantially higher in mdx:Smpd3 −/− (#238) mice than in mdx mice, while those of miR-206 were lower. mdx:Smpd3 −/− (#11) mice, on the other hand, exhibited no substantial differences in this respect from mdx mice (Additional file 18: Fig. S11).
We then investigated whether the benefits of the miRNA regulated by the nSMase2/Smpd3 gene could lead to a recovery in the survival of myogenic cells. To address this issue, we incubated H2K cells in serumdepleted medium with or without the nSMase inhibitor GW4869 (0, 5, 10, 15, 20, 25, or 30 μM). Survival rates of myoblasts incubated for 24 h with 10-30 μM of GW4869 were significantly higher than in cells incubated without GW4869 (Additional file 19: Fig. S12A).
Next, to analyze the effects of myomiRs on cell survival, we assessed the survival of H2K myotubes that had or had not been transfected with myomiRs. The survival of H2K cells transfected with miR-1 was significantly poorer than that of cells without it, whereas significantly more of the H2K cells cultured with miR-206 survived than those without it (Additional file 19: Fig. S12B). We then assessed the effect of the nSMase2/Smpd3 protein on H2K cell survival under the overexpression of myomiRs by culturing H2K cells that had or had not been transfected with miR-1, miR-133a, or miR-206 with GW4869. The survival of those transfected with miR-133a was better than that of non-transfected cells (Additional file 19: Fig. S12C). In addition, the number of EVs in the supernatant of H2K cells was significantly higher than that in the supernatant of C 2 C 12 cells (Additional file 19: Fig. S12D). The protein content of serum-derived EVs from mdx:Smpd3 −/− mice was significantly lower than that of mdx mice (Additional file 19: Fig. S12E and F).

Loss of the nSMase2/Smpd3 gene in mdx mice modulates anxiety behavior
To investigate the effects of the nSMase2/Smpd3 gene on emotional responses, mdx and mdx:Smpd3 −/− mice at 11-12 weeks of age were restrained for 10 s as an emotionally aversive stimulus. The mice were then released into a cage, and their behavior was monitored for 5 min for freezing vs locomotion. All the mdx mice froze for at least half of the 5-min period, whereas half of the mdx: Smpd3 −/− (#238) mice froze for much less than half of the time (Fig. 6a). Thus, the defense response was partially rescued in the mdx:Smpd3 −/− (#238) mice. We also investigated anxiety, emotionality, and the adaptive stress response to a novel environment using the holeboard test. The mdx mice performed significantly fewer head dips and exhibited significantly greater head-dip latency than both wt and mdx:Smpd3 −/− (#238) mice (Fig. 6b, c). The mdx mice also traveled a substantially shorter total distance than both of the other lines, although this difference was only significant in the comparison with the wt mice (Fig. 6d).
Next, to assess the role of the nSMase2/Smpd3 gene in spontaneous behavior and reactivity in a novel environment, the open-field test was performed on 10-week-old wt, mdx, mdx:Smpd3 −/− (#238), and Smpd3 −/− (#238) mice. Although the mdx and Smpd3 −/− (#238) mice moved significantly more slowly and covered less distance than the wt mice, there was no significant difference between the mdx and mdx:Smpd3 −/− (#238) mice (Fig. 6e, f). However, in the first bin, the mdx:Smpd3 −/− (#238) mice covered substantially less distance than the mdx mice (Fig. 6g), and overall, they and the Smpd3 −/− mice covered significantly less distance and moved for significantly less time than the mdx mice (Fig. 6h, i). However, all three of these lines covered significantly shorter distances and spent less time in motion than wt mice (Fig. 6h, i). There were no significant differences in movement time, total time spent in the center of the field, or average speed among any of the four lines ( Fig. 6j-l). Combined, these findings suggest that ablation of the nSMase2/Smpd3 gene in mdx mice modulates their anxiety behavior and stress response.
Roles of Bdnf, miRNA, and anxiety-associated genes in the abnormal behavior of mdx mice To analyze whether nSMase2/Smpd3 plays a role in the regulation of Bdnf expression in the hippocampus, we performed western blotting using an anti-Bdnf antibody on samples from mdx and mdx:Smpd3 −/− mice at various ages. There were no significant differences in Bdnf protein levels between mdx and mdx:Smpd3 −/− mice aged 5, 12, or 19 weeks (Additional file 20: Fig. S13A, C, and D). However, at 9 weeks, the Bdnf protein level in mdx mice was significantly lower than that of both wt and mdx:Smpd3 −/− mice (Additional file 20: Fig. S13B).
Next, to assess the effects of the nSMase2/Smpd3 protein on the transcription levels of the Bdnf isoforms in the hippocampus of mdx mice, we performed real-time qPCR using nine exon-specific primers in wt, mdx, and mdx:Smpd3 −/− mice aged 12 weeks (Additional file 20: Fig. S13E). In mdx mice, expression levels of five isoforms of Bdnf, exons I, II, V, VI, and VIII, were lower than in wt mice (Additional file 20: Fig. S13F), whereas in mdx:Smpd3 −/− mice, expression levels of all isoforms excluding exon IX were higher than in mdx mice (Additional file 20: Fig. S13G).
To assess myomiR-mediated regulation of Bdnf expression levels, we analyzed the transcriptional levels of myomiRs in the hippocampus. Although there were no significant differences in precursor-microRNA-1 levels between mdx and mdx:Smpd3 −/− mice at 5-19 weeks of age (Additional file 20: Fig. S13H), primary-microRNA-1 levels in mdx:Smpd3 −/− (#238 and #11) mice at 12 weeks were significantly lower than in mdx mice (Additional file 20: Fig. S13I). Although there were no significant differences in precursor-microRNA-133a levels between mdx and mdx:Smpd3 −/− mice at 12 weeks of age (Additional file 20: Fig. S13J), the precursor-microRNA-206 levels of mdx:Smpd3 −/− (#11) mice at this age were significantly higher than those of mdx mice (Additional file 20: Fig. S13K). To investigate the expression levels of genes induced in the hippocampus, we performed an expression analysis of 12-week-old wt, mdx, mdx:Smpd3 +/− , and mdx:Smpd3 −/− (#238 and #11) mice using real-time qRT-PCR. The expression levels of dopamine receptor d1 (Drd1), cholecystokinin (Cck), the ionotropic glutamate receptor NMDA2B (NR2b), and postsynaptic density protein 95 (PSD-95) genes in mdx: Smpd3 +/− mice were significantly lower than in mdx mice (Additional file 20: Fig. S13L, p < 0.05). In addition, the expression levels of Erg1 and Arc genes in mdx: Smpd3 −/− (#238) mice were slightly higher than those in mdx mice (Additional file 20: Fig. S13M). Expression levels of the dopamine receptor d2 (Drd2) and GIT Arf-GAP 1 (Git1) genes were higher in mdx:Smpd3 −/− (lines #238 and #11, respectively) mice than in mdx mice Fig. 6 Loss of the nSMase2/Smpd3 gene in mdx mice modulates anxiety-associated behavior in the hole-board and open-field tests. wt, mdx, and mdx:Smpd3 −/− , and Smpd3 −/− mice at 11 weeks of age were restrained for 10 s and released into a measuring field. Freezing times were measured for 5 min. a The total time spent freezing during the 5 min period is indicated as a percentage. Head-dip count (b), head-dip latency (c), and the total distance traveled (d) in the hole-board test were analyzed in 12-week-old wt, mdx, mdx:Smpd3 −/− , and Smpd3 −/− mice. Total distance (e), speed (f), movement distances in each of six time bins (g), distance per movement (h), time per movement (i), total movement duration (j), total time in the center (k), and average speed (l) were measured in the open-field test for wt, mdx, mdx:Smpd3 −/− , and Smpd3 −/− mice at 10 weeks of age. The number of animals used is indicated in parentheses. *p < 0.05, **p < 0.01, ***p < 0.001 (Additional file 20: Fig. S13M). These findings suggest that the loss of the nSMase2/Smpd3 gene may regulate the anxiety phenotype of mdx mice via the regulation of Bdnf expression.

Discussion
In this study, to investigate the role of the nSMase2/ Smpd3 gene in the dystrophic phenotypes of mdx mice, mutant mice with a deletion in the LIR of the nSMase2/ Smpd3 protein were crossed with mdx mice to produce nSMase2/Smpd3 dystrophin DKO mice. The DKO mice exhibited a reduced inflammation responses and less muscular degeneration in the early stages of the dystrophic process, but had exacerbated muscular necrosis in the later stages. We also found that the Bdnf pathway modulated the anxiety and stress responses of these DKO mice.
The nSMase/Smpd3 family of enzymes comprises four members: nSMase1, nSMase2, nSMase3, and mitochondriaassociated nSMase. With the exception of nSMase3, these enzymes possess a DNase I-type catalytic core, suggesting a common mechanism for sphingomyelin catalysis [37]. Further, it has previously been shown that the mRNA expression level of nSMase3 in brain is higher than that of nSMase1 or nSMase2 [38]. We observed a relatively large degree of variation in SMase activity in the hippocampi of line #11 mdx:Smpd3 −/− mice. This may have been due to the functional redundancy of other SMases.
We found less inflammation in the SM of mdx: Smpd3 −/− mice than in mdx mice. The acute phase of the pathology of mdx muscles (prior to four weeks of age) involves muscle inflammation with a bias toward M1 macrophages, which contributes to oxidative stress and muscle fiber lysis via the production of iNOSderived NO, and promotes inflammation and myoblast proliferation via the production of Th1 cytokines [28,39]. Later, at the age of 3 months, the arginaseexpressing M2a macrophages that compete with iNOS for arginine can begin to reduce M1 macrophage cytotoxicity, and the muscles enter the regenerative phase.
The SM of dystrophin-deficient (DMD mdx ) rats has been shown to be infiltrated by leukocytes, whose kinetics during the pathological course were parallel to those of serum CK levels in the SM, and the numbers of CD45 + mononuclear leukocytes at 4-16 weeks of age were significantly higher than in wt rats [27]. Of these increased muscle CD45 + mononuclear cells in Dmd mdx rats, approximately 90% were CD68 + macrophages. In addition, the mononuclear cells from these rats expressed higher levels of transcripts for the cytokine TNF-alpha, which is associated with early muscle damage, than the same cells from wt rats. The inflammatory cytokines TNF-alpha, IL-6, and IL-1beta lead to increased endothelium permeability and promote early recruitment of innate immune cells, such as neutrophils and monocytes, which differentiate locally into inflammatory macrophages within injured tissues. Thus, leukocyte infiltration in the muscles of DMD model animals is associated with damaged muscle fibers and elevated serum CK levels.
In our study, expression levels of CD68, CD45, IL-6, and TNF-alpha in the SM of 12-week-old mdx:Smpd3 DM mice were significantly lower than in mdx mice. This was also true for serum CK levels in 6-18-week-old mdx:Smpd3 DM mice. These findings may suggest lower numbers of infiltrating mononuclear cells in mdx:Smpd3 DM mice. In addition, the tissue recruitment of monocytes is mainly mediated by the chemokine Ccl2 and the chemokine receptor CCR2 or CCR5 via the blood circulation [40]. A deficiency in CCR2 in the SM of mdx: CCR2 KO mice markedly and persistently reduced the infiltration of Ly6C high inflammatory monocytes, which enter tissues in response to injury and differentiate into inflammatory M1 macrophages within inflamed tissues, relative to mdx controls, leading to a reduction in muscle fiber necrosis and endomysial fibrosis at 14 weeks of age [41]. On the other hand, in mdx:CCR2 KO mice the abundance of intramuscular Ly6C low macrophages, which patrol the vascular space with homing properties under steady-state conditions and comprise over 90% of resident macrophages, was significantly lower than in mdx controls during the early stages (4-9 weeks of age), but from 14 weeks to 6 months, it was similar to the mdx mice. The improvement in necrosis and fibrosis in the limb muscles of the mdx:CCR2 KO mice was not sustained during the latter phase, possibly because of unsuppressed levels of profibrotic growth factors such as osteopontin.
It is likely that a mixture of monocyte-derived and tissue-resident macrophages accumulate in chronically inflamed tissues. However, the origins of the Ly6C high and Ly6C low monocytes/macrophages in dystrophic muscles are unclear. Although bone marrow has been thought to be the principal source of monocytes, it was recently reported that splenic Ly6C high monocytes, which are outsourced from the bone marrow, contribute to recruitment and infiltration in dystrophic limb muscles, and to muscle fiber necrosis, during the early stages of the disease [42]. A reduction in infiltrated CD45 + cells in mdx mice improved muscle fiber necrosis and increased eMHC-positive regenerating fibers under a lack of splenic monocytes induced by splenectomy during the early phases of the condition. However, during the late stages, dystrophic muscle regeneration is impeded by reduced angiogenesis and increased fibrosis. This causes a delay in the phenotypic shift from proinflammatory to proregenerative macrophages, which affects the tissue cytokine environment. Thus, the optimal response to chronic tissue injury in dystrophic muscle depends on the fine balance between the phenotypes of macrophage types, and the cytokine environment may be critical for the progression of the pathology.
It has also been shown that nSMase2/Smpd3 deficiency or inhibition strongly suppresses M1 macrophage infiltration and differentiation, and inhibits inflammation, in a mouse model of atherosclerosis via Nrf2 (NF-E2-related factor 2)/HO-1 pathway activation. In this pathway, the rapid nuclear translocation and accumulation of Nrf2 protein was promoted to inhibit early cytokine response, such as IL-1beta and IL-6. This may occur via persistent Akt phosphorylation through the reduction of the ceramide-induced inhibition of PP2A (protein phosphatase 2) activity and the suppression of the expression of inflammatory and adhesion genes such as Ccl2 and ICAM-1 and/or ceramide [20,43]. Given these reports, the beneficial effects in the early stages and adverse effects in later stages in our mdx:Smpd3 DKO mice may have been caused partially by the inhibition of monocyte/macrophage recruitment and inflammatory responses. Furthermore, older mdx:Smpd3 DKO mice might display more severe dystrophic phenotypes, such as necrosis and fibrosis, which could more clearly be distinguished from the effects seen in younger mdx: Smpd3 DKO mice, such as the increased serum CK levels in 28-week-old mdx:Smpd3 DKO (#238 and #11) mice and the increased number of EBD-positive fibers in 20-week-old mdx:Smpd3 DKO (#11) mice.
Splenectomized mdx and mdx:CCR2 DM mice do not show improvements in muscle strength at 14 weeks or 6 months [41,42]. On the other hand, in our study, the running performance of the mdx:Smpd3 DM mice at 16 and even 60 weeks in the treadmill exhaustion test was better than that of the mdx mice. Combined, this evidence suggests that suppression of the inflammation of the monocytes/macrophages may not contribute to the improvement in muscle performance in the nSMase2/ Smpd3-deficient mdx mice.
In general, degenerative/necrotic lesions in the dystrophic muscles of mdx mice exist in small clusters, possibly due to an imbalanced tissue environment caused by the repeated degeneration-regeneration cycles, during which inflammatory cells such as monocytes and macrophages are recruited and infiltrate muscle cells. The pathologic changes in mdx:Smpd3 mice during the early stages were spatiotemporally restricted to specific areas, suggesting that the main cause of the pathologic changes in the SM of mdx:Smpd3 mice may be the inhibition of monocyte/macrophage recruitment into the dystrophic lesions. However, during the later stages, the dystrophic lesions in the SM of mdx:Smpd3 mice were sparse but worsened despite the improvements in muscle performance. This indicates that the changes in the dystrophic pathology within their SM, such as susceptibility to exercise-induced injury, oxidative stress, and impaired regeneration capacity, may have been caused by multifactorial pathways. Possible candidates are decreased cytosolic and mitochondrial calcium concentrations, calpain inactivation [17], and inhibition of phosphorylation of the Stat1/Stat3 transcriptional factor, in addition to the sustained inhibition of Lyc6C high monocyte/macrophage recruitment and suppressed antiinflammatory differentiation. Additionally, the dual CCR2/CCR5 chemokine receptor antagonist has been shown to reduce macrophage infiltration and decrease the prevalence of regenerated CNFs, which are thought to be a marker of previous necrosis-regeneration events, in the diaphragm of mdx mice [44]. Thus, the significant reduction in CNFs containing multiple nuclei in the mdx:Smpd3 DM mice may also have been caused by the inhibition of monocyte/macrophage infiltration.
The inhibition of TRPV2 (transient receptor potential cation channel), which is a principal Ca 2+ -entry route, leads to a sustained Ca 2+ increase and muscle degeneration in two DMD mouse models [45]. Dystrophic pathologies, such as increased abundance of CNFs, variability in fiber size, increased Ca 2+ levels in muscle fibers, elevated serum CK levels, and reduced muscle performance, are all ameliorated by the inhibition of TRPV2 in the early stages (4-10 weeks), when the degeneration-regeneration cycles are ongoing. However, in old mice (> 26 weeks), when the potential for such cycles may have been exhausted, the improvements in abundance of CNFs and fiber size variability are slight [45]. Also, since necrosis could play a critical role in mediating the myocyte and myofiber loss associated with calcium dysregulation, the inhibition of cyclophilin D which directly regulates the changes in mitochondrial permeability that depend on calcium and reactive oxygen species causes noticeable improvements in muscular dystrophies, such as reductions in CNFs, fibrosis, and myofiber necrosis [46].
Dystrophic phenotypes such as the increased abundance of CNFs, fibrosis, calpain activation, and serum CK levels that are seen in mdx mice at 6 weeks to 3 months have been shown to be improved by SM-specific overexpression of sarcoplasmic reticulum Ca 2+ ATPase 1 (SERCA1). This enzyme reverses a defect in sarcoplasmic reticulum Ca 2+ reuptake that causes dystrophic fibers and reduced total cytosolic Ca 2+ [31]. In addition, reducing sarcolipin, which is an inhibitor of SERCA and is abnormally elevated in both the slow-and the fasttwitch SM of DMD patients and animal models, mitigates dystrophic phenotypes and improves muscle regeneration via the restoration of SERCA function [47]. Increased expression of intramuscular heat shock protein 72 (Hsp72) also ameliorates the dystrophic pathway and preserves muscle strength, via an interaction with SERCA that enhances its function [48]. On the other hand, nSMase2/Smpd3 deficiency suppresses the ceramide-dependent activation of protein phosphatase 2a (PP2a) that maintains Akt phosphorylation, thereby inducing hyaluronan synthase 2 (HAS2) and Hsp72 expression [49,50]. The amelioration of dystrophic phenotypes in young mdx:Smd3 DM mice may therefore contribute to protection from mitochondrial Ca 2+ -overload-induced myofiber necrosis.
It has also been shown that C5a, sICAM-I, IL-1ra, IL-16, Ccl2, TIMP-1, and TNF-alpha levels are higher in mdx mice than in wt mice [32]. The levels of all seven of these cytokines and chemokines in the GAS muscle of the mdx:Smpd3 −/− mice were lower than in the mdx mice, as were the expression levels of TNF-alpha and IL-6. A C5a inhibitor rescued the decreased force and increased the abundance of necrotic fibers in the mdx mice via a change to fast-twitch fibers, and also increased the maturation of macrophages [32]. A coordinated balance between pro-inflammatory and antiinflammatory macrophages is important for successful muscle repair. However, in dystrophic muscle, this balance of inflammatory responses might be disrupted. Bdnf is produced by immune cells such as CD4 + and CD8 + T lymphocytes and monocytes/macrophages. It is modulated by TNF-alpha and IL-6, which are located near regenerating fibers that are positive for p75NTR, and CD56/NCAM owing to the repair of tissue in inflamed muscle [51,52]. The Bdnf-p75NTR axis positively regulates the tissue-protection response [52].
The crosstalk between immune and muscle cells, such as macrophages and SCs, can positively regulate homeostasis, proliferation, and the repair of myogenic cells via chemokines and cytokines that originate from infiltrated monocytes/macrophages [52]. The macrophages play either a supportive or a deleterious role in cells via the amplification or downregulation, respectively, of inflammatory responses that promote the elimination of myogenic debris and prevent excessive tissue damage. The role they play depends on their activation state, which is affected by changes in their environment. However, their role in either the promotion or mitigation of the pathogenesis of dystrophy is unclear. In our study, Bdnf protein expression was higher in the GAS of mdx mice than in both wt and mdx:Smpd3 −/− mice at 9 and 14 weeks of age. Thus, the infiltration of circulating monocytes might be reduced in mdx:Smpd3 −/− mice.
It has been shown that repression of Bdnf synthesis depends on cell differentiation in the SM [53]. Bdnf was highly expressed in muscle Pax7-positive SCs and myoblasts in culture, whereas its expression was absent or very low in myofibers, and was repressed after myogenic differentiation. In addition, the overall levels of Bdnf mRNA are strongly correlated with those of a progenitor marker, Pax3, in mature muscles. By reducing endogenous levels of Bdnf, myoblasts engage in early myogenic differentiation, despite the presence of growth media. This evidence indicates that the primary role of Bdnf in SM is to maintain the population of SCs by preventing their myogenic differentiation.
The expression levels of the MyoG and eMHC/Myh3 genes in the GAS of the mdx:Smpd3 −/− mice were higher in the early stages and lower in the later stages than in the mdx mice. In addition, expression levels of the early myogenic markers Myf5 and CD34 were higher in the GAS of mdx:Smpd3 −/− mice than in mdx mice. However, Pax7 expression in the GAS of mdx:Smpd3 −/− mice was significantly lower than in mdx mice. Also, it has been shown that Pax7 expression is directly targeted by the SM-specific miR-431, which promotes myogenic differentiation via the upregulation of MyoG and mitochondrial transcription factor A (Tfam) [54]. However, the expression level of Myf5 was not affected by miR-431. This suggests that Pax7 expression in mdx:Smpd3 DKO mice may be regulated by miR-431.
These results imply that the severity of the dystrophy in mdx:Smpd3 −/− mice may be ameliorated via modulation of the differentiation balance of SCs in dystrophic muscles. It has been shown that chronic overactivation of Notch signaling occurs in severely dystrophic muscles with impaired muscle regeneration [55]. In this study, mdx:Smpd3 −/− mice exhibited downregulation of the Jag1, Notch1, and Notch3 genes relative to mdx mice. Jag1, induced by IL-1beta, suppresses the muscle regeneration capacity of DMD muscles, probably through Notch3 activation [56]. This evidence suggests that nSMase2/Smpd3 regulates muscle regeneration with SCs via the Notch and Pax7 pathways.
We also analyzed the defensive behavior of the mdx: Smpd3 −/− mice and found that the abnormal behavior exhibited by mdx mice was completely reversed in half of the DKO mice observed. In addition, the mdx: Smpd3 −/− mice performed significantly fewer head-dips (a marker of anxiety behavior in the hole-board test) than the mdx mice. In the CA1 region of the hippocampus of mdx mice, abnormal synaptic plasticity caused by a reduction in gamma aminobutyric acid (GABA) efficacy has been reported [14,57]. nSMase2/Smpd3 inhibition blocks TNF-alpha-induced excitatory postsynaptic currents from CA1 pyramidal cells [29]. In our study, Bdnf protein expression in the hippocampus of mdx mice was significantly lower than in wt mice, but loss of the nSMase2/Smpd3 gene ameliorated this effect. Bdnf secretion promotes GABAergic synaptogenesis [58]. Synaptic GABA A -Rs are sensitive to benzodiazepines, drugs with robust anti-anxiety effects that bind exclusively to GABA A -Rs. They are predominantly located within nonlipid raft fractions and enhance the potentiating effects of benzodiazepines by impairing lipid raft integrity. This suggests that the localization of receptors in lipid rafts affects the potency and efficacy of neurotransmitter signaling and that this plays a role in neurological disorders [59,60]. Thus, the association of lipid rafts with GABA A -R appears to be a downregulatory mechanism for selective synaptic transmission and plasticity. In addition, the disruption of lipid rafts blocks the potentiating inhibitory effects of Bdnf in GABA A -R signaling via Bdnf-induced recruitment of the TrkB receptor into neuronal lipid rafts, creating selective synaptic plasticity [61,62].

Conclusion
In summary, our study shows that ablation of the nSMase2/Smpd3 gene in mdx mice ameliorates membrane instability in the sarcolemma, improves muscle force and performance, and reduces excess inflammation in the early stages. Furthermore, the abnormal stress response of mdx mice was modulated by the loss of the nSMase2/Smpd3 gene, possibly via Bdnf signaling. These findings suggest that signaling pathways modulated by the nSMase2/Smpd3 protein through lipid rafts might be novel therapeutic targets for DMD, via the stagespecific regulation of the expression levels of this protein and transcript, for example through exosomal transfer. genotype). (H) Expression analysis of myogenesis-related genes in the GAS muscle at the indicated ages of mdx, mdx:Smpd3 −/− (#238), and mdx:Smpd3 −/− (#11) mice using qRT-PCR.The number of animals used is indicated in parentheses. * p < 0.05, ** p < 0.01, *** p < 0.001.
Additional file 17 : Fig. S10. miRNA levels in the sera of muscular dystrophy patients and mdx mice as diagnostic markers. miR-1 (A), miR-133a (B), and miR-206 (C) levels in the serum of Duchenne muscular dystrophy (DMD) patients before (pre) and after (post) treatment with the antisense oligonucleotides NS-065/NCNP-01 that induce exon skipping to correct the frame-shift. Post-treatment expression levels expressed as fold-changes relative to pre-treatment expression levels (based on A-C) of miR-1 (D), miR-133a (E), and miR-206 (F) at doses of 1.25 mg/kg (cohort 1), 5 mg/kg (cohort 2), and 20 mg/kg (cohort 3) in patients administered weekly with NS-065/NCNP-01 for 12 weeks. Expression levels were normalized to those of miR-16. The number of patients analyzed is indicated in parentheses. Extracellular vesicles (EVs) were extracted from the sera of patients with seven types of muscle disorder (DMD, Becker muscular dystrophy [BMD], distal myopathy with rimmed vacuoles [DMRV], facioscapulohumeral muscular dystrophy [LGMD], and limb-girdle muscular dystrophy 2B [LGMD2B]) and quantified based on acetylcholinesterase (AChE) activity (G). Levels of miR-1, miR-133a, and miR-206 in the sera (H) and in EVs isolated from the sera (I) of mdx and mdx:Smpd3 +/− mice, with expression levels normalized to spiked-in cel-39. Levels of miR-1 (J), miR-206 (K), miR-133a (L), and miR-31 (M) in the sera of mdx and mdx:Smpd3 −/− mice, normalized to U6. The number of animals used is indicated in parentheses.