ANKK1 is found in myogenic precursors and muscle fibers subtypes with glycolytic metabolism

Ankyrin repeat and kinase domain containing 1 (ANKK1) gene has been widely related to neuropsychiatry disorders. The localization of ANKK1 in neural progenitors and its correlation with the cell cycle has suggested its participation in development. However, ANKK1 functions still need to be identified. Here, we have further characterized the ANKK1 localization in vivo and in vitro, by using immunolabeling, quantitative real-time PCR and Western blot in the myogenic lineage. Histologic investigations in mice and humans revealed that ANKK1 is expressed in precursors of embryonic and adult muscles. In mice embryos, ANKK1 was found in migrating myotubes where it shows a polarized cytoplasmic distribution, while proliferative myoblasts and satellite cells show different isoforms in their nuclei and cytoplasm. In vitro studies of ANKK1 protein isoforms along the myogenic progression showed the decline of nuclear ANKK1-kinase until its total exclusion in myotubes. In adult mice, ANKK1 was expressed exclusively in the Fast-Twitch muscles fibers subtype. The induction of glycolytic metabolism in C2C12 cells with high glucose concentration or treatment with berberine caused a significant increase in the ANKK1 mRNA. Similarly, C2C12 cells under hypoxic conditions caused the increase of nuclear ANKK1. These results altogether show a relationship between ANKK1 gene regulation and the metabolism of muscles during development and in adulthood. Finally, we found ANKK1 expression in regenerative fibers of muscles from dystrophic patients. Future studies in ANKK1 biology and the pathological response of muscles will reveal whether this protein is a novel muscle disease biomarker.

In previous studies we found this protein is localized in radial glial cells [7] and other neurogenic progenitors in the brain where it is differentially expressed during the cell cycle [8].The gene that codes for ANKK1 is located in a 512 Kb cluster along with neural cell adhesion molecule 1 (NCAM1), dopamine receptor D2 (DRD2) and tetratricopeptide repeat domain 12 gene (TTC12) genes [9].The NTAD gene cluster (NCAM1-TTC12-ANKK1-DRD2)shows a conserved synteny and neighborhood thus suggesting the coordinated functioning of these proteins by co-regulation of gene expression [9].
RIP4, the most similar RIP in the overall structural organization to ANKK1, is also involved in developmental processes that regulate epidermal morphogenesis and cutaneous wound repair [10].RIP2 is a checkpoint in skeletal myogenesis [11] and has to be downregulated in mouse myoblasts for correct differentiation [12].In addition, the study of other RIP proteins during myogenesis has shown changes in expression throughout the process of myotube differentiation [12].
Immunolabeling studies of ANKK1 in the central nervous system of mouse samples have shown us that it is also expressed in embryonic myotubes [13].Based on these observations, here we hypothesize that ANKK1 participates in embryonic myogenesis and adult muscle regeneration.Our findings in mice embryos, myogenic cell lines and adult muscles show that ANKK1 is located in specific myogenic precursors and in muscle fiber subtypes with glycolytic metabolism.

Animals and sample preparation
BALB/cAnNHsd mice embryos (Harlan Interfauna Ibe ´rica) or C57BL/6 mice (Jackson Laboratory Repository) at 8 weeks of age were used as tissue source.The experimental procedures were performed following the European Union Council guidelines (2010/63/EU) and Spanish regulations (RD 1201/2005) and were performed according to the Ethical Committee of Centro de Investigación Príncipe Felipe (Valencia, Spain).Mice were maintained at 21 ± 2˚C in 12 h light/dark cycles with food and water ad libitum.For adult muscle studies, C57BL/6 mice were sacrificed by cervical dislocation and gastrocnemius and flexor digitorum brevis (FDB) muscles were removed and prepared as previously described with some modifications [14][15].Gastrocnemius were fixed in 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS) for 24 h at 4˚C and embedded in paraffin blocks.Transversal tissue serial sections (5 μm) were cut on a microtome (MICROM HM340E) and mounted on poly-L-lysine coated glass slides.Single myofibers of FDB muscles were obtained after 2.5 h of enzymatic digestion at 37˚C with 0.2% collagenase type IV (MP Biomedicals) in Dulbecco's Modified Eagle's Medium (DMEM)-high glucose (Gibco) followed by gentle trituration using Pasteur pipettes rinsed with 10% horse serum (HS) (Sigma-Aldrich).To purify the myofibers from free mononucleated cells, debris and damaged myofibers, the suspension was transferred to the top of a tube containing 10 ml of DMEM-high glucose supplemented with 1% penicillin/streptomycin (Sigma-Aldrich) and 10% HS, and myofibers were allowed to settle at 1 x g during 15 min at room temperature [14].After the third purification process, the supernatant was aspirated and myofibers cultured for 24 h in 1 mg/ml matrigel (BD Biosciences) coated dishes with DMEM-high glucose supplemented with 1% penicillin/streptomycin, 1% HS and 20% fetal bovine serum (FBS) (Sigma-Aldrich).

Human samples
Samples of deltoid and quadriceps muscle tissue were obtained in the Neuromuscular Research Unit of the Instituto de Investigación Sanitaria La Fe (Valencia, Spain), collected and processed according their protocols for human studies, and approved by the Ethical Committee of Hospital Universitari i Politècnic La Fe (Valencia, Spain), in accordance with the Declaration of Helsinki (2013).A parent or guardian of the participants provided written informed consent and each subject provided written assent for the collection of samples and subsequent analyses.The muscles samples studied were from two non-dystrophic patients without a history of muscular disease (1 male and 1 female, 1 and 29 years of age respectively) and four dystrophic patients (1 male and 3 females; 8, 26, 30 and 61 years of age respectively).Muscle biopsies were immersed 30 seconds in cold isopentane and stored in liquid nitrogen until use.Transversal tissue serial sections (7 μm) were cut in a microtome, mounted on glass slides and stored at -20˚C.
To induce myoblast differentiation, C2C12 cells were changed to differentiation medium containing DMEM 2% HS [16] when they reached 80% confluence.In the case of RD, cells were changed to differentiation medium plus 2 μM insulin (Sigma-Aldrich) [17] and, after 24 h 5 μM of retinoic acid (Sigma-Aldrich) was added.The cells were harvested at different time points until 6 days of differentiation.
For the induction of glycolytic metabolism in C2C12, the cells were cultured in DMEM with glucose at high concentration (30 mM, HG) or treated with 30 μmol/L berberine (Sigma-Aldrich) as previously described [18][19].Additionally, the induction of glycolytic metabolism was performed by decreasing oxygen levels up to 2% during 24 and 48 h.
When necessary, C2C12 cells were induced to differentiate and treated with 2 ng/ml leptomycin B (LMB, Sigma-Aldrich) for 24 h before fixed or harvested [20].

Lactate assay and oxygen consumption
The induction of glycolytic metabolism was determined by measuring the generation of L-lactic acid in the culture medium following the instructions of the manufacturer of the L-lactic acid test kit (K-LATE, Biocon).The amount of light intensity absorbed was detected using the Victor 1420 Multilabel Hts counter spectrophotometer (Perkin Elmer) at 340 nm.Oxygen consumption was also measured using a Clark type oxygen electrode (Rank Brothers) at 37˚C with a magnetic stirrer as described elsewhere [21].The cells were trypsinized, counted and suspended in 1 ml of DMEM.Oxygen consumption is given in nmol O 2 min −1 10 6 cells −1 .

Western blot analysis
Western blot has been performed following previous protocols [6].Briefly, to obtain membrane and cytoplasmic protein-enriched fractions, cells were homogenized in lysis buffer (150 mM NaCl, 20 mM Tris HCl, 5 mM EDTA, 10% glycerol, 1% NP-40) containing a protease inhibitor cocktail (Complete Mini-Protease Inhibitor Cocktail, Roche) and centrifuged for 100.000x g, 4˚C.The final supernatant was used as the cytoplasmic fraction and the pellet was resuspended in lysis buffer containing 0.5% sodium dodecyl sulphate (SDS) and used as the nuclear fraction.Genomic DNA was sheared by sonication with a Microson Ultrasonic Cell Disruptor (Misonix).Equal amounts of protein samples (40 μg) quantified by Pierce BCA Protein Assay Kit (Thermo Scientific) were resolved on 9% SDS-PAGE and transferred to nitrocellulose membranes (Whatman).After overnight blocking (5% nonfat dry milk and 0.1% Tween 20 in Tris buffered saline (TBS)), the membranes were incubated with the primary antibody for 1 h: α-STk2 (1:2000) and α-STk3 (1:500) to detect ANKK1, α-beta-Actin as a reference to quantify the relative protein expression level (1:5000, A1978, Sigma-Aldrich), α-alpha-Tubulin (1:10000, T9026, Sigma-Aldrich) as a cytosolic marker, α-Lamin as a nuclear marker (1:4000, sc-7292, Santa Cruz Biotechnology) and α-MyoD (1:2000).All washing steps were performed with TBS-0.1% Tween 20.Primary antibodies were detected using goat α-mouse or goat α-rabbit antibodies coupled to peroxidase for 1 h (1:10000, Thermo Scientific).Proteins were processed for chemiluminescence with the Clarity Western ECL Substrate (Bio Rad) and scanned.Bands intensity was measured through the total area of the peaks using the ImageJ 1.45v software from National Institutes of Health.

Image acquisition
Sections from IHC were visualized with a DMR microscope while fluorescent images were taken using a SP8 scanning laser confocal microscope (Leica Microsystems).Sections of negative control tissue incubated with secondary antibodies were used for background setting previous to image acquisition.All images were captured under constant exposure time, gain, and offset.One hundred cells from 3 independent experiments were counted for each condition to calculate an average of the fluorescence observed patterns.The images were analyzed and processed (maximum intensity projections of Z-series) using ImageJ 1.45v.

Statistics
All statistics were performed using GraphPad Prism v5.01.Normal distribution of data was tested using the Kolmogorov-Smirnov test.Cell treatment experiments were analyzed using Student's t test and degrees of freedom were determined.Data are expressed as the mean ± standard deviation and p values less than 0.05 were considered significant.

ANKK1 is involved in myogenic differentiation
To study ANKK1 gene expression pattern during muscles differentiation we performed immunolabeling experiments using α-STk2 antibody (recognizes ANKK1 kinase domain) [6], and α-eMyHC in mice embryos at E14.5 and E17.5.A high level of ANKK1 signal was detected in embryonic myotubes distributed throughout the neuraxis, including the trunk and limb, as well as various cervical and facial muscles anlagen (Fig 1).The striated cardiac and the smooth muscle were negative for ANKK1 staining.

Leptomycin B inhibits the nuclear export of ANKK1 during myoblast differentiation
We have previously reported in silico evidence for nuclear localization signals (NLS) and nuclear export sequences (NES) on human ANKK1 [24].Here the in silico analysis of mice ANKK1 using the NUCDISC program (from the PSORT II server, http://psort.ims.u-tokyo.ac.jp) and NEtNES v1.1 server revealed unique and homologous consensus NLS and NES sequences when compared with the human protein (Fig 3A).The unique NLS in the mouse ANKK1 comprised amino acids 3 to 9 (PHRARRL).These in silico results suggest that nuclear exclusion of ANKK1 in myotubes could be mediated by NES.Then, we studied the effect of LMB, an inhibitor of nuclear export of NES containing proteins, on the subcellular localization of ANKK1 during differentiation.
C2C12 myoblasts were induced to undergo myotubes formation during 24 or 48 h and treated with LMB 24 h before fixed or harvested.After treatment, we observed a significant increase of ANKK1 signal in myoblasts nuclei and inhibition of the fusion of myotubes when compared to controls (Fig 3B).Specifically, C2C12 nuclei of treated myoblasts showed higher ANKK1 levels at both 24 (t: 2.728, p = 0.0117) and 48 h (t: 10.37, p <0.0001) (Fig 3C ), and a significant reduction in the fusion index (ratio between number of Actinin+ cells/total cells) at 24 (t: 2.135, p = 0.0441) and 48 h (t: 2.564, p = 0.0150) (Fig 3D).We also performed, in LMB-treated and untreated C2C12 cells, Western blot analysis of subcellular fractions using α-STk2 and α-MyoD (Fig 3E and 3F), and observed a significant nuclear ANKK1-kinase isoform increase after LMB treatment (Fig 3E) at 24 h (t: 2.505, p = 0.0336) (Fig 3F) while cytoplasmatic ANKK1 fulllength isoform levels remained unchanged in all conditions: 24 h (t: 0.3746, p = 0.7158); 48 h (t: 1.069, p = 0.3455).This result revealed that nuclear ANKK1-kinase retention by LMB in early differentiating myoblasts did not affect the full length isoform expression.In addition, ANKK1 showed more sensitivity to LMB treatment in comparison to MYOD (Fig 3E), although a significant increase of the former was evident after 48 h (t: 5.901, p = 0.0041) (Fig 3F).These findings demonstrate that LMB inhibits the nuclear export of ANKK1.In addition, ANKK1 expression pattern shows an inverse relationship to MYOD expression in early stages of differentiation.

ANKK1 is a marker of SCs and fiber subtype in adult muscles
ANKK1 expression pattern studies in muscles of adult mice were performed using FDB isolated fibers and transversal gastrocnemius serial sections.In FDB fibers, ANKK1 is located at cytoplasm and perinuclear regions, although it is totally absent in myonuclei (Fig 4).Immunolabeling experiments with α-STk2, α-Pax7 and α-MyoD showed ANKK1 expression in the nucleus and cytoplasm of all PAX7+ SCs (Fig 4A ) whereas the location of the protein was exclusively cytoplasmic in MYOD+ SCs (Fig 4B).To know whether the ANKK1-full length isoform is expressed in the cytoplasm of SCs, we used the α-STk3 antibody (Fig 4C and 4D) that recognizes the Ankyrin repeats domain of ANKK1 which is absent in ANKK1-kinase isoform.We found ANKK1 signal in the cytoplasm of PAX7+ and MYOD+ SCs (Fig 4C and 4D).Therefore, the location of ANKK1 isoforms in SCs was identical to that observed in myoblasts and cell-derived myotubes.In the same way, the study of ANKK1 in mice gastrocnemius sections using α-STk, α-STk2 and α-STk3 [6] (Fig 4E ) showed an identical expression pattern of ANKK1.Specifically, ANKK1 was expressed in the cytoplasm of a subpopulation of muscle fibers (Fig 4F).Further, fibers myonuclei were ANKK1−, while we observed some ANKK1 + positive nuclei (Fig 4F ), that would correspond to the SCs.Therefore, in adult mice ANKK1 is expressed in SCs and a subtype of muscle fibers.

ANKK1 is exclusively expressed in Fast-Twitch muscle fibers
To determine if the pattern of the differential expression of ANKK1 in adult muscle was related to fiber type, serial sections of mice transversal gastrocnemius muscles were studied using IHC staining for detection of ANKK1 (α-STk2), MyHC1 and MyHC2 (Fig 5A).We found ANKK1 signal in most of the MyHC2+ fibers (Fig 5A ) while the protein was absent in the MyHC1+ oxidative fibers.ANKK1 was also found in MyHC2+ fibers in triceps (more glycolytic muscle [26]) and soleus (more oxidative muscle [27]) (Fig 5B).
To know the influence of the metabolic environment on the selective ANKK1 expression in the muscles, we induced glycolytic metabolism in C2C12 by growing the cells with HG (30 mM glucose) or treating them with berberine (30 μM).A significant reduction in cellular oxygen consumption and an increase in lactic acid production were found in both conditions (Fig 5C).ANKK1 mRNA levels showed a significant increase with HG (t: 2.289, p = 0.0428) and berberine treatment (t: 3.834, p = 0.0021) (Fig 5D).However, Western blot of C2C12 subcellular fractions showed no significant changes for ANKK1-kinase (t: 0.9711, p = 0.3599) or fulllength isoform (t: 0.0075, p = 0.9418) after culture in HG (Fig 5E ).By contrast, in the case of berberine we observed a significant decrease of ANKK1-kinase for the nuclear enriched fraction (t: 2.626, p = 0.0392) while no changes were found for the full-length isoform (t: 0.1405, p = 0.8918) (Fig 5E).We also attempted to induce glycolytic metabolism in C2C12 cells by cultivating them under hypoxic conditions (2% oxygen) (S2 Fig) .Immunolabeling experiments with α-STk2 show a significant increase of ANKK1 in the nucleus of hypoxic cells after 24 (t: 5.802, p <0.0001) and 48 h (t: 6.995, p <0.0001, S2 Fig) .Taking altogether these results, we propose that there is a relation between ANKK1 gene expression and glycolytic metabolism that explains the specific location of the protein in Fast-Twitch muscle fibers.

ANKK1 is expressed in regenerating fibers in muscular dystrophies
Given that ANKK1 is found in muscle precursors during development and adulthood, we aimed to investigate the expression of ANKK1 in clinical samples of muscular dystrophies.We performed IHC detection of ANKK1 using a commercial antibody against a human epitope, since the antibodies used in previous experiments did not work in human adult muscle.Colocalization experiments were done in serial sections of human quadriceps and deltoid muscles using eMyHC/nMyHC as a marker of regenerating fibers.The samples were 2 controls and 4 patients with different muscular dystrophies.Specifically, we included Duchenne muscular dystrophy (DMD, MIM #310200), autosomal recessive limb-girdle muscular dystrophy 2A (LGMD2A, MIM #253600), limb-girdle muscular dystrophy 2B (LGMD2B, MIM #253601) and facioscapulohumeral dystrophy 1 (FSHD1, MIM #158900).Non-dystrophic muscle samples showed ANKK1 expression in a subpopulation of muscle fibers and the absence eMyHC/nMyHC expression (Fig 6).Regarding dystrophic patients, ANKK1 was expressed in 100% of regenerating fibers in DMD, LGMD2A and LGMD2B samples and 71.43% in FSHD1 muscle (Fig 6).

Discussion
This study shows for the first time that ANKK1 expression is not restricted to neural precursors and postmitotic neural cells but extends to the myogenic lineage.We found ANKK1 both in ANKK1 fluorescence intensity (a.u) in nucleus (N = 3) and (D) fusion index (N = 3).(E) Western blot of subcellular C2C12 fractions using α-STk2 and α-MyoD.β-ACTIN was used as control and α-TUBULIN as cytosolic marker.(F) Western blot bands were quantified using the ImageJ 1.45v software and the relative expression was represented related to β-ACTIN value.p < 0.05: Ã ; p < 0.01: ÃÃ ; p < 0.001: ÃÃÃ .a.u: arbitrary units; N: Nucleus; C: Cytoplasm.https://doi.org/10.1371/journal.pone.0197254.g003vivo in embryonic myotubes and in adult muscles and in vitro in myoblasts and fusing myotubes.Our results of the study of ANKK1 in muscles are the following: first, during the development ANKK1 is located in migrating myotubes where it shows a polarized cytoplasmic distribution; second, myoblasts and SCs, that are muscle precursors, show ANKK1 in their nuclei and cytoplasm while myotubes and muscle fibers only in their cytoplasm; third, in adulthood, ANKK1 is exclusively expressed in Fast-Twitch muscle fibers subtype; and fourth, there is a relation between glycolytic metabolism and the regulation of ANKK1 gene.
Early studies of ANKK1 in mice embryos showed that this protein is localized in the cytoplasm of the radial glia [6], which are the cells that guide migrating neurons of the developing central nervous system [28].In this work, we found that embryonic migrating myotubes also expressed ANKK1.Therefore it would be possible that ANKK1 participates in common molecular mechanisms of migration in both cell lineages.ANKK1 shows cytoplasmic location in cell-derived and in embryonic myotubes but not with the same pattern.In cell-derived myotubes ANKK1 localizes throughout the cytoplasm whereas migrating embryonic myotubes shows the protein in a pattern of apical polarization (Fig 1A).This polarized morphology in protrusions of the leading edge of myotubes clusters is characteristic of the front of the migrating cells [29][30].These findings show that ANKK1 has a role in the polarization of migrating cells during development, which is the first step of the migration process [31].It is still unknown how ANKK1 contributes to cell polarization; in any case, it is located in the front of myotubes clusters that are collectively migrating (Fig 1A -1C) or in individual migrating myotubes (Fig 1D).In addition, ANKK1 is not expressed in all embryonic myotubes, but those that show more elongated morphologies and characteristic protrusions of migration.In skeletal muscle development, migration is necessary for the correct localization of cell precursors in somites to adopt myogenic potential [32] and for the fusion of myoblasts along primary and secondary myogenesis [33].Given that cell migration is critical for normal embryonic development and functioning of an adult organism, the expression pattern of ANKK1 in the radial glia and in the embryonic myotubes would signal a role for this protein for efficient migration.
Previous studies of ANKK1 isoforms during brain development in mice had indicated that ANKK1-kinase is mainly expressed during proliferative states, whereas an inverse pattern with a prevalence of the ANKK1 full-length isoform was found in stages of neurons maturation [8].Here, we found an equivalent ANKK1 expression pattern in mice and human myogenic cell lines (C2C12 and rhabdomyosarcoma).During the proliferative stages ANKK1-kinase was found in myoblast nuclei, where it is downregulated throughout the differentiation process, while the myotubes show a prevalence of ANKK1 full-length isoform in their cytoplasm.Accordingly, we observed some ANKK1+ nuclei in cells nearby embryonic myotubes that would correspond to muscle precursors and the lack of expression at the nuclei of adult muscle fibers.Therefore, there seems to be a relationship between ANKK1 subcellular location and the cell cycle.Indeed, we previously reported that, during the cell cycle, ANKK1 is involved in G1 and M phases in different cell lines as well as in M phase, both in neurogenesis and in selfrenewal neural progenitors from mice embryo forebrain [8].Therefore it is plausible that ANKK1 expression in myoblasts nuclei in proliferative stages has an impact upon the cell cycle and its exit during muscle differentiation.Further, we also observed in differentiating C2C12 cells a reverse relationship between nuclear ANKK1 and the myogenic factor MYOD levels in a temporal coordination between cell cycle exit and differentiation.This reverse relationship was also evident after 24 h treatment with LMB that significantly retains nuclear ANKK1 while no effect was found for MYOD.We then propose that ANKK1 nucleus-cytoplasmic shuttle is carried out by exportins before the increase of MYOD in the progression of the myogenic lineage.
The ANKK1 study in adult muscle fibers shows that it is expressed exclusively in the cytoplasm of Fast-Twitch fibers, thus suggesting a linkage between the regulation of the ANKK1 gene and the metabolic environment.Fast-Twitch fibers obtain energy from stored glycogen and phosphocreatine, and they rather use a glycolytic metabolism [34][35].Here we demonstrated that the induction of a glycolytic metabolism by HG levels or hypoxic conditions in C2C12 causes an ANKK1 upregulation that could explain its expression pattern in muscles.In the case of berberine treatment, the decrease observed in nuclear ANKK1 could be explained by other pathways apart from glycolysis that are activated.For example, berberine activates myoblasts the differentiation through MYOD and p38 MAPK [36], and an inverse relationship between ANKK1 and MYOD was found.Besides, we also found ANKK1 expression in SCs that are the adult myogenic precursors [37][38].The comparison of ANKK1 isoforms expression pattern (nuclear ANKK1-kinase vs cytosolic ANKK1 full-length) between quiescent/selfrenewing and already committed SCs shows striking similarities with neural [39] and muscle cell precursors during development.Hence, it is tempting to consider that the ANKK1 expression pattern in SCs could be related to metabolic changes and/or entry or exit of these cells into the cell cycle.In any case, our results in embryonic precursors and adult muscles indicate that ANKK1 expression in the myogenic lineage is related to glycolytic metabolism.
Many similarities have been described between embryonic and adult myogenesis [40][41].The finding that ANKK1 is expressed in human SCs of healthy muscles and, in SCs and the regenerating muscle fibers of dystrophic patients, would place this protein in overall mechanisms for muscle generation/repair.From a genetic point of view, the ANKK1 gene location in the NTAD genomic cluster where the NCAM1 gene maps also supports a function of this locus upon myogenesis.In addition to the well-established role of NCAM1 in neurogenesis, this protein has also been identified in embryonic myoblasts [42], adult SCs [43] and regenerating fibers in different neuromuscular pathologies [44][45][46].Therefore, the findings of ANKK1 and NCAM1 in embryonic and adult myogenesis extend the role of the NTAD cluster beyond neurodevelopment to other cell lineages during development.Finally, since ANKK1 belongs to the RIP family and downregulation of RIP2 in myogenesis is a prerequisite for myoblasts differentiation [12], the need to know the part played by ANKK1 warrants further studies.
In conclusion, we are providing the first evidence of ANKK1 expression in the myogenic lineage during development and in adulthood.The characteristic expression pattern of ANKK1 in myogenic differentiation and regeneration would make this protein a useful marker of muscle fiber status.p < 0.05: Ã ; p < 0.01: ÃÃ ; p < 0.001: ÃÃÃ .a.u: arbitrary units.(PDF)

Fig 6 .
Fig 6.ANKK1 is expressed in regenerating fibers in neuromuscular dystrophies.ANKK1 and a mix of embryonic and neonatal myosin heavy chains (eMyHC/nMyHC) immunostaining of adult skeletal muscle sections.The samples were non-dystrophic (control) and dystrophic patients (DMD, LGDM2A, LGMD2B and FSHD).ANKK1 and eMyHC/nMyHC were studied in serial muscle sections.The black asterisk indicates the same fiber.The black arrows indicate ANKK1+ nuclei.Scale bar: 50 μm.All regenerating fibers from each tissue section were counted and analyzed.Images are representative for the group averages.https://doi.org/10.1371/journal.pone.0197254.g006