cell biochemistry and function

Cell Biochem Funct 2016; 34: 414–422.

Published online in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/cbf.3202

MicroRNA signatures predict dysregulated vitamin D receptor and

calcium pathways status in limb girdle muscle dystrophies (LGMD)
2A/2B
M. Aguennouz1*, C. Lo Giudice1, N. Licata1, C. Rodolico1, O. Musumeci1, M. Fanin2, A. Migliorato1, M.
Ragusa3, V. Macaione1, R. M. Di Giorgio1, C. Angelini2 and A. Toscano1
1
Department of Clinical and Experimental Medicine, University of Messina, Italy
2
Neurological Clinic, University of Padua, Italy
3
Department of Biomedical and Biotechnological Sciences Biology, Genetics and Bioinformatics Unit, University of Catania, Italy

miRNA expression profile and predicted pathways involved in selected limb-girdle muscular dystrophy (LGMD)2A/2B patients were
investigated. A total of 187 miRNAs were dysregulated in all patients, with six miRNAs showing opposite regulation in LGMD2A versus
LGMD2B patients. Silico analysis evidence: (1) a cluster of the dysregulated miRNAs resulted primarily involved in inflammation and
calcium metabolism, and (2) two genes predicted as controlled by calcium-assigned miRNAs (Vitamin D Receptor gene and Guanine
Nucleotide Binding protein beta polypeptide 1gene) showed an evident upregulation in LGMD2B patients, in accordance with miRNA
levels. Our data support alterations in calcium pathway status in LGMD 2A/B, suggesting myofibre calcium imbalance as a potential
therapeutic target. Copyright © 2016 John Wiley & Sons, Ltd.

key words—limb-girdle muscular dystrophy 2A/2B; microRNA expression; vitamin D Receptor; guanine Nucleotide Binding protein beta
polypeptide 1; inflammation and calcium metabolism

INTRODUCTION (CAPN3), is considered the most frequent form of recessive

Limb-girdle muscular dystrophies (LGMDs) are a heteroge- LGMD worldwide.
neous group of disorders, characterized by a progressive CAPN3 gene maps to chromosome 15q15 and is com-
weakness and wasting of the muscles of the pelvic and posed by 24 exons with an estimated size of about 53 kb.
shoulder girdle. The inheritance may be autosomal domi- Over 460 critical mutations have been described in this
nant (LGMD1) or recessive (LGMD2). To date, eight auto- gene, many of which alter the catalytic activity of the pro-
somal dominant LGMD (A–H) and 23 autosomal recessive tein.3 Other mutations may affect enzyme’s structural in-
forms (A–W) have been mapped, most of them having their tegrity, modifying its intra and intermolecular interactions
protein products identified.1 and leading to a decreased stability or altered localization.4
Clinically, LGMDs show a great phenotypic variability, Binding of calpain 3 to titin, a giant myofibrillar protein
ranging from severe forms, with onset in the first decade that serves as a scaffold for sarcomeric proteins, provides
and rapid progression, to milder forms with later onset and an example of such intermolecular interactions. CAPN3
slower progression. appears to have a very rapid turnover mediated by autoca-
To date, the main mechanisms recognized in LGMDs on- talysis, possibly reflecting the need for a precise regulation
set comprise: structural defects in the muscle membrane, of its activity. Titin binding stabilizes CAPN3 from
muscle membrane repair deficiencies and defects in sarco- autoproteolytic degradation, keeping it in an inactive state
mere remodelling, cytoskeletal structures and and preventing its rapid autolysis.5 The loss of proteolytic
cytoskeleton–membrane impaired interactions.2 activity and the loss of titin anchorage are two potential
Limb-girdle muscular dystrophy type 2A (LGMD2A) or ways by which CAPN3 mutations might result
calpainopathy, caused by mutation in the skeletal muscle- pathogenic.6
specific calcium-dependent cysteine protease calpain 3 Calpain 3 seems to exert multiple functions in the muscle
tissue: besides maintaining sarcomere integrity by the regu-
lation of sarcomeric protein turnover, calpain 3 has also
*Correspondence to: Aguennouz M’hammed, Department of Clinical and been observed in other cellular compartments, such as the
Experimental Medicine, University of Messina, Italy. subsarcolemmal membrane, the triad-associated protein
E-mail: [email protected] complex and the nucleus.2,7,8

Received 12 January 2016

Revised 12 May 2016
Copyright © 2016 John Wiley & Sons, Ltd. Accepted 1 June 2016
 MICRORNA SIGNATURES PREDICT DYSREGULATED VITAMIN D RECEPTOR AND CALCIUM PATHWAYS STATUS 415

Furthermore, Stuelsatz et al. (2010) demonstrated that Written informed consent was obtained from all subjects
calpain 3 is also important for myogenic cell fate determina- or their caregivers at the moment of diagnostic procedures,
tion, by controlling the level of MyoD protein within the even for research purpose, according to the Declaration of
myoblast population induced to differentiate. This effect Helsinki.
leads to the establishment of a pool of self-renewing reserve The diagnosis of LGMD 2A or 2B was based on fulfil-
cells.9 ment of the following criteria: (1) a clinical phenotype char-
A secondary reduction of calpain 3 can also be observed acterized by progressive muscle weakness and wasting
in LGMD2B or dysferlinopathy, caused by deficiency of affecting primarily or predominantly shoulder-girdle and
functional dysferlin protein (DYSF). Dysferlin is a 230-kD pelvic muscles, in keeping with the diagnostic criteria for
large transmembrane protein involved in the membrane- LGMD, (2) dystrophic features at the muscle biopsy; immu-
repair process, in the intracellular vesicle system and in nohistochemical (IHC) and western blot (WB) analysis were
T-tubule development in skeletal muscle.10,11 routinely carried out according to standard techniques: the
Calpain 3 is a component of the dysferlin complex where muscle sections were analysed using a monoclonal antibody
it acts as a regulatory element by cleaving AHNAK, a against dysferlin and calpain 3, followed by WB on the ho-
protein associated to the complex and involved in mogenates to confirm dysferlin and calpain 3 deficiencies.16
subsarcolemmal cytoarchitecture and membrane repair. All antibodies, including monoclonal antibody Calp
Defective calpain 3 results in AHNAK accumulation, 3d/2C4, were purchased from Novocastra. Autocatalytic ac-
detectable in skeletal muscle of calpainopathy patients, tivity of calpain3 was determined according to Fanin
suggesting a function of calpain 3 in muscle membrane procedure.17
homeostasis.2 Stored frozen muscle tissue at 80 °C were obtained from
Although significant progresses have been made in the the vastus lateralis of 15 patients affected by LGMD 2A or
understanding of muscle dysfunctions related to CAPN3 2B subdivided in: Group 1 (G1) five patients with primary
deficits, the knowledge of the underlying mechanisms has deficiency (absence of CAPN3 at WB genetically confirmed
yet to be completely clarified. for CAPN3 gene mutations); Group 2 (G2) five patients
In this sense, a substantial role of miRNAs, known for with autocatalytic activity deficiency of CAPN3 confirmed
their involvement in multiple metabolic pathways, could by genetic studies on CAPN3 gene; Group 3 (G3) five pa-
explain the variety of clinical manifestation shown by tients with CAPN3 protein secondary deficiency because
calpainopathy.12 of dysferlin gene mutations (Table 1). Muscle specimens
In fact, different miRNAs enter in the regulation of mus- from five aged sex matched and healthy patients were used
cle development: for example, miR-1 and miR-133 modu- as controls.
late skeletal muscle proliferation and differentiation in
cultured myoblasts, while miR-206 regulates connexin 43 RNA isolation and miRNA reverse transcription
(Cx43) expression during skeletal muscle development and
RNAs were isolated using mirVana miRNA Isolation Kit
mediates MyoD-dependent inhibition of the follistatin-like
(Ambion, Austin, TX) according to the manufacturer’s in-
1 (Fstl1) and utrophin (Utrn) genes in myoblasts.13–15
struction for total RNA isolation. RNA was reverse tran-
Evaluating by a bioinformatic approach the expression of
scribed using the TaqMan microRNA Reverse
specific skeletal muscle miRNAs in different groups of
Transcription Kit (Life Technologies) in combination with
patients affected by calpainopathy could highlight their
stem-loop Megaplex primer pool (Life Technologies),
epigenetic role in the pathology, thus helping to unravel
which allows simultaneous reverse transcription of 384
new potential mechanisms related to the muscle impairment.
miRNAs and endogenous controls. Briefly, 300 ng of total
RNA was supplemented with RT primer mix (10×), RT
buffer (10×), MultiScribe Reverse Transcriptase (10 U/μl),
MATERIALS AND METHODS dNTPs with dTTP (0.5 mM each), MgCl2 (3 mM) and AB
Patient samples RNase inhibitor (0.25 U/μl) in a total reaction volume of
80 μl. The cDNA synthesis was performed in a thermal cy-
A total of 15 muscle specimens of pharmacologically
cler following steps: 2 min at 16 °C, 1 min at 42 °C and 1 s
untreated patients were available for this study. All patients
at 50 °C for 40 cycles and enzyme inactivation at 85 °C for
diagnosed as affected with LGMD were recruited from our
5 min.
neuromuscular centre. The muscle biopsies were stored with
the DNA samples used in the Bank of DNA, Nerve and
Muscle Tissues, of our Department of Neurological Real-time qPCR
Sciences. The Institutional Review Board (IRB) of the Samples were assayed by TaqMan® Low Density Arrays
Department of Neurosciences of the University of Messina, (TLDA) with the aim of detecting and quantifying mature
Italy, discussed and approved this project two years ago. miRNAs. TLDA system based on Life Technologies
The IRB have allowed this project because the study 7900HT Micro Fluidic Cards contained 384 preloaded hu-
concerns a retrospective research for biomarkers on man miRNA targets, all catalogued in the miRBase data-
biological materials of patients with specific muscle disor- base, and two endogenous controls, small nucleolar RNAs
ders, which were contained in the Department Biobank. (snoRNAs), RNU48 (SNORD48) and RNU44 (SNORD44).

Copyright © 2016 John Wiley & Sons, Ltd. Cell Biochem Funct 2016; 34: 414–422.
416 m. aguennouz ET AL.

Table 1. Limb girdle muscular dystrophy 2A and 2B patients grouped according to CAPN3 deficit

CAPN3 nucleotide CAPN3 nucleotide

Patients Gender/age Age of onset Calpain 3 WB analysis * change Allele 1 change Allele 2
LGMD2A G1 94 kDa 30 kDa
LGMD2A G1-1 M/14 4 0.0 0.0 c.1063C > T c.1939G > T
LGMD2A G1-2 M/15 8 0.0 0.0 c.[883_884GA > CT; 887delA] c.2442G > A
LGMD2A G1-3 F/24 6 0.0 0.0 c.662G > T c.1333G > A
LGMD2A G1-4 M/55 45 3.6 15.8 c.1466G > A c.1711delC
LGMD2A G1-5 F/52 13 0.0 0.0 c.598_612del c.2380 + 1G > T
LGMD2A G2
LGMD2A G2-1 M/10 10 60 66 c.1469G > A, ex.11 c.1469G > A, ex.11
LGMD2A G2-2 F/7 4-5 65 28 c.479C > G, ex.3 c.1714C > T, ex.13
LGMD2A G2-3 F/30 28 67 50 c.1469G > A, ex.11 c.1469G > A, ex.11
LGMD2A G2-4 M/42 19 74 55.6 c.1469G > A, ex.11 c.1469G > A, ex.11
LGMD2A G2-5 F/49 47 35.8 30.6 c.1469G > A, ex.11 c.1469G > A, ex.11
Patients Gender/Age Dysferlin WB analysis * Calpain 3 WB analysis * DYSF nucleotide change Allele 1 DYSF nucleotide change Allele 2
LGMD2B G3 94 kDa 30 kDa
LGMD2B G3-1 M/40 0.0 24 20 c.2875C > T c.5813_5821del
LGMD2B G3-2 F/50 0.0 23.8 25 c.5200 + 1G > A c.5200 + 1G > A
LGMD2B G3-3 M/39 0.0 17.7 13.2 c.2875C > T c.5813_5821del
LGMD2B G3-4 M/44 0.0 25.6 23.8 c.895G > C c.2077delC
LGMD2B G3-5 M/26 0.0 16 15.4 c.2494C > T c.2494C > T

G1 (Primary CAPN3 deficit); G2 (Autocatalytic CAPN3 deficit); G3 (SecondaryCAPN3 deficit).

*expressed as % of residual protein at semi-quantitative WB analysis.

Each cDNA sample was diluted with nuclease free water technology as TLDA, the reaction volume is much higher
and mixed with Taqman Universal PCR Mastermix, (20 μl) and their reliability has been proven.19 Single Tube
according to manufacturer instructions. The mixture was qRT-PCR was performed according to the ΔΔCT method
finally loaded into TLDA Micro Fluidic Cards containing and in triplicate, including no template controls. Relative
384 human miRNAs probes. The card was briefly centri- changes in miRNA expression were quantified using the
fuged for 1 min at 280 g to favourite sample distribution into Real Time Fast 7900 HT system (Life Technologies), using
the multiple wells connected to the fill ports and then sealed the small nucleolar RNU48 and RNU44 (Life Technologies)
for preventing well-to-well contamination. PCR amplifica- for normalization.
tion reactions were carried out in a total volume of 100 μl,
containing 50 μl of TaqMan Master Mix 2X and 50 μl of
diluted cDNA. Cycling conditions were as follows: 2 min at Bioinformatic analysis
50 °C, 10 min at 95 °C followed by 40 cycles of 30 s at 97 °C Dysregulated miRNAs were searched on DIANA miRPath
and 1 min at 59,7 °C. All PCR reactions were performed on v.2.0 and clustered according to their targets. Interactions
Real Time Fast 7900 HT system (Life Technologies). (predicted and/or validated) were then combined with ad-
Raw Ct values were calculated using the RQ manager vanced merging and meta-analysis embedded algorithms to
software v1.2.1 (Life Technologies) with automatic baseline
find statistical relevant miRNA pathways (p < 0.05).20
and threshold settings, according to the manufacturer’s
Targets genes found on Mirpath were further analysed by
instructions.
the GeneMANIA Cytoscape plugin for expanding their in-
Fold changes (FC) in miRNA expression levels were
teraction network. Cytoscape is an open-source software
determined in each study group by the 2 ΔΔCT method.18
for visualization and analysis of molecular interaction net-
Resulting data were submitted to the ABqPCR package
works, supporting pathway integration, gene expression
(Life Technologies) in Excel to determine corresponding
profiles and other state data. Cytoscape standalone platform
p-values by Student T test. Considering a significance provides a basic set of operations; additional features, such
threshold of p < 0.05, fold-change cut-offs were set up as functional clustering and connection to databases, are
respectively at >3.0 for upregulation and <0.36 for down available as plugins.
regulation. GeneMANIA plugin identifies the most related genes to a
query gene set using a guilt-by-association approach.
Single quantitative real-time polymerase chain reaction Plugin’s default settings were modified according to a max-
To verify the accuracy of our TLDA data, we performed imum of 30 related genes, using as attributes: miRNA target
separate quantitative RT-PCR (qRT-PCR) experiments for predictions, pathway, co-localization, co-expression, physi-
representative miRNAs using single TaqMan miRNA as- cal interaction and similar protein domain.21
says (Life Technologies,) in accordance with the manufac- Another Cytoscape plugin, MCODE, with default thresh-
turer’s protocols. Single TaqMan assays rely on the same old, was then applied on the whole GeneMANIA network

Copyright © 2016 John Wiley & Sons, Ltd. Cell Biochem Funct 2016; 34: 414–422.
 MICRORNA SIGNATURES PREDICT DYSREGULATED VITAMIN D RECEPTOR AND CALCIUM PATHWAYS STATUS 417

for extracting significant gene clusters.22 Clusters were Missouri, USA) at RT for 1 h. Membrane was developed
identified as highly interconnected network’s regions. using ECL Plus Western Blotting Detection kit (Amersham
miRNA enrichment for selected genes was finally done Biosciences) following the manufacturer’s protocol.
by Cytarget linker Cytoscape plugin creating a new network Quantification of the detected proteins was carried out
for each gene and searching on Microcosm miRNA targets using the Alpha Digi Doc apparatus (Alpha Innotech Corp,
database.23 USA) for image acquisition (8 bit gray-scale), and by the
ImageJ software for densitometric analysis. The results were
Immunohistochemistry finally expressed as relative density for each sample after
GAPDH normalization, and statistical analysis was
Muscle biopsy specimens were frozen in isopentane cooled
performed using one way Anova followed by Friedman
in liquid nitrogen. Serial cryostat sections (7 μm thick) were
‘post hoc’ analysis.
mounted on poly-L-lysine-coated slides and air-dried for
30 min at room temperature (RT). They were then incubated
for 60 min at RT in mouse monoclonal antibodies against
Vitamin D Receptor VDR 1:40 (Santa Cruz Biotechnology RESULTS
Inc. CA, USA) and Guanine Nucleotide-Binding Protein 1
In this study we carried out a comparative miRNA expres-
—GNB1 at 1:40 (Abcam, Cambridge, UK). To block non-
sion profiling in human skeletal muscle of three groups of
specific binding of the antibody, sections were preincubated
patients with calpainopathy—primary deficiency (G1), auto-
with 1:10 diluted normal goat serum. Control for staining
catalytic activity deficiency (G2) and secondary deficiency
specificity was omission of the primary antibody.
(G3)—and unaffected controls, in an attempt to identify
Immunodetection was performed using a biotin–avidin
similarities and differences in two distinct forms of LGMD.
system followed by horseradish peroxidase staining with
A total of 187 miRNAs, corresponding to 48.7% of the
3.3-diaminobenzidine tetrahydrochloride. Sections were
human miRNA probes present on the panel array, were
examined and photographed with a Zeiss Primo Star
founded dysregulated at a significant level in all patients
photomicroscope (Carl Zeiss Inc, Oberkochen, Germany).
compared with controls (Table S1 in file S1).
The overall differentially expressed miRNAs were then
Western blotting analysed for selecting potential target genes and correspond-
About 40 mg of muscle specimen was homogenized in a ing pathways. To limit the number of predicted target genes
glass tube with Teflon dounce pestle in 20 volumes of a de- to the most likely candidates, we have carried out a bioinfor-
tergent saline buffer containing lysis buffer (20 mM KCl, matic approach using Mirpath and Microcosm miRNA
15% SDS and 5% β-mercaptoethanol). Samples were heated target prediction tools.
for 10 min at 100 °C and centrifuged for 10 min at 12 000 g. Seventeen miRNAs out of the 384 analysed showed the
Protein concentration of tissue homogenates was determined same trend of expression in all groups (G1, G2, G3) versus
by Lowry assay. Protein balanced samples were prepared controls: five of these miRNAs (miR-198, miR-34a, miR-
for sodium dodecyl sulphate polyacrylamide gel electropho- 432, miR-503 and miR-650), were upregulated; 12
resis (SDS-page) in twofold loading buffer containing miRNAs (miR-133a, miR-193b, miR-196a, miR-302d,
0.25 M Tris (pH 6.8), 0.2 M DTT, 10% SDS, 0.02% miR-365, miR-376b, miR-383, miR-422b, miR518e,
bromophenol blue and 20% glycerol in distilled water. Fifty miR-589, miR-615 and miR-657), were downexpressed
micrograms of proteins per line was routinely resolved by (Table S2 in file S1).
12% SDS-PAGE for 2 h at 130 V. In addition, six miRNAs (miR-154, miR-184, miR-485-
Following electrophoresis, separated proteins were later- 3p, miR-493, miR-519e and miR-654) showed an opposite
ally transferred to nitrocellulose membranes in transfer regulation in G1 and G2 versus G3 (Table S1 in file S1).
buffer containing 0.192 M glycin and 0.025 M Tris pH 8 In detail, miR-519e was downregulated in the patients G1
with 20% methanol. and G2, and upregulated in patients G3, while miR-154,
At a constant voltage of 100 V for 1 h at 4 °C, blots were miR-184, miR-485-3p and miR-654 resulted upregulated
blocked for 1 h at RT in a saturating solution containing in the groups of patients G1 and G2, and downexpressed
0.9% NaCl, 1% bovine albumin serum and 0.05 % in G3 patients.
Tween-20. Moreover, 29 miRNAs (26 up/3 down), appeared
Membranes were then incubated with antibodies against dysregulated only in the group of patient G1, 45 miRNAs
Vitamin D Receptor (VDR) (1:400 dilution; polyclonal) (40 down/5 up) only in G2 and 14 miRNAs (10 down/4
(Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA), up) only in G3 (Tables S3a,b,c in file S1).
Guanine Nucleotide-Binding Protein 1 (GNB1) (1:2500 di- By merging of G1 versus G2 TLDA expression data, a
lution; polyclonal) (Abcam, Cambridge, UK) and GAPDH miRNa in particular, miR-200, appeared highly expressed
(1:8000 dilution; polyclonal) (Santa Cruz Biotechnology, in CAPN3 primary deficit versus CAPN3 autocatalysis
Inc., Santa Cruz, CA, USA) at 4 °C overnight. Blots were deficit. From miRpath results, miR-200 appears to be
then washed and the second incubation was performed in involved in regulating extracellular matrix-receptor interac-
blocking buffer containing 1:20 000 dilution of the appropri- tions (p = 2.672864e 05) and adherents junctions
ate HRP-conjugated secondary antibody (Sigma-Aldrich, (p = 0.0004671301).

Copyright © 2016 John Wiley & Sons, Ltd. Cell Biochem Funct 2016; 34: 414–422.
418 m. aguennouz ET AL.

The remaining miRNAs showed a different pattern ex- Applying the Mcode clustering tool to the whole
pression in the three study groups and targeted numerous GeneMANIA network, we found a defined group of eight
predicted genes (mirPath) involved in various cellular pro- genes: Calcineurin subunit B type 1 (PPP3R1), ATPase
cesses, such as: endocytosis (p = 1.387705e 15), actin cy- Ca2+ Transporting Plasma Membrane 1 (ATP2B1),
toskeleton remodelling (p = 1.316438e 14), focal Sarcoplasmic/Endoplasmic Reticulum Calcium ATPase 2
adhesion (p = 5.888142e 14), calcium signalling pathway (ATP2A2), Protein Kinase C Alpha Type (PRKCA), Protein
(p = 6.715275e 12), inflammation (p = 3.288014e 12), Kinase C Beta 1 (PRKCB), Phosphorylase Kinase Subunit
ubiquitin mediated proteolysis (p = 1.157777e 11), cell cy- Beta (PHKB), Protein Phosphatase 3 Catalytic Subunit Al-
cle (p = 3.60037e 11) and apoptosis (p = 1.558425e 07), pha Isozyme (PP3CA) and Guanine Nucleotide-Binding
suggesting different pathogenic pathways. Protein 1 (GNB1).
Considering that the main pathological mechanism in This cluster was centred on GNB1, and controlled trough
calpainopathy is because of a deficit in the calcium- the same gene, by miRNAs miR-485-3p, miR-154 and miR-
dependent CAPN3 protein, we have extracted from 493 (Figure 1A).
Mirpath all the target genes belonging to calcium signal- These miRNAs, appeared significantly downregulated at
ling pathway and inserted them into the GeneMANIA TLDA gene expression analysis in LGMD2A (groups G1
Cytoscape plugin. and G2) versus LGMD2B patients (group G3).

Figure 1. (A) GNB1 cluster. miRNAs are represented by rhomboids, interacting genes by circles. The purple line indicates that the gene products are co-
expressed. Cluster Gene list: PPP3R1:Protein Phosphatase 3, Regulatory Subunit B, Alpha; PRKCB: Protein Kinase C, Beta-1; PHKB: Phosphorylase Kinase,
Beta Subunit; PPP3CA: Protein Phosphatase 3Catalytic Subunit (CALNA1); PRKCA: Protein Kinase C; ATP2A2: ATPase, Ca(2+)-Transporting, Slow-
Twitch (Sarcoplasmatic Reticulum Ca(2+)-ATPase 2; SERCA2); ATP2B1: Transporting, Plasma membrane 1; GNB1: Guanine Nucleotide-Binding Protein
Beta1. (B) MiRNAs targeting GNB1 according to Microcosm Database, The arrows indicate a direct control on the mRNA. (C) Relationship between
GNB1 and VDR through BAZ1B and miR-493: The complexity of the interaction between the various genes involved in this cluster is reported with different
coloured lines: Blue: All the genes are co-located; Purple: all products are co-expressed; Pink: physical interaction on gene product; Orange: modulating in-
teraction; Green: regulating interaction; Cluster Gene list: TBXA2R: Thromboxane A2 Receptor, Platelet; GRIK3: Glutamate Receptor, Ionotropic, Kainate
3; GNG7: Guanine Nucleotide-Binding Protein, Gamma-7; GNG13: Guanine Nucleotide-Binding Protein, Gamma-13; BAZ1B: Bromodomain Adjiacent to
Zinc Finger Domain 1B, (Williams Syndrome Transciption Factor; WSTF); GNA13: Guanine Nucleotide-Binding Protein, ALPHA-13; GNG2: Guanine Nu-
cleotide-Binding Protein, GAMMA-2, GNAQ: Guanine Nucleotide-Binding Protein, Q Polypeptide; GNAZ: Guanine Nucleotide-Binding Protein, ALPHA Z
Polypeptide; PTGIR: Prostaglandine I2 Receptor; TRIM24: Tripartite Motif-Containing Protein 24 (Transcriptional Intermediary Factor 1; TIF1); GNGT1:
Guanine Nucleotide-Binding Protein Gamma-Transducing Activity Polypeptide1; OPRM1: Opioid Receptor, Mu-1; SERINC1: Serine Incorporator 1;
P2RY1: Purinergic Receptor P2Y, G Protein-Coupled1; VDR: Vitamin D Receptor

Copyright © 2016 John Wiley & Sons, Ltd. Cell Biochem Funct 2016; 34: 414–422.
 MICRORNA SIGNATURES PREDICT DYSREGULATED VITAMIN D RECEPTOR AND CALCIUM PATHWAYS STATUS 419

After Cytarget miRNA enrichment (Microcosm these two genes in the secondary CAPN3 deficit (G3 group)
database) also mir-132, mir-598, miR-184 and miR-129 compared with G1 and G2 and approved by significative
were found regulating GNB1 (Figure 1B). MiR-184 statistical analysis (p < 0.5) (Figure 2B).
resulted upregulated in groups G1 and G2 versus G3;
mir-129 was founded only in G2 group (upregulated) and
was not detected at a significant level in G1 and G3
patients. The others two miRNAs, mir-598 and mir-132, DISCUSSION
were not included in the TLDA. Calpainopathy is a muscular disorder characterized by sym-
Another gene involved in calcium metabolism and indi- metric and progressive weakness of proximal muscles. It
rectly connected to GNB1 was found in Vitamin-D receptor shows intra- and inter-familial clinical variability. Typical
(VDR). According to the GeneMANIA analysis, VDR is in LGMD2A phenotype appears to be associated to a deficit
the same pathway and physical interacting with in the product of the calpain-3 (CAPN3) gene, a muscle-
bromodomain adjacent to zinc finger domain, 1B (BAZ1B) specific calcium-activated neutral protease.
(Figure 1C). BAZ1B encodes a member of the A secondary reduction of CAPN3 product is also possible
bromodomain protein family, involved in chromatin- in LGMD2B, another type of autosomal recessive LGMD
dependent regulation of transcription and absent in caused by mutations in the dysferlin gene (DYSF).
Williams–Beuren syndrome, a neurodevelopment disorder Although calpain-3 is known to be directly involved in
associated with hypercalcaemia.24 sarcomere remodelling, and dysferlin is proposed to play a
Interestingly, same miRNA, miR-493, regulates both role in membrane repair, the pathogenic mechanism associ-
GNB1 and BAZ1B. Similarly to other miRNAs targeting ated with deficits in these proteins is still not clear.
GNB1, miR-493 imbalanced upregulated in LGMD2A Throughout the last decade, the interest on the epigenetic
(groups G1 and G2) respect to LGMD2B (group G3) (Table mechanisms involved in gene regulation has increased enor-
S1 in file S1). mously. In particular, a class of small non-coding RNAs,
IHC analysis showed an evident expression of both pro- known as microRNAs (miRNAs), has emerged as having a
teins GNB1 and VDR in muscle fibbers of LGMD2A and central role in the post-transcriptional regulation of gene ex-
LGMD2B patients located in the nucleus, and in endomysial pression for a broad range of biological processes. On the
connective tissue (Figure 2A). There were not differences other hand, the discovery of muscle-specific miRNAs has
between the percentage of positive nuclei in all LGMDs. confirmed their importance in myogenesis, myopathies and
WB analysis of GNB1 and VDR, according with miRNA adaptation to various forms of exercise. Eisenberg (2007)
expression profile, confirmed an evident upregulation of described a total of 185 miRNAs that are up- or

Figure 2. (A) GNB1 (A) and VDR (B) antibodies immunoreactivity in endomysial connective tissue and in several muscle nuclei (arrows) in serial sections
with haematoxylin–eosin stain (C) – 280 X. (B) Western blotting study of VDR and GNB1. Representative blots from LGMD2A (G1 and G2), LGMD2B
(G3) and controls. Values in the graphs (y) are reported as relative densitometric units. Corresponding blot images on the left show an increase in VDR and
GNB1 expression in LGMD2B muscle. VDR expression and GNB1 are significant in G3 versus G1 and G2 with a significance of p < 0.5

Copyright © 2016 John Wiley & Sons, Ltd. Cell Biochem Funct 2016; 34: 414–422.
420 m. aguennouz ET AL.

downregulated in 10 major muscular disorders, and more by the dystrophic muscle to promote myoblast fusion
recently, Zacharewicz (2013) reviewed a defined group of needed for muscle regeneration.
miRNAs involved in muscle development, disease and The calcium metabolic pathway seems to be a common
function.12,25 target of two miRNAs strongly down regulated in the
In the present study, we explored the expression pattern secondary deficit, miR-184 and miR-493. As reported in
of a panel of 384 miRNA in patients affected by primary, Mirpath, mir-184 is primary involved in B cell receptor sig-
autocatalysis and secondary deficits of calpain 3. nalling pathway (p = 5.661986e 05). Genes targeted by
A total of 187 miRNAs showed to be dysregulated mir-184 include the Nuclear Factor of Activated T-Cells
compared to normal tissues and highlighted a differential (NFATC2) and V-akt murine thymoma viral oncogene ho-
expression pattern in the three study groups, suggesting that molog 2 (AKT2), both regulating and promoting acute in-
different pathways are involved in the pathogenesis and flammation.30,31 Downregulation of mir-184 in G3 group
progression of the disease. compared to G1 and G2 groups, could reflect inflammatory
Interestingly, 17 miRNAs showed the same expression trend processes exacerbated in the LGMD2B dystrophic muscle.
in all the three groups analysed. Among these, five up-regulated WB analysis of two predicted targets of mir-184 and mir-
miRNAs: miR-198, miR-34a, miR-432, miR-503 and miR-650 493, VDR and GNB1, showed an evident upregulation of
participate in growth and differentiation processes. these two genes involved in Ca2+ release in the secondary
MiR-198 and 34a act by repression of mitogenic and CAPN3 deficit.
motogenic pathways diminishing cell growth and migration, GNB1 is the beta subunit of the heterotrimeric guanine
miR-432 seems to play a not yet clear function in muscle de- nucleotide-binding proteins (G proteins). Different kinds of
velopment, mir-503 is induced during myogenesis promot- signals that trigger G protein activation lead to the dissocia-
ing cell cycle arrest via cyclin inhibition and mir-650 tion of its trimeric form into a Gα subunit and a Gβγ com-
targets proteins (cyclin-dependent kinase 1, inhibitor of plex. Whereas Gα-dependent signalling pathways involve
growth 4 and early B-cell factor 3) important in cell prolifer- the activation of soluble messenger cascades, Gβγ-mediated
ation and survival.26,27 effects typically occur through direct binding of the complex
As higher expression of miRNAs should downregulate to target effectors; one class of these targets is represented
target genes post transcriptionally, we suppose that the by voltage-gated calcium channels. In this sense, it has been
dysregulation of these miRNAs could affect regeneration demonstrated that GNB1 is capable to interfere with
mechanisms in LGMD2A and LGM2B disorders. excitation–contraction coupling in adult mouse skeletal
A total of six miRNAs (miR-154, miR-184, miR-485-3p, muscle by acting on L-type calcium channels.32
miR-493, miR-519e and miR-654) showed an opposite reg- VDR is an intracellular hormone-receptor that specifically
ulation in G1 and G2 versus G3 (Table S1 file S1). Among binds 1,25(OH)2D3 (calcitriol) and mediates its effects. In
these, miR-519e was downregulated in the patients G1 and skeletal muscle cells, VDR regulates the transcription of
G2, and upregulated in the patients G3, while miR-654 re- genes involved in calcium handling, cell differentiation
sulted upregulated in the groups G1 and G2, and and proliferation. Calcitriol rapidly promotes muscle cell
downexpressed in the group G3. Ca2+ influx by G-proteins-mediated stimulation of protein
According to the mirPath database search, both miR-519 kinase C (PKC) and protein kinase A (PKA), and subse-
and miR-654 target cyclin-dependent kinase inhibitor 1A quent activation of voltage-dependent calcium channels
(CDKN1A) gene. The expression of this gene is tightly con- (VDCC) by phosphorylation.33
trolled by the tumour suppressor protein p53, through which Elevated Ca2+ levels contribute to muscle damage; in par-
this protein mediates the p53-dependent cell cycle G1 phase ticular, increased sarcoplasmic calcium interferes with
arrest in response to a variety of stress stimuli. excitation–contraction coupling, which might account for
Cell-cycle arrest is a fundamental step during differentia- the muscular weakness typical of the dystrophic
tion of myoblasts into mature myotubes and is crucial for phenotype.34a,b
muscle regeneration. As reported by Davidovic et al., a Studies on calpain 3 protease-inactive knock-in mice have
key regulator of this event is the CDKN1A protein, which shown that calpain-3 seems to play also a role in the Ca 2+ ef-
commands myoblasts to stop proliferating.28 flux from the sarcoplasmic reticulum in a way that does not
Antithetic expression of miR-519 and miR-654 in G1 + G2 involve its proteolytic function.35 This evidence could ex-
versus G3 could represent a different way by which same plain how dysfunctional calpain 3 can be causative of muscle
target gene is regulated in LGMD2A respect to LGMD2B. weakness. On the other hand, the upregulation of proteins
MiR-200, that appeared highly expressed in primary that are involved in Ca2+ release, such as GNB1 and VDR,
CAPN3 deficit versus CAPN3 autocatalysis deficit, is in- may lead to prolonged exposure to high calcium concentra-
volved in the regulation of ECM-receptor interactions and tions with subsequent Ca2+-dependent cleavage and activa-
adherens junctions. Only one gene, fibronectin-1 (FN1), is tion of calpain 3, thus justifying its reduction in patients
reported in MirPath as targeted by miR-200 in controlling suffering from secondary CAPN3 deficits (e.g. LGMD2B).
ECM-receptor interactions. According to Vaz et al., fibro- After sarcolemmal damages that often occur during mus-
nectin removal is crucial for myoblast fusion.29 Overexpres- cle contraction functional dysferlin is accumulated at the site
sion of miR-200 in primary CAPN3 deficit, here observed, of the lesion in a calcium-dependent manner, and partici-
could be then explained as a compensatory way activated pates in patch-fusion repair; in the absence of dysferlin,

Copyright © 2016 John Wiley & Sons, Ltd. Cell Biochem Funct 2016; 34: 414–422.
 MICRORNA SIGNATURES PREDICT DYSREGULATED VITAMIN D RECEPTOR AND CALCIUM PATHWAYS STATUS 421

the membrane tear is not adequately repaired, and myofibres Furthermore, the identification of the same miRNAs in
undergo necrosis.36 other body fluids (e.g. blood or saliva) of LGMD patient
According to this, the upregulation in LGMD2B patients to validate them as ‘in vivo biomarkers’ is mandatory for ac-
of proteins participating in Ca2+ homeostasis might be celerating diagnosis, without the need for muscle biopsy.
strongly dependent on impaired membrane repair processes GC, LN participated in bioinformatic analysis and west-
because of the primary dysferlin deficiency. ern blot, and helped to draft the manuscript. RC, MA, VM
Other than participating in the repairing of the sarco- performed immunohistochemistry. MO, RC, TA, FM, AC,
lemma, different studies have evidenced the presence of DGRM obtained clinical and muscle biopsy data. AM,
dysferlin in the t-tubule membrane of mature skeletal muscle RM performed miRNA expression and data analysis. AM
fibres.37 Following experimental membrane stress in vitro, conceived of the study and its design, performed experiment
dysferlin-deficient muscle fibres undergo extensive func- coordination and drafted the manuscript.
tional and structural disruption of the t-tubules that can be
ameliorated by pharmacologically reducing external cal-
cium concentration or blocking L-type calcium channels. CONFLICT OF INTERESTS
The application of this treatment on dysferlin-deficient mice
significantly reduces eccentric contraction-induced t-tubule The authors declare that no competing interests exist.
damage, inflammation and necrosis.

REFERENCES
CONCLUSION 1. Nigro V, Savarese M. Genetic basis of limb-girdle muscular dystro-
In this study, we showed a difference in miRNA expression, phies: the 2014 update. Acta Myol 2014; 33: 1–12.
2. Huang Y, de Morree A, van Remoortere A, et al. Calpain 3 is a mod-
between LGMD2 A and LGMD2B, which represent the two ulator of the dysferlin protein complex in skeletal muscle. Hum Mol
most prevalent calpainopathies. The LGMD2A consists of Genet 2008; 17: 1855–1866.
two groups (G1 and G2), for they are considered primary, 3. Hauerslev S, Sveen M-L, Duno M, et al. Calpain 3 is important for
G1 characterized by the absence of calpain 3 and G2 with muscle regeneration: evidence from patients with limb girdle muscular
dystrophies. BMC Musculoskelet Disord 2012; 13: 43.
impaired function of calpain 3, while the group G3 4. Ermolova N, Kudryashova E, DiFranco M, et al. Pathogenity of some
(LGMD2B) is a deficiency secondary calpain 3 due to ab- limb girdle muscular dystrophy mutations can result from reduced an-
sence of dysferlin. chorage to myofibrils and altered stability of calpain 3. Hum Mol Genet
Our results show a panel of 187 (48.7%) miRNA dysreg- 2011; 20: 3331–3345.
ulated in LGMD2A/B versus controls. This suggest that 5. Garvey SM, Rajan C, Lerner AP, Frankel WN, Cox GA. The muscular
dystrophy with myositis (mdm) mouse mutation disrupts a skeletal
miRNA can have an important role in epigenetic control dis- muscle-specific domain of titin. Genomics 2002; 79: 146–149.
ease onset and, probably, also in the disease progression. 6. Kramerova I, Beckmann JS, Spencer MJ. Molecular and cellular basis
The difference in miRNA expression found between of calpainopathy (limb girdle muscular dystrophy type 2A). Biochim
LGMD2A (G1 and G2) versus LGMD2B is compatible with Biophys Acta 2007; 1772: 128–144.
7. Kramerova I, Kudryashova E, Wu B, et al. Novel role of calpain-3 in
the alteration metabolic pathways identified by bioinfor- the triad-associated protein complex regulating calcium release in skel-
matic approach, showing that some pathways are more etal muscle. Hum Mol Genet 2008; 17: 3271–3280.
expressed in LGMD2B (e.g. muscle inflammation that char- 8. Benayoun B, Baghdiguian S, Lajmanovich A, et al. NF-kappaB-de-
acterizes muscle biopsies of patient who lack dysferlin), pendent expression of the antiapoptotic factor c-FLIP is regulated by
while some pathways are common in both conditions but calpain 3, the protein involved in limb-girdle muscular dystrophy type
2A. Faseb J 2008; 22: 1521–1529.
with different degree of activation, such as intracellular reg- 9. Stuelsatz P, Pouzoulet F, Lamarre Y, et al. Down-regulation of MyoD
ulation of calcium. by calpain 3 promotes generation of reserve cells in C2C12 myoblasts.
However, we found a difference of expression within J Biol Chem 2010; 285: 12670–12683.
LGMD2A (G1 versus G2) that could account for a pheno- 10. Glover L, Brown RH, Jr. Dysferlin in membrane trafficking and patch
repair. Traffic 2007; 8: 785–794.
typic difference in patients with primary LGMD2A. 11. Klinge L, Harris J, Sewry C, et al. Dysferlin associates with the devel-
In addition in silico analysis, based on data obtained, pre- oping T-tubule system in rodent and human skeletal muscle. Muscle
dicted the involvement of two genes GNB1 and VDR that Nerve 2010; 41: 166–173.
participate in intracellular calcium management. 12. Eisenberg I, Eran A, Nishino I, et al. Distinctive patterns of microRNA
Our work, although needs further functional studies, expression in primary muscular disorders. Proc Natl Acad Sci U S A
2007; 104: 17016–17021.
shows and confirms involvement of intramyofibrillar cal- 13. Rosenberg MI, Georges SA, Asawachaicharn A, Analau E, Tapscott
cium metabolism in the aetiology of LGMD, and that could SJ. MyoD inhibits Fstl1 and Utrn expression by inducing transcription
open new potential therapeutic targets. of miR-206. J Cell Biol 2006; 175: 77–85.
14. Anderson C, Catoe H, Werner R. MIR-206 regulates connexin43 ex-
pression during skeletal muscle development. Nucleic Acids Res
2006; 34: 5863–5871.
FUTURE PERSPECTIVES 15. Chen JF, Mandel EM, Thomson JM, et al. The role of microRNA-1
and microRNA-133 in skeletal muscle proliferation and differentiation.
Further investigations will be performed (e.g. by using Nat Genet 2006; 38: 228–233.
siRNAs on cell cultures or dystrophic mice) to validate 16. Anderson LV, Davison K, Moss JA, et al. Characterization of mono-
in vivo the postulated effects of these miRNAs. clonal antibodies to calpain 3 and protein expression in muscle from

Copyright © 2016 John Wiley & Sons, Ltd. Cell Biochem Funct 2016; 34: 414–422.
422 m. aguennouz ET AL.

patients with limb-girdle muscular dystrophy type 2A. Am J Pathol 29. Vaz R, Martins GG, Thorsteinsdottir S, Rodrigues G. Fibronectin pro-
1998; 153: 1169–1179. motes migration, alignment and fusion in an in vitro myoblast cell
17. Fanin M, Nascimbeni AC, Fulizio L, et al. Loss of calpain-3 autocata- model. Cell Tissue Res 2012; 348: 569–578.
lytic activity in LGMD2A patients with normal protein expression. Am 30. Gerlach K, Daniel C, Lehr HA, et al. Transcription factor NFATc2
J Pathol 2003; 163: 1929–1936. controls the emergence of colon cancer associated with IL-6-
18. Livak KJ, Schmittgen TD. Analysis of relative gene expression data dependent colitis. Cancer Res 2012; 72: 4340–4350.
using real-time quantitative PCR and the 2( Delta Delta C(T)) 31. Li J, Kim K, Hahm E, et al. Neutrophil AKT2 regulates heterotypic
Method. Methods 2001; 25: 402–408. cell–cell interactions during vascular inflammation. J Clin Invest
19. Chen C, Ridzon DA, Broomer AJ, et al. Real-time quantification of 2014; 124: 1483–1496.
microRNAs by stem-loop RT-PCR. Nucleic Acids Res 2005; 33: e179. 32. Weiss N, Legrand C, Pouvreau S, et al. In vivo expression of G-protein
20. Vlachos IS, Kostoulas N, Vergoulis T, et al. DIANA miRPath v.2.0: beta1gamma2 dimer in adult mouse skeletal muscle alters L-type cal-
investigating the combinatorial effect of microRNAs in pathways. cium current and excitation–contraction coupling. J Physiol 2010;
Nucleic Acids Res 2012; 40: W498–504. 588: 2945–2960.
21. Warde-Farley D, Donaldson SL, Comes O, et al. The GeneMANIA pre- 33. Boland RL. VDR activation of intracellular signaling pathways in skel-
diction server: biological network integration for gene prioritization and etal muscle. Mol Cell Endocrinol 2011; 347: 11–16.
predicting gene function. Nucleic Acids Res 2010; 38: W214–W220. 34. a)a Goldstein JA, McNally EM. Mechanisms of muscle weakness in
22. Bader GD, Hogue CW. An automated method for finding molecular com- muscular dystrophy. J Gen Physiol 2010; 136: 29–34; b)b Vallejo-
plexes in large protein interaction networks. BMC Bioinformatics 2003; 4: 2. Illarramendi A, et al. Dysregulation of calcium homeostasis in muscu-
23. Mu W, Zhang W. Bioinformatic resources of microRNA sequences, lar dystrophies. Expert Rev Mol Med 2014; 16: e16.
gene targets, and genetic variation. Front Genet 2012; 3: 31. 35. Ojima K, Ono Y, Ottenheijm C, et al. Non-proteolytic functions of
24. Lu X, Meng X, Morris CA, Keating MT. A novel human gene, WSTF, calpain-3 in sarcoplasmic reticulum in skeletal muscles. J Mol Biol
is deleted in Williams syndrome. Genomics 1998; 54: 241–249. 2011; 407: 439–449.
25. Zacharewicz E, Lamon S, Russell AP. MicroRNAs in skeletal muscle 36. Bansal D, Miyake K, Vogel SS, et al. Defective membrane repair in
and their regulation with exercise, ageing, and disease. Front Physiol dysferlin-deficient muscular dystrophy. Nature 2003; 423: 168–172.
2013; 4: 266. 37. Kerr JP, Ziman AP, Mueller AL, et al. Dysferlin stabilizes stress-
26. Mraz M, Dolezalova D, Plevova K, et al. MicroRNA-650 expression is induced Ca2+ signaling in the transverse tubule membrane. Proc Natl
influenced by immunoglobulin gene rearrangement and affects the bi- Acad Sci U S A 2013; 110: 20831–20836.0.
ology of chronic lymphocytic leukemia. Blood 2012; 119: 2110–2113.
27. Elfimova N, Sievers E, Eischeid H, et al. Control of mitogenic and
motogenic pathways by miR-198, diminishing hepatoma cell growth
and migration. Biochim Biophys Acta 2013; 1833: 1190–1198.
28. Davidovic L, Durand N, Khalfallah O, et al. A novel role for the SUPPORTING INFORMATION
RNA-binding protein FXR1P in myoblasts cell-cycle progression by
modulating p21/Cdkn1a/Cip1/Waf1 mRNA stability. PLoS Genet Additional supporting information may be found in the on-
2013; 9: e1003367. line version of this article at the publisher’s web site.

Copyright © 2016 John Wiley & Sons, Ltd. Cell Biochem Funct 2016; 34: 414–422.