Collagen XXV promotes myoblast fusion during myogenic differentiation and muscle formation
Collagen XXV promotes myoblast fusion during myogenic differentiation and muscle formation
t ristan J. M. Gon?alves
t akeshi Iwatsubo
OPEN Fusion of myoblasts into multinucleated myofibers is crucial for skeletal muscle development and regeneration. However, the mechanisms controlling this process remain to be determined. Here we identified the involvement of a new extracellular matrix protein in myoblast fusion. Collagen XXV is a transmembrane-type collagen highly transcribed during early myogenesis when primary myofibers form. Limb muscles of E12.5 and E14.5 Col25a1?/? embryos show a clear defect in the formation of multinucleated myofibers. In cell culture, the cleaved soluble extracellular domain of the collagen XXV is sufficient to promote the formation of highly multinucleated myofibers. Col25a1 is transiently expressed during myogenic differentiation and Col25a1 transcripts are down-regulated in multinucleated myofibers by a muscle-specific microRNA, miR-499. Altogether, these findings indicate that collagen XXV is required in vivo and in vitro for the fusion of myoblasts into myofibers and give further evidence that microRNAs participate to the regulation of this process.
Skeletal muscle development and growth is a highly regulated process controlled by interactions between muscle
cells and their surrounding microenvironment. Skeletal muscle formation can be divided into several sequential
events, as migration of muscle precursor cells (mpc), proliferation of myoblasts, cell cycle arrest, myoblast
differentiation and fusion. These different steps are characterized by the expression of muscle specific transcription
factors, the muscle regulatory factors (MRFs), Myf5, MyoD, myogenin and MRF4. Myf5 and MyoD promote
expansion of mpc, whereas myoblast differentiation is driven by myogenin expression. During mouse
development, trunk and limb muscles arise from two waves of myogenesis involving different populations of myoblasts1.
Between E11 and E14.5, embryonic (primary) myoblasts align and fuse to form small multinucleated primary
myofibers. From E14.5 to birth, fetal (secondary) myoblasts enter the myogenic program, adhere to the primary
fibers, use them as a scaffold and fuse together to generate secondary myotubes1,2.
The different steps of myogenesis are linked to remodeling of extracellular matrix (ECM) proteins as well as
by changes in the expression pattern of several cell surface receptors, such as integrins3. Mpcs are able to regulate
their extracellular environment in adult myogenesis, by secreting microRNAs targeting synthesis of collagen by
surrounding fibroblasts4. Collagens are major ECM proteins involved in different functions, from the formation
of fibrillar networks of the ECM to cell adhesion and have been linked to muscular dystrophies5,6. Among the
collagen family, type XXV collagen is highly expressed in developing limb skeletal muscles and in
differentiating C2C12 myotubes, suggesting a role in the myogenic differentiation7. This collagen belongs to the MACIT
(membrane-associated collagens with interrupted triple helices) subset of the collagen superfamily, together with
type XIII, XVII and XXIII collagens8. This subfamily of collagen is suspected to be involved in cell adhesion.
Collagen XVII, also called BP180, is a well-characterized structural component of hemidesmosomes, anchoring
epidermal keratinocytes to the underlying basement membranes9. The other MACIT members are structurally
different from collagen XVII, each with four noncollagenous and three collagenous domains, a short intracellular
region and a furin-cleavage site10,11. Collagen XXIII is mainly expressed in epidermis and epithelia12 and likely
mediates metastasis by facilitating cell-cell and cell-matrix adhesion13. Type XIII collagen probably functions
as an adhesion molecule, since it has been localized at cell-cell and cell-matrix junction sites14. It has also been
shown to be part of the neuromuscular junctions (NMJ) and essential for the correct assembly of various
components of NMJ during the post-natal period in mice15,16.
Collagen XXV was first described as a component of senile plaques in Alzheimer?s disease brain10. The
ectodomain fragment obtained after cleavage by furin convertase was called CLAC (collagenous Alzheimer amyloid
plaque component) and transmembrane collagen XXV as the precursor of CLAC (CLAC-P). Col25a1 knock out
mice die at birth due to respiratory defects. Motoneurons fail to branch the diaphragm muscle and subsequently
degenerate, indicating that this transmembrane collagen is probably required for proper muscle innervation
during fetal mouse development7. However, Col25a1 transcripts were detected in differentiating limb myofibers
from early E11.5, long before muscle innervation, suggesting that this collagen could also be involved in early
In this study, we report the involvement of collagen XXV in early myogenesis. Col25a1?/? mice exhibit a
delay in the fusion and organization of myofibers during primary myogenesis. In culture, Col25a1 transcripts
are transiently expressed during myogenic differentiation. The cleaved soluble ectodomain of collagen XXV is
sufficient to facilitate fusion of myoblasts into myotubes, confirming the role of collagen XXV in myoblast fusion.
As myotubes mature, a muscle specific microRNA, miR-499, represses Col25a1 expression, by targeting its
transcripts. Collectively, our data reveal a new regulator of fusion of myoblasts into myotubes, collagen XXV, and
highlight the involvement of muscle specific microRNAs in this process.
Collagen XXV is required for muscle formation in vivo. Previous studies indicated that Col25a1 is
highly transcribed in developing skeletal muscle cells in limbs from E11.5 mouse embryos and not detected in
adult skeletal muscles7,10. To examine the early in vivo function of collagen XXV, we analyzed the muscle
phenotype of limb muscles from Col25a1 knock out mice at E12.5 and E14.5 (Fig.?1). Histological analysis of transverse
cross-sections of hindlimb muscles revealed the presence of myosin positive muscle cells in E14.5 Col25a1?/?
mouse embryos (Fig.?1A). Their number is similar in Col25a1?/? and wild-type (WT) E14.5 embryos (Fig.?1B).
However, muscle fibers of Col25a1?/? embryos are larger than controls and irregular in shape, whereas WT
myofibers are linear (Fig.?1C). At this embryonic stage, the proportion of myofibers with a cross-sectional area
above 80 ? m2 is much higher in the Col25a1?/? forelimb muscles than in the same muscle masses in the
control embryos. While multinucleated myofibers were present in forelimbs from WT E14.5 embryos, many of the
myosin positive longitudinal myofibers observed in the Col25a1 deficient forelimbs were much shorter, with few
myonuclei (Fig.?1D). Some of them have bulges, which could explain their larger diameter compared to controls
(Fig.?1C,D). Analysis of forelimbs from E12.5 WT and Col25a1?/? embryos strongly suggests that collagen XXV
is required for the formation of multinucleated myofibers during primary myogenesis (Fig.?1E). Multinucleated
myofibers with sarcomeres were detected in the control E12.5 embryos, whereas most of the Col25a1?/?
myofibers seem to properly align but were much shorter. This suggests that collagen XXV expression is important for
myoblasts to properly fuse, although we did detect multinucleated myofibers at later embryonic stages. As WT
myofibers, Col25a1?/? fibers had sarcomeres and expressed embryonic MyHC. The diameter of these poorly
nucleated myofibers seem to be larger, as already observed in forelimbs from E14.5 embryos (Fig.?1C?E). These in
vivo data reveal that collagen XXV is required for muscle formation during primary myogenesis.
Col25a1 is transiently expressed during myogenic differentiation. To better characterize the role of
collagen XXV during primary myogenesis, as well as in myogenic differentiation, we analyzed the precise timing
of Col25a1 expression during mouse development and in cell culture. By RT-qPCR, we confirmed that Col25a1
transcripts were abundant in limbs from WT E12.5 embryos and decreased when skeletal muscles mature, as
measured by the abundance of Myh7 transcripts, coding for the adult slow isoform of myosin heavy chain (type
I MyHC) (Fig.?2A)7. Next, we analyzed the precise timing of Col25a1 induction upon myogenic differentiation
(Fig.?2B). Primary myoblasts were isolated from P10 neonatal WT mice and induced to differentiate. RT-qPCR
analyses revealed a massive and transient induction of Col25a1 transcripts at two days of differentiation, a time
point which corresponds with the peak of expression of the differentiation marker, myogenin, and the initiation
of transcription of the embryonic isoform of MyHC, encoded by the Myh3 gene, specifically expressed in young
myotubes. These results should be confirmed at the protein levels. Unfortunately, we were not able to detect the
endogenous protein, since all antibodies tested were not specific enough. Altogether, our transcriptional
expression results of col25a1 during myogenic differentiation strongly suggest that collagen XXV is involved in the
formation of myotubes.
Collagen XXV promotes fusion of myoblasts. Once synthesized, collagen XXV is targeted at the plasma
membrane where the extracellular C-terminal domain could be cleaved by furin10. To test the localization of collagen
XXV in muscle cells, we transfected C2C12 myoblasts with expression vectors encoding either GFP alone (GFP) or
collagen XXV fused to GFP at its C-terminal end (Collagen XXV-GFP) (Fig.?3A). Collagen XXV-GFP was detected
on the surface of myoblasts, by staining live cells with an anti-GFP antibody, a method used to detect plasma
membrane proteins17 (Fig.?3Aa,Aa?). We observed that collagen XXV-GFP was concentrated on a restricted region of
the cell surface in some myoblasts (Fig.?3Ab,Ab?). To further characterize the role of collagen XXV during muscle
differentiation, 10T1/2 fibroblasts were transfected with expression vectors encoding either GFP alone or collagen
XXV-GFP (Fig.?3B). 48 h after transfection, the collagen XXV-GFP fusion protein was barely detectable in the
cellular extracts, whereas the control GFP was highly expressed in the control transfected cells (Fig.?3C). These results
suggest that the ectodomain of collagen XXV fused to GFP might be cleaved by furin and liberated from the
membrane as a secreted form in transfected 10T1/2 cells, as detected in the culture media of Col25a1-GFP transfected
cells (Fig.?3C). To determine whether the ectodomain of collagen XXV is able to regulate myogenic differentiation,
differentiating C2C12 myoblasts were cultured four days with the culture medium of GFP or Col25a1-GFP
transfected 10T1/2 fibroblasts (Fig.?3B,D). Fusion index was significantly higher when C2C12 were differentiating in the
presence of the ectodomain of collagen XXV in the medium (Fig.?3E). We observed a significant increase in large
myotubes with more than 20 nuclei, as cells were cultured with the culture medium of Col25a1-GFP transfected
10T1/2 cells compared to the myotubes monitored in control conditions (Fig.?3D,F). To test whether the ectodomain
of collagen XXV activates the differentiation program of C2C12 cells, we analyzed by RT-qPCR the expression of
MyoD, Myf5 and myogenin in C2C12 cells cultured 1 or 2 days in the conditioned media from 10T1/2 cells
previously transfected either with GFP or Col25a1-GFP (Fig.?3G). Although the amount of myod transcripts was similar
in each condition tested, the expression of the pro-differentiation myogenic factor myogenin was down-regulated
in C2C12 cells cultured 2 days in the conditioned medium of Col25a1-GFP transfected 10T1/2 cells, compared to
C2C12 cells cultured 2 days in classic differentiation medium or in conditioned media from GFP transfected 10T1/2
cells. This result suggests that the presence of the ectodomain of collagen XXV in the culture medium likely inhibits
the differentiation program of C2C12 cells. In contrast to myogenin, Myf5 is highly expressed in proliferating
myoblasts and its expression decreases as myoblasts differentiate18. We observed a significant increase of myf5 transcript
levels in C2C12 cells cultured 1 or 2 days in the presence of collagen XXV-GFP in the differentiation medium
compared to control media or conditioned media from GFP transfected 10T1/2 cells (Fig.?3G). Altogether, these results
indicate that conditioned media from col25a1-GFP transfected 10T1/2 cells promote fusion of C2C12 cells, without
the activation of differentiation.
As conditioned media from 10T1/2 cells might contain many factors which could interfere with differentiation
and fusion of C2C12 cells, we next tried to confirm the pro-fusion effect of the collagen XXV-GFP on
differentiating C2C12 cells, by minimizing the action of these factors. To do so, we used HEK293T cells to get the highest
rate as possible of collagen XXV-GFP proteins, after 48 h transfection. We then enriched collagen XXV-GFP
from conditioned media of transfected HEK293T cells by ion exchange chromatography (Fig.?4A, Sup. Fig.?1).
We identified and pooled the fractions the most enriched in collagen XXV-GFP. As negative control, we pooled
the corresponding fractions eluted from the conditioned media of GFP transfected HEK293T cells (Fig.?4A, Sup.
Fig.?1). We observed a significant increase in the number of myotubes containing more than 20 nuclei when
C2C12 cells were treated with collagen XXV-GFP enriched fractions compared to myotubes monitored in
cultures treated with the control fractions or with recombinant GFP (rGFP) (Fig.?4B?E). These results strengthen
those obtained with differentiating C2C12 cultured in conditioned media from the different transfected 10T1/2
cells, indicating that the soluble collagen XXV ectodomain is sufficient to promote the fusion of myoblasts into
myotubes. Combined with our in vivo data, these results demonstrate that collagen XXV is required for muscle
formation during mouse development, in regulating fusion of myoblasts into multinucleated myofibers.
Post-transcriptional regulation of Col25a1. Gene expression is regulated at the transcriptional level
but also post-transcriptionally by classes of non-coding RNAs19. As Col25a1 expression is transient in vitro and
in vivo (Fig.?2), we screened for microRNA binding sites within the 3?UTR of the Col25a1 gene using the miRDB
program (http://mirdb.org/). We identified miR-208a (mmu-miR-208a-3p) and miR-208b (mmu-miR-208b-3p)
as putative candidates to repress Col25a1 in young myotubes, on a seed site highly conserved in mammals
(Fig.?5A). miR-208a and miR-208b are both intronic and encoded by two distinct slow myosin genes, Myh6 and
Myh7 respectively20. This family of miRNAs has a third member, miR-499 (mmu-miR-499-5p), with a sequence
highly related to those of miR-208a and miR-208b (Fig.?5A). miR-499 is also intronic, since it is encoded by intron
19 of the slow Myh7b gene. As Myh6/miR-208a is exclusively expressed in the heart21, we focused our analyses
on Myh7/miR-208b and Myh7b/miR-499, even if miR-499 was not predicted to target the 3?UTR of the Col25a1
transcripts (Fig.?5A). To test whether these miRNAs could repress Col25a1 expression, we analyzed by RT-qPCR
the expression of miR-208b and miR-499 and their host genes, during mouse embryogenesis (Figs?2A, 5B) and
myogenic differentiation (Fig.?5C). The sequences of miR-208a and miR-208b are so close to each other that it is
impossible to discriminate them by RT-qPCR. However, we assumed that the miR-208 transcripts detected here
were miR-208b transcripts, since miR-208a is specifically expressed in the heart. During embryonic myogenesis,
miR-208b levels were relatively similar in limb muscles (Fig.?5B), despite a net increase of the transcripts of its
host gene (Fig.?2A). In contrast to miR-208b, miR-499 transcript levels exhibited a reverse expression profile
to that of Col25a1, like its host gene, Myh7b (Figs?2A, 5B). miR-499 expression was indeed about eight fold
higher in limb muscles of E18.5 WT embryos compared to limbs from E12.5 or E14.5 WT embryos, suggesting
that miR-499 is a better candidate than miR-208b to repress Col25a1 expression during muscle development.
Next, we analyzed the expression profile of miR-208b and miR-499 during differentiation of primary muscle cells
(Fig.?5C). Both miRNA expression was low 2 days after differentiation, time point when Col25a1 transcripts were
the most abundant. As Col25a1 transcripts decreased (from day 3 of differentiation), miR-208 and miR-499 levels
increased, suggesting that these two miRNAs could target Col25a1 transcripts during myogenic differentiation
To more directly establish the regulation of Col25a1 by miR-208b and miR-499, we manipulated miR-208b and
miR-499 in cultured differentiating myoblasts. Transient transfections of synthetic miR-208b, miR-499 or both
decreased significantly endogenous Col25a1 transcripts of about 30%, in differentiating primary myoblasts after 48h
transfection and differentiation (Fig.?5D). The efficiency to repress Col25a1 transcripts was similar for each microRNA
and co-transfection of both miR-208b and miR-499 did not induce further decrease in Col25a1 mRNA abundance,
suggesting a competition between these microRNAs on the 3?UTR of Col25a1. In parallel, we noticed that increased levels
of miR-499 specifically increased the amount of Myh7 transcripts, while Myh7b expression seemed to be repressed by
miR-208b (Fig.?5D), suggesting a cross-regulation of these microRNAs family members with their host genes. Next, we
tested the putative miR-208b/miR-499 binding site on the 3?UTR of the Col25a1 gene by luciferase assays. 10T1/2
fibroblasts were transiently co-transfected with miR-208b, miR-499 or both, in the presence of a miScript Target Protector
specific for the miR-208b/miR-499 binding site on the 3?UTR of the Col25a1 gene or its related negative control and
a luciferase reporter fused to the 3?UTR of the Col25a1 gene. Transient transfections of synthetic miR-499 alone or in
combination with miR-208b significantly decreased the Col25a1 3?UTR reporter activity, and this repression was
abolished in the presence of the miScript target protector specific to the miR-208b/miR-499 binding site of the 3?UTR of the
col25a1 gene (Fig.?5E). In contrast to miR-499, transfection of synthetic miR-208b alone did not significantly modify
the luciferase activity of the reporter, suggesting that miR-499 is probably the main regulator of col25a1 transcripts
(Fig.?5E). These repressions were completely abrogated when miR-208b and/or miR-499 were co-transfected with a
luciferase reporter construct harboring a mutated miR-208b/miR-499 binding site in the Col25a1 3?UTR. These results
provide evidence for a consistent post-transcriptional regulation of Col25a1 transcripts mainly by miR-499, during
muscle differentiation and development.
Implication of the extracellular matrix in primary myogenesis. The ECM has long been known to be
important for muscle integrity and function. In myogenesis, cell-ECM interactions contribute to the formation of
myotubes in the direction and organization of muscles to be. In the adult, myofibers are surrounded by an ECM
material, the basement membrane. The inner layer of the basement membrane of degenerated fibers, the basal lamina, has
recently been shown to play a critical role in muscle regeneration, since it serves as a scaffold directing myogenic
progenitors alignment for division, migration and therefore fusion into regenerating myotubes22. The basement membrane
appears only 2?3 days before birth in mice. Building the muscle tissue in the embryo also need a scaffold. During fetal
development, secondary myoblasts adhere to primary myotubes and use them as a scaffold for their alignment and
fusion into fetal myofibers23. So, primary myofibers serve as a scaffold for the remaining stages of myogenesis and set
the orientation of all muscles. In contrast to secondary myofibers, primary myoblasts fuse in an environment composed
of an interstitial ECM. Cells interact with the ECM via a variety of receptors, the integrins being the most common.
Several studies showed that ?1 integrins are essential for primary myoblasts to interact with their surrounding ECM and
properly fuse during embryonic myogenesis24?26. However, it is not clear which ligands in the ECM could be involved
in these cell-ECM interactions. Here, we provide evidence that collagen XXV is part of the ECM involved in the proper
set up of primary myofibers, since the genetic ablation of collagen XXV leads to the formation of short and misaligned
myofibers in E12.5 and E14.5 limbs (Fig.?1). Indeed, collagen XXV expression is high during primary myogenesis and
reduces as secondary myogenesis occurs (Fig.?2), suggesting an important role for collagen XXV in the early stages of
the construction of the muscle tissue. Whether this collagen is able to bind ?1 integrins during primary myogenesis
remains to be determined. As the mouse model used in our study is a whole body knock out model, we cannot exclude
that non muscle cells might express collagen XXV within the embryonic limb. However, our in vitro data clearly show
that primary myoblasts express Col25a1 transcripts as they proliferate and differentiate (Fig.?2). Furthermore, Col25a1
transcripts were not detected in fibroblast-like cells isolated from limb skeletal muscles of newborn mice (data not
shown). Therefore, we hypothesize that primary myoblasts might be the main source of collagen XXV within the limb
bud. Overall, our data combined with studies from the literature point out the crucial role of ECM in the establishment
of primary myotubes, which will be used as the scaffold for all subsequent stages of myogenesis.
Role of collagen XXV in the fusion of myoblasts. Skeletal muscle development requires migration of
muscle progenitor cells, their differentiation and fusion into myotubes before maturating. Fusion of muscle cells is
probably the most poorly understood process in this mechanism, particularly in mammals. Some transmembrane
molecules involved in the myoblast fusion process have recently been identified. Myomaker, a muscle specific
transmembrane protein, is necessary for myoblast fusion during skeletal muscle development and regeneration27. To
stimulate its fusogenic activity, it interacts with the transmembrane micropeptide Myomixer (also named Myomerger or
Minion)28. This micropeptide also promotes myoblast fusion and skeletal muscle formation during development28?30.
How this small peptide induces membrane fusion is not understood yet. Before fusion, the cells might recognize each
other and adhere31. Moreover, during myotube formation, the negatively charged phospholipid, phosphatidylserine
(PS), transiently translocates from the inner layer of the plasma membrane to the outer leaflet32. Recognition of cell
surface-exposed PS by PS receptors, such as stabilin-2, also promotes fusion with neighboring myoblasts and skeletal
muscle regeneration33. In addition, the externalization of PS controls the orientation of the fusion, thereby preventing
random fusion and leading to the proper elongation of myotubes34. In this context, we identified a completely new
regulator of fusion of myoblasts. In absence of collagen XXV, myoblast fusion is impaired in vivo at E12.5 and E14.5 (Fig.?1),
whereas primary cells collected from E18.5 col25a1?/? mice could spontaneously fuse into myotubes as efficiently as
WT ones (data not shown). Altogether, these results suggest that this collagen is rather involved in the fusion of
myoblasts into myotubes during primary myogenesis, its implication during secondary myogenesis being not essential. In
culture, increased levels of the soluble ectodomain of collagen XXV promotes fusion of myoblasts into multinucleated
myotubes, in a manner independent from the differentiation process, since the differentiation marker myogenin was
not increased by the soluble collagen XXV in the culture medium (Figs?3 and 4). Altogether, our data further point out
the role of membrane-associated molecules in myoblast fusion, mainly during primary myogenesis. As collagen XXV is
initially a transmembrane protein before being cleaved by furin (Fig.?3A), we tested whether overexpression of collagen
XXV would promote fusion of 10T1/2 cells with C2C12 cells (data not shown). We did not observe any sign of
heterologous fusion, indicating that collagen XXV is probably not a fusogenic protein, but rather involved in the recognition
or adherence of fusion competent myoblasts. The precise mechanism by which collagen XXV is involved in the fusion
process remains to be determined.
miR-499, the first microRNA involved in myogenic fusion. In this study, we also described a new type
of regulators of myogenic fusion. Indeed, by regulating the expression of Col25a1, miR-499 also participates to the
fusion process. In contrast to miR-208b, miR499 expression is significantly increasing during embryonic myogenesis,
as Col25a1 transcripts decrease (Fig.?5B). Although it was the only one from this family not to be identified in the
miRDB database as targeting mouse Col25a1 expression, miR-499 is probably the microRNA most implicated in the
Col25a1 repression during embryonic myogenesis (our results). Furthermore, miR-499 is identified in the database as
a putative regulator of human Col25a1 transcripts. Interestingly, the expression of miR-499 and its host gene, Myh7b,
is under the control of miR-208a in the heart20. This is not the case with miR-208b and Myh7b/miR499 in the skeletal
muscle tissue, since overexpression of miR-208b in differentiating primary myoblasts decreased the levels of Myh7b
transcripts, whereas miR-499 levels were not reduced (Fig.?5D). According to the miRDB database, the murine Myh7b
gene would be targeted at its 3? UTR by only three microRNAs on seed regions completely different from the
miR208b one. Therefore, our results rather suggest that one of the targets of miR-208b could be the Myh7b pre-mRNA
itself. One can hypothesize that miR-208b targets the miR-499 sequence in intron 19 to regulate transcription and/or
splicing of Myh7b, inducing downregulation of its expression. Indeed, it has been described that a specific alternative
splicing of Myh7 transcripts blocks Myh7 translation and activates Myh7b mRNA degradation35. Moreover, our results
indicate that miR-499 modulates Myh7 expression. Since overexpression of miR-499 induces Myh7 overexpression
and not a downregulation, miR-499 could stabilize Myh7 transcripts (Fig.?5D). In total, these results point out an
original cross-regulation of miR-499 and miR-208b on the expression of their host genes, Myh7 and Myh7b respectively.
Similar to these specific functions on Myh7 and Myh7b expression, these microRNAs seem to have distinct mRNA
targets, Col25a1 transcripts being a favorite target for miR-499, even if miR-208b overexpression could also repress
Col25a1 expression in cultured differentiating primary myoblasts (Fig.?5D,E). Our results suggest that miR-499
transiently participates in vitro and in vivo to myogenic differentiation by controlling myoblasts fusion as myotubes mature
and activates the expression of Myh7 and Myh7b. Thereby, this microRNA would control the amount of nuclei within
each myotube by repressing expression of collagen XXV. Whether this microRNA could control expression of other
transcripts involved in myoblast fusion remains to be determined.
Fusion of myoblasts during embryonic development, as well as in adulthood, appears to be a critical process,
which must be highly and tightly controlled by several mechanisms. Here, we show for the first time the
implication of a collagen in this process. This transmembrane collagen is transiently expressed and tightly regulated by a
muscle specific microRNA, giving the first evidence of the involvement of microRNAs in this cellular mechanism.
All methods were carried out in accordance with approved biosafety and radiation guidelines for Universit? Paris
Descartes and adhered to European and French Guidelines. All animal experiments were performed following
protocols approved by the local animal protection committee (Comit? d?Ethique en mati?re d?Exp?rimentation Animale
Paris Descartes - CEEA 34). All animal procedures adhered to the guidelines of the French Ministry of Education,
Research and Innovation, in accordance with the European Communities for Experimental animal use (2010/63/EU).
Animals. Institutional animal care and use committee at Universit? Paris Descartes, [Laboratoire
d?Exp?rimentation animale et transgen?se (LEAT) SFR Necker Broussais (A 75-14-08)], has approved protocols in accordance
with the national authority (Minist?re de l?Enseignement Sup?rieur, de la Recherche et de l?Innovation, France)
guidelines based on European Union Directive 2010/63/EU. Col25a1 null mice were described previously7.
Cell culture. Limb muscles from P10-P14 wild-type mice were dissected and subjected to a gentle collagenase
(Sigma) and dispase (Gibco) digestion. The cell suspension was filtered (40? m pore size) and incubated with the
following antibodies: anti-CD45 (1/60, Clinisciences), anti-TER119 (1/60, Becton-dikinson), anti-Sca1 (1/25,
Becton-dikinson), anti-CD34 (1/40, Becton-dikinson), anti-?7-integrin (1/30, R&D). CD34/?7-integrin double
positive cells were positively selected using a FACSAria (Becton-dikinson), while cells positive for CD31, CD45 and Sca1
were negatively selected. Sorted muscle progenitor cells were plated on gelatin-coated dishes and cultured in DMEM
(Dulbecco?s modified Eagle?s medium)/F-12 (Gibco) containing 20% FBS (Gibco), 1% Penicillin?Streptomycin (Gibco)
and 2% Ultroser (Pall). Myogenic differentiation was induced by shifting the proliferation medium to differentiation
medium (DM): DMEM/F-12(Gibco), containing 2% horse serum (Gibco) and 1% Penicillin?Streptomycin.
C2C12, 10T1/2 and HEK293T cells were cultured in high glucose DMEM (Gibco) supplemented with 20%
(for C2C12 cells) or 10% (for 10T1/2 and HEK293T cells) fetal bovine serum (FBS) (Gibco) and 1% penicillin/
streptomycin (Gibco) at 37 ?C under 5% CO2.
Cloning. The Col25a1 expressing construct was generated from a mouse cDNA library. The Col25a1 sequence
was cloned in frame into the pCI neo-EGFP reporter vector, using the following primer sequences:
Forward 5?- ATAGGCTAGCCTCGAATGTTGGTGAAGAAGCTTGC-3?;
Reverse 5?- TGGTGGCGACCGGTGGATCCCGGGACTTTTGCCAGCAGCC-3?.
The luc-Col25a1 vector contains 4kb of the 3?UTR sequence of the Col25a1 transcript, cloned into the pGl
4.23 reporter vector at the Xho I site. The putative miR-499/miR208b target site was mutated in the 3?UTR of the
Col25a1 gene by directed mutagenesis, using the following primers:
Forward 5?- CTCAATATACGAAACAATTCGAATTCAATCATTTGGGGGAC-3?, Reverse 5? GTCCCCCAAATGATTGAATTCGAATTGTTTCGTATATTGAG-3?.
The integrity of all constructs was confirmed by sequencing.
Cell transfection and luciferase reporter assays. Primary myoblasts were transfected 48 h with 10 nM of
miRIDIAN miRNA mimics for mmu-miR-208b-3p or mmu-miR-499-5p or negative control (miR-Ctrl) (Dharmacon)
using Dharmafect 3 reagent (Dharmacon). For transient transfections, 10T1/2 cells were seeded 24h on 24 well plates
before transfections at 50 to 60% confluence. For luciferase reporter assays, 10T1/2 cells were co-transfected with 4 nM
of miRNA mimics or negative control (miR-Ctrl) (Qiagen), 4 nM miScript col25a1 target protector or negative
control (Qiagen) and 0.1 ? g/well of luc-Col25a1 or mutated luc-Col25a1 plasmids using Dharmafect Duo. 0.1 ? g/well of
pCMV-Renilla plasmid were included as an internal control in all transfection assays. Cells were harvested and lysed
48h after transfection and luciferase activity was analyzed using the Dual Luciferase Reporter Assay System (Promega).
Conditioned media and in vitro differentiation assay. 10T1/2 cells were transfected 6 h with 1 ? g
GFP or Col25a1-GFP expression vectors using 10 mM PEI (Polyethylenimine) reagent. The next day,
differentiation medium (DMEM high glucose, with 2% horse serum) was added to the cells (conditioned differentiation
medium). In parallel, C2C12 cells were seeded at 2500 cells/cm2 and cultured in proliferation medium (DMEM
high glucose, with 20% FBS). The day later, the conditioned differentiation medium from transfected 10T1/2
cells was added to C2C12 cells (differentiation condition day 0). Two days later, the conditioned differentiation
medium of C2C12 was replaced by new conditioned differentiation medium from freshly transfected 10T1/2 cells
(differentiation condition day 2). C2C12 cells were fixed at differentiation condition day 4.
Enrichment of Collagen XXV-GFP soluble fractions andin vitro differentiation assay. HEK293T cells
were transiently transfected 8 h with plasmids encoding the Collagen XXV-GFP fusion protein or GFP alone using
lipofectamine 2000 (Invitrogen). The next day, the culture medium was shifted to serum-free culture medium
containing 0.02% BSA. For both conditions, conditioned media were collected 24h later, adjusted to 50 mM Tris-HCl pH8
then loaded onto a 1 mL HiTrapQ anion-exchange column (GE Healthcare) equilibrated in the same buffer. Bound
proteins were eluted in 0.5mL fractions along a NaCl gradient (0?500mM). All fractions were analyzed by immunoblot
to identify those containing the Collagen XXV-GFP fusion protein. The 3 fractions containing the highest amounts
of Collagen XXV-GFP fusion protein were pooled and frozen until used for in vitro differentiation assays. The
corresponding fractions eluted from the GFP-alone conditioned media were processed identically and used as negative
control. For the differentiation assay, C2C12 cells were seeded at 2500 cells/cm2 and cultured in proliferation medium
(DMEM high glucose, with 20% FBS). The next day, C2C12 cell medium was switched to differentiation medium,
including 10 ? L/mL of pooled fractions enriched in Collagen XXV-GFP or the same amount of the corresponding
fractions from GFP-alone conditioned media, or 0.5 ng/mL of recombinant GFP (Novus biologicals). Half of the
differentiation medium was refreshed 2 days later, prior fixation 4 days after the beginning of the treatments.
RNA extraction and quantitative RT-PCR. Total RNA from tissues or cultured cells was extracted using
the TRIzol reagent (Invitrogen), followed by a RQ1 RNase-free DNase treatment (Promega). Reverse
transcription was realized from 1 ? g total RNA using the miScript II RT kit (Qiagen) according to the manufacturer?s
protocol. Quantitative real-time PCR were performed in triplicate using SYBR green Supermix (BioRad) with a
CFX? 384 Touch qPCR system (BioRad). 26S was used as a normalizer. Primers used are listed in Table?1.
Immunocytochemistry. Cultured cells were rinsed with PBS and fixed in 2% paraformaldehyde (PFA) for
10 min at room temperature. Cells were subsequently permeabilized with 0.2% Triton X?100 for 10 minutes and
Sequence (5?-3?) Forward primer
Sequence (5?-3?) Reverse primer
incubated overnight at 4 ?C with anti-MF20 (1/100, DSHB). Nuclei were stained with Hoechst (20 ? g/mL, Sigma).
To stain live cells, the latter were first washed with PBS and incubated 30 minutes on ice in anti-GFP antibody
(Abcam, ab5450) diluted 1/800 in 3% BSA/PBS, followed by 10 minutes fixation with 4% PFA/PBS, 10 minutes
permeabilization in 0.2% Triton/PBS and incubation with secondary antibody. Nuclei were then stained with
Hoechst. All these cultures were visualized on a Zeiss ApoTome2 fluorescent microscope.
Histological analysis and immunofluorescence. Trunk and forelimbs from E12.5 and E14.5 embryos
were dissected, embedded in optimal cutting temperature (O.C.T.) and frozen in cold isopentane. 10 ? m
transverse cryostat sections were realized and fixed with acetone. After 1h in PBS with 0.1 M glycine, sections were
incubated overnight at 4 ?C with the following antibodies: anti-MF20 (1/100, DSHB), anti-embryonic MyHC
(1/40, Santa Cruz sc-53091). Nuclei were stained with Hoechst (20 ?g/mL, Sigma). Sections were visualized on a
NanoZoomer 2.0-HT slide scanner (Hamamatsu).
Western blot analysis. Total proteins were extracted from cultured cells with protein lysis buffer (2% SDS,
50 mM Tris-HCl pH 6.8, 10% glycerol, 100 mM DTT). To analyze the presence of the fusion protein collagen
XXV-GFP in the cell culture medium, transfected 10T1/2 cells were cultured overnight in a serum free medium.
Cell culture media were then collected and concentrated using Amicon 10 K (Millipore). Bio-Rad protein assay
with BSA standard was performed to measure protein concentration of samples. To detect collagen XXV-GFP
proteins in conditioned media of transfected HEK293T cells, cells were cultured overnight in a serum free
medium containing 0.02% BSA and an aliquot of the conditioned media was used for Western blotting. 20 ? g
proteins were separated by sodium dodecyl sulfate polyacrylamide gel 10% electrophoresis. The separated proteins
were transferred to a PVDF membrane before immunoblotting with the following primary antibody: anti-GFP
(1/2000, Santa Cruz sc-9996). Anti-mouse-HRP (Cell signalling 7076 S) was used as a secondary detection
antibody. Finally, immunoreactive bands were detected using the ECL chemiluminescence detection.
statistical analysis. The fusion index was calculated as the percentage of nuclei in myosin positive myotubes
with more than 3 nuclei of total nuclei. The number of nuclei in myotubes was manually counted. The results are
presented as means ? s.e.m. Statistical analyses were performed using t-test or two-way ANOVAs followed by a
two-way repeated-measures analysis of variance with a Tukey?s multiple comparisons test (Figs?3G, 4E). Statistical
significance was accepted at a P value < 0.05.
We gratefully acknowledge Sophie Berissi (Structure F?d?rative de recherche [SFR] Necker Histology) for
histology support; Nicolas Goudin (SFR Necker Cell Imaging) for assistance with microscopy; Catherine Blanc
(CyPS platform from UPMC) and J?r?me M?gret (SFR Necker Cytometry) for technical assistance. We also
thank Mario Pende for his encouragement to pursue this research in his team (Institut Necker-Enfants Malades,
INSERM, U1151, team 5, Paris, France). Financial support was provided by the Institut National de la Sant? et
la Recherche M?dicale (INSERM), the ?Association Fran?aise contre les Myopathies? (AFM 18802), the Centre
National de la Recherche Scientifique (CNRS).
A.-S.A., F.B. and T.J.M.G. designed and performed experiments. A.-S.A., F.O., S.L., T.J.M.G. and V.G. analyzed
the data. A.-S.A., S.L. and T.W. conceived the project. T.I. and T.W. generated, supplied the Col25a1?/? mice and
shared data. A.-S.A. wrote the manuscript with critical reading from F.O., S.L., T.I., T.W., and V.G.
Supplementary information accompanies this paper at https://doi.org/10.1038/s41598-019-42296-6.
Competing Interests: The authors declare no competing interests.
Publisher?s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and
Open Access This article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the
Creative Commons license, and indicate if changes were made. The images or other third party material in this
article are included in the article?s Creative Commons license, unless indicated otherwise in a credit line to the
material. If material is not included in the article?s Creative Commons license and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the
copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
1. Biressi , S. , Molinaro , M. & Cossu , G. Cellular heterogeneity during vertebrate skeletal muscle development . Dev. Biol . 308 , 281 - 293 ( 2007 ).
2. Buckingham , M. Skeletal muscle formation in vertebrates . Curr. Opin. Genet. Dev . 11 , 440 - 448 ( 2001 ).
3. Cacha?o , A. S. , Pereira , C. S. , Pardal , R. G. , Bajanca , F. & Thorsteinsd?ttir , S. Integrin repertoire on myogenic cells changes during the course of primary myogenesis in the mouse . Dev. Dyn. Off. Publ. Am. Assoc. Anat . 232 , 1069 - 1078 ( 2005 ).
4. Fry , C. S. , Kirby , T. J. , Kosmac , K. , McCarthy , J. J. & Peterson , C. A. Myogenic Progenitor Cells Control Extracellular Matrix Production by Fibroblasts during Skeletal Muscle Hypertrophy . Cell Stem Cell 20 , 56 - 69 ( 2017 ).
5. Charvet , B. et al. Knockdown of col22a1 gene in zebrafish induces a muscular dystrophy by disruption of the myotendinous junction . Dev. Camb. Engl . 140 , 4602 - 4613 ( 2013 ).
6. Telfer , W. R. , Busta , A. S. , Bonnemann , C. G. , Feldman , E. L. & Dowling , J. J. Zebrafish models of collagen VI-related myopathies . Hum. Mol. Genet . 19 , 2433 - 2444 ( 2010 ).
7. Tanaka , T. et al. CLAC-P/Collagen Type XXV Is Required for the Intramuscular Innervation of Motoneurons during Neuromuscular Development . J. Neurosci . 34 , 1370 - 1379 ( 2014 ).
8. Pihlajaniemi , T. & Rehn , M. Two new collagen subgroups: membrane-associated collagens and types XV and XVII. Prog. Nucleic Acid Res . Mol. Biol . 50 , 225 - 262 ( 1995 ).
9. Powell , A. M. , Sakuma-Oyama , Y. , Oyama , N. & Black , M. M. Collagen XVII / BP180: a collagenous transmembrane protein and component of the dermoepidermal anchoring complex . Clin. Exp. Dermatol . 30 , 682 - 687 ( 2005 ).
10. Hashimoto , T. et al. CLAC: a novel Alzheimer amyloid plaque component derived from a transmembrane precursor, CLAC-P/ collagen type XXV . EMBO J . 21 , 1524 - 1534 ( 2002 ).
11. Voermans , N. C. et al. Clinical and molecular overlap between myopathies and inherited connective tissue diseases . Neuromuscul. Disord. NMD 18 , 843 - 856 ( 2008 ).
12. Koch , M. et al. Expression of type XXIII collagen mRNA and protein . J. Biol. Chem . 281 , 21546 - 21557 ( 2006 ).
13. Spivey , K. A . et al. A role for collagen XXIII in cancer cell adhesion, anchorage-independence and metastasis . Oncogene 31 , 2362 - 2372 ( 2012 ).
14. H?gg , P. et al. Type XIII collagen: a novel cell adhesion component present in a range of cell-matrix adhesions and in the intercalated discs between cardiac muscle cells . Matrix Biol. J. Int. Soc. Matrix Biol . 19 , 727 - 742 ( 2001 ).
15. H?r?nen , H. et al. Collagen XIII secures pre- and postsynaptic integrity of the neuromuscular synapse . Hum. Mol. Genet . 26 , 2076 - 2090 ( 2017 ).
16. Latvanlehto , A. et al. Muscle-derived collagen XIII regulates maturation of the skeletal neuromuscular junction . J. Neurosci. Off. J. Soc. Neurosci . 30 , 12230 - 12241 ( 2010 ).
17. Corcoran , J. A. & Duncan , R. Reptilian reovirus utilizes a small type III protein with an external myristylated amino terminus to mediate cell-cell fusion . J. Virol . 78 , 4342 - 4351 ( 2004 ).
18. Zammit , P. S. Function of the myogenic regulatory factors Myf5, MyoD, Myogenin and MRF4 in skeletal muscle, satellite cells and regenerative myogenesis . Semin. Cell Dev. Biol . 72 , 19 - 32 ( 2017 ).
19. Gon?alves , T. J. M. & Armand , A.-S. Non-coding RNAs in skeletal muscle regeneration . Non-Coding RNA Res ., https://doi. org/10.1016/j.ncrna. 2017 . 03 .003.
20. van Rooij , E. et al. A family of microRNAs encoded by myosin genes governs myosin expression and muscle performance . Dev. Cell 17 , 662 - 673 ( 2009 ).
21. Callis , T. E. et al. MicroRNA -208a is a regulator of cardiac hypertrophy and conduction in mice . J. Clin. Invest . 119 , 2772 - 2786 ( 2009 ).
22. Webster , M. T. , Manor , U. , Lippincott-Schwartz , J. & Fan , C.-M. Intravital Imaging Reveals Ghost Fibers as Architectural Units Guiding Myogenic Progenitors during Regeneration . Cell Stem Cell 18 , 243 - 252 ( 2016 ).
23. Tajbakhsh , S. Skeletal muscle stem cells in developmental versus regenerative myogenesis . J. Intern. Med . 266 , 372 - 389 ( 2009 ).
24. Baudoin , C. , Goumans , M. J. , Mummery , C. & Sonnenberg , A. Knockout and knockin of the beta1 exon D define distinct roles for integrin splice variants in heart function and embryonic development . Genes Dev . 12 , 1202 - 1216 ( 1998 ).
25. Cacha?o , A. S. et al. Knock-in of integrin beta 1D affects primary but not secondary myogenesis in mice . Dev. Camb. Engl . 130 , 1659 - 1671 ( 2003 ).
26. Schwander , M. et al. Beta1 integrins regulate myoblast fusion and sarcomere assembly . Dev. Cell 4 , 673 - 685 ( 2003 ).
27. Millay , D. P. et al. Myomaker: A membrane activator of myoblast fusion and muscle formation . Nature 499 , 301 - 305 ( 2013 ).
28. Bi , P. et al. Control of muscle formation by the fusogenic micropeptide myomixer . Science 356 , 323 ( 2017 ).
29. Quinn , M. E. et al. Myomerger induces fusion of non-fusogenic cells and is required for skeletal muscle development . Nat. Commun . 8 ( 2017 ).
30. Zhang , Q. et al. The microprotein Minion controls cell fusion and muscle formation . Nat. Commun . 8 , 15664 ( 2017 ).
31. Kim , J. H. , Jin , P. , Duan , R. & Chen , E. H. Mechanisms of myoblast fusion during muscle development . Curr. Opin. Genet. Dev . 32 , 162 ( 2015 ).
32. van den Eijnde, S. M. et al. Transient expression of phosphatidylserine at cell-cell contact areas is required for myotube formation . J. Cell Sci . 114 , 3631 - 3642 ( 2001 ).
33. Park , S.-Y. et al. Stabilin-2 modulates the efficiency of myoblast fusion during myogenic differentiation and muscle regeneration . Nat. Commun . 7 , 10871 ( 2016 ).
34. Tsuchiya , M. et al. Cell surface flip-flop of phosphatidylserine is critical for PIEZO1-mediated myotube formation . Nat. Commun. 9 , 2049 ( 2018 ).
35. Bell , M. L. , Buvoli , M. & Leinwand , L. A. Uncoupling of expression of an intronic microRNA and its myosin host gene by exon skipping . Mol. Cell. Biol . 30 , 1937 - 1945 ( 2010 ).