To study the heterogeneity of myonuclei in different contexts, we developed 2D culture and co-culture systems and an immobilisation model in chicken embryo leading to paralyzed limb muscles.

The fusion-associated genes do not display a central location in cultured myotubes

The fusion associated genes, TMEM8C and MYOG were expressed in the middle of muscles away from tendons labelled with SCX expression, in chicken limb muscles (Fig. 1A-D, Fig. S1A-D) [23]. To monitor cell lineage progression along the myogenic program, we cultured primary quail myoblasts isolated from E10 quail limbs. After five days of culture, quail myoblasts differentiated into myotubes (Fig. 1E,F). PAX7 + and MYOG + nuclei were present in the culture assessing the myogenic progression from PAX7 + muscle progenitors (47% of all nuclei) to MYOG + differentiated cells (25.4% of all nuclei) (Fig. 1E-G). In five day myoblast cultures, the fusion index was of 48.7% (Fig. 1G). The differentiation gene, MYOD and fusion associated genes, MYOG and TMEM8C were expressed in differentiating myoblasts and myonuclei of plurinucleated myotubes of five day myoblast cultures (Fig. 1H-M, Fig. S1E-J). In contrast to limb muscles (Fig. 1A,B, Fig. S1A,B), MYOG and TMEM8C transcripts did not display any central location in quail and chicken cultured myotubes (Fig. 1H-M, Fig. S1E-J). TMEM8C, MYOG transcripts did not show any obvious regionalization and were enriched in zones with high density of myonuclei: at contact points between myotubes and at myotube tips (Fig. 1H-M,K’,L’,M’). The percentage of TMEM8C-, MYOG- and MYOD-positive myonuclei (versus total myonuclei) were approximatively equivalent around 38 to 41% (Fig. S1K-M).

Fig. 1

Loss of regionalisation of fusion-associated genes in cultured myotubes. A-D Fluorescent in situ hybridization to adjacent transverse muscle sections from E10 chicken embryos with TMEM8C (A), MYOG (B), MYOD (C) and SCX (D) probes (red), followed by an immunolabelling of myosins (green) (D), combined with DAPI staining (nuclei, blue). White dashed lines correspond to the muscle/tendon interface. E-J Primary cultures of limb myoblasts from E10 quail embryos. E,F Myoblasts and myotubes labelled with PAX7 (muscle progenitors, red) and MF20 (myosins, green) antibodies (E), or with MYOG (red) and MF20 (myosins, green) antibodies (F), combined with DAPI staining (nuclei, blue). G Percentage of PAX7-positive nuclei versus total nuclei. Percentage of MYOG-positive nuclei versus total nuclei. Fusion index. Graphs show mean ± s.d. H-J Fluorescent in situ hybridization to quail myoblast cultures with TMEM8C (H), MYOG (I) and MYOD (J) probes (red) followed by an immunolabelling of myosins (green) combined with DAPI staining (nuclei, blue). K-M Fluorescent in situ hybridization to chicken myoblast cultures with TMEM8C (K), MYOG (L) and MYOD (M) probes (red) followed by an immunolabelling of myosins (green) combined with DAPI staining (nuclei, blue). (K’-M’) are high magnification of squared regions in (K-M)

We conclude that the fusion-associated genes do not display a central location in cultured myotubes.

The presence of fibroblasts favours all steps of the muscle program and does not change the TMEM8C expression pattern in cultured myotubes

To assess the effect of fibroblasts for muscle differentiation, we monitor cell lineage progression along the myogenic program differentiation in quail myoblast / chicken fibroblast co-cultures (Fig. 2). To follow quail myoblasts and chicken fibroblasts, we used the QCPN antibody that recognized quail nuclei and not chicken nuclei [18]. After 5 days of culture (initially plated with a ratio of 80% myoblasts and 20% fibroblasts), we observed an average ratio of 54% myoblasts and 46% fibroblasts in co-cultures (Fig. 2A); indicating that fibroblasts proliferate faster than myoblasts. 41.7% of PAX7 + cells and 16.5% of MYOG + nuclei versus all nuclei (including both quail and chicken nuclei) were observed in the co-cultures (Fig. 2B-E). Because PAX7 and QCPN were both monoclonal mouse IgG1 antibodies, we could not follow PAX7 + quail nuclei. However, because there are 54% quail nuclei in co-cultures, we could estimate that the percentage of PAX7 + quail nuclei to be 90% and that of MYOG + quail nuclei to be 36% versus quail nuclei (by excluding chicken nuclei). Thus, the presence of fibroblasts induced an increase in the proportion of PAX7 + cells (p = 6,3 10–12) and MYOG + nuclei (p = 0,024) (reported to quail nuclei) in co-cultures versus mono-cultures (47% for PAX7 + cells and 25,4% for MYOG + nuclei, Fig. 1G). The fusion index was slightly increased in co-cultures (57.5%, p = 1,05 10–5) versus myoblast mono-cultures (48.7%) (Fig. 2F versus Fig. 1G). We also observed an increase of myotube area (23,8%) in co-cultures compared to myoblast mono-cultures (12%) (Fig. 2G-I). The increase in the percentage of PAX7 + cells, MYOG + nuclei, fusion index (versus quail nuclei) and myotube area/unit area in co-cultures versus mono-cultures converge to the idea that the presence of fibroblasts favours the progression of all steps of the muscle program, consistent with previous studies [31]. Consistent with the increase of MYOG + nuclei, fusion index and myotube area/unit area, the fusion gene (Fig. 2C,E-I), TMEM8C is highly expressed in co-cultures (Fig. 2J,K). However, TMEM8C displayed the same expression pattern in myotubes of co-cultures and mono-cultures, being expressed in zones with high density of myonuclei: at contact points between myotubes and at myotube tips (Fig. 2J,K).

Fig. 2

Fibroblasts promote myoblast differentiation in fibroblast/myoblast co-cultures. A Percentage of quail myoblasts and chicken fibroblasts in co-cultures after five days of cultures. Representative field of quail nuclei (QCPN antibody, red) and chicken nuclei (DAPI, blue). B Quail myoblast / chicken fibroblast co-cultures immunolabelled to PAX7 (PAX7 antibody, red) and myosins (QCPN antibody, green) combined with DAPI staining (blue). C Quail myoblast / chicken fibroblast co-cultures immunolabelled to MYOG (red) and myosins (QCPN antibody, green) combined with DAPI staining (blue). D,E Percentage of PAX7-positive nuclei (D) and of MYOG-positive nuclei (E) versus total nuclei in fibroblast/myoblast co-cultures. F Fusion index of quail cells in co-cultures. Graphs show mean ± s.d. G,H Myoblast cultures (G) and myoblast/fibroblast co-cultures (H) labelled with the MF20 antibody (myosins, green), combined with DAPI staining (nuclei, blue). I Quantification of myotube area per surface unit in myoblast cultures and myoblast/fibroblast co-cultures. Graph shows mean ± s.d. J,K Fluorescent in situ hybridization to quail myoblast cultures (J) and to quail myoblast / chicken fibroblast co-cultures (K) with TMEM8C probe (red) followed by an immunolabelling of myosins (green) combined with DAPI staining (nuclei, blue)

We conclude that the presence of fibroblasts promotes all steps of the muscle program and does not modify the TMEM8C expression pattern in myotubes of fibroblast/myoblast co-cultures.

Fibroblast nuclei are incorporated into myotubes in myoblast/fibroblast co-cultures with no obvious regionalisation

We recently identified a cellular heterogeneity of tip myonuclei with the incorporation of fibroblast nuclei into myotubes with a preferential location at muscle/tendon interface in foetal limb muscles [18]. We first wanted to assess if fibroblast nucleus incorporation occurred in myotubes in vitro. To address this question, we set myoblast/fibroblast co-culture systems using the quail/chicken system that singles out quail and chicken nuclei independently to their differentiation status. We performed myoblast/fibroblast-co-culture experiments with quail primary myoblasts and chicken primary fibroblasts from E10 limbs, initially plated with a ratio of 80% myoblasts and 20% fibroblasts, and left for five days of cultures. Primary chicken fibroblasts isolated from foetal limbs did not express PAX7 or MYOG (Fig. S2A,B). Chicken fibroblasts express the SMA myofibroblast marker, the COL12A1 connective tissue marker, and the SCX (coding for a bHLH transcription factor) and TNMD (coding for a transmembrane protein) tendon markers (Fig. S2C-F), suggestive of a fibroblast/tendon progenitor phenotype [32]. Using the QCPN antibody that recognized quail nuclei and not chicken nuclei, we followed myoblasts and fibroblasts independently of their molecular statuses along the cultures. In quail myoblast / chicken fibroblast co-cultures, we appreciated a small population of myonuclei that were not of quail origin (Fig. 3A-B). Chicken fibroblast myonuclei represented 11% of the myonuclei, and 6% of fibroblast nuclei were recruited into myotubes after 5 days of culture (Fig. 3B). To exclude any potential bias regarding the labelling of quail myonuclei with the QCPN antibody, we performed the converse co-culture experiments with chicken myoblasts and quail fibroblasts. Similarly to the chicken fibroblast nucleus incorporation into quail myotubes, we observed an incorporation of quail fibroblast nuclei within chicken myotubes (Fig. 3C-C”). We quantified 20% of quail fibroblast myonuclei out of all myonuclei into myotubes and around 8% of quail fibroblast nuclei were recruited into myotubes (Fig. 3D). The fibroblast-derived myonuclei did not appear to be regionalized along the myotubes in both types of avian co-culture experiments (Fig. 3A-A”,C–C”), while fibroblast nuclei are incorporated with a preferential location at muscle/tendon interface in foetal limbs muscles [18] and postnatal muscles [19].

Fig. 3

Recruitment of fibroblast nuclei into myotubes in fibroblast/myoblast co-cultures. (A,A’,A”) Quail myoblast and chicken fibroblast and co-cultures immunolabelled to quail nuclei (QCPN antibody, red) and myosins (MF20 antibody, green), combined with DAPI staining (nuclei, blue). A Left panel shows quail nuclei/Myosins/DAPI, while right panels (A’,A”) are high magnification squared in A, showing quail nuclei/Myosins/DAPI staining (A’) and quail nuclei/Myosins (A”). Arrows point to chicken fibroblast myonuclei. B Percentage of chicken fibroblast myonuclei versus quail and chicken myonuclei and versus total chicken nuclei. Graph shows mean ± s.d. (C,C’,C”) Chicken myoblast and quail fibroblast and co-cultures immunolabelled to quail nuclei (QCPN antibody, red) and myosins (MF20 antibody, green), combined with DAPI staining (nuclei, blue). C Left panel shows quail nuclei/Myosins/DAPI, while right panels (C’,C”) are high magnification squared in C, showing quail nuclei/Myosins/DAPI staining (C’) and Myosins/DAPI (C”). Arrows point to quail fibroblast myonuclei. D Percentage of quail fibroblast myonuclei versus chicken and quail myonuclei and versus total quail nuclei. Graph shows mean ± s.d. (E,E’,E”,F,F’,F”) Fluorescent in situ hybridization to quail myoblast / chicken fibroblast co-cultures with MYOG (E,E’,E”) or TMEM8C (F,F’,F”) probes (red) followed by an immunolabelling of myosins (green) and of quail cells (grey) combined with DAPI staining (nuclei, blue). Arrows point to chicken fibroblast myonuclei expressing MYOG (E,E’,E”) and TMEM8C (F,F’,F”)

In order to test if fibroblast recruitment was conserved in co-cultures between species, we performed co-cultures with a human myoblast cell line (line AB1079) (Fig. S3A-A”) associated with mouse or chicken fibroblasts. We first verified that myoblast fusion was possible between human and mouse myoblasts. Human myoblasts did fuse with mouse C2C12 myoblasts to form heterogeneous myotubes with human and mouse myonuclei (Fig. S3B-B”). Co-cultures with human myoblasts and mouse C3H10T1/2 fibroblasts [26] show sporadic events of incorporation of mouse fibroblast nuclei into human myotubes (Fig. S3C-C”). We also performed co-cultures of human myoblasts with primary chicken fibroblasts and observed no or very rare events of chicken fibroblast nuclei into human myotubes (Fig. S3D-D”). The presence of heterologous fibroblasts did not change the fusion index of human muscle cells (Fig. S3E-H). We conclude that fibroblast nucleus recruitment to myotubes is observed within homologous and not heterologous species.

In order to assess if fibroblast myonuclei expressed the fusion-associated genes, we sought for MYOG and TMEM8C expression in fibroblast myonuclei of quail myoblast /chicken fibroblast co-cultures. We found that fibroblastic myonuclei expressed the fusion-associated genes, MYOG and TMEM8C (Fig. 3E-E”,F-F”), indicating a reprograming of fibroblast myonuclei towards the myogenic program.

We conclude that fibroblast nuclei are recruited into myotubes in myoblast/fibroblast co-cultures, similarly to fibroblast recruitment into foetal limb muscles, although with no regionalisation and that the fibroblast-derived myonuclei are reprogrammed into the myogenic program.

BMP signalling regulates the incorporation of fibroblast nuclei into myotubes in cultures

BMP signalling has been shown to regulate fibroblast nucleus incorporation into myotubes preferentially at muscle tips during limb development consistent with active BMP signalling at muscle/tendon interface [18, 21]. BMP4 ligand was produced by tendon cells (visualized with BMP4 transcripts in tendons) (Fig. 4A, arrow), while BMP-responsive nuclei, visualized with pSMAD1/5/9, were located in tendon cells (Fig. 4B,B’, arrow) and in a subpopulation of myonuclei close to tendon, in chicken limb muscles (Fig. 4B,B’,C). The tip regionalisation of BMP-responsive nuclei observed in limb muscles was lost in myoblast cultures since pSMAD1/5/9 was observed in all myonuclei of cultured myotubes (Fig. 4D-D”). In order to assess the effect of BMP signalling in fibroblast nucleus incorporation into myotubes, we performed BMP gain-and-loss of function experiments in chicken fibroblasts and then cultured them with quail myoblasts (Fig. 4E-J). Using the RCAS-BP(A) retroviral system, we were able to infect chicken fibroblasts and not quail myoblasts [27]. BMP gain-of-function experiments were performed with BMPR1Aca/RCAS and BMP4/RCAS [21], while BMP loss-of-function experiments were performed with BMPR1Adn/RCAS and NOGGIN/RCAS [3, 21]. The BMP-treated fibroblasts were then associated with quail myoblasts as co-cultures. When BMP signalling pathway was up-regulated in chicken fibroblasts, an increase of fibroblast nucleus incorporation into myotubes was observed from 11.5% to 13.3% for BMPR1Aca and from 11.5% to 14.3% for BMP4 (Fig. 4G), while the blockade of BMP signalling in fibroblasts decreased fibroblast nucleus incorporation into myotubes from 8% to 4.4% for BMPR1Adn and from 8% to 5.5% for NOGGIN (Fig. 4J). We did not observe any obvious regionalisation of fibroblast myonuclei along the myotubes in BMP treated co-cultures (Fig. 4E,F,H,I).

Fig. 4

BMP signalling regulates the incorporation of fibroblast nuclei into myotubes. A Fluorescent in situ hybridization to longitudinal sections of limb muscles with BMP4 probe (red) followed by an immunohistochimistry with the MF20 antibody (myosins, green) combined with DAPI staining (nuclei, blue). B,B’ Immunohistochemistry to adjacent longitudinal sections of limb muscle sections with pSMAD1/5/9 (red) and MF20 (myosins, green) antibodies combined with DAPI staining (nuclei, blue). C Immunohistochemistry to transverse sections of limb muscles with pSMAD1/5/9 (red) and MF20 (myosins, green) antibodies). D,D’,D” Immunohistochemistry to quail myoblast cultures with pSMAD1/5/9 (red), PAX7 (grey) and MF20 (myosins, green) antibodies combined with DAPI staining (nuclei, blue). E-J BMPR1Aca-transfected chicken fibroblasts (E), BMP4-transfected chicken fibroblasts (F), BMPR1Adn-transfected chicken fibroblasts (H) or NOGGIN-transfected chicken fibroblasts (I) were co-cultured with quail myoblasts; and labelled with the QCPN (quail nuclei, red), MF20 antibody (myosins, green) combined with DAPI staining (nuclei, blue). G Percentage of chicken fibroblast myonuclei within myotubes in control-, BMPR1Aca-, BMP4-transfected chicken fibroblasts cultured with quail myoblasts, (BMP gain-of-function experiments). J Percentage of chicken fibroblast myonuclei within myotubes in control-, BMPR1Adn-, NOGGIN-transfected chicken fibroblasts cultured with quail myoblasts, (BMP loss-of-function experiments). (G,J) Graphs shows mean ± s.d

We conclude that BMP signalling regulate the incorporation of fibroblasts into myotubes in vitro, as observed in limb muscles [18], although with no obvious regionalisation.

Genes expressed in tip myonuclei of limb muscles behave differently in cultured myotubes

BMP-responsive myonuclei lost their tip regionalisation in cultured myotubes (Fig. 4); this prompted us to analyse in a dish, the expression of genes known to be expressed in tip myonuclei of limb muscles. We first analysed the expression of the main MTJ marker, COL22A1 [10, 11] with that of the secreted factor, FGF4 also expressed in tip myonuclei of developing chicken muscles [20]. Consistent with these studies, COL22A1 and FGF4 transcripts were located in myonuclei at muscle ends close to tendons visualized with SCX expression on transverse and longitudinal muscle sections of E10 chicken limbs (Fig. 5A-F). The expression of COL22A1 and FGF4 was lost in chicken cultured myotubes (Fig. 5G-H). No COL22A1 expression was observed in quail cultured myotubes (Fig. 5I), while being expressed in tip myonuclei of quail limb muscles (Fig. S4). The presence of chicken fibroblasts did not induce COL22A1 expression in quail myotubes in co-culture experiments (Fig. 5I,J).

Fig. 5

The COL22A1 and FGF4 expression in tip myonuclei of limb muscles is lost in cultured myotubes. A-F Gene expression for muscle tip genes (COL22A1, FGF4) and tendon gene (SCX) in adjacent transverse (A-C) and longitudinal (D-F) muscle sections from E10 chicken limbs with fluorescent in situ hybridization with COL22A1 (A,D), FGF4 (B,E) and SCX (C,F) probes (red), followed by an immunohistochemistry with the MF20 antibody to label myosins (green) combined with DAPI staining (blue). Arrows point to tendons and muscle/tendon interface. G-J Fluorescent in situ hybridization to chicken myoblast cultures (G,H), quail myoblast cultures (I) and quail myoblast / chicken fibroblast co-cultures (J) with COL22A1 (G,I,J) and FGF4 (H) probes (red) followed by an immunolabelling of myosins (green) combined with DAPI staining (nuclei, blue)

We next analysed the expression of three other genes in cultured myotubes, genes identified as being expressed in the tip domains of muscles and involved in the skeletal muscle program. NES, coding for the cytoskeletal intermediate filament Nestin, contributes to skeletal muscle homeostasis and regeneration in mice [33]; Nestin being located at muscle tips in limb muscles of newborn mice and adult rats [34, 35]. The Nestin protein has been shown recently to be accumulated at the myotendinous junction in human and horse muscles [36]. ANKRD1 (ANKyrin repeat Domain 1) coding for a muscle-ankyrin repeat protein has been described as being expressed in tip myonuclei during chicken foetal development [6] and is a marker of the MTJ Col22a1 + nucleus cluster of adult mouse muscles [15]. MEF2C (Myocyte enhancer factor 2C) codes for a transcription factor at the crossroad of transcriptional regulations in the skeletal muscle program during muscle development, homeostasis and regeneration [37]. MEF2C labels the muscle/tendon interface in axial somites during larval Xenopus development [38] and zebrafish development [39]. NES, ANKRD1 and MEF2C transcripts were regionalized in tip myonuclei of foetal limb muscles of chicken embryos (Fig. 6A-C), close to SCX transcripts in tendon (Fig. 6D). NES transcripts were also regionalized in tip myonuclei of limb muscles of quail embryos (Fig. S4). In contrast to COL22A1 and FGF4, the 3 tip genes, NES, ANKRD1, MEF2C, were expressed in cultured myotubes and displayed a regionalized expression pattern at myotube tips, while MYOG was distributed along myotubes (Fig. 6E-H). The percentages of myotubes displaying NES (83%), ANKRD1 (80%) and MEF2C (67%) expression were similar for the three genes. An interesting point is that these three genes were located in zones of low myosin expression, while MYOG transcripts were located in zones of myosin expression in cultured myotubes (Fig. 6E’-H’).

Fig. 6

The regionalized expression of NES, ANKRD1 and MEF2C in tip myonuclei of limb muscles is maintained in cultured myotubes. A-D Gene expression for muscle tip genes (NES, ANKRD1, MEF2C) and tendon gene (SCX) in transverse muscle sections from E10 chicken limbs with fluorescent in situ hybridization for NES probe (A) and with colorimetric in situ hybridization for ANKRD1 (B), MEF2C (C) and SCX (D) probes (dark grey), followed by an immunohistochemistry with the MF20 antibody to label myosins (green). E Fluorescent in situ hybridization to quail myoblast cultures with NES probe (red) followed by an immunolabelling of myosins (green) combined with DAPI staining (nuclei, blue). F-G Fluorescent in situ hybridization to chicken myoblast cultures with ANKRD1 (F), MEF2C (G), MYOG (H) probes (red) followed by an immunolabelling of myosins (green) combined with DAPI staining (nuclei, blue). E’-H’ are high magnifications of squared areas in (E–H). Arrows point to NES (E’), ANKRD1 (F’) and MEF2C (G’) expression at myotube tips in low myosins-expressing zones, while MYOG transcripts (F’) are expressed in myosins-positive zones

In order to assess if the presence of fibroblasts affects the regionalisation of tip genes in myotubes, we analyse the expression of NES, ANKRD1 and MEF2C in co-culture experiments with quail myoblasts and chicken fibroblasts. The presence of fibroblasts did not modify the tip expression of NES, ANKRD1 and MEF2C transcripts in cultured myotubes (Fig. 7A-C). Thanks to the quail/chicken system, we could assess NES expression in fibroblast-derived myonuclei. The fibroblast-derived myonuclei did or did not express NES, consistent with the absence of regionalization of fibroblast recruitment and the persistent NES expression in myotube tips of co-cultures (Fig. 7D,D’,E, F-F”).

Fig. 7

The regionalized expression of NES, ANKRD1 and MEF2C in cultured myotubes is not changed in the presence of fibroblasts. A-C Fluorescent in situ hybridization to quail myoblast / chicken fibroblast co-cultures with the NES (A), ANKRD1 (B) and MEF2C (C) probes (red), followed by an immunohistochemistry with the MF20 antibody to label myosins (green) combined with DAPI staining (blue). D,D’ High magnification of the area squared in A, showing in the top panel, (NES transcripts, red, myosins, green) and DAPI staining, blue) combined with immunolabelling of quail nuclei (grey) (D), and in the bottom panel, immunolabelling of quail nuclei (grey) combined with DAPI staining, (blue) (D’). Arrowheads point to chicken fibroblast myonuclei expressing NES, while arrows point to point to chicken fibroblast myonuclei not expressing NES. E Fluorescent in situ hybridization to quail myoblast / chicken fibroblast co-cultures with NES probes (red) followed by an immunolabelling of myosins (green) and of quail nuclei (grey), combined with DAPI staining (nuclei, blue). F,F’,F” High magnification of the area squared in E, showing NES transcripts, (red), myosins, (green) and DAPI staining (blue) combined with immunolabelling of quail nuclei (grey) (F), showing myosins, (green) and DAPI staining (blue) combined with immunolabelling of quail nuclei (grey) (F’) and showing NES transcripts, (red), DAPI staining (blue) combined with immunolabelling of quail nuclei (grey) (F”). (F,F’,F”) Arrows points to a chicken fibroblast myonuclei with low/residual levels of NES incorporated into a quail myotube

We conclude that the tip markers display three types of behaviours in cultured myotubes: loss of regionalisation (pSMAD1/5/9), loss of expression (COL22A1, FGF4), or maintenance of regionalized expression (NES, ANKRD1, MEF2C). The presence of fibroblasts does not modify the expression pattern of tip markers in cultured myotubes.

Inhibition of muscle contraction affects gene expression in tip myonuclei of limb muscles

Mechanical parameters are recognized to be an important regulator of the musculoskeletal system [8]. Muscle contractions have been shown to regulate the steps of the muscle program during foetal myogenesis [6, 23]. The absence of muscle contraction reduces the pool of muscle progenitors, while increasing their propensity to differentiate and fuse during foetal myogenesis. In order to assess the behaviour of genes expressed in tip myonuclei in absence of mechanical signals, we analysed the expression of tip genes in an unloading model during foetal myogenesis. We used the decamethonium bromide (DMB), which blocks muscle contractions and leads to limb muscle paralysis [40]. Two days after the inhibition of muscle contraction, the expression of COL22A1 and FGF4 was lost in limb muscles (Fig. 8A-D), reminiscent of the loss of COL22A1 and FGF4 expression in myotube cultures (Fig. 5G-I). The regionalisation of pSMAD1/5/9 in tip myonuclei in foetal limbs was lost in limbs of immobilized embryos (Fig. S5); reminiscent of pSMAD1/5/9 location in all myonuclei of cultured myotubes (Fig. 4D-D’’). Lastly, the regionalised expression of NES was maintained in limb muscles in the absence of muscle contraction (Fig. 8E,E’,F,F’), reminiscent of the regionalized expression of NES in cultured myotubes (Fig. 6E,E’), while the expression of ANKRD1 [6] and MEF2C (Fig. S6) was lost in limbs immobilized embryos.

Fig. 8

Inhibition of muscle contraction modified the expression of genes expressed in tip myonuclei of foetal muscles. Chicken embryos were treated with DMB at E7.5 to block muscle contraction and analysed 48 h after treatment at E9.5 (N = 3). A-D In situ hybridization to transverse limbs sections at the level of the zeugopod of control (A,C) and DMB-treated (B,D) E9.5 chicken embryos, with COL22A1 (A,B) and FGF4 (C,D) probes (blue) followed by immunostaining with the MF20 antibody to label myosins (light brown). The expression of COL22A1 and FGF4 was lost in paralyzed muscles (B,D) compared to control muscles (A,C). Note that COL22A1 expression is maintained in cartilage and perichondrium in paralyzed limbs (B) as in control limbs (A). E, E’, F,F’ Fluorescent situ hybridization to transverse limbs sections at the level of the zeugopod of control (E,E’) or DMB-treated (F,F’) E9.5 chicken embryos, with NES probe (red) followed by immunostaining with the MF20 antibody to label myosins (green) and stained with DAPI (blue). (E’,F’) are high magnification of the FCU muscles of E (control) and F (DMB-treated) limbs. u, ulna, r, radius

We conclude that the mechanical signals affect the molecular heterogeneity of myonuclei in limb foetal muscles.