In whole embryos, Sepn1 expression was detected by qRT-PCR as early as E5.5, with higher levels from E9.5 to E12.5 (six and ten fold increase, respectively). We then observed a five fold reduction of the expression from E12.5 to E18.5 (Figure 1A). This decrease was confirmed at the protein level by Western blot since SelN was reduced by more than two fold in E18.5 embryos compared to E12.5 (Figure 1B).

We then investigated Sepn1 expression in isolated quadriceps, diaphragm and non-skeletal-muscle tissues (brain, liver, kidney and heart) from late embryonic stages (E15, E18) to 15 months of age, with an earlier embryonic stage included for the heart (E13) (Figure 2A). By qRT-PCR, Sepn1 expression was detected in all tissues, albeit at barely detectable levels in liver (Figure 2A). Overall, variations in the mRNA levels observed for the different tissues were limited. Nevertheless, we observed a significant decrease in expression between E15 and E18 in the brain and in kidney. In parallel, increased levels were detected between E18 and postnatal day 1 in diaphragm, heart and kidney. Lastly, we showed a two fold increase in expression in the brain, between young (1-6 weeks) and adult mice (3-15 months) and a two fold reduction in kidney, between the first week of life and older mice (3 weeks to 15 months) (Figure 2A). Interestingly, expression in quadriceps and diaphragm did not exhibit the highest levels and appeared mostly constant. Similar results were obtained for skeletal muscles isolated from posterior (soleus + gastrocnemius) and anterior (tibialis anterior + extensor digitorum longus) hind limb compartments (data not shown).

We studied the expression of SelN at the protein level by Western blot, using proteins extracted from isolated tissues from E18 to 15 months of age. Several proteins were tested for protein loading control; actin and GAPDH exhibited the least variability between the various developmental stages (Figure 2B and data not shown). Nevertheless, quadriceps homogenates displayed an increase of both actin and GAPDH expressions after 15 days of age; for this tissue, no ideal protein normalization could be obtained. For all the tissues considered except liver, we observed a striking reduction in SelN expression between E18 and postnatal day 1 (about three fold reduction), and even more so at adult ages when it became barely detectable (less than 10% compared to E18) (Figure 2B). In protein extracts from liver, SelN was never detected, even at E18 (Figure 2B), in keeping with the transcript quantifications.

In parallel, using mouse C2C12 muscle cells, we observed a 50% decrease of Sepn1 expression in myotubes after 2 and 3 days of differentiation, compared to myoblasts (Figure 3A). This reduction was even more marked at the protein level with a five fold reduction detected in myotubes compared to myoblasts by Western blot (Figure 3B). This is in accordance with data previously obtained with human cells 16.

Since no appropriate antibody was available for immunohistochemistry, we investigated Sepn1 expression pattern during early embryogenesis by whole mount in situ hybridization (ISH). The murine Sepn1 cDNA [GenBank: NM_029100], 3461 bp in size, contains 12 exons, the in-frame UGA codon being located in exon 9, and leads to a protein of 557 residues (Figure 4A). In mouse, no sequence homology was found with the alternatively spliced third exon described in human, since this exon corresponds to a primate specific Alu sequence 11. Two Sepn1 probes (Pr1: 573 bp and Pr2: 696 bp) were synthesized from PCR fragments corresponding to the 3' region of the cDNA (Figure 4A). The corresponding murine transcript regions shared more than 85% identity with the human sequence. We performed Northern blot analysis with Pr1 and Pr2 probes, using RNA extracted from cultured human fibroblasts from a control subject and a RSMD1 patient with SEPN1 mRNA degradation due to a premature termination codon (p.L482fs 20 - Figure 4B). For both probes, one band corresponding to SEPN1 transcripts was observed in control RNA, but was undetectable in RNA extracted from the patient, thus validating the specificity of the probes (Figure 4B and data not shown). Using mouse RNA extracted from E12 and E18 whole embryos, we detected only one 3.4 kb band by Northern blot, which corresponds to the predicted transcript size for Sepn1 (Figure 4C).

Sepn1 deficient mouse embryos were used as negative controls for whole mount ISH. The recombinant gene construct introduced a frame-shift and a premature stop codon within the coding sequence, leading to a loss of the functional protein in mutants (Rederstorff et al., manuscript in preparation). As shown in Figure 5A, Sepn1 expression was barely detectable at the transcript levels in E12.5 mutant whole embryos (Sepn1^-/-) compared to wild-type (Sepn1^+/+), indicative of the Sepn1 mRNA degradation in the model. At the protein level, SelN was observed in wild-type embryos by Western blot but was undetectable in mutant samples (Figure 5B).

Identical expression patterns were obtained with Pr1 and Pr2; results presented in Figure 6 correspond to ISH performed with the Pr1 probe. At the different embryonic stages tested, no signal could be detected using either sense probes with wild-type embryos (Figure 6A, panel a and data not shown) or antisense probes with Sepn1^-/- embryos (Figure 6B, panel i and 6C, panel t). This confirmed the specificity of the staining pattern obtained.

In E8.5 whole embryos, we detected Sepn1 expression in somites and the neural tube, as well as in the mesenchyme located under the ectoderm, in the head and the branchial arches of embryos (Figure 6A, panels b-f). At E9.5, a ubiquitous-like staining was observed in whole embryos, with a stronger expression in somites (Figure 6B, panel g). Upon vibratome sectioning, the staining was associated with expression in the sub-ectodermal mesenchyme, as depicted in forebrain (Figure 6B, panel h). At E10.5 and E11.5, whole embryos appeared strongly stained in the dorsal part of the body, the limb buds, the branchial arches and the head (Figure 6C, panels j-m). Following sectioning, we observed expression in the dorsal root ganglia (drg). Furthermore, within the somites, the myotome appeared strongly stained, whereas little to no signal was detected in the dermomyotome of embryos (Figure 6C, panels n-q). Moreover, as previously seen, the mesenchyme located under the ectoderm exhibited staining, particularly in the head and limb buds (Figure 6C, panels r, s), although this signal was weak and required a longer revelation time as compared to the staining observed in the myotome and drg.

Interestingly, at all of these stages, we did not observe staining in the heart by ISH, whereas Sepn1 expression was detected as early as E13 in this tissue by qRT-PCR. This could result either from Sepn1 cardiac expression appearing after E11.5, or more likely from the higher sensitivity of the qPCR technique.

Since an early expression of Sepn1 was observed in somites and defects were reported in somite organisation and myogenic factors expression in sepn1 morpholino injected zebrafish embryos 1718, we specifically investigated these events in Sepn1^-/- embryos. At mid-gestation stages (E10 and E11.5), ISH performed against the myogenic determination factors Myf5, MyoD, and Pax3 revealed no defect in somite size and organisation in Sepn1^-/- embryos compared to wild-type (Figure 7A and data not shown). Moreover, the expression pattern of these factors was similar in both types of embryos (Figure 7A). Likewise, the expression of myosin heavy chain, a marker of differentiated muscle cells, was unaltered when visualized by immunohistochemistry in E11.5 and E13 deficient embryos (Figure 7B). These data were confirmed by qRT-PCR and Western blot analyses that demonstrated no modification of the myogenic factors expression in the deficient embryos (Figure 7C and data not shown). Lastly, the expression patterns of Scleraxis, involved in the establishment of myotendinous junctions 21, and of TrkA, a neurotrophin receptor which was used as a marker of the drg 22, appeared unmodified in Sepn1^-/- embryos by whole-mount ISH (Figure 7A).

Mouse C2C12 cells were grown in an humidified atmosphere of 5% CO₂, at 37°C, in Dulbecco's modified Eagle's medium (DMEM, Invitrogen) supplemented with 20% foetal calf serum (FCS, Invitrogen), 2 mM gentamycine (Invitrogen). Differentiation into myotubes was induced by switching the medium to DMEM supplemented with 1% FCS.

C57Bl/10 mice were euthanized with isoflurane and tissues were dissected at various ages. Embryos were obtained from gestating females of the Sepn1 knock-out lineage (Rederstorff et al., article in preparation) between embryonic days 5.5 (E5.5) and 18.5 (E18.5), the morning of the vaginal plug indicating the stage E0.5. For in situ hybridization, embryos were dissected at 4°C in phosphate buffer saline (PBS) and freed from their extra-embryonic membranes. At E9.5 and older stages, the neural tube was pierced open at the level of the brain to facilitate exchange of solutions. Embryos were fixed overnight at 4°C, in 4% paraformaldehyde (PFA), dehydrated in methanol and conserved at -20°C. DNA extraction was carried out for genotyping with the Wizard Genomic DNA Isolation System (Promega) according to the manufacturer's recommendations. All animal studies were performed in accordance with the European Union guidelines for animal care and approved by the local ethical committee on animal experimentation.

Whole embryos and isolated tissues were frozen in liquid nitrogen. Total RNA was extracted using the RNeasy Fibrous Tissue Mini Kit (Qiagen) according to the manufacturer's instructions. Total RNA was also extracted from cultured human fibroblasts (control and RSMD1 patient with a c.1446delC homozygous mutation) by Trizol (Invitrogen) procedure as previously described 11. RNA yield and quality were assessed with the Nanodrop system (Thermo Scientific) and on the 2100 Bioanalyzer (Agilent Technologies). RNA samples with a RIN (RNA Integrity Number) superior to 8 were used for quantification. cDNA was synthesized from 500 ng of RNA with random hexamer primers using the SuperScript First-Strand Synthesis System for RT-PCR (Invitrogen).

Sepn1 transcripts were quantified by real time PCR using the following primers: 5'GCTTTCCTGTAGAGATGATG3'/5'GCCCCGCCGGAGTCCTTC3'. qPCR were performed on the LightCycler480 System (Roche) using the LighCycler480 SYBR Green I Master mix (Roche). The program included an initial denaturation step of 8 min at 95°C, followed by 40 amplification cycles of denaturation at 95°C for 10 sec, hybridization at 63°C for 15 sec, and elongation at 72°C for 15 sec. Quantification of additional genes was performed similarly with the following primers: 18s: 5'ACCTGGTTGATCCTGCCAGT3'/5'CTCACCGGGTTGGTTTTGAT3'; Hprt: 5'AAGCAGATGGCCACAGAACT3'/5'ACCCCACGAAGTGTTGGATA3'. Data were analyzed using the LightCycler480 analysis software (Roche).

Two digoxigenin (DIG) labeled riboprobes (Pr1/Pr2) were generated from Sepn1 PCR fragments obtained from total murine cDNA with the following primers: Pr1: 5'TCCTAGATGAGGACGGCAAC3'/5'GGTCCTCAAAGCTGGATGAG3'; Pr2: 5'CGCCCATCCTCACTCTCCTC3'/5'TCCAGTCCACTCCCTATTCACA3'. Fragments were cloned in the pGEM-T vector (Promega). Antisense and sense probes were synthesized using the DIG RNA Labeling Kit (Roche) with T7 or SP6 RNA polymerase. Sense probes were used as negative controls. Yield and quality of the probes were estimated by electrophoresis and dot blot compared to a standard (β-actin probe - Roche). MyoD, Myf5, Scleraxis and TrkA probes, previously described 4243, were also used for whole mount ISH.

In situ hybridizations (ISH) were performed on embryos aged from E8.5 to E11.5, as previously described 43. Briefly, after rehydration, embryos were treated with 10 μg/ml proteinase K (Qiagen) at room temperature, from 10 to 30 min according to their age and fixed in 4% PFA. Embryos were incubated for more than 1 h in hybridization buffer (50% formamide, 1.3× SSC pH 5, 50 μg/ml yeast RNA, 5 mM EDTA pH 8, 0.2% Tween-20, 0.5% CHAPS, 100 μg/ml heparin) and then overnight with the appropriate probe at 0.5 to 2 μg/ml, at 69°C. Embryos were successively washed in hybridization buffer, in 1:1 hybridization buffer/MABT (100 mM maleic acid, 150 mM NaCl, 1% Tween-20), and finally in MABT. After blocking in MABT, 2% Blocking Reagent (Roche), 20% inactivated goat serum (Invitrogen) for more than 1 h at room temperature, embryos were incubated overnight at 4°C in alkaline phosphatase conjugated anti-DIG antibody (1:2000 - Roche). Revelation was obtained with the BM Purple solution (Roche) at room temperature and/or at 4°C, during different time laps depending on the probe. Vibratome sections (50 to 80 μm) were performed after inclusion of the embryos in 10% sucrose overnight at 4°C and then in 20% gelatin.

30 μg of total RNA were resolved on 1% formaldehyde agarose gel and blotted overnight onto a positively charged nylon membrane (Roche). RNA were UV cross-linked to the membrane during 3 min. Membranes were hybridized overnight at 68°C with diluted probe (100 ng/ml) in hybridization buffer (DIG Easy Hyb buffer - Roche) and then rinsed in 2× SSC, 0.1% SDS at room temperature and in 0.1× SSC, 0.1% SDS at 68°C. Membranes were blocked in Blocking Reagent (Roche) diluted in 0.1 M maleic acid, 0.15 M NaCl, pH 7.5, and then incubated with alkaline phosphatase conjugated anti-DIG antibody (Roche) in maleic acid buffer. After incubation in detection buffer (0.1 M Tris-HCl, 0.1 M NaCl, pH 9.5), signals were obtained with the CDP-star, ready-to-use buffer (Roche) on the ChemImager system (Alpha Innotech).

Frozen tissues were ground in liquid nitrogen and homogenized in protein extraction buffer (80 mM Tris-HCl, pH 6.8, 10% SDS, 120 mM sucrose, 10 mM EDTA, 1 mM PMSF and 1 mM benzamidine). Homogenates were incubated 10 min at 55°C, sonicated twice and quantified with the BCA Protein Assay (Pierce). 60 μg of proteins were electrophoretically separated in 8% polyacrylamide SDS gel and transferred to a polyvinylidene difluoride membrane (Invitrogen). Membranes were immunoprobed with the following primary antibodies: rabbit anti-SelN (ab137 16), mouse anti-tubulin α (Sigma T5168), rabbit anti-actin (Sigma A2066), rabbit anti-MyoD (Santa Cruz sc-304), rabbit anti-myogenin (Santa Cruz sc-576), rabbit anti-Myf5 (Santa Cruz sc-302) and mouse anti-MyHC (ATCC, MF20). They were then incubated with HRP-conjugated secondary antibodies (Dako). Signals were detected with chemiluminescent HRP substrate (Immobilon Western - Millipore) on the SynGene instrument (Ozyme) and quantified with the ImageJ software.

After rehydration, E12.5 whole embryos were equilibrated in 15% sucrose overnight at 4°C. They were then frozen in liquid nitrogen-cooled isopentane. Ten μm cryostat sections were fixed in 4% PFA, rinsed in 0.1 M glycine, and incubated in methanol at -20°C. They were then placed in boiling 0.01 M citric acid for 10 min to achieve antigen retrieval. Sections were blocked in 3% IgG free bovine serum albumin (BSA - Jackson ImmunoResearch) overnight at 4°C and with ChromoPure Mouse IgG, Fab Fragments (Jackson ImmunoResearch) for 30 min. Sections were then incubated in 3% IgG free BSA, 0,1% Triton X-100, with primary antibodies (MyHC: monoclonal mouse ATCC MF20-1/100; laminin: polyclonal rabbit Abcam ab11575-1/1000) for 3 h at room temperature. Alexa Fluor secondary antibodies (goat anti-mouse A568, goat anti-rabbit A488; Jackson ImmunoResearch) were incubated on slides for 90 min at room temperature. Sections were rinsed, mounted in Vectashield DAPI (Vector) and observed with an Axiophot microscope (Zeiss). Images were captured using the Metaview software (Ropper Scientific).

Results represent Means ± SD. Comparisons between two groups were performed using a Student's t-test or a Mann-Whitney rank sum test depending on the normal distribution of data, with a 0.05 level of confidence accepted for statistical significance. Multiple group comparison was performed by one way ANOVA with Dunn's pairwise procedures (p < 0.05).