Single-nucleus RNA-seq and FISH identify coordinated transcriptional activity in mammalian myofibers

Abstract

Skeletal muscle fibers are large syncytia but it is currently unknown whether gene expression is coordinately regulated in their numerous nuclei. Here we show by snRNA-seq and snATAC-seq that slow, fast, myotendinous and neuromuscular junction myonuclei each have different transcriptional programs, associated with distinct chromatin states and combinations of transcription factors. In adult mice, identified myofiber types predominantly express either a slow or one of the three fast isoforms of Myosin heavy chain (MYH) proteins, while a small number of hybrid fibers can express more than one MYH. By snRNA-seq and FISH, we show that the majority of myonuclei within a myofiber are synchronized, coordinately expressing only one fast Myh isoform with a preferential panel of muscle-specific genes. Importantly, this coordination of expression occurs early during post-natal development and depends on innervation. These findings highlight a previously undefined mechanism of coordination of gene expression in a syncytium.

Fig. 1. SnRNA-seq analysis from adult slow and fast skeletal muscles.

a Graphical scheme of the experiments used for snRNA-seq analysis of adult skeletal muscles. b Uniform Manifold Approximation and Projection (Umap) diagram from snRNA-seq from adult skeletal muscle. MuSC skeletal muscle stem cells, FAPs fibro-adipogenic progenitors, MTJ myotendinous junction, NMJ neuromuscular junction. c Umap diagram of snRNA-seq data with myonuclei only. Myonuclei are separated into five clusters: slow, fast, MTJ, NMJ, and unidentified myonuclei. The slow and fast myonuclei are called body nuclei and are surrounded by dotted lines. Five subcategories of fast myonuclei are clustered. d Heatmap of genes upregulated (yellow) and downregulated (violet) in the myonuclei expressing the different slow and fast Myh isoforms. The values correspond to z-scores of normalized counts.

Fig. 2. The different populations of myonuclei.

a Umap Plots showing the expression of several markers used to identify the different types of myonuclei. The intensity of the blue color reflects the level of expression of the gene. b Heatmap of the top 30 genes upregulated and downregulated (yellow) in body myonuclei versus unidentified myonuclei (left), in body myonuclei versus NMJ nuclei (center), and body myonuclei versus NMJ nuclei (right). The values correspond to z-scores of normalized counts. c Dot-plots from snRNA-seq experiments showing the expression of the slow and fast Myh in the different populations of myonuclei. NMJ and MTJ nuclei expressed Myh like other myonuclei.

Fig. 3. SnATAC-seq analysis from adult slow and fast skeletal muscles.

a Umap Plots from snATAC-seq from adult skeletal muscle based on chromatin accessibility. MuSC skeletal muscle stem cells, FAPs fibro-adipogenic progenitors, MTJ myotendinous junction, NMJ neuromuscular junction. b Umap plots showing the chromatin opening of slow and fast Myh genes used to identify the different types of myonuclei. The intensity of the blue color reflects the level of chromatin accessibility of the gene. c Position weight matrix (PWM) motifs showing the enrichment of TF motifs in differential peaks in the myonuclei expressing the different Myh genes. d Umap Plots showing the motif activity of TF in the different types of myonuclei. The intensity of the blue color reflects the level of TF motif enrichment in the nuclei. e PWM motif showing the enrichment of TF motifs in differential peaks in the body and MTJ myonuclei.

Fig. 4. The majority of myonuclei express only one isoform of Myh.

a Diagram of the mouse Myh fast and slow loci. b Analysis of Myh isoforms expression in myonuclei from snRNA-seq data. Each dot corresponds to a myonucleus and the x- and y-axis corresponds to the indicated Myh expression level. c RNAscope experiments on isolated fibers from EDL to visualize fast Myh4 (red), Myh1 (green) and Myh2 (yellow) pre-mRNAs. Myh pre-mRNA can be detected as two transcribed alleles (white arrowhead) or as a single allele (green arrowhead). d Same as c showing nuclei expressing at the same time different isoforms of Myh from each Myh allele and nuclei co-expressing (arrowhead) at the same time two isoforms of Myh from the same allele. e Same as c in soleus showing nuclei co-expressing at the same time Myh1 and Myh2 pre-mRNAs from each allele (arrowhead). f Percentage of nuclei expressing fMyh pre-mRNAs, in control and 1 week after denervation, in fast myofibers from EDL and soleus. g Percentage of nuclei expressing one, two, and three isoforms of pre-mRNAs of Myh in control or 1-week denervated EDL and soleus. Percentage of mono- and bi-allelic expression is also indicated. For c–e scale bar: 20 μm. For f and g, n = 3 and 20 fibers per animal. Numerical data are presented as mean ± s.e.m. *P < 0.05, **P < 0.01, ***P < 0.001. Source data for (f, g) are provided in the Source Data file.

Fig. 5. Sarcomeric and metabolic genes are co-expressed in myonuclei.

a Analysis of Atp2a2 and Myh isoform expression in myonuclei from snRNA-seq data from Fig. 1a. Each dot corresponds to one myonucleus and the axis correspond to the gene expression level. Atp2a2 is expressed only in most Myh7 and in few Myh2 nuclei. b Same as a with Tbc1d1 expression. Tbc1d1 is expressed only in fast and not in slow nuclei. c Same as a with Pvalb expression. Pvalb is expressed only in Myh1 and Myh4 nuclei. d Same as a with Pfkfb3 expression. Pfkfb3 is expressed preferentialy in Myh4 nuclei. e RNAscope on isolated fibers from EDL showing the localization of Myh4 (red) and AldoA (green) pre-mRNAs. Most myonuclei expressed at the same time Myh4 and AldoA (white arrowhead). However, some nuclei expressed Myh4 without AldoA (red arrowhead), and some nuclei are negative for both genes. f Percentage of the nuclei co-expressing Myh4 and AldoA or expressing only Myh4 or AldoA and negative nuclei in EDL fibers. g RNAscope on isolated fibers from soleus showing the localization of Myh2 (red) and Idh2 (green) pre-mRNAs. Like in e most myonuclei expressed at the same time Myh2 and Idh2 (white arrowhead). However, some nuclei expressed Myh2 without Idh2 (red arrowhead) and other nuclei are negative for both genes. h Percentage of the nuclei co-expressing Myh2 and Idh2 or expressing only Myh2 or Idh2 and negative nuclei in soleus fibers. For e.g., scale bar: 100 μm. For f, h the graphs show data pooled from three animals and fifteen fibers per animal. Numerical data are presented as mean ± s.e.m. *P < 0.05, **P < 0.01, ***P < 0.001. Source data for f, h are provided in the Source Data file.

Fig. 6. Coordination of fast Myh expression along myofibers.

a RNAscope on isolated fibers from EDL showing the localization of Myh4 (red), Myh1 (green), and Myh2 (yellow) pre-mRNAs. Up: EDL myofiber showing coordination of Myh4 expression between nuclei (pure myofiber). Down: EDL myofiber after 1 week of denervation. After denervation, myonuclei are no more coordinated in the myofiber and express either Myh4 or Myh1 (green arrowhead) or co-express both isoforms (hybrid myofiber). In most myonuclei, fMyh pre-mRNAs are no more detected (white arrowhead). b Same as a with Myh4 (yellow), Myh1 (green), and Myh2 (red) pre-mRNAs in a soleus hybrid myofiber. The white arrowhead shows a mixed Myh1, Myh2 myonucleus, whereas the red arrowhead shows a pure Myh2 myonucleus. c Percentage of the different types of fibers observed in quadriceps, EDL and soleus, and 1 week after denervation. d Percentage of hybrid (de-coordinated) fibers in control EDL and soleus and after 1 week of denervation. e SIX1 accumulates in all myonuclei of adult EDL fibers as identified by immunohistochemistry with SIX1-specific antibodies. f Percentage of the different types of EDL fibers observed in WT and Six1^flox/flox;HSA-CRE mice by RNAscope experiments. g Percentage of hybrid EDL fibers (de-coordinated) in WT and Six1^flox/flox;HSA-CRE mice. For a, b (up) scale bar: 100 μm. For b (down) scale bar: 20 μm. For c, d, n = 3 and 20 fibers per animal. For f, g the graphs show data pooled from three animals and more than fifty fibers in total. Numerical data are presented as mean ± s.e.m. *P < 0.05, **P < 0.01, ***P < 0.001. Source data for c, d, f, g are provided in the Source Data file.

Fig. 7. fMyh expression during development is regionalized.

a RNAscope against Myh3 (green) and Myh8 (red) mRNA on isolated fibers from forelimbs at 5 days post-natal (P5). The MTJ and NMJ areas showed an accumulation of Myh3 mRNA. b RNAscope against Myh4 (red), Myh1 (green), and Myh2 (yellow) pre-mRNAs on isolated fibers at E15.5, E18.5, 2 (P2), and 5 (P5) days post-natal. c Same as b. Zoom showing the increase of adult fast Myh-positive (arrowhead) nuclei after birth. d Percentage of nuclei expressing one, two, or three adult fMyh isoforms at the same time at different developmental stages. e Percentage of nuclei with bi-allelic expression of Myh genes during development. f Percentage of coordinated nuclei inside myofibers at different developmental stages. g RNAscope against Myh7 (green) and Myh8 (red) mRNAs on isolated fibers at E15.5. Myh7 mRNAs accumulate in MTJ areas of all myofibers at E15.5. For (a–c, g) scale bar: 100 μm. For d, n = 3 and 50 nuclei per animal. For e, f n = 3 and 20 fibers per animal. Numerical data are presented as mean ± s.e.m. *P < 0.05, **P < 0.01, ***P < 0.001. For a and g representative RNAscope experiments are presented, N = 3. Source data for d–f are provided in the Source Data file.

Fig. 8. Diagram of Myh expression in pure and hybrid myonuclei and myofiber.

a Diagram showing the different possibilities of Myh isoforms expression in a myonucleus from the results of our experiments. b Diagram showing the different possibilities of Myh isoforms expression in a myofiber from the results of our experiments.