Lineage tracing of nuclei in skeletal myofibers uncovers distinct transcripts and interplay between myonuclear populations

Abstract

Multinucleated skeletal muscle cells need to acquire additional nuclei through fusion with activated skeletal muscle stem cells when responding to both developmental and adaptive growth stimuli. A fundamental question in skeletal muscle biology has been the reason underlying this need for new nuclei in cells that already harbor hundreds of nuclei. Here we utilize nuclear RNA-sequencing approaches and develop a lineage tracing strategy capable of defining the transcriptional state of recently fused nuclei and distinguishing this state from that of pre-existing nuclei. Our findings reveal the presence of conserved markers of newly fused nuclei both during development and after a hypertrophic stimulus in the adult. However, newly fused nuclei also exhibit divergent gene expression that is determined by the myogenic environment to which they fuse. Moreover, accrual of new nuclei through fusion is required for nuclei already resident in adult myofibers to mount a normal transcriptional response to a load-inducing stimulus. We propose a model of mutual regulation in the control of skeletal muscle development and adaptations, where newly fused and pre-existing myonuclear populations influence each other to maintain optimal functional growth.

Fig. 1. Distinct myonuclear populations are associated with times of cell fusion.

a Uniform Manifold Approximation and Projection (UMAP) visualization represents seven color-coded myogenic nuclei clusters identified by snRNA-seq. Data were generated from tibial anterior (TA) muscles at postnatal (P) day 10. b Dot plot depicting canonical markers of the various myogenic states. Dot size denotes the percentage of nuclei expressing the gene. c PHATE trajectory illustrating myogenic populations presented in (a) (left), with Dev. 1 and Dev. 2 populations highlighted in the same PHATE plot on the right. d Heatmap of myogenic nuclei from P10 TA muscle shows a distinct transcriptional signature. e Integrated dataset displays myogenic nuclei from TA muscle at P10 and 5 months of age (Split UMAP view). f Dot plot demonstrates altered genes in Dev. myonuclei. Dot size denotes the percentage of nuclei expressing the gene. Pink bar indicates genes that are increased in Dev nuclei compared to MuSCs and decreased in other myonuclear populations at P10 and 5 months of age. Purple bar indicates genes expressed in Dev nuclei but at a reduced level compared to MuSCs. Gray bar highlights genes that are similarly expressed across myonuclear populations.

Fig. 2. A recombination-independent nuclear tracking system distinguishes newly fused nuclei and allows evaluation of gene expression.

a Schematic of the Pax7^rtTA; TRE^H2B-GFP alleles and Dox-inducible lineage tracing system. b Representative images of a single myofiber isolated from the Extensor digitorum longus (EDL) muscle at P28 after 16 days of Dox treatment. The bottom panel represents a magnified field of view of the region indicated by the box in the top panel. GFP⁺ myonuclei (white arrows) and pre-existing GFP⁻ myonuclei (arrowhead) are observed. Scale bar, 200 μm. c Representative image showing newly fused GFP⁺ myonuclei (white arrows) and pre-existing GFP⁻ myonuclei (arrowhead) (left). Quantification of the percentages of GFP⁺ myonuclei and GFP⁻ myonuclei (right) in TA muscles. Dystrophin was used to outline myofibers. (n = 4). Scale bar, 20 μm. d Representative images of smRNA-FISH for H19 and Igf2 on an EDL myofiber, taken from the same mice used in (c). Scale bar, 20 μm. e Representative images of smRNA-FISH for H19 and Igf2 (left) on Pax7^rtTA; TRE^H2B-GFP TA muscle at P28. Quantification of signal in GFP⁺ and GFP⁻ is shown on the right. Blue arrows indicate increased signal in newly fused GFP⁺ myonuclei. (n = 3–4 independent animals). Scale bar, 20 μm. f Experimental design used to delete Myomaker and block myonuclear accretion during development is shown. Myomaker^loxP/loxP; Pax7^CreER mice were treated with tamoxifen to generate fusion-incompetent mice (Mymk^scKO). Control mice were Myomaker^loxP/loxP treated with the same tamoxifen regimen. g Representative images of smRNA-FISH for H19 and Igf2 (left) from control and Mymk^scKO TA muscles. Quantification of H19 and Igf2 signal is shown on the right. (n = 4–5 independent animals). Scale bar, 20 μm. Data are presented as mean ± SD. Statistical tests used were (e) two-tailed paired t-test; (g) two-tailed unpaired t-test; *p < 0.05, **p < 0.01. Source data are provided as a Source Data file.

Fig. 3. bnRNA-seq analysis comparing newly fused myonuclei acquired during developmental and adult growth.

a Schematic diagram of the bulk nuclei RNA-seq of GFP⁺ and GFP⁻ myonuclei from Pax7^rtTA; TRE^H2B-GFP mice during postnatal development. TA and gastrocnemius muscles at P28 were used after 10 days of doxycycline treatment. b Principal component analysis of the top 500 variable genes from bnRNA-seq of newly fused myonuclei (GFP⁺) and pre-existing myonuclei (GFP⁻). c Volcano plot depicting DEGs between newly fused myonuclei (GFP⁺) and pre-existing myonuclei (GFP⁻). d Gene Ontology (GO) analysis of the DEGs from (c), showing significantly changed biological processes with a false discovery rate (FDR) < 0.01. Gene lists used for GO analysis are shown in Supplementary Data 2. e Schematic diagram for bnRNA-seq of GFP⁺ and GFP⁻ myonuclei from Pax7^rtTA; TRE^H2B-GFP mice during adult muscle overload. f Principal component analysis of bnRNA-seq data from newly fused myonuclei (GFP⁺) and existing myonuclei (GFP⁻) isolated from plantaris muscle 1 week after muscle overload, compared with sham myonuclei (Pax7^rtTA; TRE^H2B-GFP mice treated with Dox as in (e)) and developmental myonuclei datasets from (b). g Volcano plots reveal up- and down-regulated DEGs in newly fused myonuclei (GFP⁺) or pre-existing myonuclei (GFP⁻) compared with sham myonuclei. h Volcano plot reveals up- and down-regulated DEGs from comparison of GFP⁺ and GFP⁻ myonuclei.

Fig. 4. Differential contributions of newly fused and pre-existing myonuclei to adult muscle adaptations.

a Schematic diagram of the single-nuclei RNA-seq methodology applied to HSA^rtTA; TRE^H2B-GFP mice, where all myonuclei were labeled with GFP and sorted for subsequent sequencing analysis. b Integrated UMAP from plantaris muscle following overload stimulation (n = 2) or under sham condition (n = 1) (HSA^rtTA; TRE^H2B-GFP treated with Dox as depicted in (a)). c Split UMAP view of myogenic nuclei shown in (b). d Proportion of unclassified A, B, and Atf3⁺ nuclear populations. e Dot plot depicts selected genes differentially expressed in various myonuclear populations. Dot size represents the percentage of nuclei expressing the gene. Source data are provided as a Source Data file.

Fig. 5. smRNA-FISH analyses validates gene expression in newly fused and pre-existing myonuclei.

a Representative images of smRNA-FISH on sections from Pax7^rtTA; TRE^H2B-GFP muscle for selected genes enriched in newly fused (H19, Igf2, Tnnt1, Tnnt2) or pre-existing myonuclei (Atf3). Quantification of the signal for each gene is shown on the right. (n = 3–5 independent animals). Scale bar, 20 μm. b Representative images of smRNA-FISH for the indicated genes on sections from control (Mymk^loxP/loxp + tamoxifen) or Mymk^scKO mice subjected to muscle overload. Quantification of the signals are shown on the bottom. (n = 3–6 independent animals). Scale bar, 20 μm. c Representative images of smRNA-FISH for Myh3, Tnnt1, and Tnnt2 on sections from Pax7^rtTA; TRE^H2B-GFP muscle during postnatal development. (n = 3). Scale bar, 20 μm. Data are presented as mean ± SD. Statistical tests used were (a) one-way ANOVA with Tukey’s correction for multiple comparisons; (b) two-tailed unpaired t-test; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Source data are provided as a Source Data file.

Fig. 6. Comparative gene features of newly fused myonuclei.

a Representative images of smRNA-FISH for Myh3 on plantaris myofibers from Pax7^rtTA; TRE^H2B-GFP mice undergoing overload stimulation or under sham conditions. Myh3 transcription was detected in both peripheral (white arrow) and central (arrowhead) newly fused myonuclei. (n = 3–4). Scale bar, 20 μm. b Representative images of smRNA-FISH for Myh3, performed on the same mice used in (A), show that both central newly fused GFP⁺ myonuclei (arrowhead) and peripheral newly fused GFP⁺ (white arrow) are enriched for Myh3 transcription. (n = 3–4). Scale bar, 20 μm. c Heatmap of genes showing distinct transcriptional signature in nuclei of myogenic populations from plantaris muscle subjected to 1 week of overload stimulation, as used in Fig. 4c (n = 2).

Fig. 7. Fusion and myonuclear accrual is needed for pre-existing myonuclei to transcriptionally respond to muscle overload.

a Schematic diagram of the bulk nuclei RNA-seq methodology for analyzing pre-existing myonuclei adapting to muscle overload (MOV) in the absence of myonuclear accretion. Myomaker^loxP/loxP; Pax7^CreER mice were treated with tamoxifen to generate fusion-incompetent mice (Mymk^scKO). Note that three sham mice were Pax7^rtTA; TRE^H2B-GFP (+Dox) and one sham animal was Mymk^loxP/loxp (+tamoxifen). b Principal component analysis of bnRNA-seq from Mymk^scKO myonuclei isolated from plantaris muscle 1 week after muscle overload, myonuclei (GFP⁺ and GFP⁻) derived from control mice that underwent the same surgery, and myonuclei from sham mice. c Scatter plot highlighting genes upregulated in both control muscle overload (MOV) (GFP⁻ pre-existing myonuclei) and Mymk^scKO MOV myonuclei (peach dots, fusion independent) or upregulated in only GFP⁻ pre-existing myonuclei (maroon dots, fusion dependent). Gene Ontology analysis of the genes upregulated that are fusion-dependent (maroon dots) is shown on the right. d Scatter plot highlighting genes downregulated in both control muscle overload (MOV) (GFP⁻ pre-existing myonuclei) and Mymk^scKO MOV myonuclei (green dots) or downregulated in only Mymk^scKO MOV myonuclei (pink dots). Gene Ontology analysis of both sets of genes is shown on the right.