Runx1 Transcription Factor Is Required for Myoblasts Proliferation during Muscle Regeneration

Abstract

Following myonecrosis, muscle satellite cells proliferate, differentiate and fuse, creating new myofibers. The Runx1 transcription factor is not expressed in naïve developing muscle or in adult muscle tissue. However, it is highly expressed in muscles exposed to myopathic damage yet, the role of Runx1 in muscle regeneration is completely unknown. Our study of Runx1 function in the muscle's response to myonecrosis reveals that this transcription factor is activated and cooperates with the MyoD and AP-1/c-Jun transcription factors to drive the transcription program of muscle regeneration. Mice lacking dystrophin and muscle Runx1 (mdx-/Runx1f/f), exhibit impaired muscle regeneration leading to age-dependent muscle waste, gradual decrease in motor capabilities and a shortened lifespan. Runx1-deficient primary myoblasts are arrested at cell cycle G1 and consequently differentiate. Such premature differentiation disrupts the myoblasts' normal proliferation/differentiation balance, reduces the number and size of regenerating myofibers and impairs muscle regeneration. Our combined Runx1-dependent gene expression, ChIP-seq, ATAC-seq and histone H3K4me1/H3K27ac modification analyses revealed a subset of Runx1-regulated genes that are co-occupied by MyoD and c-Jun in mdx-/Runx1f/f muscle. The data provide unique insights into the transcriptional program driving muscle regeneration and implicate Runx1 as an important participant in the pathology of muscle wasting diseases.

Fig 1. Runx1 expression in response to muscle damage.

(A to D). IHC using anti- Runx1 Ab of gastrocnemius muscle from mice subjected to muscle stress. Runx1-positive cells show brown nuclear staining, scale bars, 50 μm. (A) Untreated WT mice. (B) 120 days-old tg-mSOD1 mice. (C) CTX treated WT mice. (D) 2 month old mdx mice. (E) Runx1 and Pax7 IF analysis of CTX-treated WT muscle, scale bars, 50 μm. White arrowheads indicate Runx1⁺/Pax7⁺ cells. (F) IF analysis of cultured proliferating PM using anti- Runx1 and MyoD Abs. DAPI staining was used as a nuclear marker, and myoblasts were visualized by differential interface contrast (DIC) microscopy, scale bars, 50 μm. Results from one of four different experiments with similar findings are shown.

Fig 2. Loss of Runx1 in mdx mice decreases muscle mass, muscle strength and lifespan.

(A) Scatter plot showing weight of Runx1 ^L/L (WT), Runx1 ^f/f, mdx and mdx/Runx1 ^f/f mice between 2–9 months of age (average ±SD, n = 9–28, **P<0.01). (B) Representative image of mdx and mdx/Runx1 ^f/f littermates at 7 months of age. (C) Dot plot depicting the average lean weight (as % of total body weight) of mdx and mdx/Runx1 ^f/f mice between 4–9 months of age. mdx = open circles, and mdx/Runx1 ^f/f = open squares. Mean lean weight is indicated (n = 6–22, **P<0.01, ***P<0.0001). (D) Kaplan- Meyer survival curve of Runx1 ^L/L (n = 50, blue), Runx1 ^f/f (n = 50, red), mdx (n = 46, green) and mdx/Runx1 ^f/f (n = 55, purple) (***P<0.0001). (E) Diaphragm muscle sections of mdx and mdx/Runx1 ^f/f mice stained with H&E (top) or Sirius Red (bottom) for collagen (Fibrosis), shown at x100 (left panels) or x400 (right panels). Scale bars, 200μm and 50μm for the X100 and X400 magnifications, respectively. (F) Histogram summarizing treadmill performance of mice between 2–7 months of age. Runx1 ^L/L, Runx1 ^f/f, mdx and mdx/Runx1 ^f/f mice were subjected to an exhaustion protocol (Average ±SD, n = 5–21, **P<0.01, Bonferroni corrected post-hoc comparisons). (G) 4 months old mdx and mdx/Runx1 ^f/f mice were subjected to grip strength measurements. Left and right histograms show absolute and normalized (to body weight) force comparison respectively. Values are mean ±SEM (n = 9–14 *P<0.05, **P<0.01).

Fig 3. Loss of Runx1 in mdx mice resulted in reduced muscle regeneration.

(A and B) Determination of myofibers numbers in mdx muscle. Soleus muscles from 8 weeks old mdx and mdx/Runx1 ^f/f mice were sectioned, subjected to H&E staining, and number of total and regenerating myofibers was determined. (A) Representative images of mdx and mdx/Runx1 ^f/f sections showing regenerating myofibers with central nuclei, the hallmark of regenerating myofibers, shown at x100 (top) or x200 (bottom). Scale bars, 200μm and 100μm for the x100 and x200 magnification, respectively. (B) Stacked column histograms showing the average number of regenerating myofibers (red) and normal myofibers (blue) in mdx and mdx/Runx1 ^f/f soleus muscle sections. The number of regenerating (fibers with round and central nuclei) and normal myofibers was counted in 3 H&E-stained sections per muscle and their average number calculated. Values are mean±SE (n = 9–13, ***P <0.001, unpaired student t-test). WT myofibers number is given as negative control for the presence of regenerating myofibers. (C and D) Average CSA of total or regenerating myofibers from mdx/Runx1 ^L/L (filled and open squares) and mdx/Runx1 ^f/f (filled and open triangles) of 8 weeks old mice was determined for soleus (C) and gastrocnemius (D) muscles. Values are mean± SEM (n = 7 mdx/Runx1 ^L/L, n = 6 mdx/Runx1 ^f/f;*P<0.05, unpaired student t-test). (E and F) Quantification of CSA distribution of total or regenerating myofibers in mdx/Runx1 ^L/L and mdx/Runx1 ^f/f: the percentage of mdx/Runx1 ^L/L total myofibers (filled blue columns), mdx/Runx1 ^L/L regenerating myofibers (filled red columns), mdx/Runx1 ^f/f total myofibers (empty blue columns) and mdx/Runx1 ^f/f regenerating myofibers (empty red columns) was determined for soleus (E) and gastrocnemius (F) muscles. Values are mean ± SEM (n = 7 mdx/Runx1 ^L/L, n = 6 mdx/Runx1 ^f/f; *, P < 0.05; **, P < 0.01, unpaired student t-test).

Fig 4. Runx1 attenuates PM proliferation.

(A to F) Runx1 ^L/L and Runx1^f/f PM were purified and their proliferation properties were compared. (A) Average doubling time of Runx1 ^L/L and Runx1^f/f PM cultures. Values are mean±SD (n = 4, **p <0.001). (B and C) Cell cycle analysis of proliferating PM derived from Runx1 ^L/L (B), or Runx1 ^f/f (C) mice. Cell-cycle phases G1, S, and G2/M and the relative size (in %) of PI labeled populations out of total cells are indicated. Results from one of four Runx1 ^L/L or Runx1 ^f/f different cultures with similar findings are shown. Green and red arrows indicate increase in % of G1 and decrease in % of S and G2/M of Runx1 ^f/f vs. Runx1 ^L/L PM. (D) Histograms summarizing the distribution of cell populations as analyzed in C. Values are mean±SD (n = 4, *p <0.05). (E) IF analysis of proliferating PM from Runx1^L/L and Runx1 ^f/f mice using anti-Runx1 and MHC Abs. (I-IV) Runx1 ^L/L and (V-VIII) Runx1 ^f/f at x200 magnification, scale bars, 50 μm. Results from one of four Runx1 ^L/L or Runx1 ^f/f different cultures with similar findings are shown. (F) Average fusion index of proliferating PM. Runx1 ^L/L and Runx1 ^f/f proliferating PM cultures were stained with anti-MHC Ab and DAPI and the fractions (in %) of single (blue), double (red) and multinucleated (≥ 3, green) cells are shown. Values are mean±SE (n = 4, *p <0.05). (G to J) Proliferating WT PM were infected with either Ad5CMV-eGFP or Ad-Runx1 and then grown for 24 h in differentiation medium prior to analysis. (G) IF analysis of infected PM using anti- Runx1 and MHC Abs (scale bars, 50 μm and 20 μm for x200 or x630 magnification, respectively). DAPI was used as a nuclear marker. Results from one of four different experiments with similar findings are shown. (H) Histograms showing the average fusion index of differentiating PM analyzed in (G). The fractions (in %) of single (blue), double (red) and multinucleated (≥ 3, green) cells are shown. Values are mean±SE (n = 4, **p <0.001, *p <0.05). (I and J) RT-qPCR analysis of myogenic gene expression in proliferating PM (Pro) before or 72 h post differentiation induction (Diff). PM were grown and infected as indicated in (G), RNA was purified and analyzed by TaqMan assay. Values are mean±SD (n = 3, **p <0.001).

Fig 5. Analysis of PM high confidence Runx1-regulated genes.

(A) Schematic representation of the selection procedures used to identify high-confidence Runx1-regulated genes. Each cylinder represents a gene subset, with the gene number given in brackets. I- Runx1-responsive genes were derived from Runx1 ^L/L vs. Runx1 ^f/f PM microarray expression data. II- Runx1-regulated genes were derived by cross analysis of the Runx1-responsive genes dataset and Runx1 ChIP-seq data. This gene subset represents Runx1-responsive genes that are also occupied by Runx1. III- RMJ-regulated genes are Runx1-responsive genes that are co-occupied by Runx1, MyoD and c-Jun. IV- High-confidence Runx1-regulated gene subset are RMJ-regulated genes that were also marked as having adjacent active regulatory elements by both anti histone modifications (H3K4me1 & H3K27ac) ChIP-seq and ATAC-seq. (B) Scatter plot of differentially expressed genes in WT vs. Runx1 ^f/f- PM. Gene expression level (log2 scale) in Runx1 ^f/f vs. WT PM is plotted. Significant increased or decreased genes are indicated in red or green, respectively. Filled circles indicate Runx1-responsive genes that are known to participate in myoblast proliferation or differentiation. (C) Pie chart depicting Runx1 binding sites distribution in relation to the nearest annotated TSS. Numbers represent % of bound regions. (D) Venn diagram summarizing the overlap between Runx1-ChIP-seq bound genes (ChIP) and Runx1-responsive genes, differentially expressed in Runx1 ^f/f vs. Runx1 ^L/L. Runx1-regulated genes are defined as Runx1-bound genes that were also Runx1-responsive. (E) Enriched TF motifs among Runx1-bound regions from PM ChIP-seq data. (F) Overrepresented TF modules in Runx1-bound regions from PM. Runx1 ChIP-seq data was analyzed using the module overrepresentation tool in Genomatix package (RegionMiner). The table presents the most highly enriched modules. (G) Venn diagram showing the overlap of regions bound by Runx1, MyoD and c-Jun and the common fraction of 11629 regions. (H) Cross analysis of all ChIP seq and ATAC-seq common loci with Runx1-responsive gene list (Fig 5B). Prominent genes are presented.

Fig 6. Validation of in vivo high confidence Runx1-regulated genes.

(A) Volcano plot of differentially expressed genes in soleus muscle of 8 weeks old mdx/Runx1 ^f/f vs. mdx mice. Fold expression change against p value is plotted. Significant increased or decreased genes are indicated in red or blue, respectively. Filled triangles indicate Runx1-responsive genes that are known to participate in myoblast proliferation or differentiation. (B) Venn diagram summarizing the overlap between mdx Runx1- responsive (RNA-seq) and PM RMJ- regulated gene. These genes are defined as high confidence Runx1- regulated genes in mdx myoblasts. (C to E) UCSC genome browser screenshots showing ChIP-Seq performed in PM and mdx/Runx1 ^f/f vs. mdx mice RNA- seq tracing examples of high-confidence Runx1-regulated genes. Expression of these genes was quantified by RT-qPCR of cultured Runx1-deficient or-over expressing PM, and in vivo in mdx/Runx1 ^f/f vs. mdx muscles. Values are mean±SD (n = 3). (C) Myog, encoding Myogenin (**p<0.001, *p <0.05). (D) Dlk1, encoding Delta-like 1 homolog (**p <0.001). (E) Mstn, encoding Myostatin (**p <0.001).

Fig 7. Runx1 is required for myoblast proliferation during muscle regeneration.

Schematic diagram summarizing the scenario of Runx1-regulated myoblast proliferation during muscle regeneration: (A) Following myonecrosis of WT muscle, SC are activated, Runx1 is induced and promote proliferation and prevents premature differentiation. Once the critical mass of myoblasts is reached Runx1 is destined to degradation, myoblasts differentiate to produce normal size myofibers. (B) In Runx ^f/f muscles, myoblasts lack Runx1 expression and therefore undergo premature differentiation. This leads to insufficient myoblast pool size, resulting in reduced number and size of myofibers and impaired muscle regeneration.