Inactivation of Sirt6 ameliorates muscular dystrophy in mdx mice by releasing suppression of utrophin expression

The NAD+-dependent SIRT1-7 family of protein deacetylases plays a vital role in various molecular pathways related to stress response, DNA repair, aging and metabolism. Increased activity of individual sirtuins often exerts beneficial effects in pathophysiological conditions whereas reduced activity is usually associated with disease conditions. Here, we demonstrate that SIRT6 deacetylates H3K56ac in myofibers to suppress expression of utrophin, a dystrophin-related protein stabilizing the sarcolemma in absence of dystrophin. Inactivation of Sirt6 in dystrophin-deficient mdx mice reduced damage of myofibers, ameliorated dystrophic muscle pathology, and improved muscle function, leading to attenuated activation of muscle stem cells (MuSCs). ChIP-seq and locus-specific recruitment of SIRT6 using a CRISPR-dCas9/gRNA approach revealed that SIRT6 is critical for removal of H3K56ac at the Downstream utrophin Enhancer (DUE), which is indispensable for utrophin expression. We conclude that epigenetic manipulation of utrophin expression is a promising approach for the treatment of Duchenne Muscular Dystrophy (DMD).


Supplementary Information Page
Relative numbers of Tunel + /PAX7 + MuSCs are shown on the right. (Unpaired two-tailed t-test: *p=0.0319; n=5 for each group). b EdU incorporation in PAX7 + MuSCs of TA muscles from control and mdx mice. Scale bar: 20μm. Quantifications are shown on the right (Unpaired twotailed t-test: ***p=0.0004; n=3 for each group). c Heat map of RNA-seq data visualizing upregulation of genes involved in cellular responses to stress [http://www.informatics.jax.org/go/term/ GO:0033554] in mdx compared to the control MuSCs (n=2 for each group). A list of gene names shown in the heatmap is provided in Supplementary table 1. 12-weeks old male mice were used. Data are presented as mean ± SEM (a-b). Source data are provided in the Source Data file.

Supplementary Figure 2.
Sirt6 is dispensable for skeletal muscle development and skeletal muscle homeostasis. a Schematic representation of wild-type and conditional alleles of the Sirt6 gene. Primers for genotyping are indicated by black arrows. P1/P2 primers were used to detected wild-type and floxed allele. P1/P3 primers were used to detect the deleted allele. b Genotyping by PCR to detect wild-type, floxed and deleted alleles in FACS-sorted MuSCs. c Volcano plot of differentially expressed genes in Sirt6 mKO compared to freshly isolated control MuSCs. d Top 5 biological processes (BP) upregulated in Sirt6 mKO versus wild type MuSCs. e Heat map of RNA-seq data visualizing upregulation of genes in Sirt6 mKO MuSCs contained in top 5 GO-terms (n=3). Benjamini-Hochberg test, two-sided. f Images of 12-week-old control (Sirt6 fl/fl Pax7ICN neg ) and Sirt6 mKO (Sirt6 fl/fl Pax7ICN pos ) mice. g Body weight of 12-week-old control and Sirt6 mKO male mice (Unpaired two-tailed t-test: ns p >0.05; n=6). h TA muscle weight of 12-weeks-old control and Sirt6 mKO male mice (Unpaired two-tailed t-test: ns p >0.05; n=5 for each group). i Immunofluorescence staining of PAX7+ positive MuSCs in control and Sirt6 mKO TA muscle cryo-sections. Scale bar: 20 µm. Numbers of MuSC per mm 2 section area are shown in the right panel (n=3). j Ratios of TA muscle masses after CTX injection compared to untreated samples dissected from control (n=5) or Sirt6 mKO mice (n=4) (Unpaired two-tailed t-test ns: p>0.05). Top panel: scheme to induce muscle regeneration in control and Sirt6 mKO littermates. k H&E staining of TA muscles in control (n=5) and Sirt6 mKO (n=4) mice after muscle injury. Scale bar: 20 μm. l Schematic outline of the repeated muscle regeneration assay. Ratios of TA muscle masses with and without CTX injection are shown (Unpaired two-tailed t-test: *p =0.0389 n = 4 for each group). m H&E staining of TA muscles from control and Sirt6 mKO mice after repeated injury (n=4 for each group). Scale bar: 20 μm. n Glucose tolerance test (GTT) of control (n=4) and Sirt6 mKO mice (n=3). Two-way ANOVA test with Bonferroni multiple comparisons test: ns: p>0.05. o Expression of selected glycolytic genes in freshly isolated MuSCs from control and Sirt6 mKO mice. Results were obtained by RNAseq and are presented as RPKM (Reads per kilobase of transcript per Million mapped reads) (Unpaired twotailed t-test: *p=0.0139, n=3). p RT-qPCR expression analysis of glycolytic genes in muscles from control and Sirt6 mKO mice. m36b4 was used as a reference (Unpaired two-tailed t-test: p>0.05, **p =0.0048; n=3 for each group). 12-weeks old male mice were used. Data are presented as mean ± SEM. Source data are provided in the Source Data file. Figure 3. Loss of Sirt6 attenuates persistent activation of MuSCs in mdx muscles. a RT-qPCR analysis of Sirt6 expression in MuSCs isolated from control (n=7), mdx (n=8) and Sirt6 mKO /mdx (n=4) mice. m36b4 was used as a reference gene (One-way ANOVA test with Bonferroni multiple comparisons test: *p=0.024, **p=0029, ****p<0.0001.) b RT-qPCR analysis of Myod expression in freshly isolated MuSCs from control (n=7), mdx (n=9) and Sirt6 mKO /mdx (n=4) mice. m36b4 was used as a reference gene (One-way ANOVA test with Bonferroni multiple comparisons test: *p=0.0197, ***p<0.0007). c-e Heat map of genes associated with cell cycle (c), activation markers of MuSCs (d) and cellular stress response (e) in control, mdx and Sirt6 mKO /mdx MuSCs that were up-regulated in mdx MuSCs but normalized to control level in Sirt6 mKO /mdx MuSCs (n=2). f Venn diagram to identify the overlap of upregulated genes in mdx MuSCs and in Sirt6 mKO /mdx compared to control MuSCs. g Venn diagram to identify the overlap of upregulated genes in Sirt6 mKO MuSCs and genes that are only upregulated in Sirt6 mKO /mdx but not in mdx MuSCs. h Western blot analysis of H3K56ac, H3K9ac, and H3K18ac levels in freshly isolated MuSCs from control, mdx and Sirt6 mKO /mdx mice. Histone H3 was used as a loading control. Quantification of protein bands is shown is in the lower panel (One-way ANOVA test with Bonferroni multiple comparisons test: ****p < 0.0001; n=3 for each group). i Venn diagram to identify the overlap of downregulated genes in mdx MuSCs and in Sirt6 mKO /mdx MuSCs compared to control MuSCs. 12-weeks old male mice were used. Data are presented as mean ± SEM (a, b, h). Source data are provided in the Source Data file. Figure 4. Inactivation of Sirt6 improves functions of skeletal muscles in mdx mice without apparent changes in muscle metabolism. a MRI of fat mass in 12-weekold control (n=7), mdx (n=8) and Sirt6 mKO /mdx (n=8) littermates. Quantifications of muscle and fat volume are shown in the right panel (One-way ANOVA test with Bonferroni multiple comparisons test: **p=0.0038;). b Lectin staining of TA muscle sections from control, mdx and Sirt6 mKO /mdx mice. Quantification of CSA of myofibers is shown in the right panel (Two-way ANOVA with Benjamini and Yekutieli multiple comparisons test: ns p>0.05, *p=0.0197, **p=0.0053; n=3 for each group). Scale bar: 20 µm. c Evans blue staining of muscles from control, mdx and Sirt6 mKO /mdx mice. The macroscopic view displays diaphragm muscles 16 hours after 1% Evans blue injection. Scale bar: 5 mm. d, e Food intake (d) and water intake (e) of control (n=6), mdx (n=9) and Sirt6 mKO /mdx (n=9) mice after 48 h acclimatization in PhenoMaster metabolic cages (One-way ANOVA test with Bonferroni multiple comparisons test: ns p>0.05). f Respiratory exchange ratio (RER) of control (n=3), mdx (n=7) and Sirt6 mKO /mdx (n=7) mice after 48 h acclimatization in PhenoMaster analysis metabolic cages.

Supplementary
12-16-weeks old male mice were used. Data are presented as mean ± SEM (a-b, d-f). Source data are provided as a Source Data file.

Supplementary Figure 5. SIRT6 regulates H3K56ac levels in enhancer regions of ESCs. a
Venn diagram of H3K56ac, H3K27ac (ENCODE), H3K4me1 (ENCODE) ChIP-seq data in mESCs showing that H3K56ac peaks in mESCs overlap with enhancer marks (H3K27ac and H3K4me1). b Box plot of log2-fold change (FC) representing all H3K56ac genome wide (GM) peaks and significantly deregulated (DE) H3K56ac peaks in Sirt6 KD mESCs. c Distribution of H3K56ac ChIP-seq peaks across genomic regions. d Box plots indicating that H3K56ac sites induced in Sirt6 KD mESCs correlate stronger with enhancer marks (H3K27ac, H3K4me1, and p300 (ENCODE data sets) than all H3K56ac peaks (Wilcoxon rank sum test, p<2.2x10 -16 ). e Box plots demonstrating that H3K56ac sites induced in Sirt6 KD mESCs are not correlated with promoter-associated marks such as H3K4me3 and H3K9ac and only mildly enriched at gene bodies marked by H3K36me3 (ENCODE data sets). Box plots in (b, d and e) were defined by default in R, 1.5×IQR (interquartile range). The whiskers show the minimum and maximum values. Wilcoxon rank sum test, two-sided p<2.2x1016 (n=2).

Supplementary Figure 6. SIRT6 represses activity of the DUE regulating Utrn gene expression. a Western blot analysis of utrophin expression in muscle tissues from control and
Sirt6 mKO mice. Vinculin was used as loading control. Quantification of protein bands are shown on the right (Unpaired two-tailed t-test: ***p=0.007, n=3 for each group) b RT-qPCR analysis of Utrn gene expression in freshly isolated MuSCs from control (n=7), mdx (n=12) and Sirt6 mKO /mdx (n=6) mice. m36b4 was used as a reference gene. (One-way ANOVA test with Bonferroni multiple comparisons test: ns p>0.05) c RT-qPCR analysis of Utrn gene expression in freshly isolated control (n=5) and Sirt6 mKO (n=4) MuSCs. m36b4 was used as reference gene (Unpaired two-tailed t-test: ns p>0.05). d Distribution of H3K56ac, H3K4me3 and H3K27ac in the proximal promoter region and DUE of the Utrn gene, revealing co-localization of increased H3K56ac peaks in Sirt6 mKO MuSCs with the enhancer mark H3K27ac. e BLASTN sequence alignment of DUEs in human (upper) and mouse (lower) genomes. The putative AP-1 (red line) binding site is 100% conserved. 12-16-weeks old male mice were used. Data are presented as mean ± SEM (a-c). Source data are provided in the Source Data file. Figure 7. Inactivation of Sirt6 in mdx mice results in increased Mstn expression in skeletal muscles. a Venn diagram of overlapping downregulated genes in mdx compared to control muscle and upregulated genes in Sirt6 mKO /mdx compared to mdx muscle based in RNA-seq. b Heat map showing normalization of Mstn and its receptors in Sirt6 mKO /mdx back to wild-type level when compared to mdx muscles (n=3) based on DESeq normalized counts. c RT-qPCR analysis of Mstn expression in muscles from control (n=4), mdx (n=4) and Sirt6 mKO /mdx (n=3) mice (Unpaired two-tailed t-test: *p=0.0310, *p=0.0157, **p=0.0017). m36b4 was used as reference gene. d ChIP-qPCR analysis of H3K56ac in mdx (n=8) and Sirt6 mKO /mdx (n=6) muscles and H3K27ac at enhancer/promoter regions of the Mstn gene in mdx (n=4) and Sirt6 mKO /mdx (n=3) muscles. Enrichment of H3K56ac and H3K27ac was normalized to H3. (Unpaired two-tailed t-test: *p=0.0474, **p=0.0062). e Methylationsensitive PCR to detect DNA methylation at the Mstn promoter in muscles of mdx and Sirt6 mKO /mdx (Two-way ANOVA test with Bonferroni multiple comparisons test: ** p=0.0029; n=3). 12-16-weeks old male mice were used. Data are presented as mean ± SEM (c-e). Source data are provided in the Source Data file.