Linc‐MYH configures INO80 to regulate muscle stem cell numbers and skeletal muscle hypertrophy

Figures

• 

  Figure 1. linc‐MYH is located in nuclei of proliferating and differentiated myocytes and interacts with the chromatin remodeler complex INO80 ARNA‐seq. of freshly isolated, proliferating, and differentiating satellite cells reveals an increased expression of linc‐MYH in proliferating and differentiating MuSCs, but no expression is found in freshly isolated MuSCs.B, CExpression of linc‐MYH is confined to skeletal muscle in mouse (B) and human (C) tissues.DExpression of linc‐MYH is not detected in limb buds at E10.5 (n = 3), E13.5 (n = 3), and E15.5 (n = 3). Expression of linc‐MYH in hindlimb muscle of E18.5, newborn, and 7‐day-old animals (n = 3/3/3). Expression of linc‐MYH in m. soleus (soleus), m. tibialis anterior (TA), and m. extensor digitorum longus (EDL) muscle at 3 weeks (n = 3) and 8 weeks of age (n = 5). Biological replicates were used for all PCR experiments, and data are mean ± SEM.E‐HMurine linc‐Myh is localized in nuclear extracts of C2C12 cells. Xist (nuclear) and Gapdh (mainly cytoplasmic) were used as controls (n = 2 nuclear/3 cytoplasmic biological replicates.) Data are mean ± SEM. (F‐H) RNA–FISH identifies linc‐Myh in nuclei (DAPI, blue) of C2C12 myoblasts. (G) Adipor…

• 

  Figure EV1. The linc‐MYH gene is located in the Myh3‐Myh13 gene cluster and generates two different splice variants AThe linc‐MYH gene (RefSeq ID: 2310065F04Rik; marked by red square) is located within the Myh3‐Myh13 gene cluster at mouse chromosome 11, downstream of Myh3, and upstream of Myh2 (UCSC genome browser, GRCm38/mm10).BThe linc‐MYH gene expresses two lncRNA splice variants (UCSC genes: AK010044, AK079404) that share most of the primary sequence. RNA‐seq (track: RNA‐seq from ENCODE/Caltech; C2C12 myocytes) and microarray expression data (track: Affymetrix Exon Array 1.0: normal tissues) indicate expression in skeletal muscle tissue (red bars) but not in the heart or other organs (black bars).C, DThe human genome (UCSC genome browser, GRCh37/hg19) contains an ortholog of linc‐MYH, AK097500. (D) Analysis of published human RNA‐seq data (ENCODE/Caltech) indicating that only the last five exons of AK097500 are present in human myoblasts, similar to the mouse linc‐MYH gene.E, Flinc‐MYH has a low coding potential. Comparison of the coding potential of different non‐coding RNAs (Tsix, Fendrr, Hotair), “micropeptide‐coding” RNAs (Myoregulin, Dworf), coding RNAs (Myog, Gapdh) RNAs and…

• 

  Figure 2. linc‐MYH keeps the number of MuSCs in check and limits muscle fiber growth A, BBody weight (A) of male linc‐MYH KO and ctrl mice at 10 weeks (n = 24 KO/17 WT; Student's t‐test, two‐tailed, **P < 0.01) and tibia length (B) of male linc‐MYH KO and ctrl mice of 10 weeks (n = 9 KO/13 WT; Student's t‐test, two‐tailed, not significant). Data are mean ± SEM.C, DWeight of m. tibialis anterior (TA) and m. digitorum longus (EDL) of male linc‐MYH KO and ctrl mice at 10 weeks (n = 9 KO/10 WT for TA, n = 10 KO/11 WT for EDL; Student's t‐test, two‐tailed, **P < 0.01, ***P < 0.001). Data are mean ± SEM.E–JSize distribution of fibers in cross sections of TA (E, F) and EDL (G, H) muscle in male linc‐MYH KO (blue) and ctrl (gray) animals at 10 weeks. A significant increase in fiber size was observed for TA and EDL muscles of linc‐MYH KO mice (n = 5 KO/4 WT for TA, n = 3 KO/3 WT for EDL muscle, two‐way ANOVA with Fisher's multiple comparisons; *P < 0.05, **P < 0.01; > 250 fibers per animal were counted). Data are mean ± SEM.KNumbers of myonuclei in cross sections of TA and EDL muscle (TA: n = 6 KO/6 WT, EDL n = 3 KO/3 WT animals, Mann–Whitney test one‐tailed, **P < 0.01, *P = 0.05). Data are mean ±…

• 

  Figure 3. linc‐Myh restricts proliferation of MuSCs in an INO80‐dependent manner A–DRNA in situ hybridization‐proximity ligation assay (rISH‐PLA) detects the close proximity of a specific RNA with proteins in situ modified from Roussis et al (B, C) rISH‐PLA confirms the proximity of linc‐MYH to endogenous INO80‐V5 in the nucleus of proliferating WT MuSCs (B). The specific signal is lost in linc‐MYH-deficient MuSCs (C). (n = 3 WT/3 KO biological replicates; Mann–Whitney test, *P = 0.05). Data are mean ± SEM.E–IRepresentative images of TA muscle sections of 10‐week-old mice stained for PAX7 (green) and EdU incorporation (red). MuSCs without (green arrows) and with EdU‐labeling (red arrows) are indicated. (E, F) Deletion of linc‐MYH results in increased numbers of muscle stem cells and pronounced increase in the ratio of EdU‐positive MuSCs. (G‐I) The increase in both the number of MuSCs and EdU‐positive MuSCs in linc‐MYH mutant muscle is abolished by constitutive (G: Pax7‐Crepos/INO80−/−, Ino80 KO; H: linc‐MYH−/−/Pax7-Crepos/INO80−/−, dKO) and induced (I: linc‐MYH−/−/Pax7-CreERT2pos/INO80−/−, dKO CreERT) deletion of INO80 in Pax7 expressing cells.J–LStatistical evaluation of the number of…

• 

  Figure EV2. Loss of linc‐MYH does not affect skeletal muscle regeneration A–CFiber size distribution in TA muscles of WT and linc‐MYH−/− (KO) at different time point after CTX injection. The cross‐sectional area of newly formed myofibers does not differ after one (A) or 2 weeks (B) of regeneration (two‐way ANOVA with Fisher's LSD: not significant). 4 weeks after CTX injury (C), increased numbers of myofibers with large cross‐sectional area are present in linc‐MYH−/− TA muscles (two‐way ANOVA with Fisher's LSD for column 3,000 μm²: **P < 0.003 and significant with Sidak–Holm correction; each timepoint: WT n = 4 animals, linc‐MYH KO n = 4 animals, > 297 myofibers/animal). Data are mean ± SEM.D–FImmunofluorescence images of muscle cross sections stained for WGA (green) and DAPI (blue). Centralized nuclei 2 weeks after CTX injury indicate newly formed myofibers. F, The proportion of fibers with centralized nuclei is similar between WT and linc‐MYH mutant muscle (Mann–Whitney test: not significant P = 0.999; WT n = 3 animals, linc‐MYH KO n = 3 animals, 2 weeks after injury). Data are mean ± SEM.G–IThe number of Pax7‐positive (green) MuSCs does not differ between WT (G) and linc‐MYH−/− muscles (H)…

• 

  Figure EV3. No apparent differences in MuSCs of 3‐week‐old animals AThe number of myonuclei in TA muscle is not significantly different between the WT and linc‐MYH KOs (Mann–Whitney test: not significant P = 0.7; WT n = 3 animals, linc‐MYH KO n = 3 animals). Data are mean ± SEM.B–EThe number of Pax7‐positive (green arrows in D, E) MuSCs in TA muscle does not differ between 3‐week-old linc‐MYH KO (E) compared with wild‐type (D) control animals (Mann–Whitney test: not significant P = 0.99; WT n = 3 animals, linc‐MYH KO n = 3 animals). The percentage of EdU‐positive (red arrows in D, E) MuSCs in TA muscle of linc‐MYH KO (E) compared with wild‐type controls (D) is not different in 3‐week-old animals (Mann–Whitney test: not significant P = 0.99; WT n = 3 animals, linc‐MYH KO n = 3 animals). Data are mean ± SEM.

• 

  Figure 4. Deletion of INO80 prevents linc‐MYH‐dependent changes in MuSC proliferation and gene expression A–DIsolated linc‐MYH KO MuSCs reach confluence much faster than WT MuSCs indicating higher proliferation rates. Deletion of INO80 eliminates the difference between Pax7‐Crepos/INO80−/− (Ino80 KO) and linc‐MYH−/−/linc-MYH−/−/Pax7‐Crepos/INO80−/− (dKO). Images are representative images for cultures after 13 days in culture.EDifferences in proliferation rates are abolished by additional deletion of INO80 in MuSCs. Cell confluence was monitored by a live cell imaging analysis system for 13 days (n = 7 WT/6 linc‐MYH KO/3 Ino80 KO/3 dKO‐independent preparations of MuSCs).FStatistical analysis of cell confluence after 13 days of muscle stem cell proliferation (n = 7 WT/6 linc‐MYH KO/3 Ino80 KO/3 dKO‐independent preparations of MuSCs, Student's t‐test, two‐tailed; *P < 0.05, ns = not significant). Data are mean ± SEM.G–IGene set enrichment analysis (GSEA, broadinstitute.org) of microarray data from proliferating MuSCs. Comparison of linc‐MYH−/− vs. WT proliferating MuSC transcriptomes shows enrichment of genes related to cell proliferation (G), enrichment of genes containing cis regulatory…

• 

  Figure EV4. Knock‐down of YY1 in linc‐MYH KO MuSCs Knock‐down of Yy1 in proliferating MuSCs in vitro reduces Yy1 transcript abundance in linc‐MYH KO and wild‐type controls (WT scramble shRNA n = 3 biological replicates, WT Yy1 knock‐down n = 3 biological replicates, Mann–Whitney test: significant *P = 0.05; linc‐MYH KO scramble shRNA n =3 biological replicates, linc‐MYH KO yy1 shRNA n = 3 biological replicates, Mann–Whitney test: significant *P = 0.05). Data are mean ± SEM.Knock‐down of Yy1 in linc‐MYH KO MuSCs compared with linc‐MYH KO scramble shRNA controls results in lower confluence of proliferating MuSCs (n = 3/3 with 4 technical replicates/animal for each transfection, 95 time points, two‐way ANOVA P = 0.0001) compared with their scramble controls. Similarly, knock‐down of Yy1 in WT MuSCs compared with WT scramble shRNA results in lower confluence of proliferating MuSCs. (n = 3/3 with 4 technical replicates per animal for each transfection, 95 time points, two‐way ANOVA P = 0.0001). Knock‐down of Yy1 eliminated the increased confluence of linc‐MYH KO MuSC observed in linc‐MYH KO MuSCs compared to the respective wild‐type control MuSCs.

• 

  Figure 5. linc‐MYH does not regulate the function of INO80 during replication stress A–Dγ‐H2AX staining of proliferating MuSCs from WT (A), linc‐MYH-deficient (B), Pax7‐Crepos/INO80 KO (C), and linc‐MYH/INO80 dKO (D) mice. Deletion of INO80 results in increased γ‐H2AX signals, while deletion of linc‐MYH has no effects (asterisk: low γ‐H2AX signals, arrow: γ‐H2AX-foci detected in nuclei, arrowhead: pan‐nuclear γ‐H2AX signal.E–GStatistical evaluation of the presence of γ‐H2AX signals in MuSCs of different genotypes (n = 3/3/3/3 animals used for independent isolations of MuSCs, two technical replicates for each animal, and at least 164 cells were counted, Mann–Whitney test, **P < 0.01. Data are mean ± SEM).H, IGene set enrichment analysis of microarray data obtained for proliferating MuSCs shows enrichment of genes related to activity of ATR in response to replication stress recovery (H) and enrichment of genes involved in chromosome maintenance (I) in WT compared with Pax7‐Crepos INO80−/− MuSCs. Enrichment of these or similar gene sets was not observed when comparing WT to linc‐MYH−/− MuSCs.JModel of the role of INO80 and linc‐MYH in the recognition of γ‐H2AX and DNA repair. The DNA repair…

• 

  Figure 6. linc‐MYH modulates INO80 complex composition by reducing incorporation of YY1 and WDR5 A, BCo‐Immunoprecipitation of INO80 complex subunits from TA muscles of WT and linc‐MYH KO mice followed by Western Blot analysis. (B) Statistical analysis of Co‐IP/Western Blot data. The amount of RUVBL1 and RUVBL2 in the INO80 complex is not changed after deletion of linc‐MYH, while the amounts of YY1, WDR5, and TFPT pulled down by INO80 increase (n = 3 WT/3 linc‐MYH KO animals; one‐way ANOVA, multiple comparisons by Fishers LSD test, *P < 0.05). Data are mean ± SEM.C–KProximity ligation assays between INO80 and YY1 (C, D), INO80 and WDR5 (F, G), and INO80 and RuvBl2 (I, J) in proliferating MuSCs. Red signals indicate close proximity of INO80‐V5 and the respective interacting protein. Nuclei were stained using DAPI (blue). Statistical evaluation of proximity ligation assays indicating increased interactions of INO80‐V5 and YY1 (E, n = 3/3 animals, three independent wells per animal, two‐tailed students t‐test, *P < 0.05) and of INO80‐V5 to WDR5 (H, n = 2 (WT)/2 (KO) animals, three independent wells per animal, two‐tailed students t‐test, **P < 0.01). No increase in proximity was detected…

• 

  Figure EV5. linc‐MYH interacts with specific domains of INO80 INO80 protein with functional domains labeled (NTD: N‐terminal domain, HSA: helicase‐SANT-associated domain, Snf2 ATPase: Snf2 ATPase domain). Numbers indicate positions in the sequence of the hINO80 protein (NP_060023.1).Pull‐down experiments using in vitro transcribed linc‐MYH RNA (AK079404) or control lincRNA to pull‐down human flag‐tagged INO80 protein (hINO80) or hINO80 fragments ectopically expressed in HEK293 cells. Proteins binding to linc‐MYH RNA were identified by Western blot using an anti‐FLAG antibody. The hINO80 protein binds to linc‐MYH as expected, but the interaction is lost following deletion of the NTD (ΔNTD) domain or after deletion of the NTD and HSA (ΔNTD ΔHSA) domains. Accordingly, we detect interaction of the NTD (NTD) domain and of the HSA (HSA) domain with linc‐MYH.Pull‐down experiment using in vitro transcribed linc‐MYH RNA (AK079404) to test its interaction with mouse V5‐tagged MCRS1 protein (MCRS1) ectopically expressed in HEK293 cells.Pull‐down experiments using the different in vitro transcribed exons of linc‐MYH splice variants (AK079404, AK010044) as well as four different controls to detect…