Distinct metabolic states govern skeletal muscle stem cell fates during prenatal and postnatal myogenesis

Abstract

During growth, homeostasis and regeneration, stem cells are exposed to different energy demands. Here, we characterise the metabolic pathways that mediate the commitment and differentiation of mouse skeletal muscle stem cells, and how their modulation can influence the cell state. We show that quiescent satellite stem cells have low energetic demands and perturbed oxidative phosphorylation during ageing, which is also the case for cells from post-mortem tissues. We show also that myogenic fetal cells have distinct metabolic requirements compared to those proliferating during regeneration, with the former displaying a low respiration demand relying mostly on glycolysis. Furthermore, we show distinct requirements for peroxisomal and mitochondrial fatty acid oxidation (FAO) in myogenic cells. Compromising peroxisomal but not mitochondrial FAO promotes early differentiation of myogenic cells. Acute muscle injury and pharmacological block of peroxisomal and mitochondrial FAO expose differential requirements for these organelles during muscle regeneration. Taken together, these observations indicate that changes in myogenic cell state lead to significant alterations in metabolic requirements. In addition, perturbing specific metabolic pathways impacts on myogenic cell fates and the regeneration process.

Keywords: Ageing; Metabolic state; Mitochondria; Peroxisome; Regeneration; Skeletal muscle stem cells.

Fig. 1.

Meta-analysis of transcriptomic data for metabolic pathways in quiescence. (A) Barplot showing the results of metabolic gene set enrichment analysis (GSEA) of five datasets that compare (i) old versus young QSCs, obtained from: Yennek_OvsY (orange), 26- and 2-month-old Tg:Pax7-nGFP mice (our unpublished data; GSE116586); Alonso-Martin (blue), 18- and 2-month-old Pax3^GFP/+ mice (Alonso-Martin et al. 2016); Liu (red), 24- and 2-month-old C57BL/6 mice (Liu et al. 2013); and Sousa-Victor (green), 20–24- and 2-month-old C57BL/6 mice (Sousa-Victor et al. 2014); or (ii) post-mortem versus young QSCs obtained from: Pala _PM (grey), 8–12-week-old Tg:Pax7-nGFP mice (deposited in GEO, accession number GSE116586). Only significantly enriched pathways are shown. The enrichment score was calculated by taking the log10 of the adjusted P-value from the GSEA. A negative enrichment score value (left of the dashed vertical ‘0’ line) indicates enrichment corresponding to downregulated genes; a positive enrichment score (right of the dashed vertical ‘0’ line) indicates enrichment pathways corresponding to upregulated genes. (B) Barcode plots of pathway enrichment analyses showing the genes of selected pathways from the dataset by Liu et al. comparing old and young QSCs. The x-axis indicates the order of genes corresponding to the t-statistic output obtained from the moderate t-test performed with Limma R package on this dataset. Pink areas (left) represent most upregulated genes, blue areas (right) represent most downregulated genes, grey areas (middle) represent genes that do not show substantial variations. Vertical lines correspond to the position of individual genes of the indicated pathway along the ranked list of genes. The enrichment worm-plot above each barcode plot shows the relative enrichment of the vertical bars (i.e. genes) for each part of the plot. (C) RT-qPCR analysis was performed on cells isolated from Tg:Pax7-nGFP mice and transcript values of 30 genes from four different pathways (n=4 per group) were normalised to the respective values obtained from young QSCs. The fold-change is shown as log10 value. Dots represent the mean (±s.e.m.).

Fig. 2.

Metabolic profile of QSCs. (A,B) SeaHorse assay results for OCRs of freshly isolated QSCs from (A) young (Young Q; n=6), and old (Old Q; n=4) Tg:Pax7-nGFP mice at 1×10⁵ cells/well and (B) young (n=6) and D2PM (n=4) Tg:Pax7-nGFP mice at 2×10⁵ cells/well, in real time under basal conditions and in response to mitochondrial inhibitors (O, oligomycin; F, FCCP; A, antimycin). Data are representative of at least two independent experiments. (C) Quantification of basal, minimal and maximal OCRs in old and D2PM QSCs normalised to basal respiration in young QSCs. Basal respiration is the value just before oligomycin injection, minimal respiration is the lowest value after oligomycin injection, and maximal respiration is the highest value after FCCP injection. All values were calculated after subtraction of non-mitochondrial respiration (first point after antimycin injection). (D) Relative ratio of CE [calculated as: 1−(Minimal/Basal OCR)] and SRC [calculated as: Maximal/Basal OCR] in QSCs as in C. (E) ECAR in milli pH per minute [mpH/min] under basal conditions in old, post-mortem and young QSCs freshly isolated from muscle. (F) Ratio of basal OCR to ECAR. OCR was measured at the same time as ECAR. (G) ATP production of satellite cells quantified by measuring the relative luminescence (relative luminescence unit; RLU). Luminescence of old QSCs (old Q), and QSCs D2PM and D4PM was normalised to that of young QSCs (Young Q), n=4 per group; all isolated from Tg:Pax7-nGFP mice. (H) Representative FACS profile (left) of young and old QSCs labelled with MitoTracker, and quantification of mean fluorescence intensity (right) of old QSCs (Old Q), and QSCs D2PM and D4PM normalised to that of young QSCs (Young Q ), n=6 per group. (I) Representative FACS profile (left) of young and old QSCs labelled with TMRE, and quantification of mean fluorescence intensity (right) as described for panel H (n=6 per group). Phycoerythrin (PE) and allophycocyanine (APC) fluorescent channels were used in H and I to read the fluorescent signal of MitoTracker and TMRE. Data are presented as mean (±s.e.m.). All P-values were calculated by using Mann–Whitney U or Student’s t-tests. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

Fig. 3.

Meta-analysis of transcriptomic data for metabolic pathways in proliferation. (A) Barplot showing the results of metabolic gene set enrichment analysis (GSEA) of four datasets that compare injury-activated satellite cells and QSCs from the following datasets: (i) Garcia-Prat (blue), 3 days post cardiotoxin injury of C57BL/6 mice (Garcia-Prat et al. 2016); Liu_60 h (red) and Liu_84 h (magenta; Liu et al., 2013), BaCl₂ injury of 2- to 3-month-old Pax7^CreERT2/+:R26^eYFP/+ mice); or (ii) satellite cells from P8 (proliferating during growth) and adult (quiescent) mice (ST_P8, postnatal day 8 and 6-8-week-old Pax7^nGFP/+ mice; Sambasivan et al. 2009). Only significantly enriched pathways are shown. The enrichment score was calculated by taking the log10 of the adjusted P-value from the GSEA). The positive enrichment score indicates enrichment pathways corresponding to upregulated genes. (B) Barcode plots of pathway enrichment analyses showing the genes of selected pathways from the dataset by Garcia-Prat et al. comparing activated satellite cells (ASCs) and QSCs. The x-axis indicates the order of genes corresponding to the t-statistic output obtained from the moderate t-test performed with Limma R package on this dataset. Pink areas (left) represent most upregulated genes, blue areas (right) represent most downregulated genes, grey areas (middle) represent genes that do not show substantial variations. Vertical lines correspond to the position of individual genes of the indicated pathway along the ranked list of genes. The enrichment worm-plot above each barcode plot shows the relative enrichment of the vertical bars (i.e. genes) for each part of the plot. (C) RT-qPCR analysis was performed on cells isolated from Tg:Pax7-nGFP mice and transcript values of 23 genes from three different pathways (n=4 per group) were normalised to the respective values obtained from young QSCs. The fold-change is shown as log10 value. Dots represent the mean (±s.e.m.). (D) Heatmaps show normalised gene expression in 3 samples of QSCs (columns below green boxes) and ASCs (columns below grey boxes) extracted from the dataset by Garcia-Prat (Garcia-Prat et al. 2016). Each row corresponds to a gene. Black boxes, differentially expressed genes (DEGs) expressed at significant levels (s); white boxes, DEGs expressed at not significant levels (ns). (E) RT-qPCR validation of mitochondrial (left) and peroxisomal (right) FAO gene expression, normalised to that in QSCs and performed by using cells isolated from uninjured muscles or cells isolated from notexin-treated tibialis anterior D3PI or D5PI of Tg:Pax7-nGFP mice (n=4 per group). Data are presented as mean (±s.e.m.). All P-values were calculated using Student’s t-test. *P<0.05, **P<0.01.

Fig. 4.

Satellite cells have different metabolic requirements during muscle growth and regeneration. (A) OCR was measured in real time [pmol/min] in QSCs freshly isolated from Tg:Pax7-nGFP mice at 6–8 weeks (Q), n=6; at E17.5; (n=10); at postnatal day 8 (P8), n=10; and at D3PI and D5PI, n=4 each, under basal conditions and in response to the indicated mitochondrial inhibitors (O, oligomycin; F, FCCP; A, antimycin). Data are representative of at least two independent experiments. (B) Quantification of basal, minimal and maximal OCRs in E17.5, P8, D3PI and D5PI QSCs normalised to basal respiration in young QSCs (Q). (C) Relative ratio of CE [calculated as: 1−(Minimal/Basal OCR)] and SRC [calculated as: Maximal/Basal OCR] in QSCs as described for A. (D) ECAR in milli pH per minute [mpH/min] under basal conditions in freshly isolated QSCs from muscle as in A. (E) Ratio of basal OCR to ECAR in QSCs as described for A. (F) Quantification of MitoTracker and TMRE mean fluorescence intensity normalised to that in young QSCs (Q) analysed by FACS on cells isolated from Tg:Pax7-nGFP mice (n=6 for quiescent, n=3 for other groups) as described for A. (G) By measuring the relative luminescence, ATP production in QSCs (as described for A) was quantified and normalised to that of young QSCs (Q) (n=4 per group). (H) Absolute L-carnitine concentration [ng/μl] quantified by fluorimetric assay of freshly isolated QSCs as described for A (n=4 per group). Data are presented as mean (±s.e.m.). All P-values were calculated using Mann–Whitney U or Student’s t-tests. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

Fig. 5.

Blockage of peroxisomal FAO induces differentiation on activated satellite cells. (A) Schematic of experimental design. Satellite cells were activated in vivo by injection of notexin into the tibialis anterior muscle of Tg:Pax7-nGFP mice and isolated at day 5 (D5). Cells were cultured for 48 h in presence or absence of treatment (same experimental design for drug treatment and siRNA transfection; n=4 per condition; T1, isolation; T2; analysis). Cells were pulsed with EdU for 4 h prior fixation. (B) Immunostaining for Pax7 (green), myogenin (red) and EdU (cyan) of untreated cells or cells treated with 5 µM Thioridazine or 5 µM Etomoxir for 48 h. Cells were pulsed with EdU 4 h prior to fixation. Hoechst staining shows nuclei. (C) Histogram representing percentage of cells positive for EdU and Myog (EdU⁺ and Myog⁺, respectively). (D) Expression levels of indicated genes were calculated and normalised to that of the reference gene ribosomal protein L13aRPL13a. Expression is shown as fold-change to non-treated control. Perox, peroxisomal genes; Mito, mitochondrial genes. (E) Immunostaining for EdU (green) and myogenin (red) of cells transfected with control siRNA or siRNA against catalase (siCAT), acyl-coenzyme oxidase 1 (siACOX1) and carnitine palmitoyltransferase 1b (siCPT1b). (F) Histogram representing percentage of EdU⁺ and Myog⁺ cells. (G) Expression levels of indicated genes relative to that of RPL13a and normalised to scramble control. Data are presented as mean (±s.e.m.). All P-values were calculated using Student’s t-test. *P<0.05, **P<0.01, ***P<0.001.

Fig. 6.

Impairment of FAO in vivo alters muscle regeneration. (A) Schematic of experimental design. Tg:Pax7-nGFP mice were injured with notexin injection into the tibialis anterior muscle followed by i.p. injection with NaCl (control), Thioridazine or Etomoxir (n=8 per group) every day for 10 days (T1) or for 14 days. Mice were analysed 28 days (D28) after injury (T2). Mice were injected i.p. with EdU 4 h before analysis to assess proliferation. (B) Representative images showing immunostaining for embryonic myosin heavy chain (eMyHC; red) and laminin (cyan) of tibialis anterior sections, and quantification of eMyHC-positive (eMyHC⁺) fibres per total fibres of section. (C) Representative images showing immunostaining for Pax7 (red; arrows) and EdU (Hoechst dye; cyan) at T1 (D10) of tibialis anterior sections, and quantification of Pax7-positive cells per mm² of sections (n=8). (D) Frequency of cells double positive for Pax7 and EdU (Pax7⁺EdU⁺) of total Pax7⁺ (Pax7⁺) cells on sections at T1 (D10; n=4). (E) Representative images showing immunostaining for myogenin (red; arrows) and laminin (cyan) of tibialis anterior sections, and quantification of Myog-positive (Myog⁺) cells per mm² on sections at T1 (D10) (n=8). (F) Gene expression levels of myogenin relative to that of RPL13a on isolated satellite cells normalised to NaCl control. (G) Representative FACS profile and gating strategy of antibody-stained muscle preparations from Tg:Pax7-nGFP mice at D28 of treatment (n=5). Histograms show the frequency of cells positive for Pax7 or the single surface markers calculated from the total number of live cells quantified at D10 and D28 after treatment. (H) H&E staining of skeletal muscles from mice treated with NaCl, Thioridazine and Etomoxir at D28 after injury. Histograms show the quantification of fibre cross-sectional area (CSA in µm²). (I) Pax7 gene expression relative to that of RPL13a (used as reference) of isolated satellite cells at D28 after injury, normalised to cells treated with NaCl (control). Data are presented as mean (±s.e.m.). All P-values were calculated using Student’s t-test. *P<0.05, **P<0.01, ***P<0.001. Scale bars: 100 µm (C,E,H); 200 µm (B). DNA was stained using Hoechst dye.