Retinoic acid signalling in fibro/adipogenic progenitors robustly enhances muscle regeneration

Abstract

Background: During muscle regeneration, excessive formation of adipogenic and fibrogenic tissues, from their respective fibro/adipogenic progenitors (FAPs), impairs functional recovery. Intrinsic mechanisms controlling the proliferation and differentiation of FAPs remain largely unexplored.

Methods: Here, we investigated the role of retinoic acid (RA) signalling in regulating FAPs and the subsequent effects on muscle restoration from a cardiotoxin-induced injury. Blockage of retinoic acid receptor (RAR) signalling was achieved through dominant negative retinoic acid receptor α (RARα403) expression specific in PDGFRα+ FAPs in vivo and by BMS493 treatment in vitro. Effects of RAR-signalling on FAP cellularity and muscle regeneration were also investigated in a high-fat diet-induced obese mice model.

Findings: Supplementation of RA increased the proliferation of FAPs during the early stages of regeneration while suppressing FAP differentiation and promoting apoptosis during the remodelling stage. Loss of RAR-signalling caused ectopic adipogenic differentiation of FAPs and impaired muscle regeneration. Furthermore, obesity disrupted the cellular transition of FAPs and attenuated muscle regeneration. Supplementation of RA to obese mice not only rescued impaired muscle fibre regeneration, but also inhibited infiltration of fat and fibrotic tissues during muscle repair. These beneficial effects were abolished after blocking RAR-signalling in FAPs of obese mice.

Interpretation: These data suggest that RAR-signalling in FAPs is a critical therapeutic target for suppressing differentiation of FAPs and facilitating the regeneration of muscle and other tissues.

Funding: This study was supported by grants from the National Institutes of Health (R01-HD067449 and R21-AG049976) to M.D.

Keywords: Adipogenesis; Fibro/adipogenic progenitors; Fibrosis; Muscle regeneration; Obesity; Retinoic acid signalling.

Fig. 1

Loss of retinoic acid (RA) signalling in FAPs impairs skeletal muscle regeneration. (a) A schematic showing dominant negative retinoic acid receptor α (RARα403) expression was conditionally induced in PDGFRα-expressing FAPs after administration of tamoxifen and those mice are referred as RARαDN mice. (b) Weight of the Tibialis anterior (TA) muscle after normalized to tibia length at different days post-injury (dpi). (c) H & E staining of regenerated TA muscle and the distribution of cross-sectional areas of regenerated myofibers (fibres with central nuclei) between them. Bars, 200 μm. (d) Immunoblotting of DESMIN protein in the regenerated TA muscle at 14 dpi and quantification of its relative abundance after normalized to the expression of β-ACTIN. (e) Relative mRNA expression of myogenic genes including Pax7, Myf5, Myod and Myogenin at 7 dpi. (f) Relative mRNA expression of trophic factors for myogenic cells including Igf1, Il6, Wnt1, Wnt3α and Wnt5α at 3 dpi. Results represent the means ± SEM of three mice per group at each time point. Statistics were analysed using a two-way ANOVA followed by Tukey's multiple comparison. # shows significant interaction (p<0.05) between two factors while & and $ show significant difference (p<0.05) between genotypes (WT and RARαDN) and treatments (CON and RA), irrespectively. *<0.05 and **<0.01 show significant difference between two groups.

Fig. 2

Retinoic acid (RA) signalling suppresses the adipogenic differentiation of FAPs. (a) Immunofluorescence analysis of PERILIPIN expression at 14 days post-injury (dpi) and the percentage of PERILIPIN+ areas in regenerated skeletal muscle at 7 and 14 dpi. (b) Relative mRNA expression of adipogenic markers including Fabp4, C-ebpα and Pparγ in regenerated skeletal muscle at 7 dpi. (c) Oil Red O staining of isolated FAPs and percentage of Oil Red O positive areas after adipogenic induction with different doses of RA (0, 0.1, 1 and 10 μM) treatment. (d) Oil Red O staining of isolated FAPs and percentage of Oil Red O positive areas after adipogenic induction with treatments of CON (vehicle only), 1 μM RA, 1 μM BMS493, and combined 1 μM RA + 1 μM BMS493. (e) Relative mRNA expression of adipogenic markers including Fabp4, C-ebpα and Pparγ in isolated FAPs after adipogenic induction with treatments of CON, RA, BMS493 or combined RA + BMS493. (f) Immunofluorescence analysis of PERILIPIN expression and percentage of PERLIPIN+ areas in isolated FAPs after cultured in growth media for 13 days. (g) Relative mRNA expression of adipogenic markers including C-ebpα, Pparγ and Fabp4 in isolated FAPs after cultured in growth media for 13 days. Bars, 200 μm. Results represent the means ± SEM of three mice per group at each time point. Statistics were analysed using a two-way (a, b) or one-way ANOVA (c-g). # shows significant interaction (p<0.05) between two factors while & and $ show significant difference (p<0.05) between genotypes (WT and RARαDN) and treatments (CON and RA), irrespectively. *<0.05 and **<0.01 show significant difference between two groups.

Fig. 3

Retinoic acid signalling inhibits fibrotic differentiation of FAPs. (a) Immunofluorescence images of COL1α and percentage of COL1α+ areas in regenerated muscle at 14 days post-injury (dpi). Bars, 200 μm. (b) Relative mRNA expression of fibrogenic markers including Tcf4, α-Sma, Col1α and Col3α in regenerated muscle at 7 dpi. (c) Immunofluorescence images of COL1α and percentage of COL1α+ FAPs after fibrotic induction of FAPs in vitro. Bars, 100 μm. (d) Relative mRNA expression of Col1α and Col3α in isolated FAPs after fibrotic induction of FAPs in vitro. (e) Immunofluorescence images of COL1α and percentage of COL1α+ FAPs after in vitro culturing FAPs in growth media for 7 days. Bars, 100 μm. (f) Relative mRNA expression of Col1α and Col3α after in vitro culturing FAPs in growth media for 7 days. Results represent the means ± SEM of three mice per group at each time point. Statistics were analysed using a two-way (a, b) or one-way ANOVA (c-f). # shows significant interaction (p<0.05) between two factors while & and $ show significant difference (p<0.05) between genotypes (WT and RARαDN) and treatments (CON and RA), irrespectively. *<0.05 and **<0.01 show significant difference between two groups.

Fig. 4

Retinoic acid regulates cellularity of FAPs. (a) Relative mRNA expression of the ratio of intronic variant (In) to full-length (FL) Pdgfrα transcripts (Pdgfrα In/Pdgfrα FL) during in vitro culture of confluent FAPs in growth media for 7 days. (b) Relative mRNA expression of preadipocyte genes including Pref1, Sox9 and Klf2 during in vitro culture of confluent FAPs in growth media for 7 days. (c) Immunofluorescence analysis of PDGFRα in regenerated TA muscle at 14 days post-injury (dpi) and quantification of the number of PDGFRα+/DAPI+ FAPs per field at different times post-injury. Bars, 200 μm. (d) Immunofluorescence analysis of PCNA and percentage of PCNA+/DAPI+ FAPs at 24 h after different treatments. Bars, 200 μm. (e) Immunofluorescence analysis of cleaved CASPASE3 (cCAS3) and quantifications of the percentage of cCAS3+ FAPs in regenerated muscle at 7 dpi. Bars, 100 µm. Results represent the means ± SEM of three mice per group at each time point. Statistics were analysed using a one-way (a, b, and d) or two-way (c, e) ANOVA. # shows significant interaction (p<0.05) between two factors while & and $ show significant difference (p<0.05) between genotypes (WT and RARαDN) and treatments (CON and RA), respectively. *<0.05 and **<0.01 show significant difference between two groups. For (a) and (b), significant difference compared to CON group was labelled.

Fig. 5

Supplementation of RA rescues skeletal muscle regeneration impaired due to obesity. (a) Blood glucose, insulin and calculated HOMA-IR after 5 h of fasting. n = 12. (b) Weight of regenerated Tibialis anterior (TA) muscle after normalized to tibia length at different days post-injury (dpi). (c) H & E staining of regenerated skeletal muscle, and distribution of cross-sectional areas of regenerated myofibers (fibres with central nuclei) and percentage of interstitial areas between them. Bars, 200 μm. (d) Relative mRNA expression of myogenic genes including Pax7, Myf5, Myod and Myogenin at 7 dpi. (e) Relative mRNA expression of trophic factors including Igf1, Il6, Wnt1, Wnt3α and Wnt5α at 3 dpi. Results represent the means ± SEM of three mice per group at each time point. Statistics were analysed using a one-way ANOVA. *<0.05 and **<0.01 show significant difference between two groups. For (b), significant difference compared to WT+ND group was labelled.

Fig. 6

Supplementation of RA inhibits both adipogenesis and fibrosis in the regenerated skeletal muscle of obese mice. (a) Immunofluorescence analysis of PERILIPIN and the percentage of PERILIPIN+ areas in regenerated muscle at 14 days post-injury (dpi). Bars, 200 μm. (b) Relative mRNA expression of adipogenic markers including Fabp4, C-ebpα and Pparγ in regenerated skeletal muscle at 7 dpi. (c) Masson trichrome staining of regenerated muscle at 14 dpi and percentage of fibrotic areas per field. Bars, 200 μm. (d) Relative mRNA expression of fibrogenic markers including Tcf4, α-Sma, Col1α and Col3α in regenerated TA muscle at 7 dpi. (e) Immunofluorescence analysis of PDGFRα in regenerated TA muscle at 14 days post-injury (dpi) and quantification of the number of PDGFRα+/DAPI+ FAPs per field at different times post-injury. Bars, 200 μm. (f) Immunofluorescence analysis of cleaved CASPASE3 (cCAS3) positive FAPs and quantifications of the percentage of cCAS3+ FAPs. Bars, 100 μm. Results represent the means ± SEM of three mice per group at each time point. Statistics were analysed using a one-way ANOVA. *<0.05 and **<0.01 show significant difference between two groups.