PRMT5 links lipid metabolism to contractile function of skeletal muscles

Abstract

Skeletal muscle plays a key role in systemic energy homeostasis besides its contractile function, but what links these functions is poorly defined. Protein Arginine Methyl Transferase 5 (PRMT5) is a well-known oncoprotein but also expressed in healthy tissues with unclear physiological functions. As adult muscles express high levels of Prmt5, we generated skeletal muscle-specific Prmt5 knockout (Prmt5^MKO ) mice. We observe reduced muscle mass, oxidative capacity, force production, and exercise performance in Prmt5^MKO mice. The motor deficiency is associated with scarce lipid droplets in myofibers due to defects in lipid biosynthesis and accelerated degradation. Specifically, PRMT5 deletion reduces dimethylation and stability of Sterol Regulatory Element-Binding Transcription Factor 1a (SREBP1a), a master regulator of de novo lipogenesis. Moreover, Prmt5^MKO impairs the repressive H4R3 symmetric dimethylation at the Pnpla2 promoter, elevating the level of its encoded protein ATGL, the rate-limiting enzyme catalyzing lipolysis. Accordingly, skeletal muscle-specific double knockout of Pnpla2 and Prmt5 normalizes muscle mass and function. Together, our findings delineate a physiological function of PRMT5 in linking lipid metabolism to contractile function of myofibers.

Keywords: lipid droplet; lipolysis; myofiber; posttranslational modification; protein arginine methyltransferase.

Figure 1. Muscle‐specific knockout of PRMT5 (Prmt5 ^MKO ) leads to muscle mass reduction

1. A

   Genetic targeting strategy showing skeletal muscle‐specific deletion of Prmt5 using the Cre‐LoxP recombinase under the control of the Myl1 promoter.

2. B, C

   Representative images of 2‐month‐old mice (B) and growth curves (C) of WT (n = 6) and Prmt5 ^MKO mice (n = 6).

3. D

   Food intake normalized to body weight in WT (n = 4) and Prmt5 ^MKO mice (n = 3).

4. E

   Lean and fat mass in WT (n = 4) and Prmt5 ^MKO mice (n = 4) determined by EcoMRI body composition analyzer.

5. F

   Photographs of skeletal muscles in 2‐month‐old WT and Prm5 ^MKO mice.

6. G

   The quantified weight of muscles (SOL, EDL, TA, GAS, QU) in 2‐month‐old WT (n = 4) and Prmt5 ^MKO mice (n = 4).

7. H, I

   Representative H&E (H) and dystrophin immunofluorescence of images of TA muscle cross‐sections in WT and Prmt5 ^MKO mice. Scale bar: 100 μm.

8. J, K

   Distribution of myofiber size (J) and average myofiber cross‐sectional area (CSA) (K) of WT (n = 4) and Prmt5 ^MKO TA muscles (n = 4).

9. L

   The total number of myofibers in EDL and Soleus from WT (n = 5) and Prmt5 ^MKO mice (n = 6).

Data information: The data are presented as mean ± SEM and P‐values determined by two‐tailed ANOVA unpaired t‐test based on sample sizes (biological replicates) shown in (C–E, G, J, K, L) (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001).

Figure EV1. Specific Prmt5 deletion in myocytes in Prm5 ^MKO mice

1. A

   Relative mRNA levels of Prmt isoforms from hindlimb muscles at postnatal day 4; (n = 4).

2. B

   Prmt5 mRNA levels across various human tissues.

3. C

   Violin plots showing the expression of Prmt5 and muscle satellite cell markers during differentiation.

4. D

   Relative mRNA expression of Prmt5 in skeletal muscle (SM), inguinal adipose tissue (iWAT), and brown adipose tissue (BAT) from WT and Prmt5 ^MKO mice; (n = 4, biological replicates).

5. E

   Representative images (left panel) and quantified weight (right panel) of non‐muscle tissue (iWAT and BAT) from WT and Prmt5 ^MKO mice; (n = 4, biological replicates).

6. F

   relative expression level of Atrogin‐1 and MuRF‐1 (encoded by FBX032 and Trim63 gene) in skeletal muscles from WT and Prmt5 ^MKO mice; (n = 4, technical replicates).

Data information: The data are presented as mean ± SEM in (A, D, E, F). The P‐values by two‐tailed ANOVA unpaired t‐test are indicated in (D). The total number of biologically independent samples are indicated in (A, D, E, F). ASC, activated MuSCs; COM, committed MuSCs; DIF, differentiated MuSCs; DIV, dividing MuSCs; MuSCs, muscle satellite cells; QSC, quiescent MuSCs (****P < 0.0001).

Figure EV2. Prmt5 ablation in skeletal muscle does not affect myogenesis in vivo

1. A

   Representative immunofluorescences of WGA staining in TA muscles cross‐sections at post‐natal day at 7 and 21 in WT and Prmt5 ^MKO mice. Scale bar: 100 μm.

2. B, C

   Average size (B) and total number (C) of myofiber in TA muscles at day 7 and 21 in WT (n = 5 for P7, n = 4 for P21) and Prmt5 ^MKO mice (n = 4 for P7, n = 4 for P21, biological replicates).

3. D

   A representative immunofluorescence of Pax7 and Dystrophin (left panel) and quantification of Pax7⁺ cells per area (right panel) in TA muscles from WT and Prmt5 ^MKO mice (n = 6). Scale bar: 100 μm.

4. E

   A representative immunofluorescence of Pax7 (upper panel) and quantification of Pax7⁺ cells per myofiber (bottom panel) in freshly isolated single myofibers from WT and Prmt5 ^MKO mice (n = 3).

5. F

   A representative H&E staining (left panel) and quantification of recovery rate (right panel) of TA muscles upon CTX injury at 5.5‐day and from WT (n = 4) and Prmt5 ^MKO mice (n = 3). Scale bar: 100 μm.

6. G

   A representative immunofluorescence of eMyHC and Dystrophin in injured TA muscles from WT and Prmt5 ^MKO mice. Scale bar: 100 μm.

Data information: The data are presented as mean ± SEM in (B–F). The total number of biologically independent samples are indicated in (B–F).

Figure 2. Reduced motor‐performance and muscle contractile function in 2‐month‐old Prmt5 ^MKO mice

1. A

   Grip strength tests of WT and Prmt5 ^MKO mice assessed by grip force normalized to body weight WT (n = 12) and Prmt5 ^MKO mice (n = 10).

2. B–D

   Exercise performance of maximum speed (B), running time (C), running distance (D) of WT (n = 11) and Prmt5 ^MKO mice (n = 9) measured by treadmill.

3. E, F

   Absolute (E) and specific force (F) (on the left panel) and maximal absolute and specific force (on right panel) on EDL muscle from WT (n = 4) and Prmt5 ^MKO mice (n = 3).

4. G, H

   Absolute (G) and specific force (H) (on the left panel) and maximal absolute and specific force (on right panel) on Soleus muscle from WT (n = 4) and Prmt5 ^MKO mice (n = 3). Data information: The data are presented as mean ± SEM and the P‐values by two‐tailed ANOVA unpaired t‐test based on total number of biologically independent samples indicated in (A–H) (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001).

Figure 3. Depletion of Prmt5 leads to a fiber‐type switch toward glycolytic myofibers but does not affect systemic metabolism

1. A

   Representative immunostaining of Type I, IIA, and IIB myofibers in in EDL and Soleus muscles of 2‐month‐old WT and Prmt5 ^MKO mice.

2. B, C

   Quantification of abundance of various fiber types in EDL (B) and Soleus (C) muscles of WT and Prmt5 ^MKO mice; (n = 5).

3. D

   Representative histochemical staining image of succinate dehydrogenase (SDH) enzymatic activity in EDL and Soleus muscles.

4. E, F

   Quantification of abundance of SDH‐low, ‐medium and ‐high myofibers in EDL (E) and Soleus (F) muscles of WT (n = 5) and Prmt5 ^MKO mice (n = 4).

5. G, H

   Metabolic rate of O₂ consumption (G) and CO₂ production (H) normalized to lean mass for a 48‐h cycle of 4–6‐month‐old WT (n = 8) and Prmt5 ^MKO (n = 7) mice measured by an indirect calorimetry.

Data information: The data are presented as mean ± SEM and P‐values by two‐tailed ANOVA unpaired t‐test based on total number of biologically independent samples indicated in (B, C), and (E–H) (*P < 0.05, **P < 0.01, ***P < 0.001).

Figure EV3. Systemic metabolic rate in WT and Prmt5 ^MKO mice by an indirect calorimetry

1. A

   Average day and night heat expenditure of 4–6‐month‐old WT (n = 8) and Prmt5 ^MKO (n = 7) mice over 48 h.

2. B, C

   Systemic metabolic rate of O₂ consumption (VO₂) (B) and CO₂ production (VCO₂) (C) of WT (n = 8) and Prmt5 ^MKO (n = 7) mice.

Data information: The data are presented as mean ± SEM in (A–C). The P values by two‐tailed ANOVA unpaired t test are indicated in (A). The total number of biologically independent samples are indicated in (A–C) (*P < 0.05, **P < 0.01).

Figure 4. PRMT5‐deficient skeletal muscles exhibit lower lipid content and metabolic rate

1. A, B

   Representative images of ORO staining (A) and immunofluorescence (B) in TA muscle from WT and Prmt5 ^MKO mice. Scale bar: 100 μm.

2. C

   Western blotting analysis for protein markers of lipolysis, lipogenesis, and electron transport chain (ETC) complexes in skeletal muscles of WT (n = 3) and Prmt5 ^MKO mice; (n = 3).

3. D

   Quantification of relative protein levels of ATGL, mSREBP1, CEBP‐β, FABP4, OXPHOS complexes (normalized to GAPDH) in skeletal muscles of WT and Prmt5 ^MKO mice; (n = 3).

4. E

   Seahorse analysis of oxygen consumption rate (OCR) in myotubes differentiated for 3‐days from myoblasts of 4‐week‐old WT and Prmt5 ^MKO mice; (n = 4).

5. F

   OCR was measured at basal state and after sequential addition of Oligomycin, FCCP, and Rotenone/Antimycin A to determine basal respiration (BS), proton leak (PL), ATP respiration (ATP R), maximal respiration (Max R), spare capacity (SC), and non‐mitochondrial respiration (NR); (n = 4, technical replicates).

Data information: The data are presented as mean ± SEM and P‐values by two‐tailed ANOVA unpaired t‐test based on number of biologically independent samples indicated in (D–F) (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001).

Figure 5. PRMT5 regulates lipid accumulation through SREBP1a methylation and stabilization

1. A

   C2C12 cells overexpressing PRMT5‐GFP alone or PRMT5‐GFP + SREBP1a‐Flag were immunoprecipitated with Flag antibody and blotted with SYM10, GFP, Flag, PRMT5, and Tubulin antibodies. Exo, exogenous (overexpressed); Endo, endogenous.

2. B

   C2C12 cells overexpressing SREBP1a‐Flag alone or SREBP1a‐Flag + PRMT5‐GFP were immunoprecipitated with GFP antibody and blotted with Flag, GFP, PRMT5, and Tubulin antibodies.

3. C

   Protein extracts of skeletal muscle isolated from WT and Prmt5 ^MKO mice were immunoprecipitated with SREBP1 antibody and immunoblotted with SYM10 (mSREBP1), SREBP1, PRMT5, and tubulin antibodies.

4. D

   Quantification of methylated mSREBP1 (normalized to total mSREBP1) in (C); (n = 3, biological replicates).

5. E

   HEK293 cells were transfected with PRMT5‐GFP alone or PRMT5‐GFP + SREBP1a‐Flag for 24 h, followed by cycloheximide (30 μg/ml) and protein analysis at 0, 4, 8 h. Lysates were immunoblotted with Flag, GFP, and tubulin antibodies.

6. F

   Intensity of Flag was normalized to tubulin, then normalized to 0 h; (n = 3, biological replicates).

7. G

   Relative expression of lipogenesis genes (Dgat1, Dgat2, Fabp4) in C2C12 transfected with PRMT5 and SREBP1a; (n = 4, technical replicates).

Data information: Data are presented as mean ± SEM and P‐values determined by two‐tailed ANOVA unpaired t‐test based on total number of biologically independent samples indicated in (D, F, G) (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001).

Figure EV4. PRMT5 inhibitor reduces SREBP1a dimethylation in C2C12 myoblasts

1. A

   C2C12 cells overexpressing SREBP1a‐Flag were immunoprecipitated with Flag antibody in the absence or presence of BLL 3.3, PRMT5 inhibitor, and blotted with SYM10, Flag, GAPDH.

2. B

   Immunoblotting analysis showing the contents of SREBP1 protein levels in 3‐days differentiated myotube isolated from WT and Prmt5 ^MKO mice.

3. C, D

   Relative expression of Prmt5 (C) and Srebp1 (D) in C2C12 cells overexpressing SREBP1a along or together with PRMT5 (n = 4).

Data information: The data are presented as mean ± SEM in (C, D). The P‐values by two‐tailed ANOVA unpaired t‐test are indicated in (C, D). The total number of biologically independent samples are indicated in (C, D) (*P < 0.05, **P < 0.01).

Figure 6. Epigenetic repression of Pnpla2 transcription by PRMT5 in skeletal muscle

1. A, B

   Immunoblots showing symmetric demethylation of H4R3 in skeletal muscles of WT (n = 4) and Prmt5 ^MKO mice (n = 4) (A), and quantification of H4R3Me2s normalized to total H4 (B).

2. C

   Relative Pnpla2 levels in skeletal muscles of WT and Prmt5 ^MKO mice; (n = 4)

3. D

   Relative Pnpla2 levels in Prmt5‐overexpressed C2C12 myoblasts; (n = 6).

4. E

   ChIP‐sequencing results showing the Prmt5 binding peak on the Pnpla2 promoter of 3T3‐L1 cells.

5. F, G

   Enrichment of PRMT5 (F) and H4R3Me2s (G) on the proximal promoter region of the Pnpla2 gene in myotubes; (n = 4 for F, n = 3 for G, biological replicates).

6. H, I

   Enrichment of PRMT5 (H) and H4R3Me2s (I) on the proximal promoter region of the Pnpla2 gene in skeletal muscles of WT and Prmt5 ^MKO mice (n = 4 for H, and I, biological replicates).

Data information: The data are presented as mean ± SEM and P‐values by two‐tailed ANOVA unpaired t‐test based on total number of biologically independent samples indicated in (B–D), and (F–I) (*P < 0.05).

Figure 7. Pnpla2 deletion restores lipid content and muscle function of Prmt5 ^MKO mice

1. A, B

   A representative images of whole body (A) and body weight (B) of WT (n = 6), Prmt5 ^MKO (n = 6), and Prmt5/Pnpla2 ^MKO mice (n = 6) at 8 weeks of age.

2. C

   Percentage of lean mass determined by body composition analyzer for WT (n = 6), Prmt5 ^MKO (n = 7) and Prmt5/Pnpla2 ^MKO mice (n = 6) at 8‐weeks‐old.

3. D

   Immunoblotting analysis showing PRMT5, ATGL and ETCs in muscle tissues (upper panel) and quantification of its relative protein levels in muscles of WT, Prmt5 ^MKO , and Prmt5/Pnpla2 ^MKO mice (bottom panel) (n = 2, biological replicates).

4. E, F

   Exercise performance in running distance (E) and griping strength test (F) in WT (n = 11 for E, n = 8 for F), Prmt5 ^MKO (n = 10 for E, n = 9 for F) and Prmt5/Pnpla2 ^MKO mice (n = 9 for E, n = 8 for F).

5. G, H

   Absolute force (left panel) and maximal absolute force (right panel) of EDL (G) and Soleus (H) muscles of WT (n = 4), Prmt5 ^MKO (n = 5) and Prmt5/Pnpla2 ^MKO mice (n = 5).

6. I

   Representative images of ORO staining in TA muscles under brightfield (upper panel) and fluorescent imaging (bottom panel) from WT, Prmt5 ^MKO , and Prmt5/Pnpla2 ^MKO mice. Scale bar: 50 μm.

Data information: The data are presented as mean ± SEM and P‐values by two‐tailed ANOVA unpaired t‐0test based on total number of biologically independent samples indicated in (B–H) (*P < 0.05, **P < 0.01, ***P < 0.001).

Figure EV5. Loss of Pnpla2 enlarges myofiber sizes and improves muscle function compared to Prmt5 ^MKO mice

1. A–C

   A Representative H&E staining, (A) distribution (B) and average (C) of myofiber CSA in TA muscles from WT, Prmt5 ^MKO and Prmt5/Pnpla2 ^MKO mice; (n = 5). Scale bar: 100 μm.

2. D

   Representative immunofluorescent fiber type staining (fiber type I, fiber type IIA, fiber type IIB) of 2–3‐month‐old WT, Prmt5 ^MKO and Prmt5/Pnpla2 ^MKO mice.

3. E, F

   Quantification of stained 3 fiber type in EDL (E) and Soleus (F) muscles from WT, Prmt5 ^MKO and Prmt5/Pnpla2 ^MKO mice; (n = 4).

4. G, H

   A graph of specific force (left panel) and maximal specific force (right panel) of EDL (G) and Soleus (H) muscles from WT (n = 4), Prmt5 ^MKO mice (n = 5) and Prmt5/Pnpla2 ^MKO mice (n = 5).

Data information: The data are presented as mean ± SEM in (B–H). The P‐values by two‐tailed ANOVA unpaired t‐test are indicated in (C–H). The total number of biologically independent samples are indicated in (B–H) (*P < 0.05, **P < 0.01).

Figure 8. Schematic model depicting the role of PRMT5 in the skeletal muscle

PRMT5 is highly expressed in skeletal muscle and necessary to mediate symmetric dimethylation of repressive H4R3 at the Pnpla2 gene, therefore inhibiting lipolysis. PRMT5 also methylates and stabilizes mSREBP1a to promote LD biogenesis. Genetic knockout of Prmt5 attenuates physical activity and decreases myofiber size accompanied by reduced LD deposition. The deletion of Pnpla2 in vivo can restore the decreased muscle function due to Prmt5 loss.