Reexpression of pyruvate kinase M2 in type 1 myofibers correlates with altered glucose metabolism in myotonic dystrophy

Abstract

Myotonic dystrophy type 1 (DM1) is caused by expansion of CTG repeats in the 3' UTR of the DMPK gene. Expression of CUG expansion (CUG(exp)) RNA produces a toxic gain of function by disrupting the functions of RNA splicing factors, such as MBNL1 and CELF1, leading to splicing changes associated with clinical abnormalities. Progressive skeletal muscle weakness and wasting is one of the most prominent clinical features in DM1; however, the underlying mechanisms remain unclear. Here we report that the embryonic M2 isoform of pyruvate kinase (PKM2), a key enzyme contributing to the Warburg effect in cancer, is significantly induced in DM1 tissue and mouse models owing to aberrant splicing. Expression of PKM2 in DM1 skeletal muscle is restricted to the type 1 fibers, which are particularly susceptible to wasting in DM1. Using antisense oligonucleotides to shift PKM splicing toward increased PKM2 expression, we observed increased glucose consumption with reduced oxidative metabolism in cell culture and increased respiratory exchange ratio in mice, suggesting defects in energy metabolism conferred by PKM2 expression. We propose that PKM2 expression induces changes in type 1 fibers associated with muscle atrophy and muscle weakness in DM1.

Keywords: alternative splicing; muscular dystrophy; redirected splicing; striated muscle development.

Fig. 1.

Pkm alternative splicing during mouse skeletal muscle and heart development. (A) Diagram of the mouse Pkm genomic segment indicating mutually exclusive splicing of exons 9 and 10. Constitutive exons are in black; alternative exons are in white (exon 9) or gray (exon 10). Primers used to amplify the Pkm variable region and the PstI recognition site in exon 10 are indicated. (B) RT-PCR analysis of Pkm mRNA from the indicated developmental stages of mouse skeletal muscle (Left) and heart (Right). PCR products were digested with PstI. Splicing is quantified as percent of the Pkm2 isoform (%Pkm2). (C) Western blot analysis using antibodies recognizing Pkm2, Pkm1, or both (Pkm total) isoforms. Ponceau S-stained blots are shown as loading controls.

Fig. 2.

Increased expression of PKM2 in DM1 heart and skeletal muscle. (A and B) RT-PCR analysis of PKM in control and DM1 skeletal muscle (A) and heart (B), with quantification in scatter plots. The fetal heart shows increased PKM2 compared with the adult heart, as in mice. (C) Western blot analysis of PKM isoforms and tubulin in the same skeletal muscle samples used in A and in HeLa cell extracts.

Fig. 3.

PKM2 up-regulation in DM1 muscle occurs specifically in type 1 myofibers. (A) Sections of paraffin embedded skeletal muscle tissues from control (Left) and DM1 (Right) individuals were stained with anti-PKM2 antibody. (B) Consecutive muscle sections from control (Left) and DM1 (Right) samples were stained with antibodies to fast myosin, type 1 myosin, PKM1, and PKM2. Arrowheads point to type 2 myofibers; arrows indicate type 1 myofibers. (Scale bar: 100 µm.)

Fig. 4.

Expression of CUG^exp RNA induces Pkm2 in mouse heart. (A) Diagram of the tetracycline-inducible TREDT960I transgene used to express CUG^exp RNA. (B) RT-PCR analysis of CUG^exp RNA and poly(A)-binding protein C1 (Pabc1) mRNA (loading control) in cardiac alpha-myosin heavy chain (MHC)rtTA and TREDT960I/MHCrtTA bitransgenic mice fed 6 g dox/kg food for 10 wk (Left) and in WT, homozygous (H) MHCrtTA, and double-homozygous (DH) MHCrtTA/TREDT960I bitransgenic mice with 2 g dox/kg food for 12 wk (Right). Three animals were taken off dox for 16 wk (off dox). (C) Western blot analysis of heart protein extracts from the tissue samples shown in B.

Fig. 5.

CELF1 overexpression, but not Mbnl1 depletion, induces substantial Pkm2 expression. (A) Western blot (WB) (Top) and RT-PCR (Middle) analyses in C2C12 cells after siRNA-mediated knockdown of Mbnl1 and/or Mbnl2. Serca1 splicing responded to loss of Mbnl1. (Bottom) Pkm splicing was quantified from three independent experiments. Error bars indicate SEM. *P < 0.05. (B) RT-PCR analysis of Pkm splicing in heart and skeletal muscle from Mbnl1^∆E3/∆E3 mice. (C) RT-PCR (Upper) and Western blot (Lower) analyses in heart tissue from hemizygous (single and double) MHCrtTA and MHCrtTA/TRECUGBP1 mice fed 2 g/kg dox food for 8 d. CELF1 induction is quantified as the normalized mean CELF1/GAPDH ratio from dox-unfed and dox-fed animals.

Fig. 6.

Redirecting Pkm splicing in C2C12 myotubes changes cell metabolism. (A) Schematic showing PCR primers and MOs (#1–3) complementary to the 3′ and 5′ splice sites of Pkm exon 9. The NcoI sites in exon 9 (6 nt apart) and one PstI site in exon 10 are indicated in diagrams of the PCR products. (B) Splicing analysis of Pkm in C2C12 myotubes in the presence of 10 µM control, MO #1, or MO #2. The efficiency of MO #3 is shown in Fig. S3. (C) Quantification of Pkm2 inclusion in cells treated with MOs. (D) Western blots of proteins from MO-treated C2C12 myotubes and quantification normalized to GAPDH. (E–H) OCR (E), extracellular lactate concentration (F), glucose consumption rate (G), and ATP production (H) in C2C12 myotubes treated with control or Pkm MOs. Error bars indicate SEM.*P < 0.05.

Fig. 7.

Redirecting Pkm splicing in mice alters energy expenditure. (A) Vivo-MOs (control, PKM-#1, and PKM-#2) were delivered to mice at 12 mg/kg for 2 consecutive days, and RNA was isolated from TA and soleus muscles at 72 h after the last injection. RT-PCR analysis of Pkm was performed and quantified as indicated in Fig. 6. (B) Vivo-MOs (control and #2) were delivered at 16 mg/kg twice a week for 2 wk, followed 2 wk later by Western blot analysis. (C) Daytime (Upper) and nighttime (Lower) RER of mice used in B at 2 wk after the injection were measured by indirect calorimetry analysis. Error bars indicate SEM. *P < 0.05; **P < 0.01.