eEF1A2 exacerbated insulin resistance in male skeletal muscle via PKCβ and ER stress

in Journal of Endocrinology
Correspondence should be addressed to X Du: duduyan@ccmu.edu.cn

*(senior author: Z Chen)

Restricted access

Recent studies raise the possibility that eukaryotic translation elongation factor 1 alpha (eEF1A) may play a role in metabolism. One isoform, eEF1A2, is specifically expressed in skeletal muscle, heart and brain. It regulates translation elongation and signal transduction. Nonetheless, eEF1A2’s function in skeletal muscle glucose metabolism remains unclear. In the present study, suppression subtractive hybridisation showed a decrease in Eef1a2 transcripts in the skeletal muscle of diabetic Mongolian gerbils. This was confirmed at mRNA and protein levels in hyperglycaemic gerbils, and in db/db and high-fat diet-fed mice. Further, this downregulation was independent of Eef1a2 promoter methylation. Interestingly, adeno-associated virus-mediated eEF1A2 overexpression in skeletal muscle aggravated fasting hyperglycaemia, hyperinsulinaemia and glucose intolerance in male diabetic gerbils but not in female gerbil models. The overexpression of eEF1A2 in skeletal muscle also resulted in promoted serum glucose levels and insulin resistance in male db/db mice. Up- and downregulation of eEF1A2 by lentiviral vector transfection confirmed its inhibitory effect on insulin-stimulated glucose uptake and signalling transduction in C2C12 myotubes with palmitate (PA)-induced insulin resistance. Furthermore, eEF1A2 bound PKCβ and increased its activation in the cytoplasm, whereas suppression of PKCβ by an inhibitor attenuated eEF1A2-mediated impairment of insulin sensitivity in insulin-resistant myotubes. Endoplasmic reticulum (ER) stress was elevated by eEF1A2, whereas suppression of ER stress or JNK partially restored insulin sensitivity in PA-treated myotubes. Additionally, eEF1A2 inhibited lipogenesis and lipid utilisation in insulin-resistant skeletal muscle. Collectively, we demonstrated that eEF1A2 exacerbates insulin resistance in male murine skeletal muscle via PKCβ and ER stress.

Downloadable materials

  • Table S1. Primers used in our study.
  • Table S2. Primary antibodies used in our study.
  • Table S3. 49 differentially expressed genes were characterized in spontaneous diabetic gerbils by SSH assay.
  • Table S4. dbEST ID and GenBank Accn of ESTs from SSH libraries.
  • Figure S1. CpG island prediction in the promotor region of eEF1A2 in mice.
  • Figure S2. The homology of amino acid sequences of eEF1A2 between Mongolian gerbils and mice.
  • Figure S3. The body weight of three diabetic models. After fasting for 16 h, the body weight of three diabetic animal models: diabetic Mongolian gerbils and their controls (A), db/db mice and their wild type controls (WT) (B), chow-fed or high-fat diet (HFD)-fed mice (C) was tested. 3 independent replicates, each with 3-4 animals per group. ***denotes statistical significance at P < 0.005.
  • Figure S4. Oral glucose tolerance test (OGTT) of db/db mice and gerbil models before virus injection. Male (n = 4) and female gerbils (n = 3) with severe glucose intolerance (A), male db/db mice (n = 7) (B), were injected randomly with viral particle of AAV-ZsGreen or AAV-eEF1A2 into the bilateral thigh and calf muscles. Before injection, animals were fasting for 16 h, and then OGTT was tested.
  • Figure S5. PA dose-dependently inhibited insulin-stimulated Akt phosphorylation. After differentiation and deprivation, the myotubes were preincubated with 0-1.4 mM PA for 16 h followed by 100 nM of insulin for 15 min. The cells were collected to perform western blotting to detect protein levels of p-Akt and Akt. 3 independent replicates, * and ***denotes statistical significance at P < 0.05 and P < 0.005, respectively.
  • Figure S6. The effects of eEF1A2 on PI3K and GSK3β activities in PA-treated myotubes. The myotubes were preincubated with 0.5 mM or 0.8 mM PA for 16 h, and then treated with vehicle or 100 nM insulin (Ins) for 15 min. The cells were collected for western blotting to detect protein levels of p-PI3K, PI3K (A), p-GSK3β and GSK3β (B).
  • Figure S7. The food intake of the AAV-injected db/db mice. Male db/db mice with severe glucose intolerance were randomly injected with 1.3  1011 viral particles of AAV-ZsGreen or AAV-eEF1A2 into the thigh and calf muscles of bilateral hind limbs. After 2 weeks, the daily food intake was measured, n = 7 per group.

 

      Society for Endocrinology

Related Articles

Article Information

Metrics

All Time Past Year Past 30 Days
Abstract Views 514 514 514
Full Text Views 45 45 45
PDF Downloads 12 12 12

Altmetrics

Figures

  • View in gallery

    eEF1A2 was downregulated in diabetic gerbil skeletal muscle. An enrichment analysis for DE genes from SSH was performed, and −log10 of corrected P > 1.30103 denotes corrected P < 0.05 (A). OGTT was performed on diabetic Mongolian gerbils and their controls (n = 2 male and n = 2 female gerbils per group, giving n = 4 in total, B). Two days later, the quadriceps femoris from the fasting gerbils was collected and subjected to qRT-PCR analysis. The relative expression levels of Eef1a2 were normalised to β-actin (C) as an internal control or skeletal muscle marker Actn2 to exclude the effects of non-skeletal muscle cells (D). eEF1A2 protein levels in gerbils were measured using Western blotting (E). *, ** and *** denote statistical significance at P < 0.05, P < 0.01 and P < 0.005, respectively.

  • View in gallery

    eEF1A2 expression decreased significantly in insulin-resistant murine skeletal muscle. OGTT was performed on db/db mice and their WT controls, n = 4 per group (A) and chow-fed or HFD-fed mice, n = 3 per group (B). Two days later, the quadriceps femoris from fasting db/db mice (C and D), chow-fed and HFD-fed mice (E and F) was collected and subjected to qRT-PCR analysis. The relative expression levels of Eef1a2 were normalised to β-actin (C and E) or Actn2 (D and F). eEF1A2 protein levels in db/db mice (G), chow-fed and HFD-fed mice (H) were measured using Western blotting. An immunofluorescence assay was performed on chow-fed (I) and HFD-fed mice (J) by simultaneously labelling with anti-eEF1A2 antibodies (red), anti-GLUT4 antibodies (green) and Hoechst 33342 (blue). The immunofluorescence assay conditions in the two groups were the same. Scale bar = 100 μm (top) and 25 μm (bottom). *, ** and *** denote statistical significance at P < 0.05, P < 0.01, and P < 0.005, respectively.

  • View in gallery

    Eef1a2 promoter methylation was not altered in insulin-resistant mice. After fasting for 16 h, the quadriceps femoris or hearts from db/db mice and their WT controls (A, B) and from chow-fed and HFD-fed mice (A, C) were collected to isolate the DNA. A methylation analysis of the Eef1a2 promoter in the samples was performed using the MassARRAY compact system, n = 4–16 per group. The colours of the circles (from yellow to blue) represent 0–100% methylation ratios. After fasting for 16 h, 2 g/(kg body weight) glucose and 0.75 IU/(kg body weight) insulin were administered to chow-fed (D) and HFD-fed mice (E) by gavage and intraperitoneal injection, respectively. The quadriceps femoris was collected between 0 and 120 min after administration for Western blotting, n = 3 per group. * and ** denote statistical significance at P < 0.05 and P < 0.01, respectively. A full colour version of this figure is available at https://doi.org/10.1530/JOE-19-0051.

  • View in gallery

    eEF1A2 overexpression in skeletal muscle impaired insulin sensitivity in male insulin-resistant gerbils. The age-matched gerbils with glucose intolerance were randomly injected with 1.3 × 1011 viral particles of AAV-ZsGreen or AAV-eEF1A2 into the thigh and calf muscles of the bilateral hind limbs (three to four males or females per group). After 2 weeks, the hind-limb skeletal muscle was collected for imaging of ZsGreen by means of In Vivo Multispectral System (A) or for analysis of protein expression in the quadriceps femoris by Western blotting (B). Body weight (C), fasting serum glucose levels (D), fasting serum insulin levels (E), HOMA-IR (F), OGTT in the male groups (G), OGTT in the female groups (H), OGTT area under the curve (I), ITT in the male groups (J), ITT in the female groups (K), ITT area under the curve (L) and fasting serum free FFA levels (M) were evaluated. *, ** and *** denote statistical significance at P < 0.05, P < 0.01 and P < 0.005, respectively. A full colour version of this figure is available at https://doi.org/10.1530/JOE-19-0051.

  • View in gallery

    eEF1A2 overexpression in skeletal muscle exacerbated insulin resistance in male db/db mice. Male db/db mice (n = 7) with severe glucose intolerance were randomly injected with 1.3 × 1011 viral particles of AAV-ZsGreen or AAV-eEF1A2 into the thigh and calf muscles of bilateral hind limbs. After 2 weeks, the hind-limb skeletal muscle was collected for imaging of ZsGreen on an In Vivo Multispectral System (A) or for quantification of protein expression in the quadriceps femoris by Western blotting (B). Body weight (C), fasting serum glucose levels (D), the fasting serum insulin levels (E), HOMA-IR (F), OGTT (G), its area under the curve (H), ITT (I), its area under the curve (J), the fasting serum free fatty acid (FFA) levels (K) and the fasting serum triglyceride levels (L) of the two groups were evaluated. In addition, the fat mass ratios of the AAV-ZsGreen (M) and AAV-eEF1A2 (N) groups were measured using MRI and calculated (O). *, ** and *** denote statistical significance at P < 0.05, P < 0.01 and P < 0.005, respectively. A full colour version of this figure is available at https://doi.org/10.1530/JOE-19-0051.

  • View in gallery

    eEF1A2 aggravated insulin resistance in myotubes with PA-induced insulin resistance. The C2C12 cells were transfected with an empty vector (pCDH) or the FLAG-tagged-eEF1A2 lentivirus vector (eEF1A2). The infected cells were isolated using an FACSCalibur cell sorter and cultured. The overexpression efficiency was checked for qRT-PCR (A) and Western blotting (B). After differentiation and serum deprivation, the infected myotubes were treated with 0.5 or 0.8 mM PA for 16 h followed by 100 nM of insulin for 15 min. The cells were collected for glucose uptake assays (C) and Western blotting. Membrane protein was isolated to detect protein levels of GLUT4 (D). The myotubes were preincubated with 0.5 or 0.8 mM PA for 16 h and treated with vehicle or 100 nM insulin (Ins) for 15 min. The cells were collected for Western blotting to detect protein levels of p-Akt, Akt, p-mTOR, mTOR, p-ERK and ERK (E-F). The C2C12 cells were transfected with a scramble shRNA-carrying lentivirus (SHC) or eEF1A2 shRNA-carrying lentivirus SH1 or SH2. The efficiency of RNA interference against eEF1A2 was detected by qRT-PCR (G) and Western blotting (H). The myotubes transfected with SHC or SH1 were preincubated with or without 0.8 mM PA for 16 h, and then treated with 100 nM insulin for 15 min. The cells were collected for glucose uptake assays (I) and Western blotting to assess the protein levels of p-Akt and Akt (J). *, ** and *** denote statistical significance at P < 0.05, P < 0.01 and P < 0.005, respectively. A full colour version of this figure is available at https://doi.org/10.1530/JOE-19-0051.

  • View in gallery

    eEF1A2 impaired insulin sensitivity by activating PKCβ in insulin-resistant myotubes. After differentiation and serum deprivation, pCDH-carrying, eEF1A2-overexpressing, and eEF1A2-overexpressing plus 0.1 μM PKC β inhibitor LY333531 (LY)-treated myotubes were preincubated with 0.8 mM PA for 16 h followed by 100 nM of insulin for 15 min. The cells were collected to perform Western blotting to detect amounts of p-PKCβ and PKCβ protein (A). The quadriceps femoris of db/db mice transfected with AAV-ZsGreen or AAV-eEF1A2 for 2 weeks was collected for Western blotting, n = 3 per group (B). The myotubes (with SHC or SH1) were preincubated with or without 0.8 mM PA for 16 h, and then treated with 100 nM insulin for 15 min. The cells were collected for Western blotting to measure the amounts of p-PKCβ and PKCβ protein (C). The myotubes (transfected with pCDH or FLAG-tagged eEF1A2) were preincubated with 0.8 mM PA for 16 h, and then stimulated with 100 nM insulin for 15 min. The cells were subjected to a Co-IP assay (D) and simultaneously to an immunofluorescence assay by labelling with anti-FLAG antibodies (red), anti-pPKCβ antibodies (green) and Hoechst 33342 (blue). Scale bar = 75 μm (top) and 25 μm (bottom) (E). The pCDH-transfected, eEF1A2-overexpressing, or eEF1A2-overexpressing plus 0.1 μM LY-treated myotubes were preincubated with 0.8 mM PA for 16 h followed by 100 nM insulin for 15 min. The cells were collected for a glucose uptake assay (F) and Western blotting. Membrane protein was isolated to detect protein levels of GLUT4 (G). Total protein was isolated to detect the amounts of p-Akt and Akt protein (H). *, ** and *** denote statistical significance at P < 0.05, P < 0.01 and P < 0.005, respectively.

  • View in gallery

    ER stress was involved in eEF1A2-promoted insulin insensitivity in PA-treated myotubes. After differentiation and serum deprivation, the pCDH-transfected and eEF1A2-overexpressing myotubes (A), as well as SHC- and SH1-transfected myotubes (C), were treated with 0.5 or 0.8 mM PA for 16 h followed by 100 nM insulin for 1 h. The cells were collected for qRT-PCR. The quadriceps femoris of the db/db mice transfected with AAV-ZsGreen or AAV-eEF1A2 for 2 weeks was collected for qRT-PCR, n = 7 per group (B). The myotubes (transfected with pCDH, FLAG-tagged eEF1A2 and SHC or SH1) were incubated with 0.8 mM PA for 16 h followed by 100 nM insulin for 15 min. The cells were collected for Western blotting (D). The pCDH-carrying, eEF1A2-overexpressing, eEF1A2-overexpressing plus 2 mM ER stress inhibitor 4-PBA (PBA)-treated and eEF1A2-overexpressing plus 20 mM 4-PBA-treated myotubes were preincubated with 0.8 mM PA for 16 h, and then stimulated with 100 nM insulin for 15 min. The cells were collected for a glucose uptake assay (E) and Western blotting (F). The pCDH-carrying, eEF1A2-overexpressing and eEF1A2-overexpressing plus 20 μM JNK inhibitor SP600125 (SP)-treated myotubes were incubated with 0.8 mM PA for 16 h followed by 100 nM insulin for 15 min. The cells were collected for a glucose uptake assay (G) and Western blotting (H-I). *, ** and *** denote statistical significance at P < 0.05, P < 0.01 and P < 0.005, respectively. A full colour version of this figure is available at https://doi.org/10.1530/JOE-19-0051.

  • View in gallery

    eEF1A2 inhibited lipogenesis and lipid use in insulin-resistant skeletal muscle. After differentiation and serum deprivation, the pCDH-carrying and eEF1A2-overexpressing myotubes were treated with 0.5 or 0.8 mM PA for 16 h followed by 100 nM insulin for 1 h. The cells were collected for qRT-PCR (A and B). The quadriceps femoris of db/db mice transfected with AAV-ZsGreen or AAV-eEF1A2 for 2 weeks was collected to perform qRT-PCR (n = 7 per group, C and D). The SHC-transfected or SH1-transfected myotubes were also treated with 0.8 mM PA for 16 h followed by 100 nM insulin for 1 h and were collected for qRT-PCR analysis (E and F). Legend: acetyl-CoA carboxylase α (Acc1), fatty acid synthase (Fas), stearoyl-CoA desaturase 1,2 (Scd 1,2), sterol regulatory element-binding protein 1c (Srebp 1c), lipoprotein lipase (Lpl), carnitine palmitoyl transferase 1 (Cpt1) and uncoupling protein 2 (Ucp2). * or different letters denote statistical significance at P < 0.05; ** and *** denote statistical significance at P < 0.01 and P < 0.005, respectively. A full colour version of this figure is available at https://doi.org/10.1530/JOE-19-0051.

PubMed

Google Scholar