Nitric oxide inhibition of Drp1-mediated mitochondrial fission is critical for myogenic differentiation

Abstract

During myogenic differentiation the short mitochondria of myoblasts change into the extensively elongated network observed in myotubes. The functional relevance and the molecular mechanisms driving the formation of this mitochondrial network are unknown. We now show that mitochondrial elongation is required for myogenesis to occur and that this event depends on the cellular generation of nitric oxide (NO). Inhibition of NO synthesis in myogenic precursor cells leads to inhibition of mitochondrial elongation and of myogenic differentiation. This is due to the enhanced activity, translocation and docking of the pro-fission GTPase dynamin-related protein-1 (Drp1) to mitochondria, leading also to a latent mitochondrial dysfunction that increased sensitivity to apoptotic stimuli. These effects of NO inhibition were not observed in myogenic precursor cells containing a dominant-negative form of Drp1. Both NO-dependent repression of Drp1 action and maintenance of mitochondrial integrity and function were mediated through the soluble guanylate cyclase. These data uncover a novel level of regulation of differentiation linking mitochondrial morphology and function to myogenic differentiation.

Figure 1

NO regulates myogenesis and mitochondrial network formation through cGMP. Myogenic precursor cells transfected with the red fluorescent mitochondrial protein mitoDsRed were treated with l-NAME, ODQ or vehicle (C) and differentiated by serum withdrawal for up to 12 h. (a) Mitochondrial morphology detected by transient transfection with mitoDsRed; nuclei are stained for Myo-D (blue) to distinguish myogenic cells from possible contaminating cells. Bar: 10 μm. (b) Expression of the differentiation markers Mef-2A, Myo-D, myogenin and sarcomeric myosin (MyHC) as detected by western blot analysis. (c) Expression of the mitochondrial proteins mitofusins (Mfn)-1 and 2, Opa1, Drp1, cytochrome c (CytC) and cytochrome c oxidase subunit-IV (COX-IV). Calnexin (Clx) was used as a loading control in the experiments in panels b and c. (d) NOS activity measured as the conversion of l-[³H]-arginine into l-[³H]-citrulline. The inset shows the levels of expression of nNOS during myogenic differentiation. Clx was used as a loading control. (e) Generation of cGMP. The images are from one of four independent reproducible experiments; the graphs represent the values±S.E.M. (n = 4)

Figure 2

Inhibition of myogenesis and mitochondrial network formation by NO removal are not due to toxic effects. Myogenic precursor cells were treated with l-NAME, ODQ or vehicle (C) and differentiated by serum withdrawal for up to 72 h, during which time the rate of spontaneous apoptosis was measured. (a) Phosphatidylserine exposure to the plasma membrane in 7AAD-excluding cells. (b) The number of trypan blue-excluding cells. (c) Expression of procaspase-9 and its cleaved form, and of cleaved caspase-3. The images are from one of three independent reproducible experiments; the graphs represent the values±S.E.M. (n = 3)

Figure 3

NO and cGMP regulate mitochondrial dynamics. (a) Myogenic precursor cells were co-transfected with vectors coding for mitoDsRed and cytosolic YFP (pEN1YFP) and differentiated. l-NAME, ODQ or vehicle (C) was added after 6 h of differentiation and mitochondrial morphology was examined by time-lapse microscopy at the indicated time points. The images are representative of three independent reproducible experiments. Bar: 10 μm. (b, c) Myogenic precursor cells were transfected with the mitoDsRed-coding vector and differentiated. Six hours later cells were treated with l-NAME, ODQ, DETA-NO, 8Br-cGMP or vehicle (C) in various combinations as indicated, fixed and examined after 50 min. The red and blue colors in panel b show staining of mitoDsRed and an anti-MyoD Ab, respectively. Bar: 10 μm. (c) Fifteen random fields per sample were acquired and the fragmentation index was established using the imagetool 3.0 software. The dimension ranges for fragmented and elongated were 1.20 μm ± 0.2 and >3.50 μm, respectively. The asterisks show the statistical probability versus C, crosses versus l-NAME and circles versus ODQ (P<0.001)

Figure 4

NO and cGMP stimulate myogenesis through inhibition of mitochondrial fission. (a, b) Myogenic precursor cells were transfected with the vector coding for either mitoGFP (green) or mitoDsRed (red), mixed, differentiated for 6 h and exposed for 1 h to l-NAME, ODQ, DETA-NO, 8Br-cGMP or vehicle (C) as indicated. Plasma membrane fusion was induced by addition of PEG 1500 and mitochondrial fusion events were quantified after 2, 4 and 6 h in the heteropolykaryons by measuring the fraction of mitochondria simultaneously positive for both mtGFP and mitoDsRed (colocalization index %; n = 3). Bar: 10 μM. (c) Myogenic precursor cells were transfected with vectors coding for the cytosolic marker pEYFP-N1 (green) or the dominant-negative Drp1, pEYFP-N1-DRPK38A, and differentiated in the presence of l-NAME and ODQ. Mitochondrial morphology was assessed after 6 h. Bar: 10 μM. (d) Expression of the myogenic differentiation markers Mef-2A, MyoD, myogenin and sarcomeric myosin (MyHC) was determined by western blotting in myogenic precursors transfected with either the empty pCDNA3 vector or the dominant-negative Drp1 (pcDNA3-Drp1 K38A) at the indicated time points. The result of one out of three reproducible experiments is shown

Figure 5

NO and cGMP control the activity and localization of Drp1. Myogenic precursor cells were differentiated for 6 h and treated for 1 h with l-NAME, ODQ, DETA-NO, 8Br-cGMP or vehicle (C), added as indicated. (a) Cells were fractionated and the mitochondrial (MT), microsomal (MI) and cytosolic (CI) fractions were examined for Drp1 content by western blotting, using GAPDH and COX-IV as loading controls for cytosolic and mitochondrial proteins, respectively. The results shown in the images are representative of five reproducible experiments, which are quantified in the corresponding graphs. (b) Co-immunoprecipitation of Drp1 with Fis1 was performed using the an anti-Fis1 Ab for immunoprecipitation (IP). The amount of co-immunoprecipitated Drp1 was determined by immunoblotting (IB) using a specific anti-Drp1 Ab. As control the amount of Fis1 immunoprecipitated was also checked by IB. As a further control for specificity we checked the absence of co-immunoprecipitation of Drp1 with calnexin. (c) Drp1 GTPase activity was measured in pull-down experiments using GTP-conjugated beads. The representative images shown in panels b and c are from four independent reproducible experiments. The graphs below each image report the densitometric values±S.E.M. of the relevant band from the four experiments. In all panels, the asterisks, crosses and circles show statistical probability (P<0.001), calculated versus C, l-NAME and ODQ, respectively

Figure 6

NO/cGMP triggers G-kinase-dependent phosphorylation of Drp1. (a) Myc-Drp1 was immunoprecipitated from [³²P]orthophosphate-labeled proliferating or differentiating myogenic precursor cells. The immunoprecipitates were resolved by SDS-PAGE and ³²P-labeled-Myc-Drp1 was visualized by autoradiography. (b) Myc-Drp1 was immunoprecipitated from [³²P]orthophosphate-labeled myogenic precursor cells treated with DETA-NO, 8Br cGMP with or without KT5823 and ODQ, respectively, or BAY41-2272. The immunoprecipitates were resolved by SDS-PAGE and ³²P-labeled-Myc-Drp1 was visualized by autoradiography

Figure 7

The bioenergetic consequences of regulation of fission by NO and cGMP. Myogenic precursor cells were transfected with pEYFP-N1 or pEYFP-N1-Drp1 K38A, differentiated (diff) or allowed to proliferate (pro) for 6 h, and treated as specified. (a) The cells were loaded with the mitochondrial potentiometric dye TMRM and the mitochondrial membrane potential was measured 1 h after treatment with the vehicle (C) or ODQ. The arrows indicate addition of oligomycin (O; 1 μg/ml) and FCCP (4 μM). (b, c) The cells were loaded with luciferin-luciferase and the ATP generated through oxidative phosphorylation (b) or glycolysis (c) was measured after addition of the vehicle or ODQ. (d, e) Respiratory function was measured using a high-sensitivity respirometer. The values shown are for total, oligomycin-resistant, maximal (uncoupled) and residual oxygen consumption (ROC), as well as for the ratio of oligomycin-resistant to maximal (oligomycin resistant/M) and coupled (total minus oligomycin resistant) to maximal (C/M). (f) Activity of the mitochondrial respiratory complexes measured as in panel a assessing ATP generation in cells respiring on pyruvate–malate (P + M) or glutamate–malate (G + M) (complex-I), succinate–rotenone (S + Rot) (complex-II) TMPD–ascorbate–antimycin-A (T + A + AntA) (complex-IV). The panels show values ± S.E.M. (n = 4). The double (P<0.01) and triple (P<0.001) asterisks, and the crosses show statistical probability versus control and ODQ, respectively

Figure 8

NO/cGMP protect from H₂O₂-induced apoptosis through inhibition of Drp1. Myogenic precursor cells were transfected with pcDNA3-Drp1-myc or pcDNA3-Drp1 K38A-myc and differentiated for 6 h. The cells were then exposed to H₂O₂ or vehicle in the presence or absence of l-NAME or ODQ. (a) Phosphatidylserine exposure to the plasma membrane in 7AAD-excluding cells. (b) The number of trypan blue-excluding cells. The panels show values±S.E.M. (n = 4). The asterisks and the crosses show significant differences (P<0.001) from the control of l-NAME or ODQ-treated cells expressing pcDNA-myc, and of the pCDNA3-Drp1 K38A-expressing cells from the cells expressing the empty vector, respectively