Secretion of adenylate kinase 1 is required for extracellular ATP synthesis in C2C12 myotubes

Abstract

Extracellular ATP (exATP) has been known to be a critical ligand regulating skeletal muscle differentiation and contractibility. ExATP synthesis was greatly increased with the high level of adenylate kinase 1 (AK1) and ATP synthase beta during C2C12 myogenesis. The exATP synthesis was abolished by the knock-down of AK1 but not by that of ATP synthase beta in C2C12 myotubes, suggesting that AK1 is required for exATP synthesis in myotubes. However, membrane-bound AK1beta was not involved in exATP synthesis because its expression level was decreased during myogenesis in spite of its localization in the lipid rafts that contain various kinds of receptors and mediate cell signal transduction, cell migration, and differentiation. Interestingly, cytoplasmic AK1 was secreted from C2C12 myotubes but not from C2C12 myoblasts. Taken together all these data, we can conclude that AK1 secretion is required for the exATP generation in myotubes.

Figure 1

The increase of exATP synthesis is accompanied by high expression level of AK1 and ATP synthase β during myogenesis. (A) C2C12 myotubes were differentiated to myotubes for 3 days. After incubating myoblasts and myotubes with ADP (200 µM), P_i (20 mM), and MgCl₂ (2 mM) for the indicated amounts of time, the ATP content was determined by bioluminescent luciferase assay. The ATP content was normalized by the protein concentration. (B) C2C12 myoblasts were differentiated to myotubes for the indicated amounts of time. The whole cell lysates were analyzed by immunoblotting with anti-AK1, ATP synthase β (ATPβ), caveolin-3 (Cav-3), myosin heavy chain (MHC), and GAPDH antibodies.

Figure 2

AK1 is required for exATP synthesis in myotubes. (A) Si-Control (Si-Con), Si-AK1, or Si-ATP synthase β (Si-ATPβ) was treated in myoblasts that were further differentiated for 3 days. The whole cell lysates were analyzed by immunoblotting with anti-ATP synthase β, AK1, Cav-3, and MHC antibodies. (B) Si-Con-, Si-ATPβ-, or Si-AK1-treated myotubes were analyzed by immunofluorescence with anti-MHC antibody. The myotubes were also stained with DAPI. The white bar indicates a length of 50 µm. (C) The exATP content was measured in the myotubes down-regulating AK1 or ATP synthase β after the cells were incubated with ADP (200 µM), P_i (20 mM), and MgCl₂ (2 mM) for 1 min. ATP content was normalized by the protein concentration. (D) Myotubes that had been differentiated for three days were pretreated with 100 µM Ap₅A or 20 µg/ml oligomycin for 30 min, and the exATP content was measured after the cells had been incubated with ADP (200 µM), P_i (20 mM), and MgCl₂ (2 mM) for 1 min. Ethanol was used as a vehicle for oligomycin. The ATP content was normalized by the protein concentration.

Figure 3

AK1β was localized in the lipid rafts of myoblasts and diminished during myogenesis. (A) Myoblasts and 3-days-differentiated myotubes were fractionated into cytoplasm and sarcolemma (= plasma membrane of skeletal muscle). AK1 and insulin receptor β(IRβ were analyzed by immunoblotting. WCL, whole cell lysates; S, sarcolemma Cyt, cytoplasm. (B) ATP content was measured after incubating sarcolemma with ADP (200 µM), P_i (20 mM), and MgCl₂ (2 mM) for 1 min. The ATP content was normalized by the protein concentration of sarcolemma. (C) Lipid rafts were isolated from myoblasts and myotubes, based on detergent insolubility and low density (Brown et al., 1992). After sucrose gradient ultracentrifugation, each fraction was analyzed by immunoblotting with anti-flotillin-1 and AK1 antibodies. Fraction number 1, and 2 represent non-raft fractions whereas fraction number 7, and 8 do raft fractions. P indicates pellet. Non-raft fraction (fraction number 1) and raft fraction (fraction number 8) were also analyzed by immunoblotting for side-by-side comparison of AK1 and AK1β with different molecular weights. (D) Co-localization of AK1 and Caveolin-1 in myoblasts and AK1 and Caveolin-3 in 3-days-differentiated myotubes was determined by immunofluorescence. The cells were also stained with DAPI. The white bar indicates a length of 10 µm. (E) ATP content was measured after incubating the lipid rafts of myoblasts or myotubes with ADP (200 µM), P_i (20 mM), and MgCl₂ (2 mM) for the indicated amounts of time. The ATP content was normalized by the protein concentration of the lipid rafts. (F) The lipid rafts from myoblasts or 3-days-differentated myotubes were preincubated with Ap₅A (100 µM) for 30 min, and were then incubated with ADP (200 µM), P_i (20 mM), and MgCl₂ (2 mM) for 1 min before the measurement of ATP content. The ATP content was normalized by the protein concentration of the lipid rafts.

Figure 4

The extracellular fluid of myotubes has a strong AK1 activity and contains the secreted form of AK1. (A) Myoblasts and 3-days-differentiated myotubes were incubated with HEPES buffer for 1 min. The buffer was collected and incubated with ADP (200 µM) and MgCl₂ (2 mM) for 1 min before the measurement of ATP content. It should be noted that only ADP was added into extracellular fluid in order to measure just the activity of AK1 that reacts 2ADP ↔ ATP + AMP. The ATP content was normalized by the protein concentration of the whole cell lysates. (B, C). Myotubes that had been differentiated for three days were downregulated by Si-AK1 (B) or pretreated with Ap₅A (100 µM) for 30 min (C), and were then incubated with HEPES buffer for 1 min. The buffer was collected and reacted with ADP (200 µM) and MgCl₂ (2 mM) for 1 min before the ATP content was measured. The ATP content was normalized by the protein concentration of whole cell lysates. (D) Myoblasts and 3-days-differentiated myotubes were incubated with serum-free media for the indicated amounts of time.The media were collected, and then concentrated by using Amicon Ultra. Whole cell lysates (WCL) and media were analyzed by immunoblooting with anti-AK1 and β-actin antibodies.