ISSN 00062979, Biochemistry (Moscow), 2014, Vol. 79, No. 6, pp. 555565. © Pleiades Publishing, Ltd., 2014. Original Russian Text © Yu. L. Baburina, A. E. Gordeeva, D. A. Moshkov, O. V. Krestinina, A. A. Azarashvili, I. V. Odinokova, T. S. Azarashvili, 2014, published in Biokhimiya, 2014, Vol. 79, No. 6, pp. 705717.
Interaction of Myelin Basic Protein and 2′′,3′′Cyclic Nucleotide Phosphodiesterase with Mitochondria Yu. L. Baburina*, A. E. Gordeeva, D. A. Moshkov#, O. V. Krestinina, A. A. Azarashvili, I. V. Odinokova, and T. S. Azarashvili* Institute of Theoretical and Experimental Biophysics, Russian Academy of Sciences, 142290 Pushchino, Moscow Region, Russia; fax: (4967) 330553; Email: [email protected]; [email protected] Received December 11, 2013 Revision received February 24, 2014 Abstract—The content and distribution of myelin basic protein (MBP) isoforms (17 and 21.5 kDa) as well as 2′,3′cyclic nucleotide3′phosphodiesterase (CNPase) were determined in mitochondrial fractions (myelin fraction, synaptic and non synaptic mitochondria) obtained after separation of brain mitochondria by Percoll density gradient. All the fractions could accumulate calcium, maintain membrane potential, and initiate the opening of the permeability transition pore (mPTP) in response to calcium overloading. Native mitochondria and structural contacts between membranes of myelin and mito chondria were found in the myelin fraction associated with brain mitochondria. Using Western blot, it was shown that addi tion of myelin fraction associated with brain mitochondria to the suspension of liver mitochondria can lead to binding of CNPase and MBP, present in the fraction with liver mitochondria under the conditions of both closed and opened mPTP. However, induction of mPTP opening in liver mitochondria was prevented in the presence of myelin fraction associated with brain mitochondria (Ca2+ release rate was decreased 1.5fold, calcium retention time was doubled, and swelling amplitude was 2.8fold reduced). These results indicate possible protective properties of MBP and CNPase. DOI: 10.1134/S0006297914060091 Key words: mitochondria, myelin, permeability transition pore, myelin basic protein, CNPase
Mitochondria are involved in cell death and play a key role in the survival of brain cells. Programmed death of neurons and other brain cells occurs both under acute brain injury (ischemia, trauma, CNS infections) and in chronic diseases such as Alzheimer’s and Parkinson’s and also during aging. Apoptosis, necrosis, and autophagy take part in the processes of neuronal cell death in neu rodegenerative diseases [1]; a high percentage of damaged mitochondria were found under these conditions [2, 3]. Neurodegenerative diseases are accompanied by disrup tion of the integrity of the myelin sheath – a membrane structure that protects the axon and ensures the spreading of electric impulses over long distances at high speed. Myelin membrane differs from other biological mem branes by its high lipid/protein ratio and the presence of specific proteins, myelin proteolipid protein (PLP), myelin basic protein (MBP), and 2′,3′cyclic nucleotide Abbreviations: CNPase, 2′,3′cyclic nucleotide3′phosphodi esterase; MBP, myelin basic protein; MFAM, myelin fraction associated with mitochondria; mPTP, mitochondrial perme ability transition pore; PLP, proteolipid protein. * To whom correspondence should be addressed. # Deceased.
3′phosphodiesterase (CNPase) which are the major pro teins. Interest in the localization and role of myelin pro teins outside the myelin sheath and, in particular, their interaction with mitochondria has significantly increased in recent years. The presence of myelin proteins in non myelinated tissues such as liver, kidney, pancreas, blood, and cerebral fluid has been described [47]. It has also been shown that CNPase, MBP, and PLP are able to bind to mitochondria of nonmyelinated tissues [5, 6, 8]. The role of association of these proteins with mitochondria is poorly understood both for mitochondria and for cells at present. However, the fact that these three myelin pro teins can play special, yet unknown functions being out side myelin remains indisputable [46, 8, 9]. CNPase has recently been identified in mitochondria isolated from brain, liver, and oligodendrocytes; this enzyme was shown to be involved in regulation of the nonspecific pore (mPTP). In particular, it has been shown that decrease in CNPase expression leads to the reduction of the threshold calcium concentration inducing mPTP opening, and CNPase substrates (2′,3′cAMP and 2′,3′cNADP) can accelerate mPTP opening [9]. Another myelin protein, MBP, was also found in tissues not containing myelin, i.e. in liver, kidney, and lung [5, 6]. The association of MBP
555
556
BABURINA et al.
with mitochondria was demonstrated in pancreas cells by immunoelectron microscopy [7]. MBP was detected there in the outer and inner membranes as well as in mitochondrial cristae. Association myelin PLP with mitochondria has been also demonstrated both in vitro and in vivo using confocal microscopy [8, 10]. These studies showed that PLP in COS7 cells can be embedded in the inner mitochondrial membrane. It was noted that the proteolipid Nterminus contains motifs that could be colocalized with mitochon dria. Deletion of the first 1020 amino acid residues at the proteolipid Nterminus results in the loss of the ability of PLP to interact with mitochondria [8]. In this regard, it is worth to note that PLP includes a sequence of the “c” subunit of the Fo sector of mitochondrial FoF1 ATP syn thase [11, 12], which can be involved in the functioning of the nonspecific pore [1317]. It should be noted that despite years of examination of permeability transition pore, its structural composition and mechanism of its regulation remain not fully understood. However, while investigating the regulation of the mPTP functioning, we discovered a new molecular mechanism of its regulation based on Ca2+dependent phosphorylation/dephospho rylation of mitochondrial membranebound proteins with molecular weights of 4446, 21.5, 17, and 3.5 kDa; the phosphorylation status of these proteins was shown to change depending on the induction of the nonselective pore opening [1316]. Up to now, these phosphoproteins have been identified. The phosphoprotein with molecular weight of 46 kDa was identified in highly purified rat brain mitochondria as 2′,3′cyclic nucleotide 3′phos phodiesterase (CNPase) [9], and the phosphopeptide of 3.5 kDa was detected as the hydrophobic subunit “c” of the membrane FoATP synthase [13, 16]. Two MBP phosphoisoforms (17 and 21.5 kDa) have been recently identified in preparations of brain mitochondria (which were not purified on Percoll gradient) by 2D elec trophoresis with following mass spectrometry [18]. However, the role of these proteins in mitochondria has not been still studied. At the present work, induction of mPTP opening in all fractions obtained after separation of crude mitochon dria on Percoll gradient as well as distribution there of CNPase and MBP were examined. Moreover, using myelin fraction associated with brain as source of myelin proteins, the effect of myelin proteins (CNPase and MBP) on mPTP induction in rat liver mitochondria was examined to find out whether association of CNP and MBP can modulate mPTP opening.
MATERIALS AND METHODS Isolation of rat brain mitochondria. Rat brain mito chondria were isolated by the method of Sims as modified in our laboratory [19]. Male Wistar rats (200250 g) were
used for the experiments. The rats were decapitated, and the brain was rapidly removed, cooled on ice in physio logical saline, and homogenized at 04°C in a tenfold medium volume (320 mM sucrose, 10 mM TrisHCl, pH 7.4, 0.5 mM K+EDTA, 0.5 mM K+EGTA, 0.2% BSA). The homogenate was centrifuged at 2000g for 3 min, the precipitate was removed, and the supernatant was centrifuged again for more complete precipitation of nuclei and damaged cells. The supernatant was cen trifuged at 4°C at 12,500g for 10 min to pellet the mito chondria. Mitochondrial suspension (1.5 ml) were first mixed with 3% Percoll (1 ml), layered to obtain mito chondrial fractions in the Percoll density gradient (1015 24%), and then centrifuged at 31,500g for 10 min. Precipitates of nonsynaptic mitochondria, synaptic mito chondria, and myelin fraction were washed using the extraction medium without EDTA, EGTA, and BSA at 11,500g for 10 min and resuspended in the same medium. Isolation of rat liver mitochondria by a standard method. Isolated liver was homogenized at 04°C in a ten fold medium volume (220 mM mannitol, 70 mM sucrose, 10 mM TrisHCl, pH 7.5, 1 mM EGTA, 0.05% BSA). The homogenate was centrifuged at 800g for 10 min to pellet the nuclei and damaged cells. The supernatant containing mitochondria was centrifuged for 10 min at 9000g to pre cipitate the mitochondria. The mitochondria were washed twice in medium containing EGTA and BSA for 10 min at 9000g and were resuspended in the same medium. Determination of mitochondrial functions. Changes in the mitochondrial membrane potential and calcium ion concentration were measured in a thermostatted sam ple chamber with installed TPP+ and Ca2+selective electrodes as described previously [9]. The incubation medium contained 10 mM TrisHCl, pH 7.4, 120 mM KCl, 0.4 mM KH2PO4, 5 mM potassium succinate, and 2.5 μM rotenone. Potassium succinate was used as the substrate for mitochondrial respiration. The protein con centration in the sample chamber was 1 mg/ml. Opening of the mPTP in mitochondria was induced by the addi tion of calcium to reach its threshold concentration required to initiate mPTP opening. Each Ca2+ addition was 100 nmol per mg protein. All experiments were per formed in an open sample chamber. Swelling of rat brain and liver mitochondria was fol lowed by the kinetics of light scattering by the mitochon drial suspension at wavelength 540 nm using a PE5400V spectrophotometer (Russia) at 25°C in isotonic medium containing 10 mM TrisHCl buffer, pH 7.5, 125 mM KCl, 2 mM KH2PO4, 5 mM potassium succinate, and 2.5 μM rotenone. Protein concentration in the sample chamber was 1 mg/ml for liver mitochondria and 0.4 mg/ml for brain mitochondria. Electrophoresis and Western blot. Electrophoresis under denaturing conditions was carried out in a mini chamber (Hoefer) by the Laemmli method [20]. Ten micrograms of protein was applied to each gel lane. BIOCHEMISTRY (Moscow) Vol. 79 No. 6 2014
ASSOCIATION OF MYELIN PROTEINS WITH MITOCHONDRIA Pharmacia Biotech (USA) sets containing high (14.4 97 kDa) and low molecular mass (3.4616.9 kDa) pro teins were used as markers. MBP and CNP were detected in mitochondria by the Western blot method. The protein was transferred from the gel to a nitrocellulose membrane (Sigma, USA) on a semidry transfer apparatus (BioRad, USA) using buffer containing 0.048 M TrisHCl, pH 9.2, 0.039 M glycine, 20% methanol (v/v), and 0.0357% SDS. Monoclonal mouse antiCNP antibodies (Mab 461 obtained in the laboratory of Prof. G. Reiser, dilution 1 : 10,000) and polyclonal goat antiMBP antibodies (Santa Cruz Biotechnology, USA; dilution 1 : 2000) were used for protein identification. Secondary antibodies conju gated to horseradish peroxidase were also used. Peroxidase activity was determined with the ECL chemi luminescent reagent (Pierce Chemical Co, USA) accord ing to the manufacturer’s instruction. Electron microscopy of isolated mitochondria. Brain mitochondria samples were prepared for electron microscopy by fixation for 2 h in 2.5% glutaraldehyde (AppliChem GmbH, Germany) dissolved in the isolation medium (pH 7.4), washing once after resuspension in the isolation medium, and fixation in 1% solution of osmium tetroxide (Serva Electrophoresis GmbH, Germany) in isolation medium with sucrose added to reach the iso tonicity with the previous washing solution. Precipitated samples were dehydrated without resuspension in water solutions with increasing ethanol concentrations (30, 50, 75, and 96%) and in absolute ethanol (incubation in each mixture for 1 h), in absolute acetone (three times for 1 h), and embedded in Epon 812 epoxy resin (Fluka Chemie AG, Switzerland). Ultrathin sections were prepared on an EM UC6 ultramicrotome (Leica, Germany); the samples were contrasted in saturated (4%) aqueous solution of uranyl acetate and in an alkaline solution of lead citrate (pH 12) and were examined with a Tesla BS500 electron microscope (Czechia) at accelerating voltage 90 kV. Statistical analysis. The data were analyzed and visu alized using the Sigma Plot 9.0 software. The average value of the parameters from 56 experiments was taken for statistical analysis. Significance was taken at p < 0.05 vs the control using the Student ttest.
RESULTS Since the suspension of brain mitochondria is het erogeneous and includes two mitochondrial pools (synap tic and nonsynaptic) as well as associated myelin fraction, the preparation of brain mitochondria obtained by differ ential centrifugation was subjected to fractionation on a Percoll density gradient. We examined the presence of myelin proteins in the fractions, i.e. CNPase and MBP (Fig. 1). Figure 1 shows the distribution of CNPase and MBP by Western blot analysis using antibodies highly spe cific to CNPase and polyclonal antibodies to MBP react BIOCHEMISTRY (Moscow) Vol. 79 No. 6 2014
557
ing with all four MBP isoforms. Figure 1a shows that the highest CNPase content is observed in the myelin frac tion, and its level decreases in mitochondrial fractions (synaptic and nonsynaptic). If we take the CNPase con tent in the myelin fraction as 100%, then its content in synaptic and nonsynaptic mitochondria is 21 and 17%, respectively (see table). The presence of MBP was also checked in all fractions (Fig. 1a). It is evident that two MBP isoforms (21.5 and 17 kDa) are associated with mitochondria. They can be removed to a large extent after the separation of mitochondria in the Percoll density gra dient, nevertheless, a small portion of the MBP remains strongly bound with mitochondria (8% with synaptic and 1.5% with nonsynaptic mitochondria; see table). The purity of the mitochondrial fractions were carefully stud ied earlier and confirmed by specific protein markers (synaptotagmin, COX IV) and by electron microscopy [21]. The presence of CNPase and MBP in all the frac tions of brain mitochondria indicates their high affinity to mitochondria. We studied the functioning of Ca2+ induced and CsAsensitive mitochondrial pore (mPTP) by measuring such mPTP characteristics as calcium capacity, rate of Ca2+ efflux, and Ca2+stimulated swelling of mitochondria. Figure 1b shows the average values of calcium capacity and swelling halftime (T1/2). Synaptic mitochondria were shown to have the highest T1/2 value (310 s) and calcium capacity (350 μM) in our experi ments. The swelling halftime of nonsynaptic mitochon dria was about the same (300 s), but the calcium capacity was lower by 15% (300 μM) (see table). These data are consistent with our previously described results [22]. It is interesting to note that membrane potential was regis tered in the myelin fraction associated with mitochondria (MFAM) using the TPP+electrode (not shown), and this fraction was still able to accumulate calcium despite the low calcium capacity in MFAM, which was 6.6fold lower (30 μM) than in the nonsynaptic mitochondria fraction. However, it did not undergo calciuminduced swelling. On the other hand, the experiments using a chamber with installed electrodes showed that sequential calcium addi tion to this fraction resulted in Ca2+dependent and CsA sensitive calcium release from mitochondria, which was observed in mitochondria under mPTP induction. This might indicate the presence of a small pool of intact func tioning mitochondria interacting with myelin membrane in MFAM. This possibility was confirmed by the results of electron microscopy (Fig. 2), which showed the presence of a certain number of intact mitochondria, myelin sheaths, and their fragments in MFAM, where brain mitochondria are likely to form some contact sites with them (Fig. 2). The pictures show that contacts can be formed between mitochondria and myelin membrane (Fig. 2a) as well as between mitochondria and the frag ments of myelin membrane (Fig. 2, b and c). Previously it was shown that a change in the CNPase expression level in oligodendrocytes affects the activation
558
BABURINA et al. Comparative characteristics of mitochondrial fractions Fraction
Swelling halftime, s
Calcium capacity, μmol
CNPase content, %
MBP content, %
–
18
100
100
Synaptic mitochondria
310
350
21
8
Nonsynaptic mitochondria
300
300
17.3
1.5
MFAM
of mPTP opening [9]. In the present work, the effect of myelin proteins (CNPase and MBP) on the induction of mPTP opening was studied. As there is a small amount of myelin proteins present in brain mitochondria, mito chondria isolated from liver, which has no myelin, were
used to study the effect of these proteins on the function ing of the nonselective mitochondrial pore. The ability of myelin proteins found in MFAM to associate with liver mitochondria both under control conditions (before cal cium addition) and under the conditions of nonselective
a kDa kDa
МВР
21.5 CNP
46
17
1
2
3 1
2
3
b Swelling halftime
350
500
Calcium capacity of mitochondria 300 400
300
200
T1/2, s
[Ca2+], μmol
250 200 150 100 100 50 0
0 2
3
1
Fig. 1. a) Distribution of myelin proteins (CNPase and MBP) in rat brain mitochondria: 1) myelin fraction (MFAM); 2) synaptic mitochon dria; 3) nonsynaptic mitochondria; b) functional characteristics of mitochondrial fractions obtained after separation on a Percoll density gra dient.
BIOCHEMISTRY (Moscow) Vol. 79 No. 6 2014
ASSOCIATION OF MYELIN PROTEINS WITH MITOCHONDRIA
0.1 μm
0.1 μm
a
0.1 μm
559
b
c
Fig. 2. Ultrastructure of contact sites of isolated brain mitochondria with fragments of myelin membranes. Arrows indicate. a) Contact of mitochondria with myelin membrane; b) contact of mitochondria with myelin fragments; c) contact inside and outside of a myelin fragment. Scale 0.1 μm.
pore opening (in the presence of the threshold calcium concentration) was determined. Liver mitochondria were incubated with MFAM (at protein concentration of 0.1 and 0.5 mg/ml) for 10 min. After incubation, aliquots of mitochondria were taken from the cell in two states: a) without calcium addition (control), and b) in the pres ence of threshold calcium concentrations that initiated pore opening. The samples were centrifuged, the mito BIOCHEMISTRY (Moscow) Vol. 79 No. 6 2014
chondrial precipitate was solubilized in Laemmli solu tion, and then the mitochondrial proteins were separated by SDSPAGE. The proteins were then transferred onto a nitrocellulose membrane that was treated with antiCNP and antiMBP antibodies. Figure 3a shows that MBP is absent in liver mitochondria both in control and under pore opening (lanes 1 and 2). After incubation of liver mitochondria with MFAM, association of the 17kDa
560
BABURINA et al.
MBP isoform with mitochondria was observed (lanes 3 and 4); this association was enhanced in the presence of calcium. The diagram below shows that calcium increased twofold the association of 17 kDa MBP with mitochondria in the presence of 0.1 mg/ml MFAM. Increase of the concentration of the added myelin frac tion (lanes 5 and 6) leads to higher amount of associated MBP isoform (17 kDa). In the absence of calcium, almost a fivefold increase in MBP binding was observed, which was not the case in the presence of calcium (Fig. 3a, diagram). In addition, the 21.5kDa MBP isoform appeared under incubation of 0.5 mg/ml MFAM with mitochondria. However, the level of associated myelin basic protein (17 and 21.5 kDa) does not depend on cal cium under these conditions. It should be noted that the phosphorylation status of both MBP isoforms increased under pore opening conditions [18]. Figure 3b shows the binding of CNPase from MFAM with liver mitochondria under the same conditions. Also seen, that isolated liver mitochondria originally contain a certain amount of CNPase. Addition of MFAM fraction to the mitochondr
ial suspension results in concentrationdependent bind ing of CNPase from the MFAM fraction with liver mito chondria, so that the increase in the concentration of added MFAM is accompanied by increased CNPase association with mitochondria (Fig. 3b, lanes 36). The diagram in the lower part of Fig. 3b shows that incubation of 0.1 mg/ml MFAM with mitochondria leads to twofold increase in CNPase binding, and incubation with 0.5 mg/ml MFAM results in fourfold increase in protein association. Phosphodiesterase binding with mitochon dria showed no dependence on calcium. Thus, MBP and CNPase association with mitochondria can proceed under conditions of both open and closed pore. Next we tested the effect of MBP and CNPase binding with mito chondria on the functioning of the Ca2+induced and CsAsensitive pore. To elucidate this issue, we studied the effect of added brain MFAM on Ca2+ release from liver mitochondria and Ca2+induced highamplitude swelling of liver mitochondria. Figure 4 shows that the addition of MFAM to liver mitochondria alters the rate of calcium release and the duration of the lag phase (calcium reten
a
b
kDa
kDa
21.5 46 17
МВР 1
2
Mch
3
4
5
1
4
5
6
CNP
incubation
incubation
400 Protein, % of control
Protein, % of control
3
Mch
6
500 400 300 200 100 0
2
300 200 100 0
1
2
3
4
5
6
1
2
3
4
5
6
Fig. 3. Association of myelin proteins with liver mitochondria under pore opening: a) MBP association; b) CNPase association. In the lower part of the figure the control corresponds to the values of association of the studied proteins with mitochondria when MFAM was added before calcium additions. The concentration of antiCNP antibodies was 1 : 10,000; antiMBP antibodies 1 : 2000. Mch: 1) liver mitochondria (Mch) before calcium addition; 2) Mch after mPTP induction. Incubation: 3) Mch + MFAM (0.1 mg/ml) before calcium addition (control); 4) Mch + MFAM (0.1 mg/ml) after mPTP induction; 5) Mch + MFAM (0.5 mg/ml) without calcium addition; 6) Mch + MFAM (0.5 mg/ml) after mPTP induction.
BIOCHEMISTRY (Moscow) Vol. 79 No. 6 2014
ASSOCIATION OF MYELIN PROTEINS WITH MITOCHONDRIA
a
b Lag phase 180
160
160
*
140
120
100
80
* 60
40
20
Relative values of lag phase duration, % of control
Relative values of calcium release rate, % of control
Calcium release rate 180
0
561
140
120
100
80 60
40
20
1
2
0
1
2
Fig. 4. Effect of myelin fraction of brain mitochondria on rate of Ca2+ release (a) and lag phase duration (b) under mPTP opening. Concentration of myelin fraction protein 0.1 (1) and 0.5 (2) mg/ml. Incubation time 10 min.
tion time in liver mitochondria before its release indicat ing the initiation of the pore opening). Figure 4a shows that the addition of 0.1 mg/ml protein has no effect on the rate of calcium release and the duration of the lag phase. However, the myelin fraction of brain mitochon dria at protein concentration 0.5 mg/ml reliably reduced the rate of calcium release from liver mitochondria by 32% against the control; at the same time, the period of the lag phase increased by 57% against the control (Fig. 4b). The studies of the effect of MFAM proteins on the Ca2+induced highamplitude swelling of liver mitochon dria (Fig. 5a, lower part) showed that the maximum amplitude of the swelling was observed when mitochon dria were loaded with calcium to its threshold concentra tion initiating the pore opening (control, curve 1). Incubation of liver mitochondria with MFAM led to decrease in the swelling amplitude (curves 2 and 3); the swelling amplitude decreased 2fold in the presence of 0.1 mg/ml MFAM and almost 3fold in the presence of 0.5 mg/ml of MFAM protein (Fig. 5a, upper part). Thus, by reducing the amplitude of liver mitochondria swelling, MFAM increases the resistance of liver mitochondria to BIOCHEMISTRY (Moscow) Vol. 79 No. 6 2014
the initiation of nonselective pore opening. This observa tion is consistent with the results presented in Fig. 1 and in the table, which demonstrate that mitochondria from the myelin fraction containing CNPase and MBP do not undergo swelling. These results suggest the presence of protector proteins in MFAM that prevent mPTP open ing. Because we earlier identified CNPase and MBP in this fraction, we suggested that they could fulfill this pro tective role, although still unidentified proteins can also be present in this fraction. The protective effect of CNPase and MBP was verified in experiments when CNPase and MBPcontaining MFAM was added to the suspension of mitochondria together with antibodies pro duced to these proteins (antiCNP antibodies and anti MBP antibodies). We assumed that the added antibodies would bind the proteins in the solution or on the outer mitochondrial membrane and would thus not prevent mitochondrial swelling. Simultaneous presence of the proteins and antibodies to them confirmed that CNPase and MBP increase the stability of the mitochondria. Figure 5b shows that Ca2+induced swelling of liver mito chondria (lower part of the figure, curve 1) is suppressed
562
BABURINA et al.
in the presence of added MFAM (curve 2). Incubation of calciumloaded liver mitochondria in the presence of MFAM and antiCNP antibodies (curve 3) or antiMBP antibodies (curve 4) reduced the ability of the myelin fraction to prevent mitochondrial swelling. In particular, the reduction was 34% in the case of antiCNP antibod ies (upper part of Fig. 5b, column 3) and 38% for anti MBP antibodies (upper part of Fig. 5b, column 4). This effect was CsAsensitive, i.e. no mitochondrial swelling
could be observed under similar conditions in the pres ence of CsA, the studied antibodies, and MFAM. The specificity of the protective effect of myelin proteins was confirmed by replacing antibodies to CNPase and MBP by the protein of bovine serum albumin (BSA, fraction V) at the same protein concentration. Figure 5a shows that BSA had no effect on the swelling of liver mitochondria (lower part of Fig. 5a, curves 4 and 5, and upper part of the figure, columns 4 and 5).
a
b 1.0
Swelling amplitude, absolute units
Swelling amplitude, absolute units
1.0
0.8
0.6
* 0.4
*
0.2
0.6
0.4
0.2
0 1
2
3
4
*
*
0.8
0
5
2.2
2.2
2.0
2.0
1.8
1.8
1
2
3
4
2 1.6
1.6
3
1.4
2
1.4
4
3
5 1
1.2
4
1.2 1
1.0
1.0 0
100
200
300
400
0
100
200
300
400
Time, s Fig. 5. Effect of the myelin fraction of brain mitochondria on swelling of rat liver mitochondria. Upper panel, quantitative characteristics of swelling amplitude of liver mitochondria incubated with myelin fraction of brain mitochondria; lower panel, swelling curves. a) Highamplitude calciuminduced swelling of liver mitochondria incubated with the myelin fraction of brain mitochondria. Concentration of the myelin fraction protein 0.1 and 0.5 mg/ml: 1) Mch + Ca2+; 2) Mch + MFAM (0.1 mg/ml) + Ca2+; 3) Mch + MFAM (0.5 mg/ml) + Ca2+; 4) Mch + BSA (0.1 mg/ml) + Ca2+; 5) Mch + BSA (0.5 mg/ml) + Ca2+. b) Swelling of liver mitochondria incubated with myelin fraction and antiCNP and antiMBP antibodies (Ab). Concentration of the myelin fraction protein 0.1 and 0.5 mg/ml: 1) Mch + Ca2+; 2) Mch + MFAM + Ca2+; 3) Mch + MFAM + antiCNP Ab + Ca2+; 4) Mch + MFAM + antiMBP Ab + Ca2+. Concentration of antiMBP Ab 1 : 2000, antiCNP Ab 1 : 10,000.
BIOCHEMISTRY (Moscow) Vol. 79 No. 6 2014
ASSOCIATION OF MYELIN PROTEINS WITH MITOCHONDRIA DISCUSSION Damage of the integrity of the myelin membrane is one of the factors stimulating neurodegenerative process es leading to disruption of the conductivity of nerve impulses [23]. It was noted that degradation of myelin proteins observed in acute or chronic neuropathologies could result from the activation of the nonselective mito chondrial pore. This is confirmed by the detection of myelin proteins outside myelin sheets and, in particular, in mitochondria isolated from different tissues and cells [19]. One of the major myelin proteins, CNPase, exists in two isoforms with molecular weights of 46 and 48 kDa [24]. The primary structure of myelin CNPase can be divided into two areas different in their properties. The N terminus has binding sites for nucleotide phosphates [24 26] and for ATP/GTPbinding proteins, while a catalytic site [27], tubulin and RNAbinding domain [28], and CAAX sequence (the motif for protein acylation [26]) are located at the CNPase Cterminus. However, it is not yet clear which properties of CNPase remain outside myelin. It was recently shown that CNPase identified in mito chondria can be involved in the regulation of mPTP func tioning, in particular, its substrates (2′,3′cAMP and 2′,3′cNADP) accelerate mPTP opening in the presence of threshold calcium concentrations [9]. Decrease of CNPase expression in OLN93 cells (knockdown CNP OLN93) was shown to decrease threshold calcium con centration required for the induction of pore opening. These results suggest that the increase in CNPase level can prevent mPTP opening. Since the suspension of iso lated brain mitochondria is heterogeneous, they can be purified in a Percoll density gradient, giving nonsynaptic and synaptic mitochondria as well as a fraction of mito chondriaassociated myelin (MFAM). CNPase and MBP isoforms were identified in the fractions. The highest MBP content was found in the myelin fraction with its two major bands corresponding to 17 and 21.5 kDa iso forms. In synaptic and nonsynaptic brain mitochondria, MBP was detected in small amounts, reflecting a small pool of the protein strongly associated with mitochondria (Fig. 1). There are some literature data suggesting that intravenously administered 125Ilabeled MBP can be found in liver, kidney, lung, and pancreatic islet cells [5, 6]. The binding of MBP with mitochondria was shown by immunoelectron microscopy [6]. On the other hand, an increasing number of publications report the presence of subunits of FoF1ATPase and subunits of respiratory chain complexes in myelin [2931], probably due to the presence of mitochondria or their fragments during myelin isolation [32]. Taking these data into considera tion, we examined functioning of the nonselective Ca2+ induced pore in all three fractions. We found that both types of mitochondria maintained high membrane poten tial and calcium capacity, which decreased when mito chondria were loaded with calcium. This decrease was BIOCHEMISTRY (Moscow) Vol. 79 No. 6 2014
563
cyclosporinsensitive. Furthermore, the mitochondria were capable of Ca2+induced swelling (Fig. 1). Maintained membrane potential was revealed and in the myelin fraction associated with mitochondria (MFAM) (not shown); increasing calcium concentration caused Ca2+induced calcium release. All these processes were CsAsensitive supporting involvement of mPTP phenom enon. However, despite the fact that calcium accumula tion was observed in the brain myelin fraction, it did not undergo swelling that was previously described in the studies on isolated myelin [33, 34]. The presence of a reg istered membrane potential in MFAM and the ability to accumulate calcium indicated the possible presence of a pool of intact mitochondria in MFAM. Electron microscopy data (Fig. 2) showed that a small pool of mitochondria and fragments of myelin membranes are indeed present in the myelin fraction of brain mitochon dria; brain mitochondria probably contact them similar to mitochondrial contacts with endoplasmic reticulum (forming contact sites). It should be noted that myelin membrane contacts with mitochondria can be also seen in electron micrographs in publications dedicated to the research in completely different areas [35]. Despite the fact that myelin protein association has been described in some papers, their effect on mitochondrial functions and in particular on the mPTP functioning has not been pre viously studied. Due to the fact that myelin proteins asso ciate firmly with brain mitochondria, their effect on the mPTP induction was studied with liver mitochondria, which are free of myelin. The experiments revealed that MBP and CNPase from MFAM can bind with the liver mitochondrial membrane. It should be noted that MBP interacts with lipids and can penetrate the membrane bilayer [36]. Association of these proteins with mitochon dria increases when the concentration of added myelin fraction is increased from 0.1 to 0.5 mg/ml protein. It is interesting to note that only two (out of four) minor MBP isoforms with molecular masses of 21.5 and 17 kDa have ability to bind with the mitochondrial membrane. It was previously shown that the 21.5 kDa isoform has a signal amino acid sequence corresponding to nuclear localiza tion, which can transport the protein into nuclei [37]. MBP binding with liver mitochondria has been also found under conditions of initiation of mPTP opening; the degree of association was not related with calcium con centration. However, we recently showed that the phos phorylation status of the 17 and 21.5 kDa isoforms increases in the presence of threshold concentrations of calcium [18]. Similar results were obtained for CNPase, which also binds with liver mitochondria in the presence of added MFAM fraction, despite the fact that liver mito chondria possess an independent pool of this protein [7]. Incubation of MFAM with liver mitochondria leads to additional CNPase binding. Similarly to MBP, associa tion of CNPase with liver mitochondria depends on the concentration of the added protein, but not on added cal
564
BABURINA et al.
cium. CNPase binding takes place both when the pore is closed (control) and when it is opened. Similarly to MBP, pore opening causes increase in the CNPase phosphory lation status (Baburina et al. (2014) J. Biomembr. Bioenerg., in press). The effect of MFAM on the Ca2+ induced Ca2+ release from mitochondria and Ca2+ induced high amplitude swelling were studied to deter mine the significance of MFAM binding with mitochon dria. We found that incubation of liver mitochondria with MFAM proteins reduced the rate of calcium release from liver mitochondria and increased the lag phase period (by 57%). These results indicate the protective properties of the proteins, which increase the resistance of liver mito chondria to the initiation of the pore opening in the pres ence of the myelin fraction of brain mitochondria. In this regard, it is interesting to note that one of the functions of MBP in myelin is to stabilize the myelin membrane. Since MBP and CNPase were identified in MFAM by Western blot, the protective effect of these proteins was confirmed in experiments on mitochondrial swelling in the simultaneouspresence of MFAM and antibodies against MBP and CNPase. It was found that incubation of liver mitochondria with 0.1 or 0.5 mg/ml of MFAM protein prevented mitochondrial swelling. No protective effect of MFAM could be observed when MFAM was added after incubation of mitochondria with antiMBP or antiCNP antibodies. Apparently, the binding of the antibodies with the corresponding protein on the outer membrane (and in solution) prevented them from associ ating with mitochondria, and therefore they could not prevent mitochondrial swelling. The results demonstrate that myelin proteins MBP and CNPase can execute a new unknown function being outside myelin; in particular, they can be involved in regulation of apoptosis, prevent ing cells from the initial stages of programmed cell death probably through stabilization the integrity of mitochon drial membranes. This work was supported by the Russian Foundation for Basic Research (grants 140400625, 120400671 and 130400935).
REFERENCES 1. 2.
3.
4. 5. 6.
Gorman, A. M. (2008) J. Cell Mol. Med., 12, 22632280. Calabrese, V., Scapagnini, G., Giuffrida Stella, A. M., Bates, T. E., and Clark, J. B. (2001) Neurochem. Res., 26, 739764. Aliev, G., Palacios, H. H., Walrafen, B., Lipsitt, A. E., Obrenovich, M. E., and Morales, L. (2009) Int. J. Biochem. Cell Biol., 41, 19892004. Kolehmainen, E., Knip, M., and Leppaluoto, J. (1990) Acta Physiol. Scand., 139, 493501. Kolehmainen, E. (1996) J. Neurol. Sci., 141, 1926. Kolehmainen, E., and Sormunen, R. (1998) Pancreas, 16, 176188.
7. Baburina, Yu. L., Krestinina, O. V., and Azarashvili, T. S. (2013) Neurochem. J., 7, 115. 8. Huttemann, M., Zhang, Z., Mullins, C., Bessert, D., Lee, I., Nave, K. A., Appikatla, S., and Skoff, R. P. (2009) Neuro, 1. 9. Azarashvili, T., Krestinina, O., Galvita, A., Grachev, D., Baburina, Y., Stricker, R., Evtodienko, Y., and Reiser, G. (2009) Am. J. Physiol. Cell Physiol., 296, 1428 1439. 10. Ghandour, M. S., Feutz, A. C., Jalabi, W., Taleb, O., Bessert, D., Cypher, M., Carlock, L., and Skoff, R. P. (2002) Glia, 40, 300311. 11. Knapp, P. E. (1996) Dev. Neurosci., 18, 297308. 12. Laursen, R. A., Samiullah, M., and Lees, M. B. (1983) FEBS Lett., 161, 7174. 13. Azarashvili, T. S., Tyynela, J., Odinokova, I. V., Grigorjev, P. A., Baumann, M., Evtodienko, Y. V., and Saris, N. E. (2002) J. Bioenerg. Biomembr., 34, 279284. 14. Azarashvily, T. S., Tyynela, J., Baumann, M., Evtodienko, Y. V., and Saris, N. E. (2000) Biochem. Biophys. Res. Commun., 270, 741744. 15. Krestinina, O. V., Grachev, D. E., Odinokova, I. V., Reiser, G., Evtodienko, Y. V., and Azarashvili, T. S. (2009) Biochemistry (Moscow), 74, 421429. 16. Azarashvili, T., Odinokova, I., Bakunts, A., Ternovsky, V., Krestinina, O., Tyynela, J., and Saris, N. E. (2014) Cell Calcium, 55, 6977. 17. Bonora, A., Bononi, M. E., De, C., Giorgi, M., Lebiedzinska, S., Marchi, S., Patergnani, A., Rimessi, J. M., Suski, A., Wojtala, M. R., Wieckowski, G., Kroemer, L., and Galluzzi, P. (2013) Cell Cycle, 12, 674683. 18. Krestinina, O., Makarov, P., Baburina, Yu., Gordeeva, A., and Azarashvili, T. (2013) Neurochem. J., 7, 284290. 19. Sims, N. R. (1990) J. Neurochem., 55, 698707. 20. Laemmli, U. K. (1970) Nature, 227, 680685. 21. Krestinina, O. V., Kruglov, A. G., Grachev, D. E., Baburina, Yu. L., Evtodienko, Yu. V., Moshkov, D. A., Santalova, I. M., and Azarashvili, T. S. (2010) Biochemistry (Moscow), Ser. A: Membrane and Cell Biology, 4, 180 186. 22. Azarashvili, T., Baburina, Y., Grachev, D., Krestinina, O., Evtodienko, Y., Stricker, R., and Reiser, G. (2011) Am. J. Physiol. Cell Physiol., 300, 707720. 23. Kim, T., and Pfeiffer, S. E. (1999) J. Neurocytol., 28, 281 293. 24. Lee, J., O’Neill, R. C., Park, M. W., Gravel, M., and Braun, P. E. (2006) Mol. Cell Neurosci., 31, 446462. 25. Sprinkle, T. J. (1989) Crit. Rev. Neurobiol., 4, 235301. 26. Agrawal, H. C., Sprinkle, T. J., and Agrawal, D. (1990) Biochem. Biophys. Res. Commun., 170, 817823. 27. Gravel, M., DeAngelis, D., and Braun, P. E. (1994) J. Neurosci. Res., 38, 243247. 28. Gravel, M., Robert, F., Kottis, V., Gallouzi, I. E., Pelletier, J., and Braun, P. E. (2009) J. Neurosci. Res., 87, 1069 1079. 29. Panfoli, I., Ravera, S., Bruschi, M., Candiano, G., and Morelli, A. (2011) Expert. Rev. Proteom., 8, 231239. 30. Morelli, A., Ravera, S., and Panfoli, I. (2011) Cell Biochem. Biophys., 61, 179187. 31. Ravera, S., Panfoli, I., Calzia, D., Aluigi, M. G., Bianchini, P., Diaspro, A., Mancardi, G., and Morelli, A. (2009) J. Biochem. Cell Biol., 41, 15811591. BIOCHEMISTRY (Moscow) Vol. 79 No. 6 2014
ASSOCIATION OF MYELIN PROTEINS WITH MITOCHONDRIA 32. Harris, J. J., and Attwell, D. (2013) J. Cereb. Blood Flow Metab., 33, 3336. 33. Karthigasan, J., and Kirschner, D. A. (1988) J. Neurochem., 51, 228236. 34. Rand, R. P., Fuller, N. L., and Lis, L. J. (1979) Nature, 279, 258260. 35. Zambonin, J. L., Zhao, C., Ohno, N., Campbell, G. R., Engeham, S., Ziabreva, I., Schwarz, N., Lee, S. E.,
BIOCHEMISTRY (Moscow) Vol. 79 No. 6 2014
565
Frischer, J. M., Turnbull, D. M., Trapp, B. D., Lassmann, H., Franklin, R. J., and Mahad, D. J. (2011) Brain, 134, 19011913. 36. Beest, M., and Hoekstra, D. (1993) Eur. J. Biochem., 211, 689696. 37. Smith, G. S., Seymour, L. V., Boggs, J. M., and Harauz, G. (2012) Biochem. Biophys. Res. Commun., 422, 670 675.
Interaction of Myelin Basic Protein and 2′′,3′′Cyclic Nucleotide Phosphodiesterase with Mitochondria Yu. L. Baburina*, A. E. Gordeeva, D. A. Moshkov#, O. V. Krestinina, A. A. Azarashvili, I. V. Odinokova, and T. S. Azarashvili* Institute of Theoretical and Experimental Biophysics, Russian Academy of Sciences, 142290 Pushchino, Moscow Region, Russia; fax: (4967) 330553; Email: [email protected]; [email protected] Received December 11, 2013 Revision received February 24, 2014 Abstract—The content and distribution of myelin basic protein (MBP) isoforms (17 and 21.5 kDa) as well as 2′,3′cyclic nucleotide3′phosphodiesterase (CNPase) were determined in mitochondrial fractions (myelin fraction, synaptic and non synaptic mitochondria) obtained after separation of brain mitochondria by Percoll density gradient. All the fractions could accumulate calcium, maintain membrane potential, and initiate the opening of the permeability transition pore (mPTP) in response to calcium overloading. Native mitochondria and structural contacts between membranes of myelin and mito chondria were found in the myelin fraction associated with brain mitochondria. Using Western blot, it was shown that addi tion of myelin fraction associated with brain mitochondria to the suspension of liver mitochondria can lead to binding of CNPase and MBP, present in the fraction with liver mitochondria under the conditions of both closed and opened mPTP. However, induction of mPTP opening in liver mitochondria was prevented in the presence of myelin fraction associated with brain mitochondria (Ca2+ release rate was decreased 1.5fold, calcium retention time was doubled, and swelling amplitude was 2.8fold reduced). These results indicate possible protective properties of MBP and CNPase. DOI: 10.1134/S0006297914060091 Key words: mitochondria, myelin, permeability transition pore, myelin basic protein, CNPase
Mitochondria are involved in cell death and play a key role in the survival of brain cells. Programmed death of neurons and other brain cells occurs both under acute brain injury (ischemia, trauma, CNS infections) and in chronic diseases such as Alzheimer’s and Parkinson’s and also during aging. Apoptosis, necrosis, and autophagy take part in the processes of neuronal cell death in neu rodegenerative diseases [1]; a high percentage of damaged mitochondria were found under these conditions [2, 3]. Neurodegenerative diseases are accompanied by disrup tion of the integrity of the myelin sheath – a membrane structure that protects the axon and ensures the spreading of electric impulses over long distances at high speed. Myelin membrane differs from other biological mem branes by its high lipid/protein ratio and the presence of specific proteins, myelin proteolipid protein (PLP), myelin basic protein (MBP), and 2′,3′cyclic nucleotide Abbreviations: CNPase, 2′,3′cyclic nucleotide3′phosphodi esterase; MBP, myelin basic protein; MFAM, myelin fraction associated with mitochondria; mPTP, mitochondrial perme ability transition pore; PLP, proteolipid protein. * To whom correspondence should be addressed. # Deceased.
3′phosphodiesterase (CNPase) which are the major pro teins. Interest in the localization and role of myelin pro teins outside the myelin sheath and, in particular, their interaction with mitochondria has significantly increased in recent years. The presence of myelin proteins in non myelinated tissues such as liver, kidney, pancreas, blood, and cerebral fluid has been described [47]. It has also been shown that CNPase, MBP, and PLP are able to bind to mitochondria of nonmyelinated tissues [5, 6, 8]. The role of association of these proteins with mitochondria is poorly understood both for mitochondria and for cells at present. However, the fact that these three myelin pro teins can play special, yet unknown functions being out side myelin remains indisputable [46, 8, 9]. CNPase has recently been identified in mitochondria isolated from brain, liver, and oligodendrocytes; this enzyme was shown to be involved in regulation of the nonspecific pore (mPTP). In particular, it has been shown that decrease in CNPase expression leads to the reduction of the threshold calcium concentration inducing mPTP opening, and CNPase substrates (2′,3′cAMP and 2′,3′cNADP) can accelerate mPTP opening [9]. Another myelin protein, MBP, was also found in tissues not containing myelin, i.e. in liver, kidney, and lung [5, 6]. The association of MBP
555
556
BABURINA et al.
with mitochondria was demonstrated in pancreas cells by immunoelectron microscopy [7]. MBP was detected there in the outer and inner membranes as well as in mitochondrial cristae. Association myelin PLP with mitochondria has been also demonstrated both in vitro and in vivo using confocal microscopy [8, 10]. These studies showed that PLP in COS7 cells can be embedded in the inner mitochondrial membrane. It was noted that the proteolipid Nterminus contains motifs that could be colocalized with mitochon dria. Deletion of the first 1020 amino acid residues at the proteolipid Nterminus results in the loss of the ability of PLP to interact with mitochondria [8]. In this regard, it is worth to note that PLP includes a sequence of the “c” subunit of the Fo sector of mitochondrial FoF1 ATP syn thase [11, 12], which can be involved in the functioning of the nonspecific pore [1317]. It should be noted that despite years of examination of permeability transition pore, its structural composition and mechanism of its regulation remain not fully understood. However, while investigating the regulation of the mPTP functioning, we discovered a new molecular mechanism of its regulation based on Ca2+dependent phosphorylation/dephospho rylation of mitochondrial membranebound proteins with molecular weights of 4446, 21.5, 17, and 3.5 kDa; the phosphorylation status of these proteins was shown to change depending on the induction of the nonselective pore opening [1316]. Up to now, these phosphoproteins have been identified. The phosphoprotein with molecular weight of 46 kDa was identified in highly purified rat brain mitochondria as 2′,3′cyclic nucleotide 3′phos phodiesterase (CNPase) [9], and the phosphopeptide of 3.5 kDa was detected as the hydrophobic subunit “c” of the membrane FoATP synthase [13, 16]. Two MBP phosphoisoforms (17 and 21.5 kDa) have been recently identified in preparations of brain mitochondria (which were not purified on Percoll gradient) by 2D elec trophoresis with following mass spectrometry [18]. However, the role of these proteins in mitochondria has not been still studied. At the present work, induction of mPTP opening in all fractions obtained after separation of crude mitochon dria on Percoll gradient as well as distribution there of CNPase and MBP were examined. Moreover, using myelin fraction associated with brain as source of myelin proteins, the effect of myelin proteins (CNPase and MBP) on mPTP induction in rat liver mitochondria was examined to find out whether association of CNP and MBP can modulate mPTP opening.
MATERIALS AND METHODS Isolation of rat brain mitochondria. Rat brain mito chondria were isolated by the method of Sims as modified in our laboratory [19]. Male Wistar rats (200250 g) were
used for the experiments. The rats were decapitated, and the brain was rapidly removed, cooled on ice in physio logical saline, and homogenized at 04°C in a tenfold medium volume (320 mM sucrose, 10 mM TrisHCl, pH 7.4, 0.5 mM K+EDTA, 0.5 mM K+EGTA, 0.2% BSA). The homogenate was centrifuged at 2000g for 3 min, the precipitate was removed, and the supernatant was centrifuged again for more complete precipitation of nuclei and damaged cells. The supernatant was cen trifuged at 4°C at 12,500g for 10 min to pellet the mito chondria. Mitochondrial suspension (1.5 ml) were first mixed with 3% Percoll (1 ml), layered to obtain mito chondrial fractions in the Percoll density gradient (1015 24%), and then centrifuged at 31,500g for 10 min. Precipitates of nonsynaptic mitochondria, synaptic mito chondria, and myelin fraction were washed using the extraction medium without EDTA, EGTA, and BSA at 11,500g for 10 min and resuspended in the same medium. Isolation of rat liver mitochondria by a standard method. Isolated liver was homogenized at 04°C in a ten fold medium volume (220 mM mannitol, 70 mM sucrose, 10 mM TrisHCl, pH 7.5, 1 mM EGTA, 0.05% BSA). The homogenate was centrifuged at 800g for 10 min to pellet the nuclei and damaged cells. The supernatant containing mitochondria was centrifuged for 10 min at 9000g to pre cipitate the mitochondria. The mitochondria were washed twice in medium containing EGTA and BSA for 10 min at 9000g and were resuspended in the same medium. Determination of mitochondrial functions. Changes in the mitochondrial membrane potential and calcium ion concentration were measured in a thermostatted sam ple chamber with installed TPP+ and Ca2+selective electrodes as described previously [9]. The incubation medium contained 10 mM TrisHCl, pH 7.4, 120 mM KCl, 0.4 mM KH2PO4, 5 mM potassium succinate, and 2.5 μM rotenone. Potassium succinate was used as the substrate for mitochondrial respiration. The protein con centration in the sample chamber was 1 mg/ml. Opening of the mPTP in mitochondria was induced by the addi tion of calcium to reach its threshold concentration required to initiate mPTP opening. Each Ca2+ addition was 100 nmol per mg protein. All experiments were per formed in an open sample chamber. Swelling of rat brain and liver mitochondria was fol lowed by the kinetics of light scattering by the mitochon drial suspension at wavelength 540 nm using a PE5400V spectrophotometer (Russia) at 25°C in isotonic medium containing 10 mM TrisHCl buffer, pH 7.5, 125 mM KCl, 2 mM KH2PO4, 5 mM potassium succinate, and 2.5 μM rotenone. Protein concentration in the sample chamber was 1 mg/ml for liver mitochondria and 0.4 mg/ml for brain mitochondria. Electrophoresis and Western blot. Electrophoresis under denaturing conditions was carried out in a mini chamber (Hoefer) by the Laemmli method [20]. Ten micrograms of protein was applied to each gel lane. BIOCHEMISTRY (Moscow) Vol. 79 No. 6 2014
ASSOCIATION OF MYELIN PROTEINS WITH MITOCHONDRIA Pharmacia Biotech (USA) sets containing high (14.4 97 kDa) and low molecular mass (3.4616.9 kDa) pro teins were used as markers. MBP and CNP were detected in mitochondria by the Western blot method. The protein was transferred from the gel to a nitrocellulose membrane (Sigma, USA) on a semidry transfer apparatus (BioRad, USA) using buffer containing 0.048 M TrisHCl, pH 9.2, 0.039 M glycine, 20% methanol (v/v), and 0.0357% SDS. Monoclonal mouse antiCNP antibodies (Mab 461 obtained in the laboratory of Prof. G. Reiser, dilution 1 : 10,000) and polyclonal goat antiMBP antibodies (Santa Cruz Biotechnology, USA; dilution 1 : 2000) were used for protein identification. Secondary antibodies conju gated to horseradish peroxidase were also used. Peroxidase activity was determined with the ECL chemi luminescent reagent (Pierce Chemical Co, USA) accord ing to the manufacturer’s instruction. Electron microscopy of isolated mitochondria. Brain mitochondria samples were prepared for electron microscopy by fixation for 2 h in 2.5% glutaraldehyde (AppliChem GmbH, Germany) dissolved in the isolation medium (pH 7.4), washing once after resuspension in the isolation medium, and fixation in 1% solution of osmium tetroxide (Serva Electrophoresis GmbH, Germany) in isolation medium with sucrose added to reach the iso tonicity with the previous washing solution. Precipitated samples were dehydrated without resuspension in water solutions with increasing ethanol concentrations (30, 50, 75, and 96%) and in absolute ethanol (incubation in each mixture for 1 h), in absolute acetone (three times for 1 h), and embedded in Epon 812 epoxy resin (Fluka Chemie AG, Switzerland). Ultrathin sections were prepared on an EM UC6 ultramicrotome (Leica, Germany); the samples were contrasted in saturated (4%) aqueous solution of uranyl acetate and in an alkaline solution of lead citrate (pH 12) and were examined with a Tesla BS500 electron microscope (Czechia) at accelerating voltage 90 kV. Statistical analysis. The data were analyzed and visu alized using the Sigma Plot 9.0 software. The average value of the parameters from 56 experiments was taken for statistical analysis. Significance was taken at p < 0.05 vs the control using the Student ttest.
RESULTS Since the suspension of brain mitochondria is het erogeneous and includes two mitochondrial pools (synap tic and nonsynaptic) as well as associated myelin fraction, the preparation of brain mitochondria obtained by differ ential centrifugation was subjected to fractionation on a Percoll density gradient. We examined the presence of myelin proteins in the fractions, i.e. CNPase and MBP (Fig. 1). Figure 1 shows the distribution of CNPase and MBP by Western blot analysis using antibodies highly spe cific to CNPase and polyclonal antibodies to MBP react BIOCHEMISTRY (Moscow) Vol. 79 No. 6 2014
557
ing with all four MBP isoforms. Figure 1a shows that the highest CNPase content is observed in the myelin frac tion, and its level decreases in mitochondrial fractions (synaptic and nonsynaptic). If we take the CNPase con tent in the myelin fraction as 100%, then its content in synaptic and nonsynaptic mitochondria is 21 and 17%, respectively (see table). The presence of MBP was also checked in all fractions (Fig. 1a). It is evident that two MBP isoforms (21.5 and 17 kDa) are associated with mitochondria. They can be removed to a large extent after the separation of mitochondria in the Percoll density gra dient, nevertheless, a small portion of the MBP remains strongly bound with mitochondria (8% with synaptic and 1.5% with nonsynaptic mitochondria; see table). The purity of the mitochondrial fractions were carefully stud ied earlier and confirmed by specific protein markers (synaptotagmin, COX IV) and by electron microscopy [21]. The presence of CNPase and MBP in all the frac tions of brain mitochondria indicates their high affinity to mitochondria. We studied the functioning of Ca2+ induced and CsAsensitive mitochondrial pore (mPTP) by measuring such mPTP characteristics as calcium capacity, rate of Ca2+ efflux, and Ca2+stimulated swelling of mitochondria. Figure 1b shows the average values of calcium capacity and swelling halftime (T1/2). Synaptic mitochondria were shown to have the highest T1/2 value (310 s) and calcium capacity (350 μM) in our experi ments. The swelling halftime of nonsynaptic mitochon dria was about the same (300 s), but the calcium capacity was lower by 15% (300 μM) (see table). These data are consistent with our previously described results [22]. It is interesting to note that membrane potential was regis tered in the myelin fraction associated with mitochondria (MFAM) using the TPP+electrode (not shown), and this fraction was still able to accumulate calcium despite the low calcium capacity in MFAM, which was 6.6fold lower (30 μM) than in the nonsynaptic mitochondria fraction. However, it did not undergo calciuminduced swelling. On the other hand, the experiments using a chamber with installed electrodes showed that sequential calcium addi tion to this fraction resulted in Ca2+dependent and CsA sensitive calcium release from mitochondria, which was observed in mitochondria under mPTP induction. This might indicate the presence of a small pool of intact func tioning mitochondria interacting with myelin membrane in MFAM. This possibility was confirmed by the results of electron microscopy (Fig. 2), which showed the presence of a certain number of intact mitochondria, myelin sheaths, and their fragments in MFAM, where brain mitochondria are likely to form some contact sites with them (Fig. 2). The pictures show that contacts can be formed between mitochondria and myelin membrane (Fig. 2a) as well as between mitochondria and the frag ments of myelin membrane (Fig. 2, b and c). Previously it was shown that a change in the CNPase expression level in oligodendrocytes affects the activation
558
BABURINA et al. Comparative characteristics of mitochondrial fractions Fraction
Swelling halftime, s
Calcium capacity, μmol
CNPase content, %
MBP content, %
–
18
100
100
Synaptic mitochondria
310
350
21
8
Nonsynaptic mitochondria
300
300
17.3
1.5
MFAM
of mPTP opening [9]. In the present work, the effect of myelin proteins (CNPase and MBP) on the induction of mPTP opening was studied. As there is a small amount of myelin proteins present in brain mitochondria, mito chondria isolated from liver, which has no myelin, were
used to study the effect of these proteins on the function ing of the nonselective mitochondrial pore. The ability of myelin proteins found in MFAM to associate with liver mitochondria both under control conditions (before cal cium addition) and under the conditions of nonselective
a kDa kDa
МВР
21.5 CNP
46
17
1
2
3 1
2
3
b Swelling halftime
350
500
Calcium capacity of mitochondria 300 400
300
200
T1/2, s
[Ca2+], μmol
250 200 150 100 100 50 0
0 2
3
1
Fig. 1. a) Distribution of myelin proteins (CNPase and MBP) in rat brain mitochondria: 1) myelin fraction (MFAM); 2) synaptic mitochon dria; 3) nonsynaptic mitochondria; b) functional characteristics of mitochondrial fractions obtained after separation on a Percoll density gra dient.
BIOCHEMISTRY (Moscow) Vol. 79 No. 6 2014
ASSOCIATION OF MYELIN PROTEINS WITH MITOCHONDRIA
0.1 μm
0.1 μm
a
0.1 μm
559
b
c
Fig. 2. Ultrastructure of contact sites of isolated brain mitochondria with fragments of myelin membranes. Arrows indicate. a) Contact of mitochondria with myelin membrane; b) contact of mitochondria with myelin fragments; c) contact inside and outside of a myelin fragment. Scale 0.1 μm.
pore opening (in the presence of the threshold calcium concentration) was determined. Liver mitochondria were incubated with MFAM (at protein concentration of 0.1 and 0.5 mg/ml) for 10 min. After incubation, aliquots of mitochondria were taken from the cell in two states: a) without calcium addition (control), and b) in the pres ence of threshold calcium concentrations that initiated pore opening. The samples were centrifuged, the mito BIOCHEMISTRY (Moscow) Vol. 79 No. 6 2014
chondrial precipitate was solubilized in Laemmli solu tion, and then the mitochondrial proteins were separated by SDSPAGE. The proteins were then transferred onto a nitrocellulose membrane that was treated with antiCNP and antiMBP antibodies. Figure 3a shows that MBP is absent in liver mitochondria both in control and under pore opening (lanes 1 and 2). After incubation of liver mitochondria with MFAM, association of the 17kDa
560
BABURINA et al.
MBP isoform with mitochondria was observed (lanes 3 and 4); this association was enhanced in the presence of calcium. The diagram below shows that calcium increased twofold the association of 17 kDa MBP with mitochondria in the presence of 0.1 mg/ml MFAM. Increase of the concentration of the added myelin frac tion (lanes 5 and 6) leads to higher amount of associated MBP isoform (17 kDa). In the absence of calcium, almost a fivefold increase in MBP binding was observed, which was not the case in the presence of calcium (Fig. 3a, diagram). In addition, the 21.5kDa MBP isoform appeared under incubation of 0.5 mg/ml MFAM with mitochondria. However, the level of associated myelin basic protein (17 and 21.5 kDa) does not depend on cal cium under these conditions. It should be noted that the phosphorylation status of both MBP isoforms increased under pore opening conditions [18]. Figure 3b shows the binding of CNPase from MFAM with liver mitochondria under the same conditions. Also seen, that isolated liver mitochondria originally contain a certain amount of CNPase. Addition of MFAM fraction to the mitochondr
ial suspension results in concentrationdependent bind ing of CNPase from the MFAM fraction with liver mito chondria, so that the increase in the concentration of added MFAM is accompanied by increased CNPase association with mitochondria (Fig. 3b, lanes 36). The diagram in the lower part of Fig. 3b shows that incubation of 0.1 mg/ml MFAM with mitochondria leads to twofold increase in CNPase binding, and incubation with 0.5 mg/ml MFAM results in fourfold increase in protein association. Phosphodiesterase binding with mitochon dria showed no dependence on calcium. Thus, MBP and CNPase association with mitochondria can proceed under conditions of both open and closed pore. Next we tested the effect of MBP and CNPase binding with mito chondria on the functioning of the Ca2+induced and CsAsensitive pore. To elucidate this issue, we studied the effect of added brain MFAM on Ca2+ release from liver mitochondria and Ca2+induced highamplitude swelling of liver mitochondria. Figure 4 shows that the addition of MFAM to liver mitochondria alters the rate of calcium release and the duration of the lag phase (calcium reten
a
b
kDa
kDa
21.5 46 17
МВР 1
2
Mch
3
4
5
1
4
5
6
CNP
incubation
incubation
400 Protein, % of control
Protein, % of control
3
Mch
6
500 400 300 200 100 0
2
300 200 100 0
1
2
3
4
5
6
1
2
3
4
5
6
Fig. 3. Association of myelin proteins with liver mitochondria under pore opening: a) MBP association; b) CNPase association. In the lower part of the figure the control corresponds to the values of association of the studied proteins with mitochondria when MFAM was added before calcium additions. The concentration of antiCNP antibodies was 1 : 10,000; antiMBP antibodies 1 : 2000. Mch: 1) liver mitochondria (Mch) before calcium addition; 2) Mch after mPTP induction. Incubation: 3) Mch + MFAM (0.1 mg/ml) before calcium addition (control); 4) Mch + MFAM (0.1 mg/ml) after mPTP induction; 5) Mch + MFAM (0.5 mg/ml) without calcium addition; 6) Mch + MFAM (0.5 mg/ml) after mPTP induction.
BIOCHEMISTRY (Moscow) Vol. 79 No. 6 2014
ASSOCIATION OF MYELIN PROTEINS WITH MITOCHONDRIA
a
b Lag phase 180
160
160
*
140
120
100
80
* 60
40
20
Relative values of lag phase duration, % of control
Relative values of calcium release rate, % of control
Calcium release rate 180
0
561
140
120
100
80 60
40
20
1
2
0
1
2
Fig. 4. Effect of myelin fraction of brain mitochondria on rate of Ca2+ release (a) and lag phase duration (b) under mPTP opening. Concentration of myelin fraction protein 0.1 (1) and 0.5 (2) mg/ml. Incubation time 10 min.
tion time in liver mitochondria before its release indicat ing the initiation of the pore opening). Figure 4a shows that the addition of 0.1 mg/ml protein has no effect on the rate of calcium release and the duration of the lag phase. However, the myelin fraction of brain mitochon dria at protein concentration 0.5 mg/ml reliably reduced the rate of calcium release from liver mitochondria by 32% against the control; at the same time, the period of the lag phase increased by 57% against the control (Fig. 4b). The studies of the effect of MFAM proteins on the Ca2+induced highamplitude swelling of liver mitochon dria (Fig. 5a, lower part) showed that the maximum amplitude of the swelling was observed when mitochon dria were loaded with calcium to its threshold concentra tion initiating the pore opening (control, curve 1). Incubation of liver mitochondria with MFAM led to decrease in the swelling amplitude (curves 2 and 3); the swelling amplitude decreased 2fold in the presence of 0.1 mg/ml MFAM and almost 3fold in the presence of 0.5 mg/ml of MFAM protein (Fig. 5a, upper part). Thus, by reducing the amplitude of liver mitochondria swelling, MFAM increases the resistance of liver mitochondria to BIOCHEMISTRY (Moscow) Vol. 79 No. 6 2014
the initiation of nonselective pore opening. This observa tion is consistent with the results presented in Fig. 1 and in the table, which demonstrate that mitochondria from the myelin fraction containing CNPase and MBP do not undergo swelling. These results suggest the presence of protector proteins in MFAM that prevent mPTP open ing. Because we earlier identified CNPase and MBP in this fraction, we suggested that they could fulfill this pro tective role, although still unidentified proteins can also be present in this fraction. The protective effect of CNPase and MBP was verified in experiments when CNPase and MBPcontaining MFAM was added to the suspension of mitochondria together with antibodies pro duced to these proteins (antiCNP antibodies and anti MBP antibodies). We assumed that the added antibodies would bind the proteins in the solution or on the outer mitochondrial membrane and would thus not prevent mitochondrial swelling. Simultaneous presence of the proteins and antibodies to them confirmed that CNPase and MBP increase the stability of the mitochondria. Figure 5b shows that Ca2+induced swelling of liver mito chondria (lower part of the figure, curve 1) is suppressed
562
BABURINA et al.
in the presence of added MFAM (curve 2). Incubation of calciumloaded liver mitochondria in the presence of MFAM and antiCNP antibodies (curve 3) or antiMBP antibodies (curve 4) reduced the ability of the myelin fraction to prevent mitochondrial swelling. In particular, the reduction was 34% in the case of antiCNP antibod ies (upper part of Fig. 5b, column 3) and 38% for anti MBP antibodies (upper part of Fig. 5b, column 4). This effect was CsAsensitive, i.e. no mitochondrial swelling
could be observed under similar conditions in the pres ence of CsA, the studied antibodies, and MFAM. The specificity of the protective effect of myelin proteins was confirmed by replacing antibodies to CNPase and MBP by the protein of bovine serum albumin (BSA, fraction V) at the same protein concentration. Figure 5a shows that BSA had no effect on the swelling of liver mitochondria (lower part of Fig. 5a, curves 4 and 5, and upper part of the figure, columns 4 and 5).
a
b 1.0
Swelling amplitude, absolute units
Swelling amplitude, absolute units
1.0
0.8
0.6
* 0.4
*
0.2
0.6
0.4
0.2
0 1
2
3
4
*
*
0.8
0
5
2.2
2.2
2.0
2.0
1.8
1.8
1
2
3
4
2 1.6
1.6
3
1.4
2
1.4
4
3
5 1
1.2
4
1.2 1
1.0
1.0 0
100
200
300
400
0
100
200
300
400
Time, s Fig. 5. Effect of the myelin fraction of brain mitochondria on swelling of rat liver mitochondria. Upper panel, quantitative characteristics of swelling amplitude of liver mitochondria incubated with myelin fraction of brain mitochondria; lower panel, swelling curves. a) Highamplitude calciuminduced swelling of liver mitochondria incubated with the myelin fraction of brain mitochondria. Concentration of the myelin fraction protein 0.1 and 0.5 mg/ml: 1) Mch + Ca2+; 2) Mch + MFAM (0.1 mg/ml) + Ca2+; 3) Mch + MFAM (0.5 mg/ml) + Ca2+; 4) Mch + BSA (0.1 mg/ml) + Ca2+; 5) Mch + BSA (0.5 mg/ml) + Ca2+. b) Swelling of liver mitochondria incubated with myelin fraction and antiCNP and antiMBP antibodies (Ab). Concentration of the myelin fraction protein 0.1 and 0.5 mg/ml: 1) Mch + Ca2+; 2) Mch + MFAM + Ca2+; 3) Mch + MFAM + antiCNP Ab + Ca2+; 4) Mch + MFAM + antiMBP Ab + Ca2+. Concentration of antiMBP Ab 1 : 2000, antiCNP Ab 1 : 10,000.
BIOCHEMISTRY (Moscow) Vol. 79 No. 6 2014
ASSOCIATION OF MYELIN PROTEINS WITH MITOCHONDRIA DISCUSSION Damage of the integrity of the myelin membrane is one of the factors stimulating neurodegenerative process es leading to disruption of the conductivity of nerve impulses [23]. It was noted that degradation of myelin proteins observed in acute or chronic neuropathologies could result from the activation of the nonselective mito chondrial pore. This is confirmed by the detection of myelin proteins outside myelin sheets and, in particular, in mitochondria isolated from different tissues and cells [19]. One of the major myelin proteins, CNPase, exists in two isoforms with molecular weights of 46 and 48 kDa [24]. The primary structure of myelin CNPase can be divided into two areas different in their properties. The N terminus has binding sites for nucleotide phosphates [24 26] and for ATP/GTPbinding proteins, while a catalytic site [27], tubulin and RNAbinding domain [28], and CAAX sequence (the motif for protein acylation [26]) are located at the CNPase Cterminus. However, it is not yet clear which properties of CNPase remain outside myelin. It was recently shown that CNPase identified in mito chondria can be involved in the regulation of mPTP func tioning, in particular, its substrates (2′,3′cAMP and 2′,3′cNADP) accelerate mPTP opening in the presence of threshold calcium concentrations [9]. Decrease of CNPase expression in OLN93 cells (knockdown CNP OLN93) was shown to decrease threshold calcium con centration required for the induction of pore opening. These results suggest that the increase in CNPase level can prevent mPTP opening. Since the suspension of iso lated brain mitochondria is heterogeneous, they can be purified in a Percoll density gradient, giving nonsynaptic and synaptic mitochondria as well as a fraction of mito chondriaassociated myelin (MFAM). CNPase and MBP isoforms were identified in the fractions. The highest MBP content was found in the myelin fraction with its two major bands corresponding to 17 and 21.5 kDa iso forms. In synaptic and nonsynaptic brain mitochondria, MBP was detected in small amounts, reflecting a small pool of the protein strongly associated with mitochondria (Fig. 1). There are some literature data suggesting that intravenously administered 125Ilabeled MBP can be found in liver, kidney, lung, and pancreatic islet cells [5, 6]. The binding of MBP with mitochondria was shown by immunoelectron microscopy [6]. On the other hand, an increasing number of publications report the presence of subunits of FoF1ATPase and subunits of respiratory chain complexes in myelin [2931], probably due to the presence of mitochondria or their fragments during myelin isolation [32]. Taking these data into considera tion, we examined functioning of the nonselective Ca2+ induced pore in all three fractions. We found that both types of mitochondria maintained high membrane poten tial and calcium capacity, which decreased when mito chondria were loaded with calcium. This decrease was BIOCHEMISTRY (Moscow) Vol. 79 No. 6 2014
563
cyclosporinsensitive. Furthermore, the mitochondria were capable of Ca2+induced swelling (Fig. 1). Maintained membrane potential was revealed and in the myelin fraction associated with mitochondria (MFAM) (not shown); increasing calcium concentration caused Ca2+induced calcium release. All these processes were CsAsensitive supporting involvement of mPTP phenom enon. However, despite the fact that calcium accumula tion was observed in the brain myelin fraction, it did not undergo swelling that was previously described in the studies on isolated myelin [33, 34]. The presence of a reg istered membrane potential in MFAM and the ability to accumulate calcium indicated the possible presence of a pool of intact mitochondria in MFAM. Electron microscopy data (Fig. 2) showed that a small pool of mitochondria and fragments of myelin membranes are indeed present in the myelin fraction of brain mitochon dria; brain mitochondria probably contact them similar to mitochondrial contacts with endoplasmic reticulum (forming contact sites). It should be noted that myelin membrane contacts with mitochondria can be also seen in electron micrographs in publications dedicated to the research in completely different areas [35]. Despite the fact that myelin protein association has been described in some papers, their effect on mitochondrial functions and in particular on the mPTP functioning has not been pre viously studied. Due to the fact that myelin proteins asso ciate firmly with brain mitochondria, their effect on the mPTP induction was studied with liver mitochondria, which are free of myelin. The experiments revealed that MBP and CNPase from MFAM can bind with the liver mitochondrial membrane. It should be noted that MBP interacts with lipids and can penetrate the membrane bilayer [36]. Association of these proteins with mitochon dria increases when the concentration of added myelin fraction is increased from 0.1 to 0.5 mg/ml protein. It is interesting to note that only two (out of four) minor MBP isoforms with molecular masses of 21.5 and 17 kDa have ability to bind with the mitochondrial membrane. It was previously shown that the 21.5 kDa isoform has a signal amino acid sequence corresponding to nuclear localiza tion, which can transport the protein into nuclei [37]. MBP binding with liver mitochondria has been also found under conditions of initiation of mPTP opening; the degree of association was not related with calcium con centration. However, we recently showed that the phos phorylation status of the 17 and 21.5 kDa isoforms increases in the presence of threshold concentrations of calcium [18]. Similar results were obtained for CNPase, which also binds with liver mitochondria in the presence of added MFAM fraction, despite the fact that liver mito chondria possess an independent pool of this protein [7]. Incubation of MFAM with liver mitochondria leads to additional CNPase binding. Similarly to MBP, associa tion of CNPase with liver mitochondria depends on the concentration of the added protein, but not on added cal
564
BABURINA et al.
cium. CNPase binding takes place both when the pore is closed (control) and when it is opened. Similarly to MBP, pore opening causes increase in the CNPase phosphory lation status (Baburina et al. (2014) J. Biomembr. Bioenerg., in press). The effect of MFAM on the Ca2+ induced Ca2+ release from mitochondria and Ca2+ induced high amplitude swelling were studied to deter mine the significance of MFAM binding with mitochon dria. We found that incubation of liver mitochondria with MFAM proteins reduced the rate of calcium release from liver mitochondria and increased the lag phase period (by 57%). These results indicate the protective properties of the proteins, which increase the resistance of liver mito chondria to the initiation of the pore opening in the pres ence of the myelin fraction of brain mitochondria. In this regard, it is interesting to note that one of the functions of MBP in myelin is to stabilize the myelin membrane. Since MBP and CNPase were identified in MFAM by Western blot, the protective effect of these proteins was confirmed in experiments on mitochondrial swelling in the simultaneouspresence of MFAM and antibodies against MBP and CNPase. It was found that incubation of liver mitochondria with 0.1 or 0.5 mg/ml of MFAM protein prevented mitochondrial swelling. No protective effect of MFAM could be observed when MFAM was added after incubation of mitochondria with antiMBP or antiCNP antibodies. Apparently, the binding of the antibodies with the corresponding protein on the outer membrane (and in solution) prevented them from associ ating with mitochondria, and therefore they could not prevent mitochondrial swelling. The results demonstrate that myelin proteins MBP and CNPase can execute a new unknown function being outside myelin; in particular, they can be involved in regulation of apoptosis, prevent ing cells from the initial stages of programmed cell death probably through stabilization the integrity of mitochon drial membranes. This work was supported by the Russian Foundation for Basic Research (grants 140400625, 120400671 and 130400935).
REFERENCES 1. 2.
3.
4. 5. 6.
Gorman, A. M. (2008) J. Cell Mol. Med., 12, 22632280. Calabrese, V., Scapagnini, G., Giuffrida Stella, A. M., Bates, T. E., and Clark, J. B. (2001) Neurochem. Res., 26, 739764. Aliev, G., Palacios, H. H., Walrafen, B., Lipsitt, A. E., Obrenovich, M. E., and Morales, L. (2009) Int. J. Biochem. Cell Biol., 41, 19892004. Kolehmainen, E., Knip, M., and Leppaluoto, J. (1990) Acta Physiol. Scand., 139, 493501. Kolehmainen, E. (1996) J. Neurol. Sci., 141, 1926. Kolehmainen, E., and Sormunen, R. (1998) Pancreas, 16, 176188.
7. Baburina, Yu. L., Krestinina, O. V., and Azarashvili, T. S. (2013) Neurochem. J., 7, 115. 8. Huttemann, M., Zhang, Z., Mullins, C., Bessert, D., Lee, I., Nave, K. A., Appikatla, S., and Skoff, R. P. (2009) Neuro, 1. 9. Azarashvili, T., Krestinina, O., Galvita, A., Grachev, D., Baburina, Y., Stricker, R., Evtodienko, Y., and Reiser, G. (2009) Am. J. Physiol. Cell Physiol., 296, 1428 1439. 10. Ghandour, M. S., Feutz, A. C., Jalabi, W., Taleb, O., Bessert, D., Cypher, M., Carlock, L., and Skoff, R. P. (2002) Glia, 40, 300311. 11. Knapp, P. E. (1996) Dev. Neurosci., 18, 297308. 12. Laursen, R. A., Samiullah, M., and Lees, M. B. (1983) FEBS Lett., 161, 7174. 13. Azarashvili, T. S., Tyynela, J., Odinokova, I. V., Grigorjev, P. A., Baumann, M., Evtodienko, Y. V., and Saris, N. E. (2002) J. Bioenerg. Biomembr., 34, 279284. 14. Azarashvily, T. S., Tyynela, J., Baumann, M., Evtodienko, Y. V., and Saris, N. E. (2000) Biochem. Biophys. Res. Commun., 270, 741744. 15. Krestinina, O. V., Grachev, D. E., Odinokova, I. V., Reiser, G., Evtodienko, Y. V., and Azarashvili, T. S. (2009) Biochemistry (Moscow), 74, 421429. 16. Azarashvili, T., Odinokova, I., Bakunts, A., Ternovsky, V., Krestinina, O., Tyynela, J., and Saris, N. E. (2014) Cell Calcium, 55, 6977. 17. Bonora, A., Bononi, M. E., De, C., Giorgi, M., Lebiedzinska, S., Marchi, S., Patergnani, A., Rimessi, J. M., Suski, A., Wojtala, M. R., Wieckowski, G., Kroemer, L., and Galluzzi, P. (2013) Cell Cycle, 12, 674683. 18. Krestinina, O., Makarov, P., Baburina, Yu., Gordeeva, A., and Azarashvili, T. (2013) Neurochem. J., 7, 284290. 19. Sims, N. R. (1990) J. Neurochem., 55, 698707. 20. Laemmli, U. K. (1970) Nature, 227, 680685. 21. Krestinina, O. V., Kruglov, A. G., Grachev, D. E., Baburina, Yu. L., Evtodienko, Yu. V., Moshkov, D. A., Santalova, I. M., and Azarashvili, T. S. (2010) Biochemistry (Moscow), Ser. A: Membrane and Cell Biology, 4, 180 186. 22. Azarashvili, T., Baburina, Y., Grachev, D., Krestinina, O., Evtodienko, Y., Stricker, R., and Reiser, G. (2011) Am. J. Physiol. Cell Physiol., 300, 707720. 23. Kim, T., and Pfeiffer, S. E. (1999) J. Neurocytol., 28, 281 293. 24. Lee, J., O’Neill, R. C., Park, M. W., Gravel, M., and Braun, P. E. (2006) Mol. Cell Neurosci., 31, 446462. 25. Sprinkle, T. J. (1989) Crit. Rev. Neurobiol., 4, 235301. 26. Agrawal, H. C., Sprinkle, T. J., and Agrawal, D. (1990) Biochem. Biophys. Res. Commun., 170, 817823. 27. Gravel, M., DeAngelis, D., and Braun, P. E. (1994) J. Neurosci. Res., 38, 243247. 28. Gravel, M., Robert, F., Kottis, V., Gallouzi, I. E., Pelletier, J., and Braun, P. E. (2009) J. Neurosci. Res., 87, 1069 1079. 29. Panfoli, I., Ravera, S., Bruschi, M., Candiano, G., and Morelli, A. (2011) Expert. Rev. Proteom., 8, 231239. 30. Morelli, A., Ravera, S., and Panfoli, I. (2011) Cell Biochem. Biophys., 61, 179187. 31. Ravera, S., Panfoli, I., Calzia, D., Aluigi, M. G., Bianchini, P., Diaspro, A., Mancardi, G., and Morelli, A. (2009) J. Biochem. Cell Biol., 41, 15811591. BIOCHEMISTRY (Moscow) Vol. 79 No. 6 2014
ASSOCIATION OF MYELIN PROTEINS WITH MITOCHONDRIA 32. Harris, J. J., and Attwell, D. (2013) J. Cereb. Blood Flow Metab., 33, 3336. 33. Karthigasan, J., and Kirschner, D. A. (1988) J. Neurochem., 51, 228236. 34. Rand, R. P., Fuller, N. L., and Lis, L. J. (1979) Nature, 279, 258260. 35. Zambonin, J. L., Zhao, C., Ohno, N., Campbell, G. R., Engeham, S., Ziabreva, I., Schwarz, N., Lee, S. E.,
BIOCHEMISTRY (Moscow) Vol. 79 No. 6 2014
565
Frischer, J. M., Turnbull, D. M., Trapp, B. D., Lassmann, H., Franklin, R. J., and Mahad, D. J. (2011) Brain, 134, 19011913. 36. Beest, M., and Hoekstra, D. (1993) Eur. J. Biochem., 211, 689696. 37. Smith, G. S., Seymour, L. V., Boggs, J. M., and Harauz, G. (2012) Biochem. Biophys. Res. Commun., 422, 670 675.
