FEBS 23804 FEBS Letters 475 (2000) 213^217 Protein phosphorylation/dephosphorylation in the inner membrane of potato tuber mitochondria Andrë Struglicsa;1 , Kenneth M. Fredlundb;2 , Yuri M. Konstantinovb;3 , John F. Allena;4 , Ian M. MÖllerb; * a Plant Cell Biology, Lund University, P.O. Box 7007, S-220 07 Lund, Sweden b Plant Physiology, Lund University, P.O. Box 117, S-221 00 Lund, Sweden Received 6 May 2000 Edited by Vladimir Skulachev Abstract Inside-out inner mitochondrial membranes free of matrix proteins were isolated from purified potato tuber (Solanum tuberosum L.) mitochondria and incubated with [QQ-32 P]ATP. Proteins were separated by SDS^PAGE and visualized by autoradiography. Phosphorylation of inner membrane proteins, including ATPase subunits, was strongly inhibited by the phosphoprotein phosphatase inhibitor NaF. We propose that an inner membrane phosphoprotein phosphatase is required for activation of the inner membrane protein kinase. When prelabelled inner membranes were incubated in the absence of [QQ-32 P]ATP, there was no phosphoprotein dephosphorylation unless a soluble matrix fraction was added. This dephosphorylation was inhibited by NaF, but not by okadaic acid. We conclude that the mitochondrial matrix contains a phosphoprotein phosphatase that is responsible for dephosphorylation of inner membrane phosphoproteins. ß 2000 Federation of European Biochemical Societies. Published by Elsevier Science B.V. All rights reserved. Key words: Protein phosphorylation; (Plant) mitochondrion; Inner membrane; Matrix; Protein kinase; Protein phosphatase; Protein kinase phosphatase 1. Introduction Protein phosphorylation/dephosphorylation by protein kinases and protein phosphatases is a ubiquitous mechanism in eukaryotes and prokaryotes for modulating the activity of intracellular proteins in response to extracellular signals [1]. In mitochondria, two well-characterized enzymes, the Ksubunit of pyruvate dehydrogenase and the K-subunit of branched-chain 2-oxoacid dehydrogenase complex, both located in the matrix are under the control of phosphorylation/dephosphorylation [2]. Four inner membrane phospho*Corresponding author. Fax: (46)-46-222 4113. E-mail: ian_max.moller@fysbot.lu.se 1 Present address: Active Biotech Research AB, P.O. Box 724, S-220 07 Lund, Sweden. 2 Present address: Novartis Seeds AB, P.O. Box 302, S-261 23 Landskrona, Sweden. 3 Present address: Laboratory of Plant Genetic Engineering, Siberian Institute of Plant Physiology and Biochemistry, P.O. Box 1243, Irkutsk 664033, Russia. 4 Present address: Department of Plant Biochemistry, Lund University, P.O. Box 117, S-221 00 Lund, Sweden. Abbreviations: IO-SMP, inside-out submitochondrial particles (inner membranes); TCA, trichloroacetic acid proteins have been identi¢ed by N-terminal sequencing: the 18 kDa AQDQ subunit of complex I [3]; the 17 kDa subunit IV of cytochrome c oxidase [4]; and the 22 kDa NP-subunit and the 28 kDa L-subunit of the F0 F1 -ATPase [5]. The phosphorylation of a number of unidenti¢ed proteins has also been demonstrated in bovine heart mitochondria [6,7] and in potato tuber mitochondria [5,8^10]. When inside-out submitochondrial particles (IO-SMP) from plant mitochondria are incubated with [Q-32 P]ATP, more than 20 proteins are phosphorylated mostly on serine/threonine residues [11]. The basic properties of the phosphorylation, speci¢city, time course, pH dependence, Km (ATP), have been characterized and two putative kinases/kinase subunits of 16.5 and 30 kDa identi¢ed [11]. In the present study, we show that two distinct phosphoprotein phosphatases are involved in the protein phosphorylation/dephosphorylation in IO-SMP, one bound to the mitochondrial inner membrane and another, soluble, in the matrix. 2. Materials and methods Highly puri¢ed intact mitochondria free from plastids (amyloplasts) and peroxisomes were puri¢ed from potato tubers (Solanum tuberosum L. cv. Bintje or Ukama) [12]. The mitochondria were subfractionated and inside-out inner mitochondrial membrane (IO-SMP) vesicles lacking any substrates or matrix proteins were prepared as in [10]. The matrix fraction was concentrated in an Amicon 8010 Stirred Ultra¢ltration Cell with a 10 kDa cut o¡ ¢lter (Amicon YM 10). Protein phosphorylation assays were carried out in a volume of 50 Wl containing 250 Wg IO-SMP proteins [11]. Final concentrations in the reaction mixture were as follows: 0.3 M sucrose, 50 mM HEPES^ KOH (pH 7.5), 5 mM MgCl2 , 0.1 mM CaCl2 and 0.2 mM (0.4^1.0 Ci mmol31 ) [Q-32 P]ATP. After 2^15 min incubation at room temperature (22^23³C), the reaction was stopped in either of two ways: (A) by addition of sample bu¡er [13] and thereafter boiling at 100³C for 2 min, or (B) by addition of trichloroacetic acid (TCA) (6.5% w/v ¢nal concentration), and after 30 min at room temperature the samples were pelleted by centrifugation at 13 000Ug for 6 min, resuspended in 100 Wl 50 mM HEPES^KOH (pH 7.5) and pelleted again. The washed pellets were then resuspended in the sample bu¡er. For the dephosphorylation assay, inside-out inner mitochondrial membrane proteins were ¢rst phosphorylated using [Q-32 P]ATP as described above. The membranes were then pelleted (1 min at 100 000Ug) with an Airfuge (Beckman, rotor A-100) and the supernatant containing excess ATP was discarded. The membranes were resuspended in 50 mM HEPES^KOH (pH 7.5), 5 mM MOPS (pH 7.2), 0.3 M sucrose, 125 mM mannitol, 7 mM MgCl, 0.1 mM CaCl2 and 0.4 mM EDTA in the presence or absence of 400 Wg matrix proteins in a total volume of 50 Wl. After incubation for 1^60 min, the inner membranes were pelleted as above and the supernatant, which contained matrix proteins, was removed. TCA was added to the inner membranes to a ¢nal concentration of 6.5% (w/v) and after 30 min at room temperature proteins were pelleted at 13 000Ug for 0014-5793 / 00 / $20.00 ß 2000 Federation of European Biochemical Societies. Published by Elsevier Science B.V. All rights reserved. PII: S 0 0 1 4 - 5 7 9 3 ( 0 0 ) 0 1 6 8 0 - X 214 6 min. The pellet was resuspended in 100 Wl 50 mM HEPES^KOH (pH 7.5) bu¡er and treated as above before gel electrophoresis. SDS^PAGE was performed according to [13] on a 10^15% T (C = 2.7%) gradient gel with a Protean II xi Slab Cell (Bio-Rad) ap- A. Struglics et al./FEBS Letters 475 (2000) 213^217 paratus. Gels were stained with Coomassie brilliant blue R250, destained and dried. The phosphoproteins were visualized in two ways: (1) the gels were placed on a phosphorimager plate (Molecular Dynamics) for 2^5 days. The plate was then screened in a phosphorim- Fig. 1. The e¡ect of NaF on the phosphorylation of mitochondrial inner membrane proteins. (A) Coomassie-stained gel identical to that used in B. (B) Autoradiograph of IO-SMP proteins which had been phosphorylated for 2 min in the absence or presence of NaF at di¡erent concentrations following separation using SDS^PAGE. Lane 1, no NaF; lane 2, 1 mM NaF; lane 3, 2 mM NaF; lane 4, 5 mM NaF; lane 5, 10 mM NaF; lane 6, 25 mM NaF; lane 7, 50 mM NaF. Lane 8 shows molecular mass standards (in kDa) visualized using a 35 S marker pen on the Coomassie-stained gel before autoradiography. Phosphoproteins discussed in the text are indicated by arrows and size in kDa. (C) Relative inhibition of 32 P incorporation at di¡erent NaF concentrations given as the average value for all detected phosphoproteins in IO-SMP. Results from B. A. Struglics et al./FEBS Letters 475 (2000) 213^217 ager (Molecular Dynamics SI) and the data analyzed with a Microsoft Windows NT version 3.1 program. (2) The gels were placed on a screen with autoradiographic ¢lm (Hyper ¢lm MP, Amersham) for 5^7 days at 380³C. Protein concentration was determined by use of Coomassie brilliant blue G250 according to the instructions from BioRad using IgG as the standard. All experiments were carried out at least twice, and only consistently reproducible results are presented. 3. Results and discussion Incubating IO-SMP vesicles with [Q-32 P]ATP in the presence of divalent cations results in phosphorylation of more than 20 proteins as judged by SDS^PAGE and autoradiography [5,11]. In the experiment shown in Fig. 1, the involvement of phosphatases was investigated by incubating the IO-SMP in the presence or absence of the phosphatase inhibitor NaF. The Coomassie-stained gel showed equal loading of all lanes (Fig. 1A). A comparison with the autoradiograph (Fig. 1B) showed that the polypeptide and phosphoprotein patterns were quite di¡erent. The labelling time was only 2 min, the time giving half-maximal phosphorylation [11] to maximize the sensitivity to inhibition, resulting in somewhat less 32 P incorporation than in subsequent experiments (cf. Fig. 2A). NaF at concentrations of 10 mM and above inhibited 32 P incorporation into all bands strongly (Fig. 1B). The phosphorylation of all the proteins was reduced by 75% on average at 25 mM NaF (Fig. 1C). Speci¢cally, NaF reduced the labelling of the two F0 F1 -ATPase subunits of 22 and 28 kDa [5], the prominent, but as yet unidenti¢ed, 16 kDa polypeptide, as well as the 41 kDa K-subunit of the pyruvate dehydrogenase (sometimes associating with the IO-SMP, [8]) (Fig. 1B). To test whether the e¡ect of 25 mM NaF on phosphorylation was due to increased ionic strength, we included 25 mM NaCl instead in a control experiment but it had no e¡ect on 32 P incorporation into the inner membrane proteins (results not shown). Okadaic acid, which is a phosphatase 1 and 2A inhibitor, used in mammalian systems [14], had no e¡ect on protein phosphorylation of the inner membrane of potato tuber mitochondria even at a concentration of 1 WM (results not shown). NaF has been demonstrated to be an e¡ective phosphoseryl and phosphothreonyl protein phosphatase inhibitor [15] in C Fig. 2. The e¡ect of the matrix fraction on the time-dependent dephosphorylation of inner membrane phosphoproteins. A: Autoradiograph of inner membrane phosphoproteins subjected to dephosphorylation for di¡erent times in the absence (lanes 2^6) or presence (lanes 8^12) of the matrix fraction. An inner membrane fraction was ¢rst phosphorylated for 15 min, then the ATP/ [Q-32 P]ATP was removed by centrifugation and the labelled inner membranes were incubated at room temperature for di¡erent times to become dephosphorylated. Lane 1, molecular weight standards (in kDa) visualized as in Fig. 1; lane 2, 1 min without matrix fraction; lane 3, 2 min without matrix fraction; lane 4, 7 min without matrix fraction; lane 5, 15 min without matrix fraction; lane 6, 60 min without matrix fraction; lane 7, molecular weight standards (in kDa); lane 8, 1 min with matrix fraction; lane 9, 2 min with matrix fraction; lane 10, 7 min with matrix fraction; lane 11, 15 min with matrix fraction; lane 12, 60 min with matrix fraction. The arrow indicates the position of phosphoproteins discussed in the text. The Coomassie-stained gel showed equal loading of proteins and identical polypeptide patterns in all lanes. B: Dephosphorylation pro¢le of the 16 kDa phosphoprotein. Results from A. 215 plant chloroplasts [16], and in mammalian [17] and plant [18] mitochondria. In these systems, NaF usually enhances 32 P incorporation into proteins by inhibiting protein phosphatases whose activity compete with that of the protein kinases. Since NaF inhibits protein phosphatases and not protein kinases, the decrease in 32 P incorporation in the presence of NaF (Fig. 1) was unexpected. We suggest that phosphorylation of the inner membrane proteins is dependent on a protein kinase that is activated by a protein phosphatase. The possible mechanism will be discussed later. Protein phosphatase activity was investigated as follows: the inner membrane proteins were phosphorylated using labelled ATP. After 15 min of labelling, to achieve maximum labelling of the phosphoproteins, the ATP was removed and 216 Fig. 3. Autoradiograph of the mitochondrial inner membrane phosphoproteins subjected to dephosphorylation under di¡erent conditions. Lanes 1^3, the ATP/[Q-32 P]ATP was removed after 10 min phosphorylation and the samples were further incubated for 60 min with the following additions: lane 1, matrix fraction; lane 2, bu¡er only; lane 3, matrix fraction+25 mM NaF. Lane 4, control sample labelled for 10 min with no subsequent incubation. Molecular mass standards are indicated at the left side of the ¢gure (in kDa). the incubation continued. There was no detectable dephosphorylation of the phosphoproteins even after a 60 min chase (Fig. 2A, lanes 2^6). Upon addition of the matrix fraction to inner membrane vesicles containing 32 P-labelled inner membrane proteins, dephosphorylation of all phosphoproteins in the inner membrane, including the 22 and 28 kDa subunits of the F0 F1 -ATPase, was observed (Fig. 2A, lanes 8^12). As an example, the time course of dephosphorylation of the 16 kDa inner membrane phosphoprotein, when treated in the presence or absence of a concentrated matrix fraction, is shown in Fig. 2B. The half-time of dephosphorylation of the 16 kDa phosphoprotein in the presence of matrix proteins was about 6 min. When NaF was added together with matrix proteins to prelabelled inner membrane proteins, the dephosphorylation was prevented (Fig. 3), while okadaic acid had no such e¡ect (result not shown). When a heat-denatured matrix fraction was added to phosphorylated inner membrane proteins lacking ATP, there was no detectable dephosphorylation (results not shown). There was no detectable proteolysis of 32 P-labelled IO-SMP proteins, and the protease inhibitors phenylmethylsulphonyl £uoride and E64 (L-trans-epoxysuccinyl-leucylamido-(4-guanidino)-butane) had no e¡ect on the dephosphorylation pattern of IO-SMP proteins (results not shown). These ¢ndings suggest that a phosphoprotein phos- A. Struglics et al./FEBS Letters 475 (2000) 213^217 phatase in the mitochondrial matrix is responsible for the dephosphorylation of the inner membrane phosphoproteins. There are at least two possible explanations for the NaF inhibition of protein phosphorylation in mitochondrial inner membranes (Fig. 1): (i) The substrates for the protein kinase are already phosphorylated when the IO-SMP fraction is prepared, and an active phosphatase is needed to dephosphorylate the IO-SMP proteins before they can be labelled with [Q-32 P]ATP. (ii) The protein kinase itself is inactive in the phosphorylated form and it is either phosphorylated in vivo or becomes phosphorylated upon addition of ATP. It will then require an active protein kinase phosphatase for activation of the protein kinase and phosphorylation of its substrates. Since the substrate phosphoprotein phosphatase is located in the matrix and not in the inner membrane (Fig. 2), then there must be another phosphatase that is inhibited when NaF is added to inner membranes resulting in a decrease of 32 P incorporation into IO-SMP proteins (Fig. 1). Therefore the decrease in 32 P incorporation must be the result of an inhibition of a kinase phosphatase in the IO-SMP. This supports the second proposal above: the protein kinase is in a phosphorylated and inactive form at the start of the experiment or when ATP is added to the membranes. Activation of a kinase then requires an active phosphatase. Sommarin et al. [18] and Petit et al. [9] showed that total 32 P incorporation into intact mitochondria reached a maximum incorporation after 1^2 min. A time course experiment on phosphorylation of inner membrane proteins showed that the maximum 32 P incorporation was reached only after 15^23 min [11]. The di¡erences between these results could be explained by the fact that in intact mitochondria, an equilibrium between kinase activities and phosphatase activities is rapidly reached. In the inner membrane vesicles where the substrate phosphatase is missing, an equilibrium is not reached, and thus it takes much longer to reach maximum 32 P incorporation. The half-time for the dephosphorylation of prelabelled inner membrane phosphoproteins by a phosphatase in the matrix was 6 min when 250 Wg IO-SMP proteins was incubated with 400 Wg matrix proteins in a total volume of 50 Wl (Figs. 2 and 3). The relative amounts of inner membrane and matrix proteins are close to that in intact mitochondrion (I.M. MÖller, A.G. Rasmusson and K.M. Fredlund, unpublished). However, the actual concentration of matrix proteins in the assay (8 mg ml31 ) is more than an order of magnitude below that in the intact mitochondrion where it may be as much as 500 mg ml31 [19], a concentration it is not possible to reach in vitro. Thus, the potential rates of dephosphorylation in vivo may well be much higher than those estimated here. 4. Conclusions b b Plant mitochondria contain a protein kinase phosphatase in the inner membrane which dephosphorylates, and thereby activates, the intrinsic protein kinase. Plant mitochondria contain a protein phosphatase in the matrix which can dephosphorylate all the phosphoproteins on the inner matrix surface of the inner membrane. Acknowledgements: We are grateful to Christina Nilsson for excellent technical assistance, and to the Swedish Natural Science Research A. Struglics et al./FEBS Letters 475 (2000) 213^217 Council for research grants to I.M.M. and J.F.A., the Swedish Council for Planning and Coordination of Research (FRN) to J.F.A. and to the Swedish Institute for a grant to Y.M.K. References [1] McEntyre, J. (1994) Trends Biochem. Sci. 19, 439^518. [2] Bradford, A.P. and Yeaman, S.J. (1986) in: Advances in Protein Phosphatases (Merlvede, W. and Di Salvo, J., Eds.), Vol. 3, pp. 73^106, Leuven University, Leuven. [3] Papa, S., Sardanelli, A.M., Cocco, T., Speranza, F., Scacco, S.C. and Technikova-Dobrova, Z. (1996) FEBS Lett. 379, 299^301. [4] Steenaart, N.A.E. and Shore, G.C. (1997) FEBS Lett. 415, 294^ 298. [5] Struglics, A., Fredlund, K.M., MÖller, I.M. and Allen, J.F. (1998) Biochem. Biophys. Res. Commun. 243, 664^668. [6] Ferrari, S., Moret, V. and Siliprandi, N. (1990) Mol. Cell Biochem. 97, 9^16. [7] Technikova-Dobrova, Z., Sardanelli, A.M. and Papa, S. (1993) FEBS Lett. 322, 51^55. [8] Sommarin, M., Petit, P.X. and MÖller, I.M. (1990) Biochim. Biophys. Acta 1052, 195^203. 217 [9] Petit, P.X., Sommarin, M., Pical, C. and MÖller, I.M. (1990) Physiol. Plant 80, 493^499. [10] Pical, C., Fredlund, K.M., Petit, P.X., Sommarin, M. and MÖller, I.M. (1993) FEBS Lett. 336, 347^351. [11] Struglics, A., Fredlund, K.M., MÖller, I.M. and Allen, J.F. (1999) Plant Cell Physiol. 40, 1271^1279. [12] Struglics, A., Fredlund, K.M., Rasmusson, A.G. and MÖller, I.M. (1993) Physiol. Plant 88, 19^28. [13] Laemmli, U.K. (1970) Nature 227, 680^685. [14] Cohen, P., Holmes, C.F.B. and Tsukitani, Y. (1990) Trends Biochem. Sci. 15, 98^102. [15] Brautigan, D.L. and Shriner, C.L. (1988) Methods Enzymol. 159, 339^346. [16] Bennett, J. (1980) Eur. J. Biochem. 104, 85^89. [17] Reed, L.J. and Damuni, Z. (1987) in: Advances in Protein Phosphatases (Merlvede, W. and Di Salvo, J, Eds.), Vol. 4, pp. 59^76, Leuven University, Leuven. [18] Miernyk, J.A. and Randall, D.D. (1987) Plant Physiol. 83, 311^ 315. [19] Srere, P.A. (1980) Trends Biochem. Sci. 5, 120^121.