* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download The potato tuber mitochondrial proteome
Biochemical cascade wikipedia , lookup
Oxidative phosphorylation wikipedia , lookup
Biochemistry wikipedia , lookup
NADH:ubiquinone oxidoreductase (H+-translocating) wikipedia , lookup
Plant virus wikipedia , lookup
Ancestral sequence reconstruction wikipedia , lookup
Evolution of metal ions in biological systems wikipedia , lookup
Paracrine signalling wikipedia , lookup
Gene expression wikipedia , lookup
Metalloprotein wikipedia , lookup
Signal transduction wikipedia , lookup
G protein–coupled receptor wikipedia , lookup
Expression vector wikipedia , lookup
Bimolecular fluorescence complementation wikipedia , lookup
Magnesium transporter wikipedia , lookup
Protein structure prediction wikipedia , lookup
Mitochondrion wikipedia , lookup
Interactome wikipedia , lookup
Nuclear magnetic resonance spectroscopy of proteins wikipedia , lookup
Protein purification wikipedia , lookup
Mitochondrial replacement therapy wikipedia , lookup
Two-hybrid screening wikipedia , lookup
Western blot wikipedia , lookup
Plant Physiology Preview. Published on December 18, 2013, as DOI:10.1104/pp.113.229054 1 Running head: The potato tuber mitochondrial proteome 2 3 4 5 Corresponding author: 6 7 Dr. Ian Max Møller 8 Department of Molecular Biology and Genetics 9 Science and Technology 10 Aarhus University 11 Forsøgsvej 1 12 DK-4200 Slagelse 13 Denmark 14 15 E-mail address: [email protected] 16 17 18 Research area: Cell Biology 19 1 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. Copyright 2013 by the American Society of Plant Biologists 20 Title: The potato tuber mitochondrial proteome 21 22 Salvato, Fernandaa,e), Havelund, Jesper F.b,c) , Chen, Mingjiea), Rao, R.S.P.a), Rogowska- 23 Wrzesinska, Adelinac), Jensen, Ole N.c), Gang, David R.d), Thelen, Jay J.a) and Møller, 24 Ian Maxb) 25 26 a) Department of Biochemistry and Interdisciplinary Plant Group, University of 27 Missouri-Columbia, Christopher S. Bond Life Sciences Center, Columbia, MO 65211, 28 USA 29 b) Department of Molecular Biology and Genetics, Science and Technology, Aarhus 30 University, Forsøgsvej 1, DK-4200 Slagelse, Denmark 31 c) Department of Biochemistry and Molecular Biology, University of Southern 32 Denmark, Campusvej 55, DK-5230 Odense M, Denmark 33 d) Institute of Biological Chemistry, Washington State University, Pullman, 34 Washington 99164 USA 35 e) Present address: Institute of Biology, State University of Campinas, São Paulo, Brazil 36 37 38 39 One-sentence summary: 40 A high-coverage potato tuber mitochondrial proteome uncovers many new proteins and 41 functions, especially in coenzyme and iron metabolism, and many posttranslational 42 modifications 43 2 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 44 Footnote: 45 This study was supported by grants from the Danish Council for Independent Research 46 – Natural Sciences (to IMM) and a 2012 Sabbatical fellowship (to JJT) from the OECD 47 Co-operative Research Programme: Biological Resource Management for Sustainable 48 Agricultural Systems. 49 3 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 50 ABSTRACT 51 Mitochondria are called the powerhouses of the cell. To better understand the role of 52 mitochondria in maintaining and regulating metabolism in storage tissues, highly 53 purified mitochondria were isolated from dormant potato tubers (Solanum tuberosum L. 54 cv. Folva) and their proteome investigated. Proteins were resolved by one-dimensional 55 gel electrophoresis, and tryptic peptides were extracted from gel slices and analyzed by 56 liquid chromatography-tandem mass spectrometry using an Orbitrap XL. Using four 57 different search programs, a total of 1060 non-redundant proteins were identified in a 58 quantitative manner using normalized spectral counts including as many as five-fold 59 more “extreme” proteins (low mass, high pI, hydrophobic) than previous mitochondrial 60 proteome studies. We estimate that this compendium of proteins represents a high 61 coverage of the potato tuber mitochondrial proteome (possibly as high as 85%). The 62 dynamic range of protein expression spanned 1800-fold and included nearly all 63 components of the electron transport chain, tricarboxylic acid cycle, and protein import 64 apparatus. Additionally, we identified 71 pentatricopeptide repeat proteins, 29 65 membrane carriers/transporters, a number of new proteins involved in coenzyme 66 biosynthesis and iron metabolism, the pyruvate dehydrogenase kinase, and a type 2C 67 protein phosphatase that may catalyze the dephosphorylation of the pyruvate 68 dehydrogenase complex. Systematic analysis of prominent post-translational 69 modifications (PTM) revealed that >50% of the identified proteins harbor at least one 70 modification. The most prominently observed class of PTMs was oxidative 71 modifications. This study reveals approximately 500 new or previously unconfirmed 72 plant mitochondrial proteins and outlines a facile strategy for unbiased, near- 73 comprehensive identification of mitochondrial proteins and their modified forms. 4 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 74 INTRODUCTION 75 Plant mitochondria participate in a number of processes in the plant cell depending on 76 the cell, tissue, or organ type, the developmental stage, and the environmental 77 conditions (Rasmusson et al. 2010, Millar et al. 2011). Important examples are energy 78 metabolism, photorespiration, amino acid biosynthesis, coenzyme (vitamin) 79 biosynthesis, and programmed cell death. All of these processes require that the 80 mitochondria can exchange metabolic intermediates and information with the rest of the 81 cell via membrane carriers and signaling pathways. Plant mitochondria also import 82 >95% of their proteins across the inner and outer mitochondrial membranes (IMM and 83 OMM, respectively). Finally, mitochondria are semi-autonomous organelles capable of 84 growing and dividing, and as such they perform DNA replication, DNA transcription, 85 and protein biosynthesis, in addition to protein import. 86 All of these processes require proteins and the composition of the plant mitochondrial 87 proteome therefore changes depending on the conditions (Millar et al. 2005). It has been 88 estimated that as many 2000-3000 gene products belong to the mitochondria in 89 Arabidopsis, but it is not known how many are expressed in a single tissue under a 90 specific set of conditions (Millar et al. 2005, Millar et al., 2006, Cui et al., 2011). To 91 date the most complete global proteomic study of isolated mitochondria identified 416 92 proteins in mitochondria from Arabidopsis cell cultures grown under standard 93 conditions with no photosynthesis (Heazlewood et al. 2004). In the intervening years a 94 number of more focused proteomic studies have been performed mainly on Arabidopsis 95 mitochondria swelling the number of identified plant mitochondrial proteins to 660 non- 96 redundant proteins (Heazlewood et al., 2004, Duncan et al., 2011, Klodmann et al. 97 2011, Lee et al., 2012, Tan et al., 2012). Missing from this list are many essential plant 98 mitochondrial activities including regulatory proteins, transcription factors, metabolite 99 translocators, and the wealth of tRNA synthases and pentratricopeptide repeat (PPR) 100 proteins (Havelund et al. 2013). 101 In yeast (Saccharomyces cerevisiae) about 850 proteins have been identified in isolated 102 mitochondria, which is thought to provide 85% coverage of the proteome in this species 103 (Premsler et al. 2009). A total of 1117 proteins were identified in mitochondria from the 104 nematode Caenorhabditis elegans (Li et al. 2009). The total mammalian mitochondria 105 is predicted to contain 1050-1400 proteins (Calvo and Mootha 2010) and a global 5 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 106 proteomic study of mitochondria isolated from fourteen mouse tissues using one- 107 dimensional polyacrylamide gel electrophoresis (1D-PAGE) followed by liquid 108 chromatography and two-dimensional mass spectrometry (LC-MS/MS) identified 1098 109 different gene products (Pagliarini et al. 2008). 110 Considering that the total plant mitochondrial proteome is thought to contain almost 111 twice as many proteins as the nematode and mammalian mitochondrial proteomes, it is 112 likely that an in-depth proteomic study of one type of plant mitochondria would identify 113 more than 1000 proteins or about twice the previous most complete study. It is likely 114 that newly discovered mitochondrial proteins would be of low abundance, which 115 conceivably would lead to the discovery of new pathways and regulatory proteins, as 116 well as providing additional information about established pathways. 117 In the present study, we describe the proteome of potato tuber mitochondria (POM). 118 The potato tuber is a large, relatively homogeneous storage organ and potato tuber 119 mitochondria are therefore expected to be similarly homogeneous. There are well- 120 established methods for isolating these mitochondria in a purified, intact and functional 121 form (Neuburger et al. 1982, Struglics et al. 1993, Considine et al. 2003) and their basic 122 properties are therefore well-characterized. In addition, the potato genome has recently 123 been sequenced (Xu et al. 2011), which greatly simplifies the task of identifying potato 124 proteins. Using a prefractionation by preparative SDS-PAGE followed by high-mass 125 accuracy LC-MS/MS we have achieved a high proteome coverage of 1060 proteins, 126 discovered a number of new functions, and identified a large number of 127 posttranslational modifications, particularly oxidative modifications. 128 RESULTS 129 The isolated potato tuber mitochondria are highly purified 130 The isolation of pure organellar fractions from crude source material is the critical phase 131 of any subproteome analysis. It has previously been established that very pure 132 mitochondria can be obtained from potato tubers, which only contain 0.2% peroxisomes 133 or plastids on a protein basis (Neuburger et al. 1982, Struglics et al. 1993, Considine et 134 al. 2003). We employed this established procedure for the isolation of mitochondria 135 with a recognized low contamination level. 6 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 136 We also tested the mitochondrial purity by Western blotting using subcellular marker 137 proteins. The proteins14-3-3 and anti-enolase were used as reference markers for the 138 cytosolic compartment, while alpha carboxytransferase (α-CT) and pyruvate 139 dehydrogenase E1 component subunit alpha (PDE1-α) were used as markers for the 140 plastidic and the mitochondrial compartment, respectively. Anti-PDE1-α demonstrated 141 the expected mitochondrial enrichment in the mitochondrial fraction, while there was no 142 detectable cytosolic or plastidic contamination (Figure S1). 143 The potato tuber mitochondrial proteome contains more than 1000 proteins 144 Previous Arabidopsis mitochondrial proteome studies, both shotgun and targeted, have 145 collectively identified a set of 660 non-redundant proteins (Heazlewood et al., 2004, 146 Duncan et al., 2011, Klodmann et al. 2011, Lee et al., 2012, Tan et al., 2012). These 147 studies employed primarily either two-dimensional (2D) gel electrophoresis or gel-free 148 analyses followed by tandem mass spectrometry (MS). In the present study we 149 performed one-dimensional gel electrophoresis followed by liquid chromatography- 150 tandem mass spectrometry (GeLC-LC-MS) to overcome limitations of 2D gels while 151 allowing for preparative level proteome interrogation. Using GeLC-MS, proteins from 152 each replicate were prefractionated by 12% SDS-PAGE and sectioned into 40 gel slices 153 prior to trypsin digestion, effectively reducing sample complexity prior to LC-MS/MS 154 (Figure 1). To maximize the number of assignments, four complementary database- 155 searching programs were employed simultaneously. Using a false discovery rate (FDR) 156 cutoff of 1% at the protein level, this approach resulted in 7346 assigned non-redundant 157 peptides representing 1060 protein groups including 23 that are mitochondrially 158 encoded. The average coverage was 24.9% (Table I, Table S1). 159 Figure 2 shows the overlapped peptides and proteins identified using MSGF-DB, 160 ProLuCID, Sequest, and Mascot. A comparison of these different search algorithms 161 revealed that MSGF-DB outperformed the other search engines by identifying 85% of 162 the total peptides, compared to 76% identified by ProLuCID, 46% by Sequest, and 37% 163 by Mascot. 164 165 Evaluation of the mitochondria proteome by transient fluorescence assay in tobacco epidermal cell 7 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 166 To evaluate the false positive rate of the mitochondria proteome data, 20 different 167 proteins were selected for further evaluation (Table S1, Figure 3, Figure S2). These 168 selected proteins either were of relatively low abundance and/or their mitochondrial 169 localization was at variance with previous reports. In other words, they were potential 170 contaminants. An EYFP fluorescence tag was in frame fused at their C-terminal; each 171 construct was mixed with a mitochondrial marker (CD3-986) and co-transformed into 172 tobacco leaves for subcellular localization by fluorescence microscope. Out of the total 173 20 constructs, 19 were detected in tobacco leaves at variable levels, and 18 of them 174 were confirmed to localize to mitochondria. PGSC0003DMP400005581, which encodes 175 a putative inorganic pyrophosphatase, was found to localize to chloroplasts (Figure 3). 176 Its ortholog in Arabidopsis (At5g09650) was confirmed to localize to chloroplasts 177 (Schulze et al., 2004). So this protein is likely to be a contaminant. 178 Two proteins (PGSC0003DMP400035788 and PGSC0003DMP400024128) were found 179 with dual localization. PGSC0003DMP400035788, annotated as pentatricopeptide 180 repeat-containing protein, was found to localize to both mitochondria and chloroplasts 181 (Figure S2). Its ortholog in Arabidopsis (At3g46870) was identified as a plastid protein 182 by organellar proteomics (Kleffmann et al, 2004); it also was identified as a Zn2+-IMAC 183 interaction protein in mitochondria (Tan et al, 2010). Its dual localization confirmation 184 here resolved the controversy between these two independent investigations (Kleffmann 185 et al, 2004; Tan et al, 2010). PGSC0003DMP400024128, which encodes an RNA- 186 binding protein 24-like, was found to localize to mitochondria and another unknown 187 organelle (Figure S2). This protein is low abundance and none of the subcellular 188 prediction algorithms predicted it to localize to mitochondria. 189 Two proteins (PGSC0003DMP400012523 and PGSC0003DMP400050514) were 190 observed to localize to mitochondria (Figure S2), and these observations contradict 191 previous reports. PGSC0003DMP400012523’s Arabidopsis ortholog (At1g78590) was 192 reported to localize to the cytosol although that conclusion is open to question as no 193 cytosolic marker was used (Chai et al., 2006). PGSC0003DMP400050514’s 194 Arabidopsis ortholog (At2g37250) was reported to locate to chloroplasts (Carrari et al., 195 2005). Since this Arabidopsis ortholog falls into the same phylogenetic group as the 196 potato plastidial isoform of adenylate kinase,with which it showed 76% identity, it will 197 require additional phylogenetic analysis to determine their orthologs. The existence of a 8 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 198 mitochondrial isoform of adenylate kinase in potato tuber was first reported by Roberts 199 et al. (1997). 200 In summary, these EYFP-localization results are consistent with the conclusion that the 201 potato tuber mitochondria are highly purified and that no more than 5% of the 1060 202 identified mitochondria proteins are false positives. This ratio could be well be 203 overestimated since the candidates for the localization study were purposely selected 204 based on their low abundance and their contradiction with reported results in the 205 literature. 206 Categorization of the identified proteins according to properties and function 207 Considering proteins only detected in at least two replicates, we have identified 884 208 protein groups (Table S3), of which 32% were confirmed by ortholog analysis when 209 compared to the Arabidopsis mitochondrial proteome (Heazlewood et al., 2004, Duncan 210 et al., 2011, Klodmann et al. 2011, Lee et al., 2012, Tan et al., 2012). Thus 528 new 211 non-orthologous proteins are presented in the current study, in comparison to all 212 previous Arabidopsis mitochondrial proteome investigations. As shown in Figure 4, the 213 potato mitochondrial proteins were distributed between 4 to 161 kDa in size, between - 214 0.9 to +0.3 in GRAVY score, and with pI values ranging from 4 to 11. A comparison of 215 the physical properties of the potato versus the Arabidopsis mitochondrial proteome 216 (Heazlewood et al., 2004) demonstrates that a much larger number of small proteins 217 with negative GRAVY scores have been identified in the potato mitochondrial 218 proteome (Figure 4). 219 Assigned potato mitochondrial proteins were annotated by BLAST querying against the 220 non-redundant NCBI database and were classified based on the gene ontology terms 221 related to biological process and molecular function (Figure S3). In addition, each 222 protein was assigned to thirteen general functional categories according to the 223 classification by Heazlewood et al. (2004). The distribution of proteins classified in 224 each category is compared between potato and Arabidopsis mitochondrial proteomes 225 (Table I). The largest functional classes in the potato mitochondrial proteome were 226 energy (12%), metabolism (20%), and protein synthesis (12%). Altogether these 227 proteins represent 45% of total protein abundance based on normalized spectral 228 counting (dNSAF). These proteins are mainly involved in the tricarboxylic acid cycle 9 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 229 (TCA), oxidative phosphorylation, and protein synthesis (mainly ribosomal and tRNA 230 synthetase proteins). Around 23% of the identified proteins (also, 23% of the 231 abundance) were categorized in the RNA processing, DNA synthesis and processing, 232 transcription or protein fate classes. In addition, approximately 12% of the identified 233 proteins (also 12% of the abundance) were related to transport, or 234 defense/detoxification, or signaling processes. 235 The number of proteins found in each functional category was higher for potato than for 236 Arabidopsis with the exception of structural organization, although the increase for the 237 energy category was relatively small (1.3 fold) (Table I). The most marked increase in 238 number of proteins identified was seen for RNA processing and protein synthesis where 239 7.4-fold and 8.4-fold more proteins were found in potato, but also the category transport 240 showed a large increase (3.6 fold). The relative frequency (%) of proteins in each 241 category were very similar, with the exception of the energy category, which was 242 significantly lower in potato, and RNA processing and protein synthesis categories, 243 which were significantly higher. Unclassified proteins, that is, proteins not confidently 244 assigned to any functional class, comprised 21% of the total number of identified 245 proteins (20% of the abundance) in the potato mitochondrial proteome (Table I). 246 A number of new proteins and enzymes were identified that increase our coverage of 247 known mitochondrial processes and, more interestingly, extend the list of mitochondrial 248 processes for instance with respect to coenzyme and iron metabolism. This will be 249 treated in the Discussion. 250 The abundance of mitochondrial proteins varies more than 1800-fold 251 It has been shown that in LC-MS/MS based analysis of peptide mixtures raw spectral 252 counts correlate with relative protein abundance between different samples (Liu et al., 253 2004; Stevenson et al., 2009, Matros et al., 2011). However, to compare the relative 254 abundance of different proteins within the same biological sample spectral counts for 255 each protein must be normalized. One of the most logical and promising strategies for 256 normalization is based on protein length (Zhang et al., 2010). 257 In the present work, proteins that were detected in at least two replicates (884 proteins) 258 were assigned an abundance factor (Table S3). For this, we calculated the distributed 259 normalized spectral abundance factor (dNSAF) which considers unique spectral counts 10 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 260 and distributes the shared spectral counts matched to peptides assigned to homologous 261 proteins. Spectral counts of each protein were then normalized by protein length (see 262 Methods for details). 263 Figure 5 shows the log10 transformed dNSAFs distribution of the 884 potato 264 mitochondrial proteins. Each bin corresponds to 0.25 order of magnitude difference in 265 the protein abundance. In this way, the potato mitochondrial protein abundance spans 266 three orders of magnitude ranging from -4.82 to -1.57 log10 values (Table S3) giving a 267 dynamic range of about 1800 fold as calculated by the normalized spectral count 268 (dNSAF). This is likely a low estimate of dynamic range as the most abundant proteins 269 may be underestimated. The least abundant protein detected in the potato proteome was 270 a ribonuclease (GI 255538392) and the most abundant protein was a phosphate carrier 271 protein (GI 255543593). Considering the previous mitochondrial study in Arabidopsis 272 (Heazlewood et al., 2004), the current study significantly increased coverage of both 273 low- and high-abundance proteins. For example, 53 predicted mitochondrial ribosomal 274 proteins were confidently detected in at least two replicates in the current study with an 275 average abundance of -2.68 [log10(dNSAF)] against only three ribosomal proteins 276 detected in the previous Arabidopsis study (Heazlewood et al., 2004). Also, 52 proteins 277 involved in the uptake of a variety of metabolites (ADP/ATP, dicarboxylate, 278 fumarate/succinate, phosphate, malate) including membrane carriers, transporters, and 279 porins were identified in the potato mitochondrial proteome with an average abundance 280 of -3.31, while all previous Arabidopsis studies detected 19 proteins related to the same 281 functional class. Proteins with well-known roles in primary mitochondrial metabolism 282 including proteins involved in oxidative phosphorylation via the respiratory complexes 283 and proteins from the TCA cycle were detected with an average abundance of -2.82 284 comprising around 28% of total protein abundance. We also quantified low abundance 285 proteins like those involved in RNA processing. For example, we quantified 62 286 pentatricopeptide repeat (PPR) proteins, which have almost two orders of magnitude 287 difference in expression levels (from -4.67 to -2.69 [log10(dNSAF)]) (Table S3). 288 An overview of the number and abundance of the identified proteins belonging to major 289 mitochondrial metabolic pathways are shown as heat maps in Figures 6 and 7. Figure 6 290 shows the whole mitochondrion with processes placed in the known or presumptive 291 mitochondrial subcompartment – OMM, intermembrane space (IMS), IMM, matrix – 292 while Figure 7 shows the respiratory complexes and a few associated enzymes located 11 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 293 in the IMM. Spectral counting should not be used to determine the stoichiometry of 294 subunits in a complex, particularly membrane complexes. Having said that, the apparent 295 wide variation in the amounts of the subunits, e.g. of Complex V, is mainly due to the 296 presence of low-abundance paralogs (Table S3). If we compare the relative abundances 297 of all the F1-ATPase subunits identified, they range between -3.72 and -1.96 298 [log10(dNSAF)] or almost 100-fold (Table S3). However, if we only compare the most 299 abundant paralogs of each of the five subunits identified, the range is much more 300 narrow, between -2.91 and -1.96 [log10(dNSAF)], which is about 10-fold or about the 301 same range as the expected 3:3:1:1:1 stoichiometry of the α, β, γ, δ, ε subunits (von 302 Ballmoos et al. 2009). 303 Prediction programs for mitochondrial localization recognize 63% of all identified 304 proteins 305 There are a large number of programs available to predict subcellular localization based 306 on the N-terminal region of proteins that may contain intrinsic targeting peptides 307 (Scheneider and Fechner, 2004). The protein sequences of the identified mitochondrial 308 proteome were analyzed by five different prediction programs: TargetP (Emanuelsson et 309 al., 2000), Predotar version 1.03 (Small et al. 2004), MitoProtII (Claros and Vincens, 310 1996), iPSORT (Bannai et al., 2002) and WoLF PSORT (Horton et al., 2007). Each 311 program employs different methods for subcellular prediction and usually produces 312 relatively large non-overlapping sets of results (Heazlewood et al., 2004). Each 313 prediction tool returned a variable percentage of mitochondrial positives ranging from 314 approximately 20% (WoLF PSORT) to 52% (MitoProtII) of the potato proteome (Table 315 S4). We also performed a relational evaluation of the overlapping positive predictions 316 by different combinations of the prediction tools. When we considered multiple 317 prediction tools together, the number of positive mitochondrial predictions decreased 318 (Table S4), indicating a much smaller agreement between programs. The combination 319 of MitoProt II and iPSORT provided the best pair for prediction resulting in 37% 320 positive potato mitochondrial predicted proteins (Table S4). 321 In all, 671 out of 1060 proteins (63% of the total protein set) were predicted to be 322 mitochondrial by at least one program, while 48% were predicted by two or more 323 programs (Fig. 8). From the list of 389 proteins not predicted to be mitochondrial by 324 any program, 44 proteins were confirmed OMM or IMM proteins related to the protein 12 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 325 import system and uptake of metabolites (carriers, transporters, channels, and 326 translocases) which are characterized by the absence of N-terminal targeting 327 presequences (Millar and Heazlewood, 2003; Huang et al., 2010). In addition, 31 well- 328 known mitochondrial proteins involved in the TCA cycle and the respiratory complexes, 329 as well as 11 proteins encoded by mitochondrial genes, were also negative for 330 mitochondrial localization prediction. These results show the limitations in assigning 331 mitochondrial localization based solely on bioinformatic tools. The choice of prediction 332 program can benefit different classes of proteins and the absence of targeting 333 presequences, due to genuine lack of those N-terminal targeting signaling sequences 334 (e.g. OMM or IMM proteins or mitochondrially encoded proteins), or due to incorrect 335 annotation of the N-terminus of the sequences, can result in erroneous interpretation. 336 In Figure 8, the subcellular prediction information is superimposed on the relative 337 protein abundance. The data reveal that prediction results are comparable irrespective of 338 protein abundance, suggesting that low abundance proteins must also be considered as 339 authentic mitochondrial proteins. 340 Proteins with uncleaved N-terminal sequences are sometimes further processed 341 Semi-tryptic database querying resulted in 37 matches to the protein N-terminus 342 indicating that these proteins do not contain a cleavable targeting peptide. Some of these 343 peptides had the terminal Met removed while others did not. The removal of the Met 344 can be related to the length of the side chain of the second amino acid in the sequence 345 (Gliglione and Meinnel, 2001). If the second amino acid is Ala, Gly, Pro, Ser or Thr, the 346 initiating Met is removed by a Met aminopeptidase. On the other hand, if the side-chain 347 is large, in case of Arg, Asn, Asp, Glu, Ile, Leu or Lys, the Met is retained. Only Val 348 appears to have an intermediate-side-chain specificity for Met cleavage, showing the 349 initiating Met cleaved or uncleaved. As shown in Table S5 our results confirm this 350 specificity. The Met was removed when the second side chain was Lys, Ala, Pro, Ser, 351 Thr or Val and not removed when it was Lys, Asn or Asp or Gln. Surprisingly, when the 352 second residue was Gly, some sequences had Met removed while others did not. 353 Many of the proteins identified with an uncleaved N-terminus were OMM proteins, 354 including porins, or components of the respiratory complexes present in the IMM, such 355 as NADH dehydrogenase, bc1 complex, and ATPase. Additionally, two mitochondrially 13 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 356 encoded proteins were identified with intact N-termini, a ribosomal protein 357 (STmpRH_41) and an NADH dehydrogenase subunit (STmpRH_8). 358 The majority of proteins in the potato mitochondrial proteome are post- 359 translationally modified 360 The potato mitochondrial proteome was extensively searched for different types of post- 361 translational modifications (PTMs) ( Tables S6 and S7). We identified 3,066 PTM sites 362 in a total of 556 proteins which means that 52% of the entire potato mitochondrial 363 proteome determined here exist in modified versions. Individual proteins contained up 364 to eight types of PTMs and up to 50 different modification sites. The proteins with most 365 modifications - aconitase, 2-oxoglutarate dehydrogenase, succinate dehydrogenase and 366 glycine dehydrogenase - are all part of the TCA cycle or associated with it (Table S7). 367 Mitochondria are not only the sites of oxygen consumption but also one of the sources 368 of cellular reactive oxygen species (ROS) (Møller 2001). We therefore specifically 369 focused our analyses on protein oxidation in the potato mitochondrial proteome. 370 Oxidative modifications of Arg, Pro, Lys, Met, Thr, and Trp were detected. A total of 371 505 proteins contained at least one type of oxidative modification giving a total of 2,471 372 sites of modification. As shown in Figure 9, Met sulfoxidation followed by Pro 373 oxidation (either hydroxylation or carbonylation) and Lys hydroxylation are the most 374 abundant types of PTM in the proteome. Most of the modified proteins participate in the 375 TCA cycle, respiratory complexes, protein import assembly, and protein synthesis 376 (ribosomal proteins) (Table S7). The high frequency of Met sulfoxidation is in 377 accordance with several reports that found sulfur amino acids to be more sensitive to 378 oxidation by ROS than other amino acids (Stadtman et al., 2005). 379 Carbonylated proteins include both high and low abundance proteins spanning a variety 380 of metabolic pathways (energy, metabolism, protein and DNA synthesis, protein fate 381 and transport). Among the highly abundant carbonylated proteins (abundance from - 382 2.99 to -1.56 log10[dNSAF]), up to 50 oxidation sites were mapped to a single protein, 383 glycine dehydrogenase (Table S7). Other examples were proteins involved in the TCA 384 cycle (aconitase, 2-oxoglutarate dehydrogenase, succinate dehydrogenase, citrate 385 synthase, isocitrate dehydrogenase), respiratory complexes (ATPase, cytochrome c1, 386 cytochrome c oxidase), chaperonins, carrier proteins (phosphate, oxoglutarate, 14 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 387 ADP/ATP), pyruvate dehydrogenase complex (E1 α and β subunits) and redox 388 metabolism (peroxiredoxin, Mn superoxide dismutase, monodehydroascorbate 389 reductase). Many low-abundance proteins (-4.65 to -3.0 log10[dNSAF]) also showed a 390 large number of oxidation sites, up to 25 sites. Proteins with more than ten oxidation 391 sites include FtsH proteases, monoxygenase, ATPase and branched-chain α-keto acid 392 dehydrogenase (Table S7). 393 We also detected Asn deamidation (188 proteins), Lys acetylation (31 proteins) and 394 methylation (63 proteins), and Ser/Thr phosphorylation (19 proteins). These proteins 395 showed a wide range of relative abundance (-4.56 to -1.56 log10[dNSAF]) and 396 comprised components of TCA cycle and associated enzymes, chaperone system, 397 respiratory complexes, protein import system, RNA processing and carriers proteins. 398 399 DISCUSSION 400 The discord between the number of predicted and reported plant mitochondrial proteins 401 is strikingly high – probably more so than for any other organelle. Despite this 402 discordance, no shotgun proteomic study of plant mitochondria has been published 403 since 2005. Many advances in proteomics have occurred during the intervening time, 404 including the development of best practices for unbiased protein prefractionation, 405 sensitive and higher-accuracy mass spectrometers, as well as various refinements in 406 mass spectral data querying. In this study, we leverage developments on each of these 407 fronts to offer a quantitative, unbiased, and comprehensive proteome resource for the 408 plant mitochondrial community, including about 500 new experimentally identified 409 plant mitochondrial proteins. We estimate that the coverage of the mitochondrial 410 proteome is high (possibly as high as 85%) (see Supplementary Discussion). Since the 411 theoretical size of the plant mitochondrial proteome is about twice the size of the the 412 potato tuber proteome, it is likely that the specialized and physiological condition of the 413 tissue and its mitochondria placed a limit on the number of proteins identified. Thus, we 414 can discuss not only proteins found, but also comment with some confidence on 415 proteins or protein groups underrepresented. This will be a recurrent theme in the 416 following discussion as well as in Supplementary Discussion. 417 Prediction programs 15 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 418 A comparison of five different subcellular localization prediction programs, TargetP, 419 Predotar, MitoProtII, iPSORT and WoLF PSORT), revealed a variable number of 420 mitochondrial predictions, ranging from 20 to 52% of the entire experimentally 421 determined mitochondrial proteome (Fig. 8, Table S4). These values are low compared 422 to the previous values of 70 to 90% reported in the original articles for these programs 423 (Emanuelsson et al., 2000; Small et al. 2004, Claros and Vincens, 1996). This could be 424 the result of bias in the protein training sets used in developing the programs. Also, we 425 found that the combination of multiple prediction programs showed low numbers of 426 overlapping positive predictions, indicating the divergence of the prediction methods 427 employed by each tool. Overall, 63% of the potato mitochondrial proteome was 428 predicted to be mitochondrially localized by at least one of the five prediction tools 429 (Figure 8, Table S4). However, authentic mitochondrial proteins involved in the TCA 430 cycle and the respiratory complexes as well as proteins characterized by the absence of 431 targeting presequences (OMM or IMM and mitochondrially encoded proteins) were 432 erroneously predicted to be non-mitochondrial. The number of such false negatives was 433 only partially offset by the use of different algorithms, indicating that the results 434 returned by such tools should be inspected carefully. False positives appear to be less of 435 a problem with these prediction programs. Nevertheless, even if a protein is not 436 predicted to be mitochondrial, one should not assume it is a contaminant as the accuracy 437 of these programs is both variable and limited. As most of these programs are decades 438 old, the development of new improved programs, based on more comprehensive 439 genomic sequences and proteomics data, is required to overcome the lack of accuracy 440 observed. 441 Basic metabolic processes 442 The coverage of the well-known mitochondrial functions - TCA cycle, respiratory 443 chain, transmembrane transporters, and turnover of amino acid, proteins, lipids and fatty 444 acids - was high as shown in Figures 6 and 7 and listed in Table S1. A detailed 445 discussion of the potato tuber proteins identified for each group as well as for glycolysis 446 and signaling is found in Supplementary Discussion. In the following we will discuss a 447 number of interesting protein groups linked to other important mitochondrial functions. 448 Biosynthesis of coenzymes and pyrimidines 16 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 449 Vitamins are coenzymes, which humans are unable to synthesize but acquire through 450 their food, mostly from plants. Mitochondria are involved in the biosynthesis of several 451 of the coenzymes in plant cells and we have found a number of key enzymes in these 452 processes, some of which have previously only been predicted. 453 Although NAD+ is the main pyridine coenzyme in mitochondria, a number of enzymes 454 use NADP+ (Møller and Rasmusson 1998). NAD+ uptake into plant mitochondria was 455 discovered in the 1980’s using POM (Neuburger et al. 1985), but NADP+ uptake has 456 also been inferred at least for pea leaf mitochondria (Bykova and Møller 2001). We here 457 identify an NAD+ carrier as well as an NADH kinase (Table S1). Three NAD(H) 458 kinases are found in Arabidopsis and one of them, NDK3, which appears to prefer 459 NADH as substrate over NAD+, has been reported to be localized to peroxisomes in 460 higher plants while its analogue in yeast is mitochondrial (Waller et al. 2010). The POM 461 NADH kinase is predicted to be mitochondrial by three prediction programs and is of 462 relatively low abundance (-3.87 [log10(dNSAF)]) (Tables S1 and S3, Fig. 6). Thus, in 463 order to generate the NADP+/NADPH pool inside the POM, it appears that NAD+ is 464 taken up, reduced by one of the many matrix dehydrogenases, e.g. in the TCA cycle, 465 and then phosphorylated to give NADPH. This does not imply that NADPH/NADP+ 466 ratios in the mitochondria are determined by this pathway, rather that generation of the 467 overall NADP+/NADPH pool size would be governed by the activities of the NAD+ 468 transporter and the NADH kinase. 469 L-Galactono-1,4-lactone dehydrogenase, the last enzyme in one of the ascorbate 470 biosynthetic pathways, which donates electrons to Complex IV on the outer surface of 471 the IMM (Bartoli et al. 2000), but is physically associated with Complex I (Pineau et al. 472 2008), was found at above median quantities. Five different enzymes using ascorbic 473 acid mainly for ROS detoxification were also found (see later). Clearly ascorbate must 474 be imported in some form, but the mitochondrial carrier protein family does not appear 475 to include a carrier for ascorbate or dehydroascorbate (Taylor et al. 2010, Palmieri et al. 476 2011) and no such carrier was found in our study. 477 Tetrahydrofolate (vitamin B9) and its derivatives catalyze the addition or removal of 478 C1-units, e.g., in the glycine decarboxylate reaction in the mitochondrial matrix where a 479 methyl group is transferred from one glycine molecule to another to form serine. Folates 480 consist of a pterin moiety, a p-aminobenzoate moiety and a (poly)glutamate tail and the 17 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 481 three parts are assembled in the mitochondria (Blancquaert et al. 2010). In the POM, we 482 have found three enzymes involved in folate biosynthesis, dihydropterin 483 pyrophosphokinase-dihydropteroate synthase, a molybdopterin biosynthesis protein and 484 polypolyglutamate synthase, four enzymes involved in the interconversion and transfer 485 of different C1-units, 5-formyltetrahydrofolate cycloligase, formyltetrahydrofolate 486 deformylase-like, methenyltetrahydrofolate synthase domain-containing protein-like, 5- 487 methyltetrahydropteroyltriglutamate-homocysteine methyltransferase-like and finally a 488 folate carrier responsible for exporting folate to make the coenzyme available to the rest 489 of the cell. 490 Biotin (vitamin B8) is a coenzyme involved in carboxylations and the last steps in its 491 biosynthesis are mitochondrial (Alban 2011). We found the S-adenosylmethionine 492 carrier responsible for the import of one of the precursors, but not predicted to be 493 mitochondrial by any of the prediction programs. We also detected the adrenodoxin 494 reductase involved in biotin biosynthesis and predicted to be mitochondrial by three 495 programs (Table S1). The adrenodoxin reductase uses adrenodoxin to reduce S-S 496 bridges and we find five ferredoxin analogs, which may well be adrenodoxin. Thus, a 497 significant part of the biotin biosynthesis pathway is expressed in POM. 498 Pyrimidine biosynthesis mostly takes place in the cytosol and the nucleus in mammalian 499 cells, but the fourth reaction in the pathway, catalyzed by dihydroorotate 500 dehydrogenase, takes place on the outer IMM surface (Löffler et al. 2005). Heazlewood 501 et al. (2004) found this enzyme in Arabidopsis mitochondria and in POM it is present at 502 higher than median relative abundance (Table S3) confirming that it is mitochondrial 503 and probably donates its electrons to ubiquinone as in mammalian mitochondria (Fig. 504 7). 505 The identification in this one investigation of so many enzymes involved in coenzyme 506 biosynthesis in the mitochondrion is a good demonstration of the power of applying 507 high resolution and highly sensitive mass spectrometry-based proteomic tools to 508 important questions in plant biochemistry. 509 DNA and RNA 510 We identified DNA polymerases, RNA polymerases, RNA helicase, histones, an 511 histone-modifying enzyme, a topoisomerase, transcription factors and a transcription 18 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 512 termination factor, most at low abundances (Table S1 and S3) and representing almost 513 all of the components required for DNA replication and transcription. Three of the five 514 histones identified are of a type recently reported to be present in plant mitochondria 515 (Zanin et al. 2010). 516 Pentatricopeptide repeat (PPR) proteins 517 PPR proteins are one of the most prolific protein families in plants while it is virtually 518 absent in animals (Small and Peeters 2000). The majority of the 450 PPR proteins in 519 Arabidopsis are predicted to be mitochondrial where about half are thought to be 520 involved in RNA editing and the remainder in other types of RNA processing (Fujii and 521 Small 2011). In Arabidopsis mitochondria, Heazlewood et al. (2004) found 10 PPR 522 proteins while in potato tuber mitochondria we found 71. From the total of 71 PPR 523 proteins, we relatively quantified (dNSAF) 62 PPR proteins found in at least two 524 replicates. The abundance of these PPR proteins varied by almost two orders of 525 magnitude based on dNSAF values. Only 7 PPR proteins can be considered to be 526 present at high relative abundance (-2.9 to -2.6 [log10(dNSAF)]), whereas 23 can be 527 considered medium relative abundance (-3.9 to -3.0 [log10(dNSAF)]) and 32 low 528 relative abundance (-4.67 to -4.0 [log10(dNSAF)]). 529 Since 450 PPR proteins are present in Arabidopsis (Fujii and Small 2011) and since the 530 potato tuber genome contains many more genes than the Arabidopsis genome, it is very 531 likely that at least 300 PPR proteins are targeted to the mitochondria in some potato 532 tissue under some condition. In spite of the fact that all proteins encoded in mtDNA are 533 subunits in the constitutively expressed respiratory complexes or appear to be essential 534 housekeeping proteins, the finding of “only” 71 PPR proteins indicates that many RNA- 535 related processes are inactive in POM. In mitochondria from perennial ryegrass 536 reproductive tissues, a transcript profile showed that the expression of several 537 mitochondrially expressed ribosomal proteins was very low (Islam et al. 2013). This 538 may mean that their expression, and the associated RNA-related processes including 539 PPR proteins, is dependent on the tissue and/or environmental conditions. 540 Turnover of Reactive Oxygen Species 541 The mitochondrion is one of the major sites of ROS production in the cell (Maxwell et 542 al. 1999, Møller 2001, Foyer and Noctor 2003) and it therefore also contains a number 19 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 543 of enzymes or enzyme systems capable of removing ROS. We found three of the four 544 enzymes in the ascorbate-glutathione pathway, glutathione reductase, ascorbate 545 peroxidase, and monodehydroascorbate reductase (dehydroascorbate reductase was not 546 detected) (Jimenez et al. 1998, Chew et al. 2003) with very similar relative abundances 547 (-3.9 to -3.6 [log10(dNSAF)]), except for one monodehydroascorbate reductase (GI 548 350536875) which presented higher relative abundance (-2.8 log10(dNSAF)]) (Fig. 5). 549 Also, we have identified both the classical mitochondrial Mn-SOD as well as a Cu,Zn- 550 SOD, which at least in yeast is found in the mitochondrial intermembrane space 551 (O’Brien et al. 2004). Mn- and Cu,Zn SODs showed high (-1.7 [log10(dNSAF)]) and 552 medium (-3.5 [log10(dNSAF)]) relative abundances, respectively. 553 We found thioredoxin (-3.1 [log10(dNSAF)]) and thioredoxin reductase, phospholipid 554 hydroperoxide glutathione peroxidase (-3.2 log10[dNSAF]) and peroxiredoxin (-2.6 555 log10[dNSAF])) – all in medium relative abundance. In other words, all the four 556 predicted NADPH-consuming enzyme systems involved in removing H2O2 or H2O2- 557 induced peroxidation products (Møller 2007) are expressed in POM, which appear well 558 armed to deal with oxidative stress. 559 Finally, we detected two catalases (one at low level – 4.4 and one at medium level (-2.6 560 [log10(dNSAF)]). Although POM only contain a maximum of 0.2% peroxisomes on a 561 protein basis, the catalases could be contaminants, since catalase makes up 562 approximately 50% of all the peroxisomal protein in potato tubers (Struglics et al. 563 1993). However, catalase has been reported to be present inside rat heart mitochondria 564 (Radi et al. 1991) so the parsimonious explanation of catalase inside plant mitochondria 565 cannot be excluded. 566 Metal ion turnover 567 Mitochondria contain many proteins, particularly redox active enzymes, which bind 568 metal ions often at their catalytic site. As mitochondria are also involved in the 569 biosynthesis of FeS centers and hemes, the concentration of metal ion is high in 570 mitochondria (Tan et al. 2010). However, because free metal ions can interact with 571 hydrogen peroxide and form the highly reactive and damaging hydroxyl radical, which 572 can lead to protein oxidation (see below), it is very important for the cell to keep metal 573 ions safely bound to proteins at all stages of metal ion uptake and metabolism (Kell 574 2009, 2010, Møller et al. 2011). We found a number of prominent metal-containing 20 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 575 proteins including the FeS-proteins and the Cu/Fe-containing cytochromes in the ETC, 576 but also Mn- and Cu/Zn-SOD and TCA cycle enzymes like aconitase (Fe). In addition, 577 we found the Fe-storing ferritin (low abundance) and the complete pathway for the 578 biosynthesis of FeS clusters – frataxin, iron-sulfur cluster assembly proteins, iron-sulfur 579 cluster scaffold protein, iron-sulfur cluster co-chaperone protein, cysteine desulfurase 580 and ferredoxin (all medium abundance) (Ye and Rouault 2010). 581 A majority of potato mitochondrial proteins are post-translationally modified 582 More than half the proteins detected in the potato mitochondrial proteome were post- 583 translationally modified on at least one site. And about 100 proteins had more than ten 584 PTM sites. The most modified protein, glycine dehydrogenase, had as many as 50 PTM 585 events of six different kinds including 24 Met oxidations, 3 Asp deamidations and 12 586 Pro oxidations (Table S7). In most cases the spectral counts were very low for each 587 modified peptide identified. It is likely that the coverage of modified peptides were 588 underestimated as PTMs such as Lys or Arg oxidation causes missed trypsin cleavage 589 sites resulting in larger peptides that are more difficult to ionize and fragment. 590 The large number of PTM sites on some proteins could give rise to a very large number 591 of differentially modified versions of the protein unless the PTMs are introduced in an 592 ordered rather than a random manner (Thelen and Miernyk, 2012). A relatively 593 unexplored research area is the interaction between different PTMs. At least in one case 594 we know that such an interaction takes place – The oxidation of a Met residue to Met- 595 sulfoxide next to a potential phosphorylation site abolishes phosphorylation on that site 596 and this is likely a regulatory mechanism as this first step in Met oxidation is reversible 597 (Huang et al., 2010). 598 In Supplementary Discussion we analyze phosphorylated proteins, oxidized proteins 599 and proteins with acetylated and methylated lysines. 600 Conclusions 601 A comprehensive shotgun proteomic investigation of potato tuber mitochondria led to 602 the detection of 1060 proteins, which appear to represent more than 85% of the proteins 603 expressed in the POM. A significant disparity was observed between the proteins 604 actually identified in the proteome and those predicted to be targeted to the proteome by 605 the five most commonly used prediction algorithms. Less than 50% of the proteins 21 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 606 determined to be present in the mitochondrial proteome were predicted to be thus 607 localized by at least two of these programs. Indeed, only 63% were thus predicted by at 608 least one program. This clearly demonstrates that current bioinformatics tools, while 609 often useful, can be very inadequate at predicting biological function and localization. 610 The broad coverage observed in the POM proteome allowed for investigation of 611 processes that are important for mitochondrial function, such as metabolism and 612 respiration and post-translational regulation of protein function. The ETC was well 613 covered in the POM proteome, with all complexes being well represented. In addition, 614 three different alternative NAD(P)H dehydrogenases were present at significant levels, 615 as were all members of the TCA cycle, except fumarase, suggesting that non-cyclic 616 modes of operation are likely to be functional. In addition, the biosynthesis of many 617 coenzymes was well represented in the proteome, supporting the role of the 618 mitochondrion in production of these important molecules. The general processes of 619 amino acid biosynthesis and lipid metabolism were well supported in the proteome. The 620 fact that a number of glycolytic enzymes were detected can be explained by their 621 association with the OMM. A number of other processes important to mitochondrial 622 function and replication were well supported, including transport, protein synthesis, 623 regulation of gene expression within the mitochondria, and metal ion regulation. 624 Finally, a large number of protein PTMs were found in the mitochondrial proteome, 625 with over 50% of proteins possessing such modifications and several proteins 626 possessing over a dozen modifications, such as phosphorylation and oxidative 627 modifications. The ease with which these modifications were found is a testament to the 628 power of the approach used. The fact that so many proteins contained at least one, and 629 many contained more than one PTM, suggests that this mode of functional regulation 630 plays an important role in the mitochondria. It further suggests that approaches beyond 631 transcriptional profiling will be not only desired, but required, if we are to be able to 632 better understand how mitochondria help plants respond to stress and changing 633 environments. 634 MATERIALS AND METHODS 635 Isolation of mitochondria 22 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 636 Mitochondria were isolated from potato (Solanum tuberosum L., cv. Folva) tubers by 637 the method of Considine et al. (2003) with minor modifications. All procedures were 638 done at 4°C. Peeled tubers were homogenized using a juice extractor into an equal 639 volume of extraction medium (0.6 M mannitol, 20 mM MOPS-KOH (pH 7.3), 2 mM 640 EDTA, 25 mM cysteine and 0.3 % (w/v) BSA). The homogenate was left standing for 5 641 min allowing starch to sediment before it was filtered through 2 layers of cotton towel 642 and centrifuged at 3000 g for 5 min. The supernatant was centrifuged at 18000 g for 10 643 min to recover an organelle pellet. The pellet was resuspended in mannitol buffer (0.3 644 M mannitol, 10 mM MOPS-KOH (pH 7.3), 0.1 % (w/v) BSA) and layered on top of a 645 Percoll step gradient of 50 %, 28 % and 20% in mannitol buffer. After centrifugation at 646 40000 g for 30 min the pale mitochondrial band was recovered from the 28%/50% 647 Percoll interface. After washing, the mitochondria were further purified on a second 648 self-forming Percoll gradient of 28% Percoll in sucrose buffer (0.3 M sucrose, 10 mM 649 MOPS-KOH (pH 7.3), 0.1 % (w/v) BSA). The protein concentration was estimated by 650 measuring absorbance at 280 nm using a NanoDrop 1000 Spectrophotometer (Thermo 651 Scientific). 652 Gel electrophoresis 653 One-dimensional sodium dodecyl sulfate polyacrylamide gel electrophoresis (1D-SDS– 654 PAGE, 20 × 20 cm) performed according to (Laemmli 1970) was employed as a pre- 655 fractionation step. Isolated mitochondria (250 µg) and size marker proteins (protein 656 ladder, Thermo Scientific # 26630) were dissolved in sample buffer (NuPAGE®), 657 placed at 65°C for 15 min and resolved by 12% SDS-PAGE. After protein migration, 658 proteins were stained with colloidal Coomassie Brilliant Blue. Then, each gel lane was 659 sliced in 40 segments and diced in approximately 1 mm cubes followed by gel 660 destaining and in gel digestion as described by Balbuena et al. (2011). 661 Liquid chromatography-mass spectrometry 662 A two column system was used. Both columns were in-house packed with C18 reverse 663 phase material (ReproSil, C18 AQ 3 μm (Dr. Maisch, Germany)) in fused silica. The 664 pre-column had an inner diameter of 100 μm and a length of 2 cm. The analytical 665 column had an inner diameter of 75 μm and a length of 15 cm. The flow of 250 nl/min 666 was delivered by an EASY nLC system (Thermo Fisher Scientific, Odense, Denmark) 23 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 667 and the peptides were eluted using a 21 min gradient from 0% to 40% B buffer (A 668 buffer: 0.1% formic acid, B buffer: 0.1% formic acid/90% acetonitrile, all v/v). The 669 flow from the analytical column was coupled to an LTQ-Orbitrap XL mass 670 spectrometer (Thermo Fisher Scientific). The instrument was run in positive ion mode. 671 Full scan MS spectra (from m/z 300-1800) were acquired in the Orbitrap with resolution 672 r = 60,000 at m/z 400. The up to 5 most intense ions with charge states larger than +1 673 were sequentially isolated for fragmentation in the linear ion trap using collision- 674 induced dissociation. Former target ions selected for fragmentation were dynamically 675 excluded for 30 s. For accurate mass measurements the lock mass option was enabled in 676 MS mode and the polydimethylcyclosiloxane (PCM) ions generated in the electrospray 677 process (m/z = 445.120025) were used for internal recalibration in real time, resulting in 678 a mass accuracy of approximately 2 ppm. 679 Database searching and protein identification 680 Acquired MS/MS spectra were searched using different algorithms against the potato 681 protein database (http://www.potatogenome.net). To estimate the protein false discovery 682 rate (FDR), randomized sequences were combined with the forward database in a 683 concatenated format. MASCOT server 2.3.01, SEQUEST, MS-GFDB (Kim et al., 684 2010) and ProLuCID (Xu et al., 2006) were used to interpret the data set using similar 685 parameters. MASCOT and SEQUEST were integrated within the Proteome Discoverer 686 1.3 software package (Thermo Scientific). Raw files (Thermo) were converted to ms2 687 and mzXML files prior to analysis by ProLuCID and MS-GFDB. All search engines 688 were configured to the following parameters: MW range between 200 and 2,000; 1,000 689 ppm of precursor ion tolerance; fragment tolerance 1.0 Da oxidation of methionine and 690 asparagine deamidation as variable modifications and carbamidomethylation of cysteine 691 as a static modification; two allowed missed trypsin cleavages. After searches, the 692 PSMs were filtered with a 5 ppm precursor ion tolerance, achieving a false discovery 693 rate (FDR) lower than 1% at the protein level for each biological replicate. 694 All the MS data from each gel slice was searched and merged for each replicate. The 695 results from different programs were combined in order to complement observations 696 that came from different processes of data mining, without compromising the accuracy. 697 To do this, a high stringency on precursors (5 ppm) was applied. 24 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 698 Filtered data were uploaded to the Protein Herder module within the Compass package 699 (Wenger et al., 2011) for protein grouping. At this step, all sets of indistinguishable 700 proteins, which were identified by the same peptides, were combined into protein 701 groups. 702 Purity assessment of mitochondrial preparations 703 Total protein and mitochondrial protein-enriched fractions from potato tubers were 704 separated by SDS-PAGE. From each fraction, 20 μg of proteins was loaded on the gel. 705 Blotting was perfomed as previously described by Stevenson et al. (2009) to PVDF 706 membranes (GE Healthcare Ltd, UK). Antibodies directed against 14-3-3, enolase, 707 plastidic alpha carboxytransferase (α-CT) and mitochondrial pyruvate dehydrogenase 708 E1 component subunit alpha (PDE1-α) were used to assess mitochondrial protein 709 enrichment in the preparations. Immunodetection of proteins bound to PVDF 710 membranes was performed using the colorimetric alkaline phosphatase and peroxidase 711 substrate detection system (Sigma-Aldrich). 712 Subcellular Localization Study 713 Total RNA was isolated from potato tuber by using the method described by Kumar et 714 al., (2007). A cDNA library was then synthesized by M-MLV reverse transcriptase 715 reaction (Promega, WI, US) with random hexamer. Potato cDNA was PCR amplified 716 from cDNA library by a pair of gene-specific primers; BamHI and Not I restriction 717 enzyme cutting sites were added to forward and reverse primers, respectively (Table 718 S8). The PCR fragments were cloned into pGEM-T easy vector (Promega, WI, US), the 719 insertions were dropped off by BamHI and NotI double digestion, and the insertions 720 were cloned into pE6c entry vector, such that an EYFP fluorescence tag was fused in 721 frame with potato proteins at the C-terminus (Dubin et al., 2008). The cassette was then 722 moved into the binary vector pSITE-0B by Gateway LR clonase (Chakrabarty et al., 723 2007). 724 For the co-localization assay, mitochondria marker (CD3-986) binary plasmid with CFP 725 fluorescence tag was ordered from the Arabidopsis Biological Resource Center (Nelson 726 et al., 2007). The binary constructs were transformed into Agrobacterium AGL1 strain. 727 The potato EYFP fusion construct was individually mixed with the CFP mitochondrial 728 marker and co-transformed into tobacco leaves. Transformation was performed by 25 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 729 following the procedure described by Sparkes et al. (2006). An Olympus IX70 730 microscope controlled by MetaMorph (Version 6.3 v6, Molecular Device) software was 731 used for image capture, and Chroma filter 49001 and 49003 were applied for the CFP 732 and EYFP channel, respectively. 733 Post-translational modifications 734 Multiple post-translational modifications (PTM) were searched against a concatenated 735 database containing forward and randomized protein sequences previously identified in 736 the general proteome searching approach. The searches were performed using 737 SEQUEST, MASCOT and MS-GFDB programs. For each round of PTM searching a 738 maximum of two different variable modifications (PTMs) were configured. Precursor 739 ion tolerance was set to 100 ppm and 2 missed tryptic cleavages were allowed. 740 Carbamidomethylation of cysteine was configured as a static modification and the 741 PTMs of interest were configured as variable modification. After searches, the data set 742 was filtered using 5 ppm as precursor ion tolerance, achieving FDR lower than 1%. 743 Only PTMs detected in at least 2 replicates were considered. The spectra assigned to 744 peptides with oxidation sites were searched against the database allowing all types of 745 oxidation sites at the same time to make sure that the oxidation sites were reproducible 746 compared to the first database search. 747 Semi-tryptic database searching 748 All MS data was reanalyzed using MS-GFDB program against the identified protein 749 sequences database allowing semi-tryptic peptides and 100 ppm of precursor ion 750 tolerance. Carbamidomethylation of cysteine was configured as a static modification 751 and methionine oxidation as variable modification. The PSMs obtained were filtered out 752 using 5 ppm as precursor ion tolerance, resulting in a FDR lower than 1%. 753 Prediction of Subcellular localization 754 Predictions of subcellular localization for identified proteins were taken using the full- 755 length protein sequences and five programs namely TargetP version 1.1 756 (http://www.cbs.dtu.dk/services/TargetP/, Emanuelsson et al., 2000), Predotar version 757 1.03 (http://urgi.versailles.inra.fr/predotar/predotar.html, Small et al. 2004), MitoProtII 758 (http://ihg.gsf.de/ihg/mitoprot.html, Claros and Vincens, 1996), iPSORT 26 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 759 (http://ipsort.hgc.jp/, Bannai et al., 2002) and WoLF PSORT (http://wolfpsort.org/, 760 Horton et al., 2007). 761 Molecular weight and isoelectric point 762 Molecular weight and isoelectric point of proteins were obtained using Compute pI/Mw 763 web server (http://web.expasy.org/compute_pi/, Bjellqvist et al. 1994, Gasteiger et al. 764 2005). The grand average hydropathy (GRAVY) score was computed based on Kyte- 765 Doolittle hydropathy scale (Kyte and Doolittle 1982). Bioinformatics computations 766 were performed using Python ver. 2.7. Arabidopsis mitochondrial proteome list is based 767 on Supplementary Table 1 in Heazlewood et al. (2004). 768 Protein abundance calculation 769 Total number of PSM (peptide spectrum matched) for each protein identified was 770 extracted from Protein Herder module of Compass software (Wenger et al., 2011) for 771 each biological replicate. Only proteins identified in at least two replicates were 772 considered for quantification based on the distributed normalized spectral abundance 773 factor approach (dNSAF, Zhang et al., 2010). Spectral counts (SpC) for each protein or 774 group of proteins were used to estimate protein abundance. 775 For protein abundance estimation, the PSMs produced by each program were unified by 776 an in-house developed script. Only different scans for a given protein detected by 777 multiple programs were considered for calculation of total spectral counts, which means 778 that repeated scans were considered only once in the calculation of total spectral counts. 779 Therefore, the total spectra counts of each protein was calculated by the sum of PSMs 780 (non-redundant scans) produced by different algorithms. 781 The dNSAF considers the spectral counts of shared peptides and distributes them based 782 on a distribution factor (d): 783 dk = 784 where k denotes a protein identity and n is the total number of proteins, while uSpC 785 represents the sum of spectra matched to peptides mapping uniquely to the respective uSpCk ∑uSpCn 27 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 786 protein. So, for a given protein k, d is equal to the total uSpC mapped to this protein 787 divided by the total number of uSpC from all n proteins with protein k shares peptide(s). 788 Then, the dNSAF for a given protein, k, is calculated as follows: 789 dNSAFk = [(uSpC + d × sSpC ) ÷ L]k ∑ [(uSpC + d × sSpC) ÷ L]n n 790 where sSpC corresponds to the total number of spectra from peptides shared among 791 related proteins and L represents the length of the protein. In this way, the shared 792 spectral counts (sSpC) are distributed amongst the proteins, avoiding that the same 793 spectrum will be counted multiple times. 794 Annotation and functional classification 795 Protein annotation and classification were performed based on matching the protein 796 sequences against the non-redundant NCBI database and Gene Ontology functional 797 classification using the Blast2GO tool (Conesa et al., 2005). Protein sequences were 798 aligned using the BLASTP algorithm against nr NCBI database using the following 799 parameters: report a maximum of 5 blast hits, 0.1 for the expected value and minimum 800 high scoring segment pairs (HSPs) length equal to 33. Mapping and annotation steps 801 were performed using Blast2GO default values. GOslim annotation was performed 802 using the plant slim. GO distribution graphs related to biological process and molecular 803 function were generated and analyzed at the fourth level of depth. Also, each protein 804 was assigned to general functional categories according to Heazlewood et al. (2004). 805 806 Acknowledgements 807 JFH is grateful to Drs. Katherine Beard and Lee J. Sweetlove for a study visit to their 808 laboratory at University of Oxford. 809 Authorship 810 JFH, JJT and IMM designed the research, JFH, MC and JJT performed the experiments, 811 FS, JFH, JJT, RSPR, and IMM contributed new computational tools and analyzed the 812 data, and FS, JFH, ARW, ONJ, DRG, JJT and IMM wrote the paper. 28 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 813 814 Supplemental Data 815 Supplementary Discussion 816 Figure S1: SDS-PAGE and corresponding immunoblots of enriched mitochondrial 817 protein fraction and total protein extract from potato tubers.. A, Commassie blue-stained 818 SDS-PAGE gel loaded with 20 µg of protein per lane. Positions of protein standards are 819 shown on the left (in kilodaltons). B, protein blots of mitochondrial enriched fraction 820 and total protein extract (fractions in A). The molecular masses of the protein bands are 821 shown on the left (in kilodaltons) 822 Figure S2: Subcellular localization study on a selected subset of proteins identified. 823 See Figure 2 and Table S2 for details. 824 Figure S3:. GO term distribution related to biological process and molecular function of 825 the proteins identified in the potato tuber mitochondria. 826 Table S1: List of 1060 protein groups identified by GeLC-MS/MS in isolated 827 mitochondria from Solanum tuberosum tubers. 828 Table S2: List of selected candidate proteins for validation by subcellular localization in 829 tobacco leaf. 830 Table S3. Label-free quantification based on spectral counting (884 proteins). 831 Table S4. Mitochondrial localization prediction results (for potato and Arabidopsis 832 mitochondrial proteomes) with different combinations of prediction programs. 833 Table S5: Proteins with non-cleaveable N-terminal presequences identified by LC- 834 MS/MS. 835 Table S6: Monoisotopic mass changes used to identify the post-translational 836 modifications. 837 Table S7: Posttranslational modifications detected in the mitochondrial proteome of 838 Solanum tuberosum. 29 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 839 Table S8: List of forward and reverse primer sequences used in this study for gene 840 cloning. 841 Repository data 842 Raw files as well peptide and protein identification files exported for each search engine 843 (SEQUEST, MASCOT MSGF-DB and ProLuCID) have been deposited in a public 844 repository - the ProteomeXchange Consortium 845 (http://proteomecentral.proteomexchange.org) via the PRIDE partner repository 846 (Vizcaino et al. 2013) with the dataset identifier PXD000149 available at 847 http://www.ebi.ac.uk/pride/. Phosphopeptides and respective spectra data have been 848 deposited at P3DB (Plant Protein Phosphorylation Database) available at 849 http://p3db.org/PotatoMitochondriaPhosData.php. 850 851 852 LITERATURE CITED 853 Alban C (2011) Biotin (vitamin B8) synthesis in plants. Adv Bot Res. 59: 39-66 854 Balbuena TS, Salas JJ, Martínez-Force E, Garcés R, Thelen JJ (2011) Proteome analysis 855 of cold acclimation in sunflower. J Proteome Res. 10(5): 2330-2346 856 Balbuena TS, He R, Salvato F, Gang DR,Thelen JJ (2012) Large-scale proteome 857 comparative analysis of developing rhizomes of the ancient vascular plant equisetum 858 hyemale. Front Plant Sci. 3: 131 859 Balk J, Pilon M (2011) Ancient and essential: the assembly of iron-sulfur clusters in 860 plants. Trends Plant Sci. 16: 218-226Bannai H, Tamada Y, Maruyama O, Nakai K, 861 Miyano S (2002) Extensive feature detection of N-terminal protein sorting signals. 862 Bioinformatics. 18: 298–305 863 Bartoli CG, Pastori GM, Foyer CH (2000) Ascorbate Biosynthesis in Mitochondria Is 864 Linked to the Electron Transport Chain between Complexes III and IV. Plant Physiol 865 123: 335-343 30 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 866 Bender A, Hajieva P, Moosmann B (2008) Adaptive antioxidant methionine 867 accumulation in respiratory chain complexes explains the use of a deviant genetic code 868 in mitochondria. Proc. Natl. Acad. Sci. USA 105: 16496–16501 869 Bernard DG, Chen Y, Zhao Y, Balk J (2009) An allelic mutant series of ATM3 reveals 870 its key role in the biogenesis of cytosolic iron-sulfur proteins in Arabidopsis. Plant 871 Physiol 151: 590-602 872 Bjellqvist B, Basse B, Olsen E, Celis JE (1994) Reference points for comparisons of 873 two-dimensional maps of proteins from different human cell types defined in a pH scale 874 where isoelectric points correlate with polypeptide compositions. Electrophoresis 15: 875 529–539 876 Blancquaert D, Storozhenko S, Loizeau K, De Steur H, De Brouwer V, Viaene J, 877 Ravanel S, Rébeillé F, Lambert W, Van Der Straeten D (2010) Folates and Folic 878 Acid:rom Fundamental Research Toward Sustainable Health, Critical Reviews in Plant 879 Sciences 29:1, 14-35. 880 Bonen L, Calixte S (2006) Comparative analysis of bacterial-origin genes for plant 881 mitochondrial ribosomal proteins. Mol Biol Evol 23(3): 701-712 882 Bykova NV, Møller IM (2001) Involvement of matrix NADP turnover in the oxidation 883 of NAD+-linked substrates by pea leaf mitochondria. Physiol Plant 111: 448-456 884 Bykova NV, Egsgaard H, Møller IM (2003) Identification of 14 new phosphoproteins 885 involved in important plant mitochondrial processes. FEBS Lett 540: 141-146 886 Calvo SE, Mootha VK (2010) The mitochondrial protome and human disease. Annu 887 Rev Genomics Hum Genet 22: 25-44 31 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 888 Camacho-Pereira J, Meyer LE, Machado LB, Oliveira MF, Galina A (2009) Reactive 889 oxygen species production by potato tuber mitochondria is modulated by 890 mitochondrially bound hexokinase activity.Plant Physiol 149(2): 1099-1110 891 Chai, M.F., Wei, P.C., Chen, Q.J., An, R., Chen,J., Yang, S.H., and Xue-Chen Wang, 892 X.C. (2006). NADK3, a novel cytoplasmic source of NADPH, is required under 893 conditions of oxidative stress and modulates abscisic acid responses in Arabidopsis. 894 Plant Journal 47: 665-674. 895 Chakrabarty, R., Banerjee, R., Chung, S.M., Farman, M., Citovsky, V., Hogenhout, S. A., 896 Tzfira, T., and Goodin, M. (2007). pSITE vectors for stable integration or transient expression 897 of autofluorescent protein fusions in plants: Probing Nicotiana benthamiana-viruse interactions. 898 MPMI 20: 740-750 899 Carrari, F., Coll-Garcia, D., Schauer, N., Lytovchenko, A., Palacios-Rojas, N., Balbo, I., Rosso, 900 M., and Fernie, A.R. (2005). Deficiency of a plastidial adenylate kinase in arabidopsis results in 901 elevated photosynthetic amino acid biosynthesis and enhanced growth. Plant Physiology 137: 902 70–82 903 Chaves I, Pinheiro C, Paiva JAP, Planchon S, Sergeant K, Renaut J, Graca G, Coelho 904 AV, Ricardo CPP (2009) Proteomic evaluation of wound-healing processes in potato 905 (Solanum tuberosum L.) tuber tissue. Proteomics 9: 4154–4175 906 Chew O, Whelan J, Millar AH (2003) Molecular definition of the ascorbate-glutathione 907 cycle in Arabidopsis mitochondria reveals dual targeting of antioxidant defenses in 908 plants. J Biol Chem 278: 46869–46877 909 Claros MG, Vincens P (1996) Computational method to predict mitochondrially 910 imported proteins and their targeting sequences. European Journal of Biochemistry 241: 911 779–786 32 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 912 Conesa A, Götz S, García-Gómez JM, Terol J, Talón M, Robles M (2005) Blast2GO: a 913 universal tool for annotation, visualization and analysis in functional genomics research. 914 Bioinformatics. 21(18):3674-3676. 915 Considine MJ, Goodman M, Echtay KS, Laloi M, Whelan J, Brand MD, Sweetlove LJ 916 (2003) Superoxide stimulates a proton leak in potato mitochondria that is related to the 917 activity of uncoupling protein. J Biol Chem 278: 22298–22302 918 Cui JA, Liu JH, Li YH, Shi TL (2011) Integrative Identification of Arabidopsis 919 Mitochondrial Proteome and Its Function Exploitation through Protein Interaction 920 Network. PLoS ONE 6(1): e16022 921 Danpure CJ (1997) Variable peroxisomal and mitochondrial targeting of alanine: 922 glyoxylate aminotransferase in mammalian evolution and disease. BioEssays 19: 317– 923 326 924 Day DA, Hanson JB (1977) Effect of Phosphate and Uncouplers on Substrate Transport 925 and Oxidation by Isolated Corn Mitochondria. Plant Physiol 59: 139-144 926 Demartini DR, Jain R, Agrawal GK, Thelen JJ (2011) Proteomic comparison of plastids 927 from developing embryos and leaves of Brassica napus. J Prot Res 10: 2226-2237 928 Dubin, M.J., Bowler, C., and Benvenuto, G. (2008). A modified gateway cloning 929 strategy for overexpression tagged proteins in plants. Plant Methods 4: 1-11. 930 Duncan O, Taylor NL, Carrie C, Eubel H, Kubiszewski-Jakubiak S, Zhang B, Narsai 931 R, Millar AH, Whelan J (2011) Multiple lines of evidence localize signaling, 932 morphology, and lipid biosynthesis machinery to the mitochondrial outer membrane of 933 Arabidopsis. Plant Physiol. 157 (3):1093-1113 33 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 934 Emanuelsson O, Nielsen H, Brunak S, von Heijne G (2000) Predicting subcellular 935 localization of proteins based on their N-terminal amino acid sequence. J Mol Biol. 936 300(4): 1005-1016 937 Fernie AR, Carrari F, Sweetlove LJ (2004) Respiratory metabolism: glycolysis, the 938 TCA cycle and mitochondrial electron transport. Current Opinion in Plant Biology 7: 939 254–261 940 Finkemeier I, Laxa M, Miguet L, Howden AJM, Sweetlove LJ (2011) Proteins of 941 diverse function and subcellular location are lysine acetylated in Arabidopsis. Plant 942 Physiol 155: 1779-1790 943 Foyer CH, Noctor G (2003) Redox sensing and signalling associated with reactive 944 oxygen in chloroplasts, peroxisomes and mitochondria. Physiol Plant 119: 355–364 945 Fujii S, Small I (2011) The evolution of RNA editing and pentatricopeptide repeat 946 genes. New Phytol 191: 37-47 947 Gasteiger E, Hoogland C, Gattiker A, Duvaud S, Wilkins MR, Appel RD, Bairoch A 948 (2005) Protein identification and analysis tools on the ExPASy server; (In) John MW 949 (ed): The Proteomics Protocols Handbook, Humana Press. 950 Giglione C, Meinnel T (2001) Organellar peptide deformylases: universality of the N- 951 terminal methionine cleavage mechanism. Trends Plant Sci. 6(12): 566-572 952 Graham JWA, Williams TCR, Morgan M, Fernie AR, Ratcliffe RG, Sweetlove LJ 953 (2007) Glycolytic enzymes associate dynamically with mitochondria in response to 954 respiratory demand and support substrate channeling. Plant Cell 19: 3723-3738 955 Guan KL, Xiong Y (2011) Regulation of intermediary metabolism by protein 956 acetylation. Trends Biochem. Sci 36(2): 108-116 34 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 957 Gustavsson N, Kokke BP, Härndahl U, Silow M, Bechtold U, Poghosyan Z, Murphy D, 958 Boelens WC, Sundby C (2002) A peptide methionine sulfoxide reductase highly 959 expressed in photosynthetic tissue in Arabidopsis thaliana can protect the chaperone- 960 like activity of a chloroplast-localized small heat shock protein. Plant J 29: 545–553 961 Hardin SC, Larue CT, Oh MH, Jain V, Huber SC (2009) Coupling oxidative signals to 962 protein phosphorylation via methionine oxidation in Arabidopsis. Biochem. J 422: 305- 963 312 964 Havelund JF, Thelen JJ, Møller IM (2013) Biochemistry and proteomics of 965 mitochondria from non-photosynthetic tissues. Frontiers Plant Sci 4: 51 966 Heazlewood JL, Tonti-Filippini JS, Gout AM, Day DA, Whelan J, Millar AH (2004) 967 Experimental analysis of the Arabidopsis mitochondrial proteome highlights signaling 968 and regulatory components, provides assessment of targeting prediction programs, and 969 indicates plant-specific mitochondrial proteins. Plant Cell 16: 241–256 970 Horton P, Park KJ, Obayashi T, Fujita N, Harada H, Adams-Collier CJ, Nakai K (2007) 971 WoLF PSORT: protein localization predictor. Nucleic Acids Research 35: W585–587 972 Huang Y, Houston NL, Tovar-Mendez A, Stevenson SE, Miernyk JA, Randall DD, 973 Thelen JJ (2010) A quantitative mass spectrometry-based approach for identifying 974 protein kinase –clients and quantifying kinase activity. Anal. Biochem. 402: 69-76 975 Hughes JB, Hellmann JJ, Ricketts TH, Bohannan BJ (2001) Counting the uncountable: 976 statistical approaches to estimating microbial diversity. Appl Environ Microbiol 67: 977 4399–4406 978 Ilangovan G, Venkatakrishnan CD, Bratasz A, Osinbowale S, Cardounel AJ, Zweier JL, 979 Kuppusamy P (2006) Heat shock-induced attenuation of hydroxyl radical generation 980 and mitochondrial aconitase activity in cardiac H9c2 cells. Am J Physiol Cell Physiol 981 290: C313–324 35 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 982 Islam MS, Studer B, Byrne SL, Panitz F, Farrell JD, Bendixen, C, Møller IM, Asp T 983 (2013) The genome and transcriptome of perennial ryegrass mitochondria. BMC 984 Genomics 14: 202 985 Jimenez A, Hernandez JA, del Rio LA, Sevilla F (1997) Evidence for the presence of 986 the ascorbate-glutathione cycle in mitochondria and peroxisomes of pea leaves. Plant 987 Physiol 114: 275–284 988 Junker BH, Lonien J, Heady LE, Rogers A, Schwender J (2007) Parallel determination 989 of enzyme activities and in vivo fluxes in Brassica napus embryos grown on organic or 990 inorganic nitrogen source. Phytochemistry 68: 2232-2242 991 Kell DB (2009) Iron behaving badly: inappropriate iron chelation as a major contributor 992 to the aetiology of vascular and other progressive inflammatory and degenerative 993 diseases. BMC Med Genom 2: 2 994 Kell DB (2010) Towards a unifying, systems biology understanding of large-scale 995 cellular death and destruction caused by poorly liganded iron: Parkinson's, Huntington's, 996 Alzheimer's, prions, bactericides, chemical toxicology and others as examples. Arch 997 Toxicol 84: 825-889 998 Kellems RE, Allison VF, Butow RA (1974) Cytoplasmic type 80S ribosomes associated 999 with yeast mitochondria. J Biol Chem 249, 3297-3303. 1000 Keyhani E (1973) Ribosomal granules associated with outer mitochondrial membranes 1001 in aerobic yeast cells. J Cell Biol 58, 480-484 1002 Kim S, Mischerikow N, Bandeira N, Navarro, JD, Wich L, Mohammed S, Heck AJR, 1003 Pevzner PA (2010) The Generating Function of CID, ETD and CID/ETD Pairs of 1004 Tandem Mass Spectra: Applications to Database Search. Molecular & Cellular 1005 Proteomics 9: 2840-2852 36 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 1006 Kleffmann, T., Russenberger, D., von Zychlinski, A., Christopher, W., Sjolander, K., 1007 Gruissem, W., and Baginsky, S. (2004). The Arabidopsis thaliana chloroplast proteome 1008 reveals pathway abundance and novel protein functions. Curr. Biol. 14: 354–362. 1009 Klodmann J, Braun H-P (2011) Proteomic approach to characterize mitochondrial 1010 complex I from plants. Phytochemistry 72: 1071-1080 1011 Klodmann J, Senkler M, Rode C, Braun H-P (2011) Defining the protein complex 1012 proteome of plant mitochondria. Plant Physiol 157: 587-598 1013 Koc EC, Koc H (2012) Regulation of mammalian mitochondrial translation by post- 1014 translational modifications. Biochim. Biophys. Acta 1819(9-10): 1055-1066 1015 Kristensen BK, Askerlund P, Bykova NV, Egsgaard H, Møller IM (2004) Identification 1016 of oxidised proteins in the matrix of rice leaf mitochondria by immunoprecipitation and 1017 two-dimensional liquid chromatography-tandem mass spectrometry. Phytochemistry 1018 65: 1839-1851 1019 Kyte J, Doolittle RF (1982) A simple method for displaying the hydropathic character 1020 of a protein. J Mol Biol 157:105–132 1021 Kumar, G.N. M., Iyer, S., and Knowles, N.R. (2007). Extraction of RNA from fresh, 1022 frozen, and lyophilized tuber and root tissue. J. Agric. Food Chem. 55: 1674-1678. 1023 Laemmli UK (1970) Cleavage of Structural Proteins during Assembly of Head of 1024 Bacteriophage-T4. Nature 227(5259): 680-685 1025 Lee CP, Eubel H, Solheim C, Millar AH (2012) Mitochondrial proteome heterogeneity 1026 between tissues from the vegetative and reproductive stages of Arabidopsis thaliana 1027 development. J. Proteome Res 11(6): 3326–3343 37 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 1028 Li J, Cai T, Wu P, Cui Z, Chen X, Hou J, Sie Z, Xue P, Shi L, Liu P, Yates JR,Yang F 1029 (2009) Proteomic analysis of mitochondria from Caenorhanbditis elegans. Proteomics 9: 1030 4539-4553 1031 Liepman AH, Olsen LJ (2001) Peroxisomal alanine:glyoxylate aminotransferase 1032 (AGT1) is a photorespiratory enzyme with multiple substrates in Arabidopsis thaliana. 1033 Plant J 25: 487-498 1034 Lill R, Kispal G (2001) Mitochondrial ABC transporters. Res Microbiol 152: 331-340 1035 Lister R, Carrie C, Duncan O, Ho LHM, Howell KA, Murcha MW, Whelan J (2007) 1036 Functional Definition of Outer Membrane Proteins Involved in Preprotein Import into 1037 Mitochondria. Plant Cell 19(11): 3739–3759 1038 Liu H, Sadygov RG,Yates JR III (2004 ) A model for random sampling and estimation 1039 of relative protein abundance in shotgun proteomics. Anal Chem 76: 4193– 4201. 1040 Löffler M, Fairbanks LD, Zameitat E, Marinaki AM, Simmonds HA (2005) Pyrimidine 1041 pathways in health and disease. Trends Mol Med 11: 430-437 1042 Maisonneuve E, Ducret A, Khoueiry P, Lignon S, Longhi S, Talla E, Ducan S (2009) 1043 Rules governing selective protein carbonylation. PLoS ONE 4(10):e7269 1044 Manning MC, Chou DK, Murphy BM, Payne RW, Katayama DS (2010) Stability of 1045 protein pharmaceuticals: An update. Pharmaceutical Res 27: 544-575 1046 Matros A, Kaspar S, Witzel K, Mock HP (2011) Recent progress in liquid 1047 chromatography-basedseparation and label-free quantitative plant proteomics. 1048 Phytochemistry 72(10): 963-974 1049 Maxwell DP, Wang Y, McIntosh L (1999) The alternative oxidase lowers mitochondrial 1050 reactive oxygen production in plant cells. Proc Natl Acad Sci USA 96: 8271–8276 38 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 1051 Melo AMP, Roberts TH, Møller IM (1996) Evidence for the presence of two rotenone- 1052 insensitive NAD(P)H dehydrogenases on the inner surface of the inner membrane of 1053 potato tuber mitochondria. Biochim Biophys Acta 1276: 133-139 1054 Michalecka AM, Svensson ÅS, Johansson FI, Agius SC, Johanson U, Brennicke A, 1055 Binder S, Rasmusson AG (2003) Arabidopsis genes encoding mitochondrial type II 1056 NAD(P)H dehydrogenases have different evolutionary origin and show distinct 1057 responses to light. Plant Physiol. 133: 642–652 1058 Milenkovic D, Kozjak V, Wiedemann N, Lohaus C, Meyer HE, Guiard B, Pfanner N, 1059 Meisinger, C (2004) Sam35 of the mitochondrial protein sorting and assembly 1060 machinery is a peripheral outer membrane protein essential for cell viability. J Biol 1061 Chem 279(21): 22781-22785 1062 Millar AH, Heazlewood JL (2003) Genomic and proteomic analysis of mitochondrial 1063 carrier proteins in Arabidopsis. Plant Physiology 131: 443-453 1064 Millar AH, Heazlewood JL, Kristensen BK, Braun, HP, Møller IM (2005) The plant 1065 mitochondrial proteome. Trends Plant Sci 10: 36-43 1066 Millar AH, Whelan J, Small I (2006) Recent surprises in protein targeting to 1067 mitochondria and plastids. Current Opinion in Plant Biology 9: 610-615 1068 Millar AH, Whelan J, Soole KL, Day DA (2011) Organization and regulation of 1069 mitochondrial respiration in plants. Annu Rev Plant Biol 62: 79-104 1070 Miller G, Honig A, Stein H, Suzuki N, Mittler R, Zilberstein A (2009) Unraveling 1- 1071 Pyrroline-5-Carboxylate-Proline Cycle in Plants by Uncoupled Expression of Proline 1072 Oxidation Enzymes. J Biol Chem 284(39): 26482-28492 1073 Møller IM, Rasmusson AG (1998) The role of NADP in the mitochondrial matrix. 1074 Trends Plant Sci. 3: 21-27 39 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 1075 Møller IM (2001) Plant mitochondria and oxidative stress. Electron transport, NADPH 1076 turnover and metabolism of reactive oxygen species. Annu Rev Plant Physiol Plant Mol 1077 Biol 52: 561-591 1078 Møller IM, Kristensen BK (2006) Protein oxidation in plant mitochondria detected as 1079 oxidized tryptophan. Free Radical Biology & Medicine 40: 430-435 1080 Møller IM (2007) Mitochondrial electron transport and oxidative stress. In Plant 1081 Mitochondria (D.M. Logan, ed.), pp. 185-211. Annual Plant Reviews, Blackwell 1082 Publishing. 1083 Møller IM, Rogowska-Wrzesinska A, Rao RSP (2011) Protein carbonylation and metal- 1084 catalyzed oxidation in a cellular perspective. J. Proteomics 74: 2228-2242 1085 Nelson, B.K., Cai, X., and Nebenführ, A. (2007). A multicolored set of in vivo organelle 1086 markers for co-localization studies in Arabidopsis and other plants. Plant J. 51: 1126- 1087 1136. 1088 Neuburger M, Journet E-P, Bligny R, Carde J-P, Douce R (1982) Purification of plant 1089 mitochondria by isopycnic centrifugation in density gradients of Percoll. Arch Biochem 1090 Biophys 217: 312-323 1091 Neuburger M, Day DA, Douce R (1985) Transport of NAD+ in Percoll-purified potato- 1092 tuber mitochondria - inhibition of NAD+ influx and efflux by n-4-azido-2-nitrophenyl- 1093 4-aminobutyryl-3'-NAD+. Plant Physiol 78: 405-410 1094 Pagliarini DJ, Calvo SE, Chang B, Sheth SA, Vafai SB, Ong S-E, Walford GA, Sugiana 1095 C, Boneh A, Chen WK, Hill DE, Vidal M, Evans JG, Thorburn DR, Carr SA, 1096 MoothaVK (2008) A mitochondrial protein compendium elucidates complex I disease 1097 biology. Cell 134: 112-123 40 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 1098 Palmieri F, Pierri CL, De Grassi A, Nunes-Nesi A, Fernie AR (2011) Evolution, 1099 structure and function of mitochondrial carriers: a review with new insights. Plant J 66: 1100 161-181 1101 Paschen SA, Neupert W, Rapaport D (2005). Biogenesis of beta-barrel membrane 1102 proteins of mitochondria. Trends Biochem Sci. 30(10):575-582 1103 Pastore D, Stoppelli MC, Di Fonzo N, Passarella S (1999) The existence of the K+ 1104 channel in plant mitochondria. J Biol Chem 274: 26683-26690 1105 Pineau B, Layoune O, Danon A, De Paepe R (2008) L-galactono-1,4-lactone 1106 dehydrogenase is required for the accumulation of plant respiratory complex I. J Biol 1107 Chem 283: 32500–32505 1108 Polidoros AN, Mylona PV, Arnholdt-Schmitt B (2009) Aox gene structure, transcript 1109 variation and expression in plants. Physiol Plant 137: 342–353 1110 Premsle RT, Zahedi RP, Levandrowski U, Sickmann A (2009) Recent advances in yeast 1111 organelle and membrane proteomics. Proteomics 9: 4731-4743 1112 Radi R, Turrens JF, Chang LY, Bush KM, Crapo JD, Freeman BA (1991) Detection of 1113 catalase in rat heart mitochondria. J. Biol. Chem. 266: 22028–22034 1114 Rao RS, Møller IM (2011) Pattern of occurrence and occupancy of carbonylation sites 1115 in proteins. Proteomics. 11(21):4166-4173 1116 Rasmusson AG, Møller IM (1990) NADP-Utilizing enzymes in the matrix of plant 1117 mitochondria. Plant Physiol 94:1012-1018 1118 Rasmusson AG, Møller IM (2011) Mitochondrial electron transport and plant stress. 1119 Chapter 14, pp. 357-381, In Plant Mitochondria (F. Kempken, Ed.), Springer 41 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 1120 Roberts TH, Fredlund KM, Møller IM (1995) Direct evidence for the presence of two 1121 external NAD(P)H dehydrogenases coupled to the electron transport chain in plant 1122 mitochondria. FEBS Lett. 373: 307-309 1123 Roberts J.K.M., Aubert S., Gout E., Bligny R., Douce R. (1997) Cooperation and 1124 competition between adenylate kinase, nucleoside diphosphokinase, electron transport 1125 and ATP synthase in plant mitochondria studied by P-31-nuclear magnetic resonance. 1126 Plant Physiol 113: 191–199. 1127 Schneider G, Fechner U (2004) Advances in the prediction of protein targeting signals. 1128 Proteomics 4(6): 1571-1580 1129 Schulze, S., Mant, A., Kossmann, J., Lloyd, J.M. (2004). Identification of an 1130 Arabidopsis inorganic pyrophosphatase capable of being imported into chloroplasts. 1131 FEBS Letters 565: 101-105. 1132 Schweighofer A, Kazanaviciute V, Scheikl E, Teige M, Doczi R, Hirt H, Schwanninger 1133 M, Kant M, Schuurink R, Mauch F, Buchala A, Cardinale F, Meskiene I (2007) The 1134 PP2C-Type Phosphatase AP2C1, which negatively regulates MPK4 and MPK6, 1135 modulates innate immunity, jasmonic acid, and ethylene levels in Arabidopsis. Plant 1136 Cell 19: 2213-2224 1137 Schwender J, Shachar-Hill Y, Ohlrogge JB (2006) Mitochondrial metabolism in 1138 developing embryos of Brassica napus. J Biol Chem 281(45): 34040-34047 1139 Shintani DK, Ohlrogge JB (1994) The characterization of a mitochondrial acyl carrier 1140 protein isoform isolated from Arabidopsis thaliana. Plant Physiol.104(4): 1221-1229 1141 Small I, Peeters N (2000) The PPR motif – a TPR-related motif prevalent in plant 1142 organellar proteins.Trends Biochem. Sci. 25: 46-47 42 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 1143 Small I, Peeters N, Legeai F, Lurin C (2004) Predotar: A tool for rapidly screening 1144 proteomes for N-terminal targeting sequences. Proteomics 4: 1581–1590 1145 Sparkes, I.A., Runions, J. Kearns, A., and Hawes, C. (2006). Rapid, transient expression 1146 of fluorescent fusion proteins in tobacco plants and generation of stably transformed 1147 plants. Nature Protocols 1: 2019-2025. 1148 Stadtman ER, Van Remmen H, Richardson A, Wehr NB, Levine RL (2005) Methionine 1149 oxidation and aging. Biochim Biophys Acta, Proteins Proteomics, 1703:135-140 1150 Stevenson SE, Chu Y, Ozias-Akins P, Thelen JJ (2009) Validation of gel-free, label-free 1151 quantitative proteomics approaches: applications for seed allergen profiling. J 1152 Proteomics 72(3): 555-566 1153 Struglics A, Fredlund KM, Rasmusson AG, Møller IM (1993) The presence of a short 1154 redox chain in the membrane of potato tuber peroxisomes and the association of malate 1155 dehydrogenase with the membrane. Physiol Plant 88: 19-28 1156 Szecowka M, Osorio S, Obata T, Araújo WL, Rohrmann J, Nunes-Nesi A, Fernie AR 1157 (2012) Decreasing the mitochondrial synthesis of malate in potato tubers does not affect 1158 plastidial starch synthesis, suggesting that the physiological regulation of ADP glucose 1159 pyrophosphorylase is context dependent. Plant Physiolology 160(4): 2227-38. 1160 Sweetlove LJ, Beard KF, Nunes-Nesi A, Fernie AR, Ratcliffe RG (2010) Not just a 1161 circle: flux modes in the plant TCA cycle. Trends Plant Sci 15(8): 462-470 1162 Tan Y-F, O’Toole N, Taylor NL, Millar AH (2010) Divalent metal ions in plant 1163 mitochondria and their role in interactions with proteins and oxidative stress-induced 1164 damage to respiratory function. Plant Physiol 152: 747-761 43 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 1165 Tan YF, Millar AH, Taylor NL (2012) Components of mitochondrial oxidative 1166 phosphorylation vary in abundance following exposure to cold and chemical stresses. J 1167 Proteome Res 11(7): 3860-3879 1168 Taylor NL, Howell KA, Heazlewood JL, Tan TYW, Narsai R, Huang S, Whelan J, 1169 Millar AH (2010) Analysis of the rice mitochondrial carrier family reveals anaerobic 1170 accumulation of a basic amino acid carrier involved in arginine metabolism during seed 1171 germination. Plant Physiology 154:691-704 1172 Thelen JJ, Miernyk JA (2012) The future of proteomics: where mass spectrometry 1173 should be taking us. Biochem J 444: 169-181 1174 Tomb JF,White O, Kerlavage AR, Clayton RA, Sutton GG, Fleis-chmann RD et al 1175 (1997) The complete genome sequence of the gastric pathogen Helicobacterpylori. 1176 Nature 388: 539–547 1177 Umezawa T, Sugiyama N, Mizoguchi M, Hayashi S, Myouga F, Yamaguchi-Shinozaki 1178 K, Ishihama Y, Hirayama T, Shinozaki K (2009) Type 2C protein phosphatases directly 1179 regulate abscisic acid-activated protein kinases in Arabidopsis. Proc Natl Acad Sci U S 1180 A 106 (41):17588-17593 1181 Vizcaino JA, Côté RG, Csordas A, Dianes JA, Fabregat A, Foster JM, Griss J, Alpi 1182 E, Birim M, Contell J, O'Kelly G, Schoenegger A, Ovelleiro D, Pérez-Riverol Y, 1183 Reisinger F, Ríos D, Wang R, Hermjakob H (2013) The Proteomics Identifications 1184 (PRIDE) database and associated tools: status in 2013. Nucleic Acids Res. 41(D1): 1185 D1063-1069 1186 von Ballmoos C, Wiedenmann A, Dimroth P (2009) Essentials for ATP Synthesis by 1187 F1F0 ATP Synthases. Annu. Rev Biochem 78: 649-672 1188 Wada H, Shintani D, Ohlrogge J (1997) Why do mitochondria synthesize fatty acids? 1189 Evidence for involvement in lipoic acid production. Proc Natl Acad Sci USA 94(4): 1190 1591–1596 44 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 1191 Waller JC, Dhanoa PK, Schumann U, Robert T, Mullen RT, Snedden WA (2010) 1192 Subcellular and tissue localization of NAD kinases from Arabidopsis: 1193 compartmentalization of de novo NADP biosynthesis. Planta 231: 305-317 1194 Weber APM (2004) Solute transporters as connecting elements between cytosol and 1195 plastid stroma. Curr Opin Plant Biol 7: 247-253. 1196 Wenger CD, Phanstiel DH, Lee MV, Bailey DJ, Coon JJ (2011) COMPASS: a suite of 1197 pre- and post-search proteomics software tools for OMSSA. Proteomics 11(6):1064- 1198 1674 1199 Wittmann HG (1982) Components of bacterial ribosomes. Annual Review of 1200 Biochemistry 51: 155–183 1201 Xu T, Venable JD, Park SK, Cociorva D, Lu B, Liao L, Wohlschlegel J, Hewel J,Yates 1202 JR (2006) ProLuCID, a fast and sensitive tandem mass spectra-based protein 1203 identification program. Mol Cell Proteomics. 5: S174 1204 Xu X, et al. (Potato Genome Sequencing Consortium) (2011). Genome sequence and 1205 analysis of the tuber crop potato. Nature 475: 189-195 1206 Ye H, Rouault TA (2010) Human Iron-Sulfur Cluster Assembly, Cellular Iron 1207 Homeostasis, and Disease. Biochemistry 49: 4945-4956 1208 Zanin MKB, Donohue JM, Everitt BA (2010) Evidence that core histone H3 is targeted 1209 to the mitochondria in Brassica oleracea. Cell Biol. Int. 34, 997–1003 1210 Zhang Y, Wen Z, Washburn MP, Florens L (2010) Refinements to label free proteome 1211 quantitation: how to deal with peptides shared by multiple proteins. Anal Chem 82: 1212 2272–2281 45 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 1213 Zhu W, Schlueter SD, Brendel V (2003) Refined annotation of the Arabidopsis genome 1214 by complete expressed sequence tag mapping. Plant Physiol 132: 469–484 1215 46 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 1216 FIGURE LEGENDS 1217 Figure 1: Workflow chart for the isolation of mitochondria from potato tubers 1218 followed by protein separation, identification and spectral counting-based 1219 quantification. First, mitochondria were isolated from potato tubers using Percoll 1220 gradients. Then, mitochondrial proteins were extracted and fractionated by SDS- 1221 PAGE. After gel segmentation, in gel trypsin digestion was performed and peptides 1222 were injected into an LC system coupled to a mass spectrometer. Database searching 1223 was carried out using four different algorithms (MSGF-DB, ProLuCID, Mascot, 1224 Sequest) and the potato genome database (http://www.potatogenome.net). The number 1225 of assigned MS/MS spectra for each identified protein was determined and the 1226 distributed normalized spectral abundance factor was calculated for protein abundance 1227 estimation. 1228 Figure 2: The number of peptides (A) and proteins (B) identified using different 1229 database search engines with a <1% protein FDR identification. The total number of 1230 non-redundant peptides and proteins were 7346 and 1060, respectively. 1231 Figure 3: Subcellular localization study on a selected subset of proteins identified in 1232 this study. Twenty proteins that were identified in purified potato tuber mitochondria 1233 (Table S1) were in-frame fused with EYFP fluorescence tag, then transiently co- 1234 expressed with the mitochondrial marker CD3-986 in tobacco epidermal cells. The 1235 left lane shows the EYFP fluorescence of six out of the 19 constructs where 1236 expression was detected. The center lane shows the fluorescence of the mitochondrial 1237 marker. The right lane shows an overlay between the two types of fluorescence and 1238 green colour indicates coincidence of the two probes, i.e., mitochondrial localization. 1239 The remaining 13constructs are shown in Figure S2. Size bar, 12.5 μm. 1240 Figure 4: Properties of the potato mitochondrial proteome (n = 1060) as compared to the 1241 Arabidopsis mitochondrial proteome (n = 416; [Heazlewood et al. 2004]). (A) 1242 Distribution of molecular mass (kDa), (B) isoelectric point and (C) Grand Average 1243 Hydropathy (GRAVY) score. A disproportionately large number of low-molecular- 1244 weight proteins with negative GRAVY scores were identified in the potato 1245 mitochondrial proteome. 47 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 1246 Figure 5: Abundance distribution of the potato mitochondrial proteome. Mitochondrial 1247 localization prediction results are superimposed on the protein abundance (spectral 1248 counting) [log10(dNSAF)]. Plus signs (+) indicate mitochondrial localization by as many 1249 as five prediction programs and minus sign (-) indicates localization by none. Each 1250 category on the x-axis corresponds to 0.25 order of magnitude difference in the 1251 abundance of proteins. The difference between the most abundant and the least 1252 abundant proteins was 1778 fold. The abundance calculation for all proteins is found in 1253 Table S3. 1254 Figure 6: Schematic representation showing proteins that were identified by GeLC-MS- 1255 MS in the mitochondrial preparation of potato tubers. Squares represent protein 1256 identification and abundance based on spectral counting. Each color represents a range 1257 of Log10 transformed dNSAF (bottom left). Numbers in parentheses represent number 1258 of proteins found within the same range of abundance. Respiratory chain complexes 1259 (OXPHOS) are shown in detail in Figure 6. Black text: metabolic intermediates; grey 1260 text: enzymes or proteins. Abbreviations: ABC, ABC transporter; AC, aconitase; 1261 ACOT, acyl-coenzyme A thioesterase 9; ACS, Long chain acyl-CoA synthetase 9APX, 1262 ascorbate peroxidase; AOX, alternative oxidase; ASC; ascorbate; AlaAT, alanine 1263 aminotransferase; AspAT, aspartate aminotransferase; BCKDH, branched chain alpha- 1264 keto acid dehydrogenase; CAC, carnitine/acylcarnitine carrier protein; CI-CV, 1265 respiratory chain complexes (see Figure 7); CoA DHase, isovaleryl-CoA 1266 dehydrogenase; CS, citrate synthase; DiT1, oxoglutarate malate transporter; DHA, 1267 dehydroascorbate; DHAR, dehydroascorbate reductase; DIC, Mitochondrial 1268 dicarboxylate carrier protein; D-LDH, D-lactate dehydrogenase; E1α, pyruvate 1269 dehydrogenase subunit α; E1β, pyruvate dehydrogenase subunit β; E2, 1270 dihydrolipoamide S-acetyltransferase; E3, Dihydrolipoamide dehydrogenase; ECR, 1271 trans-2-enoyl-CoA reductase; EH, Enoyl-CoA hydratase; ER, enoyl-acyl-carrier-protein 1272 reductase; ETF, electron transfer flavoprotein; FCL, 5-formyltetrahydrofolate 1273 cycloligase; FDF, formyltetrahydrofolate deformylase-like; FUM, fumarase; GABA, γ- 1274 aminobutyric acid; GABA-T, GABA transaminase; GDC, glycine decarboxylase 1275 complex; GDH, glutamate dehydrogenase; Gly-II, hydroxyacylglutathione hydrolase 3; 1276 Gly-I, lactoylglutathione lyase; GPX, glutathione peroxidase; GR, glutathione 1277 reductase; GRX, glutaredoxin; GSH; reduced glutathione; GSSG; oxidized glutathione; 1278 GST, glutathione-s-transferase; HAD, hydroxyacyl-thioester dehydratase; HMG lyase, 48 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 1279 hydroxymethylglutaryl-CoA lyase; ICDH, isocitrate dehydrogenase; IscA, iron sulfur 1280 cluster assembly protein; MCAT, acyl-carrier-protein S-malonyltransferase; MCC, 3- 1281 methylcrotonyl-CoA carboxylase; MCD, malonyl CoA decarboxylase; MCP, 1282 mitochondrial carrier proteins; MG hydratase, methylglutaconyl-CoA hydratase; MDH, 1283 malate dehydrogenase; MDHA, monodehydroascorbate; MDHAR, 1284 monodehydroascorbate reductase; MFT, mitochondrial folate transporter/carrier-like; 1285 MnSOD, superoxide dismutase; MPP, mitochondrial processing peptidase; NAD-ME, 1286 NAD-dependent malic enzyme; NDT, nicotinamide adenine dinucleotide transporter; 2- 1287 OGDH, 2-oxoglutarate dehydrogenase; PDK, pyruvate dehydrogenase kinase; PGP, 1288 phospholipid hydroperoxide glutathione peroxidase; PHB, prohibitin; Pic, phosphate 1289 carrier protein; PRDX, peroxiredoxin; PSP, presequence protease; SAM, S- 1290 adenosylmethionine; SAM-T, S-adenosylmethionine-dependent methyltransferase; 1291 SAMC, S-adenosylmethionine carrier protein; SDH, succinate dehydrogenase; SCaMC, 1292 calcium-binding mitochondrial carrier protein; SFC, Succinate/fumarate mitochondrial 1293 transporter; SHMT, serine hydroxymethyltransferase; SSA, succinic semialdehyde; 1294 SSAD, SSA dehydrogenase; sucCoA-syn, succinyl CoA synthetase; ThPP, thiamine 1295 pyrophosphate; TIMs, import inner membrane translocases; TOMs, outer membrane 1296 translocases; TPC, thiamine pyrophosphate carrier; tRNAs, tRNA synthetases TXNRD, 1297 thioredoxin reductase; TX-SS, thioredoxin oxidized; TX-SH, thioredoxin reduced; 1298 UCP, uncoupling protein; VDAC, voltage-dependent anion channel 1299 Figure 7: Schematic representation showing the proteins identified by GeLC-MS/MS 1300 related to the functionality of mitochondrial respiratory chain complexes. 1301 Abbreviations: AOX, alternative oxidase; ATPase, ATP synthase; CI, complex I; CII, 1302 complex II; CIII; complex III; CIV, complex IV; CV, complex V; Cyt b, cytochrome b; 1303 Cyt c, cytochrome c; Cyt c1, cytochrome c1; COX, cytochrome oxidase; Ddh, 1304 dihydroorotate dehydrogenase; ETF, electron transfer flavoprotein; exNDH, external 1305 NADH-ubiquinone oxidoreductase; FeS, succinate dehydrogenase iron sulfur subunit ; 1306 GAPdh, glyceraldehydes-3-phosphate dehydrogenase, GLDH, L-galactono-1,4-lactone 1307 dehydrogenase; NAD-DH, NADH dehydrogenase [ubiquinone] subcomplex; Succ DH, 1308 succinate dehydrogenase; UCP, uncoupling protein; UQox, ubiquinone oxidized; 1309 UQred, ubiquinone reduced 1310 Figure 8: Mitochondrial localization prediction (by iPSORT, MitoProt II, Predotar, 1311 TargetP and WoLF PSORT). Predicted to be mitochondrial by one (+), two (++), three 49 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 1312 (+++), four (++++) or all five (+++++) programs. Note that about 35% of the 1313 mitochondrial proteome (both potato and Arabidopsis) are not predicted to be 1314 mitochondrial by any of the five prediction programs. Combination details of the 1315 predictions are given in Supplemental Table S4. 1316 Figure 9: Post-translational modifications (PTMs) in the potato mitochondrial proteome. 1317 The list of PTM analyzed can be found in Table II. Of the 1060 potato mitochondrial 1318 protein groups, 559 (53%) had at least one PTM. Methionine oxidation followed by 1319 proline and lysine oxidation are the most abundant types of PTM. Individual proteins 1320 contained up to nine types of PTMs (Table S6). 1321 50 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 1322 Table I: Functional distribution of the mitochondrial proteins from potato and 1323 Arabidopsis based on the functional classification made by Heazlewood et al. (2004). 1324 The relative frequency (%) of proteins classified in each functional category in the 1325 Arabidopsis and potato proteome are shown. The fold change in protein numbers 1326 (potato/Arabidopsis) are in brackets. The frequency of the relative protein abundance 1327 (% dNSAF) of each functional category in the potato proteome is also shown. Arabidopsis Potato % dNSAF Function # proteins % # proteins % Energy 98 23.6 125 (1.3) 11.8 11.5 Metabolism 81 19.5 216 (2.6) 20.4 21.1 Protein fate 53 12.7 102 (1.9) 9.6 9.3 DNA synthesis and processing 9 2.2 23 (2.6) 2.2 1.8 Transcription 8 1.9 13 (1.6) 1.2 0.9 RNA processing 14 3.4 104 (7.4) 9.8 11.1 Protein synthesis 15 3.6 126 (8.4) 11.9 12.0 Defense, stress, detoxification 16 3.8 30 (1.9) 2.8 2.8 Communication/signaling 19 4.6 39 (2.1) 3.7 3.8 Transport 19 4.6 59 (3.5) 5.6 5.1 Miscellaneous 4 1.0 0 - 0.0 0 Structural organization 9 2.2 5 (0.6) 0.5 0.3 Unknown 71 17.1 218 (3.1) 20.6 20.2 Total 416 1060 (2.5) 1328 1329 1330 51 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. Mitochondrial proteome analysis 1. Isolation of mitochondria 2. Protein extraction and SDS-PAGE fractionation Percoll gradients Potato tubers Potato tuber cells 6. Spectral counting quantification (dNSAF)k = Three-step gradient Continuous gradient (SAF)k Ʃ (SAF)n 40 slices are excised from each gel lane 4. LC-MS/MS analysis 5. Database searching Correlation analysis PROLUCID MSGF-DB SEQUEST LTQ Orbitrap XL Thermo mass spectrometer Experimental MS/MS spectra MASCOT Combination of results from different search engines Theoretical MS/MS spectra Mass spectra collected Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. 3. In gel digestion Figure 2: Number of peptides (A) and proteins (B) identified using different database searching engines <1% protein FDR identification. The total number of non-redundant peptides and proteins were 7346 and 1060, respectively. Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. EYFP Marker Overlay Figure 3 Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. Figure 4: Properties of the potato mitochondrial proteome (n = 1060) as compared to the Arabidopsis mitochondrial proteome (n = 426; [Heazlewood et al. 2004]). (A) Distribution of molecular mass (kDa), (B) isoelectric point and (C) Grand Average Hydropathy (GRAVY) score. A disproportionately large number of low-molecularweight proteins with negative Gravy Scores were identified in the potato mitochondrial proteome). Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. Less More Figure 4: Abundance distribution of the potato mitochondrial proteome (Table S2). Figure 5: Mitochondrial localization prediction results are superimposed on the protein abundance (spectral counting) [log10(dNSAF)]. Plus signs (+) indicate mitochondrial localization by as many prediction programs and minus sign (-) indicates localization by none. Each category on the x-axis corresponds to 0.25 order of magnitude difference in the abundance of proteins. The difference between the most abundant and the least abundant proteins was 1778 fold. Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. Figure 6: Schematic representation showing proteins that were identified by GeLC-MS-MS in the mitochondrial preparation of potato tubers. Squares represent protein identification and abundance based on spectral counting. Each color represents a range of Log10 transformed dNSAF (bottom left legend). Numbers in parentheses represent number of proteins found within the same range of abundance. Respiratory chain complexes (OXPHOS) are showed in detail in Figure 8. Abbreviations: ABC, ABC transporter; AC, aconitase; ACOT, acyl-coenzyme A thioesterase 9; ACS, Long chain acyl-CoA synthetase 9APX, ascorbate peroxidase; AOX, alternative oxidase; ASC; ascorbate; AlaAT, alanine aminotransferase; AspAT, aspartate aminotransferase; BCKDH, branched chain alpha-keto acid dehydrogenase; CAC, carnitine/acylcarnitine carrier protein; CI-CV, respiratory chain complexes (see Figure 7); CoA DHase, isovaleryl-CoA dehydrogenase; CS, citrate synthase; DiT1, oxoglutarate malate transporter; DHA, dehydroascorbate; DHAR, dehydroascorbate reductase; DIC, Mitochondrial dicarboxylate carrier protein; D-LDH, D-lactate dehydrogenase; E1α, pyruvate dehydrogenase subunit α; E1β, pyruvate dehydrogenase subunit β; E2, dihydrolipoamide S-acetyltransferase; E3, Dihydrolipoamide dehydrogenase; ECR, trans-2-enoyl-CoA reductase; EH, Enoyl-CoA hydratase; ER, enoyl-acyl-carrier-protein reductase; ETF, electron transfer flavoprotein; FCL, 5-formyltetrahydrofolate cycloligase; FDF, formyltetrahydrofolate deformylase-like; FUM, fumarase; GABA, γ-aminobutyric acid; GABA-T, GABA transaminase; GDC, glycine decarboxylase complex; GDH, glutamate dehydrogenase; Gly-II, hydroxyacylglutathione hydrolase 3; Gly-I, lactoylglutathione lyase; GPX, glutathione peroxidase; GR, glutathione reductase; GRX, glutaredoxin; GSH; reduced glutathione; GSSG; oxidized glutathione; GST, glutathione-s-transferase; HAD, hydroxyacyl-thioester dehydratase; HMG lyase, hydroxymethylglutaryl-CoA lyase; ICDH, isocitrate dehydrogenase; IscA, iron sulfur cluster assembly protein; MCAT, acyl-carrier-protein Smalonyltransferase; MCC, 3-methylcrotonyl-CoA carboxylase; MCD, malonyl CoA decarboxylase; MCP, mitochondrial carrier proteins; MG hydratase, methylglutaconyl-CoA hydratase; MDH, malate dehydrogenase; MDHA, monodehydroascorbate; MDHAR, monodehydroascorbate reductase; MFT, mitochondrial folate transporter/carrier-like; MnSOD, superoxide dismutase; MPP, mitochondrial processing peptidase; NAD-ME, NAD-dependent malic enzyme; NDT, nicotinamide adenine dinucleotide transporter; 2-OGDH, 2-oxoglutarate dehydrogenase; PDK, pyruvate dehydrogenase kinase; PGP, phospholipid hydroperoxide glutathione peroxidase; PHB, prohibitin; Pic, phosphate carrier protein; PRDX, peroxiredoxin; PSP, presequence protease; SAM, S-adenosylmethionine; SAM-T, S-adenosylmethionine-dependent methyltransferase; SAMC, S-adenosylmethionine carrier protein; SDH, succinate dehydrogenase; SCaMC, calcium-binding mitochondrial carrier protein; SFC, Succinate/fumarate mitochondrial transporter; SHMT, serine hydroxymethyltransferase; SSA, succinic semialdehyde; SSAD, SSA dehydrogenase; sucCoA-syn, succinyl CoA synthetase; ThPP, thiamine pyrophosphate; TIMs, import inner membrane translocases; TOMs, outer membrane translocases; TPC, thiamine pyrophosphate carrier; tRNAs, tRNA synthetases TXNRD, thioredoxin reductase; TX-SS, thioredoxin oxidized; TX-SH, thioredoxin reduced; UCP, uncoupling protein; VDAC, voltage-dependent anion channel. Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. Figure 7: Schematic representation showing the proteins identified by GeLC-MS/MS related to the functionality of mitochondrial respiratory chain complex. Abbreviations: AOX, alternative oxidase; ATPase, ATP synthase; CI, complex I; CII, complex II; CIII; complex III; CIV, complex IV; CV, complex V; Cyt b, cytochrome b; Cyt c, cytochrome c; Cyt c1, cytochrome c1; COX, cytochrome oxidase; Ddh, dihydroorotate dehydrogenase; ETF, electron transfer flavoprotein; exNDH, external NADH-ubiquinone oxidoreductase; FeS, succinate dehydrogenase iron sulfur subunit ; GAPdh, Glyceraldehyde 3-phosphate dehydrogenase, GLDH, L-galactono-1,4-lactone dehydrogenase; NAD-DH, NADH dehydrogenase [ubiquinone] subcomplex; Succ DH, succinate dehydrogenase; UCP, uncoupling protein; UQox, ubiquinone oxidized; UQred, ubiquinone reduced. Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. Figure 8: Mitochondrial localization prediction (by iPSORT, MitoProt II, Predotar, TargetP and WoLF PSORT). Predicted to be mitochondrial by one (+), two (++), three (+++), four (++++) or all five (+++++) programs. Note that about 35% of the mitochondrial proteome (both potato and Arabidopsis) are not predicted to be mitochondrial by any of the five prediction programs. Combination details of the predictions are given in the Supplemental TableS3. Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved. Figure 9: Post-translational modifications (PTMs) in the potato mitochondrial proteome. The list of PTM analyzed can be found in Table 3. Of the 1060 potato mitochondrial protein groups, 559 (53%) had at least one PTM. Methionine oxidation followed by proline and lysineoxidation are the most abundant types of PTM. Individual proteins contained up to nine types of PTMs (TableS6) Downloaded from on June 14, 2017 - Published by www.plantphysiol.org Copyright © 2013 American Society of Plant Biologists. All rights reserved.