The distinct functional roles of the inner and outer chloroplast envelope are revealed by

proteomic approaches

Elain Gutierrez-Carbonell1, Daisuke Takahashi2, Giuseppe Lattanzio1, Jorge

Rodríguez-Celma1, Jürgen Soll3, Katrin Philippar3, Julia Kehr4*, Matsuo Uemura5,

Javier Abadía1, Ana Flor López-Millán1

1Plant Nutrition Department, Aula Dei Experimental Station, CSIC, P.O. Box 13034, 50080

Zaragoza, Spain; 2United Graduate School of Agricultural Sciences, Iwate University,

Morioka 020-8550, Japan; 3Plant Biochemistry and Physiology, Department Biology I,

Ludwig-Maximilians-Universität München, Großhadernerstr. 2-4, D-82152

Planegg/Martinsried, Germany; 4Center for Plant Biotechnology and Genomics (UPM-INIA),

Montegancedo Campus, Freeway M40, km. 38, 28223-Pozuelo de Alarcón, Madrid, Spain; *

Current address: Department Molecular Plant Genetics, University Hamburg, Biocenter Klein

Flottbek, Ohnhorststrasse 18, 22609 Hamburg, Germany 5Cryobiofrontier Research Center,

Faculty of Agriculture, Iwate University, Morioka 020-8550, Japan.

Corresponding author:Ana Flor López-MillánAvenida de Montañana 1005, 50059 Zaragoza, SpainTel; 34 976 716068Fax: 34 976 716145email: anaflor@eead.csic.es

Running title: Proteomic profile of inner and outer chloroplast envelope

Summary (300 words)

The protein profiles of inner (IE) and outer (OE) envelope membrane preparations from P.

sativum were studied using two different proteomic approaches, shotgun nLC-MS/MS and

two-dimensional electrophoresis. In this study, we used the relative enrichment of each

protein in the IE/OE pair of membranes to provide an integrated picture of the actual

composition and dynamics of the system. A total of 589 protein species were identified in

chloroplast envelopes using the two different proteomic approaches. From the 546 proteins

identified with the shotgun approach 321 (59%) show significant changes in allocation

between OE and IE, and from these 295 (92%) were found only with this technique, whereas

26 (8%) were also found in the 2-DE gels. Two groups of proteins were clearly separated

with the shotgun approach, with 180 and 141 of them having larger signal intensities in IE

and OE preparations, respectively. Regarding the 2-DE approach, 69 proteins were found,

with 43 not found with the shotgun approach. Functional classification of identified protein

species reveals that IE is clearly enriched in proteins related to transport, protein metabolism

and photosynthesis, whereas proteins preferentially associated to OE are more heterogeneous

and participate in a wide range of metabolic processes. Relative enrichments clearly indicate

that in the envelope the IE is the selective barrier for the transport of a wide range of

metabolites, and also plays a major role in controlling membrane protein homeostasis.

Furthermore, based on our data, metabolic processes previously described to occur in the

envelope, including chlorophyll and tocopherol biosynthesis, can be ascribed preferentially to

IE whereas others such as carotenoid or lipid biosynthesis occur in both IE and OE. Our

results allow assignation of envelope proteins to the IE and/or the OE and also revealed new

protein species in envelope preparations that had not been previously found in the

chloroplast.

Introduction

Chloroplasts, the green photosynthetic plastids responsible for energy capture in plants, are

surrounded by a double membrane system, the chloroplast envelope. The chloroplast

envelope consists of an inner (IE) and an outer (OE) membrane and controls the bidirectional

traffic flow of a diverse range of substances between the cytosol and the biochemically active

chloroplast interior, and therefore is essential for proper cell functioning. Because the

chloroplast is a semiautonomous organelle, nuclear encoded proteins need to be transported

across this membrane system and this process is achieved via the joint action of translocons

located at the OE and IE [TOC and TIC, respectively; reviewed in (1-3)]. In addition, the

envelope controls the exchange between the stroma and the cytosol of many ions and

metabolites needed for metabolic processes. For instance, inorganic nitrite and sulphate need

to cross the envelope to be reduced within the chloroplast stroma (4-6), whereas some amino

acids and fatty acids synthesized in the chloroplast need to cross the envelope for their

distribution within the cell (7). Accordingly, transporters of amino acids, dicarboxylic and

fatty acids, sugars and inorganic ions, have been found in the chloroplast envelopes of several

plant species (4, 8-11). Other organic compounds such as cofactors and vitamins are also

likely to cross the envelope, although the molecular mechanisms of transport remain

uncharacterized (12).

The envelope is not only a selective barrier between the chloroplast and the cytosol, since

several unique biosynthetic pathways are present in this membrane system [reviewed in (12,

13)]. The chloroplast envelope plays an important role in certain aspects of fatty acid and

glycerolipid metabolism, including biosynthesis and desaturation, and also contains enzymes

involved in the formation of isoprenoid plastid constituents such as carotenoids, chlorophylls

and prenyl-quinones (12). Envelopes are also involved in signaling, since they are the site for

the production of molecules such as jasmonate and abscisic acid involved in plastid-to-

nucleus communication (12). On the other hand, the presence in the envelope of antioxidant

enzymes such as ascorbate and gluthatione peroxidases has lead to hypothesize the

participation of the envelopes in redox defense mechanisms (12, 14).

A substantial body of knowledge about envelope proteins has arisen from proteomic

approaches using spinach and Arabidopsis (4, 8, 15). In spinach, a proteomic approach

identified 54 proteins, 27 of them not previously reported in this plant compartment (4).

Later, a combined proteomic and in silico study reported more than a hundred envelope

proteins in Arabidopsis (8). A curated list, including 226 envelope proteins was later

published using a combination of the data reported until then (12), and AT_Chloro, a

database dedicated to sub-plastidial localization of Arabidopsis chloroplast proteins, was later

released (1, 16). A total of 1,323 proteins were identified in chloroplasts and statistical

analysis allowed the accurate localization of 819 proteins after cross-contamination analyses

(1, 16). From this subset, “the whole envelope proteome” included 460 proteins whereas “the

specific envelope proteome” (that containing proteins with increased abundance in the

envelope when compared to other sub-plastidial compartments) included 298 proteins (1, 16).

Their functional classification indicated that unknown, metabolism, transport, translation,

chaperone, and protease protein classes comprised more than 80% of the chloroplast

envelope proteins (1, 16). Recently, the chloroplast envelope proteomes of pea, Arabidopsis

thaliana and Medicago sativa were compared, and 247 envelope proteins were identified

(17). In addition, comparative proteomics between pea and maize has shown that C3- and C4-

type envelopes have qualitatively similar but quantitatively different membrane proteins (18).

The allocation of chloroplast envelope proteins to the OE and IE has been mainly

performed based on in silico studies and extensive literature search (1, 16, 19). This approach

has some limitations, especially for the identification of OE proteins, most of which are

synthesized without a chloroplastic targeting sequence (16). Recently, the proteome of the

OE and IE of pea has been studied using nHPLC-MALDI/QTOF, and 49 and 50 proteins

were assigned to the IE and OE, respectively (17). Overall, the proteomic characterization of

the IE and OE membranes is challenging due to: i) the difficulties in obtaining highly purified

IE and OE membrane fractions, 2) the inherent problems involving the separation and

solubilization of highly hydrophobic proteins within these fractions, and 3) the relative low

abundance of the envelope proteome, which accounts for only 1-2% of the chloroplast

proteins (16). A detailed protein profile of these membrane subsystems will thus help to

clarify their respective roles in chloroplast functioning and biogenesis.

The aim of the present study was to characterize the protein profiles of IE and OE

membranes from the chloroplast envelope of Pisum sativum, using two complementary

proteomic approaches, 2-DE and label-free shotgun proteomics. 2-DE electrophoresis is

widely used in plant proteomics (20), and label free shotgun proteomics has been successfully

used to study the proteome of membranes such as that of the plant plasma membrane (21,

22). Pea was chosen as a model system because the method for the separation of IE and OE

membranes is well established (23) and because the available EST sequencing data (24)

facilitates annotation and comparison with Arabidopsis thaliana genes.

Experimental Procedures

Plant material and envelope membrane preparation

Pea plants (Pisum sativum L., cv. ‘Arvica’, Prague, Czech Republic) were grown on sand in

the greenhouse under a 16 h light (220 µmol m-2 s-1)/8 h dark regime at 21 °C. Preparations of

OE and IE membrane fractions were obtained from intact, Percoll-purified chloroplasts,

isolated from leaves of 10-day-old pea plantlets as described in (25) and (23). Chloroplasts

were ruptured in a Dounce-homogenizer after hypertonic treatment in 0.65 M sucrose on ice,

and re-adjusted to isotonic conditions by two-fold dilution. Intact chloroplasts and thylakoid

membranes were removed by low speed centrifugation (4150g, 4 °C, 10 min), followed by a

second two-fold dilution and centrifugation at 196000g, 4 °C for 1 h. The light-green

supernatant consisting of crude envelope membranes was removed, and IE and OE

membranes separated on a discontinuous sucrose gradient (0.996, 0.800 and 0.465 M) by

centrifugation (137000g, 4 °C, 3 h), resulting in an opaque (OE) and a light yellow-green (IE)

fraction. Each fraction was diluted four-fold, centrifuged (137000g, 4 ° C, 1 h), the pellet re-

suspended in 0.5 M NaiPO4 buffer (pH 7.9) and frozen in liquid N2. Final protein

concentration was approximately 2.5 µg µL-1. For details and buffer compositions during

isolation, see (23). In order to obtain a preferably broad coverage of proteins occurring in the

envelope, including integral membrane proteins as well as those associated with the

membrane, we did not carry out further purification steps. One-dimensional gels of the IE and

OE (silver-stained) are shown in Supplementary Figure 1. Specific proteins occurring in the

IE and OE preparations were detected using specific antibodies and used to assess membrane

enrichment in the different fractions.

Label free nano-liquid chromatography-tandem mass spectrometry (nLC-ESI-MS/MS)

Label free nLC-ESI-MS/MS analysis was made with four independent biological replicates

(i.e., each one from a different batch of plants) for the two membrane types (IE and OE).

Sample preparation for label free nLC-ESI-MS/MS was carried out according to (21).

Briefly, purified OE and IE fractions were subjected to 1-DE to remove non-protein

compounds. The resulting gel bands were cut into four pieces, proteins were in gel digested

with trypsin, and subsequently peptides were extracted.

Peptide solutions were concentrated in a trap column (Lcolumn Micro 0.3 x 5 mm; CERI,

Japan) using an ADVANCE UHPLC system (Michrom Bioresources, Aubum, CA, U.S.A.).

Elution was carried out with 0.1% (v/v) formic acid in ACN and concentrated peptides were

separated in a Magic C18 AQ nano column (0.1 x 150 mm; Michrom Bioresources) using a

linear gradient of ACN (from 5% to 45%) and a flow rate of 500 nL min -1. Ionization of

peptides was carried out with a spray voltage of 1.8 kV using an ADVANCE spray source

(Michrom Bioresources). Mass spectrometry analysis was carried out on an LTQ Orbitrap

XL (Thermo Fisher Scientific, Waltham, MA, U.S.A.) equipped with Xcalibur software

(version 2.0.7, Thermo Fisher Scientific). Under the data-dependent scanning mode, full scan

mass spectra were obtained in the range of 400 to 1800 m/z with a resolution of 30000.

Collision-induced fragmentation was applied to the five most intense ions at a threshold

above 500.

Mass data analysis was performed according to (22). For semi-quantitative analysis of IE

and OE membrane proteins, Progenesis LC-MS software (version 4.0, Nonlinear Dynamics,

Newcastle, U.K.) was applied to experimental data files according to the software’s

instructions. Briefly, raw MS/MS files were imported to Progenesis software. In this process,

compared peptides were analyzed with ANOVA (p<0.05). The resultant peptides were

analyzed and identified with the Mascot search engine (version 2.3.02, Matrix Science,

London, U.K.) by comparison with the NCBInr Green Plants database (version 20120616;

18538238 sequences; 6354146082 residues). Search parameters were: peptide mass tolerance

±5 ppm, MS/MS tolerance ±0.6 Da, one allowed missed cleavage, allowed fixed modification

carbamidomethylation (Cys), and variable modification oxidation (Met) and peptide charges

were set to +1, +2 and +3. Positive identification was assigned with at least two unique top-

ranking peptides with scores above the threshold level (p < 0.05). Only proteins identified in

all four biological replicates were considered for this analysis. The list of peptides identified

and quantified by shotgun proteomic and Progenesis LC-MS analyses are listed in

Supplementary Table 1.

2-DE

Gels were made from two independent IE and OE membrane preparations (i.e., each from a

different batch of plants) with four technical replicates each (n=8). Preliminary 2-DE

experiments were carried out using a first dimension IEF separation with a linear pH gradient

3–10; in these conditions most of the spots were located in the central region of the 2-DE gel

in the OE preparations and evenly distributed throughout the gel for the IE preparations

(results not shown); therefore, to prevent protein co-migration and improve resolution a

narrower pH gradient (5-8) was chosen for the OE samples. First dimension IEF separations

were carried out in a Protean IEF Cell (BioRad, Hercules CA, USA) on 7 cm ReadyStrip IPG

Strips (BioRad) with linear pH gradients 5-8 for OE and 3-10 for IE preparations. Strips were

rehydrated for 16 h at 20ºC in 125 µL of rehydration buffer containing 80 µg (for OE) or 45

µg (for IE) of envelope proteins and a trace of bromophenol blue, and then transferred onto a

strip electrophoresis tray. IEF was run at 20ºC, for a total of 14000 V h (20 min with a 0-250

V linear gradient, 2 h with a 250-4000 V linear gradient and 4000 V until 10000 V h). After

IEF, strips were equilibrated for 10 min in equilibration solution I [6 M urea, 0.375 M Tris-

HCl, pH 8.8, 2% (w/v) SDS, 20% (v/v) glycerol, 2% (w/v) DTT] and for another 10 min in

equilibration solution II [6 M urea, 0.375 M Tris-HCl pH 8.8, 2% (w/v) SDS, 20% (v/v)

glycerol, 2.5% (w/v) iodoacetamide].

For the second dimension SDS-PAGE, equilibrated IPG strips were placed on top of

vertical 12% SDS-polyacrylamide gels (8 x 10 x 0.1 cm) and sealed with melted 0.5%

agarose in 50 mM Tris-HCl, pH 6.8, containing 0.1% SDS. SDS-PAGE was carried out at 20

mA per gel for approximately 1.5 h, until the bromophenol blue reached the plate bottom, in

a buffer containing 25 mM Tris, 192 mM glycine, and 0.1% SDS, at 4 ºC. Gels were

subsequently stained with Coomassie-blue G-250 (Serva, Barcelona, Spain).

Gel Image and statistical analysis

Stained gels were scanned with an Epson Perfection 4990 Photo scanner at 600 dpi (Epson

corporation, Barcelona, Spain). Experimental Mr values were calculated by mobility

comparisons with Precision Plus protein standard (BioRad) run in a separate lane on the SDS-

gel, and pI was determined by using a 3-10 linear scale for IE and 5-8 for OE preparations.

Spot detection, gel matching and interclass analysis were performed with PDQuest 8.0

software (BioRad). First, normalized spot volumes based on total intensity of valid spots

were calculated for each 2-DE gel and used for statistical calculations of protein abundance;

for all spots present in the gels, pI, Mr, and normalized volumes (mean values, SD, and CV)

were determined. Only spots consistently present in 100% of the replicates (all eight gels)

were considered as consistent and used in further analysis. The spots were also manually

checked, and a consistent reproducibility between normalized spot volumes was found in the

different replicates.

Protein in gel digestion of 2-DE spots

Spots were excised automatically using a spot cutter EXQuest (BioRad), transferred to 500

µL Protein LoBind Eppendorf tubes, destained in 400 µL of 40% [v/v] acetonitrile (ACN)

and 60% [v/v] 200 mM NH4HCO3 for 30 min and dehydrated in 100% ACN for 10 min. Gel

pieces were dried at room temperature and then in gel digested with 15 µL Trypsin solution

(Sequencing Grade Modified Trypsin V511, Promega, Madison, WI, US; 0.1 µg μL−1 in 40

mM NH4HCO3/9% ACN). After incubation for 5 h at 37 °C, the reaction was stopped by

adding 1 μL of 1% TFA. The peptide solution was finally analyzed using mass spectrometry

(MS).

Protein identification by nano-liquid chromatography-tandem mass spectrometry (nLC-ESI-

MS/MS) of 2-DE spots

Peptides present in 5 µL of sample were pre-concentrated on line onto a 300 µm i.d. x 5 mm,

5 µm particle size ZORBAX 300SB-C18 trap column (Agilent Technologies, Waldbronn,

Germany), using a 100 μL min-1 flow rate of 3% ACN, 0.1% formic acid, in a nano-HPLC

system 1200 series (Agilent Technologies). Backflow elution of peptides from the trap

column was carried out, and separation done with a 75 µm i.d. x 150 mm, 3.5 µm particle

size ZORBAX 300SB-C18 column (Agilent Technologies), using a 300 nL min-1 nano-flow

rate and a 55 min linear gradient from solution 97% A (0.1% formic acid) to 90% of solution

B (90% ACN, 0.1% formic acid). The nano-HPLC was connected to a HCT Ultra high-

capacity ion trap (Bruker Daltoniks, Bremen, Germany) using a PicoTip emitter (50 μm i.d.,

8 μm tip i.d., New Objective, Woburn, MA, USA) and an on line nano-electrospray source.

Capillary voltage was -1.8 kV in positive mode and a dry gas flow rate of 10 L min -1 was

used with a temperature of 180 °C. The scan range used was from 300 to 1500 m/z. The mass

window for precursor ion selection was ±0.2 Da and the rest of parameters were those

recommended by the manufacturer for MS/MS proteomics work. Peak detection,

deconvolution and processing were performed with Data Analysis 3.4 (Bruker Daltoniks).

Protein identification was carried out using the mascot search engine (Matrix Science;

London, UK) and the non-redundant databases NCBInr 20120204 (17172511 sequences;

5891506632 residues) and Plants_EST EST_110 (151512996 sequences; 26700120026

residues). Search parameters were: monoisotopic mass accuracy, peptide mass tolerance ±0.2

Da, fragment mass tolerance ±0.6 Da, one allowed missed cleavage, allowed fixed

modification carbamidomethylation (Cys), and variable modification oxidation (Met).

Positive identification was assigned with Mascot scores above the threshold level (p<0.05), at

least 2 identified peptides with a score above homology, 10% sequence coverage and similar

experimental and theoretical molecular weight and pI values. We used the GO biological

process annotation (http://www.geneontology.org) of the individual identified proteins for

classification.

In Silico analysis of subplastidial localization of identified pea proteins

The UniProt entry and the GO biological process annotation were obtained from the NCBI

entries using the PIR id Mapping tool (pir.georgetown.edu/pirwww/search/idmapping.shtml).

Results and Discussion

Protein profiling of membrane preparations enriched in IE and OE using shotgun

proteomics and 2-DE

The protein profiles of IE and OE membrane preparations from P. sativum were studied using

a shotgun nLC-MS/MS approach. A total of 546 protein species (NCBI entries) occurred in

all four biological replicates of at least one membrane type and were positively identified.

Due to the high sensitivity of the methodology used, all but one of the proteins identified

(545) were detected both in the IE and OE, although the relative abundances in each

membrane fraction were very different in many cases. To determine the relative enrichment

of proteins in the IE and OE fractions, a semi-quantitative analysis was carried out using

peptide signal intensity data to calculate the percentage of abundance (as [signal in IE/(signal

in IE + signal in OE)]). From the 546 identified protein species, 321 showed significant

differences in signal intensity between the IE and OE preparations (ANOVA, p<0.05;

Supplementary Table 2). Two groups of proteins were clearly separated, with 180 and 141 of

them having larger signal intensities in IE and OE preparations, respectively (Figure 1). The

225 proteins not showing significant differences in signal intensity between the IE and OE

preparations are listed in Supplementary Table 3.

The protein profiles of membrane preparations enriched in chloroplast inner (IE) and outer

envelope (OE) from Pisum sativum were also studied using 2-DE (IEF-SDS-PAGE

electrophoresis). Two biological and four technical replicates were analyzed for each

membrane type. Typical real scans of 2-DE gels obtained from IE and OE membrane protein

extracts are shown in Figure 2A,B, respectively. The average number of detected spots (mean

± SD; n=8) in gels obtained from IE and OE membranes was 182 ± 8 and 189 ± 6,

respectively. The number of consistent spots (those present in 100% of the gels of each

membrane type) was 160 and 181 for the IE and OE preparations, respectively. All consistent

spots were analyzed by nLC-MS/MS, and 92 and 59 spots in the IE and OE membrane

preparations were unambiguously identified as protein species. The location of these spots is

shown in the composite averaged maps in Supplementary Figure 2, and details on the protein

identifications are shown in Supplementary Table 4. These proteins corresponded to 50 and

33 NCBI protein species in IE and OE, respectively. The total number of protein species

identified with the 2-DE approach was 69, because 14 of the proteins were found in both IE

and OE.

In summary, a total of 589 protein species (NCBI entries) were identified in chloroplast

envelopes using the two different proteomic approaches (Supplementary Tables 2-4). The

overlaps between the two techniques are shown using Venn diagrams (Figure 3A). From the

546 proteins identified with the shotgun approach 321 show significant changes (59%), and

from these 295 (92%) were found only with this technique, whereas 26 (8%) were also found

in the 2-DE gels (14 in IE, 5 in OE and 7 in both). Regarding the 2-DE approach, 69 proteins

were found, with 43 not found in the shotgun approach: 22 in IE, 14 in OE and 7 in both

membrane preparations.

Proteins were also assigned to Arabidopsis gene identifiers (AGI codes; Supplementary

Tables 2-4), and a total of 417 AGI were found using the combination of proteomic

approaches. The corresponding Venn diagrams are shown in Figure 3B. From the 392 AGI

identified with the shotgun approach 229 show significant changes (58%), and from these

199 (87%) were found only with this technique, whereas 30 (13%) were also found in the 2-

DE gels (13 in IE, 6 in OE and 11 in both). Regarding the 2-DE approach, 48 proteins were

found, with 18 not found in the shotgun approach: 12 in IE, 5 in OE and 1 in both membrane

preparations (Figure 3B).

The 247 non redundant AGI identified in the two chloroplast envelope membrane

preparations (229 by shotgun and 18 by 2-DE; Supplementary Tables 2, 4) were contrasted

with the localization proposed in the At_CHLORO database. Among them, 71 were

described in the AT_CHLORO database as envelope-enriched proteins, whereas 47 were

assigned to the envelope and others (thylakoid or stroma), 51 were ascribed to other

subplastidial compartments and 78 more were not previously accounted for as chloroplastic

proteins (Figure 4).

Delving into the IE functions using shotgun proteomics and 2-DE

Proteins (NCBI entries) were then classified according to their functional category (Figure 5).

Transport (44 proteins; detected only by shotgun), photosynthesis light harvesting related

processes (45 proteins; 39 by shotgun and 6 by 2-DE), and proteolysis (30 proteins; 27 by

shotgun and three by 2-DE) were the largest functional categories of proteins associated

preferentially to the IE (Figure 5; Supplementary Tables 2, 4). Also, 38 proteins (34 in

shotgun and four detected only in 2-DE) had unknown functions and higher abundances in

the IE. These 38 proteins yielded 27 AGIs, and from these seven were ascribed in the

AT_CHLORO database to the envelope, whereas one was ascribed to thylakoids. The

remaining 19 AGIs were not present in the AT_CHLORO database.

The IE has a major role in transport

IE membrane preparations were significantly enriched in proteins related to transport, with 44

proteins ascribed to this category (in OE, only seven proteins were found in this category).

This is indicative of a major role of IE in active, carrier or pump-mediated transport

processes. Most of them were identified as envelope proteins in the AT_CHLORO database

(Supplementary Table 2). Among identified proteins, transporters for a wide array of

substrates, including lipids, nucleotides, dicarboxylates, SAM, auxin and inorganic ions were

enriched in the IE preparations. Ten proteins, corresponding to seven AGIs, were ABC

transporters, with four of them belonging to the I (ABCI10 -At4g33460-, 13 -At1g65410- and

14 -At1g19800-) and A (ATH -At5g64940-) subclasses, with known functions in lipid

transport (26). For instance, ABCI13 participates in transfer of ER-derived lipids to the

plastid galactoglycerolipid biosynthetic machinery (27). The remaining three ABC proteins

belonged to the B subclass (ABCB 25 -At5g58270-, 26 -At1g70610- and 28 -At4g25450-)

with roles in transport of lipids, pigments and Fe-sulfur clusters. One of them, ABCB25

(STA1), is functionally equivalent to yeast ATM1, required for the export of Fe-sulfur

clusters from the mitochondrial matrix (28). Five proteins, yielding three AGIs, were

phosphate transporters (PHT4-1 -At2g29650-, PHT2-1 -At3g26570- and PHT4-4 -

At4g00370-) whereas two more (PPT1 -At5g33320- and TPT -At5g46110-) were

PEP/phosphate and triose-phosphate/phosphate translocators, respectively (Supplementary

Table 2). The later one (TPT) is involved in the export of triose-phosphates generated in the

Calvin cycle to the cytosol for sucrose synthesis and has been previously found in envelopes

(4, 16, 29). Two dicarboxylate transporters (DIT1 -At5g12860- and DIT2-1 -At5g64290-)

and one protein from the major facilitator superfamily (At5g59250) involved in carbohydrate

transport were also enriched in IE. Four more proteins were orthologs to two SAM

transporters in Arabidopsis (SAMTL -At2g35800- and SAMT1 -At4g39460-); an impaired

function of SAMT1 leads to a decreased accumulation of prenyl lipids and affects

chlorophyll (Chl) biosynthesis (30). Nine cation transporters were also enriched in IE

preparations, corresponding to five AGIs of the Arabidopsis K (KEA2 -At4g00630-), Mg

(MGT10 -At5g22830-), Fe (TIC21/PIC1 -At2g15290-) and Cu (HMA1 -At4g37270- and

HMA6 -At4g33520-) transporters (Supplementary Table 2). KEA has been involved in pH

and cation homeostasis, PIC1 transports Fe to the chloroplast (31) and the HMAs have a role

in Zn detoxification in addition to Cu transport (32). Two more proteins (yielding one AGI -

At3g13070-) similar to CorC, an Mg and Co efflux protein of E. coli, were more abundant in

IE. Nucleotide transporters were also enriched in IE preparations. These include six proteins,

two of them with homology to the Arabidopsis AATP1 -At1g80300-, an ATP/ADP

antiporter, and two with homology to ATBT1 -At4g32400-, a nucleotide exporter to the

cytosol. Two more proteins with homology to MSL2 -At5g10490-, an ion channel involved

in plastid division, and RUS2 -At2g31190-, an auxin transporter, and one predicted protein

(B9H046 -At1g31410-) involved in polyamine transport, were enriched in IE.

With regard to the protein import machinery, as expected IE preparations were enriched

(with relative abundances of 60-85%) in translocons TIC 22 -At4g33350-, TIC 40 -

At5g16620-, TIC 55 -At2g24820-, TIC 55-IV -At4g25650-, TIC 62 -At3g18890- and TIC

110 -At1g06950-, previously described as IE proteins in other plant species including pea (1-

3) (Supplementary Table 2). Two orthologs of the SecA pathway, SecA1 -At4g01800- and

SecA2 -At1g21650-, were also enriched in IE preparations (ca. 85%) suggesting a role of this

pathway in protein import in the IE as it occurs in thylakoids (33). All the identified proteins

by 2-DE in this category were also present in the shotgun results.

The IE has a role in membrane protein homeostasis

Thirty one identified proteins (three of them only present in 2-DE) corresponding to 20 AGIs

were plastid proteases, including members of the well-characterized families FtsH (13) and

Clp (4) (Supplementary Tables 2 and 4). The FtsH (filamentation temperature sensitive)

family of membrane bound proteases was enriched in IE as revealed by the shotgun analysis.

Eighteen identified proteins, yielding eleven AGIs, had large signal intensities in the IE,

including among others: FtsH 2 -At2g30950-, 5 -At5g42270-, 7 -At3g47060-, 9 -At5g58870-,

11 -At5g53170-, 12 -At1g79560- and FtsHi1 -At4g23940- (Supplementary Table 2). In

addition, two AGIs only identified in 2-DE were FtsH2 -At2g30950- and FtsZ2-1 -

At2g36250- (Supplementary Table 4). Members of this family contain an ATPase -AAA+

type and a Zn(II)-metalloprotease domains, with the exception of the chloroplast specific

FtsHi’s that lacks the Zn motif (34). Excepting FtsH2 and FtsZ2-1 described in thylakoids, all

the other identified members were previously ascribed to the envelope (16, 19). All FtsH

proteases are nuclear encoded and transported to the stroma via TIC and TOC translocons.

FtsH5 is integrated in thylakoids via the Sec pathway in a NTP-dependent way (33).

Orthologs of SecA1 and SecA2 were also enriched in IE (ca. 85%) suggesting that a similar

mechanism could occur in envelopes. These proteases play a role in membrane protein

maintenance, being responsible for removal of misfolded or damaged proteins (34, 35),

therefore suggesting that IE has a role in membrane protein homeostasis.

Seven identified proteins of the Clp/Hsp100 (caseinolytic protease) family, corresponding

to four AGIs, were enriched in IE (Supplementary Table 2). Among them, we found two of

the five catalytic ClpP units of the ClpPR protease complex (1 -AtCg00670- and 4 -

At5g45390-) and two chloroplastic chaperones, ClpC1 -At5g50920- and ClpD -At5g51070-.

Clp/Hsp100 chaperones act as independent units or with the proteolytic core to degrade

irreversibly damaged proteins (36), and their enrichment in IE further supports a role in

protein homeostasis. The Clp chaperones are also implicated in protein import along with

Hsp70 chaperones, by pulling precursors in as the transit peptide enters the organelle (37).

All of the Clp members found in IE have been previously described in envelopes (4, 16).

Two more identified proteins, a stromal processing peptidase (SPP -At5g42390-)

responsible for cleavage of transit peptides (38) and a chloroplast processing enzyme (SPPA1

-At1g73990-) were also enriched in IE preparations (Supplementary Table 2). None of them

have been previously reported in envelopes. SPP has been allocated in the stroma (38)

whereas SPPA1, with a yet unknown function, has been described in thylakoids (39). A 17.6

kDa heat shock protein was identified only in 2-DE gels of IE membranes (Supplementary

Table 4).

The IE is enriched in proteins related to the thylakoid

IE preparations contained proteins involved in the light-dependent photosynthesis reactions

(Supplementary Table 2). Thirty-nine identified proteins in this class, yielding 31 AGIs, had

relative abundances in IE in the range of 67 to 91%. Nine and 16 of them were ascribed to

envelopes and thylakoids, respectively, whereas six are not present in AT_CHLORO

(Supplementary Table 2). Among identified proteins, we found the subunits A, B, C, D2, F,

G and N of the PSI reaction center (RC) (PSAA, PSAB, PSAC, PSAD2, PSAF, PSAG,

PSAN, respectively), the subunits A, B, C and E of the PSII RC (PSBA, PSBB, PSBC,

PSBE, PSBO2, PSBP1, PSBQ2, respectively), the subunits O2, P1 and Q2 of the oxygen

evolving complex (PSBO1, PSBP1 and PSQ2, respectively), the subunits 1, 3 and 4 of the

light harvesting complex (LHC) I (LHCA1, LHCA3, LHCA4, respectively), the subunits 2, 3

and 4 of the LHC II (LHCB2.1, LHCB3 and LHCB 4.2, respectively), the subunits α, β, γ and

∂ of the CF1 complex (ATPA, ATPB, ATPC1, ATPD, respectively) and the subunits a and b

of the CF0 complex of the chloroplastic ATPase (ATPI and F, respectively). Two more were

identified as Cyt f and plastocyanin (PETA and PETE1, respectively) and two are involved in

the PSII stability (HCF136, CYP38). In addition, six proteins yielding four AGIs related to

thylakoids were identified only in IE 2-DE gels, including subunit 1 of the LHC II (LHCB1),

a ferredoxin-NADP reductase and the subunits α and ε of ATP synthase (Supplementary

Table 4).

In contrast, only two proteins, the subunit D1 of the PSI RC (PSAD1) and the subunit R of

the PSII RC (PSBR) were enriched in OE preparations as revealed in shotgun and none was

detected only in 2-DE. Although these results could indicate that some degree of cross-

contamination between thylakoids and IE exist in the samples, we should keep in mind that

thylakoids are generated by invaginations of IE and therefore it is plausible to find

components of the light-dependent reactions in IE.

The IE has a role in chlorophyll and tocopherol biosynthesis

While it has been previously described that the chloroplast envelope contains enzymes

involved in the formation of Chl and tocopherol (12), our results support that these processes

occur preferentially in the IE (Supplementary Table 2). Nine proteins of tetrapyrrole

metabolism, yielding six AGIs, and two of tocopherol biosynthesis were enriched in IE

preparations, with abundances in the range 70-90%. These included two protoporphyrinogen

oxidases (PPOX -At4g01690- and PPO2 -At5g14220-) from the common tetrapyrrol

pathway, pheophorbide A oxygenase (PAO -At3g44880-) of the Chl breakdown pathway and

three proteins of the Chl synthesis pathway: Mg-protoporphyrin IX monomethylester cyclase

(CHL27 -At3g56940-), divinyl protochlorophyllide reductase (PCB2 -At5g18660-), and

SAM Magnesium-protoporphyrin O-methyltransferase (CHLM -At4g25080-). One more, an

uroporphyrinogen decarboxylase (UPD2 -At2g40490-) involved in Chl synthesis, was only

detected in IE 2-DE gels (Supplementary Table 4). The two proteins involved in tocopherol

biosynthesis were tocopherol cyclase (VTE1 -At4g32770-) and methyltransferase (VTE4 -

At1g64970-). Nine of these proteins had been previously ascribed to envelopes, whereas

CHLM and PCB2 have not been found previously in this compartment and UPD2 was

assigned to the stroma (Supplementary Tables 2 and 4). The 2-DE analysis revealed two

protein species yielding one AGI (glutamate 1-semialdehyde aminotransferase -At3g48730-)

related to the tetrapyrrole metabolism, which was only detected in OE preparations

(Supplementary Table 4).

Envelopes are also involved in carotenoid biosynthesis (12, 16) and three enzymes of this

pathway were found in this study, a phytoene desaturase (PDS3 -At4g14210-) and a carotene

mono-oxygenase (CYP97C1 -At3g53130-), which were more abundant in IE, and a

carotenoid dioxygenase (CCD1 -At3g63520-), which was more abundant in OE

(Supplementary Table 2). Therefore, it is difficult is to ascribe this pathway to one of the

membranes since enzymes involved in their biosynthesis were found in both IE and OE.

The IE has a role in lipid metabolism

Nine proteins ascribed to lipid metabolism, yielding seven AGIs, had abundances in IE

ranging from 62 to 85% (Supplementary Table 2). Identified proteins were part of the early

steps of fatty acid (CAC3 -At2g38040- and AAE16 -At3g23790-), phospholipid (PGP1 -

At2g39290-) and sulpholipid (SQD2 -At5g01220-) biosynthesis as well other lipid related

processes (CYP74B2 -At4g15440-, CYP97B3 -At4g15110- and LOX6 -At1g67560-). Three

of them (AAE16, CAC3, and CYP97B3) were previously described in envelopes, whereas

the remaining six were not in the AT_CHLORO database.

Delving into the OE composition by shotgun proteomics and 2-DE

While the IE membrane preparations are clearly enriched in proteins belonging to three major

metabolic pathways, proteins preferentially associated to OE participate in a wide range of

metabolic processes. Carbohydrate metabolism (38 proteins; 32 by shotgun and six by 2-DE),

protein transport (19 proteins; 18 by shotgun and one by 2-DE), lipid metabolism (19

proteins; 18 by shotgun and 1 by 2-DE), and protein synthesis (15 proteins detected by

shotgun) were the largest metabolic categories within proteins preferentially associated to

OE. Proteins involved in other metabolic processes such as transcription, signal transduction

and Fe homeostasis were also enriched in OE (Figure 5). Thirteen proteins (11 revealed by

shotgun and two found only in 2-DE), yielding ten AGIs, had unknown functions, and from

these three were ascribed to the envelope in AT_CHLORO, whereas the remaining were not

present in this database. A membrane-associated 30 kDa protein of unknown function and

ascribed to the envelope in AT_CHLORO was detected in IE and OE preparations only in the

2-DE proteomic approach (Supplementary Table 4).

The OE roles in carbohydrate metabolism

Proteins of the light-dependent phase of photosynthesis were not particularly abundant in OE,

but those related to the dark-phase did show a preferential accumulation. Twenty-three

proteins with nine AGIs involved in the three phases of the Calvin cycle were more abundant

in OE (78-99%) (Supplementary Table 2). All enzymes participating in carboxylation and

reduction were found in OE, including both subunits of Rubisco (RBCS1A -At1g67090- and

RBCL -AtCg00490-), phosphoglycerate kinase (At1g56190) and two isoforms of

glyceraldehydo-3-phosphate dehydrogenase (GAPA-2 -At1g12900- and GAPB -

At1g42970-). Two proteins were involved in the regeneration phase (fructose-bisphosphate

aldolase 1 -At4g38970- and phosphoribulokinase -At1g32060-) and two members of the

chaperonin 60 family (CPN60B1 -At1g55490- and CPN60B3 -At5g56500-) involved in the

correct folding of Rubisco (40) were also enriched in OE. Four of the nine AGIs have been

found in Arabidopsis envelopes whereas the remaining ones were ascribed to the stroma in

AT_CHLORO. All of them, except for RBCL, are nuclear encoded and most have been

annotated as precursors, suggesting that their presence in OE reflects their transport to the

stroma. A preferential allocation of these proteins in OE may indicate that OE is a limiting

step for their transport to the stroma, and might also suggest that they are not functionally

active in envelopes. Six proteins involved in carbohydrate metabolism were only detected in

2-DE gels of OE; these proteins corresponded to three AGIs: RBCL -AtCg00490-, fructose-

bisphosphate aldolase -At4g38970- and a transketolase -At3g60750- (Supplementary Table

4). Four more proteins yielding two AGIs: RBCL -AtCg00490- and Rubisco activase (RCA -

At2g39730-) were only detected by 2-DE in both IE and OE, whereas six more yielding four

AGIs (RCA -At2g39730-, CPN60A1 -At2g28000-, CPN60B1 -At1g55490- and PGK1 -

At3g12780-) were only detected in 2-DE gels of IE (Supplementary Table 4).

Nine more proteins involved in carbohydrate metabolism, yielding six AGIs, were also

more abundant (86-97%) in OE (Supplementary Table 2). These included two hexokinases

(HXK1 -At4g29130- and HXK2 -At2g19860-) involved in glycolysis, a transaldolase of the

pentose phosphate shunt -At1g12230-, an aldehyde dehydrogenase -At1g44170-, a

chloroplastic precursor of the carbonic anhydrase (CA1 -At3g01500-) and a β-glucosidase -

At3g06510-. Most of them were identified as envelope proteins in At_CHLORO. In contrast,

only two proteins related to carbohydrate metabolism were more abundant in IE: a plastid-

NAD dependent malate dehydrogenase (At3g47520) and a xyloglucan galactosyltransferase

(KAM1 -At2g20370-).

The OE roles in transport

In contrast with the IE results, only seven proteins related to transport were enriched in OE

(Supplementary Table 2), with three of them yielding the same AGI (ABCG7 -At2g01320-).

Interestingly, the G subclass of ABC transporters contains “reverse orientation” transporters

functioning in lipid export (26), which would imply that the OE is responsible for export

whereas the IE controls the import of lipids into the chloroplast. The remaining four proteins

were the already characterized OE channel pore proteins of 16 -At2g28900-, 21 -At1g76405-,

24 -At1g45170- and 37 -At2g43950- kDa, with functions in aminoacid, phosphorylated

carbohydrate, triose-P, dicarboxylic acid and sugar transport (41-44).

OE preparations were enriched (with relative abundances of 80-99%) in proteins

belonging to the OE translocon, including eighteen proteins with ten AGIs (Supplementary

Table 2). These included the known translocon proteins 34 -At5g05000-, 64-III -At3g17970-,

75-III -At3g46740-, 75-V -At5g19620-, 90 -At5g20300-, 120 -At3g16620-, 132 -At2g16640-

and 159 -At4g02510- and two members of the HSP70 chaperon family (HSP-70 -At3g12580-

and HSP70-1 -At5g02500-). One more protein, an ortholog of the Arabidopsis TIC110 -

At1g06950-, was only identified in 2-DE OE preparations (Supplementary Table 4).

The OE is enriched in ribosomal proteins

Ten identified proteins enriched in OEs, yielding nine AGIs, were constituents of plastid

ribosomal subunits including the RPS4 -AtCg00380-, 8 -AtCg00770- and 17 -At1g79850- of

the 30S subunit and RPL1 -At3g63490-, 2.2 -AtCg01310-, 6 -At1g05190-, 13 -At1g78630-,

14 -AtCg00780-, and 27 -At5g40950- of the 50S subunit. Five more OE-enriched proteins

were related to protein translation, including two elongation factors (SVR3 -At5g13650- and

RABD8 -At4g20360-), one peptidyl-tRNA hydrolase (PTH2 -At5g16870-), and two proteins

involved in ribosomal RNA metabolism, the CSP41B endonuclease -At1g09340- (45) and

RBP31 -At4g24770- (Supplementary Table 2). The genes of the plastidic translational

machinery are distributed between the plastid and the nucleus, with the genes for

processing/modification enzymes, aminoacyl tRNA synthetases, and 60% of the RPs located

in the nuclear genome (46). Six of the identified ribosomal constituents and the PTH2 are

nuclear encoded; this makes tempting to speculate that their enrichment in OE may be a

consequence of their translocation process that could be a bottleneck in their movement into

the plastid. However, the presence of four plastid-encoded RPs in OE opens the possibility

for other explanations.

The 2DE analysis revealed four proteins related to protein synthesis only detected by this

technique. Among these, two (a RPL1 -At3g63490- and a TUFA -At4g20360- elongation

factor) were found in both IE and OE preparations whereas the other two (RPL10 -

At5g13510- and a tRNA Ile-lysidine synthetase -At3g24560-) were found in OE membranes

(Supplementary Table 4).

The OE roles in lipid metabolism

Eighteen proteins with high abundances in OE (67-97%) were ascribed to lipid metabolism

(Supplementary Table 2). These were represented by nine AGIs involved in fatty acid

synthesis, including the three components of the pyruvate dehydrogenase complex (pyruvate

dehydrogenase E1 -At2g34590-, dihydrolipoamide acetyl transferase -At1g34430- and

lipoamide dehydrogenase -At3g16950-), an acetylCoA carboxylase (CAC2 -At5g35360-), a

dephospho CoA Kinase (ATCOE -At2g27490-), and a long chain fatty acid synthase (LACS9

-At1g77590-), PECT1 -At2g38670- involved in phospholipid synthesis, a lipid associated

protein (PAP -At3g26070-) and a cytochrome b5 reductase (CBR1 -At5g17770-) that

provides electrons for fatty acid desaturases (47). Also, a protein similar to the

dihydrolipoamide acetyl transferase was only found in 2-DE of OE preparations

(Supplementary Table 4). These results indicate that both membranes have an active role in

lipid metabolism.

The OE roles in protein homeostasis

In contrast to the IE results, only seven proteins involved in proteolysis were enriched in OE

preparations (Supplementary Table 2). These corresponded to four AGIs, including ClpP5 -

At1g02560-, two metallo-endopeptidases (At5g56730 and At1g67690) and UBQ11 -

At4g05050-. The later two have not been previously described to occur in the chloroplast

envelope.

The OE roles in aminoacid synthesis and redox defense mechanisms

Five proteins yielding four AGIs involved in cystein (ACS1 -At2g43750-), arginine

(At4g24830), threonine (TS1 -At4g29840-) and glutamate biosynthesis (GLU2 -At2g41220-)

had abundances in OE ranging from 71 to 89% (Supplementary Table 2). One more cysteine

synthase (OASA1 -At4g14880-) was identified only by 2-DE in OE preparations

(Supplementary Table 4). All of them, with the exception of GLU2 were ascribed to the

stroma in AT-CHLORO.

Five proteins, yielding four AGIs related to redox metabolism were also more abundant in

OE. These included two peroxiredoxins (2CPB -At5g06290- and PEX22 -At3g21865-),

ascorbate peroxidase (APX3 -At4g35000-) and monodehydroascorbate reductase (MDAR4 -

At3g27820-) related to redox defense.

The OE roles in signal transduction, transcription and Fe homeostasis

Eight proteins with abundances between 80 and 93% in OE were related to signal

transduction (Supplementary Table 2). Two of them matched phototropin 2 (PHOT2 -

At5g58140-), a blue light photoreceptor involved in chloroplast movement that has been

recently found in OE (48) and two more matched WAV2 -At5g20520-, a protein involved in

the regulation of external stimuli (49). The remaining four proteins belong to the rab

(RAB8A-like -At5g59840-, RAB1B -At4g17170-) and rho (MIRO1 -At5g27540- and

MIRO2 -At3g63150-) subfamilies of the small GTPase superfamily. The small rab-GTPases

are usually associated with vesicular trafficking, whereas the rho subfamily participates in

cytoskeleton reorganization (50).

Eight proteins yielding eight AGIs and with abundances in OE from 65 to 98% were

related to transcription (Supplementary Table 2). These included two subunits of the RNA

polymerase (RPOC1 -AtCg00180-, RPOA -AtCg00740-), three transcription regulators

(PTAC3 -At3g04260-, 14 -At4g20130- and 16 -At3g46780-) and three histones (HTA3 -

At1g54690-, HTA12 -At5g02560-, and HTB9 -At3g45980-). A MADS box protein involved

in regulation of transcription was identified only in 2-DE gels of OE preparations

(Supplementary Table 4).

Finally, three proteins with two AGIs (FER2 -At3g11050- and FER4 -At2g40300-) had

higher abundance in OE.

Conclusions and final remarks

In summary, our combined proteomic approach has allowed some progress towards the

characterization of the protein profiles of the inner (IE) and outer (OE) chloroplast envelope

membranes of P. sativum. Envelope preparations necessarily contain intrinsic proteins, but

also include proteins being processed or transported at the time of isolation (since a

proteomic profile is always a snapshot of a dynamic process), as well as a small degree of

contamination from neighboring compartments. In this study we have used the relative

enrichment of each protein in the IE/OE pair of membranes to provide an integrated picture

of the actual composition and dynamics of the system. The IE is clearly enriched in proteins

related to transport, protein metabolism and photosynthesis, whereas proteins preferentially

associated to OE are more heterogeneous and participate in a wide range of metabolic

processes. Relative enrichments clearly indicate that in the envelope the IE is the selective

barrier for the transport of a wide range of metabolites, and also plays a major role in

controlling membrane protein homeostasis, given the high prevalence of metallo-proteases

involved in the removal of damaged proteins. Furthermore, based on our data, metabolic

processes previously described to occur in the envelope, including chlorophyll and tocopherol

biosynthesis, can be ascribed preferentially to IE whereas others such as carotenoid or lipid

biosynthesis occur in both IE and OE. Data also reveal that the IE preparations were

particularly enriched in proteins involved in the light-dependent photosynthetic processes

whereas proteins related to the dark-phase of photosynthesis showed a preferential

accumulation in the OE.

Also, whereas results show a good overlap with the AT_CHLORO database, they add a

new layer of information by allowing further assignation of envelope proteins to the IE and/or

the OE. Furthermore, this combined approach has revealed 78 AGIs (32% of the total number

found) in envelope preparations that had not been previously found in the chloroplast. These

included some protein species involved in chlorophyll and tocopherol biosynthesis, peptide

processing enzymes, lipid metabolism and other categories. Finally, it should be mentioned

that even though shotgun is a technique more powerful than 2-DE for membrane protein

profiling, both are complementary, with many proteins being detected in only one of the two

methodologies.

Acknowledgements

Supported in part by the Spanish Ministry of Economy and Competitivity (MINECO;

projects AGL2010-16515 and AGL2012-31988, co-financed with FEDER), the Aragón

Government (Group A03), and by a Japan Society for the Promotion of Sciences Research

Fellowship for Young Scientists (#247373 to D.T.) and Grants-in-Aid for Scientific Research

from the Ministry of Education, Culture, Sports, Sciences and Technology, Japan

(#22120003 and #24370018 to M.U)..

References

1. Bruley, C., Dupierris, V., Salvi, D., Rolland, N., and Ferro, M. (2012) AT_CHLORO: A Chloroplast Protein Database Dedicated to Sub-Plastidial Localization. Frontiers in plant science 3, 205

2. Hiltbrunner, A., Bauer, J., Alvarez-Huerta, M., and Kessler, F. (2001) Protein translocon at the Arabidopsis outer chloroplast membrane. Biochemistry and cell biology = Biochimie et biologie cellulaire 79, 629-635

3. Jarvis, P., and Soll, J. (2002) Toc, tic, and chloroplast protein import. Biochimica et biophysica acta 1590, 177-189

4. Ferro, M., Salvi, D., Rivière-Rolland, H., Vermat, T., Seigneurin-Berny, D., Grunwald, D., Garin, J., Joyard, J., and Rolland, N. (2002) Integral membrane proteins of the chloroplast envelope: Identification and subcellular localization of new transporters. Proceedings of the National Academy of Sciences of the United States of America 99, 11487-11492

5. Leustek, T., and Saito, K. (1999) Sulfate transport and assimilation in plants. Plant physiology 120, 637-644

6. Mourioux, G., and Douce, R. (1979) [Sulfate transport across the limiting double membrane or envelope, of spinach chloroplasts]. Biochimie 61, 1283-1292

7. Ortiz-Lopez, A., Chang, H. C., and Bush, D. R. (2000) Amino acid transporters in plants. Biochimica et Biophysica Acta - Biomembranes 1465, 275-280

8. Ferro, M., Salvi, D., Brugière, S., Miras, S., Kowalski, S., Louwagie, M., Garin, J., Joyard, J., and Rolland, N. (2003) Proteomics of the chloroplast envelope membranes from Arabidopsis thaliana. Molecular & cellular proteomics : MCP. 2, 325-345

9. Renne, P., Dressen, U., Hebbeker, U., Hille, D., Flugge, U. I., Westhoff, P., and Weber, A. P. (2003) The Arabidopsis mutant dct is deficient in the plastidic glutamate/malate translocator DiT2. Plant J 35, 316-331

10. Shikanai, T., Muller-Moule, P., Munekage, Y., Niyogi, K. K., and Pilon, M. (2003) PAA1, a P-type ATPase of Arabidopsis, functions in copper transport in chloroplasts. The Plant cell 15, 1333-1346

11. Weber, A. P., Schwacke, R., and Flugge, U. I. (2005) Solute transporters of the plastid envelope membrane. Annual review of plant biology 56, 133-164

12. Rolland, N., Ferro, M., Seigneurin-Berny, D., Garin, J., Block, M., and Joyard, J. (2009) The chloroplast envelope proteome and lipidome. In: Sandelius, A. S., and Aronsson, H., eds., pp. 41-88

13. Block, M. A., Douce, R., Joyard, J., and Rolland, N. (2007) Chloroplast envelope membranes: a dynamic interface between plastids and the cytosol. Photosynthesis research 92, 225-244

14. Rolland, N., Ferro, M., Seigneurin-Berny, D., Garin, J., Douce, R., and Joyard, J. (2003) Proteomics of chloroplast envelope membranes. Photosynthesis research 78, 205-230

15. Froehlich, J. E., Wilkerson, C. G., Ray, W. K., McAndrew, R. S., Osteryoung, K. W., Gage, D. A., and Phinney, B. S. (2003) Proteomic study of the Arabidopsis thaliana chloroplastic envelope membrane utilizing alternatives to traditional two-dimensional electrophoresis. Journal of proteome research 2, 413-425

16. Ferro, M., Brugiere, S., Salvi, D., Seigneurin-Berny, D., Court, M., Moyet, L., Ramus, C., Miras, S., Mellal, M., Le Gall, S., Kieffer-Jaquinod, S., Bruley, C., Garin, J., Joyard, J., Masselon, C., and Rolland, N. (2010) AT_CHLORO, a comprehensive chloroplast proteome database with subplastidial localization and curated information on envelope proteins. Mol Cell Proteomics 9, 1063-1084

17. Simm, S., Papasotiriou, D. G., Ibrahim, M., Leisegang, M. S., Muller, B., Schorge, T., Karas, M., Mirus, O., Sommer, M. S., and Schleiff, E. (2013) Defining the core proteome of the chloroplast envelope membranes. Frontiers in plant science 4, 11

18. Brautigam, A., Hoffmann-Benning, S., and Weber, A. P. (2008) Comparative proteomics of chloroplast envelopes from C3 and C4 plants reveals specific adaptations of the plastid envelope to C4 photosynthesis and candidate proteins required for maintaining C4 metabolite fluxes. Plant physiology 148, 568-579

19. Ferro, M., Salvi, D., Brugiere, S., Miras, S., Kowalski, S., Louwagie, M., Garin, J., Joyard, J., and Rolland, N. (2003) Proteomics of the chloroplast envelope membranes from Arabidopsis thaliana. Mol Cell Proteomics 2, 325-345

20. Jorrin-Novo, J. V., Maldonado, A. M., Echevarria-Zomeno, S., Valledor, L., Castillejo, M. A., Curto, M., Valero, J., Sghaier, B., Donoso, G., and Redondo, I. (2009) Plant proteomics update (2007-2008): Second-generation proteomic techniques, an appropriate experimental design, and data analysis to fulfill MIAPE standards, increase plant proteome coverage and expand biological knowledge. Journal of proteomics 72, 285-314

21. Li, B., Takahashi, D., Kawamura, Y., and Uemura, M. (2012) Comparison of plasma membrane proteomic changes of Arabidopsis suspension-cultured cells (T87 Line) after cold and ABA treatment in association with freezing tolerance development. Plant & cell physiology 53, 543-554

22. Takahashi, D., Li, B., Nakayama, T., Kawamura, Y., and Uemura, M. (2013) Plant plasma membrane proteomics for improving cold tolerance. Frontiers in plant science 4, 90

23. Waegemann, K., and Soll, J. (1995) Characterization and isolation of the chloroplast protein import machinery. Methods in cell biology 50, 255-267

24. Franssen, S. U., Shrestha, R. P., Brautigam, A., Bornberg-Bauer, E., and Weber, A. P. (2011) Comprehensive transcriptome analysis of the highly complex Pisum sativum genome using next generation sequencing. BMC genomics 12, 227

25. Keegstra, K., and Yousif, A. E. (1986) Isolation and characterization of chloroplast envelope membranes. Methods in Enzymology, pp. 316-325

26. Verrier, P. J., Bird, D., Burla, B., Dassa, E., Forestier, C., Geisler, M., Klein, M., Kolukisaoglu, U., Lee, Y., Martinoia, E., Murphy, A., Rea, P. A., Samuels, L., Schulz, B., Spalding, E. J., Yazaki, K., and Theodoulou, F. L. (2008) Plant ABC proteins - a unified nomenclature and updated inventory. Trends in Plant Science 13, 151-159

27. Xu, C., Fan, J., Froehlich, J. E., Awai, K., and Benning, C. (2005) Mutation of the TGD1 chloroplast envelope protein affects phosphatidate metabolism in Arabidopsis. The Plant cell 17, 3094-3110

28. Chen, S., Sánchez-Fernández, R., Lyver, E. R., Dancis, A., and Rea, P. A. (2007) Functional characterization of AtATM1, AtATM2, and AtATM3, a subfamily of Arabidopsis half-molecule ATP-binding cassette transporters implicated in iron homeostasis. Journal of Biological Chemistry 282, 21561-21571

29. Cho, M. H., Jang, A., Bhoo, S. H., Jeon, J. S., and Hahn, T. R. (2012) Manipulation of triose phosphate/phosphate translocator and cytosolic fructose-1,6-bisphosphatase, the key components in photosynthetic sucrose synthesis, enhances the source capacity of transgenic Arabidopsis plants. Photosynthesis research 111, 261-268

30. Bouvier, F., Linka, N., Isner, J. C., Mutterer, J., Weber, A. P. M., and Camara, B. (2006) Arabidopsis SAMT1 defines a plastid transporter regulating plastid biogenesis and plant development. The Plant cell 18, 3088-3105

31. Duy, D., Stübe, R., Wanner, G., and Philippar, K. (2011) The chloroplast permease PIC1 regulates plant growth and development by directing homeostasis and transport of iron. Plant physiology 155, 1709-1722

32. Kim, Y. Y., Choi, H., Segami, S., Cho, H. T., Martinoia, E., Maeshima, M., and Lee, Y. (2009) AtHMA1 contributes to the detoxification of excess Zn(II) in Arabidopsis. Plant Journal 58, 737-753

33. Rodrigues, R. A. O., Silva-Filho, M. C., and Cline, K. (2011) FtsH2 and FtsH5: Two homologous subunits use different integration mechanisms leading to the same thylakoid multimeric complex. Plant Journal 65, 600-609

34. Wagner, R., Aigner, H., and Funk, C. (2012) FtsH proteases located in the plant chloroplast. Physiologia Plantarum 145, 203-214

35. Kato, Y., and Sakamoto, W. (2010) New insights into the types and function of proteases in plastids. International Review of Cell and Molecular Biology, pp. 185-218

36. Kim, J., Rudella, A., Rodriguez, V. R., Zybailov, B., Olinares, P. D. B., and Van Wijk, K. J. (2009) Subunits of the plastid CIpPR protease complex have differential contributions to embryogenesis, plastid biogenesis, and plant development in arabidopsis. The Plant cell 21, 1669-1692

37. Bruch, E. M., Rosano, G. L., and Ceccarelli, E. A. (2012) Chloroplastic Hsp100 chaperones ClpC2 and ClpD interact in vitro with a transit peptide only when it is located at the N-terminus of a protein. BMC Plant Biology 12

38. Trösch, R., and Jarvis, P. (2011) The stromal processing peptidase of chloroplasts is essential in Arabidopsis, with knockout mutations causing embryo arrest after the 16-cell stage. PLoS ONE 6

39. Wetzel, C. M., Harmacek, L. D., Yuan, L. H., Wopereis, J. L. M., Chubb, R., and Turini, P. (2009) Loss of chloroplast protease SPPA function alters high light acclimation processes in Arabidopsis thaliana L. (Heynh.). Journal of Experimental Botany 60, 1715-1727

40. Kim, S. R., Yang, J. I., and An, G. (2013) OsCpn60α1, encoding the plastid chaperonin 60α subunit, is essential for folding of rbcL. Molecules and Cells 35, 402-409

41. Pohlmeyer, K., Soll, J., Grimm, R., Hill, K., and Wagner, R. (1998) A high-conductance solute channel in the chloroplastic outer envelope from pea. The Plant cell 10, 1207-1216

42. Pohlmeyer, K., Soll, J., Steinkamp, T., Hinnah, S., and Wagner, R. (1997) Isolation and characterization of an amino acid-selective channel protein present in the chloroplastic outer envelope membrane. Proceedings of the National Academy of Sciences of the United States of America 94, 9504-9509

43. Böllter, B., Soll, J., Hill, K., Hemmler, R., and Wagner, R. (1999) A rectifying ATP-regulated solute channel in the chloroplastic outer envelope from pea. EMBO Journal 18, 5505-5516

44. Goetze, T. A., Philippar, K., Ilkavets, I., Soll, J., and Wagner, R. (2006) OEP37 is a new member of the chloroplast outer membrane ion channels. Journal of Biological Chemistry 281, 17989-17998

45. Beligni, M. V., and Mayfield, S. P. (2008) Arabidopsis thaliana mutants reveal a role for CSP41a and CSP41b, two ribosome-associated endonucleases, in chloroplast ribosomal RNA metabolism. Plant Mol Biol 67, 389-401

46. Yamaguchi, K., and Subramanian, A. R. (2000) The plastid ribosomal proteins. Identification of all the proteins in the 50 S subunit of an organelle ribosome (chloroplast). Journal of Biological Chemistry 275, 28466-28482

47. Kumar, R., Wallis, J. G., Skidmore, C., and Browse, J. (2006) A mutation in Arabidopsis cytochrome b5 reductase identified by high-throughput screening differentially affects hydroxylation and desaturation. Plant Journal 48, 920-932

48. Kong, S. G., Suetsugu, N., Kikuchi, S., Nakai, M., Nagatani, A., and Wada, M. (2013) Both phototropin 1 and 2 localize on the chloroplast outer membrane with distinct localization activity. Plant and Cell Physiology 54, 80-92

49. Mochizuki, S., Harada, A., Inada, S., Sugimoto-Shirasu, K., Stacey, N., Wada, T., Ishiguro, S., Okada, K., and Sakai, T. (2005) The arabidopsis wavy growth 2 protein modulates root bending in response to environmental stimuli. The Plant cell 17, 537-547

50. Bourne, H. R., Sanders, D. A., and McCormick, F. (1990) The GTPase superfamily: A conserved switch for diverse cell functions. Nature 348, 125-132

Figure legends

Figure 1. Allocation of protein species (NCBI entries) between OE and IE using shotgun

proteomics. Only protein species showing significant differences in allocation between the IE

and OE membrane preparations (p<0.05) are shown.

Figure 2. 2-DE IEF-SDS PAGE protein profile maps of inner (IE) and outer (OE) chloroplast

preparations from pea. Proteins were separated in the first dimension in a linear 3-10 (IE) and

5-8 (OE) gel strips (7 cm) and in the second dimension in 12% acrylamide vertical gels (8 x 8

cm) and stained with Coomassie-blue G-250. Scans of typical gels are shown in A (IE) and

B, (OE).

Figure 3. Venn diagrams showing partial overlaps of identified protein species between the

shotgun and 2-DE techniques, using (A) the NCBI identifiers and (B) the corresponding AGI

blast results.

Figure 4. Localization of identified envelope proteins (using their AGI) compared to the

AT_CHLORO database (16) of Arabidopsis thaliana chloroplastic proteins. A reference non-

redundant Arabidopsis proteome was built by carrying out BLAST searches of the identified

(NCBI) envelope proteins of pea in the Arabidopsis genome database (TAIR). Duplicate

identifiers were deleted and only E-values lower that e-30 were considered.

Figure 5. (A) Distribution of identified pea protein species (NCBI) by shotgun proteomics or

only by 2-DE according to their GO:P annotation. (B) Distribution of identified pea protein

species according to their GO:P annotation in OE and IE.

Legends of Supplementary Tables

Supplementary Table 1. List of peptides identified and quantified by shotgun proteomics

and Progenesis LC-MS analyses.

Supplementary Table 2. Protein species showing significant differences in abundance

between IE and OE membrane preparations (ANOVA, p<0.05) using shotgun proteomics.

Grey and white backgrounds indicate proteins identified in IE and OE, respectively. Columns

are: (A) NCBI entry, (B) protein description, (C) plant species, (D) UniProt entry, (E)

percentage of abundance (as [signal in IE/(signal in IE + signal in OE)]), (F) corresponding

AGI, (G) localization using the AT_CHLORO database, (H) AGI name, (I) E blast value, (J)

pI, (K) molecular weight, (L) number of identified peptides, (M) peptides used for

quantitation, (N) confidence score, (O) Anova p, (P) signal mean in IE, (Q) signal SD in IE,

(R) signal mean in OE, (S) signal SD in OE, (T) GO Slim process (U) GO Slim function and

(V) GO Slim component. Positive identification was assigned with at least two unique top-

ranking peptides with scores above the threshold level (p < 0.05). Only proteins identified in

all four biological replicates were considered for this analysis. For all identified proteins,

blast searches were run against the TAIR database and the closest match (using a threshold

value of e-50) was assigned to the protein species. NCBI entries corresponding to AGIs not

present in the AT_CHLORO database are highlighted in red.

Supplementary Table 3. Protein species showing no significant differences in abundance

between IE and OE membrane preparations (ANOVA, p<0.05) using shotgun proteomics.

Analysis details and column description is as in Supplementary Table 2

Supplementary Table 4. Proteins identified in 2-DE IEF-SDS PAGE gels. Grey and white

backgrounds indicate proteins identified in IE and OE, respectively. Columns are: (A) spot

number in the gels, (B) protein description, (C) found in shotgun, (D) UniProt entry, (E)

percentage of abundance (as [signal in IE/(signal in IE + signal in OE)]), (F) corresponding

AGI, (G) localization using the AT_CHLORO database, (H) AGI name, (I) E blast value, (J)

Mascot score/pep/ion, (K) sequence coverage (%), (L) theoretical molecular weight/pI, (M)

experimental molecular weight/pI, (N) GO process, (O) GO:P identifier, (P) inferred

function. Positive identification was retained with Mascot scores above the threshold level

(p≤0.05), at least 2 identified peptides with a score (pep/ion) above homology, at least 10%

sequence coverage and similar experimental and theoretical MW and pI. For all identified

proteins, blast searches were run against the TAIR database and the closest match (using a

threshold value of e-50) was assigned to the protein species. NCBI entries corresponding to

AGIs not present in the AT_CHLORO database are highlighted in red.

Legends of Supplementary Figures

Supplementary Figure 1. One-dimensional gels of the IE and OE (silver-stained). Specific

proteins occurring in the IE and OE preparations were detected using specific antibodies and

used to assess membrane enrichment in the different fractions.

Supplementary Figure 2. Virtual composite images containing all spots present in the real

2-DE IEF-SDS PAGE gels are shown in (A) for IE and (B) for OE. Identified proteins are

marked.