2070

DOI 10.1002/pmic.200600953

Proteomics 2007, 7, 20702082

RESEARCH ARTICLE

Proteomics of ionically bound and soluble extracellular

proteins in Medicago truncatula leaves
Nelson C. Soares1*, Rita Francisco1*, Cndido P. Ricardo1, 2 and Phil A. Jackson1
1
2

Plant Biochemistry, Instituto de Tecnologia Qumica e Biolgica, Oeiras, Portugal

Instituto Superior de Agronomia, Lisboa, Portugal

A large proportion of the apoplast proteome resides in the intercellular fluid (IF) or is ionically
bound (IB) to the wall matrix. A combined analysis of IF and IB proteins of the Medicago truncatula leaf apoplast was performed. 2-DE analyses demonstrated the reproducible presence of 220
IF and 84 IB proteins in the apoplast. These two protein populations were largely distinct; 22
proteins could be spatially matched, but MALDI-TOF/TOF analyses suggested a considerably
smaller number had common identities. MALDI-TOF/TOF characterisation identified 81 distinct proteins. Analyses of selected IF proteins (45) indicated 17 distinct proteins with mainly
defence-related functions, whereas analyses of IB proteins (70) identified 63 distinct proteins of
diverse natures, including proteins of non-canonical natures. The presence of non-canonical
proteins in IB extracts is discussed in the light of evidence supporting a low level of contamination of purified walls from symplastic proteins. This work indicates that IB and IF proteins are
functionally distinct fractions of the apoplast. The data obtained complements earlier studies of
the Medicago proteome and therefore will be useful in future studies investigating the role of
apoplastic proteins in plant processes.

Received: November 30, 2006

Revised: February 28, 2007
Accepted: March 6, 2007

Keywords:
Apoplast / Intercellular fluid / Ionically bound wall proteins / MALDI-TOF/TOF / Medicago truncatula

Introduction

The apoplast contains a dynamic composition of proteins

involved in a myriad of functions including cell expansion
[1], growth cessation [2], intercellular adhesion [3], signalling
[4], fruit maturation [5, 6] and the response to biotic [7, 8] and
abiotic stresses [9]. Despite the central role apoplastic proteins play in these processes, the apoplast proteome is less
well characterised than that of other sub-cellular compart-

Correspondence: Dr. Phil Jackson, Instituto de Tecnologia

Quimica e Biologica, Av. da Republica, Apartado 127, Oeiras,
2780157, Portugal
E-mail: phil@itqb.unl.pt
Fax: 135-1-21-443-3644
Abbreviations: ECM, extracellular matrix; IB, ionically bound; IF,
intercellular fluid; OEE, oxygen evolving enhancer; PR, pathogen-related; SI, saline infiltrates

2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

ments [10], partly due to the difficulty in isolating a representative fraction of apoplastic proteins demonstrably free
from intracellular contamination.
Many studies have used non-disruptive methods as a
convenient means to isolate secreted proteins from cell cultures free from intracellular contaminants [11, 12], but for
more complex tissues, vacuum-infiltration techniques are
required to extract apoplastic proteins without incurring cellular disruption [9, 13, 14]. A major problem associated with
this technique is that the protein yield is generally low [7, 10,
14, 15]. In addition, infiltrates prepared with single buffers
fail to provide information describing possible differences in
protein solubility in the apoplast, which can provide insight
into protein function. Furthermore, potentially large differences between the quantities of soluble and more tightly
bound forms could lead to difficulty in spot detection which
could be avoided by the use of fractionated samples.
* These authors contributed equally to this work.

www.proteomics-journal.com

Proteomics 2007, 7, 20702082

In some cases, sequential vacuum-infiltration with different salts and chelating agents could result in isolation of
specific fractions of wall proteins. However, in Arabidopsis,
sequential extractions have been shown to result in increased
wall disruption and a greater risk of contamination from the
symplastic space [11]. Thus, in order to isolate more tightly
bound proteins from the wall, it is often necessary to employ
disruptive techniques to obtain purified walls for sequential
extraction as suggested by Chivasa et al. (2002) [16]. On the
other hand, cell-disruptive methods could result in wall contamination through the electrostatic binding of intracellular
proteins with the extracellular matrix (ECM), which is largely
negatively charged [17]. Several studies using disruptive
techniques have in fact not only identified several classical
cell wall proteins but also unexpected proteins that generally
have been considered to be non-secretory or traditionally
associated with organelles other than the wall [8, 16, 1820].
The presence of these proteins in the wall is a current subject
of controversial discussion. Nonetheless, while the possibility of intracellular contamination has not been excluded in
many cases, proteins of unexpected nature have also been
isolated from the cell wall using undisruptive techniques
[1114], suggesting they are targeted to the ECM in vivo. In
this study, we provide corroborative evidence supporting
these novel proteins as true cell wall proteins.
The intercellular fluid (IF) is important for the transport
of essential solutes in extracellular spaces. As the outermost
cellular compartment, it also serves to release IF, or guttation
fluid at sites of damage, leading to the transport of, e.g.
pathogen-related (PR) proteins to the wounded surface to
help impede pathogen ingress. In agreement, many studies
have found several PR-proteins present in the IF [2123], and
PR-proteins represent the major quantitative changes in soluble proteins during defensive responses [24].
The intercellular spaces are also likely to contain signalling proteins which can interact with specific receptors at the
ECM to initiate signal cascades in cell-cell communication
and other events. In Arabidopsis, it has been demonstrated
that the stem cell-specific protein, CLAVATA3 (CLV3), is soluble in the extracellular space where it is required for the
activation of the CLV1/CLV2 receptor complex at the plasma
membrane of the underlying cells [4]. In maize, the small
extracellular and hydrophilic protein, embryo surrounding
region, shares a conserved region with CLV3, and is also
thought to participate in protein-protein interactions [25].
It might also be expected that the protein population
more tightly bound to the ECM to be more enriched in proteins functionally dedicated to this compartment. Matrixmodifying proteins such as, e.g. expansin [26], xyloglucan
endotransglycosidase/ hydrolase [1] or extensin [27], and
those thought to be involved in wall/cytoplasm communication (e.g. Arabinogalactan-rich proteins [28]), might be
expected to be largely restricted to the ECM where they are
functionally relevant.
Nevertheless, a few examples have been documented of
apoplastic proteins which can be found both bound to cell
2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Plant Proteomics

2071

walls and as soluble forms in the IF. During the ripening of

both tomatoes [29] and melon fruit [6], there is an increase in
water soluble pectins and a decrease in CDTA (cyclohexanetrans-1,2-diaminetetra-acetate)- and Na2CO3-soluble pectin
fractions, leading to the release of proteins bound to such
fractions into the IF. In lupin, a basic peroxidase can be
found freely soluble in the extracellular spaces of meristems,
but strongly associated to the cell wall in post meristematic
phases of vegetative growth [30]. A further acidic peroxidase
of lupin shows similar changes in solubility in post elongation phases of etiolated hypocotyls [31].
Documented cases of such changes in protein attachment have been scarce; thus, the extent of this phenomenon
has not been studied in any detail. The anionic nature of the
primary wall is essentially due to galacturonic acids, whose
charge promotes wall swelling and the retention of appropriately charged proteins [32]. Its charge is heavily and
reversibly influenced by changes in physiological pH and
apoplastic ion content [33], which can occur during development and the response to stress. Furthermore, developmentally regulated pectin methylestarases demethylate
nascent homogalacturonic acid in a block-wise or punctuate
action [34], and can therefore determine the pattern and capacity of wall charge and resulting protein retention. In
addition, the analysis of the secretome in the culture medium of Arabidopsis suspension cells revealed the presence of a
range of enzymes and structural proteins, as well as a variety
of proteins involved in diverse cellular functions, including
signalling [35], suggesting that proteins in the IF can share
some functions with proteins bound to the ECM.
To date, it remains unclear whether proteins of the IF and
those ionically bound (IB) to the ECM are largely distinct, or
whether a large number of apoplastic proteins can exist in
both electrostatically bound and freely soluble forms. Therefore, proteomics studies aimed at identifying differentially
regulated extracellular proteins should also be aware that
changes in protein level may, in part, be related with changes
in protein solubility. Here we report our data obtained from
the study of both IB and freely soluble proteins of the apoplast in Medicago truncatula leaves.
Due to its fairly small diploid genome, Medicago is
rapidly becoming well established as a model legume, e.g.
[20, 3639], and has become increasingly utilised in genomic
and proteomic studies. Recently, a 2-DE-based proteomic
study was applied to stem tissue of Medicago sativa and
identified 87 distinct proteins bound to the ECM [20]. Here,
we have used 2-DE and MALDI-TOF/TOF characterisation to
complement this earlier work by looking at both IB proteins
and proteins freely soluble in the IF of Medicago truncatula
leaf. We report that there exists a clear functional compartmentalisation between these protein fractions, and that IB
proteins appear to be the most functionally diverse. This
combined approach therefore provides a more detailed
description of the apoplast proteome of Medicago, and will
therefore assist future proteomic studies in this important
model legume.
www.proteomics-journal.com

2072

N. C. Soares et al.

Materials and methods

2.1 Plant material

Callus from Medicago truncatula cv. Jemalong was induced
from leaf explants, placed on Murashige and Skoog-based
medium [40] containing (per liter) 20 g sucrose, 100 mg
casein hydrolysate, 5 g PVP 40 T, 1 mg 2,4-dichlorophenoxyacetic acid, 0.5 mg kinetin, 2 g Gelright (Sigma, Germany)
and 5 mg of DTT (Merck, Germany). Selected callus were
subsequently maintained at 247C in the dark and subcultured every three weeks on the same medium.
M. truncatula cv. Jemalong seeds were imbibed on watersoaked filter paper for 48 h prior to planting in a substrate
composed of sterilised sand:peat:soil (1:1:1), and growth in a
25/197C, light/day cycle with a photoperiod of 12 h and a
light intensity of 250 mmol.m-2.s-1. Leaves were harvested for
extract preparation after 4245 days.
2.2 Protein extraction and enzyme assays
IB proteins were extracted from tissue homogenates as
described in Jackson et al. (2001) [41]. Briefly, plant material
(leaves or callus) were homogenised in 2 mL/g (fresh
weight) of sodium acetate buffer (15 mM, pH 4.5) and crude
cell wall fractions isolated by centrifugation at 45006g for
5 min. The crude cell wall fraction was washed twice by centrifugation at 45006g in 2 mL/g (fresh weight) of 1% Triton
X-100 followed by three washes in 2 mL/g (fresh weight) of
sodium acetate buffer (pH 4.5). The resultant cell wall pellet
was then resuspended in 1 mL/g (fresh weight) of 1 M KCl
for a maximum of 5 min before centrifugation at 45006g to
yield the saline extract. Saline eluates of M. truncatula callus
were obtained by extensively washing small callus pieces in
sodium acetate buffer (15 mM, pH 4.5), then gently agitating
in 2 mL/g (fresh weight) of 1 M KCl in 15 mM Na acetate
(pH 4.5) for 10 min. Both saline extracts and eluates were
ultrafiltered through a 0.45 mM filter (Schleicher & Schuell,
Germany) and concentrated by pressure-assisted filtration
through a 10 kDa cut-off membrane (Diaflow, Amicon,
USA). The extracts were than processed using the 2-DE
Cleanup Kit (Amersham, USA) as per the manufacturers
instructions.
To extract soluble, apoplastic proteins, sectioned leaf
pieces were washed with distilled water and vacuum infiltrated with sodium acetate buffer (15 mM, pH 4.5), for three
periods of 30 s and carefully blotted before centrifugation at
14806g for 15 min at 47C to collect the infiltrate, which was
clarified through a 0.45 mM filter (Schleicher & Schuell) and
concentrated in a speed-vac (Savant, USA). Where referred to
in the text, leaves were infiltrated with the same buffer containing 1 M KCl and proteins collected as described above.
In leaf extracts, malate dehydrogenase activity (EC
1.1.1.37) was assayed as a marker of cytosolic contamination
in IF and saline extracts of purified cell walls as described by
Lopez-Millan et al. (2000) [42].
2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Proteomics 2007, 7, 20702082

2.3 2-DE
Protein concentration in the extracts was determined with a
BioRad protein assay kit (BioRad, Germany), using a modified Bradford assay [43] as suggested by Ramagli et al. (1999)
[44].
For IEF, the IPGphor system was used (Amersham Biosciences) with 310 non-linear (NL) pH gradient strips (IPG
strips, Amersham Biosciences; Sweden). Proteins were
solubilised in 8 M urea, 2% w/v CHAPS, 40 mM DTT and
0.5% v/v IPG buffer (310 NL (Amersham Biosciences). IEF
was carried out at 30 V for 12 h, followed by 250 V for 1 h,
500 V for 1.5 h, 1000 V for 1.5 h, a gradient to 8000 V over
1.5 h and 8000 V for a further 4 h, all at 207C. Prior to the
second dimension SDS-PAGE, the focused IPGstrips were
equilibrated for 2615 min in 50 mM Tris-HCl (pH 8.8) buffer containing 6 M urea, 30% v/v glycerol, 2% w/v SDS and a
trace of Bromophenol Blue. DTT at 1% w/v was added to the
first equilibration step and 2.5% w/v iodoacetamide to the
second. SDS-PAGE was performed on 12 or 15% polyacrylamide gels [45]. For analytical 2-DE gels, silver-staining was
performed according to Blum et al. (1987) [46] and gels were
loaded with 40 mg of total protein. In all cases the experiments used at least three replicate protein samples prepared
from separate biological samples. For preparative gels, MScompatible silver staining [47] was utilised and the gels
loaded with 150 mg of total protein. Gels were scanned using
the ImageQuant v3.3 densitometer (Molecular Dynamics)
and were analysed by the Image Master Platinum software
v.5.0 (Amersham Biosciences).
2.4 Trypsin digestion of proteins and characterisation
by MALDI-TOF/TOF
Selected spots were excised from the gels and destained
with a solution containing 20% w/v sodium thiosulphate
and 1% w/v potassium ferricyanide for 5 min. The supernatant was removed and the gel spots were washed twice
with 25 mM ammonium bicarbonate in 50% v/v ACN for
20 min. The gel spots were then washed in ACN, dried in a
speed-vac (Savant, USA) and digested overnight with 20 mg/
mL of trypsin in 25 mM ammonium bicarbonate at 377C.
Tryptic peptides were passed through C18 Zip-Tips and
mixed with 5 mg/mL of an CHCA as matrix and subject to
MALDI-TOF/TOF analysis (4700 Proteomics Analyzer,
Applied Biosystems).
2.5 Database queries and protein identification
Databases (Est-others, NCBInr, NRDB, ) were queried with
either MASCOT data files obtained from MALDI -TOF/TOF
mass spectral data or the compiled partial de novo sequences
obtained per spot, using MASCOT software available at
www.matrixscience.com, or MS-Blast [48] at http://
dove.embl-heidelberg.de/Blast2, respectively. All the protein
www.proteomics-journal.com

Proteomics 2007, 7, 20702082

Plant Proteomics

2073

identification reported had MASCOT or MS-Blast scores

greater than the accepted significant threshold (at the p,0.05
level).

Results

3.1 Comparing proteins of the IF and IB proteins of

the M. truncatula leaf apoplast
An analysis of 2-D gels of leaf apoplastic proteins revealed
that both the IF and IB fraction of the cell wall contained a
considerable number of proteins (Figs. 1A and 1B). Analyses
of four replicate biological samples of each fraction revealed
that 220 spots could be reproducibly detected in the IF, while
the IB fraction contained 84 reproducible spots. These analyses also showed that the IF contained spots with a wide
range of pI and molecular weight (A), while 2-DE gels of the
IB fraction demonstrated a clear predominance of acidic
(pI 3.56) proteins (B), although a number of highly basic
proteins remain unresolved proximal to the cathode
(.pI 9.5).
We wished to determine if the IF and IB fractions contained proteins in common, or whether these two fractions
could be considered as largely distinct populations. We
therefore intended to identify proteins of the IF and IB which
could be spatially matched in 2-D gels and assay for their
common identity by MALDI-TOF/TOF.
Due to the large qualitative differences in spot patterns
presented by IF and IB fractions, direct 2-D gel alignments
could not be made with confidence. Saline infiltrates (SI)
of leaves prepared with 1 M KCl (Fig. 1C) did however,
appear to contain prominent spots of both the IF (spot
nos. 69, 71, 83, and 97) and IB fractions (spot nos. 4, 33,
and 44). The identities of these spots in 1 M KCl infiltrates
with their counterparts in IF and IB fractions were confirmed by MALDI-TOF/TOF analyses (Tables 1 and 2).
Obtaining the relative spatial locations of several prominent IB and IF proteins enabled the generation of a composite reference gel comprising both fractions to facilitate
their 2-D alignments.
Alignments of 2-D gels revealed that 22 proteins in the IF
and IB fractions could be spatially matched. However, subsequent MALDI-TOF/TOF characterisation of six putative
matches demonstrated them to have differing homologies
from their spatial partner (72 and 47; 81 and 59, and 83 and
61; Tables 1 and 2, respectively). A further three apparent
matches (spots 108 and 32; 79 and 40, 98 and 43; Fig. 1)
contained at least one protein which showed no homology to
data-base entries, but nevertheless produced de novo sequences which differed. A single pair (IF spot 103 and IB
spot 33) shared homology to a germin-like protein (accession
no. CAC34417; see Tables 1 and 2). These results demonstrate that IF and IB proteins consist largely of different
populations of proteins.
2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Figure 1. 2-DE of leaf apoplastic proteins. (A) Soluble IF proteins.

(B) IB proteins. (C) vacuum-infiltrates of leaves with 1 M KCl (SI).
In this gel, spots that are spatially mapped to spots in IB or IF gels
are numbered. Numbered spots (in C) whose identity with IF (gel
A) or IB (gel B) proteins was confirmed by MALDI-TOF/TOF are
identified by an asterisk. All gels were loaded with 40 mg total
protein.

www.proteomics-journal.com

2074

N. C. Soares et al.

Proteomics 2007, 7, 20702082

Table 1. MALDI-TOF/TOF characterisation of leaf IF proteins. Databases were queried with either MASCOT data files or the compiled partial
de novo sequences. In all cases, peptide charges were 11

Spot
No.

kDa

69
69SIe)
70
71
71SI
72
73
81
83
83SI
85
87
93
94
97
97SI
100
103
105
106
109
112
114

34.9
34.9
31.6
33.3
33
34.7
29.0
39.5
39.2
39.2
32.6
28.8
37.1
37.1
28.6
28.3
26.4
26.2
23.5
23.0
21.1
15.5
17.2

pI

4.0
4.0
4.0
4.3
4.3
4.8
4.6
5.9
5.95
5.95
6.4
6.2
7.8
8.4
7.45
7.45
7.4
6.0
6.2
6.6
7.3
6.7
4.3

Homologue

b21,3 Glucanase
b21,3 Glucanase
b21,3 Glucanase
Thaumatin-like protein
Thaumatin-like protein
b21,3 Glucanase
Class II chitinase
Class II chitinase
Class Ib chitinase
Class Ib chitinase
Endochitinase
Chitinase
b21,3 Glucanase
b21,3 Glucanase
NtPRp27-like protein
NtPRp27-like protein
Thaumatin-like protein
Germin-like protein
Class Ib chitinase
Thaumatin-like protein
Class Ib chitinase
b21,3 Glucanase
Putative glycine-rich

Accession
No.a)

Classb)

Q8GT15
Q8GT15
Q6S9W0
AY035168
AY035168
Q9ZP12
P29024
Q9SDY6
Q7X9F6
Q7X9F6
AAD34596
CAA64868
AF435088
O23473
AY185207
AY185207
O04364
CAC34417
Q42428
O04364
Q42428
AF239617
Q6Z498

D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
OR
D
D
D
D
S

MALDI-TOF/TOFc)

de novo Sequencingd)

Score

Matched
peptides

Score

Matched
peptides

2
2
2

2
2
2

2
2
65
2
2
2
2
154
2
2
2
2
2
81
114
2
129
2
2

2
2
1
2
2
2
2
1
2
2
2
2
2
1
1
2
1
2
2

264
350
256
161
161
224
2
532
369
492
365
2
124
310
241
459
102
2
2
102
2
355
208

3
4
4
2
3
4
2
5
6
7
5
2
2
5
4
9
2
2
2
2
2
4
5

a) Accession number of the homologue in NCBI database.

b) The protein class is abbreviated as follows: D, Defence; S, Structural; OR, Oxidoreductase.
c) Proteins were identified from the NCBInr database.
d) Result obtained using MS Blast with the NCBInr database.
e) Protein isolated by saline infiltration and excised from the respective gel (SI).

3.2 Relative abundance of IB and IF proteins in the

leaf apoplast
Although the above results successfully demonstrate that IB
and IF fractions are qualitatively different, the relative abundance of these fractions in the apoplast (per g (Fwt)) has been
difficult to estimate. This is largely due to the difficulties in
obtaining quantitative yields of IF proteins [49], whereas IB
proteins can be extracted more efficiently from purified cell
walls, allowing more realistic estimates of their abundance.
However, the relative abundance of IF and IB proteins in
the apoplast might be calculated using 2-D analyses of infiltrates prepared with 1 M KCl (SI), which can be expected to
contain both IF and IB proteins in levels which reflect their
relative abundance in vivo. The spot volumes of selected proteins in gels of isolated IF or IB extracts can be compared to
their spot volumes in combined fractions prepared by
vacuum-infiltration with 1 M KCl.
Therefore, four proteins exclusive to, and prominent in
the IF, and a further four proteins from the IB were chosen.
In gels of SI proteins, spot volumes of IF proteins were, on
2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

average, reduced to ca. 60% of their spot volumes observed in

gels of aqueous infiltrates (Fig. 2A), whereas in these same
gels, spot volumes of IB proteins were reduced further (on
average, to ca. 40%; see Fig. 2B). This indicates that the relative abundance of IF: IB proteins in the Medicago leaf apoplast was approximately 3:2. Based on the yield of IB proteins
from purified cell walls (33 mg/g (Fwt)), we estimate the IF
contains ca. 1.5 times more protein, at ca. 50 mg/(Fwt).
3.3 Assay for intracellular contamination in
apoplastic extracts
We have assayed malate dehydrogenase activity in extracts as
a marker of cytosolic contaminants. Table 3 shows that in all
apoplastic extracts, the specific malate dehydrogenase activities were very low relative to that present in cytosolic extracts,
indicating negligible cytosolic contamination. However, the
various washing steps involved in purifying cell walls for the
preparation of IB fractions are likely to remove soluble, cytosolic marker enzymes such as malate dehydrogenase, while
leaving behind charged symplastic proteins which could
www.proteomics-journal.com

Plant Proteomics

Proteomics 2007, 7, 20702082

2075

Table 2. MALDI-TOF/TOF characterisation of leaf IB proteins. The databases were queried with either MASCOT data files or the compiled
partial de novo sequences. In all cases peptide charges were 11

Spot
No.

kDa

pI

1
2
3

24.1
21.5
18.8

4.0
4.2
4.2

4
4SIe)
5
6
7
8
9

16.3
16.1
14.7
18.0
20.6
20.3
22.8

4.3
4.3
4.3
4.6
4.8
5.1
5.4

10
11
12

21.2
23.1
23.1

5.5
5.2
5.6

13
14

14.0
16.5

5.0
5.3

15
16

18.8
20.6

5.5
5.6

17
18

18.3
15.6

5.7
5.6

19
20
21
22

16.0
15.6
15.6
21.5

5.8
6.0
6.5
8.5

23

20.3

8.8

24

18.8

9.5

25
26

30.1
31.0

4.9
5.0

27

32.8

4.9

28

28.8

5.3

29

29.2

5.5

30

26.8

5.6

31

28.8

5.6

33
33SI
34
35
36
37

29.2
29.2
26.0
35.3
36.7
33.3

6.0
6.0
6.0
4.9
4.8
5.3

Homologue

Putative stress inducible

ABC-a.a type transport
N-acetylglucosamine
transferase I
Plastocyanin
Plastocyanin
Glycine-rich protein 1.8
PR-10
Peroxidase precursor
Putative glycine rich
Putative cyclosporin
A-binding
Superoxide dismutase
Superoxide dismutase
Lysine-rich wall
structural protein
Thioredoxin
Oxygen evolving
enhancer protein 1
ABC sugar transporter
a-galactoside-binding
protein
Transketolase precursor
Assimilatory nitrate
reductase
Rubisco small chain 3C
Thioredoxin
Rubisco small chain 3C
Protease serine 2
isoform B
Folypolyglutamatesynthase
Oxygen evolving
enhancer protein 3
Glycine-rich protein
Putative NAD-dependent
oxireductase
Sensory box histidine
kinase
Oxygen evolving
enhancer protein 1
Oxygen evolving
enhancer protein 1
3-dehydroquinate
synthase
Mannose-1-phosphateguanyltranferase
Germin like protein
Germin like protein
Putative secreted protein
RNA-binding protein
ATP synthase beta
Peroxidase

2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Accession
No.a)

Classb)

NP_910497
BAD04724
AAD03022

MALDI-TOF/TOFc)

de novo Sequencingd)

Score

Matched
peptides

Score

Matched
peptides

D
T/B
GA

2
2
2

2
2
2

62
70
111

1
1
2

CAA26709
CAA26709
P10496
CAA67375
AAB41811
AAP54051
CAC81066

EPC
EPC
S
D
OR
S
T/B

2
2
2
114
2
2

2
2
2
1
2
2

216
229
510
2
332
475
285

3
3
5
2
4
9
7

AAC14127
AAB05888
XP_465138

OR
OR
S

2
127
2

2
2
2

194
2
167

3
2
4

CAA53900
AB043960

OR
EPC

72

2
196

2
3

Q8ZDQ7
CAC49966

T/B
T/B

2
2

2
2

64
106

1
2

AAD10219
AAG05168

Misc
ID

225
2

4
2

2
64

2
1

BG449895
AW775706
AW775375
AAL14244

N/A
OR
N/A
Misc

322
84
147
2

5
1
4
2

2
2

2
2

118

BAC46013

Misc

66

AJ498300

EPC

422

AY389763*
CAD76101

S
OR

2
2

2
2

130
65

3
1

AAR34625

ID

68

AW559699

EPC

CAA78043

EPC

Q6L1G8

88

191

Misc

161

BAC08163

GA

65

CAC34417
CAC34417
BAC74347
AW696867
AAK72818
AAB41811

OR
OR
U
Misc
EPC
OR

284
217

2
2

174
2
2

2
2
2

2
2
173
2
173
324

2
2
4
2
3
5

www.proteomics-journal.com

2076

N. C. Soares et al.

Proteomics 2007, 7, 20702082

Table 2. Continued

Spot
No.

kDa

pI

38
39

31.0
35.1

5.4
5.8

40
41

34.6
30.1

5.7
5.9

42
43
44
44SI
45
46

32.8
33.8
32.4
32
33.6
40.3

6.3
6.8
7.4
7.4
9.5
5.4

47
48
49

36.4
36.7
34.8

5.5
5.6
5.6

50

42.8

6.0

51

44.7

6.5

52

47.4

6.8

53
54
55
56
57
58

45.3
51.0
50.2
50.2
53.3
51.0

9.5
4.8
5.3
5.5
5.5
5.6

59
60
61

46.7
48.8
46.7

5.8
5.9
6.0

62
63

55.7
54.1

6.2
6.5

64

54.8

6.7

65
66
67

52.1
63.5
40.3

7.8
5.5
5.3

68

40.2

5.3

Homologue

Ferritin
S-Adesonylmethionine
synthase
Triophosphate isomerase
Epithelial cell adhesion
molecule
Unknown protein
Germin-like protein
Germin-like protein
Germin-like protein
Germin-like protein
Oxygen evolving
enhancer protein 1
Harpin binding protein 1
Type I anti-freeze protein
Vacuolar sorting-associated protein
ATP synthase (beta
subunit)
ATP synthase (beta
subunit)
Arabinogalactan protein
AGP2
Peroxidase
Threonine dehydratase
Hypothetical protein
Phosphoribulosekinase
Hypothetical protein
Putative ABC transporter
ATP-binding
Sensor histidine kinase
Unknown protein
Ferredoxin-NADP(H)
oxidoreductase
Unknown protein
D-3-phosphoglycerate
dehydrogenase
ABC transporter
ATP-binding
Peroxidase
ATP synthase
Oxygen evolving
enhancer protein 1
Oxygen evolving
enhancer protein 1

Accession
No.a)

Classb)

CAA65771
Q5P2V5

MALDI-TOF/TOFc)

de novo Sequencingd)

Score

Matched
peptides

Score

Matched
peptides

T/B
Misc

92
2

2
2

2
60

2
1

AAT46998
AB161197

EPC
ID

311
2

4
2

2
63

2
1

AACY01181595
P94040
AW776317
AW776317
P94040
P12359

U
OR
OR
OR
OR
EPC

2
2
166
325
2
2

2
2
2
2
2
2

67
75
2
2
75
311

1
1
2
2
1
5

BF006528
AAR22529
CAA81876

D
D
Misc

184
2
2

3
2
2

2
113
134

2
3
3

AAD46914

EPC

519

AAD46914

EPC

358

AAB35284

143

BI272831
Q63BA8
AAG39973
CAA72118
AE013598
Q5Z2A9

OR
Misc
U
EPC
U
T/B

456
2
2
88
2
2

6
2
2
4
2
2

2
72
63
2
62
70

2
1
1
2
1
1

Q6HC69
AACY01426312
CV283317

ID
U
OR

2
2
170

2
2
3

102
406
2

2
3
2

AAB41813
BAC07877

U
EPC

2
2

2
2

84
67

1
1

Q9A7G4

T/B

67

CAA62226
AAD46914
AW317313

OR
EPC
EPC

2
380
227

2
3
2

197
2
2

4
2
2

AW776405

EPC

388

a) Accession number for the homologue in NCBI database.

b) The protein class is abbreviated as follows: D, Defence; T/B, Transport/binding; GA, Glycoactive; EPC, Energy production and conversion; S, Structural; OR, Oxidoreductase; Misc, Miscellaneous; ID, Interaction domain; N/A, Not applicable; U, Unknown.
c) All proteins were identified from MASCOT data files using the Est-others database, except for spots 38, 51, 56, 66, 67 which were identified using NCBInr.
d) All proteins were identified by MS Blast using NCBInr database.
e) Protein isolated by saline infiltration and excised from the respective gel (SI).

2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.proteomics-journal.com

Plant Proteomics

Proteomics 2007, 7, 20702082

2077

Figure 2. Assaying the relative abundance of IF and IB proteins in the leaf apoplast. (A) Spot volumes of four proteins in the IF prepared by
vacuum-infiltration with aqueous buffer (grey bars) and vacuum-infiltrates prepared with 1 M KCl (white bars). (B) Spot volumes of three IB
proteins extracted from purified leaf cell walls (grey bars) or through vacuum infiltration with 1 M KCl (white bars). Error bars represent the
SD based on three biological replicates. In all cases, 40 mg of proteins were loaded onto gels and the normalised spot volumes expressed as
a % of total spot volume.
Table 3. Activities of the cytosolic marker enzyme, malate dehydrogenase, in soluble (cytosolic) and apoplastic extracts
of callus material and leaves. Specific activity is given as
units activity mg-1 protein6102 (1 U = 1 mmol/s)

Material

Leaf

Callus

Fraction
Cytosol (soluble)
Intercellular fluid (aq.)
Intercellular fluid
(1 M KCl)
Purified cell walls
Cytosol (soluble)
Saline eluate
Purified cell walls

Specific activity
malate dehydrogenase
24.30 6 3.00
0.01 6 0.00
0.13 6 0.01
0.20 6 0.02
25.0 6 1.00
0.58 6 0.04
0.43 6 0.09

bind to the negatively charged cell wall through electrostatic

interactions. This contamination of IB fractions would be
undetectable by marker enzyme assays.
A lack of such contaminants in the IB fraction might be
indicated by demonstrating that all IB proteins obtained in
saline extracts of purified walls can also be liberated from the
cell wall by infiltration with 1 M KCl, which can be demonstrably performed without incurring cellular disruption and
the liberation of symplastic contaminants (Table 3). Interestingly, this was shown for a non-canonical protein, plastocyanin (Table 2, spots 4 and 4SI) which is present in both
extracts.
However, as demonstrated above (Fig. 1C and Fig. 2),
many IB proteins are poorly represented in 2-D gels of SI
proteins, due to the predominance of IF proteins in this
fraction. Consequently, the majority (54/84) of IB proteins
can not be verified as true, wall-bound proteins using this
approach. Sequential infiltrations with aqueous and saline
buffers could yield an enriched fraction of IB proteins which
would overcome this limitation. However, sequential infiltrations can lead to increased contamination from the symplast [13], and we therefore avoided this technique.
2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Although we were unable to directly assay for symplastic

contaminants in our leaf IB extracts, we have used Medicago
callus to determine if the binding of charged, symplastic
proteins to cell walls was necessarily a consequence of our
methods of wall purification and IB extraction per se. Callus
from Medicago is a convenient material for this purpose,
since it can be sequentially extracted by surface elution with
aqueous and saline buffers to provide an enriched source of
IB proteins without causing cellular disruption (Table 3). The
saline eluate can be directly compared with saline extracts of
purified callus walls by 2-DE.
This comparison revealed that saline extracts of purified
callus walls contained 34% less protein spots than that in
saline eluates of intact cells. This is likely due to protein loss
by solubilisation during the several washing steps involved
in cell wall purification. However, 92% of the spots present in
saline extracts of purified callus walls could be closely spatially matched to 2-DE gels of eluates from intact callus cells
(see Fig. 3), strongly suggesting their common identity. This
was confirmed for six selected pairs of spatially mapped proteins by MALDI-TOF/TOF (Table 4). This indicates that only
the remaining 8% unmatched IB proteins could be considered as potential contaminants from the symplast. Interestingly, similar to that observed for the non-canonical plastocyanin found in both 1 KCl infiltrates and purified walls of
leaves, eluates from intact callus cells and saline extracts
from purified cell walls both contained a protein with
homology to malate dehydrogenase (Table 4, spots 125E and
125H). Together, this data supports a low degree of contamination by symplastic proteins in saline extracts of purified walls, yet nevertheless, provides evidence for the presence of non-canonical proteins in this fraction.
3.4 Characterisation of apoplastic proteins
Some of the proteins spots belonging to the same fraction
shared deduced homology with the same sequence in the
data base. These spots demonstrated only minor shifts in
www.proteomics-journal.com

2078

N. C. Soares et al.

Proteomics 2007, 7, 20702082

Figure 3. An assay for contamination by electrostatically charged proteins of the symplast during callus cell wall purification. (A) 2-DE of
proteins extracted from intact cells eluates with 1 M KCl and (B) from purified cell walls. Arrowheads or circular heads indicate the spots
present exclusively in gel (A) and (B), respectively. The number of proteins considered as potential contaminants was estimated from the
number of spots exclusively present in (B). Gels were loaded with 40 mg of total protein.

Table 4. MALDI-TOF/TOF characterisation of spatially matched proteins in gels of saline eluates of intact callus and of purified callus cell
walls. The databases were queried with either MASCOT data files or the compiled de novo sequences. In all cases peptide charges
were 11

MASCOTa)

MS-Blastb)

Spot
No.

kDa

pI

Homologue

Accession
No.c)

Score

Matched
peptides

Score

Matched
peptides

122Ed)
122He)
123E
123H
124E
124H
125E
125H
126E
126H
127E
127H

16.2
16.2
32.4
32.6
39.2
39.2
46.6
46.8
57.3
57.5
53.8
53.8

3.8
3.8
5.6
5.6
6.5
6.5
5.8
5.8
5.5
5.5
5.7
5.7

PR-10
PR-10
Glycine rich protein
Glycine rich protein
Secretory peroxidase
Secretory peroxidase
Malate dehydrogenase
Malate dehydrogenase
Secretory peroxidase
Secretory peroxidase
Hypothetical protein
Hypothetical protein

BE239885
BE239885
Q39682
Q39682
BF644273
BF644273
BF006134
BF006134
AW559660
AW559660
CAE60064
CAE60064

266
93
2
2
241
218
140
178
156
99
2
2

3
1
2
2
3
3
2
2
1
1
2
2

2
2
416
272
2
2
2
2
2
2
117
244

2
2
4
3
2
2
2
2
2
2
3
1

a)
b)
c)
d)
e)

All the proteins were identified by using the Est-others database

All proteins were identified using the NCBInr data base.
Accession number for the homologue in NCBI database.
Protein isolated by surface elution of intact callus cells.
Protein isolated by saline extraction of purified cell walls.

molecular weight and/or pI (e.g. spots 105 and 109, spots 100
and 106; spots 43 and 45, spots 50 and 51; see also Tables 1
and 2). These observations are a common feature in 2-DE
gels [7, 20], possibly arising from heterogeneity in PTMs.
Other spots shared homology to the same protein families,
but with different corresponding gene sequences, indicating
the presence of multi-gene families in the apoplast.
3.4.1 IF proteins
Of the 45 IF protein spots (see Fig. 1A) selected for MALDITOF/TOF analysis, 19 showed homology with known proteins. For the remaining proteins, although it was possible to
2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

obtain de novo sequence information from good spectra, no

significant similarity was obtained to any protein in the current databases. Similarly low percentages have also been
reported for soluble apoplastic proteins in leaves of Oryza
sativa (39%) and Arabidopsis thaliana (56%) using a similar
2-DE approach [22].
Table 1 reports the list of the IF proteins identified. With
the exception of protein spot 111, which shows homology to
a glycine-rich protein, all are potentially defence-related,
including chitinases, thaumatin-like proteins, b-1,3 glucanases and a germin-like oxidoreductase. A protein with
similarity to a tobacco PR protein (NtPRp27) was also present (see Fig. 4A).
www.proteomics-journal.com

Plant Proteomics

Proteomics 2007, 7, 20702082

2079

mutase (2), thioredoxin (2) and others. A further major

group consisted of miscellaneous proteins (8), including a
protein with homology to a serine protease and synthases of
3-dehydroquinate, folypolyglutamate and S-adesonylmethionine. A fourth group (7) comprises proteins involved
in the transport and binding of solutes. In this group, four
out of seven were ABC transporter binding proteins. Proteins with homology to the glycoactive enzymes mannose
1-phosphate-guanyltransferase and an N-acetylglucosaminyl-transferase I were also detected.
Three glycine rich proteins, a putative lysine-rich cell
wall structural protein and an arabinogalactan-rich protein
comprised a group of wall structural proteins. Obvious
homology with defence-related proteins were limited to an
anti-freeze protein, PR-10, harpin-binding protein and a
putative stress-inducible protein. A further four proteins
showed homology with proteins containing important interaction domains.

Discussion

4.1 IB and IF proteins of leaves are largely distinct

populations

Figure 4. A functional classification of 45 proteins from the

intercellular fluid (A) and of 65 of the most abundant ionically
bound wall proteins (B). N.B. spots homologous to the small
subunit of rubisco (Table 2) were not included in this classification.

3.4.2 IB proteins
MALDI-TOF/TOF analyses of IB proteins revealed that 59
proteins demonstrate homology with known proteins,
whereas six have homology with unknown or hypothetical
proteins (Table 2). In addition, two spots had no homology
with any protein in the current data-base.
The results demonstrate that IB cell wall proteins of
Medicago are diverse in nature, and can be conveniently
divided into nine groups (see Fig. 4B). The most representative group is related with energy production and conversion (15). Energy production-related proteins included
those with homology with chloroplastic oxygen-evolving
enhancer proteins 1 and 3 (7). Proteins associated with
energy conversion include ATP synthase-like proteins (4),
triosephosphate isomerase (1) and D-3-phosphoglycerate
dehydrogenase (1). The second most abundant group
revealed homology with oxidoreductases (14), including
germin-like proteins (4), peroxidase (4), superoxide dis 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

It is generally understood that secreted apoplastic proteins

can be retained in the wall or liberated as soluble components of the IF. However, there is little information concerning the possible overlap of these two protein populations, or
the extent of their functional compartmentalisation. Physiological factors such as pH [50] or ionic content [33] of the
apoplastic fluid could influence the extent of wall retention
of particular proteins [33, 51], suggesting a dynamic relationship. Conversely, particular proteins may lack the appropriate surface motifs for retention and remain permanently
soluble.
We have compared the freely soluble/loosely-bound
group of apoplastic proteins (IF) with those retained in the
wall by electrostatic interaction (IB proteins). Interestingly,
the predominantly acidic pI of Arabidopsis rosette IF proteins led to the suggestion that low pI would favour protein solubility by reducing their electrostatic interaction
with matrix components [11]. However, we find that Medicago leaves contain a considerable number of both basic IF
and acidic IB proteins, which supports other studies [16,
20] which indicate the pI of an apoplastic wall protein is
unreliable as an indicator of its matrix retention or solubility.
Quantitatively, IB proteins were seen to be considerably
less abundant in the apoplast than IF proteins. Qualitatively,
spot patterns of IF and IB proteins from leaves were substantially different, sharing only 22 spatially matched spots
out of a total of 304. However, MALDI-TOF/TOF characterisation of seven putative spot matches validated the identity
of only a single match, suggesting the overlap could be considerably less (,7%).
www.proteomics-journal.com

2080

N. C. Soares et al.

4.2 The presence of non-canonical proteins in the

apoplast
The characterisation of selected IB proteins revealed them to
consist of several expected cell wall proteins, but also noncanonical proteins. High proportions of unexpected proteins
have been reported in many recent cell wall proteomic studies (up to 50% of proteins, [16, 19, 20, 52], 49%, this work),
which cannot easily be explained by contamination.
During the course of this work, Boudart et al. (2005) [53]
reported that the use of 0.3 M mannitol can reduce cytosolic
contamination which otherwise occurs during sequential,
vacuum-infiltrations for the isolation of IB proteins from
Arabidopsis. Here, we have extracted IB proteins from purified walls and demonstrated negligible contamination by
soluble, symplastic proteins with assays of malate dehydrogenase activity (Table 3) and by the presence of only trace
amounts of rubisco in our 2-DE gels (see spots 19 and 21
(Fig. 1B)), an alternative indicator of leaf wall contamination
[10]. However, the negatively charged plant cell wall [17] is
potentially susceptible to additional contamination from
charged, symplastic proteins during wall isolation [16, 18, 20,
54], for which there is no straight forward assay. Here we
have demonstrated that wall purification need not result in
significant contamination from such proteins, since .92%
of the proteins isolated from purified callus cell walls could
be verified in eluates from intact callus cells. Although the
content of IB proteins in callus and leaves are demonstrably
different (Figs. 1 and 3), it seems reasonable to suggest that
purified cell walls can provide an enriched and useful source
of IB proteins.
Interestingly, a protein with homology to malate dehydrogenase was isolated from both purified callus cell walls
and surface eluates of intact callus cells. Similarly, a protein
homologous to a plastocyanin was isolated from both purified leaf cell walls and saline infiltrates of leaves. Together,
these results support other reports which suggest that the
majority of proteins in saline extracts of purified walls are
directed there in vivo, and that these include non-canonical
proteins. [18].
4.3 Characterisation of leaf apoplastic proteins
A large number of the selected IF proteins did not show significant homology with any protein in the database. This
could be due to the incomplete sequencing of the Medicago
genome, although low percentages of identifiable proteins
(3956%) in the IF have also been reported in Oryza sativa
and Arabidopsis thaliana, which have been sequenced in
entirety [22]. Nevertheless, a representative sample of 19 IF
proteins (out of 45) allowed us to obtain a reasonable qualitative perspective of these proteins. The high number of
chitinases and b-1,3 glucanases present (Table 1) suggest the
Medicago IF is a highly glycolytic compartment [55] containing proteins with mainly defence-related functions, similar
to that observed in other species [21, 23, 24].
2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Proteomics 2007, 7, 20702082

The IB wall proteins described here were extracted with

1 M KCl, but nevertheless, show some similarity with CaCl2
and LiCl extracts of purified walls of M. sativa stem tissue
[20]. This earlier study identified 87 proteins with homology
to different data-base entries. Our 1 M KCl extracts contained
63 apparently unique IB proteins in Medicago leaves. It is of
interest that whereas Watson et al. (2004) [20] identified chitinases and b-1,3 glucanases ionically attached to the wall, we
find these proteins exclusively in the IF. In addition, our cell
wall extracts contain proteins not previously reported in the
Medicago cell wall, including glycine-rich proteins (3), an
arabinogalactan protein, ABC-type transporters (5) and a
number of proteins of varied functions (22). This work
therefore provides complementary information of the Medicago apoplast proteome.
The largest group of wall proteins were related with
energy production and conversion, including ATP synthaselike proteins, triosephosphate isomerase and D-3-phosphoglycerate dehydrogenase. ATP is now recognised as an
essential metabolite in cell walls [56] and the occurrence of
protein phosphorylation events in the apoplast is gaining
support from a growing number of cell wall proteomic studies [8, 12, 20, 56], including the demonstration of D-3phosphoglycerate dehydrogenase, also identified in this
study, in the maize cell wall [7]. These apoplastic proteins are
thought to be involved in extracellular phosphorylation networks leading to the regulation of many genes in response to
changes in the environment. Potentially related proteins
include two identified ATP-binding ABC-type transporters
family. These have been located in the cell walls of the algae
Haematococcus pluvialis [57] and are thought to mediate ATP
release from cells [5860].
Some IB proteins demonstrated partial homology with
oxygen evolving enhancer (OEE) subunits (1 and 3), which
have been previously detected in the Medicago apoplast [49].
These OEE subunits are associated with PSII in the chloroplastic lumen and are highly conserved. Interestingly, the
homology of proteins with OEE1 represents far less coverage
(spot 28 presents only 5% and spot 67 has only 38% matching coverage) than would be expected for this highly conserved protein, and the remaining peptide sequences
showed no homology with any entry in the data-base. This
strongly suggests the ECM contains some chimeric proteins
with OEE-like domains.
Another OEE subunit, OEE2, is known to be phosphorylated by wall-associated kinase 1 in a manner regulated by
the extracellular and glycine-rich protein AtGRP-3, and it has
been suggested that WAK1/AtGRP-3 phosphorylation of
OEE2 leads to the induction of defence genes [61]. Hence, it
is tempting to suggest that these novel cell wall proteins
share a conserved OEE domain and are involved in wall-signalling events.
Defence-related proteins were predominate in the IF, but
were also found ionically bound to the cell wall. A harpinbinding protein was detected, which is thought to interact
with harpins of pathogenic bacterial pili to initiate the
www.proteomics-journal.com

Proteomics 2007, 7, 20702082

hypersensitive response in many species [62]. Proteins homologous to a type I anti-freeze, PR-10 and a stress-inducible
protein were also present. Several oxidoreductases were
identified, including SOD, peroxidase and germin-like proteins, which could contribute to the regulation and/or generation of ROS during the oxidative burst, thereby contributing to the mobilisation of defence genes [63].
4.4 Concluding remarks
In summary, we have demonstrated that purified cell walls
can be a convenient source of IB proteins which can be isolated with minimal contamination, yet which contain noncanonical proteins. Our comparison of IF and IB proteins
indicated them to be largely distinct protein populations.
With the exception of defence-related proteins which are
present in both IF and IB proteins, these two protein fractions appear to represent distinct functional compartments
of the apoplast.
This study therefore suggests that the combined study of
IF and IB proteins will enable a more comprehensive overview of the role of apoplastic proteins in plant processes.

Plant Proteomics

2081

[12] Kwon, H.-K., Yokoyama, R., Nishitani, K., Plant Cell Physiol.
2005, 46, 843857.
[13] Boudart, G., Jamet, E., Rossignol, M., Lafitte, C. et al., Proteomics 2005, 5, 212221.
[14] Zhu, J., Chen, S., Alvarez, S., Asirvatham, V. S. et al., Plant
Physiol. 2006, 140, 311325.
[15] Jackson, P., Paulo, S., Ricardo, C. P. P., in: Obinger, C., Burner,
U., Ebermann, R., Penel, C., Greppins, H. (Eds.), Plant peroxidases, Biochemistry and Physiology, Universit de Genve press, Vienna 1996, pp. 255258.
[16] Chivasa, S., Ndimba, B. K., Simon, W. J., Robertson, D. et al.,
Electrophoresis 2002, 23, 17541765.
[17] Shomer, I., Novacky, A. J., Pike, S. M., Yermiyahu, U. et al.,
Plant Physiol. 2003, 133, 411422.
[18] Bayer, E. M., Bottrill, A. R., Walshaw, J., Vigouroux, M. et al.,
Proteomics 2006, 6, 301311.
[19] Slabas, A. R., Ndimba, B., Simon, W. J., Chivasa, S., Biochem. Soc. Trans. 2004, 32, 524528.
[20] Watson, B. S., Lei, Z., Dixon, R. A., Sumner, L. W., Phytochemistry 2004, 65, 17091720.
[21] Regalado, A. P., Ricardo, C. P. P., Plant Physiol. 1996, 110,
227232.
[22] Haslam, R. P., Downie, A. L., Raveton, M., Gallardo, K. et al.,
Ann. Appl. Biol. 2003, 143, 8191.
[23] Will, T., van Bel, A. J. E., J. Exp. Bot. 2006, 57, 729737.

Nelson C. Soares acknowledges a personal grant from the

Fundaco de Cincia e Tecnologia (ref. SFRH/ BD/ 6487/
2001). This work was also supported by the Instituto de Biologia
Experimental e Tecnolgica within the Plant-IBET protocol. We
gratefully acknowledge the gift of Medicago leaf callus from Prof.
Pedro Fevereiro (ITQB).

[24] Stintzi, A., Heitz, T., Prasad, V., Wiedemann-Merdinoglu, S.

et al., Biochimie 1993, 75, 687706.
[25] Bonello, J.-F., Sevilla-Lecoq, S., Berne, A., Risueno, M.-C. et
al., J. Exp. Bot. 2002, 53, 15591568.
[26] Cosgrove, D. J., Plant Physiol. 2001, 125, 131134.
[27] Jackson, P. A., Galinha, C. I. R., Pereira, C. S., Fortunato, A. et
al., Plant Physiol. 2001, 127, 10651076.
[28] Majewska-Sawka, A., Nothnagel, E. A., Plant Physiol. 2000,
122, 310.

References

[1] Osato, Y., Yokoyama, R., Nishitani, K., J. Plant Res. 2006, 119,
153162.
[2] Price, N. J., Pinheiro, C., Soares, C. M., Ashford, D. A. et al., J.
Biol. Chem. 2003, 278, 4138941399.

[29] Carrington, C., Greve, L. C., Labavitch, J. M., Plant Physiol.

1993, 103, 429434.
[30] Jackson, P., Ricardo, C. P. P., Aust. J. Plant Physiol. 1998, 25,
261269.
[31] Barcel, A. R., Pedreo, A., Muoz, R., Sabater, F., Physiol.
Plant. 1989, 75, 267274.

[3] Hall, Q., Cannon, M. C., Plant Cell 2002, 14, 11611172.

[32] MacDougall, A., Brett, G., Morris, V., Rigby, N. et al., Carbohydr. Res. 2001, 335, 115126.

[4] Rojo, E., Sharma, V. K., Kovaleva, V., Raikhel, N. V. et al.,

Plant Cell 2002, 14, 969977.

[33] MacDougall, A. J., Parker, R., Selvendran, R. R., Plant Physiol. 1995, 108, 16791689.

[5] Irifune, K., Nishida, T., Egawa, H., Nagatani, A., Plant Cell
Rep. 2004, 23, 333338.

[34] Willats, W. G. T., Orfila, C., Limberg, G., Buchholt, H. C. et al.,

J. Biol. Chem. 2001, 276, 1940419413.

[6] Rose, J. K. C., Hadfield, K. A., Labavitch, J. M., Bennett, A. B.,

Plant Physiol. 1998, 117, 345361.

[35] Oh, I. S., Park, A. R., Bae, M. S., Kwon, S. J. et al., Plant Cell
2005, 17, 28322847.

[7] Chivasa, S., Simon, W. J., Yu, X.-L., Yalpani, N. et al., Proteomics 2005, 5, 48944904.

[36] Asirvatham, V., Watson, B., Sumner, L., Proteomics 2002, 2,

960968.

[8] Ndimba, B. K., Chivasa, S., Hamilton, J. M., Simon, W. J. et

al., Proteomics 2003, 3, 10471059.

[37] Imin, N., De Jong, F., Mathesius, U., van Noorden, G. et al.,
Proteomics 2004, 4, 18831896.

[9] Dani, V., Simon, W. J., Duranti, M., Croy, R. R. D., Proteomics
2005, 5, 737745.

[38] Gallardo, K., Le Signor, C., Vandekerckhove, J., Thompson,

R. D. et al., Plant Physiol. 2003, 133, 664682.

[10] Lee, S.-J., Saravanan, R. S., Damasceno, C. M. B., Yamane,

H. et al., Plant Physiol. Biochem. 2004, 42, 979988.

[39] Watson, B. S., Asirvatham, V. S., Wang, L., Sumner, L. W.,

Plant Physiol. 2003, 131, 11041123.

[11] Borderies, G., Jamet, E., Claude, L., Miche, l. R. et al., Electrophoresis 2003, 24, 34213432.

[40] Murashige, T., Skoog, F., Physiol. Plantarum 1962, 5, 473

497.

2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.proteomics-journal.com

2082

N. C. Soares et al.

Proteomics 2007, 7, 20702082

[41] Jackson, P., Paulo, S., Brownleader, M. D., Freire, P. et al.,

Aust. J. Plant Physiol. 1999, 26, 313326.

[54] Jamet, E., Canut, H., Boudart, G., Pont-Lezica, R. F., Trends
Plant Sci. 2006, 11, 3339.

[42] Lopez-Millan, A. F., Morales, F., Abadia, A., Abadia, J., Plant
Physiol. 2000, 124, 873884.

[55] Dietz, K.-J., Sauter, A., Wichert, K., Messdaghi, D. et al., J.

Exp. Bot. 2000, 51, 937944.

[43] Bradford, M., Anal. Biochem. 1976, 72, 248254.

[56] Chivasa, S., Ndimba, B. K., Simon, W. J., Lindsey, K. et al.,

Plant Cell 2005, 17, 30193034.

[44] Ramagli, L., Methods Mol. Biol. 1999, 112, 99103.

[45] Laemmli, U. K., Nature 1970, 227, 680685.
[46] Blum, H., Beier, H., Gross, H., Electrophoresis 1987, 8, 9399.
[47] Pandey, A., Andersen, J., Mann, M., Sci. STKE 2000, 2000,
PL1.
[48] Shevchenko, A., Sunyaev, S., Loboda, A., Bork, P. et al., Anal.
Chem. 2001, 73, 19171926.
[49] Abbasi, F., Komatsu, S., Proteomics 2004, 4, 20722081.
[50] Cosgrove, D., Plant Mol. Biol. 1999, 50, 391417.
[51] Gao, D., Knight, M. R., Trewavas, A. J., Sattelmacher, B. et
al., Plant Physiol. 2004, 134, 898908.
[52] Blee, K. A., Wheatley, E. R., Bonham, V. A., Mitchell, G. P. et
al., Planta 2001, 212, 404415.
[53] Aebersold, R., Mann, M., 2003, 422, 198207.

2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

[57] Wang, S.-B., Hu, Q., Sommerfeld, M., Chen, F., Proteomics
2004, 4, 692708.
[58] Roman, R. M., Wang, Y., Lidofsky, S. D., Feranchak, A. P. et
al., J. Biol. Chem. 1997, 272, 2197021976.
[59] Abraham, E., Prat, A., Gerweck, L., Seneveratne, T. et al.,
PNAS 1993, 90, 312316.
[60] Thomas, C., Rajagopal, A., Windsor, B., Dudler, R. et al.,
Plant Cell 2000, 12, 519534.
[61] Yang, E. J., Oh, Y. A., Lee, E. S., Park, A. R. et al., Biochem.
Biophys. Res. Commun. 2003, 305, 862868.
[62] Li, C.-M., Brown, I., Mansfield, J., Stevens, C. et al., EMBO J.
2002, 21, 19091915.
[63] Lamb, C., Dixon, R., Annu. Rev. Plant Physiol. Plant Mol.
Biol. 1997, 48, 251275.

www.proteomics-journal.com