Báo cáo khoa học: Identification, cloning and characterization of two thioredoxin h isoforms, HvTrxh1 and HvTrxh2, from the barley seed proteome pot

Identification, cloning and characterization of two thioredoxin
h isoforms, HvTrxh1 and HvTrxh2, from the barley seed proteome
Kenji Maeda, Christine Finnie, Ole Østergaard and Birte Svensson
Department of Chemistry, Carlsberg Laboratory, Copenhagen, Denmark
Two thioredoxin h isoforms, HvTrxh1 and HvTrxh2, were
identified in two and one spots, respectively, in a proteome
analysis of barley (Hordeum vulgare) seeds based on 2D gel
electrophoresis and MS. HvTrxh1 was observed in 2D gel
patterns of endosperm, aleurone layer and embryo of
mature barley seeds, and HvTrxh2 was present mainly in the
embryo. During germination, HvTrxh2 decreased in abun-
dance and HvTrxh1 decreased in the aleurone layer and
endosperm but remained at high levels in the embryo. On the
basis of MS identification of the two isoforms, expressed
sequence tag sequences were identified, and cDNAs enco-
ding HvTrxh1 and HvTrxh2 were cloned by RT-PCR. The
sequences were 51% identical, but showed higer similarity
to thioredoxin h isoforms from other cereals, e.g. rice Trxh
(74% identical with HvTrxh1) and wheat TrxTa (90%
identical with HvTrxh2). Recombinant HvTrxh1, HvTrxh2
and TrxTa were produced in Escherichia coli and purified
using a three-step procedure. The activity of the purified
recombinant thioredoxin h isoforms was demonstrated
using insulin and barley a-amylase/subtilisin inhibitor as
substrates. HvTrxh1 and HvTrxh2 were also efficiently
reduced by Arabidopsis thaliana NADP-dependent thio-
redoxin reductase (NTR). The biochemical properties of
HvTrxh2 and TrxTa were similar, whereas HvTrxh1 had
higher insulin-reducing activity and was a better substrate
for Arabidopsis NTR than HvTrxh2, with a K
m

of 13 l
M
compared with 44 l
M
for HvTrxh2. Thus, barley seeds
contain two distinct thioredoxin h isoforms which differ
in temporal and spatial distribution and kinetic proper-
ties, suggesting that they may have different physiological
roles.
Keywords: barley seed; disulfide reduction; proteomics;
recombinant proteins; thioredoxin h.
Thioredoxins are protein disulfide reductases of molecular
mass  12 kDa [1]. The conserved active-site sequence
WCGPC forms a disulfide bond in the oxidized form of the
protein. The reduced, dithiol form of thioredoxin can
modulate the activity of a variety of target proteins by
reduction of their disulfide bonds. Plants contain several
forms of thioredoxin which differ in their subcellular
localization and thus in the target proteins with which they
interact [2,3]. Thioredoxins f and m are found in chloro-
plasts and are involved in regulation of photosynthetic
enzymes [4], whereas thioredoxin h is primarily cytosolic,
and has also been identified in rice as one of the major
protein components of phloem sap [5]. A mitochondrial
thioredoxin system has been identified in Arabidopsis [6].
The catalytic system comprises, in addition to thioredoxin,
an electron donor and a thioredoxin reductase, which are
required for regeneration of the reduced form of thio-
redoxin. Thioredoxins f and m are reduced by ferredoxin
via ferredoxin-dependent thioredoxin reductase [2]. Thio-

redoxin h is reduced by NADPH via NADP-dependent
thioredoxin reductase (NTR) [7].
Thioredoxin h has been found to have an important
influence on seed germination in barley and other plants
[8,9]. Among the known target proteins of thioredoxin h in
seeds are storage proteins such as hordeins in barley and
glutenins and gliadins in wheat, which are deposited in
disulfide-bound complexes and are mobilized during the
germination process. Reduction by thioredoxin renders
them more soluble and susceptible to proteolytic degrada-
tion [10]. Some a-amylase/trypsin inhibitor proteins [11]
are also known targets of thioredoxin h, as is the barley
a-amylase/subtilisin inhibitor (BASI) [12]. Studies of germi-
nating transgenic barley seeds overexpressing wheat thio-
redoxin h revealed that limit dextrinase activity was
increased [13], and that a-amylase activity increased earlier
than in normal seeds [14]. This implicates thioredoxin h in
regulation of the mobilization of starch reserves during seed
germination.
Multiple forms of thioredoxin h exist in plants; at least
five isoforms have been identified in Arabidopsis [15] and
three in wheat [16,17]. Whether these isoforms have
different specificities or functions in the plant is not known.
However, yeast complementation studies have provided
evidence for different target specificities of Arabidopsis
Correspondence to C. Finnie, Department of Chemistry, Carlsberg
Laboratory, Gamle Carlsberg Vej 10, DK-2500 Valby, Copenhagen,
Denmark. Fax: + 45 33274708, Tel.: + 45 33275304,
E-mail:
Abbreviations: BASI, barley a-amylase/subtilisin inhibitor; DTNB,

5,5¢-dithio-bis-(2-nitrobenzoic acid); EST, expressed sequence tag;
NTR, NADP-dependent thioredoxin reductase; TC, tentative
consensus; TIGR, The Institute for Genomic Research.
Proteins and enzymes: Arabidopsis NADP-dependent thioredoxin
reductase (Q39243) (EC 1.8.1.9); barley a-amylase/subtilisin inhibitor
(P07596); wheat thioredoxin h TrxTa (O64394); barley thioredoxin h1
HvTrxh1 (AY245454); barley thioredoxin h2 HvTrxh2 (AY245455).
(Received 10 March 2003, revised 23 April 2003,
accepted 28 April 2003)
Eur. J. Biochem. 270, 2633–2643 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03637.x
thioredoxin h isoforms [18]. Three forms of thioredoxin h
have been observed in Western blots with barley seed
extracts [9], and a gene has been identified in the wild barley
Hordeum bulbosum that resembles a subset of thioredoxin h
sequences [19]. However, until now, barley thioredoxin h
has not been characterized at the protein level and target
proteins in barley have been identified using thioredoxins
from other organisms.
Proteomics, based on the techniques of 2D gel electro-
phoresis and MS, offers the opportunity to study the
appearance patterns of many proteins simultaneously, and
to identify proteins of interest. Recently, these techniques
have been applied to the study of seed development and
germination in several plants including Arabidopsis [20,21],
wheat [22] and barley [23–25]. We used 2D gel electropho-
resis to identify thioredoxin h forms in barley seeds and
characterize their patterns of appearance in the seed tissues
and during germination. Identification of thioredoxin h
forms in mature seed extracts by MS peptide mapping
enabled cloning of the corresponding genes encoding two

thioredoxin h forms with different appearance patterns.
Recombinant proteins were produced and found also to
differ in their biochemical properties. This is the first
characterization of barley thioredoxin h proteins, and this
comparative study of their properties extends our know-
ledge of the cereal thioredoxin h family.
Materials and methods
Materials
Rabbit antibody to wheat thioredoxin h was kindly sup-
plied by B. Buchanan (UC Berkeley, CA, USA). Secondary
antibodies were from Dako A/S. Primers were from DNA
Technology (Aarhus, Denmark). Pfu DNA polymerase
and BL21(DE3)Gold were from Stratagene. Restriction
enzymes, DNA ligase and BL21(DE3)pLysS were from
Promega. The pETtrxTa expression system was kindly
provided by M. Gautier (INRA, Montpellier, France).
Purified Arabidopsis thaliana NTR and Populus tremula
thioredoxin h were kindly supplied by J P. Jacquot (INRA,
Nancy, France). Bovine pancreas insulin, monobromo-
bimane, isopropyl thio-b-
D
-galactoside, 5,5¢-dithiobis-(2-
nitrobenzoic acid) (DTNB) and NADPH were from Fluka.
Plant material and protein extraction
Spring barley (Hordeum vulgare cv. Barke) was field grown
in Fyn, Denmark, in the 2000 season. Seeds were micro-
malted according to standard procedures [23]. Micromalted
seeds were frozen in liquid nitrogen and freeze-dried before
milling and extraction.
Barley seeds were dissected as previously described [25].

Dissected tissues from five seeds were freeze-dried before
extraction. Proteins were extracted from 4 g milled seeds in
20 mL extraction buffer (5 m
M
Tris/HCl, pH 7.5, 1 m
M
CaCl
2
)for30minat4°C, as previously described [24].
Dissected tissues from five seeds were extracted in the same
buffer for 30 min at 4 °C as previously described [25]. After
centrifugation to remove debris, supernatants containing
soluble proteins were transferred to clean tubes and stored
at )80 °C until required.
2D gel electrophoresis
Proteins contained in 250 lLand100lL mature seed
extract were applied to the gels for colloidal Coomassie
staining and 2D Western blotting, respectively. Duplicate
gels were run containing proteins from 100 lLdissected
seed extracts. By loading equal volumes, a similar ratio
is seen between the proteins from each tissue on the
dissected seed gels as the whole seed gels [25]. Germi-
nated seed gels were also loaded with proteins from an
equal volume of extract (100 lL), as the soluble protein
content of germinated seeds is increased as the result of
mobilized storage proteins. Thus, by loading equal
volumes, proteins that remain at a constant level during
germination will also appear at similar intensity on the
2D gels.
Proteins were precipitated with 4 vol. acetone at )20 °C

for 24 h and resuspended in reswelling buffer [8
M
urea, 2%
(w/v) CHAPS, 0.5% (v/v) IPG buffer 4–7 (Amersham
Biosciences), 20 m
M
dithiothreitol and a trace of bromo-
phenol blue]. IEF was carried out using 18 cm immobilized
linear pH gradient (IPG) strips, pI 4–7, run on an IPGphor
(Amersham Biosciences) as previously described [24].
Second-dimension SDS/polyacrylamide gels (12–14%,
18 · 24 cm; Amersham Biosciences) were run on a Phar-
macia Multiphor II according to the manufacturer’s
recommendations. Gels were stained with silver nitrate
[26] or colloidal Coomassie blue [27].
For immunodetection of thioredoxin h, the 2D gel was
blotted on to a nitrocellulose membrane in 10 m
M
CAPS,
pH 11.0, using a Multiphor II NovaBlot unit (Amersham
Biosciences). Immunodetection was carried out according
to standard protocols, using a 1 : 2000 dilution of rabbit
anti-(wheat thioredoxin h) serum. Goat anti-rabbit horse-
radish peroxidase-conjugated secondary IgGs were used at
a 1 : 5000 dilution, and the signal was detected by enhanced
chemiluminescence [28].
In-gel digestion and MALDI-TOF MS
Spots were excised from the colloidal Coomassie-stained gel
and subjected to in-gel trypsin digestion [29]. Tryptic
peptides were desalted and concentrated on a home-made

5 mm nano-column [30] as previously described [24].
Peptides were eluted with 0.8 lL matrix (20 mgÆmL
)1
a-cyano-4-hydroxycinnamic acid in 70% acetonitrile/0.1%
trifluoroacetate) and deposited directly on to the MALDI
target.
A Bruker REFLEX III MALDI-TOF mass spectrometer
(Bruker-Daltonics, Bremen, Germany) in positive ion
reflector mode was used to analyse tryptic peptides. The
m/z software (Proteometrics, New York, NY, USA) was
used to analyse spectra. Spectra were calibrated using
trypsin autolysis products (m/z 842.51 and m/z 2211.10) as
internal standards. To identify proteins, the SwissProt and
NCBI nonredundant sequence databases and the NCBI
expressed sequence tag (EST) databases were searched with
peptide masses using the Mascot (rix
science.com) server. Tentative consensus (TC) sequences
corresponding to identified EST sequences were obtained by
searching the Institute for Genomic Research (TIGR)
sequence database (www.tigr.org/tdb/tgi/hvgi).
2634 K. Maeda et al.(Eur. J. Biochem. 270) Ó FEBS 2003
Cloning and sequencing of barley thioredoxin h isoforms
Embryos were dissected from five seeds after 24 h micro-
malting, and total RNA was extracted using the RNAeasy
kit (Qiagen) according to the manufacturer’s recommenda-
tions, for use as a template in RT-PCR.
Hvtrxh1. The coding sequence of HvTrxh1 was amplified
by RT-PCR from barley embryo RNA using the primers
trxh8 (TT
CATATGGCCGCCGAGGAGGGAG) and

trxh9 (GG
GGATCCTAACCGGGCAATCACTCTTC).
The primers were designed on the basis of the sequence of
TC44851 and introduce an NdeI restriction site (underlined
in trxh8) at the start codon (bold) and a BamHI restriction
site (underlined in trxh9) after the stop codon. Reverse
transcription and the following PCR was carried out on
total RNA using an RT-PCR kit (Qiagen) and a PTC-200
Peltier Thermal Cycler (MJ Research). The RT-PCR
product was cloned into pCR4-TOPO (Invitrogen) to give
pCR-h1.
Hvtrxh2. Primers trxh5 (TTGAATTCGCGTGAGAAA
TAAGCCGAGT) and trxh6 (TTCTGCAGTCTTCTT
GAGAGGACCTTTT), based on TC45680, were used for
amplification of the HvTrx2 coding sequence as above. A
second PCR with Pfu DNA polymerase and primers trxh1
(TT
CATATGGCGGCGTCGGCAACGGCG) and trxh2
(GG
GGATCCTGAGCGGCAATTTTATTTAGGCG)
was used to introduce an NdeI restriction site (underlined in
trxh1) at the start codon (bold) and a BamHI restriction site
(underlined in trxh2) after the stop codon of HvTrxh2. The
resulting PCR product was cloned into pCR Blunt II–
TOPO (Invitrogen) to give pCR-h2.
Construction of expression vectors. Inserts were isolated
from pCR-h1 and pCR-h2 by digestion with Nde1and
BamHI and ligated into the pET11a expression vector
linearized with Nde1andBamHI, to give pETHvTrxh1 and
pETHvTrxh2, respectively. The sequences of the inserts

were determined on both strands and found to be as
expected from the identified TC sequences, and confirmed
that the cloning junctions were correct. Accession numbers
for HvTrxh1 and HvTrxh2 sequences are AY245454 and
AY245455, respectively.
Expression and purification of recombinant barley
thioredoxin isoforms
Saturated cultures of Escherichia coli BL21(DE3)Gold
transformed with pETHvTrxh1 or pETHvTrxh2 were
diluted 100-fold in 2 L Luria–Bertani medium and grown
at 37 °C. After reaching an A
600
of 0.6, the cultures were
induced with 100 l
M
isopropyl thio-b-
D
-galactoside for 3 h.
Cells were harvested and stored at )20 °C until use. With
the same procedure, wheat TrxTa was expressed in
BL21(DE3)pLysS transformed with pETtrxTa [17].
HvTrxh1, HvTrxh2 and TrxTa were purified by a
procedure for TrxTa [17] with minor modifications. Har-
vested cells were resuspended in 100 mL 50 m
M
Tris/HCl/
1m
M
EDTA, pH 8.0 and lysed by passage three times
through a French Press. The lysate was sonicated to shear

nucleic acids and centrifuged to remove insoluble material.
The supernantant was heat-treated for 7 min at 65 °Cand
centrifuged to remove aggregated material. The supernatant
was filtered and loaded on a HiLoad 26/10 Q Sepharose
High Performance column (Amersham Biosciences) equi-
librated with 30 m
M
Tris/HCl, pH 8.0. Proteins were eluted
by a linear gradient from 0 to 700 m
M
NaCl in the same
buffer. Thioredoxin h-containing fractions were detected by
dot blotting 0.5 lL on to a nitrocellulose membrane.
Immunodetection was carried out as described above,
except that an alkaline phosphatase-conjugated swine anti-
rabbit secondary IgG was used, and signal was detected
using a 5-bromo-4-chloro-3-indolyl phosphate/nitroblue
tetrazolium tablet (Sigma Chemical, St Louis, MO, USA).
Thioredoxin-containing fractions were pooled and concen-
trated using a Centriprep YM3 (Millipore, Bedford, MA,
USA) to  2 mL. The concentrated samples were then
loaded on a HiLoad 16/60 Superdex 75 prep grad column
(Amersham Biosciences) equilibrated with 30 m
M
Tris/HCl,
pH 8.0, and eluted at a flow rate 0.2 mLÆmin
)1
. Thio-
redoxin-containing fractions were detected by dot-blotting
as above and pooled.

Purified proteins were quantified by amino-acid analysis.
Amino-acid amounts in protein hydrolysates were deter-
mined using Biochrom 20 (Pharmacia Biotech).
N-Terminal sequencing was performed using a 477A
protein sequencer equipped with a 120A phenylthiohydan-
toin analyzer (Applied Biosystems). Liquid chromatogra-
phy MS spectra of  100 pmol intact purified proteins were
obtained using an HP 1100 LC/MSD (Hewlett-Packard).
For SDS/PAGE, 3 lg HvTrxh1 or HvTrx2 was loaded on
to a NuPAGE Bis/Tris 4–12% gel (Invitrogen) in the
presence of 0.5 m
M
dithiothreitol. For native IEF, 1 lg
HvTrxh1 or HvTrxh2 was loaded on to an IEF PhastGel
covering the range pI 4.0–6.5 (Amersham Biosciences).
Molecular modelling and sequence analysis
The 3D structures of HvTrxh1 and HvTrxh2 were modelled
using the SWISS-MODEL server (ssmodel.
unibas.ch/) and Swiss PDB-viewer [31], with the crystal
structure of thioredoxin h from Chlamydomonas reinhardtii
(PDB accession 1EP7) [32] as a template. The modelled
structures covered residues 5–115 and 11–121 of HvTrxh1
and HvTrxh2, respectively.
The seqboot, protpars and consense programs in the
PHYLIP
3.5 package [33] were used to generate an unrooted
consensus tree based on an alignment of 46 plant thio-
redoxin h sequences from the NCBI sequence database. In
addition to HvTrxh1 and HvTrxh2, sequences were from
Arabidopsis thaliana (Ata, Q39241; Atb, NP_175128; Atc,

NP_199112; Atd, S58119; Ate, NP_173403; Atf, Q8L907;
Atg, Q39239; Ath, NP_190672; Ati, NP_188415), Brassica
napus (Bna, Q42388; Bnb, Q39362), Brassica oleracea
(Q9FQ63), Brassica rapa (Bra, O64432; Brb, Q8GZT3),
Curcurbita maxima (Q8H9E2), Fagopyrum esculentum (Fe,
Q96419), H. bulbosum (Hb, T50864), Hordeum vulgare
(Hva, Q8GZR4), Leymus chinensis (Lc, AAO16555),
Lolium perenne (Lp, T50865), Nicotiana tabacum (Nta,
Q8H6 · 3); Ntb, Q07090; Ntc, P29449), Oryza sativa (Osa,
Q42443; Osb,Q9FRT3;Osc, AAO37523; Osd,Q9AS75;
Ose,Q8H6· 4), Phalaris coerulescens (Pc, T50862), Picea
mariana (Pm, O65049), Prunus persica (Pp, Q93WZ3),
Ó FEBS 2003 Barley seed thioredoxin h isoforms (Eur. J. Biochem. 270) 2635
Pisum sativum (Psa, Q8GUR8; Psb, Q9AR82; Psc,
Q93 · 24; Psd, Q8GUR9), Populus tremula (Pt, Q8S3L3),
Ricinus communis (Rc, Q43636), Secale cereale (Sc, T50863),
Triticum aestivum (Taa, Q8GVD3; Tab, O64394; Tac,
Q9LDX4; Tad,Q8H6· 0) Triticum turgidum ssp. durum
(Td, O64395) and Zea mays (Zm, Q8H6 · 5). SwissProt/
TREMBL accession numbers are given where available,
otherwise EMBL/GenBank accession numbers are used.
Enzyme assays
Insulin assay [34]. A 1-mL reaction mixture contained
100 m
M
potassium phosphate, pH 7.0, 0.2 m
M
EDTA,
1 mg insulin, 1 l
M

HvTrxh1, HvTrxh2, or TrxTa, and
0.33 m
M
dithiothreitol. The reactions were initiated by
addition of dithiothreitol. Reactions proceeded at room
temperature (22 °C) and were followed by measuring A
650
on a Lambda 2 spectrophotometer (Perkin–Elmer).
NTR assay [35]. A 200-lL reaction mixture contained
50 m
M
Tris/HCl, pH 8.0, 150 l
M
NADPH, 100 l
M
DTNB, 77 n
M
A. thaliana NTR [36], and 1–16 l
M
HvTrxh1, HvTrxh2, TrxTa, or Thioredoxin h-1 from
P. tremulus [35]. Reactions proceeded at room temperature
and were followed by measuring A
405
on a MRX Revelation
absorbance reader (Dynex Technologies).
Reduction of BASI. A 50-lL reaction mixture contained
50 m
M
Tris/HCl, pH 7.5, 0.6 m
M

NADPH, 3.9 lg Arabi-
dopsis NTR and either 10 lg recombinant BASI [37] or
10 lg insulin. The mixtures were incubated for 10 min at
room temperature with 0.4 nmol HvTrxh1 or HvTrxh2.
Free thiol groups were then labelled by the addition of
0.2 lmol monobromobimane in 10 lL acetonitrile. After
incubation for 10 min at room temperature, proteins were
precipitated with 80% (v/v) acetone before being loaded on
a NuPAGE Bis-Tris 4–12% gel (Invitrogen). The labelled
proteins were visualized by being photographed under near
UV light.
Results
Identification of thioredoxin h in the barley seed
proteome
To locate spots containing thioredoxin h in the barley seed
proteome, a 2D gel with proteins extracted from mature
barley seeds was electroblotted and probed with antibodies
raised against wheat thioredoxin h. The antibody recog-
nized three spots with an approximate molecular mass of
12 kDa and pI 5.0 (Fig. 1A).
Protein spots excised from a colloidal Coomassie-stained
gel of mature barley seed proteins were digested with trypsin
and analysed by peptide mass mapping using MALDI-TOF
MS. From the position of the spots recognized by the
antibody, spots 296, 297 and 313 (Fig. 1A) were considered
likely to contain thioredoxin h. The approximate molecular
mass and pI of the proteins in these spots were determined
from their positions on the 2D gel to be 12.2 kDa/pI 5.0,
12.5 kDa/pI 5.0 and 11.3 kDa/pI 5.0, respectively. Peptide
mass data obtained for these spots did not lead to

identifications in searches of the NCBI or SwissProt
nonredundant databases, suggesting that the sequence data
for these proteins were not present. However, by searching
the NCBI EST sequence database, matches were obtained
against barley EST accessions BE230983 (both spots 296
and 297) and BF626734 (spot 313).
BLAST
searches using
these EST sequences indicated that both encoded proteins
with homology to thioredoxin h.
As ESTs can contain sequence errors and may not be full-
length, the matched EST sequences were used to search the
TIGR database of TC sequences from barley. TC sequences
were identified that contained the EST sequences and
apparently encoded full-length proteins. This was supported
by the fact that in-frame stop codons were present upstream
of the ATG start codon in each case. The thioredoxin
isoforms predicted to be encoded by these sequences were
designated HvTrxh1 (TC44851; corresponding to spots 296
and 297) and HvTrxh2 (TC45680; spot 313).
Theoretical tryptic digests of these sequences were used to
determine the sequence coverage obtained from peptide
mass mapping of the three spots. Peptide mass data from
spot 313 resulted in 39% sequence coverage with seven
matched peptides (Fig. 2A). Another TC sequence
(TC45681) also matched the peptide masses from spot
313; this encoded a single amino-acid substitution at the
C-terminus of the protein (A119 to G). As the peptide
covering this region was not observed in the mass spectrum,
it was not possible to distinguish between these almost

identical variants. The peptide maps obtained for spots 296
and 297 were highly similar. The peptides matching the
barley thioredoxin h TC sequence were the same in both
cases, resulting in 64% sequence coverage with 12 matched
peptides (Fig. 2A). It was therefore not possible from the
peptide mass data to explain the molecular mass difference
between the two spots on the 2D gel.
Although the same EST sequence was matched for
both spots, it is possible that spots 296 and 297 contain
thioredoxin h isoforms with sequence differences in
regions not covered by the peptide maps. However, no
other barley TC sequences were found in the TIGR
database that matched the peptide mass data for these
spots. A few peptides originating from another protein
(barley dimeric a-amylase inhibitor; P13691) were present
in the spectrum obtained for spot 297 and in smaller
amounts in the spectrum for spot 296. A spot containing
this protein (O. Østergaard, C. Finnie, S. Melchior,
P. Roepstorff & B. Svensson, unpublished results) forms a
horizontal smear overlapping with spot 297 (Fig. 1A, spot
318; the smear is particularly noticeable in the silver-
stained gel), and this electrophoretic smear is probably
the source of the peptides of barley dimeric a-amylase
inhibitor. However, most of the protein in spots 297 and
296 was found to be thioredoxin h.
Previously, evidence has been presented for variation in
the distribution of thioredoxin h forms in seed tissues
[9,16]. To obtain more detailed information about the
origins of HvTrxh1 and HvTrxh2, the occurrence of the
thioredoxin h spots in protein extracts from dissected

barley seeds was analysed by 2D gel electrophoresis
(Fig. 1B). The three thioredoxin h spots are distributed
differently in the tissues of mature seeds (Fig. 1B). Spot
297 is present in extracts made from dissected endosperm,
aleurone layer and embryo. Spot 296 is present in
2636 K. Maeda et al.(Eur. J. Biochem. 270) Ó FEBS 2003
endosperm and embryo, but is much less abundant in
aleurone layer extracts. Spot 313 (containing HvTrxh2) is
more abundant in extracts from the embryo than from
the other tissues. In mature seeds, all three forms of
thioredoxin h were observed on silver-stained 2D gels
(Fig. 1A). Extracts from whole seeds made during
micromalting showed that spots 297 and 313 were
decreased in abundance after 3 and 6 days of germination,
whereas spot 296, containing HvTrxh1, remained at a
high level even after 6 days (Fig. 1C). Analysis of
dissected seed extracts made after 6 days of micromalting
showed that spot 296 remained abundant in the embryo
(Fig. 1C) but was not detectable either in the aleurone
layer or endosperm (not shown). The total amount of
thioredoxin h has been shown to increase in the embryo
and decrease in the endosperm during germination [9], in
agreement with these observations.
Cloning and sequence analysis of barley thioredoxin h
isoforms
Based on the identified EST sequences, specific primers were
designed for cloning of the two thioredoxin h isoforms.
Both transcripts were isolated by RT-PCR using RNA
isolated from barley embryos after one day of germination.
Nucleotide sequencing demonstrated that the isolated

clones were identical with the TC sequences for HvTrxh1
and HvTrxh2 identified on the basis of peptide mass data.
The predicted amino-acid sequences of the proteins were
51% identical. The two thioredoxins were more similar to
thioredoxin h sequences from other plants than to each
other (Fig. 2A), as is also the case for Arabidopsis thio-
redoxins h [15]. HvTrxh1 was 74% identical with a thio-
redoxin h identified in rice as an abundant phloem sap
protein (Q42443) [5]. HvTrxh2 was 53% identical with this
ES
AL
EM
B
Seed 6dSeed 3d
EM 6d
C
A
Western
297
296
313
Coomassie Silver
318
Fig. 1. Barley thioredoxin h forms visualized on 2D gels. Sections of 2D gels from 11 to 16 kDa and pI 4.85–5.25 are shown. The positions of Trx h
spots are indicated by circles. (A) Identification of Trx h in seed extracts by Western blotting. Corresponding spots 296, 297 and 313 on a colloidal
Coomassie-stained gel were confirmed by MS to contain Trx h, and were also observed by silver staining. Spot 318 contains barley dimeric
a-amylase inhibitor BDAI-1. (B) Tissue distribution of Trx h forms analysed using extracts from dissected barley seeds. AL, Aleurone layer; ES,
starchy endosperm; EM, embryo. Proteins are visualized by silver staining. (C) Fate of Trx h forms during micromalting analysed using extracts
from whole seeds after 3 and 6 days micromalting, and from embryo (EM) after 6 days micromalting. Proteins are visualized by silver staining.
Ó FEBS 2003 Barley seed thioredoxin h isoforms (Eur. J. Biochem. 270) 2637

protein, but 90% identical with wheat TrxTa (O64394) [17]
and 78% identical with rice rTrxh2 (Q9FRT3) [38].
HvTrxh1 shared only 49% and 53% identity, respectively,
with these proteins.
The TIGR database contains additional barley sequences
(TC44856 and TC56664) encoding thioredoxin h. These
thioredoxin h forms have an elongated N-terminal
sequence, and TC56664 is very similar to the H. bulbosum
sequence (T50864) identified from mature pollen and
described as a member of a thioredoxin h subgroup [19].
The isoform encoded by TC44856 is 35% and 38%
identical, respectively, with HvTrxh1 and HvTrxh2. The
calculated molecular mass and pI for these predicted
proteins are 14.5 kDa/pI 5.9 and 14.4 kDa/pI 5.2, respect-
ively. No spots in this area, however, were observed on the
2D Western blot with the antibody to wheat thioredoxin h.
This suggests that the antibody does not recognize these
isoforms because several of the ESTs contained in the TC
HvTrxh2
MAAS_____ATAAAVAA_EVISVHSLEQWTMQIEEANTAKKLVVIDFTAS
WCGPC
RIMAP
54
Ta

O64394
AATAT VG-G A

60
Os

Q9FRT3
AS____ Q-EGT__ AI DE I S

I 54
HvTrxh1
__________M EEG-__ AC-TKQEFDTHMANGKDTG I

-VI
48
Os

Q42443
__________M EEGV__ AC-NKDEFDA-MTK-KE-G-V-I

-FI 48
VFADLAKKFPNAVFLK
VDVDELKPIAEQFSVEAMPTFLFMKEGDVKDRVVGAIKEELTAKVGLHAAAQ______ 122
I A T Q_______ 127
HT M-D—-AS-LE—-M-________ 120
EY G-I DV AYN I-
D-EKV-S GR-DDIHT-IVALMGSAST____ 118
EY G EV KYN I-D-AEA-K R-DD-QNTIVK-VGATAASASA 122
A
Active site
R101
B
N
R
I/M

HvTrxh1
HvTrxh2
Ta
a
Os
a
Os
b
Td
Ta
b
Ta
c
Ta
d
Hv
a
Sc
Lp
Hb
Os
c
C
K
N
A
Zm
Os
e
Os

d
Lc
Pc
Ata
Atb
Atc
Atd
Ate
Atf
Atg
Bna
Bo
Bnb
Bra
Brb
Pm
Fe
Cm
Nta
Pt
Ath
Ati
Rc
Ntb
Ntc
Pp
Psa
Psb
Psc
Psd

*
Fig. 2. Sequence alignment of plant thioredoxin h sequences (A), modelled structure of barley HvTrxh1 (B), using the structure of C. reinhardtii Trx h
(PDB accession 1EP7) as a template, and consensus tree of 46 plant thioredoxin h sequences (C), using the alignment region delineated by vertical lines
in (A). (A) Barley HvTrxh1 and HvTrxh2 sequences aligned with TrxTa from wheat (Ta, O64394) and two Trxh sequences from rice (Os, Q9FRT3
and Q42443). Only residues differing between sequences are shown. Dashes indicate identity between sequences and underscores indicate gaps
introduced into the alignment. Peptides observed in mass spectra for spots 297 (HvTrxh2) and 313 (HvTrxh1) are boxed. The conserved active-site
sequence is in bold. The residues corresponding to R101 in HvTrxh1 are marked by an arrow. Vertical lines delineate the region used for the analysis
in (C). (B) Residues differing in HvTrxh1 and HvTrxh2 are shaded. Side chains are shown for the active-site cytseines and R101, which is replaced
by isoleucine in HvTrxh2. (C) HvTrxh1 and HvTrxh2 are indicated. Cereal sequences are in bold and indicated by filled circles. Species are identified
by initials and isoforms by italicised suffix. Accession numbers are given in Materials and methods. The identity of the residue corresponding to
R101 in HvTrxh1 is given for each of the circled clusters. The cluster marked with an asterisk corresponds to the previously described subgroup of
thioredoxin h sequences [19].
2638 K. Maeda et al.(Eur. J. Biochem. 270) Ó FEBS 2003
sequences originate from cDNA libraries from developing
or germinating barley seeds. Alternatively, these isoforms
may not be present in mature seeds. It is currently not
known whether this subclass of thioredoxin h has a specific
function [19].
As no structure is yet available for plant thioredoxin h,
the structures of HvTrxh1 and HvTrxh2 were modelled,
using the crystal structure of C. reinhardtii thioredoxin h
[32] as a template (Fig. 2B). The structure of C. reinhardtii
thioredoxin h is highly superimposable on that of E. coli
thioredoxin, despite only limited sequence identity.
Most of the sequence variation between HvTrxh1 and
HvTrxh2 occurs in the N-terminal and C-terminal parts.
Thus, although HvTrxh1 and HvTrxh2 have an overall
identity of 51%, residues 29–99 of HvTrxh1 and HvTrxh2
(HvTrxh1 numbering) are 75% identical. The residues
differing in HvTrxh1 and HvTrxh2 were mapped on to the

modelled structure of HvTrxh1 (Fig. 2B). These are mainly
distributed away from the active site. The loops close to the
active site consist nearly exclusively of conserved residues.
The most significant difference between the two isoforms in
this region is an arginine at residue 101 of HvTrxh1, located
in the loop before the C-terminal a-helix. This residue is
replaced by isoleucine in HvTrxh2 (Figs 2A.B), which
would lead to a local charge difference on the surface of
the proteins  12 A
˚
from the active-site disulfide bridge.
A comparison of the other related thioredoxin sequences
(Fig. 2A) shows that the rice thioredoxin most similar to
HvTrxh1 also has arginine at this position, whereas the
sequences more similar to HvTrxh2 have isoleucine or
methionine. This trend was supported by a more detailed
analysis of 46 plant thioredoxin h sequences (Fig. 2C) in
which cereal sequences cluster into different groups where
the equivalent residue is R in HvTrxh1-like sequences,
hydrophobic (I or M) in HvTrxh2-like sequences, and N
in the thioredoxin h subgroup previously defined [19]. Non-
cereal thioredoxin h sequences formed additional clusters
that also correlated with the identity of this residue (A, N or
K; Fig. 2C).
Expression and purification of recombinant thioredoxin h
Expression vectors of recombinant barley thioredoxin h
isoforms were constructed for production of nontagged full-
length thioredoxin h isoforms based on the previous
expression and purification system developed for wheat
thioredoxin h (TrxTa) [17]. HvTrxh1, HvTrxh2 and TrxTa

were thus produced in E. coli, and both recombinant barley
thioredoxin h isoforms were recognized by the antibody to
wheat thioredoxin h. The proteins were purified as des-
cribed in Materials and methods. The yields of purified
proteins were 8 mgÆL
)1
culture for HvTrxh1, 3 mgÆL
)1
culture for HvTrxh2, and 1.5 mgÆL
)1
culture for TrxTa.
This compares well with the previously published yield of
5mgÆL
)1
for TrxTa produced from the same expression
system [17].
Characterization of purified proteins
Masses of the recombinant proteins were obtained using
liquid chromatography MS. A single peak corresponding to
an average mass of 12 621.23 Da was obtained for purified
HvTrxh1. This value was not in agreement with the
predicted mass of the full-length protein (12 754.70 Da)
but matched the predicted mass of the protein lacking the
N-terminal methionine and with the active-site cysteines in
oxidized form (12 621.49 Da). N-Terminal sequencing
confirmed that the N-terminal methionine had been cleaved
off. For HvTrxh2, two peaks were obtained by liquid
chromatography MS. One peak (13 032.75 Da) matched
the predicted mass of the protein lacking the N-terminal
methionine and with oxidized active-site cysteines

(13 033.17 Da). The second peak (12 961.53 Da) was
71 Da smaller, indicating that a fraction of the purified
protein was missing methionine and alanine from the
N-terminus. This was confirmed by the N-terminal
sequences of HvTrxh2, determined to be AASATAAAVA
and ASATAAAVAA. Multiple peaks were observed in
the mass spectrum of TrxTa. The largest mass obtained
was 12 675.91 Da, matching the predicted mass of TrxTa
with oxidized active-site cysteines and lacking 10 residues
at the N-terminus (12 675.80 Da). Additional values of
12 574.22 Da, 12 503.50 Da and 12 432.42 Da were
observed that corresponded to cleavage of the N-terminal
segment at residues 11, 12 and 13, respectively.
SDS/PAGE and IEF supported the purity of the recom-
binant proteins. In SDS/PAGE, HvTrxh1 and HvTrxh2 had
apparent molecular masses of  10 kDa and 9 kDa, respect-
ively (Fig. 3A). These values are slightly lower than the
predicted or MS determined masses and the apparent
molecular masses of the thioredoxin h isoforms as observed
in 2D gels of barley protein extracts. Isoelectric points of 5.2
for HvTrxh1 (predicted pI 5.09) and 5.3 for HvTrxh2
(predicted pI 5.2) were obtained by native IEF (Fig. 3B).
Faint bands were observed at slightly lower pI in both cases.
Activity of thioredoxin h isoforms
Insulin reduction. The recombinant barley thioredoxin h
isoforms were analysed for their ability to reduce insulin [34]
and compared with the wheat thioredoxin TrxTa (Fig. 4A).
AB
6.0
14.4

21.5
31.0
36.5
55.4
66.3
97.4
kDa

1

2
5.20
4.55
4.15
5.85
6.55
pI

1

2
Fig. 3. Purified recombinant barley thioredoxin h. (A) SDS/polyacryl-
amide gel loaded with 3 lg purified recombinant thioredoxins. (B)
Native IEF gel loaded with 1 lg purified recombinant thioredoxins.
Lane 1, HvTrxh1; Lane 2, HvTrxh2. Gels are stained with Coomassie
blue. Molecular mass and pI markers are indicated.
Ó FEBS 2003 Barley seed thioredoxin h isoforms (Eur. J. Biochem. 270) 2639
Reaction mixtures containing 1 l
M
TrxTa, HvTrxh1 or

HvTrxh2 showed faster aggregation of insulin compared
with a control containing only dithiothreitol, demonstrating
that all three recombinant proteins could catalyse insulin
reduction. The activity of HvTrxh1 under these conditions
was 1.7 A
650
Æs
)1
Ælmol
)1
, whereas HvTrxh2 and TrxTa were
more similar to each other, with activities of 1.3 and
1.2 A
650
Æs
)1
Ælmol
)1
, respectively.
Reduction of thioredoxin h by NTR. Recombinant
HvTrxh1, HvTrxh2 and TrxTa were tested and compared
as substrates for NTR from A. thaliana [36] (Fig. 4B). The
reduction of thioredoxin h was followed by measuring the
increase in A
405
caused by reduction of DTNB [35].
A. thaliana NTR was able to reduce all three thioredoxin h
proteins. The resulting V
max
and K

m
values for HvTrxh1
were 2.5 nmolÆs
)1
Ænmol
)1
and 13 l
M
. Again, HvTrxh2 and
TrxTa resembled each other with V
max
of 2.0 and
1.5 nmolÆs
)1
Ænmol
)1
and K
m
of 44 l
M
and 37 l
M
, respect-
ively. In the same assay, a K
m
value of  8 l
M
was obtained
for thioredoxin h from P. tremulus (data not shown). For
comparison, a K

m
value of  1.5 l
M
was reported for this
protein under similar assay conditions [35]. The slight
discrepancy between the two studies may be explained by
slight differences in the analysis of the amounts of protein
applied as amino-acid analysis was used for calculation of
the protein concentration in the present study.
Reduction of BASI. Previously, BASI has been identified
as a possible target of thioredoxin h, on the basis of its
reduction by recombinant wheat thioredoxin h [12]. To
determine whether both endogenous thioredoxins were able
to reduce BASI, purified recombinant BASI [37] was
incubated with HvTrxh1 or HvTrxh2 in the presence of
Arabidopsis NTR and NADPH. The reduced cysteines in
BASI were subsequently fluorescence labelled with mono-
bromobimane. The proteins were separated by SDS/PAGE
and the labelled products were visualized under UV light
(Fig. 4C). Both HvTrxh1 and HvTrxh2 were able to reduce
BASI and insulin in this system.
Discussion
Between one and three bands in 1D Western blots of protein
extracts from barley seed tissues were recognized by the
antibody to wheat thioredoxin h [9]. The question was
raised whether these bands represented different thio-
redoxin h isoforms. In the present study using 2D Western
blotting, three spots were also observed and these were used
to locate thioredoxin h-containing spots on Coomassie-
stained and silver-stained 2D gels. Owing to the limited

amount of barley sequence information in the databases, it
was necessary to use EST sequence data for identification of
0
0.4
0.8
1.2
1.6
2.0
010203040
HvTrxh2
HvTrxh1
TrxTa
control
time (min)
A
650
A
0
0.02
0.04
0.06
0.08
0.10
0 5 10 15 20
[trxh] (µM)
V (∆A
405
/min)
HvTrxh2
HvTrxh1

TrxTa
B
Trxh
BASI
NTR
Insulin

1 2 3 4 5
C
Fig. 4. Activity measurements of thioredoxin h. (A) Time course of
insulin reduction by purified recombinant thioredoxins. (B) Reduction
of recombinant thioredoxins by Arabidopsis NTR, monitored via the
reduction of DTNB. (h)HvTrxh1;(s)HvTrxh2;(n)TrxTa;(·)
control (without addition of thioredoxin). (C) Reduction of BASI and
insulin by barley thioredoxin h isoforms. BASI (lanes 1–3) or insulin
(lanes 4 and 5) were incubated with NTR and NADPH without the
addition of thioredoxin h (lane 1) or together with HvTrxh1 (lanes 2
and 4) or HvTrx2 (lanes 3 and 5). Free thiols were labelled with
monobromobimane, and reaction mixtures were run on SDS/PAGE
and visualized under UV. The positions of NTR, BASI, thioredoxin
and insulin bands are indicated.
2640 K. Maeda et al.(Eur. J. Biochem. 270) Ó FEBS 2003
these spots. This led to the identification of two new
thioredoxin h isoforms, designated HvTrxh1 and HvTrxh2.
From MALDI-TOF peptide mass mapping data, it is
probable that the two spots of higher molecular mass
contain the same isoform (HvTrxh1), whereas the spot of
lower molecular mass contains HvTrxh2. Further analysis,
involving more advanced MS techniques, will be required to
explain the appearance of HvTrxh1 in two spots.

Silver-stained 2D gels with extracts from the aleurone
layer, endosperm and embryo show in detail how HvTrxh1
and HvTrxh2 are distributed in these tissues. HvTrxh1 from
all three tissues is observed. The relative intensity of the two
HvTrxh1 spots from the different tissues varies (Fig. 1),
suggesting that the degree of modification giving rise to the
two spots differs among these tissues. HvTrxh2 shows a
different distribution to HvTrx1, being most abundant in
extracts from embryo. Thioredoxin h has previously been
reported to decrease in endosperm and increase slightly in
the embryo in wheat and barley during germination [9,16].
This study confirms these observations, but demonstrates
that the two barley isoforms display different temporal
patterns of appearance. Whereas HvTrx2 decreases in
abundance during micromalting, and HvTrxh1 also decrea-
ses in abundance in the aleurone layer and endosperm, the
lower form of HvTrxh1 remains at a high level in
germinated embryo. These results suggest that the two
isoforms are differentially regulated in the seed tissues and
may have different physiological roles.
Higher plants are characterized by having many thio-
redoxin h isoforms [3,15], typically with divergent sequences
that share higher sequence identity between than within
species. The two barley isoforms identified here are also more
similar to other cereal thioredoxin h isoforms than to each
other, and at least two other barley thioredoxin h sequences
with low identity with HvTrxh1 and HvTrxh2 are present
in the TIGR barley sequence database. As only HvTrxh1
and HvTrxh2 have been identified so far in the barley
seed proteome, the other thioredoxin h isoforms may not

be present in high abundance in mature seeds. The question
remains whether this sequence diversity of thioredoxin h
isoforms in barley and other plants reflects differences in
their biochemical properties or physiological roles.
The coding sequences for HvTrxh1 and HvTrxh2 were
successfully cloned on the basis of the EST sequences
identified by peptide mass mapping of spots on 2D gels. The
purification procedure developed for TrxTa [17] was also
suitable for HvTrxh1 and HvTrxh2. Both isoforms were
confirmed to have protein disulfide reductase activity using
insulin as a substrate, and both were efficiently reduced by
Arabidopsis NTR, as might be expected from the high
sequence identity between NTR from Arabidopsis and
wheat [39]. TrxTa and HvTrxh2, which share the highest
sequence identity, showed very similar properties in both
assays. The properties of HvTrxh1 and HvTrxh2 were,
however, more divergent. Under the conditions used,
HvTrxh1 was both more active in insulin reduction and a
better substrate for Arabidopsis NTR (with  threefold
lower K
m
). Thioredoxins h1 and h2 from P. tremula (with
31% sequence identity) are also reported to have different
biochemical properties. The K
m
of A. thaliana NTR was
1.5 and 12 l
M
for P. tremula thioredoxin h1 and h2,
respectively [35,40].

The structure/function relationships will need to be
analysed to understand the differences in biochemical
properties between HvTrxh1 and HvTrxh2. The residues
that differ between HvTrxh1 and HvTrxh2 are mainly at
the N-termini and C-termini of the proteins, and they are
mostly located away from the active site in the modelled
3D structure. However, R101 of HvTrxh1, replaced by
I105 in HvTrxh2, is located 12 A
˚
fromtheactivesite.
This charge difference on the surface of the protein may
influence the interaction with target proteins. Hetero-
logous expression of chimeric Arabidopsis thioredoxin h
isoforms in yeast suggested that the residues necessary for
interaction with different targets are located in different
regions of the proteins [41]. Interestingly, the N-terminal
sequence of HvTrxh1 contains the sequence MAAEE
(Fig. 2), which has been suggested in the rice phloem
sap thioredoxin h (accession Q42443) to be involved in
phloem trafficking of the protein [38]. By alanine
scanning, a four-amino-acid region of the rice protein
containing arginine at the equivalent position to R101 in
HvTrxh1 was also found to affect phloem trafficking [38].
Mutagenesis studies will be required to determine the
importance of this and other residues in HvTrxh1 and
HvTrxh2.
Identification of target proteins of thioredoxin h has been
reported in several plants including peanuts [42] and barley
embryo [9]. As barley thioredoxin h has not been available
until now, target proteins of thioredoxin h in barley,

including BASI [12], have previously been identified using
heterologous thioredoxin h. It is now demonstrated here
that BASI can be reduced by both endogenous thio-
redoxin h isoforms. Heterologous expression of Arabidopsis
thioredoxin h isoforms in yeast showed that different
thioredoxin h isoforms can interact with different proteins
[18]. Further work is required to determine whether the
different properties of barley HvTrxh1 and HvTrxh2 are
reflected in different target specificities.
In conclusion, the identified isoforms of barley thio-
redoxin h, HvTrxh1 and HvTrxh2, show differences in
tissue distribution, temporal appearance, and biochemical
properties. Further characterization of both proteins and
identification of specific interaction partners will contribute
to a fuller understanding of the functions of thioredoxin h
in barley seeds.
Acknowledgements
Mette Hersom Bien, Sidsel Ehlers, Lone H. Sørensen and Pia Breddam
are gratefully acknowledged for technical assistance, Sejet Plantbreed-
ing for seed material and Jørgen Larsen and Ella Meiling (Carlsberg
Research Laboratory) for micromalted samples. We thank Professor
B. Buchanan (UC Berkeley) for rabbit anti-(wheat thioredoxin h) IgG,
Dr M. Gautier (INRA, Montpellier, France) for pETtrxTa expression
system, and Professor J P. Jacquot (INRA, Nancy, France) for
purified A. thaliana NTR, and P. tremula thioredoxinh,B.C.
Bønsager (Carlsberg Laboratory) for recombinant BASI, and Karen
Skriver (University of Copenhagen) and Kristian Sass Bak-Jensen
(Carlsberg laboratory) for helpful discussions. K.M. is supported by a
scholarship from Carlsbergs Mindelegat for Brygger J. C. Jacobsen.
O.Ø. is supported by a Ph.D. fellowship (no. EF803) from the Danish

Academy of Technical Sciences. The project is supported by the Danish
Research Agency’s SUE programme (samarbejde mellem sektorforsk-
ning, universitet og erhverv) grant no. 9901194.
Ó FEBS 2003 Barley seed thioredoxin h isoforms (Eur. J. Biochem. 270) 2641
References
1. Holmgren, A. (1985) Thioredoxin. Annu. Rev. Biochem. 54,
237–271.
2. Schu
¨
rmann, P. & Jacquot, J.P. (2000) Plant thioredoxin systems
revisited. Annu. Rev. Plant Biol. 51, 371–400.
3. Baumann, U. & Juttner, J. (2002) Plant thioredoxins: the multi-
plicity conundrum. Cell. Mol. Life Sci. 59, 1042–1057.
4. Buchanan, B.B. (1991) Regulation of CO
2
assimilation in oxygenic
photosynthesis: the ferredoxin/thioredoxin system. Perspective
on its discovery, present status, and future development. Arch.
Biochem. Biophys. 288, 1–9.
5. Ishiwatari, Y., Honda, C., Kawashima, I., Nakamura, S., Hirano,
H., Mori, S., Fujiwara, T., Hayashi, H. & Chino, M. (1995)
Thioredoxin h is one of the major proteins in rice phloem sap.
Planta 195, 456–463.
6. Laloi, C., Rayapuram, N., Chartier, Y., Grienenberger, J.M.,
Bonnard, G. & Meyer, Y. (2001) Identification and characteriza-
tion of a mitochondrial thioredoxin system in plants. Proc. Natl
Acad. Sci. USA 98, 14144–14149.
7. Florencio, F.J., Yee, B.C., Johnson, T.C. & Buchanan, B.B. (1988)
An NADP/thioredoxin system in leaves: purification and char-
acterization of NADP-thioredoxin reductase and thioredoxin h

from spinach. Arch. Biochem. Biophys. 266, 496–507.
8. Yano, H., Wong, J.H., Cho, M.J. & Buchanan, B.B. (2001) Redox
changes accompanying the degradation of seed storage proteins in
germinating rice. Plant Cell Physiol. 42, 879–883.
9. Marx, C., Wong, J.H. & Buchanan, B.B. (2003) Thioredoxin and
germinating barley: targets and protein redox changes. Planta 216,
454–460.
10. Kobrehel,K.,Wong,J.H.,Balogh,A.,Kiss,F.,Yee,B.C.&
Buchanan, B.B. (1992) Specific reduction of wheat storage
proteins by thioredoxin h. Plant Physiol. 99, 919–924.
11. Kobrehel, K., Yee, B.C. & Buchanan, B.B. (1991) Role of the
NADP/thioredoxin system in the reduction of alpha-amylase and
trypsin inhibitor proteins. J. Biol. Chem. 266, 16135–16140.
12. Jiao, J., Yee, B.C., Wong, J.H., Kobrehel, K. & Buchanan, B.B.
(1993) Thioredoxin-linked changes in regulatory properties of
barley a-amylase/subtilisin inhibitor protein. Plant Physiol. Bio-
chem. 31, 799–804.
13. Cho,M.J.,Wong,J.H.,Marx,C.,Jiang,W.,Lemaux,P.G.&
Buchanan, B.B. (1999) Overexpression of thioredoxin h leads to
enhanced activity of starch debranching enzyme (pullulanase) in
barley grain. Proc.NatlAcad.Sci.USA96, 14641–14646.
14. Wong, J.H., Kim, Y.B., Ren, P.H., Cai, N., Cho, M.J., Hedden,
P., Lemaux, P.G. & Buchanan, B.B. (2002) Transgenic barley
grain overexpressing thioredoxin shows evidence that the starchy
endosperm communicates with the embryo and the aleurone.
Proc.NatlAcad.Sci.USA99, 16325–16330.
15. Rivera-Madrid, R., Mestres, D., Marinho, P., Jacquot, J.P.,
Decottignies, P., Miginiac-Maslow, M. & Meyer, Y. (1995) Evi-
dence for five divergent thioredoxin h sequences in Arabidopsis
thaliana. Proc.NatlAcad.Sci.USA92, 5620–5624.

16. Serrato, A.J., Crespo, J.L., Florencio, F.J. & Cejudo, F.J. (2001)
Characterization of two thioredoxins h with predominant locali-
zation in the nucleus of aleurone and scutellum cells of germi-
nating wheat seeds. Plant Mol. Biol. 46, 361–371.
17. Gautier, M.F., Lullien-Pellerin, V., de Lamotte-Guery, F.,
Guirao, A. & Joudrier, P. (1998) Characterization of wheat
thioredoxin h cDNA and production of an active Triticum
aestivum protein in Escherichia coli. Eur. J. Biochem. 252,
314–324.
18. Mouaheb, N., Thomas, D., Verdoucq, L., Monfort, P. & Meyer,
Y. (1998) In vivo functional discrimination between plant thio-
redoxins by heterologous expression in the yeast Saccharomyces
cerevisiae. Proc.NatlAcad.Sci.USA95, 3312–3317.
19. Juttner, J., Olde, D., Langridge, P. & Baumann, U. (2000) Cloning
and expression of a distinct class of plant thioredoxins. Eur. J.
Biochem. 267, 7109–7117.
20. Gallardo, K., Job, C., Groot, S.P., Puype, M., Demol, H.,
Vandekerckhove, J. & Job, D. (2001) Proteomic analysis of
Arabidopsis seed germination and priming. Plant Physiol. 126,
835–848.
21. Gallardo, K., Job, C., Groot, S.P., Puype, M., Demol, H.,
Vandekerckhove, J. & Job, D. (2002) Proteomics of Arabidopsis
seed germination. A comparative study of wild-type and
gibberellin-deficient seeds. Plant Physiol. 129, 823–837.
22. Skylas, D.J., Mackintosh, J.A., Cordwell, S.J., Basseal, D.J.,
Walsh, B.J., Harry, J., Blumenthal, C., Copeland, L., Wrigley,
C.W. & Rathmell, W. (2000) Proteome approach to the char-
acterisation of protein composition in the developing and mature
wheat-grain endosperm. J. Cereal Sci. 32, 169–188.
23. Østergaard, O., Melchior, S., Roepstorff, P. & Svensson, B. (2002)

Initial proteome analysis of mature barley seeds and malt.
Proteomics 2, 733–739.
24. Finnie, C., Melchior, S., Roepstorff, P. & Svensson, B. (2002)
Proteome analysis of grain filling and seed maturation in barley.
Plant Physiol. 129, 1308–1319.
25. Finnie, C. & Svensson, B. (2003) Feasability study of a tissue-
specific approach to barley proteome analysis: aleurone layer,
endosperm, embryo and single seeds. J. Cereal Sci. (in press).
26. Heukeshoven, J. & Dernick, R. (1988) Improved silver
staining procedure for fast staining in PhastSystem Development
Unit I. Staining of sodium dodecyl sulfate gels. Electrophoresis 9,
28–32.
27. Rabilloud, T. & Charmont, S. (2000) Detection of proteins on
two-dimensional electrophoresis gels two-dimensional gel electro-
phoresis and identification methods. In Proteome Research
(Rabilloud, T., ed.), pp. 109–110. Springer Verlag, Berlin/Heidel-
berg.
28. Thorpe, G.G.G. & Kricka, L.J. (1986) Enhanced chemilumines-
cent reactions catalysed by horseradish peroxidase. Methods
Enzymol. 133, 311–353.
29. Shevchenko, A., Wilm, M., Vorm, O. & Mann, M. (1996) Mass
spectrometric sequencing of proteins: silver-stained polyacryl-
amide gels. Anal. Chem. 68, 850–858.
30. Gobom, J., Nordhoff, E., Mirgorodskaya, E., Ekman, R. &
Roepstorff, P. (1999) Sample purification and preparation tech-
nique based on nano-scale reversed-phase columns for the sensi-
tive analysis of complex peptide mixtures by matrix-assisted laser
desorption/ionization mass spectrometry. J. Mass Spectrom. 34,
105–116.
31. Guex, N. & Peitsch, M.C. (1997) SWISS-MODEL and the Swiss-

PdbViewer: an environment for comparative protein modeling.
Electrophoresis 18, 2714–2723.
32. Menchise, V., Corbier, C., Didierjean, C., Saviano, M., Benedetti,
E., Jacquot, J.P. & Aubry, A. (2001) Crystal structure of the wild-
type and D30A mutant thioredoxin H of Chlamydomonas
reinhardtii and implications for the catalytic mechanism. Biochem.
J. 359, 65–75.
33. Felsenstein, J. (1993) PHYLIP (Phylogeny Inference Package),
Version 3.5c. Distributed by the Author. Department of Genetics,
University of Washington, Seattle.
34. Holmgren, A. (1979) Thioredoxin catalyzes the reduction of
insulin disulfides by dithiothreitol and dihydrolipoamide. J. Biol.
Chem. 254, 9627–9632.
35. Behm, M. & Jacquot, J.P. (2000) Isolation and characterization
of thioredoxin h from poplar xylem. Plant Physiol. Biochem. 38,
363–369.
36. Jacquot, J.P., Rivera-Madrid, R., Marinho, P., Kollarova, M.,
Mare
´
chal, P.L., Miginiac-Maslow, M. & Meyer, Y. (1994)
Arabidopsis thaliana NAPHP thioredoxin reductase cDNA
2642 K. Maeda et al.(Eur. J. Biochem. 270) Ó FEBS 2003
characterization and expression of the recombinant protein in
Escherichia coli. J. Mol. Biol. 235, 1357–1363.
37. Bønsager, B.C., Prætorius-Ibba, M., Nielsen, P.K. & Svensson, B.
(2003) Purification and characterization of the B-trefoil fold
protein barley a-amylase/subtilisin inhibitor overexpressed in
Escherichia coli. Protein Express. Purif. (in press).
38. Ishiwatari, Y., Fujiwara, T., McFarland, K.C., Nemoto, K.,
Hayashi, H., Chino, M. & Lucas, W.J. (1998) Rice phloem

thioredoxin h has the capacity to mediate its own cell-to-cell
transport through plasmodesmata. Planta 205, 12–22.
39. Serrato, A.J., Pe
´
rez-Ruiz, J.M. & Cejudo, F.J. (2002) Cloning
ofthioredoxinhreductaseandcharacterizationofthethioredoxin
reductase-thioredoxin h system from wheat. Biochem. J. 367,
491–497.
40. Gelhaye, E., Rouhier, N., Laurent, P., Sautie
`
re, P.E., Martin, F. &
Jacquot, J.P. (2002) Isolation and characterization of an extended
thioredoxin h from poplar. Physiol. Plant. 114, 165–171.
41. Bre
´
he
´
lin, C., Mouaheb, N., Verdoucq, L., Lancelin, J.M. &
Meyer, Y. (2000) Characterization of determinants for the speci-
ficity of Arabidopsis thioredoxin h in yeast complementation.
J. Biol. Chem. 275, 31641–31647.
42. Yano, H., Wong, J.H., Lee, Y.M., Cho, M.J. & Buchanan, B.B.
(2001) A strategy for the identification of proteins targeted by
thioredoxin. Proc. Natl Acad. Sci. USA 98, 4797–4799.
Ó FEBS 2003 Barley seed thioredoxin h isoforms (Eur. J. Biochem. 270) 2643