Báo cáo khoa học: Molecular cloning and characterization of two soybean protein disulfide isomerases as molecular chaperones for seed storage proteins doc

Molecular cloning and characterization of two soybean
protein disulfide isomerases as molecular chaperones for
seed storage proteins
Shinya Kamauchi
1,
*
,
†, Hiroyuki Wadahama
1,
*, Kensuke Iwasaki
1
, Yumi Nakamoto
2
,
Keito Nishizawa
2
, Masao Ishimoto
2
, Teruo Kawada
1
and Reiko Urade
1
1 Graduate School of Agriculture, Kyoto University, Uji, Japan
2 National Agricultural Research Center for Hokkaido Region, Sapporo, Japan
Secretory, organelle and membrane proteins are folded
with the assistance of molecular chaperones and other
folding factors in the endoplasmic reticulum (ER). In
many cases, protein folding in the ER is accompanied
by N-glycosylation and the formation of disulfide bonds
[1]. The directed formation of disulfide bonds in a nas-
cent polypeptide chain is thought to be catalyzed by

protein disulfide isomerase (PDI; EC 5.3.4.1) and PDI-
related family proteins that belong to the thioredoxin
superfamily [2–4]. Animal PDI has been shown to act
not only as a thiol-oxidoreductase enzyme, but also as a
molecular chaperone [5]. PDI is thought to bind poly-
peptides through its hydrophobic region and to form,
break and isomerize disulfide bonds in these polypep-
Keywords
endoplasmic reticulum; molecular
chaperone; protein disulfide isomerase;
soybean; storage protein
Correspondence
R. Urade, Graduate School of Agriculture,
Kyoto University, Uji, Kyoto 611-0011, Japan
Fax: +81 774 38 3758
Tel: +81 774 38 3757
E-mail:
†Present address
Osaka Bioscience Institute, Suita, Japan
*These authors contributed equally to this
work
Database
The nucleotide sequence data for the cDNA
of GmPDIL-1 and GmPDIL-2 and genomic
GmPDIL-1 and GmPDIL-2 are available in
the DDBJ ⁄ EMBL ⁄ GenBank databases under
accession numbers AB182628, AB185851,
AB300660 and AB300661, respectively
(Received 26 December 2007, revised 22
February 2008, accepted 18 March 2008)

doi:10.1111/j.1742-4658.2008.06412.x
Protein disulfide isomerase family proteins play important roles in the fold-
ing of nascent polypeptides and the formation of disulfide bonds in the
endoplasmic reticulum. In this study, we cloned two similar protein disul-
fide isomerase family genes from soybean leaf (Glycine max L. Merrill. cv
Jack). The cDNAs encode proteins of 525 and 551 amino acids, named
GmPDIL-1 and GmPDIL-2, respectively. Recombinant versions of GmP-
DIL-1 and GmPDIL-2 expressed in Escherichia coli exhibited oxidative
refolding activity for denatured RNaseA. Genomic sequences of both
GmPDIL-1 and GmPDIL-2 were cloned and sequenced. The comparison of
soybean genomic sequences with those of Arabidopsis, rice and wheat
showed impressive conservation of exon–intron structure across plant spe-
cies. The promoter sequences of GmPDIL-1 apparently contain a cis-acting
regulatory element functionally linked to unfolded protein response. GmP-
DIL-1, but not GmPDIL-2, expression was induced under endoplasmic
reticulum-stress conditions. GmPDIL-1 and GmPDIL-2 promoters contain
some predicted regulatory motifs for seed-specific expression. Both proteins
were ubiquitously expressed in soybean tissues, including cotyledon, and
localized to the endoplasmic reticulum. Data from coimmunoprecipitation
experiments suggested that GmPDIL-1 and GmPDIL-2 associate with pro-
glycinin, a precursor of the seed storage protein glycinin, and the a¢-subunit
of b-conglycinin, a seed storage protein found in cotyledon cells under con-
ditions that disrupt the folding of glycinin or b-conglycinin, suggesting that
GmPDIL-1 and GmPDIL-2 are involved in the proper folding or quality
control of such storage proteins as molecular chaperones.
Abbreviations
AZC,
L-azetidine-2-carboxylic acid; DSP, dithiobis(succinimidylpropionate); ER, endoplasmic reticulum; ERSE, endoplasmic reticulum stress-
responsive element; PDI, protein disulfide isomerase; PVDF, poly(vinylidene difluoride).
2644 FEBS Journal 275 (2008) 2644–2658 ª 2008 The Authors Journal compilation ª 2008 FEBS

tides [6]. In plants, a genome-wide search of Arabidop-
sis thaliana identified a set of 22 PDI orthologs sepa-
rated into 10 phylogenetic groups [7]. Groups I and II
have two thioredoxin domains and show structural simi-
larities to PDI of other eukaryotes. Genes encoding
group I or II PDIs are present in several plants [8–16].
Among them, group I PDIs of rice and Rubiaceae were
shown to affect folding of proglutelin in vivo and cystine
knot defense proteins in vitro [15,17].
In soybean cotyledon cells, large quantities of stor-
age protein are synthesized in the ER during seed
development to reserve carbon and nitrogen for germi-
nation and early growth [18]. Primary soybean seed
storage proteins are globulins called glycinin and
b-conglycinin. They are folded and assemble into tri-
mers in the ER, and are then transported to, and
deposited in, protein storage vacuoles [19]. Glycinin is
synthesized as a precursor subunit that undergoes two
proteolytic processing steps; the first is a removal of an
N-terminal signal peptide in the ER, and the second is
fragmentation of the precursor into 40 kDa acidic and
20 kDa basic subunits in protein storage vacuoles
[20,21]. The second processing step is required for
assembly into hexamers [22]. A major glycinin,
A1aB1b, possess two intramolecular disulfide bonds,
Cys12–Cys45 and Cys88–Cys298, which are required
for hexamer assembly and structural stability [23–25].
Folding and the formation of the disulfide bonds of
glycinin are predicted to be facilitated by some PDI
family members. Previously, we identified novel PDI

family proteins belonging to group IV (GmPDIS-1
and GmPDIS-2) and group V (GmPDIM), and
showed that GmPDIS-1 and GmPDIM associated with
proglycinin in the ER [26,27]. However, involvement
of other PDI family proteins in the folding of the stor-
age proteins remains a topic of investigation.
In this study, we isolated cDNA clones and genomic
sequences encoding soybean group I and II PDI family
members. We present the tissue distribution, cellular
localization and modulation of expression of the
proteins encoded by genes from each of these two
groups during soybean seed development. We provide
evidence of an association between GmPDIL-1 or
GmPDIL-2 and proglycinin or b-conglycinin during
the course of the folding process of these proteins.
Results
Cloning and expression of GmPDIL-1 and
GmPDIL-2
To clone the soybean ortholog of Arabidopsis
PDI-like 1-1 and PDI-like 1-3 categorized in groups I
and II [7], a blast search was performed using the
nucleotide sequence of PDI-like 1-1 or PDI-like 1-3
cDNA from the Institute for Genomic Research Soy-
bean Index. As a result, tentative consensus sequences,
TC188262 from PDI-like 1-1 and TC176115 from
PDI-like 1-3, were found. Using primer sets designed
from their nucleotide sequences, we cloned cDNAs
derived from young soybean leaves by RT-PCR. These
cDNAs encoded proteins, named GmPDIL-1 and
GmPDIL-2, which were 525 and 551 amino acids long,

respectively (supplementary Figs S1 and S2). Both pro-
teins possess a putative N-terminal secretory signal
sequence and two thioredoxin-like motifs with a
CGHC active site. Arginines (Arg128 and Arg482 of
GmPDIL-1 and Arg163 and Arg505 of GmPDIL-2)
known to be involved in the regulation of the active
site redox potential in human PDI [28,29] were con-
served. In addition, glutamic acid residues (Glu67 and
Glu412 of GmPDIL-1, and Glu95 and Glu434 of
GmPDIL-2), suggested to facilitate the release of the
active site from a mixed disulfide with substrate [30],
were also conserved. GmPDIL-1 and GmPDIL-2 pos-
sessed C-terminal, KDEL-related sequences that func-
tion in ER retention [31,32]. The amino acid sequence
identity of GmPDIL-1 and GmPDIL-2 to each other,
minus the signal peptides, was 30%.
Recombinant GmPDIL-1 and GmPDIL-2 proteins
were expressed in Escherichia coli and purified
(Fig. 1A,B). Both recombinant proteins were soluble
and eluted in a monomeric form from a gel filtration
column (data not shown). To examine the helical con-
tent, far-UV CD was performed. Both GmPDIL-1 and
GmPDIL-2 yielded CD spectra that reflected folded
globular protein, and the calculated a-helical content
was 34% and 28% for PDIL-1 and PDIL-2, respec-
tively (data not shown). The domain structures of
GmPDIL-1 and GmPDIL-2 were predicted to form a
linear sequence of four domains in an a–b–b¢–a¢ orien-
tation beginning at the region of conserved domain
sequence homology. We subjected the recombinant

GmPDIL-1 and GmPDIL-2 proteins to limited prote-
olysis with either trypsin or V8 protease to determine
their domain boundaries. After proteolysis for various
time periods, the native recombinant proteins were
gradually degraded, resulting in the generation of
smaller-sized peptide fragments (data not shown). The
sites of proteolytic cleavage were determined to be
Lys77, Lys152, Lys162 and Glu39 of GmPDIL-1 by
N-terminal sequencing of the trypsin peptide fragments
and the V8 protease peptide fragments, respectively. In
addition, four cleavage sites were identified by measur-
ing the masses of the peptide fragments by MALDI-
TOF MS. Among the cleavage sites, four resided in
S. Kamauchi et al. Two soybean protein disulfide isomerases
FEBS Journal 275 (2008) 2644–2658 ª 2008 The Authors Journal compilation ª 2008 FEBS 2645
two narrow regions, overlapping the putative bound-
ary regions in GmPDIL-1 between a and b and b¢ and
a¢ (Fig. 1C). In the case of GmPDIL-2, Lys175, Glu43,
Glu62, Glu68 and Glu135 were identified by N-termi-
nal sequencing of the enzymatic digest fragments,
respectively, and Glu541 was identified by MALDI-
TOF MS. These cleavage sites were located in regions
of N-terminal extension, in the C-terminal tail, and in
the putative boundary regions between a and b
(Fig. 1D). As lysine, arginine or glutamic acid residues
are present in the putative boundary region between b
and b¢ in GmPDIL-1, and between b and b¢ and b¢
and a¢ in GmPDIL-2, the structure of these regions
may be protease resistant.
The activity of recombinant GmPDIL-1 and GmP-

DIL-2, which catalyze oxidative refolding of reduced
and denatured RNaseA, was examined. The specific
activities of GmPDIL-1 and GmPDIL-2 were 472 and
300 mmol RNaseAÆmin
)1
Æmol
)1
, respectively (Fig. 2A).
Several mammalian and yeast PDI family proteins are
known to function as molecular chaperones [5]. We
measured the molecular chaperone activity, which pre-
vents the aggregation of unfolded rhodanese. Aggrega-
tion occurred over 14 min without PDI, but was
inhibited by GmPDIL-2 in a concentration-dependent
manner (Fig. 2C). In the presence of 2.4 lm GmPDIL-
2 (molar ratio of 6 : 1 to rhodanese), 30% of the
rhodanese aggregation was inhibited for at least
14 min. GmPDIL-1 exhibited slight, but significant,
chaperone activity at a molar ratio of 6 : 1 to rhoda-
nese (Fig. 2B).
Cloning of GmPDIL-1 and GmPDIL-2 genomic
sequences
Genomic sequences encoding GmPDIL-1 or GmPDIL-
2 were cloned and sequenced. The alignment and
comparison with the corresponding cDNA showed
that GmPDIL-1 and GmPDIL-2 were composed of
10 and 12 exons, respectively (supplementary Fig. S3).
A
C
D

B
Fig. 1. Prediction of the GmPDIL-1 and GmPDIL-2 domain structures. Recombinant GmPDIL-1 (A) and GmPDIL-2 (B) in E. coli (lane 1) were
purified by His-tag column chromatography (lane 2), followed by gel filtration chromatography (lane 3). Proteins in each effluent were sepa-
rated by 10% SDS ⁄ PAGE and stained with Coomassie Blue. (C,D) Schematic representation of cleavage sites in recombinant GmPDIL-1 (C)
and GmPDIL-2 (D) by limited proteolysis with trypsin and V8 protease. The upper line represents recombinant protein. The boxes below indi-
cate the domain boundaries predicted by an NCBI conserved domain search. The arrows indicate the determined cleavage sites. Black
boxes in domains a and a¢ represent the CGHC motif. A closed circle with a bar represents an N-glycosylation consensus site. SP, signal
peptide.
Two soybean protein disulfide isomerases S. Kamauchi et al.
2646 FEBS Journal 275 (2008) 2644–2658 ª 2008 The Authors Journal compilation ª 2008 FEBS
Nucleotide sequences of the ORF of both PDI genes
were exactly the same as those of cDNAs cloned
in this study. Comparisons of soybean genomic
sequences of GmPDIL-1 or GmPDIL-2 with those of
Arabidopsis (AGI number At1g77510), rice (MOsDb
number Os04g35600) and wheat (accession number
AJ277377) [33] or Arabidopsis (AGI number
At3g54960) and rice (MOsDb number Os02g01010)
showed a significant conservation of the exon–intron
structure across these species. The exon–intron
structure across GmPDIL-1 and GmPDIL-2 was
considerably different. All introns of GmPDIL-1 and
GmPDIL-2 matched the branchpoint consensus
sequence in plants (YTNAN) upstream of the 3¢-site.
The second, third, fifth, sixth and seventh introns
of GmPDIL-1 had the plant branchpoint consensus
sequence TTNAN, whereas the first, fourth, eighth
and ninth introns of GmPDIL-1 had the consensus
sequence CTNAN [34]. In the case of GmPDIL-2,
10 of the 11 introns had TTNAN, and the eighth

intron possessed CTNAN.
Promoter regions of around 1 kb and 1.5 kb
upstream of each start codon of GmPDIL-1 and GmP-
DIL-2 were analyzed. A search for elements upstream
of the coding region of GmPDIL-1 in the database of
plant promoters (PLACE: />PLACE/) detected ER stress-responsive element (ER-
SEs), CCAAT-N
9
-CCACG [35], and a number of
cis-acting regulatory elements involved in endosperm-
specific gene regulation, including G-box, DPBF core
Dc3, E-box, SEF 1 motif and SEF 4 motifs (Table 1).
In the promoter region of GmPDIL-2, cis-acting regu-
latory elements involved in the regulation of endo-
sperm-specific genes, AACA motif, DPBF core Dc3,
E-box, RY-repeat and SEF 1 motif, were found
(Table 2). However, no ER stress regulatory element
was found.
GmPDIL-1 mRNA, but not GmPDIL-2 mRNA, is
upregulated by ER stress
Expression of genes encoding ER-resident proteins is
known to be upregulated by the accumulation of
unfolded protein in the ER (i.e. ER stress) in plant
cells [36,37]. As the consensus sequence of the ERSE
was found in the promoter region of GmPDIL-1, the
potential for GmPDIL-1 response to ER stress was
Fig. 2. Activity of recombinant GmPDIL-1 and GmPDIL-2. (A) Oxi-
dative refolding activity of the recombinant GmPDIL-1, GmPDIL-2,
GmPDIS-1, GmPDIS-2 and GmPDIM. Activity was assayed by the
measurement of RNase activity produced through the regeneration

of the active form of reduced RNaseA. Data represent the
mean ± SD for four experiments. The data for GmPDIS-1, GmP-
DIS-2 and GmPDIM are from Wadahama et al. [26,27]. (B,C) Chap-
erone activity of GmPDIL-1 (B) and GmPDIL-2 (C). The aggregation
of rhodanese (0.4 l
M) was measured without (solid circles) or with
0.8 l
M GmPDIL-2 (open triangles), 2.4 lM GmPDIL-1 (open
squares) or 2.4 l
M GmPDIL-2 (open squares). Each value repre-
sents the mean of three experiments. Bars represent SD. The
statistical significance of difference was determined between
aggregations in the reaction with and without GmPDIL1 or
GmPDIL-2 by the unpaired Student t -test. *P < 0.01.
S. Kamauchi et al. Two soybean protein disulfide isomerases
FEBS Journal 275 (2008) 2644–2658 ª 2008 The Authors Journal compilation ª 2008 FEBS 2647
examined. First, we performed microarray analysis of
transcripts from soybean cotyledons treated with or
without an ER-stress inducing reagent, tunicamycin,
on a GeneChip (Affymetrix, Santa Clara, CA, USA)
designed from soybean-expressed sequence tags. Many
(488) probe sets showed a mean variation of ‡ 4-fold
under tunicamycin treatment (supplementary
Table S1). In total, 178 genes, including some encoding
PDI family members, GmPDIL-1, GmPDIS-1 [26] and
GmPDIM [27], were identified by homology search of
these probe sets. Probes sets for GmPDIL-2 exhibited
little variation (data not shown). To confirm the induc-
tion of GmPDIL-1 mRNA by ER stress, the cotyledon
was treated with or without tunicamycin or l-azeti-

dine-2-carboxylic acid (AZC), and mRNA levels of
GmPDIL-1 and GmPDIL-2 were measured by real-
time RT-PCR. Expression of GmPDIL-1 was upregu-
lated by treatments with tunicamycin and AZC in a
similar manner to expression of the well-known
unfolded protein response genes BiP and calreticulin
(Fig. 3). On the other hand, the expression of GmP-
DIL-2 was not affected by treatment with either
tunicamycin or AZC.
Tissue distribution and cellular localization of
GmPDIL-1 and GmPDIL-2
Anti-GmPDIL-1 and anti-GmPDIL-2 sera specifically
recognized recombinant GmPDIL-1 and GmPDIL-2,
respectively (Fig. 4A, lanes 1 and 4). Anti-GmPDIL-1
and anti-GmPDIL-2 sera did not immunoreact with
recombinant GmPDIL-2 and GmPDIL-1, respectively
(data not shown). Anti-GmPDIL-1 serum immunore-
acted with bands of 60 and 63 kDa (Fig. 4A, lane 2),
whereas anti-GmPDIL-2 serum immunoreacted with a
single 72 kDa band in western analysis of cotyledon
proteins (Fig. 4A, lane 5). These bands were not
detected with anti-GmPDIL-1 and anti-GmPDIL-2
sera pretreated with purified recombinant GmPDIL-1
and GmPDIL-2, respectively (Fig. 4A, lanes 3 and 6),
suggesting that such antibodies specifically immunore-
acted with GmPDIL-1 and GmPDIL-2 in the cotyle-
don. GmPDIL-1 and GmPDIL-2 have two and five
consensus sequences for N-glycosylation, respectively
(Fig. 1C,D). When cotyledon proteins were digested
with glycosidase F, the mass of bands that immuno-

reacted with GmPDIL-1 or GmPDIL-2 sera became
Table 1. Putative regulatory motifs found within the promoter sequences of GmPDIL-1. Base substitutions are in lower-case letters.
Motif Consensus sequence Function Strand
Distance
from ATG Sequence
ERSE CCAAT-N9-CCACG Putative cis-acting element involved in
unfolded protein response
+ 150–168 CgAAT-gatatttcg-CCACG
) 115–133 CgAAT-ctcatgtcc-CCACG
CACGTG
motif (G-box)
CACGTG Essential for expression of b-phaseolin
gene during embryogenesis in bean,
tobacco and Arabidopsis
+ 68–73 CACGTG
) 68–73 CACGTG
DPBFcore Dc3 ACACNNG bZIP transcription factors, Dc3
promoter-binding factor-1 and fator-2
binding core sequence; found in the
carrot Dc3 gene promoter; Dc3
expression is normally
embryo-specific, and also can
be induced by abscisic acid
) 68–73 ACACgtG
+ 350–356 ACACagG
E-box CANNTG E-box of napA storage protein gene of
Brassica napus. Sequence is also
known as RRE (R response element).
Conserved in many storage protein
gene promoters

+ 68–73 CAcgTG
+ 442–447 CAaaTG
+ 515–520 CAacTG
) 68–73 CAcgTG
) 459–464 CAttTG
) 795–798 CAaaTG
) 948–953 CAacTG
SEF 1 motif ATATTTAWW Sequence found in 5¢-upstream region
of soybean b-conglicinin gene
) 649–657 ATATTTAat
) 701–709 ATATTTAta
SEF 4 motif RTTTTTR Sequence found in the 5¢-upstream
region of soybean b-conglycinin gene
+ 812–818 aTTTTTa
+ 860–866 aTTTTTa
) 290–296 aTTTTTa
) 724–730 aTTTTTa
) 801–807 aTTTTTg
) 812–818 aTTTTTa
Two soybean protein disulfide isomerases S. Kamauchi et al.
2648 FEBS Journal 275 (2008) 2644–2658 ª 2008 The Authors Journal compilation ª 2008 FEBS
smaller, shifting from 60 and 63 kDa to 59 and
62 kDa, or from 72 kDa to 63 kDa (Fig. 4B), suggest-
ing N-glycosylation of these proteins in soybean.
Isoelectric points (pI) of GmPDIL-1 and GmPDIL-2
deglycosylated with glycosidase F were determined by
two-dimensional electrophoresis (Fig. 4C). Two spots
of 60 and 63 kDa with similar pI values (5.0 and 5.1),
corresponding to bands detected on SDS ⁄ PAGE, were
detected with anti-GmPDIL-1 serum in cotyledon pro-

teins. These pI values are almost consistent with that
of the recombinant GmPDIL-1. A single spot was
detected with anti-GmPDIL-2 serum. The pI of the
spot (4.9) was consistent with that of the recombinant
GmPDIL-2. GmPDIL-1 and GmPDIL-2 were found
Table 2. Putative regulatory motifs found within the promoter sequences of GmPDIL-2. Base substitutions are in lower-case letters.
Motif
Consensus
sequence Function Strand
Distance
from ATG Sequence
AACA motif AACAAAC Core of AACA motifs found in rice glutelin genes, involved in
controlling endosperm-specific expression
+ 870–876 AACAAAC
) 1521–1527 AACAAAC
DPBFcore Dc3 ACACNNG bZIP transcription factors, Dc3 promoter-binding factor-1 and
factor-2 binding core sequence; found in the carrot Dc3 gene
promoter; Dc3 expression is normally embryo-specific, and
also can be induced by abscisic acid
+ 591–597 ACACgtG
) 170–176 ACACttG
) 756–762 ACACaaG
) 838–844 ACACagG
E-box CANNTG E-box of napA storage protein gene of Brassica napus.
Sequence is also known as RRE (R response element).
Conserved in many storage protein gene promoters
+ 81–86 CAacTG
+ 592–597 CAcgTG
) 90–95 CAacTG
) 171–176 CActTG

) 529–534 CAttTG
) 564–569 CAatTG
) 793–798 CAatTG
) 1458–1463 CAtaTG
RY repeat CATGCAY RY repeat in seed storage protein genes in legumes such
as soybean
+ 728–733 CATGCA
SEF 1 motif RTTTTTR Sequence found in the 5¢-upstream region of soybean
b-conglycinin gene
+ 421–427 aTTTTTa
+ 1185–1191 gTTTTTa
+ 1206–1212 aTTTTTa
) 288–294 aTTTTTa
) 622–628 aTTTTTg
) 642–648 aTTTTTa
) 784–790 aTTTTTa
) 1185–1191 aTTTTTa
Fig. 3. Response of GmPDIL-1 and GmP-
DIL-2 gene expression to ER stress. Cotyle-
dons from 137–142 mg or 210–263 mg
beans were divided into two halves and
incubated in the absence or presence of
tunicamycin (TM) for 24 h (A) or AZC for
18 h (B), respectively. The mRNAs of GmP-
DIL-1, GmPDIL-2, BiP or calreticulin (CRT)
were quantified by real-time RT-PCR. Each
value was standardized by correcting for
actin mRNA. Fold expression change was
calculated as the ratio of mRNA in the sam-
ples treated with the stress reagent to that

in the untreated sample. Data represent the
mean ± SD for three experiments. Data for
*BiP and *CRT are from Wadahama et al.
[26].
S. Kamauchi et al. Two soybean protein disulfide isomerases
FEBS Journal 275 (2008) 2644–2658 ª 2008 The Authors Journal compilation ª 2008 FEBS 2649
to be expressed in roots, stems, trifoliolate leaves, flow-
ers and cotyledons by immunodetection (Fig. 4D).
Expression in leaves decreased during leaf expansion.
Levels of GmPDIL-1 increased until the seeds grew
to 70 mg (Fig. 5B). Thereafter, the level remained
almost constant. Thus, GmPDIL-1 may be expressed
to enhance the machinery for the folding of seed stor-
age proteins. However, this event appeared to be inde-
pendent of transcriptional regulation, as the amounts
of GmPDIL-1 mRNA did not correlate with the levels
of GmPDIL-1 protein expression (Fig. 5A). Levels of
GmPDIL-2 and GmPDIL-2 mRNA did not correlate
with the synthesis of storage proteins (Fig. 5C,D).
GmPDIL-1 and GmPDIL-2 have both an N-termi-
nal signal sequence for targeting to the ER and a
C-terminal ER retention sequence (KDEL). To confirm
localization of GmPDIL-1 and GmPDIL-2 to the ER,
microsomes prepared from cotyledon cells were sepa-
rated by sucrose gradient in the presence of MgCl
2
or
EDTA and analyzed by western blotting (Fig. 6A).
Peaks for GmPDIL-1, GmPDIL-2 and BiP, well
known as an ER lumen protein, were detected at a

density of 1.21 gÆmL
)1
in the presence of MgCl
2
.In
the presence of EDTA, which causes release of ribo-
somes from the rough ER, the peaks of GmPDIL-1,
GmPDIL-2 and BiP displayed a similar shift to the
lighter sucrose fractions (density of 1.16 gÆmL
)1
). This
suggests localization of GmPDIL-1 and GmPDIL-2 to
the rough ER. Then, to confirm residence of GmP-
DIL-1 and GmPDIL-2 in the ER lumen, microsomes
prepared from cotyledon cells were treated with pro-
teinase K in the absence or presence of Triton X-100.
A
B
D
C
Fig. 4. Expression of GmPDIL-1 and GmPDIL-2 in soybean tissues. (A) Cross-reactivity of antiserum prepared against recombinant GmPDIL-
1 or GmPDIL-2 with recombinant GmPDIL-1 (20 ng) (lane 1) and GmPDIL-2 (20 ng) (lane 4) and the cotyledon proteins (10 lg) (lanes 2, 3, 5
and 6). Anti-GmPDIL-1* and anti-GmPDIL-2* represent anti-GmPDIL-1 serum (1 lL) and anti-GmPDIL-2 serum (1 lL) treated with the recom-
binant GmPDIL-1 (4 lg) and the recombinant GmPDIL-2 (10 lg), respectively. (B) GmPDIL-1 and GmPDIL-2 were N-glycosylated proteins in
soybean. The proteins extracted from the cotyledon were treated without (lane 1) or with (lane 2) glycosidase F. The cotyledon proteins
(10 lg) and recombinant GmPDIL-1 (20 ng) or GmPDIL-2 (20 ng) (lane 3) were separated by SDS ⁄ PAGE and immunostained with anti-GmP-
DIL-1 or anti-GmPDIL-2 serum. (C) Separation of recombinant GmPDIL-1 and GmPDIL-2 and GmPDIL-1 and GmPDIL-2 expressed in the soy-
bean cotyledon by two-dimensional electrophoresis. (D) Detection of GmPDIL-1 and GmPDIL-2 in soybean tissues. Thirty micrograms of
protein extracted from the cotyledon (80 mg bean), root, stem, 3 cm leaf, 6 cm leaf, 9 cm leaf and flower were separated by 10%
SDS ⁄ PAGE and immunostained with anti-GmPDIL-1 serum or anti-GmPDIL-2 serum.

Two soybean protein disulfide isomerases S. Kamauchi et al.
2650 FEBS Journal 275 (2008) 2644–2658 ª 2008 The Authors Journal compilation ª 2008 FEBS
Both GmPDIL-1 and GmPDIL-2 were resistant to
protease treatment in the absence of detergent, and
were degraded in the presence of Triton X-100
(Fig. 6B), suggesting that GmPDIL-1 and GmPDIL-2
are luminal proteins.
Association of GmPDIL-1 and GmPDIL-2 with
proglycinin and b-conglycinin a¢-subunit in
cotyledon cells
GmPDIL-1 and GmPDIL-2 were shown to have oxi-
dative folding activity in vitro and to localize to the
cotyledon ER, suggesting that they may function in
proglycinin folding that is accompanied by the forma-
tion of intramolecular disulfide bonds. We sought to
detect an association between GmPDIL-1 or GmP-
DIL-2 and glycinin in cotyledon cells by immunopre-
cipitation from a cotyledon microsomal extract
pretreated with the protein crosslinker dithiobis(succin-
imidylpropionate) (DSP). First, we confirmed the
immunoprecipitation of GmPDIL-1 and GmPDIL-2
from the microsomal extract with anti-GmPDIL-1
serum and anti-GmPDIL-2 serum, respectively
(Fig. 7A,B). Each immunoprecipitant was analyzed by
western blotting with anti-GmPDIL-1 or anti-GmP-
DIL-2 serum. GmPDIL-1 and GmPDIL-2 were immu-
noprecipitated irrespective of crosslinking treatment.
Following metabolic labeling of nascent proteins with
[
35

S]methionine and [
35
S]cysteine, glycinin was immu-
noprecipitated with anti-(glycinin acidic subunit) serum
and detected by fluorography (Fig. 7C, lanes 3 and 4).
Most of the label was found in proglycinin. After
labeling, microsomes from the cotyledons were
crosslinked, solubilized, and immunoprecipitated with
preimmune serum, anti-GmPDIL-1 serum or anti-
GmPDIL-2 serum. The immunoprecipitants were trea-
ted with dithiothreitol to reduce the disulfide bonds
formed by crosslinking, and subjected to a second
immunoprecipitation with anti-(glycinin acidic subunit)
serum. No band was observed in the preimmune serum
sample (Fig. 7C, lanes 1 and 2). Little proglycinin was
detected in the immunoprecipitant with anti-GmPDIL-1
A C
D
B
Fig. 5. Expression of GmPDIL-1 and GmP-
DIL-2 in soybean cotyledons during matura-
tion. mRNA of GmPDIL-1 (A) and GmPDIL-2
(C) was quantified by real-time RT-PCR.
Each value was standardized with actin
mRNA. Values are calculated as a percent-
age of the highest value obtained during
maturation. Bars represent SD of four
experiments. Thirty micrograms of proteins
extracted from the cotyledons was sepa-
rated by 10% SDS ⁄ PAGE and immuno-

stained with anti-GmPDIL-1 serum (B) and
anti-GmPDIL-2 serum (D).
A
B
Fig. 6. Localization of GmPDIL-1 and GmPDIL-2 in the ER lumen.
(A) Microsomes were isolated from cotyledons (100 mg beans) and
fractionated on isopicnic linear sucrose gradients in the presence of
MgCl
2
or EDTA. Proteins from each gradient fraction were analyzed
by western blotting with anti-GmPDIL-1 serum, anti-GmPDIL-2
serum, and anti-BiP serum, respectively. The top of the gradient is
on the left. Density (gÆmL
)1
) is indicated at the top. (B) Microsomes
were treated without (lanes 1 and 2) or with (lanes 3 and 4) pro-
teinase K in the absence (lanes 1 and 3) or presence (lanes 2 and
4) of Triton X-100. Micosomal proteins (10 lg) were analyzed by
western blotting with anti-GmPDIL-1 serum and anti-GmPDIL-2
serum, respectively.
S. Kamauchi et al. Two soybean protein disulfide isomerases
FEBS Journal 275 (2008) 2644–2658 ª 2008 The Authors Journal compilation ª 2008 FEBS 2651
serum or anti-GmPDIL-2 serum (data not shown).
Sufficient detection of proglycinin associated with
GmPDIL-1 and GmPDIL-2 was difficult, as only a
few proglycinin molecules may associate transiently
with PDI proteins in the ER. Then, similar experi-
ments were performed with dithiothreitol-treated coty-
ledons. Unfolded proglycinin may increase in the ER
in the presence of dithiothreitol. Some proglycinin was

detected in the immunoprecipitant from the cotyledon
untreated with DSP, but much more proglycinin was
detected in the immunoprecipitant with anti-GmPDIL-1
and anti-GmPDIL-2 sera from the cotyledon treated
with DSP (Fig. 7C, lanes 5–8). These results suggest
that GmPDIL-1 and GmPDIL-2 molecules associate
with unfolded proglycinin in the lumen of the ER in
the presence of dithiothreitol.
Recombinant proteins, especially GmPDIL-2, exhib-
ited chaperone activity in vitro, raising the question of
whether they act as chaperones for proteins such as
b-conglycinin, which have no intramolecular disulfide
bonds. We then looked for an association between
GmPDIL-1 or GmPDIL-2 and b-conglycinin a¢-sub-
unit. [
35
S]b-conglycinin a¢-subunit was confirmed to be
immunoprecipitated with anti-(b-conglycinin a¢-sub-
unit) serum (Fig. 7D, lanes 3 and 4). b-Conglycinin
a¢-subunit was hardly detected in the immunoprecipi-
tant with anti-GmPDIL-1 serum or anti-GmPDIL-2
serum (data not shown). Then, the immunoprecipita-
tion was performed with the cotyledons treated with
tunicamycin. Tunicamycin may increase unfolded
b-conglycinin a¢-subunit in the ER, as the folding effi-
ciency of glycoproteins such as b-conglycinin a¢-sub-
unit may be lowered by inhibition of N-glycosylation.
b-Conglycinin a¢-subunit was detected in the immuno-
precipitants with anti-GmPDIL-1 serum only from the
cotyledons treated with DSP (Fig. 7D, lane 6). Some

b-conglycinin a¢-subunit was detected in the immuno-
precipitants with anti-GmPDIL-2 serum from cotyle-
dons untreated with DSP (Fig. 7D, lane 7). Much
more b-conglycinin a¢-subunit was detected in the
immunoprecipitants with anti-GmPDIL-2 serum from
the cotyledons treated with DSP (Fig. 7D, lane 8).
These results suggest that GmPDIL-1 and GmPDIL-2
associate with b-conglycinin a¢-subunit in the lumen of
the ER in the presence of tunicamycin.
A
C
B
D
Fig. 7. Coimmunoprecipitation of GmPDIL-1 or GmPDIL-2 and pro-
glycinin or b-conglycinin a¢-subunit. Confirmation of immunoprecipi-
tation of GmPDIL-1 and GmPDIL-2 with each specific antibody.
Microsomes were isolated from cotyledons (150 mg beans) and
treated with (+) or without ()) DSP. Proteins were extracted and
immunoprecipitated with anti-GmPDIL-1 serum or anti-GmPDIL-2
serum. The proteins extracted from the ER (lane 1) and the immu-
noprecipitants (lanes 2 and 3) were analyzed by western blotting
with anti-GmPDIL-1 serum (A) or anti-GmPDIL-2 serum (B). Aster-
isks indicate rabbit serum immunoglobulins recovered by the first
immunoprecipitation in the immunoprecipitant. (C,D) Coimmunopre-
cipitation experiments. Cotyledons were pretreated with dithiothrei-
tol (C) or tunicamycin (D) and labeled with Pro-mix
L-[
35
S] in vitro
labeling mix for 1 h. After labeling, microsomes were isolated and

treated with (+) or without ()) DSP. The extracts from the micro-
somes were subjected to immunoprecipitation with preimmune
serum (lanes 1 and 2), anti-(glycinin acidic subunit) serum (C,
lanes 3 and 4) anti-(b-conglycinin a¢-subunit) serum (D, lanes 3 and
4), anti-GmPDIL-1 serum (lanes 5 and 6), or anti-GmPDIL-2 serum
(lanes 7 and 8). The precipitants were subjected to a second immu-
noprecipitation with anti-(glycinin acidic subunit) serum (C) or anti-
(b-conglycinin a¢-subunit) serum (D). The final precipitants were
subjected to SDS ⁄ PAGE and analyzed by fluorography. The position
of proglycinins (pro11S) or b-conglycinin a¢-subunit (7S-a¢) is indi-
cated on the right.
Two soybean protein disulfide isomerases S. Kamauchi et al.
2652 FEBS Journal 275 (2008) 2644–2658 ª 2008 The Authors Journal compilation ª 2008 FEBS
Discussion
In this study, we cloned and characterized the cDNAs
of GmPDIL-1 and GmPDIL-2 as members of the PDI
family. The amino acid sequences and putative domain
structures of GmPDIL-1 and GmPDIL-2 were similar
to each other, and both recombinant proteins exhibited
thiol-oxidoreductase and chaperone activities. Antise-
rum against recombinant GmPDIL-1 immunoreacted
with two soybean proteins of comparable masses that
had similar pIs. It is unclear whether either protein
was the product of the GmPDIL-1 gene or of different
genes. As the masses of both bands appeared to be
smaller to similar extents after treatment with glycosi-
dase F, it is unlikely that the differences in size were
due to differences in glycosylation. In other plant spe-
cies, gene duplications are found at high frequencies.
Previously, we found other soybean PDI family genes,

those encoding GmPDIS-1 and GmPDIS-2, which
might have arisen by gene duplication [26]. Kainuma
et al. purified a 63 kDa soybean protein from cotyle-
dons and characterized it as a PDI family protein [38].
N-terminal amino acid sequences of the peptide frag-
ments from the purified 63 kDa protein were analyzed,
and the sequence of a 63 amino acid fragment was
determined. Within the sequence, 58 amino acids were
identical to the amino acid sequence of GmPDIL-1,
suggesting that PDIL-1 and the 63 kDa PDI are
homologous proteins encoded by different genes.
Therefore, it seems likely that the doublet band may
be the 63 kDa PDI and GmPDIL-1.
GmPDIL-1, GmPDIS-1 and GmPDIM mRNAs,
but not GmPDIL-2 and GmPDIS-2 mRNAs, were ele-
vated after ER stress [26,27]. Expression of the Arabid-
opsis orthologs of GmPDIL-1, GmPDIS-1 and
GmPDIM has been revealed to be induced by ER
stress by DNA microarray analysis [36,37,39]. In the
promoter regions of GmPDIL-1, GmPDIM [27], and
their Arabidopsis orthologs, consensus sequences of
ERSE were found. Consensus sequences of ERSE were
frequently found in the promoter region of other Ara-
bidopsis genes responsive to ER stress [36,37]. In addi-
tion, a novel Arabidopsis transcription factor,
AtbZIP60, has been shown to activate transcription
from ERSE [40]. Therefore, genes of these PDI family
members may be unfolded protein-responsive genes
that play important roles in maintaining homeostasis
of the ER under conditions of stress.

The consensus sequences for seed-specific expression
were found in the promoter regions of both GmPDIL-1
and GmPDIL-2. However, the mRNA expression
patterns of GmPDIL-1 and GmPDIL-2 were different,
suggesting that the expression of these genes in cotyle-
dons is regulated differently and varies from that
observed for storage proteins [18]. A large amount of
soybean storage proteins is synthesized in cotyledon
cells during seed maturation [18], suggesting that abun-
dant, nascent, unfolded proteins are translocated to
the ER lumen. A rapid increase in the workload of the
ER, as a result of the synthesis of storage proteins,
may elicit an ER stress response. However, it seems
unlikely that ER stress arises during the normal matu-
ration process of soybean seeds as decreases in the
mRNA levels of GmPDIL-1, GmPDIS-1 [26] and
GmPDIM [27] were observed during the accumulation
of the storage proteins. Expression of certain seed stor-
age proteins changed as a function of sulfur supply.
Under conditions of no sulfur, expression of glycinin,
a sulfur-rich storage protein, was decreased. In
contrast, expression of the b-conglycinin b-subunit, a
sulfur-poor storage protein, was elevated [41]. Sulfur
regulation by such proteins is mediated by O-acetyl-
l-serine levels [42]. On the other hand, the levels of
GmPDIL-1 and GmPDIL-2 mRNAs were not affected
by the level of sulfur (supplementary Fig. S4), suggest-
ing that the levels of these mRNAs are not regulated
in a manner responsive to the expression levels of stor-
age proteins. The protein levels of both GmPDIL-1

and GmPDIL-2 were also differentially regulated in
cotyledons during seed development. Protein levels of
GmPDIL-1, GmPDIS-1 and GmPDIM dramatically
increased during seed maturation, but GmPDIL-2 and
GmPDIS-2 were expressed at low levels during the
same stage. These results may reflect the importance of
GmPDIL-1 in seed maturation.
In general, the PDI family proteins catalyze the for-
mation of disulfide bonds on nascent polypeptide
chains in the ER. Hence, GmPDIL-1 and GmPDIL-2
may support proglycinin folding that accompanies the
formation of disulfide bonds in the ER of cotyledon
cells. However, the association of GmPDIL-1 or GmP-
DIL-2 and proglycinin was barely detectable under
normal conditions, whereas this association was
detected in the presence of dithiothreitol, which inhib-
its disulfide bond formation in the ER and may cause
the accumulation of unfolded proglycinin. As the
active sites of the PDI family proteins are reduced in
the ER in the presence of dithiothreitol, neither GmP-
DIL-1 nor GmPDIL-2 forms a mixed disulfide bond
with the cysteine residues of proglycinin. Therefore,
GmPDIL-1 and GmPDIL-2 could noncovalently asso-
ciate with proglycinin in the presence of dithiothreitol,
suggesting that GmPDIL-1 and GmPDIL-2 may func-
tion as molecular chaperones for proglycinin rather than
thiol-oxidoreductases. The chaperone activity of GmP-
DIL-1 for rhodanese was low (Fig. 2B). GmPDIL-1
S. Kamauchi et al. Two soybean protein disulfide isomerases
FEBS Journal 275 (2008) 2644–2658 ª 2008 The Authors Journal compilation ª 2008 FEBS 2653

may recognize proglycinin structures other than the
unfolded rhodanese. The results obtained with anti-
GmPDIL-1 serum must be interpreted cautiously.
Associations of GmPDIL-1 and a protein similar to
GmPDIL-1 with proglycinin may be detected, as anti-
GmPDIL-1 serum immunoreacted with two similar
cotyledon proteins (Fig. 4A). Future investigations uti-
lizing antibodies specific for individual proteins will
clarify these results. An association of GmPDIL-1 or
GmPDIL-2 with b-conglycinin a¢-subunit was also
detected in the presence of tunicamycin, which inhibits
N-glycosylation and may cause the accumulation of
unfolded b-conglycinin a¢-subunit in the ER. As
mature b-conglycinin a¢-subunit possesses no disulfide
bonds, GmPDIL-1 and GmPDIL-2 may act as molec-
ular chaperones for b-conglycinin a¢-subunit [43,44]. It
is uncertain whether GmPDIL-1 and GmPDIL-2 func-
tion in the folding of both proglycinin and b-conglyci-
nin a¢-subunit. Alternatively, GmPDIL-1 and
GmPDIL-2 may be involved in the degradation of
unfolded glycinin and b-conglycinin a¢-subunit through
an ER-associated degradation process.
GmPDIL-1 and GmPDIL-2 are similar proteins and
are coexpressed in the ER lumen. In addition, three
other PDI family members, GmPDIS-1, GmPDIS-2,
and GmPDIM, are also expressed in the same com-
partment. However, their expression levels were differ-
entially regulated in the cotyledon. These proteins may
play different physiological roles in soybean embryo-
genesis. An association of GmPDIS-1 and GmPDIM,

but not of GmPDIS-2, GmPDIL-1, or GmPDIL-2,
with either proglycinin or b-conglycinin was detected
under normal conditions [26,27]. Among soybean PDI
family members, GmPDIS-1 and GmPDIM may pri-
marily function in the folding of seed storage proteins.
In the ER, molecular chaperones appear to collaborate
in response to substrate proteins. Mapping of associa-
tion sites of chaperones and PDI family proteins along
individual substrate polypeptides will be necessary for
clarification of the mechanism of action of this protein
family.
Experimental procedures
Plants
Soybean (Glycine max L. Merrill. cv. Jack) seeds were
planted in 5 L pots and grown in a controlled environmen-
tal chamber at 25 °C under 16 h day ⁄ 8 h night cycles.
Roots were collected from plants 10 days after seeding.
Flowers, leaves and stems were collected from plants
45 days after seeding. All samples were immediately frozen
and stored in liquid nitrogen until use.
Cloning of GmPDIL-1 and PDIL-2
GmPDIL-1 and GmPDIL-2 cDNA cloning was performed
by RT-PCR. Soybean trifoliolate center leaves were frozen
under liquid nitrogen and then ground into a fine powder
with a micropestle SK-100 (Tokken, Inc., Chiba, Japan).
Total RNA was isolated using the RNeasy Plant Mini kit
(Qiagen Inc., Valencia, CA, USA) according to the manu-
facturer’s protocol. The amplification of cDNA from total
RNA was performed with a High Fidelity RNA PCR kit
(TaKaRa Bio Inc., Shiga, Japan), using the following oligo-

nucleotide primers: 5¢-GTCTGTGGTACCTCCTTCAAAA
CCCCCTCCT-3¢ and 5¢-CGTGATGGTACCGGGTGTG
CAACCCACATGTA-3¢ for GmPDIL-1, and 5¢-CAGCCC
GCAGTTGAAAGTCAACCAAGTC-3¢ and Oligo dT-
Adaptor primer FB (TaKaRa Bio Inc.) for GmPDIL-2.
Amplicons were digested with EcoRI or XhoI, respectively,
and subcloned into pUC118 (TaKaRa Bio Inc.), which had
been cleaved with either EcoRI or XhoI, respectively.
Inserts were sequenced using the fluorescence dideoxy chain
termination method on an ABI PRISM 3100-Avant
Genetic Analyzer (Applied Biosystems, Foster City, CA,
USA).
Cloning of GmPDIL-1 and GmPDIL-2 genomic
sequences
Genomic fragments encoding GmPDIL-1 and GmPDIL-2
were isolated from the transformation-competent artificial
chromosome (TAC; pYLTAC7) library of the soybean vari-
ety ‘Misuzudaizu’ by three-dimensional screening [45,46].
Screening was performed by PCR using the oligonucleotide
primers 5¢-CCCAATTTGGAAGCTGATCACAT-3¢ and
5¢-CTTCCTTGGTCCTACCCCCTTCGT-3¢ for GmPDIL-1,
and 5¢-AGCCCGAGGTGGACGAGAAGG-3¢ and 5¢-T
TTGCCACATCAGGATCCACAGTTT-3¢ for GmPDIL-2,
and sequencing.
His-tagged expression plasmid construction
Expression plasmids encoding His-tagged GmPDIL-1 and
GmPDIL-2 minus the putative signal peptides were con-
structed as follows. DNA fragments were amplified from
cDNAs of GmPDIL-1 and GmPDIL-2 by PCR using the
oligonucleotide primers 5¢-GACGACGACAAGATGGAG

GAATCATCGGAGAAAGAGTTC-3¢ and 5¢-GAGGAGA
AGCCCGGTTCAAAGCTCATCTTTTCCTTTTTC-3¢ for
GmPDIL-1, and 5¢-GACGACGACAAGATGCTCACCGA
CGACGAGGACC-3¢ and 5¢-GAGGAGAAGCCCGGTTC
ATAATTCATCCTTCACATC-3¢ for GmPDIL-2. Ampli-
fied DNA fragments were subcloned into pET46Ek ⁄ LIC
(EMD Biosciences, Inc., San Diego, CA, USA). The
recombinant proteins have the His-tag linked to the
N-terminus.
Two soybean protein disulfide isomerases S. Kamauchi et al.
2654 FEBS Journal 275 (2008) 2644–2658 ª 2008 The Authors Journal compilation ª 2008 FEBS
Expression and purification of recombinant
GmPDIL-1 and GmPDIL-2
E. coli BL21(DE3) cells were transformed with the expres-
sion vectors described above. The expression of recombi-
nant proteins was induced by the addition of 0.4 mm
isopropyl thiogalactoside at 30 °C for 4 h; expressed recom-
binant proteins were soluble. Extraction and purification of
recombinant proteins was performed as described previ-
ously [26]. The amino acid sequences of the recombinant
proteins were confirmed by N-terminal sequencing using a
Procise Protein Sequencer 492 (Applied Biosystems).
Determination of protein concentration
The concentrations of purified recombinant GmPDIL-1
and GmPDIL-2 were determined by absorbance at
280 nm using the molar extinction coefficients calculated
according to the modified Gill and von Hippel method
[47]. Extinction coefficients of 38 850 m
)1
Æcm

)1
and
40 340 m
)1
Æcm
)1
were used for recombinant GmPDIL-1
and GmPDIL-2, respectively. The concentration of the
proteins extracted from soybean tissues was measured by
an RC DC protein assay (Bio-Rad Laboratories, Hercules,
CA, USA).
Peptide mapping of GmPDIL-1 and GmPDIL-2
Purified recombinant GmPDIL-1 and GmPDIL-2 (50 lg)
were digested with either trypsin (1 lg) (Sigma-Aldrich
Inc., St Louis, MO, USA) in 100 mm Tris ⁄ HCl buffer
(pH 8.0) at 25 °C for 30 min or V8 protease (2 lg) (Sigma-
Aldrich Inc.) in 100 mm Tris ⁄ HCl buffer (pH 8.0) at 25 °C
for 60 min (GmPDIL-1) or 30 min (GmPDIL-2). The pep-
tides produced were separated by SDS ⁄ PAGE (15% gel)
[48], transferred to a poly(vinylidene difluoride) (PVDF)
membrane (Bio-Rad Laboratories), and stained with Pon-
ceau S. N-terminal amino acid sequencing of each peptide
was carried out. Mass values of the peptides produced by
limited proteolysis were determined by MALDI-TOF MS
on an AXIMA-CFR MALDI-TOF MS plus (Shimadzu
Biotech, Kyoto, Japan).
RNaseA refolding assay
PDI activity was assayed by the measurement of RNase
activity produced through the regeneration of the active
form from reduced RNaseA as described previously [26].

Chaperone activity assessment
Chaperone activity was assayed as described previously
[49]. Aggregation of rhodanese (0.4 lm, Sigma-Aldrich Inc.)
during refolding was measured spectrophotometrically at
320 nm (25 °C) in the absence or presence of GmPDIL-1
and GmPDIL-2.
Antibodies
Antibodies against GmPDIL-1 and GmPDIL-2 were pre-
pared with recombinant GmPDIL-1 and GmPDIL-2 by
Operon Biotechnologies, K.K. (Tokyo, Japan). Anti-BiP,
anti-(glycinin A1aB1b acidic subunit) and anti-(b-conglyci-
nin a¢-subunit) sera have been described previously [26].
Western blot analyses
Western blot analysis was performed essentially as
described previously [26]. Briefly, proteins were extracted
from the frozen tissues by boiling in SDS ⁄ PAGE buffer.
To cleave N-glycans of the proteins, proteins were extracted
from the cotyledons in 0.2% SDS ⁄ 0.1 m Tris ⁄ HCl
(pH 8.6) ⁄ 1% Nonidet P-40. Proteins (0.4 mg) were treated
with 10 mU of glycosidase F (Sigma-Aldrich Inc.) at 37 °C
for 16 h. Proteins were subjected to SDS ⁄ PAGE, and trans-
ferred to a PVDF membrane. For two-dimensional electro-
phoresis, SDS was removed from the samples with the 2D
clean-up kit (GE Healthcare UK Ltd, Chalfont St Giles,
UK). Cotyledon proteins (100 lg), or recombinant
GmPDIL-1 (1 lg) and GmPDIL-2 (1 lg), were subjected to
isoelectric focusing, carried out on a Protean IEF Cell
(Bio-Rad Laboratories), using the 7 cm ReadyStrip IPG
Strip. IEF strips were then subjected to SDS ⁄ PAGE and
transferred to a PVDF membrane. Membranes were probed

first with specific antibodies, and then with a horseradish
peroxidase-conjugated IgG secondary antibody (Promega
Corporation, Madison, WI, USA), using Western Lightning
Chemiluminescence Reagent (Perkin Elmer Life Science,
Boston, MA, USA).
Real time RT-PCR
Measurement of mRNA was performed as described previ-
ously [26]. Briefly, 250 lgÆmL
)1
tunicamycin or 50 lm AZC
(Sigma-Aldrich Inc.) was applied to the inner surface of the
divided half of the cotyledon and incubated at 25 °C. Total
RNA was isolated using RNeasy Plant Mini (Qiagen).
Quantification of mRNA was carried out by real-time
RT-PCR with a Thermal Cycler Dice Real Time System
(TaKaRa Bio Inc.). Forward primers 5¢-GACCTGTTATC
CAACCGTGTACTTCAGGT[FAM]C-3¢ and 5¢-CGGAAC
GCCAATTCATTCTCTTC[FAM]G-3¢, and reverse primers
5¢-GCAGGTTTGTCCCGGTTCT-3¢ and 5¢-GCTCAAGG
GCGAAGACGTAA-3¢ (Invitrogen Corporation, Carlsbad,
CA, USA), were used for detection of the mRNAs of
GmPDIL-1 and GmPDIL-2, respectively. Primers for quan-
tification of actin, BiP and calreticulin mRNA have been
described previously [26].
S. Kamauchi et al. Two soybean protein disulfide isomerases
FEBS Journal 275 (2008) 2644–2658 ª 2008 The Authors Journal compilation ª 2008 FEBS 2655
Proteinase K treatment of microsomes
Microsomes were prepared from cotyledons as described
previously [26], and treated with 0.5 mgÆmL
)1

proteinase K
(Sigma-Aldrich Inc.) in the presence or absence of 1%
Triton X-100 for 5 min at 4 °C. Proteins were precipitated
with 10% trichloroacetic acid for 30 min at 4 °C, and ana-
lyzed by western blotting.
ER fractionation
Slices of cotyledons were homogenized by 20 strokes of a
Dounce homogenizer in ice-cold 100 mm Tris ⁄ HCl
(pH 7.8) ⁄ 10 mm KCl containing 12% (w ⁄ v) sucrose and
either 5 mm MgCl
2
or 5 mm EDTA. The homogenates
were centrifuged for 10 min at 1000 g,at4°C. Then,
600 lL of the supernatant was loaded on a 12 mL linear
21–56% (w ⁄ w) sucrose gradient made in the same buffer.
After centrifugation at 154 400 g for 2 h at 4 °C, frac-
tions (each 1 mL) were collected and assayed by western
blotting.
Labeling of cotyledons
Six pairs of cotyledons were isolated, halved and labeled
flat-side up in a Petri dish at 25 °C for 1 h with a mixture
of 1.48 MBqÆ4mL
)1
of Pro-mix l-[
35
S] in vitro labeling
mix (37 TBqÆmmol
)1
) (GE Healthcare UK Ltd) and 6 mL
of FN Lite [50]. For treatments of cotyledons under ER

stress conditions, a cotyledon was treated with or without
250 lgÆmL
)1
tunicamycin or 1 mm dithiothreitol at 25 °C
for 5 h and labeled. The cotyledons were rinsed three
times with FN Lite containing 10 mm cold methionine
and cysteine. Radiolabeled cotyledons were rinsed at 4 °C
with 20 mm sodium pyrophosphate buffer (pH 7.5) con-
taining 0.3 m mannitol (buffer A), and the flat side of
each cotyledon was sectioned. Slices were homogenized by
20 strokes of a Dounce homogenizer at 4 °C in 3 mL of
buffer A with or without 1 mgÆmL
)1
DSP. The homoge-
nate was placed on ice for 2 h, crosslinking was termi-
nated by the addition of 2 mm glycine for 30 min on ice,
and microsomes were prepared as described previously
[26].
Immunoprecipitation
Immunoprecipitation was carried out as described previ-
ously [27]. Immunoprecipitation was first carried out at
4 °C for 16 h with preimmmune serum, and anti-GmPDIL-1,
anti-GmPDIL-2, anti-(glycinin acidic subunit) or anti-
(b-conglycinin a¢-subunit) sera, which were affinity-purified.
The immunoprecipitate was dissolved in 2% SDS and
0.4 m dithiothreitol. The second immunoprecipitation was
carried out with anti-(glycinin acidic subunit) serum or
anti-(b-conglycinin a¢-subunit) serum at 4 °C for 16 h.
Antigen–antibody complexes were analyzed by SDS ⁄
PAGE, and radiolabeled proteins were detected by fluoro-

graphy with ENLIGHTNING (Perkin Elmer Life Sciences).
A part of the immunoprecipitant obtained by the first
immunoprecipitation with anti-GmPDIL-1 or anti-GmP-
DIL-2 serum was subjected to SDS ⁄ PAGE, transferred to a
PVDF membrane, and probed with anti-GmPDIL-1 or
anti-GmPDIL-2 serum, respectively.
Acknowledgements
We thank Ms Masatoshi Izumo for technical support
in identifying the C-terminal amino acid of the peptide
produced by protease digestion. We thank Ms Akie
Ko for assaying the oxidative refolding activity. This
work was supported by a grant from the Program for
Promotion of Basic Research Activities for Innovative
Biosciences and a Grant-in-Aid for Exploratory
Research from the Ministry of Education, Culture,
Sports, Science and Technology of Japan (18658055).
References
1 Helenius A & Aebi M (2004) Roles of N-linked glycans
in the endoplasmic reticulum. Annu Rev Biochem 73,
1019–1049.
2 Freedman RB, Hirst TR & Tuite MF (1994) Protein
disulphide isomerase: building bridges in protein fold-
ing. Trends Biochem Sci 19, 331–336.
3 Creighton TE, Zapun A & Darby NJ (1995) Mecha-
nisms and catalysts of disulfide bond formation in
proteins. Trends Biotechnol 13, 18–23.
4 Gilbert HF (1998) Protein disulfide isomerase. Methods
Enzymol 290, 26–50.
5 Puig A & Gilbert HF (1994) Protein disulfide isomerase
exhibits chaperone and anti-chaperone activity in the

oxidative refolding of lysozyme. J Biol Chem 269, 7764–
7771.
6 Tian G, Xiang S, Noiva R, Lennarz WJ & Schindelin
H (2006) The crystal structure of yeast protein disulfide
isomerase suggests cooperativity between its active sites.
Cell 124, 1085–1088.
7 Houston NL, Fan C, Xiang QY, Schulze JM, Jung R
& Boston RS (2005) Phylogenetic analyses identify 10
classes of the protein disulfide isomerase family in
plants, including single-domain protein disulfide isomer-
ase-related proteins. Plant Physiol 137, 762–778.
8 Shorrosh BS & Dixon RA (1991) Molecular cloning of
a putative plant endomembrane protein resembling ver-
tebrate protein disulfide-isomerase and a phosphatidyl-
inositol-specific phospholipase C. Proc Natl Acad Sci
USA 88, 10941–10945.
Two soybean protein disulfide isomerases S. Kamauchi et al.
2656 FEBS Journal 275 (2008) 2644–2658 ª 2008 The Authors Journal compilation ª 2008 FEBS
9 Shorrosh BS, Subramaniam J, Schubert KR & Dixon
RA (1993) Expression and localization of plant protein
disulfide isomerase. Plant Physiol 103, 719–726.
10 Chen F & Hayes PM (1994) Nucleotide sequence and
developmental expression of duplicated genes encoding
protein disulfide isomerase in barley (Hordeum vulgare
L.). Plant Physiol 106, 1705–1706.
11 Shimoni Y, Segal G, Zhu XZ & Galili G (1995) Nucle-
otide sequence of a wheat cDNA encoding protein
disulfide isomerase. Plant Physiol 107, 281.
12 Shimoni Y, Zhu XZ, Levanony H, Segal G & Galili G
(1995) Purification, characterization, and intracellular

localization of glycosylated protein disulfide isomerase
from wheat grains. Plant Physiol 108, 327–335.
13 Coughlan SJ, Hastings C & Winfrey RJ Jr (1996)
Molecular characterisation of plant endoplasmic reticu-
lum. Identification of protein disulfide-isomerase as the
major reticuloplasmin. Eur J Biochem 235, 215–224.
14 Huang D-J, Chen H-J & Lin Y-H (2005) Isolation and
expression of protein disulfide isomerase cDNA from
sweet potato (Ipomoea batatas [L.] Lam ‘Tainong 57’)
storage roots. Plant Sci 169, 776–784.
15 Gruber CW, Cemazar M, Clark RJ, Horibe T, Renda
RF, Anderson MA & Craik DJ (2007) A novel plant
protein-disulfide isomerase involved in the oxidative
folding of cystine knot defense proteins. J Biol Chem
282, 20435–20446.
16 Xu ZJ, Ueda K, Masuda K, Ono M & Inoue M (2002)
Molecular characterization of a novel protein disulfide
isomerase in carrot. Gene 284, 225–231.
17 Takemoto Y, Coughlan SJ, Okita TW, Satoh H,
Ogawa M & Kumamaru T (2002) The rice mutant
esp2 greatly accumulates the glutelin precursor and
deletes the protein disulfide isomerase. Plant Physiol
128, 1212–1222.
18 Goldberg RB, Barker SJ & Perez-Grau L (1989) Regu-
lation of gene expression during plant embryogenesis.
Cell 56, 149–160.
19 Mu
¨
ntz K (1998) Deposition of storage proteins. Plant
Mol Biol 38, 77–99.

20 Ereken-Tumer N, Richter JD & Nielsen NC (1982)
Structural characterization of the glycinin precursors.
J Biol Chem 257, 4016–4018.
21 Tumer NE, Thanh VH & Nielsen NC (1981) Purifica-
tion and characterization of mRNA from soybean
seeds. Identification of glycinin and beta-conglycinin
precursors. J Biol Chem 256, 8756–8760.
22 Dickinson CD, Hussein EH & Nielsen NC (1989) Role
of posttranslational cleavage in glycinin assembly. Plant
Cell 1, 459–469.
23 Jung R, Nam YW, Saalbach I, Mu
¨
ntz K & Nielsen NC
(1997) Role of the sulfhydryl redox state and disulfide
bonds in processing and assembly of 11S seed globulins.
Plant Cell 9, 2037–2050.
24 Utsumi S, Gidamis AB, Mikami B & Kito M (1993)
Crystallization and preliminary X-ray crystallographic
analysis of the soybean proglycinin expressed in Escher-
ichia coli. J Mol Biol 233, 177–178.
25 Adachi M, Okuda E, Kaneda Y, Hashimoto A, Shutov
AD, Becker C, Mu
¨
ntz K & Utsumi S (2003) Crystal
structures and structural stabilities of the disulfide
bond-deficient soybean proglycinin mutants C12G and
C88S. J Agric Food Chem 51, 4633–4639.
26 Wadahama H, Kamauchi S, Ishimoto M, Kawada T &
Urade R (2007) Protein disulfide isomerase family pro-
teins involved in soybean protein biogenesis. FEBS J

274, 687–703.
27 Wadahama H, Kamauchi S, Nakamoto Y, Nishizawa
K, Ishimoto M, Kawada T & Urade R (2008) A novel
plant protein disulfide isomerase family homologous to
animal P5: molecular cloning and characterization as a
functional protein for folding of soybean seed storage
proteins. FEBS J 275, 399–410.
28 Lappi AK, Lensink MF, Alanen HI, Salo KE, Lobell
M, Juffer AH & Ruddock LW (2004) A conserved argi-
nine plays a role in the catalytic cycle of the protein
disulphide isomerases. J Mol Biol 335, 283–295.
29 Ellgaard L & Ruddock LW (2005) The human protein
disulphide isomerase family: substrate interactions and
functional properties. EMBO Rep 6, 28–32.
30 LeMaster DM, Springer PA & Unkefer CJ (1997) The
role of the buried aspartate of Escherichia coli thiore-
doxin in the activation of the mixed disulfide intermedi-
ate. J Biol Chem 272, 29998–30001.
31 Munro S & Pelham HR (1986) An Hsp70-like protein
in the ER: identity with the 78 kd glucose-regulated
protein and immunoglobulin heavy chain binding
protein. Cell 46, 291–300.
32 Munro S & Pelham HR (1987) A C-terminal signal
prevents secretion of luminal ER proteins. Cell 48,
899–907.
33 Ciaffi M, Paolacci AR, Dominici L, Tanzarella OA &
Porceddu E (2001) Molecular characterization of gene
sequences coding for protein disulfide isomerase (PDI)
in durum wheat (Triticum turgidum ssp. durum). Gene
265, 147–156.

34 Simpson CG, Clark G, Davidson D, Smith P & Brown
JW (1996) Mutation of putative branchpoint consensus
sequences in plant introns reduces splicing efficiency.
Plant J 9, 369–380.
35 Yoshida H, Haze K, Yanagi H, Yura T & Mori K
(1998) Identification of the cis-acting endoplasmic retic-
ulum stress response element responsible for transcrip-
tional induction of mammalian glucose-regulated
proteins. Involvement of basic leucine zipper transcrip-
tion factors. J Biol Chem 273, 33741–33749.
36 Martı
`
nez IM & Chrispeels MJ (2003) Genomic analysis
of the unfolded protein response in Arabidopsis shows
S. Kamauchi et al. Two soybean protein disulfide isomerases
FEBS Journal 275 (2008) 2644–2658 ª 2008 The Authors Journal compilation ª 2008 FEBS 2657
its connection to important cellular processes. Plant Cell
15, 561–576.
37 Kamauchi S, Nakatani H, Nakano C & Urade R
(2005) Gene expression in response to endoplasmic
reticulum stress in Arabidopsis thaliana. FEBS J 272,
3461–3476.
38 Kainuma K, Ookura T & Kawamura Y (1995) Purifica-
tion and characterization of protein disulfide isomerase
from soybean. J Biochem 117, 208–215.
39 Urade R (2007) Cellular response to unfolded proteins
in the endoplasmic reticulum of plants. FEBS J 274,
1152–1171.
40 Iwata K & Koizumi N (2005) An Arabidopsis transcrip-
tion factor, AtbZIP60, regulates the endoplasmic reticu-

lum stress response in a manner unique to plants. Proc
Natl Acad Sci USA 102, 5280–5285.
41 Hirai MY, Fujiwara T, Chino M & Naito S (1995)
Effects of sulfate concentrations on the expression of a
soybean seed storage protein gene and its reversibility
in transgenic Arabidopsis thaliana. Plant Cell Physiol 36,
1331–1339.
42 Kim H, Hirai MY, Hayashi H, Chino M, Naito S &
Fujiwara T (1999) Role of O-acetyl-l-serine in the coor-
dinated regulation of the expression of a soybean seed
storage-protein gene by sulfur and nitrogen nutrition.
Planta 209, 282–289.
43 Beachy RN, Jarvis NP & Barton KA (1981) Biosynthe-
sis of subunits of the soybean 7S storage protein. J Mol
Appl Genet 1, 19–27.
44 Coates JB, Medeiros JS, Thanh VH & Nielsen NC
(1985) Characterization of the subunits of beta-conglyci-
nin. Arch Biochem Biophys 243, 184–194.
45 Liu YG, Shirano Y, Fukaki H, Yanai Y, Tasaka M,
Tabata S & Shibata D (1999) Complementation of
plant mutants with large genomic DNA fragments by a
transformation-competent artificial chromosome vector
accelerates positional cloning. Proc Natl Acad Sci USA
96, 6535–6540.
46 Harada K & Xia Z (2004) Soybean genomics: effects to
reveal the complex genome. Breed Sci 54, 215–224.
47 Pace CN, Vajdos F, Fee L, Grimsley G & Gray T
(1995) How to measure and predict the molar absorp-
tion coefficient of a protein. Protein Sci 4, 2411–2423.
48 Laemmli UK (1970) Cleavage of structural proteins

during the assembly of the head of bacteriophage T4.
Nature 227, 680–685.
49 Song JL & Wang CC (1995) Chaperone-like activity of
protein disulfide-isomerase in the refolding of rhoda-
nese. Eur J Biochem 231, 312–316.
50 Samoylov VM, Tucker DM & Parrott WA (1998) Soy-
bean [Glycine Max (L.) Merrill] embryogenic cultures:
the role of sucrose and total nitrogen content on prolif-
eration. In Vitro Cell Dev Biol Plant 34, 8–13.
Supplementary material
The following supplementary material is available
online:
Fig. S1. Multiple amino acid sequence alignment of
GmPDIL-1, AtPDIL1-2 (Arabidopsis), OsPDIL1-1
(rice) and ZmPDIL1-1 (maize).
Fig. S2. Multiple amino acid sequence alignment of
GmPDIL-2, AtPDIL1-3 (Arabidopsis), OsPDIL1-4
(rice) and ZmPDIL1-4 (maize).
Fig. S3. Comparison of intron–exon structures of
GmPDIL-1 and GmPDIL-2 from the different species.
Fig. S4. Expression of GmPDIL-1 and GmPDIL-2 in
soybean cotyledons under sulfur deficiency.
Table S1. Upregulated genes selected for DNA micro-
array analysis.
This material is available as part of the online article
from
Please note: Blackwell Publishing are not responsible
for the content or functionality of any supplementary
materials supplied by the authors. Any queries (other
than missing material) should be directed to the corre-

sponding author for the article.
Two soybean protein disulfide isomerases S. Kamauchi et al.
2658 FEBS Journal 275 (2008) 2644–2658 ª 2008 The Authors Journal compilation ª 2008 FEBS