Molecular cloning and characterization of soybean protein disulfide isomerase family proteins with nonclassic active center motifs Kensuke Iwasaki 1 , Shinya Kamauchi 1, *, Hiroyuki Wadahama 1 , Masao Ishimoto 2 , Teruo Kawada 1 and Reiko Urade 1 1 Division of Food Science and Biotechnology, Graduate School of Agriculture, Kyoto University, Uji, Japan 2 National Agricultural Research Center for Hokkaido Region, Sapporo, Japan Introduction Secretory, organelle and membrane proteins are synthe- sized and folded with the assistance of molecular chap- erones and other folding factors in the endoplasmic reticulum (ER). In many cases, the process of protein folding is accompanied by N-glycosylation and the for- mation of disulfide bonds [1]. Disulfide bonds are essential for structural stabilization and for regulation of the functions of many secretory and plasma mem- brane proteins [2,3]. The formation and isomerization of disulfide bonds are catalyzed by protein disulfide isomerase (PDI) and other PDI family proteins located in the ER [4,5]. PDI has two thioredoxin domains containing the redox active site CGHC (a and a¢) and two inactive domains (b and b¢) [6]. Other PDI family Keywords cotyledon; disulfide bond; endoplasmic reticulum; protein disulfide isomerase; soybean 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: urade@kais.kyoto-u.ac.jp *Present address Osaka Bioscience Institute, Suita, Japan Database The nucleotide sequence data for the cDNA of GmPDIL-3a, GmPDIL-3b and genomic GmPDIL-3b are available in the DDBJ ⁄ EMBL ⁄ GenBank databases under accession numbers AB189468, AB189469 and AB303863, respectively (Received 2 April 2009, revised 14 May 2009, accepted 29 May 2009) doi:10.1111/j.1742-4658.2009.07123.x Protein disulfide isomerase (PDI) and other PDI family proteins are mem- bers of the thioredoxin superfamily and are thought to play important roles in disulfide bond formation and isomerization in the endoplasmic reticulum (ER). The exact functions of PDI family proteins in plants remain unknown. In this study, we cloned two novel PDI family genes from soy- bean leaf (Glycine max L. Merrill cv. Jack). The cDNAs encode proteins of 520 and 523 amino acids, and have been denoted GmPDIL-3a and GmP- DIL-3b, respectively. GmPDIL-3a and GmPDIL-3b are the first plant ER PDI family proteins reported to contain the nonclassic redox center motif CXXS ⁄ C, and both proteins are ubiquitously expressed in the plant body. However, recombinant GmPDIL-3a and GmPDIL-3b did not function as oxidoreductases or as molecular chaperones in vitro, although a proportion of each protein formed complexes in both thiol-dependent and thiol-inde- pendent ways in the ER. Expression of GmPDIL-3a and GmPDIL-3b in the cotyledon increased during seed maturation when synthesis of storage proteins was initiated. These results suggest that GmPDIL-3a and GmPDIL-3b may play important roles in the maturation of the cotyledon by mechanisms distinct from those of other PDI family proteins. Structured digital abstract l MINT-7137566: Bip (uniprotkb:Q587K1), GmPDIL-3b (genbank_nucleotide_g:51848586) and GmPDIL-3a (genbank_nucleotide_g: 51848584) colocalize (MI:0403)bycosedimentation through density gradients ( MI:0029) Abbreviations ER, endoplasmic reticulum; PDI, protein disulfide isomerase; PDILT, testis-specific protein disulfide isomerase-like protein; PVDF, poly(vinylidene difluoride). 4130 FEBS Journal 276 (2009) 4130–4141 ª 2009 The Authors Journal compilation ª 2009 FEBS members contain one or more thioredoxin domains [7]. PDI family proteins containing the redox active center transfer the disulfide bond between the two cysteine residues of their active site to the substrate protein [8]. Recently, it has been shown that PDI family proteins containing nonclassic redox motifs, such as yeast Eug1p and mammalian testis-specific PDI-like protein (PDILT) and ERp44, may function in protein folding, retention or transport of ER proteins, or regulation of ER calcium channel activity [9–12]. In plants, a set of 22 orthologs of known PDI family proteins was discovered by a genome-wide search of Arabidopsis thaliana, and was separated into 10 phylo- genetic groups [13]. However, little is known about the physiological roles of plant PDI family members. Stud- ies investigating their contribution to protein folding, transport and quality control are only now beginning. Soybean seeds contain large amounts of protein, especially in their cotyledon cells, where large quanti- ties of storage proteins such as glycinin and b-conglyci- nin are synthesized and folded in the ER during seed development [14,15]. PDI family proteins are predicted to function in collaboration with other molecular chaperones during the folding of these proteins. Previ- ously, we identified and characterized five soybean PDI family proteins belonging to group I (GmPDIL-1), group II (GmPDIL-2), group IV (GmPDIS and GmP- DIS-2), and group V (GmPDIM) [16–18], which all contain two classic CGHC motifs. All of these proteins had thiol oxidoreductase activity in vitro and were ubiquitously expressed in the body of the plant. GmP- DIL-1, GmPDIM and GmPDIS-1 are unfolded protein response genes, and were upregulated by the accumula- tion of unfolded proteins in the ER. GmPDIS-1 and GmPDIM associate with proglycinin (glycinin precur- sor prior to proteolytic processing), and GmPDIL-1 and GmPDIL-2 associate with proglycinin and b-con- glycinin in the ER, suggesting that they may play important roles in folding and in formation and rearrangement of disulfide bonds in the storage proteins. Group III PDI family proteins have not been stud- ied. Putative amino acid sequences obtained from Arabidopsis genome sequence predict the typical PDI domain structure a–b–b¢–a¢, but that both the a-domain and a¢-domain contain nonclassic CXXS ⁄ C motifs as opposed to the more traditional CGHC sequence. In this study, we describe soybean group III PDI family ER proteins, namely GmPDIL-3a and GmPDIL-3b, and identify nonclassic redox center CXXS ⁄ C motifs in each. Characterization of GmP- DIL-3a and GmPDIL-3b and changes in their expression during seed development are described. In addition, our data suggest that GmPDL-3a and GmPDIL-3b form protein complexes in both thiol- dependent and thiol-independent ways in the ER. Results cDNA cloning of GmPDIL-3a and GmPDIL-3b In order to clone the soybean orthologs of Arabidopsis PDI-like1-5 and PDI-like1-6 (group III PDIs) [13], we first obtained their nucleotide sequences from the Insti- tute for Genomic Research Soybean Index and used them in blast searches. We identified the tentative consensus sequence TC183516, and primer sets were designed on the basis of this sequence. Two cDNAs were cloned using RNA extracted from young soybean leaves by 5¢-RACE and 3¢-RACE using these primers. Genomic GmPDIL-3b was cloned and sequenced, whereas the genomic sequence data of GmPDIL-3a were obtained from phytozome v3.1.1 (Department of Energy Joint Genome Institute and the Center for Integrative Genomics, http://www.phytozome.net/soy- bean#A). Comparison of the genomic sequences of GmPDIL-3a and GmPDIL-3b with those of the Ara- bidopsis and rice orthologs showed conservation of exon ⁄ intron structure across these plant species (Fig. S1). Two AUG codons (AUG1 and AUG2) were found upstream of the putative functional translation initia- tion codon (AUG3). Initiation of translation from AUG3 produces 520 or 523 amino acid proteins named GmPDIL-3a and GmPDIL-3b, respectively, in both mRNAs (Figs 1A and S2). To determine whether AUG3 in both mRNAs was the authentic initiation codon, in vitro translation reactions were performed using in vitro transcribed wild-type mRNA, or an mRNA containing an AUG codon mutant(s). A 54 kDa polypeptide was generated when wild-type GmPDIL-3a mRNA was used (Fig. 1B, lane 2), but was not detected when a mutant GmPDIL-3a mRNA that contained AGG in place of AUG3 was used (lane 6). However, this polypeptide was translated when both AUG1 and AUG2 were changed to AGG, confirming that neither AUG1 nor AUG2 is the authentic initiation codon (Fig. 1B, lanes 3–5). A lar- ger amount of the 54 kDa polypeptide was generated from the GmPDIL-3a mRNA with the AUG2 muta- tion (Fig. 1B, lanes 4 and 5) than from that contain- ing only the AUG1 mutation (lane 3) or wild-type mRNA (lane 2), suggesting that initiation events can also begin at AUG2, but are unproductive because of the stop codon located just upstream of AUG3. On the other hand, AUG1 in GmPDIL-3a mRNA may K. Iwasaki et al. Novel plant PDI family proteins FEBS Journal 276 (2009) 4130–4141 ª 2009 The Authors Journal compilation ª 2009 FEBS 4131 not be used as efficiently for translation initiation, and therefore may not interfere with translation from AUG3. For GmPDIL-3b, a 59 kDa polypeptide was generated from wild-type mRNA, and from mRNA that contained AGG in place of both AUG1 and AUG2 (Fig. 1C, lanes 2–5), whereas it was not trans- lated from mutant mRNA in which both AUG3 and AUG4 were changed to AGG (lane 6). The amino acid sequence identity shared between GmPDIL-3a and GmPDIL-3b, excluding the signal peptides, was 92%. The structure of GmPDIL-3a and GmPDIL-3b was predicted to contain the four domains a–b–b¢–a¢ (Fig. 1D). GmPDIL-3a and GmPDIL-3b have two predicted thioredoxin domains between amino acids 65–164 and 403–481, and 68–167 and 406–508, respectively, corresponding to the a-domain and a¢-domain of PDI [7]. Notably, both GmPDIL-3a and GmPDIL-3b lack the two classic PDI redox-active CGHC motifs within the a-domain and a¢-domain. Instead, they both contain the sequence CPRS in the a-domain and CMNC or CINC in the a¢-domain. GmPDIL-3a and GmPDIL-3b contain a C-terminal KDEL sequence that probably functions in ER reten- tion ⁄ retrieval [19], and one putative N-glycosylation site. Recombinant GmPDIL-3a and GmPDIL-3b have neither thiol oxidoreductase nor chaperone activities in vitro Many types of PDI family proteins have oxidative refolding activity on unfolded polypeptides and ⁄ or the ability to reduce disulfide bonds [7,8]. To determine whether GmPDIL-3a or GmPDIL-3b possesses these activities, recombinant mature forms of each pro- tein were expressed in Escherichia coli and purified (Fig. S3A,B). Both proteins were soluble and eluted in a monomeric form from a gel filtration column (data not shown). It was confirmed by far-UV CD experi- ments that the two proteins were folded (Fig. S3C). A B C D Fig. 1. Identification of the initiation codons in GmPDIL-3a and Gm-PDIL-3b mRNAs. (A) Schematic representation of the structure of GmPDIL-3a and GmPDIL-3b mRNAs. The putative ORFs (gray boxes) of GmPDIL-3a and GmPDIL-3b are indicated. AUG1, AUG2, AUG3 and AUG4 indicate the first, second, third and fourth AUG codons from the 5¢-termini, respectively. Crosses indicate ter- mination codons. (B) In vitro translation of GmPDIL-3a. Translation reactions were per- formed without (lane 1) or with (lane 2) 1 lg of wild-type GmPDIL-3a mRNA or mutant GmPDIL-3a mRNA, of which the first (lane 3), second (lane 4), first and second (lane 5) or third AUG (lane 6) was replaced with AGG. Products were separated by SDS ⁄ PAGE and detected by fluorography. (C) In vitro translation of GmPDIL-3b. Trans- lation reactions were performed without (lane 1) or with (lane 2) 1 lg of wild-type GmPDIL-3b mRNA, or with mutant GmP- DIL-3b mRNA, of which the first (lane 3), second (lane 4), first and second (lane 5) or third and fourth AUGs (lane 6) were replaced with AGG. (D) Putative domain structure of GmPDIL-3a and GmPDIL-3b. The boxes indicate the domain boundaries predicted by an NCBI conserved domain search. Black boxes in domain-a and domain-a¢ represent the CPRS and CXXC motifs. A closed circle with a bar represents an N-glycosylation consensus site. SP, signal peptide. Novel plant PDI family proteins K. Iwasaki et al. 4132 FEBS Journal 276 (2009) 4130–4141 ª 2009 The Authors Journal compilation ª 2009 FEBS As shown in Fig. 2A,B, neither protein was able to catalyze the oxidation of thiol residues on the synthetic peptide and the oxidative refolding of reduced and denatured RNase A. In addition, neither protein reduced the disulfide bond in insulin (Fig. 2C). As it has been reported that mammalian PDI family pro- teins function together with other PDI family proteins or with molecular chaperones to effectively fold nas- cent proteins [20–23], we next tested the ability of GmPDIL-3a and GmPDIL-3b to work in concert with the other soybean PD1 proteins Gm-PDIL-1 and GmPDIL-2 [16]. However, as shown in Fig. 2B, GmP- DIL-3a and GmPDIL-3b had no stimulatory effect on the oxidative refolding of RNase A by GmPDIL-1 and GmPDIL-2 when mixed together, further confirming that the functional properties of GmPDIL-3a and GmPDIL-3b are probably unique. Among other soybean PDI proteins, GmPDIL-1 and GmPDIL-2 function as molecular chaperones, and prevent the aggregation of unfolded rhodanese [16]. We next tested whether GmPDIL-3a and GmPDIL-3b function in a similar manner. Aggregation of unfolded rhodanese occurred over 14 min in the absence of PDI, and was partially inhibited by GmPDIL-2 (Fig. 2D). On the other hand, GmPDIL-3a and GmP- DIL-3b did not inhibit the aggregation of rhodanese, even at concentrations up to 1.2 lm (3 : 1 molar ratio, PDI to rhodanese), suggesting that they do not func- tion as molecular chaperones like other PDI family proteins. A B C D Fig. 2. GmPDIL-3a and GmPDIL-3b have neither oxidoreductase nor chaperone activity. (A) Thiol oxidase activities of recombinant GmPDIL- 3a (open circles), GmPDIL-3b (solid circles) and bovine PDI (solid squares) were assayed using the synthetic peptide as a substrate as described in Experimental procedures. (B) Oxidative refolding activity of recombinant GmPDIL-3a (3a), GmPDIL-3b (3b), GmPDIL-1 (L-1), L-1 plus 3a or 3b, GmPDIL-2 (L-2), or L-2 plus 3a or 3b. Activity was assayed by measuring the RNase activity produced through the regeneration of the active form of reduced RNase A. Data represent the mean ± standard deviation for three experiments. (C) Thiol reductase activities of recombinant GmPDIL-3a (open circles), GmPDIL-3b (solid circles) and bovine PDI (solid squares) were assayed using insulin as a substrate. (D) Chaperone activities of recombinant GmPDIL-3a, GmPDIL-3b and GmPDIL-2 were assayed by measuring the aggregation of rhodanese in the absence (open triangles) or presence of GmPDIL-3a (open circles), GmPDIL-3b (solid circles), or GmPDIL-2 (solid squares). K. Iwasaki et al. Novel plant PDI family proteins FEBS Journal 276 (2009) 4130–4141 ª 2009 The Authors Journal compilation ª 2009 FEBS 4133 Expression of GmPDIL-3a and GmPDIL-3b in soybean tissue We next prepared antiserum directed against recombi- nant GmPDIL-3a and a synthetic peptide containing sequences found in GmPDIL-3b, but not in GmPDIL- 3a. Anti-GmPDIL-3a serum recognized both recombi- nant GmPDIL-3a and GmPDIL-3b (Fig. 3A, lanes 1 and 2), whereas anti-GmPDIL-3b serum reacted exclu- sively with recombinant GmPDIL-3b (Fig. 3A, lanes 4 and 5). Anti-GmPDIL-3a serum reacted with both a 55 kDa and a 59 kDa band in western blot analysis of cotyledon proteins (Fig. 3A, lane 3). Both the 55 and 59 kDa bands were N-glycosylated, as digestion experiments using glycosidase F resulted in the bands shifting to 53 and 57 kDa, respectively (Fig. 3B). The cotyledon proteins that were deglycosylated with glycosidase F and detected with the serum were characterized by two-dimensional gel electrophoresis and western blot analysis. Two spots of 53 and 57 kDa, with isoelectric points of 5.3 and 5.1, respec- tively, were detected with anti-GmPDIL-3a serum (Fig. 3C, upper panel). The isoelectric point of the 53 kDa spot (5.3) was identical to a pI value calcu- lated from the amino acid sequence of GmPDIL-3a, and the isoelectric point of the 57 kDa spot (5.1) was consistent with that from the amino acid sequence of GmPDIL-3b. GmPDIL-3b antiserum detected the 57 kDa spot, but not the 53 kDa spot (Fig. 3C, lower panel), suggesting that the 53 and 57 kDa spots are GmPDIL-3a and GmPDIL-3b, respectively. Samples from different parts of the soybean plant were then prepared and analyzed by western immunoblot. GmPDIL-3a and GmPDIL-3b were expressed in roots, stems, trifoliolate leaves, flowers, and cotyledons (Fig. 3D), suggesting that it is a ubiquitously expressed protein that probably performs a function common to all of these tissues. Both GmPDIL-3a and GmPDIL-3b have N-termi- nal signal sequences that target these proteins to the A B D C Fig. 3. Expression of GmPDIL-3a and GmPDIL-3b in soybean tissues. (A) Purified recombinant GmPDIL-3a (lanes 1 and 4), GmPDIL-3b (lanes 2 and 5) and proteins extracted from the cotyle- don (lane 3) were analyzed by western blot using anti-GmPDIL-3a serum (lanes 1–3) or anti-GmPDIL-3b serum (lanes 4 and 5). (B) GmPDIL-3a and GmPDIL-3b are N-glycosylated in soybean. The proteins extracted from the cotyledon were treated without (lane 1) or with (lane 2) glycosidase F. Proteins were analyzed by western blot using anti-GmPDIL-3a serum. (C) Cotyledon proteins were treated with glycosidase F, separated by two-dimensional electro- phoresis, and analyzed by western blot using anti-GmPDIL-3a serum (upper panel) or anti-GmPDIL-3b serum (lower panel). pI, iso- electric point. (D) Thirty micrograms of protein extracted from the cotyledon (80 mg bean) (lane 1), root (lane 2), stem (lane 3), leaf (lane 4) and flower (lane 5) were analyzed by western blot using anti-GmPDIL-3a serum. A B Fig. 4. Localization of GmPDIL-3a and GmPDIL-3b in the ER lumen. (A) Microsomes were isolated from cotyledons (100 mg bean), and were fractionated on isopyknic sucrose gradients in the presence of MgCl 2 or EDTA. Proteins from each fraction were ana- lyzed by western blot using anti-GmPDIL-3a serum or anti-BiP serum. The top of the gradient is on the left, and density (gÆmL )1 ) is indicated on the top. (B) Microsomes were treated without (lanes 1 and 2) or with (lanes 3 and 4) proteinase K, in the absence (lanes 1 and 3) or presence (lanes 2 and 4) of Triton X-100. Microsomal proteins (10 lg) were analyzed by western blot using anti-GmPDIL- 3a serum. Novel plant PDI family proteins K. Iwasaki et al. 4134 FEBS Journal 276 (2009) 4130–4141 ª 2009 The Authors Journal compilation ª 2009 FEBS ER, and a C-terminal ER retention sequence (KDEL). To confirm localization of GmPDIL-3a and GmPDIL- 3b to the ER, microsomes were prepared from cotyle- don cells and were separated by sucrose gradient centrifugation in the presence of MgCl 2 or EDTA, and fractions were collected and analyzed by western blot (Fig. 4A). Peaks corresponding to GmPDIL-3a and GmPDIL-3b were detected at a density of 1.21 gÆmL )1 in the presence of MgCl 2 . In the presence of EDTA, which releases ribosomes from the rough ER [24], the peaks of GmPDIL-3a and GmPDIL-3b were shifted to the lighter sucrose fractions and therefore had a reduced density of 1.16 gÆmL )1 , suggesting that GmP- DIL-3a and GmPDIL-3b localize to the rough ER. To confirm the presence of GmPDIL-3a and GmPDIL-3b in the ER lumen, microsomes were prepared from cotyledon cells and were treated with proteinase K in the absence or presence of Triton X-100. Both GmP- DIL-3a and GmPDIL-3b were resistant to protease treatment in the absence of detergent, and were degraded when detergent was added (Fig. 4B), suggest- ing that they are both luminal proteins. Generally, PDI family proteins play important roles in folding and quality control of nascent polypeptides [25]. In soybean cotyledon, large amounts of seed stor- age proteins such as glycinin and b-conglycinin are synthesized and translocated into the ER lumen during the maturation stage of embryogenesis. Therefore, we next measured the mRNA and protein levels of GmP- DIL-3a and GmPDIL-3b by real-time RT-PCR and western blotting, respectively, during different stages of development. The amounts of pro-b-conglycinin and proglycinin are considered to be nearly equivalent to the synthesis levels of both b-conglycinin and glycinin, as pro-b-conglycinin and proglycinin are transient pro- tein forms that are present in the ER prior to process- ing in the protein storage vacuoles. The synthesis of proglycinin and pro-b-conglycinin was initiated when the seeds achieved a mass of 50 mg (Fig. 5A, lanes 2 and 4). The amount of GmPDIL-3a and GmPDIL-3b proteins increased until the seeds grew from 40 to 80 mg (Fig. 5A, lane 1). Thereafter, the level remained constant. This event correlated with the amount of GmPDIL-3a mRNA, although the amount of GmP- DIL-3b mRNA was not consistent with the amount of GmPDIL-3b protein expression (Fig. 5B). Expression of many ER-resident proteins can be upregurated by ER stress in plant cells [26–28]. There- fore, we next measured the amounts of GmPDIL-3a and GmPDIL-3b mRNA in cotyledon cells under stress by treatment with tunicamycin or dithiothreitol. The amount of neither mRNA was affected by either treatment, whereas the mRNA of BiP, which is a representative unfolded protein response gene, was dramatically upregulated (data not shown). These data suggest that expression of neither GmPDIL-3a nor GmPDIL-3b is influenced by cellular stress. GmPDIL-3a and GmPDIL-3b form protein complexes in the ER Many ER proteins form complexes with other ER resi- dent proteins, and associate with nascent polypeptides during folding [22,29]. We next determined whether GmPDIL-3a or GmPDIL-3b forms complexes in the ER. Cotyledon proteins were extracted with digitonin A B Fig. 5. Expression of GmPDIL-3a and GmPDIL-3b in soybean coty- ledons during maturation. (A) Cotyledon proteins (25 lg) were sepa- rated by SDS ⁄ PAGE and immunostained with anti-GmPDIL-3a serum (lane 1), anti-pro-b-conglycinin a¢ serum (lane 2), anti-b-con- glycinin a¢ serum (lane 3), and anti-glycinin acidic subunit serum (lanes 4 and 5). 3a, GmPDIL-3a; 3b, GmPDIL-3b; Pro 7S-a¢, pro- b-conglycinin a¢; 7S-a¢, mature-b-conglycinin a¢; Pro 11S, proglyci- nin; and 11S-A, mature glycinin acidic subunit. (B) GmPDIL-3a mRNA (upper panel) and GmPDIL-3b mRNA (lower panel) were quantified by real-time RT-PCR. Each value was normalized by dividing it by that for actin mRNA. Values were calculated as a percentage of the highest value obtained during maturation. Data represent the mean ± standard deviation of four experiments. K. Iwasaki et al. Novel plant PDI family proteins FEBS Journal 276 (2009) 4130–4141 ª 2009 The Authors Journal compilation ª 2009 FEBS 4135 and were separated by blue native PAGE, which pro- vides analysis of native proteins and protein complexes [30]. Blue native gels were subjected to SDS⁄ PAGE as a second-dimension separation, followed by western blot analysis (Fig. 6A). Multiple complexes containing GmPDIL-3a or GmPDIL-3b with molecular sizes larger than those of monomeric GmPDIL-3a or GmPDIL-3b were detected in the region of 130– 300 kDa (Fig. 6A). A proportion of GmPDIL-3a or GmPDIL-3b in these complexes was detected as mixed disulfides. When western blots were performed in the presence of N-ethylmaleimide to trap any disulfide- bound intermediates under nonreducing conditions, trace amounts of GmPDIL-3a and GmPDIL-3b mole- cules were found to be engaged in intermolecular, disulfide-linked complexes of approximately 130 kDa (Fig. 6B, lane 3). As these mixed disulfide bonds disap- peared under reducing conditions (Fig. 6B, lane 1), it is likely that GmPDIL-3a or GmPDIL-3b interacts with proteins in the ER through a redox-dependent mechanism. When nonreducing experiments were per- formed after crosslinking treatment of associated pro- teins with dithiobis(succinimidyl propionate), the 130 kDa complexes decreased in abundance, whereas complexes ranging in size from 200 kDa to greater than 250 kDa appeared (Fig. 6B, lane 4). This could suggest that a proportion of the 130 kDa disulfide- linked complexes associated noncovalently with other proteins. Partner proteins for GmPDIL-3a or GmP- DIL-3b in these complexes remain to be identified. Discussion In this report, we characterized new members of the plant PDI family, which we now refer to as GmP- DIL-3a and GmPDIL-3b. The conserved exon struc- ture of the GmPDIL-3a and GmPDIL-3b genomic sequences suggests that both genes may have arisen by gene or chromosome duplication. The conservation of GmPDIL-3a and GmPDIL-3b orthologs in higher plants suggests that they play important physiological roles in these systems. In the cotyledon, maximal expression of GmPDIL-3a and GmPDIL-3b in the late stage of seed development suggests that they per- form a unique role in folding or in accumulation of storage proteins, which are synthesized during this stage. Both GmPDIL-3a and GmPDIL-3b have the same domain architecture, a–b–b ¢–a¢, as the soybean group I and group II PDI family proteins GmPDIL-1 and GmPDIL-2. GmPDIL-3a and GmPDIL-3b share 30% identity with GmPDIL-2, but contain the non- classic redox motif CXXS ⁄ C as opposed to the more common CGHC motif. Atypical CXXS ⁄ C motifs in thioredoxin domains have been noted in some PDI family proteins of yeast and animals [9–12], but this is the first report to confirm expression of such pro- teins in the ER of plants. A CXXS motif and a CXXC motif in the N-terminal and C-terminal thi- oredoxin domains and the surrounding sequences are extremely conserved between plant orthologs of GmP- DIL-3a and GmPDIL-3b, suggesting an important functional role for these regions. PDI requires both cysteine residues present in the redox active site for oxidase activity, but the N-terminal cysteine is suffi- cient for isomerase function [31,32]. Recombinant GmPDIL-3a and GmPDIL-3b showed no oxidase activity in vitro, although they have a CXXC motif in A B Fig. 6. GmPDIL-3a or GmPDIL-3b form protein complexes in a thiol-dependent or thiol-independent manner in the ER. (A) Cotyle- don proteins (100 mg bean) were extracted with digitonin and ana- lyzed by two-dimensional electrophoresis on blue native (BN) PAGE and SDS ⁄ PAGE and western blot using anti-GmPDIL-3a serum. (B) Cotyledon proteins (100 mg bean) treated with (lanes 2 and 4) or without (lanes 1 and 3) dithiobis(succinimidyl propionate) (DSP) were lysed in the presence of N-ethylmaleimide and analyzed by 10% reducing (R) or nonreducing (NR) SDS ⁄ PAGE and western blot using anti-GmPDIL-3a serum. Novel plant PDI family proteins K. Iwasaki et al. 4136 FEBS Journal 276 (2009) 4130–4141 ª 2009 The Authors Journal compilation ª 2009 FEBS their a¢-domain. Additionally, GmPDIL-3a and GmP- DIL-3b showed no reductase activity. Replacement of the second and third amino acids in classic redox- active CGHC motifs with methionine or isoleucine and asparagine in GmPDIL-3a and GmPDIL-3b may be the cause of the lack of such enzymatic activities. Alternatively, the lack of other amino acids, such as arginine, which is important for the regulation of the active site redox potential in human PDI [8,33], may cause the lack of enzymatic activity. Mammalian PDILT, which has the same domain structure as PDI, but lacks oxidoreductase activity, has been dem- onstrated to have chaperone activity in vitro [34]. As PDILT forms a complex with the calnexin homolog calmegin in vitro, this protein is thought to function as a redox-inactive chaperone for glycoprotein folding in testis. However, neither GmPDIL-3a nor GmP- DIL-3b showed chaperone activity in vitro, although it was demonstrated that they formed noncovalent complexes with unidentified proteins in the ER. In addition, interaction between GmPDIL-3a or GmP- DIL-3b and storage proteins such as proglycinin and b-conglycinin, and other ER molecular chaperones such as calnexin, calreticulin, BiP and PDI family proteins, in vivo was not detected (data not shown), suggesting that GmPDIL-3a and GmPDIL-3b may not act as chaperones in the ER. A proportion of GmPDIL-3a and GmPDIL-3b formed mixed disulfide complexes with an unidenti- fied protein in the ER. Mammalian PDI family pro- tein ERp44 forms transient intermolecular bonds with substrate proteins or with the disulfide donor Ero1s. ERp44 cannot be an oxidoreductase, because it has CRFS instead of CGHC. However, the cyste- ine in this motif forms transient mixed disulfide bonds with IgM subunits, adiponectin, and formyl- glycine-generating enzyme, which are devoid of ER retention signals, to regulate their transport [35–38]. ERp44 also functions to retain Ero1a and Ero1b into the ER by forming a mixed disulfide bond and by controlling the ratio of redox isoforms of Ero1a [12,38,39]. Instead, GmPDIL-3a and GmPDIL-3b may function as retention or redox devices, like ERp44, rather than as chaperones. In any case, iden- tification of partner proteins in the mixed disulfide complex and noncovalent complexes of GmPDIL-3a and GmPDIL-3b will be required to establish their physiological function. Little is known about the coordinated function of ER chaperones in the plant. Previously, we observed that at least four types of PDI family proteins (GmP- DIL-1, GmPDIL-2, GmPDIM, and GmPDIS) were expressed ubiquitously in the plant body [16–18]. Thus, it may be difficult to substitute other PDI family pro- teins for GmPDIL-3a or Gm-PDIL-3b, as they proba- bly have unique functions in the plant. The details of how PDI family proteins contribute to ER function and protein folding are beginning to emerge, and, importantly, knowledge concerning GmPDIL-3a and GmPDIL-3b can now be applied to the understanding of how divergent PDI family proteins contribute to quality control in the ER, and how this process influ- ences vital plant function. Experimental procedures Plants Soybean seeds (Glycine max L. Merrill cv. Jack) were planted in 5 L pots and grown in a controlled environment 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-3a and GmPDIL-3b Cloning of the cDNAs for GmPDIL-3a and GmPDIL-3b was performed by 3¢-RACE and 5¢-RACE. Soybean trifolio- late center leaves were frozen under liquid nitrogen and then ground into a fine powder with an SK-100 micropestle (Tokken, Inc., Chiba, Japan). Total RNA was isolated using the RNeasy Plant Mini kit (Qiagen Inc., Valencia, CA, USA), according to the manufacturer’s protocol. mRNA was isolated from total RNA with the PolyATtract mRNA Isolation System (Promega Corporation, Madison, WI, USA). The 3¢-RACE method was performed using the SMART RACE cDNA Amplification kit (Clontech Labora- tories, Inc., Mountain View, CA, USA), according to the manufacturer’s protocol, using the primer 5¢-ACTCTCC TGAATCTTGTTAAC-3¢. The amplified DNA fragments were subcloned into the pT7Blue vector (TaKaRa Bio Inc., Shiga, Japan), and the inserts in the plasmid vectors were sequenced using the fluorescence dideoxy chain termination method and an ABI PRISM 3100-Avant Genetic Analyzer (Applied Biosystems, Foster City, CA, USA). The 5¢-RACE method was performed using the primer 5¢-GAAGCGTGG GGTAGTCATTCACTTGCAG-3¢, which was designed on the basis of the sequence obtained by 3¢-RACE. The ampli- fied DNA fragments were subcloned into the pT7Blue vector and sequenced as described above. In vitro translation Plasmids containing the cDNA fragments encoding GmP- DIL-3a or GmPDIL-3b with mutations in the ATG codons K. Iwasaki et al. Novel plant PDI family proteins FEBS Journal 276 (2009) 4130–4141 ª 2009 The Authors Journal compilation ª 2009 FEBS 4137 were constructed as follows. DNA fragments with mutations were amplified from cDNA of wild-type GmPDIL-3a or GmPDIL-3b as template by PCR, using a mutagenic primer and a forward primer (Table S1). Then, the second PCR was performed using the reaction mixture of the first PCR and a reverse primer (Table S1). Wild-type and mutagenic DNA fragments were subcloned at the SpeI restriction site into pT7Blue (TaKaRa Bio Inc.) and sequenced. Plasmids were linearized by digestion with KpnI, and were tran- scribed in vitro using a RiboMax Large Scale RNA produc- tion systems kit (Promega Corporation). In vitro translation reactions were performed in in a total volume of 25 lL containing 1 lg of mRNA, 555 kBq of L-[ 35 S] in vitro cell labeling mix (37 TBqÆmmol )1 ; GE Healthcare BioSciences Corporation, Piscataway, NJ, USA), 80 lm cysteine ⁄ methi- onine-free amino acid mixture, 0.8 units of RNasin ribo- nuclease inhibitor, 120 mm potassium acetate, and 12.5 lL of wheat germ extract (Promega Corporation) at 25 °C for 90 min. Proteins were separated by 10% SDS ⁄ PAGE, and were detected by fluorography with ENLIGHTNING (Perkin Elmer Life Sciences, Boston, MA, USA). Construction of expression plasmids Expression plasmids encoding mature GmPDIL-3a (Thr24– Leu520) and GmPDIL-3b (Ser27–Leu523), excluding the putative signal peptides, were constructed as follows. DNA fragments were amplified from GmPDIL-3a and GmPDIL- 3b cDNAs by PCR using the primers 5¢-GACGACGACA AGATGGAGGTTAAGGATGAGTTG-3¢ and 5¢-GAG GAGAAGCCCGGTCTATAACTCATCTTTGAGTAC-3¢ for GmPDIL-3a, and 5¢ -GACGACGACAAGATGGAGG TTGAGGATGAGTTGG-3¢ and 5¢-GAGGAGAAGCCCG GTTCATAACTCATCTTTGACGAC-3¢ for GmPDIL-3b. Amplified DNA fragments were subcloned into pET46Ek ⁄ LIC (EMD Biosciences, Inc., San Diego, CA, USA) and sequenced. The recombinant proteins have the His-tag linked to the N-terminus. Expression and purification of recombinant GmPDIL-3a and GmPDIL-3b E. coli BL21(DE3) cells were transformed with the expres- sion vectors described above. The expression of recombi- nant GmPDIL-3a was induced in the presence of 0.4 mm isopropyl thio-b-d-galactoside at 4 °C for 5 days, whereas the expression of recombinant GmPDIL-3b was induced in the presence of 0.4 mm isopropyl thio-b-d-galactoside at 30 °C for 4 h. Extraction and purification of recombinant proteins was performed as described previously [18]. The concentration of each protein was determined by measuring the absorbance at 280 nm using the molar extinction coeffi- cient of 31 830 m )1 Æcm )1 for both proteins, as calculated according to the modified method of Gill and von Hippel [40]. The concentration of the proteins extracted from soybean tissues was measured by RC-DC protein assay (Bio-Rad Laboratories, Hercules, CA, USA). Enzymatic activity assays Thiol oxidative refolding activity was assayed as previously described by measuring RNase activity produced through the regeneration of the active form from reduced and denatured RNase A in the presence of 0.5 lm recombinant GmPDIL- 3a, GmPDIL-3b, GmPDIL-1, or GmPDIL-2 [41,42]. Thiol reductase activity was measured as previously described, where the glutathione-dependent reduction of insulin was measured by Morjana and Gilbert [43]. Briefly, 50 lgof bovine PDI (TaKaRa Bio Inc.), recombinant GmPDIL-3a or recombinant GmPDIL-3b was incubated in 1 mL of 0.2 m sodium phosphate buffer (pH 7.5) containing 5 mm EDTA, 3.7 mm reduced glutathione, 0.12 mm NADPH, 16 U of glu- tathione reductase (Sigma-Aldrich Inc., St Louis, MO, USA) and 30 lm insulin (Sigma-Aldrich Inc.) at 25 °C, and absor- bance was monitored at 340 nm. Oxidase activity was assayed using a synthetic peptide, NH 2 -NRCSQGSCWN- COOH, as described by Alanen et al. [44]. Briefly, 0.5 lm bovine PDI, recombinant GmPDIL-3a or recombinant GmPDIL-3b was incubated in 0.2 m sodium phosphate ⁄ citrate buffer (pH 6.5), 2 mm reduced glutathione, 0.5 mm oxidized glutathione and 5 lm synthetic peptide at 25 °C, and fluorescence was monitored at 350 nm with excitation at 280 nm on a Hitachi F-3000 fluorescence spectrophotometer (Hitachi Ltd, Tokyo, Japan). Chaperone activity assays Chaperone activity was assayed as described previously [45]. Briefly, aggregation of 0.4 lm rhodanese (Sigma- Aldrich Inc.) during refolding was measured spectrophoto- metrically at 320 nm (25 °C) in the absence or presence of 1.2 lm recombinant GmPDIL-3a, GmPDIL-3b, and GmPDIL-2. Antibody production Guinea pig antiserum specific for GmPDIL-3a and rabbit antiserum specific for a GmPDIL-3b peptide were prepared using recombinant GmPDIL-3a and the synthetic peptide GSVTEAEKFLRKY, which was conjugated to keyhole limpet hemocyanin by Operon Biotechnologies, K.K. (Tokyo, Japan). Antisera specific for pro-b-conglycinin, b-conglycinin, glycinin acidic subunit and BiP were prepared as described previously [18]. Western blot analysis Proteins were extracted from plant tissues by boiling in SDS ⁄ PAGE buffer [46]. To cleave N-linked glycans, Novel plant PDI family proteins K. Iwasaki et al. 4138 FEBS Journal 276 (2009) 4130–4141 ª 2009 The Authors Journal compilation ª 2009 FEBS proteins were extracted from the cotyledons in buffer containing 0.02% SDS, 0.1 m Tris ⁄ HCl (pH 8.6), and 1% Nonidet P-40. Approximately 0.4 mg of protein was treated with 10 mU of glycosidase F (Sigma-Aldrich Inc.) at 37 °C for 16 h. For crosslinking of proteins, slices of cotyledons were homogenized with 20 strokes of a Dounce homo- genizer at 4 °C in 3 mL of buffer containing 20 mm Hepes (pH 7.2), 150 mm NaCl, 1% protease inhibitor cocktail for plant cells (Sigma-Aldrich Inc.), and 20 mm N-ethylmale- imide, in the presence or absence of 20 mm dithiobis(succin- imidyl propionate). The homogenate was placed on ice for 2 h, and crosslinking was terminated by the addition of 20 mm glycine for 20 min on ice. Proteins were then sus- pended in SDS ⁄ PAGE buffer and subjected to SDS ⁄ PAGE [46], and were transferred to poly(vinylidene difluoride) (PVDF) membranes. For two-dimensional separation by isoelectric focusing and SDS ⁄ PAGE, SDS was removed from the samples with the 2D clean-up kit (GE Healthcare UK Ltd). Proteins (100 lg) were applied to 7 cm Ready- Strip IPG Strips (Bio-Rad Laboratories), and isoelectric focusing was performed using a Protean IEF Cell (Bio-Rad Laboratories). The strips were then subjected to SDS ⁄ PAGE, and proteins on the gel were transferred to PVDF membranes. For two-dimensional electrophoresis of blue native PAGE [30] and SDS ⁄ PAGE, slices of cotyledons were homogenized with 20 strokes of a Dounce homo- genizer in ice-cold buffer containing 50 mm Bis-Tris (pH 7.2), 50 mm NaCl containing 10% (w ⁄ v) glycerol, 0.001% ponceau S, and 1% digitonin. After standing at 4 °C for 1 h, the homogenate was centrifuged for 30 min at 14 000 g. Five per cent Coomassie Brilliant Blue G-250 solution was added to the supernatant to a final concentra- tion of 0.25%, and the supernatant was subjected to 3–12% polyacrylamide gradient gel electrophoresis accord- ing to the manufacturer’s protocol for the Native PAGE Novex Bis-Tris Gel System (Invitrogen Corporation, Carls- bad, CA, USA). Blue native PAGE gels were then sub- jected to SDS ⁄ PAGE, and proteins on the gel were transferred to PVDF membranes. Membranes were incu- bated with primary antibody, followed by a horseradish peroxidase-conjugated IgG secondary antibody (Promega Corporation), and were developed using Western Lightning Chemiluminescence Reagent (Perkin Elmer Life Science) as previously described [18]. Real-time RT-PCR Measurement of mRNA was performed as described previ- ously [18]. Briefly, total RNA was isolated using the RNeasy Plant Mini kit (Qiagen Inc.). mRNA was quanti- fied by real-time RT-PCR with a Thermal Cycler Dice Real Time System (TaKaRa Bio Inc.). Forward primers 5¢-CG TTTGAAGGGTGAGGAGGAAAA[FAM]G-3¢ and 5¢-CA CAAGAGAGTTCTGCGATAACCTTG[FAM]G-3¢ (Invi- trogen Corporation) and reverse primers 5¢-AAGTAGGCA ACAAAACAACG-3¢ and 5¢-GTTTTCCCGACAATAA- CATGG-3¢ were used for detecting GmPDIL-3a and GmPDIL-3b, respectively. Primers for quantification of actin mRNA were as described previously [18]. Each value was normalized by dividing it by that for actin mRNA. Proteinase K treatment of microsomes Microsomes were prepared from cotyledons as described previously [17], and treated with 0.5 mgÆmL )1 proteinase K (Sigma-Aldrich Inc.) in the presence or absence of 1% Tri- ton X-100 for 5 min at 4 °C. Proteins were precipitated with 10% trichloroacetic acid for 30 min at 4 °C and analyzed by western blot. ER fractionation Slices of cotyledons were homogenized with 20 strokes of a Dounce homogenizer in ice-cold buffer containing 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. Homo- genates were centrifuged for 10 min at 1000 g at 4 °C. Next, 600 lL of the supernatant was loaded onto a 12 mL linear 21–56% (w ⁄ w) sucrose gradient prepared in the same buffer. Samples were centrifuged at 154 400 g for 2 h at 4 °C, and 1 mL fractions were collected and assayed by western blot. Acknowledgements We thank Dr M. Kito for critical reading of the man- uscript and warm encouragement. 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 & Aeb M (2004) Roles of N-linked glycans in the endoplasmic reticulum. Annu Rev Biochem 73, 1019–1049. 2 Wittrup KD (1995) Disulfide bond formation and eukaryotic secretory productivity. Curr Opin Biotechnol 6, 203–208. 3 Hogg PJ (2002) Biological regulation through protein disulfide bond cleavage. Redox Rep 7, 71–77. 4 Freedman RB, Hirst TR & Tuite MF (1994) Protein disulphide isomerase: building bridges in protein folding. Trends Biochem Sci 19, 331–336. 5 Creighton TE, Zapun A & Darby NJ (1995) Mecha- nisms and catalysts of disulfide bond formation in proteins. Trends Biotechnol 13, 18–23. K. Iwasaki et al. Novel plant PDI family proteins FEBS Journal 276 (2009) 4130–4141 ª 2009 The Authors Journal compilation ª 2009 FEBS 4139 [...]... (2008) Molecular cloning and characterization of two soybean protein disulfide isomerases as molecular chaperones for seed storage proteins FEBS J 275, 2644–2658 17 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. .. Lyles MM & Gilbert HF (1991) Catalysis of the oxidative folding of ribonuclease A by protein disulfide isomerase: dependence of the rate on the composition of the redox buffer Biochemistry 30, 613–619 43 Morjana NA & Gilbert HF (1991) Effect of protein and peptide inhibitors on the activity of protein disulfideisomerase Biochemistry 30, 4985–4990 Novel plant PDI family proteins 44 Alanen HI, Salo KE, Pirneskoski... dependence of the peptide thiol -disulfide oxidase activity of six members of the human protein disulfide isomerase family Antioxid Redox Signal 8, 283–291 45 Kikuchi M, Doi E, Tsujimoto I, Horibe T & Tsujimoto Y (2002) Functional analysis of human P5, a protein disulfide isomerase homologue J Biochem 132, 451–455 46 Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage... Comparison of intron–exon structures of GmPDIL-3a, GmPDIL-3b and orthologs of other plants Fig S2 Multiple amino acid sequence alignment of GmPDIL-3a, GmPDIL-3b and orthologs of other plants Fig S3 Expression and CD analysis of recombinant GmPDIL-3a and GmPDIL-3b Table S1 List of primers for ATG-mutagenesis of GmPDIL-3a and GmPDIL-3b This supplementary material can be found in the online version of this... activities of protein disulfide isomerase family in protein folding Cell Stress Chaperones 10, 211–220 Ellgaard L & Frickel EM (2003) Calnexin, calreticulin, and ERp57: teammates in glycoprotein folding Cell Biochem Biophys 39, 223–247 Meunier L, Usherwood YK, Chung KT & Hendershot LM (2002) A subset of chaperones and folding enzymes form multiprotein complexes in endoplasmic reticulum to bind nascent proteins. .. reticulum protein that is functionally related to protein disulfide isomerase Mol Cell Biol 12, 4601–4611 10 Nørgaard P & Winther JR (2001) Mutation of yeast Eug1p CXXS active sites to CXXC results in a dramatic increase in protein disulfide isomerase activity Biochem J 358, 269–274 11 van Lith M, Hartigan N, Hatch J & Benham AM (2005) PDILT, a divergent testis-specific protein disulfide isomerase with a...Novel plant PDI family proteins K Iwasaki et al 6 Gilbert HF (1998) Protein disulfide isomerase Methods Enzymol 290, 26–50 7 Hatahet F & Ruddock LW (2007) Substrate recognition by the protein disulfide isomerases FEBS J 274, 5223– 5234 8 Ellgaard L & Ruddock LW (2005) The human protein disulphide isomerase family: substrate interactions and functional properties EMBO Rep 6, 28–32... characterization as a functional protein for folding of soybean seed-storage proteins FEBS J 275, 399–410 18 Wadahama H, Kamauchi S, Ishimoto M, Kawada T & Urade R (2007) Protein disulfide isomerase family proteins involved in soybean protein biogenesis FEBS J 274, 687–703 19 Munro S & Pelham HR (1987) A C-terminal signal prevents secretion of luminal ER proteins Cell 48, 899–907 20 Satoh M, Shimada A, Kashiwai A,... 223–231 Walker KW, Lyles MM & Gilbert HF (1996) Catalysis of oxidative protein folding by mutants of protein disulfide isomerase with a single active- site cysteine Biochemistry 35, 1972–1980 Laboissiere MC, Sturley SL & Raines RT (1995) The essential function of protein- disulfide isomerase is to unscramble non-native disulfide bonds J Biol Chem 270, 28006–28009 Lappi AK, Lensink MF, Alanen HI, Salo KE, Lobell... Phylogenetic analyses identify 10 classes of the protein disulfide isomerase family in plants, including single-domain protein disulfide isomerase- related proteins Plant Physiol 137, 762–778 14 Goldberg RB, Barker SJ & Perez-Grau L (1989) Regulation of gene expression during plant embryogenesis Cell 56, 149–160 15 Muntz K (1998) Deposition of storage proteins Plant ¨ Mol Biol 38, 77–99 16 Kamauchi S, Wadahama . Molecular cloning and characterization of soybean protein disulfide isomerase family proteins with nonclassic active center motifs Kensuke. Urade R (2008) Molecular cloning and characterization of two soybean protein disulfide isomerases as molecular chaperones for seed storage proteins. FEBS