Molecular cloning, expression and characterization of protein disulfide isomerase from Conus marmoreus Zhi-Qiang Wang 1 , Yu-Hong Han 1,2 , Xiao-Xia Shao 1 , Cheng-Wu Chi 1,2 and Zhan-Yun Guo 1 1 Institute of Protein Research, Tongji University, Shanghai, China 2 Institute of Biochemistry and Cell Biology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Graduate School of the Chinese Academy of Sciences, Shanghai, China In eukaryotes, all proteins that travel along or reside in the secretory pathway are folded in the endoplas- mic reticulum (ER). As one of the most important post-translational modifications, the disulfide bonds are formed in the ER lumen, where oxidoreductases catalyze the reaction and serve as disulfide donors [1]. The archetypical oxidoreductase in the ER lumen is protein disulfide isomerase (PDI). In the oxidized state, PDI functions as a disulfide donor for its client proteins. In the reduced state, PDI catalyzes reduction and isomerization of pre-existing disulfides. The abil- ity of PDI to function as a reductase, an oxidase and an isomerase ensures PDI’s ability to serve as a major catalyst for disulfide formation in vivo [2,3]. Moreover, PDI also acts as a chaperone for substrates during catalysis [4]. Conotoxins are small, cysteine-rich peptides pro- duced by marine cone snails [5]. Although their amino acid sequences are hypervariable, they can form spe- cific disulfide patterns that are essential for their bio- logical activities. It is believed that cone snails possess evolving mechanisms to ensure efficient folding of conotoxins in vivo, but these mechanisms are not fully understood yet. Keywords conotoxin; disulfide isomerization; oxidative folding; protein disulfide isomerase Correspondence Z Y. Guo, Institute of Protein Research, Tongji University, 1239 Siping Road, Shanghai 200092, China Fax: +86 21 65988403 Tel: +86 21 65985167 E-mail: zhan-yun.guo@mail.tongji.edu.cn C W. Chi, Shanghai Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences, 320 YueYang Road, Shanghai 200031, China Fax: +86 21 54921011 Tel: +86 21 54921165 E-mail: chi@sunm.shcnc.ac.cn (Received 2 March 2007, revised 8 July 2007, accepted 18 July 2007) doi:10.1111/j.1742-4658.2007.06003.x The oxidative folding of disulfide-rich conotoxins is essential for their biological functions. In vivo, disulfide bond formation is mainly catalyzed by protein disulfide isomerase. To elucidate the physiologic roles of pro- tein disulfide isomerase in the folding of conotoxins, we have cloned a novel full-length protein disulfide isomerase from Conus marmoreus. Its ORF encodes a 500 amino acid protein that shares sequence homology with protein disulfide isomerases from other species, and 70% homology with human protein disulfide isomerase. Enzymatic analyses of recombi- nant C. marmoreus protein disulfide isomerase showed that it shared functional similarities with human protein disulfide isomerase. Using conotoxins tx3a and sTx3.1 as substrate, we analyzed the oxidase and isomerase activities of the C. marmoreus protein disulfide isomerase and found that it was much more efficient than glutathione in catalyzing oxi- dative folding and disulfide isomerization of conotoxins. We further dem- onstrated that macromolecular crowding had little effect on the protein disulfide isomerase-catalyzed oxidative folding and disulfide isomerization of conotoxins. On the basis of these data, we propose that the C. mar- moreus protein disulfide isomerase plays a key role during in vivo folding of conotoxins. Abbreviations cPDI, Conus marmoreus protein disulfide isomerase; ER, endoplasmic reticulum; hPDI, human protein disulfide isomerase; GSH, reduced glutathione; GSSG, oxidized glutathione; nTx3.1, native Tx3.1; PDI, protein disulfide isomerase; RNase A, bovine pancreatic RNase A; sTx3.1, swap Tx3.1. 4778 FEBS Journal 274 (2007) 4778–4787 ª 2007 The Authors Journal compilation ª 2007 FEBS One mechanism exploited by cone snails to ensure efficient folding of conotoxins is adding necessary folding information to the mature polypeptides. For example, the C-terminal Gly that is used for amidation of the mature form of x-conotoxin containing an ami- dated C-terminus, as isolated from the venom of Conus magnus (x-MVIIA) can significantly increase the folding yield [6]. The carboxylation of Glu residues can also improve the folding yield, because the resultant c-carboxyglutamates can bind Ca 2+ and facilitate dis- ulfide pairing [5]. Other post-translational modifications, such as O-glycosylation, bromination of tryptophan, hydroxylation of proline, and l-tod-epimerization, may also facilitate the folding of conotoxins [7]. Another mechanism to improve the folding of cono- toxins in vivo is utilization of the molecular chaperones and foldases [5,8]. Many chaperones and foldases are present in the ER lumen [9,10]. Among them, PDI (EC 5.3.4.1) is a unique and multifunctional enzyme that exhibits disulfide reductase, oxidase and isomerase activities, as well as chaperone activity, and its concen- tration in the ER lumen can be as high as  200 lm [2,3]. Previous results indicated that PDI was the major soluble protein in the ER, and was expressed throughout the whole length of Conus venom ducts. Two full-length cDNAs encoding two PDI isoforms have been isolated from Conus textile [5,11]. However, the enzymatic properties of Conus PDI have not been thoroughly investigated, especially using the endoge- nous conotoxins as substrates. In this article, we report gene cloning, recombinant expression and enzymatic activity analyses of a novel Conus marmoreus PDI (cPDI). Our results strongly suggest that cPDI might play a key role during in vivo folding of conotoxins. Results Molecular cloning of a novel PDI from C. marmoreus PDI is an abundant protein in the venom ducts of C. textile, from which two PDIs have been cloned [5,11]. In present work, we cloned a novel PDI from C. marmoreus (GenBank accession number DQ486867). The 1742 bp full-length cDNA includes a 3¢-UTR and a polyadenylation consensus sequence (AATTATAA) located 12 nucleotides upstream of the polyA tail. Its 1500 bp ORF encodes a 500 amino acid protein (Fig. 1A) that shares sequence homology with PDIs from other species. A signal peptide (17 amino acids) predicted by the signalp program [12] is present at its N-terminus, and a typical ER retention signal, RDEL, is present at its C-terminus. The mature cPDI protein has a calculated molecular mass of 54 913.7 Da and an isoelectric point of 4.6. The cPDI contains four thioredoxin domains and an acidic C-ter- minal tail (a, b, b¢, a¢ and c). Two thioredoxin active sites (WCGHCK) are found in the a and a¢ domains, respectively. The cPDI shares 94% amino acid sequence identity with its homologs C. textile PDI 1 and C. textile PDI 2 [5], 70% identity with human PDI (hPDI), and 42% with yeast PDI. An unrooted neighbor-joining phylogenetic tree was obtained by comparing the deduced amino acid sequences of different PDIs from fungi to mammals using the mega (Molecular Evolutionary Genetic Analysis Software, Version 3.1) program, bootstrap: 1000 replications (Fig. 1B). Enzymatic activities of cPDI The cPDI was recombinantly expressed in Escherichia coli as a soluble cytoplasmic protein, recovered from a soluble cell extract, and purified to homogeneity. Its purity was approximately 95% as judged by SDS ⁄ PAGE. The expression yield was about 50 mgÆL )1 . As shown in Table 1, the enzymatic activities of cPDI were analyzed using various substrates and com- pared with those of hPDI. The recombinant cPDI and hPDI shared similar reductase, oxidase and isomerase activities. PDI exhibits both chaperone and antichaperone activities when catalyzing the refolding of reduced ⁄ denatured lysozyme in Hepes buffer [13]. As shown in Fig. 2, we analyzed the chaperone and antichaperone activities of cPDI. Without PDI, the final refolding yield of the reduced ⁄ denatured lysozyme reached approximately 40%. At low concentrations, both cPDI and hPDI decreased the refolding yield of lysozyme (antichaperone activity). At high concentrations, both cPDI and hPDI increased the refolding yield of lyso- zyme (chaperone activity). Thus, cPDI and hPDI shared similar chaperone and antichaperone activities. In summary, the cPDI cloned from C. marmoreus had similar foldase and chaperone activities as hPDI, suggesting that the biological functions of PDI are highly conserved during evolution. cPDI-catalyzed oxidative folding of tx3a During oxidative folding, oxidized PDI catalyzes disul- fide formation through transferring its active site’s disulfide to dithiols of reduced polypeptide. We ana- lyzed the oxidase activity of cPDI using reduced tx3a (20 lm) as substrate. As shown in Fig. 3, the conotoxin Z Q. Wang et al. Characterization of PDI from Conus marmoreus FEBS Journal 274 (2007) 4778–4787 ª 2007 The Authors Journal compilation ª 2007 FEBS 4779 A B Fig. 1. Comparison of amino acid sequences of PDIs from C. marmoreus and other species. (A) Multiple sequence alignment of PDIs from human, yeast, C. marmoreus,andC. textile. Identical or similar residues are shaded in black or gray. The potential N-terminal signal peptides are boxed; the thioredoxin active sites are underlined; and the ER-retention motif is indicated by underlined dashes. (B) A phylogenetic tree con- structed on the basis of the amino acid sequences of different PDIs, listed with GenBank accession numbers. The scale bar represents 0.1 units. Table 1. Enzymatic activities of cPDI compared with those of hPDI. PDI Reduction activity a [10 2 · (DAÆmin )1 ÆlM PDI )1 )] %of cPDI Oxidase activity b [10 2 · (lM )1 Æmin )1 )] %of cPDI Isomerization activity b (lMÆmin )1 ÆlM PDI )1 ) %of cPDI Conus 5.74 ± 0.36 100 9.24 ± 0.83 100 1.11 ± 0.05 100 Human 5.38 ± 0.52 94 10.18 ± 0.57 110 1.17 ± 0.10 105 a The disulfide reduction activity assay (thiol-protein oxidoreductase) was performed in 0.2 M sodium phosphate buffer (pH 7.5) containing 8m M GSH, 30 lM insulin, 120 lM NADPH, 0.5 units of glutathione reductase, 5 mM EDTA, and 0.7 lM PDI. The absorbance decrease at 340 nm was monitored. b Refolding of the reduced RNase A (final concentration 8.4 lM) was carried out in the refolding buffer (0.1 M Tris ⁄ Cl, pH 8.0, 0.2 mM GSSG, 1 mM GSH, 2 mM EDTA, 4.5 mM cCMP) and catalyzed by appropriate concentrations of PDI (0–10 lM). The absorbance increase at 296 nm was monitored. Characterization of PDI from Conus marmoreus Z Q. Wang et al. 4780 FEBS Journal 274 (2007) 4778–4787 ª 2007 The Authors Journal compilation ª 2007 FEBS tx3a is a 16 residue peptide containing three disulfide bonds [14,15]. Figure 4A shows the HPLC profiles of the cPDI-catalyzed (1 lm) oxidative refolding of reduced tx3a at different refolding stages. The refold- ing was finished after 2 h, and the folding yield was over 90%. As shown in Fig. 4B, when the concentra- tion of cPDI increased, the refolding of reduced tx3a accelerated accordingly. Thus cPDI can catalyze the oxidative folding of reduced conotoxins in vitro.Itis logical to expect that this process also occurs in vivo. The calculated oxidase activity of cPDI was measured as the initial rate of decrease of reduced tx3a (Table 2). Besides cPDI, both oxidized glutathione (GSSG) and molecular oxygen can also oxidize dithiols to form disulfide bonds [16], and both of them are present in the ER lumen. To compare the roles of these different oxidants during the folding of conotoxins, the refolding of reduced tx3a was carried out in three different sys- tems (Fig. 5). In all these systems, the final refolding yields were approximately 90 ± 5%. When molecular oxygen dissolved in the buffer was used as an oxidant, the folding of reduced tx3a was barely detectable at the start and finally reached equilibrium 10 h later (Fig. 5A). As with air oxidation of reduced bovine pan- creatic RNase A (RNase A) [17], a significant lag time ( 60 min) caused by formation of partially and⁄ or fully oxidized intermediates was observed when molec- ular oxygen was used as an oxidant (Fig. 5A). The refolding intermediates were collected, alkylated by N-ethylmaleimide or 4-vinylpyridine, and analyzed by MS, which revealed that these intermediates were com- plex mixtures of one, two or three disulfide isomers Fig. 2. The effect of PDIs on the refolding of lysozyme. The oxida- tive refolding of the denatured ⁄ reduced lysozyme was carried out in the refolding buffer (0.1 M Hepes, pH 7.0, 2 mM EDTA, 5 mM MgCl 2 ,20mM NaCl, 1 mM GSSG, 2 mM GSH) and catalyzed by dif- ferent concentrations of PDI. In the refolding reaction mixture, the final concentration of reduced lysozyme was 10 l M. The refolding was carried out at room temperature for 2 h, and then the lyso- zyme activity was measured. The refolding yields were calculated from the activity recovery on the basis of a standard curve. Fig. 3. The amino acid sequences of tx3a and Tx3.1. tx3a and Tx3.1 have identical disulfide linkages, indicated by connection lines. The asterisk indicates C-terminal amidation. AB Fig. 4. The oxidase activity of cPDI determined by using reduced tx3a as substrate. (A) The HPLC profiles of the tx3a refolding mixture. The refolding of reduced tx3a (20 l M) was performed in the refolding buffer (0.1 M Tris ⁄ Cl, 1 mM EDTA, pH 7.5) containing 0.1 mM GSSG and 1 l M cPDI. At the indicated times, the reaction mixture was acidified and immediately analyzed by RP-HPLC. N indicates the native tx3a. (B) The time course of tx3a refolding catalyzed by different concentrations of cPDI. The amount of reduced tx3a was calculated from its elution peak area. The original data were fitted by Y(t) ¼ e –kt · 100%, where Y is the percentage of the linear form, and t is the refolding time. Z Q. Wang et al. Characterization of PDI from Conus marmoreus FEBS Journal 274 (2007) 4778–4787 ª 2007 The Authors Journal compilation ª 2007 FEBS 4781 (data no shown). When GSSG was used as an oxidant, the folding process was expedited, and the lag time was diminished ( 6 min), as shown in Fig. 5B. When cPDI (at a final concentration of 2 lm) was added to the GSSG refolding system, the refolding process further accelerated (Fig. 5B). The calculated molar specific oxi- dase activities of cPDI and GSSG are listed in Table 2: cPDI was about 268-fold more effective than GSSG in promoting disulfide formation. cPDI-catalyzed disulfide isomerization of swap Tx3.1 Besides having an oxidase function, PDI can also cata- lyze the isomerization of non-native disulfide bonds [2,3]. To date, there are no reports of the PDI-cata- lyzed disulfide isomerization of conotoxins. Here, we used a homogeneous non-native conotoxin isomer, swap Tx3.1 (sTx3.1), to study the isomerase activity of cPDI. As shown in Fig. 3, Tx3.1 is an 18 amino acid conotoxin with three disulfide bonds [14]. During the oxidative refolding of reduced Tx3.1, two major fold- ing products, native Tx3.1 (nTx3.1) and sTx3.1, were formed at the final folding stage (first trace of Fig. 6A). The molecular mass of sTx3.1 as measured by MS was identical to that of nTx3.1. After modifica- tion by N-ethylmaleimide under denatured condition (6 m urea), its molecular mass did not change. Thus, sTx3.1 was a fully oxidized isomer that had similar thermodynamic stability to the native form. We purified sTx3.1 and used it as the cPDI substrate in the isomerase activity assay. The traces b–e in Fig. 6A Table 2. Oxidase and isomerase activities of cPDI measured using conotoxins as substrates. Oxidase a,b Moles of reduced tx3a per mole of oxidase per min (· 10 )2 ) Isomerase b,c Moles of nTx3.1 per mole of isomerase per min (· 10 )3 ) cPDI (1–8 l M) 115.13 ± 9.08 cPDI (0.5–4 lM) 208.17 ± 21.61 GSSG (0.1–1 m M) 0.43 ± 0.04 GSH (0.25–1 mM) 0.12 ± 0.01 a The oxidase activity assay was performed in the refolding buffer (0.1 M Tris ⁄ Cl, 1 mM EDTA, pH 7.5) containing different concentrations of GSSG or cPDI. The refolding mixture of reduced tx3a was analyzed by HPLC using the conditions described in the legend of Fig. 4. b Oxidase or isomerase is a broad definition [16], including any compound that is capable of promoting disulfide formation and isomerization. c The isomerase activity assay was performed in the refolding buffer (0.1 M Tris ⁄ Cl, 1 mM EDTA, pH 7.5) containing different concentrations of GSH or cPDI. The refolding mixture of sTx3.1 was analyzed by HPLC as described in the legend of Fig. 6. Fig. 5. The tx3a refolding catalyzed by different oxidants. (A) Refolding carried out in buffer A (0.1 M Tris ⁄ Cl, pH 7.5, 1 mM EDTA). The filled and open circles denote native and linear tx3a, respectively. (B). Refolding carried out in buffer B (buffer A plus 1 m M GSH and 1 mM GSSG) and in buffer C (buffer B plus 2 l M cPDI). Filled circles, native tx3a in the presence of PDI; open circles, linear tx3a in the presence of PDI; filled squares, native tx3a in the absence of PDI; open squares, linear tx3a in the absence of PDI. At different reaction times, the refolding mixture was acidified and immediately analyzed by RP-HPLC. The amounts of native and linear tx3a were calculated from their elution peak areas. The data are the average of three independent experiments. For the rate of decrease of the linear form, the original data were fitted by Y(t) ¼ e –kt · 100%, where Y is the percentage of the linear form, and t is the refolding time. For the rate of increase of the native form in the presence of cPDI, the original data were fitted by Y(t) ¼ Y max (1 ) e –kt ) · 100%, where Y is the percentage of the native form, and t is the refolding time. For the rate of increase of the native form in the absence of cPDI, the original data were fitted by Y(t) ¼ Y max ⁄ [1 + e –k(t ) a) ] · 100%, where Y is the percentage of the native form, and t is the refolding time. Characterization of PDI from Conus marmoreus Z Q. Wang et al. 4782 FEBS Journal 274 (2007) 4778–4787 ª 2007 The Authors Journal compilation ª 2007 FEBS represent the disulfide isomerization process of sTx3.1 (20 lm) catalyzed by cPDI (at a final concentration of 1 lm). cPDI could accelerate disulfide reshuffle, but it could not shift the equilibrium between nTx3.1 and sTx3.1, which had similar thermodynamic stability. Thus, cPDI could not convert all of the sTx3.1 to the native form. When the concentration of cPDI was increased, the disulfide isomerization process signifi- cantly accelerated (Fig. 6B). The calculated molar spe- cific isomerase activity of cPDI was expressed as the initial rate of increase of nTx3.1 (Table 2). Besides PDI, it is known that GSH, an abundant redox molecule in the ER lumen, can also catalyze disulfide isomerization [16]. As shown in Fig. 7, we compared the isomerase activities of reduced glutathi- one (GSH) and cPDI, and the result showed that cPDI was much more efficient than GSH as an isomerase. The half-times of sTx3.1 disappearance in the presence or absence of cPDI were approximately 3.9 min and 13.5 min, respectively. The half-times of nTx3.1 appearance in the presence or absence of cPDI were about 4.9 min and 21.9 min, respectively. The molar specific isomerase activity of cPDI was about 1700-fold higher than that of GSH (Table 2). Effect of macromolecular crowding on the PDI-catalyzed folding The intracellular environment is highly crowded, con- sisting of various proteins and other macromolecules, the concentration being about 80–300 g ÆL )1 [8]. Thus, the protein folding catalyzed by PDI in vivo occurs in a crowded environment. To mimic the scenario of in vivo PDI-catalyzed folding, the reactions of PDI-cat- alyzed oxidative folding of reduced tx3a (Fig. 8A) and PDI-catalyzed isomerization of sTx3.1 (Fig. 8B) were carried out in a crowded environment, using Ficoll 70 as crowding agent. As shown in Fig. 8, the crowding had little effect on PDI-catalyzed disulfide formation or isomerization of conotoxins. Our results are similar to those obtained with hirudin, which is a 65 amino acid peptide containing three disulfide bonds [18]. Discussion In the present work, we cloned a novel PDI from C. marmoreus. The cPDI shares high sequence homol- ogy with PDIs from C. textile and other organisms. It also has similar biological functions as hPDI, including disulfide reductase, oxidase and isomerase activities, as well as chaperone and antichaperone activities. The high sequence and function conservations support the hypothesis that all of the current PDIs evolved from a common ancestral enzyme [19]. We further analyzed the enzymatic activities of cPDI using its potential endogenous substrates, namely reduced tx3a and sTx3.1. Both tx3a and Tx3.1 belong to the M-1 branch of the M-superfamily. The different branches in the M-superfamily possess different disul- fide linkages [14]. The oxidative folding properties of AB Fig. 6. The isomerase activity of cPDI determined by using sTx3.1 as substrate. (A) (a) The HPLC profile of the refolding mixture of reduced Tx3.1. The refolding was carried out in the refolding buffer (50 m M NH 4 CO 3 , pH 8.0, 5 mM GSH, 0.5 mM GSSG) for 8 h. nTx3.1 and sx3.1 are designated as N and S, respectively. (b–e) The HPLC profiles of sTx3.1 refolding mixtures. The disulfide isomerization of sTx3.1 (20 l M) was carried out in the refolding buffer (0.1 M Tris ⁄ Cl, pH 7.5, 1 mM EDTA, 0.1 mM GSH) and catalyzed by 1 lM cPDI. At different reaction times, the reaction mixture was acidified and immediately analyzed by RP-HPLC. (B) The time course of sTx3.1 isomerization catalyzed by different concentrations of cPDI. The isomerization of sTx3.1 was performed in the refolding buffer (0.1 M Tris ⁄ Cl, pH 7.5, 1 mM EDTA, 0.1 m M GSH) and catalyzed by different concentrations of cPDI. The amount of nTx3.1 was calculated from its elution peak area on HPLC. The original data were fitted by Y(t) ¼ Y max (1 ) e –kt ) · 100%, where Y is the percentage of the native form, and t is the refolding time. Z Q. Wang et al. Characterization of PDI from Conus marmoreus FEBS Journal 274 (2007) 4778–4787 ª 2007 The Authors Journal compilation ª 2007 FEBS 4783 four M-4 branch conotoxins have been thoroughly investigated, and two distinct folding mechanisms have been unveiled [20]. Through comparison of folding kinetics and thermodynamics, the folding mechanism of tx3a and Tx3.1 was found to be similar to that of conotoxins GIIIA and RIIIK [20]. Their refolding follows a slow-rearrangement mechanism, where the partially ⁄ fully oxidized folding intermediates are formed quickly and then converted to the native form slowly. The activity analyses demonstrate that cPDI can greatly accelerate both oxidative folding and disulfide isomerization of conotoxins. The calculated molar spe- cific oxidase and isomerase activities of cPDI are much Fig. 7. The disulfide isomerization of sTx3.1 catalyzed by GSH or by cPDI. (A) The time course of nTx3.1 appearance. (B) The time course of sTx3.1 disappearance. The disulfide isomerization of sTx3.1 was performed in buffer A (0.1 M Tris ⁄ Cl, pH 7.5, 1 mM EDTA, 1 mM GSH) (open circles) or in buffer B (buffer A plus 2 l M cPDI) (filled circles). At different reaction times, the refolding mixture was acidified and immediately analyzed by HPLC. The amounts of nTx3.1 and sTx3.1 were calculated from their elution peak areas. The data are the average of three independent experiments. For the rate of increase of the native form, the original data were fitted by Y(t) ¼ Y max (1 ) e –kt ) · 100%, where Y is the percentage of the native form, and t is the refolding time. For the rate of decrease of the linear form, the original data were fitted by Y(t) ¼ [Y max +(1) Y max )e –kt ] · 100%, where Y is the percentage of the linear form, and t is the refolding time. Fig. 8. Effects of macromolecular crowding on the PDI-catalyzed folding of conotoxins. (A) The oxidative folding of reduced tx3a in the absence (open squares) or presence (filled squares) of crowding agent. The refolding was performed in refolding buffer (0.1 M Tris ⁄ Cl, pH 7.5, 1m M EDTA, 1 mM GSH, 1 mM GSSG, 2 lM cPDI) in the presence or absence of 200 gÆL )1 Ficoll 70. At different reaction times, the refolding mixture was acidified and immediately analyzed by HPLC. The amounts of native and linear tx3a were calculated from their elution peak areas. (B) The isomerization of sTx3.1 in the absence (open squares) or presence (filled squares) of crowding agent. The isomerization was per- formed in refolding buffer (0.1 M Tris ⁄ Cl, pH 7.5, 1 mM EDTA, 0.1 mM GSH, 2 lM cPDI) in the presence or absence of 200 gÆL )1 Ficoll 70. The amounts of nTx3.1 and sTx3.1 were calculated from their elution peak areas. The data are the average of three independent experiments. For the rate of increase of the native form in the presence of cPDI, the original data were fitted by Y(t) ¼ Y max (1 ) e –kt ) · 100%, where Y is the percentage of the native form, and t is the refolding time. For the rate of decrease of the linear form, the original data were fitted by Y(t) ¼ [Y max +(1) Y max )e –kt ] · 100%, where Y is the percentage of the linear form, and t is the refolding time. Characterization of PDI from Conus marmoreus Z Q. Wang et al. 4784 FEBS Journal 274 (2007) 4778–4787 ª 2007 The Authors Journal compilation ª 2007 FEBS higher than those of glutathione (Table 2). The con- centrations of GSH and GSSG in the ER lumen are in the millimolar range, whereas the concentration of PDI is about 200 lm (about 10-fold lower than the concentration of glutathione). This work provides direct evidence that the molar specific oxidase and isomerase activities of cPDI are much higher (268-fold and 1500-fold, respectively) than those of glutathione; hence, the total oxidase and isomerase activities in the ER lumen should be dominated by cPDI. We therefore propose the hypothesis that PDI plays a key role dur- ing in vivo folding of conotoxins. Experimental procedures Materials The plasmid encoding mature hPDI was a generous gift from L. W. Ruddock (Biocenter, University of Oulu, Finland). Ni 2+ -chelating Sepharose Fast Flow resin and Q-Sepharose Fast Flow resin were obtained from Amer- sham Biosciences (Arlington Heights, IL, USA). The RACE kit was obtained from Invitrogen (Carlsbad, CA, USA). Lysozyme, Micrococcus lysodeikticus and RNase A were products of Sigma (St Louis, MO, USA). Other reagents were of analytical grade. Gene cloning of cPDI The full-length cDNA of cPDI was amplified by RT-PCR from total RNAs isolated from the venom ducts of C. marmoreus. The 3¢-end fragment was amplified using a 3¢-RACE adapter primer and a degenerate primer based on the conserved amino acid sequence (WCGHCK) of the thioredoxin-like active site found in other PDIs. The 3¢-RACE product was gel-purified, cloned into pGEM-T easy vector, and sequenced. The nested PCR primers for 5¢-RACE amplification were based on the 3¢-end sequence, and the 5¢-end fragment was amplified using the nested primer and the 5¢-RACE adapter primer. The 5¢-RACE product was also gel-purified, cloned into pGEM-T easy vector, and sequenced. Primers for amplifying the full- length cDNA were designed on the basis of these RACE products. The full-length cDNA of the cPDI was inserted into an expression vector, pET24a, which contains an N-terminal His 6 tag. Expression and purification of cPDI and hPDI The expression plasmid of cPDI was transformed into BL21 (DE3) cells. The transformed E. coli cells were cultured in LB medium containing 25 lgÆmL )1 kanamycin at 37 °C, and the expression was induced by standard procedures. Thereafter, the cells were harvested, resuspended in buffer A (20 mm phosphate buffer, pH 7.5, 0.5 m NaCl) and lysed by sonication. After centrifugation (12 000 g,4°C, 20 min; Hitachi Himac CR22G centrifuge, rotor 46), the superna- tant was loaded onto an Ni 2+ -chelating Sepharose Fast Flow column pre-equilibrated with buffer A. The column was extensively washed with buffer A, and then the nonspe- cifically bound proteins were eluted with buffer B (buffer A plus 20 mm imidazole). Finally, the recombinant cPDI was eluted from the column with buffer C (buffer A plus 250 mm imidazole). The eluted cPDI was dialyzed against 20 mm phosphate buffer (pH 7.5) at 4 °C, and subsequently applied to a Q-Sepharose Fast Flow column pre-equili- brated with 20 mm phosphate buffer (pH 7.5). cPDI was eluted from the ion exchange column using a linear NaCl gradient (0–1 m). The cPDI fraction was collected, analyzed by SDS ⁄ PAGE, dialyzed against distilled water, and stored at ) 80 °C for later use. Expression and purification of human PDI were per- formed as described previously [21], and its purity was ana- lyzed by SDS ⁄ PAGE. Enzymatic activity assays of PDI The thiol-protein oxidoreductase activity of PDI was mea- sured as described previously, using insulin as substrate [22]. The assay was performed in 0.2 m sodium phosphate buffer (pH 7.5) containing 8 mm GSH, 30 lm insulin, 120 lm NADPH, 0.5 units of glutathione reductase, 5 mm EDTA, and 0.7 lm PDI. The assay mixture (without insu- lin and PDI) was equilibrated at 25 °C, and the NADPH oxidation rate was recorded against a reference cuvette con- taining NADPH, EDTA and buffer only. Subsequently, insulin was added, and a stable nonenzymatic rate was recorded. Finally, PDI was added, and the total NADPH oxidation rate was recorded. The oxidase and isomerase activities of PDI were mea- sured using the refolding assay of fully reduced RNase A as previously described [17,23,24]. Briefly, it was performed in the assay solution (0.1 m Tris ⁄ Cl, pH 8.0, 0.2 mm GSSG, 1mm GSH, 2 mm EDTA, 4.5 mm cCMP) at 25 °C. After preincubation, the fully reduced RNase A (8.4 lm) and dif- ferent concentrations of PDI (0–10 lm) were added to the assay solution to initiate refolding. The formation of active RNase A was measured spectrophotometrically by monitor- ing cCMP hydrolysis at 296 nm. During the oxidative refolding, the reduced RNase A was quickly converted to inactive oxidized forms by the oxidase activity of PDI, and these inactive oxidized forms were then slowly converted to active native form by the isomerase activity of PDI [23]. The lag time before appearance of the active RNase A indicates the oxidase activity, which corresponds to the x-intercept of the RNase activity plot. The oxidase activity matches the slope of the linear plot of reciprocal of lag times against the PDI concentrations in units of lm )1 Æmin )1 . The Z Q. Wang et al. Characterization of PDI from Conus marmoreus FEBS Journal 274 (2007) 4778–4787 ª 2007 The Authors Journal compilation ª 2007 FEBS 4785 isomerase activity was determined from the linear increase of enzymatically active RNase A after the lag time. The chaperone and antichaperone activities of PDI were analyzed using the renaturation of reduced ⁄ denatured lyso- zyme [13]. The lysozyme activity was measured at 30 °Cby following the absorbance decrease at 450 nm of the M. lyso- deikticus suspension (0.25 mgÆmL )1 in 67 mm sodium phosphate buffer, pH 6.2, and 0.1 m NaCl). Peptide synthesis Conotoxins tx3a and Tx3.1 [14,15] were chemically synthe- sized by using the Fmoc method on an ABI 433 A peptide synthesizer. The crude reduced peptides were purified by C 18 reversed-phase HPLC and lyophilized. The identity of each peptide was confirmed by MS. The molecular masses of linear tx3a and Tx3.1 were 1660.5 Da (theoretical value: 1660.8 Da) and 2158.6 Da (theoretical value: 2158.44 Da). Preparation of homogeneous sTx3.1 Refolding of the linear Tx3.1 was carried out in the refold- ing buffer (50 mm NH 4 CO 3 ,5mm GSH, 0.5 mm GSSG, pH 8.0) at 25 °C for 8 h, and the refolding mixture was analyzed by RP-HPLC. Two major disulfide isomers, nTx3.1 and sTx3.1, were collected and lyophilized. PDI-catalyzed oxidative folding and isomerization of conotoxins The oxidative folding or isomerization of conotoxins was performed in the refolding buffer (0.1 m Tris ⁄ Cl, pH 7.5, plus appropriate concentrations of GSH, GSSG and cPDI as indicated in the figure legends) at ambient temperature (23–25 °C). The folding reaction was initiated by adding the peptide stock solution to the final concentration of 20 lm. At different reaction times, the folding reaction was quenched by adding formic acid to the final concentration of 8%. The reaction mixture was immediately analyzed using C 18 analytical RP-HPLC. The amounts of linear form, native form and swap form were calculated from their integrated elution peak areas, and the PDI’s oxidase and isomerase activities were expressed as the initial rate of decrease of linear tx3a and the initial rate of increase of nTx3.1, respectively. To investigate the effect of macromo- lecular crowding on the PDI-catalyzed oxidative folding or disulfide isomerization of conotoxins, the folding reaction was carried out as described above, except that 200 gÆL )1 of Ficoll 70 was added to the refolding buffer. Acknowledgements The authors wish to acknowledge Professors C. C. Wang, D. F. Cui, Q. Y. Dai and L. W. Ruddock for their generous support for this work. This work was supported by the National Basic Research Program of China (2004CB719904). References 1 Ellgaard L & Ruddock LW (2005) The human protein disulphide isomerase family: substrate interactions and functional properties. EMBO Rep 6 , 28–32. 2 Wilkinson B & Gilbert HF (2004) Protein disulfide isomerase. Biochim Biophys Acta 1699, 35–44. 3 Kersteen EA & Raines RT (2003) Catalysis of protein folding by protein disulfide isomerase and small-mole- cule mimics. Antioxid Redox Signal 5, 413–424. 4 Song JL & Wang CC (1995) Chaperone-like activity of protein disulfide-isomerase in the refolding of rhoda- nese. Eur J Biochem 231, 312–316. 5 Bulaj G, Buczek O, Goodsell I, Jimenez EC, Kranski J, Nielsen JS, Garrett JE & Olivera BM (2003) Efficient oxidative folding of conotoxins and the radiation of venomous cone snails. Proc Natl Acad Sci USA 100 (Suppl. 2), 14562–14568. 6 Price-Carter M, Gray WR & Goldenberg DP (1996) Folding of omega-conotoxins. 2. Influence of precursor sequences and protein disulfide isomerase. Biochemistry 35, 15547–15557. 7 Craig AG, Bandyopadhyay P & Olivera BM (1999) Post-translationally modified neuropeptides from Conus venoms. Eur J Biochem 264, 271–275. 8 van den Berg B, Ellis RJ & Dobson CM (1999) Effects of macromolecular crowding on protein folding and aggregation. EMBO J 18, 6927–6933. 9 Gething MJ & Sambrook J (1992) Protein folding in the cell. Nature 355, 33–45. 10 Ellis J (1987) Proteins as molecular chaperones. Nature 328, 378–379. 11 Garrett JE, Buczek O, Watkins M, Olivera BM & Bulaj G (2005) Biochemical and gene expression analyses of conotoxins in Conus textile venom ducts. Biochem Biophys Res Commun 328, 362–367. 12 Nielsen H, Engelbrecht J, Brunak S & von Heijne G (1997) Identification of prokaryotic and eukaryotic signal peptides and prediction of their cleavage sites. Protein Eng 10, 1–6. 13 Song J, Quan H & Wang C (1997) Dependence of the anti-chaperone activity of protein disulphide isomerase on its chaperone activity. Biochem J 328 (3), 841–846. 14 Han YH, Wang Q, Jiang H, Liu L, Xiao C, Yuan DD, Shao XX, Dai QY, Cheng JS & Chi CW (2006) Charac- terization of novel M-superfamily conotoxins with new disulfide linkage. FEBS J 273, 4972–4982. 15 Corpuz GP, Jacobsen RB, Jimenez EC, Watkins M, Walker C, Colledge C, Garrett JE, McDougal O, Li W, Gray WR et al. (2005) Definition of the M-conotoxin Characterization of PDI from Conus marmoreus Z Q. Wang et al. 4786 FEBS Journal 274 (2007) 4778–4787 ª 2007 The Authors Journal compilation ª 2007 FEBS superfamily: characterization of novel peptides from mol- luscivorous Conus venoms. Biochemistry 44, 8176–8186. 16 Lu BY & Chang JY (2005) Assay of disulfide oxidase and isomerase based on the model of hirudin folding. Anal Biochem 339, 94–103. 17 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. 18 Chang JY, Lu BY & Lai PH (2006) Oxidative folding of hirudin in human serum. Biochem J 394, 249–257. 19 Kanai S, Toh H, Hayano T & Kikuchi M (1998) Molecular evolution of the domain structures of protein disulfide isomerases. J Mol Evol 47, 200–210. 20 Fuller E, Green BR, Catlin P, Buczek O, Nielsen JS, Olivera BM & Bulaj G (2005) Oxidative folding of conotoxins sharing an identical disulfide bridging frame- work. FEBS J 272, 1727–1738. 21 Lappi AK, Lensink MF, Alanen HI, Salo KE, Lobell M, Juffer AH & Ruddock LW (2004) A conserved arginine plays a role in the catalytic cycle of the protein disulphide isomerases. J Mol Biol 335, 283– 295. 22 Lambert N & Freedman RB (1983) Kinetics and specificity of homogeneous protein disulphide- isomerase in protein disulphide isomerization and in thiol-protein-disulphide oxidoreduction. Biochem J 213, 235–243. 23 Lyles MM & Gilbert HF (1991) Catalysis of the oxida- tive folding of ribonuclease A by protein disulfide iso- merase: dependence of the rate on the composition of the redox buffer. Biochemistry 30, 613–619. 24 Wilkinson B, Xiao R & Gilbert HF (2005) A structural disulfide of yeast protein-disulfide isomerase destabilizes the active site disulfide of the N-terminal thioredoxin domain. J Biol Chem 280, 11483–11487. Z Q. Wang et al. Characterization of PDI from Conus marmoreus FEBS Journal 274 (2007) 4778–4787 ª 2007 The Authors Journal compilation ª 2007 FEBS 4787 . Molecular cloning, expression and characterization of protein disulfide isomerase from Conus marmoreus Zhi-Qiang Wang 1 , Yu-Hong. physiologic roles of pro- tein disulfide isomerase in the folding of conotoxins, we have cloned a novel full-length protein disulfide isomerase from Conus marmoreus.