BiP and Immunoglobulin Light Chain Cooperate to Control the Folding of Heavy Chain and Ensure the Fidelity of Immunoglobulin Assembly

The immunoglobulin (Ig) molecule is composed of two identical heavy chains and two identical light chains (H2L2). Transport of this heteromeric complex is dependent on the correct assembly of the component parts, which is controlled, in part, by the association of incompletely assembled Ig heavy chains with the endoplasmic reticulum (ER) chaperone, BiP. Although other heavy chain-constant domains interact transiently with BiP, in the absence of light chain synthesis, BiP binds stably to the first constant domain (CH1) of the heavy chain, causing it to be retained in the ER. Using a simplified two-domain Ig heavy chain (VH-CH1), we have determined why BiP remains bound to free heavy chains and how light chains facilitate their transport. We found that in the absence of light chain expression, the CH1 domain neither folds nor forms its intradomain disulfide bond and therefore remains a substrate for BiP. In vivo, light chains are required to facilitate both the folding of the CH1 domain and the release of BiP. In contrast, the addition of ATP to isolated BiP-heavy chain complexes in vitro causes the release of BiP and allows the CH1 domain to fold in the absence of light chains. Therefore, light chains are not intrinsically essential for CH1 domain folding, but play a critical role in removing BiP from the CH1 domain, thereby allowing it to fold and Ig assembly to proceed. These data suggest that the assembly of multimeric protein complexes in the ER is not strictly dependent on the proper folding of individual subunits; rather, assembly can drive the complete folding of protein subunits.     

Heavy chain binding protein recognizes incompletely disulfide-bonded forms of vesicular stomatitis virus G protein

To investigate the function of heavy chain binding protein (BiP, GRP 78) in the endoplasmic reticulum, we have characterized its interaction with a model plasma membrane glycoprotein, the G protein of vesicular stomatitis virus. We used a panel of well characterized mutant G proteins and immunoprecipitation with anti-BiP antibodies to determine if BiP interacted with newly synthesized G protein and/or mutant G proteins retained in the endoplasmic reticulum. We made three major observations: 1) BiP bound transiently to folding intermediates of wild-type G protein which were incompletely disulfide-bonded; 2) BiP did not bind stably to all mutant G proteins which remain in the endoplasmic reticulum; and 3) BiP bound stably only to mutant G proteins which do not form correct intrachain disulfide bonds.     

Role of glycoprotein PE2 in formation and maturation of the Sindbis virus spike.

Sindbis virus envelope assembly is a multistep process resulting in the maturation of a rigid, highly ordered T=4 icosahedral protein lattice containing 80 spikes composed of trimers of E1-E2 heterodimers. Intramolecular disulfide bonds within E1 stabilize E1-E1 associations required for envelope formation and maintenance of the envelope's structural integrity. The structural integrity of the envelope protein lattice is resistant to reduction by dithiothreitol (DTT), indicating that E1 disulfides which stabilize structural domains become inaccessible to DTT at some point during virus maturation. The development of E1 resistance to DTT occurs prior to the completion of E1 folding and is temporally correlated with spike assembly in the endoplasmic reticulum. From these data we have predicted that in the final stages of spike assembly, E1 intramolecular disulfides, which stabilize the structural integrity of the envelope protein lattice, are buried within the spike and become inaccessible to the reductive activity of DTT. The spike is formed prior to the completion of E1 folding, and we have suggested that PE2 (the precursor to E2) may play a critical role in E1 folding after PE2-E1 oligomer formation has occurred. In this study we have investigated the role of PE2 in E1 folding, oligomer formation, and development of E1 resistance to both protease digestion and reduction by DTT by using a Sindbis virus replicon (SINrep/E1) which allows for the expression of E1 in the presence of truncated PE2. Through pulse-chase analysis of both Sindbis virus- and SINrep/E1-infected cells, we have determined that the folding of E1 into a trypsin-resistant conformation and into its most compact and stable form is not dependent upon association of E1 with PE2. However, E1 association with PE2 is required for oligomer formation, the export of E1 from the endoplasmic reticulum, and E1 acquisition of resistance to DTT.     

Transient aggregation of nascent thyroglobulin in the endoplasmic reticulum: relationship to the molecular chaperone, BiP

Because of its unusual length, nascent thyroglobulin (Tg) requires a long time after translocation into the endoplasmic reticulum (ER) to assume its mature tertiary structure. Thus, Tg is an ideal molecule for the study of protein folding and export from the ER, and is an excellent potential substrate for molecular chaperones. During the first 15 min after biosynthesis, Tg is found in transient aggregates with and without interchain disulfide bonds, which precede the formation of free monomers (and ultimately dimers) within the ER. By immunoprecipitation, newly synthesized Tg was associated with the binding protein (BiP); association was maximal at the earliest chase times. Much of the Tg released from BiP by the addition of Mg-ATP was found in aggregates containing interchain disulfide bonds; other BiP-associated Tg represented non-covalent aggregates and unfolded free monomers. Importantly, the immediate precursor to Tg dimer was a compact monomer which did not associate with BiP. The average stoichiometry of BiP/Tg interaction involved nearly 10 BiP molecules per Tg molecule. Cycloheximide was used to reduced the ER concentration of Tg relative to chaperones, with subsequent removal of the drug in order to rapidly restore Tg synthesis. After this treatment, nascent Tg aggregates were no longer detectable. The data suggest a model of folding of exportable proteins in which nascent polypeptides immediately upon translocation into the ER interact with BiP. Early interaction with BiP may help in presenting nascent polypeptides to other helper molecules that catalyze folding, thereby preventing aggregation or driving aggregate dissolution in the ER.     

Mapping the major interaction between binding protein and Ig light chains to sites within the variable domain.

Newly synthesized Ig chains are known to interact in vivo with the binding protein (BiP), a major peptide-binding chaperone in the endoplasmic reticulum. The predominant interactions between the light chain and BiP are observed early in the folding pathway, when the light chain is either completely reduced, or has only one disulfide bond. In this study, we describe the in vitro reconstitution of BiP binding to the variable domain of light chains (VL). Binding of deliberately unfolded VL was dramatically more avid than that of folded VL, mimicking the interaction in vivo. Furthermore, VL binding was inhibited by addition of ATP, was competed with excess unlabeled VL, and was demonstrated with several different VL proteins. Using this assay, peptides derived from the VL sequence were tested experimentally for their ability to bind BiP. Four peptides from both beta sheets of VL were shown to bind BiP specifically, two with significantly higher affinity. As few as these two peptide sites, one from each beta sheet of VL, are sufficient to explain the association of BiP with the entire light chain. These results suggest how BiP directs the folding of Ig in vivo and how it may be used in shaping the B cell repertoire.     

Disulfide bridge-mediated folding of Sindbis virus glycoproteins.

The Sindbis virus envelope is composed of 80 E1-E2 (envelope glycoprotein) heterotrimers organized into an icosahedral protein lattice with T=4 symmetry. The structural integrity of the envelope protein lattice is maintained by E1-E1 interactions which are stabilized by intramolecular disulfide bonds. Structural domains of the envelope proteins sustain the envelope's icosahedral lattice, while functional domains are responsible for virus attachment and membrane fusion. We have previously shown that within the mature Sindbis virus particle, the structural domains of the envelope proteins are significantly more resistant to the membrane-permeative, sulfhydryl-reducing agent dithiothreitol (DTT) than are the functional domains (R. P. Anthony, A. M. Paredes, and D. T. Brown, Virology 190:330-336, 1992). We have used DTT to probe the accessibility of intramolecular disulfides within PE2 (the precursor to E2) and E1, as these proteins fold and are assembled into the spike heterotrimer. We have determined through pulse-chase analysis that intramolecular disulfide bonds within PE2 are always sensitive to DTT when the glycoproteins are in the endoplasmic reticulum. The reduction of these disulfides results in the disruption of PE2-E1 associations. E1 acquires increased resistance to DTT as it folds through a series of disulfide intermediates (E1alpha, -beta, and -gamma) prior to assuming its native and most compact conformation (E1epsilon). The transition from a DTT-sensitive form into a form which exhibits increased resistance to DTT occurs after E1 has folded into its E1beta conformation and correlates temporally with the dissociation of BiP-E1 complexes and the formation of PE2-E1 heterotrimers. We propose that the disulfide bonds within E1 which stabilize the protein domains required for maintaining the structural integrity of the envelope protein lattice form early within the folding pathway of E1 and become inaccessible to DTT once the heterotrimer has formed.     

Involvement of Endoplasmic Reticulum Chaperones in the Folding of Hepatitis C Virus Glycoproteins

The hepatitis C virus (HCV) genome encodes two envelope glycoproteins (E1 and E2) which interact noncovalently to form a heterodimer (E1-E2). During the folding and assembly of HCV glycoproteins, a large portion of these proteins are trapped in aggregates, reducing the efficiency of native E1-E2 complex assembly. To better understand this phenomenon and to try to increase the efficiency of HCV glycoprotein folding, endoplasmic reticulum chaperones potentially interacting with these proteins were studied. Calnexin, calreticulin, and BiP were shown to interact with E1 and E2, whereas no interaction was detected between GRP94 and HCV glycoproteins. The association of HCV glycoproteins with calnexin and calreticulin was faster than with BiP, and the kinetics of interaction with calnexin and calreticulin were very similar. However, calreticulin and BiP interacted preferentially with aggregates whereas calnexin preferentially associated with monomeric forms of HCV glycoproteins or noncovalent complexes. Tunicamycin treatment inhibited the binding of HCV glycoproteins to calnexin and calreticulin, indicating the importance of N-linked oligosaccharides for these interactions. The effect of the co-overexpression of each chaperone on the folding of HCV glycoproteins was also analyzed. However, the levels of native E1-E2 complexes were not increased. Together, our data suggest that calnexin plays a role in the productive folding of HCV glycoproteins whereas calreticulin and BiP are probably involved in a nonproductive pathway of folding.     

Adenosine nucleotides and the regulation of GRP94-client protein interactions

The molecular chaperone heat shock protein 90 (Hsp90) serves essential roles in the regulation of signaling protein function, trafficking, and turnover. Hsp90 function is intimately linked to intrinsic ATP binding and hydrolysis activities, the latter of which is under the regulatory control of accessory factors. Glucose-regulated protein of 94 kDa (GRP94), the endoplasmic reticulum Hsp90, is highly homologous to cytosolic Hsp90. However, neither accessory factors nor adenosine nucleotides have been clearly implicated in the regulation of GRP94-client protein interactions. In the current study, the structural and regulatory consequences of adenosine nucleotide binding to GRP94 were investigated. We report that apo-GRP94 undergoes a time- and temperature-dependent tertiary conformational change that exposes a site(s) of protein-protein interaction; ATP, ADP, and radicicol markedly suppress this conformational change. In concert with these findings, ATP and ADP act identically to suppress GRP94 homooligomerization, as well as both local and global conformational activity. To identify a role(s) for ATP or ADP in the regulation of GRP94-client protein interactions, immunoglobulin (Ig) heavy chain folding intermediates containing bound GRP94 and immunoglobulin binding protein (BiP) were isolated from myeloma cells, and the effects of adenosine nucleotides on chaperone-Ig heavy chain interactions were examined. Whereas ATP elicited efficient release of BiP from both wild-type and mutant Ig heavy chain intermediates, GRP94 remained in stable association with Ig heavy chains in the presence of ATP or ADP. On the basis of these data, we propose that structural maturation of the client protein substrate, rather than ATP binding or hydrolysis, serves as the primary signal for dissociation of GRP94-client protein complexes.     

Folding of rabies virus glycoprotein: epitope acquisition and interaction with endoplasmic reticulum chaperones.

Four well-characterized monoclonal antibodies (MAbs) directed against rabies virus glycoprotein (G) were used to study G folding in vivo. Two of the MAbs were able to immunoprecipitate incompletely oxidized folding intermediates. The two others recognized G only after folding was completed. By using these MAbs, the ability of G to undergo low-pH-induced conformational changes during folding was also investigated. It appeared that some domains acquire this ability before folding is completed. In addition, interactions between unfolded G and some of the molecular chaperones were analyzed. Unfolded G was associated with BiP and calnexin. Association with BiP was maximal immediately after the pulse, whereas association with calnexin was maximal after 5 to 10 min of chase. The effects of tunicamycin and castanospermine on chaperone binding and folding were also studied. In the presence of both drugs, calnexin binding was reduced, consistent with the view that calnexin specifically recognizes monoglucosylated oligosaccharides, but some residual binding was still observed, indicating that calnexin also recognizes the polypeptide chain. In the presence of both drugs, association with BiP was increased and prolonged and folding was impaired. However, the global effects of the drugs were different, since folding was much more efficient in the presence of castanospermine than in the presence of tunicamycin. Taken together, these results provide the basis to draw a schematic view of rabies virus glycoprotein folding.     

Relaxation of replication control in chaperone-independent initiator mutants of plasmid P1.

Escherichia coli chaperones DnaJ, DnaK and GrpE increase P1 plasmid initiator binding to the origin by promoting initiator folding. The binding allows initiation and also promotes pairing of origins which is believed to control initiation frequency. Chaperone-independent DNA binding mutants are often defective in replication control. We show here that these mutants have increased rates of association for DNA binding and defects in origin pairing. The increases in association rates were found to be due either to increased protein folding into active forms or to increases in the association rate constant, kon. Since the dissociation rate constants for DNA release with these mutants are not changed, it is unlikely that the DNA binding domain is affected. The pairing domain may thus control replication and modulate DNA binding. The role of the pairing domain in DNA binding can be significant in vivo as the selection for chaperone-independent binding favors pairing-defective mutants.     

The formation of intramolecular disulfide bridges is required for induction of the Sindbis virus mutant ts23 phenotype.

The Sindbis virus envelope protein spike is a hetero-oligomeric complex composed of a trimer of glycoprotein E1-E2 heterodimers. Spike assembly is a multistep process which occurs in the endoplasmic reticulum (ER) and is required for the export of E1 from the ER. PE2 (precursor to E2), however, can transit through the secretory pathway and be expressed at the cell surface in the absence of E1. Although oligomer formation does not appear to be required for the export of PE2, there is evidence that defects in E1 folding can affect PE2 transit from the ER. Temperature-sensitive mutant ts23 of Sindbis virus contains two amino acid substitutions in E1, while PE2 and capsid protein have the wild-type sequence; however, at the nonpermissive temperature, both E1 and PE2 are retained within the ER and can be isolated in protein aggregates with the molecular chaperone GRP78-BiP. We previously demonstrated that the temperature sensitivity for ts23 was lost as oligomer formation took place at the permissive temperature, suggesting that temperature sensitivity is initiated early in the process of viral spike assembly (M. Carleton and D. T. Brown, J. Virol. 70:952-959, 1996). Experiments described herein investigated the defects in envelope protein maturation that occur in ts23-infected cells and which result in retention of both envelope proteins in the ER. The data demonstrate that in ts23-infected cells incubated at the nonpermissive temperature, E1 folding is disrupted early after synthesis, resulting in the rapid incorporation of both E1 and PE2 into disulfide-stabilized aggregates. Furthermore, the aberrant E1 conformation which is responsible for induction of the ts phenotype requires the formation of intramolecular disulfide bridges formed prior to E1 association with PE2 and the completion of E1 folding.     

Posttranslational folding of vesicular stomatitis virus G protein in the ER: involvement of noncovalent and covalent complexes

In this study, we show that posttranslational folding of Vesicular Stomatitis virus G protein subunits can involve noncovalent, multimeric complexes as transient intermediates. The complexes are heterogeneous in size (4-21S20,W), contain several G glycopolypeptides, and are associated with BiP/GRP78. The newly synthesized, partially intrachain disulfide-bonded G proteins enter these complexes immediately after chain termination, and are released 1-4 min later as fully oxidized, trimerization-competent monomers. These monomers are properly folded, judging by their binding of conformation-specific mAbs. When the G protein is translated in the presence of DTT, it remains reduced, largely unfolded and aggregated in the ER, but it can fold successfully when the DTT is removed. In this case, contrary to normal folding, the aggregates become transiently disulfide cross-linked. We also demonstrated that the fidelity of the folding process is dependent on metabolic energy. Finally, we established that the G protein of the folding mutant of the Vesicular Stomatitis virus, ts045, is blocked at a relatively late step in the folding pathway and remains associated with oligomeric, BiP/GRP78-containing folding complexes.     

Grp94 Works Upstream of BiP in Protein Remodeling Under Heat Stress

Hsp90 and Hsp70 are highly conserved molecular chaperones that promote the proper folding and activation of substrate proteins that are often referred to as clients. The two chaperones functionally collaborate to fold specific clients in an ATP-dependent manner. In eukaryotic cytosol, initial client folding is done by Hsp70 and its co-chaperones, followed by a direct transfer of client refolding intermediates to Hsp90 for final client processing. However, the mechanistic details of collaboration of organelle specific Hsp70 and Hsp90 are lacking. This work investigates the collaboration of the endoplasmic reticulum (ER) Hsp70 and Hsp90, BiP and Grp94 respectively, in protein remodeling using in vitro refolding assays. We show that under milder denaturation conditions, BiP collaborates with its co-chaperones to refold misfolded proteins in an ATP-dependent manner. Grp94 does not play a major role in this refolding reaction. However, under stronger denaturation conditions that favor aggregation, Grp94 works in an ATP-independent manner to bind and hold misfolded clients in a folding competent state for subsequent remodeling by the BiP system. We also show that the collaboration of Grp94 and BiP is not simply a reversal of the eukaryotic refolding mechanism since a direct interaction of Grp94 and BiP is not required for client transfer. Instead, ATP binding but not hydrolysis by Grp94 facilitates the release of the bound client, which is then picked up by the BiP system for subsequent refolding in a Grp94-independent manner.     

The nucleotide exchange factors Grp170 and Sil1 induce cholera toxin release from BiP to enable retrotranslocation

Cholera toxin (CT) intoxicates cells by trafficking from the cell surface to the endoplasmic reticulum (ER), where the catalytic CTA1 subunit hijacks components of the ER-associated degradation (ERAD) machinery to retrotranslocate to the cytosol and induce toxicity. In the ER, CT targets to the ERAD machinery composed of the E3 ubiquitin ligase Hrd1-Sel1L complex, in part via the activity of the Sel1L-binding partner ERdj5. This J protein stimulates BiP''s ATPase activity, allowing BiP to capture the toxin. Presumably, toxin release from BiP must occur before retrotranslocation. Here, using loss-and gain-of-function approaches coupled with binding studies, we demonstrate that the ER-resident nucleotide exchange factors (NEFs) Grp170 and Sil1 induce CT release from BiP in order to promote toxin retrotranslocation. In addition, we find that after NEF-dependent release from BiP, the toxin is transferred to protein disulfide isomerase; this ER redox chaperone is known to unfold CTA1, which allows the toxin to cross the Hrd1-Sel1L complex. Our data thus identify two NEFs that trigger toxin release from BiP to enable successful retrotranslocation and clarify the fate of the toxin after it disengages from BiP.     

A phaseolin domain involved directly in trimer assembly is a determinant for binding by the chaperone BiP

The binding protein (BiP; a member of the heat-shock 70 family) is a major chaperone of the endoplasmic reticulum (ER). Interactions with BiP are believed to inhibit unproductive aggregation of newly synthesized secretory proteins during folding and assembly. In vitro, BiP has a preference for peptide sequences enriched in hydrophobic amino acids, which are expected to be exposed only in folding and assembly intermediates or in defective proteins. However, direct information regarding sequences recognized in vivo by BiP on real proteins is very limited. We have shown previously that newly synthesized monomers of the homotrimeric storage protein phaseolin associate with BiP and that phaseolin trimerization in the ER abolishes such interactions. Using different phaseolin constructs and green fluorescent protein (GFP) fusion proteins, we show here that one of the two alpha-helical regions of polypeptide contact in phaseolin trimers (35 amino acids located close to the C terminus and containing three potential BiP binding sites) effectively promotes BiP association with phaseolin and with secretory GFP fusions expressed in transgenic tobacco or in transfected protoplasts. We also show that overexpressed BiP transiently sequesters phaseolin polypeptides. We conclude that one of the regions of monomer contact is a BiP binding determinant and suggest that during the synthesis of phaseolin, the association with BiP and trimer formation are competing events. Finally, we show that the other, internal region of contact between monomers is necessary for phaseolin assembly in vivo and contains one potential BiP binding site.     

Kinetic model of BiP- and PDI-mediated protein folding and assembly

A mechanism for heavy chain binding protein (BiP)- and protein disulfide isomerase (PDI)- mediated protein folding and assembly has been proposed. It considers BiP chaperoning action and PDI catalytic activity. A kinetic model has been developed based on the proposed mechanism. The model was used for quantifying the influence of polypeptide concentration and ratio, and the effect of BiP and PDI concentration on the kinetics of folding and assembly. An optimum value for polypeptide concentration that minimizes assembly times was found, and different kinetic behaviors were identified for polypeptide concentrations higher or lower than the optimum. Pulse-chase experiments and the dependence of assembly time on unassembled polypeptides ratio predicted by the model are similar to those found during in vitro and in vivo folding and assembly of antibodies and human chorionic gonadotropin (hCG), as well as bovine pancreatic trypsin inhibitor (BPTI) folding. The model also explains the increase in folding and assembly rates during overexpression of BiP and PDI.     

In vivo expression of mammalian BiP ATPase mutants causes disruption of the endoplasmic reticulum.

BiP possesses ATP binding/hydrolysis activities that are thought to be essential for its ability to chaperone protein folding and assembly in the endoplasmic reticulum (ER). We have produced a series of point mutations in a hamster BiP clone that inhibit ATPase activity and have generated a species-specific anti-BiP antibody to monitor the effects of mutant hamster BiP expression in COS monkey cells. The enzymatic inactivation of BiP did not interfere with its ability to bind to Ig heavy chains in vivo but did inhibit ATP-mediated release of heavy chains in vitro. Immunofluorescence staining and electron microscopy revealed vesiculation of the ER membranes in COS cells expressing BiP ATPase mutants. ER disruption was not observed when a "44K" fragment of BiP that did not include the protein binding domain was similarly mutated but was observed when the protein binding region of BiP was expressed without an ATP binding domain. This suggests that BiP binding to target proteins as an inactive chaperone is responsible for the ER disruption. This is the first report on the in vivo expression of mammalian BiP mutants and is demonstration that in vitro-identified ATPase mutants behave as dominant negative mutants when expressed in vivo.     

In vivo analysis of the lumenal binding protein (BiP) reveals multiple functions of its ATPase domain

The endoplasmic reticulum (ER) chaperone binding protein (BiP) binds exposed hydrophobic regions of misfolded proteins. Cycles of ATP hydrolysis and nucleotide exchange on the ATPase domain were shown to regulate the function of the ligand-binding domain in vitro. Here we show that ATPase mutants of BiP with defective ATP-hydrolysis (T46G) or ATP-binding (G235D) caused permanent association with a model ligand, but also interfered with the production of secretory, but not cytosolic, proteins in vivo. Furthermore, the negative effect of BiP(T46G) on secretory protein synthesis was rescued by increased levels of wild-type BiP, whereas the G235D mutation was dominant. Unexpectedly, expression of a mutant BiP with impaired ligand binding also interfered with secretory protein production. Although mutant BiP lacking its ATPase domain had no detrimental effect on ER function, expression of an isolated ATPase domain interfered with secretory protein synthesis. Interestingly, the inhibitory effect of the isolated ATPase was alleviated by the T46G mutation and aggravated by the G235D mutation. We propose that in addition to its role in ligand release, the ATPase domain can interact with other components of the protein translocation and folding machinery to influence secretory protein synthesis.     

GroEL-assisted dehydrogenase folding mediated by coenzyme is ATP-independent.

It has been commonly accepted that GroEL functions as a chaperone by modulation of its affinity for folding intermediates through binding and hydrolysis of ATP. However, we have found that NAD, as a coenzyme of d-glyceraldehyde-3-phosphate dehydrogenase (GAPDH), also stimulates the discharge of GAPDH folding intermediate from its stable complex with GroEL formed in the absence of ATP and assists refolding with the same yield as ATP/Mg(2+) does. The reactivation further increases when ATP is also present, but addition of Mg(2+) has no more effect. NADP, a coenzyme of glucose-6-phosphate dehydrogenase, also releases its folding intermediates from GroEL and increases reactivation. Different from ATP, NAD triggers the release of GAPDH intermediates bound by GroEL via binding with GAPDH itself but not with GroEL, and the released intermediates all folded to native molecules without the formation of aggregation. The collaborative effects of coenzyme and GroEL mediate GroEL-assisted dehydrogenase folding in an ATP-independent way.     

Peptide models of four possible insulin folding intermediates with two disulfides

The single-chain insulin (PIP) can spontaneously fold into native structure through preferred kinetic intermediates. During refolding, pairing of the first disulfide A20-B19 is highly specific, whereas pairing of the second disulfide is likely random because two two-disulfide intermediates have been trapped. To get more details of pairing property of the second disulfide, four model peptides of possible folding intermediates with two disulfides were prepared by protein engineering, and their properties were analyzed. The four model peptides were named [A20-B19, A7-B7]PIP, [A20-B19, A6-B7]PIP, [A20-B19, A6-A11]PIP, and [A20-B19, A7-A11]PIP according to their remaining disulfides. The four model peptides all adopt partially folded structure with moderate conformational differences. In redox buffer, the disulfides of the model peptides are more easily reduced than those of the wild-type PIP. During in vitro refolding, the reduced model peptides share similar relative folding rates but different folding yields: The refolding efficiency of the reduced [A20-B19, A7-A11]PIP is about threefold lower than that of the other three peptides. The present results indicate that the folding intermediates corresponding to the present model peptides all adopt partially folded conformation, and can be formed during PIP refolding, but the chance of forming the intermediate with disulfide [A20-B19, A7-A11] is much lower than that of forming the other three intermediates.