Ultrafast conformational dynamics in cyclic azobenzene peptides of increased flexibility

Structural changes of peptides containing the azobenzene dye 4-aminomethyl-phenylazobenzoic acid (AMPB) are studied with ultrafast spectroscopy. AMPB peptides are a new class of molecules where the photoisomerizable dye azobenzene is linked to the peptide moiety via a flexible methylene spacer. The ultrafast reactions in the femtosecond to nanosecond time domain are investigated for the optical switch AMPB, a linear and cyclic octapeptide, and a bicyclic octapeptide containing an additional disulfide bridge. These molecules with increasing conformational constraints are studied for the cis to trans and the trans to cis photoreactions. For the cis to trans reaction the isomerization of the chromophore occurs fast in the 1-ps range, whereas it is slower (10-ps range) in the trans to cis reaction. In all peptides the structural changes of the chromophore lead to modifications in the peptide structure in the 10-ps-1-ns time range. The results indicate that the chromophore AMPB acts simultaneously as a fast molecular switch and as a sensor for initial conformational dynamics in the peptide. Experiments in the mid-infrared range where the structural changes of the peptide backbone are directly observed demonstrate that the essential part of the structural dynamics in the bicyclic AMPB peptide occurs faster than 10 ns.     

Photomodulation of conformational states. IV. Integrin-binding RGD-peptides with (4-aminomethyl)phenylazobenzoic acid as backbone constituent

In previous studies we have investigated octapeptides backbone-cyclized by (4-amino)phenyl azobenzoic acid (APB) or (4-aminomethyl)phenylazobenzoic acid (AMPB) and containing the active-site sequence Cys-Ala-Thr-Cys-Asp from the thioredoxin reductase. The conformational and redox properties of these peptides were strongly dependent on the isomeric state of the azobenzene chromophore. Using the same approach we were successful in constructing photoresponsive ligands for alphavbeta3 integrin containing the Arg-Gly-Asp (RGD) sequence as binding motif. For achieving maximal conformational restriction of the peptide a reduced ring size compared to our previous azobenzene peptides was employed in the cyclic peptide c[Asp-D-Phe-Val-AMPB-Lys-Ala-Arg-Gly-]. Conformational properties of the trans and cis isomers of this peptide in solution were investigated by CD and NMR and were found to differ markedly from the thioredoxin derived azobenzene peptides. In a second peptide, c[Asp-D-Phe-Val-Lys-AMPB-Ala-Arg-Gly-], shifting the position of the chromophore lead to a marked decrease in affinity. With the availability of the x-ray structure of a cyclic RGD-pentapeptide bound to alphavbeta3 integrin (PDB entry 1L5G) modeling of possible bound conformations for trans and cis isomers of both azobenzene peptides was possible. Notably, both peptides in either isomeric form share the same overall conformation in the bound state according to our molecular dynamics simulations.     

Photomodulation of conformational states. II. Mono- and bicyclic peptides with (4-aminomethyl)phenylazobenzoic acid as backbone constituent

It has been reported that backbone cyclization of octapeptides with the photoresponsive (4-aminomethyl)phenylazobenzoic acid imparts sufficient restraints to induce and stabilize ordered conformations of the peptide backbone in both the cis- and trans-azo-isomers (L. Ulysse, J. Cubillos, and J. Chmielewski, Journal of the American Chemical Society, 1995, Vol. 117, pp. 8466-8467). Correspondingly, the active-site octapeptide fragment H-Ala-Cys-Ala-Thr-Cys-Asp-Gly-Phe-OH [134-141] of thioredoxin reductase, with its high preference for a 3(10)-helix turn conformation centered on the Thr-Cys sequence, was backbone cyclized with this azobenzene moiety in the attempt to design a photoresponsive system where the conformational states of the peptide backbone are dictated by the configuration of the azobenzene and can be further modulated by the disulfide bridge. Nuclear magnetic resonance conformational analysis of the monocyclic compound clearly revealed the presence of two conformational families in both the cis- and trans-azo configuration. Of the higher populated conformational families, the structure of the trans-isomer seems like a pretzel-like folding, while the cis-isomer relaxes into a significantly less defined conformational state that does not exhibit any regular structural elements. Further restrictions imparted by disulfide bridging of the peptide moiety leads to an even better defined conformation for the trans-azo-isomer, whereas the cis-isomer can be described as a frustrated system without pronounced energy minima and thus with little conformational preferences. Our findings would suggest that this photoresponsive peptide template may not be of general usefulness for light-induced conformational transitions between two well-defined conformational states at least under the experimental conditions employed, even in the bicyclic form. However, trans --> cis isomerization of the bicyclic peptide is accompanied by a switch from a well-defined conformation to an ensemble of possible conformations.     

Photomodulation of conformational states. III. Water-soluble bis-cysteinyl-peptides with (4-aminomethyl) phenylazobenzoic acid as backbone constituent

In previous studies we have shown that light-induced cis/trans isomerization of the azobenzene moiety in cyclo-[Ala-Cys-Ala-Thr-Cys-Asp-Gly-Phe-AMPB] [AMPB: (4-aminomethyl)phenylazobenzoic acid] leads both in the monocyclic and in the oxidized bicyclic form to markedly differentiated conformational states in DMSO, a fact that lends itself for photomodulation of the redox potential of such bis-cysteinyl-peptides. For this purpose water-soluble systems are required, and this was achieved by replacing three residues outside the Cys-Ala-Thr-Cys active-site motif of thioredoxin reductase with lysines. The resulting cyclo-[Lys-Cys-Ala-Thr-Cys-Asp-Lys-Lys-AMPB] fully retains its photoresponsive properties in water as well assessed by uv and CD measurements. Paralleling results of the previously investigated azobenzene-containing cyclic peptides, the trans --> cis isomerization of the water-soluble monocyclic and oxidized bicyclic peptide is accompanied by a marked transition from a well-defined conformation to an ensemble of possible conformations. However, the conformational preferences are very dissimilar from those of the DMSO-soluble peptides. In fact, hydrogen bonds as well as secondary structure elements were found that change in the mono- and bicyclic peptide upon irradiation. The photo switch between different turn types and hydrogen bonding networks offers the structural rational for the significantly differentiated redox potentials, but also the possibility of monitoring by femtosecond uv-vis and ir spectroscopy fast and ultra fast backbone rearrangement processes following the electronic trans --> cis isomerization.     

Native-like cyclic peptide models of a viral antigenic site: finding a balance between rigidity and flexibility

Antigenic site A of foot-and-mouth disease virus (serotype C) has been reproduced by means of cyclic versions of peptide A15, YTASARGDLAHLTTT, corresponding to residues 136-150 of envelope protein VP1. A structural basis for the design of the cyclic peptides is provided by crystallographic data from complexes between the Fab fragments of anti-site A monoclonal antibodies and A15, in which the bound peptide is folded into a quasi-cyclic pattern. Head-to-tail cyclizations of A15 do not provide peptides of superior antigenicity. Internal disulfide cyclization, however, leads to analogs which are recognized as one to two orders of magnitude better than linear A15 in both ELISA and biosensor experiments. CD and NMR studies show that the best antigen, CTASARGDLAHLTT-Ahx-C (disulfide), is very insensitive to environment-induced conformational change, suggesting that cyclization helps to stabilize a bioactive-like structure.     

The CXC motif: a functional mimic of protein disulfide isomerase

Protein disulfide isomerase (PDI) utilizes the active site sequence Cys-Gly-His-Cys (CGHC; E degrees ' = -180 mV) to effect thiol-disulfide interchange during oxidative protein folding. Here, the Cys-Gly-Cys-NH(2) (CGC) peptide is shown to have a disulfide reduction potential (E degrees ' = -167 mV) that is close to that of PDI. This peptide has a thiol acid dissociation constant (pK(a) = 8.7) that is lower than that of glutathione. These attributes endow the CGC peptide with substantial disulfide isomerization activity. Escherichia coli thioredoxin (Trx) utilizes the active site sequence Cys-Gly-Pro-Cys (CGPC; E degrees ' = -270 mV) to effect disulfide reduction. Removal of the proline residue from the Trx active site yields a CGC active site with a greatly destabilized disulfide bond (E degrees ' >or= -200 mV). The DeltaP34 variant retains high conformational stability and remains a substrate for thioredoxin reductase. In contrast to the reduced form of the wild-type enzyme, the reduced form of DeltaP34 Trx has disulfide isomerization activity, which is 25-fold greater than that of the CGC peptide. Thus, the rational deletion of an active site residue can bestow a new and desirable function upon an enzyme. Moreover, a CXC motif, in both a peptide and a protein, provides functional mimicry of PDI.     

Synthesis and application of an azobenzene amino acid as a light-switchable turn element in polypeptides

The synthesis of an azobenzene amino acid (aa) for use as a photo-inducible conformational switch in polypeptides is described. The compound can be easily incorporated into an aa sequence by solid-phase peptide synthesis using standard 9-fluorenylmethoxycarbonyl methods. A reversible conformational change of the peptide backbone is induced by switching between the cis and trans configurations of the azobenzene moiety by irradiation with light of suitable wavelength. Thermal cis --> trans isomerization of this azobenzene aa is slow, enabling detailed structural investigations of the modified peptides, e.g., using NMR techniques. The total time for the synthesis of the photoswitch is typically 4 d, with an overall yield of 40-50%.     

Synthetic peptide models for the redox-active disulfide loop of glutaredoxin. Conformational studies

Two cyclic peptide disulfides (Sequence: see text). (X = L-Tyr or L-Phe) have been synthesized as models for the 14-membered redox-active disulfide loop of glutaredoxin. 1H NMR studies at 270 MHz in chloroform solutions establish a type I beta-turn conformation for the Pro-X segment in both peptides, stabilized by a 4----1 hydrogen bond between the Cys(1) CO and Cys(4) NH groups. Nuclear Overhauser effects establish that the aromatic ring in the X = Phe peptide is oriented over the central peptide unit. In dimethyl sulfoxide solutions two conformational species are observed in slow exchange on the NMR time scale, for both peptides. These are assigned to type I and type II beta-turn structures with -Pro-Tyr(Phe)- as the corner residues. The structural assignments are based on correlation of NMR parameters with model 14-membered cyclic cystine peptides with Pro-X spacers. Circular dichroism studies based on the -S-S- n-omega* transition suggest a structural change in the disulfide bridge with changing solvent polarity, establishing conformational coupling between the peptide backbone and the disulfide linkage in these systems.     

Multiple loop conformations of peptides predicted by molecular dynamics simulations are compatible with nuclear magnetic resonance

The affinity and selectivity of protein-protein interactions can be fine-tuned by varying the size, flexibility, and amino acid composition of involved surface loops. As a model for such surface loops, we study the conformational landscape of an octapeptide, whose flexibility is chemically steered by a covalent ring closure integrating an azobenzene dye into and by a disulfide bridge additionally constraining the peptide backbone. Because the covalently integrated azobenzene dyes can be switched by light between a bent cis state and an elongated trans state, six cyclic peptide models of strongly different flexibilities are obtained. The conformational states of these peptide models are sampled by NMR and by unconstrained molecular dynamics (MD) simulations. Prototypical conformations and the free-energy landscapes in the high-dimensional space spanned by the phi/psi angles at the peptide backbone are obtained by clustering techniques from the MD trajectories. Multiple open-loop conformations are shown to be predicted by MD particularly in the very flexible cases and are shown to comply with the NMR data despite the fact that such open-loop conformations are missing in the refined NMR structures.     

An analysis of the 225-230-nm CD band of elapid toxins

Spectroscopic measurements of three representative elapid toxins are presented. An analysis of the data leads to the conclusion that the 228-nm CD band in this class of proteins originates largely in the disulfide chromophore. The intensity of this CD band is sensitive to conformational change associated with the disulfide group.     

Molecular characterization of the principal substrate binding site of the ubiquitous folding catalyst protein disulfide isomerase

Disulfide bond formation in the endoplasmic reticulum of eukaryotes is catalyzed by the ubiquitously expressed enzyme protein disulfide isomerase (PDI). The effectiveness of PDI as a catalyst of native disulfide bond formation in folding polypeptides depends on the ability to catalyze disulfide-dithiol exchange, to bind non-native proteins, and to trigger conformational changes in the bound substrate, allowing access to buried cysteine residues. It is known that the b' domain of PDI provides the principal peptide binding site of PDI and that this domain is critical for catalysis of isomerization but not oxidation reactions in protein substrates. Here we use homology modeling to define more precisely the boundaries of the b' domain and show the existence of an intradomain linker between the b' and a' domains. We have expressed the recombinant b' domain thus defined; the stability and conformational properties of the recombinant product confirm the validity of the domain boundaries. We have modeled the tertiary structure of the b' domain and identified the primary substrate binding site within it. Mutations within this site, expressed both in the isolated domain and in full-length PDI, greatly reduce the binding affinity for small peptide substrates, with the greatest effect being I272W, a mutation that appears to have no structural effect.     

Photoinduced conformational dynamics of a photoswitchable peptide: a nonequilibrium molecular dynamics simulation study

Employing nonequilibrium molecular dynamics simulations, a comprehensive computational study of the photoinduced conformational dynamics of a photoswitchable bicyclic azobenzene octapeptide is presented. The calculation of time-dependent probability distributions along various global and local reaction coordinates reveals that the conformational rearrangement of the peptide is rather complex and occurs on at least four timescales: 1) After photoexcitation, the azobenzene unit of the molecule undergoes nonadiabatic photoisomerization within 0.2 ps. 2) On the picosecond timescale, the cooling (13 ps) and the stretching (14 ps) of the photoexcited peptide is observed. 3) Most reaction coordinates exhibit a 50-100 ps component reflecting a fast conformational rearrangement. 4) The 500-1000 ps component observed in the simulation accounts for the slow diffusion-controlled conformational equilibration of the system. The simulation of the photoinduced molecular processes is in remarkable agreement with time-resolved optical and infrared experiments, although the calculated cooling as well as the initial conformational rearrangements of the peptide appear to be somewhat too slow. Based on an ab initio parameterized vibrational Hamiltonian, the time-dependent amide I frequency shift is calculated. Both intramolecular and solvent-induced contributions to the frequency shift were found to change by < or = 2 cm(-1), in reasonable agreement with experiment. The potential of transient infrared spectra to characterize the conformational dynamics of peptides is discussed in some detail.     

Analysis of peptide design in four-, five-, and six-helix bundle template assembled synthetic protein molecules

Four-, five-, and six-helix bundle template assembled synthetic proteins (TASPs) have been synthesized using disulfide bonds between cavitand templates and peptides, and characterized in terms of stability and structural specificity. The peptide sequence (CGGGEELLKKLEE LLKKG) used was originally designed for a four-helix bundle. The TASPs were analyzed using CD spectroscopy, chemical denaturation studies, NMR spectroscopy, sedimentation equilibria studies, and hydrophobic dye binding studies to determine the effect of a single peptide sequence when incorporated into bundles with different numbers of helices. If the design was indeed idealized for a four-helix bundle, then the five- and six-helix bundles should be less stable and manifest lower conformational specificity. The TASPs all demonstrated high stability and cooperative unfolding. However, the four-helix bundle was found to be significantly more stable and nativelike compared to the five- and six-helix bundles. This suggests that the peptide sequence is specific to the four-helix bundle, as designed. This result demonstrates the ability to design de novo proteins with specified structure, not just generic stability.     

Structure and energy landscape of a photoswitchable peptide: a replica exchange molecular dynamics study

A replica exchange molecular dynamics (REMD) simulation of a bicyclic azobenzene peptide in explicit dimethyl sulfoxide solution is presented in order to characterize the conformational structures and energy landscape of a photoswitchable peptide. It is shown that an enhanced-sampling technique such as the REMD method is essential to obtain a converged conformational sampling of the peptide at room temperature. This is because conventional MD simulations of less than approximately 100-ns length are either trapped in local minima (at 295 K) or-if run at high temperature-do not resemble the room-temperature REMD results. Calculating various nuclear Overhauser effects (NOEs) and (3)J-couplings, a good overall agreement between the REMD simulations and the NMR experiments of Renner et al. (Biopolymers 2000;54:501-514) is found. In particular, the REMD study confirms the general picture drawn by Renner et al. that the trans-isomer of the azobenzene peptide exhibits a well-defined structure, while the cis-isomer is a conformational heterogeneous system; that is, the trans-isomer occurs in 2 well-defined conformers, while the cis-isomer represents an energetically frustrated system that leads to an ensemble of conformational structures. Employing a principal component analysis of the REMD data, the free energy landscape of the systems is studied at various temperatures. The implications for the folding and unfolding pathways of the system are discussed.     

Photomodulation of conformational states. I. Mono- and bicyclic peptides with (4-amino)phenylazobenzoic acid as backbone constituent

The thioredoxin reductase active-site fragment H-Ala-Cys-Ala-Thr-Cys-Asp-Gly-Phe-OH [134-141], which is known for its high tendency to assume an almost identical conformation as in the intact enzyme, was backbone cyclized with the photoresponsive (4-amino)phenylazobenzoic acid (APB) to produce a monocyclic and disulfide-bridged bicyclic APB-peptide. Light-induced reversible cis/trans isomerization occurs at identical extents in both the linear and the two cyclic forms. Nuclear magnetic resonance conformational analysis clearly revealed that in the bicyclic APB-peptide both as a trans- and cis-azo-isomer the constraints imparted by the bicyclic structure do not allow the molecule to relax into a defined low energy conformation, thus making the molecule a frustrated system that flip-flops between multiple conformational states. Conversely, the monocyclic APB peptide folds into a well-defined lowest energy structure as a trans-azo-isomer, which upon photoisomerization to the cis-azo configuration relaxes into a less restricted conformational space. First femtosecond spectroscopic analysis of the dynamics of the photoreaction confirm a fast first phase on the femtosecond time scale related to the cis/trans isomerization of the azobenzene moiety followed by a slower phase in the picosecond time scale that involves an adjustment of the peptide backbone. Due to the well- defined photoresponsive two-state transition of this monocyclic peptide molecule, it represents a model system well suited for studying the ultrafast dynamics of conformational transitions by time-resolved spectroscopy.     

Preferred conformations of cyclic Ac-Cys-Pro-Xaa-Cys-NHMe peptides: a model for chain reversal and active site of disulfide oxidoreductase

The conformational study on cyclic Ac-Cys-Pro-Xaa-Cys-NHMe (Ac-CPXC-NHMe; X=Ala, Val, Leu, Aib, Gly, His, Phe, Tyr, Asn and Ser) peptides has been carried out using the Empirical Conformational Energy Program for Peptides, version 3 (ECEPP/3) force field and the hydration shell model in the unhydrated and hydrated states. This work has been undertaken to investigate structural implications of the CPXC sequence as the chain reversal for the initiation of protein folding and as the motif for active site of disulfide oxidoreductases. The backbone conformation DAAA is commonly the most feasible for cyclic CPXC peptides in the hydrated state, which has a type I beta-turn at the Pro-Xaa sequence. The proline residue and the hydrogen bond between backbones of two cystines as well as the formation of disulfide bond appear to play a role in stabilizing this preferred conformation of cyclic CPXC peptides. However, the distributions of backbone conformations and beta-turns may indicate that the cyclic CPXC peptide seems to exist as an ensemble of beta-turns and coiled conformations in aqueous solution. The intrinsic stability of the cyclic CPXC motif itself for the active conformation seems to play a role in determining electrochemical properties of disulfide oxidoreductases.     

Catalysis of protein disulfide bond isomerization in a homogeneous substrate

Protein disulfide isomerase (PDI) catalyzes the rearrangement of nonnative disulfide bonds in the endoplasmic reticulum of eukaryotic cells, a process that often limits the rate at which polypeptide chains fold into a native protein conformation. The mechanism of the reaction catalyzed by PDI is unclear. In assays involving protein substrates, the reaction appears to involve the complete reduction of some or all of its nonnative disulfide bonds followed by oxidation of the resulting dithiols. The substrates in these assays are, however, heterogeneous, which complicates mechanistic analyses. Here, we report the first analysis of disulfide bond isomerization in a homogeneous substrate. Our substrate is based on tachyplesin I, a 17-mer peptide that folds into a beta hairpin stabilized by two disulfide bonds. We describe the chemical synthesis of a variant of tachyplesin I in which its two disulfide bonds are in a nonnative state and side chains near its N and C terminus contain a fluorescence donor (tryptophan) and acceptor (N(epsilon)-dansyllysine). Fluorescence resonance energy transfer from 280 to 465 nm increases by 28-fold upon isomerization of the disulfide bonds into their native state (which has a lower E(o') = -0.313 V than does PDI). We use this continuous assay to analyze catalysis by wild-type human PDI and a variant in which the C-terminal cysteine residue within each Cys-Gly-His-Cys active site is replaced with alanine. We find that wild-type PDI catalyzes the isomerization of the substrate with kcat/K(M) = 1.7 x 10(5) M(-1) s(-1), which is the largest value yet reported for catalysis of disulfide bond isomerization. The variant, which is a poor catalyst of disulfide bond reduction and dithiol oxidation, retains virtually all of the activity of wild-type PDI in catalysis of disulfide bond isomerization. Thus, the C-terminal cysteine residues play an insignificant role in the isomerization of the disulfide bonds in nonnative tachyplesin I. We conclude that catalysis of disulfide bond isomerization by PDI does not necessarily involve a cycle of substrate reduction/oxidation.     

Inhibition of the Functional Interplay between Endoplasmic Reticulum (ER) Oxidoreduclin-1α (Ero1α) and Protein-disulfide Isomerase (PDI) by the Endocrine Disruptor Bisphenol A

Bisphenol A (BPA) is an endocrine disruptor that may have adverse effects on human health. We recently isolated protein-disulfide isomerase (PDI) as a BPA-binding protein from rat brain homogenates and found that BPA markedly inhibited PDI activity. To elucidate mechanisms of this inhibition, detailed structural, biophysical, and functional analyses of PDI were performed in the presence of BPA. BPA binding to PDI induced significant rearrangement of the N-terminal thioredoxin domain of PDI, resulting in more compact overall structure. This conformational change led to closure of the substrate-binding pocket in b′ domain, preventing PDI from binding to unfolded proteins. The b′ domain also plays an essential role in the interplay between PDI and ER oxidoreduclin 1α (Ero1α), a flavoenzyme responsible for reoxidation of PDI. We show that BPA inhibited Ero1α-catalyzed PDI oxidation presumably by inhibiting the interaction between the b′ domain of PDI and Ero1α; the phenol groups of BPA probably compete with a highly conserved tryptophan residue, located in the protruding β-hairpin of Ero1α, for binding to PDI. Consistently, BPA slowed down the reoxidation of PDI and caused the reduction of PDI in HeLa cells, indicating that BPA has a great impact on the redox homeostasis of PDI within cells. However, BPA had no effect on the interaction between PDI and peroxiredoxin-4 (Prx4), another PDI family oxidase, suggesting that the interaction between Prx4 and PDI is different from that of Ero1α and PDI. These results indicate that BPA, a widely distributed and potentially harmful chemical, inhibits Ero1-PDI-mediated disulfide bond formation.     

Protein disulfide isomerase is cleaved by caspase-3 and -7 during apoptosis

Apoptotic signals are typically accompanied by activation of aspartate-specific cysteine proteases called caspases, and caspase-3 and -7 play crucial roles in the execution of apoptosis. Previously, using the proteomic approach, protein disulfide isomerase (PDI) was found to be a candidate substrate of caspase-7. This abundant 55 kDa protein introduces disulfide bonds into proteins (via its oxidase activity) and catalyzes the rearrangement of incorrect disulfide bonds (via its isomerase activity). PDI is abundant in the ER but is also found in non-ER locations. In this study we demonstrated that PDI is cleaved by caspase-3 and -7 in vitro. In addition, in vivo experiment showed that it is cleaved during etoposide-induced apoptosis in HL-60 cells. Subcellular fractionation showed that PDI was also present in the cytosol. Furthermore, only cytosolic PDI was clearly digested by caspase-3 and -7. It was also confirmed by confocal image analysis that PDI and caspase-7 partially co-localize in both resting and apoptotic MCF-7 cells. Overexpression of cytosolic PDI (ER retention sequence deleted) inhibited cell death after an apoptotic stimulus. These data indicate that cytosolic PDI is a substrate of caspase-3 and -7, and that it has an anti-apoptotic action.     

Domain architecture of protein-disulfide isomerase facilitates its dual role as an oxidase and an isomerase in Ero1p-mediated disulfide formation

Native disulfide bond formation in eukaryotes is dependent on protein-disulfide isomerase (PDI) and its homologs, which contain varying combinations of catalytically active and inactive thioredoxin domains. However, the specific contribution of PDI to the formation of new disulfides versus reduction/rearrangement of non-native disulfides is poorly understood. We analyzed the role of individual PDI domains in disulfide bond formation in a reaction driven by their natural oxidant, Ero1p. We found that Ero1p oxidizes the isolated PDI catalytic thioredoxin domains, A and A' at the same rate. In contrast, we found that in the context of full-length PDI, there is an asymmetry in the rate of oxidation of the two active sites. This asymmetry is the result of a dual effect: an enhanced rate of oxidation of the second catalytic (A') domain and the substrate-mediated inhibition of oxidation of the first catalytic (A) domain. The specific order of thioredoxin domains in PDI is important in establishing the asymmetry in the rate of oxidation of the two active sites thus allowing A and A', two thioredoxin domains that are similar in sequence and structure, to serve opposing functional roles as a disulfide isomerase and disulfide oxidase, respectively. These findings reveal how native disulfide folding is accomplished in the endoplasmic reticulum and provide a context for understanding the proliferation of PDI homologs with combinatorial arrangements of thioredoxin domains.