HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 278, Issue 49, 49478-49486, December 5, 2003

This Article Abstract Full Text (PDF) All Versions of this Article: 278/49/49478    most recent M309015200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Choi, J. Articles by Shin, W. Search for Related Content PubMed PubMed Citation Articles by Choi, J. Articles by Shin, W. Social Bookmarking                      What's this?

Crystal Structure of Escherichia coli Thiol Peroxidase in the Oxidized State

INSIGHTS INTO INTRAMOLECULAR DISULFIDE FORMATION AND SUBSTRATE BINDING IN ATYPICAL 2-CYS PEROXIREDOXINS^*

Jongkeun Choi, Soonwoong Choi, Jungwon Choi¶, Mee-Kyung Cha||, Il-Han Kim||, and Whanchul Shin**

From the School of Chemistry and Center for Molecular Catalysis, Seoul National University, Seoul 151-742, Korea, the ^¶Department of Chemistry, The University of Suwon, Suwon 445-743, Korea, and the ^||National Creative Research Initiatives Center for Antioxidant Proteins, Department of Biochemistry, Paichai University, Taejon 302-735, Korea

Received for publication, August 14, 2003 , and in revised form, September 18, 2003.

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS AND DISCUSSION REFERENCES

 
Thioredoxin-dependent thiol peroxidase (Tpx) from Escherichia coli represents a group of antioxidant enzymes that are widely distributed in pathogenic bacterial species and which belong to the peroxiredoxin (Prx) family. Bacterial Tpxs are unique in that the location of the resolving cysteine (C_R) is different from those of other Prxs. E. coli Tpx (EcTpx) shows substrate specificity toward alkyl hydroperoxides over H₂O₂ and is the most potent reductant of alkyl hydroperoxides surpassing AhpC and BCP, the other E. coli Prx members. Here, we present the crystal structure of EcTpx in the oxidized state determined at 2.2-Å resolution. The structure revealed that Tpxs are the second type of atypical 2-Cys Prxs with an intramolecular disulfide bond formed between the peroxidatic (C_P, Cys⁶¹) and resolving (Cys⁹⁵) cysteine residues. The extraordinarily long N-terminal chain of EcTpx folds into a -hairpin making the overall structure very compact. Modeling suggests that, in atypical 2-Cys Prxs, the C_R-loop as well as the C_P-loop may alternately assume the fully folded or locally unfolded conformation depending on redox states, as does the C_P-loop in typical 2-Cys Prxs. EcTpx exists as a dimer stabilized by hydrogen bonds. Its substrate binding site extends to the dimer interface. A modeled structure of the reduced EcTpx in complex with 15-hydroperoxyeicosatetraenoic acid suggests that the size and shape of the binding site are particularly suited for long fatty acid hydroperoxides consistent with its greater reactivity.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS AND DISCUSSION REFERENCES

 
Peroxiredoxins (Prxs)¹ are a large, diverse, and ubiquitous family of antioxidant enzymes. Prxs reduce deleterious H₂O₂ or alkyl hydroperoxides utilizing the thiol group of a cysteine residue (1, 2), and in some cases are involved in the decomposition of highly toxic peroxynitrite (3). Some Prxs also control cytokine-induced peroxide levels that mediate redox-sensitive signal transduction (2, 4-6). All Prxs, belonging to the thioredoxin-fold superfamily, contain a conserved peroxidatic Cys (C_P) in the N-terminal portion and share the same peroxidatic active site structure (7). The C_P residue is oxidized by peroxides to a cysteine sulfenic acid (C_P-SOH) intermediate. Prxs are classified into either the 2-Cys or 1-Cys types, based on whether they contain the resolving Cys (C_R) residue or not. In 2-Cys Prxs, the C_P-SOH and C_R-SH react and form a disulfide (C_P-S-S-C_R). The stable disulfide form is then reduced by one of several cell-specific disulfide oxidoreductases (e.g. thioredoxin (Trx), tryparedoxin, AhpD, or AhpF), completing the catalytic cycle (2, 7). Unlike the 2-Cys Prxs, the catalytic mechanism of 1-Cys Prxs still remains to be elucidated (8, 9). The 2-Cys Prxs have been further subdivided into either typical or atypical types depending on the location of the C_R residue. In typical 2-Cys Prxs, the C_P-SOH reacts with the C_R residue located in the C-terminal arm from the other subunit of the antiparallel (₂) homodimer, resulting in two symmetrical active sites per dimer (10-13). In contrast, the C_R residue in atypical 2-Cys Prxs resides within the same subunit (14, 15).

Thus far, seven crystal structures have been reported for the Prx members in various redox states. The structures of rat PrxI (HBP23) (10) and Salmonella typhimurium AhpC (StAhpC) (13) were determined in the oxidized state with a disulfide bond. The remaining structures were determined in states that do not contain a disulfide. They include human PrxV (15), Crithida fasciculate TryP (12) and StAhpC (16), all in the reduced (C_P-SH or Ser-OH in StAhpC) state, human PrxII (TPxB) (11) in the overoxidized sulfinic acid (C_P-SO₂H) state, and human PrxVI (hORF6) (8) in the C_P-SOH state. PrxV and PrxVI are atypical 2-Cys and 1-Cys Prxs, respectively, while the others (AhpC, PrxI, PrxII and TryP) are typical 2-Cys Prxs. Among these, only the structure of StAhpC has been determined in both the reduced and oxidized states (13, 16). Structural studies on Prxs, especially those of the typical 2-Cys type, have yielded valuable insights into their modes of action in detoxification or signaling.

Crystal structures revealed that the five-residue segment containing the C_P residue, referred to as the C_P-loop, exists in two stable conformations depending on the redox states of Prxs (7). The C_P-loop adopts a fully folded (FF) conformation forming the first turn of a long -helix in all non-disulfide forms, whereas it is locally unfolded (LU) from the -helix in the disulfide forms. These structural data suggest that the following sequence of events occurs during a catalytic cycle of typical 2-Cys Prxs. First, the reduced Prx in an FF form reacts with the peroxide to become a C_P-SOH intermediate still in FF. Then a local unfolding of the C_P-loop and a concomitant conformational change of the long C-terminal arm of the other subunit occur for the C_P-SOH and C_R-SH to react and form a disulfide bond. Finally, the exposed disulfide in an LU form reacts with the reductant to regenerate the reduced Prx in an FF form. In the typical 2-Cys Prxs, the homodimers assemble into toroid-shaped ((₂)₅) decamers when reduced or overoxidized, whereas the decamers dissociate into dimers when oxidized (11, 13). Wood et al. (13) found that the C_P-loop disrupts the interfacial interactions between adjacent dimers when it adopts the LU form, and suggested that the C_P-loop acts as a molecular switch responsible for redox-dependent oligomerization. From structural data, they could also clarify the origin of different sensitivities to oxidative inactivation between the prokaryotic (much less sensitive) and eukaryotic 2-Cys Prxs. They identified two sequence motifs unique to the eukaryotic Prxs and suggested that these motifs act as a molecular switch that determines whether H₂O₂ acts as a deleterious oxidant or beneficial signal in eukaryotes (16). Contrary to the case of typical 2-Cys Prxs, structural data for the atypical 2-Cys Prxs are scarce and the structure of the reduced PrxV is the only one available at present.

Escherichia coli has three Prx family members of peroxidases that are highly diverged from one another. They include the periplasmic thiol peroxidase (Tpx, p20, scavengase) (17, 18) and a weakly active bacterioferritin-comigratory protein (BCP) (19), and the alkyl hydroperoxide reductase peroxidase component (AhpC) (20). Both Tpx and BCP specifically require Trx as a reductant, while AhpC requires AhpF. E. coli proteomics studies showed that, during the growth phase in minimal media and without induction by H₂O₂, Tpx (1.6 µM) is 3.5-fold less abundant than AhpC, and 2.5-fold more abundant than BCP (21). Homologues of E. coli Tpx (EcTpx) are widely distributed in both Gram-negative and Gram-positive eubacterial species, most of which are pathogenic (22). EcTpx has three Cys residues (Cys⁶¹, Cys⁸², and Cys⁹⁵), among which Cys⁶¹ was identified as the C_P residue earlier (23). Other Cys residues do not align with the C_R residues of other 2-Cys Prxs. Very recently, Baker and Poole (24) firmly established Cys⁹⁵ as the C_R residue. They confirmed the formation of an intramolecular disulfide bond, but did not explicitly describe Tpx as an atypical 2-Cys Prx probably because the location of the C_R residue is quite different from that of archetypal PrxV. They found that EcTpx is homodimeric in solution independent of redox states although the dimerization is not required for catalytic reaction. They also found that EcTpx shows substrate specificity toward alkyl hydroperoxides over H₂O₂ and, on the basis of catalytic efficiency, is the most potent reductant of alkyl hydroperoxides surpassing AhpC and BCP. In this study, we have determined the crystal structure of EcTpx in its oxidized state, demonstrating that Tpx is a second type of atypical 2-Cys Prx. It is the first structural example that shows the intramolecular disulfide bond formed in a Prx. The structure also helps us to explain why EcTpx exists as a stable dimer and shows the substrate preference for alkyl hydroperoxides with greater catalytic efficiency.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS AND DISCUSSION REFERENCES

 
Crystallization and Data Collection—EcTpx was overexpressed, purified and crystallized as reported previously (25). Crystals were obtained at room temperature using the hanging-drop vapor-diffusion method. For the crystallization, 2 µl of reservoir solution (10% PEG4000, 20% 2-propyl alcohol, 50 mM Na-HEPES pH 7) and 2 µl of protein solution (10 mg/ml) were mixed and equilibrated against the reservoir. Rod-shaped small crystals appeared within 3 days. Seeding techniques with macroseeds were essential to obtain large single crystals. With this procedure, crystals grew routinely to their full size of 0.10 x 0.15 x 0.85 mm. The crystals belong to the orthorhombic space group P2₁2₁2₁, with cell constants of a = 38.97, b = 58.97, and c = 127.59 Å. There are two molecules of EcTpx in the asymmetric unit. For phasing, three heavy atom derivatives were used. They were prepared by soaking the crystals for 7, 9, and 1 day, respectively, into solutions of mercury acetate, mercury chloride, and ethyl mercury chloride in which each heavy atom compound was dissolved in mother liquor at a final concentration of 3 mM. Both native and derivative crystals were flash-frozen at 100 K in a nitrogen stream before data collection, using 50% PEG4000 as a cryo-protectant. Data for native crystals were collected at = 1.12714 Å using a MacScience 2030 image plate detector at the beamline 6B of the Pohang Light Source, Korea. Data for derivative crystals were measured on a Rigaku rotating anode (CuK = 1.5418 Å)-Raxis IV image plate system. Each data set was processed and scaled using the programs DENZO and SCALEPACK (26).

Structure Determination and Refinement—The structure of EcTpx was determined by the multiple isomorphous replacement with anomalous scattering (MIRAS) technique. Location of the heavy atom sites, refinement of heavy atom parameters and calculation of the initial phases were performed with the program SOLVE (27), including the anomalous scattering data in the resolution range of 30.0-3.0 Å. The initial MIRAS phases had a mean figure-of-merit of 0.60. Phases were improved with solvent flattening by using the program RESOLVE (28). Model building was conducted using the program O (29), based on a poly(A) model generated by the program RESOLVE. The model was refined with the program CNS (30) progressively increasing the resolution to 2.2-Å. Five percent of the data were randomly selected to be set aside for the R_free calculation. Water molecules were picked up from the difference map on the basis of the peak heights and distance criteria, and were discarded when thermal parameters after refinement were above 60 Ų. A summary of data collection, phasing and refinement statistics is shown in Table I. The stereochemical analysis using the program PROCHECK (31) showed that one residue is in generously allowed regions, and there is no residue in disallowed regions. Recent mass spectrometric analysis of EcTpx indicated the absence of the initiating methionine predicted by the tpx open reading frame (24). In accordance with this data, there is clearly no electron density for this residue. The three C-terminal residues of Leu¹⁶⁶, Lys¹⁶⁷, and Ala¹⁶⁸ are disordered in the molecules A and B. The final model contains residues 2-165 in each EcTpx molecule, and 353 water molecules. The R-value of the final model is 18.8% (R_free = 24.6%) at 2.2-Å resolution. Structural figures were generated using the programs MOLSCRIPT (32) and RASTER3D (33).

	RESULTS AND DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS AND DISCUSSION REFERENCES

 
Overall Structure of Tpx—The crystal structure of EcTpx determined to 2.2-Å resolution (Fig. 1a) revealed that the enzyme is in the oxidized disulfide form, a stable intermediate in the catalytic cycle. EcTpx has a compact globular structure with a typical fold of the Prx family that is constructed around a seven-stranded twisted -sheet surrounded by four -helices (12). When the C_P residues are aligned, the N-terminal chain of EcTpx is longer by about 15 residues than those of other Prxs. This extra chain folds into a -hairpin composed of two antiparallel -strands that are designated N1 and N2, respectively. Other strands are designated following those in other Prxs. The N-terminal -hairpin is connected to the 1 by a 14 residue-long chain that is kinked at Gly²². The 2-strand is followed by a short segment of 3₁₀-helix, then two units (3-1-4-2-5). A -hairpin (6 and 7), connected to 5 by the 3 and a long loop, completes the -sheet and leads to the 4, then finally to a disordered C-terminal chain. The Prx family belongs to the Trx superfamily. A Trx fold in the EcTpx structure is defined by a five-stranded -sheet (5-4-3-6-7) that matches with 1-3-2-4-5 of Trx and by four flanking -helices. The peroxidatic Cys⁶¹ of EcTpx is located on the C_p-loop before 1 and the resolving Cys⁹⁵ is on a chain, designated the C_R-loop, which connects the 2 and 5. The remaining Cys⁸² is located on the 4 and is quite distant (10 Å) from the active site.

View larger version (56K): [in this window] [in a new window]   FIG. 1. Structure of E. coli thiol peroxidase. a, ribbon diagram showing secondary structure elements with labels. -Helices, 310-helix, and -strands are shown in light brown, dark yellow, and light blue, respectively. Redox-active disulfide is depicted as ball-and-stick models. The peroxidatic (CP) and resolving (CR) cysteine sulfur atoms are colored in yellow and orange, respectively, throughout the figures. b, stereoview of the superposed EcTpx monomers. The C-C links of molecule A are colored according to the mean B-factors of the residues (more reddish for higher values and more bluish for lower values), and those of molecule B are colored in light gray. Residues conserved in bacterial Tpxs are shown with balls in red (strictly conserved residues with labels), light green (conserved in more than 90% of 47 known Tpxs) and light blue (homologously conserved). c, stereoview of the 2Fo-Fc electron density map for the redox-active disulfide of molecule A. The density is contoured at 1.0 level. This figure was made using the program PyMOL (www.pymol.org/).

At present 47 bacterial Tpxs can be identified in the most recent sequence databases by BLAST searches. These Tpxs are highly homologous (29-88% identity and 49-94% similarity) to each other with similar molecular sizes. The amino acid sequences align very well with little gaps or insertions (data not shown). All Tpxs contain seven strictly conserved residues. These include Gly²², Pro⁵⁴, Thr⁵⁸, Ser⁸⁴, and Arg¹³³ as well as the peroxidatic Cys⁶¹ and resolving Cys⁹⁵ (in EcTpx numbering). Additionally, 20 residues are conserved in more than 90% of 47 Tpxs and 21 residues are homologously conserved. In total, 48 residues, amounting to about 30% of the entire sequence, are highly conserved. These conserved residues are denoted by color in the EcTpx sequence (Fig. 2a), and their positions are depicted on the EcTpx structure (Fig. 1c). About 80% of them are situated on the central -sheet and on one side of this -sheet containing the active site. Since the conserved residues are located at either structurally or functionally important positions, it seems that all bacterial Tpxs assume an essentially identical fold and share the functional characteristics of EcTpx. Structural differences, if present, may be limited to some loop regions and the C terminus.

View larger version (64K): [in this window] [in a new window]   FIG. 2. Comparison of sequences and structures of Prxs. a, structure-based sequence alignment of selected Prxs. The catalytic cysteine residues in each Prx are masked in color (yellow for CP and orange for CR). Residues strictly conserved in all Prxs are boxed. In the sequence of EcTpx, conserved residues in bacterial Tpxs are denoted with the same color as in Fig. 1b. The secondary structural elements of oxidized EcTpx and reduced StAhpC are shown on top and at the bottom, respectively, of the aligned sequences. The figure was prepared using the program ALSCRIPT (40). b, stereoview of the superposed Prx structures. C-traces of EcTpx, PrxV, oxidized AhpC and reduced AhpC are shown in green, red, blue, and purple, respectively. Only structurally variant regions are colored. Strictly conserved residues are depicted as ball-and-stick models, with oxygen in red and nitrogen in blue. Oxygen atom of the C46S mutant of reduced AhpC, which corresponds to the CP sulfur atom, is depicted as a large half sphere in red. Labels of the -helices are those for EcTpx.

The superposed crystal structures of oxidized EcTpx, reduced PrxV and StAhpC in the reduced (C46S mutant) and oxidized states are shown in Fig. 2b. In these structures, the central -sheets align very well and the conserved Arg as well as the Pro residue occupy essentially the same position. However, several regions show considerable conformational variations that depend on the types of the Prxs as well as their redox states. In the reduced PrxV and StAhpC, the C_P-loops commonly assume the FF conformation and the C_P-loop-1-helix regions align well. In the oxidized StAhpC, the C_P-loop assumes the LU conformation exposing the intersubunit disulfide to the surface, but its 1-helix still coincides well with those of the reduced forms. In the oxidized EcTpx, the C_P-loop assumes an LU conformation that is significantly different from that in the oxidized StAhpC. Its 1-helix is shorter with three less residues than and thus is considerably shifted from those in PrxV and StAhpC. The C(C_P) atoms in the FF and LU forms of StAhpC are separated by 6.0 Å, while those in the oxidized EcTpx and StAhpC are separated by 5.4 Å.

In addition to the C_P-loop-helix region, the three Prxs show considerable conformational variations at three different regions. One is the chain between 2 and 5 (in EcTpx numbering), which is designated L because it flanks the central C_P-loop on the left from the viewpoint shown in Fig. 2b. The L chain contains the C_R residue of EcTpx. Another is the chain between 3 and 6, which is designated R likewise and contains the C_R residue of PrxV. The other is the chain between 7 and 4, which is designated S because it is involved in the formation of the substrate binding site in all three Prxs. The C_R residue of StAhpC resides on the long C-terminal arm that undergoes a large conformational change during catalysis, as shown in Fig. 2b.

Disulfide Formation in Atypical 2-Cys Prxs—Intramolecular (or intrasubunit) disulfide has not been observed in Prxs before the present structure. The crystal structure of human PrxV determined in the reduced state has been the only available one for the atypical 2-Cys Prx. PrxV has three Cys residues (Cys⁴⁸, Cys⁷³, and Cys¹⁵²) and mutational studies identified Cys⁴⁸ and Cys¹⁵² as the C_P and C_R residues, respectively (14). Its structure revealed that the C_P and C_R residues are very far apart (13.8 Å) and it has been suggested that large conformational changes are required for the formation of a disulfide bond (15). Sometimes PrxV is referred to as the 1-Cys Prx, probably taking into consideration the fact that Cys¹⁵² is not always conserved in its homologues (37).

To examine how the intramolecular disulfide bonds can be formed in Prxs, we have modeled the structures of the reduced EcTpx and oxidized PrxV, and compared them with their crystal structures. In the modeled structure of reduced EcTpx, the active site has the same geometry as those of the reduced PrxV and StAhpC, and the C_P-loop assumes the FF conformation. The model suggests that the L chain containing the C_R loop, which is bowed forming a disulfide bond in the oxidized state, changes its conformation in the reduced state so that the C_R-loop also assumes the FF conformation making a helical turn. Consequently, the C_R-loop moves away from the 1 while extending into the 2-helix, and the C atoms of the C_P and C_R become separated by 12.9 Å in the reduced state. This separation is comparable to the separation of 13.8 Å observed in the crystal structure of reduced PrxV. The movement of the L chain in EcTpx is local and rather small, compared with the large conformational change that the C-terminal chain containing the C_R residue undergoes in StAhpC to regulate the oligomeric state as well as forming a disulfide bond.

It can be readily expected that similar local conformational changes may occur while the reduced PrxV becomes a disulfide form. Indeed, modeling of the oxidized PrxV structure indicates that the disulfide bond can be formed comfortably if its C_P- and C_R-loops assume the LU and FF conformations, respectively. The difference between EcTpx and PrxV is that, in the oxidized disulfide state, the C_R-loop in the R chain assumes the FF conformation in PrxV, whereas the C_R-loop in the L chain assumes the LU conformation in EcTpx. Taken together, it seems that the atypical 2-Cys Prxs commonly employ local conformational changes (either helix-to-loop or loop-to-helix) of both C_P- and C_R-loops to change their redox states during a catalytic cycle. In all reduced Prxs, the C_P-loops assume the same FF conformation. In the oxidized Prxs, however, the C_P-loops have different LU conformations depending on the types of Prx due to differences in spatial locations of the C_R residues. To create a disulfide bond from the reduced state, the C_P-loop should unwind, rotating the C(C_P) atom 120° about the 1 helical axis in EcTpx, 90° in StAhpC, and 60° in PrxV.

Classification of 2-Cys Prxs—Thus far, the Prxs have been classified as 1-Cys or 2-Cys types depending on the number of catalytic Cys residues, and the 2-Cys Prxs are further classified as typical or atypical types depending on whether the disulfide bond is formed between the two molecules or within a molecule (2, 5). The typical 2-Cys Prxs are the largest subfamily, comprising bacterial AhpC, mammalian subfamilies PrxI-IV, some plant Prxs, Tpx of yeast, and the tryparedoxin peroxidase of kinetoplastida. PrxV of atypical 2-Cys type is seen in mammals, bacteria, lower fungi and higher plants. Sometimes the Prx family is classified into five major molecular clades, treating bacterial Tpx and BCP as distinct subfamily members (2).

BCP homologues have been found in archaea, bacteria, plants and animals, and are least characterized among the Prx members (2). Like EcTpx, E. coli BCP contains three Cys residues (Cys⁴⁵, Cys⁵⁰, and Cys⁹⁹) and shows peroxidase activity with the substrate preference of linoleic acid hydroperoxides over H₂O₂ (19). The C45S mutant completely loses peroxidase activity, while the C50S mutant retains partial activity, indicating that Cys⁴⁵ is the C_P residue. Cys⁴⁵ is conserved in all bacterial BCPs, while Cys⁵⁰ is conserved only in some of them. Whether Cys⁵⁰ of E. coli BCP is catalytic or not remains to be determined. Another BCP with two Cys residues was identified from a flowering plant Sedum lineare (38). Mutational studies explicitly identified Cys⁴⁴ as the C_P and Cys⁴⁹ as the C_R residues. S. lineare BCP was specifically designated PrxQ since the name BCP provides no functional insights. BLAST searches indicate that there may be either 1-Cys or 2-Cys types of BCP homologues and most of the 2-Cys BCPs may contain the C_PTXXAC_R motif. Modeling of the 2-Cys BCP structure indicates that both the C_P and C_R residues may be located on the 1-helix in the reduced state and, upon oxidation, the C_P-loop would unwind to the LU conformation while the C_P residue forms a disulfide bond with the C_R residue still on the 1-helix.

Both Tpx and 2-Cys BCP are apparently distinct from PrxV, but they are also of atypical type in the context of the above scheme classifying the 2-Cys Prxs. These three Prxs differ from each other in the locations of their resolving C_R residues. Therefore, it may be more appropriate to subdivide the atypical 2-Cys Prxs instead of classifying Tpx and BCP as independent molecular clades of the Prx family. From the structural viewpoint, it seems that the spatial locality of the C_R residue in 2-Cys Prxs is limited to one of the four different locations around the C_P, including the 1-helix, the L and R chains flanking the 1, and the C-terminal chain that comes from the other subunit and is situated on top of the C_P-loop, as schematically shown in Fig. 3. Then the disulfide bridge formed may be labeled an intersubunit for AhpC, an L-intrasubunit for Tpx, an R-intrasubunit for PrxV, and an intrachain for BCP. Considering the three-dimensional structures of Prxs, there would be no other distinct location available for the C_R residue, and all 2-Cys Prxs should uniquely belong to one of these four classes. Thus we propose to classify the atypical 2-Cys Prxs into three subfamilies, designating Tpx the L-atypical type; PrxV, the R-atypical type; and 2-Cys BCP the C-atypical type (C stands for chain).

View larger version (30K): [in this window] [in a new window]   FIG. 3. A schematic diagram showing that 2-Cys Prxs can be classified into four molecular clades. Arrows indicate the directions of the disulfide formation. Shaded cylinders denote the -helices and white cylinders denote the region that contains the CR residue and undergoes a loop-to-helix or a helix-to-loop transition depending on redox states. HOOR denotes the substrate.

In the crystal, the two EcTpx molecules in the asymmetric unit are related by 2-fold non-crystallographic symmetry (NCS), forming a dimer stabilized by hydrogen bonding and hydrophobic interactions (Fig. 4a). The two active sites are on one side of the dimer and are separated by about 15 Å. Five separate loop regions from each monomer form the monomermonomer interface. The Arg¹¹⁰ residues from both monomers play an important role in stabilizing the dimer. The N atom of Arg¹¹⁰ of one monomer forms hydrogen bonds with two main-chain O atoms (Pro¹²⁵ and Lys¹²⁸) of the other monomer, which results in clamping either flank of the dimer. The N1 atom of Arg¹¹⁰ forms hydrogen bonds with carboxyl O of Asp³⁷ and main-chain O of Leu³⁵ in the same monomer. Arg⁹³ also contributes to the stabilization of the dimer. Pairs of Arg⁹³ and Asp⁵⁷ are clustered around the 2-fold axis, with an unsymmetrical pattern in their interactions. The N and N1 atoms of Arg^93A are hydrogen bonded to the carboxyl O atoms of Asp^57A, and its N and N1 atoms to the carboxyl O atoms of Asp^57B. Arg^93B protrudes to the solvent region, the N atom forming a hydrogen bond with the main-chain O of Phe^89B. Besides these hydrogen bonding interactions, two hydrophobic patches related by 2-fold NCS enhance the stability of the dimer. Each hydrophobic cluster involves the residues of Leu³⁵, Leu⁸⁷, Phe⁸⁹ and Phe¹⁰⁹ from one monomer and Leu¹²⁷ and Leu¹³⁰ from the other. There are several water molecules between the two hydrophobic clusters. The modeled structure of reduced EcTpx indicates that the conformations of the C_P- and C_R-loops of each monomer are significantly changed from those in the crystal structure, but the overall structure of the dimer remains essentially unchanged consistent with results from analytical ultracentrifugation studies. Arg⁹³ and Arg¹¹⁰ as well as Asp⁵⁷ are conserved in more than 75% of 46 known bacterial Tpxs and, in the remaining Tpxs, are replaced by residues such as Asn, Tyr, Gln, or Lys, all with hydrogen bonding potential. Thus it is expected that most, if not all, bacterial Tpxs may exist as dimers.

View larger version (65K): [in this window] [in a new window]   FIG. 4. Quaternary structures and putative substrate binding sites. a, ribbon diagram of the EcTpx dimer, with monomers in different colors and looking down the dimer axis. Residues involved in the monomer-monomer interaction are depicted as ball-and-stick models. Dotted lines denote hydrogen bonds. Atoms are colored as described in Figs. 1 and 2. b, ribbon diagram showing only four monomers in the decameric reduced StAhpC. c, electrostatic potential surface map of the oxidized EcTpx dimer, colored with regions of positive charge in blue and negative charge in red. The view is rotated by about 90° horizontally from that in a. A boxed depression is a putative substrate binding site. The disulfide cystine is shown in yellow. This figure was made using the program DS ViewerPro (www.accelrys.com/). d, stereoview of a binding site from the modeled structure of the reduced EcTpx dimer in complex with 15-HPETE (light green). Chains in different monomers are shown in different colors. e, stereoview of a putative binding site from the crystal structure of the reduced StAhpC. Three monomers may be involved in the formation of the binding site. f, stereoview of a putative binding site from the crystal structure of the reduced PrxV with a bound benzoic acid (light green).

Substrate Binding Sites of 2-Cys Prxs—A solvent accessible surface diagram of the EcTpx dimer in the oxidized state is shown in Fig. 4c. It clearly shows a large depression that can be readily recognized as the putative substrate binding site. The cleft becomes considerably narrower in the reduced state since the C_P-loop moves inward forming a helical turn. The binding site in the modeled structure of reduced EcTpx is shown in Fig. 4d. Also shown together is a putative model of 15-HPETE that is docked into the binding site. In this study, modeling was done rather extensively for the apo enzyme, but not for the complex. A large flexible substrate such as 15-HPETE has many rotatable bonds and docking of 15-HPETE into EcTpx involves an exploration of the multidimensional configurational hyperspace, which is a formidable task. In this study, docking was performed manually although the optimization of the initial model was done extensively. Thus the geometry of the binding site itself seems correct, but the configuration of the bound substrate may merely represent a plausible binding mode. Nevertheless the model of the complex appears reasonable on the whole and helps to delineate the nature of the binding site of EcTpx.

The binding cleft of EcTpx is approximately L-shaped with the active site pocket at one end of the cleft and is extended to the monomer-monomer interface. Four different chains of the monomer A and two chains of the monomer B are involved in the formation of the binding site of monomer A. The cleft is shaped mainly by eight hydrophobic residues, which include Phe⁷ on the N-terminal -hairpin, Pro⁵⁴ conserved at the active site, Pro¹²⁶, Leu¹²⁷ and Leu¹³⁰ on the S-loop (see Fig. 2b), and Ile¹⁵³ on the R-chain of the monomer A, and Leu³⁵ on the -hairpin and Phe⁸⁹ on 2 of the monomer B. The binding cleft is highly hydrophobic and is of the right size to accommodate long fatty acid hydroperoxides such as physiological 15-HPETE, not to mention small alkyl hydroperoxides such as t-butyl hydroperoxide. The extra N-terminal -hairpin that is lacking in other Prxs plays a critical role in the formation of this well-defined binding cleft. The binding site of EcTpx seems self-explanatory regarding its substrate preference toward alkyl hydroperoxides over H₂O₂ (17, 24).

The putative substrate binding sites of StAhpC (16) and PrxV (15), both in the reduced state, are illustrated in Fig. 4, e and f, respectively, for comparison. The figures clearly show that the three Prxs significantly differ from each other in the shape and size of the putative binding site. In StAhpC, the binding site of the monomer A is formed by the monomers A, A', and B. StAhpC lacks a chain corresponding to the N-terminal -hairpin of Tpx and its S-loop is shorter by three residues than that of EcTpx. Therefore the binding site of StAhpC is wider than that of EcTpx and its overall shape is not so clearly defined as in EcTpx. The binding pocket in reduced StAhpC is partly covered by a long C-terminal chain from the monomer A', which changes its position upon oxidation. In fact, the chain after Cys¹⁶⁵ is disordered in the crystal structure of oxidized StAhpC. StAhpC and EcTpx have Pro³⁹ (Pro⁵⁴), Phe⁷⁶ (Phe⁸⁹), and Leu¹¹⁶ (Leu¹³⁰) residues in common, but StAhpC lacks the residues equivalent to Phe⁷ and Pro¹²⁶ of EcTpx. Moreover, StAhpC has Glu¹¹⁴ in place of Leu¹²⁷, Gln¹³⁸ in place of Ile¹⁵³, and Gln¹³ and His⁷⁵ near the equivalent position of Leu³⁵ of EcTpx. Consequently, the binding site of StAhpC is much less hydrophobic than that of EcTpx.

Human PrxV, being a monomeric protein, has a smaller binding site with a simpler shape compared with EcTpx and StAhpC. Interestingly, in PrxV, a long S-loop plays an important role in shaping the binding site. The S-loop of PrxV is longer by four and seven residues, respectively, than those of EcTpx and StAhpC. It contains three hydrophobic residues of Leu¹¹⁶, Ile¹¹⁹, and Phe¹²⁰, which are located near the equivalent positions of Leu³⁵ and Phe⁸⁹ in the monomer B of EcTpx. PrxV has Leu¹¹² and Thr¹⁴⁷, respectively, in place of Leu¹²⁷ and Ile¹⁵³ of EcTpx. In Fig. 4f is shown a benzoic acid that is bound to the small, hydrophobic pocket of PrxV in the crystal. E. coli BCP is also a monomeric protein and has an S-loop with the same length as that of EcTpx. These findings suggest that the binding site of BCP may be even smaller than that of PrxV.

On the basis of catalytic efficiency, EcTpx is the most potent reductant of alkyl hydroperoxides surpassing AhpC and BCP, the other E. coli Prx members (24). The properties of the binding sites may be the major factor that affects the reactivities of these Prxs, since they share the same catalytic mechanism with the same active site. In AhpC, the binding site may exist in its entirety only in the reduced decameric form that requires assembly of the oxidized dimers during the catalytic cycle. This redox-regulated adaptation of oligomeric states may act as a deleterious factor for its reactivity toward lipid hydroperoxides. In contrast, EcTpx exists as a stable dimer and the preformed binding site may always maintain its integrity, which could be a favorable factor contributing to its catalytic efficiency. These structural characteristics fully support the previous notion that the central role of Tpx in vivo involves the reduction of complex physiological lipid hydroperoxides, whereas AhpC mainly reduces H₂O₂ and also regulates H₂O₂-mediated signal transduction (20, 24).

In summary, the present crystal structure of EcTpx in the oxidized state shows, for the first time, the intrasubunit disulfide bond formed in a Prx, demonstrating that Tpx is the second type of atypical 2-Cys Prx. The extra N-terminal chain uniquely present in pathogenic bacterial Tpxs folds into a -hairpin making the overall structure very compact, and plays an important role in shaping the well-defined substrate binding site. Comparison of the crystal and modeled structures suggests that the chain containing the C_R residue as well as the C_P-loop may undergo redox-dependent loop-helix conformational change in atypical 2-Cys Prxs. EcTpx exists as a hydrogen-bonded dimer and, unlike typical 2-Cys Prxs, its C_P-loop does not interfere with dimer formation assuming LU conformation in the oxidized state. The substrate binding site of EcTpx extends to the dimer interface, and its size and shape are particularly suited for long fatty acid hydroperoxides, in agreement with its greater reactivity compared with other Prxs.

	FOOTNOTES

 
The atomic coordinates and structure factors (code 1QXH) have been deposited in the Protein Data Bank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/).

^* This work was supported by a grant from the Korea Science and Engineering Foundation through the Center for Molecular Catalysis at Seoul National University and in part by a grant from the 21C Frontier R&D Program, Ministry of Science and Technology, Korea, through the Center for Biological Modulators. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

A recipient of a fellowship from the Brain Korea 21 Program.

^** To whom correspondence should be addressed. Tel.: 82-2-880-6656; Fax: 82-2-889-5304; E-mail: nswcshin{at}plaza.snu.ac.kr.

¹ The abbreviations used are: Prxs, peroxiredoxins; Tpx, thiol peroxidase; EcTpx, E. coli thiol peroxidase; AhpC, alkyl hydroperoxide reductase peroxidase component; StAhpC, S. typhimurium AhpC; Trx, thioredoxin; BCP, bacterioferritin-comigratory protein; C_P, peroxidatic cysteine; C_R, resolving cysteine; C_P-SOH, cysteine sulfenic acid; 15-HPETE, 15-hydroperoxyeicosatetraenoic acid; FF, fully folded; LU, locally unfolded; MIRAS, multiple isomorphous replacement with anomalous scattering; NCS, non-crystallographic symmetry; MD, molecular dynamics; r.m.s., root mean square.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS AND DISCUSSION REFERENCES

 

1. Chae, H. Z., Robinson, K., Poole, L. B., Church, G., Storz, G., and Rhee, S. G. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, 7017-7021[Abstract/Free Full Text]
2. Hofmann, B., Hecht, H. J., and Flohé, L. (2002) Biol. Chem. 383, 347-364[CrossRef][Medline] [Order article via Infotrieve]
3. Bryk, R., Griffin, P., and Nathan, C. (2000) Nature 407, 211-215[CrossRef][Medline] [Order article via Infotrieve]
4. Forman, H. J., and Cadenas, E. (1997) Oxidative Stress and Signal Transduction, Chapman and Hall, New York
5. Rhee, S. G., Kang, S. W., Chang, T.-S., Jeong, W., and Kim, K. (2001) IUBMB Life 52, 35-41[Medline] [Order article via Infotrieve]
6. Fujii, J., and Ikeda, Y. (2002) Redox Rep. 7, 123-130[CrossRef][Medline] [Order article via Infotrieve]
7. Wood, Z. A., Schröder, E., Harris, J. R., and Poole, L. B. (2003) Trends Biochem. Sci. 28, 32-40[CrossRef][Medline] [Order article via Infotrieve]
8. Choi, H. J., Kang, S. W., Yang, C. H., Rhee, S. G., and Ryu, S. E. (1998) Nat. Struct. Biol. 5, 400-406[CrossRef][Medline] [Order article via Infotrieve]
9. Wang, X., Phelan, S. A., Forsman-Semb, K., Taylor, E. F., Petros, C., Brown, A., Lerner, C. P., and Paigen, B. (2003) J. Biol. Chem. 278, 25179-25190[Abstract/Free Full Text]
10. Hirotsu, S., Abe, Y., Okada, K., Nagahara, N., Hori, H., Nishino, T., and Hakoshima, T. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, 12333-12338[Abstract/Free Full Text]
11. Schröder, E., Littlechild, J. A., Lebedev, A. A., Errington, N., Vagin, A. A., and Isupov, M. N. (2000) Structure (Lond.) 8, 605-615[Medline] [Order article via Infotrieve]
12. Alphey, M. S., Bond, C. S., Tetaud, E., Fairlamb, A. H., and Hunter, W. N. (2000) J. Mol. Biol. 300, 903-916[CrossRef][Medline] [Order article via Infotrieve]
13. Wood, Z. A., Poole, L. B., Hantgan, R. R., and Karplus, P. A. (2002) Biochemistry 41, 5493-5504[CrossRef][Medline] [Order article via Infotrieve]
14. Seo, M. S., Kang, S. W., Kim, K., Baines, I. C., Lee, T. H., and Rhee, S. G. (2000) J. Biol. Chem. 275, 20346-20354[Abstract/Free Full Text]
15. Declercq, J. P., Evrard, C., Clippe, A., Stricht, D. V., Bernard, A., and Knoops, B. (2001) J. Mol. Biol. 311, 751-759[CrossRef][Medline] [Order article via Infotrieve]
16. Wood, Z. A., Poole, L. B., and Karplus, P. A. (2003) Science 300, 650-653[Abstract/Free Full Text]
17. Cha, M. K., Kim, H. K., and Kim, I. H. (1995) J. Biol. Chem. 270, 28635-28641[Abstract/Free Full Text]
18. Cha, M. K., Kim, H. K., and Kim, I. H. (1996) J. Bacteriol. 178, 5610-5614[Abstract/Free Full Text]
19. Jeong, W., Cha, M. K., and Kim, I. H. (2000) J. Biol. Chem. 275, 2924-2930[Abstract/Free Full Text]
20. Costa Seaver, L., and Imlay, J. A. (2001) J. Bacteriol. 183, 7173-7181[Abstract/Free Full Text]
21. Link, A. J., Robison, K., and Church, G. M. (1997) Electrophoresis 18, 1259-1313[CrossRef][Medline] [Order article via Infotrieve]
22. Wan, X. Y., Zhou, Y., Yan, Z. Y., Wang, H. L., Hou, Y. D., and Jin, D. Y. (1997) FEBS Lett. 407, 32-36[CrossRef][Medline] [Order article via Infotrieve]
23. Zhou, Y., Wan, X. Y., Wang, H. L., Yan, Z. Y., Hou, Y. D., and Jin, D. Y. (1997) Biochem. Biophys. Res. Commun. 233, 848-852[CrossRef][Medline] [Order article via Infotrieve]
24. Baker, L. M. S., and Poole, L. B. (2003) J. Biol. Chem. 278, 9203-9211[Abstract/Free Full Text]
25. Choi, J., Choi, S., Choi, J., Cha, M. K., Kim, I. H., and Shin, W. (2003) Acta Crystallogr. Sect. D 59, 1064-1066[CrossRef][Medline] [Order article via Infotrieve]
26. Otwinowski, Z., and Minor, W. (1997) Methods Enzymol. 276, 307-326
27. Terwilliger, T. C., and Berendzen, J. (1999) Acta Crystallogr. Sect. D 55, 849-861[CrossRef][Medline] [Order article via Infotrieve]
28. Terwilliger, T. C. (2000) Acta Crystallogr. Sect. D 56, 965-972[CrossRef][Medline] [Order article via Infotrieve]
29. Jones, T. A., Zou, J. Y., Cowan, S. W., and Kjeldgaard, M. (1991) Acta Crystallogr. Sect. A 47, 110-119
30. Brünger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse-Kunstleve, R. W., Jiang, J. S., Kuszewski, J., Nilges, M., Pannu, N. S., Read, R. J., Rice, L. M., Simonson, T., and Warren, G. L. (1998) Acta Crystallogr. Sect. D 54, 905-921[CrossRef][Medline] [Order article via Infotrieve]
31. Laskowski, R. A., MacArthur, M. W., Moss, D. S., and Thornton, J. M. (1993) J. Appl. Crystallogr. 26, 283-291[CrossRef]
32. Kraulis, P. J. (1991) J. Appl. Crystallogr. 24, 946-950[CrossRef]
33. Merritt, E. A., and Murphy, M. E. P. (1994) Acta Crystallogr. Sect. D 50, 869-873[CrossRef][Medline] [Order article via Infotrieve]
34. Case, D. A., Pearlman, D. A., Caldwell, J. W., Cheatham III, T. E., Wang, J., Ross, W. S., Simmerling, C. L., Darden, T. A., Merz, K. M., Stanton, R. V., Cheng, A., Vincent, J. J., Crowley, M., Tsui, V., Gohlke, H., Radmer, R. J., Duan, Y., Pitera, J., Massova, I., Seibel, G. L., Singh, U. C., Weiner, P. K., and Kollman, P. A. (2002) AMBER 7, University of California, San Francisco
35. Cho, S. W., Lee, S., and Shin, W. (2001) J. Mol. Biol. 314, 863-878[CrossRef][Medline] [Order article via Infotrieve]
36. Kabsch, W., and Sander, C. (1983) Biopolymers 22, 2577-2637[CrossRef][Medline] [Order article via Infotrieve]
37. Kim, S. J., Woo, J. R., Hwang, Y. S., Jeong, D. G., Shin, D. H., Kim, K., and Ryu, S. E. (2003) J. Biol. Chem. 278, 10790-10798[Abstract/Free Full Text]
38. Kong, W., Shiota, S., Shi, Y., Nakayama, H., and Nakayama, K. (2000) Biochem. J. 351, 107-114[CrossRef][Medline] [Order article via Infotrieve]
39. Chang, T.-S., Jeong, W., Choi, S. Y., Yu, S., Kang, S. W., and Rhee, S. G. (2002) J. Biol. Chem. 277, 25370-25376[Abstract/Free Full Text]
40. Barton, G. J. (1993) Protein Eng. 6, 37-40[Free Full Text]

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				M. Murcia, J. D. Faraldo-Gomez, F. R. Maxfield, and B. Roux Modeling the structure of the StART domains of MLN64 and StAR proteins in complex with cholesterol J. Lipid Res., December 1, 2006; 47(12): 2614 - 2630. [Abstract] [Full Text] [PDF]

				V. Noguera-Mazon, J. Lemoine, O. Walker, N. Rouhier, A. Salvador, J.-P. Jacquot, J.-M. Lancelin, and I. Krimm Glutathionylation Induces the Dissociation of 1-Cys D-peroxiredoxin Non-covalent Homodimer J. Biol. Chem., October 20, 2006; 281(42): 31736 - 31742. [Abstract] [Full Text] [PDF]

				M. Trujillo, P. Mauri, L. Benazzi, M. Comini, A. De Palma, L. Flohe, R. Radi, M. Stehr, M. Singh, F. Ursini, et al. The Mycobacterial Thioredoxin Peroxidase Can Act as a One-cysteine Peroxiredoxin J. Biol. Chem., July 21, 2006; 281(29): 20555 - 20566. [Abstract] [Full Text] [PDF]

				M.-K. Cha, W.-C. Kim, C.-J. Lim, K. Kim, and I.-H. Kim Escherichia coli Periplasmic Thiol Peroxidase Acts as Lipid Hydroperoxide Peroxidase and the Principal Antioxidative Function during Anaerobic Growth J. Biol. Chem., March 5, 2004; 279(10): 8769 - 8778. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 278/49/49478    most recent M309015200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Choi, J. Articles by Shin, W. Search for Related Content PubMed PubMed Citation Articles by Choi, J. Articles by Shin, W. Social Bookmarking                      What's this?