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

This Article Abstract Full Text (PDF) Alert me when this article is cited 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 Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Powlowski, J. Articles by Sahlman, L. Search for Related Content PubMed PubMed Citation Articles by Powlowski, J. Articles by Sahlman, L. Social Bookmarking                      What's this?

J Biol Chem, Vol. 274, Issue 47, 33320-33326, November 19, 1999

Reactivity of the Two Essential Cysteine Residues of the Periplasmic Mercuric Ion-binding Protein, MerP^*

Justin Powlowski and Lena Sahlman

From the Department of Chemistry and Biochemistry, Concordia University, Montreal, Quebec H3G 1M8, Canada

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Reactivities of the two essential cysteine residues in the heavy metal binding motif, MTC₁₄AAC₁₇, of the periplasmic Hg²⁺-binding protein, MerP, have been examined. While Cys-14 and Cys-17 have previously been shown to be Hg²⁺-binding residues, MerP is readily isolated in an inactive Cys-14-Cys-17 disulfide form. In vivo results demonstrated that these cysteine residues are reduced in the periplasm of Hg²⁺-resistant Escherichia coli. Denaturation and redox equilibrium studies revealed that reduced MerP is thermodynamically favored over the oxidized form. The relative stability of reduced MerP appears to be related to the lowered thiol pK_a (5.5) of the Cys-17 side chain. Despite its much lower pK_a, the Cys-17 thiol is far less accessible than Cys-14, reacting 45 times more slowly with iodoacetamide at pH 7.5. This is reminiscent of proteins such as thioredoxin and DsbA, which contain a similar C-X-X-C motif, except in those cases the more exposed thiol has the lowered pK_a. In terms of MerP function, electrostatic attraction between Hg²⁺ and the buried Cys-17 thiolate may be important for triggering the structural change that MerP has been reported to undergo upon Hg²⁺ binding. Control of cysteine residue reactivity in heavy metal binding motifs may generally be important in influencing specific metal-binding properties of proteins containing them.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Bacterial resistance to mercuric ion is mediated by the polypeptides encoded by mer operons. Although the specific number of required polypeptides varies according to species, all operons appear to encode: one or more regulatory proteins (MerR, MerD); a periplasmic mercuric ion-binding protein (MerP); one or more integral membrane proteins thought to be required for mercuric ion transport (MerT, MerC); and a cytoplasmic mercuric ion reductase (MerA), which reduces intracellular Hg²⁺ ion to a volatile form, Hg^o (reviewed in Refs. 1 and 2).

The mer operon-encoded periplasmic protein, MerP, from the transposon Tn21 has a molecular mass of 7500 Da after removal of a periplasmic signal sequence (1, 3, 4). There are no tryptophan or histidine residues in the mature protein and only two cysteine residues, at positions 14 and 17. MerP has previously been shown to specifically bind Hg²⁺ in the presence of external thiols via these two cysteine residues (4, 5). The recently published structure of MerP revealed that Cys-14 is surface-exposed and Cys-17 is buried inside the protein; once Hg²⁺ is bound to the thiols, they are both surface-exposed (6). The structural change that accompanies Hg²⁺ binding has been proposed to be important for the interaction of Hg²⁺-loaded MerP with the inner membrane transport protein(s). Although it has been postulated that the function of MerP is to transfer Hg²⁺ to the mer operon-encoded integral membrane proteins for passage across the inner membrane (7), current evidence suggests that MerP is dispensable for transport. Instead, it may function as a mercuric ion "sponge" to protect components of the periplasm from the toxic effects of this heavy metal (8, 9).

MerP has been shown to exist in vitro in oxidized (disulfide) and reduced (dithiol) forms, but the redox state of the protein in the bacterial periplasm has never been reported. In order to bind Hg²⁺ via its thiols, MerP should be present in the reduced form. However, it is generally accepted that the periplasm is an oxidizing environment, with proteins such as DsbA and DsbB catalyzing disulfide bond formation in many bacterial periplasmic proteins (reviewed in Ref. 10). In this investigation, the redox state in vivo and the reactivity of the thiol groups of MerP have been investigated.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Chemicals-- All chemicals were of the highest purity available. ²⁰³HgCl₂ was purchased from Amersham Pharmacia Biotech (Stockholm, Sweden).

Bacterial Strains and Manipulations-- The gene, merP from transposon Tn21, was expressed in E. coli BL21(DE3) (11) from the T7 promoter of pCA (4), or using the plasmid pDU1003, which contains the complete Hg²⁺-inducible mer operon (12). Expression of the MerP variants C14A, C14S, C17A, and C17S was as described previously (5). The C17D variant was constructed using the QuikChange method (Stratagene). The template was pCAmerP, and the mutagenic primers were 5'-GACTTGCGCCGCGGACCCGATCACAGTC-3' and its complement, obtained from Biocorp Inc. (Montreal, Canada). Nucleotide sequence analysis of the complete gene was performed at the Sheldon Biotechnology Center (Montreal, Canada), and confirmed the presence of the desired mutation.

In all cases, cultures were grown in LB medium, plus the appropriate antibiotic, to an A₆₀₀ of 0.8-1.0 before induction. Induction from T7-based plasmids was achieved by adding isopropyl-1-thio--D-galactopyranoside to a concentration of 0.5 mM, followed by further growth for 3 h at 37 °C. For induction from pDU1003, HgCl₂ was added to an initial concentration of 20 µM, and a second aliquot (20 µM) was added after an additional 1 h of growth at 37 °C.

Release of periplasmic proteins was achieved either using chloroform (13) or lysozyme-EDTA (14). In order to examine the redox state of MerP, iodoacetate was included to trap free thiols at all stages of the release procedure (14). Alternatively, cultures were processed by precipitation of cellular proteins using trichloroacetic acid, followed by dissolution of the precipitated proteins in 6 M urea, containing iodoacetic acid, essentially as described (15).

Proteins and Protein Modification-- MerP and all of the variants were purified in the absence of added cysteine, as described previously; chromatography on hydroxylapatite was usually included as a final step (4, 5). Since all of these proteins were from recombinant sources never exposed to mercuric ion, residual bound Hg²⁺ is not a concern in any experiments done with purified MerP or variants.

When required, wild-type MerP was reduced either using dithiothreitol (DTT)¹ or tris-(2-carboxyethyl)phosphine (TCEP), under conditions described in the text. Modification of MerP by iodoacetate was accomplished at room temperature in the presence of sodium iodoacetate (0.1 M) in 0.1 M Tris-HCl, pH 8.1. The modification reactions were generally run in the dark for 10-30 min, and then placed on ice before analysis. In some cases, urea was included during the modification reaction.

Protein concentrations were estimated by the method of Gill and von Hippel (16).

Thiol Group Titration-- The pK_a values of cysteine thiol groups were monitored by the change in absorbance at 240 nm accompanying ionization (17). Over the pH range tested, the oxidized form of MerP exhibited no absorbance change at 240 nm, indicating that the changes observed during pH titration were due to thiol group ionization. In preparation for these experiments, samples were reduced at room temperature for 1-1.5 h with a 10-fold excess of TCEP. Excess reductant was removed by gel filtration on Econo-Pac10DG columns (Bio-Rad) equilibrated with 10 mM acetate buffer, pH 5.0, containing 0.2 mM EDTA. Samples were then concentrated using Centricon-3 (Amicon) ultrafiltration. Absorbance measurements were made by diluting a 30-50-fold concentrated solution of protein into 0.1 M acetate-MES-Tris buffers of varying pH (18) containing 0.1 M KCl. Water was added to give final buffer and KCl concentrations of 0.05 M each, and protein concentrations in the range of 50-120 µM. Spectra for each sample, contained in a microcell, were scanned immediately in the region 220-400 nm. A separate base line was recorded for each buffer prior to scanning the protein solution.

Thiol pK_a values were estimated by non-linear least-squares fitting of plots of extinction (240 nm) versus pH using equations for 1 or 2 pK_a values as supplied with Grafit 3 (Erithacus Software, Middlesex, United Kingdom).

Determination of Redox Potential-- The redox equilibrium constant of MerP with glutathione, K_ox, is shown in Equation 1.

(Eq. 1)

GSH and GSSG are the oxidized and reduced forms of glutathione, respectively. K_ox was estimated after incubation of MerP with various ratios of oxidized and reduced glutathione, essentially as described elsewhere (19). Briefly, MerP_ox (90 µM) was incubated anaerobically for 24 h in 0.1 M potassium phosphate buffer, pH 7.5, containing EDTA (1 mM), KCl (0.1 M), GSH (10.4 mM), and various concentrations of GSSG. Oxidized and reduced forms of MerP were quantitated by HPLC (see below) after quenching to pH 2 with HCl.

Data were fitted to Equation 2 (19, 20).

(Eq. 2)

where R = [GSH]/[GSSG], and K_mix is the equilibrium constant between glutathione and the MerP-glutathione mixed disulfide. Only small quantities (10%) of mixed disulfide accumulated, and a K_mix of 0.56 was estimated from a plot of [MerP-SSG]/MerP_red versus [GSSG]/[GSH], as described in Ref. 21.

The standard redox potential was calculated using the Nernst equation and a standard redox potential of +0.24 V (22) for the glutathione redox pair.

Quantitation of Oxidized, Reduced, and Mixed Disulfides of MerP-- Different forms of MerP were quantitated using HPLC. The reduced, oxidized, and modified forms of the protein were separated on a Vydac C-18 Protein and Peptide column (catalog no. 218TP54) at a flow rate of 1.5 ml/min. Samples were acidified to a pH of approximately 2 with HCl, and injected onto the column, which was equilibrated with 0.1% trifluoroacetic acid/water (68%):0.1% trifluoroacetic acid/acetonitrile (32%). After 2 min a linear gradient was run over 20 min to 0.1% trifluoroacetic acid/water (63%):0.1% trifluoroacetic acid/acetonitrile (37%), and then the column was washed for 3 min with this solvent. Quantitation was achieved by calculating peak areas after detection at 215 nm.

Denaturation of MerP-- MerP was denatured with different concentrations of guanidine HCl in 0.1 M Tris-SO₄ buffer, pH 7.5, for 18 h at room temperature. The protein concentration was approximately 15 µM; guanidine hydrochloride stock solution concentrations were determined by refractometry (23). For denaturation of MerP_red, 2 mM DTT was added to reduce a stock solution of protein. Samples were transferred to a glove box, whereupon the protein was diluted in 0.1 M Tris-SO₄ buffer, pH 7.5, containing 1-2 mM DTT and different concentrations of guanidine HCl. The samples were left for 18 h under anaerobic conditions. Samples of single cysteine variants of MerP were prepared similarly, but were not anaerobic. CD spectra were recorded at 23 °C using a Jasco J-710 spectropolarimeter. Each spectrum was an average of five accumulations, with step size 0.2 nm and bandwidth 1 nm.

Data analysis was carried out by non-linear fitting, using Sigmaplot 4 (SPSS Scientific, Chicago, IL), of measured ellipticities at different guanidine HCl concentrations as described (24-26). Fraction folded was calculated using the fitted parameters (25).

Kinetics of Iodoacetamide Reactions with MerP and Variants-- Reactions between iodoacetamide and MerP were carried out anaerobically at 25 °C in 0.1 M phosphate buffer, pH 7.5, containing 1 mM EDTA and 0.1 M KCl. MerP or variants (75-160 µM), prepared as described for the pH titration experiments, were incubated with various concentrations of iodoacetamide. At different times, reactions were quenched by acidification with HCl to a pH of approximately 2, and analyzed immediately by HPLC as described above.

Data were analyzed using non-linear least-squares fitting to a single exponential using Sigmaplot 4. In the case of wild-type MerP_red, the pseudo-first order rate constants obtained at 3 and 30 mM iodoacetamide were divided by iodoacetamide concentration to estimate second order rate constants; data for the variants were obtained at several iodoacetamide concentrations. The pH-independent rate constant, k_S-, for reactivity of a thiolate with iodoacetamide was estimated using Equation 3.

(Eq. 3)

Analytical Techniques-- Quantitation of protein thiol groups was performed using Ellman's reagent (27). UV-visible absorbance spectra were acquired using a Philips 8715 spectrophotometer.

Electrospray ionization mass spectrometry was performed using a Finnigan SSQ 7000 single quadrupole mass spectrometer interfaced to a liquid chromatograph. Samples were introduced via a C-18 (5 µm) column (1 × 10 mm) at flow rate of 80 µl/min. Samples were loaded in 15% acetonitrile, 0.05% trifluoroacetic acid, and after washing the column for 3 min after injection, proteins were eluted with 70% acetonitrile, 0.05% trifluoroacetic acid. Data were analyzed using software supplied with the instrument.

Polyacrylamide Gels-- Native gels were run in the absence of SDS, as described previously (4, 28). Since MerP is positively charged at pH 7.5, current flow was reversed, so the samples were run from the anode toward the cathode. The purity of purified MerP samples was checked using SDS-polyacrylamide gel electrophoresis (29).

Binding Studies-- Binding of Hg²⁺ to MerP (9 µM) was measured as described previously (4), except the buffer used was 50 mM sodium acetate, pH 4.0. Briefly, this binding assay is based on incubation of MerP with ²⁰³HgCl₂ solutions of varying concentration with a 4:1 ratio of cysteine:Hg to minimize nonspecific binding. Ultrafiltration in microconcentrators (Microsep) followed by liquid scintillation counting of ²⁰³Hg in the upper and lower chambers allowed the concentration of MerP-bound Hg²⁺ to be determined; corrections for nonspecific binding to the concentrator were determined for each HgCl₂ concentration by identical ultrafiltration experiments carried out in the absence of MerP (4).

Data from two separate experiments were analyzed using Grafit with non-linear least-squares fitting to the following equation to estimate the capacity for binding Hg²⁺ (C), and the apparent dissociation constant (K_d).

(Eq. 4)

y is the concentration of bound Hg²⁺, and t the total concentration of added Hg²⁺. This equation uses total added Hg²⁺, rather than the amount of free versus bound, since the presence of cysteine in the assay makes it difficult to estimate the concentration of free Hg²⁺.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Electrophoretic Methods for Detecting Oxidized and Reduced Forms of MerP-- The in vivo oxidation state of MerP has never been reported, although only the reduced form can bind Hg²⁺via Cys-14 and Cys-17 (4-6). These cysteine residues can form a disulfide bond since reduced (MerP_red) and oxidized (MerP_ox) forms have been obtained by isolation of the protein in the presence or absence of cysteine in purification buffers (4). As is shown in Fig. 1, purified MerP_red (lane 1) and MerP_ox (lane 4) have different mobilities on a native gel; the reduced form migrates more slowly toward the cathode (bottom) in this gel system, possibly because of thiol group ionization at neutral pH (see below). However, in order to preserve and identify MerP_red in periplasmic extracts, it is necessary to derivatize the free thiols prior to electrophoresis.

View larger version (108K): [in this window] [in a new window]   Fig. 1.   Native polyacrylamide gel electrophoresis of different forms of purified MerP: proteins were run toward the cathode (bottom). Lane 1, MerPred; lane 2, MerPred reacted with iodoacetate in the presence of 5.6 M urea; lane 3, MerPred reacted with iodoacetate; lane 4, MerPox. Reactions of MerP (28 µM) with iodoacetate (0.1 M) were conducted at room temperature for 10 min, after which samples were placed on ice briefly before electrophoresis. MerPred was prepared by pre-incubation with a 50-fold excess of DTT.

Reaction with iodoacetate or iodoacetamide to trap reduced thiols and produce proteins with altered electrophoretic mobility has been used for a number of other thiol-containing redox proteins (see, e.g., Refs. 15 and 21). Reaction of MerP_red with iodoacetamide resulted in the conversion of the reduced form to a species that migrated at the position of the oxidized protein, as would be expected by blockage of an ionized thiol group(s) with the uncharged acetamide group (data not shown). On the other hand, reaction of reduced MerP with iodoacetate resulted in production of two slower migrating forms (Fig. 1, lane 3). Only the most slowly migrating form was observed over longer periods of reaction time (data not shown), or when MerP was reacted with iodoacetate in the presence of urea (Fig. 1, lane 2). Therefore, it appears that the two thiols on the protein are modified at quite different rates. The observed change in mobility is consistent with the expected addition of negative charge to MerP upon carboxymethylation. No change in the mobility of oxidized MerP was observed in the presence of iodoacetate, as would be expected if only the protein thiols were reacting (data not shown).

The identities of the faster and slower migrating carboxymethylated forms were examined using electrospray mass spectrometry. A sample prepared in the presence of urea and showing only the upper band on a native gel exhibited a single major peak with a molecular mass of 7,590; this molecular mass corresponds to that of doubly carboxymethylated MerP (7,472 (observed) + (59 × 2)). In a sample containing mostly the faster migrating carboxymethylated form a species corresponding to singly carboxymethylated MerP (M_r = 7531) was the major product, and the doubly carboxymethylated form was the minor product.

In the next section, results are described using iodoacetate to trap and identify the reduced form of MerP from the periplasm of whole cells. The initial observations suggesting differing accessibility and pK_a values of the two thiols will be addressed further in later sections.

Redox State of MerP in Vivo-- In these experiments, cultures were exposed to iodoacetate to trap the reduced form of MerP, which was then released from the periplasm in the presence of iodoacetate and analyzed using polyacrylamide gel electrophoresis. Similar methods have been used for other periplasmic thiol-containing proteins (see, e.g., Refs. 14 and 30). When MerP was expressed alone from the T7 promoter, samples taken at various times throughout the induction period were all mostly in the oxidized form (Fig. 2A). Similar results were obtained when HgCl₂ (20 µM) was added together with the inducer (data not shown), indicating that the presence of Hg²⁺ is not sufficient to maintain the reduced form. However, MerP was mainly in the reduced form in periplasmic extracts from Hg²⁺-resistant cells harboring the complete operon (Fig. 2B, far left lane). In this strain, MerP remained reduced for up to 90 min after Hg²⁺ had been removed from the culture (Fig. 2B). Similar results for each strain were obtained using an alternative method (15) in which cellular proteins were precipitated with trichloroacetic acid to prevent possible nonspecific redox reactions after cell disruption (results not shown). It should be noted that the levels of MerP expression in the strains used for the experiments shown in Fig. 2 are comparable, despite the differences in the plasmid constructs. From these experiments it can be concluded that MerP is in the reduced form in the periplasm of Hg²⁺ resistant cells, and that one or more of the mer operon-encoded proteins may be involved in keeping its thiols reduced.

View larger version (29K): [in this window] [in a new window]   Fig. 2.   Iodoacetate-trapped MerP from the periplasm of cells expressing only MerP (A) and cells expressing the whole mer operon (B). Samples were prepared from equal quantities of periplasmic fractions released in the presence of chloroform and iodoacetate, as described under "Materials and Methods." Migration positions of wild-type MerP (ox, red) and carboxymethylated derivatives (ac, ac2) are indicated. A, E. coli BL21(DE3) harboring pCA(merP) were induced with isopropyl-1-thio--D-galactopyranoside and samples were removed at 0 min (lane 1), 15 (lane 2), 30 (lane 3), 60 (lane 4), and 90 (lane 5) min after induction. B, E. coli BL21(DE3) harboring pDU1003 were induced with HgCl2 for 90 min, at which point a sample was taken (0 min). The remaining cells were harvested by centrifugation, and resuspended in fresh medium at 37 °C with (+), or without (), HgCl2 and including chloramphenicol (100 µg/ml) to prevent new protein synthesis. Samples were taken at the times indicated after the 0-min sample.

Redox Properties of Purified MerP-- The apparent stability of the reduced form of MerP in the oxidizing environment of the periplasm prompted us to examine the redox potential of the protein. Oxidized MerP was incubated anaerobically with various ratios of oxidized and reduced glutathione, and the ratios of oxidized and reduced MerP were estimated using reverse-phase HPLC after acid quench (Fig. 3). The equilibrium constant (K_ox) with glutathione was estimated to be 27 mM, which corresponds to a redox potential of 190 mV. This value is approximately midway between the redox potentials of oxidizing proteins such as DsbA (approximately 100 mV) (21, 31), and reductants such as thioredoxin (270 mV) (32).

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Determination of redox equilibrium between MerP and glutathione. Experimental conditions and equation of the fitted line (solid) are indicated under "Materials and Methods."

Thiol Group pK_a Values-- The reactivities of the two thiol groups in MerP should be governed by their respective pK_a values, since the ionized thiol is more reactive than the protonated form (33). Native gel electrophoresis results (see above) suggested that at least one of the two thiol groups is ionized at neutral pH. The mobilities of single cysteine variants of MerP were thus examined using the native gel electrophoresis system, both before and after reaction with iodoacetate (Fig. 4) or iodoacetamide (data not shown). Since mass spectrometry indicated modification of some variants by cysteine, or dimerization (data not shown), all samples were reduced using DTT or TCEP prior to electrophoresis or modification. The majority of the reduced C14S sample ran at the position of the reduced wild-type sample (Fig. 4, lane 1), while reduced C17S mostly ran at the position of the oxidized native MerP (Fig. 4, lane 2). Thus, it appears that it is Cys-17 which is ionized at the pH of the gel (gel buffer pH = 7.0). Consistent with this, C17D, unlike C17S, migrates at the position of reduced MerP (Fig. 4, lane 3).

View larger version (83K): [in this window] [in a new window]   Fig. 4.   Electrophoretic mobility of MerP variants (lanes 1-3) and their carboxymethylated (CM) derivatives (lanes 4-6). Electrophoresis and modification conditions were as described in Fig. 1 except that the proteins were initially reduced using a 10-fold excess of TCEP, and a 40-fold excess of DTT was added prior to electrophoresis. Lanes 1 and 4, C14S and CM-C14S; lanes 2 and 5, C17S and CM-C17S; lanes 3 and 6, C17D and CM-C17D. Migration positions of wild-type MerP (ox, red) and carboxymethylated derivatives (ac, ac2) are indicated.

The electrophoretic mobilities of unmodified proteins can be compared with the mobilities of carboxymethylated samples (Fig. 4, lanes 4-6). The carboxymethylated and non-carboxymethylated forms of C14S have the same mobilities (Fig. 4, lanes 1 and 4), as would be expected if Cys-17 is already ionized. Carboxymethylation of Cys-14 in C17S results in a complete shift to the position of reduced MerP (Fig. 4, lane 5), as would be expected for the conversion of an unionized thiol to the negatively charged carboxymethylated derivative. Finally, carboxymethylated C17D migrates at the same position as dicarboxymethylated MerP_red (Fig. 4, lane 6). These observations add support for the conclusion that Cys-17 is ionized at neutral pH, whereas Cys-14 is not.

Since MerP has only two thiol groups, estimation of the pK_a values is possible by pH titration and monitoring the appearance of the thiolate forms at 240 nm (  4,000 M¹ cm¹) (17). Controls using oxidized MerP and variants in which each thiol had been replaced by alanine or serine allowed specific contributions of each thiol group to the absorbance changes to be assessed. Representative results of these titrations are shown in Fig. 5, and pK_a values are summarized in Table I. The native reduced protein showed two ionizations with pK_a values of 5.5 and 9.16 (Fig. 5A); oxidized MerP showed no appreciable change in absorbance over this pH range, indicating that ionization of the single tyrosine residue does not contribute to the observed absorbance changes (data not shown). The C14A and C14S variants showed single absorbance changes titrating with pK_a values of 6.08 and 5.8, respectively, while the C17A and C17S variants also revealed single thiol ionizations with pK_a values of 7.7 and 7.4, respectively (Table I and Fig. 5B). Thus, it appears that the thiol with a pK_a of 5.5 in native MerP is Cys-17, while that with a pK_a of 9.2 is Cys-14, although the pK_a of this thiol appears to be considerably perturbed in the C17A and C17S variants.

View larger version (22K): [in this window] [in a new window]   Fig. 5.   pH titrations of wild-type (A) and variant (B) MerP. The variant proteins in B are: C14S (), C17S (), and C17D (). Protein concentrations varied between 50 and 115 µM.

A possible reason for the lowered pK_a values of Cys-14 in the C17A and C17S variants is the absence of a negative charge at residue 17 at neutral pH. Results obtained for the C17D variant showed a single thiol titrating with a pK_a of 9.1 (Fig. 5B), confirming the conclusion that it is Cys-14 which has this pK_a, and suggesting that the presence of a negative charge at position 17 is necessary to maintain the native-like conformation around Cys-14 in reduced MerP. In turn, this could affect the pK_a of Cys-14.

Kinetics of MerP Thiol Reaction with Iodoacetamide-- In the initial electrophoretic analysis of modified MerP (Fig. 1), it was apparent that the two thiol groups reacted with iodoacetate at quite different rates. This observation is consistent with the NMR structure of reduced MerP, which shows that Cys-14 is on the surface of the protein while Cys-17 is buried (6). However, in addition to accessibility, the reactivities of the thiol will also depend on the fraction in the thiolate form. Since Cys-17 has a much lower pK_a than Cys-14, it was of interest to determine which of the two thiols reacts rapidly with a small thiol reagent such as iodoacetamide.

The kinetics of reaction of C17D and C14S with iodoacetamide are shown in Fig. 6. From these data, it is clear that Cys-14 reacts much more rapidly (k_app = 0.95 M¹ s¹, Fig. 6B) with this small neutral thiol reagent than Cys-17 (k_app = 0.021 M¹ s¹, Fig. 6A), which is fully ionized at the pH of the experiment. Comparable apparent second order rate constants of 1.5 and 0.029 M¹ s¹ were observed for conversion of MerP_red first to singly and then to doubly modified protein. Correcting for the percentage of each thiol in the anion form to obtain k_S-, the pH-independent rate constant for reaction of the thiolate form, gave a value for Cys-14 that is 1800 times higher than that for Cys-17 in the single cysteine variants. This indicates that Cys-17 is sterically inaccessible and very slow to interact with external reagents despite the fact that, at pH 7.5, it is in the very reactive thiolate form and Cys-14 is not.

View larger version (30K): [in this window] [in a new window]   Fig. 6.   Kinetics of reaction between iodoacetamide and C14S (top) or C17D (bottom) at pH 7.5 and 25 °C. Reactions were as described under "Materials and Methods." Second-order rate constants were estimated from the slopes of the plots shown in the insets.

Stabilities of Oxidized, Reduced, and Variant MerP-- The influence of structural differences between MerP_ox and MerP_red on protein stability were probed using guanidinium hydrochloride denaturation studies (Fig. 7). The thermodynamic parameters obtained assuming a two-state transition are summarized in Table I. The thermodynamic parameters obtained for denaturation of MerP_ox are in excellent agreement with those reported previously (26). Changes in free energy of unfolding, G_u^50%, were calculated by multiplying the average m-value for the two forms of MerP by the difference in transition midpoints ([guanidinium hydrochloride]_1/2) between the variant and wild-type proteins (34). MerP_red unfolded with a [guanidinium hydrochloride]_1/2 significantly higher than that observed for the oxidized form (Fig. 7), with a G_u^50% of 0.85 kcal/mol. MerP_red is thus thermodynamically more stable than MerP_ox.

View larger version (21K): [in this window] [in a new window]   Fig. 7.   Denaturation of wild type and variant MerP by guanidinium hydrochloride. Unfolding was monitored by circular dichroism spectroscopy, and curve-fitting to a two-state transition was as described under "Materials and Methods." All transitions were reversible (data not included).

The variant MerPs were generally less stable than wild-type MerP_red (Table I). The C17A and C17S variants were, like MerP_ox, 0.7-0.8 kcal/mol less stable than MerP_red, but the C17D variant was not destabilized relative to MerP_red. These results support the notion that the negative charge at position 17 is important in maintaining a conformation more like the reduced than the oxidized form of MerP. Also consistent with this, the stability of C14S, where ionized Cys-17 is present, is almost identical to that of MerP_red. However, the stability of C14A is anomalously low, suggesting that the hydrophobic alanine residue is not well tolerated at position 14.

Binding of Hg²⁺ at Low pH-- The finding that one of the MerP thiol groups is ionized at neutral pH prompted us to examine whether Hg²⁺ binding is affected when this thiol group is protonated. Titration of reduced MerP (9 µM) with Hg²⁺ was carried out in the presence of external cysteine, as described previously (4) except at pH 4 (Fig. 8). The apparent K_d was 4.7 ± 1.9 µM, with a total binding capacity of 6.0 µM ± 0.6 µM (0.7 mol/mol protein). These results are similar to those reported for pH 7.3, where the apparent K_d and binding capacity using 10 µM MerP were 3.7 ± 1.3 and 8.8 ± 0.6 µM (0.88 mol/mol protein), respectively (4).

View larger version (17K): [in this window] [in a new window]   Fig. 8.   Binding of Hg2+ to MerP at pH 4. The assay contained the following components: 50 mM sodium acetate buffer, pH 4.0, 9 µM reduced MerP, HgCl2 (including 203HgCl2) in varying concentrations, and cysteine in a 4-fold excess over HgCl2. The amount of bound versus total Hg2+ (solid line) was estimated using the fitting procedure described in the text.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The role of periplasmic MerP in mercuric ion resistance is postulated to be scavenging of mercuric ion via the heavy metal binding motif, GMTC₁₄XXC₁₇, for later transfer to Hg²⁺ translocating membrane proteins. Consistent with this role, MerP with both Cys-14 and Cys-17 reduced has been shown to bind mercuric ion with high affinity, even in the presence of excess free cysteine (4-6). Since MerP is readily isolated in the oxidized form, with a disulfide formed between Cys-14 and Cys-17, the reactivity of these thiol groups is an important determinant of the Hg²⁺ binding role of MerP.

Since various redox catalysts (the Dsb proteins) active with protein thiols/disulfides are present in the bacterial periplasm (reviewed in Ref. 10), a relevant question is whether MerP in vivo exists in the dithiol, Hg²⁺-binding form. Trapping experiments established that in the absence of expression of the other proteins of the mer operon, periplasmic MerP was mainly oxidized, while in cells expressing the complete operon, MerP was mainly in the reduced form. Thus, in mercuric ion-resistant cells, periplasmic MerP indeed exists in the Hg²⁺-binding dithiol form despite the presence of DsbA, a disulfide bond-forming catalyst. Our experiments indicate that Hg²⁺ is not required to maintain MerP in the reduced form. The observation that MerP was mainly in the oxidized form when expressed alone suggests the possibility that association with other mer operon proteins is important to preserve the reduced form, but this has not been confirmed.

The reduced form of MerP was found to be thermodynamically more stable than the oxidized form. While disulfide bonds in proteins are generally considered to be a stabilizing influence, in some cases they are destabilizing. A well known example of a destabilizing disulfide bond is the one found in DsbA (21). Unfolding experiments showed stabilization of MerP_red relative to the oxidized form, with G_u^50% = 0.85 kcal/mol. Unfolded MerP_ox is theoretically 1.86 kcal/mol less stable than the unfolded reduced form, assuming that the only difference between the unfolded forms (35) is the reduced entropy associated with the 4-amino acid loop present in MerP_ox. Taking this into account, the overall difference in free energy between the reduced and oxidized forms of MerP is calculated to be about 2.7 kcal/mol.

A number of factors appear to be important for stabilizing MerP_red relative to the oxidized form. One of these is the unusually low pK_a, 5.5, of the cysteine thiol at position 17. Its absence in C17S and C17A resulted in variants with stabilities similar to MerP_ox rather than to MerP_red. Interestingly, the Cys-14 thiol pK_a was also perturbed when Cys-17 was replaced by alanine or serine; this may be a result of structural changes occurring upon loss of the negative charge at position 17. Consistent with these notions, a variant in which Cys-17 was replaced with aspartate, preserving the negative charge, was almost as stable as the reduced form, and the Cys-14 thiol pK_a was not affected. Furthermore, the observation by NMR spectroscopy (36) that low pH alters the structure of MerP_red, but not MerP_ox, provides additional support for the proposal that a negative charge at position 17 helps to maintain the structure. The lowering to 5.5 of the pK_a of the Cys-17 thiol from a more typical 8.7 would be expected to stabilize MerP_red by 4.3 kcal/mol, which is greater than the observed value of 2.7 kcal/mol. However, the conformational change in going from reduced to oxidized MerP involves movement of Cys-17 from a buried to exposed position and other associated structural changes that amount to more than simple removal of a thiolate (6, 36). Thus, the low pK_a of Cys-17 is not the sole determinant of the relative stabilities of MerP_ox and MerP_red.

Like MerP, proteins such as thioredoxin, protein-disulfide isomerase and DsbA contain a pair of redox-active cysteines separated by two amino acid residues. The ability of proteins to oxidize reduced glutathione, as reflected in the equilibrium constant, K_ox, provides a way of comparing their redox properties (19). The K_ox of 27 mM obtained for MerP is between those for DsbA (approximately 0.1 mM) (21, 31), a strongly oxidizing protein, and thioredoxin (10 M) (37), a strongly reducing protein. Furthermore, studies of DsbA have indicated that there is a relationship between K_ox and the pK_a of the leaving group thiol in the disulfide. The experimentally obtained K_ox for MerP is very similar to what what would be expected for a leaving group pK_a of 5.5 (see Fig. 4 in Ref. 38), so this relationship appears to extend to MerP as well. However, it does not necessarily hold for all proteins with a CXXC motif (39).

Lowering of cysteine thiol pK_a values in a -C-X-X-C- motif has been reported for proteins such as DsbA, for which a pK_a of approximately 3.5 has been reported (40). The low pK_a thiols of both DsbA and MerP are situated at the amino-terminal ends of -helices. Stabilization of the negatively charged thiolate anion by the helix dipole lowers the pK_a (41), but in model -helical peptides this can only account for a lowering of the thiol pK_a by up to 1.6 units (42). It is interesting to note that it is only in the reduced form of MerP that Cys-17 is part of the -helix, whereas in the oxidized and Hg²⁺-bound forms the -helix starts at residues 18 and 19, respectively (6, 36). A second thiolate-stabilizing feature of both thioredoxin and DsbA is hydrogen bonding of backbone amides to the sulfur of the low-pK_a thiol (43, 44). In reduced MerP, the amide of M12 appears to be close enough to hydrogen bond with, and stabilize the thiolate of MerP. In the set of 20 NMR structures of reduced MerP, the S-amideN_Met12 and S-amideH_Met12 average distances are 3.24 and 2.61 Å, respectively, within the limits of 3.25-3.55 and 2.3-2.6 Å, respectively, expected for S-N H-bonds (45, 46). Unlike in DsbA, where His-32 appears to help stabilize the thiolate anion at C30 (38, 44), there are no charged residues in the vicinity of Cys-17 in MerP.

Another key feature of Cys-17 is its inaccessibility as measured by reactivity with iodoacetamide, a neutral solvent-borne thiol-reactive reagent. Thus, the intrinsic reactivity of Cys-17 thiolate with iodoacetamide is 1800 times lower than Cys-14 thiolate. This is consistent with the structure of MerP_red, which shows that Cys-14 is exposed on the surface while Cys-17 is buried (6). As discussed earlier, the low pK_a of the Cys-17 thiol appears to help stabilize MerP_red relative to MerP_ox, maintaining the protein in a form competent for Hg²⁺ binding. By keeping ionized Cys-17 away from the surface of the protein, undesirable reactions of this group, such as thiol exchange or oxidation, would also be minimized; surface-exposed Cys-14, which is not ionized at neutral pH, is not so susceptible to undesirable side reactions. Furthermore, by keeping the thiol groups well apart in the reduced protein, conversion to the oxidized form would be minimized.

A C-X-X-C motif with a low pK_a thiol has been well characterized in redox proteins such as thioredoxin and DsbA. In these proteins it is the exposed thiol in the cysteine redox pair that has a lowered pK_a (43, 44). This is undoubtedly related to the disulfide-exchange mechanism of these proteins, where a nucleophilic cysteine initiates attack on the polypeptide substrate (47). By contrast, the MerP C-X-X-C motif appears to be specialized for metal binding.

While the separation of thiol groups may be important in keeping MerP reduced and competent to bind Hg²⁺, how can both thiols become coordinated to mercuric ion as has been indicated by NMR experiments (6)? A comparison of the structures of free and Hg²⁺-liganded MerP indicates that Cys-17 migrates to the surface of the protein, in the process partially unravelling the amino terminus of the -helix of which it is a part (6); a similar structural change is observed after formation of MerP_ox (36). A possible mechanism for this might be that Hg²⁺ initially binds to Cys-14 at the surface of the protein and then attracts the negatively charged thiol from its inaccessible position. The structural change that results may be important to allow loaded MerP to dock with one of the Hg²⁺-transporting proteins (6, 7), but experimental evidence for docking is currently lacking.

The importance of Cys-17 thiol deprotonation for equilibrium binding of Hg²⁺ was investigated by examining mercuric ion binding at pH 4, where this residue is mostly un-ionized. These binding studies, carried out in the presence of competing cysteine ligand to minimize nonspecific binding, indicated little difference in binding parameters at pH 4 versus pH 7.3. This is not too surprising, since dissociation constants for Hg²⁺(thiol)₂ complexes are on the order 10⁴⁰ over a range of pH values (48). Therefore, it follows that the experimentally determined apparent K_d values are dominated by competitive Hg²⁺ binding to cysteine and MerP thiols, and would be unperturbed by the relatively insignificant contribution of the equilibrium between protonated and deprotonated thiols at pH 4 versus pH 7.3. In other words, extremely tight binding of Hg²⁺ to the deprotonated form would shift the equilibrium from the protonated form to the Hg complex. Indeed, NMR data for bidentate binding of Hg²⁺ to glutathione thiols showed that binding was tight over the pH range 1-13 (49). It therefore may be concluded that a fully deprotonated Cys-17 thiol is not essential for equilibrium Hg²⁺ binding, although a role for the Cys-17 thiolate in influencing Hg²⁺-binding kinetics cannot be ruled out by our data. Instead, its influence on MerP structure, and possibly structural change upon Hg²⁺ binding, are much more significant.

The low pK_a of Cys-17 may also play an important role in release of Hg²⁺ from MerP to the mercuric ion transport proteins, MerT and/or MerC. NMR studies have demonstrated that Hg²⁺ is rapidly exchanged among thiol ligands, such as those in Hg(glutathione)₂, via transient formation of an Hg(thiol)₃ complex (50). If transfer of Hg²⁺ from MerP to a thiol pair on MerT or MerC occurs via an Hg(thiol)₃ complex, the low pK_a of Cys-17 relative to the other two thiol pK_a values would favor it as a leaving group. However, it must be noted that despite its attractiveness as a hypothesis, evidence is currently lacking for direct transfer of Hg²⁺ between thiol pairs on different Mer proteins.

It is interesting to note that, although the structure of the metal-binding domain of the Menkes copper-transporting ATPase is very similar to that of MerP, no conformational change was observed upon metal ion binding to a Menkes domain (51). Furthermore, Cys-17 in the Menkes protein is exposed to solvent, and is thus unlikely to have a perturbed pK_a as in MerP. Use of the MTCXXC heavy metal binding motif to engineer metal binding sites into proteins thus may need to take into account the context of the cysteine residues to help control metal binding reactivity.

	ACKNOWLEDGEMENTS

We thank Craig Fenwick for mass spectrometry measurements and Simon Silver for a gift of pDU1003. Jack and Judy Kornblatt made many useful comments on the manuscript.

	FOOTNOTES

^* This work was supported by an operating grant from the Natural Sciences and Engineering Research Council of Canada (to J. P.). This work was initiated during a sabbatical visit (by J. P.) to the Department of Biochemistry, Umeå University, Sweden.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed: Dept. of Chemistry and Biochemistry, Concordia University, 1455 de Maisonneuve Blvd. W., Montreal, Quebec H3G 1M8, Canada. Tel.: 514-848-8727; Fax: 514-848-2868; E-mail: powlow@vax2.concordia.ca.

	ABBREVIATIONS

The abbreviations used are: DTT, dithiothreitol; MerP_ox, oxidized MerP (disulfide form); MerP_red, reduced MerP (dithiol form); MES, 2-(N-morpholino)ethanesulfonic acid; TCEP, tris-(2-carboxyethyl)phosphine; HPLC, high performance liquid chromatography.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us