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

J. Biol. Chem., Vol. 275, Issue 25, 18801-18809, June 23, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/25/18801    most recent M001001200v1 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 Hillas, P. J. Articles by Ortiz de Montellano, P. R. Search for Related Content PubMed PubMed Citation Articles by Hillas, P. J. Articles by Ortiz de Montellano, P. R. Social Bookmarking                      What's this?

The AhpC and AhpD Antioxidant Defense System of Mycobacterium tuberculosis^*

Patrick J. Hillas, Federico Soto del Alba, Julen Oyarzabal^§, Angela Wilks, and Paul R. Ortiz de Montellano^¶

From the Department of Pharmaceutical Chemistry, School of Pharmacy, University of California, San Francisco, California 94143-0446

Received for publication, February 7, 2000, and in revised form, March 30, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The peroxiredoxin AhpC from Mycobacterium tuberculosis has been expressed, purified, and characterized. It differs from other well characterized AhpC proteins in that it has three rather than one or two cysteine residues. Mutagenesis studies show that all three cysteine residues are important for catalytic activity. Analysis of the M. tuberculosis genome identified a second protein, AhpD, which has no sequence identity with AhpC but is under the control of the same promoter. This protein has also been cloned, expressed, purified, and characterized. AhpD, which has only been identified in the genomes of mycobacteria and Streptomyces viridosporus, is shown here to also be an alkylhydroperoxidase. The endogenous electron donor for catalytic turnover of the two proteins is not known, but both can be turned over with AhpF from Salmonella typhimurium or, particularly in the case of AhpC, with dithiothreitol. AhpC and AhpD reduce alkylhydroperoxides more effectively than H₂O₂ but do not appear to interact with each other. These two proteins appear to be critical elements of the antioxidant defense system of M. tuberculosis and may be suitable targets for the development of novel anti-tuberculosis strategies.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Tuberculosis, caused by opportunistic infection by Mycobacterium tuberculosis, is a leading cause of death (1). Worldwide, infection rates are increasing although in the United States the rate of tuberculosis infection has begun to decrease after an increase in the late 1980s (2). A very alarming observation is the appearance of M. tuberculosis strains resistant to many of the front-line compounds, including isoniazid, that are currently utilized to treat the disease (3). Middlebrook and co-workers (4) observed in the 1950s that M. tuberculosis strains resistant to isoniazid were devoid of catalase/peroxidase activity. This circumstantial link between peroxidase and isoniazid activities was placed on a molecular footing by Heym et al. (5, 6), who showed that isoniazid-resistant M. tuberculosis strains had deletions or mutations in the katG gene that encodes for the KatG catalase/peroxidase. The importance of KatG in the action of isoniazid was confirmed by the demonstration that transformation of Escherichia coli or Mycobacterium smegmatis, both of which are isoniazid resistant, with the katG gene rendered them sensitive to isoniazid (7). Furthermore, Johnson and Schultz (8) showed that isoniazid is oxidized by the M. tuberculosis KatG catalase/peroxidase to a number of chemically reactive products. These combined results imply that isoniazid is a prodrug that must be processed into its active form by the bacterial cell.

The critical target of activated isoniazid is not yet clear. Evidence exists that this role is played by the inhA gene product, an enoyl reductase (9), and/or by KasA, a -ketoacyl synthase (10). Both of these enzymes are involved in the biosynthesis of mycolic acid, an essential constituent of the M. tuberculosis cell wall. It is not yet clear whether one or both of these proteins is the principal target, or whether there are additional targets for activated isoniazid (11, 12).

In view of the requirement for the activation of isoniazid by the KatG catalase/peroxidase, one strategy used by the organism to overcome its sensitivity to isoniazid is to suppress the catalase/peroxidase activity through mutation of the katG gene. However, the survival of the bacterium requires that it compensate in some manner for loss of the catalase/peroxidase, as it must still attenuate the oxidative stress caused by peroxides and other reactive oxygen species. M. tuberculosis primarily resides in the macrophages of the host, where it is subjected to a highly oxidative environment (13, 14). This environment includes peroxides formed by the oxidative burst, species such as peroxynitrite formed by the inducible nitric oxide synthase, and the alkyl peroxides that result from exposure of unsaturated lipids to oxidative stress. Analysis of the genes induced in isoniazid-resistant M. tuberculosis indicates that one of the mechanisms used by the organism to compensate for loss of the KatG antioxidant activity is to up-regulate the ahpC gene product, which codes for a non-hemoprotein alkylhydroperoxidase (15-19). Incubation of M. tuberculosis expressing elevated levels of the alkylhydroperoxidase with isoniazid has shown that the drug is not activated by this enzyme (15). AhpC thus differs from KatG in its interactions with isoniazid.

Very little is known about the M. tuberculosis alkylhydroperoxidase in terms of its structure or catalytic mechanism, as the protein has not yet been purified and investigated. The most studied member of the alkylhydroperoxidase family is the enzyme from Salmonella typhimurium (20, 21). This protein contains two cysteine sulfhydryls that catalyze the reduction of peroxides to the corresponding alcohols and water with concomitant oxidation of the cysteine residues to give a disulfide bond (Scheme 1). A sulfenic acid derivative of one of the two sulfhydryl groups is thought to be a transient intermediate in the formation of the disulfide link (22, 23). The catalytic cycle is completed by reduction of the disulfide bond using AhpF, a flavoprotein reductase (Scheme 2) (21). The protein required to reduce the AhpC disulfide bond is not the same in all organisms. Thus, in yeast the alkylhydroperoxidase is coupled to thioredoxin and thioredoxin reductase rather than to an AhpF-like protein (24).

View larger version (6K): [in this window] [in a new window]   Scheme 1.

View larger version (9K): [in this window] [in a new window]   Scheme 2.

The M. tuberculosis genome has been sequenced in its entirety (25). No ahpF gene appears to be present, although a number of flavoproteins of unknown function are found in the genome. In S. typhimurium the ahpF gene is found immediately downstream of the ahpC gene. The corresponding position in the M. tuberculosis genome is occupied by a gene that, because of its position in the sequence, was named ahpD. AhpD exhibits no sequence similarity with either ahpC or ahpF and its function is unknown. We report here the first expression and purification of the M. tuberculosis AhpC and AhpD proteins and their initial structural and catalytic characterization.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials

Oligonucleotide synthesis and DNA sequencing were performed by the Biomolecular Resource Center of the University of California, San Francisco. A Perkin-Elmer 480 DNA thermal cycler was used for PCR¹ experiments. The pGem-T and pET23a plasmids were from Novagen (Madison, WI). Plasmid pACYC was from New England Biolabs (Beverly, MA). Restriction enzymes and Vent DNA polymerase were purchased from New England Biolabs and Promega (Madison, WI). Plasmids were purified using the Qiagen (Chatsworth, CA) Quick-Prep kit. E. coli strain BL21(DE3) was from Novagen and strain DH5 from Life Technologies, Inc. (Gaithersburg, MD). Q-Sepharose Fast Flow was from Amersham Pharmacia Biotech and the Ni-NTA-agarose resin was from Qiagen. PEI was from Research Biotechnologies, Inc. (Natick, MA). LB medium was from Life Technologies, Inc. All other chemical reagents were purchased from Sigma. An Amersham Pharmacia Biotech Sephadex 200 column, connected to an Amersham Pharmacia Biotech PCC-500 FPLC system, was used to determine the native aggregation state of each protein. A Hewlett-Packard HP-8452 UV-visible spectrophotometer was used for all spectroscopic measurements. The plasmid encoding for the S. typhimurium AhpF (pAF1) was generously provided by Leslie B. Poole (26). The plasmids encoding the M. tuberculosis thioredoxin and thioredoxin reductase were a gift from Brigitte Wieles (27, 28). All expression plasmids were introduced into competent BL21(DE3) E. coli.

Construction of the AhpC Expression Vector

The ahpC gene was generously provided by Clifton E. Barry as a 1.3-kilobase NotI-PstI fragment in pMH91 (16). The open reading frame for the ahpC gene was amplified by PCR with the following primers (forward: 5'-CGCTAGGTACCATATGCCACTGCTAACCATTGGC-3'; reverse: 5'-TCTAGAGGATCCTTAGGCCGAAGCCTTGAGGAG-3'). The primers coded for an NdeI restriction site at the ATG codon and a BamHI site following the stop (TAA) codon. The reaction contained 50 ng of pMH91, 50 pmol each of the primers, 1 mM dNTPs, and 10 units of Vent DNA polymerase (NEB) in a final volume of 50 µl of 20 mM Tris-HCl (pH 8.8), 10 mM KCl, 10 mM (NH₄)₂SO₄, 2 mM MgSO₄, and 0.1% Triton X-100. The annealing and extension cycles were as follows: 94 °C for 10 min (1 cycle), 94 °C for 1 min, 50 °C for 1 min, 72 °C for 1 min (10 cycles), 94 °C for 1 min, 65 °C for 1 min, 72 °C for 1 min (20 cycles), and 72 °C for 10 min (1 cycle). Following gel purification of the amplified product the 0.59-kilobase gene was digested with NdeI and BamHI, ligated into pGem-T, and subcloned into pET 23a. This plasmid was called pEahpC.

A second vector was constructed in which a 6-His tag was added to the 3' end of the ahpC gene using the primer 5'-TCTAGACTCGAGGGCCGAAGCCTTGAGGAG-3'. This primer encoded for an XhoI site that removed the stop codon. The PCR conditions were identical to the above reaction. This plasmid was called pEahpC-histag.

Isolation of AhpD from M. tuberculosis Genomic DNA and Construction of the AhpD Expression Vector

The open reading frame for the ahpD gene was amplified by PCR from M. tuberculosis genomic DNA (provided by Clifton E. Barry) using the following primers (forward: 5'-GATCTGGTTGCCCGGGAACATATGAGTATAGAAAAGCTC-3'; reverse: 5'-GGCGTCATGGCGTCGACACACTTAGCTTGGGCTTAGTGCCTCGGTTGTGCC-3').

The primers coded for an NdeI restriction site at the ATG codon and a SalI site following the stop (TAA) codon. The reaction contained 50 ng of M. tuberculosis genomic DNA, 50 pmol each of the primers, 2 mM dNTPs, and 10 units of Vent DNA polymerase in a final volume of 100 µl of 20 mM Tris-HCl (pH 8.8), 10 mM KCl, 10 mM (NH₄)₂SO₄, 2% dimethyl sulfoxide, and 0.1% Triton X-100. The annealing and extension cycles were as follows: 90 °C for 10 min (1 cycle), 90 °C for 1 min, 72 °C for 1 min, 60 °C for 1 min (30 cycles), and 72 °C for 10 min (1 cycle). Following gel purification of the amplified product the 0.55-kilobase gene was digested with NdeI and SalI. The ahpD gene was then inserted into pACYC using the available NdeI and SalI sites. This plasmid was called pACahpD.

AhpC and AhpD Mutagenesis

The single cysteine to serine mutants of the 3 cysteine residues of AhpC and 2 cysteine residues of AhpD were made using the QuikChange Site-directed Mutagenesis Kit from Stratagene (La Jolla, CA). The PCR conditions were as follows: 50 ng of pEahpC (without his-tag) or pACahpD, 125 ng of each primer, 50 µM each dNTP, 2.5 units of Pfu polymerase, and 5 µl of 10× Pfu buffer in a total volume of 50 µl. The cycling parameters were: 95 °C for 30 s (1 cycle), 95 °C for 30 s, 55 °C for 1 min, and 68 °C for 8.4 min (16 cycles). After amplification the PCR mixture was incubated with 20 units of DpnI for 1 h at 37 °C and then 1 µl was used to transform 50 µl of DH5 cells. The mutations where confirmed by DNA sequencing.

The primers used for the mutagenesis were: C61S, forward primer: 5'-TTCACGTTCGTGTCCCCTACCGAG-3', reverse primer: 5'-CTCGGTAGGGGACACGAACGTGAA-3'; C174S, forward primer: 5'-GACGAGCTGTCCGCATGCAACTGG-3', reverse primer: 5'-CCAGTTGCATGCGGACAGCACGTC-3'; C176S, forward primer: 5'-GAGCTGTGCGCATCCAACTGGCGC-3', reverse primer: 5'-GCGCCAGTTGGATGCGCACAGCAC-3'.

In the case of ahpD mutagenesis, the C129S primers were: forward: 5'-GCGATCAACGGGTCCTCGCATTGCCTC-3', reverse: 5'-GAGGCAATGCGAGGACCC-GTTGATCGC-3'. The primers for the C132S mutations were: forward: 5'-GGGTGCTCGCATTCCCTCGTCGCCCAC-3', reverse: 5'-GTGGGCGACGAGGGAATGCGAGCACCC-3'. The underlined codons represent the cysteine mutation, with the boldface letters indicating the nucleotide changed to facilitate the mutation.

Bacterial Cell Growth

Bacterial growth was carried out at 37 °C in LB medium containing 100 µg/ml ampicillin (for pEahpC) or 50 µg/ml chloramphenicol (for pACahpD and pAF1). One colony was used to inoculate 50 ml of LB medium containing the appropriate antibiotic, and the culture was incubated for 10 h. The culture was used to inoculate a 1-liter culture of LB containing the appropriate antibiotic at a ratio of 10 ml/liter. When the A₆₀₀ value of the culture reached 0.7-1.0, isopropyl--D-thiogalactopyranoside was added to a final concentration of 0.5 mM for pEahpC and pAF1, and 0.2 mM for pACahpD. Incubation was continued for 3-3.5 h at 37 °C for pEahpC and pAF1, and 20 °C for pACahpD. Cells were harvested by centrifugation at 5000 × g for 45 min, 4 °C, and stored at 20 °C overnight.

Protein Purification

AhpC-- Cells were suspended in a 4-fold excess (with respect to the initial weight of cells) of lysis buffer (50 mM KP_i, pH 7.0, 1.0 mM DTT, 1.0 mM EDTA, 44 µg/ml phenylmethanesulfonyl fluoride, 2 µg/ml leupeptin, 2 µg/ml pepstatin, 5% glycerol, and 5% lysozyme). The solution was stirred 60 min at 4 °C. The cells were then sonicated using a Branson sonicator with 4 bursts of 30 s at 45 W with 30 s intervals. The cell debris was precipitated by centrifugation at 27,000 × g for 60 min at 4 °C. The supernatant was removed and PEI was added to a final concentration of 0.005%. The solution was stirred for 15 min at 4 °C and then centrifuged at 27,000 × g for 15 min at 4 °C. The PEI supernatant was then loaded onto the Q-Sepharose Fast Flow column (1.5 × 12 cm) equilibrated in 50 mM KP_i, pH 7.0, 1.0 mM DTT, 1.0 mM EDTA, and 5% glycerol. After loading, the resin was washed with the same buffer for 10 column volumes, followed by a wash with buffer containing 0.2 M KCl for 10 column volumes. The protein was eluted with a gradient from 0.2 to 0.4 M KCl in 50 mM KP_i, pH 7.0, 1.0 mM DTT, 1.0 mM EDTA, 5% glycerol. The protein eluted at approximately 0.25 M KCl. Fractions containing pure AhpC, as assessed by denaturing 20% polyacrylamide gels, were pooled, concentrated in an Amicon ultrafiltration cell using a YM10 membrane, and dialyzed against 20 mM KP_i, pH 7.0, 50 mM KCl, 0.1 mM EDTA, and 5% glycerol (3 × 2 liter). The protein was stored at 70 °C until used.

AhpD-- Cells were suspended in a 6-fold excess (with respect to the initial weight of cells) of lysis buffer. The solution was stirred 60 min at 4 °C. A PEI supernatant was prepared and loaded onto the Q-Sepharose column as described for AhpC. After loading, the resin was washed with the same buffer for 20 column volumes. The protein was eluted with a gradient from 0 to 0.1 M KCl in 50 mM KP_i, pH 7.0, 1.0 mM DTT, 1.0 mM EDTA, 5% glycerol. The protein eluted at approximately 0.03 M KCl. Fractions containing pure AhpD, as assessed by denaturing 20% polyacrylamide gels, were pooled, concentrated, and dialyzed against 50 mM KP_i, pH 7.0, 100 mM KCl, 0.1 mM EDTA, and 5% glycerol (3 × 2 liter). The protein was stored at 70 °C until used.

AhpF-- This enzyme was purified according to the protocol of Poole and Ellis with slight modifications (30). Nucleic acids were removed with 0.005% PEI, and the ammonium sulfate precipitation steps were omitted.

Thioredoxin and Thioredoxin Reductase-- These enzymes were expressed and purified according to the protocol of Zhang et al. (31) with a poly-histidine tag on each protein to facilitate purification.

Gel Permeation Chromatography

Protein size determination was performed using an Amersham Pharmacia Biotech Sephadex 200 FPLC column. The column was equilibrated in 50 mM KP_i, pH 7.0, 0.1 mM EDTA, 100 mM KCl, and 5% glycerol at a flow rate of 0.5 ml/min. Approximately 0.5 mg of protein was injected on the column. Protein elution was monitored at a wavelength of 280 nm. Data was collected and processed using the Virtual Bench (National Instruments) software.

AhpF-dependent Activity Assays

Rates of hydroperoxide reduction were determined anaerobically in a coupled assay with AhpF, monitoring the decrease in absorbance at 340 nm due to NADH oxidation. The assays typically contained 2 mM hydroperoxide substrate in 100 mM KP_i, pH 7.0, 1 mM EDTA, 0.25 mM NADH, and 20 µM either AhpC or AhpD, and 10 µM AhpF. Background NADH oxidation due to AhpF was monitored, then the hydroperoxide substrate was added and the enzymatic rate was observed. For steady-state kinetic assays, the substrate concentration was varied, and data was fit to the equation: v = V_max [S]/(K_m + [S]).

DTT-dependent Activity Assays

The rate of DTT oxidation catalyzed by AhpC or AhpD in the presence of the peroxide substrate was measured by monitoring the change in absorbance at 310 nm due to formation of the DTT disulfide (32). A Cary 1E spectrophotometer was used to obtain this data. The buffer and the water used for the assays were Chelex-pretreated as recommended by the supplier. Typical conditions for the assays were: 100 mM KP_i, pH 7.0, 1 mM EDTA, and 10 mM DTT in a 1-ml quartz cuvette at 25 °C (maintained with a circulating water bath). The initial rate of DTT oxidation was obtained by calculating the slope over the first 11 s after addition and mixing of the peroxide. The initial rates were corrected for the background oxidation of DTT by the peroxides in the absence of the enzyme.

HPLC Analysis of Cumene Hydroperoxide Products from AhpC and AhpD

A 50-µl solution containing an equimolar mixture (0.23 µmol) of cumene hydroperoxide and either AhpC or AhpD was equilibrated in KP_i, pH 7.0, buffer for 120 min. The reaction was then quenched by addition of an equal volume (50 µl) of a solution of 6% acetic acid in acetonitrile. The protein that precipitated was removed by centrifugation, and the supernatant was injected onto a Hewlett-Packard 1090 HPLC system equipped with an Axxiom ODS (4.6 × 250 mm) reverse-phase HPLC column. The products were separated using 20% acetonitrile and 80% water at a flow rate of 1 ml/min. The detector was set at 260 nm. Control reactions were performed in the same buffer (50 mM KP_i, pH 7.0, 100 mM KCl, 0.1 mM EDTA, and 5% glycerol) but without the enzyme. Product peaks were identified by comparison with authentic standards under identical elution conditions.

Synthesis of Hydroperoxide Substrates

Hydroperoxides were generated using the Schenck reaction of singlet oxygen with unactivated olefins bearing allylic hydrogens (33, 34). In general, a solution of the olefin (3 mmol) and the sensitizer tetraphenylporphine (14.8 mg, 10 mM) in 25 ml of CCl₄ was irradiated with a Sylvania 750 W lamp at 0 °C. A slow stream of oxygen was bubbled through the stirred solution for 5-6 h. The solvent was then removed in vacuo, and the hydroperoxide products were purified by column chromatography followed by crystallization or distillation, as appropriate. The products were identified by comparison of their physical properties and spectra with those in the literature and/or by reduction to the corresponding alcohols with triphenylphosphine (not shown). The structures of the hydroperoxide products are shown in Fig. 1.

View larger version (10K): [in this window] [in a new window]   Fig. 1.   Structures of the synthetic alkylhydroperoxides and the structure numbers used to identify them in the text and Table I.

5-Hydroperoxy-6-methyl-6-hepten-2-one (1) and 6-hydroperoxy-6-methyl-4-hepten-2-one (2) were obtained from the oxidation of 6-methyl-5-hepten-2-one in a 2:1 ratio, respectively, and in 96% yield: colorless oil, bp 68-71 (~5 mm Hg), R_f = 0.59 (silica gel, 1:10 ethyl acetate:hexane); IR (KBr) 3439 (-OOH) and 1678 cm¹ (C=O); ¹H NMR (400 MHz): (1) 1.36 (s, 3H, CH₃), 1.70 (m, 2H, CH₂), 1.73 (s, 3H, CH₃), 1.92 (m, 2H, CH₂), 4.37 (d, 1H, ³J_HH = 11.1 Hz, CHOOH), 4.91 (s, 1H, =CH₂), and 4.94 ppm (s, 1H, =CH₂); (2) 1.24 (s, 3H, CH₃), 1.33 (s, 3H, CH₃), 2.22 (s, 3H, CH₃), 3.28 (d, 2H, ³J_HH = 6.5 Hz, CH₂), 6.30 (d, 1H, ³J_HH = 16.0 Hz, CH), and 6.58 ppm (dd, 1H, ³J_HH = 16.0, and ³J_HH = 6.6 Hz, CH).

Trans-pinocarveylhydroperoxide (3) was obtained in 89% yield by irradiation of -pinene as described by Schenck et al. (35) and subsequently Capdeville and Maumy (36): yellow oil, R_F = 0.59 (benzene); IR (KBR) 3395 cm¹ (OOH); EIMS m/z 168; ¹H NMR (400 MHz): 0.66 (s, 3H, CH₃), 1.26 (s, 3H, CH₃), 1.48 (d, 1H, ³J_HH = 9.8 Hz, CH), 1.93 (m, 2H, CH₂), 2.23 (m, 1H, CH₂), 2.32 (m, 1H, CH₂), 2.47 (m, 1H, CH), 4.61 (d, 1H, ³J_HH = 8.2 Hz, CHOOH), 4.99 (s, 1H, =CH₂), 5.12 (s, 1H, =CH₂), and 7.97 ppm (s, 1H, OOH). 7-Hydroperoxy-3-hydroxycholest-6-ene (6) resulted from the photosenzitized oxidation of cholesterol as reported by Beckwith et al. (37). It was obtained in 32% yield as a white solid (recrystallization from benzene) m.p. 157-158 °C (literature 152-153 °C (37), 154-156.5 °C (38)), R_F = 0.69 (silica gel, 1:1 benzene:ethyl acetate): IR (KBr) 3335 cm¹ (-OOH); ¹H NMR (400 MHz): 0.65 (s, 3H, CH₃), 0.85 (d, 3H, ³J_HH = 6.6 Hz, CH₃), 0.90 (d, 3H, ³J_HH = 6.3 Hz, CH₃), 0.98 (s, 3H, CH₃), 1.11-1.39 (m, 8H), 1.48 (m, 6H), 1.54 (m, 4H), 1.57 (m, 2H), 1.86 (m, 3H), 1.97 (d, 2H, ³J_HH = 12.6 Hz, CH₂), 2.37 (m, 2H), 3.61 (m, 1H, CHOH), 4.15 (t, 1H, ³J_HH = 4.0 Hz, CHOOH), and 5.71 ppm (dd, 1H, ³J_HH = 4.8 and ⁴J_HH = 1.6 Hz, =CH-).

Trans-9-hydroperoxyoctadec-10-enoic acid (4) and trans-10-hydroperoxyoctadec-8-enoic acid (5) were obtained in a ratio of 1:0.8, respectively, from oleic acid as reported by Porter and Wujek (39): yellow oil obtained in 60% yield, R_F = 0.33 (silica gel, ethyl acetate). An additional purification step was required for these compounds using C18 reverse phase HPLC, with a 30-78% acetonitrile gradient over 30 min with a flow rate of 1 ml/min. ¹H NMR (400 MHz): (4) 0.89 (m, 3H, CH₃), 1.28 (broad peak, 18H), 1.63 (m, 2H, CH₂), 2.08 (m, 2H, CH₂), 2.35 (m, 2H, CH₂), 4.27 (q, 1H, ³J_HH = 7.8 Hz, CHOOH), 5.36 (dd, 1H, ³J_HH = 8.0 and ³J_HH = 15.0 Hz, CH), and 5.76 ppm (dt, 1H, ³J_HH = 6.6 and ³J_HH = 15.1 Hz, CH) (4 + 5). ¹³C NMR: (4)  = 179.9 (COOH), ₁₁ = 137.2, ₁₀ = 128.4, ₉ = 87.1; (5)  = 179.5 (COOH), ₈ = 136.7, ₉ = 128.8, ₁₀ = 87.1.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Overexpression and Purification of AhpC and AhpD-- Both proteins were independently overexpressed in E. coli strain BL21(DE3), using a pET vector for AhpC and a pACYC vector for AhpD. Lower yields or proteolyzed proteins were obtained when a pET23a or pUC19 vector was used, or when efforts were made to express the protein in the DH5 or XL1-BLUE strains of E. coli. Large quantities of AhpD (>50%) were lost as insoluble inclusion bodies when isopropyl--D-thiogalactopyranoside-induced expression was performed above 25 °C. Expressions were therefore carried out at a lower temperature to minimize this problem, although some loss of protein still occurred. Attempts to refold the precipitated protein were unsuccessful. In contrast, AhpC was produced in high amounts, and no inclusion bodies were observed even when the protein was expressed at 37 °C. A single ion-exchange chromatographic protocol was sufficient to purify the two enzymes. Both proteins were judged to be >95% pure by denaturing SDS-polyacrylamide gel electrophoresis (Fig. 2). Due to the absence of a strong chromophore in either protein, enzyme concentrations were determined from the molar absorption coefficients using the method of Pace et al. (40). For AhpC, the calculated  (280) is 25,170 M¹ cm¹, and for AhpD, the calculated  (280) is 15,720 M¹ cm¹.

View larger version (74K): [in this window] [in a new window]   Fig. 2.   Scan of a 12% polyacrylamide gel containing AhpC and AhpD. The proteins are: lane 1, soluble fraction of AhpD-containing cell lysate; lane 2, purified AhpD; lane 3, purified AhpC; lane 4, soluble fraction of AhpC-containing cell lysate. Molecular weight standards are on the far right. Each lane contained approximately 2-4 µg of enzyme.

Native Structure of AhpC and AhpD-- Size exclusion chromatography indicated that AhpC was primarily present as a higher-order oligomer (10-12 mer) (Fig. 3). A minor peak at approximately 15 min in the AhpC chromatogram indicated that a small amount of the dimer was also present. This dimer could represent either modified protein that is unable to oligomerize or the fraction of the protein present as the dimer in a dimer-oligomer equilibrium. A dimeric rather than oligomeric species was observed for AhpD (Fig. 3). No monomer was observed with either protein. The elution profiles of AhpC and AhpD were unchanged in the presence of reductant (1 mM DTT).

View larger version (26K): [in this window] [in a new window]   Fig. 3.   The elution profiles of AhpC (···), AhpD (- - -), and a mixture of the two proteins () from the size exclusion column. The column was equilibrated in 50 mM KPi (pH 7.0), 100 mM KCl, 0.1 mM EDTA, and 5% glycerol, at a flow rate of 0.5 ml/min. Inset, calibration of the size exclusion column using carbonic anhydrase (15.7 ml), bovine serum albumin (14.0 ml), alcohol dehydrogenase (12.0 ml), and apoferritin (10.5 ml). The void volume was estimated using dextran blue (8.0 ml), and the total volume was estimated using flavin mononucleotide (19.5 ml).

Attempts to identify an interaction between AhpC and AhpD were unsuccessful. Native gel electrophoresis, gel permeation chromatography (Fig. 3), and kinetic studies (see below) provided no evidence for any change in the protein oligomeric state or activity when the proteins were mixed. AhpC and AhpD appear to function as completely independent proteins.

The Catalytic Activities of AhpC and AhpD-- To identify the products formed by AhpC and AhpD, a high concentration of each of the two enzymes was incubated with cumene hydroperoxide under single turnover conditions and the products formed were determined by HPLC analysis. Under these conditions, cumene hydroperoxide was converted by both AhpC (Fig. 4) and AhpD (not shown) exclusively to the cumyl alcohol.

View larger version (19K): [in this window] [in a new window]   Fig. 4.   Product formation from cumene hydroperoxide in single turnover studies with AhpC in the absence of either AhpF/NADH or DTT. Product profile in the presence () and absence (- - -) of AhpC. The peaks at approximately 21.5, 24.2, and 27.0 min correspond, respectively, to cumyl alcohol, acetophenone, and cumene hydroperoxide.

Normal catalytic turnover of AhpC and AhpD requires reduction of the disulfide bond presumed to be formed in the reaction with the peroxide. In the absence of information on the endogenous electron donor that reduces the disulfide bond, enzymatic activity was first assessed using an NADH-coupled assay in which S. typhimurium AhpF was used to reduce the oxidized forms of AhpC and AhpD. Fig. 5 shows a representative trace at 340 nm for the AhpF-dependent turnover of AhpC and AhpD under anaerobic conditions. All the assay components except for the hydroperoxide substrate were present at time 0 in order to assess the background NADH consumption due to the reduction by AhpF of traces of oxygen or other alternative electron acceptors in the medium. The substrate was then added and the resulting NADH loss observed. The background NADH oxidation was consistently less than 10% of the total observed activity. Incubation of AhpF and NADH alone with cumene hydroperoxide followed by HPLC analysis of the products shows that the hydroperoxide is reduced by this flavoprotein to acetophenone rather than cumyl alcohol (not shown). Acetophenone is presumably formed by reaction of the hydroperoxide with trace metals in the medium that are reduced by electron transfer from the AhpF flavin group, as found previously with the thioredoxin and thioredoxin reductase from M. tuberculosis (31).

View larger version (19K): [in this window] [in a new window]   Fig. 5.   Activity assay following addition of NADH (upper panel) or DTT (lower panel). Upper panel, substrate was added to the anaerobic incubation ~150 s (indicated by arrow) after the background consumption of NADH due to AhpF alone was determined. Trace 1 is the activity observed for AhpC, trace 2 is the activity observed for AhpD. Lower panel, the solid line indicates DTT oxidation in the absence, and the dotted line oxidation in the presence, of AhpC. Conditions: 2 mM tert-butylhydroperoxide, 1 mM EDTA, 20 µM either AhpC or AhpD, and either 250 µM NADH and 10 µM AhpF (upper panel) or 10 mM DTT (lower panel) in 100 mM Kpi (pH 7.0), 25 °C, in a total volume of 1 ml.

The AhpF-supported catalytic activities of AhpC and AhpD are shown in Fig. 6 (columns 2 and 3). Addition of both proteins to the same assay resulted in an additive rather than synergistic effect on the measured activity (Fig. 6, column 4). This finding confirms that AhpC and AhpD act independently as hydroperoxidases. As previously reported, some residual hydroperoxidase activity is observed in the presence of S. typhimurium AhpF alone (Fig. 6, column 1) (30). As noted above, HPLC analysis of the product formed from cumene hydroperoxide indicates that this residual AhpF activity involves a homolytic rather than heterolytic cleavage of the hydroperoxide. The M. tuberculosis thioredoxin and thioredoxin reductase were also tested as possible redox partners for AhpC and AhpD because the corresponding proteins are the natural partners for AhpC in yeast (24). However, only background activity was observed with AhpC or AhpD when incubated with NADH and either thioredoxin reductase alone or thioredoxin and thioredoxin reductase (Fig. 6, columns 5 and 6). As a further control, it was shown that addition of the hydroperoxide alone to NADH under our assay conditions did not result in NADH consumption (not shown).

View larger version (11K): [in this window] [in a new window]   Fig. 6.   AhpC and AhpD activities in the presence of various components. Each protein was present (+) or absent (). Conditions: 2 mM tert-butyl hydroperoxide, 0.25 mM NADH, and 5 µM AhpC, 5 µl AhpD, and 5 µM AhpF were incubated in 100 mM KPi (pH 7.0), 25 °C, in a total volume of 1 ml. Anaerobic background NADH consumption was monitored before enzymatic turnover was initiated by addition of the hydroperoxide substrate. Results are the average of at least three trials. Trx corresponds to thioredoxin and TR to thioredoxin reductase. The values are the mean of three experiments ± S.D.

The enzymatic activity increased as the protein concentration was increased and was saturable (Fig. 7). Increasing the concentration of AhpC at a fixed AhpF concentration showed that the maximum activity was achieved at something above a 3:1 AhpC:AhpF ratio. The AhpD activity was only saturated at something in excess of a 6:1 AhpD:AhpF ratio. Increasing the concentration of AhpF at a fixed concentration (10 µM) of AhpC gave a similar activity profile to that shown in Fig. 7 (not shown). These results clearly show that both AhpC and AhpD can form a catalytically competent hydroperoxide reductase system with AhpF and NADH.

View larger version (12K): [in this window] [in a new window]   Fig. 7.   AhpC (upper panel) and AhpD (lower panel) activity as a function of protein concentration. Conditions: 10 µM AhpF, 1 mM tert-butylhydroperoxide, 250 µM NADH, 1 mM EDTA, and varying concentrations of AhpC or AhpD in 100 mM KPi (pH 7.0), 25 °C, in a total volume of 1 ml.

The AhpF-supported activity of the AhpC system is pH-dependent (Fig. 8). Using KP_i buffer over the pH range from 4.5 to 10.0, the highest activity was observed at pH 8.0 to 8.5. Below pH 8.0 and above pH 9, there was a gradual decrease in the AhpC activity. In the case of AhpD, the activity was constant between pH 5 and 8, slightly increased between pH 8 and 9, and then markedly decreased. These results contrast with those for the S. typhimurium AhpC, for which a bell-shaped pH dependence with a maximum at pH 7.0 was observed (30). The decreases in the activity presumably stem from the protonation (at lower pH) or deprotonation (at higher pH) of critical active site residues, although the protonation or deprotonation could impact either the catalytic process itself or the protein-protein interactions required for efficient electron transfer from AhpF to AhpC or AhpD. However, the data appear to reflect changes in the AhpC and AhpD components of the reactions as the activities increase linearly with the concentration of AhpC or AhpD at pH values (6.0, 7.0, 8.0, and 9.0) throughout the range explored in the pH profiles.

View larger version (11K): [in this window] [in a new window]   Fig. 8.   AhpC activity () and AhpD activity () as a function of pH. Conditions: 10 µM AhpF, 5 mM tert-butylhydroperoxide, 250 µM NADH, 1 mM EDTA, and 20 µM either AhpC or AhpD in 100 mM KPi, 25 °C. Background NADH consumption was monitored before enzymatic turnover was initiated by addition of the hydroperoxide substrate. The values are the mean of two independent measurements. The error bars indicate the range of the two values.

DTT Oxidation Assays-- As AhpF is not the natural partner for AhpC and AhpD, we have evaluated their catalytic activities using a second system in which DTT is used to reduce the putative disulfide bond formed in their reactions with peroxides. In this assay, the peroxidase activity of AhpC and AhpD was evaluated using tert-butyl hydroperoxide as the substrate. The oxidation of DTT in the presence of tert-butyl hydroperoxide is catalyzed by AhpC (Fig. 5) in a reaction that follows Michaelis-Menten kinetics (Fig. 9). A linear dependence was observed for the oxidation of DTT as a function of AhpC concentration, in agreement with a catalytic role for AhpC in the DTT oxidation process (Fig. 9, inset). An analysis of the pH dependence of the reaction of AhpC exhibited an optimum at a pH of approximately 7.5 (not shown). A similar study of the activity of AhpD was not possible because the enzyme activity with this reducing agent is so low that the only pH dependence that was observed was a continuous increase in activity above pH 8.0 due to deprotonation of the DTT (not shown).

View larger version (20K): [in this window] [in a new window]   Fig. 9.   DTT oxidation as a function of the tert-butylhydroperoxide concentration. The incubations, carried out at 25 °C, contained 5 µM AhpC, 100 mM KPi (pH 7.0), 1 mM EDTA, and 10 mM DTT in addition to the indicated tert-butylhydroperoxide concentration. The inset shows the rate of DTT oxidation as a function of the AhpC concentration. Each point in the plot represents the average of four measurements ± S.E of the mean.

Substrate Specificity of AhpD and AhpC-- Alkylhydroperoxides of varying carbon length and functionality were synthesized and tested as substrates for AhpC and AhpD (Table I). The steady-state kinetic parameters for the alkylhydroperoxides were determined at pH 7.0, 25 °C, using the NADH-coupled assay. Cholesterol hydroperoxide, with K_m values of 91 and 132 µM for AhpC and AhpD, respectively, is the best substrate for AhpC but only the second best substrate for AhpD. The best substrate for AhpD is cumene hydroperoxide with a K_m of 50 µM. For each enzyme, approximately a 10-fold difference exists between the lowest and highest k_cat values (Table I). Although lipophilicity appears to contribute to the affinity of the substrate for both AhpC and AhpD, the number of carbons in the substrate is not the only parameter that determines its activity because oleic acid hydroperoxide (C₁₈ chain) is a poor substrate for both enzymes. In accord with this finding, plots of the log K_m values of the substrates as a function of their hydrophobicity showed that the two parameters are not simply correlated (not shown). These studies establish, however, that although the two enzymes differ somewhat in their specificity both are able to catalyze the reduction of a variety of substrate structures. It is of interest that H₂O₂ is not the best substrate for either enzyme.

AhpC and AhpD Active Site Mutants-- There are three cysteine residues in AhpC (Cys-61, Cys-174, and Cys-176) and two in AhpD (Cys-129 and Cys-132). To ascertain the roles of the cysteine residues in the two enzymes, each cysteine residue was mutated to a serine and the resulting protein was expressed, purified, and tested for catalytic activity. The three AhpC mutant proteins were expressed at high levels, but the two AhpD mutant proteins were expressed at significantly lower levels than the wild-type enzyme. Furthermore, the AhpD mutants were subject to proteolytic degradation and the mutant proteins purified by the same protocol as the wild-type enzyme were contaminated with small amounts of these proteolytic fragments (not shown). In all cases, the catalytic activities of the mutated proteins were considerably lower than those of the wild-type proteins whether activity was evaluated in the AhpF- or DTT-dependent assay system. Mutation of each of the three cysteines in AhpC decreased the AhpF dependent activity with respect to cumene hydroperoxide to 12-25% of that of the wild-type enzyme, but for none of them was complete loss of activity observed (Fig. 10). The DTT dependent assay with tert-butylhydroperoxide as the substrate, which gives a much higher absolute level of activity, indicated that the C61S mutant is essentially inactive whereas the C174S and C176S mutants retain ~10% of the wild-type activity. The results from the two AhpC assays are thus qualitatively similar, although the C61S mutant is completely inactive in the DTT assay but appears to retain a small degree of activity in the AhpF assay. These results suggest that all three cysteine residues are involved in the function of AhpC, either in a structural or catalytic role, with Cys-61 as the critical catalytic residue.

View larger version (34K): [in this window] [in a new window]   Fig. 10.   Activity of AhpC and its mutant proteins supported by either AhpF or DTT. Each data point represents the average of at least three trials. Upper panel, AhpF-dependent activity of AhpC assayed as follows: 2 mM cumene hydroperoxide, 250 µM NADH, 1 mM EDTA, 20 µM AhpC, and 10 µM AhpF incubated in 100 mM KPi (pH 7.0), 25 °C, in a total volume of 1 ml. Lower panel, DTT-dependent activity of AhpC assayed under the following conditions: 2 mM tert-butylhydroperoxide, 10 mM DTT, 1 mM EDTA, and 20 µM AhpC in 100 mM KPi (pH 7.0), 25 °C. The values are the mean of three to five measurements ± S.D.

In the AhpF-dependent assay with cumene hydroperoxide as substrate, no activity (i.e. <3%) was observed with the AhpD C132S mutant and only 5% of the wild-type activity with the C129S mutant (Fig. 11). In the DTT assay, the C132S mutation also caused the greatest loss of activity. However, the total AhpD-dependent peroxidase activity in the DTT assay was low and difficult to quantitate due to the background DTT activity, so that the DTT data is less reliable for AhpD than the AhpF-dependent data. The results indicate that Cys-132 in AhpD is critical for function of the enzyme, at least when catalytic turnover is coupled to AhpF, but that Cys-129 is also catalytically important.

View larger version (40K): [in this window] [in a new window]   Fig. 11.   Activity of AhpD and its mutant proteins supported by either AhpF or DTT. Each data point represents the average of three independent determinations. Upper panel, AhpF-dependent activity of AhpD assayed as follows: 2 mM cumene hydroperoxide, 250 µM NADH, 1 mM EDTA, 20 µM AhpD, and 10 µM AhpF incubated in 100 mM KPi (pH 7.0), 25 °C, in a total volume of 1 ml. Lower panel, DTT dependent activity of AhpD assayed under the following conditions: 2 mM tert-butylhydroperoxide, 10 mM DTT, 1 mM EDTA, and 20 µM AhpD in 100 mM KPi (pH 7.0), 25 °C. The values are the mean of three to five measurements ± S.D.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

We report here the first purification, and the initial structural and catalytic characterization, of the heterologously expressed M. tuberculosis AhpC and AhpD proteins. AhpC has sequence homology to the 2 cysteine containing AhpC proteins from other organisms. It has a higher similarity to the Gram-positive (Mycobacterium bovis, Mycobacterium smegmatis, C. diphtheriae, and S. viridosporus) than to the Gram-negative (E. coli, Salmonella typhimurium, Bacillus subtilis, and Staphylococcus aureus) AhpC proteins. In contrast, AhpD has no sequence similarity to AhpC or to any other AhpC-like protein. The ahpD gene has been detected in other mycobacteria and in Streptomyces viridosporus. In M. tuberculosis, the ahpD gene is found immediately downstream from the ahpC gene, in the position occupied by the ahpF gene in S. typhimurium (Scheme 3). No ahpF gene has so far been found in the Gram-positive genomes.

View larger version (7K): [in this window] [in a new window]   Scheme 3.

AhpC and AhpD are overexpressed heterologously in E. coli in soluble form and can be purified in high yield to near homogeneity using one ion exchange column (Fig. 2). Size-exclusion chromatography indicates that AhpC is found in solution as a higher order aggregate, possibly in equilibrium with a small amount of the dimer (Fig. 3). The formation of higher order aggregates was recently reported for the AhpC from Amphibacillus xylanus (41). As the oligomeric structure of the M. tuberculosis enzyme was not disrupted by the addition of reductant, it does not appear that the aggregates are held together by disulfide bonds. In contrast, AhpD is a dimeric species under both reducing and oxidizing conditions (Fig. 3). The AhpD dimer thus also appears to be held together by protein-protein interactions rather than by disulfide bonds. The data on these M. tuberculosis proteins does not preclude the presence of disulfide bonds, but simply suggests that the oligomeric states of the proteins do not depend critically on such bonds.

The AhpC from M. tuberculosis is exceptional in that it contains three (Cys-61, Cys-174, and Cys-176) rather than the one or two catalytic cysteine residues usually found in AhpC proteins. Cys-61 corresponds to the residue that is highly conserved in all AhpC proteins, in accord with the finding in the DTT assay system that the C61S mutant was inactive (Fig. 10). The second conserved cysteine residue could be either Cys-174 or Cys-176. However, both of these cysteine residues appear to be important for the catalytic function of the enzyme although the role for a third cysteine residue in catalysis is unclear. Only two cysteines are required for the accepted catalytic sequence (Scheme 1), one to react with the hydroperoxide to form the sulfenic acid, and the second to reduce the sulfenic acid with concomitant formation of a disulfide bond. In the AhpC proteins with only one cysteine residue, the sulfenic acid is directly reduced by an exogenous reducing agent. The recently reported crystal structure of a two-cysteine hydroperoxidase revealed a head-to-tail dimer with intersubunit disulfide bonds (42), as postulated by Ellis and Poole (20), but the structure did not include the 24 C-terminal residues in which the third cysteine would be located in the M. tuberculosis AhpC. The third cysteine in the M. tuberculosis protein could conceivably react with the initially formed disulfide to translocate the disulfide bond to a site that is more accessible to the external reducing agent, or could have some other structural or catalytic role. The key point is that both Cys-174 and Cys-176 facilitate the catalytic reaction, and thus that the third cysteine also participates in the catalytic process. The demonstration that the third cysteine in M. tuberculosis AhpC is catalytically relevant suggests that the third cysteine may also play a role in the function of the other AhpC proteins in the data base that have a third cysteine, although the highly variable location of this cysteine makes it unlikely that this will always be true.

The AhpD from M. tuberculosis is a unique protein that contains two cysteine residues, Cys-129 and Cys-132. A gene coding for this protein has so far only been detected in mycobacteria and in S. viridosporus (29). There is no sequence similarity between AhpD and AhpC or any other related protein. The AhpD cysteine residues are located within a CXXC motif, a motif seen in electron transport proteins, but there is no other sequence similarity to electron transport proteins such as thioredoxin. AhpD is thus the first member of a new class of proteins of heretofore unknown function. The ahpD gene is found immediately downstream of ahpC, in the position occupied by ahpF in the S. typhimurium genome. There are 11 nucleotides between the stop codon for ahpC and the start codon for ahpD, indicating that both proteins are under control of the same promoter. A similar ahpC/ahpD gene arrangement is found in S. viridosporus, although ahpD has not been characterized in that organism (29). This gene arrangement would be useful if both proteins are defense enzymes utilized by the bacteria to protect the organism from reactive oxygen species.

We have demonstrated here that AhpD, in conjunction with AhpF or DTT, catalyzes the reduction of peroxides to alcohols. Thus, although the protein has no sequence relationship with the large AhpC family, it appears to have a similar function. Mutation of Cys-132 to a serine in AhpD resulted in complete loss of activity when activity was assayed with cumene hydroperoxide in the AhpF-dependent system. A lower loss of activity was measured in the DTT assay system, but the activity in this system above the DTT-mediated background was so low that the DTT data for the AhpD system is unreliable. Similar mutation of Cys-129 caused a 95% loss of the AhpD activity in the AhpF assay system (Fig. 11). If one assumes a peroxiredoxin mechanism similar to that proposed for AhpC, these results suggest that Cys-132 reacts with hydroperoxides to give a protein sulfenic acid and the hydroperoxide-derived alcohols (Scheme 1). Cys-129 then participates in the reaction by converting the sulfenic acid to a disulfide. In the absence of Cys-132 catalysis is prevented, but the sulfenic acid obtained from Cys-132 can presumably be directly reduced at a slower rate by AhpF in the C129S mutant.

The two M. tuberculosis proteins have pH rate profiles in the AhpF assay with optima around pH 8-9 (Fig. 8), in contrast to the optimum of pH 7.0 for the S. typhimurium protein. It is not yet possible to tell whether these optima reflect deprotonation of the active site thiols, but it is clear that the pH dependence is not due to ionization of the hydroperoxide substrates, since the pK_a values for these compounds are in the 10-12 range.

In the absence of the endogenous M. tuberculosis redox partner, activity assays were developed for AhpC and AhpD using the S. typhimurium AhpF or DTT as a surrogate redox partner (Figs. 5 and 6). The activity in these assays depends on the protein concentration, as shown in Fig. 7 for the AhpF assay, and for AhpC was saturable at or above a 3:1 ratio of AhpC:AhpF. However, saturation of the AhpD activity only occurred above a 6:1 ratio of the two proteins. The two enzymes appear to act independently, as no interaction between the proteins was detected by physical methods and addition of both proteins to an assay mixture did not give activity above the sum of the activities of the two separate enzymes (Fig. 6).

The fact that the AhpF from S. typhimurium catalyzes the reduction of AhpC and AhpD suggests that a similar flavoprotein might be present in M. tuberculosis. However, a BLAST search of the proteins in the M. tuberculosis genome yielded no matches to the S. typhimurium AhpF, although a number of flavoproteins of unknown function are identified. Assays utilizing the M. tuberculosis thioredoxin and thioredoxin reductase showed that these proteins are not the native redox partners for AhpC or AhpD (not shown). Instead, they simply function as an NADH oxidase system in the presence or absence of a hydroperoxide substrate (31). However, only one of the three thioredoxin isoforms present in the M. tuberculosis was tested, so it is possible that another isoform may couple with AhpC or AhpD. In any case, the endogenous redox partners for M. tuberculosis AhpC and AhpD have yet to be identified.

Analysis of the products formed from cumene hydroperoxide in single turnover experiments with AhpC and AhpD demonstrated that these two proteins exclusively reduce cumene hydroperoxide to the cumyl alcohol product (Fig. 4). AhpF, however, supports a background reaction in which electrons from the flavin are used to homolytically cleave the peroxide bond to give acetophenone as the final product. However, in the presence of AhpC or AhpD, the AhpF-dependent homolytic reaction is a minor side reaction.

A number of hydroperoxides were synthesized in order to evaluate the activities of AhpC and AhpD with respect to a range of substrate sizes and shapes (Fig. 1). The peroxides were reduced by AhpC and AhpD with k_cat/K_m values that differed by a factor of up to ~100 (Table I). Cholesterol hydroperoxide had the lowest K_m for AhpC and cumene hydroperoxide for AhpD, but cholesterol had the highest k_cat/K_m value for both AhpC and AhpD. In accord with our observation that relatively lipophilic compounds have lower K_m values, the recently reported crystal structure of a dimeric hydroperoxidase has shown that the active site is primarily lined by hydrophobic side chains (Phe, Val, and Ala) (42). The finding that both AhpC and AhpD reduced all the hydroperoxides shows that both proteins are broad specificity hydroperoxidases, in accord with their proposed role as enzymes that protect the organism against oxidative stress.

In conclusion, this is the first report of the expression and purification of the M. tuberculosis AhpC, and the first report on the purification and function of any AhpD. These proteins form a peroxiredoxin system distinct from the AhpC systems of E. coli and S. typhimurium in that the M. tuberculosis AhpC is a three- rather than two-cysteine protein, and AhpD is a previously uncharacterized protein so far found only in mycobacteria and S. viridosporus (29). Mutagenesis studies indicate that all three cysteine residues in the AhpC are important for its function, as are both cysteines in AhpD. The catalytic role of the third cysteine in the M. tuberculosis AhpC implies that alkylhydroperoxidase catalysis is more complicated in some proteins than suggested by the work on the AhpC proteins with only one or two cysteine residues. Substrate specificity studies indicate that AhpC and AhpD are broad specificity alkylhydroperoxidases and as such are important components of the M. tuberculosis antioxidant defense system. These two proteins are potential targets for the development of novel approaches for the treatment of M. tuberculosis infections.

	ACKNOWLEDGEMENTS

We thank Clifton E. Barry III, Briggite Wieles, and Leslie B. Poole for their gift of plasmids.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grant GM56531.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.

Supported by a postdoctoral fellowship from the Basque Government.

^§ Supported by a predoctoral fellowship from the Mexican Government through CONACYT.

^¶ To whom correspondence should be addressed: School of Pharmacy, University of California, San Francisco, CA 94143-0446. Tel.: 415-476-2903; Fax: 415-502-4728 or 415-476-0688; E-mail: ortiz@cgl.ucsf.edu.

Published, JBC Papers in Press, April 13, 2000, DOI 10.1074/jbc.M001001200

	ABBREVIATIONS

The abbreviations used are: PCR, polymerase chain reaction; DTT, dithiothreitol; PEI, polyethyleneimine; KP_i, potassium phosphate buffer; HPLC, high pressure liquid chromatography.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES