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

J. Biol. Chem., Vol. 277, Issue 46, 43608-43614, November 15, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/46/43608    most recent M206724200v1 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 Berthold, D. A. Articles by Nordlund, P. Search for Related Content PubMed PubMed Citation Articles by Berthold, D. A. Articles by Nordlund, P. Social Bookmarking                      What's this?

EPR Studies of the Mitochondrial Alternative Oxidase

EVIDENCE FOR A DIIRON CARBOXYLATE CENTER^*

Deborah A. Berthold, Nina Voevodskaya, Pål Stenmark, Astrid Gräslund, and Pär Nordlund

From the Department of Biochemistry and Biophysics, Stockholm University Svante Arrhenius väg 16, S-106 91 Stockholm, Sweden

Received for publication, July 8, 2002, and in revised form, August 28, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The alternative oxidase (AOX) is a ubiquinol oxidase found in the mitochondrial respiratory chain of plants as well as some fungi and protists. It has been predicted to contain a coupled diiron center on the basis of a conserved sequence motif consisting of the proposed iron ligands, four glutamate and two histidine residues. However, this prediction has not been experimentally verified. Here we report the high level expression of the Arabidopsis thaliana alternative oxidase AOX1a as a maltose-binding protein fusion in Escherichia coli. Reduction and reoxidation of a sample of isolated E. coli membranes containing the alternative oxidase generated an EPR signal characteristic of a mixed-valent Fe(II)/Fe(III) binuclear iron center. The high anisotropy of the signal, the low value of the g-average tensor, and a small exchange coupling (J) suggest that the iron center is hydroxo-bridged. A reduced membrane preparation yielded a parallel mode EPR signal with a g-value of about 15. In AOX containing a mutation of a putative glutamate ligand of the diiron center (E222A or E273A) the EPR signals are absent. These data provide evidence for an antiferromagnetic-coupled binuclear iron center, and together with the conserved sequence motif, identify the alternative oxidase as belonging to the growing family of diiron carboxylate proteins. The alternative oxidase is the first integral membrane protein in this family, and adds a new catalytic activity (ubiquinol oxidation) to this group of enzymatically diverse proteins.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Alternative oxidase (AOX)¹ activity, first described as oxygen consumption in the presence of cyanide, was reported in the early 1900s from studies on lily pollen and was later found in high levels in the thermogenic inflorescences of the plant family Araceae (reviewed in Refs. 1-3). It has subsequently been found in many plant tissues and in many fungal (4) and protist species (5). AOX plays a key role in the respiration of the trypanosome (6), the causative agent of African sleeping sickness. Because the enzyme is absent from animal mitochondria, there has been interest in AOX as a target for anti-parasite therapeutics.

The AOX does not pump protons, and therefore the energy derived from the oxidation of ubiquinol is not conserved as ATP. The metabolic rationale behind the presence of AOX can vary with organism. In the mammalian bloodstream stage of trypanosomes, where AOX is the only terminal oxidase, its dissipation of excess reducing equivalents balances the glycolytic reactions occurring in the glycosome. In an unusual group of plants belonging to the family Araceae, the AOX allows the rapid uncoupled respiration, which leads to heating of the inflorescence (2). In more typical plants, there is evidence that AOX is a response to oxidative stress (7). It has been suggested that here the presence of AOX serves to prevent overreduction of the quinone pool and in so doing minimizes the release of reactive oxygen species. Evidence consistent with this role has been obtained in cultured tobacco cells (8). In keeping with this role, in yeast the presence of AOX correlates with the absence of an alternative means to dissipate excess reducing equivalents, i.e. aerobic fermentation (4).

Despite its wide occurrence in mitochondria and the early observation of the activity, the nature of the active site of AOX has remained elusive. In the early 1970s a specific inhibitor was reported, and this discovery allowed AOX activity to be further specified as a cyanide-resistant, salicylhydroxamic acid-sensitive ubiquinol oxidase (9). However, continued efforts at spectroscopic detection of the enzyme were not successful. Detergent-solubilized, partial-purified preparations of this integral membrane protein yielded no EPR signal upon reduction or oxidation, and no visible absorption spectrum above 320 nm (10). Experimental evidence for the identity of the cofactor emerged when iron was shown to be required for the appearance of activity from the AOX apoprotein in whole-cell experiments with the yeast Pichia anomala (formerly Hansenula anomala) (11) and more recently for the trypanosome AOX activity in heme-deficient E. coli (12).

In the late 1980s a partial purification of AOX allowed the generation of an antibody (13, 14), which permitted the first cloning of the gene from the thermogenic aroid Sauromatum guttatum (15). The further cloning of AOX from a variety of species revealed several conserved Glu-Xaa-Xaa-His motifs. This prompted Siedow et al. (16) to postulate that AOX was a diiron carboxylate protein. Modification of that original hypothesis using additional sequence data and evolutionary considerations has generated a model of AOX as a diiron carboxylate protein that interacts with one leaflet of the membrane bilayer as an interfacial integral membrane protein (17, 18). To date it has not been possible to experimentally confirm the identity of the metal center because of the continued difficulties in purifying AOX in a stabilized form and with a yield sufficient for spectroscopic methods (6, 10, 13, 19-21).

Here we have expressed the A. thaliana alternative oxidase AOX1a to high levels in E. coli membranes. Using these membranes, we present the first EPR spectra of AOX. The spectra are characteristic of a mixed-valent-coupled Fe(II)/Fe(III) state and identify AOX as a diiron carboxylate protein. We have also mutated several putative iron ligands and show loss of this distinctive mixed-valent EPR signal, in agreement with the assignment of the EPR signal to the diiron center in the AOX.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Mutagenesis and Expression-- The A. thaliana alternative oxidase gene AOX1a was amplified with polymerase chain reaction (PCR) using pAOX (22) as a template, and the primers (cggaattcagcatgcatatggctagcacgatt) and (caggatccgtcgactcaatgatacccaat), and cloned into the EcoRI-BamHI sites of pMALc2 (New England BioLabs) such that the coding region of AOX (omitting the signal sequence) was fused to the C terminus of MBP. The AOX gene in this construct, pAtAOc2, was sequenced to ensure that no mutations were introduced and subsequently transferred to pMALc2X (New England BioLabs) with an EcoRI-BamHI digest to form pAtAOc3. The AOX gene was split into two cassettes for mutagenesis at the KpnI site. pNNK8 was described previously (23). pKH9 contains the KpnI-HindIII fragment from pAtAOc3 in pUC119. Mutagenesis was performed by recombination PCR (24), using two overlapping oligonucleotides specific to the -lactamase gene (gacttggttgaatattcaccagtc and gactggtgaatattcaaccaaagtc), and two oligonucleotides specific to each mutation: E222A (gcagagaatgCgagaatgcatcttatg and cattctcGcattctctgcttcc); H225A (gagagaatgGCtcttatgacattcatg and cataagaGCcattctctcattc); E273A (gtaccttgCGgaagaagcgatcc and gcttcttcCGcaaggtaccgag); and H327A (gaggctcatGCGcgtgatgtaaacc and catcacgCGCatgagcctcgtc). Following sequencing, the mutated cassettes were reintroduced into pAtAOc3, and the plasmids sequenced again to verify the transfer.

When low level expression of AOX was desired, a spontaneous mutant of pAtAOc2 having an attenuated expression level was used. The mutations in this vector, pAtAOcM, are found outside the AOX-coding region. The isolation and characterization of pAtAOcM will be described elsewhere (25).

For expression, pAtAOc3 was transformed into E. coli C43 (DE3). The cells were grown at 37 °C with shaking in LB media containing 60 µg/ml ampicillin and 40 µM FeSO₄ to an OD₆₀₀ of 0.5, at which point the temperature was dropped to 18 °C. After 1 h, 0.1 mM isopropyl-1-thio--D-galactopyranoside (IPTG) was added to induce expression. The cells were harvested 15-18 h after induction. pAtAOcM was transformed into E. coli GO103 (a gift from R. B. Gennis, University of Illinois), grown at 37 °C in the same medium, and harvested after a 6-h induction with 0.1 mM IPTG.

For complementation studies, the mutations were transferred to plasmid pAtAOn3, which is based upon pAtAOmKX (26), but has an altered 3'-non-coding region containing the rrnB terminator and yields roughly double the level of AOX expression. To construct pAtAOn3, a fragment containing the terminator region was transferred from pAtAOc2 to pAtAOmKX by digestion with KpnI and SacI. The activity of the AOX mutants in pAtAOn3 was assessed by testing their ability to restore aerobic respiration in heme-deficient E. coli as described previously (26). Plates were incubated at 37 °C for 3-4 days. Colonies expressing wild type AOX were visible after 2 days.

Membrane Isolation and Characterization-- The E. coli cell pellet was resuspended at an OD₆₀₀ of 50-100 in 50 mM potassium phosphate, pH 7.0, 5 mM dithiothreitol, and 10 mM pyruvate, with 0.5 mM phenylmethylsulfonyl fluoride added immediately prior to breakage. The cells were broken using one pass through a French Pressure Cell (18,000 psi). Unbroken cells and debris were removed by centrifugation (15 min, 10,000 × g), and the supernatant was removed and centrifuged a second time. The supernatant of the second centrifugation was centrifuged (2 h, 100,000 × g) to pellet the membrane fraction. The pellet was homogenized in a minimal volume of 15% sucrose, 50 mM potassium phosphate, 1 mM EDTA, 10 mM pyruvate, pH 7.0, and frozen at 80 °C. Prior to EPR analysis, the membranes were thawed and diluted into 50 mM Tris, 5 mM pyruvate, pH 7.5, and ultracentrifuged as above. The pellet was homogenized in 15% sucrose, 50 mM Tris, 5 mM pyruvate, pH 7.5 and either used for EPR immediately, or stored at 80 °C until use. In later experiments, 50 mM Tris, pH 7.5 replaced the potassium phosphate buffer throughout the preparation, and EDTA was omitted from the 80 °C storage buffer. Protein was determined using Petersen's modification of the Lowry method (27), with bovine serum albumin as the standard. The protein concentration in the E. coli C43 (DE3) membrane preparations ranged from 73-78 mg protein/ml, and that in the E. coli GO103 preparation (containing the low expression level of AOX) was 58 mg protein/ml. For preparation of EPR samples, the membrane suspension was diluted 2-fold. All protein and spin concentrations given in the text refer to the original undiluted sample. Denaturing gel electrophoresis was carried out using a modified Laemmli system containing 2.5 M urea as previously described (10), except that the protease inhibitor was omitted, and dithiothreitol (0.1 M) replaced -mercaptoethanol as the reducing agent.

Redox Titrations-- Oxidation-reduction potential titrations were carried out in an apparatus similar to that described by Dutton (28). The sample (1.6-1.8 ml) was stirred in a vessel maintained at 0 °C. The redox potential was measured using a platinum electrode with a calomel reference. Seven mediators were used: indigo carmine (E'_o = 125 mV), methylene blue (E'_o = 10 mV), duroquinone (E'_o = 10 mV), thionine (E'_o = 56 mV), phenazine methosulfate (PMS) (E'_o = 80 mV), 1,2-naphthoquinone (E'_o = 143 mV), N,N,N',N'-tetramethyl phenylenediamine (E'_o = 276 mV) in a mixture containing 50 µM each. The membrane suspension was reduced with 10 µl of 200 mM dithionite (DT) in 1 M Tris buffer (pH 7.5) or 30 µl of 200 mM NADH in 50 mM Tris buffer (pH 7.5) while flushing with argon. After stabilization of redox potential of the reduced membrane suspension at ~300mV (in the case of DT) and 150mV (in the case of NADH) reoxidation was performed either by titration with 200 mM K₃Fe(CN)₆ or by flushing with air. Samples were transferred to EPR tubes using a Hamilton syringe and frozen in cold isopentane.

EPR Spectroscopy-- EPR spectra were recorded on a Bruker ESP 300 X-band spectrometer with an Oxford Instruments ESR9 helium cryostat. Spin quantitation was made by comparison with a standard solution of 1 mM CuSO₄ in 50 mM EDTA, by double integration of spectra recorded at nonsaturating microwave power levels using standard Bruker software.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

A single gene product was shown to be sufficient to give an active AOX when it was reported that the AOX1a gene from an A. thaliana library restored aerobic respiration in a heme-deficient E. coli mutant (22). We undertook extensive screening to identify an inducible high level expression system. The optimal expression construct was found to be an MBP fusion under a tac promoter. Fig. 1 depicts a Coomassie Blue-stained SDS-polyacrylamide gel of isolated membranes showing the level of protein expressed from this plasmid, pAtAOc3, following IPTG induction. Using scanning densitometry we estimated that the MBP-AOX fusion was 25% of the total protein or 250 µM AOX monomer. This MBP-AOX fusion is active, insofar as uninduced (leaky) expression from pAtAOc3 complements heme-deficient E. coli, restoring the capacity for aerobic respiration.

View larger version (113K): [in this window] [in a new window]   Fig. 1.   Expression of the MBP-alternative oxidase fusion in isolated E. coli membranes. Isolated membranes (2 µg of protein) were run on a 12% acrylamide, 2.5 M urea SDS-polyacrylamide gel and stained with Coomassie Blue. Lane 1, protein size standards labeled in kDa; lane 2, E. coli C43 (DE3) lacking the AOX expression plasmid; lane 3, E. coli C43 (DE3)/pAtAOc3; lane 4, E. coli GO103/pAtAOcM; lanes 5-8, isolated membranes from E. coli C43 (DE3) containing the expression plasmid pAtAOc3 with the following mutations: lane 5, E222A; lane 6, H225A; lane 7, E273A; lane 8, H327A. The calculated mass of the MBP-AOX fusion is 77 kDa.

Since the sequence motif suggested a diiron carboxylate site in AOX (16-18), we used EPR to search for signals that might arise from the overexpressed AOX protein in the isolated membranes. We began with redox conditions similar to those used to observe the mixed-valent paramagnetic site in mouse ribonucleotide reductase protein R2 (29). EPR spectra were recorded at low temperature (5 K). The EPR sample was made strictly anaerobic by repeatedly adding argon and evacuating. DT (2 mM) was added directly to the EPR tube with the mediator PMS (1 mM). Only an EPR signal at g = 1.94 was seen and no mixed-valent signal (Fig. 2a). The g = 1.94 signal is typical for the iron-sulfur centers in the respiratory chain of E. coli, which are paramagnetic in the reduced form. Fig. 2, b-d show EPR spectra from samples prepared in an oxygen electrode chamber without initial removal of oxygen, where they were reduced by DT and PMS, and then exposed to air with increasing incubation times. The spectra show a characteristic mixed-valent (Fe(II)/Fe(III)) EPR signal that should originate from the metal site of AOX with g-values 1.86, 1.67, 1.53. The mixed-valent signal appears in increasing amount with time of aerobic incubation. Fig. 3 depicts a time course of sample preparation showing the oxygen concentration and quantitation of the mixed-valent EPR signal.

View larger version (13K): [in this window] [in a new window]   Fig. 2.   EPR spectra of the AOX expressed in E. coli membranes (a-e) and E. coli membranes without AOX (f). Sample a was made strictly anaerobic then reduced by DT (2 mM) and PMS (1 mM). Samples b-f were reduced with DT and PMS without prior removal of oxygen in the oxygen electrode chamber, then b was frozen, e was kept anaerobic for 30 min, and c, d, and f were exposed to air for 20 (c and f) or 30 min (d). Instrument parameters are as follows: microwave frequency, 9.62 GHz; microwave power, 100 milliwatts; modulation frequency, 100 kHz; modulation amplitude, 1 mT; temperature, 5 K.

View larger version (13K): [in this window] [in a new window]   Fig. 3.   Time course of AOX sample preparation showing formation of the mixed-valent state. E. coli membranes containing the overexpressed alternative oxidase (76 mg of protein/ml) were added to the oxygen electrode sample chamber, held at 7 °C, with stirring. Phenazine methosulfate (PMS, 2.5 mM) and dithionite (DT) were added at the times indicated. Oxygen was re-introduced by removing the chamber lid. Samples were removed to an EPR tube made anaerobic by flushing with argon and frozen immediately using isopentane. The time of removal and the spin concentration of each sample are shown by bars. One sample (indicated by an asterisk) was removed to the EPR tube and held anaerobic on ice for the time indicated (dashed line, labeled -O2), and then frozen. Instrument parameters are as follows: microwave frequency, 9.62 GHz; microwave power, 100 milliwatts; modulation frequency, 100 kHz; modulation amplitude, 1 mT; temperature, 5 K. The ordinate on the left (O2 concentration) is derived from the initial calibration of the electrode (100% saturation = 376 µM at 7 °C) but over the course of the measurement the sensitivity of the electrode declined from interaction with the oxidation products of DT. Note that the lag in oxygen reappearance is due to the residual DT in the chamber, as well as the slow diffusion across the sample interface owing to the viscosity of the membrane sample. The spin concentration of the sample begins to increase with time after oxygen is reintroduced to the chamber and becomes maximal near the point where the oxygen begins to accumulate.

A control sample with overexpressed AOX prepared was removed from the electrode chamber prior to the introduction of oxygen and was kept anaerobically on ice for 30 min. Only a small EPR signal from the mixed-valent iron center was observed (Fig. 2e), which is even smaller than that from the initial sample (Fig. 2f). Another control sample, consisting of E. coli membranes, which lacked AOX, was treated under the same conditions as Fig. 2c (reduction by DT and PMS, and exposure to oxygen for 20 min). This sample gave no evidence of a mixed-valent EPR signal (Fig. 2f and Table IA).

Further characterization of the conditions producing the mixed-valent EPR signal are shown in Table 1, B-E. The membranes were reduced under an argon atmosphere, and the redox potential was monitored throughout the preparation. The redox state of the Fe/S centers in the E. coli membranes, as determined by the g = 1.94 and the g = 2.03 EPR signals, were used as internal monitors for the redox state. Interestingly, with overexpressed AOX, the mixed-valent signal was initially observed in membranes prior to any treatment at a level of about 25% of the maximal observable signal. The maximal signal (about 11 µM) was obtained after reduction by DT and reoxidation by air. We observed this maximum signal (with a range from 9.6 to 12 µM) in six independent experiments, which used three separate membrane preparations. In contrast, when the sample was reduced and reoxidized with ferricyanide rather than air, only a small signal was observed (which may be due to imperfect anaerobic transfer of the sample to the EPR tube). This result demonstrates the specific requirement for oxygen and not simply an oxidant for the generation of the mixed-valent EPR signal. When NADH was used to reduce the sample, only a partial reduction of the Fe/S centers was obtained, and in this instance we did not see the maximal appearance of the mixed-valent signal. We also found that no mediators were required for the appearance of the signal, regardless of whether DT or NADH was used as reductant.

A very high level of MBP-AOX protein is expressed from pAtAOc3 under conditions of IPTG induction. For comparison, we investigated a preparation of E. coli GO103/pAtAOcM membranes containing a low AOX expression level. In this preparation, the MBP-AOX band is not immediately identifiable on the Coomassie Blue-stained gel (Fig. 1, lane 4). Preliminary results from an immunoblot indicate that the level of expression is 4-7-fold less than the sample expressed from pAtAOc3 (data not shown). In these membranes, we observed a maximal mixed-valent spin concentration of 2 µM (Table 1C). This concentration is 6-fold less than what is observed for the high expression level samples.

In order to observe the effect of a disrupted diiron center on the appearance of the mixed-valent EPR signal, we mutated each of four putative iron ligands (Glu-222, His-225, Glu-273, and His-327) individually to alanine. The mutants were tested for their ability to restore aerobic respiration to the heme-deficient E. coli SASX41BD. Using this complementation assay, it was observed that none of the four mutants were active. When either of the two histidine ligands His-225 or His-327 were mutated to alanine, the protein was destabilized such that little or no protein was observed on a Coomassie Blue-stained SDS gel (Fig. 1, lanes 6 and 8). Mutation of the glutamate residues Glu-222 or Glu-273 to alanine did allow a level of protein expression similar to the native MBP-AOX fusion (Fig. 1, lanes 5 and 7). Reduction and reoxidation of the E. coli membranes containing either of these two mutants yielded no detectable mixed-valent EPR signal (Table 1, D and E), in agreement with the proposed role of these residues as iron ligands.

The EPR signal at g < 2 observed after the reduction and reoxidation of AOX is highly characteristic of an antiferromagnetic-coupled Fe(II)/Fe(III) site (30-33). It is typically observed only at very low temperatures due to rapid relaxation. The temperature dependence of the microwave saturation parameter P_1/2 of the signal showed a linear dependence on 1/T as predicted by theory (32, 34). (Fig. 4.) From this relationship, the exchange integral was found to be J = 5.5 ± 1 cm¹.

View larger version (14K): [in this window] [in a new window]   Fig. 4.   Dependence of the EPR half-saturation power P1/2 on the inverse absolute temperature for the mixed-valent state of AOX. The P1/2 was determined by a nonlinear least squares fitting of the experimental data as described by Fox et al. (48).

We also observed the signal of the fully reduced diiron site in parallel mode EPR (microwave frequency 9.35 GHz) at g = 15, and this signal disappeared upon oxidation when the mixed-valent signal increased (Fig. 5).

View larger version (12K): [in this window] [in a new window]   Fig. 5.   EPR spectra of AOX protein expressed in E. coli membranes recorded in parallel mode. Sample a was reduced to 320 mV by DT, and sample b was reduced by DT and exposed to air. Instrument parameters are as follows: microwave frequency, 9.35 GHz; microwave power, 50 milliwatts; modulation frequency, 100 kHz; modulation amplitude, 1 mT; temperature 5 K.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The present study gives the first experimental evidence for a coupled binuclear iron center at the active site of AOX, made possible by a high level of expression of AOX in E. coli. Careful attempts made earlier to locate an EPR signal in the enzyme in whole mitochondria had been unsuccessful (35-37), as were EPR analyses of partially purified preparations (10, 38).

The oxidized form of a typical diiron carboxylate protein has an antiferromagnetic-coupled pair of high-spin ferric ions and is therefore EPR silent. The coupling is mediated by a bridging ligand, typically an oxo or hydroxo group. The fully reduced form has a pair of ferrous ions, weakly if at all coupled, and is likewise EPR silent, except in parallel mode EPR where a high g-value signal may be seen (39, 40). The antiferromagnetic-coupled Fe(II)/Fe(III) state has a spin 1/2 and should give rise to an EPR spectrum with g-value components below 2 (29, 41, 42). These EPR signals are now observed in AOX: a mixed-valent signal under oxidizing conditions and a g = 15 signal under reducing conditions. The mixed-valent signal is present when AOX is expressed in the membrane preparation and is lacking when AOX is absent. The observation of the mixed valent signal shows that AOX contains a diiron carboxylate site, and further support is provided by the g = 15 signal, which is observed under fully reducing conditions. This EPR signal is typical of a site containing two exchange-coupled high spin ferrous ions. Such systems having S = 4 can show an integer spin EPR signal, which arises from a M_S = 8 transition. This transition is usually forbidden but is allowed in a parallel mode EPR experiment (31, 39). Normally this EPR signal cannot be quantitated since it may be only partially observable in the EPR spectrum.

The EPR observations on the native samples are further substantiated by the results of the mutants; when either of two putative iron ligands are changed from glutamate to alanine, no mixed-valent iron signal is seen. We found that mutation of any of the four putative ligands (Glu-222, His-225, Glu-273, His-327) resulted in a loss of AOX activity, as assessed by their inability to restore aerobic respiration to a heme-deficient strain of E. coli. This is consistent with a recent report of mutation of the residues in the trypanosome AOX corresponding to Glu-222, His-225, and His-327 (12). Here, too, each substitution resulted in a loss of AOX activity.

The mixed-valent state of AOX was observed after partial (1-electron) oxidation of the fully reduced state, in contrast to other diiron carboxylate proteins, such as ribonucleotide reductase and methane monooxygenase, where the mixed-valent state is generated by a one-electron reduction of the diferric center (29, 43). The observed g-value components of the AOX signal (1.86, 1.67, 1.53) are very similar to the mixed-valent signal from other characterized diiron carboxylate proteins and model compounds, as summarized in Table II. For AOX, the large g-anisotropy (see Davydov et al., Ref. 33), the small value of the exchange coupling constant (J), the fact that the EPR signal is seen only at very low temperature, and the lack of a visible absorption at wavelengths above 320 nm (10), all point to the presence of a hydroxo (rather than oxo) bridge between the two irons. In fact, the parameters also leave open the possibility that the bridging ligand may be an aquo group.

View this table: [in this window] [in a new window]   Table II g-tensor components, average g-values (gav), and values of the exchange coupling integral (J) for mixed-valent forms (Fe(II)/Fe(III)) of diiron proteins and model compounds

The formation of the mixed-valent EPR signal requires a reaction with molecular oxygen. The appearance of the mixed-valent signal as a function of time (Fig. 4) shows that it appears under conditions where the oxygen concentration is quite low, due to oxygen scavenging by residual DT and the endogenous E. coli ubiquinol oxidases. From this we infer that the reaction with oxygen leading to the mixed-valent state of AOX must be efficient and may also be quite rapid.

The observation that the mixed-valent signal can be generated in the absence of mediators following reduction by NADH, a physiological substrate, could suggest that we have trapped a catalytic intermediate on the natural reaction pathway. At this point this interpretation is inconsistent with what is known about the interaction of dioxygen with the binuclear iron sites in soluble diiron carboxylate enzymes. There oxygen undergoes a two-electron reduction upon binding to the diferrous center, forming a peroxy intermediate (31). It is difficult to reconcile such a mechanism with the one-electron transfer required to generate the mixed-valent Fe(II)/Fe(III) state we observe here with the AOX.

One should also consider the possibility that the mixed-valent state, which consistently accumulates to a maximum of about 5% of the estimated concentration of the AOX monomer, is a consequence of damaged or malformed iron centers. Incomplete iron incorporation may be a realistic concern in a system where the protein is expressed at such a high level (here to an estimated 25% of the total membrane protein). Therefore we also measured the mixed-valent EPR signal in a membrane preparation in which AOX is expressed at a 4-7-fold lower level of expression. The results showed an approximate 6-fold decrease in the concentration of the maximal mixed-valent signal (Table 1C). Thus, the appearance of the mixed-valent signal is not related to the very high level of AOX expression.

The appearance of the mixed-valent state of the AOX could possibly represent a physiologically relevant reversible inactivation pathway, in which the inactivated AOX can be reactivated in the presence of sufficient reductant. Activation of the plant AOXs through reduction of the disulfide bridge that links the subunits of the homodimer, followed by interaction of the free cysteine with pyruvate, is well established (44). Our membrane preparations have been isolated in the presence of dithiothreitol and kept in the presence of pyruvate throughout, in order to keep the enzyme in an activated state. Inactivation by air-oxidation to form the disulfide bridge is known to be slow in the absence of pyruvate and presumably slower yet in its presence. Therefore, an inactivation mechanism through formation of the disulfide-bridge is unlikely to correlate to a putative inactivation mechanism involving oxidation of the iron site. There has been one report proposing an additional reductive mechanism of AOX activation, independent of the disulfide/pyruvate mechanism, in soybean cotyledon mitochondria (45). A lag in the appearance of AOX activity followed the rapid reduction of the ubiquinone pool, suggesting the presence of a second mechanism of reductive activation. Whether there is any relationship between this potential activation mechanism and our observations on the disappearance of the mixed-valent state with reduction and its reappearance with oxidation, is a question for future studies.

AOX shares some features with two proteins recently predicted to also contain a coupled diiron center, based on sequence motif: the plastid terminal oxidase (Im) (46) and demethyoxyquinone hydroxylase (Coq7/clk-1) (47). Like AOX, each of these enzymes interacts with a quinone substrate and is predicted to be bound to the membrane as an interfacial integral membrane protein. Together with AOX, the plastid terminal oxidase and demethoxyquinone hydroxylase may comprise a new group of diiron proteins.

The demonstration of EPR signals characteristic of a hydroxo-bridged diiron center, together with the sequence motif, firmly establishes AOX as a member of the family of diiron carboxylate proteins. AOX is the first integral membrane protein with such a center, and as a ubiquinol oxidase adds a new catalytic activity to this family of proteins.

	ACKNOWLEDGEMENTS

We thank Prof. P. Brzezinski and G. Gilderson for the use of the redox potentiometric apparatus and Prof. K. K. Andersson for helpful discussions.

	FOOTNOTES

^* This work was supported by grants from the Swedish Research Council (to P. N. and A. G.), the European Union (EU-TMR Contract Number FMRX-CT98-0207) (to P. N. and A. G.), and the Wenner-Gren Center Foundation (to N. V.).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 Biochemistry and Biophysics, Stockholm University, Svante Arrhenius väg 12, S-106 91 Stockholm, Sweden. Tel.: 46-8-16-2715; Fax: 46-8-15-3679; E-mail: berthold@dbb.su.se.

Published, JBC Papers in Press, September 4, 2002, DOI 10.1074/jbc.M206724200

	ABBREVIATIONS

The abbreviations used are: AOX, alternative oxidase; DT, sodium dithionite; MBP, maltose-binding protein; PMS, phenazine methosulfate; IPTG, isopropyl-1-thio--D-galactopyranoside; EPR, electron paramagnetic resonance.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				A. Fu, S. Park, and S. Rodermel Sequences Required for the Activity of PTOX (IMMUTANS), a Plastid Terminal Oxidase: IN VITRO AND IN PLANTA MUTAGENESIS OF IRON-BINDING SITES AND A CONSERVED SEQUENCE THAT CORRESPONDS TO EXON 8 J. Biol. Chem., December 30, 2005; 280(52): 42489 - 42496. [Abstract] [Full Text] [PDF]

				J. B. Vicente and M. Teixeira Redox and Spectroscopic Properties of the Escherichia coli Nitric Oxide-detoxifying System Involving Flavorubredoxin and Its NADH-oxidizing Redox Partner J. Biol. Chem., October 14, 2005; 280(41): 34599 - 34608. [Abstract] [Full Text] [PDF]

				S. Padilla, T. Jonassen, M. A. Jimenez-Hidalgo, D. J. M. Fernandez-Ayala, G. Lopez-Lluch, B. Marbois, P. Navas, C. F. Clarke, and C. Santos-Ocana Demethoxy-Q, An Intermediate of Coenzyme Q Biosynthesis, Fails to Support Respiration in Saccharomyces cerevisiae and Lacks Antioxidant Activity J. Biol. Chem., June 18, 2004; 279(25): 25995 - 26004. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 277/46/43608    most recent M206724200v1 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 Berthold, D. A. Articles by Nordlund, P. Search for Related Content PubMed PubMed Citation Articles by Berthold, D. A. Articles by Nordlund, P. Social Bookmarking