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

J. Biol. Chem., Vol. 277, Issue 18, 16220-16228, May 3, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/18/16220    most recent M200198200v1 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 Oh, J.-I. Articles by Kaplan, S. Search for Related Content PubMed PubMed Citation Articles by Oh, J.-I. Articles by Kaplan, S. Social Bookmarking                      What's this?

Oxygen Adaptation

THE ROLE OF THE CcoQ SUBUNIT OF THE cbb₃ CYTOCHROME c OXIDASE OF RHODOBACTER SPHAEROIDES 2.4.1^*

Jeong-Il Oh and Samuel Kaplan

From the Department of Microbiology and Molecular Genetics, University of Texas Health Science Center, Medical School, Houston, Texas 77030

Received for publication, January 8, 2002, and in revised form, February 14, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The cbb₃ cytochrome c oxidase of Rhodobacter sphaeroides consists of four nonidentical subunits. Three subunits (CcoN, CcoO, and CcoP) comprise the catalytic "core" complex required for the reduction of O₂ and the oxidation of a c-type cytochrome. On the other hand, the functional role of subunit IV (CcoQ) of the cbb₃ oxidase was not obvious, although we previously suggested that it is involved in the signal transduction pathway controlling photosynthesis gene expression (Oh, J. I., and Kaplan, S. (1999) Biochemistry 38, 2688-2696). Here we go on to demonstrate that subunit IV protects the core complex, in the presence of O₂, from proteolytic degradation by a serine metalloprotease. In the absence of CcoQ, we suggest that the presence of O₂ leads to the loss of heme from the core complex, which destabilizes the cbb₃ oxidase into a "degradable" form, perhaps by altering its conformation. Under aerobic conditions the absence of CcoQ appears to affect the CcoP subunit most severely. It was further demonstrated, using a series of COOH-terminal deletion derivatives of CcoQ, that the minimum length of CcoQ required for stabilization of the core complex under aerobic conditions is the amino-terminal ~48-50 amino acids.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The purple nonsulfur photosynthetic bacterium, Rhodobacter sphaeroides, contains a branched respiratory electron transport chain that is terminated with at least two cytochrome c oxidases and at least one functional quinol oxidase (Qxt) (1). The aa₃- and cbb₃-type cytochrome c oxidases, both belonging to the heme-copper oxidase superfamily, catalyze the reduction of molecular oxygen to water using electrons derived from the oxidation of cytochromes c₂ and c_y (2-4). In R. sphaeroides the aa₃ oxidase is the major cytochrome c oxidase under highly aerobic conditions, whereas the cbb₃ oxidase is the predominant and perhaps exclusive cytochrome c oxidase under oxygen-limiting and anaerobic conditions, respectively (5).

The bacterial aa₃ cytochrome c oxidase has been demonstrated both genetically and from x-ray crystallographic analyses to consist of four subunits (6, 7). Subunits I and II form the functional unit containing all of the redox centers. These consist of a low spin heme and a binuclear center composed of a high spin heme and Cu_B in subunit I, and Cu_A in subunit II, as well as amino acid residues required for enzyme activity and proton translocation. Subunit III appears not to be essential for catalytic function of the enzyme because a functional two-subunit enzyme consisting of subunits I and II has been demonstrated (8). It was suggested that the removal of subunit III leads to "suicide inactivation" of the enzyme during turnover (9). The function of the smallest subunit IV is unknown. Deletion of the gene encoding subunit IV (CtaH in Paracoccus denitrificans) was shown to have no effect on either the integrity of the enzyme complex or its spectral and enzymatic properties (7).

The cbb₃ cytochrome c oxidase encoded by the ccoNOQP (fixNOQP) operon is also composed of four subunits (10, 11). The ccoN (fixN) gene encodes the catalytic subunit of the oxidase, which is homologous to subunit I of the aa₃ oxidase. The ccoO (fixO) and ccoP (fixP) genes encode membrane-bound mono- and diheme cytochromes c. On the basis of determined redox potentials, it was suggested that electrons are transferred from cytochrome c to CcoN via CcoP and CcoO in that order (12). CcoN, CcoO, and CcoP subunits are all required for catalytic activity of the cbb₃ oxidase (13, 14). Elimination of any redox prosthetic center by site-directed mutagenesis was shown to lead to the loss of enzyme activity and a defect in enzyme assembly (14). The CcoQ (FixQ) is the smallest subunit of the cbb₃ oxidase, consisting of 48-73 amino acids depending upon its source. Its primary structure shows a basal level of homology to subunit IV (CtaH) of the aa₃ oxidase in the membrane-spanning helix region. In-frame deletion of the ccoQ gene was demonstrated to affect neither the catalytic properties nor the assembly of the enzyme complex in R. sphaeroides when examined in cells grown anaerobically (5). This is also true for Bradyrhizobium japonicum (13).

We have shown the cbb₃ oxidase to play an additional role as an redox sensor in a signal transduction pathway. Electron flow through the cbb₃ oxidase is postulated to generate a signal under aerobic conditions, which is inhibitory to and mediated by the PrrBA two-component activation system and which leads to the repression of photosynthesis (PS)¹ gene expression (14, 15). Although a ccoQ in-frame deletion mutant of R. sphaeroides grown under anaerobic conditions retains a catalytically intact cbb₃ oxidase, under highly aerobic conditions it produces spectral complexes accompanied by the aerobic derepression of PS genes, as observed for Cco null mutants (5). Based upon these observations, we initially suggested that CcoQ was involved in the process of actually transducing the inhibitory signal from the cbb₃ oxidase to the PrrBA two-component system (5). However, further study of the role of CcoQ reveals a more subtle mode of action, leading us to suggest that the CcoQ subunit stabilizes the cbb₃ oxidase complex under aerobic conditions and, in its absence, the core complex is destabilized, thereby removing the inhibitory signal. We report these findings here.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Strains, Plasmids, and Growth Conditions-- The bacterial strains and plasmids used in this study are listed in Table I. R. sphaeroides and Escherichia coli strains were grown as described previously (5).

DNA Manipulations and Conjugation Techniques-- Standard protocols or manufacturer's instructions were followed for recombinant DNA manipulations. Mobilization of plasmids from E. coli strains into R. sphaeroides strains was carried out as described elsewhere (16).

Construction of the CBB3 Mutant and Plasmids-- To construct the CBB3 mutant, a 1400-bp StuI fragment containing a 3'-portion of ccoN, all of ccoO and ccoQ, and a 5'-portion of ccoP, was deleted by the restriction of pCCO1 with StuI and subsequent recircularization, yielding the plasmid pCBB31. A 3.2-kb EcoRV-SacI fragment from pCBB31 was cloned into the suicide vector pLO1 digested with PmeI and SacI. The resulting plasmid pCBB32 was transferred from E. coli strain S17-1 to R. sphaeroides 2.4.1 by conjugation. Heterogenotes of R. sphaeroides, generated by a single recombination event, were selected for kanamycin resistance on Sistrom's medium A (SIS) agar plates incubated under aerobic conditions. Isogenic homogenotes were obtained from the heterogenotes after a second recombination selecting for sucrose resistance on SIS agar plates containing 15% (w/v) sucrose, 1% (w/v) Bactotrypsin, 0.5% (w/v) yeast extract, 0.5% (w/v) NaCl, and 0.5% (v/v) Me₂SO under dark anaerobic conditions. The allelic exchange was verified by PCR.

For pUI2803NHIS, a 0.8-kb fragment containing the 3' portion of ccoN with a tail of six histidine codons immediately upstream of its stop codon was generated by a two-round recombination PCR using Pfu Turbo polymerase (Stratagene, La Jolla, CA). With pCCO2 as the template, two primary PCR reactions were performed with the primers HIS+ (5'-GTTCCCGCCACCACCATCACCATCACTGAGTGAAGATAAGGGGACA-3') and OUT+ (5'-CGAATAGCGCTCGCCGACGCGGGC-3'), and HIS (5'-TTCACTCAGTGATGGTGATGGTGGTGGCGGGAACGGCCGCGGCGCG-3') and OUT (CTACGGCATGTCGACCTTCGAGGG-3') to generate two DNA fragments containing a 34-bp overlapping region. The two primary PCR products were used as templates for the secondary PCR, which was performed using the primers OUT+ and OUT. A 0.8-kb PCR product was restricted with ApaI and cloned into pCCO2 digested by ApaI, replacing its 0.8-kb wild-type ApaI fragment. The resulting plasmid pCCO2NHIS with the correct orientation of the 0.8-kb ApaI fragment was digested with BamHI and EcoRI, and a 4.7-kb BamHI-EcoRI fragment was ultimately cloned into pRK415 restricted with the same enzymes, giving the plasmid pUI2803NHIS.

For the construction of pUI2803NHIS-CCOQ, a 0.83-kb SacI-StuI fragment from pCCOQ1 was cloned into the vector portion of p19CCOQP, in which the 0.85-kb wild-type SacI-StuI fragment was removed, to give the plasmid pCCOOP. A 2-kb EcoRI-SacI fragment was cloned into the vector portion of pUI2803NHIS restricted with the same enzymes, yielding the plasmid pUI2803NHIS-CCOQ.

Site-directed Mutagenesis-- To obtain a set of truncated forms of CcoQ, a series of nonsense mutations was introduced into the plasmid-borne ccoQ gene using the QuikChange site-directed mutagenesis kit (Stratagene) with the template plasmid pOYQ. Synthetic deoxyoligonucleotides 34 bases long containing the TGA stop codon in place of Ser-60, Asp-56, Thr-51, or Arg-48 codons in the middle of their sequences were employed to mutagenize the corresponding codons. Following verification of the mutations by DNA sequencing, a 490-bp SalI-StuI fragment containing the mutated ccoQ gene was cloned into pBBR1MCS2 restricted with EcoRV and SalI. The resulting plasmids pBBRQTGA1-pBBRQTGA4 (see Table I) were introduced into R. sphaeroides CCOQ in which 171 bp of the ccoQ gene was deleted in-frame by gene replacement.

Purification of the His₆-tagged cbb₃ Cytochrome c Oxidase-- The R. sphaeroides CBB3 strain carrying either pUI2803NHIS or pUI2803NHIS-CCOQ was grown semi-aerobically or under anaerobic dark Me₂SO conditions to the late exponential phase in SIS supplemented with 2 µg of tetracycline per ml. For semi-aerobic growth, where the expression of the ccoNOQP operon is highest, 1-liter flasks were filled with 700 ml of SIS and shaken at 150 rpm on a rotary shaker. Two-liter cultures were harvested, resuspended in 50 ml of buffer A (20 mM Tris-HCl, pH 8.0) containing 1 mM phenylmethylsulfonyl fluoride, and disrupted by two passages through a French pressure cell at 120 megapascals. Following DNase treatment (150 units of DNase (Promega, Madison, WI) in the presence of 10 mM MgCl₂ for 30 min at room temperature), cell-free crude extracts were obtained by centrifugation at 20,000 × g for 15 min at 4 °C two times. Membrane fractions were isolated by ultracentrifugation of crude extracts at 150,000 × g for 1 h at 4 °C. After the membrane fraction (pellet) was washed twice with buffer A, the membranes were solubilized in 15 ml of buffer A containing 1% (w/v) n-dodecyl -D-maltoside (DM), 1 mM phenylmethylsulfonyl fluoride, and 100 mM NaCl, and then centrifuged at 150,000 × g for 1 h at 4 °C. The supernatant was taken as solubilized membrane proteins and used for affinity chromatography. To reduce the concentration of DM, 12 ml of buffer A containing 100 mM NaCl was added to the solubilized membrane protein and then imidazole was added to a final concentration of 5 mM. 2 ml of the 50% (v/v) nickel-nitrilotriacetic acid HIS-bind slurry (Novagen, Madison, WI) was added to 30 ml of solubilized membrane protein and mixed gently by shaking at 4 °C for 2 h. The protein-resin mixture was loaded into a column, and the column was washed with 10 volumes of buffer B (buffer A with 100 mM NaCl and 0.01% (w/v) DM) containing 5 mM imidazole followed by 6 volumes of buffer B containing 60 mM imidazole. The red-colored cbb₃ oxidase was finally eluted with 250 mM imidazole in buffer B. The fractions containing the cbb₃ oxidase were dialyzed overnight against 2 liter of buffer A containing 0.01% (w/v) DM to remove imidazole and NaCl. The desalted cbb₃ oxidase was concentrated by means of ultrafiltration (membrane YM10, Millipore Co., Bedford, MA).

Spectroscopic and Immunoblotting Analyses-- Both heme b and heme c contents were determined by using the pyridine hemochrome difference spectra and the cognate calculation matrix as described by Berry and Trumpower (17).

The reduced plus CO minus reduced difference spectra and the oxidized minus oxidized plus CN spectra were obtained as described previously (7, 18). The CO and CN reactivity of the cbb₃ oxidase reflecting the content of the high spin heme b within the purified enzyme was estimated for the following wavelength pairs: 415.5 and 436 nm (CO binding), and 404 and 420 nm (CN binding) (19). All spectra were recorded using the purified three- and four-subunit cbb₃ oxidase equilibrated in buffer A containing 0.01% (w/v) DM. The B800-850 and B875 spectral complex levels were determined spectrophotometrically as described elsewhere (5). Preparation of solubilized membrane proteins, SDS-PAGE, and Western blotting were performed as described previously (14), except that samples were denatured for 40 min at room temperature in SDS loading buffer prior to electrophoresis.

Enzyme Assays, Protein Determination, and Heme Staining-- Preparation of crude cell extracts and determination of -galactosidase activities were performed as described previously (5). Cytochrome c oxidase activities were measured spectrophotometrically with reduced horse heart cytochrome c (5). Protein concentration was determined by the bicinchoninic acid protein assay (Pierce) using bovine serum albumin as the standard protein.

After SDS-PAGE the c-type cytochromes were visualized via their intrinsic peroxidase activity, using 3,3',5,5'-tetramethyl benzidine and H₂O₂ (20).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CcoQ and Its Effect on the Activity and Integrity of the Core Oxidase-- In good agreement with our previous report (5), the levels of the CcoO and CcoP subunits detected in the CCOQ in-frame deletion mutant were similar to those found in the wild-type strain of R. sphaeroides 2.4.1 when both were grown anaerobically in the dark with Me₂SO as an external electron acceptor (Fig. 1A). Surprisingly, when grown under either aerobic (30% O₂) or semi-aerobic (2% O₂) conditions, the CCOQ mutant showed significant decreases in both the CcoO and especially the CcoP polypeptides in comparison with the wild type (Fig. 1A). In the case of CcoP, the polypeptide was undetectable by means of immunoblotting. When assayed using reduced cytochrome c, the cytochrome c oxidase activity of the CCOQ mutant was only marginally lower than that detected in the wild type 2.4.1 when both strains were grown under anaerobic dark Me₂SO conditions (Fig. 1B). The CCON mutant when used as a negative control grown under the same conditions showed no cytochrome c oxidase activity, which is consistent with previous results revealing that the cbb₃ oxidase appears to be the exclusive cytochrome c oxidase activity present in R. sphaeroides grown under anaerobic conditions (5). In contrast, the CCOQ mutant grown under semi-aerobic conditions possessed only 37% of the cytochrome c oxidase activity found in the wild type grown under the same conditions. The negative control strain CCON contained only marginal cytochrome c oxidase activity, which is most likely attributable to some residual aa₃ cytochrome c oxidase. When normalized to the cytochrome c oxidase activity present in the CCON mutant, the cbb₃ oxidase activity in the CCOQ mutant amounted to ~30% of that detected in the wild type, when both strains were grown under semi-aerobic conditions and reduced cytochrome c was employed in the assay of cytochrome c oxidase.

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Effect of ccoQ disruption on the integrity and activity of the cbb3 oxidase under various growth conditions. A, Western blot analysis using a polyclonal antibody against CcoO and CcoP. 50 µg of solubilized membrane protein isolated from R. sphaeroides strains grown under aerobic (30% O2), semi-aerobic (2% O2), or anaerobic dark Me2SO conditions (DMSO) was loaded onto each lane. B, cytochrome c oxidase activity was determined using cell-free crude extracts of R. sphaeroides strains grown semi-aerobically (2% O2) or anaerobically in the dark with Me2SO (Dark DMSO). The specific activity is expressed as micromoles/min/mg of protein. C, the CBB3 strains containing either pUI2803NHIS (lane 1) or pUI2803NHIS-CCOQ (lane 2) were grown under semi-aerobic conditions and 50 µg of each solubilized membrane protein was employed for Western blotting. The CcoN polypeptide was detected using anti-His4 monoclonal antibody (Qiagen) and the CcoO and CcoP proteins using the polyclonal antibody of CcoO and CcoP. Cells were grown aerobically by sparging with 30% O2, 69% N2, 1% CO2 to an A600 of 0.9-1.0 or semi-aerobically by sparging 2% O2, 97% N2, 1% CO2 to an A600 of 0.3-0.4.

When the cco operon with an in-frame deletion in ccoQ was overexpressed in the CBB3 mutant under 2% O₂ conditions using the plasmid pUI2803NHIS-CCOQ (see Table I for strain and plasmid definition), the negative effect of the ccoQ deletion on the accumulation of the CcoP and CcoO polypeptides in membrane fractions was partially overcome (Fig. 1C). This was anticipated based upon the known increase in cbb₃ oxidase when the intact ccoNOQP operon is expressed in trans (14). However, the levels of CcoO and CcoP in membranes were still less than those observed in the control strain CBB3 with pUI2803NHIS carrying the entire ccoNOQP operon. In this experiment we were able to probe the level of CcoN in the membrane by means of immunoblotting using anti-His₄ antibody. The removal of CcoQ affected to a lesser extent the accumulation of CcoN in the membrane as compared with the accumulation of CcoO and CcoP. Taken together, the results presented here demonstrate that removal of the CcoQ subunit affects both the activity and integrity of the cbb₃ oxidase under aerobic growth conditions, but not under anaerobic conditions. The magnitude of the effect, because of the removal of CcoQ, on the cbb₃ subunits under aerobic conditions, appears to be in the order of CcoP > CcoO > CcoN, as judged by the immunoblotting results (Fig. 1, A and C).

Characterization of the Purified Three-subunit cbb₃ Oxidase Lacking CcoQ-- The intact four-subunit cbb₃ oxidase (holoenzyme) and three-subunit cbb₃ oxidase (core enzyme) lacking CcoQ were purified from semi-aerobically grown CBB3 carrying pUI2803NHIS and pUI2803NHIS-CCOQ, respectively, as described under "Experimental Procedures." As shown in Fig. 2A, two bands of apparent molecular masses of 45 and 31 kDa, which correspond to the CcoN and CcoO subunits of the cbb₃ oxidase, respectively, were clearly visible in the stained SDS-PAGE gel. The CcoP band was more diffuse and less well stained than the other bands. SDS-PAGE using a Tris-Tricine gel revealed a protein band with a molecular mass of ~11 kDa in the purified holoenzyme, which was missing from the purified core enzyme, suggesting that it is the CcoQ polypeptide with a theoretical molecular mass of 7778 Da (Fig. 2A). Because of the difference in purity and heme content of the different oxidase preparations, the amounts of the core enzymes and holoenzymes used for comparative biochemical analyses were adjusted so that the band intensity of the catalytic CcoN subunit of each preparation was visually the same following Coomassie Blue staining of the SDS-PAGE gel. This is unavoidable because of the instability of the core enzyme. When approximately the same amount of CcoN was applied to an SDS-PAGE gel, the band intensity of CcoO from the core enzyme was similar to that of the holoenzyme (Fig. 2A). However, immunoblotting showed that the purified core enzyme contained a lower amount of CcoP than did the holoenzyme.

View larger version (34K): [in this window] [in a new window]   Fig. 2.   Properties of the purified core enzyme and holo-oxidases. A, 12 µg of the holoenzyme (lane 1) and 17 µg of the core enzyme (lane 2) purified from semi-aerobically grown cells were used for SDS-PAGE, giving the same band intensity for the CcoN polypeptide. Following SDS-PAGE the gel was vertically cut into three slices containing the same samples. One gel slice was stained with Coomassie Brilliant Blue, a second heme-stained with 3,3',5,5'-tetramethyl benzidine and H2O2, and the third used for Western blotting using a polyclonal antibody against CcoP. To resolve the small CcoQ polypeptide by SDS-PAGE, Tris-Tricine SDS-PAGE was performed (Tris-tricine). Quantitation of band intensity was performed using the program NIH Imager 1.62. The background-corrected values are placed below the photos. The value of each subunit of the holoenzyme is set at 100, and the relative values are expressed for the subunits of the core enzyme. B, cytochrome c oxidase activity and Km for reduced horse heart cytochrome c were measured spectrophotometrically. Both heme b and heme c contents were determined by using the pyridine hemochrome difference spectra. CO and CN reactivities (the ability to bind CO and CN) of the holo- and core enzymes were obtained employing the reduced plus CO minus reduced difference spectra and the oxidized minus oxidized plus CN difference spectra, respectively. All spectra were recorded in buffer A containing 0.01% (w/v) DM at room temperature using 100 µg/ml holoenzyme and 141 µg/ml core enzyme, which contained the same amount of CcoN. Except for the Km values, the values of the holoenzyme are set at 100 and the relative values are expressed for the core enzyme. Cbb3, purified holoenzyme; Cbb3-CCOQ, purified core enzyme. C, heme staining of the core oxidase purified from cells grown under anaerobic dark Me2SO conditions (lane 3). Holo- (lane 1) and core (lane 2) enzymes purified from semi-aerobically grown cells were subjected to heme staining as the controls. Protein concentration of each enzyme preparation was adjusted to give the same band intensity of CcoN after Coomassie Blue staining. Coomassie blue, Coomassie Blue-stained gel, Heme, heme-stained gel.

To compare the heme content of each subunit of each form of the oxidase, heme-staining analysis was performed. Protein samples were denatured under mild heating conditions prior to subjecting them to SDS-PAGE to visualize the CcoN subunit, which contains noncovalently bound b-type hemes. Heme staining showed that the CcoP and CcoN subunits of the core enzyme contained significantly decreased amounts of heme compared with the intact holoenzyme. A substantial decrease in the heme content of the CcoP subunit is undoubtedly caused, in part, by the reduced level of the CcoP apoprotein in the core enzyme. The heme content of CcoP was reduced to a greater extent than the CcoP apoprotein, as judged by comparison of the heme-stained gel and immunoblot. The heme content of CcoO in the core enzyme was also decreased when compared with the holoenzyme, but appeared to be less susceptible to loss than observed for the heme of CcoN and CcoP (Fig. 2A). In addition to the three heme-containing subunits of the cbb₃ oxidase, two bands of molecular masses 26 and 24 kDa were also observed after heme staining. The 24-kDa band is cytochrome c_y because heme staining showed this band to be missing in membranes derived from a cycY null mutant (data not shown). The 26-kDa band is either a degradation product of CcoP or a CcoP conformer, because this band could not be detected by heme staining of membrane fractions from the CBB3 mutant (data not shown) and it also cross-reacted with CcoP-specific antibody (Fig. 2A).

The core enzyme exhibited ~37% of the cytochrome c oxidase activity of the holoenzyme (Fig. 2B). In 20 mM potassium phosphate buffer (pH 7.5), the core enzymes and holoenzymes showed virtually the same K_m value for reduced horse heart cytochrome c (3.4 and 3.3 µM, respectively), indicating that both oxidase preparations have the same binding affinity for reduced cytochrome c regardless of the presence or absence of CcoQ. This observation rules out the possibility that reduced cytochrome c can donate electrons to both CcoO and CcoP. Both the heme b and heme c contents of the core enzyme and holoenzymes were determined by pyridine hemochrome analysis. The heme b content in the core enzyme was 71% of that observed for the holoenzyme. Similarly, the core enzyme contained 87% of the heme c detected in the holoenzyme. The heme c content of the core enzyme appears to be overestimated because the heme c content determined by spectroscopic analysis represents not only the heme c present in the CcoO and CcoP subunits, but also that of the copurified cytochrome c_y. Because of a lower purification yield of the core oxidase complex resulting from a lower cellular level of the core enzyme over the holoenzyme, contamination of the core oxidase with cytochrome c_y was always greater than that of the holoenzyme complex (Fig. 2A). The difference in the heme b content of the purified core enzyme and holo-oxidases estimated by heme-staining analysis appears greater than that determined by spectroscopic analysis. Because heme b is noncovalently bound to the CcoN subunit, there is a possibility that CcoN of the core enzyme may lose heme more readily during SDS-PAGE than does the holoenzyme. We should also point out that heme staining is at best semiquantitative, especially in the case of noncovalently bound heme. The content of the functional high spin heme b₃ was estimated by spectroscopic changes concomitant with CO binding to the reduced oxidases and CN binding to the oxidized enzymes. Reduced heme b₃ of the core enzyme bound 78% of the amount of CO observed for the holoenzyme. Similarly, the oxidized core enzyme showed 76% of the CN binding as compared with the holoenzyme. Both ligand-binding experiments indicated that the CcoN subunit of the core enzyme retained 76-78% of the functional high spin heme b when compared with the holoenzyme. The fact that the total heme b content of the core enzyme is 71% of that in the holoenzyme also suggests that both high and low spin hemes in CcoN of the core enzyme are compromised.

As shown in Fig. 2C, when the core enzyme was purified from the CBB3 with pUI2803NHIS-CCOQ grown anaerobically with Me₂SO, it contained significantly more heme b and heme c than the same core enzyme purified from semi-aerobically grown cells, again indicating that a likely role for CcoQ is to protect the heme moieties of the cbb₃ oxidase under aerobic conditions. Even in anaerobically grown cells, the heme content of the core enzyme was still lower than that in the holoenzyme. Because the purification of the oxidase was carried out in the presence of air, there is a likely possibility that some heme moieties were lost during purification. This possibility is addressed below.

Removal of CcoQ and the Stability of the cbb₃ Oxidase under Aerobic Conditions-- The induction of the ccoNOQP operon encoding the cbb₃ oxidase under semi-aerobic and anaerobic conditions is mediated by the global anaerobic activator FnrL (21). The fact that the CCOQ mutant showed similar levels of cbb₃ oxidase activity as the wild type under anaerobic dark Me₂SO conditions makes it highly unlikely that a decrease in cbb₃ oxidase activity in the CCOQ mutant grown under semi-aerobic conditions resulted from a decrease in the transcriptional activity of the ccoNOQP operon in the mutant. The decline in cbb₃ activity in the mutant appears to come from either the instability of, or an assembly defect in, the cbb₃ oxidase in the presence of O₂. To examine the former possibility, cell-free crude extracts of the wild-type and CCOQ mutant strains grown under anaerobic dark Me₂SO conditions were prepared, and the stability of the cbb₃ oxidase in crude extracts exposed to air was determined by measuring cbb₃ oxidase activity over time. As shown in Fig. 3, cbb₃ oxidase activity in cell-free crude extracts of the wild type remained unaltered after a 3-day incubation. In contrast, a relatively steep decline in cbb₃ activity was observed for the CCOQ mutant and after 3 days the activity was reduced to 39% of the initial activity and was apparently still declining. This result strongly suggested that the presence of O₂ directly leads to instability of the cbb₃ oxidase in the absence of the CcoQ subunit.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Effect of the absence of CcoQ on the stability of the cbb3 oxidase. Cell-free crude extracts of the wild-type (2.4.1) and CCOQ mutant strains were prepared from the cells grown anaerobically in the dark with Me2SO. 50 µl of crude extracts containing chloramphenicol and tetracycline to a final concentration of 0.005% (w/v) were placed in 1.5-ml Eppendorf tubes and incubated at room temperature. Crude extracts were assayed for cytochrome c oxidase activity at the times indicated. The enzyme activity is expressed as micromoles/min/mg of protein. All values provided are the average of two independent determinations.

To further investigate a possible role for CcoQ in protecting the cbb₃ oxidase from "oxidative" damage, the following analyses were performed. The core enzyme was purified from cells grown anaerobically with Me₂SO, and as a control the intact holoenzyme was isolated from semi-aerobically grown cells. As shown in Fig. 4A, the holoenzyme remained relatively stable as judged by both heme and Coomassie Blue staining of CcoN, CcoP, and CcoO after a 3-day incubation. The CcoN subunit of the core enzyme appeared relatively stable. In contrast, the heme content of CcoP and CcoO and the amount of the CcoO polypeptide of the core enzyme were decreased in a time-dependent manner to a significantly greater extent than observed for the holoenzyme. In particular, the heme content of CcoP was most severely affected. These results suggest that the loss of heme from the purified core enzyme is most likely associated with degradation of the CcoP and CcoO apoproteins, which is consistent with in vivo results (see Fig. 1, A and C).

View larger version (51K): [in this window] [in a new window]   Fig. 4.   In vitro stability analyses of the holoenzymes and core enzymes. A, 10 µl of each aliquot of the purified core oxidase (Cbb3-CCOQ, 17 µg) and holo-oxidase (Cbb3, 12 µg) was placed in the 1.5-ml Eppendorf tube and incubated over a period of 3 days at room temperature. The incubation times are indicated at the top of the gel figures. Samples were frozen rapidly at the indicated time points by placing them in ethanol-dry ice slurry and stored at 20 °C until SDS-PAGE analysis. The samples were subjected to SDS-PAGE followed by heme and Coomassie Blue staining. B, 10 µl of each aliquot of the purified core enzyme (Cbb3-CCOQ, 17 µg) and holo-oxidase (Cbb3, 12 µg) was placed in the 1.5-ml Eppendorf tube and incubated for 3 days at room temperature. C, control samples before incubation; none, control samples incubated for 3 days without treatment with argon and ascorbate; Ar, argon was flushed; ascorbate, sodium ascorbate was added to the samples to a final concentration of 10 mM; P inhibitors, the mixture of protease inhibitors was added to the samples (antipain-dihydrochloride, 50 µg/ml; bestatin, 40 µg/ml; Pefabloc SC, 500 µg/ml; pepstatin, 0.7 µg/ml). When the core enzyme was treated with each protease inhibitor, the samples were incubated for 4 days. Cysteine, Metal, Serine, Aspartate, the corresponding protease inhibitor was added to the samples to a final concentration as described above. C, the purified holo-oxidase (12 µg), core oxidase (17 µg), and a mixture of both purified oxidases (Holo+Core) were incubated for 4 days at room temperature. Samples were subjected to SDS-PAGE followed by heme staining.

To ascertain whether exposure to air (most probably O₂) led to the instability of CcoP and CcoO of the core enzyme, we performed the same experiment in the absence of air. To remove air, the 1.5-ml incubation tube was stoppered with a rubber septum, and gently flushed with argon gas for 2 min. Sodium ascorbate (pH 8.0) was also used to a final concentration of 10 mM to remove O₂ dissolved in the purified oxidase preparations because ascorbate serves as an electron donor for the cbb₃ oxidase during its turnover. After a 3-day incubation in the presence of argon and ascorbate, heme staining of the core enzyme revealed a strong CcoO band as well as a strong 26-kDa band. The 26-kDa band appeared to be derived from CcoP because this band cross-reacted with the CcoP antibody (see Fig. 2A), and it also appeared concomitantly with the disappearance of the CcoP band. The intact holoenzyme showed the same band pattern as the core enzyme when incubated in the presence of argon and ascorbate for 3 days. Although we do not know what caused the change in the migration behavior of CcoP on SDS-PAGE, the reducing conditions resulting from both the removal of air and the presence of ascorbate stabilized both CcoP and CcoO of the core enzyme. Neither argon flushing nor the addition of ascorbate alone was sufficient to stabilize the core enzyme, as judged from the intensity and location of the heme-stainable bands on the gel. As shown in Fig. 4B, a mixture of protease inhibitors protected the core enzyme even in the presence of air, indicating that proteolysis is responsible for the instability of the core enzyme under oxidizing conditions. To more accurately assess which type of protease was involved in degradation of the core enzyme under oxidizing conditions, we employed several individual protease inhibitors for the stability test (Fig. 4B). After a 4-day incubation, heme staining showed that bestatin and Pefabloc SC, which are inhibitors of metalloprotease and serine protease, respectively, stabilized the core enzyme. In contrast, neither pepstatin (aspartate protease inhibitor) nor antipain-dihydrochloride (cysteine protease inhibitor) showed any protection of the core enzyme under oxidizing conditions. The difference in stability between the holo- and core oxidases could be the result of a greater contamination by protease(s) of the purified core oxidase preparation compared with the purified holo-oxidase preparation. If this was the case, mixing the two preparations might be expected to affect the stability of the holo-oxidase. As shown in Fig. 4C, a mixture of the holo- and core oxidases remained as stable as the holo-oxidase alone, after a 4-day incubation. The core enzyme complex in the mixture appeared to be protected by the presence of the holoenzyme. In contrast, CcoP of the core oxidase was undetectable and the CcoO subunit was nearly absent after the same period of incubation of the core oxidase. This result suggests that it is not simply the difference in the amount of protease contamination in each of the purified preparations that results in the different stability of the holo- and core oxidases under oxidizing conditions. The protection effect of the holo-oxidase on the core oxidase leads us to speculate that the holo- and core oxidases might associate to form a supramolecular complex and the presence of the CcoQ subunit even in a stoichiometry of less than 1 relative to the core complex, affords protection of the core complex. Taken together, these results clearly indicate that the CcoQ subunit protects the core complex of the cbb₃ oxidase from degradation under oxidizing conditions and that a serine metalloprotease is most likely involved in the proteolytic degradation of the cbb₃ core complex in the absence of CcoQ. We can exclude the possibility that the core enzyme is destabilized in the process of catalytic turnover because the purified oxidase contains no electron donor.

Delineation of the Functional Portion of CcoQ-- When the amino acid sequence of R. sphaeroides CcoQ was multiply aligned with those CcoQ homologues from various organisms, only the central portions of the CcoQ homologues were found to be well conserved (Fig. 5A). This central region also shows some homology to the smallest subunit (CtaH) of the aa₃ cytochrome c oxidase of P. denitrificans (7). According to the resolved crystal structure of the P. denitrificans aa₃ oxidase, this region comprises the membrane-spanning region of CtaH and the COOH terminus of CtaH faces the periplasm, implying that CcoQ might have a similar topology in the cytoplasmic membrane as CtaH (6).

View larger version (54K): [in this window] [in a new window]   Fig. 5.   Multiple alignment of the CcoQ homologues and CtaH of P. denitrificans (A) and determination of the minimal length of CcoQ capable of complementing the CcoQ mutant of R. sphaeroides (B). Identical or conservatively substituted residues are highlighted by a gray background. The numbers on the right of the aligned sequences indicate the length of the polypeptides deduced from the corresponding nucleotide sequences. pBBRQ is the pBBR1MCS2-based plasmid harboring the wild-type ccoQ gene. pBBRQ-TGA1 through pBBRQ-TGA4 are derivatives of pBBRQ, and each contains the ccoQ gene with a nonsense mutation encoding the truncated form of CcoQ. a, levels of spectral complex in R. sphaeroides strains grown under aerobic conditions. b, -galactosidase activities in aerobically grown R. sphaeroides strains bearing the puf::lacZ transcriptional fusion plasmid (pUI1830). c, cytochrome c oxidase activities in R. sphaeroides strains grown semi-aerobically. Strains were grown aerobically (30% O2) to an A600 of 0.3-0.4 or semi-aerobically (2% O2) to an A600 of 0.5-0.6. P.d., P. denitrificans; R.c., Rhodobacter capsulatus; V.c., Vibrio cholerae; B.j., B. japonicum; S.m., S. meliloti; R.s., R. sphaeroides.

Among the CcoQ homologues, the CcoQ polypeptide of R. sphaeroides deduced from its nucleotide sequence is longer than the others (Fig. 5A). To define the COOH-terminal boundary of the functional portion of CcoQ as judged by cbb₃ stabilization, a series of COOH-terminal deletion derivatives of CcoQ were constructed by replacing the codons for Ser-60, Asp-56, Thr-51, or Arg-48 of CcoQ by the TGA stop codon. The ccoQ gene carrying each nonsense mutation was introduced into the CCOQ mutant using the vector pBBR1MCS2, and its functionality was assessed by complementation.

Cytochrome c oxidase activity was determined in the mutant strains as well as the control strains grown under 2% O₂ conditions, where the cbb₃ oxidase is the predominant cytochrome c oxidase and where the removal of CcoQ severely affects cbb₃ oxidase activity. As shown in Fig. 5B, the negative control strain CCOQ (pBBR1MCS2) showed approximately half of the cytochrome c oxidase activity present in the control strain CCOQ (pBBRQ). The deletion of 8 and 12 amino acids from the COOH terminus of CcoQ did not affect cbb₃ oxidase activity in CCOQ (pBBRQ-TGA1) and CCOQ (pBBRQ-TGA2), respectively. In contrast, the cbb₃ oxidase activity in the CCOQ (pBBRQ-TGA4) strain showed oxidase levels similar to those of the negative control strain CCOQ (pBBR1MCS2), indicating that the removal of 20 amino acids from the COOH terminus of CcoQ makes CcoQ nonfunctional. The CCOQ (pBBRQ-TGA3) strain, in which 17 amino acids were removed from the COOH terminus of CcoQ, retained 81% of the cytochrome c oxidase activity detected in the positive control strain CCOQ (pBBRQ) and is therefore considered to be partially functional.

We previously demonstrated that the extent of electron flow through the cbb₃ oxidase is inversely related to the level of spectral complexes as well as to the expression of PS genes, which are under control of the PrrBA two-component system (14). The positive control strain CCOQ (pBBRQ), as well as the mutant strains CCOQ (pBBRQ-TGA1) and CCOQ (pBBRQ-TGA2) containing the fully active cbb₃ oxidase, produced only basal levels of the light harvesting complexes under 30% O₂ conditions like the wild type 2.4.1 (pBBR1MCS2). In contrast, substantial levels of the light harvesting complexes were synthesized in CCOQ (pBBRQ-TGA4) and the negative control strain CCOQ (pBBR1MCS2). Only marginal increases in spectral complex levels were observed in CCOQ (pBBRQ-TGA3).

As anticipated, the puf operon, which is regulated by the PrrBA two-component system, was derepressed 4.3-fold in the negative control strain CCOQ (pBBR1MCS2) grown under 30% O₂ conditions when compared with the CCOQ (pBBRQ) strain grown under the same conditions. Similarly, the CCOQ (pBBRQ-TGA4) strain showed a 3.8-fold increase in puf expression. The CCOQ (pBBRQ-TGA3) exhibited only marginal derepression in puf expression under the same growth conditions compared with the control strain CCOQ (pBBRQ). The introduction of the B. japonicum fixQ gene into the CCOQ mutant led to complementation of the CcoQ-minus phenotype as judged by puf operon expression under aerobic conditions (data not shown).

Taken together, the results clearly suggest that the minimum length of CcoQ required for protection of the cbb₃ terminal oxidase is 48-50 amino acids, which is consistent with the fact that the CcoQ homologues of P. denitrificans and Sinorhizobium meliloti consist of 48 and 50 amino acid residues, respectively. This observation also suggests that the CcoQ subunit may play a similar role in these strains. These results also confirmed that the "functional state" of the cbb₃ oxidase is inversely proportional to the extent of aerobic derepression of those PS genes that are regulated by the PrrBA two-component system.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Bacterial quinol and cytochrome c oxidases belonging to the heme-copper superfamily are generally composed of four different subunits. The smallest subunit IV differs in size and membrane topology depending on the type of oxidase. The aa₃ cytochrome c oxidase from P. denitrificans contains subunit IV (CtaH) with a single membrane-spanning helix (6). This is also the case for the cbb₃ cytochrome c oxidase (11). In contrast, subunit IV of the well studied bo₃ quinol oxidase of E. coli has three membrane-spanning helices and is larger than those corresponding to the aa₃ and cbb₃ cytochrome c oxidases (22, 23). According to the three-dimensional crystal structures of the P. denitrificans aa₃ cytochrome c oxidase (6) and the E. coli bo₃ quinol oxidase (24), subunit IV is positioned in the cleft between subunits I and III. However, the crystal structure of the mitochondrial aa₃ cytochrome c oxidase consisting of 13 different subunits reveals no obvious counterpart to the bacterial subunit IV (25).

The removal of subunit IV from the bo₃ quinol oxidase leads to inactivation of the enzyme and defects in the heme-copper binuclear center of subunit I (26). Likewise, a deletion of the Bacillus subtilis qoxD encoding subunit IV of the aa₃ menaquinol oxidase was shown to significantly reduce respiration and proton pumping (27). In contrast, deletion of ctaH in P. denitrificans was reported to have no effect on the spectral and enzymatic properties of the aa₃ cytochrome c oxidase (7).

In this report, we first demonstrated that subunit IV (CcoQ) of the cbb₃ cytochrome c oxidase plays a role in protecting the cbb₃ oxidase under aerobic conditions. The rationale for this argument is as follows. (i) Subunit abundance and cbb₃ oxidase activity in the CCOQ mutant were significantly reduced under aerobic growth conditions (2 and 30% O₂), but not under anaerobic dark Me₂SO or photosynthetic conditions (5) when compared with measurements of the wild-type strain grown under the same conditions. (ii) The activity of the cbb₃ oxidase in cell-free crude extracts of the CCOQ mutant was unstable when exposed to air, whereas the cbb₃ activity in cell-free crude extracts of the wild type was stable under the same experimental conditions. (iii) When the purified core enzyme and holoenzymes were compared regarding their stability in the presence of air, the absence of CcoQ was directly related to the stability of both the CcoP and CcoO subunits, as well as the heme content of these subunits. Reducing conditions resulting from the removal of air and the addition of ascorbate stabilized the hemes of the core enzyme.

The absence of CcoQ appears to affect the CcoP subunit of the cbb₃ oxidase most severely under oxidizing conditions, as judged by the following observations. (i) The most significant reduction in the heme content was observed for CcoP of the purified core enzyme (see Fig. 2, A and C). The stoichiometry of CcoP to CcoN and CcoO in the purified core enzyme is altered compared with that of the holoenzyme (see Fig. 2A). (ii) The level of CcoP in the membranes of the CCOQ mutant grown under aerobic conditions was most severely affected (see Fig. 1A). (iii) In vitro stability tests showed that the loss of heme from CcoP of the purified core enzyme occurred to the greatest extent compared with the other subunits (see Fig. 4A).

In addition, our data suggest that a serine metalloprotease activity is related to the instability of the cbb₃ core complex lacking the CcoQ subunit. The proteolytic degradation of the core enzyme occurred only in the presence of O₂, indicating that the presence of O₂ triggers the degradation process of the core enzyme lacking the CcoQ subunit. At this point we are faced with the question of why the core enzyme is degraded only under oxidizing conditions. A plausible answer can be drawn from the observation that, although the purified core oxidase consists of the same three subunits having the same molecular mass as those of the holoenzyme, it shows decreased levels of heme b as well as heme c when compared with the same amount of the holoenzyme (see Fig. 2, A and C). This finding leads us to hypothesize that, in the presence of O₂, the loss of the heme moieties in the core oxidase occurred prior to degradation of the apoproteins, i.e. the loss of heme would be expected to convert the cbb₃ oxidase into a conformationally altered, degradable protein, which presumably becomes susceptible to proteolysis.

The complementation experiments using a series of COOH-terminal deletion derivatives of CcoQ demonstrated that the minimum length of CcoQ required for stabilizing the cbb₃ oxidase under aerobic conditions is ~48-50 amino acids. All CcoQ homologues that have been so far identified consist of 48-73 amino acids, implying that the minimum size of the functional CcoQ homologues may correspond to the 48-amino acid CcoQ homologues when these are multiply aligned (see Fig. 5A). As shown in Fig. 5A, there are a number of conserved amino acid residues in the first 48 amino acids. It is possible that an analysis of these may shed light on the mechanism of CcoQ protection of the core complex, perhaps through an alteration of its binding to the core enzyme.

The cbb₃ cytochrome c oxidase of R. sphaeroides has multiple functions, i.e. as a terminal oxidase in the presence of O₂ (19), as an O₂/redox sensor (1, 14) and finally as a conduit for reducing power under anaerobic conditions (5). Under aerobic conditions the cbb₃ oxidase generates the signal that is transmitted to the PrrBA two-component system, most likely via PrrC, to repress PS gene expression (1, 15). We have previously demonstrated that the cbb₃ oxidase does not sense O₂ itself, but it is the extent of electron flow through the oxidase that serves to generate the inhibitory signal (14). The greater the volume of electron flow through the cbb₃ oxidase, the stronger the inhibitory signal generated by the oxidase to shift the equilibrium of PrrB activity toward the phosphatase mode and away from kinase mode, resulting in the absence of activation of PS genes. Therefore, under aerobic conditions where the substrate for the cbb₃ oxidase (O₂) is sufficient, PS genes are not induced. The complementation experiment using the truncated forms of CcoQ reconfirms that the functional state (activity) of the cbb₃ oxidase is inversely related to the extent of aerobic derepression of those PS genes that are under the control of the PrrBA two-component system.

Although the CCOQ mutant has near-normal cbb₃ oxidase activity under anaerobic photosynthetic and dark Me₂SO conditions, it produces significant levels of the spectral complexes under highly aerobic conditions like the various Cco null mutant strains (5). This phenotype originally suggested to us that the CcoQ subunit was part of the signaling pathway from the cbb₃ oxidase to the PrrBA two-component system. However, noting the stabilizing effect of CcoQ on the cbb₃ oxidase in the presence of O₂, as described here, provides an alternative explanation. The cbb₃ oxidase, as we have shown, is severely affected in both its cellular levels and activity in the CCOQ mutant grown under aerobic conditions. This destabilization of the altered cbb₃ oxidase attenuates the inhibitory signal to the PrrBA two-component system, resulting in the aerobic formation of the spectral complexes.

The cbb₃ cytochrome c oxidase is considered to be a primitive form of terminal oxidase. It is suggested that the cbb₃ oxidase evolved from the anaerobic enzyme, NO reductase on the basis of the similarity in the catalytic subunit and heme composition (3). The high affinity of the cbb₃ oxidase for O₂ leads us to hypothesize that this form of oxidase evolved during the transition period from an anoxygenic to an oxygenic atmosphere to enable the more efficient aerobic respiration. However, such an "anaerobic" enzyme system might have been sensitive to oxygen even at low levels. Therefore, subunit IV might have been "recruited" to protect the enzyme complex from oxidative damage. If the aa₃ cytochrome c oxidase evolved from a cbb₃-type oxidase, subunit IV of the bacterial aa₃ oxidase may be an "evolutionary remnant."

	ACKNOWLEDGEMENT

We thank Dr. Hans-Martin Fischer for providing the plasmid pRJ4504 carrying the fixQ gene of B. japonicum.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grant GM15590 (to S. K.).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 Microbiology and Molecular Genetics, University of Texas Health Science Center, Medical School, 6431 Fannin, Houston, TX 77030. Tel.: 713-500-5502; Fax: 713-500-5499; E-mail: samuel.kaplan@uth.tmc.edu.

Published, JBC Papers in Press, February 25, 2002, DOI 10.1074/jbc.M200198200

	ABBREVIATIONS

The abbreviations used are: PS, photosynthesis; DM, n-dodecyl -D-maltoside; SIS, Sistrom's medium A; Tricine, N-[2-hydroxy-1,1-bis(hydroxymethyl)ethyl]glycine.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit