Expression, Purification, and Characterization of Recombinant Forms of Membrane-Bound Cytochromec-550nm fromBacillus subtilis

Protein Expression and Purification 15, 69 –76 (1999) Article ID prep.1998.1001, available online at http://www.idealibrary.com on

Expression, Purification, and Characterization of Recombinant Forms of Membrane-Bound Cytochrome c-550nm from Bacillus subtilis Pamela S. David,* Megan R. Morrison,* Sui-Lam Wong,† and Bruce C. Hill*,1 *Department of Biochemistry, Queen’s University, Kingston, Ontario, Canada K7L 3N6; and †Department of Biological Sciences, University of Calgary, 2500 University Drive N.W., Calgary, Alberta, Canada T2N 1N4

Received July 7, 1998, and in revised form October 13, 1998

Bacillus subtilis expresses a cytochrome c-550nm that participates in respiratory electron transfer and is an integral membrane protein. Analysis of the B. subtilis cytochrome c-550nm amino acid sequence predicts a single N-terminal transmembrane helix attached to a water-soluble heme binding domain [C. von Wachenfeldt and L. Hederstedt (1990) J. Biol. Chem. 265, 13939 – 13948]. We have purified cytochrome c-550nm from wildtype B. subtilis and B. subtilis transformed with the shuttle vector pHP13 containing the gene for B. subtilis cytochrome c-550nm (cccA). In B. subtilis transformed with pHP13/cccA there is better than eightfold more membrane-bound cytochrome c-550nm than in wildtype B. subtilis. The overexpressed cytochrome c-550nm can be purified by chromatography on hydroxylapatite and Q-Sepharose media. A six-histidine tag has been added to the C-terminus of cytochrome c-550nm from B. subtilis as a further aid for purification. This strain produces cytochrome c-550nm to a level fourfold greater than wild type and allows for one-step purification using metal affinity chromatography. UV-Vis spectroscopy detects no change in the heme C spectrum due to the addition of six histidines. Neither form of B. subtilis cytochrome c-550nm is stable in its reduced state in aerated buffer, unless EDTA is added. The two forms, wild-type and his-tagged, of cytochromes c have similar midpoint redox potentials of 195 and 185 mV, respectively, and are equally good substrates for B. subtilis cytochrome c oxidase. We conclude that the addition of the histidine tag eases the purification of cytochrome c-550nm from B. subtilis plasma membranes and that the additional metal binding site does not compromise the stability or functional properties of the protein. © 1999 Academic Press

1 To whom correspondence should be addressed. Fax: (613) 5452497. E-mail: [email protected].

1046-5928/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.

Mitochondrial cytochrome c is one of the best-studied proteins (2). These well-known cytochromes c are water-soluble proteins that are essential as electron carriers from the cytochrome bc1 complex to the cytochrome c oxidase complex in the mitochondrial respiratory chain. Mitochondrial cytochrome c has a single heme prosthetic group that is covalently attached to a protein with a molecular weight of about 12 kDa. The protein component has a highly basic character that is conferred by an abundance of lysine residues. Furthermore, the distribution of these lysines on the surface of cytochrome c gives the molecule a dipole moment and is important in the functional orientation for electron transfer. It is proposed that reduction and oxidation of the heme occur via an edge of the heme that is exposed on the protein surface (3). This exposed edge is surrounded by a set of lysine residues that aid in the formation of protein–protein complexes between cytochrome c and cytochrome bc1 (4) and cytochrome c oxidase (5). Prokaryotic cytochromes c are not as well described. In Bacillus subtilis three c-type cytochromes have been identified: cytochrome c-550nm (6) and the cytochrome c domains in cytochrome bc1 (7) and cytochrome caa3 (8) complexes. Cytochrome c-550nm may be functionally analogous to water-soluble mitochondrial cytochrome c. Structurally, however, B. subtilis cytochrome c-550nm is only 14% identical when aligned with horse heart cytochrome c (9). Hydropathy analysis of B. subtilis cytochrome c-550nm indicates the presence of a single transmembrane domain that confers its integral membrane nature. This analysis predicts that the heme C group is covalently attached to an aqueous domain that is located on the external surface of the B. subtilis plasma membrane. In addition, the heme-binding domain of B. subtilis cytochrome c-550nm has an altered balance of charged amino acids giving an acidic 69

70

DAVID ET AL.

pI of 5.44 compared to basic mitochondrial cytochrome c (e.g., horse cytochrome c, pI 9.59). This is of particular interest because of the recognized role of surface lysine residues in the interaction of mitochondrial cytochrome c with its redox partners (4,5). The electron transport chain in B. subtilis is branched with the split occurring at the level of its quinol, menaquinone-7. Menaquinol is able to donate electrons directly to menaquinol oxidase, an aa3-type menaquinol:oxygen oxidoreductase (10). In addition, menaquinol is presumably an electron donor to the cytochrome c oxidase pathway via a cytochrome bc1 complex (7). In B. subtilis, our working model of the cytochrome c oxidase pathway consists of the cytochrome bc1 complex, cytochrome c-550nm, and the cytochrome c oxidase, cytochrome caa3. B. subtilis cytochrome c-550nm is proposed to play a role in electron transfer from the cytochrome bc1 complex to cytochrome caa3, by analogy with the mitochondrial electron transport chain. However, the presence of a covalently bound cytochrome c domain in the B. subtilis cytochrome c oxidase complex suggests a different interaction between this complex and cytochrome c-550nm compared to their mitochondrial counterparts. Recent studies have shown that overexpression of B. subtilis cytochrome c-550nm is possible from a plasmid-encoded copy (11). We have purified cytochrome c-550nm from an overexpressing strain of B. subtilis and have used PCR-based mutagenesis to add a six-histidine tail onto the C-terminal end of B. subtilis cytochrome c-550nm. We are able to purify up to 10 mg of cytochrome c-550nm per liter from these overexpressing strains and this will allow for a more thorough physical characterization of this member of the cytochrome c family. MATERIALS AND METHODS

The restriction enzyme HindIII and Chelating Fast Flow Sepharose were purchased from Pharmacia Biotech (Baie d’Urfe, PQ, Canada). All other restriction enzymes, DNA-modifying enzymes, Taq DNA polymerase, ultrapure agarose, and low-melting-point agarose (electrophoresis grade) were purchased from Gibco BRL (Gaithersburg, MD). The deoxynucleotide triphosphates were purchased from Perkin–Elmer Cetus (Norwalk, CT). All other chemicals were of analytical grade from commercial suppliers. The B. subtilis cccA gene in the Escherichia coli/B. subtilis shuttle vector pHP13 was a kind gift from Dr. L. Hederstedt (University of Lund, Sweden) (11). The six-histidine tag was added onto B. subtilis cccA via PCR (12). The primer pair used was a 47-base-pair forward primer (59-GGCAAGCTTGAGCTCCCCTTTATTTTACTGAAAAATGATGTCATTTGC) and a 63-basepair reverse primer (59-CGGGTACCGGATCCCTATTA-

ATGGTGATGGTGATGGTGTTTAATTTTTGACACCCACTCTGCC). The forward primer contains a HindIII and SacI site and the reverse primer contains the six histidine codons (underlined) plus two stop codons and BamHI and KpnI sites. The primers were synthesized at the Core Facility for Protein/DNA Chemistry at Queen’s University. PCR contained 0.5 mg of pHP13/ cccA as template, 0.5 mM each of the two primers, 1.5 mM MgCl2, and 200 mM dNTPs in a 50-ml reaction volume. The hot-start method of PCR was used (13). The cycle conditions were 95°C for 1 min, 60°C for 1 min, and 72°C for 1 min, 25 cycles, with a final 5-min elongation at 72°C using a Techne Progene thermocycler. The PCR-mutagenized cccA gene and plasmid pHP13 (14) were restriction digested using HindIII and BamHI. The restricted DNA was run on a lowmelting-point agarose gel and ligated overnight. Ligated DNA was transformed into E. coli DH5a made competent using the CaCl2 method (15). Plasmid purification from E. coli was carried out as in (15). Plasmid transformation into B. subtilis was according to (16,17). B. subtilis plasmid purification was carried out as in (18). The DNA sequence of histidine-tagged cccA was confirmed using the ABI PRISM dye terminator kit as performed by the Core Facility at Queen’s University. The expression of cytochrome c-550nm from the plasmid pHP13/cccA or phiscccA was in B. subtilis DB104 (his nprR2 nprE18 apr 3) (19). Strains generated were BSC1 (pHP13/cccA in DB104) and BSCH1 (phiscccA in DB104). All B. subtilis strains were grown on superrich media in 1-liter cultures at 37°C with overnight shaking at 250 rpm in 2.8-liter Fernbach flasks (20). Cultures of strains containing either plasmid were supplemented with 15 mg chloramphenicol per milliliter. B. subtilis cultures were harvested and membrane proteins extracted as in (21). Spectra were recorded on a Shimadzu UV-160 spectrophotometer or a Hewlett–Packard 8452A diode-array spectrophotometer. The protein concentration was determined (22) and purified protein samples were run on SDS–PAGE (21). Detection of the heme C moiety of cytochrome c on SDS–PAGE was via a peroxidase reaction (23). Redox potentiometry was carried out in a sealed double-arm cuvette constructed for this experiment (24). The reaction mix contained 3 mM cytochrome c plus 20 mM phenazine methosulfate and 20 mM diaminodurine in 50 mM NaH2PO4 at pH 7.0 with 0.5 mg dodecyl maltoside per milliliter and was flushed continuously with argon. Voltage readings were recorded with an Ag/AgCl electrode connected to a Corning Model 255 ion meter. Reactions were carried out at 25°C and titrated with either sodium dithionite or potassium ferricyanide.

B. subtilis CYTOCHROME c-550nm

FIG. 1. Representation of plasmid phiscccA. The 680-base-pair histidine-tagged PCR fragment of cytochrome c was cloned into the plasmid pHP13 using HindIII and BamHI. This altered gene is under the control of the native cytochrome c promoter. Also indicated are the two antibiotic resistance genes, erythromycin and chloramphenicol, and the B. subtilis and E. coli origins of replication. The plasmid is 5580 base pairs in length.

RESULTS

Cloning, DNA Sequence, and Expression Two primers were constructed to allow the addition of a hexahistidine tag to the C-terminus of cytochrome c-550nm. The reverse primer contained six histidine codons, two stop codons, and two restriction enzyme sites. The forward primer contained two restriction enzyme sites. The DNA sequence for the histidine codons is an alternating series of two codons used in B. subtilis for histidine. The reverse primer was constructed based on the plasmid pUSH2 (25). The 680base-pair PCR product was cloned into the shuttle vector pHP13 using the HindIII and BamHI sites in the forward and reverse primers, respectively (Fig. 1). Purified plasmid phiscccA was transformed into B. subtilis DB104. The sequence of cccA in plasmid phiscccA was identical to that published with the addition of the six-histidine tag (1). Reduced minus oxidized difference spectra, in the visible region, of membrane extracts from B. subtilis DB104, BSC1, and BSCH1 show the relative levels of cytochrome c-550nm in each strain (Fig. 2). There are three prominent peaks in this spectral region that arise from the cytochromes in the membrane extract. The peak centered at 520 nm has contributions from the b-bands of each of the B-, C-, and A-type hemes of the cytochromes. The peak near 600 nm is the main visible transition, or a-band, of the cytochromes containing A-type heme (i.e., cytochrome aa3 and cytochrome caa3). The band at 560 nm is the a-band of

71

B-type heme such as found in the cytochrome bc1 complex. The peak at 550 nm is the a-band of the cytochromes c including cytochrome c-550nm. This peak also includes contributions from the cytochrome c domain of cytochrome caa3 and also from cytochrome c1 of the cytochrome bc1 complex. Strains of B. subtilis transformed with the plasmid pHP13/cccA express extra cytochrome c-550nm. The BSC1 strain which carries the wild-type cytochrome c-550nm gene on the pHP13 vector expresses about threefold higher total cytochrome c per gram of membrane protein. The content of cytochrome c in wild-type membranes is composed of at least three c-type cytochromes, and about one-third of this is cytochrome c-550nm. We estimate, therefore, better than eightfold overexpression of cytochrome c-550nm. This level of overexpression is similar to that obtained with this plasmid in the B. subtilis strain 3G18 (6). The strain carrying the histidinetagged form of cytochrome c-550nm, BSCH1, also overexpresses cytochrome c-550nm relative to wild type but less than BSC1 (see Fig. 2). Purification Purification of cytochrome c-550nm from wild-type B. subtilis is difficult and requires large volumes of culture, e.g., 300 liters, to obtain small amounts of partially pure material (1). Our own purification attempts beginning with 4 liters of wild-type B. subtilis yielded only 0.8 mg of pure cytochrome c-550nm. We have, therefore, sought an alternative approach to obtaining larger amounts of purified cytochrome c-550nm that would allow more detailed physical analysis. The

FIG. 2. Dithionite-reduced minus oxidized difference spectra of B. subtilis membrane extracts. The membranes are extracted in 50 mM Tris–HCl, 1 M NaCl, 1 mM EDTA, 5% (v/v) Triton X-100, pH 7.4. The volume of the extraction buffer was adjusted according to the wet weight of cells obtained in the culture. The spectra are designated as —, BSC1; – - –, BSCH1; and - - -, DB104.

72

DAVID ET AL.

TABLE 1 Purification of Plasmid Expressed Cytochrome c from 4-Liter Cultures of B. substilis

Membrane extract Step 1c Step 2d Step 3e

Sample

Total protein (mg)

Cyto c f (nmol)

Cyto c/protein (nmol/mg)

Fold purification

Percentage yield

BSC1a BSCH1b BSC1 BSCH1 BSC1 BSCH1 BSC1 BSCH1

854 1414 202 37 105 24 49.6 10.4

2790 1225 1863 1040 1779 830 1752 421

3.26 0.87 9.21 28 16.9 34 35.3 40.4

1 1 2.8 32 5.2 40 10.9 46.4

100 100 67 85 67 68 63 34

a

Plasmid-expressed cytochrome c. Histidine-tagged plasmid expressed cytochrome c. c Step 1 for BSC1 is a hydroxylapatite column; for BSCH1, a Ni21-bound metal affinity column. d Step 2 for BSC1 is a Q-Sepharose column; for BSCH1, a Ni21-bound metal affinity column. e Step 3 for BSC1 is four Q-Sepharose columns; for BSCH1, a Ni21-bound metal affinity column. f Cyto c is cytochrome c. b

development of strains of B. subtilis overexpressing cytochrome c-550nm has enhanced its purification. Table 1 shows the amount and purity levels of cytochrome c-550nm at different stages in its isolation from strains BSC1 and BSCH1. The amount of heme C can be readily determined at each stage of the purification from the visible spectrum. Purification of cytochrome c-550nm from BSC1 is accomplished by passage over hydroxylapatite followed by several rounds of anionexchange chromatography on Q-Sepharose. Typically five individual Q-Sepharose columns are required for complete purification. By comparison, purification of histidine-tagged cytochrome c-550nm from BSCH1 yields a product that is completely free of other cytochromes in a single stage of metal ion affinity chromatography using nickel-charged Chelating Fast Flow Sepharose. The use of other forms of chromatography than that outlined in Table 1 with either preparation does not improve the purifications. The loss of cytochrome c-550nm at each stage is perhaps higher than anticipated. We ascribe this to the fact that we are purifying an integral membrane protein and as such the purification is of protein– detergent micelles and these are probably heterogeneous in their chemical and physical properties. We have not found detergent conditions that give a better result than those reported, although it is plausible that more favorable detergent conditions could exist. A typical chromatogram for passage of a membrane extract from BSCH1 over a nickel column is shown in Fig. 3. In this chromatogram the absorbance at 280 and 405 nm is shown over the course of the elution profile. A fraction of proteins in the membrane extract do not bind to the Ni21 column and these include heme B- and heme C-containing cytochromes. Two peaks are eluted when the imidazole concentration is increased. The

first peak appears at 80 mM imidazole and includes all of the heme A-containing proteins. Chromatography of wild-type B. subtilis membrane extracts on Ni21charged metal ion affinity resin is identical to this point and we are now using this step in our procedure for purification of cytochrome aa3 and cytochrome caa3. The second peak from the BSCH1 strain elutes at 200 mM imidazole and contains only c-type cytochrome. Purification of the material in this fraction to homogeneity, as defined by a single band on SDS– gel electrophoresis, requires two more passages over the metal affinity column. The overall yield of cytochrome c-550nm per volume of culture is higher from BSC1, but obtaining the homogeneously pure material is easier from BSCH1 (see Table 1). Characterization Sodium dodecyl sulfate–polyacrylamide gel electrophoresis of horse cytochrome c and B. subtilis cytochrome c-550nm, in the native and histidine-tagged forms, are compared in Fig. 4. A single band is seen for each of these proteins with apparent molecular weights based on their relative mobilities of 13,400 kDa for horse, 14,000 kDa for native B. subtilis, and 14,400 kDa for histidine-tagged B. subtilis. Although the apparent molecular weights do not agree exactly with the known molecular weights from the protein sequence, they are in the proper relative order. The size difference between native and histidine-tagged B. subtilis cytochrome c-550nm, attributed to the added six histidine residues, is easily distinguished. Neither B. subtilis proteins run as a sharp band when compared to horse cytochrome c and this is also true of cytochrome c-550nm purified from wild-type B. subtilis. When duplicate SDS–PAGE gels are stained independently for

B. subtilis CYTOCHROME c-550nm

73

FIG. 3. Elution profile of membrane extract of BSCH1 off the Ni21-charged metal ion affinity column. The column was run in 1% (v/v) Triton X-100, 20 mM NaH2PO4, 200 mM NaCl. —, A405 nm absorbance; - - -, A280 nm absorbance; – – –, imidazole (scale on right-hand axis).

the presence of heme and protein, the staining patterns overlap. This indicates that the heme, as expected, remains covalently attached to the protein and that the diffuse electrophoretic mobility is an intrinsic property of the B. subtilis cytochrome c-550nm and not due to the presence of contaminating proteins. The two forms of B. subtilis cytochrome c-550nm have nearly identical spectra in the reduced and oxidized states indicating that the addition of the histi-

FIG. 4. SDS–PAGE of cytochromes c. Lane 1, molecular weight standards (ovalbumin, carbonic anhydrase, trypsin inhibitor, lysozyme, aprotinin, insulin b chain); lane 2, 2 mg horse cytochrome c; lane 3, 8 mg B. subtilis cytochrome c-550nm; lane 4, 8 mg histidinetagged B. subtilis cytochrome c-550nm.

dine tag does not alter the environment of the heme. The spectrum of horse cytochrome c is slightly different in both reduced and oxidized states from the B. subtilis proteins. In addition, a band at 695 nm is present in both versions of B. subtilis cytochromes c-550nm which indicates the presence of a methionine residue as the sixth axial ligand to heme C (26,27). Potentiometric titration was used to determine the midpoint potential for the two B. subtilis cytochromes c-550nm and compare them to horse cytochrome c (Fig. 5). The slopes are close to 60 mV per decade of redox change consistent with a single electron reduction event. The midpoint potentials are 195 and 185 mV for native and histidine-tagged B. subtilis cytochromes c-550nm, respectively. These values are close to those reported previously for B. subtilis cytochrome c-550nm and are 70 to 80 mV more electronegative than horse heart cytochrome c at 265 mV (6,28,29). It is a characteristic feature of respiratory cytochromes c that they are stable in the reduced state in the presence of oxygen (30,31). We have assessed the autooxidation rate of both forms of B. subtilis cytochrome c-550nm as isolated. In these assays B. subtilis cytochrome c-550nm is initially reduced with ascorbate, the excess reductant removed by gel filtration, and the stability of the reduced form assessed spectrophotometrically. The histidine-tagged B. subtilis cytochrome c-550nm, as isolated, was unstable in the reduced state with a half-life for autooxidation of 2–5 min. Native B. subtilis cytochrome c-550nm, as isolated, was more stable with a half-life for autooxidation of 50 min. Both B. subtilis cytochromes c-550nm could

74

DAVID ET AL.

with cytochrome caa3. Figure 6B compares the activity of bovine cytochrome c oxidase using either horse or B. subtilis cytochromes c. The B. subtilis cytochromes c-550nm are poor substrates compared to horse cytochrome c and indistinguishable from one another. DISCUSSION

FIG. 5. Redox titration of cytochromes c. In each case 3 mM cytochrome c was reduced by the addition of dithionite and then titrated back to the oxidized state by small additions of potassium ferricyanide. The buffer was 50 mM NaH2PO4, pH 7.4, with 0.5 mg of dodecyl maltoside per milliliter. Horse cytochrome c, squares; B. subtilis cytochrome c-550nm, circles; histidine-tagged B. subtilis cytochrome c-550nm, triangles.

be stabilized by addition of EDTA to the incubation medium. In the presence of EDTA they have identical autooxidation half-lives of 140 min. EDTA also effects the autooxidation of horse cytochrome c. Horse cytochrome c in the absence of EDTA has a half-life for autooxidation of 120 min (30) which is increased to 600 min by addition of EDTA. The effect of EDTA implies that autooxidation of each of the cytochromes c is catalyzed by contaminating metal ions. The increased rate of autooxidation of histidine-tagged cytochrome c-550nm in the absence of EDTA could be due to the presence of a metal ion bound by the histidine tag. The autooxidation half-lives of both forms of B. subtilis cytochrome c-550nm in the presence of EDTA reached the same level of stability, but are still less stable than that of the horse protein. In the presence of EDTA, B. subtilis cytochrome c-550nm is stable enough for use as a substrate for B. subtilis cytochrome c oxidase, the cytochrome caa3 complex. Figure 6A compares the rate of enzyme-catalyzed oxidation of horse cytochrome c and the two B. subtilis cytochromes c-550nm at a single concentration of cytochrome c and the same concentration of cytochrome caa3. The two forms of B. subtilis cytochrome c-550nm are oxidized at about the same rate and that is 100-fold faster than oxidation of horse cytochrome c under the same conditions. The B. subtilis cytochrome c oxidase therefore does not distinguish between the two forms of B. subtilis cytochrome c-550nm due to the addition of the histidine tag and only utilizes horse cytochrome c poorly. The histidine tag does not appear to interfere with the interaction of cytochrome c-550nm

B. subtilis is able to produce plasmid-encoded cytochrome c-550nm that is fully folded and correctly targeted to the plasma membrane. The plasmid-based cytochrome c-550nm is produced eightfold above wildtype B. subtilis on a heme to protein basis. The addition of the C-terminal six-histidine tag to the same plasmid-encoded B. subtilis cytochrome c-550nm decreases the level of cytochrome c-550nm production to twofold over the wild-type level. Purification of histi-

FIG. 6. Oxidation of reduced cytochromes c by cytochrome c oxidase. (A) The reaction was carried out at room temperature using 100 nM cytochrome caa3 in 1% (v/v) Triton X-100, 50 mM NaH2PO4, 5 mM EDTA, pH 7.4. Horse cytochrome c, triangles; B. subtilis cytochrome c-550nm, circles; histidine-tagged B. subtilis cytochrome c-550nm, squares. (B) The reaction was carried out at room temperature using 29 mM cytochrome c, 25 nM bovine cytochrome c oxidase in 0.5 mg dodecyl maltoside per milliliter, 50 mM NaH2PO4, 5 mM EDTA, pH 7.4. Horse cytochrome c, triangles; B. subtilis cytochrome c-550nm, circles; histidine-tagged B. subtilis cytochrome c-550nm, squares.

B. subtilis CYTOCHROME c-550nm

dine-tagged cytochrome c-550nm is more efficient than purification of cytochrome c-550nm from wild-type B. subtilis DB104, or B. subtilis BSC1, expressing the plasmid-encoded cytochrome c-550nm. However, it is possible to produce a greater amount of purified cytochrome c-550nm from the strain of B. subtilis BSC1 expressing the plasmid-encoded cytochrome c-550nm. We are able to purify fourfold more plasmid-encoded native cytochrome c-550nm than histidine-tagged cytochrome c-550nm. Both of the plasmid-based preparations of cytochrome c-550nm are a great improvement over the purification from wild-type B. subtilis. The C-terminus of the protein was chosen as the location for the histidine tag as this terminus is extracellular and would not interfere with the signal sequence of cytochrome c-550nm located at the N-terminus. Placing the histidine tag at the C-terminus, however, puts the histidines in close proximity to the site of covalent attachment of heme C. The added histidine residues have the potential for creating a new metal binding site on the protein that might interfere with the properties of the heme group. The added C-terminal histidine tag does not alter the visible heme C spectrum in purified cytochrome c-550nm, nor the 695-nm band, indicating that the ligation of the heme iron by methionine is not perturbed. The redox midpoint potential was also largely unaffected by the addition of the histidine residues. However, the histidine-tagged cytochrome c-550nm was found initially to be highly susceptible to autooxidation. This instability of the reduced state of the histidine-tagged cytochrome c-550nm could be largely relieved by the addition of EDTA to the incubation medium. The non-histidine-tagged cytochrome c-550nm, although initially more stable than the histidinetagged version, was also found to be stabilized in its reduced state by the presence of EDTA. At optimal EDTA concentrations, the two forms of B. subtilis cytochrome c-550nm have similar autooxidation rates. The heme C domain of B. subtilis cytochrome c-550nm is smaller than mitochondrial cytochrome c and thus the heme of B. subtilis cytochrome c-550nm may have a greater exposure to solvent. The exposure of heme C in horse cytochrome c is 7.5% of the heme macrocycle, as determined from the crystal structure (3). This amounts to less than 0.5% of the total surface of the protein. A greater exposure of the heme in B. subtilis cytochrome c-550nm could explain the greater instability of B. subtilis cytochrome c-550nm to autooxidation relative to mitochondrial cytochrome c as well as the lower midpoint redox potential of the B. subtilis proteins (32). The attainment of conditions to stabilize reduced cytochrome c-550nm has allowed us to assay the relative activity of the different forms of B. subtilis cytochrome c-550nm as substrates for B. subtilis and bovine cyto-

75

chrome c oxidases. The native and histidine-tagged B. subtilis cytochromes c-550nm are indistinguishable as substrates for B. subtilis cytochrome c oxidase indicating that the histidine tag does not interfere with the interaction of these two proteins. B. subtilis cytochrome c-550nm is a much better substrate for B. subtilis cytochrome c oxidase than for the bovine enzyme, and the opposite is true for horse cytochrome c. This is somewhat anticipated because the B. subtilis cytochrome c heme-containing domain has an acidic pI and thus would be negatively charged at neutral pH, whereas horse cytochrome c is positively charged. Mitochondrial cytochrome c is proposed to interact electrostatically with its oxidase. The poor cross-reactivity seen here with these two cytochromes is consistent with the importance of the electrostatic contributions to the interactions with their respective oxidases. ACKNOWLEDGMENTS We are grateful to Professor Lars Hederstedt, University of Lund, Sweden, for supplying us with the plasmid-encoded copy of cccA. This work was supported by an operating grant from the Medical Research Council of Canada to B.C.H.

REFERENCES 1. von Wachenfeldt, C., and Hederstedt, L. (1990) Bacillus subtilis 13-kilodalton cytochrome c-550 encoded by cccA consists of a membrane-anchor and a heme domain. J. Biol. Chem. 265, 13939 –13948. 2. Pettigrew, G. W., and Moore, G. R. (1987) “Cytochromes c. Biological Aspects,” Springer-Verlag, Berlin. 3. Bushnell, G. W., Louie, G. V., and Brayer, G. D. (1990) Highresolution three-dimensional structure of horse heart cytochrome c. J. Mol. Biol. 214, 585–595. 4. Broger, C., Nalecz, M. J., and Azzi, A. (1980) Interaction of cytochrome c with cytochrome bc1 complex of the mitochondrial respiratory chain. Biochim. Biophys. Acta 592, 519 –527. 5. Reider, R., and Bosshard, H. R. (1980) Comparison of the binding sites for cytochrome c oxidase, cytochrome bc1 and cytochrome c1. J. Biol. Chem. 255, 4732– 4739. 6. von Wachenfeldt, C., and Hederstedt, L. (1993) Physico-chemical characterization of membrane-bound and water-soluble forms of Bacillus subtilis cytochrome c-550. Eur. J. Biochem. 212, 499 – 509. 7. Yu, J., Hederstedt, L., and Piggot, P. J. (1995) The cytochrome bc complex (menaquinone:cytochrome c reductase) in Bacillus subtilis has a nontraditional subunit organization. J. Bacteriol. 177, 6751– 6760. 8. Saraste, M., Nakari, T., Metso, T., Jalli, T., Lauraeus, M., and van der Oost, J. (1991) The Bacillus subtilis cytochrome-c oxidase. Eur. J. Biochem. 195, 517–525. 9. Huang, X. (1994) On global sequence alignment. Comput. Appl. Biosci. 10, 227–235. 10. Lemma, E., Schagger, H., and Kroger, A. (1993) The menaquinol oxidase of Bacillus subtilis W23. Arch. Microbiol. 159, 574 –578. 11. von Wachenfeldt, C., and Hederstedt, L. (1990) Bacillus subtilis holo-cytochrome c-550 can be synthesized in aerobic Escherichia coli. FEBS Lett. 270, 147–151. 12. Higuchi, R. (1989) Using PCR to engineer DNA, in “PCR Tech-

76

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

DAVID ET AL. nology, Principles and Applications for DNA Amplification” (Erlich, H. A., Ed.), pp. 61–70, Stockton Press, New York. Erlich, H. A., Gelfand, D., and Sninsky, J. J. (1991) Recent advances in the polymerase chain reaction. Science 252, 1643– 1651. Haima, P., Bron, S., and Venema, G. (1987) The effect of restriction on shotgun cloning and plasmid stability in Bacillus subtilis Marburg. Mol. Gen. Genet. 209, 335–342. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1987) “Molecular Cloning: A Laboratory Manual,” Cold Spring Harbor Laboratory Press, Plainview, NY. Gryczan, T. J., Contente, S., and Dubnau, D. (1978) Characterization of Staphylococcus aureus plasmids introduced by transformation into Bacillus subtilis. J. Bacteriol. 134, 318 –329. Dubnau, D., and Davidoff-Abelson, R. (1971) Fate of transforming DNA following uptake by competent Bacillus subtilis. J. Mol. Biol. 56, 209 –221. Bron, S. (1990) Plasmids, in “Molecular Biological Methods for Bacillus” (Harwood, C. R., and Cutting, S. M., Eds.), pp. 75–139, Wiley, Chichester. Kawamura, F., and Doi, R. H. (1984) Construction of Bacillus subtilis double mutant deficient in extracellular alkaline and neutral protease. J. Bacteriol. 160, 442– 444. DeVrij, W., Azzi, A., and Konings, W. N. (1983) Structural and functional properties of cytochrome c oxidase from Bacillus subtilis W23. Eur. J. Biochem. 131, 97–103. Henning, W., Vo, L., Albanese, J., and Hill, B. C. (1995) Highyield purification of cytochrome aa3 and cytochrome caa3 oxidases from Bacillus subtilis plasma membranes. Biochem. J. 309, 279 –283. Smith, P. K., Krohn, R. I., Hermanson, G. T., Mallia, A. K., Gartner, F. H., Provenzo, M. D., Fujimoto, E. K., Goeke, N. M.,

23.

24.

25.

26.

27.

28.

29. 30. 31. 32.

Olsen, B. J., and Klenk, D. C. (1985) Measurement of protein using bicinchoninic acid. Anal. Biochem. 150, 76 – 85. MacDonnel, A., and Staehelin, L. A. (1981) Detection of cytochrome f, a c-class cytochrome with diaminobenzidine polyacrylamide gels. Anal. Biochem. 40, 177–180. Dutton, P. L. (1978) Redox potentiometry: Determination of midpoint potentials of oxidation-reduction components of biological electron-transfer systems. Methods Enzymol. LIV, 411– 435. Schon, U., and Schumann, W. (1994) Construction of His6-tagging vectors allowing single-step purification of GroES and other polypeptides produced in Bacillus subtilis. Gene 147, 91–94. McDonald, C. C., Phillips, W. D., and Vinogradov, S. S. (1969) Proton magnetic resonance evidence for methionine–iron coordination in mammalian-type ferrocytochrome c. Biochem. Biophys. Res. Commun. 36, 442– 449. Harbury, H. A., Cronin, J. R., Fanger, M. W., Hettinger, T. P., Murphy, A. J., Myer, Y. P., and Vinogradov, S. N. (1965) Complex formation between methionine and a heme peptide from cytochrome c. Proc. Natl. Acad. Sci. USA 54, 1658 –1666. DeVrij, W., van den Burg, B., and Konings, W. N. (1987) Spectral and potentiometric analysis of cytochromes from Bacillus subtilis. Eur. J. Biochem. 166, 589 –595. Hagen, W. R. (1989) Direct electron transfer of redox proteins at the bare glassy carbon electrode. Eur. J. Biochem. 182, 523–530. Ivanetich, K. M., Bradshaw, J. J., and Kaminsky, L. S. (1976) Methionine sulfoxide cytochrome c. Biochemistry 15, 1144 –1153. Keilin, D. (1930) Cytochrome and intracellular oxidase. Proc. R. Soc. London B 106, 418 – 444. Stellwagen, E. (1978) Haem exposure as the determinate of oxidation-reduction potential of haem proteins. Nature 275, 73– 74.