- Email: [email protected]
Direct expression of mature bovine adrenodoxin in Escherichia coli
ARCHIVES
OF BIOCHEMISTRY
AND
BIOPHYSICS
Vol. 295, No. 1, May 15, pp. 126-131,1992
Direct Expression of Mature Bovine Adrenodoxin in Escherichia co/l’ Marie-France Palin, Luc Berthiaume, Jean-Guy Lehoux, Michael R. Waterman,* and Jurgen Syguschl DGpartement de Biochimie, Faculte’ de Mkdecine, Universite’ de Sherbroohe, Quebec, Canada, JlH 5N4; and *Department of Biochemistry, University of Texas, South Western Medical Center, Dallas, Texas 75235
Received September 24, 1991, and in revised form January
24, 1992
Site-directed mutagenesis was utilized to enable direct expression of the mature form of bovine adrenodoxin cDNA using the pKK223-3 expression vector in Escherichia coli. Expression was under control of the “tat” promoter and resulted in a direct expression of soluble mature bovine adrenodoxin (>16 mg per liter). Chromatographic behavior of recombinant adrenodoxin did not differ from that reported for mature native adrenodoxin. The purified recombinant protein was identical to native mitochondrial adrenodoxin on the basis of molecular weight, NH2 terminal sequencing and immunoreactivity. E. coli lysates were brown in color, and the purified protein possessed a visible absorbance spectra identical to native bovine adrenodoxin consistent with incorporation of a [2Fe-2S] cluster in uiuo. Recombinant bovine adrenodoxin was active in cholesterol side-chain cleavage when reconstituted with adrenodoxin reductase and cytochrome P45Oscc and exhibited kinetics reported for native bovine adrenodoxin. The presence of the adrenodoxin amino terminal presequence does not appear to be essential for correct folding of mature recombinant adrenodoxin in E. coli. This expression system should prove useful for overexpression of adrenodoxin mutants in future structure/function studies. The approach described herein can potentially be used to directly express the mature form of any protein in bacteria. o 1992 Academic Press,
Inc.
Adrenal mitochondria catalyze several key oxidative steps in steroid hormone biosynthesis including the oxidative side-chain cleavage of cholesterol and llfl-hydroxylation of various steroids (l-3). These reactions take place intramitochondrially, utilize molecular oxygen, and necessitate an electron transport chain which transfers electrons from NADPH (1). The cholesterol side-chain 1 To whom correspondence
should be addressed.
enzyme cleavage complex is common to a variety of steroidogenesis tissues (4) and its product pregnenolone is the common precursor to a number of steroid hormones (5). The conversion of cholesterol into pregnenolone is rate-limiting in steroidogenesis (5) and is thus the regulatory step for control of steroid hormone levels. Adrenal ferredoxins, adrenodoxins, are small iron-sulfur proteins (- 14 kDa) which transfer reducing equivalents from an NADPH-dependent flavoprotein, adrenodoxin reductase, to a hemoprotein, cholesterol side-chain cleavage cytochrome P450 (P45Osc~)~ or llfi-hydroxylase cytochrome P450 (P4501 Ifi). Adrenodoxin and adrenodoxin reductase are soluble components of the mitochondrial matrix (4) while P45Oscc and P45011/3 are integral proteins of the inner mitochrondrial membrane (6). The soluble adrenodoxin acts as a one-electron shuttle. It associates with NADPH-reduced adrenodoxin reductase (7) from which it accepts an electron, dissociates, and reassociates with the membrane bound cytochrome P450. Donation of one electron to the cytochrome is followed by dissociation and reassociation with adrenodoxin reductase to initiate a subsequent cycle of electron transfer. The protein-protein interactions which make up the adrenodoxin shuttle mechanism have been extensively investigated kinetically and structurally (5). Electrostatic interactions have been implicated in mediating both the adrenodoxin-adrenodoxin reductase complex (8, 9), and the adrenodoxin-cytochrome P450 complex (10,ll) formation. On the basis of chemical modification and crosslinking studies, several carboxylic amino acid residues capable of modulating the interaction of adrenodoxin with its electron transfer partners (12) have been identified in ’ Abbreviations used: P45Oscc, product of the CYPllA gene specific for cholesterol side chain cleavage; P45011@, product of the CYPllB gene specific for ll@ hydroxylation; ADX, adrenodoxin; FPLC, fast protein liquid chromatography; DYT media, 16 g Bactotryptone, 10 g yeast extract, 10 g NaCl per liter; EDTA, ethylenediaminetetraacetic acid; cDNA, complementary deoxyribonucleic acid; RF, replicative form; SDS, sodium dodecyl sulfate.
126 All
0003.9861/92 $3.00 Copyright 0 1992 by Academic Press, Inc. rights of reproduction in any form reserved.
EXPRESSION
OF RECOMBINANT
adrenodoxin. Furthermore, Coghlan and Vickery have demonstrated that homologous carboxylic amino acid residues Asp-76 and Asp-79 are critical for binding offerredoxin to ferredoxin reductase and P45Oscc (13). Complementary genetic studies on adrenodoxin to delineate residues implicated in binding or electron transfer remain to be carried out. Site-directed mutagenesis studies, designed to elucidate the functional role of specific adrenodoxin amino acid residues, necessitate a high level expression system for bovine adrenodoxin. Such overexpression of sufficient protein would also be of advantage for crystallographic studies by those wishing to characterize the structural perturbations in adrenodoxin mutants at the atomic level. Adrenodoxin mRNA, which is encoded by a nuclear gene, is translated on cytoplasmic ribosomes as a higher molecular weight precursor (14). This precursor is subsequently processed upon import into mitochondria by an endoproteolytic cleavage which removes an N-terminal amphiphilic presequence (15). In order to express the mature form of adrenodoxin in Escherichia coli, adrenodoxin cDNA was modified by “site-directed” mutagenesis at the corresponding endoproteolytic cleavage site in the precursor form. The modification was designed to permit removal of the precursor cDNA sequence and to allow for direct expression of mature adrenodoxin. Direct expression of mature recombinant adrenodoxin has been achieved in E. coli. Functionality of the mature recombinant adrenodoxin suggests that in vivo incorporation of the (2Fe-2s) prosthetic group does not depend on the presence of the amphiphilic N-terminal presequence. The use of site-directed mutagenesis to eliminate the N-terminal presequence of adrenodoxin represents a particularly facile method which can be employed for direct expression of the mature form of any recombinant protein. EXPERIMENTAL
PROCEDURES
Materials. Plasmid pcDADX and antibodies against native bovine adrenodoxin were obtained as described (16-18). Expression vector pKK223-3, M13mp9 RF DNA, DNA restriction and modifying enzymes, chromatographic gels, and columns were purchased from Pharmacia. Radio-1251-labeled protein A, [3H]cholesterol, and [“Clpregnenolone were bought from DuPont. Plastic silica gel plates (0.25 mm thickness) for thin layer chromatography were purchased from Macherey-Nagel. Construction of pKADX. A lOOO-base-pair (bp) fragment containing the entire bovine adrenodoxin coding sequence was obtained by BamHI digestion of plasmid pcDADX. The M13mp9 vector RF DNA was linearized with BamHI and treated with alkaline phosphatase. The lOOObp insert was ligated into the M13mp9 vector and used to transform competent E. coli TG-1 cells. The recombinant White transformants plaques were selected, amplified and RF DNA extracted, using standard alkaline lysis protocol (19). Positive transformants were then screened, using appropriate restriction enzyme digestion. Site directed mutagenesis was carried out according to the protocol developed by Eckstein (20, 21) and available from Amersham Radiochemicals. A mutagenic oligonucleotide SGCGTATCGGGGCGAGCGCAGGAATTCATGAGCAGCTCAGAAGATAAAAT3’ which contained an EcoRI restriction site plus an initiation codon (underlined) was annealed to the single-stranded template of the recombinants pre-
BOVINE
ADRENODOXIN
127
viously constructed. Mutants were screened by DNA sequencing using the dideoxy sequencing method (22). DNA sequences of mutants having both the inserted EcoRI restriction site plus the initiation codon at position 338 of bovine adrenodoxin cDNA were further digested with EcoRI and HindI restriction enzymes. The EcoRI/Z-ZindIII DNA fragment, coding for mature adrenodoxin protein, was ligated into pKK223-3 plasmid that had been previously cleaved with EcoRI and HindIII. The ligated mixture was used to transform competent E. coli JM 83 cells. Expression andpwification. Transformed cells were grown overnight at 37°C in 3 ml DYT medium supplemented with ampicillin (50 gg/ ml). This culture was used to inoculate 1 liter of the same medium and cells were grown by vigorous agitation for 24 h at 37°C. All subsequent purification steps were carried out at 4°C. Cells were harvested by centrifugation at SOOOgfor 10 min. The cellular pellet was resuspended in I$ vol of 19 mM Tris-HCl (pH 7.6) containing 1 mM EDTA and 15% sucrose. The cells were lysed by addition of lysosyme (25 pg/ml) and 0.02% Triton X-100 (w/v) final concentration. Recombinant bovine adrenoxin was recovered as a brown supernatant after centrifugation of cellular debris at 12,000g for 15 min. The brown supernatant was dialyzed four times against 1 liter of buffer A (50 mM K2HPOI, pH 7.6) containing 150 mM KC1 and then loaded onto an anion exchanger, DEAE-Sepharose, previously equilibrated in the same buffer. The absorbed proteins were then washed with buffer A containing 170 mM KCl. The brown adrenodoxin-containing band was eluted, using buffer A made up to 500 mM KCl. Prior to absorption onto a semipreparative FPLC phenyl-Superose column (50 ml), the brown eluate was dialyzed four times against 1 liter of buffer B (20 mM Tris-HCl, pH 7.6) containing 1.75 M ammonium sulfate. Recombinant adrenodoxin was eluted using a decreasing ammonium sulfate linear gradient made up in buffer B. The brown eluate fraction obtained from the hydrophobic interaction chromatography step was further purified by gel filtration, using a Superose 12 chromatographic column equilibrated with 50 mM phosphate buffer, pH 7.6, containing 150 mM NaCl. Purity of recombinant adrenodoxin was determined by the absorbance ratio (AdJAm) at each purification step. Crystallization was achieved by dropwise addition of a 3.8 M ammonium sulfate solution to a 1% protein solution until slight turbidity was visible. At 4’C, tabular crystals grew to about 0.5 mm in longest dimension within a week. Electrophoresis and immunoblotting. Aliquots corresponding to the various purifications steps were analyzed by denaturing gel electrophoresis (23). The protein concentrations of each fraction were determined by the bicinchoninic acid method from Pierce Laboratories. Proteins were either stained with Coomassie brilliant blue R-250 or were transferred onto nitrocellulose membranes (0.22 Hm). The transferred proteins were incubated with antibodies for 2 h at room temperature. Immunoreactive proteins were visualized by autoradiography following incubation with iz51-labeled Protein A for 1 h at room temperature. Amino acid sequence analysis. NH&erminal amino acid sequence was determined by Dr. C. Lazure (Institut Clinique de Recherche de Montreal) using a 470 A Applied Biosystems gas-phase sequenator. Absorption spectra of recombinant bovine adrenodoxin diluted in 30 mM KZHPOI (pH 7.6) were recorded at room temperature using a PerkinElmer X 5 spectrophotometer. Enzyme assays. Adrenodoxin reductase (24) and P45Oscc (25) were isolated from bovine adrenal cortex, as previously described. Cytochrome P45Oscc concentrations were determined by carbon monoxide difference spectra assuming a molar extinction coefficient difference A~-90 = 91 of adrenodoxin reductase and recommM-’ * cm- i (26). Concentrations binant adrenodoxin were determined spectrophotometrically, using molar extinction coefficients of t4m = 10.9 (mM*cm)-’ and e4i4 = 11 (mM * cm)-‘, respectively (27, 28). Cholesterol side-chain cleavage assays were performed similar to those previously described (29). The reaction mixtures (300 ~1) consisted of adrenodoxin reductase (0.33 PM), recombinant bovine adrenodoxin
128
PALIN
ET AL.
(variable amounts), P45Oscc (23 nM), and [3H]cholesterol (0.4 mM, 0.2 PCi), made up in 30 mM KxHPO, buffer, pH 7.6, containing 0.3% Tween 20. The unlabeled cholesterol was initially dissolved in hot ethanol and then diluted into the phosphate buffer to which radiolabeled cholesterol was added. The final concentration of ethanol in the reaction mixture was 0.2% (v/v). Each end point assay was initiated by addition of NADPH (0.4 mM final concentration) at 37°C. The reactions were stopped with 0.8 ml methanol and twice extracted with 0.8 ml chloroform. Steroids were separated by ascending thin-layer chromatography (29) using plastic silica gel plates and developed by sulfuric acid spraying. The gel zones corresponding to relative migration of authentic cholesterol and pregnenolone were cut out and counted for radioactivity. Reaction velocities were estimated on the basis of [sH]pregnenolone production as a function of time. Each velocity measurement utilized at least three regularly spaced end points to estimate the rate of [3H]pregnenolone production. All assays showed linear [3H]pregnenolone production for a time period exceeding 4 min. Cytochrome c reduction (1) was assayed in 30 mM KzHPOI (pH 7.5) containing 0.3% Tween 20 at room temperature. Reaction mixtures (1 ml) contained 0.1 pM adrenodoxin reductase, 20 ~1 of 20 mg/ml horse heart cytochrome c (Sigma, type III), and recombinant adrenodoxin (variable amount). The reactions were initiated by addition of 0.6 mM NADPH. Reduction was monitored at 550 nm, and electron transfer activity was calculated on the basis of a molar extinction coefficient of tssc,= 20 (mM . cm))’ for cytochrome c. 1”site
directed
mutagenesis
RESULTS
Construction
of Expression Plasmid pKADX
The cloning strategy used to express recombinant bovine adrenodoxin in E. coli is shown in Fig. 1. Bovine adrenodoxin cDNA obtained following BamHI digestion of plasmid pcDADX (17,18) was first subcloned into the M13mp9 vector (Ml3 ADX). To eliminate the adrenodoxin presequence coding region, the single-stranded Ml3 ADX was used as template, and a sequence corresponding to an EcoRI restriction site and an initiation codon encoded by the 49-bp oligonucleotide was inserted by sitedirected mutagenesis between nucleotides 337 and 338 of the adrenodoxin cDNA. The mutant cDNA encoding only the initiation codon and mature bovine adrenodoxin was selected by EcoRI/HindIII digestion and ligated into the EcoRI/HindIII restriction sites of pKK223-3 expression vector, creating plasmid pKADX. The pKK223-3 vector contains a strong trp-lac (tat) promoter (30), followed by a multiple cloning site derived from pUC-8 (31), which permits proper insertion of cDNA to allow its expression. In addition to possessing a strong transcriptional rrnB terminator sequence (32), termination of protein synthesis was provided by a UAA codon located on the inserted bovine adrenodoxin cDNA. Clones were selected for ampicillin resistance, and restriction enzyme digestion confirmed the desired length of insertion of bovine adrenodoxin cDNA. Expression and Purification Adrenodoxin
of Recombinant
Bovine
Plasmid pKADX was used to transform E. coli strain JM 83. This strain has a lac repressor defective phenotype (A lac pro), resulting in constitutive expression of the
FIG. 1. Construction of the mature recombinant bovine adrenodoxin vector, pKADX. cDNA, cDNA of bovine adrenodoxin; Amp, ampicillin; Pkc, hybrid promoter from trp and lac promoters; E and H, restriction sites for EcoRI and HindIII; large E, additional EcoRI restriction site and initiation codon created by site-directed mutagenesis; C.I.P., calf intestine phosphatase.
recombinant protein. Pellets of transformed cells were brown, suggesting synthesis of an iron-sulfur protein. Recombinant bovine adrenodoxin was purified from the soluble fraction of E. coli lysates by anion exchange and hydrophobic interaction chromatography, as previously described for mature native adrenodoxin (33). The recombinant adrenodoxin represented -2.5% of the total soluble protein. A 24-h culture of E. coli yielded at least
EXPRESSION
OF RECOMBINANT
15 mg per liter of purified soluble recombinant adrenodoxin. The degree of purity of recombinant bovine adrenodoxin corresponding to various chromatographic steps are shown in Fig. 2a. Hydrophobic interaction chromatography, using the phenyl-Superose column, yielded a single band of -14 kDa on SDS gels which corresponds on the basis of relative molecular weight to native bovine adrenodoxin. Gel filtration also yielded a relative molecular weight of -14 kDa corresponding to monomeric adrenodoxin. The purity of recombinant adrenodoxin was also confirmed by running the protein on a higher percentage (18%) polyacrylamide-SDS gel (data not shown). Antibodies raised against mature native bovine adrenodoxin reacted, as shown in Fig. 2b, with the 14-kDa band on denaturing electrophoretic gels, which is consistent with the identity of the purified recombinant protein being mature adrenodoxin. The absorption spectrum of purified recombinant bovine adrenodoxin (oxidized) is shown in Fig. 3. This spectrum has absorption maxima present at 322,414, and 455 nm, characteristic of native bovine adrenodoxin (34). Automated Edman amino acid sequencing of recombinant adrenodoxin yielded a 20-amino acid sequence identical to the NH, terminal sequence of mature bovine adrenodoxin (17,35), confirming expression of the mature form of bovine adrenodoxin in E. coli. The absorbance ratio (A414/A280)was 0.86, indicating a high degree of purity for the recombinant protein (33). Crystals of the mature recombinant adrenodoxin are deep brown-red in colour and show the same morphology (results not shown) as that reported for crystals of mature native bovine adrenodoxin obtained under similar crystallization conditions (33).
B (KDajA
a
b
c
d
e
f
i**dw 31,0-
e
21,5-
am
14,4-
-
_ asm -
-al!w-
FIG. 2. Purification of recombinant bovine adrenodoxin from E. coli JM 83 (A) 15% polyacrylamide-SDS gel electrophoresis of the various stages of purification. The gel was stained with Coomassie blue. Lane a, molecular weight standards (10 ng). Lane b, lysate of recombinant pKADX plasmid expressed in E. coli JM83 (50 pg). Lane c, fraction eluted with 500 mM KC1 from DEAE-Sepharose at pH 7.4 (20 ng); lane d, flowthrough fraction eluted with ammonium sulfate gradient from phenyl-Superose column pH 7.4 (2 pg). Lane e, eluate from Superose 12 gel filtration (4 fig). Lane f, native bovine adrenodoxin (4 pg). (B) Corresponding Western blot of lane e, using polyclonal antibodies directed against native bovine adrenodoxin.
BOVINE
129
ADRENODOXIN
WAVELENGT H (n m) FIG.
3.
Absorption
spectra of recombinant
bovine adrenodoxin.
Kinetic Assays Kinetic analyses of electron transfer by recombinant bovine adrenodoxin were carried out in the presence of saturating amounts of adrenodoxin reductase, using two different assay systems. Reciprocal plots, corresponding to cholesterol side-chain cleavage activity and cytochrome c reduction activity, are shown in Figs. 4A and 4B, respectively. Kinetics of electron transfer by recombinant bovine adrenodoxin to P45Oscc is characterized by a turnover of 49 nmol pregnenolone/min/nmol P45Oscc and a Km of 0.96 PM, while turnover of 460 mine1 and K,,, of 38 nM characterized electron transfer kinetics to cytochrome c. These results are within the range reported for native bovine adrenodoxin (12,36) and establish that the recombinant adrenodoxin contains a functional (2Fe-2s) center and a properly folded structure capable of mediating electron transfer from NADPH to P45Oscc. DISCUSSION
A “site-directed” mutagenesis approach has enabled the direct expression of mature recombinant bovine adrenodoxin in E. coli. The recombinant adrenodoxin is expressed as a soluble protein (2.5% of total soluble protein) and corresponds to a yield in excess of 15 mg of purified recombinant adrenodoxin per liter of E. coli culture. The purification of a comparable amount of native adrenodoxin would require an equivalent of at least 40 bovine adrenal cortices (33). Direct expression of adrenodoxin in E. coli involves fewer purification steps, when compared to the purification scheme of recombinant human ferredoxin protein (37). In this protocol, two additional steps are required to yield a homogenous mature protein that is free of the leader peptide. Losses in these additional steps could account for the apparent lower yield of recombinant human ferrodoxin (5 mg/liter), even though the fusion protein corresponded initially to -12% of the total soluble protein. The rationale of incorporating the
130
PALIN
t 0
1
2 1/adrenodoxin
3
4
ET AL.
0
(pM)-’
100 l/Adrenodoxin
200 (PM)
300
-1
FIG. 4. Kinetic studies on recombinant adrenodoxin. (A) Reciprocal plot showing kinetics of cholesterol transformation to pregnenolone for various concentrations of recombinant bovine adrenodoxin. (B) Reciprocal plot showing kinetics corresponding to cytochrome c reductase activity by recombinant bovine adrenodoxin. Cytochrome c reduction was monitored at 550 nm. Error bars reflect maximum difference for duplicate measurements and are shown whenever these exceeded the size of the plotted symbols.
first 31 amino acids of the X cI1 protein in order to express human ferredoxin as a cleavable fusion protein in E. coli was based on the premise that the leader sequence would augment stability of the recombinant protein (37). On the basis of our study, the addition of a presequence does not appear to be essential for stability, since from both NH2 terminal sequence and Western blot analysis, proteolysis of the recombinant bovine adrenodoxin was not detectable, even upon prolonged incubation (24 h) in E. coli. Codon usage corresponding to high abundance tRNA species in E. coli has been postulated as a requirement for efficient high level expression of recombinant myoglobin and recombinant human hemoglobin (38,39). High level expression of recombinant adrenodoxin, and the unrelated protein recombinant anaerobic maize aldolase, is obtained even though >50% of codons in the cDNA sequences correspond to usage of tRNA species that are of low abundance in E. coli (40). A high degree of optimal codon usage is therefore not essential for overexpression of all soluble recombinant proteins in E. cob. The characteristic brown color of bovine adrenodoxin is visible in E. cob cells expressing the recombinant protein and suggests assembly of iron-sulfur clusters into recombinant adrenodoxin. This assembly was confirmed by the absorption spectrum of purified bovine adrenodoxin which was indistinguishable from the native protein, suggesting a properly assembled (2Fe-2s) type cluster. Recombinant bovine adrenodoxin directly expressed in E. coli is able to reduce both P45Oscc and cytochrome c. The kinetic parameters are consistent with a fully competent electron transfer capacity by recombinant adrenodoxin and suggest that the recombinant protein possesses a tertiary structure folded similarly to the native structure and into which the (2Fe-2s) type cluster has been incorporated identically to that of the native protein. Full activity and chromatographic behavior identical to
that of the native protein indicates that post-translational modifications are not required for adrenodoxin function, unless they are also common to E. coli. The finding that absence of the presequence results in a fully functional protein implies that the presequence is not required for proper folding of the mature apoprotein with the (2Fe2s) cluster. The expression system allows for efficient production of homogenous preparations of mature adrenodoxin in quantities sufficient to carry out extensive biochemical and crystallographic studies. The absence of significant amino acid sequence homology of the vertebrate ferredoxins, with algae and bacterial ferredoxins (41), represents an intriguing paradigm as to the evolution of function by divergent amino acid sequences. Structural studies on mammalian ferredoxins have yet to be initiated to determine the similarity of the molecular architecture of the vertebrate ferredoxins with respect to other ferredoxins (42, 43) of unrelated function. ACKNOWLEDGMENTS We are grateful to Dr. Claude Lazure for NHr-terminal amino acid sequence analysis. We thank Mr. Jacques Lehoux for technical assistance and Mrs. Catherine Masse for her expert secretarial assistance. This research was supported by the Medical Research Council of Canada, MRC Grant MT-8088 to J. Sygusch, MRC Grant MT-4653 to J. G. Lehoux, a F.C.A.R. award to M. F. Palin, and a FRSQ scholarship to L. Berthiaume.
REFERENCES 1. Omura, T., Sanders, E., Estabrook, R. W., Cooper, D. Y., and Rosenthal, 0. (1966) Arch. Biochem. Btiphys. 117, 660-673. 2. Takikawa, O., Gomi, T., Suhara, K., Itigaki, E., Takemori, S., and Katagiri, M. (1978) Arch. Biochem. Biophys. 190,300-306. 3. Sato, H., Ashida, N., Suhara, K., Itagaki, E., Takemori, S., and Katagiri, M. (1978) Arch. Biochem. Biophys. 190,307-314.
EXPRESSION 4. Nakamura, Y., Otsuka, H., and Tamaoki, phys. Acta 122,34-42. 5. Lambeth, J. D., Seybert, D. W., Lancaster, Kamin, H. (1982) Mol. Cell. Biochem. 45, 6. Churchill, P. F., and Kimura, T. (19’79) J.
OF RECOMBINANT
B. (1966) Biochem. BioJ. R., Salerno, J. C., and 13-31. Biol. Chem. 254,10443-
10448. 7. Chu, J. W., and Kimura, T. (1973) J. Biol. Chem. 248,5183-5187. 8. Lambeth, J. D., McCaslin, D. R., and Kamin, H. (1976) J. Biol.
Chem.251,7545-7550. 9. Lambeth,
J. D., and Kamin,
H. (1979) J. Biol. Chem. 254,
2766-
2774. 10. Katagiri, M., Takikawa, O., Sato, H., and Suhara, K. (1977) B&hem. Biophys. Res. Commun. 77,804-809. 11. Seybert, D. W., Lancaster, J. R., Lambeth, J. D., and Kamin, H. (1979) J. Biol. Chem. 254, 12,08812,098. 12. Lambeth, J. D., Geren, L. M., and Millet, F. (1984) J. Biol. Chem. 259, 10,025-10,029. 13. Coghlan, V. M., andvickery, L. E. (1991) J. Biol. Chem. 266,18,60618,612. 14. Nabi, N., and Omura, T. (1980) &o&cm. Biophys. Res. Commun.
97,680-686.
ADRENODOXIN
24. Hiwatashi, A., Ichikawa, Y., Maruya, (1976) Biochemistry 15,3082-3090.
131 N., Yamano, T., and Aki, K.
25. Katagiri,
26. 27. 28. 29.
M., Takemori, S., Itagaki, E., and Suhara, K. (1978) in Methods in Enzymology (Fleischer, S., and Packer, L., Eds.), Vol. 52, pp. 124-132, Academic Press, San Diego. Omura, T., and Sato, R. (1964) J. Biol. Chem. 239,2370-2378. Chu, J., and Kimura, T. (1973) J. Biol. Chem. 248, 2089-2094. Huang, J. J., and Kimura, T. (1973) Biochemistry 12,406-409. Monnier, N. Defaye, G., and Chambaz, E. M. (1987) Eur. J. Biochem.
169,147-153. 30. De Boer, H. A., Comstock, L. J., and Vasser, M. (1983) Proc. Natl. Acad. Sci. USA 80, 21-25. 31. Vieira, J. and Messing, J. (1982) Gene 19, 259-268. 32. Brosius, J., Dull, T. J., Sleeter, D. D., and Noller, H. F. (1981) J. Mol. Biol. 148, 107-127. 33. Suhara, K., Takemori, S., and Katagiri, M. (1972) Biochem. Biophys. Acta. 263,272-278. 34. Suzuki, K., and Kimura, T. (1965) Biochem. Biophys. Res. Commun.
19,340-345. 35. Tanaka, M., Haniu, M., Yasunobu, K. T., and Kimura, T. (1973) J. Biol. Chem. 248,1141-1157.
15. Matocha, M. F., and Waterman, M. R. (1985) J. Biol. Chem. 260, 12,259-12,265. 16. Kramer, R. E., Du Bois, R. N., Simpson, E. R., Anderson, C. M., Kashiwagi, K., Lambeth, J. D., Jefcoate, C. R., and Waterman, M. R. (1982) Arch. Biochem. Biophys. 215,478-485. 17. Okamura, T., John, M. E., Zuber, M. X., Simpson, E. R., and Waterman, M. R. (1985) Proc. Natl. Acad. Sci. USA 82, 5705-5709. 18. Okayama, H., and Berg, P. (1983) Mol. Cell. Biol. 3, 280-289. 19. Maniatis, T., Fritsch, E., and Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. 20. Taylor, J. W., Schmidt, W., Cosstick, R., Okruszek, A., and Eckstein, F. (1985) Nucleic Acids Res. 13,8749-8764. 21. Taylor, J. W., Ott, J., and Eckstein, F. (1985) Nucleic Acids Res.
13,8765-8785. 22. Sanger, F., Nicklen, S., and Coulson, A. R. (1977) Proc. Natl. Acad. Sci. USA 74,5463-5467. 23. Laemmli, U. K. (1970) Nature (London)
BOVINE
227,680-685.
36. Hanukoglu, I., Spitsberg, V., Bumpus, J. A., Dus, K. M., and Jefcoate, C. R. (1981) J. Biol. Chem. 256,4321-4328. 37. Coghlan, V. M., and Vickery, L. E. (1989) Proc. N&l. Acad. Sci. USA 86,835-839. 38. Springer, B. A., and Sligar, S. G. (1987) Proc. Natl. Acad. Sci. USA
84,8961-8965. 39. Hoffman, S. J., Looker, D. L., Roehrich, J. M., Cozart, P. E., Durfee, S. L., Tedesco, J. L., and Stetler, G. L. (1990) Proc. Natl. Acacl. Sci. USA 87,8521-8525. 40. Ikemura, T. (1981) J. Mol. Biol. 151,389-409. 41. Meyer, J., Bruschi, M. H., Bonicel, J. J., and Bovier-Lapierre, G. E. (1986) Biochemistry 25,6054-6061. 42. Fukuyama, K., Hase, T., Matsumoto, S., Tsukihara, T., Katsube, Y., Tanaka, N., Kakudo, M., Wada, K., and Matsubara, H. (1980) Nature (London) 286, 522-524. 43. Sussman, J. L., Brown, J. H., and Shokam, M. (1986) in Iron-Sulfur Protein Research (Matsubara, H., Katsube, Y., and Wada, K., Eds.), pp. 69-82, Springer, Berlin.
OF BIOCHEMISTRY
AND
BIOPHYSICS
Vol. 295, No. 1, May 15, pp. 126-131,1992
Direct Expression of Mature Bovine Adrenodoxin in Escherichia co/l’ Marie-France Palin, Luc Berthiaume, Jean-Guy Lehoux, Michael R. Waterman,* and Jurgen Syguschl DGpartement de Biochimie, Faculte’ de Mkdecine, Universite’ de Sherbroohe, Quebec, Canada, JlH 5N4; and *Department of Biochemistry, University of Texas, South Western Medical Center, Dallas, Texas 75235
Received September 24, 1991, and in revised form January
24, 1992
Site-directed mutagenesis was utilized to enable direct expression of the mature form of bovine adrenodoxin cDNA using the pKK223-3 expression vector in Escherichia coli. Expression was under control of the “tat” promoter and resulted in a direct expression of soluble mature bovine adrenodoxin (>16 mg per liter). Chromatographic behavior of recombinant adrenodoxin did not differ from that reported for mature native adrenodoxin. The purified recombinant protein was identical to native mitochondrial adrenodoxin on the basis of molecular weight, NH2 terminal sequencing and immunoreactivity. E. coli lysates were brown in color, and the purified protein possessed a visible absorbance spectra identical to native bovine adrenodoxin consistent with incorporation of a [2Fe-2S] cluster in uiuo. Recombinant bovine adrenodoxin was active in cholesterol side-chain cleavage when reconstituted with adrenodoxin reductase and cytochrome P45Oscc and exhibited kinetics reported for native bovine adrenodoxin. The presence of the adrenodoxin amino terminal presequence does not appear to be essential for correct folding of mature recombinant adrenodoxin in E. coli. This expression system should prove useful for overexpression of adrenodoxin mutants in future structure/function studies. The approach described herein can potentially be used to directly express the mature form of any protein in bacteria. o 1992 Academic Press,
Inc.
Adrenal mitochondria catalyze several key oxidative steps in steroid hormone biosynthesis including the oxidative side-chain cleavage of cholesterol and llfl-hydroxylation of various steroids (l-3). These reactions take place intramitochondrially, utilize molecular oxygen, and necessitate an electron transport chain which transfers electrons from NADPH (1). The cholesterol side-chain 1 To whom correspondence
should be addressed.
enzyme cleavage complex is common to a variety of steroidogenesis tissues (4) and its product pregnenolone is the common precursor to a number of steroid hormones (5). The conversion of cholesterol into pregnenolone is rate-limiting in steroidogenesis (5) and is thus the regulatory step for control of steroid hormone levels. Adrenal ferredoxins, adrenodoxins, are small iron-sulfur proteins (- 14 kDa) which transfer reducing equivalents from an NADPH-dependent flavoprotein, adrenodoxin reductase, to a hemoprotein, cholesterol side-chain cleavage cytochrome P450 (P45Osc~)~ or llfi-hydroxylase cytochrome P450 (P4501 Ifi). Adrenodoxin and adrenodoxin reductase are soluble components of the mitochondrial matrix (4) while P45Oscc and P45011/3 are integral proteins of the inner mitochrondrial membrane (6). The soluble adrenodoxin acts as a one-electron shuttle. It associates with NADPH-reduced adrenodoxin reductase (7) from which it accepts an electron, dissociates, and reassociates with the membrane bound cytochrome P450. Donation of one electron to the cytochrome is followed by dissociation and reassociation with adrenodoxin reductase to initiate a subsequent cycle of electron transfer. The protein-protein interactions which make up the adrenodoxin shuttle mechanism have been extensively investigated kinetically and structurally (5). Electrostatic interactions have been implicated in mediating both the adrenodoxin-adrenodoxin reductase complex (8, 9), and the adrenodoxin-cytochrome P450 complex (10,ll) formation. On the basis of chemical modification and crosslinking studies, several carboxylic amino acid residues capable of modulating the interaction of adrenodoxin with its electron transfer partners (12) have been identified in ’ Abbreviations used: P45Oscc, product of the CYPllA gene specific for cholesterol side chain cleavage; P45011@, product of the CYPllB gene specific for ll@ hydroxylation; ADX, adrenodoxin; FPLC, fast protein liquid chromatography; DYT media, 16 g Bactotryptone, 10 g yeast extract, 10 g NaCl per liter; EDTA, ethylenediaminetetraacetic acid; cDNA, complementary deoxyribonucleic acid; RF, replicative form; SDS, sodium dodecyl sulfate.
126 All
0003.9861/92 $3.00 Copyright 0 1992 by Academic Press, Inc. rights of reproduction in any form reserved.
EXPRESSION
OF RECOMBINANT
adrenodoxin. Furthermore, Coghlan and Vickery have demonstrated that homologous carboxylic amino acid residues Asp-76 and Asp-79 are critical for binding offerredoxin to ferredoxin reductase and P45Oscc (13). Complementary genetic studies on adrenodoxin to delineate residues implicated in binding or electron transfer remain to be carried out. Site-directed mutagenesis studies, designed to elucidate the functional role of specific adrenodoxin amino acid residues, necessitate a high level expression system for bovine adrenodoxin. Such overexpression of sufficient protein would also be of advantage for crystallographic studies by those wishing to characterize the structural perturbations in adrenodoxin mutants at the atomic level. Adrenodoxin mRNA, which is encoded by a nuclear gene, is translated on cytoplasmic ribosomes as a higher molecular weight precursor (14). This precursor is subsequently processed upon import into mitochondria by an endoproteolytic cleavage which removes an N-terminal amphiphilic presequence (15). In order to express the mature form of adrenodoxin in Escherichia coli, adrenodoxin cDNA was modified by “site-directed” mutagenesis at the corresponding endoproteolytic cleavage site in the precursor form. The modification was designed to permit removal of the precursor cDNA sequence and to allow for direct expression of mature adrenodoxin. Direct expression of mature recombinant adrenodoxin has been achieved in E. coli. Functionality of the mature recombinant adrenodoxin suggests that in vivo incorporation of the (2Fe-2s) prosthetic group does not depend on the presence of the amphiphilic N-terminal presequence. The use of site-directed mutagenesis to eliminate the N-terminal presequence of adrenodoxin represents a particularly facile method which can be employed for direct expression of the mature form of any recombinant protein. EXPERIMENTAL
PROCEDURES
Materials. Plasmid pcDADX and antibodies against native bovine adrenodoxin were obtained as described (16-18). Expression vector pKK223-3, M13mp9 RF DNA, DNA restriction and modifying enzymes, chromatographic gels, and columns were purchased from Pharmacia. Radio-1251-labeled protein A, [3H]cholesterol, and [“Clpregnenolone were bought from DuPont. Plastic silica gel plates (0.25 mm thickness) for thin layer chromatography were purchased from Macherey-Nagel. Construction of pKADX. A lOOO-base-pair (bp) fragment containing the entire bovine adrenodoxin coding sequence was obtained by BamHI digestion of plasmid pcDADX. The M13mp9 vector RF DNA was linearized with BamHI and treated with alkaline phosphatase. The lOOObp insert was ligated into the M13mp9 vector and used to transform competent E. coli TG-1 cells. The recombinant White transformants plaques were selected, amplified and RF DNA extracted, using standard alkaline lysis protocol (19). Positive transformants were then screened, using appropriate restriction enzyme digestion. Site directed mutagenesis was carried out according to the protocol developed by Eckstein (20, 21) and available from Amersham Radiochemicals. A mutagenic oligonucleotide SGCGTATCGGGGCGAGCGCAGGAATTCATGAGCAGCTCAGAAGATAAAAT3’ which contained an EcoRI restriction site plus an initiation codon (underlined) was annealed to the single-stranded template of the recombinants pre-
BOVINE
ADRENODOXIN
127
viously constructed. Mutants were screened by DNA sequencing using the dideoxy sequencing method (22). DNA sequences of mutants having both the inserted EcoRI restriction site plus the initiation codon at position 338 of bovine adrenodoxin cDNA were further digested with EcoRI and HindI restriction enzymes. The EcoRI/Z-ZindIII DNA fragment, coding for mature adrenodoxin protein, was ligated into pKK223-3 plasmid that had been previously cleaved with EcoRI and HindIII. The ligated mixture was used to transform competent E. coli JM 83 cells. Expression andpwification. Transformed cells were grown overnight at 37°C in 3 ml DYT medium supplemented with ampicillin (50 gg/ ml). This culture was used to inoculate 1 liter of the same medium and cells were grown by vigorous agitation for 24 h at 37°C. All subsequent purification steps were carried out at 4°C. Cells were harvested by centrifugation at SOOOgfor 10 min. The cellular pellet was resuspended in I$ vol of 19 mM Tris-HCl (pH 7.6) containing 1 mM EDTA and 15% sucrose. The cells were lysed by addition of lysosyme (25 pg/ml) and 0.02% Triton X-100 (w/v) final concentration. Recombinant bovine adrenoxin was recovered as a brown supernatant after centrifugation of cellular debris at 12,000g for 15 min. The brown supernatant was dialyzed four times against 1 liter of buffer A (50 mM K2HPOI, pH 7.6) containing 150 mM KC1 and then loaded onto an anion exchanger, DEAE-Sepharose, previously equilibrated in the same buffer. The absorbed proteins were then washed with buffer A containing 170 mM KCl. The brown adrenodoxin-containing band was eluted, using buffer A made up to 500 mM KCl. Prior to absorption onto a semipreparative FPLC phenyl-Superose column (50 ml), the brown eluate was dialyzed four times against 1 liter of buffer B (20 mM Tris-HCl, pH 7.6) containing 1.75 M ammonium sulfate. Recombinant adrenodoxin was eluted using a decreasing ammonium sulfate linear gradient made up in buffer B. The brown eluate fraction obtained from the hydrophobic interaction chromatography step was further purified by gel filtration, using a Superose 12 chromatographic column equilibrated with 50 mM phosphate buffer, pH 7.6, containing 150 mM NaCl. Purity of recombinant adrenodoxin was determined by the absorbance ratio (AdJAm) at each purification step. Crystallization was achieved by dropwise addition of a 3.8 M ammonium sulfate solution to a 1% protein solution until slight turbidity was visible. At 4’C, tabular crystals grew to about 0.5 mm in longest dimension within a week. Electrophoresis and immunoblotting. Aliquots corresponding to the various purifications steps were analyzed by denaturing gel electrophoresis (23). The protein concentrations of each fraction were determined by the bicinchoninic acid method from Pierce Laboratories. Proteins were either stained with Coomassie brilliant blue R-250 or were transferred onto nitrocellulose membranes (0.22 Hm). The transferred proteins were incubated with antibodies for 2 h at room temperature. Immunoreactive proteins were visualized by autoradiography following incubation with iz51-labeled Protein A for 1 h at room temperature. Amino acid sequence analysis. NH&erminal amino acid sequence was determined by Dr. C. Lazure (Institut Clinique de Recherche de Montreal) using a 470 A Applied Biosystems gas-phase sequenator. Absorption spectra of recombinant bovine adrenodoxin diluted in 30 mM KZHPOI (pH 7.6) were recorded at room temperature using a PerkinElmer X 5 spectrophotometer. Enzyme assays. Adrenodoxin reductase (24) and P45Oscc (25) were isolated from bovine adrenal cortex, as previously described. Cytochrome P45Oscc concentrations were determined by carbon monoxide difference spectra assuming a molar extinction coefficient difference A~-90 = 91 of adrenodoxin reductase and recommM-’ * cm- i (26). Concentrations binant adrenodoxin were determined spectrophotometrically, using molar extinction coefficients of t4m = 10.9 (mM*cm)-’ and e4i4 = 11 (mM * cm)-‘, respectively (27, 28). Cholesterol side-chain cleavage assays were performed similar to those previously described (29). The reaction mixtures (300 ~1) consisted of adrenodoxin reductase (0.33 PM), recombinant bovine adrenodoxin
128
PALIN
ET AL.
(variable amounts), P45Oscc (23 nM), and [3H]cholesterol (0.4 mM, 0.2 PCi), made up in 30 mM KxHPO, buffer, pH 7.6, containing 0.3% Tween 20. The unlabeled cholesterol was initially dissolved in hot ethanol and then diluted into the phosphate buffer to which radiolabeled cholesterol was added. The final concentration of ethanol in the reaction mixture was 0.2% (v/v). Each end point assay was initiated by addition of NADPH (0.4 mM final concentration) at 37°C. The reactions were stopped with 0.8 ml methanol and twice extracted with 0.8 ml chloroform. Steroids were separated by ascending thin-layer chromatography (29) using plastic silica gel plates and developed by sulfuric acid spraying. The gel zones corresponding to relative migration of authentic cholesterol and pregnenolone were cut out and counted for radioactivity. Reaction velocities were estimated on the basis of [sH]pregnenolone production as a function of time. Each velocity measurement utilized at least three regularly spaced end points to estimate the rate of [3H]pregnenolone production. All assays showed linear [3H]pregnenolone production for a time period exceeding 4 min. Cytochrome c reduction (1) was assayed in 30 mM KzHPOI (pH 7.5) containing 0.3% Tween 20 at room temperature. Reaction mixtures (1 ml) contained 0.1 pM adrenodoxin reductase, 20 ~1 of 20 mg/ml horse heart cytochrome c (Sigma, type III), and recombinant adrenodoxin (variable amount). The reactions were initiated by addition of 0.6 mM NADPH. Reduction was monitored at 550 nm, and electron transfer activity was calculated on the basis of a molar extinction coefficient of tssc,= 20 (mM . cm))’ for cytochrome c. 1”site
directed
mutagenesis
RESULTS
Construction
of Expression Plasmid pKADX
The cloning strategy used to express recombinant bovine adrenodoxin in E. coli is shown in Fig. 1. Bovine adrenodoxin cDNA obtained following BamHI digestion of plasmid pcDADX (17,18) was first subcloned into the M13mp9 vector (Ml3 ADX). To eliminate the adrenodoxin presequence coding region, the single-stranded Ml3 ADX was used as template, and a sequence corresponding to an EcoRI restriction site and an initiation codon encoded by the 49-bp oligonucleotide was inserted by sitedirected mutagenesis between nucleotides 337 and 338 of the adrenodoxin cDNA. The mutant cDNA encoding only the initiation codon and mature bovine adrenodoxin was selected by EcoRI/HindIII digestion and ligated into the EcoRI/HindIII restriction sites of pKK223-3 expression vector, creating plasmid pKADX. The pKK223-3 vector contains a strong trp-lac (tat) promoter (30), followed by a multiple cloning site derived from pUC-8 (31), which permits proper insertion of cDNA to allow its expression. In addition to possessing a strong transcriptional rrnB terminator sequence (32), termination of protein synthesis was provided by a UAA codon located on the inserted bovine adrenodoxin cDNA. Clones were selected for ampicillin resistance, and restriction enzyme digestion confirmed the desired length of insertion of bovine adrenodoxin cDNA. Expression and Purification Adrenodoxin
of Recombinant
Bovine
Plasmid pKADX was used to transform E. coli strain JM 83. This strain has a lac repressor defective phenotype (A lac pro), resulting in constitutive expression of the
FIG. 1. Construction of the mature recombinant bovine adrenodoxin vector, pKADX. cDNA, cDNA of bovine adrenodoxin; Amp, ampicillin; Pkc, hybrid promoter from trp and lac promoters; E and H, restriction sites for EcoRI and HindIII; large E, additional EcoRI restriction site and initiation codon created by site-directed mutagenesis; C.I.P., calf intestine phosphatase.
recombinant protein. Pellets of transformed cells were brown, suggesting synthesis of an iron-sulfur protein. Recombinant bovine adrenodoxin was purified from the soluble fraction of E. coli lysates by anion exchange and hydrophobic interaction chromatography, as previously described for mature native adrenodoxin (33). The recombinant adrenodoxin represented -2.5% of the total soluble protein. A 24-h culture of E. coli yielded at least
EXPRESSION
OF RECOMBINANT
15 mg per liter of purified soluble recombinant adrenodoxin. The degree of purity of recombinant bovine adrenodoxin corresponding to various chromatographic steps are shown in Fig. 2a. Hydrophobic interaction chromatography, using the phenyl-Superose column, yielded a single band of -14 kDa on SDS gels which corresponds on the basis of relative molecular weight to native bovine adrenodoxin. Gel filtration also yielded a relative molecular weight of -14 kDa corresponding to monomeric adrenodoxin. The purity of recombinant adrenodoxin was also confirmed by running the protein on a higher percentage (18%) polyacrylamide-SDS gel (data not shown). Antibodies raised against mature native bovine adrenodoxin reacted, as shown in Fig. 2b, with the 14-kDa band on denaturing electrophoretic gels, which is consistent with the identity of the purified recombinant protein being mature adrenodoxin. The absorption spectrum of purified recombinant bovine adrenodoxin (oxidized) is shown in Fig. 3. This spectrum has absorption maxima present at 322,414, and 455 nm, characteristic of native bovine adrenodoxin (34). Automated Edman amino acid sequencing of recombinant adrenodoxin yielded a 20-amino acid sequence identical to the NH, terminal sequence of mature bovine adrenodoxin (17,35), confirming expression of the mature form of bovine adrenodoxin in E. coli. The absorbance ratio (A414/A280)was 0.86, indicating a high degree of purity for the recombinant protein (33). Crystals of the mature recombinant adrenodoxin are deep brown-red in colour and show the same morphology (results not shown) as that reported for crystals of mature native bovine adrenodoxin obtained under similar crystallization conditions (33).
B (KDajA
a
b
c
d
e
f
i**dw 31,0-
e
21,5-
am
14,4-
-
_ asm -
-al!w-
FIG. 2. Purification of recombinant bovine adrenodoxin from E. coli JM 83 (A) 15% polyacrylamide-SDS gel electrophoresis of the various stages of purification. The gel was stained with Coomassie blue. Lane a, molecular weight standards (10 ng). Lane b, lysate of recombinant pKADX plasmid expressed in E. coli JM83 (50 pg). Lane c, fraction eluted with 500 mM KC1 from DEAE-Sepharose at pH 7.4 (20 ng); lane d, flowthrough fraction eluted with ammonium sulfate gradient from phenyl-Superose column pH 7.4 (2 pg). Lane e, eluate from Superose 12 gel filtration (4 fig). Lane f, native bovine adrenodoxin (4 pg). (B) Corresponding Western blot of lane e, using polyclonal antibodies directed against native bovine adrenodoxin.
BOVINE
129
ADRENODOXIN
WAVELENGT H (n m) FIG.
3.
Absorption
spectra of recombinant
bovine adrenodoxin.
Kinetic Assays Kinetic analyses of electron transfer by recombinant bovine adrenodoxin were carried out in the presence of saturating amounts of adrenodoxin reductase, using two different assay systems. Reciprocal plots, corresponding to cholesterol side-chain cleavage activity and cytochrome c reduction activity, are shown in Figs. 4A and 4B, respectively. Kinetics of electron transfer by recombinant bovine adrenodoxin to P45Oscc is characterized by a turnover of 49 nmol pregnenolone/min/nmol P45Oscc and a Km of 0.96 PM, while turnover of 460 mine1 and K,,, of 38 nM characterized electron transfer kinetics to cytochrome c. These results are within the range reported for native bovine adrenodoxin (12,36) and establish that the recombinant adrenodoxin contains a functional (2Fe-2s) center and a properly folded structure capable of mediating electron transfer from NADPH to P45Oscc. DISCUSSION
A “site-directed” mutagenesis approach has enabled the direct expression of mature recombinant bovine adrenodoxin in E. coli. The recombinant adrenodoxin is expressed as a soluble protein (2.5% of total soluble protein) and corresponds to a yield in excess of 15 mg of purified recombinant adrenodoxin per liter of E. coli culture. The purification of a comparable amount of native adrenodoxin would require an equivalent of at least 40 bovine adrenal cortices (33). Direct expression of adrenodoxin in E. coli involves fewer purification steps, when compared to the purification scheme of recombinant human ferredoxin protein (37). In this protocol, two additional steps are required to yield a homogenous mature protein that is free of the leader peptide. Losses in these additional steps could account for the apparent lower yield of recombinant human ferrodoxin (5 mg/liter), even though the fusion protein corresponded initially to -12% of the total soluble protein. The rationale of incorporating the
130
PALIN
t 0
1
2 1/adrenodoxin
3
4
ET AL.
0
(pM)-’
100 l/Adrenodoxin
200 (PM)
300
-1
FIG. 4. Kinetic studies on recombinant adrenodoxin. (A) Reciprocal plot showing kinetics of cholesterol transformation to pregnenolone for various concentrations of recombinant bovine adrenodoxin. (B) Reciprocal plot showing kinetics corresponding to cytochrome c reductase activity by recombinant bovine adrenodoxin. Cytochrome c reduction was monitored at 550 nm. Error bars reflect maximum difference for duplicate measurements and are shown whenever these exceeded the size of the plotted symbols.
first 31 amino acids of the X cI1 protein in order to express human ferredoxin as a cleavable fusion protein in E. coli was based on the premise that the leader sequence would augment stability of the recombinant protein (37). On the basis of our study, the addition of a presequence does not appear to be essential for stability, since from both NH2 terminal sequence and Western blot analysis, proteolysis of the recombinant bovine adrenodoxin was not detectable, even upon prolonged incubation (24 h) in E. coli. Codon usage corresponding to high abundance tRNA species in E. coli has been postulated as a requirement for efficient high level expression of recombinant myoglobin and recombinant human hemoglobin (38,39). High level expression of recombinant adrenodoxin, and the unrelated protein recombinant anaerobic maize aldolase, is obtained even though >50% of codons in the cDNA sequences correspond to usage of tRNA species that are of low abundance in E. coli (40). A high degree of optimal codon usage is therefore not essential for overexpression of all soluble recombinant proteins in E. cob. The characteristic brown color of bovine adrenodoxin is visible in E. cob cells expressing the recombinant protein and suggests assembly of iron-sulfur clusters into recombinant adrenodoxin. This assembly was confirmed by the absorption spectrum of purified bovine adrenodoxin which was indistinguishable from the native protein, suggesting a properly assembled (2Fe-2s) type cluster. Recombinant bovine adrenodoxin directly expressed in E. coli is able to reduce both P45Oscc and cytochrome c. The kinetic parameters are consistent with a fully competent electron transfer capacity by recombinant adrenodoxin and suggest that the recombinant protein possesses a tertiary structure folded similarly to the native structure and into which the (2Fe-2s) type cluster has been incorporated identically to that of the native protein. Full activity and chromatographic behavior identical to
that of the native protein indicates that post-translational modifications are not required for adrenodoxin function, unless they are also common to E. coli. The finding that absence of the presequence results in a fully functional protein implies that the presequence is not required for proper folding of the mature apoprotein with the (2Fe2s) cluster. The expression system allows for efficient production of homogenous preparations of mature adrenodoxin in quantities sufficient to carry out extensive biochemical and crystallographic studies. The absence of significant amino acid sequence homology of the vertebrate ferredoxins, with algae and bacterial ferredoxins (41), represents an intriguing paradigm as to the evolution of function by divergent amino acid sequences. Structural studies on mammalian ferredoxins have yet to be initiated to determine the similarity of the molecular architecture of the vertebrate ferredoxins with respect to other ferredoxins (42, 43) of unrelated function. ACKNOWLEDGMENTS We are grateful to Dr. Claude Lazure for NHr-terminal amino acid sequence analysis. We thank Mr. Jacques Lehoux for technical assistance and Mrs. Catherine Masse for her expert secretarial assistance. This research was supported by the Medical Research Council of Canada, MRC Grant MT-8088 to J. Sygusch, MRC Grant MT-4653 to J. G. Lehoux, a F.C.A.R. award to M. F. Palin, and a FRSQ scholarship to L. Berthiaume.
REFERENCES 1. Omura, T., Sanders, E., Estabrook, R. W., Cooper, D. Y., and Rosenthal, 0. (1966) Arch. Biochem. Btiphys. 117, 660-673. 2. Takikawa, O., Gomi, T., Suhara, K., Itigaki, E., Takemori, S., and Katagiri, M. (1978) Arch. Biochem. Biophys. 190,300-306. 3. Sato, H., Ashida, N., Suhara, K., Itagaki, E., Takemori, S., and Katagiri, M. (1978) Arch. Biochem. Biophys. 190,307-314.
EXPRESSION 4. Nakamura, Y., Otsuka, H., and Tamaoki, phys. Acta 122,34-42. 5. Lambeth, J. D., Seybert, D. W., Lancaster, Kamin, H. (1982) Mol. Cell. Biochem. 45, 6. Churchill, P. F., and Kimura, T. (19’79) J.
OF RECOMBINANT
B. (1966) Biochem. BioJ. R., Salerno, J. C., and 13-31. Biol. Chem. 254,10443-
10448. 7. Chu, J. W., and Kimura, T. (1973) J. Biol. Chem. 248,5183-5187. 8. Lambeth, J. D., McCaslin, D. R., and Kamin, H. (1976) J. Biol.
Chem.251,7545-7550. 9. Lambeth,
J. D., and Kamin,
H. (1979) J. Biol. Chem. 254,
2766-
2774. 10. Katagiri, M., Takikawa, O., Sato, H., and Suhara, K. (1977) B&hem. Biophys. Res. Commun. 77,804-809. 11. Seybert, D. W., Lancaster, J. R., Lambeth, J. D., and Kamin, H. (1979) J. Biol. Chem. 254, 12,08812,098. 12. Lambeth, J. D., Geren, L. M., and Millet, F. (1984) J. Biol. Chem. 259, 10,025-10,029. 13. Coghlan, V. M., andvickery, L. E. (1991) J. Biol. Chem. 266,18,60618,612. 14. Nabi, N., and Omura, T. (1980) &o&cm. Biophys. Res. Commun.
97,680-686.
ADRENODOXIN
24. Hiwatashi, A., Ichikawa, Y., Maruya, (1976) Biochemistry 15,3082-3090.
131 N., Yamano, T., and Aki, K.
25. Katagiri,
26. 27. 28. 29.
M., Takemori, S., Itagaki, E., and Suhara, K. (1978) in Methods in Enzymology (Fleischer, S., and Packer, L., Eds.), Vol. 52, pp. 124-132, Academic Press, San Diego. Omura, T., and Sato, R. (1964) J. Biol. Chem. 239,2370-2378. Chu, J., and Kimura, T. (1973) J. Biol. Chem. 248, 2089-2094. Huang, J. J., and Kimura, T. (1973) Biochemistry 12,406-409. Monnier, N. Defaye, G., and Chambaz, E. M. (1987) Eur. J. Biochem.
169,147-153. 30. De Boer, H. A., Comstock, L. J., and Vasser, M. (1983) Proc. Natl. Acad. Sci. USA 80, 21-25. 31. Vieira, J. and Messing, J. (1982) Gene 19, 259-268. 32. Brosius, J., Dull, T. J., Sleeter, D. D., and Noller, H. F. (1981) J. Mol. Biol. 148, 107-127. 33. Suhara, K., Takemori, S., and Katagiri, M. (1972) Biochem. Biophys. Acta. 263,272-278. 34. Suzuki, K., and Kimura, T. (1965) Biochem. Biophys. Res. Commun.
19,340-345. 35. Tanaka, M., Haniu, M., Yasunobu, K. T., and Kimura, T. (1973) J. Biol. Chem. 248,1141-1157.
15. Matocha, M. F., and Waterman, M. R. (1985) J. Biol. Chem. 260, 12,259-12,265. 16. Kramer, R. E., Du Bois, R. N., Simpson, E. R., Anderson, C. M., Kashiwagi, K., Lambeth, J. D., Jefcoate, C. R., and Waterman, M. R. (1982) Arch. Biochem. Biophys. 215,478-485. 17. Okamura, T., John, M. E., Zuber, M. X., Simpson, E. R., and Waterman, M. R. (1985) Proc. Natl. Acad. Sci. USA 82, 5705-5709. 18. Okayama, H., and Berg, P. (1983) Mol. Cell. Biol. 3, 280-289. 19. Maniatis, T., Fritsch, E., and Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. 20. Taylor, J. W., Schmidt, W., Cosstick, R., Okruszek, A., and Eckstein, F. (1985) Nucleic Acids Res. 13,8749-8764. 21. Taylor, J. W., Ott, J., and Eckstein, F. (1985) Nucleic Acids Res.
13,8765-8785. 22. Sanger, F., Nicklen, S., and Coulson, A. R. (1977) Proc. Natl. Acad. Sci. USA 74,5463-5467. 23. Laemmli, U. K. (1970) Nature (London)
BOVINE
227,680-685.
36. Hanukoglu, I., Spitsberg, V., Bumpus, J. A., Dus, K. M., and Jefcoate, C. R. (1981) J. Biol. Chem. 256,4321-4328. 37. Coghlan, V. M., and Vickery, L. E. (1989) Proc. N&l. Acad. Sci. USA 86,835-839. 38. Springer, B. A., and Sligar, S. G. (1987) Proc. Natl. Acad. Sci. USA
84,8961-8965. 39. Hoffman, S. J., Looker, D. L., Roehrich, J. M., Cozart, P. E., Durfee, S. L., Tedesco, J. L., and Stetler, G. L. (1990) Proc. Natl. Acacl. Sci. USA 87,8521-8525. 40. Ikemura, T. (1981) J. Mol. Biol. 151,389-409. 41. Meyer, J., Bruschi, M. H., Bonicel, J. J., and Bovier-Lapierre, G. E. (1986) Biochemistry 25,6054-6061. 42. Fukuyama, K., Hase, T., Matsumoto, S., Tsukihara, T., Katsube, Y., Tanaka, N., Kakudo, M., Wada, K., and Matsubara, H. (1980) Nature (London) 286, 522-524. 43. Sussman, J. L., Brown, J. H., and Shokam, M. (1986) in Iron-Sulfur Protein Research (Matsubara, H., Katsube, Y., and Wada, K., Eds.), pp. 69-82, Springer, Berlin.