Heterologous synthesis of cytochrome c¢ by Escherichia coli is not dependent on the System I cytochrome c biogenesis machinery Hiroki Inoue 1 , Satoshi Wakai 1 , Hirofumi Nishihara 2 and Yoshihiro Sambongi 1 1 Graduate School of Biosphere Science, Hiroshima University, Japan 2 Faculty of Agriculture, Ibaraki University, Japan Introduction Cytochromes c¢ are classified as class II cytochromes c according to Ambler [1], and are found in the peri- plasm of certain Gram-negative Alpha-, Beta- and Gammaproteobacteria. Recent biochemical and genetic analyses have demonstrated that cytochromes c¢ mainly play roles in the cellular metabolism of nitric oxide [2], which is an electron acceptor in denitrifying bacteria and is also implicated as a signaling molecule in a wide range of organisms. The structure of cytochromes c¢ exhibits clear differ- ences from that of the well-known Ambler’s class I cytochromes c. The class I cytochromes c are spherical proteins with a hexacoordinate heme covalently bound near their N-termini. In contrast, cytochromes c¢, con- sisting of approximately 130 residues, contain a penta- coordinate heme located towards the C-terminus of a four-helix bundle protein. Escherichia coli cyto- chrome b 562 (EC b 562 ), a 106-residue protein, also has a four-helix bundle structure with a noncovalently bound heme [3]. Despite the sequence difference between cytochromes b 562 and c¢, the four helices of each nearly spatially coincide when the respective heme groups are superimposed [4]. Although knowledge concerning the function and structure of cytochromes c¢ has accumulated, their biogenesis remains unclear. In general, covalent heme Keywords cytochrome c biogenesis; cytochrome c¢; Escherichia coli; heterologous synthesis; System I Correspondence Y. Sambongi, Graduate School of Biosphere Science, Hiroshima University, 1-4-4 Kagamiyama, Higashi-Hiroshima, Hiroshima 739-8528, Japan Fax: +81 824 24 7924 Tel: +81 824 24 7924 E-mail: sambongi@hiroshima-u.ac.jp (Received 28 February 2011, revised 5 April 2011, accepted 27 April 2011) doi:10.1111/j.1742-4658.2011.08155.x Hydrogenophilus thermoluteolus cytochrome c¢ (PHCP) has typical spectral properties previously observed for other cytochromes c¢, which comprise Ambler’s class II cytochromes c. The PHCP protein sequence (135 amino acids) deduced from the cloned gene is the most homologous (55% iden- tity) to that of cytochrome c¢ from Allochromatium vinosum (AVCP). These findings indicate that PHCP forms a four-helix bundle structure, similar to AVCP. Strikingly, PHCP with a covalently bound heme was heterologously synthesized in the periplasm of Escherichia coli strains deficient in the DsbD protein, a component of the System I cytochrome c biogenesis machinery. The heterologous synthesis of PHCP by aerobically growing E. coli also occurred without a plasmid carrying the genes for Ccm pro- teins, other components of the System I machinery. Unlike Ambler’s class I general cytochromes c, the synthesis of PHCP is not dependent on the System I machinery and exhibits similarity to that of E. coli periplasmic cytochrome b 562 , a 106-residue four-helix bundle. Database The sequence data reported here have been deposited in the DDBJ database under accession no. AB617519. Abbreviations AVCP, Allochromatium vinosum cytochrome c¢; Ccm, cytochrome c maturation; Dsb, disulfide bond formation; EC b 562 , Escherichia coli cytochrome b 562 ; PHCP, Hydrogenophilus thermoluteolus cytochrome c¢;PHc 552 , Hydrogenophilus thermoluteolus cytochrome c 552 . FEBS Journal 278 (2011) 2341–2348 ª 2011 The Authors Journal compilation ª 2011 FEBS 2341 attachment to class I cytochromes c is catalyzed by the cellular machinery, resulting in cytochrome c biogene- sis [5]. For example, in some Gram-negative bacteria, such as E. coli, the System I cytochrome c biogenesis machinery, consisting of some disulfide bond forma- tion (Dsb) and cytochrome c maturation (Ccm) proteins, is responsible for the biogenesis of a wide variety of both endogenous and exogenous class I cy- tochromes c [6]. Successful heterologous synthesis of several cytochromes c¢ has been reported using aerobi- cally growing E. coli with co-expressed ccm genes from a plasmid [7–9]. However, a variant of EC b 562 , which has been mutated so as to bind heme covalently like cytochromes c, can be formed as a holo-protein with- out co-expressed ccm genes from a plasmid [10,11]. Although the heme-binding mode of the resulting EC b 562 variant differs from that with co-expressed ccm genes from a plasmid, its holo-formation is obvi- ous. This prompted us to re-examine the heterologous synthesis of cytochromes c¢ with or without co- expressed ccm genes from a plasmid. In addition, the effects of Dsb proteins on cytochrome c¢ synthesis have not been examined to date. In this study, we examined the heterologous synthesis of cytochrome c¢ proteins by E. coli strains deficient in the DsbD protein and co-expressing or not co-express- ing ccm genes from a plasmid. For this purpose, we first purified and characterized Hydrogenophilus thermoluteo- lus cytochrome c¢ (PHCP). Secondly, the PHCP gene was cloned for sequence and expression analyses. Heter- ologous synthesis of the PHCP protein by the E. coli strains was investigated in direct comparison with that of H. thermoluteolus cytochrome c 552 (PH c 552 ), which is a typical class I cytochrome c that has been demon- strated to be System I dependent with regard to its bio- genesis in E. coli [12,13]. Our results provide information on the biogenesis of cytochromes c¢, which has not been studied systematically. Results Purification of the PHCP protein The PHCP protein was purified to homogeneity by col- umn chromatography, as illustrated on an SDS ⁄ PAGE gel (Fig. 1). The estimated molecular weight of the PHCP protein on the gel was 13 kDa, which was close to that of other cytochromes c¢ isolated from various bacteria. The N-terminal amino acid sequence of PHCP was determined up to the 30th residue, as illus- trated in Fig.2.Ablast search indicated that the pro- tein sequence determined up to the 30th residue was homologous to that of other cytochromes c¢ isolated from other bacteria. Thus, at this stage of the present work, we concluded that the purified PHCP protein was a novel cytochrome c¢ isolated from H. thermolute- olus. Spectral properties of the authentic PHCP protein Visible absorption spectra of the authentic PHCP pro- tein purified from H. thermoluteolus were obtained to examine the local heme environment in the protein interior. The spectra of the oxidized and reduced PHCP were essentially the same as those reported for other cytochromes c¢ (Fig. 3A), indicating that the heme environment in the PHCP protein was similar to that in others. Specifically, a Soret band at 425 nm was observed for the reduced form of PHCP, which is characteristic of a pentacoordinate heme with a His residue as an axial ligand [14]. Furthermore, a peak around 630 nm was observed for the oxidized form of PHCP, indicating that the position of the sixth ligand to the heme iron is empty, as discussed for other cyto- chromes c¢ [14]. In addition, the a-band in the pyridine hemochrome spectrum of reduced PHCP corresponded to 550 nm, which is indicative of the covalent bonding of heme vinyl groups to the protein via two thioether linkages. A far-UV CD spectrum (190–260 nm) was obtained to examine the secondary structure of the PHCP Mw (kDa) 1 2 3 4 75 25 20 15 10 5 5 Fig. 1. Purification of the Hydrogenophilus thermoluteolus cyto- chrome c¢ (PHCP) protein. Lane 1, total soluble extract of H. therm- oluteolus cells; lane 2, HiTrap Q batch elution with 0.2 M NaCl; lane 3, HiTrap Q linear gradient elution with 0–0.2 M NaCl; lane 4, HiTrap SP flow-through elution; lane 5, Sephadex 75 elution. The arrow indicates the position of PHCP. One to ten micrograms of protein were loaded per lane, and the gel was stained with Coo- massie Brilliant Blue. Biogenesis of cytochrome c¢ H. Inoue et al. 2342 FEBS Journal 278 (2011) 2341–2348 ª 2011 The Authors Journal compilation ª 2011 FEBS protein. From the ellipticity peak height of the PHCP protein at 222 nm (Fig. 3B), its helical content was cal- culated to be 60.3% [15]. This value is close to the a-helical content of Allochromatium vinosum cyto- chrome c¢ (AVCP), i.e. 63.0%, which was calculated directly from its primary (Fig. 2) and three-dimen- sional [16] structures. Cloning of the PHCP gene PCR with mixed primers PHcp01fw and PHcp01rv, using H. thermoluteolus chromosomal DNA as a tem- plate, gave a DNA fragment of approximately 360 bp, which was then cloned into the pUC19 vector. At least five independent clones were sequenced, and the amino acid sequence (25th to 125th residues, Fig. 2) deduced from the DNA was homologous to the sequences of cytochromes c¢ deposited previously in the database. Using the inverse PCR method, we obtained a single 6.5-kbp DNA fragment from an SphI-digested H. thermoluteolus chromosomal DNA library. DNA sequencing of the fragment revealed that the product contained the 5¢ and 3¢ ends of the PHCP gene plus putative promoter, Shine–Dalgarno and transcriptional terminator sequences. From the deduced sequence, the mature PHCP was found to consist of 135 amino acids, and the N-terminal Asp was preceded by a Sec- dependent periplasmic targeting signal peptide of 19 amino acid residues (Fig. 2). This indicates that the PHCP protein is synthesized as a precursor, and that its signal peptide is cleaved off during translocation to the periplasm of H. thermoluteolus cells. From the amino acid sequence deduced from the cloned PHCP gene, the heme-binding motif observed in general cytochromes c, Cys–X–X–Cys–His, was found to be located close to the C-terminus of the PHCP protein, which is conserved in other biochemi- cally characterized cytochromes c¢ (Fig. 2). The mature PHCP protein exhibited overall sequence identity of 54.8% to AVCP, this being the highest identity among the homologs in the genome database. Heterologous synthesis of the PHCP and PH c 552 proteins by E. coli The cloned PHCP gene, together with the typical class I PH c 552 gene as a reference control, was exam- ined with regard to its heterologous expression in various E. coli strains by means of heme-specific stain- ing of SDS ⁄ PAGE gels. On such gels, when stained materials are observed at positions coinciding with those of PHCP and PH c 552 , the proteins each have a covalently attached heme, which is defined here as completion of cytochrome c synthesis. The PHCP protein was heterologously synthesized in the periplasm of anaerobically growing E. coli dsbD 20 30 40 50 60 70 80 90 100 110 120 130 1 10 mkriamitaltlcaaaahaDALKPEDKVKFRQAS mkhvlastaaglmalgl-assaiaAGLSPEEQIETRQAG mkklstlaalacmtvgsll-atsaqaQFAKPEDAVKYRQSA mrrvllatlmaalpaaaMAADAEHVVEARKGY (1) H. thermoluteolus (2) A. vinosum (3) A. xylosoxidans (4) R. sphaeroides YTTMAWNMGKIKAMVVDGTMPFSQTQVSAAANVIAAIANSGMGALYSPDTLGVVGFKKSR YEFMGWNMGKIKA-NLEGE YNAAQVEAAANVIAAIANSGMGALYGPGTDKNVGDVKTR LTLMASHFGRMTP-VVKGQAPYDAAQIKANVEVLKTLSAL-PWAAFGPGTEGG-D FSLVALEFGPLAAM-AKGEMPYDAAAAKAHASDLVTLTKYDPSDLYAPGTSAD-DVKGTA LKENFFQEQDEVRKIATNFVEQANKLAEVAAMGDKDEIKAQFGEVGKACKACHEKFREEE VKPEFFQNMEDVGKIAREFVGAANTLAEVAATGEAEAVKTAFGDVGAACKSCHEKYRAK- ARPEIWSDAASFKQKQQAFQDNIVKLSAAADAGDLDKLRAAFGDVGASCKACHDAYRKKK AKAAIWQDADGFQAKGMAFFEAVAALEPAAGAGQKE-LAAAVGKVGGTCKSCHDDFRVKR * ** * * * * * * Fig. 2. Multiple sequence analysis of biochemically characterized cytochrome c¢ proteins. Experimentally determined signal peptides are depicted in lower case letters. The numbering on the Hydrogenophilus thermoluteolus cytochrome c¢ (PHCP) sequence is that of the mature protein. The sequence of PHCP was chemically determined up to the 30th residue in this study and confirmed by the protein sequence deduced from the cloned gene. The stretches of the PHCP amino acid sequence used for the design of the PCR primers are underlined with arrows indicating the 5¢ to 3¢ direction. The sequences of biochemically characterized cytochromes c¢ were obtained from a database: (2) locus tag of Alvin_2765 of Allochromatium vinosum DSM180; (3) accession number P00138 of Achromobacter xylosoxidans NCIMB11015; (4) locus tag of RSP_0474 of Rhodobacter sphaeroides 2.4.1. The consensus cytochrome c Cys–X–X–Cys–His heme-binding motif is close to the C-terminus of each protein. Gaps in the alignment are indicated by dashes. Identical residues to those in PHCP are highlighted in gray. Helical regions determined from the crystal structure of A. vinosum cytochrome c¢ (AVCP) are underlined. A residue occupying the empty sixth ligand to the heme iron and hydrophobic residues in contact with the heme in the AVCP structure are indicated by asterisks above the sequence (see details in the Discussion section). H. Inoue et al. Biogenesis of cytochrome c¢ FEBS Journal 278 (2011) 2341–2348 ª 2011 The Authors Journal compilation ª 2011 FEBS 2343 null mutant strain RI242, whereas the PH c 552 protein was not (Fig. 4). The isogenic wild-type E. coli RI89 strain with the intact DsbD protein was able to heter- ologously synthesize the PHCP and PH c 552 proteins, confirming that the observed difference between the two proteins in the RI242 strain is a result of the absence of the DsbD protein. PHCP was also synthesized as a holo-protein in the periplasm of aerobically growing E. coli JCB387 cells not harboring the pEC86 plasmid carrying the ccm genes (Fig. 4). In contrast, when it did not harbor the plasmid, the E. coli JCB387 strain was not able to produce PH c 552 aerobically. These results indicate that the present growth conditions in the absence of pEC86 do not confer the cytochrome c biogenesis ability to the PH c 552 protein. This is possibly a result of the shortage of Ccm proteins, because the expression of ccm genes is repressed under aerobic growth condi- tions. In the presence of the pEC86 plasmid, both the PHCP and PH c 552 proteins were heterologously syn- thesized in the periplasm of E. coli JCB387 cells (Fig. 4). Judging from the staining intensity, the level of production of the PHCP protein in the presence of the pEC86 plasmid was significantly lower than that without the plasmid, indicating that the co-expression of plasmid-borne ccm genes represses PHCP overpro- duction by aerobically growing E. coli cells. A similar difference in the PHCP production level was observed in the early and late logarithmic and stationary phases of E. coli JCB387 cells with and without the pEC86 plasmid. Spectral properties of PHCP heterologously synthesized by E. coli The visible absorption spectra of periplasmic extracts containing the PHCP protein heterologously synthe- sized by E. coli RI242 and JCB387 without pEC86 were the same as those observed for the authentic Wavelength (nm) Oxidized PHCP Reduced PHCP Absorbance (A.U.) 300 400 425 630 500 600 700 0.0 0.1 0.2 0.3 0.4 0.5 0.6 200 220 240 260 –20 –10 0 10 20 30 40 [θ] × 10 –3 (deg·cm 2 ·dmol –1 ) Wavelength (nm) 222 B A Fig. 3. Spectral analysis of the authentic Hydrogenophilus thermo- luteolus cytochrome c¢ (PHCP) protein: (A) visible absorption spec- tra; (B) CD spectra. Specific wavelengths referred to in the text are indicated by arrows in (A) and (B). 20 –pEC86 c' c'c 552 c 552 c' c 552 c' c 552 Mw (kDa) RI242 JCB387 15 10 * 5 RI89 +pEC86 Fig. 4. Heterologous synthesis of cytochromes c by Escherichia coli strains. Periplasmic extracts (equivalent to 5 · 10 8 cells) of the E. coli RI242 and RI89 strains, and the JCB387 strain without (indi- cated by –) or with (indicated by +) the pEC86 plasmid carrying the ccm genes, were analyzed by heme staining after SDS ⁄ PAGE. In each lane of the gel, periplasmic extracts from the E. coli cells transformed with the Hydrogenophilus thermoluteolus cyto- chrome c¢ (PHCP) and H. thermoluteolus cytochrome c 552 (PH c 552 ) genes are indicated as c¢ and c 552 , respectively. The arrow and arrowhead indicate the positions of the PHCP and PH c 552 proteins, respectively. The band denoted by the asterisk on the right-hand side is the result of nonspecific staining of the extracts containing the PH c 552 protein. Biogenesis of cytochrome c¢ H. Inoue et al. 2344 FEBS Journal 278 (2011) 2341–2348 ª 2011 The Authors Journal compilation ª 2011 FEBS purified protein in both the oxidized and reduced states (Fig. 3A). These findings indicate that the heme is correctly incorporated into the apo-form of PHCP heterologously synthesized by E. coli, even without the DsbD protein and without co-expression of the ccm genes from the pEC86 plasmid. In addition, the a-band in the pyridine hemochrome spectra of the same periplasmic extracts with dithionite corresponded to 550 nm, as observed for the authentic PHCP pro- tein, indicating the covalent attachment of the heme to the protein through two thioether bonds. Discussion In this study, we attempted to determine whether or not Ambler’s class II cytochromes c¢ are synthesized by the System I cytochrome c biogenesis machinery. For this purpose, we first performed spectral analysis of the authentic PHCP protein, aiming at the predic- tion of its structure, which is the final state of biogene- sis. Secondly, the PHCP gene was cloned to gain sequence information and to examine its heterologous expression in E. coli strains with reference to PH c 552 , which has been characterized as a System I-dependent cytochrome c. Spectral properties of the authentic PHCP protein The visible absorption and CD spectral features of the PHCP protein indicate that its local heme environment and helical content are similar to those found in typi- cal cytochromes c¢. In the four-helix bundle structure of general cytochromes c¢, access to the sixth ligand position with regard to the heme iron is hindered primarily by the side-chains of aromatic or nonaromat- ic hydrophobic residues. Such a responsible residue is Tyr16 in the crystal structure of the AVCP protein [16]. The same residue is also conserved in the PHCP protein (Fig. 2). Other residues responsible for the maintenance of the hydrophobic environment around the heme in AVCP are Met19, Gly20, Met23, Tyr61, Val76, Phe80, Val87, Val95 and Val120 (PHCP numbering, Fig. 2), which directly face the heme group [16]. Of these nine residues, seven are identical in PHCP, the other two, Gly20 and Val76, in AVCP being homolo- gously replaced by Ala20 and Leu76, respectively, in PHCP. These sequence similarities, together with the spectral properties observed for the PHCP and AVCP proteins, indicate that the former has a three-dimen- sional structure comprising a four-helix bundle, as demonstrated for other cytochromes c¢, including the latter. Heterologous synthesis of PHCP by E. coli The E. coli System I cytochrome c biogenesis machin- ery, consisting of the Dsb and Ccm proteins, is respon- sible for the synthesis of class I cytochromes c even from various exogenous sources [6]. Normally, the E. coli chromosomal ccm genes are not aerobically expressed. Therefore, through co-expression of the ccm genes in the pEC86 plasmid, together with various class I cytochrome c genes, holo-cytochromes c can be successfully overproduced by aerobically growing E. coli cells. In previous studies, it has been shown that co-expression of the ccm genes in the pEC86 plasmid is required for the heterologous expression of class II cytochromes c¢ by E. coli [7–9], predicting that cyto- chrome c¢ biogenesis is System I dependent. However, systematic studies on the effects of the ccm and dsb genes with reference controls have not been performed. It is clear from our results that the co-expression of the ccm genes in the pEC86 plasmid and the presence of the DsbD protein are not necessarily required for the heterologous synthesis of the PHCP protein by E. coli, unlike that of class I cytochromes c, including the PH c 552 protein. Similarity to and differences from periplasmic EC b 562 Previously, the c-type heme-binding Cys–X–X–Cys– His motif was introduced into periplasmic EC b 562 in order to determine whether or not the resulting variant is synthesized as a holo-protein with a covalently bound heme [10]. Even without the DsbD protein or without co-expression of the ccm genes in a plasmid, the EC b 562 variant can be formed as a holo-protein with a covalently bound heme [11]. Although the pro- duction level and heme-binding mode of the EC b 562 variant under these conditions differ from those with the dsbD gene product or with the co-expression of the ccm genes in a plasmid, holo-protein synthesis clearly occurs with such an imperfect System I cytochrome c biogenesis machinery. Therefore, the EC b 562 variant resembles the PHCP protein in terms of biogene- sis, which is different from System I cytochrome c biogenesis. The above EC b 562 variant with the c-type heme-bind- ing Cys–X–X–Cys–His motif was further modified so as to add extra Cys residues around the motif. The result- ing variants were examined for heterologous synthesis by E. coli JCB387 with or without the pEC86 plasmid, it being shown that co-expression of the ccm genes in the plasmid caused enhanced levels of production of the variants [17]. These observations are not consistent with H. Inoue et al. Biogenesis of cytochrome c¢ FEBS Journal 278 (2011) 2341–2348 ª 2011 The Authors Journal compilation ª 2011 FEBS 2345 those in the present study, in which the co-expression of the ccm genes in the pEC86 plasmid was found to result in a low level of production of PHCP (Fig. 4). There is presently no explanation as to why the production levels differ between the PHCP protein and the EC b 562 variant. Further experiments on the two proteins with the same growth medium and aerobicity are required for a clear comparison, which will provide information on the function of Ccm proteins. Structural implication for PHCP synthesis Although the sequence identity is low between the PHCP protein and the EC b 562 variant, they may have the same architecture, comprising a four-helix bundle structure, indicating that their folding mechanisms, including heme attachment, are conserved, as sug- gested previously [18]. A large portion of the EC b 562 protein can fold in the absence of heme to yield its apo-form with an empty heme-binding site [19]. Should such a folding process in apo-EC b 562 also occur in apo-PHCP, the latter protein may incorporate free heme, which is then spontaneously bound in a System I-independent manner. Although no direct evi- dence for this is available, hydrophobic interactions within apo-PHCP may facilitate protein folding in the absence of heme, as observed for Aquifex aeolicus class I cytochrome c 555 , whose apo-form is exception- ally folded [20,21]. It would be of interest to investi- gate further the biogenesis mechanism for PHCP with regard to the relation to its structural features in conjunction with a mutagenesis study. Materials and methods Purification of PHCP from H. thermoluteolus Hydrogenophilus thermoluteolus TH-1 [22] was cultured at 45 °C in an inorganic medium under H 2 :O 2 :CO 2 (75 : 15 : 10). The constituents of this medium have been given previously [23]. The H. thermoluteolus cells (30 g wet weight) were resuspended in 210 mL of 10 mm Tris ⁄ HCl (pH 8.0). The cells were then disrupted with a French pres- sure cell, followed by centrifugation (200 000 g) to obtain a total soluble extract. The resulting soluble extract was dialyzed against 10 mm Tris ⁄ HCl (pH 8.0) at 4 °C, and then loaded onto a Hi- Trap Q anion-exchange column (diameter, 1.4 cm; height, 3 cm; GE Healthcare, Tokyo, Japan) that had been equili- brated with 10 mm Tris ⁄ HCl (pH 8.0). Batch elution was carried out with 50 mL of the same buffer containing 0, 0.2 or 1.0 m NaCl, a red-colored fraction containing the PHCP protein being eluted with 0.2 m NaCl. The red-colored fraction was further dialyzed against 10 mm Tris ⁄ HCl (pH 8.0), and then loaded onto the same column that had been equilibrated with the same buffer. Proteins were eluted with a linear gradient of NaCl (0–0.2 m). The resulting red fraction was dialyzed against 25 mm sodium acetate (pH 5.5), and then loaded onto a HiTrap SP cation- exchange column (diameter, 1.4 cm; height, 3 cm; GE Healthcare) that had been equilibrated with the same buf- fer. The fraction containing the PHCP protein flowed through, and was finally separated by gel filtration on a column of Sephadex 75 (diameter, 1.6 cm; height, 60 cm; GE Healthcare) that had been equilibrated with 25 mm sodium acetate (pH 5.5). Characterization of the purified PHCP protein Protein purity during the column chromatography steps was checked by SDS ⁄ PAGE and staining with Coomassie Brilliant Blue. The gels were also subjected to heme stain- ing, proteins with covalently bound heme being stained to detect cytochrome c specifically [24]. The band correspond- ing to the PHCP protein on a gel was blotted onto a polyv- inylidene fluoride membrane (Millipore, Tokyo, Japan) for direct protein sequencing analysis with an automatic protein sequencer (Applied Biosystems, Tokyo, Japan). The protein concentrations of the crude extracts were deter- mined with a protein assay kit (Bio-Rad, Tokyo, Japan) with bovine serum albumin as a standard. For the purified PHCP protein, the concentrations were determined spectro- photometrically using the extinction coefficient at 205 nm caused by the peptide bond [25]. Visible absorption and CD spectra of the purified authentic PHCP protein in 10 mm potassium phosphate buffer (pH 7.0) were obtained with JASCO V-530 and JASCO J-820 spectrometers, respectively, at 25 °C. The PHCP protein was air oxidized or reduced with a grain of sodium dithionite. The protein concentrations were 6 and 20 lm for visible absorption and CD spectral analysis, respectively. Pyridine hemochrome spectra were obtained according to the method described by Bartsch [26]. Isolation of full-length DNA encoding the PHCP protein In order to clone the PHCP gene and to determine the complete DNA sequence, we used the PCR method. From N-terminal sequence information on the PHCP protein up to the 30th residue, we designed 512 mixed forward primers (PHcp01fw) corresponding to the resulting PHCP protein sequence Glu-Asp-Lys-Val-Lys-Phe-Arg-Glu-Ala (5th to 14th residues of the mature PHCP sequence, see Fig. 2), and 18 432 mixed reverse primers (PHcp01rv) correspond- ing to the well-conserved cytochrome c¢ sequence Cys-Lys- Ala-Cys-His-Asp-X-Tyr-Arg (124th to 132nd residues in the case of PHCP, Fig. 2; X denotes any residue), and used Biogenesis of cytochrome c¢ H. Inoue et al. 2346 FEBS Journal 278 (2011) 2341–2348 ª 2011 The Authors Journal compilation ª 2011 FEBS them to amplify H. thermoluteolus chromosomal DNA with Ex Taq polymerase (Takara, Shiga, Japan). The DNA frag- ment obtained from PCR was sequenced and found to code a part of the PHCP protein. We next used the inverse PCR method to obtain the entire PHCP gene. DNA fragments that had been prepared by digestion of H. thermoluteolus chromosomal DNA with sev- eral restriction enzymes separately were self-ligated and then used as the first PCR templates with a gene-specific reverse primer, PHcp03rv, corresponding to the PHCP protein sequence of the 31st to 40th residues (Fig. 2), and a gene-spe- cific forward primer, PHcp03fw, corresponding to the sequence of the 45th to 53rd residues. The resulting PCR products were then used as the second PCR templates with a gene-specific nested reverse primer, PHcp04rv, correspond- ing to the protein sequence of the 27th to 31st residues, and a gene-specific nested forward primer, PHcp04fw, corre- sponding to the sequence of the 106th to 111th residues. Heterologous synthesis of the PHCP and PH c 552 proteins by E. coli Escherichia coli DH5a was used for the maintenance and propagation of all plasmids. The E. coli RI89, RI242 and JCB387 strains were examined with regard to the synthesis of exogenous PHCP and PH c 552 proteins. The RI89 strain is a parental strain of RI242, which is a dsbD null mutant [27], and the JCB387 strain is usually used for heterologous synthesis of cytochromes c in our laboratory [28]. These strains were transformed with pKK223-3 derivatives carry- ing the PHCP or PH c 552 gene (ampicillin resistance). The original signal sequence of PHCP was replaced with that of Pseudomonas aeruginosa cytochrome c 551 to target the PHCP apo-protein to the E. coli periplasm by the PCR method described previously for PH c 552 [12]. The resulting PHCP gene was flanked by artificially introduced restriction sites (EcoRI, 5¢ and SalI, 3¢), and then inserted into the corresponding sites of pKK223-3. The E. coli JCB387 strain was further co-transformed with pEC86 [29], which carries the E. coli cytochrome c maturation genes ccmABCDEFGH (chloramphenicol resistance). The transformed E. coli RI89, RI242 and JCB387 cells were grown in LB liquid medium containing appropriate antibiotics overnight at 37 °C. The resulting precultures of RI89 and RI242 cells were each inoculated into 50 mL of minimal medium supplemented with 0.4% (v ⁄ v) glycerol as a carbon source, and with nitrite and fumarate as substrates for respiration, in a screw capped bottle, which was then incubated anaerobically for 24 h at 37 °C [30]. The preculture of the JCB387 strain was inoculated into 20 mL of the same minimal medium supplemented with 0.4% (v ⁄ v) glycerol in a 50-mL flask, which was then incu- bated aerobically for 16 h at 37 °C [28]. The growing E. coli cells at the late logarithmic phase were harvested. Periplasmic extracts of these cells were obtained by the cold osmotic shock method [31], and then subjected to SDS ⁄ PAGE, followed by heme staining of the gels in order to detect holo-cytochromes c [24]. The same extracts were subjected to visible absorption spectral analysis, as carried out for the purified PHCP protein. Reagents Restriction enzymes, T4 DNA ligase and other reagents for DNA handling were purchased from Takara. All other chemicals used were of the highest grade commercially available. Acknowledgements We wish to thank D. Miyake, R. Sano and S. Fujii (Hiroshima University) for technical assistance. This work was partly supported by a Grant-in-Aid for Scientific Research on Innovative Areas (No. 20118005) from the Ministry of Education, Culture, Sports, Science and Technology of Japan. References 1 Ambler RP (1982) The structure and classification of cytochromes c.InFrom Cyclotrons to Cytochromes (Kaplan NO & Robinson A eds), pp 263–280. Academic Press, New York. 2 Ascenzi P, Santucci R, Coletta M & Polticelli F (2010) Cytochromes: reactivity of the ‘dark side’ of the heme. Biophys Chem 152, 21–27. 3 Hamada K, Bethge PH & Mathews SF (1995) Refined structure of cytochrome b 562 from Escherichia coli at 1.4 A ˚ resolution. J Mol Biol 247, 947–962. 4 Weber PC, Salemme FR, Mathews FS & Bethge PH (1981) On the evolutionary relationship of the 4-a-heli- cal heme proteins. J Biol Chem 256, 7702–7704. 5 Page MD, Sambongi Y & Ferguson SJ (1998) Contrast- ing routes of c-type cytochrome assembly in mitochon- dria, chloroplast and bacteria. Trends Biochem Sci 23, 103–108. 6 Sambongi Y, Uchiyama S, Kobayashi Y, Igarashi Y & Hasegawa J (2002) Cytochrome c from a thermophilic bacterium has provided insights into the mechanism of protein maturation, folding, and stability. Eur J Bio- chem 269, 3355–3361. 7 Harris RL, Barbieri S, Paraskevopoulos K, Murphy LM, Eady RR, Hasnain SS & Sawers RG (2010) Characterization of cycP gene expression in Achromo- bacter xylosoxidans NCIMB 11015 and high-level heterologous synthesis of cytochrome c¢ in Escherichia coli. J Mol Microbiol Biotechnol 18, 102–108. 8 Evers TH & Merkx M (2005) Successful recombinant production of Allochromatium vinosum cytochrome c¢ H. Inoue et al. Biogenesis of cytochrome c¢ FEBS Journal 278 (2011) 2341–2348 ª 2011 The Authors Journal compilation ª 2011 FEBS 2347 requires coexpression of cmm genes in heme-rich Escherichia coli JCB712. Biochem Biophys Res Commun 327, 668–674. 9 McGuirl MA, Lee JC, Lyubovitsky JG, Thanyakoop C, Richards JH, Gray HB & Winkler JR (2003) Cloning, heterologous expression, and characterization of recom- binant class II cytochromes c from Rhodopseudomonas palustris. Biochim Biophys Acta 1619, 23–28. 10 Allen JW, Barker PD & Ferguson SJ (2003) A cyto- chrome b 562 variant with a c-type cytochrome CXXCH heme-binding motif as a probe of the Escherichia coli cytochrome c maturation system. J Biol Chem 278, 52075–52083. 11 Barker PD, Nerou EP, Freund SM & Fearnley IM (1995) Conversion of cytochrome b 562 to c-type cyto- chromes. Biochemistry 34, 15191–15203. 12 Ichiki S, Nakamura S, Ohkubo T, Kobayashi Y, Haseg- awa J, Uchiyama S, Nishihara H, Mizuta K & Sam- bongi Y (2005) Cloning, expression, crystallization and preliminary X-ray characterization of cytochrome c 552 from a moderate thermophilic bacterium, Hydrogeno- philus thermoluteolus. Acta Crystallogr 61, 395–398. 13 Kojima N, Yamanaka M, Ichiki S & Sambongi Y (2005) Unexpected elevated production of Aquifex aeoli- cus cytochrome c 555 in Escherichia coli cells lacking disulfide oxidoreductases. Biosci Biotechnol Biochem 69, 1418–1421. 14 Kruglik SG, Lambry JC, Cianetti S, Martin JL, Eady RR, Andrew CR & Negrerie M (2007) Molecular basis for nitric oxide dynamics and affinity with Alcaligenes xylosoxidans cytochrome c¢. J Biol Chem 282, 5053–5062. 15 Chen YH, Yang JT & Martinez HM (1972) Determina- tion of the secondary structures of proteins by circular dichroism and optical rotatory dispersion. Biochemistry 11, 4120–4131. 16 Ren Z, Meyer T & McRee DE (1993) Atomic structure of a cytochrome c¢ with an unusual ligand-controlled dimer dissociation at 1.8 A ˚ resolution. J Mol Biol 234, 433–445. 17 Allen JW, Sawyer EB, Ginger ML, Barker PD & Fer- guson SJ (2009) Variant c-type cytochromes as probes of the substrate specificity of the E. coli cytochrome c maturation (Ccm) apparatus. Biochem J 419, 177–184. 18 Goldenberg DP (1999) Finding the right fold. Nat Struct Biol 6, 987–990. 19 Feng Y, Sliger SG & Wand AJ (1994) Solution structure of apocytochrome b 562 . Nat Struct Biol 1, 30–35. 20 Yamanaka M, Mita H, Yamamoto Y & Sambongi Y (2009) Heme is not required for Aquifex aeolicus cyto- chrome c 555 polypeptide folding. Biosci Biotechnol Biochem 73, 2022–2025. 21 Yamanaka M, Masanari M & Sambongi Y (2011) Con- ferment of folding ability to a naturally unfolded apocy- tochrome c through introduction of hydrophobic amino acid residues. Biochemistry 50, 2313–2320. 22 Hayashi NR, Ishida T, Yokota A, Kodama T & Igarashi Y (1999) Hydrogenophilus thermoluteolus gen. nov., sp. nov., a thermophilic, facultatively chemolitho- autotrophic, hydrogen-oxidizing bacterium. Int J Syst Bacteriol 49, 783–786. 23 Goto E, Kodama T & Minoda Y (1978) Growth and taxonomy of thermophilic hydrogen bacteria. Agric Biol Chem 42, 1305–1308. 24 Goodhew CF, Brown KR & Pettigrew GW (1986) Haem staining in gels, a useful tool in the study of bac- terial c-type cytochromes. Biochim Biophys Acta 852, 288–294. 25 Scopes RK (1974) Measurement of protein by spectro- photometry at 205 nm. Anal Biochem 59, 277–282. 26 Bartsch RG (1971) Cytochromes: bacterial. Methods Enzymol 23, 344–363. 27 Rietsch A, Belin D, Martin N & Beckwith J (1996) An in vivo pathway for disulfide bond isomerization. Proc Natl Acad Sci USA 93, 13048–13053. 28 Oikawa K, Nakamura S, Sonoyama T, Ohshima A, Kobayashi Y, Takayama SJ, Yamamoto Y, Uchiyama S, Hasegawa J & Sambongi Y (2005) Five amino acid residues responsible for the high stability of Hydroge- nobacter thermophilus cytochrome c 552 : reciprocal mutation analysis. J Biol Chem 280, 5527–5532. 29 Arslan E, Schulz H, Zufferey R, Kunzler P & Tho ¨ ny- Meyer L (1998) Overproduction of the Bradyrhizobium japonicum c-type cytochrome subunits of the cbb 3 oxi- dase in Escherichia coli. Biochem Biophys Res Commun 251, 744–747. 30 Sambongi Y & Ferguson SJ (1996) Mutants of Escheri- chia coli lacking disulphide oxidoreductases DsbA and DsbB cannot synthesise an exogenous monohaem c-type cytochrome except in the presence of disulphide com- pounds. FEBS Lett 398, 265–268. 31 Sambongi Y, Stoll R & Ferguson SJ (1996) Alteration of haem-attachment and signal-cleavage sites for Para- coccus denitrificans cytochrome c 550 probe pathway of c-type cytochrome biogenesis in Escherichia coli. Mol Microbiol 19, 1193–1204. Biogenesis of cytochrome c¢ H. Inoue et al. 2348 FEBS Journal 278 (2011) 2341–2348 ª 2011 The Authors Journal compilation ª 2011 FEBS . Heterologous synthesis of cytochrome c by Escherichia coli is not dependent on the System I cytochrome c biogenesis machinery Hiroki Inoue 1 , Satoshi. cytochromes c , including the latter. Heterologous synthesis of PHCP by E. coli The E. coli System I cytochrome c biogenesis machin- ery, consisting of the