792
Biochemical Society Transactions (2005) Volume 33, part 4
Covalent cofactor attachment to proteins: cytochrome c biogenesis J.M. Stevens*1 , T. Uchida†, O. Daltrop* and S.J. Ferguson* *Department of Biochemistry, University of Oxford, South Parks Road, Oxford OX1 3QU, U.K., and †Okazaki Institute for Integrative Bioscience, National Institutes of Natural Sciences, Myodaiji, Okazaki, Aichi 444-8787, Japan
Abstract Haem (Fe-protoporphyrin IX) is a cofactor found in a wide variety of proteins. It confers diverse functions, including electron transfer, the binding and sensing of gases, and many types of catalysis. The majority of cofactors are non-covalently attached to proteins. There are, however, some proteins in which the cofactor binds covalently and one of the major protein classes characterized by covalent cofactor attachment is the c-type cytochromes. The characteristic haem-binding mode of c-type cytochromes requires the formation of two covalent bonds between two cysteine residues in the protein and the two vinyl groups of haem. Haem attachment is a complex post-translational process that, in bacteria such as Escherichia coli, occurs in the periplasmic space and involves the participation of many proteins. Unexpectedly, it has been found that the haem chaperone CcmE (cytochrome c maturation), which is an essential intermediate in the process, also binds haem covalently before transferring the haem to apocytochromes. A single covalent bond is involved and occurs between a haem vinyl group and a histidine residue of CcmE. Several in vitro and in vivo studies have provided insight into the function of this protein and into the overall process of cytochrome c biogenesis.
Covalent cofactor attachment Proteins have acquired a wide variety of cofactors that provide them with functions they would not otherwise be able to perform. Cofactors vary in their complexity from single atoms (such as Fe, Zn, Mg, Cu and Co) to, for example, Fe-S clusters, to much more complex organic and organometallic molecules such as haem, chlorophyll, flavin, ascorbate, cobalamin, PQQ (pyrroloquinoline quinone), NADH, FAD and many others. In many of these cases, the cofactors are non-covalently attached to the polypeptides. However, there are a number of cases where the cofactor is attached to the protein through covalent bonds. In some flavoproteins, the flavin cofactor is covalently attached to the protein, for example in succinate dehydrogenase [1]. The flavin is attached to tyrosine, cysteine or (most commonly) histidine residues and this occurs by an autocatalytic pathway, at least in some cases [2]. Biotinylation of proteins on specific lysine residues, on the other hand, requires the participation of additional proteins [3]. The pyridoxal phosphate cofactor in alliinase is covalently attached to a lysine residue [4], as is the case with 5-aminolevulinate synthase [5] and other proteins. In some cases, covalently linked cofactors are protein-derived and are formed by post-translational amino acid modification; an example is TTQ (tryptophan tryptophylquinone) formed in methylamine dehydrogenase [6]. A classical and widely studied example of covalent cofactor attachment is the c-type cytochrome class of proteins, in Key words: CcmE, chaperone, cofactor, covalent bond, cytochrome c, haem. Abbreviations used: Ccm, cytochrome c maturation; RR, resonance Raman. 1 To whom correspondence should be addressed (email
[email protected]).
C 2005
Biochemical Society
which the cofactor haem is attached to the protein through two thioether bonds [7]. Haem is a widely distributed cofactor found in proteins with diverse functions such as the binding of gases (haemoglobin), electron transport in the respiratory chain (cytochrome c) and many others. In addition to being non-covalently bound to many of these proteins, haem is found to attach covalently to others with different modes of attachment. In mammalian peroxidases, haem is covalently attached to the protein through ester bonds to haem methyl groups in an autocatalytic process [8]. Other covalent tetrapyrrole adducts have been observed: for example, a haem–myoglobin complex that is formed during the reductive debromination of BrCCl3 , in which a histidine residue becomes covalently attached to a haem vinyl group [9]. There are various cases in which autocatalytic reactions cause non-physiological modifications of haem, as has been observed in cytochromes P450 [8]. In general, it is thought that covalent cofactor attachment ensures proximity and retention of the cofactor, and there may be other advantages and considerations in specific cases. In the case of haem, various arguments have been proposed and they are reviewed elsewhere [10–12]. Currently, the most convincing argument is that very dense haem packing can only be achieved when the haems are covalently attached to the protein, with as many as over 30 per polypeptide. The toxicity of free haem, due to its ability to produce hydroxyl radicals and to bind to membranes, could also be a consideration. We discuss below the process of covalent attachment of haem to cytochromes, known as cytochrome c maturation (Ccm), and focus also on the bacterial haem chaperone
Coenzymology: biochemistry of vitamin biogenesis and cofactor-containing enzymes
CcmE, which itself unexpectedly binds haem covalently before transferring the haem to apocytochromes.
Overview of cytochrome c biogenesis The biogenesis process to produce c-type cytochromes involves the formation of two thioether bonds between the haem vinyl groups and the cysteine residues of a CXXCH motif in the protein. The stereochemistry of the attachment is universally conserved and generally requires the involvement of other proteins. Cytochromes c are essential proteins in virtually all organisms as are, consequently, their dedicated biogenesis systems. Unexpectedly, it has been found that there are three different systems involved in this process in different organisms [10,11,13] and evidence suggesting the existence of a fourth such system has recently been presented [14]. The simplest system involves a single protein (haem lyase) and is found in certain types of mitochondria. Thylakoids, Grampositive and some Gram-negative bacteria contain a different system, which involves at least four essential proteins. The most complex system (called System I) occurs in plant mitochondria and some Gram-negative bacteria and involves the Ccm proteins, which are best characterized in Escherichia coli. The eight essential Ccm proteins (CcmA–CcmH) are associated with the cytoplasmic membrane and function in the haem attachment process, which occurs in the periplasm. Both the haem and the apocytochrome are produced in the cytoplasm and so require transmembrane transport. The mode of transport of the protein has been described and involves the Sec system [15], whereas the mechanism for delivery of haem to the periplasm is unknown. CcmA and CcmB form an ABC transporter of an unknown substrate that is essential for the maturation process. CcmC is involved in haem delivery to the haem chaperone CcmE [16], which is described in more detail below. CcmD is a small membrane protein that facilitates the interactions of other Ccm proteins [17]. CcmF appears to facilitate the transfer of haem from CcmE to the cysteine residues of the apocytochrome [18], which should be in the reduced form. Disulphide reduction and/or isomerization is controlled by the proteins CcmG and CcmH and involves the Dsb (disulphide bond) proteins [19]. There are several steps in the maturation process in which reductant is required. Reduced haem has been shown to be required for the in vitro attachment of haem to apocytochromes [20] as well as to CcmE [21]. It may be that the reductant is provided by the Ccm proteins themselves, through a route that is independent of the protein DsbD [22], which is known to transport reducing equivalents to the periplasm. Unlike the haem lyase cytochrome c biogenesis pathway, in which the lyase has restricted specificity, the Ccm system has been used to produce a wide variety of cytochromes c from various sources, notably the recent attachment of haem to a 12 amino acid peptide [23]. It is noteworthy that other phenotypes have been identified for ccm genes in different organisms, suggesting that these proteins have other functions in addition to their role in cytochrome c biogenesis. The nitrogen-fixing Gluconacetobacter diazotrophicus displays
a deficiency in indole-3-acetic acid biosynthesis due to mutation of its ccm genes [24]. CcmF appears to be involved in manganese oxidation in Pseudomonas putida [25] and the pathogenicity of Legionella pneumophila [26]. CcmC, in particular, appears to have more diverse functions and has been implicated in siderophore production [27] and haem biosynthesis in Pseudomonas fluorescens [28].
The central role of the haem chaperone CcmE in cytochrome c biogenesis It has been found that metals are chaperoned in the cellular environment, presumably due to toxicity, scarcity or other factors. Copper chaperones have been identified and described [29] as well as metallochaperones that deliver nickel to enzymes requiring it as a cofactor [30]. There is, however, only a single example of a haem chaperone that has been identified: CcmE [31]. This protein is unique as it binds haem covalently through a histidine residue; the binding is transient as the haem is then transferred to apocytochromes [31]. We have shown that haem attachment to CcmE and haem transfer from it to an apocytochrome can occur in vitro [21]; both of these processes are enhanced by the presence of a His6 tag on CcmE. Very recently, the structure of the haem-bound peptide of CcmE was solved and it showed an unexpected linkage between the Nδ1 of the histidine (His130 in E. coli) and the β-carbon of a haem vinyl group [32]. The vinyl group (2- or 4-vinyl) involved in the bond formation was not identified. It has also been shown that CcmE binds haem both in vivo and in vitro when the histidine residue is replaced with a cysteine [33,34]. The structure of the holoprotein has not been solved, but two structures of the apo-form of CcmE are available [35,36]. The protein comprises a β-barrel core and a flexible C-terminal domain with the haem-binding histidine residue at the junction of these two domains. For the E. coli protein, it was possible to model a haem-binding site on the protein surface. This consisted of a patch of hydrophobic residues and two basic residues to interact with the haem propionates, and placed the 2-vinyl group of the haem close to His130 [36]. The role of the C-terminal domain has been investigated in vivo [37] but its precise function remains to be elucidated. CcmE is structurally related to the OB (oligonucleotide/ oligosaccharide-binding) proteins, which bind and transfer non-protein molecules. Since only the structure of the apoprotein is available, insight into the nature of the haem ligation in CcmE has come from spectroscopic studies, as described below.
The haem-binding site of CcmE It has been shown that apo-CcmE has a hydrophobic site to which ANS (8-anilinonaphthalene-l-sulphonic acid) can bind [21], and a hydrophobic surface on the protein has been identified [34]. The structures of the apo-forms of CcmE did not, however, provide obvious clues regarding the nature of the haem ligands. We have undertaken a detailed RR (resonance Raman) study of CcmE containing covalently C 2005
Biochemical Society
793
794
Biochemical Society Transactions (2005) Volume 33, part 4
attached haem [38]. RR spectra of the ferric form indicate that the haem is in a five-coordinate high-spin state, which converts, upon reduction, to a six-coordinate low-spin form. Isotopic shift analysis of the CO-bound form allowed us to assign the Fe–CO and C–O stretching modes, which were compared with those of other haem-containing proteins, and suggested that the trans ligand to the CO was a neutral histidine residue. We proposed previously that Tyr134 could act as a haem ligand due to its position, four residues away from the haem-binding histidine (HDENY in CcmE), since the histidine in the CXXCH motif is always a haem ligand in c-type cytochromes [21]. We replaced this residue with Phe and studied its RR spectra. Near-UV RR spectra showed a band in the wild-type protein at 600 cm−1 that was assigned to the Fe-O (Tyr) stretching mode, and was found to be absent from the Y134F (Tyr134 → Phe) variant. Differences in tyrosine-specific bands in UV-excited RR spectra of the wild-type and Y134F proteins provided further evidence for tyrosine ligation. It was thus concluded that Tyr134 was a ligand to the haem iron in both oxidation states. Data on the CO-bound form suggested that the sixth ligand in the ferrous form was a histidine residue. Additional information on the interaction of haem with CcmE was obtained from the RR studies. The haem peripheral modes in the holoprotein were examined and compared with those observed in CcmE containing non-covalently bound haem (CcmE can be produced in its apo-form, which forms a complex with haem added in vitro). The bending modes of the two vinyl groups of haem are observed at different frequencies in myoglobin and other haem proteins. It was therefore not unexpected to find that there was only one such vinyl mode for holo-CcmE. However, it was also found that there was a single mode for the non-covalent haem protein complex, making it difficult to draw any conclusions. However, two vinyl stretching modes of the non-covalent complex were observed separately and only one was observed for the holoprotein, confirming that one of the vinyl groups is modified in holo-CcmE. Comparison of our data with the effect of removal of haem vinyl groups (by substitution with, for example, formyl groups) on other haem resonances in other proteins suggested that the 2-vinyl group of the haem in holo-CcmE is modified. The evidence for co-ordination of the haem iron by Tyr134 is consistent with the binding of His130 to the 2-vinyl group of haem in the HDENY motif of CcmE, because, in a c-type cytochrome, the first cysteine residue in the CXXCH motif forms a bond with the haem 2-vinyl group and the histidine ligates the iron. Although CcmE has been the subject of a number of in vivo and in vitro studies, it remains unclear why the covalent bond to haem is formed in this protein. An attractive hypothesis is one in which the stereochemistry of haem attachment to apocytochromes is controlled by CcmE, if it forms a bond specifically to one of the haem vinyl groups. It is not yet clear if this is the case. It seems, however, that the correct stereochemistry of haem attachment to cytochromes is conferred by the Ccm system in vivo [39], although the mechanism remains unclear. C 2005
Biochemical Society
It has been suggested that radical chemistry may be involved in the function of CcmE. The involvement of a radical has been proposed in the attachment of His130 to the β-carbon of the haem vinyl group [32], as has also been considered for the attachment of a histidine residue to the haem in an algal haemoglobin [40], which forms a bond to the α-carbon of the 2-vinyl group. Further investigation will be required to elucidate the precise role of CcmE and the molecular details of cytochrome c biogenesis in general. This work was supported by a grant from the BBSRC (C20071) to S.J.F. O.D. is a Junior Research Fellow of Christ Church, Oxford. J.M.S. acknowledges funding from EMBO and thanks Professor L. Thony¨ Meyer and Drs M. Braun, U. Ahuja and O. Christensen for valuable discussions.
References 1 Yankovskaya, V., Horsefield, R., Tornroth, S., Luna-Chavez, C., Miyoshi, H., Leger, C., Byrne, B., Cecchini, G. and Iwata, S. (2003) Science 299, 700–704 2 Fraaije, M.W., van Den Heuvel, R.H., van Berkel, W.J. and Mattevi, A. (2000) J. Biol. Chem. 275, 38654–38658 3 Chapman-Smith, A. and Cronan, Jr, J.E. (1999) Trends Biochem. Sci. 24, 359–363 4 Kuettner, E.B., Hilgenfeld, R. and Weiss, M.S. (2002) J. Biol. Chem. 277, 46402–46407 5 Ferreira, G.C., Neame, P.J. and Dailey, H.A. (1993) Protein Sci. 2, 1959–1965 6 Davidson, V.L. (2004) Arch. Biochem. Biophys. 428, 32–40 7 Moore, G.R. and Pettigrew, G.W. (1990) Cytochromes c: Evolutionary, Structural, and Physicochemical Aspects, Springer-Verlag, New York 8 Colas, C. and de Montellano, P.R.O. (2003) Chem. Rev. 103, 2305–2332 9 Osawa, Y., Highet, R.J., Bax, A. and Pohl, L.R. (1991) J. Biol. Chem. 266, 3208–3214 10 Allen, J.W., Daltrop, O., Stevens, J.M. and Ferguson, S.J. (2003) Philos. Trans. R. Soc. Lond. B Biol. Sci. 358, 255–266 11 Stevens, J.M., Daltrop, O., Allen, J.W. and Ferguson, S.J. (2004) Acc. Chem. Res. 37, 999–1007 12 Barker, P.D. and Ferguson, S.J. (1999) Structure Fold. Des. 7, R281–R290 13 Thony-Meyer, L. (1997) Microbiol. Mol. Biol. Rev. 61, 337–376 14 Allen, J.W., Ginger, M.L. and Ferguson, S.J. (2004) Biochem. J. 383, 537–542 15 Thony-Meyer, L. and Kunzler, P. (1997) Eur. J. Biochem. 246, 794–799 16 Ren, Q. and Thony-Meyer, L. (2001) J. Biol. Chem. 276, 32591–32596 17 Ahuja, U. and Thony-Meyer, L. (2005) J. Biol. Chem. 280, 236–243 18 Ren, Q., Ahuja, U. and Thony-Meyer, L. (2002) J. Biol. Chem. 277, 7657–7663 19 Fabianek, R.A., Hennecke, H. and Thony-Meyer, L. (2000) FEMS Microbiol. Rev. 24, 303–316 20 Daltrop, O., Allen, J.W., Willis, A.C. and Ferguson, S.J. (2002) Proc. Natl. Acad. Sci. U.S.A. 99, 7872–7876 21 Daltrop, O., Stevens, J.M., Higham, C.W. and Ferguson, S.J. (2002) Proc. Natl. Acad. Sci. U.S.A. 99, 9703–9708 22 Stevens, J.M., Gordon, E.H. and Ferguson, S.J. (2004) FEBS Lett. 576, 81–85 23 Braun, M. and Thony-Meyer, L. (2004) Proc. Natl. Acad. Sci. U.S.A. 101, 12830–12835 24 Lee, S., Flores-Encarnacion, M., Contreras-Zentella, M., Garcia-Flores, L., Escamilla, J.E. and Kennedy, C. (2004) J. Bacteriol. 186, 5384–5391 25 de Vrind, J.P., Brouwers, G.J., Corstjens, P.L., den Dulk, J. and de Vrind-de Jong, E.W. (1998) Appl. Environ. Microbiol. 64, 3556–3562 26 Polesky, A.H., Ross, J.T., Falkow, S. and Tompkins, L.S. (2001) Infect. Immun. 69, 977–987 27 Baysse, C., Budzikiewicz, H., Uria Fernandez, D. and Cornelis, P. (2002) FEBS Lett. 523, 23–28 28 Baysse, C., Matthijs, S., Schobert, M., Layer, G., Jahn, D. and Cornelis, P. (2003) Microbiology 149, 3543–3552 29 Huffman, D.L. and O’Halloran, T.V. (2001) Annu. Rev. Biochem. 70, 677–701
Coenzymology: biochemistry of vitamin biogenesis and cofactor-containing enzymes
30 Rosenzweig, A.C. (2002) Chem. Biol. 9, 673–677 31 Schulz, H., Hennecke, H. and Thony-Meyer, L. (1998) Science 281, 1197–1200 32 Lee, D., Pervushin, K., Bischof, D., Braun, M. and Thony-Meyer, L. (2005) J. Am. Chem. Soc. 127, 3716–3717 33 Stevens, J.M., Daltrop, O., Higham, C.W. and Ferguson, S.J. (2003) J. Biol. Chem. 278, 20500–20506 34 Enggist, E., Schneider, M.J., Schulz, H. and Thony-Meyer, L. (2003) J. Bacteriol. 185, 175–183 35 Arnesano, F., Banci, L., Barker, P.D., Bertini, I., Rosato, A., Su, X.C. and Viezzoli, M.S. (2002) Biochemistry 41, 13587–13594
36 Enggist, E., Thony-Meyer, L., Guntert, P. and Pervushin, K. (2002) Structure (Camb.) 10, 1551–1557 37 Enggist, E. and Thony-Meyer, L. (2003) J. Bacteriol. 185, 3821–3827 38 Uchida, T., Stevens, J.M., Daltrop, O., Harvat, E.M., Hong, L., Ferguson, S.J. and Kitagawa, T. (2004) J. Biol. Chem. 279, 51981–51988 39 Allen, J.W., Barker, P.D. and Ferguson, S.J. (2003) J. Biol. Chem. 278, 52075–52083 40 Vu, B.C., Jones, A.D. and Lecomte, J.T. (2002) J. Am. Chem. Soc. 124, 8544–8545 Received 5 April 2005
C 2005
Biochemical Society
795
