The cytochrome c family is widespread and includes the mitochondrial cytochromes c with their essential functions in respiration. This soluble electron transfer protein found in the intermembrane space (IMS), and referred to as “cytochrome c” has, in the last decade, been found to play an additional key role in signaling apoptosis. In bacteria, cytochromes c perform a broad range of functions in catalysis and electron transfer. Some are analogous to mitochondrial, single-heme cytochromes, whereas others are multiheme proteins with, in some cases, large numbers of heme groups per polypeptide. Cytochromes of the c-type also perform electron transfer reactions in photosynthesis.
Cytochromes c are assembled in the biological compartments in which they function, that is on the p-side of energy transduction membranes. One exception to this is the novel c-type component of the cytochrome b6f complex that is produced on the n-side of some photosynthetic membranes (as reviewed in ref. (1)). Holocytochromes c are produced from the apo-form of the polypeptide and the cofactor Fe-protoporphyrin IX (b-heme, shown in Fig. 1), both of which need to be transported across a membrane. Once the apocytochrome is delivered to that site, the two cysteine residues to which the heme will become covalently bound (usually a CXXCH motif) can become oxidized to form a disulfide bond; specific proteins (thiol-disulfide oxidoreductases) are required, in some cases at least, to reverse this process. The heme iron also needs to be reduced before the attachment reaction and heme chaperones are sometimes involved in heme delivery. Finally, the thioether bonds between the heme and the protein are formed, a process also requiring maturation proteins.
The composition and features of the biogenesis systems vary dramatically between different organisms and organelles (as shown in Fig. 2); the diversity of the different systems and the details of how they function have been reviewed recently (1–4). The proteins involved in these machineries are listed according to their functions (where known) in Table 1. Widespread agreement on the nomenclature of cytochrome c biogenesis proteins has not been reached. The type of maturation apparatus and its location in different organisms is complicated and sometimes unpredictable; a general guideline for this is presented in Fig. 3 for all six biogenesis systems. A more detailed analysis can be found in refs. (4–6). Briefly, bacteria utilize Systems I and II to mature their cytochromes c. These are both multicomponent systems located in the cytoplasmic membranes. Bacterial cytochromes are highly diverse in structure (Fig. 1) and function, possibly accounting for the complexity of their biogenesis apparatuses. Many eukaryotes, including yeast, animals and some protozoa, use a single type of protein to produce cytochromes c in their mitochondria. Plants have the most complex collection of cytochrome c assembly systems; they use System I in their mitochondria, System II on the p-side of their thylakoid membranes and System IV on the n-side. That the plant cell originated from two endosymbiotic events presumably accounts for their use of several distinct cytochrome assembly systems. These systems attach heme to cytochromes that are soluble (e.g. mitochondrial cytochrome c) or membrane-bound (e.g. cytochrome c1 in Complex III).
Table 1. The proteins comprising cytochrome c biogenesis Systems I, II, and III and their functional classification, when known
Many unknowns remain but recent developments have helped to elucidate some key elements of cytochrome assembly in the different systems. Progress has been hampered by the fact that many of the cytochrome c assembly proteins are membranous, complicating overexpression and direct characterization. In the case of System I, most studies have been performed on Rhodobacter capsulatus and Escherichia coli. For System II, representative organisms include Chlamydomonas reinhardtii, Bacillus subtilis, Bordetella pertussis, and Wolinella succinogenes. Yeast species have been used in experiments on System III, and what is known about System IV has been elucidated in C. reinhardtii. Reference is made in this review to as many key works as possible, subject to the limit on the allowed number of references.
Apocytochrome Transport and Chaperoning
It has been established for some time that mitochondrial cytochrome c is imported into the IMS via a unique mechanism that does not involve a classical targeting signal. The protein that performs the heme attachment reaction is also involved in the import of the apocytochrome (in the case of mitochondria that contain System III). The protein is referred to as both holocytochrome c synthase (HCCS) and heme lyase. The former nomenclature is preferred because it describes more clearly the function of the protein. Good evidence exists to demonstrate a complex between HCCS and apocytochrome (7) leading to a model in which the holocytochrome becomes trapped in the IMS by adopting its folded conformation following covalent heme attachment. Some organisms contain a HCCS for each of their mitochondrial c-type cytochromes (the soluble “cytochrome c” and the cytochrome c1 subunit of Complex III). Cytochrome c1 is imported via a different mechanism; it has a bipartite signal sequence with a matrix import signal resulting in the protein being inserted into the inner mitochondrial membrane (IMM). When both termini are in the IMM, HCCS (or HCC1S when present) attaches heme, with subsequent cleavage of the N-terminal anchor, leaving the final product anchored via its C-terminus. Agreement on the details of these processes has not yet been reached; the apocytochrome might be delivered all the way to the matrix and then re-exported (8) or a stop-transfer mechanism could be involved (9). It is not known whether the cytochrome c biogenesis proteins of mitochondria that do not use System III (i.e. System I or V) are involved in cytochrome import (10).
Cytochromes c that are destined for the chloroplast lumen contain targeting signal peptides; these vary according to whether the apocytochromes are encoded on the nuclear or the plastid genomes (11). Unlike the mitochondrial cytochromes, the import mechanisms are not dependent on the heme attachment reaction (12). The import of the chloroplast c-type cytochromes, c6 and f, has been reviewed in ref. (13).
It is accepted that bacterial apocytochromes are transported to their extracytoplasmic destinations by the Sec pathway (14). Sec transports unfolded proteins across the membrane when the protein carries a characteristic targeting sequence with the nascent protein being maintained in the unfolded state by SecB before translocation. What happens to bacterial cytochromes when they are delivered across the membrane is not known. The proteins could enter different pathways including interaction with the cytochrome c biogenesis proteins or with thiol-disulfide oxidoreductases which would insert disulfide bonds into the heme binding motifs (see next section). Given that heme would need to be attached to an unfolded substrate (at least around the heme-binding motif) it is not surprising that transient binding of the apocytochrome might be required. Some evidence for chaperone activities has been presented for the Ccm (cytochrome c maturation) proteins of System I. The protein CcmI (which is the C-terminal domain of the protein CcmH in certain bacterial species, Fig. 2) has tetratricopeptide repeat motifs that are characteristically involved in protein-protein interactions. Recent progress has been made in showing that apocytochrome binds to CcmI and a chaperone function has been proposed (15).
In some cases, cytochromes are assembled in comparatively oxidizing subcellular environments (the periplasm of Gram-negative bacteria or the chloroplast lumen) exposing the cysteine thiols in the heme-binding motif to the possibility of disulfide bond formation. It was even thought that formation of a disulfide in the CXXCH motif of the apocytochrome could be an obligate intermediate en route to heme attachment. The bacterial periplasmic protein DsbA is highly oxidizing and inserts disulfide bonds into extracytoplasmic proteins (16). It has been demonstrated that DsbA is not essential for cytochrome c biogenesis (17) although it does oxidize apocytochromes c. A reductive pathway exists to reverse the oxidation reaction. The proteins DsbD and CcdA transfer reducing power across the membrane (18). DsbC is a thiol-disulfide isomerase that corrects non-native disulfide bonds in proteins with more than two cysteines. This function could be relevant to multiheme cytochromes with their multiple CXXCH motifs. A recent advance in understanding the thiol-disulfide pathway in bacteria was publication of the structure of the protein DsbB. Its function is to reoxidize DsbA and to transfer the electrons it acquires to the respiratory chain by interaction with quinones in the cytoplasmic membrane (19, 20). A related oxidizing protein, called BdbD, occurs in Gram-positive bacteria like B. subtilis. The protein ResA performs the reducing function in organisms that utilize System II (21), see Figs. 2 and 3, and acquires reductant from the membranous CcdA/DsbD. The structures and mechanisms of action of the Dsb proteins are better understood than any of the cytochrome biogenesis proteins to which reductant is transferred.
CcmG, a thioredoxin-like protein (22), is thought to deliver specifically reducing power specifically to oxidized apocytochromes in System I (23) as shown in Fig. 2, although its role may not be so straightforward (24). The role of CcmH has not been fully elucidated. The E. coli protein CcmH is expressed as two separate proteins in many other organisms, which are named CcmH (corresponding to the N-terminal domain of E. coli CcmH (N-CcmH)) and CcmI (the equivalent of the C-terminal domain in E. coli (C-CcmH)). N-CcmH has a three-helix bundle structure with a conserved pair of cysteines and has been proposed to have a reductive function (25, 26). Interaction between CcmH and apocytochromes has been demonstrated (27) but the route of reductant transfer has not been established. The most likely route is DsbD, CcmG, CcmH, and then apocytochrome. However, both cysteines of CcmH are not essential under all conditions (28).
In plants ccmH is an essential gene as shown in Arabidopsis thaliana studies (29). It might be argued that CcmH is part of a system that is needed to reverse disulfide bond formation within the CXXCH motif of the apocytochrome in the plant IMS. On the other hand, the plant IMS has been regarded as a reducing environment and as yet no non-specific disulfide bond forming system has been identified in plant mitochondria. A disulfide-bond forming protein has been identified in the chloroplast lumen; its substrate specificity remains to be determined (30).
In mitochondria that use System III (Fig. 3), no universal additional proteins appear to be needed for cytochrome c production. The role of HCCS in import and binding of the apocytochrome as it is delivered to the IMS could protect the protein from any unwanted side reactions. The flavoprotein Cyc2p has a role in cytochrome c assembly in yeast; it interacts with HCCS and apocytochromes and is postulated to act as a source of reductant for the heme iron (31) but is not present in non-fungal mitochondria. Our knowledge of thiol redox chemistry in mitochondrial cytochrome c biogenesis is described elsewhere (32).
Heme Transport and Provision
It is not known for any of the biogenesis systems how the heme cofactor is transported across the relevant membranes. In bacteria, heme biosynthesis occurs in the cytoplasm and the product must be transported across the cytoplasmic membrane. Many studies have sought to identify how this occurs, considering options like passive diffusion and candidates proteins within the biogenesis sets of apparatus. An obvious possibility was the proteins CcmA and B (Fig. 2) in System I which have the features of an ABC-transporter (ATP-binding cassette). It has been clearly shown, however, that these proteins do not export heme, as heme accumulates in abundance in the periplasm of E.coli when the ATP hydrolysis function of CcmA is abolished by mutation (33, 34). In addition, a periplasmic b-type cytochrome is produced even in the absence of the Ccm genes (35). CcmAB are thought rather to perform a conformationally driven function in shuttling heme between CcmC and CcmE. For System I, it is established that heme is delivered from the membrane protein CcmC to the heme chaperone CcmE. A complex between these two proteins, involving heme, has been characterized and the amino acids that ligate the heme iron on CcmC were identified (36, 37). CcmE is a unique heme chaperone that binds heme covalently via a histidine residue before the heme is transferred to the apocytochrome (38). This is unexpected for a chaperone protein and the importance of the covalent but transient linkage remains unknown. CcmE binds heme non-covalently and then covalently in vitro (39) with a tyrosine ligating the heme iron (40, 41). The heme-binding region of the protein is flexible (42–44), as would be expected from its function. Recent progress has been made in establishing the functional relevance of the covalent bond, which is between a histidine N and the β-carbon of one of the heme vinyl groups (45). A complex between the membrane-anchored CcmE and a variant cytochrome c was trapped in vivo and characterized (46). The effects on the formation of the complex caused by changes in the other Ccm proteins indicate that the covalent bond is on-pathway and physiologically relevant.
A further complication in understanding the chemistry of heme provision in System I arose from the observation that a variant form of the system (System I*) occurs in some organisms (archaea and sulfate-reducing bacteria). The CcmE protein in the variant system has a conserved cysteine in place of the classical histidine; the cysteine binds heme covalently (47) and the structure of the apoprotein is very similar to that of E. coli CcmE (48). It is surprising that thiol-based heme chemistry (which is not uncommon in nature) can function analogously to the histidine-based reactions with heme observed in the CcmE protein of E. coli and many other organisms. When the histidine residue is replaced with a cysteine in E. coli CcmE, cytochrome maturation is abolished, although the variant form binds heme covalently (42, 49).
Evidence has been presented suggesting that proteins in System II (Fig. 2) function in heme transport. Some bacteria produce a fusion protein of ResB and ResC as a single large polypeptide. ResBC from Helicobacter hepaticus was analyzed following expression in E. coli and it was found that the protein contained reduced heme (50). Amino acids that are suggested to ligate the heme iron were identified and it was hypothesized that histidine residues form a channel for heme transport across the membrane. Proteolytic degradation of the protein complicates the interpretation of some of these observations. A second study on System II proteins provided further evidence of heme binding by ResBC (51); protein degradation was also found to occur. Interestingly, covalent binding of heme to a cysteine residue of ResB was found although the protein performed its function even in the absence of that (poorly conserved) residue. Another study of the function of ResBC in E. coli showed that the formation of a b-type cytochrome was possible (not so when System I proteins were present), also suggesting a possible role for the System II proteins in heme transport (52). It would be good to have further evidence that ResBC function in heme transport.
Covalent Heme Attachment to Cytochromes c
Recent progress has been made in elucidating the function of the protein CcmF in System I (Fig. 2). Because it is large, membranous and has no homology to proteins of known function, CcmF had been perhaps the least well understood of the Ccm proteins. It was purified and found to contain a b-type heme (53); putative heme-coordinating histidines were identified amongst the transmembrane helices (54). The protein was able to oxidize quinols, suggesting that the b-heme transfers electrons to the Ccm pathway, possibly to reduce the heme that is covalently attached to CcmE. It is notable that CcmF in plants consists of two or three polypeptides, a feature that is likely to give insight into the function of the larger fused form found in other System I-containing organisms (55). CcmF is regarded as the “heme lyase” protein of the Ccm system (56), although no mechanism has been proposed. A paralogue of CcmF (NrfE) is required for heme attachment to a cytochrome c (NrfA) containing a CXXCK motif (57). How the System II proteins catalyze thioether bond formation between heme and the apocytochrome cysteines is also not known.
Recent work has investigated substrate recognition by the System III protein HCCS (Fig. 2). Several lines of evidence indicate that the N-terminal region (the residues before the CXXCH, 13 amino acids in the case of horse cytochrome c) of the apocytochrome is all that is recognized by the synthase (58–60) with a conserved phenylalanine residue being of particular importance (58, 61). This is significant for the function of HCCS in capturing the apocytochrome as it is delivered to the IMS, a process known to be essential for apocytochrome import, as well as for the actual heme attachment. It has been suggested that heme binds to HCCS via cysteine residues in CP motifs (62). The latter are found in a variety of proteins as so-called heme regulatory motifs. It has been shown in a recombinant system testing HCCS function in the E. coli cytoplasm, that the two CP motifs of a yeast HCCS are not essential for the heme attachment reaction (63); the motifs may perform another function within mitochondria. HCCS is also active when directed to the E. coli periplasm, where it attaches heme exclusively to mitochondrial cytochrome c (64). Until HCCS is purified and characterized directly, its precise function will remain unknown (See Note Added in Proof).
Systems IV, V, and VI
A additional cytochrome maturation system occurs in all organisms that perform oxygenic photosynthesis (Fig. 3). Apart from the cytochrome f component, which is actually a c-type cytochrome formed by System II, the cytochrome b6f complex contains a very unusual and essential heme that is attached to the cytochrome b6 protein via a single thioether bond to a cysteine residue. Four proteins (CCB1-4, System IV) are essential for the maturation of this protein and these all occur in the photosynthetic membrane on the stromal side. Their localization and interaction properties have been analyzed but the function of the individual proteins is not yet known (the four proteins are therefore not in Table 1). Interestingly, the Firmicutes, for example B. subtilis, have a related single-cysteine attached cytochrome but do not contain the System IV proteins, suggesting that yet another set of proteins may be present here (System VI); see Fig. 3.
Another group of organisms, the Euglenozoa, contain single-cysteine attached cytochromes c. It is not yet clear why these organisms have this unusual heme attachment, as the CXXCH form of the cytochrome can be matured and utilized by them (65). The structure of Euglenozoan cytochromes is highly similar to classical respiratory cytochromes, despite having a single bond between the heme and the protein (66). However, this phylum does not appear to contain any of the aforementioned cytochrome maturation systems that would be required to make their unusual mitochondrial cytochromes c. They must therefore contain a distinct cytochrome c biogenesis apparatus that remains to be identified ((67), System V).
Although cytochromes were discovered in the century before last, and were biochemically studied in the 1920s, the mechanism of covalent attachment of the essential heme cofactor to cytochromes c remains unresolved. From the simplest (System III) to the most complex (System I) cytochrome c biogenesis systems, understanding of the functions of many of the proteins has not yet been achieved. Transport and delivery of heme to the cellular site of cytochrome c assembly remains an important unknown. Genomics has served to confound this field by revealing the complexity of the systems and their unexpected and somewhat unpredictable distributions in nature. There are few similarities among proteins in the different systems; understanding why they are so different is interesting from an evolutionary perspective. Structural information on the constituent proteins will certainly assist, along with better descriptions of protein-protein interactions. New technological developments in studying protein complexes, as well as single-molecule studies of membrane proteins in particular, will bring advances in the coming years and shed light on these surprisingly complicated and diverse post-translational modification systems. The acquisition of molecular-level information could be significant because animals use entirely distinct cytochrome c biogenesis pathways than many of our pathogens, raising therapeutic possibilities.
Note Added in Proof
Ater submission of this manuscript, the first purification (from membrane extracts of E. coli following heterologous expression) of the System III protein holocytochrome c synthase was reported (San Francisco, B., Bretsnyder, E. C. and Kranz, R. G. (2012) Human mitochondrial holocytochrome c synthase's heme binding, maturation determinants, and complex formation with cytochrome c. Proc. Natl. Acad. Sci. USA. Nov 12. doi:10.1073/pnas.1213897109).
This work was supported by the Biotechnology and Biological Sciences Research Council (Grant BB/H017887/1). The authors thank Dr James W.A. Allen for invaluable input into Fig. 3.