Evolutionary relationships between heme-binding ferredoxin α + β barrels

Abstract

Background: The α + β barrel superfamily of the ferredoxin-like fold consists of a functionally diverse group of evolutionarily related proteins. The barrel architecture of these proteins is formed by either homo-/hetero-dimerization or duplication and fusion of ferredoxin-like domains. Several members of this superfamily bind heme in order to carry out their functions.

Results: We analyze the heme-binding sites in these proteins as well as their barrel topologies. Our comparative structural analysis of these heme-binding barrels reveals two distinct modes of packing of the ferredoxin-like domains to constitute the α + β barrel, which is typified by the Type-1/IsdG-like and Type-2/OxdA-like proteins, respectively. We examine the heme-binding pockets and explore the versatility of the α + β barrels ability to accommodate heme or heme-related moieties, such as siroheme, in at least three different sites, namely, the mode seen in IsdG/OxdA, Cld/DyP/EfeB/HemQ and siroheme decarboxylase barrels.

Conclusions: Our study offers insights into the plausible evolutionary relationships between the two distinct barrel packing topologies and relate the observed heme-binding sites to these topologies.

Keywords: Barrel packing; Heme; Hemoprotein; Iron metabolism; Protein evolution.

Fig. 1

Representative structures of the heme-binding ferredoxin α + β barrel superfamily. (a) Ribbon diagram of the structures of OxdA (PDB identifier 3A16), Cld (PDB identifier 3NN1), MhuD (PDB identifier 3HX9), IsdG (PDB identifier 2ZDP), DyP (PDB identifier 3VXJ) and Siroheme decarboxylase (PDB identifier 4UN1). The individual ferredoxin-like folds in all structures have been colored from their N- to C-termini in chainbow coloring scheme of PyMOL and are faded to highlight the position of heme (magenta). Elements that do not constitute a part of the core of the ferredoxin-like fold are colored white. (b) A comparison of heme binding sites using topology diagrams. (c) Orientation of heme-moieties in different proteins of the ferredoxin α + β barrel superfamily. Here, the porphyrin ring of heme and siroheme substituents is shown as a rectangle and the extensions from the rectangle denote the propionate side chains

Fig. 2

A comparative view of the packing topologies of the ferredoxin α + β barrels. (a-f) Left-side diagrams show a top view of the barrels with β-strands indicated as triangles and α-helices as circles. Right-side diagrams show an opened-up view of the barrel with two ferredoxin-like domains placed side-by-side to show β-strand interactions. Only one pair of interacting β-strands can be seen as the other pair is formed by the peripheral-β-strands. The linear arrangement of the secondary structure elements is shown above the opened-up view, with β-strands indicated as arrows and α-helices as rectangles. Each barrel is constituted of four β-α-β units from two ferredoxin-like domains, colored blue, green, yellow and red. Heme is shown as a black star in all but H. thermophilus siroheme decarboxylase where it is shown as a white dotted-outlined star. N- and C-termini of the protein chains are labeled ‘N’ and ‘C’ in black while those of individual ferredoxin-like domains in barrels with duplicated domains are in pink

Fig. 3

Structural and sequence features of the heme-binding region in IsdG-like, OxdA, Cld/Dyp/EfeB/HemQ, and siroheme decarboxylase proteins. (a) Structure superimposition stereo diagram of the C_α backbone of IsdG-like ferredoxin-like monomer (PDB identifier 2ZDP; cyan), the C-terminal domain of OxdA-like (PDB identifier 3A16_A; red) and the C-terminal domain of Cld/DyP/EfeB/HemQ-like (PDB identifier 3NN1_A; green). (b) Stereo diagram of a ferredoxin-like domain of siroheme decarboxylase (PDB identifier 4UN1_A). In panels (a, b), heme and side chain of the heme-binding histidine is shown in stick form. c Structure-based MSA of IsdG, OxdA, Cld/Dyp/EfeB/HemQ and siroheme decarboxylase. PDB identifier/UniProt ID, organism name, start, and end for each sequence are mentioned. The region of circular permutation is indicated by ‘|’ and the sequence number around this region is colored red. The secondary structure diagram for a ferredoxin-like domain is indicated above the alignment, where arrows represent β-strands and rectangles indicate α-helices. The number of omitted residues for large insertions is boxed in green. The sequence region(s) corresponding to disorder in structure or structurally not superimposable regions is italicized. Organism abbreviations: Sa-Staphylococcus aureus, Mt-Mycobacterium tuberculosis, Bs-Bacillus subtilis, Bc-Bacillus cereus, Dr-Deinococcus radiodurans, Bt-Bacteroides thetaiotaomicron, Ap-Acetobacter pasteurianus, Aa-Alicyclobacillus acidocaldarius, Te-Taylorella equigenitalis, Cn-Candidatus nitrosoarchaeum, Al-Acholeplasma laidlawii, Bp-Beggiatoa sp., Sc-Streptococcus criceti, Er-Erysipelothrix rhusiopathiae, Ss-Staphylococcus saprophyticus, Ns-Novosphingobium sp., Kr-Ktedonobacter racemifer, Mu-Methyloversatilis universalis, At-Agrobacterium tumefaciens, Bm-Burkholderia multivorans, Sp-Streptomyces sp., Re-Rhodococcus erythropolis, Sw-Shewanella woodyi, Sl-Streptomyces albus, Am-Actinosynnema mirum, Vd-Verticillium dahliae, Vp-Variovorax paradoxus, Ko-Klebsiella oxytoca, Pm-Patulibacter medicamentivorans, Gf-Gibberella fujikuroi, Fo-Fusarium oxysporum, Pd-Penicillium digitatum, Bf-Botryotinia fuckeliana, Nf-Neosartorya fischeri, Nw-Nitrobacter winogradskyi, Tt-Thermus thermophilus, Ci-Candidatus Nitrospira, So-Shewanella oneidensis, Ba-Bjerkandera adusta, Ec-Escherichia coli, Dd-Desulfovibrio desulfuricans, Ht-Hydrogenobacter thermophilus