Evolutionary Persistence of the Molybdopyranopterin-Containing Sulfite Oxidase Protein Fold

Gregory J Workun; Kamila Moquin; Richard A Rothery; Joel H Weiner

doi:10.1128/MMBR.00041-07

. 2008 Jun;72(2):228–248. doi: 10.1128/MMBR.00041-07

Evolutionary Persistence of the Molybdopyranopterin-Containing Sulfite Oxidase Protein Fold

Gregory J Workun ¹, Kamila Moquin ¹, Richard A Rothery ¹, Joel H Weiner ^1,^*

PMCID: PMC2415749 PMID: 18535145

Abstract

Summary: The importance of molybdoenzymes is exemplified both by the debilitating and fatal human diseases caused by their deficiency and by their persistence throughout evolution. Here, we show that the protein fold of the molybdopyranopterin-containing domain of sulfite oxidase (the SUOX fold) can be found in all three domains of life. Analyses of sequence data and protein structure comparisons (secondary structure matching) show that the SUOX fold is found in enzymes that have quite distinct macromolecular architectures comprising one or more domains and sometimes subsidiary subunits. These are summarized as follows: (i) animal SUOXs that contain an N-terminal cytochrome b₅ domain and an SUOX fold fused to a C-terminal dimerization domain; (ii) plant SUOX that contains an SUOX fold fused to a C-terminal dimerization domain; (iii) the YedY protein from Escherichia coli, which comprises only the SUOX fold; (iv) the sulfite dehydrogenase from Starkeya novella that contains the SUOX fold, a dimerization domain, and an additional c-type cytochrome subunit; and (v) the plant-type nitrate reductases, exemplified by that of Pichia angusta, that contain an N-terminal SUOX fold, a dimerization domain, a cytochrome b₅ domain, and a C-terminal NADH binding flavin adenine dinucleotide-containing domain. We used the primary sequences of the proteins containing an SUOX fold to mine 559 sequences of related proteins. A phylogeny of a nonredundant subset of these sequences was generated, and the resultant clades were categorized by sequence motif analyses in the context of the available protein structures. Based on the motif analyses, cladistics, and domain conservations, we are able to postulate a plausible pathway of SUOX fold enzyme evolution.

INTRODUCTION

Enzymes containing the metal molybdenum or tungsten at their active sites are essential for nearly all living organisms and catalyze primarily oxo transfer reactions for the metabolism or catabolism of nitrogen, sulfur, and carbon compounds (20, 34-37, 46, 57, 58). For all examples other than those belonging to the nitrogenase class (18, 69), the Mo or W is coordinated by a pyranopterin (PPT) cofactor (35, 37). This coordination is supplemented by an amino acid side chain such as those of Cys, Sec, Ser, or Asp residues and/or by the coordination of single oxygen or sulfur atoms in the form of oxo or sulfido groups. Molybdoenzymes have been assigned to four classes. The first class is the archaeal aldehyde oxidoreductase class that contains a bis-PPT cofactor that coordinates a W atom at its active site (10, 66). This class lacks a protein W ligand. The second class is the xanthine dehydrogenase class, examples of which include the eukaryotic xanthine dehydrogenase class and Desulfovibrio gigas aldehyde oxidoreductase (36). This class also lacks a protein-Mo ligand. The third class is the prokaryotic dimethyl sulfoxide reductase class that contains a Mo-bis(pyranopterin guanine dinucleotide) (Mo-bisPGD) cofactor (5, 39, 75, 80, 81) with the full range of protein-Mo ligands described above. The fourth class is the sulfite oxidase (SUOX) class that contains a Mo-PPT cofactor and a Cys protein-Mo ligand (46, 47). The plant-type assimilatory nitrate reductase (NIA) also belongs to the SUOX class (8, 22) and functions in nitrate assimilation (56). Animal SUOX enzymes funnel electrons derived from sulfite oxidation to cytochrome c within the intermembrane space of the mitochondrion. Plant SUOX enzymes are located in the peroxisome and transfer sulfite-derived electrons directly to oxygen, generating hydrogen peroxide (30, 31). There are also a large number of archaeal and bacterial enzymes of currently unknown function that are predicted to contain an SUOX fold.

REACTIONS CATALYZED BY SUOX FOLD ENZYMES

The catalytic mechanisms of the SUOX enzymes have been reviewed extensively (20, 30, 31, 46) and will not be dealt with in detail herein. Animal SUOX catalyzes the following reaction in the mitochondrial intermembrane space with cytochrome c as the electron acceptor:

Plant SUOX catalyzes a similar reaction but with oxygen as an electron acceptor (30-32):

The reaction catalyzed by plant NIA is as follows:

In each case, the oxidation (ox) of sulfite or the reduction (red) of nitrate occurs at the Mo atom of a Mo-PPT cofactor (20, 22, 34, 37, 46). In fully oxidized SUOX and NIA, Mo(VI) coordination is provided by a pair of pterin-derived dithiolene sulfurs, the sulfur of a Cys residue, and two oxo groups (14, 20, 22). A single electron reduction generates the Mo(V) form, which retains the dithiolene coordination and has one oxo group and one hydroxyl or water ligand. The fully reduced Mo(IV) form has coordination very similar to that of the Mo(V) form (14, 33). The proposed SUOX reaction cycle envisions bond formation between the sulfite lone pair and a Mo(VI) oxo group and oxidative oxo transfer to generate the product sulfate and the Mo(IV) form of the enzyme. Thus, oxo transfer occurs from the Mo atom to the substrate. The NIA reaction would proceed in the reverse direction, with oxo transfer occurring from the substrate nitrate to the Mo atom.

Of the bacterial enzymes predicted to share the SUOX fold, the best characterized is SUOX from Starkeya novella. This enzyme has been described as being a “sulfite dehydrogenase” in the literature (11). It is a heterodimer comprising a SUOX fold protein (SorA) and a cytochrome c subunit (SorB) (41, 42). SorAB is periplasmically localized and is predicted to transfer electrons from sulfite oxidation to the membrane-bound respiratory chain via at least one soluble cytochrome c (43, 92).

Another partially characterized bacterial SUOX fold enzyme system is encoded by the Escherichia coli yedYZ operon. YedY is a periplasmically localized SUOX fold protein that appears to interact transiently with membrane-bound cytochrome b (YedZ) (7, 53). The physiological substrate for YedY is currently unknown, although it demonstrates reduction of a range of S- and N-oxides. Data on two further bacterial SUOX enzymes are also available. These enzymes are the cytoplasmically located SUOX from Deinococcus radiodurans (12) and the periplasmically localized SUOX from Thermus thermophilus (13). Neither of these enzymes appears to have subsidiary cytochrome subunits or domains, although the latter enzyme is predicted to be periplasmically localized and to be redox coupled to the cytoplasmic membrane via cytochrome c.

SUOX FOLD ENZYMES IN HUMAN DISEASE AND GEOCHEMICAL CYCLES

Human SUOX deficiency is typically inherited as a recessive autosomal trait for which there is no known therapy and typically results in death in infancy (38). Its severity is due to the critical role of SUOX in degrading Cys and Met amino acids, and deficiency results in the excretion of large amounts of S-sulfo-l-cysteine, sulfite, and thiosulfate in the urine (65). The intracellular accumulation of sulfite is thought to result in sulfitolysis of disulfide bonds that results in compromised protein stability (20). Point mutations linked to SUOX deficiency include the following: R160Q, A208D, S370Y, and G473D (47). The R160Q mutation increases the K_m for sulfite and decreases the k_cat, resulting in a 1,000-fold decrease in catalytic efficiency (23). Structural analysis of chicken liver SUOX reveals the importance of three Arg residues (47). R138, R190, and R450 contribute to a positively charged binding pocket, which stabilizes substrate/product binding, and these residues will be referred to herein as the Arg triad (see Active-Site Sequence Conservation Suggests a Simplified Classification of SUOX Fold Enzymes). The effects of an R160Q mutant in humans, corresponding to R138Q in chicken SUOX, have recently been studied by X-ray absorption spectroscopy (15) and reveal an increase in coordination number for the Mo, from 5 to 6. In the case of the R138Q mutant in chicken SUOX, X-ray crystallography revealed that the side chain nitrogen of the Gln appears to be within the coordination sphere of the Mo (44).

Plant SUOX functions in sulfite detoxification and has been implicated in the adaptation to elevated sulfur dioxide levels (“acid rain”) (30, 32). Atmospheric sulfur dioxide is converted to sulfite as follows:

SUOX is thus of importance to biosphere sulfur cycling and adaptation to industrial pollution. Sulfate is the major anion within the xylem system of plants and is absorbed by the root system for biosynthetic incorporation into Cys/Met residues and also into other compounds such as glucosinolates (6).

One novel aspect of plant SUOX is its reactivity toward oxygen, which results in the generation of potentially toxic H₂O₂ during enzyme turnover (20, 67). This problem is alleviated by the cellular localization of plant SUOX to the peroxisome organelle, which contains relatively large concentrations of peroxidase/catalase enzymes (52).

Plant NIA enzymes play a crucial role in nitrogen uptake and assimilation into plant proteins and play a role in the global nitrogen cycle (56, 72). They play a role complementary to that of the distinct nitrate reductases found in bacterial systems, whose activities are dependent on the Mo-bisPGD cofactor (73, 75).

EMERGING STRUCTURES REVEAL MULTIDOMAIN ARCHITECTURES OF SUOX FOLD ENZYMES

The importance of SUOX fold enzymes in human disease, plant sulfite detoxification, agriculture, adaptation to pollution, and global geochemical cycles has prompted efforts to obtain atomic-resolution structural data by protein crystallography and X-ray absorption spectroscopy methods. Structures for animal, plant, and bacterial SUOX enzymes, as well as the Mo-PPT binding SUOX fold fragment of a plant-type NIA enzyme, have been reported (14, 15, 33, 43, 44, 47, 82). Analyses of these structures reveal that animal SUOX enzymes comprise a heme b₅ domain, a Mo-PPT binding SUOX fold, and a dimerization domain, with this order being reflected in the primary sequence of the protein. The available plant-type NIA structure constitutes a fragment of the complete enzyme and includes the SUOX domain and the dimerization domain (22). Plant SUOX comprises a SUOX fold and a dimerization domain, and it lacks a heme-containing domain or accessory subunit and is localized to the peroxisome organelle (30, 31, 57). The periplasmic SUOX from S. novella has a Mo-PPT binding catalytic subunit (SorA) comprised of an SUOX fold and a dimerization domain. SorA forms a heterodimer with an additional subunit, SorB, which coordinates a c-type heme (41, 43).

Genome and protein structure database mining has identified a range of bacterial enzymes sharing sequence and inferred structural similarity to the well-characterized members of the SUOX fold-containing family. In the case of E. coli, YedY, a SUOX fold protein, has been identified, and its structure has been solved by protein crystallography (7, 53). Intriguingly, it is comprised of only the Mo-PPT binding SUOX fold itself. The redox partner of YedY is YedZ, a hydrophobic monoheme cytochrome b (midpoint potential [E_m] value −8 mV) (7) with six predicted transmembrane (TM) segments (17) (see Differential Redox Partners) that presumably couples YedY to the versatile redox sink of the membrane-intrinsic quinone pool. It was demonstrated using PSI-BLAST searches, which identify homologs with low similarity to bait sequences, that a plasma membrane antigen that is overexpressed in prostate cancers is related to YedZ (77, 89).

Another variation of the multidomain architecture observed in the SUOX fold enzyme is provided by the plant-type NIA enzymes that are also represented in yeasts such as Pichia angusta (22). These retain the SUOX fold and the dimerization domain, but the heme b₅ domain follows the dimerization domain in the primary sequence. Also, it contains an additional flavin adenine dinucleotide (FAD)-containing NADH binding domain at its C terminus, which, in the case of Zea mays, has had its structure solved by protein crystallography (54).

Our interest in the structure and function of the SUOX fold proteins stems from our discovery of the YedYZ system of E. coli (7, 53). The physiologically relevant reaction catalyzed at the Mo-PPT of this enzyme remains unresolved. As discussed below (see SUOX Fold Protein Cladistics), YedY and its predicted redox partner, YedZ, are highly conserved across a broad range of primarily proteobacterial species, so it likely catalyzes a conserved interconversion within their metabolomes.

SECONDARY STRUCTURE MATCHING SEARCHES REVEAL SIMILARITIES BETWEEN PROTEINS OF LOW SEQUENCE SIMILARITY

The availability of several hundred genome sequences presents an excellent opportunity to analyze SUOX fold proteins in all domains of life. As of August 2007, the complete sequences of 46 archaeal, 523 bacterial, and 65 eukaryotic genomes were available (listed on the Genomes Online Database at http://www.genomesonline.org). This is reflected in the total number of protein sequences available, which exceeds 4.4 million as of August 2007 (listed on the European Bioinformatics Institute [EBI] server at http://www.ebi.ac.uk/swissprot/sptr_stats/index.html). Remarkable strides have also been made in protein structure determinations, with approximately 47,000 structures being available as of August 2007 (listed on the Protein Data Bank [PDB] server at http://www.pdb.org). The sheer size of available sequence and structural bioinformatic data sets has catalyzed the emergence of a diverse array of analysis methods. One such approach is the use of secondary structure matching (SSM) algorithms to compare a protein structure of interest with either a set of protein structures or the entire PDB archive (49). We have recently used this technique to gain new insights into the complex iron-sulfur molybdoenzyme family (75).

In order to study the relationships between the SUOX fold proteins, we used the EBI SSM server (at www.ebi.ac.uk/msd-srv/ssm/ssmstart.html) (49) to search the entire PDB archive (at www.pdb.org). Because of its simplified domain composition, the structure of E. coli YedY (PDB accession number 1XDQ) (chain A, 2.6-Å resolution) was used as bait (53). Table 1 summarizes the results of this “three-dimensional BLAST search”. YedY shows striking structural similarity to the following enzymes: plant SUOX (from Arabidopsis thaliana [PDB accession number 1OGP] [2.6-Å resolution]) (82), chicken SUOX (from Gallus gallus [accession number 1SOX] [1.9-Å resolution]) (47), NIA from P. angusta (accession number 2BII) (1.7-Å resolution) (22), and SUOX from S. novella (accession number 2BLF) (1.8-Å resolution) (41). Table 1 reveals that between 66 and 70% of maximum possible C-α positions overlap between the bait and hit structures.

TABLE 1.

SSM results by using the structure of E. coli YedY (PDB accession number 1XDQ) as bait

PDB accession no.	Description	Resolution (Å)	RMSD^a (Å)	No. of identified backbone C-α atoms	No. of overlapping C-α atoms between the two structures^b	ID_C-α^c	No. of gaps^d
1XDQ	E. coli YedY	2.6	0	262	262	100	0
1OGP	Arabidopsis thaliana SUOX	2.6	2.02	388	185	71	11
2BLF	Starkeya novella SUOX	1.8	1.85	373	173	66	11
2BII	Pichia angusta NIA	1.7	2.00	415	182	69	10
1SOX	Gallus gallus SUOX	1.9	2.03	463	181	69	14

Open in a new tab

Deviation between PDB accession number 1XDQ and the indicated structure.

Number of overlapping C-α atoms between the two structures with the quoted RMSD.

ID_C-α, C-α position identity expressed as a percentage of the maximum possible.

Number of gaps represented in the sequences of the structural alignment.

Figure 1 illustrates the similarities and differences among the five structures. Each of the hits is significantly larger than the bait structure (YedY [PDB accession number 1XDQ]), with the latter comprising a Mo-PPT-coordinating catalytic core and being considered to be a prototypical SUOX fold (see Evolutionary Pathway of the SUOX Fold Proteins). The SUOX fold of the enzymes shown in Fig. 1 comprises a mixed α and β structure organized as two to three β-sheets and 9 to 12 α-helices (47, 53, 82). Each of the four hit proteins contains the distinctive seven-stranded β-barrel structure characteristic of an SUOX dimerization domain (47). SUOX from G. gallus also contains an N-terminal heme b₅ domain. SUOX from S. novella (SorAB) has the additional cytochrome c subunit (SorB) (11, 43, 92) that plays a role similar to that of the heme b₅ domain of G. gallus SUOX (41). Plant SUOX lacks an additional heme-containing subunit or domain (82). Finally, P. angusta NIA resembles the enzyme found in plants, which comprises a SUOX fold followed by a dimerization domain, a heme b₅ domain, and a FAD-containing NADH binding domain. The structure determined by Fischer et al. (22) represents only the N-terminal SUOX fold and the dimerization domain. A structure is available for the NADH binding domain from the NIA enzyme from Z. mays at a resolution of 2.5 Å (54). With the exception of YedY, each of the SUOX fold enzymes has a well-defined substrate for which a wealth of kinetic data has been collected.

FIG. 1. — Unique structures sharing the SUOX fold. (a) Plant SUOX (PDB accession number 1OGP). (b) Chicken SUOX (accession number 1SOX). (c) *E. coli* YedY (accession number 1XDQ). (d) *P. angusta* NIA (accession number 2BII). (e) *S. novella* SUOX (accession number 2BLF). The positions of the PPT binding, dimerization, and b₅-binding domains are labeled PPT, d, and b₅, respectively. The figure was created with the PYMOL molecular visualization package (DeLano Scientific LLC, Palo Alto, CA).

SSM GENERATES AN IMPROVED SEQUENCE ALIGNMENT OF SUOX FOLD ENZYMES

The availability of five experimentally determined protein structures containing the SUOX fold presents an opportunity to investigate the use of SSM matching to generate an improved sequence alignment of these proteins. Using CLUSTALW (87, 88), we generated an alignment inconsistent with the structural overlap implicit in Fig. 1 (data not shown). Figure 2 shows a graphical alignment of structural conservation and sequence within the group of five structures that is based on SSM analyses. It also shows the approximate positions in the primary sequences of five critical active-site residues (see below), the heme b₅ domain, and the dimerization domain. Also shown is the N-terminal tat leader that directs YedY and other cofactor-containing enzymes to the periplasmic compartment of E. coli and other bacteria (1, 3, 90). Figure 3 presents the textual alignment resulting from the SSM approach. Within the multiple alignment, a total of 181 C-α positions (out of a total possible of 262, the number of C-α positions in the YedY structure [PDB accession number 1XDQ]) overlap with an overall root mean square deviation (RMSD) of 1.82 Å. A total of 15 residues are absolutely conserved within the five structures, including the Mo-coordinating Cys residue (YedY-C102, aligned to position 88 in Fig. 3), and the positions of these residues within the structure of YedY (accession number 1XDQ) are indicated in Fig. 4. Remarkably, except for the Cys protein-Mo ligand, only three of these conserved residues, corresponding to YedY-K207, YedY-R194, and YedY-G202, are within 4 Å of the Mo-PPT cofactor. In the case of the two inferred cationic side chains (K207 and R194), the side chains are within 3 Å of the phosphate oxygens of the phosphate group of the cofactor, presumably interacting electrostatically with it. As described below, none of these residues appear to be implicated in the active-site funnels of the SUOX fold-containing family. Two further residues appear to be close to the Mo-PPT shown in Fig. 4, Val-133 and Pro-198; however, this is an artifact of the projection used, with these residues being ∼14 Å and ∼11.9 Å away, respectively.

FIG. 2. — Comparison of secondary structure alignment with sequence features in the five proteins sharing the SUOX fold. The blocks of the upper bar graph represent segments of sequences from the five structures that overlap over 181 C-α positions (out of a total possible of 262) with an RMSD of 1.82 Å in the multiple alignment. It can also be described as the sequence that constitutes the core of the SUOX fold. Gaps represent sequence segments that either do not align or are absent in one or more of the proteins. The lower five bar graphs represent the alignment of the five sequences based on the SSM alignment. In these cases, gaps indicate the absence of sequence. The presence of the heme b₅ domain at the N terminus of chicken SUOX (PDB accession number 1SOX) is indicated, as is the presence of the dimerization domain in all the structures except for that of YedY (accession number 1XDQ). YedY and *S. novella* also possess an N-terminal *tat* leader that is absent from the mature (crystallized) protein. See the text for a discussion of critical Arg and other residues within the five proteins.

FIG. 3. — Sequence alignment based on SSM matching analyses of proteins in the PDB sharing the SUOX fold. The following proteins were selected for analysis: plant SUOX (PDB accession number 1OGP) (82), chicken SUOX (accession number 1SOX) (47), *E. coli* YedY (accession number 1XDQ) (53), *Pichia* nitrate reductase (accession number 2BII) (22), and *S. novella* sulfite dehydrogenase (accession number 2BLF) (41). Capitalized single-letter abbreviations represent the 181 residues whose C-α positions overlap with an RMSD of 1.82 Å. Structural overlap is also indicated in the consensus sequence by use of the “☰” symbol. The 15 absolutely conserved residues are indicated. For the sake of brevity, the N-terminal 110 residues of chicken SUOX (accession number 1OGP) and the C-terminal 50 residues of *Pichia* NIA are not shown. The positions of the conserved substrate-binding Arg residues and the Cys protein-Mo ligand are indicated.

FIG. 4. — Locations of the 15 conserved residues within the family of structurally characterized SUOX fold proteins. Numbering is done according to *E. coli* YedY (PDB accession number 1XDQ), and the conserved residues are displayed, with the backbone rendered in PYMOL in transparent cartoon format. Of the 15 conserved residues, only Cys102 (protein-Mo ligand), Gly202, Lys207, and Arg194 are within 4 Å of the Mo-PPT.

Given the complementary nature of the SSM and sequence-based bioinformatics methods, we compared the sequence identity data obtained from the SSM alignment (Fig. 3) with data obtained from pairwise sequence comparisons using the EMBOSS STRETCHER program (70). Figure 5 shows the relationship between pairwise RMSD values obtained from SSM analyses and equivalent pairwise sequence identities. It is apparent that sequences that are deemed quite dissimilar in conventional bioinformatic analyses, with sequence identities of ≤20%, can overlap, with strikingly low RMSD values of ≤2 Å. For example, E. coli YedY (PDB accession number 1XDQ) and P. angusta NIA (accession number 2BII) overlap, with an RMSD of 2 Å, but are only 11.5% identical in sequence. At the other extreme, plant (accession number 1OGP) and chicken (accession number 1SOX) SUOXs overlap, with an RMSD of 1.02 Å, and are 38.3% identical in sequence. However, when identity is defined as a percentage of possible C-α overlap, these two pairs exhibit 69% and 92% identities, respectively. Overall, these observations illustrate the power of the SSM methodology in exploring structural and sequence identities within protein structure families.

FIG. 5. — Relationship between sequence identity and SSM RMSD values within the SUOX family. Data for the plot of RMSD versus sequence identities (squares) were obtained by doing pairwise comparisons using the SSM server and the EMBOSS STRETCHER program. A plot of the percentage of C-α overlap at the reported RMSD versus RMSD (triangles) is also shown.

ACTIVE-SITE SEQUENCE CONSERVATION SUGGESTS A SIMPLIFIED CLASSIFICATION OF SUOX ENZYMES

Figure 6 shows a simplified view of the active-site region of structurally characterized members of the SUOX fold family. Each of the enzymes has a conserved Cys residue that provides the protein-Mo ligand, with the rest of the Mo coordination sphere being provided by the dithiolene sulfurs of the PPT and two oxo groups (14, 15). With the exception of YedY, each of the enzymes has at least two Arg residues in its active site, and these have been proposed to stabilize the negative charge of the substrate (sulfite and nitrate) or product (sulfate and nitrite) (21, 44, 47). An additional feature in terms of sequence conservation is the presence of a Gly residue following the Cys. A Glu appears at this position in YedY. This rather simplistic description of active-site residues enables the categorization of the SUOX fold enzymes based on active-site conservation and greatly simplifies the interpretation of the cladistics presented below (see SUOX Fold Protein Cladistics). Genuine SUOX enzymes can be coded as R C G R R: they retain the “arginine triad,” have the conserved Cys as a protein-Mo ligand, and retain the active-site Gly. A critical non-active-site feature of enzymes containing the SUOX fold is the presence or absence of a tat leader (3, 90). YedY, for example, can be coded tat N C E | |, indicating that it has a tat leader in its immature form, an Asn at the first Arg triad position (“N”), a Cys protein-Mo ligand, a Glu (“E”) at the position occupied by a Gly in many of the other SUOX fold proteins, and an absence of residues corresponding to the final two Arg residues of the Arg triad (“| |”). This coding will be discussed in more detail when the cladistics of the SUOX fold family of proteins are considered below.

FIG. 6. — Location of conserved residues within the active sites of the SUOX fold proteins. The five protein structures were structurally aligned using a combination of SSM server output and the ALIGN function of the PYMOL molecular graphics package. The plant (PDB accession number 1OGP) and chicken (accession number 1SOX) enzymes each contain the three Arg residues of the Arg triad, and these appear to stabilize the substrate (SO₃²⁻) and/or product (SO₄²⁻) in the active site. *Pichia* NIA (accession number 2BII) contains only one conserved active-site Arg, and YedY (accession number 1XDQ) contains none.

SUOX FOLD PROTEIN CLADISTICS

Our observations on the structural conservation of the SUOX fold in organisms ranging from E. coli to G. gallus prompted us to investigate the full range of organisms in which it occurs. Because of the relatively limited number of SUOX fold structures from a broad taxonomic diversity of organisms, we relied upon a traditional sequence-based bioinformatics approach. The sequences of the proteins sharing significant structural similarity with E. coli YedY were each used as bait to search the UNIPROT sequence database using the BLASTP algorithm at the EBI Wu-BLAST2 server (www.ebi.ac.uk/blast2/; developed by W. R. Gish). Accession numbers from each of the searches were merged, and duplicate entries were eliminated. The EMBOSS SEQRET program (70) was used to retrieve the entire sequence set corresponding to the hits in the Wu-BLAST2 searches. This generated a sequence set of 559 sequences that were used to generate a CLUSTALW alignment (87, 88), and obvious outliers were removed by inspection. In order to generate a suitable data set for evolutionary analyses, the sequences were filtered pairwise to remove sequences with greater than 60% identity to any other sequence in the data set (using the EMBOSS STRETCHER program) (70). This approach yielded a data set comprising 137 distinct sequences. Dendrograms grouping the SUOX-related proteins into discrete clades were generated using the TREEVIEW program (http://taxonomy.zoology.gla.ac.uk/rod/treeview.html; R. D. M. Page). In order to simplify this process, the clades were assigned on the basis of the simplified classification scheme outlined above (see Active-Site Sequence Conservation Suggests a Simplified Classification of SUOX Fold Enzymes), generating the unrooted tree presented in Fig. 7. Sequences corresponding to members of the individual clades were subjected to individual CLUSTALW alignments. Tentative functional assignments for each branch were based on BLASTP searches carried out against the annotated SWISSPROT protein sequence database hosted at the National Center for Biotechnology Information (at http://www.ncbi.nlm.nih.gov/) using the BLASTCL3 network client (available at http://www.ncbi.nlm.nih.gov/), which accesses the NCBI BLASTP server.

Inspection of Fig. 7 in combination with the simplified SUOX fold classification described above (see Active-Site Sequence Conservation Suggests a Simplified Classification of SUOX Fold Enzymes) enabled us to split the tree into eight distinct clades. The taxonomies of the organisms corresponding to each protein were extracted for each sequence from the UNIPROT flat-file entry. Our final data set included sequences from 77 Proteobacteria species, 19 Actinobacteria species, 11 eukaryote species, 9 Archaea species, 4 Cyanobacteria species, 4 Deinococcus species, 2 Acidobacteria species, 2 Chloroflexi species, 2 Firmicutes species, 1 Aquificae species, 1 Bacteriodetes, and 1 Planctomycetes species. Thus, the SUOX fold appears in organisms encompassing the entire span of evolution in the three domains of life.

One characteristic of the SUOX fold proteins that will be discussed in more detail below (see Correlation between the Presence of a tat Leader Sequence and a Diderm Bacterial Cell Envelope Morphology) is the correlation between clade assignment, the presence of a tat leader, and bacterial cell envelope morphology. Because a significant number of species possessing outer membranes are reported as lacking one on the basis of the Gram stain, we are reluctant to use the “gram positive”/“gram negative” terminology herein. We instead choose to use the terms monoderm (lacking an outer membrane) and diderm (possessing an outer membrane) (29). Species with outer membranes were identified on the basis of morphological studies or the presence of readily identifiable outer membrane proteins encoded by their genomes. In the context of this work, the following domains/phyla were deemed to be monoderms: Archaea, Firmicutes, Actinobacteria, and Planctomycetes. The following bacterial phyla were deemed to be diderms: Proteobacteria, Deinococcus, Chloroflexi, Cyanobacteria, Bacteriodetes, Aquificae, and Acidobacteria.

Diderm bacterial species are gray in Fig. 7, while monoderms are black. Eukaryotic species are green, and the Archaea are blue. It is clear that the archaeal sequences are located in clades 7 and 8, whereas the eukaryotic species are located in clade 5. Each clade is distinguished by a group of conserved motifs listed in Table 2 and discussed below.

TABLE 2.

Conserved motifs within individual clades of proteins containing the SUOX fold

Clade^a	No. of sequences^b	Function^c	Simplified classification^d	Arginine triad (no. of arginine residues)^e	Dimerization domain^f	Heme binding domain^g	FAD domain^h	Conserved motifsⁱ
1	30	YedY-1	tat N C E \| \|	No	No	No	No	NFYEF, RCVExW, PYxExL, GAPxR, PWKYGFK, YGFxANVNP, NGY
2	15	YedY-2	tat N C E \| \|	No	No	No	No	CVEGW, HPQT, KLGYK, GxPxR
3	24	SUOX	tat R C G R \|	Yes (2)	Yes	Yes (subunit)	No	CxGNxR⁹, GxPxR
4	3	SUOX	tat R C S R \|	Yes (2)	Yes	Yes (subunit)	No	CSGNGR, GxPxR^j
5	16	SUOX, NIA	Δtat R C G R R	Yes (3)	Yes	Yes/no	Yes/no	PxNxEP, RNH, CAGNRR, HGxPxR, GARxVKW (I/L), RV(E/D)(V/L)S, WNxRxG
6	15	SUOX, NIA	tat R C (A/G) \| R	Yes (2)	Yes	Yes (subunit)	No	RxHxG, HGxV, FxECxxN, WTGV, EgxDx(S/A), QNGE, EQGYPxR, GLAWSGxG, SRxxDxTGY
7	22	SUOX, YedY, NIA	Δtat \| C T \| \|	No	No	No	No	FHCVTxWS, GxPxR, GYWE
8	12	SUOX	Δtat R C S \| \|	Yes (1)	Yes	No	No	DFYR(V/I)D, CVSN, HGYPVR, GxAWA, PxGA (S/T)G

Open in a new tab

Branches identified in Fig. 7.

Number of sequences in each clade of Fig. 7.

Function identified by BLAST searches of sequences against the curated SWISSPROT database.

Simplified sequence classifications as described in the text. This is the basis for assignment to the individual clades shown in Fig. 7.

The arginine triad stabilizes the product/substrate at the active site of SUOX (47). The number of Arg residues is given in parentheses.

The dimerization domain, if present, is located at the C terminus of the SUOX domain.

The heme binding domain can either comprise a separate subunit (as in the SUOX from S. novella, clade 3), can comprise a separate domain at the N terminus of the SUOX fold (as in plant SUOX, clade 5), or can be located C terminal to the dimerization domain (as in plant NIA, clade 5).

The FAD domain is located at the C terminus of the plant-type nitrate reductase.

ⁱ

Underlined motifs imply conservation across a few clades. Motifs in boldface type imply conservation across all clades.

Motifs for clade 4 are underreported. As clade 4 contains only three members, the sequences are highly conserved across their entire lengths. As such, only common motifs found in other clades are reported therein.

Clade 1: Bona Fide YedY Sequences (YedY-1)

We have arbitrarily set clade 1 to be the branch containing YedY from E. coli and its close homologs. Sequences in this clade are derived from diderm species, with all the sequences possessing a tat leader (1, 78, 79, 91). Since in the case of E. coli, YedY is a soluble periplasmic enzyme, there is presumably an evolutionary advantage in having an outer membrane to prevent its loss to the external environment. The correlation between the presence of soluble tat-exported proteins and bacterial cell wall morphology has been noted elsewhere (75) and will be discussed in more detail below (see Correlation between the Presence of a tat Leader Sequence and a Diderm Bacterial Cell Envelope Morphology). Clade 1 is composed mainly of sequences from the Proteobacteria, with 24 entries, along with two for the Cyanobacteria, one for the Acidobacteria, one for Deinococcus, and one for Chloroflexi.

The sequence motifs conserved in the YedY-1 clade are indicated in Table 2 and will be discussed in detail below (see “Conserved Motifs in YedY”).

Clade 2: YedY-2 Sequences That Are Closely Related to YedY-1

Clade 2 diverges from clade 1 but is very similar to it. It is distinguished by the presence of two unique conserved motifs, HxQT and KLGYK (Table 2). This branch is comprised mostly of Proteobacteria, with 11 entries in total. The remaining four homologs are found in two Acidobacteria sequences and two Cyanobacteria sequences. As is the case for clade 1, these organisms are all diderms, and each protein is predicted to be a tat-exported soluble periplasmic protein.

Clade 3: tat-Exported Periplasmic SUOX

Clade 3, along with clade 7, contains sequences from both monoderm and diderm bacterial species. The majority (14/24) of the sequences in clade 3 possess a tat leader peptide, indicating that these are secreted into the periplasm or extracellular space. Members of clade 3 are identified as being bacterial SUOX enzymes in BLASTP searches against the SWISSPROT database and include the structurally characterized SorAB SUOX of S. novella (41). Each member of the clade possesses the first two Arg residues of the Arg triad. Members of this clade containing an outer membrane are 16 Proteobacteria, one Acidobacteria, one Deinococcus, and one Bacteriodetes species. Monoderm members comprise four Actinobacteria sequences and one Firmicutes sequence.

Clade 4: a Second Clade of tat-Exported Periplasmic SUOX

Clade 4 contains the fewest members of those represented in Fig. 7 and has a distinct simplified classification (Table 2). Its three members all belong to the Proteobacteria. Like clade 3, these sequences possess the first two Arg residues of the Arg triad.

Clade 5: Eukaryotic SUOX and NIA Enzymes

Clade 5 includes the archetypal eukaryotic SUOX and NIA enzymes, including plant SUOX (PDB accession number 1OGP) (82), chicken SUOX (accession number 1SOX) (47), and P. angusta NIA (accession number 2BII) (22). All members of this clade lack a tat leader but contain the catalytic Arg triad. This clade is comprised of 12 eukaryotic proteins, six of which belong to the Metazoa, three of which belong to the Fungi, two of which belong to the Viridiplantae, and one of which belongs to the Mycetozoa. The remaining species belong to the Actinobacteria, with three entries, and the Proteobacteria, with one entry. As will be discussed below, it is notable that this clade contains sequences in which there is a variability of domain order. For example, in chicken SUOX, the heme b₅ domain precedes the SUOX fold, but in the plant-type nitrate reductases (including that of P. angusta), it follows the dimerization domain. It is absent in the plant-type SUOX. These observations suggest that “domain shuffling” may have occurred recently on the evolutionary timescale.

Clade 6: Bacterial Periplasmic SUOX and NIA Enzymes

Clade 6 encompasses enzymes tentatively identified in BLASTP searches against the curated SWISSPROT database as bacterial SUOXs and NIAs. Fourteen Proteobacteria and one Bacteriodetes species are represented. All the homologs possess a tat leader and are predicted to be exported into the periplasm. In this group, the first and third Arg residues of the Arg triad are conserved.

Clade 7: Cytoplasmic Bacterial SUOX/YedY/NIA

The bacterial species observed in clade 7 are the most diverse of those represented in Fig. 7. The sequences lack a tat leader, and they also completely lack the Arg residues of the Arg triad. On this basis, although tentatively identified functionally as being SUOX/YedY/NIA enzymes, it is unlikely that they interact with simple inorganic anionic substrates. Species lacking an outer membrane are represented by six Archaea, three Actinobacteria, one Firmicutes, and one Planctomycetes species. Diderm bacterial species are represented by six Proteobacteria, two Deinococcus, one Chloroflexi, one Aquificae, and one Acidobacteria species. The presence of archaeal sequences indicates that this clade arose very early on in evolution.

Clade 8: Cytoplasmic SUOX Enzymes from Monoderm Bacteria

Uniquely, clade 8 exclusively contains sequences from monoderm bacteria and Archaea. The first Arg of the Arg triad is retained, and the active-site Gly is replaced by a Ser residue. Nine sequences are from the Actinobacteria, and three are from the Archaea. The lack of a tat leader indicates that the mature proteins are cytoplasmically localized.

PRESENCE AND FUNCTION OF THE DOMAINS OBSERVED IN STRUCTURALLY CHARACTERIZED PROTEINS CONTAINING THE SUOX FOLD

Three structurally identifiable domains are observed in the eukaryotic SUOX/NIA clade (clade 5) (Fig. 7). These are the critical Mo-PPT binding SUOX domain, the heme b₅ domain, and the dimerization domain. Within the subgroup of proteins of known structure, these are best represented by the SUOX enzyme from G. gallus (47). In G. gallus SUOX, the domain order is heme b₅, Mo-PPT, and then dimerization. In the plant-type NIA, the domain order is Mo-PPT, dimerization, heme b₅, and then an FAD-containing NADH-oxidizing domain (8). The structure of the NADH-oxidizing flavoprotein domain from maize has been solved to a resolution of 2.5 Å (PDB accession number 1CNF) (54). As can be seen in Fig. 1 and 2, the SUOX fold is common to all members and defines the family. The variability of the presence and locations of the other domains within the SUOX fold enzymes can provide important clues as to the evolutionary relationships among the sequences represented in Fig. 7.

The dimerization domain can act as the interface between SUOX fold monomers, and many enzymes containing the SUOX fold have been shown to be dimers (20). However, the role of the dimerization domain in enzyme function remains enigmatic. While the monomeric forms of SUOX enzymes are catalytically active, the dimerized form can be threefold more active (12). Mutations that cause an SUOX deficiency in humans have been identified in the dimerization domain as well as in the SUOX fold, highlighting the importance of the latter (46).

As will be discussed below (see Evolutionary Pathway of the SUOX Fold Proteins), the appearance of a heme-containing domain or subunit appears to be a critical evolutionary step following the emergence of the dimerization domain. In the case of S. novella SUOX (SorAB), a separate heme c-containing subunit (SorB) is observed in the protein structure (PDB accession number 2BLF) in a conformation that provides a center-to-center distance (Fe to Mo) of 16.6 Å (41). Crucially, the edge-to-edge distances between these cofactors are only 7.0 Å from one of the heme propionates to the PPT and 8.5 Å from the same propionate to the Mo atom of the Mo-PPT. Kappler et al. (41, 42) identified a plausible pathway of bond-orbital and H bonds that could serve as a conduit between the two redox-active cofactors.

In the case of chicken SUOX, the situation is quite different. The heme-containing b₅ domain is located at the N terminus of the protein, but the edge-to-edge distance between the two cofactors is 21.3 Å, and the center-to-center distance is 25.8 Å (47). These distances are beyond the 14-Å maximum that has been proposed to be the maximum for kinetically competent intercenter electron transfer (61-64, 68). It has been proposed that there is a hinge region comprising approximately 10 amino acids between the heme b₅ domain and the Mo-PPT/dimerization SUOX fold that permits domain rearrangement, allowing the heme to approach the Mo-PPT to permit catalytically competent electron transfer rates (67). This hypothesis is based on the heme domain interacting electrostatically with the Arg triad in the absence of substrate or product, thus facilitating interdomain electron transfer.

In the case of NIA enzymes such as that from P. angusta, the presence of a similar hinge region between the dimerization domain and the heme b₅ domain that is localized C terminal to the dimerization domain has been proposed (8). Indirect evidence for the role of this hinge in activating plant NIA has been obtained from studies using spinach (Spinacea oleacea) NIA as a model system. This enzyme shares the same domain order within its primary sequence as that observed in the P. angusta enzyme, with the heme b₅ domain following the SUOX fold. Spinach NIA is inactivated in the dark via a process in which a Ser residue (Ser543) in the hinge region connecting the Mo-PPT dimerization domain with the heme b₅ domain is phosphorylated (16, 60), followed by binding of the NIA inhibitor protein (24, 55, 60). Binding of the NIA inhibitor protein presumably constrains the hinge function and results in a Mo-PPT-to-heme distance well beyond the 14-Å limit for catalytically competent intercenter electron transfer. Thus, in both animal SUOX and plant NIA enzymes, available data suggest a role for hinge regions in facilitating and regulating interdomain electron transfer. Interestingly, heme c of S. novella SUOX is located in close juxtaposition not only to the Mo-PPT cofactor but also to the substrate binding funnel, suggesting that perhaps its location is functionally equivalent to that of the heme b₅ domain of chicken SUOX (67).

PROTEIN-PPT INTERACTIONS

Figure 8 shows the structure and selected nomenclature of the organic moiety of the Mo-PPT cofactor that is bound by the SUOX fold. It is the product of a complex biosynthetic pathway that has been extensively reviewed (57, 58). It incorporates a large number of noncarbon atoms able to interact with a range of side chain functionalities of the protein scaffold. As discussed below, many of the conserved motifs identified in Table 2 contribute to Mo-PPT coordination in a clade-specific manner.

FIG. 8. — Structure of the PPT moiety of the Mo-PPT cofactor. The correct naming conventions for individual atoms of the PPT are detailed. This nomenclature is used for interactions between the protein and the PPT discussed in the text.

It is notable that the E_m values for the Mo(IV/V) and Mo(V/VI) couples vary considerably between the enzymes sharing the SUOX fold. In chicken SUOX, the E_m values have been reported to be −86 mV and 131 mV for the Mo(IV/V) and Mo(V/VI) couples, respectively (86), and in S. novella, they are reported to be −46 mV and 160 mV (42). Thus, a partial consensus suggests that the true SUOXs have widely separated E_m values that render the Mo(V) electron paramagnetic resonance spectrum readily observable. In the case of the NIA enzymes, the potentials are closer together: the reported values for the spinach enzyme (Spinacea oleacea) are −6 mV and 2 mV (45), and in green algae (Chlorella vulgaris), the reported values are −54 mV and −34 mV (84). Thus, in the NIA enzymes, the Mo(V) is less easily observed by electron paramagnetic resonance than are true SUOX enzymes. Finally, YedY from E. coli exhibits only the Mo(IV/V) transition, and this occurs with an E_m of 132 mV (7). Thus, in YedY, the Mo(V) state is the most accessible of all the SUOX fold enzymes studied to date. Because of the observed complexities of the Mo coordination environment in enzymes having the SUOX fold (20, 34, 35, 37, 46), the factors determining the Mo(IV/V) and Mo(V/VI) midpoint potentials are poorly understood. It is possible that the coordination of the organic component of the Mo-PPT cofactor may play a crucial role in defining the Mo(IV/V) and Mo(V/VI) midpoint potentials.

Conserved Motifs in YedY

The conserved motifs of E. coli YedY (of the YedY-1 clade) are concentrated mostly around the PPT cofactor (53), and these motifs participate primarily in its coordination. Of primary importance is ⁴⁵NFYEF⁴⁹ (Fig. 9). Asn45 interacts with the phosphate group of the PPT via its backbone amide nitrogen (3.7 Å). Its side chain nitrogen also interacts with the Mo atom (3.6 Å). The backbone amide nitrogen of Tyr47 hydrogen bonds to the 6-keto oxygen of the pterin. Glu48 appears to stabilize the 7-amine nitrogen as well as the amine attached to C-8 of the pterin using both of its side chain carboxylate oxygens. This Glu residue is not conserved in any of the other clades presented in this study. An additional conserved motif, ¹⁵⁹PYxExL¹⁶⁴, is also unique to the YedY-1 clade, and some detail of this is also shown in Fig. 9. While most of its residues do not interact with Mo-PPT, Tyr160 appears to “stack” with the PPT ring system in a manner similar to that observed between a Tyr residue (Tyr1005) and one of the pterin rings of the Mo-bisPGD cofactor of FdnGHI (40). A similar phenomenon occurs with the Tyr47 residue of the YedY-1 ⁴⁵NFYEF⁴⁹ motif, but in this case, there is a greater interplane angle between the Phe ring and the two coplanar rings of the PPT. The side chains of both Tyr47 and Tyr160 are located within ∼4 Å of the PPT ring system. Interestingly, the absolutely conserved Lys207 serves an essential role in the coordination of the PPT, as it interacts with the C-8 amine and 9-imino nitrogen of the pterin ring as well as with two oxygens of the phosphate group (Fig. 4). This Lys, while absolutely conserved within the structurally characterized SUOX fold proteins, is not surrounded by other conserved residues within an identifiable motif.

FIG. 9. — Active-site motifs of YedY. Conserved motifs uncovered by the CLUSTALX alignment were analyzed in the structure of YedY. The conserved motif NFYEF from clade 1 makes six potential interactions with Mo-PPT. The conserved motif PYxExL contributes an aromatic ring (Tyr160 in YedY) to an aromatic base-stacking formation consisting of F45, Y160, and the outer ring of the pterin group. The cysteine-containing motif is not visualized in this figure. All distance values are within the accepted values for hydrogen bonding.

Three additional motifs also appear in the general area of the Mo-PPT active site of YedY. These motifs are ¹⁰¹RCVExW¹⁰⁶, ¹⁹⁸PWKYGFK²⁰⁴, and ²³¹YGFxANVNP²³⁹. Notwithstanding the role of the conserved Cys residue in Mo coordination and catalysis, Tyr231 (7.2 Å), Tyr47 (7.6 Å), and the previously mentioned Asn45 (3.6 Å) appear to be close enough to the Mo and its adjacent substrate binding site to have an influence on the catalysis of the as-yet-unknown interconversion catalyzed by members of the YedY-1 clade.

Another YedY-1 motif identified in Table 2, ²⁷¹NGY²⁷³, is found on the surface of YedY. While the crystal structure of YedY was found as a pentamer, the ²⁷¹NGY²⁷³ domain does not occur between monomer surfaces. The ¹⁹⁰GAPIR¹⁹⁴ motif, which is found conserved as GxPxR across all of the clades, is discussed below (see “Structural Motifs in Clades of Unknown Structure and Function”). Overall, we conclude that at least two of the conserved motifs of the YedY-1 clade are involved in PPT coordination.

Conserved Motifs in P. angusta NIA and Plant and Chicken SUOX (Clade 5)

The known structures of proteins represented in clade 5 contain motifs that serve a function similar to those found in clade 1 but which have distinct sequences. In the following text, residues are numbered according to chicken SUOX numbering, although the motifs are also conserved in plant SUOX and P. angusta NIA. The function of the YedY-1 ⁴⁵NFYEF⁴⁹ motif is assumed by the ¹³⁸RNH¹⁴⁰ motif in the clade 5 enzymes. While YedY-1 ⁴⁵NFYEF⁴⁹ appears to interact with Mo-PPT at up to six positions (Fig. 9), the ¹³⁸RNH¹⁴⁰ motif appears to make only two interactions (Fig. 10). The first is with the PPT 6-keto oxygen group via a potential hydrogen bond between the 6-keto oxygen and the backbone amide nitrogen of the histidine (2.8 Å). The 6-keto oxygen is also only 2.6 Å from one of the imidazole nitrogens of the His140 side chain and could hydrogen bond with it in its protonated form. The backbone carbonyl oxygen of Arg138 is located 3.9 Å from one of the phosphate oxygens. Note that Arg138 is equivalent to Arg1 of the Arg triad (Fig. 6).

FIG. 10. — RNH motif of chicken SUOX. Conserved motifs uncovered by the CLUSTALX alignment were analyzed in the structure of chicken SUOX. The RNH motif found in the active site of clade 5 enzymes contributes to the coordination of the PPT (four potential interactions), including R138, a crucial residue in the Arg triad. The product of sulfite oxidation, sulfate, is in close proximity to R138.

A second motif, ²⁹⁶GARxxKW(I/L)³⁰³, also interacts with the PPT (Fig. 11) but on the opposite side of the ring system compared to the interactions observed with the ¹³⁸RNH¹⁴⁰ motif. The ²⁹⁶GARxxKW(I/L)³⁰³ motif includes the absolutely conserved Lys residue equivalent to YedY-K207 (Fig. 4). The amide bond nitrogen of this Lys301 interacts with the N-9 of the PPT ring system (Fig. 8). Its side chain amine interacts with the PPT phosphate group. The amide bond carbonyl oxygen interacts with the amine attached to the C-8 position in the PPT ring system. These three interactions are absolutely conserved among the structurally characterized SUOX fold enzymes. Of the other residues within the ²⁹⁶GARxxKW(I/L)³⁰³ motif, it is the backbone carbonyl oxygen of the first “x” residue (Ser299) that appears to be of greatest importance in coordinating the PPT. Its carbonyl oxygen appears to form a hydrogen bond with the N-10 nitrogen.

In the clade 5 proteins, the ¹⁸⁵CAGNRR¹⁹⁰ motif is equivalent to the YedY-1 ¹⁰¹RCVExW¹⁰⁶ motif (Fig. 3 and Table 2). In both cases, the Cys residue is the protein-Mo ligand, and in the clade 5 ¹⁸⁵CAGNRR¹⁹⁰ motif, the second Arg residue (Arg190) is Arg2 of the Arg triad (Fig. 3 and 6). The clade 5 ¹¹⁴PxNxEP¹¹⁹ motif functions to coordinate a portion of the ¹³⁸RNH¹⁴⁰ motif (not shown), with a hydrogen bond between the side chain oxygen of the ¹¹⁴PxNxEP¹¹⁹ Asn and one of the side chain nitrogens of the ¹³⁸RNH¹⁴⁰ Arg. When looking at a surface representation of chicken SUOX, the ¹³⁸RNH¹⁴⁰ motif is found at the active-site funnel (not shown). Additionally, the ¹¹⁴PxNxEP¹¹⁹ motif is located adjacent to it. The final conserved motif of this family, ³⁸¹RV(E/D)(V/L)S³⁸⁵, is found in the dimerization domain.

Conserved Motifs in S. novella SUOX

Clade 3, containing S. novella SUOX, shows a marked lack of conserved motifs compared to the other clades. In this case, only two motifs can be found (Table 2), and they resemble motifs of the previously described clades. Due to the lack of motifs, analysis of individual conserved residues was performed. In many cases, the conserved residues were related to those found in the other clades as well, providing insight into exactly which residues are of importance in each motif. As is the case with the other clades containing members of known structure, the active-site Cys is found in the same position, as well as an active-site Tyr and the Lys bearing multiple interactions with the PPT. While the P. angusta, plant, and chicken enzymes all have a conserved motif in their dimerization domains, the S. novella SUOX has only three absolutely conserved residues in the dimerization domain. These residues (Trp307, Gly346, and Gln349), while not in an identifiable motif, are localized to the protein surface. Additional semiconserved Trp and Leu residues are found in this region. However, these residues are not found on the surface of the dimerization domain. There are two additional conserved Gly residues, which are not found near the surface, but these may be to add flexibility to the dimerization domain, as they are both found near the “joint” of this domain.

Structural Motifs in Clades of Unknown Structure and Function

Across all known structures, there exists a surprisingly low level of conservation of residues surrounding the active site (Fig. 4). The conserved motifs of the remaining branches can be seen in Table 2. Unfortunately, no crystal structures are available for enzymes found in these branches, and as such, an analysis of the potential functions of these motifs cannot be done. This problem is compounded by a relative lack of experimental data on these enzymes.

Aside from the integral Cys residue, all of the known structures contain a Lys residue that coordinates the PPT as well as an Arg residue that interacts with the phosphate group of the PPT (Fig. 4). This Arg is part of a motif (GxPxR) that is conserved across all of the clades of our analysis. However, most branches differ in regard to the residues preceding and internal to this motif. Table 2 shows that in clades 5 and 8, GxPxR is preceded by a His. In the structures of clade 5 enzymes, this His hydrogen bonds to the same phosphate oxygen that is being coordinated by the Arg of the GxPxR motif. In YedY, as well as most of the clade 1 enzymes, the GxPxR motif is usually preceded by an Asn. Much like the His residues, this Asn interacts with the phosphate group. The backbone carbonyl group of this residue also appears to interact with the side chain oxygen of the Arg. The Gly and Pro that are internal in this motif most likely give the motif a characteristic “U shape.” The Pro in the center of the motif causes a turn in the backbone, while the Gly gives freedom to the preceding residue, allowing it to adopt a proper orientation. Thus, it would appear that the GxPxR motif offers extensive coordination of the phosphate group.

DIFFERENTIAL REDOX PARTNERS

The redox partners of the clade 5 enzymes appear to be well established. As described above, the heme b₅ domain of animal SUOX appears to swivel toward the Mo-PPT active site to accept electrons from the Mo-PPT cofactor. Plant SUOX is localized to the peroxisomes and appears to donate electrons directly to oxygen to form peroxide (32). Plant NIA also appears to undergo domain movement to facilitate electron transfer to its heme. In the case of the structurally characterized clade 3 SorAB enzyme from S. novella, an accessory subunit that coordinates a heme c appears to be the primary electron acceptor from the Mo-PPT active site, with no domain or subunit movement being required for enzyme turnover. Little is known about the redox partners of members of the remaining clades of SUOX fold enzymes.

As stated above, the second gene in the E. coli yedYZ operon, yedZ, encodes a TM heme-containing protein (7). In order to gain further insights into YedZ, E. coli YedZ was used as bait in a BLASTP search of the UNIPROT database, and we obtained a sequence data set comprising 107 sequences after filtering pairwise with a 95% identity cutoff filter. These sequences were aligned using CLUSTALX (87, 88), and the resultant alignment was analyzed using the EMBOSS TMAP program (70) to generate the TM topology shown in Fig. 12, which is essentially in agreement with that proposed by Drew et al. (17). Inspection of Fig. 12 reveals the placement of sequential Arg residues (Arg77 and Arg78) toward the cytoplasmic end of TM helix 3. In order to verify the predicted inclusion of these residues within the hydrophobic core of YedZ, we subjected the entire sequence set to analysis using the TMHMM server (version 2; https://www.cbs.dtu.dk/services/TMHMM/) (50, 85). The TM topology predicted for the YedZ sequence set by the TMHMM server is essentially identical to that predicted by the EMBOSS TMAP program.

FIG. 12. — Predicted TM topology and distribution of conserved residues in the YedZ family. A sequence data set of 107 YedZ sequences was subjected to hydropathy analyses as described in the text. The positions of the absolutely conserved residues within the bacterial YedZ family are shown. The numbering of the TM helices used in the text is indicated beneath the topology model.

Twenty-four residues are highly conserved within the YedZ family, including the two Arg residues mentioned above. Most significant are three His residues that are likely candidates for heme-iron coordination. His91 (E. coli numbering) and His164 are located toward the periplasmic ends of TM helices 3 and 5, and these are the most likely to be involved in the coordination of heme b observed by Brokx et al. (7). Hydrophobic b-type cytochromes typically have conserved Arg residues whose side chains interact with one or more of the heme propionates. Examples of this include mitochondrial and bacterial complex II (93), E. coli nitrate reductase A (4, 5), and the cytochrome bc₁ complex (94). In the case of YedZ, a candidate Arg for interacting with the heme propionates is located toward the periplasmic end of TM helix 4 (Arg115). The identification of residues involved in quinol binding and redox reactions is more problematic, with a cluster of conserved residues being located at the cytoplasmic ends of TM helices 4 and 5, including His151 and the nearby Tyr155.

Of the 107 YedZ sequences obtained in our analyses, 98 of the sequences are proteobacterial in origin, 5 are from the Chloroflexi, 2 are from the Acidobacteria, and two are from the Deinococci. All these species are diderms. When sequences are mined using E. coli YedY as bait, a data set of 114 sequences is obtained following the use of a 95% cutoff filter. Of these, 103 are proteobacterial, 5 are from the Chloroflexi, 4 are from the Cyanobacteria, and two are from the Deinococci, and all the phyla represented are diderms. Thus, with the exception of the representation of cyanobacterial species in the YedY sequence set, there is a strong correlation between the phylogenic representations of YedY- and YedZ-type subunits, with both sets being largely proteobacterial in origin. In the case of YedY, each sequence bears a tat leader at its N terminus, and the status of this subunit as a soluble periplasmic enzyme renders the presence of an outer membrane necessary to prevent its loss to the bulk milieu.

CORRELATION BETWEEN THE PRESENCE OF A tat LEADER SEQUENCE AND A DIDERM BACTERIAL CELL ENVELOPE MORPHOLOGY

The correlation between the observation of YedY sequences and a diderm cell envelope morphology of the species in which they are found prompted us to investigate the correlation between the localization of bacterial SUOX-type enzymes and the presence of an outer membrane. We examined this issue in each of the clades shown in Fig. 7, with clades possessing a tat leader being deemed to contain periplasmically localized proteins. Table 3 shows the results of our analyses of the species encoding the 137 sequences represented by the eight clades shown in Fig. 7. There is a strong correlation between the presence of a tat leader and the sequence coming from a diderm species. This is clearly demonstrated by clades 1, 2, 4, and 6. In each case, all of the sequences come from species possessing an outer membrane. In clade 3, 19 out of 24 sequences come from diderm species, with 14 out the 24 sequences having tat leaders. The monoderm species of clade 3 are four actinobacterial species (Arthrobacter aurescens, Saccharopolyspora erythraea, Streptomyces ambofaciens, and Rubrobacter xylanophilus) and one Firmicutes species (Bacillus sp. strain NRRL B-14911).

TABLE 3.

Correlation between periplasmic localization and presence of an outer membrane

Clade^a	Function^b	Total no. of species^c	No. of species containing an OM^d	tat leader^e	Major group (no. of species)
1	YedY-1	30	30 (0)	Yes	Proteobacteria (25)
2	YedY-2	15	15 (0)	Yes	Proteobacteria (11)
3	SUOX	24	19 (5)	Yes	Proteobacteria (16)
4	SUOX	3	3 (0)	Yes	Proteobacteria (3)
5	SUOX/NIA	16	1 (15)^f	No	Eukaryotes (12)
6	SUOX/NIA	15	15 (0)	Yes	Proteobacteria (15)
7	SUOX/YedY/NIA	22	11 (11)	No	Archaea (6), proteobacteria (5)
8	SUOX	12	0 (12)	No	Actinobacteria (9)
Total	SUOX fold	137	93

Open in a new tab

Clades are indicated in Table 2 and Fig. 7.

Function identified by BLAST searches of sequences against the curated SWISSPROT database, except in cases where function has been assigned by experiment.

Total number of species represented by each clade shown in Fig. 7.

Species from bacterial phyla possessing an outer membrane (OM). Numbers in parentheses indicate the number of additional entries in each clade that are comprised of archaea, monoderm bacteria, and eukaryotes.

The presence of a tat leader was determined by inspection of the sequence alignments of sequences assigned to each clade.

Inclusion of clade 5 is done for the sake of completeness; however, four prokaryotic species are represented in it.

The clades dominated by sequences lacking a tat leader have a more balanced distribution between diderm and monoderm species. Species represented in clade 7 are equally distributed between the cell envelope morphologies, and those represented in clade 8 are exclusively from monoderm species (Actinobacteria). The proteins represented in clades 7 and 8 are predicted to be soluble cytoplasmic enzymes, and thus, no correlation between their occurrence and the cell envelope morphology of the species in which they are found is expected. The presence of an outer membrane in the species represented in clades 1, 2, 4, and 6 presumably serves to prevent the loss of the tat-targeted SUOX-type proteins to the bulk milieu. Similar correlations have been observed in analyses of other closely related families of proteins that can be either cytoplasmically located or targeted for export by the tat system (75).

CORRELATION BETWEEN THE PRESENCE OF SUOX FOLD PROTEINS AND YedZ

Our analyses of SUOX fold protein and YedZ sequence data prompted us to investigate the correlation between their occurrences within the genomes analyzed herein. This issue is important because YedZ may interact with a variety of non-SUOX-fold periplasmic proteins. We addressed this issue by comparing the unfiltered list of SUOX fold proteins (559 entries) with an unfiltered list of YedZ homologs (171 entries). All species containing both a SUOX fold protein and a YedZ homolog were compiled, and the former species were sorted into branches based on the presence of the conserved motifs described above. The results show a correlation between the appearance of YedZ homologs and the species represented in clades 1, 2, 4, and 6 (Fig. 7).

Utilizing an approach using both the accession numbers of the proteins and an operon searching server (http://string.embl.de/newstring_cgi/show_input_page.pl#), we identified an additional membrane-bound cytochrome b family whose genes appear adjacent to those of a subset of the SUOX fold proteins. This family, referred to herein as FdnI-like, is represented by Q2N9I4_ERYLH (UNIPROT accession no. for Erythrobacter litoralis), which has a TM topology similar to that found in the four TM members of the diheme FdnI family (40, 75). The distributions of the two types of hydrophobic cytochromes b that may interact with SUOX fold enzymes are listed in Table 4.

TABLE 4.

Distribution of differential redox partners

Clade	No. of species with cell wall morphology of^a:		No. of species of domain:			tat^b	No. of species with cytochrome family of type^c:
	No. of species with cell wall morphology of^a:		No. of species of domain:				b		c
	Monoderm	Diderm	Bacteria	Archaea	Eukyaryotes		YedZ	FdnI	1	2	3	4	5
1		30	30			Yes	25
2		15	15			Yes	8	5
3	5	19	24			Yes	10	5	11	4	2	1
4		3	3			Yes	2	1				3
5	3	1	4		12	No
6		15	15			Yes	5	3					13
7	11	11	16	6		No
8	12		9	3		No

Open in a new tab

Numbers relate to the species represented in Fig. 7.

tat, targeted to the tat translocon in the majority of cases.

See the text for more information on the cytochrome families.

DISTRIBUTION OF ALTERNATIVE CYTOCHROME REDOX PARTNERS

Compared to many other redox enzyme systems, enzymes of the SUOX class present relatively simple electron transfer relay architectures. These enzymes do not contain chains of [Fe-S] clusters, and their relays typically encompass the Mo-PPT, a heme, and, in the NIA enzymes, an extra NADH binding FAD domain. As mentioned above, a hinge region in animal SUOX enzymes appears to allow the heme b₅ domain to approach the SUOX fold domain and accept electrons from it. S. novella SUOX, rather than having a heme-containing domain, forms a heterodimer with a cytochrome c subunit (SorB). A preliminary BLAST search of the sequence of this SorB subunit returned hits identified as being other cytochrome c subunits. Akin to the YedZ-like protein analysis, we cross-referenced the accession numbers, taxonomies, and operon positions of the cytochrome c proteins against our phylogeny of SUOX fold proteins. In many cases, a cytochrome c gene could be found in an operon with a SUOX protein. We thus determined that only 11 members of our phylogeny contained a SorB-type cytochrome c subunit, and all of these were in clade 3. As seen in Table 2, not only members of clade 3 but also the members of clades 4 and 6 contain separate cytochrome c subunits that are associated with a SUOX fold protein. By searching by accession number and gene position, we identified an additional four unique classes of cytochromes c. These are exemplified by proteins with UNIPROT accession no. A5V4U9_9SPHN (Sphingomonas wittichii RW1), A4AQV2_FLAO (Flavobacteriales bacterium), A3SHJ5_9RHOB (Roseovarius nubinhibens), and Q9LAH4_THINO (S. novella). The distribution of these families across our phylogeny is indicated in Table 4. Thus, there are five families of cytochromes c that can be predicted to interact directly or indirectly with their respective SUOX fold proteins. It is also notable that these cytochromes c appear in operons with SUOX fold proteins bearing a tat leader, suggesting that they form part of electron transfer systems coupling the periplasmic SUOX fold proteins to the membrane-intrinsic quinone pool. However, with the exception of SorB (family 1), there is no confirmation that these proteins form stable heterodimeric complexes with their putative SUOX fold partners.

Four of the cytochrome c families appear in members of clade 3, which agrees with the observation that clade 3 contains a greater taxonomic diversity than clades 4 and 6 do. Both clades 4 and 6 claim a single cytochrome c family. The distribution of these cytochrome c proteins is shown in Table 4. Cytochromes c appear in three clades totaling 41 members. However, six of the members (five from clade 3 and one from clade 6) are missing this subunit.

PRESENCE OF MULTIPLE REDOX SYSTEMS

Above (see Correlation between the Presence of SUOX Fold Proteins and YedZ and Distribution of Alternative Cytochrome Redox Partners), we detailed the presence of both soluble and membrane-bound redox partners of the SUOX fold proteins. Interestingly, and as detailed in Table 4, these two systems overlap in several taxonomies. Inspection of our unfiltered list of SUOX fold proteins (559 entries) reveals that many species contain multiple (at least three) SUOX fold proteins located in the different clades shown in Fig. 7. The SUOX fold proteins from clades 1 and 2 appear to be associated exclusively with YedZ-like or FdnI-like redox partners. Among these, the majority are members of the YedZ family. Also, there are only seven proteins without a predicted redox partner. Clade 3 shows a large diversity of potential redox partners, with no apparent trend. All of the members of clade 4 contain both redox systems, having an exclusive cytochrome c family (family 4) and containing YedZ-like or FdnI-like proteins. Finally, clade 6 also contains an exclusive cytochrome c family (family 5). Of the clades predicted to have cytochrome redox partners, clade 6 contains the lowest preponderance of YedZ-like or FdnI-like proteins, where almost 50% of the members do not encode such a protein. Although YedZ homologs or FdnI homologs appear in five of the eight clades of our phylogeny, it is interesting that they are never found in archaea or eukaryotes.

Overall, the presence of SUOX-associated cytochromes correlates with their localizations (Table 4). The SUOX fold proteins predicted to be cytoplasmically located do not appear to have these redox partners and presumably instead rely on cytoplasmic electron donors such as members of the various classes of soluble ferredoxins (59). This is clearly the case with the proteins represented in clades 7 and 8 (Fig. 7). In general, the various classes of cytochromes c play an essentially ubiquitous role in periplasmic electron transfer in diderm species and in the mitochondrial intermembrane space. The presence of a YedZ-type or FdnI-type membrane-bound cytochrome b correlates with proteins represented in the tat-targeted clades shown in Fig. 7. This supports the proposed role of these proteins in coupling SUOX fold-type proteins to the versatile electron sink of the membrane-bound quinone pool. It is also possible that some of the c-type cytochromes represented in Table 4 may shuttle electrons between the SUOX fold proteins or complexes and YedZ or FdnI-type membrane-bound cytochromes b. These cytochromes c may also couple to other membrane-bound redox complexes containing distinct hemes c, hemes b, and [Fe-S] clusters. Consideration of such systems is beyond the scope of this work (for reviews, see references 2, 71, 74, and 75).

EVOLUTIONARY PATHWAY OF THE SUOX FOLD PROTEINS

Our observation of a large number of SUOX fold proteins across a broad taxonomic diversity presents an opportunity to analyze sequence motif conservation with the aim of presenting an evolutionary pathway encompassing the species from which our sequence data were obtained. As described below, these analyses bear interesting comparisons with the theoretical evolutionary tree proposed by Gupta (27, 28) as well as with evolutionary insights provided by other studies (26, 48, 51, 83).

Motifs

Each of the clades presented in our phylogeny contained at least two relatively conserved motifs (Table 2). In addition, all of the clades, with the exception of clade 3, contain unique sequence motifs. Because of their lack of persistence throughout archeal, bacterial, and eukaryotic species, sequence conservation within the various heme-containing domains or subunits does not provide insights into the SUOX fold class. A similar conclusion can be made about the NADH binding flavoprotein domain of the plant-type NIA enzymes.

In the SUOX fold proteins, the motifs CVExW, CxGNxR, and GxPxR give insight into regions of great importance to the function and structure of the enzyme. As stated above (see Protein-PPT Interactions), the first two motifs, CVExW and CxGNxR, coordinate the Mo, whereas the GxPxR motif aids in coordinating the phosphate of the PPT. The CVExW and CxGNxR motifs are of great functional importance and are found in seven of the eight clades. The Cys included in these motifs has been identified as being critical for sulfite oxidation because of its role in Mo coordination. These active-site motifs imply two separate evolutionary pathways for the SUOX enzymes, which will be elaborated on below (see “Domains”). The only proteins that do not contain either of these motifs are those found in clade 6, which possess an FxECxxN motif instead. Additional information on active-site evolution can be found in the GxPxR motif. The presence of the GxPxR motif throughout all clades of our phylogeny indicates that while this motif may appear to have a only single interaction with the PPT, the U-shaped turn that it elicits may prove to be structurally important to the active site of this enzyme as well.

The dimerization domain is also found to contain conserved motifs. Of the five structures covered in this review, all contain a dimerization domain, with the exception of YedY. Also, as listed in Table 2, clade 3 does not contain a variety of conserved motifs. Taking this information together, the proteins of known structures represented in clade 5 were prime candidates for analyzing conserved motifs in the dimerization domain. Using this approach, we identified that the RV(E/D)(V/L)S motif was found in the dimerization domain. Along with this conserved motif, there is an abundance of conserved Trp residues in the dimerization domain. We extrapolated this information to clades that do not contain proteins of known structure. We determined the dimerization domain to be an area with a large number of conserved Trp residues found toward the C terminus of the protein. This approach has also identified the GLAWSGxG and SRxxDxTGY motifs, of clade 6 members, as being found in the dimerization domain. In the clades containing proteins without the dimerization domain, conserved motifs exist on the surface of the protein (i.e., NGY in YedY). While YedZ is the proposed redox partner of YedY, this surface motif may allow YedY to interact with other proteins, including additional redox partners.

Domains

Based on the assumption that evolution usually proceeds from less complicated to more complicated architectures, inspection of Fig. 7 and Table 2 suggests that the original SUOX ancestor arose as a monomeric Mo-PPT-containing protein of a species lacking an outer membrane that is represented in clade 7 or clade 8.

The first major event to occur following the emergence of the SUOX fold domain is the addition of a dimerization domain to its C terminus. This represents a fundamental split in the evolution of SUOX fold-containing proteins because, as will be seen below, the SUOX fold domain continues separately through bacterial evolution and is retained as the YedY protein in a large number of proteobacterial species. The first proteins to contain a dimerization domain are in the Actinobacteria species represented in clade 3 and in clade 8. It is notable that the actinobacterial and firmicute sequences represented in clade 7 lack this domain. This pathway would ultimately lead to the inclusion of a heme-binding domain as well as the FAD domain (Fig. 13). In the eukaryotic enzymes (clade 5), there appears to be a significant amount of diversity in the presence and sequence of domains. Some eukaryotic organisms, such as fungi and amoebae, contain only the SUOX fold domain and the dimerization domain. The plant SUOX enzyme shown in Fig. 1 is also included in this group, which finds its place in Fig. 13 by the branch labeled as “no additional domains.” The domain shuffling and omissions observed in eukaryotes is not surprising considering that in prokaryotic systems, heme-containing redox partners are found in operons both upstream and downstream of the SUOX protein. As such, a gene fusion of a heme-containing domain to either end is possible.

FIG. 13. — Potential evolution of the domains found within the SUOX fold protein family. Evolution moves from the simplest form (bottom) to the most complex (top). The dotted outline encompasses a variety of enzymes found in clade 5. The light gray model serves as a hypothetical intermediate between known enzymes. Domains are labeled where appropriate, and lined patterns depict the respective domains in the absence of a label. Clade numbers to the side of the domains indicate the clades where these domain combinations are found. The detached heme b/c portion indicates a separate gene upstream/downstream of the SUOX fold protein that acts as an interacting subunit.

Following the path of those SUOX fold proteins lacking the dimerization domain, bacterial species that evolved later and are found in clade 7 also synthesize a protein that lacks a dimerization domain. This protein most likely is the progenitor to the proteins found in clades 1 and 2. As clades 1 and 2 are primarily proteobacterial, they evolved later than did half of the species found in clade 7. Additional evidence is found in the motif surrounding the Cys residue that coordinates the Mo in those proteins lacking the dimerization domain. Clade 7 contains the CVTxWS motif that is similar to the CVExW and CVEGW motifs of clades 1 and 2, respectively. It is also notable that while members of clades 1 and 2 are predicted to be exclusively periplasmic by virtue of their tat leaders, those of clade 7 are predicted to be cytoplasmically localized.

The evolution of the eukaryotic CxGNxR motif is likely the result of a split occurring in clade 8, where the CVSN Mo-coordinating motif is observed. This clade has the dimerization domain but lacks a tat leader. It is likely that the CxGNxR motif evolved from the clade 8 CVSN motif, and it persists in clades 5 (CAGNRR), 4 (CSGNGR), and 3 (CxGNxR), ultimately occurring in the eukaryotic SUOX fold enzymes of clade 5.

Clade 6 appears to have a unique Mo-coordinating motif, FxECxxN, that can be considered to be weakly related to the CAGNRR motif in the sense that there are two residues between the Cys and the Asn. Like clades 1 and 2, this clade is dominated by proteobacterial species, but unlike them, it possesses the dimerization domain.

Correlation with the Evolution of Life

The evolution of the SUOX fold family members can be delineated by taking the appearance of distinct domains and motifs and correlating these with current research regarding the evolution of the three major domains of life (27, 28, 51, 83). Figure 14 shows a potential evolutionary path of the SUOX fold proteins overlaid with a potential evolution of life. In Fig. 14, the two separate evolutionary paths are labeled by the presence of Cys-containing active-site motifs. As the members of clade 6 show a faint resemblance to the CxGNxR motif, they have been included in this evolutionary branch.

FIG. 14. — Evolution of the SUOX family over the evolution of life. The evolutionary path of the SUOX family is overlaid onto a theoretical evolutionary tree of life. Each path is labeled by a conserved active-site motif. The boxes at the end of each branch indicate the clade members (C) that belong to that evolutionary branch as well as the domains present in the SUOX fold proteins (D). Numbers in parentheses listed beside each phylum represent the numbers of genomes solved compared to the numbers of times that the phylum appears in all clades (i.e., number of genomes solved/number of entries in clades).

Approximately half of the SUOX fold proteins observed in our study belonged to the Proteobacteria. This observation implies two different conclusions. The first is that these systems are heavily weighted toward the Proteobacteria. The second is that the completed genomes for bacterial species favors the Proteobacteria. If more sequences of proteins for Proteobacteria species are known, there is a higher chance of these sequences being included in sequence data sets. We analyzed the Genomes Online Database (http://www.genomesonline.org) and determined that there are far more completed Proteobacteria genomes than those of other prokaryotic groups, comprising 277 of the 523 genome sequences. Using our unfiltered list of 559 SUOX fold protein sequences, we compared the number of entries to the number of completed genomes (data not shown). The results show that the apparent favoring of the Proteobacteria is created by their larger number of complete genomes. This effect is present in the Proteobacteria (168 entries over 278 completed genomes), Actinobacteria (31/44), Bacteriodetes (4/11), Chloroflexi (4/4), and Aquificae (1/1), etc. The only exception is the phylum Firmicutes, which accounts for only four entries over 122 complete genomes. This information is conveyed in Fig. 14, where the number of completed genomes is found above the respective branches.

Our observation of the relative lack of SUOX fold enzymes in the Firmicutes has significant evolutionary implications. Many current theories of evolution place the monoderm bacteria as the first prokaryotes, evolving from the “last universal ancestor” (48). Other theories suggest that diderm cells were the first to evolve, arguing that a bilayered “obcell” was the progenitor of the prokaryotes (9). Our analysis supports the theory that the Archaea evolved from monoderm bacteria. In this model, Firmicutes species would contain the original ancestor of the SUOX fold enzyme family. This may explain our observation of only two firmicute species being represented in Fig. 7 despite their large number of sequenced genomes (122 competed genomes). The two species are in clade 3 (Bacillus sp. strain NRRL B-14911) and clade 7 (Bacillus licheniformis). These two entries contain different Cys-containing motifs, where one contains the CxGNxR (clade 3) motif and the other contains the CVTxW motif. In this context, two possibilities exist for SUOX fold evolution. The first is that the Firmicutes species contained a single gene that underwent gene duplication and gave rise to two separate SUOX-like gene products. The second is that two separate genes arose in the Firmicutes, each containing the SUOX fold (convergent evolution). Either of these possibilities can account for the different Cys-containing motifs. In theory, the archaeal enzymes arose from the CVTxW-containing enzyme of the Firmicutes. This CVTxW-containing enzyme would thus be the starting point of the evolutionary branch that contains no additional domains (Fig. 13 and 14). Likewise, the CxGNxR enzyme would be starting point of the evolutionary branch that has the appearance of the dimerization domain. In Fig. 14, we do not suggest that the Firmicutes and Actinobacteria evolved separately from one another, as may be interpreted by these phyla being at the same evolutionary level. While Gupta (29) suggested that the Firmicutes evolved first, a study by Lake et al. (51) suggests that diderms may have evolved from the Firmicutes and that the Archaea may have evolved from the Actinobacteria. Therefore, in Fig. 14, these two phyla have been placed based on the evolution of the SUOX family.

CONCLUSIONS AND OUTLOOK

The explosion of bioinformatic data in recent years has been complemented by an abundance of structural data that enables emerging bioinformatic approaches to be used to gain new insights into SUOX archetypes and closely related proteins. In this review, we used complementary bioinformatic approaches to address the importance and evolutionary persistence of the SUOX fold. These proteins play critical roles in sulfur and nitrogen metabolism and participate in two critical geochemical cycles that are critical for maintaining the integrity and sustainability of the biosphere. We have established that SUOX fold proteins exist in “primordial” archaeal, actinobacterial, or firmicute species and that they persisted throughout evolution to appear in higher animal and plant species.

We have merged structural and sequence bioinformatic data to correlate cladistics analyses with the importance of conserved sequence motifs within eight clades of SUOX fold enzymes. Interestingly, members of many of these clades may have substrates that are not simple inorganic anions such as sulfite or nitrite. It is therefore likely that many SUOX fold enzymes contribute to unanticipated levels of metabolic diversity in a very broad range of bacterial species. Our analyses of sequence conservation should provide interesting targets for site-directed mutagenesis studies that will enable the relationship between protein sequence, Mo-PPT redox chemistry, and substrate specificity to be rigorously explored.

Acknowledgments

We thank Paul Stothard of the Canadian Bioinformatics Help Desk for authorship of a PERL script that removes closely related entries from within a sequence data set. We thank Craig Knox, also of the Canadian Bioinformatics Help Desk, for writing a PERL script that extracts taxonomies from SWISSPROT and/or TREMBL flat-file entries. We also thank Dean Schieve for installing and maintaining the EMBOSS package used during the course of this work.

This work was funded by the Canadian Institutes of Health Research, the Natural Sciences and Engineering Research Council, and the National Institutes of Health. Infrastructure funding was provided by the Canada Foundation for Innovation. J.H.W. holds a Canada Research Chair in Membrane Biochemistry.

REFERENCES

1.Berks, B. C. 1996. A common export pathway for proteins binding complex redox cofactors? Mol. Microbiol. 22393-414. [DOI] [PubMed] [Google Scholar]
2.Berks, B. C., S. J. Ferguson, J. W. B. Moir, and D. J. Richardson. 1995. Enzymes and associated electron transport systems that catalyze the respiratory reduction of nitrogen oxides and oxyanions. Biochim. Biophys. Acta 123297-123. [DOI] [PubMed] [Google Scholar]
3.Berks, B. C., F. Sargent, and T. Palmer. 2000. The Tat protein export pathway. Mol. Microbiol. 35260-274. [DOI] [PubMed] [Google Scholar]
4.Bertero, M. G., R. A. Rothery, N. Boroumand, M. Palak, F. Blasco, N. Ginet, J. H. Weiner, and N. C. Strynadka. 2005. Structural and biochemical characterization of a quinol binding site of Escherichia coli nitrate reductase A. J. Biol. Chem. 28014836-14843. [DOI] [PubMed] [Google Scholar]
5.Bertero, M. G., R. A. Rothery, M. Palak, C. Hou, D. Lim, F. Blasco, J. H. Weiner, and N. C. Strynadka. 2003. Insights into the respiratory electron transfer pathway from the structure of nitrate reductase A. Nat. Struct. Biol. 10681-687. [DOI] [PubMed] [Google Scholar]
6.Bones, A. M., and J. T. Rossiter. 2006. The enzymic and chemically induced decomposition of glucosinolates. Phytochemistry 671053-1067. [DOI] [PubMed] [Google Scholar]
7.Brokx, S. J., R. A. Rothery, G. Zhang, D. P. Ng, and J. H. Weiner. 2005. Characterization of an Escherichia coli sulfite oxidase homologue reveals the role of a conserved active site cysteine in assembly and function. Biochemistry 4410339-10348. [DOI] [PubMed] [Google Scholar]
8.Campbell, W. H. 2001. Structure and function of eukaryotic NAD(P)H:nitrate reductase. Cell. Mol. Life Sci. 58194-204. [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Cavalier-Smith, T. 2001. Obcells as proto-organisms: membrane heredity, lithophosphorylation, and the origins of the genetic code, the first cells, and photosynthesis. J. Mol. Evol. 53555-595. [DOI] [PubMed] [Google Scholar]
10.Chan, M. K., S. Mukund, A. Kletzin, M. W. Adams, and D. C. Rees. 1995. Structure of a hyperthermophilic tungstopterin enzyme, aldehyde ferredoxin oxidoreductase. Science 2671463-1469. [DOI] [PubMed] [Google Scholar]
11.Charles, A. M., and I. Suzuki. 1965. Sulfite oxidase of a facultative autotroph, Thiobacillus novellus. Biochem. Biophys. Res. Commun. 19686-690. [DOI] [PubMed] [Google Scholar]
12.D'Errico, G., A. Di Salle, F. La Cara, M. Rossi, and R. Cannio. 2006. Identification and characterization of a novel bacterial sulfite oxidase with no heme binding domain from Deinococcus radiodurans. J. Bacteriol. 188694-701. [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Di Salle, A., G. D'Errico, F. La Cara, R. Cannio, and M. Rossi. 2006. A novel thermostable sulfite oxidase from Thermus thermophilus: characterization of the enzyme, gene cloning and expression in Escherichia coli. Extremophiles 10587-598. [DOI] [PubMed] [Google Scholar]
14.Doonan, C. J., U. Kappler, and G. N. George. 2006. Structure of the active site of sulfite dehydrogenase from Starkeya novella. Inorg. Chem. 457488-7492. [DOI] [PubMed] [Google Scholar]
15.Doonan, C. J., H. L. Wilson, K. V. Rajagopalan, R. M. Garrett, B. Bennett, R. C. Prince, and G. N. George. 2007. Modified active site coordination in a clinical mutant of sulfite oxidase. J. Am. Chem. Soc. 1299421-9428. [DOI] [PubMed] [Google Scholar]
16.Douglas, P., N. Morrice, and C. MacKintosh. 1995. Identification of a regulatory phosphorylation site in the hinge 1 region of nitrate reductase from spinach (Spinacea oleracea) leaves. FEBS Lett. 377113-117. [DOI] [PubMed] [Google Scholar]
17.Drew, D., D. Sjostrand, J. Nilsson, T. Urbig, C. N. Chin, J. W. de Gier, and G. von Heijne. 2002. Rapid topology mapping of Escherichia coli inner-membrane proteins by prediction and PhoA/GFP fusion analysis. Proc. Natl. Acad. Sci. USA 992690-2695. [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Einsle, O., F. A. Tezcan, S. L. Andrade, B. Schmid, M. Yoshida, J. B. Howard, and D. C. Rees. 2002. Nitrogenase MoFe-protein at 1.16 Å resolution: a central ligand in the FeMo-cofactor. Science 2971696-1700. [DOI] [PubMed] [Google Scholar]
19.Felsenstein, J. 1992. Estimating effective population size from samples of sequences: a bootstrap Monte Carlo integration method. Genet. Res. 60209-220. [DOI] [PubMed] [Google Scholar]
20.Feng, C., G. Tollin, and J. H. Enemark. 2007. Sulfite oxidizing enzymes. Biochim. Biophys. Acta 1774527-539. [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Feng, C., H. L. Wilson, J. K. Hurley, J. T. Hazzard, G. Tollin, K. V. Rajagopalan, and J. H. Enemark. 2003. Essential role of conserved arginine 160 in intramolecular electron transfer in human sulfite oxidase. Biochemistry 4212235-12242. [DOI] [PubMed] [Google Scholar]
22.Fischer, K., G. G. Barbier, H. J. Hecht, R. R. Mendel, W. H. Campbell, and G. Schwarz. 2005. Structural basis of eukaryotic nitrate reduction: crystal structures of the nitrate reductase active site. Plant Cell 171167-1179. [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Garrett, R. M., J. L. Johnson, T. N. Graf, A. Feigenbaum, and K. V. Rajagopalan. 1998. Human sulfite oxidase R160Q: identification of the mutation in a sulfite oxidase-deficient patient and expression and characterization of the mutant enzyme. Proc. Natl. Acad. Sci. USA 956394-6398. [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Glaab, J., and W. M. Kaiser. 1995. Inactivation of nitrate reductase involves NR-protein phosphorylation and subsequent ‘binding’ of an inhibitor protein. Planta 195514-518. [Google Scholar]
25.Gonnet, G. H., M. A. Cohen, and S. A. Benner. 1992. Exhaustive matching of the entire protein sequence database. Science 2561443-1445. [DOI] [PubMed] [Google Scholar]
26.Gupta, R. S. 2001. The branching order and phylogenetic placement of species from completed bacterial genomes, based on conserved indels found in various proteins. Int. Microbiol. 4187-202. [DOI] [PubMed] [Google Scholar]
27.Gupta, R. S. 2005. Molecular sequences and the early history of life, p. 160-183. In J. Sapp (ed.), Microbial phylogeny and evolution: concepts and controversies. Oxford University Press, New York, NY.
28.Gupta, R. S. 2000. The phylogeny of proteobacteria: relationships to other eubacterial phyla and eukaryotes. FEMS Microbiol. Rev. 24367-402. [DOI] [PubMed] [Google Scholar]
29.Gupta, R. S. 1998. What are archaebacteria: life's third domain or monoderm prokaryotes related to gram-positive bacteria? A new proposal for the classification of prokaryotic organisms. Mol. Microbiol. 29695-707. [DOI] [PubMed] [Google Scholar]
30.Hänsch, R., C. Lang, H. Rennenberg, and R. R. Mendel. 2007. Significance of plant sulfite oxidase. Plant Biol. (Stuttgart) 9589-595. [DOI] [PubMed] [Google Scholar]
31.Hänsch, R., C. Lang, E. Riebeseel, R. Lindigkeit, A. Gessler, H. Rennenberg, and R. R. Mendel. 2006. Plant sulfite oxidase as novel producer of H₂O₂: combination of enzyme catalysis with a subsequent non-enzymatic reaction step. J. Biol. Chem. 2816884-6888. [DOI] [PubMed] [Google Scholar]
32.Hänsch, R., and R. R. Mendel. 2005. Sulfite oxidation in plant peroxisomes. Photosynth. Res. 86337-343. [DOI] [PubMed] [Google Scholar]
33.Harris, H. H., G. N. George, and K. V. Rajagopalan. 2006. High-resolution EXAFS of the active site of human sulfite oxidase: comparison with density functional theory and X-ray crystallographic results. Inorg. Chem. 45493-495. [DOI] [PubMed] [Google Scholar]
34.Hille, R. 2002. Molybdenum and tungsten in biology. Trends Biochem. Sci. 27360-367. [DOI] [PubMed] [Google Scholar]
35.Hille, R. 2002. Molybdenum enzymes containing the pyranopterin cofactor: an overview. Met. Ions Biol. Syst. 39187-226. [PubMed] [Google Scholar]
36.Hille, R. 2005. Molybdenum-containing hydroxylases. Arch. Biochem. Biophys. 433107-116. [DOI] [PubMed] [Google Scholar]
37.Hille, R. 1996. The mononuclear molybdenum enzymes. Chem. Rev. 962757-2816. [DOI] [PubMed] [Google Scholar]
38.Johnson, J. L., and M. Duran. 2001. Molybdenum cofactor deficiency and isolated sulfite oxidase deficiency, p. 3163-3177. In C. R. Scryver (ed.), The metabolic and molecular bases of inherited disease. McGraw-Hill, New York, NY.
39.Jormakka, M., B. Byrne, and S. Iwata. 2003. Protonmotive force generation by a redox loop mechanism. FEBS Lett. 54525-30. [DOI] [PubMed] [Google Scholar]
40.Jormakka, M., S. Tornroth, B. Byrne, and S. Iwata. 2002. Molecular basis of proton motive force generation: structure of formate dehydrogenase-N. Science 2951863-1868. [DOI] [PubMed] [Google Scholar]
41.Kappler, U., and S. Bailey. 2005. Molecular basis of intramolecular electron transfer in sulfite-oxidizing enzymes is revealed by high resolution structure of a heterodimeric complex of the catalytic molybdopterin subunit and a c-type cytochrome subunit. J. Biol. Chem. 28024999-25007. [DOI] [PubMed] [Google Scholar]
42.Kappler, U., S. Bailey, C. Feng, M. J. Honeychurch, G. R. Hanson, P. V. Bernhardt, G. Tollin, and J. H. Enemark. 2006. Kinetic and structural evidence for the importance of Tyr236 for the integrity of the Mo active site in a bacterial sulfite dehydrogenase. Biochemistry 459696-9705. [DOI] [PubMed] [Google Scholar]
43.Kappler, U., B. Bennett, J. Rethmeier, G. Schwarz, R. Deutzmann, A. G. McEwan, and C. Dahl. 2000. Sulfite:cytochrome c oxidoreductase from Thiobacillus novellus. Purification, characterization, and molecular biology of a heterodimeric member of the sulfite oxidase family. J. Biol. Chem. 27513202-13212. [DOI] [PubMed] [Google Scholar]
44.Karakas, E., H. L. Wilson, T. N. Graf, S. Xiang, S. Jaramillo-Busquets, K. V. Rajagopalan, and C. Kisker. 2005. Structural insights into sulfite oxidase deficiency. J. Biol. Chem. 28033506-33515. [DOI] [PubMed] [Google Scholar]
45.Kay, C. J., M. J. Barber, B. A. Notton, and L. P. Solomonson. 1989. Oxidation—reduction midpoint potentials of the flavin, haem and Mo-pterin centres in spinach (Spinacia oleracea L.) nitrate reductase. Biochem. J. 263285-287. [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Kisker, C., H. Schindelin, D. Baas, J. Retey, R. U. Meckenstock, and P. M. Kroneck. 1998. A structural comparison of molybdenum cofactor-containing enzymes. FEMS Microbiol. Rev. 22503-521. [DOI] [PubMed] [Google Scholar]
47.Kisker, C., H. Schindelin, A. Pacheco, W. A. Wehbi, R. M. Garrett, K. V. Rajagopalan, J. H. Enemark, and D. C. Rees. 1997. Molecular basis of sulfite oxidase deficiency from the structure of sulfite oxidase. Cell 91973-983. [DOI] [PubMed] [Google Scholar]
48.Koch, A. L. 2003. Were gram-positive rods the first bacteria? Trends Microbiol. 11166-170. [DOI] [PubMed] [Google Scholar]
49.Krissinel, E., and K. Henrick. 2004. Secondary-structure matching (SSM), a new tool for fast protein structure alignment in three dimensions. Acta Crystallogr. D Biol. Crystallogr. 602256-2268. [DOI] [PubMed] [Google Scholar]
50.Krogh, A., B. Larsson, G. von Heijne, and E. L. Sonnhammer. 2001. Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes. J. Mol. Biol. 305567-580. [DOI] [PubMed] [Google Scholar]
51.Lake, J. A., C. W. Herbold, M. C. Rivera, J. A. Servin, and R. G. Skophammer. 2007. Rooting the tree of life using nonubiquitous genes. Mol. Biol. Evol. 24130-136. [DOI] [PubMed] [Google Scholar]
52.Lopez-Huertas, E., F. J. Corpas, L. M. Sandalio, and L. A. Del Rio. 1999. Characterization of membrane polypeptides from pea leaf peroxisomes involved in superoxide radical generation. Biochem. J. 337531-536. [PMC free article] [PubMed] [Google Scholar]
53.Loschi, L., S. J. Brokx, T. L. Hills, G. Zhang, M. G. Bertero, A. L. Lovering, J. H. Weiner, and N. C. Strynadka. 2004. Structural and biochemical identification of a novel bacterial oxidoreductase. J. Biol. Chem. 27950391-50400. [DOI] [PubMed] [Google Scholar]
54.Lu, G., Y. Lindqvist, G. Schneider, U. Dwivedi, and W. Campbell. 1995. Structural studies on corn nitrate reductase: refined structure of the cytochrome b reductase fragment at 2.5 Å, its ADP complex and an active-site mutant and modeling of the cytochrome b domain. J. Mol. Biol. 248931-948. [DOI] [PubMed] [Google Scholar]
55.MacKintosh, C., and S. E. Meek. 2001. Regulation of plant NR activity by reversible phosphorylation, 14-3-3 proteins and proteolysis. Cell. Mol. Life Sci. 58205-214. [DOI] [PMC free article] [PubMed] [Google Scholar]
56.Marquez, A. J., M. Betti, M. Garcia-Calderon, P. Pal'ove-Balang, P. Diaz, and J. Monza. 2005. Nitrate assimilation in Lotus japonicus. J. Exp. Bot. 561741-1749. [DOI] [PubMed] [Google Scholar]
57.Mendel, R. R. 2007. Biology of the molybdenum cofactor. J. Exp. Bot. 582289-2296. [DOI] [PubMed] [Google Scholar]
58.Mendel, R. R., and F. Bittner. 2006. Cell biology of molybdenum. Biochim. Biophys. Acta 1763621-635. [DOI] [PubMed] [Google Scholar]
59.Meyer, J. 2001. Ferredoxins of the third kind. FEBS Lett. 5091-5. [DOI] [PubMed] [Google Scholar]
60.Moorhead, G., P. Douglas, N. Morrice, M. Scarabel, A. Aitken, and C. MacKintosh. 1996. Phosphorylated nitrate reductase from spinach leaves is inhibited by 14-3-3 proteins and activated by fusicoccin. Curr. Biol. 61104-1113. [DOI] [PubMed] [Google Scholar]
61.Moser, C. C., T. A. Farid, S. E. Chobot, and P. L. Dutton. 2006. Electron tunneling chains of mitochondria. Biochim. Biophys. Acta 17571096-1109. [DOI] [PubMed] [Google Scholar]
62.Moser, C. C., J. M. Keske, K. Warncke, R. S. Farid, and P. L. Dutton. 1992. Nature of biological electron transfer. Nature 355796-802. [DOI] [PubMed] [Google Scholar]
63.Moser, C. C., C. C. Page, and P. L. Dutton. 2006. Darwin at the molecular scale: selection and variance in electron tunnelling proteins including cytochrome c oxidase. Philos. Trans. R. Soc. Lond. B Biol. Sci. 3611295-1305. [DOI] [PMC free article] [PubMed] [Google Scholar]
64.Moser, C. C., C. C. Page, R. Farid, and P. L. Dutton. 1995. Biological electron transfer. J. Bioenerg. Biomem. 27263-274. [DOI] [PubMed] [Google Scholar]
65.Mudd, S. H., F. Irreverre, and L. Laster. 1967. Sulfite oxidase deficiency in man: demonstration of the enzymatic defect. Science 1561599-1602. [DOI] [PubMed] [Google Scholar]
66.Mukund, S., and M. W. W. Adams. 1990. Characterization of a tungsten-iron-sulfur protein exhibiting novel spectroscopic and redox properties from the hypothermophilic archaebacterium Pyrococcus furiosus. J. Biol. Chem. 26511508-11516. [PubMed] [Google Scholar]
67.Pacheco, A., J. T. Hazzard, G. Tollin, and J. H. Enemark. 1999. The pH dependence of intramolecular electron transfer rates in sulfite oxidase at high and low anion concentrations. J. Biol. Inorg. Chem. 4390-401. [DOI] [PubMed] [Google Scholar]
68.Page, C. C., C. C. Moser, and P. L. Dutton. 2003. Mechanism for electron transfer within and between proteins. Curr. Opin. Chem. Biol. 7551-556. [DOI] [PubMed] [Google Scholar]
69.Peters, J. W., K. Fisher, and D. R. Dean. 1995. Nitrogenase structure and function: a biochemical-genetic perspective. Annu. Rev. Microbiol. 49335-366. [DOI] [PubMed] [Google Scholar]
70.Rice, P., I. Longden, and A. Bleasby. 2000. EMBOSS: the European Molecular Biology Open Software Suite. Trends Genet. 16276-277. [DOI] [PubMed] [Google Scholar]
71.Richardson, D. J. 2000. Bacterial respiration: a flexible process for a changing environment. Microbiology 146551-571. [DOI] [PubMed] [Google Scholar]
72.Richardson, D. J. 2001. Introduction: nitrate reduction and the nitrogen cycle. Cell. Mol. Life Sci. 58163-164. [DOI] [PMC free article] [PubMed] [Google Scholar]
73.Richardson, D. J., B. C. Berks, D. A. Russell, S. Spiro, and C. J. Taylor. 2001. Functional, biochemical and genetic diversity of prokaryotic nitrate reductases. Cell. Mol. Life Sci. 58165-178. [DOI] [PMC free article] [PubMed] [Google Scholar]
74.Richardson, D. J., and N. J. Watmough. 1999. Inorganic nitrogen metabolism in bacteria. Curr. Opin. Chem. Biol. 3207-219. [DOI] [PubMed] [Google Scholar]
75.Rothery, R. A., G. J. Workun, and J. H. Weiner. 2007. The prokaryotic complex iron-sulfur molybdoenzyme family. Biochim. Biophys. Acta. [Epub ahead of print.] doi: 10.1016/j.bbamem.2007.09.002. [DOI] [PubMed]
76.Saitou, N., and M. Nei. 1987. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4406-425. [DOI] [PubMed] [Google Scholar]
77.Sanchez-Pulido, L., A. M. Rojas, A. Valencia, A. C. Martinez, and M. A. Andrade. 2004. ACRATA: a novel electron transfer domain associated to apoptosis and cancer. BMC Cancer 498. [DOI] [PMC free article] [PubMed] [Google Scholar]
78.Sargent, F. 2007. Constructing the wonders of the bacterial world: biosynthesis of complex enzymes. Microbiology 153633-651. [DOI] [PubMed] [Google Scholar]
79.Sargent, F., B. C. Berks, and T. Palmer. 2002. Assembly of membrane-bound respiratory complexes by the Tat protein-transport system. Arch. Microbiol. 17877-84. [DOI] [PubMed] [Google Scholar]
80.Schindelin, H., C. Kisker, J. Hilton, K. V. Rajagopalan, and D. C. Rees. 1996. Crystal structure of DMSO reductase: redox-linked changes in molybdopterin coordination. Science 2721615-1621. [DOI] [PubMed] [Google Scholar]
81.Schneider, F., J. Löwe, R. Huber, H. Schindelin, C. Kisker, and J. Knäben. 1996. Crystal structure of dimethyl sulfoxide reductase from Rhodobacter capsulatus at 1.88 Å resolution. J. Mol. Biol. 26353-69. [DOI] [PubMed] [Google Scholar]
82.Schrader, N., K. Fischer, K. Theis, R. R. Mendel, G. Schwarz, and C. Kisker. 2003. The crystal structure of plant sulfite oxidase provides insights into sulfite oxidation in plants and animals. Structure 111251-1263. [DOI] [PubMed] [Google Scholar]
83.Skophammer, R. G., C. W. Herbold, M. C. Rivera, J. A. Servin, and J. A. Lake. 2006. Evidence that the root of the tree of life is not within the Archaea. Mol. Biol. Evol. 231648-1651. [DOI] [PubMed] [Google Scholar]
84.Solomonson, L. P., M. J. Barber, W. D. Howard, J. L. Johnson, and K. V. Rajagopalan. 1984. Electron paramagnetic resonance studies on the molybdenum center of assimilatory NADH:nitrate reductase from Chlorella vulgaris. J. Biol. Chem. 259849-853. [PubMed] [Google Scholar]
85.Sonnhammer, E. L. L., G. von Heijne, and A. Krogh. 1998. A hidden Markov model for predicting transmembrane helices in protein sequences. Proc. Int. Intell. Syst. Mol. Biol. 6175-182. [PubMed] [Google Scholar]
86.Spence, J. T., C. A. Kipke, J. H. Enemark, and R. A. Sunde. 1991. Stoichiometry of electron uptake and the effect of anions and pH on the molybdenum and heme reduction potentials of sulfite oxidase. Inorg. Chem. 303011-3015. [Google Scholar]
87.Thompson, J. D., T. J. Gibson, F. Plewniak, F. Jeanmougin, and D. G. Higgins. 1997. The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 254876-4882. [DOI] [PMC free article] [PubMed] [Google Scholar]
88.Thompson, J. D., D. G. Higgins, and T. J. Gibson. 1994. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 224673-4680. [DOI] [PMC free article] [PubMed] [Google Scholar]
89.von Rozycki, T., M. R. Yen, E. E. Lende, and M. H. Saier, Jr. 2004. The YedZ family: possible heme binding proteins that can be fused to transporters and electron carriers. J. Mol. Microbiol. Biotechnol. 8129-140. [DOI] [PubMed] [Google Scholar]
90.Weiner, J. H., P. T. Bilous, G. M. Shaw, S. P. Lubitz, L. Frost, G. H. Thomas, J. A. Cole, and R. J. Turner. 1998. A novel and ubiquitous system for membrane targeting and secretion of cofactor-containing proteins. Cell 9393-101. [DOI] [PubMed] [Google Scholar]
91.Wu, L.-F., B. Ize, A. Chanal, Y. Quentin, and G. Fichant. 2000. Bacterial twin-arginine signal peptide-dependent protein translocation pathway: evolution and mechanism. J. Mol. Microbiol. Biotechnol. 2179-189. [PubMed] [Google Scholar]
92.Yamanaka, T., T. Yoshioka, and K. Kimura. 1981. Purification of sulphite-cytochrome c reductase of Thiobacillus novellus and reconstitution of its sulphite oxidase system with the purified constituents. Plant Cell Physiol. 22613-622. [Google Scholar]
93.Yankovskaya, V., R. Horsefield, S. Törnroth, C. Luna-Chavez, H. Miyoshi, C. Léger, B. Byrne, G. Cecchini, and S. Iwata. 2003. Architecture of succinate dehydrogenase and reactive oxygen species generation. Science 299700-704. [DOI] [PubMed] [Google Scholar]
94.Zhang, Z., L. Huang, V. Shulmeister, Y. Chi, K. K. Kim, L. Hung, A. R. Crofts, E. A. Berry, and S. Kim. 1998. Electron transfer by domain movement in cytochrome bc₁. Nature 392677-684. [DOI] [PubMed] [Google Scholar]

[r1] 1.Berks, B. C. 1996. A common export pathway for proteins binding complex redox cofactors? Mol. Microbiol. 22393-414. [DOI] [PubMed] [Google Scholar]

[r2] 2.Berks, B. C., S. J. Ferguson, J. W. B. Moir, and D. J. Richardson. 1995. Enzymes and associated electron transport systems that catalyze the respiratory reduction of nitrogen oxides and oxyanions. Biochim. Biophys. Acta 123297-123. [DOI] [PubMed] [Google Scholar]

[r3] 3.Berks, B. C., F. Sargent, and T. Palmer. 2000. The Tat protein export pathway. Mol. Microbiol. 35260-274. [DOI] [PubMed] [Google Scholar]

[r4] 4.Bertero, M. G., R. A. Rothery, N. Boroumand, M. Palak, F. Blasco, N. Ginet, J. H. Weiner, and N. C. Strynadka. 2005. Structural and biochemical characterization of a quinol binding site of Escherichia coli nitrate reductase A. J. Biol. Chem. 28014836-14843. [DOI] [PubMed] [Google Scholar]

[r5] 5.Bertero, M. G., R. A. Rothery, M. Palak, C. Hou, D. Lim, F. Blasco, J. H. Weiner, and N. C. Strynadka. 2003. Insights into the respiratory electron transfer pathway from the structure of nitrate reductase A. Nat. Struct. Biol. 10681-687. [DOI] [PubMed] [Google Scholar]

[r6] 6.Bones, A. M., and J. T. Rossiter. 2006. The enzymic and chemically induced decomposition of glucosinolates. Phytochemistry 671053-1067. [DOI] [PubMed] [Google Scholar]

[r7] 7.Brokx, S. J., R. A. Rothery, G. Zhang, D. P. Ng, and J. H. Weiner. 2005. Characterization of an Escherichia coli sulfite oxidase homologue reveals the role of a conserved active site cysteine in assembly and function. Biochemistry 4410339-10348. [DOI] [PubMed] [Google Scholar]

[r8] 8.Campbell, W. H. 2001. Structure and function of eukaryotic NAD(P)H:nitrate reductase. Cell. Mol. Life Sci. 58194-204. [DOI] [PMC free article] [PubMed] [Google Scholar]

[r9] 9.Cavalier-Smith, T. 2001. Obcells as proto-organisms: membrane heredity, lithophosphorylation, and the origins of the genetic code, the first cells, and photosynthesis. J. Mol. Evol. 53555-595. [DOI] [PubMed] [Google Scholar]

[r10] 10.Chan, M. K., S. Mukund, A. Kletzin, M. W. Adams, and D. C. Rees. 1995. Structure of a hyperthermophilic tungstopterin enzyme, aldehyde ferredoxin oxidoreductase. Science 2671463-1469. [DOI] [PubMed] [Google Scholar]

[r11] 11.Charles, A. M., and I. Suzuki. 1965. Sulfite oxidase of a facultative autotroph, Thiobacillus novellus. Biochem. Biophys. Res. Commun. 19686-690. [DOI] [PubMed] [Google Scholar]

[r12] 12.D'Errico, G., A. Di Salle, F. La Cara, M. Rossi, and R. Cannio. 2006. Identification and characterization of a novel bacterial sulfite oxidase with no heme binding domain from Deinococcus radiodurans. J. Bacteriol. 188694-701. [DOI] [PMC free article] [PubMed] [Google Scholar]

[r13] 13.Di Salle, A., G. D'Errico, F. La Cara, R. Cannio, and M. Rossi. 2006. A novel thermostable sulfite oxidase from Thermus thermophilus: characterization of the enzyme, gene cloning and expression in Escherichia coli. Extremophiles 10587-598. [DOI] [PubMed] [Google Scholar]

[r14] 14.Doonan, C. J., U. Kappler, and G. N. George. 2006. Structure of the active site of sulfite dehydrogenase from Starkeya novella. Inorg. Chem. 457488-7492. [DOI] [PubMed] [Google Scholar]

[r15] 15.Doonan, C. J., H. L. Wilson, K. V. Rajagopalan, R. M. Garrett, B. Bennett, R. C. Prince, and G. N. George. 2007. Modified active site coordination in a clinical mutant of sulfite oxidase. J. Am. Chem. Soc. 1299421-9428. [DOI] [PubMed] [Google Scholar]

[r16] 16.Douglas, P., N. Morrice, and C. MacKintosh. 1995. Identification of a regulatory phosphorylation site in the hinge 1 region of nitrate reductase from spinach (Spinacea oleracea) leaves. FEBS Lett. 377113-117. [DOI] [PubMed] [Google Scholar]

[r17] 17.Drew, D., D. Sjostrand, J. Nilsson, T. Urbig, C. N. Chin, J. W. de Gier, and G. von Heijne. 2002. Rapid topology mapping of Escherichia coli inner-membrane proteins by prediction and PhoA/GFP fusion analysis. Proc. Natl. Acad. Sci. USA 992690-2695. [DOI] [PMC free article] [PubMed] [Google Scholar]

[r18] 18.Einsle, O., F. A. Tezcan, S. L. Andrade, B. Schmid, M. Yoshida, J. B. Howard, and D. C. Rees. 2002. Nitrogenase MoFe-protein at 1.16 Å resolution: a central ligand in the FeMo-cofactor. Science 2971696-1700. [DOI] [PubMed] [Google Scholar]

[r19] 19.Felsenstein, J. 1992. Estimating effective population size from samples of sequences: a bootstrap Monte Carlo integration method. Genet. Res. 60209-220. [DOI] [PubMed] [Google Scholar]

[r20] 20.Feng, C., G. Tollin, and J. H. Enemark. 2007. Sulfite oxidizing enzymes. Biochim. Biophys. Acta 1774527-539. [DOI] [PMC free article] [PubMed] [Google Scholar]

[r21] 21.Feng, C., H. L. Wilson, J. K. Hurley, J. T. Hazzard, G. Tollin, K. V. Rajagopalan, and J. H. Enemark. 2003. Essential role of conserved arginine 160 in intramolecular electron transfer in human sulfite oxidase. Biochemistry 4212235-12242. [DOI] [PubMed] [Google Scholar]

[r22] 22.Fischer, K., G. G. Barbier, H. J. Hecht, R. R. Mendel, W. H. Campbell, and G. Schwarz. 2005. Structural basis of eukaryotic nitrate reduction: crystal structures of the nitrate reductase active site. Plant Cell 171167-1179. [DOI] [PMC free article] [PubMed] [Google Scholar]

[r23] 23.Garrett, R. M., J. L. Johnson, T. N. Graf, A. Feigenbaum, and K. V. Rajagopalan. 1998. Human sulfite oxidase R160Q: identification of the mutation in a sulfite oxidase-deficient patient and expression and characterization of the mutant enzyme. Proc. Natl. Acad. Sci. USA 956394-6398. [DOI] [PMC free article] [PubMed] [Google Scholar]

[r24] 24.Glaab, J., and W. M. Kaiser. 1995. Inactivation of nitrate reductase involves NR-protein phosphorylation and subsequent ‘binding’ of an inhibitor protein. Planta 195514-518. [Google Scholar]

[r25] 25.Gonnet, G. H., M. A. Cohen, and S. A. Benner. 1992. Exhaustive matching of the entire protein sequence database. Science 2561443-1445. [DOI] [PubMed] [Google Scholar]

[r26] 26.Gupta, R. S. 2001. The branching order and phylogenetic placement of species from completed bacterial genomes, based on conserved indels found in various proteins. Int. Microbiol. 4187-202. [DOI] [PubMed] [Google Scholar]

[r27] 27.Gupta, R. S. 2005. Molecular sequences and the early history of life, p. 160-183. In J. Sapp (ed.), Microbial phylogeny and evolution: concepts and controversies. Oxford University Press, New York, NY.

[r28] 28.Gupta, R. S. 2000. The phylogeny of proteobacteria: relationships to other eubacterial phyla and eukaryotes. FEMS Microbiol. Rev. 24367-402. [DOI] [PubMed] [Google Scholar]

[r29] 29.Gupta, R. S. 1998. What are archaebacteria: life's third domain or monoderm prokaryotes related to gram-positive bacteria? A new proposal for the classification of prokaryotic organisms. Mol. Microbiol. 29695-707. [DOI] [PubMed] [Google Scholar]

[r30] 30.Hänsch, R., C. Lang, H. Rennenberg, and R. R. Mendel. 2007. Significance of plant sulfite oxidase. Plant Biol. (Stuttgart) 9589-595. [DOI] [PubMed] [Google Scholar]

[r31] 31.Hänsch, R., C. Lang, E. Riebeseel, R. Lindigkeit, A. Gessler, H. Rennenberg, and R. R. Mendel. 2006. Plant sulfite oxidase as novel producer of H₂O₂: combination of enzyme catalysis with a subsequent non-enzymatic reaction step. J. Biol. Chem. 2816884-6888. [DOI] [PubMed] [Google Scholar]

[r32] 32.Hänsch, R., and R. R. Mendel. 2005. Sulfite oxidation in plant peroxisomes. Photosynth. Res. 86337-343. [DOI] [PubMed] [Google Scholar]

[r33] 33.Harris, H. H., G. N. George, and K. V. Rajagopalan. 2006. High-resolution EXAFS of the active site of human sulfite oxidase: comparison with density functional theory and X-ray crystallographic results. Inorg. Chem. 45493-495. [DOI] [PubMed] [Google Scholar]

[r34] 34.Hille, R. 2002. Molybdenum and tungsten in biology. Trends Biochem. Sci. 27360-367. [DOI] [PubMed] [Google Scholar]

[r35] 35.Hille, R. 2002. Molybdenum enzymes containing the pyranopterin cofactor: an overview. Met. Ions Biol. Syst. 39187-226. [PubMed] [Google Scholar]

[r36] 36.Hille, R. 2005. Molybdenum-containing hydroxylases. Arch. Biochem. Biophys. 433107-116. [DOI] [PubMed] [Google Scholar]

[r37] 37.Hille, R. 1996. The mononuclear molybdenum enzymes. Chem. Rev. 962757-2816. [DOI] [PubMed] [Google Scholar]

[r38] 38.Johnson, J. L., and M. Duran. 2001. Molybdenum cofactor deficiency and isolated sulfite oxidase deficiency, p. 3163-3177. In C. R. Scryver (ed.), The metabolic and molecular bases of inherited disease. McGraw-Hill, New York, NY.

[r39] 39.Jormakka, M., B. Byrne, and S. Iwata. 2003. Protonmotive force generation by a redox loop mechanism. FEBS Lett. 54525-30. [DOI] [PubMed] [Google Scholar]

[r40] 40.Jormakka, M., S. Tornroth, B. Byrne, and S. Iwata. 2002. Molecular basis of proton motive force generation: structure of formate dehydrogenase-N. Science 2951863-1868. [DOI] [PubMed] [Google Scholar]

[r41] 41.Kappler, U., and S. Bailey. 2005. Molecular basis of intramolecular electron transfer in sulfite-oxidizing enzymes is revealed by high resolution structure of a heterodimeric complex of the catalytic molybdopterin subunit and a c-type cytochrome subunit. J. Biol. Chem. 28024999-25007. [DOI] [PubMed] [Google Scholar]

[r42] 42.Kappler, U., S. Bailey, C. Feng, M. J. Honeychurch, G. R. Hanson, P. V. Bernhardt, G. Tollin, and J. H. Enemark. 2006. Kinetic and structural evidence for the importance of Tyr236 for the integrity of the Mo active site in a bacterial sulfite dehydrogenase. Biochemistry 459696-9705. [DOI] [PubMed] [Google Scholar]

[r43] 43.Kappler, U., B. Bennett, J. Rethmeier, G. Schwarz, R. Deutzmann, A. G. McEwan, and C. Dahl. 2000. Sulfite:cytochrome c oxidoreductase from Thiobacillus novellus. Purification, characterization, and molecular biology of a heterodimeric member of the sulfite oxidase family. J. Biol. Chem. 27513202-13212. [DOI] [PubMed] [Google Scholar]

[r44] 44.Karakas, E., H. L. Wilson, T. N. Graf, S. Xiang, S. Jaramillo-Busquets, K. V. Rajagopalan, and C. Kisker. 2005. Structural insights into sulfite oxidase deficiency. J. Biol. Chem. 28033506-33515. [DOI] [PubMed] [Google Scholar]

[r45] 45.Kay, C. J., M. J. Barber, B. A. Notton, and L. P. Solomonson. 1989. Oxidation—reduction midpoint potentials of the flavin, haem and Mo-pterin centres in spinach (Spinacia oleracea L.) nitrate reductase. Biochem. J. 263285-287. [DOI] [PMC free article] [PubMed] [Google Scholar]

[r46] 46.Kisker, C., H. Schindelin, D. Baas, J. Retey, R. U. Meckenstock, and P. M. Kroneck. 1998. A structural comparison of molybdenum cofactor-containing enzymes. FEMS Microbiol. Rev. 22503-521. [DOI] [PubMed] [Google Scholar]

[r47] 47.Kisker, C., H. Schindelin, A. Pacheco, W. A. Wehbi, R. M. Garrett, K. V. Rajagopalan, J. H. Enemark, and D. C. Rees. 1997. Molecular basis of sulfite oxidase deficiency from the structure of sulfite oxidase. Cell 91973-983. [DOI] [PubMed] [Google Scholar]

[r48] 48.Koch, A. L. 2003. Were gram-positive rods the first bacteria? Trends Microbiol. 11166-170. [DOI] [PubMed] [Google Scholar]

[r49] 49.Krissinel, E., and K. Henrick. 2004. Secondary-structure matching (SSM), a new tool for fast protein structure alignment in three dimensions. Acta Crystallogr. D Biol. Crystallogr. 602256-2268. [DOI] [PubMed] [Google Scholar]

[r50] 50.Krogh, A., B. Larsson, G. von Heijne, and E. L. Sonnhammer. 2001. Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes. J. Mol. Biol. 305567-580. [DOI] [PubMed] [Google Scholar]

[r51] 51.Lake, J. A., C. W. Herbold, M. C. Rivera, J. A. Servin, and R. G. Skophammer. 2007. Rooting the tree of life using nonubiquitous genes. Mol. Biol. Evol. 24130-136. [DOI] [PubMed] [Google Scholar]

[r52] 52.Lopez-Huertas, E., F. J. Corpas, L. M. Sandalio, and L. A. Del Rio. 1999. Characterization of membrane polypeptides from pea leaf peroxisomes involved in superoxide radical generation. Biochem. J. 337531-536. [PMC free article] [PubMed] [Google Scholar]

[r53] 53.Loschi, L., S. J. Brokx, T. L. Hills, G. Zhang, M. G. Bertero, A. L. Lovering, J. H. Weiner, and N. C. Strynadka. 2004. Structural and biochemical identification of a novel bacterial oxidoreductase. J. Biol. Chem. 27950391-50400. [DOI] [PubMed] [Google Scholar]

[r54] 54.Lu, G., Y. Lindqvist, G. Schneider, U. Dwivedi, and W. Campbell. 1995. Structural studies on corn nitrate reductase: refined structure of the cytochrome b reductase fragment at 2.5 Å, its ADP complex and an active-site mutant and modeling of the cytochrome b domain. J. Mol. Biol. 248931-948. [DOI] [PubMed] [Google Scholar]

[r55] 55.MacKintosh, C., and S. E. Meek. 2001. Regulation of plant NR activity by reversible phosphorylation, 14-3-3 proteins and proteolysis. Cell. Mol. Life Sci. 58205-214. [DOI] [PMC free article] [PubMed] [Google Scholar]

[r56] 56.Marquez, A. J., M. Betti, M. Garcia-Calderon, P. Pal'ove-Balang, P. Diaz, and J. Monza. 2005. Nitrate assimilation in Lotus japonicus. J. Exp. Bot. 561741-1749. [DOI] [PubMed] [Google Scholar]

[r57] 57.Mendel, R. R. 2007. Biology of the molybdenum cofactor. J. Exp. Bot. 582289-2296. [DOI] [PubMed] [Google Scholar]

[r58] 58.Mendel, R. R., and F. Bittner. 2006. Cell biology of molybdenum. Biochim. Biophys. Acta 1763621-635. [DOI] [PubMed] [Google Scholar]

[r59] 59.Meyer, J. 2001. Ferredoxins of the third kind. FEBS Lett. 5091-5. [DOI] [PubMed] [Google Scholar]

[r60] 60.Moorhead, G., P. Douglas, N. Morrice, M. Scarabel, A. Aitken, and C. MacKintosh. 1996. Phosphorylated nitrate reductase from spinach leaves is inhibited by 14-3-3 proteins and activated by fusicoccin. Curr. Biol. 61104-1113. [DOI] [PubMed] [Google Scholar]

[r61] 61.Moser, C. C., T. A. Farid, S. E. Chobot, and P. L. Dutton. 2006. Electron tunneling chains of mitochondria. Biochim. Biophys. Acta 17571096-1109. [DOI] [PubMed] [Google Scholar]

[r62] 62.Moser, C. C., J. M. Keske, K. Warncke, R. S. Farid, and P. L. Dutton. 1992. Nature of biological electron transfer. Nature 355796-802. [DOI] [PubMed] [Google Scholar]

[r63] 63.Moser, C. C., C. C. Page, and P. L. Dutton. 2006. Darwin at the molecular scale: selection and variance in electron tunnelling proteins including cytochrome c oxidase. Philos. Trans. R. Soc. Lond. B Biol. Sci. 3611295-1305. [DOI] [PMC free article] [PubMed] [Google Scholar]

[r64] 64.Moser, C. C., C. C. Page, R. Farid, and P. L. Dutton. 1995. Biological electron transfer. J. Bioenerg. Biomem. 27263-274. [DOI] [PubMed] [Google Scholar]

[r65] 65.Mudd, S. H., F. Irreverre, and L. Laster. 1967. Sulfite oxidase deficiency in man: demonstration of the enzymatic defect. Science 1561599-1602. [DOI] [PubMed] [Google Scholar]

[r66] 66.Mukund, S., and M. W. W. Adams. 1990. Characterization of a tungsten-iron-sulfur protein exhibiting novel spectroscopic and redox properties from the hypothermophilic archaebacterium Pyrococcus furiosus. J. Biol. Chem. 26511508-11516. [PubMed] [Google Scholar]

[r67] 67.Pacheco, A., J. T. Hazzard, G. Tollin, and J. H. Enemark. 1999. The pH dependence of intramolecular electron transfer rates in sulfite oxidase at high and low anion concentrations. J. Biol. Inorg. Chem. 4390-401. [DOI] [PubMed] [Google Scholar]

[r68] 68.Page, C. C., C. C. Moser, and P. L. Dutton. 2003. Mechanism for electron transfer within and between proteins. Curr. Opin. Chem. Biol. 7551-556. [DOI] [PubMed] [Google Scholar]

[r69] 69.Peters, J. W., K. Fisher, and D. R. Dean. 1995. Nitrogenase structure and function: a biochemical-genetic perspective. Annu. Rev. Microbiol. 49335-366. [DOI] [PubMed] [Google Scholar]

[r70] 70.Rice, P., I. Longden, and A. Bleasby. 2000. EMBOSS: the European Molecular Biology Open Software Suite. Trends Genet. 16276-277. [DOI] [PubMed] [Google Scholar]

[r71] 71.Richardson, D. J. 2000. Bacterial respiration: a flexible process for a changing environment. Microbiology 146551-571. [DOI] [PubMed] [Google Scholar]

[r72] 72.Richardson, D. J. 2001. Introduction: nitrate reduction and the nitrogen cycle. Cell. Mol. Life Sci. 58163-164. [DOI] [PMC free article] [PubMed] [Google Scholar]

[r73] 73.Richardson, D. J., B. C. Berks, D. A. Russell, S. Spiro, and C. J. Taylor. 2001. Functional, biochemical and genetic diversity of prokaryotic nitrate reductases. Cell. Mol. Life Sci. 58165-178. [DOI] [PMC free article] [PubMed] [Google Scholar]

[r74] 74.Richardson, D. J., and N. J. Watmough. 1999. Inorganic nitrogen metabolism in bacteria. Curr. Opin. Chem. Biol. 3207-219. [DOI] [PubMed] [Google Scholar]

[r75] 75.Rothery, R. A., G. J. Workun, and J. H. Weiner. 2007. The prokaryotic complex iron-sulfur molybdoenzyme family. Biochim. Biophys. Acta. [Epub ahead of print.] doi: 10.1016/j.bbamem.2007.09.002. [DOI] [PubMed]

[r76] 76.Saitou, N., and M. Nei. 1987. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4406-425. [DOI] [PubMed] [Google Scholar]

[r77] 77.Sanchez-Pulido, L., A. M. Rojas, A. Valencia, A. C. Martinez, and M. A. Andrade. 2004. ACRATA: a novel electron transfer domain associated to apoptosis and cancer. BMC Cancer 498. [DOI] [PMC free article] [PubMed] [Google Scholar]

[r78] 78.Sargent, F. 2007. Constructing the wonders of the bacterial world: biosynthesis of complex enzymes. Microbiology 153633-651. [DOI] [PubMed] [Google Scholar]

[r79] 79.Sargent, F., B. C. Berks, and T. Palmer. 2002. Assembly of membrane-bound respiratory complexes by the Tat protein-transport system. Arch. Microbiol. 17877-84. [DOI] [PubMed] [Google Scholar]

[r80] 80.Schindelin, H., C. Kisker, J. Hilton, K. V. Rajagopalan, and D. C. Rees. 1996. Crystal structure of DMSO reductase: redox-linked changes in molybdopterin coordination. Science 2721615-1621. [DOI] [PubMed] [Google Scholar]

[r81] 81.Schneider, F., J. Löwe, R. Huber, H. Schindelin, C. Kisker, and J. Knäben. 1996. Crystal structure of dimethyl sulfoxide reductase from Rhodobacter capsulatus at 1.88 Å resolution. J. Mol. Biol. 26353-69. [DOI] [PubMed] [Google Scholar]

[r82] 82.Schrader, N., K. Fischer, K. Theis, R. R. Mendel, G. Schwarz, and C. Kisker. 2003. The crystal structure of plant sulfite oxidase provides insights into sulfite oxidation in plants and animals. Structure 111251-1263. [DOI] [PubMed] [Google Scholar]

[r83] 83.Skophammer, R. G., C. W. Herbold, M. C. Rivera, J. A. Servin, and J. A. Lake. 2006. Evidence that the root of the tree of life is not within the Archaea. Mol. Biol. Evol. 231648-1651. [DOI] [PubMed] [Google Scholar]

[r84] 84.Solomonson, L. P., M. J. Barber, W. D. Howard, J. L. Johnson, and K. V. Rajagopalan. 1984. Electron paramagnetic resonance studies on the molybdenum center of assimilatory NADH:nitrate reductase from Chlorella vulgaris. J. Biol. Chem. 259849-853. [PubMed] [Google Scholar]

[r85] 85.Sonnhammer, E. L. L., G. von Heijne, and A. Krogh. 1998. A hidden Markov model for predicting transmembrane helices in protein sequences. Proc. Int. Intell. Syst. Mol. Biol. 6175-182. [PubMed] [Google Scholar]

[r86] 86.Spence, J. T., C. A. Kipke, J. H. Enemark, and R. A. Sunde. 1991. Stoichiometry of electron uptake and the effect of anions and pH on the molybdenum and heme reduction potentials of sulfite oxidase. Inorg. Chem. 303011-3015. [Google Scholar]

[r87] 87.Thompson, J. D., T. J. Gibson, F. Plewniak, F. Jeanmougin, and D. G. Higgins. 1997. The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 254876-4882. [DOI] [PMC free article] [PubMed] [Google Scholar]

[r88] 88.Thompson, J. D., D. G. Higgins, and T. J. Gibson. 1994. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 224673-4680. [DOI] [PMC free article] [PubMed] [Google Scholar]

[r89] 89.von Rozycki, T., M. R. Yen, E. E. Lende, and M. H. Saier, Jr. 2004. The YedZ family: possible heme binding proteins that can be fused to transporters and electron carriers. J. Mol. Microbiol. Biotechnol. 8129-140. [DOI] [PubMed] [Google Scholar]

[r90] 90.Weiner, J. H., P. T. Bilous, G. M. Shaw, S. P. Lubitz, L. Frost, G. H. Thomas, J. A. Cole, and R. J. Turner. 1998. A novel and ubiquitous system for membrane targeting and secretion of cofactor-containing proteins. Cell 9393-101. [DOI] [PubMed] [Google Scholar]

[r91] 91.Wu, L.-F., B. Ize, A. Chanal, Y. Quentin, and G. Fichant. 2000. Bacterial twin-arginine signal peptide-dependent protein translocation pathway: evolution and mechanism. J. Mol. Microbiol. Biotechnol. 2179-189. [PubMed] [Google Scholar]

[r92] 92.Yamanaka, T., T. Yoshioka, and K. Kimura. 1981. Purification of sulphite-cytochrome c reductase of Thiobacillus novellus and reconstitution of its sulphite oxidase system with the purified constituents. Plant Cell Physiol. 22613-622. [Google Scholar]

[r93] 93.Yankovskaya, V., R. Horsefield, S. Törnroth, C. Luna-Chavez, H. Miyoshi, C. Léger, B. Byrne, G. Cecchini, and S. Iwata. 2003. Architecture of succinate dehydrogenase and reactive oxygen species generation. Science 299700-704. [DOI] [PubMed] [Google Scholar]

[r94] 94.Zhang, Z., L. Huang, V. Shulmeister, Y. Chi, K. K. Kim, L. Hung, A. R. Crofts, E. A. Berry, and S. Kim. 1998. Electron transfer by domain movement in cytochrome bc₁. Nature 392677-684. [DOI] [PubMed] [Google Scholar]

PERMALINK

Evolutionary Persistence of the Molybdopyranopterin-Containing Sulfite Oxidase Protein Fold

Gregory J Workun

Kamila Moquin

Richard A Rothery

Joel H Weiner

Abstract

INTRODUCTION

REACTIONS CATALYZED BY SUOX FOLD ENZYMES

SUOX FOLD ENZYMES IN HUMAN DISEASE AND GEOCHEMICAL CYCLES

EMERGING STRUCTURES REVEAL MULTIDOMAIN ARCHITECTURES OF SUOX FOLD ENZYMES

SECONDARY STRUCTURE MATCHING SEARCHES REVEAL SIMILARITIES BETWEEN PROTEINS OF LOW SEQUENCE SIMILARITY

TABLE 1.

FIG. 1.

SSM GENERATES AN IMPROVED SEQUENCE ALIGNMENT OF SUOX FOLD ENZYMES

FIG. 2.

FIG. 3.

FIG. 4.

FIG. 5.

ACTIVE-SITE SEQUENCE CONSERVATION SUGGESTS A SIMPLIFIED CLASSIFICATION OF SUOX ENZYMES

FIG. 6.

SUOX FOLD PROTEIN CLADISTICS

FIG. 7.

TABLE 2.

Clade 1: Bona Fide YedY Sequences (YedY-1)

Clade 2: YedY-2 Sequences That Are Closely Related to YedY-1

Clade 3: tat-Exported Periplasmic SUOX

Clade 4: a Second Clade of tat-Exported Periplasmic SUOX

Clade 5: Eukaryotic SUOX and NIA Enzymes

Clade 6: Bacterial Periplasmic SUOX and NIA Enzymes

Clade 7: Cytoplasmic Bacterial SUOX/YedY/NIA

Clade 8: Cytoplasmic SUOX Enzymes from Monoderm Bacteria

PRESENCE AND FUNCTION OF THE DOMAINS OBSERVED IN STRUCTURALLY CHARACTERIZED PROTEINS CONTAINING THE SUOX FOLD

PROTEIN-PPT INTERACTIONS

FIG. 8.

Conserved Motifs in YedY

FIG. 9.

Conserved Motifs in P. angusta NIA and Plant and Chicken SUOX (Clade 5)

FIG. 10.

FIG. 11.

Conserved Motifs in S. novella SUOX

Structural Motifs in Clades of Unknown Structure and Function

DIFFERENTIAL REDOX PARTNERS

FIG. 12.

CORRELATION BETWEEN THE PRESENCE OF A tat LEADER SEQUENCE AND A DIDERM BACTERIAL CELL ENVELOPE MORPHOLOGY

TABLE 3.

CORRELATION BETWEEN THE PRESENCE OF SUOX FOLD PROTEINS AND YedZ

TABLE 4.

DISTRIBUTION OF ALTERNATIVE CYTOCHROME REDOX PARTNERS

PRESENCE OF MULTIPLE REDOX SYSTEMS

EVOLUTIONARY PATHWAY OF THE SUOX FOLD PROTEINS

Motifs

Domains

FIG. 13.

Correlation with the Evolution of Life

FIG. 14.

CONCLUSIONS AND OUTLOOK

Acknowledgments

REFERENCES

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases