Evolution of the soluble nitrate reductase: defining

122

Biochemical Society Transactions (2006) Volume 34, part 1

Evolution of the soluble nitrate reductase: defining the monomeric periplasmic nitrate reductase subgroup B.J.N. Jepson*1 , A. Marietou†, S. Mohan†, J.A. Cole†, C.S. Butler‡ and D.J. Richardson* *Centre for Metalloprotein Spectroscopy and Biology, School of Biological Sciences, University of East Anglia, Norwich NR4 7TJ, U.K., †School of Biosciences, University of Birmingham, Edgbaston, Birmingham B15 2TT, U.K., and ‡Institute for Cell and Molecular Biosciences, Medical School, University of Newcastle upon Tyne, Catherine Cookson Building, Framlington Place, Newcastle upon Tyne NE2 4HH, U.K.

Abstract Bacterial nitrate reductases can be classified into at least three groups according to their localization and function, namely membrane-bound (NAR) or periplasmic (NAP) respiratory and cytoplasmic assimilatory (NAS) enzymes. Monomeric NASs are the simplest of the soluble nitrate reductases, although heterodimeric NASs exist, and a common structural arrangement of NAP is that of a NapAB heterodimer. Using bioinformatic analysis of published genomes, we have identified more representatives of a monomeric class of NAP, which is the evolutionary link between the monomeric NASs and the heterodimeric NAPs. This has further established the monomeric structural clade of NAP. The operons of the monomeric NAP do not contain NapB and suggest that other redox partners are employed by these enzymes, including NapM or NapG predicted proteins. A structural alignment and comparison of the monomeric and heterodimeric NAPs suggests that a difference in surface polarity is related to the interaction of the respective catalytic subunit and redox partner.

Introduction Bacterial nitrate reductases are molybdoenzymes that can catalyse the two-electron reduction of nitrate to nitrite. They can be classified into three groups according to their localization and function, namely membrane-bound (NAR) or periplasmic respiratory (NAP) or cytoplasmic assimilatory (NAS) enzymes [1,2]. Bacterial respiratory membrane-bound nitrate reductases are generally integral membrane protein complexes with the active site located on the cytoplasmic face of the cytoplasmic membrane. The soluble nitrate reductases present in either the periplasm (NAP) or the cytoplasm (NAS) have adapted to their respective environment. Assimilatory nitrate reduction has been described in various bacterial sources [3], including several strains of cyanobacteria [4], Klebsiella oxytoca [5], Bacillus subtilis [6] and Azotobacter vinelandii [7]. In A. vinelandii and cyanobacteria, the enzymes are monomeric cytoplasmic or thylakoid-associated proteins that contain molybdenum, iron and acid-labile sulphide, and use flavodoxin, in the case of A. vinelandii [7], or ferredoxin and flavodoxin, in the case of cyanobacteria [8,9], as the physiological electron donor. The synthesis of assimilatory nitrate reductases in bacteria is insensitive to aerobiosis but inhibited by ammonium, contrasting with both the respiratory nitrate reductase groups [3,4]. In periplasmic niKey words: assimilatory nitrate reductase (NAS), Desulfovibrio desulfuricans, nitrate reductase evolution, periplasmic nitrate reductase (NAP), soluble nitrate reductase. Abbreviations used: Mo-bis-MGD, Mo-bis-molybdopterin guanine dinucleotide; NAP, periplasmic respiratory nitrate reductase; NapA, catalytic subunit of the periplasmic nitrate reductase; NapB, di-c-haem cytochrome redox partner of NapA; NAR, membrane-bound nitrate reductase; NAS, cytoplasmic assimilatory nitrate reductase. To whom correspondence should be addressed (email [email protected]).

1

C 2006

Biochemical Society

trate reductases, electrons from quinol are generally passed through one or two cytochrome c-containing proteins [NapC and NapB (di-c-haem cytochrome redox partner of NapA)] to the catalytic subunit, NapA, that contains a Mo-bis-MGD (Mo-bis-molybdopterin guanine dinucleotide) cofactor and a [4Fe-4S] cluster [10–13]. Structural studies on the periplasmic nitrate reductase (NapA from Desulfovibrio desulfuricans and the NapAB complex from Rhodobacter sphaeroides) and the membrane-bound nitrate reductases (NarGHI from Escherichia coli) have established that these nitrate reductases are quite distinct [12–14]. For example, in the periplasmic enzymes, cysteine provides a thiol ligand to the molybdenum ion, whereas aspartate provides one or two oxygen ligands to the molybdenum ion in the membrane-bound enzymes [11–15]. In evolutionary terms, the periplasmic and assimilatory enzymes are most closely related and this is reflected in similar spectroscopic properties between the two enzymes, although the predicted structural similarity has yet to be established [2,7,16,17].

Evolution of periplasmic nitrate reductases: monomeric or heterodimeric Bioinformatic analyses reveal that the periplasmic nitrate reductase is phylogenetically widespread in proteobacteria [2], but detailed biochemical and spectroscopic studies have been restricted to enzymes from relatively few species. The catalytic subunit (NapA) from periplasmic nitrate reductases from the α-proteobacteria Paracoccus pantotrophus and R. sphaeroides forms a very tight complex with a di-c-haem subunit (NapB). This heterodimeric complex resists separation

The 11th Nitrogen Cycle Meeting 2005

by a range of chromatographic methods that disrupt weak protein–protein interactions [13,18,19] and a redox-dependent K D of 0.5 nM has been determined for the reduced ˚ (1 A ˚ = NapAB complex of the R. sphaeroides [13]. A 3.2 A 0.1 nm) resolution structure of this enzyme suggests that tight complex formation is contributed to by two loops at the Nand C-terminal extremities of NapB that adopt an extended conformation and embrace the NapA subunit [13]. However, ˚ resolution structure another structure of NapA, the 1.9 A from the δ-proteobacterium D. desulfuricans, is that of a monomer [12]. The various NapA structures can be compared to give some insight into the evolution of periplasmic nitrate reductases. The current consensus is that NAP evolved from cytoplasmic eubacterial assimilatory nitrate reductases, which are ferredoxin-, flavodoxin- or NADH-dependent [8,9]. During evolution, these cytoplasmic enzymes picked up a signal peptide and were translocated to the periplasmic compartment. Here, the nitrate reductase would need a new electron donor that could allow it to derive electrons from the respiratory electron transport chain. The monomeric NapA of the δ-proteobacterium D. desulfuricans is currently the closest known relative of the assimilatory nitrate reductase and targeted sequencing of the D. desulfuricans nap gene cluster has confirmed the absence of a napB gene [20]. This adds weight to the subclassification of periplasmic nitrate reductases into monomeric NapA and heterodimeric NapAB members [13]. Our bioinformatic analysis of recently released bacterial genome sequences now allows us to add another three NapAs to the NapB-independent group: Symbiobacterium thermophilum, a non-culturable obligate commensal Gram-positive thermophile [23]; Desulfitobacterium hafniense (http://genome.jgi-psf.org/draft microbes/ desha/desha.home.html), a Gram-positive organism important in environmental dehalogenation [21] and Anaeromyxobacter dehalogenans (http://genome.jgi-psf.org/draft microbes/anade/anade.home.html), a facultatively anaerobic δ-proteobacterium also important in environmental dehalogenation [22] (Figure 1). A feature of our original analysis of the D. desulfuricans nap gene cluster was the presence of a gene encoding a small (∼10 kDa) tetrahaem cytochrome c, called NapM. It is intriguing to note that homologues of this protein can be identified in the D. hafniense and A. dehalogenans nap clusters (note that in the D. hafniense protein, two of the haem-binding motifs are not conserved). The A. dehalogenans napMA genes cluster, and are likely to be operonic with nrfHA genes encoding a tetrahaem membraneanchored protein of the NapC family and a pentahaem cytochrome c nitrite reductase. This is intriguing because the gene designated genetically napC in the D. desulfuricans nap cluster encodes a membrane-anchored tetrahaem cytochrome that biochemically clusters more closely with NrfHs than NapCs. Because the NapAs of D. desulfuricans, S. thermophilum, A. dehalogenans and D. hafniense are predicted to be monomeric and with NapB absent, there is the potential for promiscuity of the redox partner. NapM could be a direct electron

donor when present, but it is also notable that napG is a common feature in the D. desulfuricans, S. thermophilum and D. hafniense nap clusters. This gene encodes a putative 4[4Fe-4S] ferredoxin. 4[4Fe-4S] proteins commonly donate electrons directly to the Mo-bis-MGD subunits of respiratory reductases, such as the DMSO reductase, tetrathionate reductase or membrane-bound nitrate reductases [14,15], and so it is perfectly possible that NapG is the periplasmic electron donor to NapA in D. desulfuricans, S. thermophilum and D. hafniense. However, in contrast with the analogous NapG protein of E. coli, the D. desulfuricans, S. thermophilum and D. hafniense NapGs do not have signal peptides, and so if they are to be electron donors to NapA, they must be co-exported through the TAT (twin-arginine translocation) system with NapA. This is quite common for the 4[4Fe-4S] ferredoxin subunits of other periplasmic oxidoreductases, for example the FdnH subunit of formate dehydrogenase and the SerB subunit of selenate reductase [24,25]. Many nap gene clusters encoding heterodimeric NapAB (i.e. encode napB) have retained napG, including that of E. coli, and dual pathways of electron transfer from quinol to NapA in this organism are known to operate, though it is not clear if NapG can serve as a direct electron donor to NapA [26]. In structural terms, the NapB-dependent NapA can be distinguished from the monomeric NapB-independent NapA from D. desulfuricans, S. thermophilum, H. desufitibacterium and A. dehalogenans on the basis of its smaller size, since the mature proteins of the latter group are approx. 720 rather than 800 amino acids in length, by virtue of the absence of two polypeptide insertions. This makes these proteins similar in size to the simplest cytoplasmic assimilatory nitrate reductases of cyanobacteria which also lack the same insertions. An evolutionary comparison of the heterodimeric and monomeric NapAs together with various genes of the monomeric assimilatory nitrate reductases highlights the distinctive groups of the soluble nitrate reductase family (Figure 1). The tree agrees with the current consensus that the NAPs and NASs evolved from an original, presumably cytoplasmic assimilatory nitrate reductase, most similar to the NAS found in cyanobacteria, typified by the NarB enzyme of Synechococcus. The original NAS enzyme then evolved into different structural forms allowing the utilization of various electron donors and ultimately recruited a signal peptide so that we see the appearance of the periplasmic nitrate reductase. The NapB-independent enzymes from D. desulfuricans, S. thermophilum, D. hafniense and A. dehalogenans clearly form a distinct clade to the NapB-dependent enzymes and are most similar to the NAS enzymes, and so would be predicted to be more ancient than the NAPs that interact with NapB. This would seem to indicate that the NAP enzyme has evolved to initially cross the inner membrane to the positive potential side of the membrane (periplasm in Gram-negative bacteria) and then to interact with a specific redox partner, NapB. Therefore at some point in evolution NapA has recruited NapB as a redox partner, but the primordial NapB must have had a role in another electron transport system(s) at the time C 2006

Biochemical Society

123

124


Figure 1 Evolutionary comparison of Nap (A) Organization of nap gene clusters in various bacteria. (B) Phylogenetic tree compiled using TreeTop prediction program (Topological algorithm) using a multiple sequence alignment of NapA compiled using ClustalW from the following bacterial species: Nas Bs, NAS from B. subtilis; Nas Rc, R. capsulatus; Nas Ko, K. oxytoca; NarB Syn, Synechococcus PCC 7942; NapA Rs, R. sphaeroides; Nap Pp, P. pantotrophus; NapA Ec, E. coli; NapA Dd, D. desulfuricans; NapA Sth, S. thermophilum; Nap Dh, D. hafniense; Nap Ad, A. dehalogenans; Nap Sm, S. meliloti; Nap Cn, Cupriavidus necator; Nap Bb, Bordetella bronchiseptica; Nap Sf, Shewanella frigidimarina; Nap Bj, Bradyrhizobium japonicum; Nap Pa, Pseudomonas aeruginosa; Nap Vv, Vibrio vulnificus; Nap Hi, Haemophilus influenzae; Nap Da, Dechloromoas aromatica.

and so the interaction was initially not specific to NapA. As the NapA and NapB proteins co-evolved, the interaction became stronger such that a tight heterodimeric NapAB complex, typified in the α-proteobacteria R. sphaeroides and P. pantotrophus, was the final outcome. It may then also be C 2006

Biochemical Society

notable that in R. sphaeroides and P. pantotrophus, the napA and napB genes overlap in a way that is indicative of translational coupling to yield a 1:1 ratio of translated polypeptide. The structure of the heterodimeric NapAB of R. sphaeroides reveals a surface Tyr58 that lies between the buried haem 2 of

The 11th Nitrogen Cycle Meeting 2005

Figure 2 Electrostatic surface representations of NAP structures

to create a predicted structural conformation used in the surface potential

Electrostatic surface of the NapB interaction face of the various NAP structures. Electrostatic surfaces were calculated using the program GRASP

calculation. The haems of NapB are shown in green where they interact with NapA of R. sphaeroides.

[27] and displayed in PYMOL (DeLano Scientific). The ‘surface potentials’ of colour property from scale −10 (red, negatively charged) to +10 (blue, positively charged). Co-ordinates for D. desulfuricans NAP can be found under Protein Data Bank code 2NAP. Co-ordinates for R. sphaeroides NapA can be found under Protein Data Bank code 1OGY. The structures of D. desulfuricans and R. sphaeroides NAPs were fitted using R

SwissPdbViewer 3.0 (‘Magic Fit’), and P. pantotrophus (GenBank R accession number Q56350), S. thermophilum (GenBank accession R number YP 074746), A. dehalogenans (GenBank accession num R

ber ZP 00398606), and D. hafniense (GenBank accession number ZP 00558071) protein sequences were threaded on to this alignment

NapB and the [4Fe-4S] cluster of NapA. This residue could play an important role in electron transfer and is absent from D. desulfuricans NapA, consistent with this protein not being NapB-dependent. In the R. sphaeroides NapAB, the interface between the NapA and NapB is largely defined by H-bond interactions between residues that are less conserved in D. desulfuricans, again consistent with the NapB dependence and independence of the two monomeric enzymes. For both the D. desulfuricans and R. sphaeroides NapAs, the Mo-bis-MGD cofactor lies buried at the base of putative substrate and product entry and egress funnel that is lined with charged residues provided by domains II and III (Asp158 , Glu159 , Arg382 and Glu383 ). However, significant differences are apparent in the electrostatic surface potential of the monomeric and heterodimeric NapAs that could influence the relative strength of intermolecular interaction within the NapAB electron transfer complex in the different enzymes. A comparison of these for the NapAs from D. desulfuricans, R. sphaeroides, P. pantotrophus and S. thermophilum, D. hafniense and A. dehalogenans (the latter four constructed using SwissPdbViewer 3.0 ‘Magic Fit’ alignment) has been undertaken. This is illustrated for the surface of NapA that, in the case of the R. sphaeroides and P. pantotrophus enzymes, interacts with NapB (as indicated by the haem motifs in the Figure) and surrounds the iron–sulphur cluster (Figure 2). It is apparent that the surfaces of NapAs from D. desulfuricans, S. thermophilum, D. hafniense and A. dehalogenans exhibit a qualitatively similar electrostatic profile and are more hydrophobic in nature than the equivalent surfaces of R. sphaeroides and P. pantotrophus enzymes, which are more polar and anionic. In fact the predicted surfaces of the NapAs from the Gram-negative δ-proteobacterium representatives, D. desulfuricans and A. dehalogenans, appear to be slightly positive in nature. The difference in the surface polarity may explain the presence of an alternative electron donor in the case of monomeric nitrate reductases in both Grampositive and Gram-negative bacteria respectively, e.g. NapM or NapG. However, the predicted site of binding of NapM and NapG would be at a similar site to that of the NapA–NapB interface due to the availability of the [4Fe4S] cluster and so a comparison of the surfaces is valid. In conclusion, we have identified two new members of the monomeric form of NAP that, together with those identified in D. desulfuricans and S. thermophilum [20], leads to the confirmation of this structural clade of the soluble nitrate reductase. The members of this group are the potential evolutionary link between the monomeric NAS and heterodimeric NAPs. This structural group highlights that other redox partners such as NapM or NapG could have been utilized before the specific interaction with NapB to reduce nitrate in the periplasm. C 2006

Biochemical Society

125

126


This work was supported by grants P13842 and P11528 from the U.K. Biotechnology and Biological Sciences Research Council to D.J.R. and C.S.B. and by EC Programme Framework V contract EVKI-CT200000054 from the European Commission.

References 1 Stolz, J.F. and Basu, P. (2002) ChemBioChem 3, 198–206 2 Richardson, D.J., Berks, B.C., Russell, D.A., Spiro, S. and Taylor, C.J. (2001) Cell. Mol. Life Sci. 58, 165–178 3 Lin, J.T. and Stewart, V. (1998) Adv. Microb. Physiol. 39, 1–30 4 Flores, E. and Herrero, A. (1994) in The Molecular Biology of Cyanobacteria (Bryant, D.A., ed.), pp. 487–517, Kluwer Academic Publishers, Dordrecht, The Netherlands 5 Lin, J.T., Goldman, B.S. and Stewart, V. (1993) J. Bacteriol. 175, 2370–2378 6 Ogawa, K., Akagawa, E., Yamane, K., Sun, Z.W., LaCelle, M., Zuber, P. and Nakano, M.M. (1995) J. Bacteriol. 177, 1409–1413 7 Gangeswaran, R. and Eady, R.R. (1996) Biochem. J. 317, 103–108 8 Mikami, B. and Ida, S. (1984) Biochim. Biophys. Acta 791, 294–304 9 Hirasawa, M., Rubio, L.M., Griffin, J.L., Flores, E., Herrero, A., Li, J., Kim, S.K., Hurley, J.K., Tollin, G. and Knaff, D.B. (2004) Biochim. Biophys. Acta 1608, 155–162 10 Breton, J., Berks, B.C., Reilly, A., Thomson, A.J., Ferguson, S.J. and Richardson, D.J. (1994) FEBS Lett. 345, 76–80 11 Butler, C.S., Charnock, J.M., Bennett, B., Sears, H.J., Reilly, A.J., Ferguson, S.J., Garner, C.D., Lowe, D.J., Thomson, A.J., Berks, B.C. et al. (1999) Biochemistry 38, 9000–9012 12 Dias, J.M., Than, M.E., Humm, A., Huber, R., Bourenkov, G.P., Bartunik, H.D., Bursakov, S., Calvete, J., Caldeira, J., Carneiro, C. et al. (1999) Structure 7, 65–79

C 2006

Biochemical Society

13 Arnoux, P., Sabaty, M., Alric, J., Frangioni, B., Guigliarelli, B., Adriano, J.M. and Pignol, D. (2003) Nat. Struct. Biol. 10, 928–934 14 Jormakka, M., Richardson, D., Byrne, B. and Iwata, S. (2004) Structure 12, 95–104 15 Bertero, M.G., Rothery, R.A., Palak, M., Hou, C., Lim, D., Blasco, F., Weiner, J.H. and Strynadka, N.C. (2003) Nat. Struct. Biol. 10, 681–687 16 Jepson, B.N., Anderson, L.J., Rubio, L.M., Taylor, C.J., Butler, C.S., Flores, E., Herrero, A., Butt, J.N. and Richardson, D.J. (2004) J. Biol. Chem. 31, 32212–32218 17 Butt, J.N., Anderson, L.J., Rubio, L.M., Richardson, D.J., Flores, E. and Herrero, A. (2002) Bioelectrochemistry 56, 17–18 18 Richardson, D.J., McEwan, A.G., Page, M.D., Jackson, J.B. and Ferguson, S.J. (1990) Eur. J. Biochem. 194, 263–270 19 Berks, B.C., Richardson, D.J., Robinson, C., Reilly, A., Aplin, R.T. and Ferguson, S.J. (1994) Eur. J. Biochem. 220, 117–124 20 Marietou, A., Richardson, D., Cole, J. and Mohan, S. (2005) FEMS Microbiol. Lett. 248, 217–225 21 Damborsky, J. (1999) Folia Microbiol. (Praha) 44, 247–262 22 Sanford, R.A., Cole, J.R. and Tiedje, J.M. (2002) Appl. Environ. Microbiol. 68, 893–900 23 Ueda, K., Yamashita, A., Ishikawa, J., Watsuji, T.O., Ikeda, H., Hattori, M. and Beppu, T. (2004) Nucleic Acids Res. 32, 4937–4944 24 Sargent, F., Berks, B.C. and Palmer, T. (2002) Arch. Microbiol. 178, 77–84 25 Schroder, I., Rech, S., Krafft, T. and Macy, J.M. (1997) J. Biol. Chem. 272, 23765–23768 26 Brondijk, T.H., Fiegen, D., Richardson, D.J. and Cole, J.A. (2002) Mol. Microbiol. 44, 245–255 27 Bailey, S. (1994) Acta Crystallogr. D: Biol. Crystallogr. 50, 760–763 Received 23 September 2005