Limits...
Homology modeling of dissimilatory APS reductases (AprBA) of sulfur-oxidizing and sulfate-reducing prokaryotes.

Meyer B, Kuever J - PLoS ONE (2008)

Bottom Line: These structural alterations correlated with the protein phylogeny (three major phylogenetic lineages: (1) SRP including LGT-affected Archaeoglobi and SOB of Apr lineage II, (2) crenarchaeal SRP Caldivirga and Pyrobaculum, and (3) SOB of the distinct Apr lineage I) and the presence of potential APS reductase-interacting redox complexes.The almost identical protein matrices surrounding both [4Fe-4S] clusters, the FAD cofactor, the active site channel and center within the AprB/A models of SRP and SOB point to a highly similar catalytic process of APS reduction/sulfite oxidation independent of the metabolism type the APS reductase is involved in and the species it has been originated from.Based on the comparative models, there are no significant structural differences between dissimilatory APS reductases from SRP and SOB; this might be indicative for a similar catalytic process of APS reduction/sulfite oxidation.

View Article: PubMed Central - PubMed

Affiliation: Max Planck Institute for Marine Microbiology, Bremen, Germany.

ABSTRACT

Background: The dissimilatory adenosine-5'-phosphosulfate (APS) reductase (cofactors flavin adenine dinucleotide, FAD, and two [4Fe-4S] centers) catalyzes the transformation of APS to sulfite and AMP in sulfate-reducing prokaryotes (SRP); in sulfur-oxidizing bacteria (SOB) it has been suggested to operate in the reverse direction. Recently, the three-dimensional structure of the Archaeoglobus fulgidus enzyme has been determined in different catalytically relevant states providing insights into its reaction cycle.

Methodology/principal findings: Full-length AprBA sequences from 20 phylogenetically distinct SRP and SOB species were used for homology modeling. In general, the average accuracy of the calculated models was sufficiently good to allow a structural and functional comparison between the beta- and alpha-subunit structures (78.8-99.3% and 89.5-96.8% of the AprB and AprA main chain atoms, respectively, had root mean square deviations below 1 A with respect to the template structures). Besides their overall conformity, the SRP- and SOB-derived models revealed the existence of individual adaptations at the electron-transferring AprB protein surface presumably resulting from docking to different electron donor/acceptor proteins. These structural alterations correlated with the protein phylogeny (three major phylogenetic lineages: (1) SRP including LGT-affected Archaeoglobi and SOB of Apr lineage II, (2) crenarchaeal SRP Caldivirga and Pyrobaculum, and (3) SOB of the distinct Apr lineage I) and the presence of potential APS reductase-interacting redox complexes. The almost identical protein matrices surrounding both [4Fe-4S] clusters, the FAD cofactor, the active site channel and center within the AprB/A models of SRP and SOB point to a highly similar catalytic process of APS reduction/sulfite oxidation independent of the metabolism type the APS reductase is involved in and the species it has been originated from.

Conclusions: Based on the comparative models, there are no significant structural differences between dissimilatory APS reductases from SRP and SOB; this might be indicative for a similar catalytic process of APS reduction/sulfite oxidation.

Show MeSH

Related in: MedlinePlus

Three-dimensional structure of AprB from A. fulgidus (A, B) and selected, homology modeling-based AprB models from Allochromatium vinosum (C, D), Pelagibacter ubique (E, F), Pyrobaculum calidifontis (G, H), Desulfotomaculum reducens (I, J), Desulfovibrio vulgaris (K, L), Chlorobaculum tepidum (M, N) and Thiobacillus denitrificans (O, P).Protein molecular surface colored by calculated electrostatic potential are shown in panels A, C, E, G, I, K, M, O (electric charge at the molecular surface is colored with a red (negative), white (neutral), and blue (positive) color gradient; electric field extending into the solvent is shown); the differently present, negatively charged loops in the models of SRP and SOB are marked by yellow color (the additional loop of Thiobacillus denitrificans is shown in violet); the electron-transferring Trp-B43/-B48 is marked by yellow color (Pyrobaculum calidifontis: Trp-substituting Ala-B43 is highlighted in G and H). Protein molecular surface colored by calculated solvent accessibility are shown in panels B, D, F, H, J, L, N, P.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC2211403&req=5

pone-0001514-g004: Three-dimensional structure of AprB from A. fulgidus (A, B) and selected, homology modeling-based AprB models from Allochromatium vinosum (C, D), Pelagibacter ubique (E, F), Pyrobaculum calidifontis (G, H), Desulfotomaculum reducens (I, J), Desulfovibrio vulgaris (K, L), Chlorobaculum tepidum (M, N) and Thiobacillus denitrificans (O, P).Protein molecular surface colored by calculated electrostatic potential are shown in panels A, C, E, G, I, K, M, O (electric charge at the molecular surface is colored with a red (negative), white (neutral), and blue (positive) color gradient; electric field extending into the solvent is shown); the differently present, negatively charged loops in the models of SRP and SOB are marked by yellow color (the additional loop of Thiobacillus denitrificans is shown in violet); the electron-transferring Trp-B43/-B48 is marked by yellow color (Pyrobaculum calidifontis: Trp-substituting Ala-B43 is highlighted in G and H). Protein molecular surface colored by calculated solvent accessibility are shown in panels B, D, F, H, J, L, N, P.

Mentions: Besides their overall conformity in the secondary structure element positioning, the first and the second segment of the comparative models showed significant differences in distinct loop regions resulting from insertions and deletions (see Fig. 3 and 4; for details see supplementary data material Table S1 and Figure S1). Interestingly, these structural differences reflected the AprBA-based phylogeny in its separation into three major phylogenetic AprBA clusters: (1) the SRB including the LGT-affected Archaeoglobi and SOB of Apr lineage II, (2) the Caldivirga-Pyrobaculum group of putative crenarchaeal sulfate-reducers, and (3) the SOB of distinct Apr lineage I (see Fig. 2A). Furthermore, the presence/absence of certain loops among the AprB models of each cluster correlated with the absence/presence of the Qmo redox complex and AprM protein encoding genes in the investigated SRP and SOB genomes [16], [17] and might reflect their different functional linkage to the electron transport chain in the membrane. (1) In all comparative AprB models of SRB, A. fulgidus and the SOB of Apr lineage II, there is a flexible loop conservatively located between Cys-B13 and Gly-B19 which comprised predominantly charged amino acids of the following general sequence, K G X D/E K/R (see Fig. 3 and 4; for sequence details see supplementary data material Table S1). This loop is absent in the beta-subunit models of SOB from Apr lineage I and the Caldivirga-Pyrobaculum group (see Fig. 3B–D and supplementary data material Figure S1). Especially the conserved Lys-B14 and Glu/Asp-B17 are located at exposed, solvent-accessible positions in this loop (see Fig. 4 and supplementary data material Figures S1, panels G/H); they might be responsible for the functional docking of the putative redox partner QmoABC/QmoAB(-HdrBC) to the AprB protein surface adjacent to the [4Fe-4S] cluster II. The Qmo complex presumably operates as physiological electron carrier between the membrane-integral quinone/quinol pool and the cytoplasmic APS reductase in the SRP as well as in the sulfite-oxidizing Chlorobiaceae and Beta- and Gammaproteobacteria that harbor a SRB-related enzyme (SOB Apr lineage II) [17]. Thiobacillus denitrificans possessed a second exposed loop of six amino acids positioned between the beta-alpha-beta structure motifs of the ferredoxin-like segment; however, its functional or structural role is not apparent. (2) The crenarchaeal Caldivirga maquilingensis and putative sulfate-reducing Pyrobaculum spp. did not harbor qmo homologous sequences in their genomes [17]; in agreement to the previous proposal, an elongated loop between secondary structure element 2 and 3 was absent in their AprB comparative models (see Fig. 3D and supplementary material Figure S1). The putative physiological electron donor for the dissimilatory APS reductase in these species is unknown; however, naphtoquinones have been described from Pyrobaculum species [41], [42]. The electrostatic potential at the protein surface of the crenarchaeal AprB models differed significantly from the other SRP and SOB which might be an indication that the potential interacting redox partner and the electron transfer process of the aforementioned are different and unrelated to the latter (see Fig. 4 and supplementary data material Figures S1, panels E/F). Interestingly, both Pyrobaculum APS reductases missed the strictly conserved, electron transfer-relevant Trp-B48 that was substituted by an Ala-B43 residue at the corresponding AprB model position (see Fig. 4 panel G for the Pyrobaculum calidifontis AprB model). In all other Apr comparative models, the Trp-B48/43 was located between the S3 of cluster I and the methyl C8M of FAD (distance 12.4 Å) and in van der Waals contact with both redox centers; its indole ring was locked in its position by Thr-A233 and Arg-A232. The electron transfer function of trytophan residues located between two redox centers has been frequently documented [43], [44]; an analogous functional role, however, has not been reported for alanine residues. In consequence, the Pyrobaculum spp. APS reductases might have lost their structural ability for electron transfer between both subunits (essential for APS reduction) and, thus, their enzymes will not be functional anymore (see also the structural comparison of the alpha-subunits). Indeed, no cultivated Pyrobaculum species has been described to be capable of dissimilatory sulfate reduction [45], [46]; notably, the aprBA sequences of the Pyrobaculum aerophilum genome are frame-shifted [47] which might be a result of elevated mutation rate in irrelevant enzymes/metabolic pathways. (3) Like the beta-subunits of the crenarchaeal putative sulfate-reducers, the AprB comparative models of the SOB Apr lineage-I sulfur-oxidizers did not contain an elongated loop between secondary structure element 1 and 2 (see Fig. 3B and C and supplementary material Figure S1). Consistently, the genomes of the respective species did not include qmo homologues; however, their aprBA genes were always preceded and co-transcribed by a membrane-integral protein encoding gene, aprM. Interestingly, all AprB models of the SOB Apr lineage I possessed in the antiparallel beta-sheet segment an elongated loop by amino acid insertion between the secondary structure element 8 and 9 that comprised predominantly negatively charged residues (see Fig. 4 panels C and E; supplementary data material Table S1 and Figure S1, compare panels E/F). The latter were exposed to the solvent (see Fig. 4 panels D and F and supplementary data material Figure S1, compare panels G/H) and might function as an interface region for docking the AprBA enzyme to the AprM protein that anchors the dissimilatory APS reductase to the membrane and enables its physical contact to the yet unknown electron receptor. The presumed differing functional linkage of the cytoplasmic SOB Apr lineage I-type APS reductases to the membrane (not involving Qmo complex homologues) was also reflected in the deviating electrostatic potential at the protein surface when compared to the SRB-type AprB models (supplementary data material Figure S1, compare panels E/F).


Homology modeling of dissimilatory APS reductases (AprBA) of sulfur-oxidizing and sulfate-reducing prokaryotes.

Meyer B, Kuever J - PLoS ONE (2008)

Three-dimensional structure of AprB from A. fulgidus (A, B) and selected, homology modeling-based AprB models from Allochromatium vinosum (C, D), Pelagibacter ubique (E, F), Pyrobaculum calidifontis (G, H), Desulfotomaculum reducens (I, J), Desulfovibrio vulgaris (K, L), Chlorobaculum tepidum (M, N) and Thiobacillus denitrificans (O, P).Protein molecular surface colored by calculated electrostatic potential are shown in panels A, C, E, G, I, K, M, O (electric charge at the molecular surface is colored with a red (negative), white (neutral), and blue (positive) color gradient; electric field extending into the solvent is shown); the differently present, negatively charged loops in the models of SRP and SOB are marked by yellow color (the additional loop of Thiobacillus denitrificans is shown in violet); the electron-transferring Trp-B43/-B48 is marked by yellow color (Pyrobaculum calidifontis: Trp-substituting Ala-B43 is highlighted in G and H). Protein molecular surface colored by calculated solvent accessibility are shown in panels B, D, F, H, J, L, N, P.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC2211403&req=5

pone-0001514-g004: Three-dimensional structure of AprB from A. fulgidus (A, B) and selected, homology modeling-based AprB models from Allochromatium vinosum (C, D), Pelagibacter ubique (E, F), Pyrobaculum calidifontis (G, H), Desulfotomaculum reducens (I, J), Desulfovibrio vulgaris (K, L), Chlorobaculum tepidum (M, N) and Thiobacillus denitrificans (O, P).Protein molecular surface colored by calculated electrostatic potential are shown in panels A, C, E, G, I, K, M, O (electric charge at the molecular surface is colored with a red (negative), white (neutral), and blue (positive) color gradient; electric field extending into the solvent is shown); the differently present, negatively charged loops in the models of SRP and SOB are marked by yellow color (the additional loop of Thiobacillus denitrificans is shown in violet); the electron-transferring Trp-B43/-B48 is marked by yellow color (Pyrobaculum calidifontis: Trp-substituting Ala-B43 is highlighted in G and H). Protein molecular surface colored by calculated solvent accessibility are shown in panels B, D, F, H, J, L, N, P.
Mentions: Besides their overall conformity in the secondary structure element positioning, the first and the second segment of the comparative models showed significant differences in distinct loop regions resulting from insertions and deletions (see Fig. 3 and 4; for details see supplementary data material Table S1 and Figure S1). Interestingly, these structural differences reflected the AprBA-based phylogeny in its separation into three major phylogenetic AprBA clusters: (1) the SRB including the LGT-affected Archaeoglobi and SOB of Apr lineage II, (2) the Caldivirga-Pyrobaculum group of putative crenarchaeal sulfate-reducers, and (3) the SOB of distinct Apr lineage I (see Fig. 2A). Furthermore, the presence/absence of certain loops among the AprB models of each cluster correlated with the absence/presence of the Qmo redox complex and AprM protein encoding genes in the investigated SRP and SOB genomes [16], [17] and might reflect their different functional linkage to the electron transport chain in the membrane. (1) In all comparative AprB models of SRB, A. fulgidus and the SOB of Apr lineage II, there is a flexible loop conservatively located between Cys-B13 and Gly-B19 which comprised predominantly charged amino acids of the following general sequence, K G X D/E K/R (see Fig. 3 and 4; for sequence details see supplementary data material Table S1). This loop is absent in the beta-subunit models of SOB from Apr lineage I and the Caldivirga-Pyrobaculum group (see Fig. 3B–D and supplementary data material Figure S1). Especially the conserved Lys-B14 and Glu/Asp-B17 are located at exposed, solvent-accessible positions in this loop (see Fig. 4 and supplementary data material Figures S1, panels G/H); they might be responsible for the functional docking of the putative redox partner QmoABC/QmoAB(-HdrBC) to the AprB protein surface adjacent to the [4Fe-4S] cluster II. The Qmo complex presumably operates as physiological electron carrier between the membrane-integral quinone/quinol pool and the cytoplasmic APS reductase in the SRP as well as in the sulfite-oxidizing Chlorobiaceae and Beta- and Gammaproteobacteria that harbor a SRB-related enzyme (SOB Apr lineage II) [17]. Thiobacillus denitrificans possessed a second exposed loop of six amino acids positioned between the beta-alpha-beta structure motifs of the ferredoxin-like segment; however, its functional or structural role is not apparent. (2) The crenarchaeal Caldivirga maquilingensis and putative sulfate-reducing Pyrobaculum spp. did not harbor qmo homologous sequences in their genomes [17]; in agreement to the previous proposal, an elongated loop between secondary structure element 2 and 3 was absent in their AprB comparative models (see Fig. 3D and supplementary material Figure S1). The putative physiological electron donor for the dissimilatory APS reductase in these species is unknown; however, naphtoquinones have been described from Pyrobaculum species [41], [42]. The electrostatic potential at the protein surface of the crenarchaeal AprB models differed significantly from the other SRP and SOB which might be an indication that the potential interacting redox partner and the electron transfer process of the aforementioned are different and unrelated to the latter (see Fig. 4 and supplementary data material Figures S1, panels E/F). Interestingly, both Pyrobaculum APS reductases missed the strictly conserved, electron transfer-relevant Trp-B48 that was substituted by an Ala-B43 residue at the corresponding AprB model position (see Fig. 4 panel G for the Pyrobaculum calidifontis AprB model). In all other Apr comparative models, the Trp-B48/43 was located between the S3 of cluster I and the methyl C8M of FAD (distance 12.4 Å) and in van der Waals contact with both redox centers; its indole ring was locked in its position by Thr-A233 and Arg-A232. The electron transfer function of trytophan residues located between two redox centers has been frequently documented [43], [44]; an analogous functional role, however, has not been reported for alanine residues. In consequence, the Pyrobaculum spp. APS reductases might have lost their structural ability for electron transfer between both subunits (essential for APS reduction) and, thus, their enzymes will not be functional anymore (see also the structural comparison of the alpha-subunits). Indeed, no cultivated Pyrobaculum species has been described to be capable of dissimilatory sulfate reduction [45], [46]; notably, the aprBA sequences of the Pyrobaculum aerophilum genome are frame-shifted [47] which might be a result of elevated mutation rate in irrelevant enzymes/metabolic pathways. (3) Like the beta-subunits of the crenarchaeal putative sulfate-reducers, the AprB comparative models of the SOB Apr lineage-I sulfur-oxidizers did not contain an elongated loop between secondary structure element 1 and 2 (see Fig. 3B and C and supplementary material Figure S1). Consistently, the genomes of the respective species did not include qmo homologues; however, their aprBA genes were always preceded and co-transcribed by a membrane-integral protein encoding gene, aprM. Interestingly, all AprB models of the SOB Apr lineage I possessed in the antiparallel beta-sheet segment an elongated loop by amino acid insertion between the secondary structure element 8 and 9 that comprised predominantly negatively charged residues (see Fig. 4 panels C and E; supplementary data material Table S1 and Figure S1, compare panels E/F). The latter were exposed to the solvent (see Fig. 4 panels D and F and supplementary data material Figure S1, compare panels G/H) and might function as an interface region for docking the AprBA enzyme to the AprM protein that anchors the dissimilatory APS reductase to the membrane and enables its physical contact to the yet unknown electron receptor. The presumed differing functional linkage of the cytoplasmic SOB Apr lineage I-type APS reductases to the membrane (not involving Qmo complex homologues) was also reflected in the deviating electrostatic potential at the protein surface when compared to the SRB-type AprB models (supplementary data material Figure S1, compare panels E/F).

Bottom Line: These structural alterations correlated with the protein phylogeny (three major phylogenetic lineages: (1) SRP including LGT-affected Archaeoglobi and SOB of Apr lineage II, (2) crenarchaeal SRP Caldivirga and Pyrobaculum, and (3) SOB of the distinct Apr lineage I) and the presence of potential APS reductase-interacting redox complexes.The almost identical protein matrices surrounding both [4Fe-4S] clusters, the FAD cofactor, the active site channel and center within the AprB/A models of SRP and SOB point to a highly similar catalytic process of APS reduction/sulfite oxidation independent of the metabolism type the APS reductase is involved in and the species it has been originated from.Based on the comparative models, there are no significant structural differences between dissimilatory APS reductases from SRP and SOB; this might be indicative for a similar catalytic process of APS reduction/sulfite oxidation.

View Article: PubMed Central - PubMed

Affiliation: Max Planck Institute for Marine Microbiology, Bremen, Germany.

ABSTRACT

Background: The dissimilatory adenosine-5'-phosphosulfate (APS) reductase (cofactors flavin adenine dinucleotide, FAD, and two [4Fe-4S] centers) catalyzes the transformation of APS to sulfite and AMP in sulfate-reducing prokaryotes (SRP); in sulfur-oxidizing bacteria (SOB) it has been suggested to operate in the reverse direction. Recently, the three-dimensional structure of the Archaeoglobus fulgidus enzyme has been determined in different catalytically relevant states providing insights into its reaction cycle.

Methodology/principal findings: Full-length AprBA sequences from 20 phylogenetically distinct SRP and SOB species were used for homology modeling. In general, the average accuracy of the calculated models was sufficiently good to allow a structural and functional comparison between the beta- and alpha-subunit structures (78.8-99.3% and 89.5-96.8% of the AprB and AprA main chain atoms, respectively, had root mean square deviations below 1 A with respect to the template structures). Besides their overall conformity, the SRP- and SOB-derived models revealed the existence of individual adaptations at the electron-transferring AprB protein surface presumably resulting from docking to different electron donor/acceptor proteins. These structural alterations correlated with the protein phylogeny (three major phylogenetic lineages: (1) SRP including LGT-affected Archaeoglobi and SOB of Apr lineage II, (2) crenarchaeal SRP Caldivirga and Pyrobaculum, and (3) SOB of the distinct Apr lineage I) and the presence of potential APS reductase-interacting redox complexes. The almost identical protein matrices surrounding both [4Fe-4S] clusters, the FAD cofactor, the active site channel and center within the AprB/A models of SRP and SOB point to a highly similar catalytic process of APS reduction/sulfite oxidation independent of the metabolism type the APS reductase is involved in and the species it has been originated from.

Conclusions: Based on the comparative models, there are no significant structural differences between dissimilatory APS reductases from SRP and SOB; this might be indicative for a similar catalytic process of APS reduction/sulfite oxidation.

Show MeSH
Related in: MedlinePlus