Insight into the molecular mechanism of the sulfur

Journal of Molecular Modeling (2018) 24:117 https://doi.org/10.1007/s00894-018-3652-5

ORIGINAL PAPER

Insight into the molecular mechanism of the sulfur oxidation process by reverse sulfite reductase (rSiR) from sulfur oxidizer Allochromatium vinosum Semanti Ghosh 1 & Angshuman Bagchi 1 Received: 11 December 2017 / Accepted: 9 April 2018 # Springer-Verlag GmbH Germany, part of Springer Nature 2018

Abstract Sulfur metabolism is one of the oldest known biochemical processes. Chemotrophic or phototrophic proteobacteria, through the dissimilatory pathway, use sulfate, sulfide, sulfite, thiosulfate or elementary sulfur by either reductive or oxidative mechanisms. During anoxygenic photosynthesis, anaerobic sulfur oxidizer Allochromatium vinosum forms sulfur globules that are further oxidized by dsr operon. One of the key redox enzymes in reductive or oxidative sulfur metabolic pathways is the DsrAB protein complex. However, there are practically no reports to elucidate the molecular mechanism of the sulfur oxidation process by the DsrAB protein complex from sulfur oxidizer Allochromatium vinosum. In the present context, we tried to analyze the structural details of the DsrAB protein complex from sulfur oxidizer Allochromatium vinosum by molecular dynamics simulations. The molecular dynamics simulation results revealed the various types of molecular interactions between DsrA and DsrB proteins during the formation of DsrAB protein complex. We, for the first time, predicted the mode of binding interactions between the cofactor and DsrAB protein complex from Allochromatium vinosum. We also compared the binding interfaces of DsrAB from sulfur oxidizer Allochromatium vinosum and sulfate reducer Desulfovibrio vulgaris. This study is the first to provide a comparative aspect of binding modes of sulfur oxidizer Allochromatium vinosum and sulfate reducer Desulfovibrio vulgaris. Keywords Reverse sulfite reductase (rSiR) . Allochromatium vinosum . Anoxygenic phototroph . Sulfur oxidation . Molecular dynamics simulation . Cofactor binding

Introduction The sulfur cycle is an ancient biogeochemical cycle. Inorganic sulfur is found in sedimentary rocks or by weathering of rocks buried deep in the oceanic sediments in salt form and also through decomposition of organic substances. Reduced sulfur compounds are oxidized by phototrophic and chemotrophic sulfur oxidizing bacteria (SOB) and used as a photosynthetic electron donor during anoxygenic photosynthesis [1, 2]. The dominant anoxygenic phototroph Allochromatium vinosum

Electronic supplementary material The online version of this article (https://doi.org/10.1007/s00894-018-3652-5) contains supplementary material, which is available to authorized users. * Angshuman Bagchi [email protected]; [email protected] 1

Department of Biochemistry and Biophysics, University of Kalyani, Nadia, Kalyani 741235, India

(A. vinosum) belongs to gamma-proteobacteria and has a wide range of habitats. This purple sulfur bacteria grows on marine as well as freshwater photo-anoxic ecosystems where sulfidic water or sediments are present [3]. A. vinosum is able to grow in intertidal mud flats with high redox versatility and as a result undergoes diverse metabolic pathways like chemolithoautotrophically or photolithoautotrophically. This unique metabolic feature is expressed by A. vinosum and Allochromatium minutissimum among all anoxygenic phototrophs [3, 4]. Sulfur globules are formed by purple sulfur bacteria A. vinosum as an obligatory intermediate, which is further oxidized by dissimilatory sulfite reductase (dSiR) [5]. The key enzymes of dissimilatory as well as assimilatory sulfur metabolism are sulfite reductase enzymes (SiRs). SiRs are diverged into different types, i.e., monomeric assimilatory SiR (aSiR), dissimilatory sulfite reductase (dSiR), and assimilatory nitrite (aNiR) reductase [6]. The dSiR is one of the oldest known redox enzyme complexes, and its presence was known even before the advent of oxygenic photosynthesis. Dissimilatory sulfite reductase (dSiR) is a crucial enzyme for

117

J Mol Model (2018) 24:117

Page 2 of 10

two extremely ancient processes, i.e., sulfite oxidation and sulfate reduction. dSiRs belong to a redox enzyme superfamily with siroheme-[4Fe-4S] coupled cofactor including assimilatory sulfite (aSiR) and nitrite (aNiR) reductase and many other types of sulfite reductases [7–9]. dSiRs are isolated and characterized from several sulfate reducers Desulfovibrio vul gar is (D. vu lgar is), A rc h a e o g l o b u s f u l g i d u s, Desulfovibrio gigas, and Desulfomicrobium norvegicum [6, 10–12]. The first X-ray structures of dSiRs were determined from the two sulfate-reducing organisms, Desulfovibrio vulgaris (D. vulgaris) at 2.10 Å [10] and Archaeoglobus fulgidus at 2.04 Å resolutions [6]. Both structures showed an α2β2 arrangement with similar overall folds, containing eight [4Fe–4S] such clusters. Four sirohemes are present in Archaeoglobus fulgidus dSiR, only two of which are catalytically active [6]. On the other hand, two sirohemes, viz., two sirohydrochlorins (i.e., the demetalled form of siroheme) are present in D. vulgaris and are called Dvir [10]. In Dvir, the α2β2 subunits of hetero-tetramer (known as DsrAB complex) are bound to the γ2 subunit, i.e., DsrC protein [10]. DsrA and DsrB proteins of Dvir (PDB ID: 2V4J) possess similar domains (A1/B1, A2/B2, A3/B3) but N- and C-terminal tails are distinct in the two proteins. In the A2 domain, four conserved cystein residues are present, viz., Cys177, Cys183, Cys221, and Cys225, which are conserved in all dSiRs forming CX5CX nCX3C motif for [4Fe-4S] cluster coordination. Similarly, in the B2 domain, Cys151, Cys188, Cys189, and Cys193 form the conserved CXnCCX3C motif for Fe-S cluster. dSiR from A. vinosum also contains siro(heme)amide-[4Fe-4S] as prosthetic group [13], but their position is still unknown. In sulfur oxidizers this enzyme is present in reversely operating form, i.e., reverse sulfite reductase, rSiR, and oxidize sulfide to sulfite and then to sulfate to generate electrons for CO2 reduction during their photoautotrophic sulfur metabolism [5, 14]. Various DsrAB mutants of A. vinosum were generated, and it was revealed that the mutants could oxidize sulfide to sulfur but further intracellular sulfur oxidation to form sulfate was totally abolished [5]. So, this DsrAB enzyme is essential for the oxidation of periplasmic sulfur globules [15]. DsrAB from sulfur oxidizer A. vinosum reduces the DsrC and DsrC transfers electrons to the membrane complex DsrMKJOP [15]. However, no three dimensional crystal structures of the DsrAB protein complex from A. vinosum are available to date. To compare sulfur oxidation and reduction mechanisms, the three dimensional structure of the DsrAB protein complex was built using the amino acid sequences of DsrA and DsrB protein from A. vinosum [16]. All the dSiRs so far have been studied from sulfur reducing bacteria and archea but the molecular insight of the reverse sulfite reductase, rSiR from sulfur oxidizing microorganisms, has not yet been explored. As the mechanism of the sulfur oxidation process is different from the sulfate reduction reactions, this structure function study would be useful to predict the hitherto unknown molecular mechanism of the sulfur

oxidation process from the sulfur oxidizing bacteria A. vinosum. Furthermore, the organism A. vinosum is important in biotechnological processes because it is able to generate bio-hydrogen from sulfur pollutants, and thereby would help to make an environment free from sulfur containing pollutants.

Methods Sequence alignment Due to unavailability of the 3D structure of DsrAB from A. vinosum, the binding interface residues of siroheme or the de-metalleted form of siroheme can not be predicted. However, we performed multiple sequence analysis using Clustal Omega server [17] taking the amino acid sequences of the DsrA and DsrB proteins with the amino acid sequences of the other well established dSiRs for which the structures are available in PDB. The analysis revealed the putative cysteins responsible for the coordination of the Fe-S clusters.

Formation of DsrAB protein complex ZDOCK was previously used to build the three dimensional structure of the DsrAB protein complex [16]. Further proteinprotein docking between DsrA and DsrB were performed in ZDOCK server and ZDOCK software suite of Discovery Studio 2.5 (DS 2.5) for consensus results. In order to get the consensus conformation of the DsrAB protein complex, the coordinates of the DsrA and DsrB proteins were taken from the stable region of the molecular dynamics (MD) simulation trajectories using the GROMOS clustering method, and these structures were further used for docking with help from the ZDOCK server [18]. We also used the ZDOCK [19] standalone tool present in Discovery Studio 2.5 for docking. The docked complexes were further re-ranked using ZRANK [20], and the best pose as obtained from ZRANK was analyzed further for comparative studies. The 3D coordinates of backbone atoms of DsrAB protein complexes from A. vinosum obtained from ZDOCK were superimposed onto the PDB coordinates of DsrAB protein complex from sulfate reducers using the DS 2.5 platform. The root mean squared deviations (RMSDs) were measured for the DsrAB protein complexes between the sulfate reducers and sulfur anion oxidizers for structural comparison. The final docking outputs of the DsrAB complex from both docking methods were compared with the structures of the docked complexes from the previous experiments to get the effect of protein dynamics on the binding interface of the DsrAB protein complex. The docked protein complexes were energy minimized using steepest descent followed by the conjugate gradient algorithm of DS 2.5 using the virtual water solvent system after applying the CHARMM

J Mol Model (2018) 24:117

force field. The final energy minimized complexes were further prepared for MD simulations.

Molecular dynamics simulation of DsrAB protein complex from A. vinosum To study the macromolecular interactions between DsrA and DsrB proteins we simulated DsrA, DsrB, and DsrAB protein complexes. Multiple replicates of the DsrA, DsrB, and DsrAB complex were prepared. For the DsrAB protein complex, three docked poses of DsrAB complexes were simulated in GROMACS as three replicates. All structures were corrected using the clean protein tool in the Discovery Studio (DS) 2.5 package (Accelrys Inc., San Diego, CA). The preparation of all molecular files and MD simulations were conducted using the previously described energy minimized files. The all-atom CHARMM 27 force field was assigned to all molecules for topology generation, and the explicit extended simple point charge (SPC/E) SPC216 water model was applied to solvate the molecules. A triclinic box was generated with a minimum of 1.0 nm distance from the edges of the box to maintain periodic boundary conditions throughout the simulations. Adequate counter ions (Cl− and Na+) were added to the solvent to keep the system neutral at physiological ionic strength (0.10 M of salt concentration). Steepest descent minimization was performed in GROMACS [21] until the maximum force reached below 1000.0 kJ mol−1 nm−1 and converged to the machine precision to relieve steric clashes and inappropriate geometries. The detailed composition of each replica system regarding box volume, counter ions, and minimization steps are provided in Suppl. Table 1. All bond lengths were constrained using the LINCS method [22], and long-range electrostatic interactions were calculated with the smooth particle mesh Ewald (PME) method [23]. Before the MD simulations, the systems were equilibrated using positionrestrained (PR) for 100 ps of isochoric-isothermal (NVT) equilibration at 300 K. This was followed by an equilibration under an isothermal-isobaric (NPT) ensemble for 100 ps at the same temperature and 1.0 bar of pressure without position restraints. In NVT and NPT ensembles, the short range nonbonded interactions were defined as van der Waals (VDW) and electrostatic interactions for particles within 1.0 nm. The well equilibrated systems DsrA, DsrB, and DsrAB were used to run 100 ns production MD in three replicates using velocity rescale thermostat and Parrinello-Rahman barostat; LINCS and PME treatments were also implemented as described in NVT and NPT ensembles. Snapshots of the trajectory were saved every 2 ps.

Analysis The structural features of all the proteins were calculated from their respective trajectories with the help of GROMOS

Page 3 of 10 117

clustering, and the middle most structures of the largest clusters were taken as the representative structures (RS) for DsrA, DsrB, and DsrAB proteins. The ‘trjconv’ and ‘tpbconv’ tools within GROMACS were used to strip out coordinates and correct for periodicity. In order to identify the amino acid residues at the protein–protein interface, interacting binding residues and hydrogen bond analysis were performed in DS 2.5. Non-bonded interaction energies in short range, i.e., van der Waals and electrostatic energy values, were calculated (throughout the simulations) for the binding interface residues using g_energy tools of GROMACS. Total interaction energies were the summation of individual van der Waals and electrostatic energy values of each residue plotted in Microsoft Excel. H-bond calculations of the interacting amino acids were measured throughout the simulation by g_hbond tools of GROMACS and also calculated using VMD. The most stable conformation of the DsrAB complex of pose 1 obtained from g_cluster tools of GROMACS, as obtained using the GROMOS clustering method, was used for binding free energy calculation and superimposition of the backbone atoms with the DsrAB protein complex of D. vulgaris in DS 2.5. To mimic the cellular environment in the binding free energy calculation, implicit or continuum electrostatic models were applied as a solvent where the complex was surrounded by a thermally averaged homogenously polarizable medium. The dielectric constant and degree of polarizability were set to 1 and non-bonded radius distances were in ranges 10–12 Å. The GROMACS suite of tools along with a secondary structure recognition algorithm (DSSP) [24] was used for MD simulation analysis. The core region of the protein complex was analyzed using Protein Core Composition Analyzer (ProCoCoA) [25]. The graphs were prepared in Microsoft Excel, and for structure visualization, DS Visualizer and VMD were employed.

Results and discussion Comparison of binding interface The basic aim of the work is to differentiate between the structural aspects of the molecular mechanisms of sulfur oxidation and sulfate reduction reactions. In order to do so, we compared the binding interfaces of the simulated DsrAB structure from A. vinosum and the crystal structure of DsrAB protein complex from D. vulgaris. The binding interfaces of the two protein complexes are markedly different as presented in Table 1 and Suppl. Fig. 1. It is quite apparent from the analysis of the binding interfaces of the DsrAB protein complexes that in sulfur oxidizer A. vinosum the binding interface is highly charged and the interface is stabilized mainly by charged interactions. On the other hand, the binding interface of the DsrAB protein

117

J Mol Model (2018) 24:117

Page 4 of 10

Table 1 Binding interface details of DsrA & DsrB proteins from A. vinosum and D. vulgaris DsrA

DsrB

A. vinosum

D. vulgaris

Lys83, Lys89, Asp148, Glu149, Asp156, Arg188, Asn192, Asp337 and Glu339 Arg60, Arg87, Glu203, Arg204 and Arg210

Gln11, Ser14, Pro16, Trp17, Asp23, Glu27, Tyr109, Tyr110, Arg116, Asp120, Asp123, Thr130, Met132, Glu151, Asp163, Lys180, Arg182, Lys260, Asn262, Arg269, Glu279, Arg283, Asp337 Ser6, Ile21, Gly22, Arg24, Phe26, Trp41, His44, Glu45, Ile46, Arg71, Met73, Ser74, Ile75, Arg79, Asp83, Asp86, Glu102, Ser127, Lys163, Asn291, Arg312, Lys333 and Arg336

complex in sulfate reducer D. vulgaris contains some nonpolar amino acids like Phe26, Trp41, Met73, Ile75 from DsrB and Pro16, Trp17, Thr130 from DsrA. The binding interface of the DsrAB protein in D. vulgaris is more stable than that of A. vinosum as apparent from the fact that more amino acids are present in the binding interface of D. vulgaris compared to A. vinosum. The DsrAB protein complex with the help of the DsrC protein in sulfur oxidizers functions by accepting electrons from sulfur substrates, and thereby they themselves get reduced. In order to carry out the electron transport processes, the DsrAB protein complex in sulfur oxidizer A. vinosum requires a highly positively charged environment. Since the electron transport occurs via the DsrAB protein complex in sulfur oxidizer A. vinosum, the binding interface of the DsrAB protein was found to contain a large number of positively charged amino acids. On the other hand, in sulfate reducer D. vulgaris, the sulfur substrates are ultimately converted to neutral sulfur. This requires the binding interface of the DsrAB protein complex in sulfur reducer D. vulgaris to have some non-polar amino acids that are able to bind the uncharged sulfur substrate. It may also be considered that in sulfur reducer D. vulgaris the inorganic sulfur atoms obtained after the reduction process may remain attached to the DsrAB protein complex. However, in sulfur oxidizer A. vinosum, the negatively charged sulfate ions obtained after the oxidation process require a highly positively charged environment to remain bound to the DsrAB protein complex. The presence of positively charged amino acids in the binding interface of the DsrAB protein complex in sulfur oxidizer A. vinosum not only stabilizes the DsrAB protein complex but also helps to bind the negatively charged sulfate to the protein. It is also noteworthy that the siroheme cofactors are present in the protein–protein interface of the DsrA and DsrB proteins with a minimum difference of ~14 Å, in the DsrAB protein complex in D. vulgaris. The cofactor is known to stabilize the DsrAB protein complex in D. vulgaris [10]. The catalytically active two sirohemes are surrounded by the amino acid residues of DsrA, DsrB, and three residues including the terminally conserved cysteine from the C-terminal part of DsrC. On the other hand, the amino acid residues from DsrA and DsrB involved in binding with sulfur substrates in D. vulgaris are also involved in binding with the sirohemes. Among the

amino acid residues of DsrA, the positively charged residues, like Lys 215, Lys 217, Arg231 from DsrA of D. vulgaris involved in the catalytic mechanism of sulfur reduction processes are also involved in binding the siroheme carboxylates through H-bonds or salt bridges [10]. On the other hand, no such phenomenon was observed in the case of sulfur oxidizer A. vinosum. Structure based sequence alignment results (Suppl. Figs. 2 and 3) revealed that the CX5CXnCX3C motif from the A2 domain of A. vinosum was conserved among all Dsr proteins and Cys170, Cys176, Cys214, and Cys218 were putative cysteins of A. vinosum responsible for binding with [4Fe-4S] cluster 1, i.e., part of siroheme-[4Fe-4S] cofactor. However, in the A3 domain of A. vinosum, another CX3CXnCX2C motif resposible for [4Fe-4S] cluster 2 coordination was found to be present (Cys264, Cys288, Cys291) where the second cystein, Cys268, was replaced by alanine in DsrA of A. vinosum. A similar structural arrangement was observed in A. fulgidus. In DsrB another conserved sequence motif CXnCCX3C was present which suggests that Cys133, Cys170, Cys171, and Cys175 from DsrB of A. vinosum were responsible for iron-sulfur cluster coordination. Similarly, in the B3 domain, the [4Fe-4S] cluster 2 coodinated by different conserved CXnCX2CX2C motif suggests Cys211, Cys234, Cys237, and Cys240 residues from DsrB of A. vinosum were responsible for siroheme binding. Interestingly, no catalytic residues in A. vinosum were found to be involved in binding with the siroheme as in the case of D. vulgaris.

Comparison of DsrAB complex from A. vinosum with other sulfate reducers The three dimensional structures of the DsrAB protein complex obtained from ZDOCK protein-protein docking methods were compared with the DsrAB structures from sulfate reducers available in the Protein Data Bank (PDB). The analyses of the RMS deviations of the backbone atoms among them revealed that there were huge structural changes present between the DsrAB protein complexes, and the RMSDs vary within the range 23–28 Å (Table 2). All poses of the DsrAB complex from sulfur oxidizer A. vinosum showed high conformational deviations from the DsrAB protein complexes of the sulfate reducers signifying the differences between sulfur

J Mol Model (2018) 24:117 Table 2 reducers

Page 5 of 10 117

Structural comparison between sulfur oxidizer and sulfate

ZDOCK poses of DsrAB from A. vinosum

PDB ID (sulfate reducers)

RMSD in Å

Pose 1

2V4J 2XSJ 3MM5 3OR1 3OR2 2V4J 2XSJ 3MM5 3OR1 3OR2 2V4J 2XSJ 3MM5 3OR1 3OR2

28.034 28.003 27.082 27.904 27.917 23.873 23.711 23.857 23.413 23.442 26.317 25.972 25.533 26.302 26.335

Pose 2

Pose 3

PDB code: 3MM5 – Archaeoglobus fulgidus; 2XSJ – Desulfomicrobium norvegicum; 2V4J – Desulfovibrio vulgaris; 3OR2 – DsrII of Desulfovibrio gigas; 3OR1 – DsrI of Desulfovibrio gigas. Poses 1 and 2 represent the structure of the DsrAB protein complexes from A. vinosum obtained from two different ZDOCK results in DS 2.5 and pose 3 is the structure of DsrAB protein complex from the ZDOCK online server

oxidizers and reducers. We also compared the reaction mechanisms of sulfur anion oxidation and reduction processes in A. vinosum (sulfur anion oxidizer) and D. vulgaris (sulfate anion reducer) [16]. We observed that the binding interface of A. vinosum contains five positively charged basic amino acid residues in the sulfur anion binding interface. It was observed that A. vinosum binds most strongly with thiosulfate and sulfide. These two ions are comparatively smaller in sizes and thus have high surface charge densities. Therefore, the presence of the five positively charged amino acid residues in the binding interface facilitates binding of the sulfur anions in A. vinosum. On the other hand, D. vulgaris binds maximally with sulfate and sulfite. These two ions are comparatively larger in sizes and therefore have lower surface charge densities. The binding interface of D. vulgaris contains three positively charged amino acid residues sufficient for interaction with the comparatively larger sulfate and sulfite. It is also noteworthy that the final product of sulfur anion oxidation is sulfate, which requires the addition of oxygen atoms to the sulfur anions bound to the sulfur anion binding interface of DsrAB protein complex in A. vinosum. Thus, the highly positively charged environment in the sulfur anion binding interface of DsrAB protein complex in A. vinosum, due to the presence of five positively charged amino acid residues, is essential for the removal of oxygen from water.

Stability of DsrAB protein complex and prediction of core center The progress of the MD simulation run was determined by plotting the change in RMSD of backbone atoms with time scale (Fig. 1). Free DsrA proteins from the first and second replicates had higher RMS deviations (range: 0.8–1.0 nm) than free DsrB protein and DsrA protein bound to the DsrAB complex (Fig. 1a). The third replicate of DsrA protein showed a similar trend in the pattern of RMS deviations as mentioned before. DsrA became more stable after associating with DsrB as the deviation lowered after formation of the DsrAB complex. The DsrAB complex (pose 1) stabilized with time and showed the maximum standard deviation of the RMSD value as 0.7 nm. We then docked the structures of DsrA and DsrB proteins using the middle-most structures of the proteins from equilibrated trajectories of the first replicates of DsrA and DsrB proteins to form the DsrAB protein complex to compare the RMS deviations with the DsrAB docked complex as mentioned [16]. The docked pose (pose 3) of the DsrAB protein complex obtained from the ZDOCK online server showed the RMS deviation around 0.5 nm whereas pose 2 obtained from the ZDOCK suite of DS 2.5 showed the lowest RMS deviation around 0.3 nm. Therefore, the protein dynamics of DsrA

Fig. 1 a Structural stability analysis using root mean square deviations (RMSD) of backbone atoms of DsrA, DsrB, and DsrAB proteins. Three replicates of DsrA, DsrB, and three docked poses of DsrAB denoted as A1, A2, A3; B1, B2, B3; and AB1, AB2, AB3 respectively. b Amino acid fluctuation analysis by plotting root mean square fluctuations (RMSF) of C-alpha atoms of DsrA, DsrB, and DsrAB proteins

117

Page 6 of 10

and DsrB provided the stable conformations that ultimately reduced the fluctuation on DsrAB protein complex formation. The root mean square fluctuations (RMSFs) of the amino acid residues of the DsrA and DsrB proteins revealed that DsrA and DsrB fluctuated less when the DsrAB protein complex was formed (Fig. 1b). Amino acid residues of N-terminus, spanning the amino acid residues 230–360 and the Cterminal region (amino acid residues 390–410) of DsrA fluctuate less when the DsrA is present in the DsrAB protein complex. Similarly, the N and C-terminal regions and amino acid residues 190–250 from DsrB protein show fewer fluctuations in the DsrAB protein complex during the course of the MD simulation run. The RMSFs were justified by the secondary structure alternation throughout MD simulation in DSSP profiles (Fig. 2, Suppl. Fig. 4 and Suppl. Fig. 5). The amino acid residues of N-terminal regions spanning the amino acid residues 10–40 of DsrA, the helices joined by coils and turns, were converted into bend and turn regions

J Mol Model (2018) 24:117

in the DsrAB protein complex. The amino acid residues 75– 85 consisting of fluctuating turn regions in the DsrA protein were turned into prevalent bends in due course of the MD run. Another fluctuating region of DsrA, viz., the amino acid residues 230–260, became more stable in the DsrAB complex as coil, turn, and bend regions in these parts of the protein convert into β-bridges and helices. The C-terminal part, 390–402, a coil and fluctuating bend region, was converted into a more stable super secondary structure turn in the DsrAB protein complex. All of which indicated good protein stability of DsrA over the course of the simulation after complex formation. In the case of DsrB, the coil region of the C-terminus and the region spanning the amino acid residues 57–62 were converted into a bend depicted by lower RMSF values. The fluctuating region of DsrB, which is the region spanning the amino acid residues 190–250, became more stable after complex formation. In the C-terminal region of DsrB, the helices were

Fig. 2 Secondary structure alteration throughout MD simulation using DSSP profiles of DsrAB protein complex

J Mol Model (2018) 24:117

interrupted by the turn formation, which became a more stable helix region after joining with DsrA. InterPro search of the amino acid residues of DsrA protein revealed that the amino acid residues spanning the region 6–161 consist of a nitrite/ sulfite reductase ferredoxin like domain (IPR005117), and the amino acid residues spanning the region 161–381 show a nitrite/sulfite reductase [4Fe-4S] domain (IPR006067). It was also observed from sequence alignment of the DsrA and DsrB proteins from A. vinosum with those of D. vulgaris that mostly the basic amino acid residues, viz., Arg77, Arg95, Arg165, Lys206, Lys208, Lys210, Arg224, His359, and Arg361 from DsrA protein and Arg60, His132, and Asp 134 from DsrB were involved in binding the sulfur anions [16]. However, none of these amino acid residues were supposed to be involved with siroheme in A. vinosum as in the case of D. vulgaris. Moreover, no experimental structure of siroheme-[4Fe-4S] cofactor bound DsrAB from A. vinosum was present. This was also exemplified by the structural comparison of the DsrAB protein complex of A. vinosum with that of D. vulgaris (Fig. 3) where the RMS deviations between the backbone atoms of the proteins were 31.469 Å. These amino Fig. 3 Structural superimposition of backbone atoms DsrAB protein from A. vinosum and D. vulgaris. DsrA and DsrB are red and magenta color from D. vulgaris, and DsrA and DsrB are blue and sky blue color from A. vinosum. c DsrAB complex from A. vinosum and D. vulgaris and their color code and representation forms as follows: DsrA_A.vinosum in blue color tube representation, DsrB_ A.vinosum in red color with schematic form, DsrA_D.vulgaris in cyan color in stick form, and DsrB_D.vulgaris in violet solid ribbon form. N-terminus and Cterminus of all proteins show they were optimally aligned

Page 7 of 10 117

acid residues lie in the fluctuating regions of the DsrA and DsrB proteins in A. vinosum as observed by the RMSF and DSSP profiles of the MD simulation run. So, it can be concluded that this particular site of DsrA is responsible for the catalytic activity of the DsrAB enzyme complex for sulfur anion oxidations. The core region of the DsrAB protein complex as obtained by ProCoCoA server revealed the following amino acid residues: DsrA- Phe 91, Trp 116, Leu 123, Ile 124, Asn 132, Ile 136, Phe 146, Phe 147, Ile 150, Gly 154, Gly 158, Ala 167, Gly172, Ala 186, Leu 190, Phe 207, Phe 209, Phe 227, Cys 362, Met 365, Phe 374. DsrB- Ser 88, Val 113, Ser 120, Val 121, Gly 129, Cys 133, Val 143, Val 144, Met 147, Leu 151, Phe155, Val 164, Ser 169, Ile 183, His 187, Arg 204, Ser 206, Lys 224, Leu 257, Tyr 303, Ala 307, ASN 349. This core region of the protein consisting of mostly hydrophobic residues was the most stable region according to RMSF and DSSP profiles that also confirmed the aforementioned conserved basic amino acid residues lie on the flexible region and are responsible for sulfur anion oxidations.

117

J Mol Model (2018) 24:117

Page 8 of 10

Binding interactions between DsrA and DsrB proteins in DsrAB protein complex Protein–protein interaction depends on some bonded and nonbonded interactions. To make a stable protein-protein complex as well as proper folding of proteins several H-bonds, nonbonded electrostatic, and van-der Waals (vdW) interactions are required as shown in Fig. 4 and Suppl. Fig. 6. Among the amino acid residues of DsrA, the participating residues Arg188, Asp156, Glu149, Asp148, Lys89, Glu339, Asp337, Lys83, and Asn192 had more than −60 kJ mol−1 total electrostatic and vdW energy terms (Fig. 4a). It was quite obvious that these residues were also maximally involved in hydrogen bond formation with DsrB (Fig. 4b). Similarly, the amino acid residues Arg210, Glu203, Arg204, Arg60, and Arg87 from DsrB were maximally involved in interactions with DsrA throughout MD simulation (Suppl. Fig. 6). In order to analyze the binding interactions between DsrA and DsrB proteins in a more proper way, hydrogen bond pairing during all frames of the MD production run between DsrA and DsrB proteins were calculated in VMD with bond pair distances within 3.5 Å and angles 35° as represented in Table 3. It was found that several amino acid residues of DsrA and DsrB formed hydrogen bonds in the DsrAB protein complex, and a single amino acid participated in hydrogen bond formations with more than two amino acid residues, viz., Arg188 of DsrA formed hydrogen-bonds with Glu203 and Arg210 of DsrB. Fig. 4 Total interaction energies (a) and average number of Hbonds (b) of amino acid residues of DsrA involved in the DsrAB complex throughout MD simulation

Maximum occupancies of the following amino acids were Arg188, Asp156, and Glu 149 of DsrA and Arg204, Arg210, and Arg188 of DsrB. The solvent accessible surface area of the DsrAB complex became higher after simulation. The sum of solvent accessible surface area of DsrA and DsrB proteins was 454.47 nm2 whereas for the DsrAB complex it was 457.77 nm2. The spatial changes denote the DsrAB complex became more accessible for sulfur anions or DsrC binding. The binding free energy for the stable DsrAB complex was −2098.276 kJ mol−1, which also denotes a stable protein complex. It is desirable that the DsrAB protein complex remains stable as without the presence of the DsrAB protein in sulfur oxidizers, the sulfur anion oxidation process would not be possible. The results of the MD simulation run clearly indicated that the DsrAB protein complex was held together by strong intermolecular non-covalent forces. These strong binding interactions between the DsrA and DsrB proteins make the complex highly stable and fit for possible involvement in sulfur anion oxidation reactions.

Conclusions The molecular details of the binding interactions of DsrAB protein complex were elucidated for sulfur anion reducing bacterial systems, but no such works are available for sulfur anion oxidizers. In our previous work, we compared the binding interactions of sulfur substrates with DsrA and DsrB

J Mol Model (2018) 24:117 Table 3 H-bonds analysis between DsrA and DsrB throughout MD simulation

Page 9 of 10 117

Donor

Acceptor

Occupancy

A_TYR119-Side-OH A_ARG188-Side-NH2 A_ARG188-Side-NE B_ARG210-Side-NH1 B_ARG210-Side-NH2

B_PHE84-Main-O B_GLU203-Side-OE1 B_GLU203-Side-OE2 A_ASP156-Side-OD2 A_ASP156-Side-OD1

86.34% 80.24% 79.63% 71.75% 71.41%

B_ARG60-Side-NH2 A_ARG188-Side-NH2 B_ARG210-Side-NH1 B_ARG210-Side-NH2 B_ARG204-Side-NH2 A_ARG188-Side-NE B_GLN17-Side-NE2 B_ARG204-Side-NH2 A_LYS89-Side-NZ B_ARG210-Side-NE B_ARG60-Side-NH2 B_TYR18-Side-OH B_ARG87-Side-NH1 B_ARG204-Side-NH1 B_ARG204-Side-NH2 B_ARG204-Side-NH2 B_GLN17-Main-N B_ARG204-Side-NH1

A_GLU149-Side-OE2 B_GLU203-Side-CD A_ASP156-Side-CG A_ASP156-Side-CG A_ASP337-Side-OD2 B_GLU203-Side-CD A_ASP31-Main-O A_GLU339-Side-OE2 B_SER10-Main-O A_ASN192-Side-OD1 A_GLU149-Side-CD A_ARG37-Main-O A_ASP148-Side-OD1 A-GLU339-Side-CD A_ASP337-Side-OD1 A_GLU339-Side-CD A_ASN43-Side-OD1 A_GLU339-Side-OE1

68.90% 56.55% 48.51% 47.59% 44.24% 43.88% 38.91% 36.35% 35.84% 34.46% 34.26% 32.90% 32.08% 32.01% 31.10% 30.22% 30.14% 27.93%

B_ARG204-Side-NH2 A_LYS89-Side-NZ B_ARG210-Side-NH1 B_ARG204-Side-NH1 B_GLY176-Main-N B_ARG87-Side-NH1 B_ARG87-Side-NH1 B_GLN128-Side-NE2 B_ARG210-Side-NH2 B_ARG210-Side-NH2 B_ARG210-Side-NH1 B_ARG60-Side-NH1 B_ARG60-Side-NH2 A_ARG188-Side-NH2 B_ARG204-Side-NH2 A_ARG188-Side-NE A_THR42-Side-OG1 A_GLU90-Main-N

A_GLU339-Side-OE1 B_ILE8-Main-O A_ASN192-Side-OD1 A_GLU339-Side-OE2 A_GLY69-Main-O A_ASP148-Side-CG A_ASP148-Side-OD2 A_ASP152-Side-OD1 A_ASP156-Side-OD2 A_ASN192-Side-OD1 A_ASP156-Side-OD1 A_GLU149-Side-OE2 A_GLU149-Side-OE1 B_GLU203-Side-OE2 A_ASP337-Side-CG B_GLU203-Side-OE1 B_ASP14-Main-O B_CYS12-Side-SG

27.70% 27.39% 25.24% 24.29% 23.48% 22.80% 21.82% 21.09% 20.48% 20.13% 19.93% 19.02% 18.66% 18.56% 17.61% 17.21% 16.95% 16.80%

A_LYS185-Side-NZ B_GLN128-Side-NE2 A_ASN43-Side-ND2 B_TRP16-Main-N B_ARG204-Side-NH1 B_ARG87-Side-NH2

B_GLU203-Side-OE1 A_GLU149-Side-OE1 B_GLN17-Side-OE1 A_ASN43-Side-OD1 A_ASP337-Side-OD2 A_ASP148-Side-OD1

16.26% 14.93% 14.77% 14.66% 14.62% 11.87%

Amino acid residues of DsrA and DsrB are represented as A and B respectively

117

J Mol Model (2018) 24:117

Page 10 of 10

proteins from sulfur anion oxidizers and reducers. In this work, we tried to establish the modes of binding interactions between DsrA and DsrB protein to form the stable DsrAB protein complex formation as it is a very vital process in sulfur oxidation. Moreover, we tried to illustrate the structural details of the formation of DsrAB protein complex in a sulfur anion oxidizer and compare the differences in binding patterns of DsrAB protein complexes in sulfur oxidizer A. vinosum and reducer D. vulgaris to understand their reverse mode of redox reactions. The structural insights and differences of DsrAB in A. vinosum also provide an idea for its diverse anoxic habitat and metabolic versatility. So, far this is the first such effort to analyze the binding interactions in a sulfur oxidizer and compare the differences between a sulfur oxidizer and reducer. Thus, our work may be useful for future biochemical studies to elucidate the molecular mechanism of sulfur anion oxidation process in a sulfur oxidizer. Acknowledgments Ms. Semanti Ghosh is thankful to the University of Kalyani, West Bengal, India for the fellowship support. We would like to thank the Bioinformatics Infrastructure Facility and also the DST-PURSE program 2012–2016 at the Department of Biochemistry and Biophysics, University of Kalyani for the infrastructural support. The authors are thankful to the Dept of Biotechnology, Govt. of India (project No. BT/ PR6869/BID/7/417/2012) and (SAN no. 102/IFD/SAN/1824/20152016) for the instrumental support.

8.

9.

10.

11.

12.

13.

14.

15.

Funding The authors are thankful to the Dept of Biotechnology, Govt. of India (SAN no. 102/IFD/SAN/1824/2015-2016) for the instrumental support. 16.

Compliance with ethical standards Conflict of interest None.

References Sorokin DY, Berben T, Melton ED, Overmars L, Vavourakis CD, Muyzer G (2014) Microbial diversity and biogeochemical cycling in soda lakes. Extremophiles 18:791–809 2. Brune DC (1995) Sulfur oxidation by phototrophic bacteria. Biochim Biophys Acta 975:189–221 3. Weissgerber T, Zigann R, Bruce D, Chang Y, Detter JC, Han C, Hauser L, Jeffrie CD, Land M, Munk AC, Tapia R, Dahl C (2011) Complete genome sequence of Allochromatium vinosum DSM 180T. Stand Genomic Sci 5:311–330 4. Kämpf C, Pfennig N (1980) Capacity of chromatiaceae for chemotrophic growth. Specific respiration rates of Thiocystis violacea and Chromatium vinosum. Arch Microbiol 127:125–135 5. Pott AS, Dahl C (1998) Sirohaem-sulfite reductase and other proteins encoded in the dsr locus of Chromatium vinosum are involved in the oxidation of intracellular sulfur. Microbiology 144:1881–1894 6. Schiffer A, Parey K, Warkentin E, Diederichs K, Huber H, Stetter KO, Kroneck PMH, Ermler U (2008) Structure of the dissimilatory sulfite reductase from the hyperthermophilic archaeon Archaeoglobus fulgidus. J Mol Biol 379:1063–1074 7. Crane BR, Getzoff ED (1996) The relationship between structure and function for the sulfite reductases. Curr Opin Struct Biol 6:744–756

17.

18.

1.

19. 20. 21.

22. 23. 24.

25.

Loy A, Duller S, Wagner M (2007) Evolution and ecology of microbes dissimilating sulfur compounds: insights from siroheme sulfite reductases. In: Dahl C, Friedrich C (eds) Microbial sulfur metabolism. Springer, Heidelberg, pp 46–59 Dhillon A, Goswami S, Riley M, Teske A, Sogin ML (2005) Domain evolution and functional diversification of sulfite reductases. Astrobiology 5(1):18–29 Oliveira TF, Vonrhein C, Matias PM, Venceslau SS, Pereira IA, Archer M (2008) The crystal structure of Desulfovibrio vulgaris dissimilatory sulfite reductase bound to DsrC provides novel insights into the mechanism of sulfate respiration. J Biol Chem 283: 34141–34149 Hsieh YC, Liu MY, Wang VC, Chiang YL, Liu EH, Wu WG, Chan SI, Chen CJ (2010) Structural insights into the enzyme catalysis from comparison of three forms of dissimilatory sulphite reductase from Desulfovibrio gigas. Mol Microbiol 78(5):1101–1116 Oliveira TF, Franklin E, Afonso JP, Khan AR, Oldham NJ, Pereira IAC, Archer M (2011) Structural insights into dissimilatory sulfite reductases: structure of desulforubidin from Desulfomicrobium norvegicum. Front Microbiol 2:71 Lübbe YJ, Youn H-S, Timkovich R, Dahl C (2006) Siro (haem) amide in Allochromatium vinosum and relevance of DsrL and DsrN, a homolog of cobyrinic acid a, c-diamide synthase, for sulphur oxidation. FEMS Microbiol Lett 261:194–202 Dahl C, Engels S, Pott-Sperling AS, Schulte A, Sander J, Lübbe Y, Oliver Deuster O, Brune DC (2005) Novel genes of the dsr gene cluster and evidence for close interaction of Dsr proteins during sulfur oxidation in the phototrophic sulfur bacterium Allochromatium vinosum. J Bacteriol 187:1392–1404 Cort JR, Selan U, Schulte A, Grimm F, Kennedy MA, Dahl C (2008) Allochromatium vinosum DsrC: solution-state NMR structure, redox properties and interaction with DsrEFH, a protein essential for purple sulfur bacterial sulfur oxidation. J Mol Biol 382(3): 692–707 Ghosh S, Bagchi A (2015) Comparative analysis of the mechanisms of sulfur anion oxidation and reduction by dsr operon to maintain environmental sulfur balance. Comput Biol Chem 59: 177–184 Sievers F, Wilm A, Dineen D, Gibson TJ, Karplus K, Li W, Lopez R, McWilliam H, Remmert M, Söding J, Thompson JD, Higgins DG (2011) Fast, scalable generation of high-quality protein multiple sequence alignments using Clustal Omega. Mol Syst Biol 7:539 Pierce BG, Wiehe K, Hwang H, Kim BH, Vreven T, Weng Z (2014) ZDOCK server: interactive docking prediction of protein-protein complexes and symmetric multimers. Bioinformatics 30(12): 1771–1773 Chen R, Weng Z (2002) Docking unbound proteins using shape complementarity, desolvation, and electrostatics. Proteins 47:281–294 Pierce B, Weng Z (2007) ZRANK: reranking protein docking predictions with an optimized energy function. Proteins 67:1078–1086 Hess B, Kutzner C, Van Der Spoel D, Lindahl E (2008) GROMACS 4: algorithms for highly efficient load-balanced and scalable molecular simulation. J Chem Theory Comput 4:435–447 Hess B (2007) P-LINCS: a parallel linear constraint solver for molecular simulation. J Chem Theory Comput 4:116–122 Essmann U, Perera L, Berkowitz ML, Darden T, Lee H et al (1995) A smooth particle mesh Ewald method. J Chem Phys 103:8577–8593 Kabsch W, Sander C (1983) Dictionary of protein secondary structure: pattern recognition of hydrogen-bonded and geometrical features. Biopolymers 22:2577–2637 Bottini S, Bernini A, De Chiara M, Garlaschelli D, Spiga O, Dioguardi O, Vannuccini E, Tramontano A, Niccolai N (2012) ProCoCoA: a quantitative approach for analyzing protein core composition. Comput Biol Chem 43:29–34